DEPARTMENT OF ARCHAEOLOGY Document a Praehist orica XLIX Univerza v Ljubljani Filozofska fakulteta Oddelek za arheologijo Document a Praehist orica XLIX Urednik/Editor Mihael Budja Ljubljana 2022 DOCUMENTA PRAEHISTORICA XLIX (2022) Urednika/Editors: Prof. Dr. Mihael Budja, urednik/editor, Mihael.Budja@ff.uni-lj.si Bojan Kambiè, tehnièni urednik/technical editor, bojan.kambic@amis.net Uredniški odbor/Editorial board: Maja Andriè, Institute of Archaeology, ZRC SAZU, Ljubljana, Slovenia Mihael Budja, University of Ljubljana, Faculty of Arts, Slovenia Canan Çakirlar, University of Groningen, Faculty of Arts, Netherlands Ekaterina Dolbunova, The State Hermitage Museum, The department of archaeology of Eastern Europe and Siberia, Saint-Petersburg, Russian Federation Ya-Mei Hou, Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, Beijing, China Dimitrij Mlekuž Vrhovnik, University of Ljubljana, Faculty of Arts, Slovenia; Institute for the Protection of the Cultural Heritage of Slovenia, Ljubljana, Slovenia Simona Petru, University of Ljubljana, Faculty of Arts, Slovenia Žiga Šmit, University of Ljubljana, Faculty of mathematics and physics, Slovenia Manca Vinazza, University of Ljubljana, Faculty of Arts, Slovenia Katherine Willis, University of Oxford, United Kingdom Andreja Žibrat Gašpariè, Agricultural Institute of Slovenia, Slovenia To delo je ponujeno pod licenco Creative Commons Priznanje avtorstva-Deljenje pod enakimi pogoji 4.0 Mednarodna licenca. This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. Založila/Published by: Založba Univerze v Ljubljani/University of Ljubljana Press Za založbo/For the publisher: Prof. Dr. Gregor Majdiè, rektor Univerze v Ljubljani/ Prof. Dr. Gregor Majdiè, The Rector of the University of Ljubljana Izdala/Issued by: Znanstvena založba Filozofske fakultete, Oddelek za arheologijo/ University of Ljubljana Press, Faculty of Arts; Department of Archaeology Za izdajatelja/For the issuer: Prof. Dr. Mojca Schlamberger Brezar, dekanja Filozofske fakultete Univerze v Ljubljani/Prof. Dr. Mojca Schlamberger Brezar, The Dean of the Faculty of Arts, University of Ljubljana Naslov uredništva/Address of Editorial Board: Oddelek za arheologijo, Filozofska fakulteta, Univerza v Ljubljani, Aškerèeva 2, 1001 Ljubljana, p.p. 580, tel.: +386 12411570 Spletni naslov/Website: https://journals.uni-lj.si/DocumentaPraehistorica Prelom/DTP: Cambio d.o.o., Ljubljana Tisk/Printed by: Birografika BORI d.o.o., Ljubljana Naklada/Circulation: 250 izvodov/copies Cena/Price: 56,13 EUR Natisnjeno s podporo Javne agencije za raziskovalno dejavnost Republike Slovenije. Funded by the Slovenian Research Agency. Documenta Praehistorica je vkljuèena v Evropski referenèni seznam za humanistiène in družbene vede (ERIH PLUS), Scopus in DOAJ. Sodeluje v omrežju CrossRef (http://www.crossref.org/). The Documenta Praehistorica is indexed in the European Reference Index for the Humanities and Social Sciences (ERIH PLUS), Scopus and DOAJ. The journal participates in CrossRef (http://www.crossref.org/). E-izdaja: Publikacija je v digitalni obliki prosto dostopna na https://journals.uni-lj.si/DocumentaPraehistorica. E-edition: The publication is freely available in digital form at https://journals.uni-lj.si/DocumentaPraehistorica. ISSN 1408–967X (Print) ISSN 1854–2492 (Online) CONTENTS Cédric Bodet The Neolithic dualist scheme Hüreyla Balci, Rana Özbal, Fokke Gerritsen, and Erkan Fidan 28 Different manifestations of Neolithization in Northwest Anatolia| An archaeobotanical review from Barcin Höyük and Bahçelievler, Turkey Ana .uri.i. 46 What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans Kelly Reed, Petra Radakovi., Sara Essert, and Dinko Tresi. Pavi.i. 68 Late Bronze Age food storage in Lower Cerova;ka Cave, Croatia> the archaeobotanical evidence Laura Dietrich, Oliver Dietrich, and Julia Meister 84 Hermetic cereal storage in the Bronze Age> evidence from the Gáva culture settlement at Rotbav, Transylvania Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, 94 and Dimitra Kotsachristou Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of Pontokomi-Souloukia Nelson J. Almeida, André Texugo, and Ana Catarina Basílio 124 ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance Robin Brigand, Jérôme Dubouloz, and Olivier Weller 150 Colonization dynamics of LBK farmers in Europe under geostatistics test Franti.ek Trampota, Petr Pajdla 194 Neolithic settlement structures in Central Europe> case study of East Bohemia and the Morava River catchment Andrea Villani 214 Disuse of spaces and discard of artefacts during the abandonment of Erimi-Laonin tou Porakou Mohsen Heydari Dastenaei, Mohsen Dana 230 Zoning analysis of Iron Age sites using Analytic Hierarchical Process (AHP) methods in the Middle Atrak River Basin, Northeast of Iran Basak Boz 244 A possible case of ‘accompanying dead’ in the second half of the 6th millennium cal BC at Ug¢ urlu\Gökçeada, Turkey Noémi Beljak Pa.inová, Alena Bistáková 260 Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe Judit Regenye, Krisztián Oross, Eszter Bánffy, Elaine Dunbar, Ronny Friedrich, Alex Bayliss, Nancy Beavan, Bisserka Gaydarska, and Alasdair Whittle 280 Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary Federico Bernardini, Giacomo Vinci, Vanja Macovaz, Andrea Baucon, Angelo De Min, Stefano Furlani, and Sne.ana Smoli. 300 Prehistoric stone disks from entrances and cemeteries of north-eastern Adriatic hillforts Natalia Petrova, Hojjat Darabi 318 Analysis of Neolithic pottery technology along the Iranian Zagros foothills Andrey Mazurkevich, Ekaterina Dolbunova, Yolaine Maigrot, and Veronika Filippova 344 The end of the earliest ceramic traditions> Dnieper-Dvina region became part of the Circum-Baltic space at the turn of the 6th to 5th millennium BC Nina Kyparissi-Apostolika, Evita Kalogiropoulou, Dimitris Roussos, Niki Saridaki, Odysseas Metaxas, Georgia Kotzamani, Katerina Trantalidou, and Yorgos Facorellis 364 A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida Manca Vinazza, Matej Dolenec 388 Pottery firing in the Early Iron Age in western Slovenia Vasile Opris, Bogdan Manea, Mircea Lechintan, Roxana Bugoi, Florin Constantin, Theodor Ignat, and Catalin Lazar 406 Between object and subject> multiple approaches to a prehistoric human-shaped pot from Romania Jarrad W. Paul, Coskun Sivil, and Özlem Çevik 422 The Neolithic worked bone assemblage from Ulucak Höyük, Western Anatolia Valdis Berzin., Agnese .akare 434 Pattern and variation in jewellery production sequences> analysis of 4th millennium BC amber assemblages from the Latvian coast Dimitra Malamidou, Zoi Tsirtsoni, and Markos Vaxevanopoulos 450 The emergence of metal use in Greek Eastern Macedonia during the Neolithic period (late 6th–5th millennia BC) Oliver Dietrich, Laura Dietrich, and Botond Rezi 470 ‘Scrap metal hoards’ of the Later Urnfield Period in the Carpathian Basin> a case study on a hoard from Dezmir, Romania Documenta Praehistorica XLIX (2022) The Neolithic dualist scheme Cédric Bodet Mug¢ la Sitki Koçman University, Kötekli, TR< cedric.bodet@yahoo.com, cedricbodet@mu.edu.tr ABSTRACT – The monumental twin steles of Göbekli Tepe are one in a long series of isomorphic compositions in Neolithic symbolism. Seemingly tracing back to the Palaeolithic, symmetry likely played a fundamental role for prehistoric societies. Ethnographers showed how hunter-gatherer ideo­logy (mythology, totemism, etc.) is often structured around a dualistic worldview (male/female; sum-mer/winter etc.) taking root in the kinship system through a division of the community into exoga-mic subgroups. It is this dualism that is argued to be embodied in the twin steles. The advent of au­tonomous agricultural lineages could explain why this timeless principle appears with such promi­nence in the Neolithic. KEY WORDS – Neolithic symbolism; exogamy; kinship structure; hunter-gatherer ideology; incipient farmer Neolitski dualizem IZVLE.EK – Monumentalna kamnita dvoj.ka na najdi..u Göbekli Tepe predstavljata eno od izomorf­nih kompozicij neolitske simbolike. Simetrija je, domnevno .e vse od paleolitika, igrala klju.no vlogo v prazgodovinskih dru.bah. Etnografi so pokazali, kako je lovsko-nabiralni.ka ideologija (mitolo­gija, totemizem itd.) strukturirana na osnovi dualisti.nega svetovnega nazora (mo.ki/.enska, polet-je/zima itd.), ki temelji na sistemu delitve dru.be v eksogamne sorodstvene podskupine. Prav ta dua­lizem naj bi poosebljala kamnita dvoj.ka. S pomo.jo pojava avtonomnih rodov poljedelcev morda lahko razlo.imo, zakaj je bil ta brez.asni princip tako pomemben v neolitiku. KLJU.NE BESEDE – simbolika v neolitiku; eksogamija; sorodstvena struktura; lovsko-nabiralni.ka ideologija; prvi poljedelci Introduction The Neolithic site of Göbekli Tepe in Northern Me­sopotamia has raised much interest, but the most relevant questions have mostly remained unans­wered. One reason is that the field of prehistory alone is rather ill-adjusted to properly address mat­ters of ideology. The ethnology of comparable ex­tant societies offers an alternative means of explo­ration (Forest 1992.28–31; Yakar 2005.111–112), as it can reveal the concepts conveyed through the symbolism. The present discussion concentrates on the most imposing feature of the Southwest Asian Neolithic symbolic world, that is, the monumental pairs of twin steles standing in the centre of the PPNA stone enclosures (A to H) of Göbekli Tepe III (Fig. 1). How­ever, these must be conceived just as one particular case among the numerous contemporary isomorphic (architectural, geometric and iconographic) repre­sentations identified throughout (Fig. 2) the Neoli­thic period (Peters et al. 2005.31–32; Stordeur 2003): symmetric clay poles, parallel lines painted on floor, geometric figures on walls, antithetic or converging animals, twin figurines, couple of human skulls, symmetrical partition of communal buildings or of entire sites. A non-exhaustive list of such items DOI> 10.4312\dp.49.2 The Neolithic dualist scheme is presented elsewhere (Bodet 2021; forthcoming). The current paper concentrates on the likely mean­ing enclosed in these isomorphic symbols, or rather, on the ideology and the social structure they reflect. Ethnoarchaeological analogy presents various prob­lems (David, Kramer 2001.51–54), but the symbol in question here appears remarkably central to both archaeological and ethnographic societies. If the analogy proves appropriate enough, the analysis may somehow make the archaeological data ‘speak’ (Gould 1978.250), thereby unlocking some of the meaning enclosed in the symbolism (Wilson 2020.6). Problematic and methodology The fact that the Neolithic symbolic material posses­ses such a meaning is made explicit by the planned and recurrent arrangement in which its various ele­ments are invariably found. The symmetrical so-called ‘pillars’ systematically hold a central position in the communal buildings of a large number of spa­tially and chronologically separated sites in North­ern Mesopotamia, the Levant and all the way to Late Neolithic Central Anatolia1 . Moreover, this element often appears in association with the same set of fi­gurative elements of strong symbolic connotation Fig. 1. Ground plan of Göbekli Tepe. GT_Gesamtplan_2014 (central area truncated), by Klaus Schmidt and Jens Notroff. © Deutsches Archäologisches Institut, Göbekli Tepe Projekt. 1 Little noticed, the female figure in Çatalhöyük gives birth by unifying two parallel pillars (Forest 1993.7), while (symbolically sig­nificant) parallel lines are recurrent in wall paintings. Again, see Cédric Bodet (2021; forthcoming) for a more detailed pre­sentation. Cédric Bodet (Henderson 1964.154), in particular predators, snakes, birds (probably with a psy­chopomp function), bucrania, anthropomorphic statues or smaller side-steles. The order­ing and redundancy of these symbolic compositions are doubtlessly not arbitrary and must correspond to a prede­fined logical system, convey­ing a particular message left to be deciphered (Stordeur 2003.32; Testart 1987a.171). This message carried by such structures, together with the fact that nothing indicates that they supported anything (Jeu­nesse 2020), is, in passing, the reason why the word ‘stele’ is preferred here to that of ‘pillar’ generally used in the literature. In the same trend of thought, the term ‘temple’ is ill-fated to designate these enclo­sures, as these ceremonies probably do not imply a ‘cult of deities’ (Testart 2006a), which only arises when required for the ideological unification of large urban congregations during the much later Obeid/ Uruk horizons (Forest 1996a). Though rarely or too briefly (Voigt 2002.254) menti­oned (Roger Matthews 2003.37 is a significant ex­ception), the decipherment of this symbolic message was successfully initiated nearly three decades ago by Jean-Daniel Forest (1993), providing a solid foun­dation on which further elaboration ought now to proceed. Always placed in the centre, the twin composition is suspected to symbolize the highest sphere of Neoli­thic ideology. It is a simple symbol with which re­searchers are doomed to start with in order to pati­ently reconstruct the meaning it may hold in the so­cial structure (Durkheim 1937.42–45). One precon­dition is not to underestimate the capacity of early communities to express abstract themes through corresponding symbols. A symbol is a signifier standing for a signified. Re­peated over and over again in places dedicated to communal matters, the signified in question must indeed be very significant for the community. This element can only be described, for now, as a sym­bol of symmetry, but it makes sense within a system of thought (ideology) deriving from a correspond-ing social context (Yakar 2005.111). Disconnected from this context, the symbol loses its meaning. This implies that in order to decipher this symmetrical symbol, the social context first ought to be recon­structed, at least in broad strokes. This social coherence is an indispensable basis to start with, but without an intermediary reference, without a Rosetta Stone infusing the structure with meaning, the decipherment will be left to hollow speculations (Schmandt-Besserat 2013.xxv). It hap­pens that, for the present concern, a reference exists. The latter is not a similar iconographic element but an abstract concept, and the correspondence appears too striking and the analogy too compatible not to be considered. Dualism as a prehistoric principle This investigation was originally inspired by Alain Te­start’s (1985) meticulous analysis of what he calls the ‘primeval communist’ societies. These are a pristine form of hunter-gatherers, prior to the advent of bows and arrows. This long Palaeolithic dawn of humani­ty appears, though with much caution, accessible through the abundant ethnographic documentation of Australian Aboriginal societies (Testart 1988.12). Any analogy led through the spectrum of a narrow technological comparison is doomed to failure, and absolutely no cultural comparison is attempted here; the social structure, because of its theoretical and universal nature, is the only element considered. ‘Uni­versal’ is here to be understood in the sense that every human community necessarily has an economy, a kinship system, rules, customs and an ideology: these are the main structures that concern us here. The Neolithic dualist scheme Australian Aborigines are the only living mirror of Pre-Mesolithic-type of societies. A coherent and all-encompassing theoretical reconstitution of their so­cial structure was achieved by Testart, as stable and lengthy as the Palaeolithic period itself, as it is known from archaeology at least. But the ultimate reason why Aborigines are of interest to us here is that, together with Neolithic societies, they clearly appear to hold dualism as a keystone of their ideo­logical construction and that only a living society can reveal its meaning. A word of warning, however. This is not simply about making “connection between two entirely different societies on the basis that they use sym­metrical symbols in their ideology”, as one revie­wer of an earlier version of this paper suspected. There could be a slight chance that isomorphism may relate to something utterly different in the two societies although as dualism and exogamy are ex­tremely widespread (near universal) in the ethno­graphic record, there is a much higher chance that the Neolithic isomorphism is a delayed expression of the Palaeolithic dualism. Moreover, the modern understanding of dualism, “two irreducible, hetero­geneous principles” (Britannica.com) must be here understood as being thoroughly complementary in their opposition, which is what we will try to report. Ethnographers have been struck by the extreme at­tention given to kinship by all traditional societies, and in particular to one critical point (Spencer, Gil-len 1899; Frazer 1910; Howitt 1905): such societies are always divided into several subgroups, at least two, exchanging sexual mates every generation for the sake of procreation. This ‘artificial’ social divi­sion is at the root of exogamy (‘marrying outside’), an absolutely fundamental principle from which later social and kinship systems evolved (Freud 2010[1913].39–53, 255–256; Lévi-Strauss 1967.80– 97; Testart 1985; 1988). One direct consequence of this law is that, as Robert S. Walker et al. (2011.1) say, “arranged marriages [necessarily among rel­atives] are inferred to go back at least to first mo­dern human migrations out of Africa”. The archaeological data offers monumental evi­dence to support the idea that this arrangement per­petuated at least until the Neolithic. It is certainly biased and erroneous to designate a society by what it does not have (bows and arrows). What Palaeolithic-type societies do have, and even more so than later hunter-gatherers, are relations of production entirely based on and shaped by an elabo­rate kinship system. In these small communities turn­ed inward, ‘elementary’ to use Claude Lévi-Strauss’ (1967) term or ‘universal’ to use Alan Barnard’s (1978.69–71; 2020.50–53), everyone is somehow related to everyone else, and this kinship or mari­tal relation dictates the modalities of their social in­teraction in every way. The way individuals respect, joke, avoid, command/obey, punish, teach/learn, give/receive, conduct ceremonials (initiation and fu­neral rites), and, most importantly, the way they marry, are thus prescribed primarily by the subgroup the people in interaction belong to (Malinowski 1926; Radcliffe-Brown 1952.90–104; Woodburn 1982; Ghasarian 1996.152–159,185–197; Walker et al. 2011.2; Bird-David 2019.15–16). “The fundamental feature in the organisation of (…) Australian tribes, is the division of the tribe into two exogamous inter-marrying groups. These two divisions may become further broken up, but even when more than two are now present we can still recognise their former existence” (Spencer, Gillen 1899.55). Even though odd numbers may also be found (as a result of historically induced dispari­ties), ‘primeval communist’ societies are often sep­arated into parallel subgroups or phratries: eight sub­sections, four sections, or, for the most genuine case, two moieties (halves) (Barnard 2020.52). Dualism is thus generally considered to be the most original and purest form of this form of social organization (Cook 2003.65; Freud 2010[2013].50–51; Testart 1978.15–22; 1985.478–479). But whatever the num­ber of subgroups, this plurality is necessarily reduc­ed in conceptual terms to the number two, because it is the number par excellence that embodies the concept of ‘differentiation’ (Girard 1972.87–92), making (equal) exchange possible. Dualism is much more than a marital arrangement. Organically articulated to the economic system, it re­flects on the symbolic sphere: totemism, mythology, rituals, etc. (Testart 1985.451–489). It finds in nature an obvious mode of expression, through fixed oppositions such as day (sun) and night (moon), winter (cold, wet) and summer (hot, dry), and, more particularly, males and females, the interdependence of which is naturally indispensable for the perpetu­ation of the cycle of life and death, oppositions them­selves seen through their own interdependence. This fecund sexual opposition was suspected by An-dré Leroi-Gourhan (1964.108), among others (Te-start 2006b.26), for Upper Palaeolithic societies, and Cédric Bodet by James Mellaart (1967.48, note 27–28), Ian Hod-der or Forest (Matthews 2003.46) for Neolithic ones. Dualism is also the principle behind the famous Yin (female/earth/moon/water) and Yang (male/sky/ sun/fire) of the Chinese tradition (Granet 1929. 225). A similar symbolic partition of fundamental opposite elements is still present today, for exam­ple, in the arrangement of the Berber house in north­ern Africa (Bourdieu 1980). The philosopher Vol-kert Haas likewise refers to the concept of separa­tion of the ‘undifferentiated’ cosmos in primeval times into two sets of opposite but mutually inter­dependent elements, in particular above-heaven-male and below-earth-female (Becker et al. 2012. 30). This widespread differentiation is personified in the antithetic heroes, twins or brothers/sisters in many founding myths all over the world (Girard 1972.247–248). Among Australian Aborigines, this binary interdepen­dence becomes a ubiquitous principle encompassing inorganic elements like mountains, water holes, stars or meteorological events (storms, rainbows). The en­tire world is thus systematically divided into sepa­rated but interdependent halves, a reflection of the society itself, as ideology generally does (Testart 1985.467–489). Dualism thus does not appear as a cultural but as a structural element deeply wired in the constitution of (all?) early human societies. Barnard (pers. com.) tells me that this dualist divi­sion of the society “is true for Aboriginal Australia, but not necessarily for hunter-gatherers in gener­al”. This is a crucial point because it shows the chro­nological and structural evolution from ‘primeval communists’, for which Australian Aborigines are the sole ethnographic representatives, towards ‘la­ter’ hunter-gatherers like the !Kung San, in which relations of production seem to have been altered by a certain spur of individualization (see below; Testart 1985.56–60; 1987b). This evolution would explain the distinction between the ‘socio-centric system’ and the ‘ego-centric system’ made by Alan Barnard (1978.77), as well as, the full “correlation between the system of kin categorization as a whole and the rules of marriage” (ibid. 75) that characterizes the Australians but is not found among the San. All this tends to show how ‘primeval com­munism’ could represent the genuine social back­ground, characterized by “a lack of ambiguity of categorization” (ibid.) and from which later devel­opments are likely to have derived. Among these later developments there is the Neo­lithic period. Right in the centre of the a priori my­sterious symbolic repertoire on display at Göbekli Tepe, there is a pair of huge parallel stone slabs standing majestically, seemingly conveying an abs­tract statement (Becker et al. 2012.14). They ap­pear as nothing but a material representation of this universal dualist scheme. This is the hypothesis that will presently be explored by trying to understand what this dualism is really about. It goes without saying that Australian Aborigines have absolutely nothing to do with Göbekli Tepe, it is just that they seem to share a similar social struc­ture, thus opening the door to a possible analogy, which now needs to be questioned. A major obsta­cle first ought to be removed: if the aspect and cen­trality of the ‘primeval communist’ principle and the Mesopotamian Neolithic symbol present a strik­ing similarity, these societies must be somehow struc­turally compatible for this analogy to function. The analogy Ideology is a central social organ in close interaction with the relations of production (Giddens 1971. 42), which implies that, for the analogy to be ac­ceptable, the mode of production of the societies in question ought to be comparable2 . From this point of view, primeval hunter-gatherers appear starkly different from Neolithic proto-farmers (Willcox, Stordeur 2012.112; Asouti, Fuller 2013.308). How­ever, a sociological rule needs to be considered here: if technical and economic changes can diffuse ra­pidly, their repercussions for ideology (and, subse­quently, for symbolism) are always very much de­layed. This fact has been well attested by anthropo­logists studying the appearance of agriculture. “Peo­ple can hold on to ideologies (mode of thought) re­flecting foraging for generations, even when their systems of production have undergone transition”; “relations of production among proto-agricultur­alists (…) tend to retain the structures of a hunter-gatherer habitus” (Barnard 2007.8,14, quoted by Asouti, Fuller 2013.300). An ideological structure should not be seen here as a conscious and planned construction, but, indeed, somehow like Pierre Bourdieu’s (1980) habitus, that is, continuously shaped by an everlasting accumu­lation of practice and experience. The ideology of the earliest farmers is thus likely to be largely inhe­rited from a Palaeolithic background, built over hun­ 2 In Marxian terminology, “the gathering of food (hunting included) is a form of economic production” (Ingold 1980.83). The Neolithic dualist scheme dreds of thousands of years: Forest (2006.126) thus states without hesitation that “in the case of the Pre-pottery Neolithic, the analogical basis to take into consideration is of course the ‘primeval com­munism’”. This implies that Neolithic people are like­ly to have possessed a cosmogony organized around some form of totemism, animism or shamanism (Le-wis-Williams 2002.132; Bischoff 2002.237; Yakar 2005.112). From there, it is difficult to support the idea that Neolithic people started to cultivate grains because they would have begun (why?) to conceive differ­ently (how?) their relation to deities (are there any?). This is where Jacques Cauvin’s (1997) famous mo­del is problematic, and why the chain of causal ef­fects may benefit if reversed (Testart 1998.27). The Palaeolithic ‘middle range’ The idea that the dualist ideology could be present in Neolithic symbolism as a continuum of a much older tradition would certainly gain some weight if found directly in the Palaeolithic period proper. As mentioned above, Leroi-Gourhan (1964) pointed out such reciprocal dichotomy in the Franco-Cantabrian cave paintings, Lascaux in particular. Horned ani­mals (placed on protruding parts of the cave wall) are supposed to represent a male abstraction (horns being an obvious phallomorphic symbol, see Han­sen 2017), while hornless animals (horses, in con­cave spaces) a female one, their interaction leading to fertility. These images are moreover painted in the very depth of caves, an obvious symbol of Mo­ther Earth’s vagina and womb, where not only hu­mans but all organic forms come to life. “The earth would have been considered the source of all life’s elements” writes Yak Yakar (2005.111–112), speci­fying that such “communal fertility-related rituals may have originated in the period before farm­ing became the principal subsistence economy”. Dimitrij Mleku. Vrhovnik (2021.3) further says that a cave is “a womb and a tomb at the same time”, which would fit well with the above-mentioned idea that life is conceived as taking place in an eternal cycle where death is its inevitable opposite (Gibson 2009[2010.23]). Jean-Loic Le Quellec (2015.259–260) comes to the same conclusion after his comprehen­sive studies of ancient myths from around the world, where humans and animals emerge from a hole in the ground, making the underground at the same time the place where life originates and where the deceased return. The cave paintings thus seem to have put into action symbolically the “structuring principle (of) vitality (fecundity, life-force)” (Ver­hoeven 2002.244), astonishingly resembling the ‘tao’ of the Chinese (Granet 1929.293). But if all forms of life are concerned, it is certainly the community that is primarily envisaged by the principle of fertility. The dualist principle seems to have been known by the Neanderthals as well (Fig. 3): the symbolic com­position found at the bottom of the Bruniquel cave (Jaubert et al. 2016) represents two piles of stalag­mites (another phallomorphic symbol) in one circle (a shape often connected with maternal womb and fertility, see Haland 2017.166), making the com­position strikingly similar to the circle surrounding the twin steles of Göbekli Tepe III (Figs. 2, 3). Since absolutely no cultural connection can be established between these cultures, such similarity can, here again, only make sense if dualism and exogamy are understood as extremely widespread principles among early humans (Freud 2010). Because the animal species painted in the Palaeoli­thic caves are represented as isolated groups, Testart (1985.276–290; 2012.254–267) sees them as totems. His analysis is based on the identification, mostly by James George Frazer (1910) and Alfred William Ho-witt (1905) (who were also Sigmund Freud’s main anthropological sources when he wrote his famous Totem and Taboo), of a strong correlation between totemism and exogamy. Totems are natural species (animals, sometimes plants) representing a specific social subgroup, as if the natural world, classified into species, was called in to naturally classify the community among separate groups. Both Leroi-Gour­han and Testart thus consider that Palaeolithic so­cieties likely knew some form of ‘classification’, in Lewis Henry Morgan’s (1871) use of the term (see also Radcliffe-Brown 1972.98–103; Bloch 1983.8– 13), so as to ensure the practice of exogamy. And we saw that every classification ultimately and theore­tically resumes as a two-fold division, that is, dualism. In the absence of writing, how are people to express what matters most to them, that is, the (male) soci­ety, the (female) engendering principle and exo­gamy, if not by using elements with readily identi­fiable characteristics, such as, respectively, the horn, the circle and isomorphic forms? It is remarkable how early anthropologists from all corners of the world like Morgan in North America, Marce Granet in China, Spencer and Gillen in Austra­lia, Marshall Sahlins in Oceania, Claude Meillassoux Cédric Bodet in Africa or Marcel Mauss and Bronis³aw Malinowski in Me­lanesia recognized related practices of inter-clan exoga­my, cross-cousin marriage or ‘classificatory’ structures to describe the internal organi­zation of pre-state societies, an organization so different from their own Western ‘com­plex’ type of kinship (where marriage is practiced with the outside world, as opposed to closed-in and ‘elementary’, sy­stems to use Lévi-Strauss ter­minology). This all-encompas­sing dualist classification appears as the principle according to which early human society coped with the distribution of sexual mates in order to ensure its own perpetuation. There is thus nothing surpris­ing in finding it all the way to the Neolithic, before the Agricultural Domestic system altered it profound­ly (infra). There seems to linger in Western thought an ethno­centric reflex to consider pre-state societies as un­familiar with elaborate forms of conceptualization (Asouzu 2007.192). The ethnography of hunter-ga­therer societies largely suggests the contrary (Bar­nard 2020). It is much beyond the scope of this pa­per to develop the Palaeolithic symbolic world, but it was essential to show that the Neolithic dualist system is a natural offspring of a much more ancient and complex ideological background (Verhoeven 2001.84). We are now ready to investigate more precisely what this ‘dualist scheme’ is all about. The ‘primeval communist’ social structure and its persistence in the Early Neolithic The relations of production According to the analysis that Testart (1985) pro­posed of at least certain genuine (i.e. matrilineal) Australian tribes (in the southeast, especially), the hunter is not supposed to eat the prey he has killed, but to give it away to the community. “For example, in south-west Victoria, the hunter is said to receive nothing, and his brothers are treated in the same way (Howitt 1904.765)” (Testart 1987b.296). This is the basic opposition this author makes between ‘primeval communists’ and later hunters who usual­ly distribute their prey according to a pattern which “leaves no doubt about the sharer’s close kinship ties” (Bird-David 2019.17–19). Indeed, “possession of a kill in a hunting society confers not the right to its consumption but the privilege of performing its distribution” (Ingold 1980.158, citing Dowling 1968.505). Who, then, appropriates the prey among ‘primeval communists’? It is often individuals be­longing to the social group opposite to that of the hunter (Testart 1988.10). According to Morgan’s (1871) ‘classificatory system’, these groups are de­fined by filiation and generation, thus grouping all siblings in the same class (Radcliffe-Brown 1952). Because of exogamy, the opposite moiety is the one where the hunter finds his spouse. His prey may then go to his spouse’s parents. “Among the Ngatat­jara, the parents-in-law take first and the broth­ers last. (…) Among the Maljangaba of New South Wales, the tribe is divided into matrilineal moi­eties and a man gets very little meat from his ma­ternal kin because they belong to the same kinship group as he does. He receives much more from his father, since he is not maternal kin (Beckett I967. 459)” (Testart 1987b.296). There may be as many rules as there are societies, but it is significant that if the exchange of meat proceeds according to the kinship system among hunter-gatherers, it is in par­ticular the non-producers who generally appropriate the product among ‘primeval communists’. Generally speaking, the producer is never the consumer and the consumer never the producer (Testart 1987b. 294). Because the rule applies to every hunter, the latter eventually always gets his share, and, if the production is denied to the producer, it is, in the end, to the benefit of the society as a whole. Reciprocity as a rule of traditional economies is also well known in the ethnography of Melanesia (Malinowski 1926. 33; Mauss 1924) and elsewhere (Barnard 2020. 31), but it is in Australia that this form of exchange appears the most equalitarian. Comprehensive and equal internal cooperation has thus been identified The Neolithic dualist scheme as the dominant mode of production (Testart 1985. 115,169). The universality of this rule can be questioned for the early stages of humanity, but the general para­digm seems to be that sharing follows the division traced by the kinship pattern in the social body (Speth 2010.xiiiv). This could help shed light on specific archaeological traces. For example, food exchange can be inferred from such data in the PPNA sites of the Northern Levant, where harvested grains are assumed to have been stored in commu­nal buildings (Stordeur 2000.3; 2003.20; 2012), more or less symmetrically divided into two equal parts along the axis of the building (Stordeur et al. 2001.32–33, Fig. 5/1). Though this is nothing but the author’s speculation, it could be that this sym­metrical division of the village granary was made according to the kinship division within the com­munity for matters of exchange (each subgroup pro­ducing and storing for the other). This division goes much beyond economic matters. At Faynan, in the southern Levant, the communal buildings are cha­racterized by “general symmetry to the structure along an axis formed by a deep trough” (Mithen et al. 2011.354). Sometimes, such complementary division has been identified at the level of the entire site: “the small settlement at Qermez Dere had been laid out in two contrasting halves that per­formed complementary functions. Part way through its life, the village was re-formed, but once again in two complementary halves” (Watkins 2006.16). Leaving aside the case of Asikli, which has a street dividing the village (Özbasaran et al. 2012. 140) but in less clearly symmetrical parts, Hodder (2012.304) identified such an arrangement at Çatal­höyük, with “a large dip or trough across the mid­dle, dividing it into two hills. The mound does seem to have developed in two halves (north and south). (…) In addition we have found some dif­ferences in the genetic make-up of the humans buried in the two halves”. Finally, in the late Neo­lithic (but in fact, contemporary with the local emer­gence of the farming system), Ulf-Dietrich Schoop (2005.49), concerning the “lines of parallel houses facing one another” in Hacilar and several other western Anatolian Late Neolithic sites, writes that “this brings to mind the social organization known in the ethnographic record as the ‘moiety system’, in which a community views itself consisting as two competitive3 halves. I do not wish to elabo­ rate on this, for at the present state of investiga­tion it would only be grasping at straws”4 . This pre­caution appears academically wise, but the straws grasped here are arguably nothing but the very root of prehistoric ideology. A slow change can be detected in the following PPNB period at Çayönü, with granaries attached to every house from the Grill Building phase onward indi­cating “that economic emphasis may have been shifted from community to family based produc­tion and consumption” (Yakar 2003.442). This re­organization indeed seems to reflect a slow trend to­wards an economic (and marital) autonomy of lin­eages, following a “segmentation and separation of balanced components arranged in relation to each other” (Hodder 2020.49–51), and possibly leading in late Neolithic Çatalhöyük to “the House as a his­torical and genealogical social unit” (Kuijt 2018. 584; infra) based on a line of ancestors. This late Neolithic fission into autonomous families apparent­ly emphasizes, by contrast, the closed-in reciprocal pattern that was arguably still strong in the earliest Neolithic. This economic evolution also seems supported by the genetics of wild food. The PPNA plant material, though anthropologically managed, is not morpho­logically domestic yet. The subsequent physical do­mestication, eventually including animals, implies that the originally loose farming mode of produc­tion is, relatively speaking, gaining in intensity dur­ing the PPNB (Zeder 2011.230; Willcox, Stordeur 2012.112; Asouti, Fuller 2013.329). This trend goes well with the idea of weak communal production (PPNA) gradually intensifying (PPNB) towards the specific interests of each lineage (PN), in particular for the constitution of bride-prices (herd animals), suspected elsewhere (Bodet 2019b) to have begun in these latter periods. This morphological evolution of resources emphasizes, by contrast again, the eco­nomically loose, reciprocal and equal form of food exchange expected to characterize the earliest Neoli­thic groups. The relations of reproduction Economic reciprocity appears to reflect marital pat­terns. Among hunter-gatherers “marriage prescrip­tions commonly involve real or classificatory cross-cousins and (…) exchange between two kin line­ages” (Walker et al. 2011.4). The parents of ‘cross­ 3 We will see that the division is, at this stage at least, not about competition at all, and in fact quite the contrary (infra). 4 I am thankful to Çiler Çilingiroglu (pers. com.) for bringing this reference to my attention. Cédric Bodet cousins’ are the children of a brother and sister; because the social affiliation comes from either the father or the mother, cross-cousins necessarily be­long to a different ‘class’ and are expected to marry. This is just one straightforward example; there are many possible types of marital alliance, with many more subgroups but the founding exogamic princi­ple remains everywhere the same: the hunter does not ‘consume’ his sister but the sister of a hunter from the opposite moiety (hence the famous ‘ex­change of sisters’). Close-kin mating is often thought to have been pro­hibited to lessen biological complications coming from consanguine mating, but there is no genetic difference between a cross- (prescribed mate) and a parallel-cousin (proscribed). The key point is that healthy mate circulation in the long term implies not only separate groups (at least two), but reci­procity among them. It is obvious that this social di­vision is not prompted by emulation or rivalry but, quite on the contrary, by the welfare of the entire society. “A social benefit results from an exogamic marriage (…), the law of exogamy is omnipresent (…). It is the archetype of all other manifestations at the basis of reciprocity, it provides the funda­mental and immutable rule ensuring the existence of the group as group” (Lévi-Strauss 1967.551). Ex­change and the social classification permitting it thus appear again ultimately as modes of (re)pro­duction. The tight infrastructural correspondence be­tween the kinship system and the economy seems to strengthen the fundamental role played by reci­procity and dualism in the ideology (Lévi-Strauss 1967.48–170; Bloch 1983.9–10). As for the reflection of these principles in archaeo­logy, the internal subdivision of the PPNA commu­nities suspected above may be continuing in PPNB Çayönü, where the two large buildings just north of the Plaza could house the elder(s) of each moiety. Another contemporary hint is found in Nevali Çori with “two groups of houses with different orienta­tion (…) that could have belonged to two groups of families with different lineage” (Yakar 2003.443), two groups expectedly related through permanent intermarriages. Whatever the case, all hunter-gather­er societies seem to know one form or another of kinship classification (Ghasarian 1996.31; Walker et al. 2011.1): there seems to be no viable reason not to expect a similar system of reciprocal mate ex­change among Neolithic communities. And this could be what the twin steles state out-loud. Totemism Some form of totemism seems rather common among hunter-gatherer communities, though its ab­sence among the San shows it is not universal (Te-start 2006c.149; Barnard 2020.46). It is neverthe­less thought by a number of specialists cited by Freud (2010.42, note 2), in particular Frazer (1910), to have been very widespread at an original stage which would correspond to ‘primeval communism’. There is, again, no a priori reason to exclude its pre­sence in the Neolithic, as strongly suggested by the twin stele arrangement. The totem consist of natu­ral species, usually an animal, considered to be the ancestor of a clan or tribe. Individuals maintain a very specific relationship with it, being strictly for­bidden to consume it, except once a year during the ritual known as ‘Intichiuma’, aimed at magically in­creasing the totemic resource for the opposite moi­ety to consume (Spencer, Gillen 1899.169). This is the symbolic projection of the rule of reciprocity re­viewed for production (the hunter not eating his prey but hunting it for the other moiety to consume) and for reproduction (individuals not marrying within their own subgroup, but ‘producing’ children to be ‘consumed’ by the opposite one). By way of animal species, totemism can be conceived as a sym­bolic representation of the dualistic kinship system ongoing in the community (Freud 2010.203–204, 255–256). Göbekli Tepe, sometimes seen as a place of inter-clanic reunions for very extended kinship groups (Schmidt 2001.52–53; Belfer-Cohen, Goring-Mor­ris 2002; Peters, Schmidt 2004.210–212), would fit well as a place where Intichiuma-type ceremonies were taking place. In fact, if contemporary levels are to be found, the high number of large early Neoli­thic sites, like Karahantepe, recovered within a ra­dius of about 20km all around Göbekli Tepe (see be­low), may have composed this population (Bodet 2019a). For Hans Georg K. Gebel (pers. com.), “the ideolo­gy of the early Göbekli Tepe Culture represents a symbolically sustained system needed to serve the integration of growing group numbers (…). Mu­tual understanding and conflict management of groups not knowing each other were reached by commonly accepted strong and binding ideologies and conventions. One may speak of ideocratic ter­ritories mediated through the fixed image pro­grams”. A common ideological background indeed certainly played an important binding role among all these communities, and, looking at the homoge­ The Neolithic dualist scheme nous symbolic program on the entire site, this bind­ing element must have been related to an isomor­phic, i.e. dualist conception, shared long before the construction of Göbekli Tepe. The Neolithic iconography could fit neatly with the totemic analogy, by supposing, like the excavators of Göbekli Tepe and other specialists (Peters, Schmidt 2004.209–212; Kornienko 2018.17–18), that the animals carved on the steles represented totems. However, for Forest (2006.134) this is pro­bably not the case, because these elements interfere with each other, being complementary or synony­mous, so as to convey a message, while totemism simply classifies in purely equal terms. Totemism and message/law indeed stand at different levels in the ideological structure, but are not at all incom­patible (Testart 1985.510). The numerous animals represented on the side steles could fall under Fo­rest’s warning (Bodet 2021), while the few ones on the central steles, absolutely alone, may be more in tune with a totemic classification. Though intrinsi­cally related in form and signification, we will see that side and central steles may hold a different symbolic value. The rich animal repertoire represented on the side steles is very similar to that deciphered by Forest (1993; 2003) at Çatalhöyük: bulls, predators, psy­chopomp birds, found to symbolize the society, death of the body and transportation of the soul, respectively. Göbekli also has many snakes, under­stood as lineages (Forest 2006). The same geomet­ric elements, thought to symbolize the two moieties or the sexual mates they exchange, are also found on both sites (and many others): parallel lines, zig­zags, triangles, and chevrons. The several side steles could then represent the subgroups the community is composed of, linked to each other by the circu­lar stone wall, forming a large matrimonial self-suf­ficient unit. The non-totemic (message-delivering) symbolic animals carved on them would suggest the endless (feminine) cycle of life (snakes) and death (birds, predators) in which these groups were in­volved (Bodet 2021). On the other hand, if the cen­tral parallel steles stand for the dualist subparts of the (masculine) society (product of the feminine principle), the isolated animals carved on them may indeed be totemic. For example, the reflecting foxes on the twin steles of enclosure B (Peters, Schmidt 2004.184) would represent the two inseparable but distinct subparts of the same totemic clan (see Ma­linowski 1926 for compatible ethnographic exam­ples). A look at a tightly interwoven subject – mythology – will allow a more comprehensive understanding of the dualistic nature of totemism. The mythology The species included in the totemic partition have a correspondence in the mythology. The main char­acters of the Australian Aboriginal myths of the Dream Time (Testart 1978; 1985.390–395), with a correspondence worldwide (Girard 1972), are gene­rally divided in two types: the violator and the coun-ter-violator. For anything to happen in the founding myths, the fundamental rule (exogamy) must be violated, which invariably entails a counter-violation. The widespread myth of the eagle (the hunter, the creator, the counter-violator) and the crow (the sca­venger, the trickster, the violator) illustrates this point. The crow steals the fire from the eagle (viola­tion), and, as he escapes, he drops the fire, allowing humans to capture it and cook food. The eagle aven­ges himself by causing a huge fire that threatens humans (counter-violation). Some variants of this myth are about stealing water (violation) and so permitting life but provoking floods when uncon­trolled (counter-violation). Myths thus explain the origin of the society as an interaction between two opposite but interdependent poles (Testart 1978.95, 118–125; 1985.384–387,432–444). Just like later religions, mythology aims in the end at securing the social order through ideology. Forest (1993.17–21; 2006.134) reads the elements in the iconography of Çatalhöyük as principles con­veying a message: life and death on a vertical line, two exogamic moieties on a horizontal one, all inter­secting in a cross pattern to permit the existence of the community, thus recalling its fundamental rule (exogamy). This message is essentially the same as the one present in myths. It is open to question whe­ther the animals in the iconography, strikingly simi­lar in Göbekli Tepe and Çayönü despite a wide chro­nological and spatial gaps, directly represent mythi­cal characters. But in the end it matters little, since they are likely to stand for the same opposition among interdependent elements (as an image, again, of the subgroups of the society). Just like in many myths of the primeval world, a dualistic oppo­sition can be suspected with a certain degree of con­fidence among the antithetic heroes of Neolithic my­thology, symbols of a fertile opposition. René Girard (1972.88–95) presents a somewhat dif­ferent interpretation of early myths and rituals, fo­cusing on sacrifice as catharsis, expelling the tensions Cédric Bodet accumulated within the community on an innocent victim, but the ultimate goal remains to prevent the ‘divided’ community from the risk of becoming “un­differentiated”. Jungian psychology also notes the case of twin snakes in mythology. “These are the fa­mous Naga serpents of ancient India; and we find them in Greece as the entwined serpents on the end of the staff belonging to the god Hermes5 . An early Grecian herm is a stone pillar (…). On one side, are the intertwined serpents (in the act of se­xual union) and on the other an erect phallus: we can draw certain conclusions about the function of the herm as a symbol of fertility. (…) But Her­mes is (also) Trickster (…) the leader of souls to and from the underworld” (Henderson 1964.155). The two snakes appear as the dualist lineages whose union alone can engender society. The blood ideology Based on ethnographic data, Chris Knight et al. (1995.89,93–97) have proposed a Palaeolithic “sym­bolically structured sexual division of labour” where, notably, the recurrent use of red ochre would be utilized in menarchal rituals to symbolize fertility. This interpretation could fit well with Te­start’s (1985.345–475) reconstitution of the prime­val mind, according to which the female compen­sation for the masculine blood-soiled meat brought by the hunter is the feminine blood-soiled newborn child. The widespread presence of red paint on the floor of special buildings all through the Neolithic of south­western Asia (Gökce 2021) supports the idea that blood played a fundamental symbolic role. At Çayö­nü, actual traces of human (and animal) blood have been detected on a one-ton slab in the courtyard of the so-called ‘Skull-building’ (Özbek 2004.20). Given what is known about these societies, notably the classification according to generations, it seemed li­kely to Forest (1996b) that this blood was that of initiation rites. The blood of circumcision (symboli­zing the first hunt?), equivalent to that of the (first) menstruations (Doyle 2005.280; Knight et al. 1995. 95), can be understood as a separation between two crucial statuses: not simply synchronically between male and female (or between their respective moi­eties), but diachronically between consumers (chil­dren) and (re)producers (adults). Through its intimacy with both life and death, blood is thus considered by Testart to have played a syn­thetic role in the primeval communist ideology. A 5 This is the caeduces, still symbol of modern medicine. symbol of order and life when running in the closed system of the veins (=exogamic rule respected), blood represents chaos and death when running out of a disrupted vein system (=exogamic rule violat­ed). Just like for the mythical figures (divided into violators and counter-violators), for the natural spe­cies in totemism (divided among social subgroups) and for society (divided into parallel moieties), blood, the one and same blood, is artificially divided and separated into distinct but mutually interdepen­dent classes so as to promote their mutual interde­pendence and strengthen the unity of the whole. Synthesis: the dualist scheme Because the data mobilized here is not archaeologi­cal in nature, it is perhaps not superfluous at this point to synthesize what we have proposed. The kin­ship dual classification of the society is reflected in the economy and ideological structures, such as tote-mism, mythology and blood ideology. Beyond a me­chanical Marxist view that would present the social superstructure as invariably determined by the eco­nomic infrastructure, kinship thus appears to domi­nate the primeval communist relations of produc­tion, making the reproductive infrastructure the very root of dualism. Yet, as Alan Barnard (1978.78–79) writes, “Australian systems differ from other uni­versal systems in that Australian universality is not confined to kinship”, it is “closely connected with totemism and with other aspects of cosmol­ogy” (which) “divide the universe -nature and cul­ture alike- into named categories [which] repre­sent a concept of world order in which kinship is only a part”. It thus seems that what determines kin­ship and all other structures is, in fact, the exogamic principle which must be conceived as an overarch­ing pattern imposed on the entire social fabric. Op­posing sets of the natural world like male/female, sun/moon, winter/summer, water/fire, dry/wet or life/death are ‘given’ to humans, who use them as symbols to express and justify the only opposition on which they have a hold, the division of society itself into exogamic lineages or moieties. In other words, just as in later religious systems (Forest 1996a), the cosmos is mobilized to promote, through its own perpetual and fecund oppositions, the perpe­tuation of the fragile opposition between lineages or sub-clans in order for the society as a whole to re­produce. It is this all-encompassing exogamic divi­sion that is termed the ‘dualist scheme’ by Testart (1985.207–218, 477–515), and which, as we intend to show, was still very vivid in the Neolithic. The Neolithic dualist scheme The Neolithic dualistic symbolism Two parallel steles seem to be the symbol chosen by Neolithic people to represent the concept of exoga­my/dualism. Given the pervasive twin steles present in a large number of sites (Bodet forthcoming), it appears that early Neolithic communities felt the urge, maybe more than Palaeolithic ones, to recall and impose this principle. Is it possible to be more precise as to the message conveyed and to the cause of the monumentality given at this precise moment to an immemorial principle? The stele as a symbol of the lineage In the communal building of Nevali Çori (devoted to reunions, given the bench running at the base of the surrounding wall), Alexis McBride (2013.54) pro­posed that the anthropomorphic steles inserted in the bench are ‘participants’ along with the real hu­mans seating there. This makes sense indeed, but the verticality of the stele must be meaningful. Ac­cording to several researchers, like Forest (1993.7), Tatiana V. Kornienko (2018.17) or Christian Jeunes­se (2020.54), the stele stands for a genealogical line of ancestors related over time, that is, a lineage, built up generation after generation. This interpre­tation is well supported by ethnographic observa­tions: in the American North-West coast, the “totem poles, house posts, memorial posts (…) record the household’s lineage” (Banning 2011.626). The end­less continuity in time of the lineage could be the reason why certain Neolithic steles are reused at the same location phase after phase (Watkins 1996), while others are buried or ritually broken. Seeing side steles as lineages implies that each cir­cle could represent a larger social group, like a tribe or clan, divided into a number of subsections, an idea already alluded to, in one way or another, by several authors (Belfer-Cohen, Goring-Morris 2002; Yakar 2013.438; Hodder 2020.50). Beyond the re­presentation of the cycle of life and death as suggest­ed above, the surrounding wall of the Göbekli Tepe enclosures could bind the lineages in an endless cir­cle of marital exchange, a stone materialization of a closed-in ‘generalized-type’ of kinship pattern6 (Lé­vi-Strauss 1967), where lineage A gives a mate to lineage B, B to C… back to A (Bodet 2012). There would then be between four and twelve subsections for each tribe; interestingly, twelve is also the num­ber of subgroups chosen by Freud (2010[1915].51) to present a typical totemic society. The intercon­nected side steles would then stand, like a temenos wall, as a transition between the real world and the sphere of pure abstraction, which, we will now see, seems to take stage at the centre of the circle. The reciprocal relationship Central and side steles have the same monolithic structure and same morphology, the latter being simply smaller and less well executed. For the sym­bolic program in question, they are likely to be a re­lated signifier standing for a related signified, but on a different scale. If side steles may represent the actual lineages making up the community, the cen­tral twin steles would then represent two parallel lineages, but on a purely conceptual level. They seem to stand for an abstract idea, an allegory of the exo­gamic rule, the active (feminine) principle of ferti­lity. Parallel, isomorphic and face-to-face, the twin steles seem to express a bilateral relation of strong symmetry, where each subgroup is at the same time the donor and recipient of a (marital) transaction. Behind the exogamic rule (Forest 1996b.29), the composition seems to express the type of social re­lation that exogamy entails, one of pure reciprocity. As Lévi-Strauss (1967.97) puts it, the dualist prin­ciple is itself only a modality of the principle of rec­iprocity. This emphasis on the relation itself appears hindered precisely by the fact that the twin steles do not ge­nerally enter in relation with each other, just like parallel lines painted on floors (also recalling line­ages). The relation between the steles is suggested elsewhere: a low bench, a slight clay lip or a slab set on edge at Qermez Dere, Beidha, Çayönü (Skull Buil­ding), Musular and at late Göbekli Enclosure A (E PPNB) (Watkins 1996; Makarewicz, Finlayson 2018; Erim-Özdogan 2001.208; Özbasaran et al. 2012; Schmidt 2001.50 respectively). In the enclosures of Göbekli III, twin steles can be said to be connected by the ground, Mother Earth, of which we saw the importance for the concept of fertility. But there is more. The lack itself of any obvious connection be­tween the parallel steles implies exactly the con­trary: perpetually reflecting each other (as well as the – totemic? – decorations carved on them), one is nothing but the permanent counterpart of the other, each stele fundamentally dependent on the other to exist. The intrinsic relation between the two steles is conspicuous by its absence. 6 https://www.britannica.com/topic/kinship/Alliance-theory – for an introduction in English to the work of Lévi-Strauss. An inspired interpretation of his work has also been proposed by Barnard (1978). Cédric Bodet The composition seems to be bluntly saying: a moi­ety is the mirror of its counterpart, and only the two together, as equal partners in the (marital) exchange, can engender the society altogether, which, in turn, can only exist divided into equal subparts intrinsi­cally bound to each other. But if twin steles repre­sent an abstract sphere of symbolization with such majesty, it is exactly because this notion is not sim­ply an allegory: reciprocity must be conceived as a law governing social conduct. In other words, the full message conveyed by the twin composition ap­pears as such: ‘marital reciprocity must be respected for the sake of the whole society’. We will later see why this antediluvian rule took such a ‘monumental’ urgency in the Neolithic. The twin steles as a symbol of fertile regenera­tion Let us first complete our reflection on the striking fact that the most central place of the entire com­position at Göbekli Tepe (see also Jerf el-Ahmar) is the space left ‘religiously’ empty between the twin steles (Fig. 4). This vacant space must have been filled with meaning in the eyes of the audience. We saw that at Çatalhöyük two symbolic pillars are con­nected by the limbs of a feminine figure giving birth to a bull (which is a good enough reason to see this representation as a metaphor and not as a realistic scene). Again, according to Forest (see English sum­maries in Bodet 2012.7 and 2021.149–151), this is not a woman, but the personification of the princi­ple of regeneration, engendering the society. On the same line of thought, the space between the twin steles at Göbekli can be understood as a threshold to life (a symbolic vulva?) and death (a symbolic ‘swallowing’ mouth), be­cause it is a representation of the va­gina dentalia, the ‘toothed vagina’, a widespread mythological female principle of both regeneration and destruction among traditional socie­ties (Forest 1993.22; Ross 2021). This invisible principle of the regene­ration of society and all life forms would then be put into action by the mutual interaction existing between the exogamic moieties represented by the two monoliths. This is a prac­tical illustration of Trevor Watkins’ (2006.21–22) statement that “archi­tecture is a specially powerful mode of external symbolic storage”. And these symbols convey a specific me­aning. The Neolithic twin steles seem to state: ‘the eternal cycle of life and death can only be put into action by the principle of pure reciprocity’. Such pre­historic capacity of abstraction can only be a surprise to ethnocentric prejudices. McBride (2013.59) further suggests that the partic­ipants in a ritual or ceremony may have been asked to walk through the central steles at Göbekli Tepe. The idea deserves attention. This particular space, here putatively interpreted as the principle of rege­neration of both life and death, would indeed be the ideal place to have adolescents pass through dur­ing their initiation ceremony, initiation being con­ceived in many societies as the death of the child and rebirth as an adult (Weiss 1966.72; Henderson 1964.120–121; Forest 1996b.28). According to Max Weber (1920[1996.184]), initiated aristocrats in Chi­na or India call themselves ‘the twice-born’. The link between dualism, totemism and initiation is further supported by the fact that in order to become hun­ters, young Aborigines are systematically initiated by the opposite moiety. The same holds true for funer­als and Intichiuma ceremonies (Testart 1987b.299). Regeneration may also be the main principle dis­played in the Franco-Cantabrian caves (also an ideal place for ceremonies of initiation). In that case, the cave paintings would ‘magically’ assist these princi­ples by being represented in the ‘womb of Mother Earth’ (Henderson 1964.146–153), the latter also being a recurrent theme in Carl Jung’s ‘collective un­conscious’. An analogy can also be made with the Turkish custom of Hidirellez, still performed today, The Neolithic dualist scheme where the drawing of babies and cradles on the sand or earth in springtime (when nature comes back to life) is believed to enhance fertility7 . The be­lief that the mere representation of symbols has a ‘magical’ active power is universal: it is the same when Christians hang a cross in their homes or Mus­lims a picture of the Mecca. This (and probably not a belief in a deity) would also explain why so many crude female figurines (representing the same alle­gory of fertility) were so common in domestic con­texts in the Levant or Central Anatolia (Cauvin 1997.46–49), as well as many hand-size T-shaped fi­gurines in the Urfa area (Hodder 2020). This same belief could finally explain why symbolic enclosures in sites like Göbekli Tepe, Nevali Çori, Sefertepe, Ka­rahantepe (Moetz-Çelik 2012.699) and as far away as Çayönü, Qermez Dere (Upper/Middle Tigris ba­sin), Jerf el Ahmar (Middle Euphrates) and Beidha (Jordan), were carefully buried upon abandonment8 , arguably so as to preserve the active principle of fer­tility they enclosed. Epilogue: a likely cause for the monumentality This investigation must now be placed in its socio­economic context, thereby answering a last impor­tant question: why such monumentality, especially if, as discussed above, the displayed concept had been a basic one for tens of millennia? We must come back to the idea that in spite of its ‘universality’ (Malinowski 1926; Lévi-Strauss 1967. 3–29,49), the division of society into subgroups is not founded in nature, it is a social product; this im­plies that nothing can physically guarantee the re­spect of exogamy, and that its importance must be permanently reinforced in the community, in parti­cular to the newly initiated generations. In hunter-gatherer societies marriage is not left to the free-will of individuals; it is codified by tradition and contracted among more or less closely related individuals, like cross-cousins (Lévi-Strauss 1967; Walker et al. 2011; Barnard 1978; Ghasarian 1996. 147–174; Bird-David 2011). The distribution of se­xual mates is thus regulated so as to avoid a dange­rous anarchy for the entire community. Once set­tled, communities naturally continued this immemo­rial tradition of ‘prescriptive’ (or pre-arranged) mat­ing, every new generation being bound to stay with­in the village so as to comply with this systematic exchange. Coupled with the fact that settled life and farming naturally lead to a strong demographic growth (Bellwood 2005.61–64), the consequence of this alliance system is that this growth is largely local. This age-old inward-looking ‘elementary-type’ (whether ‘restricted’ or ‘generalized’) of marital al­liance rule (i.e. among the subgroups of the tribe) is most likely the ultimate cause for the appearance of Late Neolithic mega-sites like Çatalhöyük, Halula, Ain Ghazal, Shu’eib or Basta (Forest 1993; Bodet 2019a; this probably goes also for Neolithic mega-sites elsewhere like the Trypillian sites). Thus, if at Göbekli the symbolic emphasis is monumentally placed on the community altogether, in the much la­ter Çatalhöyük horizon the same concern gradually shifts towards the intimacy of the (autonomous) li­neage itself, or “multiple single-family households”, to use Kuijt’s (2018.565,584) words. Such evolution did not go without problems. In farm­ing families the elder son traditionally inherits from his father not only the estate, land and animals, but also a decisional power over his younger brothers and sisters, especially in terms of alliances. This leads to a growing internal stress with younger individu­als searching to withdraw from the domination of their elders by splitting from the group, a situation well described in comparable ethnographic cases (Sahlins 1961.324–327; Meillassoux 1991.51–52, 122–124). For Çatalhöyük, Forest (1996b.5) devised a similar incongruous situation, all the more so that farming allows for (and is much more efficient with) small producing units (nuclear families) spread over the landscape, each family/farm on a separate piece of land. The resolution of this inextricable situation had to wait for the abandonment of the prescribed ‘elementary-type’ alliance system (among related in­dividuals). This is indeed what seems to have hap­pened nearly everywhere by 6500/6000 BCE at the latest, as suggested by the (gradual, then total) de­sertion of all mega-sites, followed by the establish­ment of gradually smaller farmsteads spread around, wherever land allowed for farming and herding. Huge Çatalhöyük East thus gives way to relatively smaller Çatalhöyük West (though this site still re­mains rather large); relatively small Musular, found­ed towards the end of large Asikli, may represent an earlier (Late Neolithic), because eastern, example of the same process of site segmentation. This trend will continue throughout the early Chalcolithic, rea­ching its apex with the Halaf culture (Forest 1996a. 27–35). 7 See ethnographical support for this idea in the authors’ video (in Turkish): https://youtu.be/hTl3eG6wTqM 8 This could still hold true even if the burying of Göbekli Tepe was initiated by slope-sliding and inundation (Kinzel, Clare 2020. 33). Cédric Bodet The expected tense social context just referred to is crucial to understand the monumentality of the ste­les of Göbekli Tepe. In charge of maintaining the moral conduct of the society, the elders are con­stantly recalling, imposing and teaching the new (initiated) generations about the old traditions, in particular the reciprocal exchange of mates. They would have been particularly anxious and careful to avoid any disruption in the smooth circulation of women, and therefore, to counter, with the help of symbolism, the splitting of younger people from the larger kin-group. Indeed, the elders of each social subgroup, responsible for giving mates to the other subgroup(s), could not fulfil their duties if these younger individuals had left. From their point of view, such a fission would invariably lead to an out­break of the inextricable internal violence that is known to have scared these societies so much (Gi­rard 1972). The fear of the loss of reciprocal exchange and of the cohesion of the community is hypothesized to have pushed the elders to express the old rule of (exogamic) alliances with much force and promi­nence by ordering the erection of the monumental twin steles. It is in this sense that we propose to illu­strate Thomas Zimmerman’s (2020.14–15) intuition that the symbolic program of Göbekli Tepe reflects much more a Palaeolithic cultural collapse than the advent of a new one, the way Cauvin (1997.50–55) sees it. “It is doubtful that the supernatural world-order envisaged by earlier hunter-gatherers would have been entirely altered by new spiritual con­cepts”; (the Neolithic) “repertories of symbols (…) seem to have their origins in earlier periods” (Ya-kar 2005.111). “The establishment of such symbolic systems, or the externalisation and canonisation of symbols, is not the result of a cognitive process but rather the result of a basic need, the need to sustain a current life mode by coping with newly arising social and ideological challenges of fast growing social aggregates in the Upper Mesopotamian gras­slands”: this statement of Gebel (2013.40) applies very well to our views, provided that the “basic need” in question is first and foremost that of a su­stained marital alliance system. The monumental isomorphic steles understood as an enforcement of respect for the old reciprocal al­liance rule, can thus be seen as a form of propagan­da, erected in the face of the threat of being aban­doned at a time when Agricultural Domestic lineages were gaining economic and marital autonomy. These ‘monumental’ fears were indeed justified as, in spite of all these efforts, elementary alliances will prove obsolete by the Chalcolithic. This necessarily implies that nuclear families (a married couple and children), breaking free from their larger family groups, proceeded to ‘complex’ types of marriages taking the form of ‘contracts’ (hence called ‘allian­ces’) among unrelated larger families, probably se­cured by material transactions like the bride-price (herd animals) (Bodet 2019b). In such a context, exogamy became reduced to the prohibition of in­cest (Ghasarian 1996; Forest 1996a) and dualism naturally lost its ground as a principle of alliance to­gether with its ideological relevance. In Gebel’s words (pers. com.), “dualist schemes may even become extinct in early productive environments when strong relational ordering principles help or suf­fice to organize lineages and the societies they are part of”. Synthesis: Dualism as a Neolithic scheme By shedding light on the archaeological data using ethnographic social structures, we have here attempt­ed to review how the Neolithic revolution transform­ed the Palaeolithic society into an Agricultural Do­mestic one. We were greatly helped in this task by the Neolithic symbolism on which social changes were invariably projected. This structural evolution is synthesized in Table 1 and Figure 5. We started our investigation on the premise that hunter-gatherer societies do not marry with outsi­ders and are, as a rule, divided into (at least two) subgroups (moieties or lineages) as a direct outcome of the universal rule of exogamy so as to secure the distribution of mates and reproduction of closed so­cieties: the hundreds of early societies reviewed in The Elementary Structures of Kinship by Lévi-Strauss (1967) as well as general handbooks (Gha­sarian 1996) or articles (Barnard 1971; Walker et al. 2011) on early kinship make this point clear. This seems to apply during much of the prehistoric peri­od, as symbolic representations in the depths of Mid­dle and Upper Palaeolithic caves seem to suggest. Throughout the Mesopotamian and Anatolian Neo­lithic, this tradition continues with an impressive se­ries of isomorphic representations, in particular twin steles. The message seemingly conveyed by this dua­list symbolism can be read as follows: ‘only the rec­iprocal (marital) relationship ongoing between the moieties (lineages) composing the society can allow for the society to reproduce safely’. Beyond the kin­ship pattern, reciprocity encompasses all other as­ The Neolithic dualist scheme Period Economic Social st. Relations of Dualist Compatible structure Kinship production symbolism Ethnog. Paleolithic communism Nomadic Closed-in Primitive Neandertal Aust. Abori Hunter-gather Moiety sys communism Bruniquel Spencer&Gill. Late Up. Pal. Meso-Epip. Hunter-gath. Classificat. Egalitarian Franco­ !Kung, Inuit w\ bows\arrows Morgan (Woodburn) Cantabrian Barnard PPNA (N. Mesop) Pre-dom. Agric. Lineage Communal Monumental Trobriand (Willcox) formation reciprocity Göbekli Malinowski PPNB\PN (C. Anat) Domestic Agric. Segment. Lineage Domestic Baruya (Peters, Zeder) lineage based Çatal Godelier Chalco (Halaf) ‘Agricultural Do­ ‘Complex’ Domestic Lineage-base Gouro mestic’ (Sahlins) (open) hierarchy Latmos Meillassoux Tab. 1. (Very) rough evolution of prehistoric social structures (a preliminary attempt). pects of the hunter-gatherer social structure, first and foremost its economy. Testart has therein come to the conclusion that dualism generally represents a fundamental scheme determining the structural and ideological composition of pre-state communi­ties. It is suggested here that this dualistic ideology finds its most phenomenal transcription in the mo­numental central twin steles of Göbekli Tepe, at a time when it was in danger of being supplanted by the advent of a whole new social and ideological or­der, the Agricultural Domestic System, founded on autonomous (unrelated) agricultural lineages orga­ nizing marital alliances (with bride-pri­ces) freely among themselves. As agri­culture diffuses towards the west, and is appropriated by local hunter-gather­ers, the same emphasis on dualism ap­pears, this time on a more modest scale but widespread in every domestic con­text, as in Çatalhöyük. This latter social system (lineages, autonomous in mari­tal terms) was destined, a few millennia later, to aggregate hierarchically into city-states on the pattern of the status differentiation between elder males and women/youngsters in the agricultural family itself (Forest 1996a; Meillassoux 1991). According to the dictionary, a scheme is “a large-scale systematic arrangement for attaining some particular object or putting a particular idea into effect”9 . For hunter-gatherers, the ‘arrangement’ is the reciprocal partition, and the ‘ob­ject put into effect’ the regeneration of the community. The symbolic reperto­ire of Göbekli Tepe and of a large num­ ber of sites up until Çatalhöyük could show that this scheme may apply all the way to incipient farmers. It is now possible to clarify a theoretical problema­tic raised above, and state that there is no coinci­dence in recognizing dualism in the ideology of so­cieties as geographically, chronologically and cultu­rally distinct as the Neanderthals, Australian Abori­gines, Magdalenians, Ancient Chinese, SW Asian Neo­lithic or early historical Mesopotamians. This conver­gence becomes structurally logical when the ideolo­gy of these societies is ultimately determined by Fig. 5. The Neolithic Dualist scheme seen as an evolution of so­cial structures (synthesizing graphic). Dualist and closed-in (‘elementary’) Palaeolithic societies confronted with growing autonomous Neolithic farming lineages trigger an ideological conservative reaction (monumental: Göbekli III, omnipresent: Çatalhöyük), but finally evolve towards open (‘complex’) Chalco­lithic Domestic lineages. 9 https://www.encyclopedia.com/science-and-technology/computers-and-electrical-engineering/computers-and-computing/scheme Cédric Bodet strictly equal relations of (re)production (compen-of this remarkably stable social construction. But the sating for low productive forces). The human consti-ultimate illustration for the central position held by tution, physically weak but with a very high poten-dualism in prehistory, and without which dualism tial for intra-specific communication, is such that the would have never occurred to the author’s mind as survival of the species is mostly dependent on the a way to enlighten the Neolithic ideology, is the set process of exchange between individuals. This per-of central monumental twin steles of Göbekli Tepe. petual need to distribute food and especially mates within the subgroups of the community and across Conclusion generations required the adoption of a form of so­cial conduct based on pure reciprocity. Dualism ap-The following quote on the Australian Aboriginal so-pears as the ideological result (and not the cause ex-cial and ideological structure seems appropriate to nihilo) of this chain of causal factors. This is the conclude this paper: “to affirm that appropriation reason why dualism should not be seen as a mere is the fact of the community as a whole only, to af-cultural tradition which, among others, would have firm that the latter is an inseparable totality, it was been miraculously preserved until the Neolithic. It first necessary to break it into two, into two parts can be presumed to have been ‘socially selected’ in each closely dependent on the other. Each part was the Darwinian sense of the expression, that is, un-conceived from the start as part of the whole” (Te-consciously over countless generations, for having start 1985.478). It is remarkable and fortunate that provided humanity with the highest, maybe the these lines were written about a decade before the only, probability of survival. excavation of Göbekli Tepe. Today, the twin steles stand as a monumental confirmation of Testart’s Evaluation (1988) audacious intuition that Australian Aborigi-It should be noted here that Testart is suspected of nes likely reflect an extended Palaeolithic ideologi-having somehow distorted the ethnographical facts, cal background. although this is certainly due to the goal he set him­self: not that of describing specific communities, but ACKNOWLEDGEMENTS instead uncovering the purely theoretical structure that binds them all. He was thus able to reconstruct I am extremely grateful to Anna Belfer-Cohen and a coherent social system where every structure (pro-Alan Barnard for their patient reviews of early ver­duction, reproduction, ideology) is absolutely in tune sions of this paper and for their support. Jak Yakar, Barbara Helwing, Çiler Çiligiroglu and Hans-Georg with all the others. The best clue to support the co- Gebel all provided constructive criticisms that im­ herence of this reconstitution is that everywhere so- proved the manuscript. I would also like to thank The cieties reproduced successfully the hunting-gather- Deutsches Archaologisches Institut and Christian ing way of life throughout the entire Palaeolithic pe- Jaubert for letting me reproduce their pictures. The riod (one to two million years?). We saw how this ideas presented here are my own only but are in- success was achieved through the total annihilation debted to the insights of late Jean-Daniel Forest. This of individual interests to the benefit of the whole, paper is dedicated to his memory. and that dualism has been identified as the keystone . References Asouti E., Fuller D. Q. 2013. A contextual Approach to the emergence of Agriculture in Southwest Asia: reconstructing Early Neolithic Plant-Food production. Current Anthro­pology 54(3): 299–331. https://doi.org/10.1086/670679 Asouzu I. I. 2007. Ibuanyidanda: New Complementary Ontology: Beyond World-Immanentism, Ethnocentric Reduction and Impositions. LIT Verlag. Zürich. Banning E. B. 2011. So fair a house: Göbekli Tepe and the identification of temples in the Neolithic of the Near East. Current Anthropology 52(5): 619–560. https://doi.org/10.1086/661207 Barnard A. 1978. Universal systems of kin categorization. African Studies 37(1): 69–81. 2007. From Mesolithic to Neolithic modes of thought. In A. Whittle, V. Cummings (eds.), Going Over. The Me-solithic-Neolithic Transition in North-West Europe. Proceedings of the British Academy 144. Oxford Uni­versity Press. Oxford: 5–19. DOI: 10.5871/bacad/9780197264140.001.0001 2020. Hunters and Gatherers: What Can We Learn from Them. (Hearing Others’ Voices). Balestier Press. London, Singapore. The Neolithic dualist scheme Belfer-Cohen A., Goring-Morris A. N. 2002. Recent Deve­lopments in Near Eastern Neolithic Research. Paléorient 28(2): 143–148. Bellwood P. 2005. First farmers. The origin of agricul­tural societies. Blackwell. Victoria. Becker N., Dietrich O., Götzelt T., Köksal-Schmidt Ç., Not-roff J., and Schmidt K. 2012. Materialien zur Deutung der zentralen Pfeilerpaare des Göbekli Tepe und weiterer Orte des obermesopotamischen Frühneolithikums. Zeitschrift für Orient-Archäologie (5): 14–43. Beckett J. 1967. Marriage, circumcision and avoidance among the Maljangaba of northwest New South Wales. Mankind 6: 456–464. Binford L. 1983. In pursuit of the past: Decoding the Archaeological Record. University of California Press. Berkley. Bird-David N. 2019. Where have all the kin gone? On hun­ter-gatherer sharing, kinship and scale. In N. Lavi, D. E. Friesem (eds.), Towards a Broader View of Hunter-Ga­therer Sharing. McDonald Institute. Cambridge: 15–24. Bischoff D. 2002. Symbolic Worlds of Central and South­east Anatolia in the Neolithic. In F. Gérard, L. Thissen (eds.), The Neolithic of Central Anatolia. Ege Yayinlari. Istanbul: 237–244. Bloch M. 1983. Marxism and anthropology, a history of relationship. Oxford Univesity Press. Oxford Bodet C. 2012 Kinship patterns as a doorway to appre­hend the symbolic and social structure of Göbekli Tepe communities. Ethneo 3(2): 1–17. 2019a. Early Animal Production for Marital Trade: A Neolithic Bride-Price? In O. Dumankaya (ed.), Pro­duction and trade through the ages: from Prehistory to the Byzantine Period. Bilgin Kültür Sanat yayinlari. Ankara: 95–122. 2019b. Why are there no Neolithic mega-sites in the Anatolian Euphrates basin? A socio-archaeological ap­proach to the marital structure of early farming socie­ties. Arkhaia Anatolika 2: 21–50. 2021. Deciphering Göbekli Tepe: The Hidden Potential of Archaeo-Ethnology. In I. Akkas, M. Karakoç (eds.), Etnoarcheology Research In Anatolia. From the prehi­storic times to the present. Doruk. Istanbul: 147–165. Forthcoming. Twin Steles, Isomorphic symbolism and the Neolithic Dualist Identity. TAG-III Turkey- proce­edings of the 2021 conference. Bourdieu P. 1980. Le sens pratique. Edition de minuit. Paris. Cauvin J. 1997. Naissance des divinités, Naissance de l’agriculture. Flammarion. Paris. Cook M. 2003. A Brief History of the Human Race. Gran-ta books. London. David N., Kramer C. 2001. Ethnoarchaeology in action. Cambridge University Press. Cambridge. Dowling J. H. 1968. Individual ownership and the sharing of game in hunting societies. American anthropologist 70: 502–507. Doyle D. 2005. Ritual Male Circumcision: A Brief History. Journal of Royal Colledge of Physicians of Edinburgh 35: 279–285. Durkheim E. 1937. Les regles de la méthode sociologi­que. Presse Universitaire de France. Paris. Erim-Özdogan A. 2011. Çayönü. In M. Özdogan, N. Basge­len, and P. Kuniholm (eds.), The Neolithic in Turkey, The Tigris Basin. Archaeology and art. Istanbul: 185–269. Forest J.-D. 1992. L’archéologie et l’ethnologie ou la né­cessité de mélanger les genres. In Ethnoarchéologie: justi­fication, problemes, limites, XIIe Rencontres d’archéo­logie et d’histoire d’Antibes. Éditions de Association pour la Promotion et la Diffusion des Connaissances Archéolo­giques. CNRS. Juan-les-Pins. Paris: 25–32. 1993. Çatal Höyük et son décor: pour le déchiffrement d’un code symbolique. Anatolia Antiqua 2: 1–42. 1996a. Mésopotamie, l’apparition de l’Etat. Editions Paris-Méditerranée. Paris. 1996b. Le PPNB de Çayönü et de Nevali Çori: pour une approche archéo-ethnologique de la néolithisation du Proche-Orient. Anatolia Antiqua 4: 1–31. 2003. Çatal Höyük et son décor: pour le déchiffrement d’un code symbolique. In J. Guilaine (ed.), Arts et Sym-boles du Néolithique a la Protohistoire. Errance. Paris: 41–58. 2006. Le processus de néolithisation proche-oriental: pour une archéologie sans frontieres. Syria 83: 125–138. Frazer J. G. 1910. Totemism and exogamy. Macmillan. London. Freud S. 2010 [1913]. Totem et Tabou. Points. Paris. Cédric Bodet Gebel H.-G. K. 2013. The territoriality of early neolithic symbols and ideocracy. Neo-Lithics 2(13): 39–41. Ghasarian C. 1996. Introduction a l’étude de la paren­té. Seuil. Paris. Gibson C. 2009. How to read symbols. A crash course in the meaning of symbols in art. London. Herbert Press. Giddens A. 1971. Capitalism and modern social theory, an analysis of the writings of Marx, Durkheim and Max Weber. Cambridge University Press. Cambridge. Girard R. 1972. La violence et le sacré. Hachette. Paris. Gökce N. 2021. Neolitik yasam biçiminin yayiliminda kireç tabanli yapilar sorunu. Unpublished MA thesis. Ege University. Izmir. Gould R. 1978. Beyond analogy in ethnoarchaeology. In R. Gould (ed.), Explorations in ethnoarchaeology. Uni­versity of New Mexico Press. Albuquerque: 249–293. Granet M. 1929. La civilisation chinoise. La vie publi­que et la vie privée. La renaissance du livre. Paris. Haland E. J. 2017. Greek festivals, modern and ancient. A comparison of female and male values. Cambridge scholars publishing. Cambridge. Hansen S. 2017. Phallomorphic antler tools. In B. Hel­lwing, T. Aliyev, B. Lyonnet, F. Guliyev, S. Hansen, and G. Mirtskhulava (eds.), The Kura projects, new research on the later Prehistory of the southern Caucasus. Dietrich Reimer Verlag. Berlin: 269–271. Henderson J. L. 1964. Ancient myths and modern man. In C. G. Jung (ed.), Man and his symbols. Dell. London: 95– 155. Hodder I. 2012. Çatalhöyük. A summary of recent work concerning architecture. In B. Sögüt (ed.), Ahmet A. Tir­pan’a Armagan/Festschrift for Ahmet A. Tirpan. Ege · Yayinlari. Istanbul: 303–314. 2020. From communal to segmentary: an alternative view of Neolithic ‘monuments’ in the Middle East. In A. B. Gebauer, L. Sorensen, A. Teather, and A. C. Valera (eds.), Monumentalising life in the Neolithic. Oxbow. Oxford: 49–51. Howitt A. W. 1904. The native tribes of Southeast Austra­lia. MacMillan. London. Ingold T. 1980. Hunters, patoralists and ranchers. Cam­bridge University Press. Cambridge. Jaubert J., Verheyden S., Genty D., + 16 authors, and San­tos F. 2016. Early Neanderthal constructions deep in Bru­niquel Cave in southwestern France. Nature 534: 111– 114. https://doi: 10.1038/nature18291 Jeunesse C. 2020. Elite houses or specialized buildings? In A. B. Gebauer, L. Sorensen, A. Teather, and A. C. Valera (eds.), Monumentalising life in the Neolithic. Oxbow. Oxford: 53–56. Kinzel M., Clare L. 2020. Monumental compared to what? A Perspective from Göbekli Tepe. In A. B. Gebauer, L. So­rensen, A. Teathe, and A. C. Valera (eds.), Monumenta­lising life in the Neolithic. Oxbow. Oxford: 29–48. Knight C., Power C., and Watts I. 1995. The Human Sym­bolic Revolution: A Darwinian Account. Cambridge Ar­chaeological Journal 5(1): 75–114. https://doi.org/10.1017/S0959774300001190 Kornienko T. V. 2018. On the Interpretation of Stelae in the Cult Complexes of Northern Mesopotamia During the Pre-Pottery Neolithic. Archaeology, Ethnology and An­thropology of Eurasia 46(4): 13–21. DOI: 10.17746/1563-0110.2018.46.4.013-021 Kuijt I. 2018. Material geographies of House Societies: Re­considering Neolithic Çatalhöyük, Turkey. Cambridge Ar­chaeological Journal 28(4): 565–590. https://doi.org/10.1017/S0959774318000240 Le Quellec J.-L. 2015. Peut-on retrouver les mythes préhi­storiques? L’exemple des récits anthropogoniques. Comp-tes rendus des séances de l’Académie des Inscriptions et Belles-Lettres 1: 235–266. Leroi-Gourhan A. 1964. Les religions de la préhistoire. Presses Universitaires de France, PUF. Paris. Lévi-Strauss C. 1967. Les structures élémentaires de la parenté. Mouton de Gruyter. Paris. Lewis-Williams D. 2002. The mind in the cave. Conscious­ness and the Origins of Art. Thames & Hudson. London. Makarewicz C. A., Finlayson B. 2018. Constructing com­munity in the Neolithic of southern Jordan: Quotidian prac­tice in communal architecture. PLoS ONE 13(6): e0193712. https://doi.org/10.1371/journal.pone.0193712 Malinowski B. 1926. Crime and Custom in Savage So­ciety. Littlefield, Adam & Co. London. Matthews R. 2003. The Archaeology of Ancient Mesopo­tamia, Theories and approaches. Routledge. London. The Neolithic dualist scheme Mauss M. 1924. Essai sur le don. Forme et raison de l’échange dans les sociétés primitives. L’année Sociolo­gique (1). Presses Universitaires de France, PUF. Paris. McBride A. 2013. Performance and participation: multi-sensual analysis of Near-Eastern Pre-Pottery Non-Domes­tic Architecture. Paléorient 39(2): 47–67. Meillassoux C. 1991. Femmes, greniers et capitaux. Paris. L’Harmattan. Mellaart J. 1967. Çatal Höyük, a Neolithic Town in Ana­tolia. Thames & Hudson. London Mithen S. J., Finlayson B., Smith S., Jenkins E., Najjar M., and Maricevi. D. 2011. An 11 600 year-old communal structure from the Neolithic of southern Jordan. Antiquity 85: 350–364. https://doi.org/10.1017/S0003598X00067806 Mleku. Vrhovnik D. 2021. Approaching weird: psychoana­lysis and archaeology of caves. Accademia Letters 190: 1–5. https://doi.org/10.20935/AL190. Moetz F. K., Çelik B. 2012. T-Shaped Pillar Sites in the Landscape around Urfa. In R. Matthews, J. Curtis (eds.), Proceedings of the 7th International Congress on the Ar­chaeology of the Ancient Near East, vol. 1. Harrassowitz Verlag. Wiesbaden: 695–709. Morgan L. H. 1871. Systems of Consanguinity and Affi­nity of the Human Family. Smithsonian Institution. Wa­shington. Özbasaran M., Duru G., Kayacan N., Erdogu B., and Bui­tenhuis H. 2012. Asikli (and) Musular the 8th millenni­um cal BC Satellite site of Asikli. In M. Özdogan, N. Basge­len, P. Kuniholm (eds.), Neolithic in Turkey-Central Tur­key. New Excavations and New Research. Central Turkey. Archaeology and Art Publications. Istanbul: 135–180. · Özbek M. 2004. Çayönü’nde Insan. Arkeoloji ve Sanat · Yayinlari. Istanbul. Peters J., von den Driesch A., Pöllath N., and Schmidt K. 2005. Birds in the megalithic art of the Pre-Pottery Neo­lithic Göbekli Tepe, Southeast Turkey. In G. Grupe, J. Peters (eds.), Feathers, grit and symbolism: Birds and humans in the ancient Old and New World (ICAZ 5). Documenta archaeobiologiae 3. Verlag Marie Leidorf. Rah-den/Westf.: 223–234. Peters J., Schmidt K. 2004. Animals in the symbolic world of Pre-Pottery Neolithic Göbekli Tepe, Southeastern Tur­key: a preliminary assessment. Anthropozoologica 39(1): 179–218. Radcliffe-Brown A. R. 1952. Structure and function in primitive society. Cohen & West. London. Sahlins M. 1961. The Segmentary Lineage and Predatory Expansion. American Anthropologist 63: 322–345. 1972. Stone Age economics. Aldine de Gruyter. New York. Schmandt-Besserat D. 2013. Preface and acknowledg­ments. In D. Schmandt-Besserat (ed.), Symbols at ‘Ain Ghazal – Ain Ghazal Excavation Reports 3. Bibliotheca neolithica Asiae meridionalis et occidentalis et Yarmouk University. Monograph of the Faculty of Archaeology and Anthropology, Yarmouk University. Ex Oriente. Berlin: xxv–xxvi. Schmidt K. 2001. Göbekli Tepe, Southeastern Turkey. A preliminary report on the 1995–1999 excavations. Palé-orient 26(1): 45–54. Schoop U.-D. 2005. The Late escape of the Neolithic from the Central Anatolian plain. In C. Lichter (ed.), How Did Farming Reach Europe? Anatolian-European relations from the second half of the 7th through the first half of the 6th millennium cal BC. Proceedings of the internatio­nal workshop, Istanbul, 20–22 May 2004. Byzas 2. Veröf­fentlichungen des Deutschen Archäologischen Instituts Istanbul. Istanbul: 41–58. Spencer B., Gillen F. J. 1899. The native tribes of Central Australia. Macmillan. London. Stordeur D. 2000. Jerf el Ahmar et l’émergence du Néoli­thique au Proche Orient. In J. Guilaine (ed.), Premiers paysans du monde. Errance. Paris: 33–60. 2003. Symbole et Imaginaire des premieres cultures néolithiques du Proche Orient (Haute et Moyenne val-lée de l’Euphrate). In J. Guilaine (ed.), Arts et Symboles du Néolithique a la Protohistoire. Errance. Paris: 15–37. 2012. Les villages et l’organisation des groupes au Néo­lithique précéramique A. L’exemple de Jerf el-Ahmar, Syrie du nord. In J. L. Fenollos (ed.), Du village Néoli­thique a la ville Syro-Mésopotamienne. Bibliotheca eu­phratica 1. Universidad da Coruna & Sociedad Luso-Gallega de Estudios Mesopotámicos. Ferrol: 35–54. Stordeur D., Brenet M., Der Aprahamian G., and Roux J.­ C. 2001. Les bâtiments communautaires de Jerf el-Ahmar et Mureybet horizon PPNA (Syrie). Paléorient 26(1): 29– 44. https://www.persee.fr/doc/paleo_0153-9345_2000_ num_26_1_4696 Testart A. 1978. Des classifications dualistes en Austra-lie. Essai sur l’évolution de l’organisation sociale. Edi­ Cédric Bodet tions de la maison des sciences de l’homme. Université de Lille III. Lille. 1985. Le communisme primitif. I- Economie et idéo­logie. Maison des sciences et de l’homme. Paris. 1987a. Deux modelés du rapport entre l’homme et l’animal dans les systemes de représentations. Etudes Rurales 107(8): 171–193. 1987b. Game sharing systems and kinship systems among hunter-gatherers. Man 22: 287–304. 1988. Some Major Problems in the Social Anthropolo­gy of Hunter-Gatherers. Current Anthropology 29(1): 1–31. https://www.journals.uchicago.edu/doi/abs/10. 1086/203612?journalCode=ca 1998. Révolution, révélation ou évolution sociale. A pro-pos du livre de Jacques Cauvin: Naissance des divinités, Naissance de l’agriculture. Les nouvelles de l’archéo­logie 72: 25–29. 2006a. Des dons et des dieux. Errance. Paris. 2006b. Interprétation symbolique et interprétation re-ligieuse en archéologie. Exemple du taureau a Çatal Hö­yük. Paléorient 32(2): 23–57. 2006c. Review of Magic practices and rituals in the Near Eastern Neolithic. Paléorient 32(1): 149–150. 2012. Avant l’histoire. L’évolution des sociétés huma­ines de Lascaux a Carnac. Gallimard. Paris. Verhoeven M. 2002. Ritual and ideology in the Pre-Pot­tery Neolithic B of the Levant and South east Anatolia. Cambridge Archaeological Journal 12(2): 233–258. https://doi.org/10.1017/S0959774302000124 2001. The birth of a concept and the origins of the Neo­lithic: a history of prehistoric farmers in the Near East. Paléorient 37(1): 75–87. Voigt M. 2002. Çatalhöyük in context: Ritual at Early Neo­lithic Sites in Central and Eastern Turkey. In I. Kuijt (ed.), Life in Neolithic Farming Communities Social Organi­zation, Identity, and Differentiation. Kluwer. New York: 253–293. Walker R. S., Hill K. R., Flinn M. V., and Ellsworth R. M. 2011. Evolutionary History of Hunter-Gatherer Marriage Practices. PLoS ONE 6(4): 1–6. https://doi.org/10.1371/journal.pone.0019066 Watkins T. 1996. The origin of household in North Meso­potamia. In K. R. Veenhof (ed.), From houses to house­holds in ancient Mesopotamia. Papers Read at the 40th Rencontre Assyriologique Internationale, Leiden, July 5–8, 1993. Uitgaven van het Nederlands Historisch-Archa- · eologisch Instituut te Istanbul: 78. Nederlands Historisch-Archaeologisch Instituut te Istanbul. Istanbul: 79–88. 2006. Architecture and the Symbolic Construction of New Worlds. In E. B. Banning, M. Chazan (eds.), Do­mesticating Space: Construction, Community, and Cosmology in the Late Prehistoric Near East. Studies in Early Near Eastern Production, Subsistence, and En­vironment 12. Ex Oriente. Berlin: 15–24. Weber M. 1921. Wirtschaft und Gesellschaft, [1996. Socio-logie des religions]. Gallimard. Paris. Weiss C. 1966. Motives for Male Circumcision among Pre­literate and Literate Peoples. The Journal of Sex Research 2(2): 69–88. Willcox G., Stordeur D. 2012. Large-scale cereal process­ing before domestication during the tenth millennium cal. BC in northern Syria. Antiquity 86: 99–114. https://doi.org/10.1017/S0003598X00062487 Wilson M. 2020. Symbols in art. Thames and Hudon. Lon­don. Yakar J. 2003. Defining social complexity in Anatolian communities of hunter-gatherer cultivators. In M. Özdo­gan, H. Hauptmann, and N. Basgelen (eds.), Köyden ken-te. Yakindoguda ilk yerlesimler 2/From Village to Cities. Early Villages in the Near East 2. Arkeoloji Sanat Yayin­ · lari. Istanbul: 437–446. 2005. The language of symbols in prehistoric Anatolia. Documenta Praehistorica 32: 111–121. https://doi.org/10.4312/dp.32.7 Zeder M. A. 2011. The Origins of Agriculture in the Near East. Current Anthropology 52(4): 221–235. http://www.jstor.org/stable/10.1086/659307 Zimmerman T. 2020. Ein Untergang im Morgenland? Gö­bekli Tepe als Fallstudie für die Vollendung und das Ver­löschen einer späteiszeitlichen Jäger- und Sammlerkultur. The Journal of the Oswald Spengler Society 3: 7–25. http://hdl.handle.net/11693/76053 back to contents Documenta Praehistorica XLIX (2022) Different manifestations of Neolithization in Northwest Anatolia| An archaeobotanical review from Barcin Höyük and Bahçelievler, Turkey Hüreyla Balci 1–2, Rana Özbal2, Fokke Gerritsen3, and Erkan Fidan 4 hureyla.balci@istanbul.edu.tr 1 Istanbul University, Istanbul, TR 2 Koç University, Istanbul, TR< rozbal@ku.edu.tr 3 Netherlands Institute in Turkey & Universiteit Leiden, Leiden, NL< fa.gerritsen@nit-istanbul.org 4 Bilecik S¸eyh Edebali University, Bilecik, TR< erkan.fidan@bilecik.edu.tr ABSTRACT – The Neolithic way of life was first established in Northwest Anatolia before the middle of the 7th millennium BC. The recently excavated sites of Barcin Höyük and Bahçelievler have yield­ed archaeological evidence for the earliest Neolithic levels in the region and provide new archaeo-botanical datasets. To compare different adaptations to the changes brought on by the Neolithization processes, we studied 348 archaeobotanical samples from Phases VIe and VId1 at Barcin and 63 samples from the contemporaneous levels, Phase 6 and Phase 5, at Bahçelievler. The economic plants include hulled and naked six-row barley, einkorn, emmer, bread/hard wheat, small-sized naked wheat, lentil, bitter vetch, pea, chickpea, flax, hazelnut, bramble, and pistacia. Our analyses show small but significant differences between the sites in the selected economic plant ranges, among the cereals, pulses as well as gathered plants. KEY WORDS – archaeobotany; Neolithization; Northwest Anatolia; Barcin Höyük; Bahçelievler Razli;ne oblike neolitizacije v severozahodni Anatoliji| Arheobotani;ni pregled v Barcin Höyüku in Bahçelievlerju v Tur;iji IZVLE.EK – Neolitski na.in .ivljenja se je v severozahodni Anatoliji uveljavil pred sredino 7. tiso.­letja pr. n. .t. Nedavno raziskani najdi..i Barcin Höyük in Bahçelievler sta prinesli arheolo.ke do-kaze in arheobotani.ne zbirke podatkov o najzgodnej.em neolitiku v regiji. Primerjavo razli.nih pri­lagoditev spremembam, ki so jih prinesli procesi neolitizacije, smo opravili s pomo.jo analiz 348 ar­heobotani.nih vzorcev iz naselbinskih faz VIe in VId1 v Barcinu in 63 vzorcev iz so.asnih faz 6 in 5 v Bahçelievlerju. Gospodarske rastline so olu..eni in goli .estvrstni je.men, enozrnica, dvozrnica, kru.na .ita, drobnozrnata p.enica golica, le.a, grenka gra.ica, grah, .i.erika, lan, le.nik, robida in pistacija. Na.e analize ka.ejo na majhne, a pomembne razlike med najdi..i v izboru gospodarskih rastlin, tako med .iti in stro.nicami kot tudi nedomesticiranimi rastlinami. KLJU.NE BESEDE – arheobotanika; neolitizacija; severozahodna Anatolija; Barcin Höyük; Bahçe­lievler Introduction Archaeological evidence indicates that sedentism, first half of the seventh millennium BC. How this domesticated plants, and herded animals made their spread occurred and how societies adapted to Neo-way from the Fertile Crescent and Central Anatolia lithic lifestyles exhibit great variability. Studies on towards the Aegean and the Marmara Region in the the Neolithization processes suggest that while some DOI> 10.4312\dp.49.19 Different manifestations of Neolithization in Northwest Anatolia| An archaeobotanical review from Barcin Höyük and Bahçelievler, Turkey communities established Neolithic habits from the outset, others fused two diverse ways of life; a hun­ter-gatherer lifestyle with agriculture (for Europe see Robb 2013; Zvelebil 2001). Recent studies on modes of subsistence illustrate heterogeneous and complex processes and a mosaic of adaptations (Ivanova et al. 2018; Jovanovi. et al. 2021; Kotzamani, Livar­da 2018; Zeder 2011). These data challenge the idea of the spread of a uniform ‘Neolithic Package’, but much remains to be done to understand how the processes took place from region to region. Macro-botanical and micro-botanical analyses can be im­portant to understand the variability of Neolithic adaptations and subsistence strategies during this process of expansion and colonization. How processes of expansion took place in Northwest Anatolia remains an important question given that this region was among the first territories that Neo­lithic pioneers coming from the core regions of Neo­lithization encountered (Fig. 1a). While all early settlements in the region display an established Neo­lithic way of life, it is still unclear whether these Neolithic societies incorporated Mesolithic foragers present in the region. Hypotheses have been formu­lated about a merging of forager and farmer groups in Northwest Anatolia (M. Özdogan 2013; 2014), but the supporting evidence at hand is far from con­crete. Differences in architectural styles and mater­ial culture have led to theories regarding the pres­ence and continuity of local pre-Neolithic communi­ties at some sites (Düring 2013; Özbal, Gerritsen 2019; E. Özdogan 2016). While such a mosaic mo­del in the Neolithization process is well-documented for Europe (Zvelebil 2001), Northwest Anatolia faces a general lack of data with regard to Mesolithic lifeways, except, poten­tially, Agaçli to the north of Istan­bul (Gatsov 2001; Gatsov, Özdogan 1994; Özdogan, Gatsov 1998). Re­cent aDNA studies have shown that early Neolithic populations in West Anatolia and the first farmers in Eu­rope belong to the same gene pool (Hofmanova et al. 2016; Lazaridis et al. 2016; Mathieson et al. 2015; 2018). Much less clear at present is the genetic history of Anatolia be­fore and during the period of initial Neolithization, but there are indica­tions for complex processes during and after the Late Glacial that in­clude genetic bottlenecks, admixture from outside the region and regio­nal heterogenization (Kilinç et al. 2016; Marchi et al. 2022; Yaka et al. 2021). Botanical remains provide an important dataset through which the process of Neolithization can be studied. After all, what people cultivated and gath­ered must be viewed as a reflection of their life­style choices and could provide important insights on the Neolithization process. Botanical remains can act as a proxy, not only for the reconstruction of the local environmental or ecological situation but also for the incorporation and transmission of cultivated plants (Balci 2018; Gaastra et al. 2019; Kotzamani, Livarda 2018; Krauß et al. 2017; Marinova, Krauß 2014; Popova, Marinova 2007). How much did peo­ple engage in and exploit their local environments, especially in the incipient phases of occupation? Is there a predominant dependence on farmed Neoli­thic founder crops or do we find evidence for the persistent utilization of local gathered resources? In what ways could the ratio between the wild and do­mesticated correlate with the habits of migrant far­mers and local hunter-gatherers? To explore these questions regarding Neolithization, this article makes use of two new Neolithic archaeo-botanical datasets from the contemporaneous North­west Anatolian sites of Barcin Höyük and Bahçeliev­ler. These sites are less than 40km apart and appear to be in the same vegetational zone, making them ideal case studies for a comparative analysis of ma-cro-botanical data. The site of Barcin Höyük (Bursa) was excavated between 2007 and 2015 (Gerritsen, Özbal 2019); Bahçelievler (Bilecik) between 2019 and 2021 (Fidan 2020; Kolankaya-Bostanci, Fidan Hüreyla Balci, Rana Özbal, Fokke Gerritsen, and Erkan Fidan 2021). Both sites have levels dating to the first half of the seventh millennium BC and yield evidence for the earliest Neolithic communities in their respective sub-regions. Investigating the plant remains from the early and comparable levels of each site provides a first-hand way to observe similarities and differ­ences in subsistence strategies. This, in turn, offers a window into their relative reliance on local resour­ces and/or introduced founder crops. Our aim is to furnish our interpretation on the Neolithization pro­cess of Northwest Anatolia with new, first-hand data. What subsistence strategies did the inhabitants of each site adhere to, especially when it comes to plant use? When establishing Neolithic settlements where farming became the prominent form of sub­sistence, how much of the local flora was utilized? Different manifestations in the same region? A case study of Barcin Höyük and Bahçelievler Northwest Anatolia includes the region to the south­east of the Marmara Sea that extends from the Bos­phorus to the Eskisehir Region. The latter provides direct access to the Anatolian Plateau. In the past, as today, this region represented a diverse vegetatio­nal, geographical, and palaeogeographical structure consisting of coasts, mountain thresholds, moun­tains, plains, and valleys irrigated by rivers (Atalay, Mortan 1997; Clare, Weninger 2014; Kayan 2014; Roberts 2014). Climatically, it has mild/Mediterra­nean conditions (Clare, Weninger 2014). Given the humid climate of Northwest Anatolia around 11 000 BC, the predominant tree species were birch, oak, pine, and juniper. These species also formed the main tree taxa during the Late Glacial period (Ro­berts 2014.Fig.1a). A general vegetation history covering most of the Holocene indicates that the lowland hills and moun­tain slopes of this region were covered with wood­lands dominated by deciduous oak from about 10000 to 6500 BC (Bottema, Woldring 1995; Bot-tema et al. 2001; Kayan, Woldring 2002). Pollen studies from a location in the Yenisehir lake basin near Barcin Höyük revealed that the vegetation was also comprised of fir (Abies), pine (Pinus), elderber­ry (Sambucus), hornbeam (Carpinus), hazelnut (Corylus), beech (Fagus), cedar (Cedrus), linden (Ti­lia), and elm (Ulmus) (Bottema et al. 2001). A re­cent wood charcoal study by Schroedter and Nelle on data obtained from the Late Neolithic layers at the site of Aktopraklik – located near Lake Ulubat, fifty kilometres to the West of the Yenisehir Plain – yielded oak, pine, mock privet, and pistacia as well (Schroedter, Nelle 2015). Despite the geographical proximity, the latter two species are not documented in the pollen study from Lake Yenisehir (Bottema et al. 2001; Schroedter, Nelle 2015.92). Today, about a third of the region remains covered with forests (Atalay, Mortan 2011; Roberts 2014). Due to the felling of oak and red pine forests in hi­storical times, dense maquis shrubland covers the landscape. Vegetation includes species such as rock rose (Cistus creticus), hazel (Corylus), tree heath (Erica arborea), prickly juniper (Juniperus oxyce­drus), mock privet/green olive tree (Phillyrea lati­folia), pistacia (Pistacia terebinthus), and plum (Prunus) (Atalay, Mortan 2011.153). Archaeologically, the Istanbul region, the Yenisehir Plain, the foothills overlooking Lake Ulubat, and the Bilecik-Eskisehir region fall within what has tradi­tionally been called the Fikirtepe Culture zone (Fig. 1a) (Özdogan 2014). Evidence for agriculture and animal husbandry is most prevalent here, but minor amounts of hunting, gathering and fishing are also evident across the communities of the Neolithic and Chalcolithic in the Fikirtepe Culture zone at sites in­cluding Pendik, Fikirtepe, Yenikapi in Istanbul pro­vince, Barcin Höyük, Mentese, Aktopraklik, Ilipinar in Bursa province, and Bahçelievler in Bilecik and Keçiçayiri in Eskisehir provinces in Northwest Ana­tolia (Arbuckle et al. 2014; Balci 2018; Balci et al. 2019; Boessneck, von den Driesch 1979; Budd et al. 2013; 2018; 2020; Buitenhuis 2008; Cappers 2008; 2014; Çakirlar 2013; 2015; Galik 2013; Gou- . richon, Helmer 2008; Izdal Çaydan 2018; Karul 2011; 2017; Kiziltan, Polat 2013, Kiziltan 2013; Kolankaya-Bostanci, Fidan 2021; Özdogan 1983; Sari, Akyol 2019; Thissen et al. 2010; Ulas 2020; Würtenberger 2012). A noteworthy element with regard to the architec­ture is that we see variability across sites. While those like Barcin Höyük (Gerritsen, Özbal 2016), Mentese (Roodenberg et al. 2003), Ilipinar X-IX (Ro­odenberg 2008), and Aktopraklik B (Karul 2010) display rectilinear architecture, others including Ak­topraklik C (Karul 2011; Karul, Avci 2011), and Bahçelievler (Fidan 2020; Kolankaya-Bostanci, Fi­dan 2021) yield evidence for round semi-subterra­nean structures. Both round and rectangular buil­dings have been discovered and excavated at Yeni­kapi (Kiziltan, Polat 2013) and Pendik (Harmanka-ya 1983; Özdogan 2013; Pasinli et al. 1993), while Yarimburgaz (Özdogan 2013) stands alone as a slightly later cave settlement. If architecture is a phy­ Different manifestations of Neolithization in Northwest Anatolia| An archaeobotanical review from Barcin Höyük and Bahçelievler, Turkey sical manifestation of world views and lifestyles (Lefebvre 1991), then the variability observed across Northwest Anatolia in the seventh and sixth millen­nia BC may be noteworthy. The contrast that the sites of Barcin Höyük and Bahçelievler show with regard to architecture, with the former yielding rec­tangular and the latter round structures, juxtaposes these two pioneering Neolithic sites. This allows us to consider any notable differences in botanical re­mains in a larger context. We may ask whether di­vergences in assemblages may reflect indications of diverse representations of lifestyle preferences or whether they are, in fact, a result of micro-regional adaptations. Are there indications that we are deal­ing with immigrant farmers at one community and a representation of local hunting and gathering com­munities who adopted agriculture in another? Barcin Höyük and Bahçelievler were inhabited par­tially contemporaneously, as demonstrated by both absolute dates and material assemblage comparisons (Fidan 2020; Gerritsen, Özbal 2013a; 2013b; Özbal, Gerritsen 2019). Bahçelievler Phase 6 is likely con­temporary with Phase VIe at Barcin Höyük, while Bahçelievler Phase 5 corresponds timewise with Barcin Höyük Phase VId1. The later levels at both sites, beyond the scope of this paper, show paralle­lisms with the Fikirtepe culture (Fig. 1b). Barcin Höyük general background Barcin Höyük is located in the Yenisehir Plain, Bur­sa, and was excavated between 2005–2015. The Neolithic levels are separated into seven distinct phases from the uppermost VIa to the lowest VIe (Gerritsen et al. 2013a; Gerritsen, Özbal 2019). The most important result of the Barcin Höyük excava­tions is possibly its contribution to the reconstruc­tion of a continuous developmental sequence for the Neolithic of the Marmara Region. The stratigraphic sequence from the site, supported by 80+ radiocar­bon dates, enables us to restructure the period from the first half of the seventh millennium to the begin­ning of the sixth millennium BC with associated ma­terial culture, architecture, and subsistence strate­gies (Gerritsen, Özbal 2016; 2019; Özbal, Gerritsen 2019). Excavations at Barcin Höyük yielded rectangular houses. This article discusses Phases VIe and VId1, for which the botanical remains have been extensi­vely studied (Balci et al. 2019). While two posthole structures dating to the earliest phase (VIe) were unearthed, excavations also brought to light a row of four slightly smaller structures dating to VId1, the overlying phase (Gerritsen, Özbal 2016; Özbal, Ger-ritsen 2019; van den Bos 2021). Courtyards were discovered north and south of the structures in both phases. Posthole architecture was the primary buil­ding technique in Phase VIe, but in VId1 consider­ably smaller posts set into foundation trenches were used instead (van den Bos 2021.168). Most notably, there are differences between the two phases with regards to material culture as well. The scarcity of archaeological materials in the earliest layers, in­cluding pottery, is noteworthy. By Phase VId1 the range of objects available increases both in quanti­ty and variability (Gerritsen, Özbal 2016; Özbal, Gerritsen 2019). Bahçelievler site general background The site of Bahçelievler was discovered on an empty land parcel between apartment buildings in the city centre of Bilecik. The Neolithic settlement was lo­cated on the eastern bank of a small stream that has subsequently dried up. The Neolithic layers have been divided into seven different phases, from Pha­ses 8 to 2. Preliminary radiocarbon dates suggest that the earliest levels of Bahçelievler correspond to the first half of the seventh millennium BC (Fidan 2020). The exact dates are difficult to ascertain given Fig. 1b. Comparative chronological table for Bah-çelievler and Barcin. Hüreyla Balci, Rana Özbal, Fokke Gerritsen, and Erkan Fidan the problems with the calibration curve, but the ear­liest dates fall between 7192–7052 BCE.1 Excava­tions at Bahçelievler in Phases 3–8 yielded oval/ round structures with diameters of 3–5m and walls up to 45–50cm thick in some structures. The walls were strengthened in some instances by mud or mudbrick, and post-holes traces are visible in some walls as well as clusters of small pebbles (Fidan 2020.36). Workshop and courtyard areas were disco­vered between the structures, yielding most of the artifact assemblages with the exception of stone tools, which for the large majority come from inside the buildings (Kolankaya-Bostanci, Fidan 2021.102). Materials and methods The macro-botanical samples collected at both sites were floated in water, not more than two litres at a time, by means of manual flotation in buckets. Chiffon fabric was used for drying the light materi­al and a 1mm mesh was used for the heavy materi­al during the flotation for collecting and drying. The dried samples were sifted through steel test sieves of 0.24<0.5<1.0<2.0<3.0mm and placed inside zip­ped plastic bags and centrifuge tubes for sorting. A triocular 0.6–4x stereo zoom microscope was used for identification and photography. The plant re­mains were compared with plant catalogues to aid with the determination of genus and species (Boj­nansky, Fargasova 2007; Cappers et al. 2012; Cap-pers, Bekker 2013; Cappers et al. 2016; Neef et al. 2012). For Barcin Höyük, a systematic sampling strategy was applied to the site during the excavations. A to­tal of 163 samples corresponding to 480 litres of soil from Phase VIe and 185 samples corresponding to 580 litres of soil from Phase VId1, all collected during the 2013–2015 seasons, have been analysed within the scope of this study. The samples repre­sent different contexts including layers, surfaces, platforms, pits, foundation trenches, pyrotechnic fea­tures, postholes, and burials. The frequent burned contexts at Barcin Höyük facilitated excellent pre­servation of plant remains as well as substantial amounts of wood charcoal. There is no particular context in which we find a high percentage of plant remains in Barcin Höyük except a single burned store of lentils from structure 2a in level VId1 that yielded around 28 000 seeds. However, no special wild plant group was found among the samples (e.g., Fairbairn et al. 2007). A large proportion of the wild plants consists of field grass/weeds. The wild plant group is part of another study (in prep.). The archaeobotanical samples were studied by the first author in several places including the Barcin Höyük Excavation House in Yenisehir, Bursa, the Netherlands Institute in Turkey in Istanbul, and the Koç University Archaeology Laboratory in Istanbul under the supervision of René Cappers of the Uni­versity of Groningen. For Bahçelievler, a total of 134 archaeobotanical samples corresponding to 650 litres of soil sampled from the Neolithic phases during the 2019, 2020, and 2021 seasons were analysed. Included here in this study are 40 samples (248 litres) from Phase 6 and 23 samples (108 litres) from Phase 5. The sam­ples were taken from surfaces, courtyards, hearths, and burials, yielding great variability in the num­ber of archaeobotanical samples for each phase. The preservation of the plant remains was notably po­orer than at Barcin Höyük, probably due to the lack of burned deposits, but it may also reflect the circum­stances of plant use at the site. Most of the cereal re­mains were fragmented, making species identifica­tion difficult and wood charcoal remains remain li­mited. The archaeobotanical samples have been stu­died in the Bilecik Museum by the first author. Archaeobotanical results from Barcin Höyük and Bahçelievler Overall, the crop range between the two sites is si­milar. Both sites display an increase in quantity and variety of plant remains from the earliest phases Barcin VIe and Bahçelievler 6 to the subsequent phases Barcin VId1 and Bahçelievler 5 (Fig. 2). This may be to some extent a result of factors like preser­vation and sample numbers, but despite these issues a remarkable increase in the variety of cereals and pulses at both sites is noted over time. At Barcin Höyük Phase VIe, investigations yielded economic plants from the grass (Poaceae) family which constitute the main cereal group. This includ­es six-row barley – hulled and naked (Hordeum vul­gare ssp. vulgare L.), einkorn wheat (Triticum mo-nococcum ssp. monococcum L.), emmer wheat (Tri­ticum turgidum ssp. dicoccon (Schrank) Schübl.) and bread/hard wheat (Triticum aestivum L./du­rum Desf.). Among the pulses (Fabaceae), lentils (Lens culinaris Medik.), peas (Pisum sativum L.) and bitter vetch (Vicia ervilia L.) were identified. . 1 The 14C results of the settlement are being prepared for publication by Erkan Fidan and TÜBITAK MAM. Different manifestations of Neolithization in Northwest Anatolia| An archaeobotanical review from Barcin Höyük and Bahçelievler, Turkey In this phase, excavations yielded only fifteen pulse fragments, while flax (Linum usitatissimum L.) was represented by a mere seven seed fragments (Fig. 2). In summary, barley (hulled and naked), einkorn, and emmer hulled wheats, bread/hard wheat (na­ked), lentils, bitter vetch, peas and flax represent the main documented economic plants from VIe at Barcin Höyük. In Phase VId1, in contrast, we find an expansion of types and a greater variety than in VIe. The main cereals remain identical with the Phase VIe but we also begin to find a small-sized naked wheat type (not exactly defined wheat species between Triti-cum ssp. aestivum/ssp. durum and T. turgidum ssp. dicoccon) added to the cereal range in this phase. Likewise, we see a real presence of pulses – especial­ly lentils – of the pulse family. Identified species are similar to those from Phase VIe, but we find that the chickpea (Cicer arietinum) begins to appear among the pulses range in this Phase. Flax is also present as observed in VIe in small quantities. Phase VId1 also yields species gathered from the surround­ings including 22 fruits of hazelnut (Corylus avella­na L.) and two fruitlets of bramble (Rubus). As mentioned above, the plant preservation at Bah-çelievler is poor compared to at Barcin Höyük, and many samples yielded hardly any remains. Phases 8 and 7 at Bahçelievler with a total of four and 21 plant remains, respectively, are not considered in this paper because the botanical yields are too low to make meaningful interpretations (Fig. 2). The lack of botanical remains in the two lowest phases at Bahçelievler may be a result of sampling sizes, pre­servation and restricted exposures of the excava­tions, but could potentially reflect the limited use of farming plants. Instead, this paper focuses on Pha­ses 6 and 5 where the counts are not only adequate but the dates for these levels align well with Barcin Höyük’s Phases VIe and VId1. Thirty-eight of the samples, mostly coming from Trench B3 and dating to Phases 6 and 5, show somewhat higher concentra­tions. In Phase 6, the cereals include six-row barley – naked/hulled (Hordeum vulgare ssp. vulgare), ein­korn (Triticum monococcum ssp. monococcum), emmer (Triticum turgidum ssp. dicoccon), and bread/hard wheat (Triticum ssp. aestivum/durum). However, the einkorn wheat is only represented by two fragmented grains. Likewise, two seeds of len­til (Lens culinaris) have been identified for Phase 6 at Bahçelievler. Though minimal, this phase also yielded evidence for gathering with two fruits of pi-stacia (Pistacia ssp.). In the subsequent Phase 5, however, six-row barley – hulled/naked (Hordeum vulgare ssp. vulgare), emmer wheat (Triticum turgidum ssp. dicoccon), and bread/hard wheat (Triticum ssp. aestivum/ durum) were identified within the grass family among the main economic plants, while this time einkorn wheat (Triticum monococcum ssp. mono-coccum) was represented by a single grain. Pulses remain poorly represented and lentil (Lens culina­ris) continues to be represented by no more than two seeds. While this points to the presence of the species, it may not effectively show that this species had a significant role in the diet, at least within the excavated contexts. At the same time, however, ga­thered plants may suggest a different exploitation strategy of the immediate landscape in this phase. Pistacia, which was found in negligible quantities in Phase 6, becomes represented by 77 fruits from five different samples from the courtyard areas of Trench B3, suggesting a much larger emphasis on gathering by Phase 5. The gathered plant remains also include two grape seeds. Discussion Comparing the results for Barcin Höyük and Bahçelievler As at Barcin Höyük, the results also show an increase in botanical remains through time at Bahçelievler as well. While by Bahçelievler Phase 5 the variety of economic plants parallels that at Barcin Höyük, there are a few elements that show dissimilarity (Figs. 3–4). Barcin Höyük yielded small-sized naked wheat and flax, both of which were lacking at Bah-çelievler, and the presence of einkorn wheat, rep­resented by only two grains at Bahçelievler Phase 6 is debatable. In addition, the cereals remain the do­minant group of edible plants at both sites when compared with other plant remains, where it com­prised 95% of the assemblage at Barcin Höyük (Fig. 5a) and 72% at Bahçelievler (Fig. 5b). The pulse group comes second and retains a minor place, es­pecially at Bahçelievler. However, the most meaningful results that differen­tiate the sites derive from gathered plants. Though still preliminary, the results raise the question as to whether gathering at Bahçelievler contributed to the diet in a more substantial way than at Barcin Hö­yük. Pistacia, a gathered resource, comes second in quantity after the cereal remains, suggesting that it played a significant role at this site (Fig. 5b). Pista­cia is represented in this area within a range of trees with edible fruits including pistacia/terebinth (Pi­ Hüreyla Balci, Rana Özbal, Fokke Gerritsen, and Erkan Fidan Fig. 2. The quantitative comparison of crop plants between Phases VIe and VId1 at Barcin and contemporary Phases 6 and 5 at Bahçelievler, as well as theearliest Phases 8 and 7 at Bahçelievler. * represented as two samples, this total includes a lentil storage unit (which yielded around 20 000 seeds of lentil)and an associated collapse context (which yielded around 8000 seeds of lentil). ** rachis fragments include partly glume bases, spikelet forks, rachis, andrachis internodes. Different manifestations of Neolithization in Northwest Anatolia| An archaeobotanical review from Barcin Höyük and Bahçelievler, Turkey stacia terebinthus L.). At Barcin Höyük, in contrast, gathering remains almost trivial, and hazelnut and bramble fragments, especially when compared to the high quantity of samples, remain negligible (Fig. 5a). On the other hand, it may not always be consi­stent to emphasize the importance of a species based on the number of remains discovered, given that a range of criteria including preservation, fruit mor­phology,2 food preparation, and consumption, may affect the ultimate proportions. Nonetheless, the use of economic plants remains notably important at both sites in the earliest phases. Local adaptation: pulses Both Barcin Höyük and Bahçelievler yielded small quantities of pulses in their earliest phases, suggest­ing that pulses may be rare in general in the region in the first half of the seventh millennium BC. At Barcin Höyük, only fifteen pulse seeds were found in Phase VIe (Fig. 2), strikingly low, especially given the rich array of cereal remains recovered from the same contexts. However, we do see a significant in­crease in pulses by VId1 when we find a store of them in situ clustered in a lentil storage bin as well as from several other contexts. Bahçelievler, like­wise, yields a similar picture with regards to pulses where they remain conspicuously lower in quantity when compared with cereals (Fig. 6). There may be several reasons underlying the near lack of pulses in the earliest phases of these sites. The earliest in­habitants, whether incipient pioneer settlers or de­scendants of local foragers, might initially have briefly experimented with pulses but may instead have chosen to target cereal cultivation during the first occupation Phase VIe. On the other hand, the rarity of pulse species has also been interpreted as a result of preservation-de­pendent factors, and the scarcity of pulses might be a result of post-depositional processes specific to the species (Cappers 2008; Kotzamani, Livarda 2018; Marinova, Popova 2008). If taphonomic, the chal­lenge is to explain the significant difference in the pulse ratio between Phases VIe and Vd1 at Barcin Fig. 4. a small-sized naked wheat; b flax; c hazelnut; d bramble (a, b, c, d from Phase VId1, Barcin Hö­yük), e pistacia, f grape (e, f from Phase 5, Bahçelievler). 2 For example, the number of fruitlet endocarps for someone who eats five brambles would be c. 300–350. Post-depositional dis­persal might dilute the number concentrated in feces (personal communication with René Cappers). Hüreyla Balci, Rana Özbal, Fokke Gerritsen, and Erkan Fidan Höyük. Except for the burned store of lentils in structure 2a in Barcin VId1, we know that there is no significant difference in terms of the preservation conditions across the site. Aside from the store, 182 pulse seeds were discovered in the 580 litres sorted for Barcin VId1 across a range of 35 different con­texts. But only 14 pulse seeds were documented for the 480 litres analysed for Phase VIe. Preservation-related factors are often suggested to diminish the importance of pulses in the diet, but the discovery of a dense store of lentils in Phase VId1 questions the assumption that they were insignificant. The pulse spectrum at Barcin Höyük is paralleled at Bah-çelievler, where we see an increase in quantities over time. A question that comes to mind is whether this increase is a result of the changes in social behavi-our, the household structure, and/or the subsistence strategies of the inhabitants which may collective­ly have contributed to major shifts in the exploited species. Limited exposures and the low level of pre­servation of plant remains at Bahçelievler make it difficult to make a direct quantitative comparison, unlike at Barcin Höyük. However, it can be suggest­ed that the first settlers of Barcin Höyük must have applied different strategies regarding the growing, storing and processing of pulses. Local adaptation: gathered plants A major factor differentiating the sites of Barcin Hö­yük and Bahçelievler with regard to their subsistence strategies lies in the approaches that their inhabi­tants took with regard to gathered plants. Barcin Hö­yük lacks the general exploitation of edible fruits. Analyses only documented a single fruit of hazel in level VIe, though this number approaches 22 fruits by Phase VId1, which come from seven different con­texts. The presence of hazelnut increases in the la­ter levels of the site (Balci et al. in prep.). This could be considered an indication of how people interact­ed with their immediate environment. At Barcin Hö­yük, the exploitation of gathered plants was quite limited, and instead, cultivated, and harvested agri­cultural plants were favoured. At Bahçelievler, on the other hand, as demonstrated by the courtyard area of Trench B3 in Phase 5, the gathered plant pi-stacia and most likely Pistacia terebinthus was col­lected, where it comprised 25% of the assemblage demonstrating definitive utilization of this species in the diet. An interesting aspect of the gathered plant remains found at both sites is that they are typically from re­stricted numbers of contexts in comparison with eco­nomic plant groups such as cereals and pulses. This raises the question of full-time exploitation. Unlike cereals, which are particularly hardy and are excep­tionally suited for long-term storage, gathered plants are typically seasonal, and thus collection and ex­ploitation times must have been limited. In addi­tion, the location of the consumption of gathered plants such as hazelnut/pistacia and bramble/grape show differences with regard to depositional proces­ses. In this context, we can ask whether the lack of hazelnut at Bahçelievler, and, despite the large sam­ple sizes, the complete absence of pistacia at Barcin Höyük, was a result of sub-regional vegetation boun­daries. While pistacia was not documented in the pollen study from Lake Yenisehir (Bottema et al. 2001; Schroedter, Nelle 2015.92), the presence of this species is well attested in the Late Neolithic la­yers at Aktopraklik (Schroedter, Nelle 2015) and in the early Chalcolithic layers from Ilipinar X (Cap­pers 2008) which are 75 and 40km away, respecti­vely. Tim M. Schroedter and Oliver Nelle (2015) sug­gest that pistacia is a plant that thrives in open Me­diterranean type environments with shrub-like ve­getation. Barcin Höyük was located in a valley bot­tom with ample potential for agriculture while Bah-çelievler was in an upland region, so the differences in the setting may have contributed to the micro-en­ Different manifestations of Neolithization in Northwest Anatolia| An archaeobotanical review from Barcin Höyük and Bahçelievler, Turkey vironmental juxtaposition. On the other hand, we think that the nearby slopes along the edges of the Yenisehir plain could have been used for agri­culture as well (Balci 2018). Hazelnuts often thrive in open woodlands, which likely de­scribes the situation for Bar­cin Höyük. We cannot rule out that the differences across the sites with regard to their reliance on gathered plants was a result of micro-climatic and vegetational as­pects and hence different methods of adapting to the environment. The data from Barcin Höyük does not point to an intense reliance on other micro-cli­matologically suitable gathered plants. It is there­fore possible that part of the divergence may be a result of the ways in which the residents of each site interacted with their immediate surroundings and exploited the local vegetation. A comparison of subsistence strategies in the region Most of the Neolithic sites in Northwest Anatolia have levels dating to the end of the first half and second half of the seventh millennium, and yield evidence for what appears to be the earliest Neoli­thic inhabitants in their respective sub-regions, sup­porting our interpretation of the Neolithization pro­cess for Northwest Anatolia. Bahçelievler (Balci, in prep.), Barcin Höyük (under study by Cappers, Bal­ci; Balci 2018; Balci et al. 2019), Aktopraklik (Ka­rul 2017; Kabukçu et al. in prep.), Mentese, and Ili­pinar (Van Zeist et al. 1995b; Cappers 2008; 2014), Pendik (Ulas 2020), Fikirtepe, Yenikapi (Ulas 2020), and Neolithic Yarimburgaz provide insights on the Neolithic way of life across the Eastern Marmara Re­gion (Fig. 7). Even though not all excavations have yielded archaeobotanical data such as Fikirtepe, we do have ample data on their subsistence economies. In general, a terrestrial diet, rather than an aquatic or wild game-based one, is predominant at the in­land settlements of Bahçelievler, Barcin Höyük, Ba­sal Mentese, Aktopraklik C, and Ilipinar X (Arbuc­kle et al. 2014; Balci et al. 2019; Buitenhuis 2008; Budd et al. 2013; 2018; 2020; Cappers 2008; Galik . 2013; Gourichon, Helmer 2008; Izdal-Çaydan 2018; Karul 2017; Kolankaya-Bostanci, Fidan 2021). In addition, based on the presence of marine-based and hunted foods at coastal sites like Pendik and Fikirtepe, it is possible to interpret these subsi-stence practices as a continuation of Mesolithic cus­toms (Boessneck, von den Driesch 1979; Çilingi­roglu 2005; Düring 2011; Evershed et al. 2008; Öz­dogan 1983b; 2010; 2011; 2013; Röhrs, Herre 1961; Thissen 1999; Thissen et al. 2010). Burhan Ulas’ (2020) study on plant subsistence in Pendik also supports this suggestion. Though a coastal site, Yenikapi presents a different picture than the agri­cultural communities at Fikirtepe and Pendik, which also appear to have practiced fishing and hunting, probably because Yenikapi primarily represents the sixth millennium and is thus later (Kiziltan, Polat 2013; Ulas 2020). In Bursa province, archaeobotanical data has been obtained from Barcin Höyük, Aktopraklik, and Ili­pinar. At Neolithic Aktopraklik C, we know of the presence of six-row barley, emmer, lentils, bitter vetch, and flax (Karul 2017). At Ilipinar, excava­tions yielded 24 samples from the earliest Phase X and 20 samples from the overlying Phase IX dating to just after the turn of the sixth millennium BCE. The data suggests that barley, emmer, small-sized wheat, einkorn, lentil, bitter vetch, grass peas, peas, flax, figs, and bramble were used as economic plant species in the two earliest phases. However, the ear­liest Phase X only yielded a single non-economic plant (Cappers 2008). In Istanbul province archaeobotanical data has been obtained from both Pendik on the Asian side and Yenikapi on the European side. At Pendik, archaeo-botanical analyses yielded limited results. These comprised only a few cereals including a single em-mer grain, and a single barley grain as well as only a couple of pulses, including one-seed of a grass pea and one-seed of a pea. In addition, seven seeds of flax and two fruitlets of bramble were identified within the economic plant data. Other identified plant remains are included in the wild plant group (Ulas 2020.30–31). At Yenikapi, there is a higher va­riety in the economic plant range. The cereal group includes emmer, einkorn, bread/hard wheat, and Hüreyla Balci, Rana Özbal, Fokke Gerritsen, and Erkan Fidan Different manifestations of Neolithization in Northwest Anatolia| An archaeobotanical review from Barcin Höyük and Bahçelievler, Turkey new glume wheat, T. spelta, T. compactum, which is a species related to bread/hard wheat (Ulas 2020. 32). Overall, this yields a different picture than the general regional crop range. The pulse group in­cludes lentils, chickpeas, grass peas, bitter vetch, peas, and faba beans (Vicia faba var. minor, Ulas 2020). In addition, a range of gathered plants in­cluding figs, grapes, and bramble have been docu­mented (Ulas 2020.32–33). Consequently, we find notable dissimilarities in foun­der crops across sites. While it is possible to talk about a transition to farming and husbandry in the region, simply applying a universal ‘Neolithic Pack­age’ idea does not embody the complexity that is present across different sites. From this point of view, inhabitants at most sites within the region knew and practiced agriculture, and some also seem to have had a keen understanding of the immediate environment. Overall, each settlement appears to have opted to apply individual behaviours at a small scale. In addition, it has been shown that dairy products typically comprise a significant amount of the diet for many Marmara Region residents during the Neolithic Period (Evershed et al. 2008; Özbal et al. 2013; Thissen et al. 2010). Meat would naturally also have contributed to the nutrition needs, but it is still generally thought that economic plants and mostly cereals formed the largest percentage of these communities’ diets, because they were also intensive farmers. We know this especially from the carbon and nitrogen isotope analyses on bone col­lagen from individuals at Aktopraklik and Barcin Höyük from the work of Chelsea Budd et al. (2013; 2018; 2020). Fish consumption also seems to play a notable role at some sites, but it mostly appears as a supplementary food in the diet. Given this com­plex picture, our archaeological interpretation must, for the moment, remain incomplete and perhaps in­accurate until multi-proxy subsistence research is car­ried out for each settlement. Conclusions The main aims of this study have been to use the archaeobotanical datasets from Barcin Höyük and Bahçelievler to discern variability in Neolithization processes in Northwest Anatolia in the seventh mil­lennium BC, and to compare local community-based adaptations with the macro-regional phenomenon of Neolithization. A careful study of the datasets from each site shows various nuances in the specific economic plant packages, which can be clustered under four groups. First, cereals such as barley, ein­korn, emmer, and bread/hard wheat, are identified with certainty for Phase VIe at Barcin Höyük, but barley and einkorn are not favoured in the contem­poraneous Phase 6 at Bahçelievler. Second, the data did yield some differences in the presence of flax and a small-sized naked wheat, which are both pre­sent in Phase VId1 at Barcin Höyük but have not been found at Bahçelievler. Third, the presence of pulses such as lentils, peas, chickpeas, and bitter vetch differs between phases at both sites. Although lentils, bitter vetch, and peas were identified in near­ly negligible amounts in Phase VIe at Barcin Höyük, these pulses become common in the subsequent phase and chickpeas also emerge within the local inventory at this point. At Bahçelievler, on the other hand, lentils and a single seed of bitter vetch were identified in Phase 6, while lentils continue to be the only identified pulse species in the overlying Phase 5. Finally, the presence of gathered plants, conspicuously lacking from the earliest phases, ap­pears in the subsequent levels of both Barcin Hö­yük and Bahçelievler. Hazelnut and bramble are found at Barcin Höyük while pistacia and grape oc­cur at Bahçelievler. Despite the proximity of the two sites, the results show distinct local food practices and potential re-interpretations of the process of Neolithization. Based on the current evidence, the pioneer settlers at Barcin Höyük appear to have brought their full subsistence package with them. The data from Bah-çelievler also suggest a reliance on non-local econo­mic plants. However, the inhabitants of Bahçelievler appear to show more readiness to exploit local wild resources and to integrate gathered plants into the local subsistence strategies in Phases 6 and 5. We observed that the economic plant range in both sites remains limited, especially when compared with the later levels. While some of the differences observed between Barcin Höyük and Bahçelievler may be related to sub-regional climatic variability, local geographical conditions, or vegetational diffe­rences and the particulars of the plant economies at each site were the outcome of the choices made by their respective communities, based likely on lo­cal cultural preferences and social practices. Suggesting that this results from the divergent path­ways that the inhabitants of these sites took in the process of Neolithization admittedly requires a large leap. Whether the reliance on gathering at Bahçe­lievler, with its semi-subterranean round houses, Hüreyla Balci, Rana Özbal, Fokke Gerritsen, and Erkan Fidan was a remnant of a practice the inhabitants held onto since the pre-Neolithic periods is difficult to ascertain and cannot be addressed confidently with macro-botanical data alone. Nonetheless, we can at least propose that the behaviour that emerges from the choices that the inhabitants of each site made were due to a complex set of habits and environ­mental circumstances. This notion poses new ques­tions about this region’s transition to the Neolithic. Overall, though, at both sites the majority of the botanical remains, and hence the main subsistence strategy, remains one that is based on the cultiva­tion and dominance of economic plants, yet there are clearly unique ways in which the inhabitants of each site perceived and incorporated wild resources within their diet. ACKNOWLEDGEMENTS The excavations at Barcin Höyük were funded by the Netherlands Organization for Scientific Research (NWO) and the Netherlands Institute in Turkey. The Bahçelievler excavations were funded by the Bilecik Municipality. Hüreyla Balci conducted the macro-botanical analysis between the seasons 2013–2015 at Barcin Höyük, and between 2019–2021 at Bahçelievler. Our thanks go to Prof. Dr. René Cappers from the University of Groningen, the Netherlands, for his supervision of the analysis of the Barcin Höyük samples, and his corrections, sugges­tions, and comments on this paper. Other thanks go to Cavit Özcan and Mine Sanli from Bilecik Seyh Edebali University, and Ayse Hacibektasoglu from Istanbul University for helping with the flotation and preparing the samples at Bahçelievler. And final thanks go to Barcin Höyük and Bahçelievler Team Members for their hard work during the excavation seasons. This paper constitutes a part of HB’s doctoral research. . References Arbuckle B. S., Whitcher Kansa S., Kansa E., +19 authors, and D. Würtenberger. 2014. Data Sharing Reveals Com­plexity in the Westward Spread of Domestic Animals ac­ross Neolithic Turkey. PLoS ONE 9(6): 1–11. https://doi.org/10.1371/journal.pone.0099845 . Atalay I., Mortan K. 2011. Türkiye Bölgesel Cografyasi. . Inkilap Publications. Istanbul. Balci H. 2018. Kuzeybati Anadolu’da Tarimin Baslan­gici: Barcin Höyük Bitki Kalintilari Üzerine Bir Deger- . lendirme. MA thesis. Social Sciences Institute. Istanbul . University. Istanbul. Balci H., Cappers R. T. J., Gerritsen F., and Özbal R.. 2019. Barcin Höyük’te Bitki Seçimi: 2013–2015 Yili Arkeobotanik Sonuçlarinin Degerlendirilmesi. Arkeome­tri Sonuçlari Toplantisi 34: 333–352. Boessneck J., Von den Driesch A. 1979. Die Tierknochen­funde aus der neolithischen Siedlung auf dem Fikirte­pe bei Kadiköy am Marmarameer. Institut für Palaeo­anatomie, Domestikationsforschung und Geschichte der Tiermedizin der Universität München. München. Bojnansky V., Fargasova A. 2007. Atlas of Seeds and Fruits of Central and East-European Flora. The Carpa­thian Mountains Region. Springer. Dordrecht. https://doi.org/10.1007/978-1-4020-5362-7 Bottema S., Woldring H. 1995. The Prehistoric Environ­ . ment of the Lake Iznik Area; a Palynological Study. In J. J. Roodenberg (ed.), The Ilipinar Excavations I: Five Sea­sons of Fieldwork in NW Anatolia 1987–91. Nederlands Historisch-Archaeologisch Institut te Istanbul 72. Neder-lands Historisch-Archaeologisch Instituut te Istanbul. Lei­den: 8–16. Bottema S., Woldring H., and Kayan I. 2001. The Late Quaternary Vegetation History of Western Turkey. In J. J. Roodenberg, L. C. Thissen (eds.), The Ilipinar Excava­tions II. Uitgaven van het Nederlands Historisch-Archaeo­logisch Instituut te Istanbul 93. Nederlands Instituut voor het Nabije Oosten. Leiden: 327–354. Budd C., Lillie M., Alpaslan-Roodenberg S., Karul N., and Pinhasi R. 2013. Stable isotope analysis of Neolithic and Chalcolithic populations from Aktopraklik, northern Ana­tolia. Journal of archaeological science 40(2): 860– 867. https://doi.org/10.1016/j.jas.2012.09.011 Budd C., Karul N., Alpaslan-Roodenberg S., Galik A., Schul-ting R., and Lillie M. 2018. Diet uniformity at an early farming community in northwest Anatolia (Turkey): car­bon and nitrogen isotope studies of bone collagen at Aktopraklik. Archaeological and Anthropological Scien­ces 10(8): 2123–2135. https://doi.org/10.1007/s12520-017-0523-4 Budd C., A. Galik, S. Alpaslan-Roodenberg, R. Schulting, and Lillie M. 2020. Early farmers in northwest Turkey: First dietary isotopes study of human diet at Neolithic Bar­cin Höyük. Journal of Archaeological Science: Reports 31: 1–8. https://doi.org/10.1016/j.jasrep.2020.102288 Different manifestations of Neolithization in Northwest Anatolia| An archaeobotanical review from Barcin Höyük and Bahçelievler, Turkey Buitenhuis H. 2008. Ilipinar: the faunal remains from the late Neolithic and earlyChalcolithic levels. MOM Éditions 49(1): 299–322. https://www.persee.fr/doc/mom_ 1955­ 4982_2008_act_49_1_2711 Cappers R. T. J. 2008. Plant Remains from the Late Neoli­thic and Early Chalcolithic levels. In J. J. Roodenberg, S. Alpaslan Roodenberg (eds.), The Ilipinar Excavations. Volume III. With contributions on Hacilartepe and Men-tese. Uitgaven van het Nederlands Instituut voor het Na-bije Oosten te Leiden. Nederlands Instituut voor het Nabije Oosten. Leiden: 117–148. 2014. Archaeobotanical Evidence of Agriculture in Neo­lithic Turkey. In M. Özdogan, N. Basgelen, and P. Kuni­holm (eds.), The Neolithic in Turkey: 10500–5200 BC Environment Settlement, Flora, Fauna, Dating, Sym­bols of Belief, With Views from North, South, East, And West. Volume 6. Archaeology and Art Publications. . Istanbul: 205–222. Cappers R. T. J., Bekker R. M., and Jans J. E. A. 2012. Di­gital Seed Atlas of the Netherlands. Barkhuis & Univer­sity of Groningen Library. Groningen. Cappers R. T. J., Bekker R. M. 2013. A manuel for the Identification of Plant Seeds and Fruits. Barkhuis & Uni­versity of Groningen Library. Groningen. Cappers R. T. J., Neef R., Bekker R. M., Fantone F., and Okur Y. 2016. Digital Atlas of Traditional Agricultural Practices and Food Processing. Volume 1–3. Barkhuis & University of Groningen Library. Groningen. Clare L., Weninger B. 2014. The Dispersal of Neolithic Lifeways: Absolute Chronology and Rapid Climate Change in Central and West Anatolia. In. M. Özdogan, N. Basge­len, and P. Kuniholm (eds.), The Neolithic in Turkey: 10500–5200 BC Environment Settlement, Flora, Fau­na, Dating, Symbols of Belief, With Views from North, South, East, And West. Volume 6. Archaeology and Art . Publications. Istanbul: 1–65. Cristiani E., Radini A., Edinborough M., and Bori. D. 2016. Dental calculus reveals Mesolithic foragers in the Balkans consumed domesticated plant foods. Proceedings of the National Academy of Sciences of the United States of America 113–37: 10298–10303. https://doi.org/10.1073/pnas.1603477113 Çakirlar C. 2013. Rethinking Neolithic subsistence at the gateway to Europe with new archaeozoological evidence from Istanbul. In M. Groot, D. Lentjes, and J. Zeiler (eds.), Barely Surviving or More than Enough? The environ­mental archaeology of subsistence, specialization, and surplus food production. Sidestone Press. Leiden: 59–79. 2015. Adaptation, identity, and innovation in Neolithic and Chalcolithic western Anatolia (6800–3000 cal. BC). The evidence from aquatic mollusc shells. Quaternary International 390: 117–125. https://doi.org/10.1016/j.quaint.2015.05.008 Çevik Ö., Abay E. 2016. Neolithisation in Aegean Turkey: Towards a More Realistic Reading. Anatolian Metal 7: 187–197. Çilingiroglu Ç. 2005. The concept of “Neolithic package”: considering its meaning and applicability. Documenta Praehistorica 32: 1–13. https://doi.org/10.4312/dp.32.1 Douka K., Efstratiou N., Hald M. M., Henriksen P. S., and Karetsou A. 2017. Dating Knossos and the arrival of the earliest Neolithic in the southern Aegean. Antiquity 91 (356): 304–321. https://doi.org/10.15184/aqy.2017.29 Düring B. S. 2011. The prehistory of Asia Minor: from complex hunter-gatherers to early urban societies. Cam­bridge University Press. Cambridge. 2013. Breaking the Bond: Investigating the Neolithic Expansion in Asia Minor in the Seventh Millennium BC. Journal World Prehistory 26: 75–100. https://doi.org/10.1007/s10963-013-9065-6 Evershed R. P., Payne S., Sherratt A. G., +18 authors, and Burton M. M. 2008. Earliest date for milk use in the Near East and southeastern Europe linked to cattle herding. Nature 455.7212: 528–531. https://doi.org/10.1038/nature07180 Fairbairn A., Martinoli D., Butler A., and Hillman G. 2007. Wild plant seed storage at Neolithic Çatalhöyük East, Tur­key. Vegetation History and Archaeobotany 16(6): 467– 479. https://doi.org/10.1007/s00334-006-0069-3 Fidan E. 2020. Fikirtepe Culture And Before: Preliminary Results From The Rescue Excavation of Bilecik Bahçeliev­ler. Archaeology and Art 163: 29–38. Gaastra J. S., de Vareilles A., and Vander Linden M. 2022. Bones and Seeds: An Integrated Approach to Understand­ing the Spread of Farming across the Western Balkans. Environmental Archaeology 27: 44–60. https://doi.org/10.1080/14614103.2019.1578016 Galik A. Barcin Höyük Zooarchaeology Data. Open Con­text. https://opencontext.org/tables/23d7c8387a870c5 6fd4b5d47500f6311 Gatsov I. 2001. Epipalaeolithic/Mesolithic, Neolithic pe­riods chipped-stone assemblages from southern Bulgaria and northwest Turkey: Similarities and differences. Tur­ Hüreyla Balci, Rana Özbal, Fokke Gerritsen, and Erkan Fidan kiye Bilimler Akademisi Arkeoloji Dergisi (TUBA-AR) 4: 101–112. Gatsov I., Özdogan M. 1994. Some Epipalaeolithic sites from NW Turkey: Agaçli, Domali, and Gümüsdere. Anato­lica 20: 97–120. Gerritsen F., Özbal R. 2016. Barcin Höyük and the pre-Fikirtepe Neolithization of the Eastern Marmara Region. Anatolian Metal 7: 199–208. 2019. Barcin Höyük, a seventh millennium settlement in the Eastern Marmara region of Turkey. Documenta Praehistorica 46: 58–67. https://doi.org/10.4312/dp.46.4 Gerritsen F., Özbal R., and Thissen L. 2013a. Barcin Hö­yük: The Beginnings of Farming in the Marmara Region. In M. Özdogan, N. Basgelen, and P. Kuniholm (eds.), The Neolithic in Turkey: New Excavations-New Research, . Northwestern Turkey and Istanbul. Volume 5. Archaeo- . logy and Art Publications. Istanbul: 93–112. 2013b. The Earliest Neolithic Levels at Barcin Höyük. Anatolica XXXIX: 53–92. Gourichon L., Helmer D. 2008. Etude de la faune néolithi­que de Mentese (Turquie). In J. J. Roodenberg, S. Alpa­slan Roodenberg (eds.), Life and Death in a Prehistoric Settlement in Northwest Anatolia. The Ilipinar Excava­tions. Volume III. With contributions on Hacilartepe and Mentese. Uitgaven van het Nederlands Instituut voor het Nabije Oosten te Leiden. Nederlands Instituut voor het Nabije Oosten. Leiden: 435–448. Harmankaya S. 1983. Pendik Kazisi 1981. In IV. Kazi sonuçlari toplantisi, 8–12 Subat 1982. Hacettepe sosyal ve idari billmler döner serma ye isletmesi tesislerinde ba­silmistir. Eski Eserler ve Müzeler Genel Müdürlügü. Anka­ra: 25–30. Hofmanová Z., Kreutzer S., Hellenthal G., +35 authors, and Burger J. 2016. Early farmers from across Europe directly descended from Neolithic Aegeans. Proceedings of the National Academy of Sciences 113(25): 6886– 6891. https://doi.org/10.1073/pnas.152395111 Horejs B. 2019. Long and short revolutions towards the Neolithic in western Anatolia and Aegean. Documenta Praehistorica 46: 68–83. https://doi.org/10.4312/dp.46.5 Ivanova M., de Cupere B., Ethier J., and Marinova E. 2018. Pioneer farming in southeast Europe during the early sixth millennium BC: Climate-related adaptations in the exploi­tation of plants and animals. PLOS One 13: e0197225. https://doi.org/10.1371/journal.pone.0197225 . Izdal-Çaydan B. 2018. Kuzeybati Anadolu’da Neolitik Dönem Geçim Ekonomisinde Hayvanlarin ve Hayvansal .. Ürünlerin Yeri. MA thesis. Istanbul University. Istanbul. Jovanovi. J., Power R. C., de Becdeliévre C., Goude G., and Stefanovi. S. 2021. Microbotanical evidence for the spread of cereal use during the Mesolithic-Neolithic transition in the Southeastern Europe (Danube Gorges): Data from den­tal calculus analysis. Journal of Archaeological Science 125: 1–14. https://doi.org/10.1016/j.jas.2020.105288 Karul N. 2010. A New Prehistoric Settlement in Northwest Turkey: Aktopraklik Höyük. In P. Matthiae, F. Pinnock, L. Nigro, and N. Marchetti (eds.), Proceedings of the 6th In­ternational Congress on the Archaeology of the Ancient Near East May, 5th 10th 2009. “Sapienza” – Universita di Roma. Volume 3. Islamic Session. Poster Session. The Ceremonial Precinct of Canaanite Hazor. Harrassowitz Verlag. Wiesbaden: 285–390. 2011. The Emergence of Neolithic Life in South and East Marmara Region. In R. Krauß (ed.), Beginnings – new research in the appearance of the Neolithic be­tween Northwest Anatolia and the Carpathian basin: papers of the international workshop, 8th–9th April 2009, Istanbul. Menschen – Kulturen – Traditionen. Bd. 1. Deutsches Archäologisches Institut. Forschungsclus­ter 1: Von der Sesshaftigkeit zur Komplexen Gesel­lschaft: Siedlung, Wirtschaft, Umwelt. Verlag Marie Lei-dorf GmbH. Rahden/Westf.: 57–65. 2017. Aktopraklik: Tasarlanmis Prehistorik Bir Köy. . Ege Yayinlari. Istanbul. Karul N., Avci M. B. 2011. Neolithic Communities in the Eastern Marmara Region: Aktopraklik C. Anatolica 37: 1–15. . Kayan I. 2014. Paleogeography of the Coastal Regions of Turkey During the Neolithic Period. In M. Özdogan, N. Basgelen, and P. Kuniholm (eds.), The Neolithic in Tur­key: 10500–5200 BC Environment Settlement, Flora, Fauna, Dating, Symbols of Belief, With Views From North, South, East, And West. Volume 6. Archaeology . and Art Publications. Istanbul: 95–123. Kayan I., Woldring H. 2001. The Late Quaternary vege­tation history of western Turkey. In J. J. Roodenberg, L. C. Thissen (eds.), The Ilipinar Excavations II. Uitgaven van het Nederlands Historisch-Archaeologisch Instituut te Is­tanbul 93. Nederlands Instituut voor het Nabije Oosten. Leiden: 327–354. Kilinç G. M., Omrak A., Özer F., +23 authors, and Göther­ström A. 2016. The Demographic Development of the First Farmers in Anatolia. Current Biology 26(19): 2659– 2666. https://doi.org/10.1016/j.cub.2016.07.057 Different manifestations of Neolithization in Northwest Anatolia| An archaeobotanical review from Barcin Höyük and Bahçelievler, Turkey Kiziltan Z. 2013. Pendik Höyük Kazilari. Aktüel Arkeoloji 34: 32–36. Kiziltan Z., Polat M. A. 2013. The Neolithic at Yenikapi: Marmaray-Metro Project Rescue Excavations. In M. Özdo­gan, N. Basgelen, and P. Kuniholm (eds.), The Neolithic in Turkey: New Excavations – New Research, North­western Turkey and Istanbul. Volume 5. Archaeology . and Art Publications. Istanbul: 113–165. Kolankaya-Bostanci N., Fidan E. 2021. Bilecik Bahçeliev- . ler Neolitik Çag Yontmatas Topluluguna Ait Ilk Degerlen­dirmeler. Anadolu Arastirmalari 24: 93–116. https://doi.org/10.26650/anar.2021.24.935202 Kotzamani G., Livarda A. 2018. People and plant entan­glements at the dawn of agricultural practice in Greece. An analysis of the Mesolithic and early Neolithic archaeo-botanical remains. Quaternary International 496: 80– 101. https://doi.org/10.1016/j.quaint.2018.04.044 Krauß R., Marinova E., de Brue H., and Weninger B. 2017. The rapid spread of early farming from the Aegean into the Balkans via the Sub-Mediterranean-Aegean Vegetation Zone. Quaternary International 496: 24–41. https://doi.org/10.1016/j.quaint.2017.01.019 Lazaridis I., Nadel D., Rollefson G., +49 authors, and Reich D. 2016. Genomic insights into the origin of farming in the ancient Near East. Nature 536(7617): 419–424. https://doi.org/10.1038/nature19310 Lefebvre H., Nicholson-Smith D. 1991. The production of space. Blackwell. Oxford. Marchi N., Winkelbach L., Schulz I., +27 authors, and Ex-coffier L. 2022. The Genomic Origins of the World’s First Farmers. Cell 185: 1842–1859. https://doi.org/10.1016/j.cell.2022.04.008 Marinova E., Popova T. 2008. Cicer arietinum (chickpea) in the Neolithic and Chalcolithic of Bulgaria: implications for cultural contacts with the neighbouring regions? Vege­tation History and Archaeobotany 17 (Suppl 1): 73–80. https://doi.org/10.1007/s00334-008-0159-5 Marinova E., Krauß R. 2014. Archaeobotanical evidence on the Neolithisation of Northeast Bulgaria in the Balkan-Anatolian context: chronological framework, plant eco­nomy, and land use. Bulgarian e-Journal of Archaeology 4: 179–194. https://be-ja.org/index.php/journal/ar ticle/ view/be-ja-4-2-2014-179-194 Mathieson I., Lazaridis I., Rohland N., +34 authors, and Reich D. 2015. Genome-wide patterns of selection in 230 ancient Eurasians. Nature 528(7583): 499–503. https://doi.org/10.1038/nature16152 Mathieson I., Alpaslan-Roodenberg S., Posth C., +114 au­thors, and Reich D. 2018. The genomic history of south­eastern Europe. Nature 555: 197–203. https://doi.org/10.1038/nature25778 Neef R., Cappers R. T. J., and Bekker R. M. 2012. Digital Atlas of Economic Plants in Archaeology. Barkhuis & University of Groningen Library. Groningen. Özbal H., Thissen L., Dogan T., Gerritsen F. A., Özbal R. D., and Türkekul Biyik A. 2013. Neolitik Bati Anadolu ve Mar­mara yerlesimleri çanak çömleklerinde organik kalinti analizleri. In H. Dönmez, Ö. Ötgün (eds.), Arkeometri So­nuçlari Toplantisi. Mugla Sitki Koçman Üniversitesi Basi­mevi. Ankara: 105–114. Özbal R., Gerritsen F. 2019. Farmer-Forager Interactions the Central/Western Anatolian Farming Frontier: Neolithi­zation of Northwest Anatolia: Reassessing the Evidence. In M. Brami, B. Horejs (eds.), Proceedings of the interna­tional workshop held at the 10ICAANE Conference in Vienna, April 2016. Oriental and European Archaeology. OREA 12. Österreichische Akademie der Wissenschaften­Philosophisch-historische Klass. Austrian Academy of Sci­ences Press. Vienna: 181–209. Özdogan E. 2016. Diversity and Homogeneity among the Early Farming Communities of Western Anatolia. Docu­menta Praehistorica 43: 265–282. https://doi.org/10.4312/dp.43.13 Özdogan M. 1983. Pendik. A Neolithic site of Fikirtepe cul­ture in the Marmara region. In R. M. Boehmer, H. Haupt­mann (eds.), Beiträge zur Altertumskunde Kleinasiens. Festschrift fur Kurt Bittel. Beiträge zur Altertumskunde Kleinasiens. Philipp von Zabern in Wissenschaftliche Buch­gesellschaft. Mainz, 401–411. 2010. Westward expansion of the Neolithic way of life: Sorting the Neolithic package into distinct packages. Near Eastern Archaeology in the Past, Present and Fu­ture. Heritage and Identity 1: 883–897. 2011. Archaeological evidence on the westward expan­sion of farming communities from eastern Anatolia to the Aegean and the Balkans. Current Anthropology 52 (S4): 415–430. https://doi.org/10.1086/658895 2013. Neolithic Sites in the Marmara Region: Fikirtepe, Pendik, Yarimburgaz, Toptepe, Hoca Çesme, and Asagi Pinar. In M. Özdogan, N. Basgelen (eds.), Neolithic in Turkey. Volume 5. Archaeology and Art Publications. . Istanbul: 167–269. 2014. A new look at the introduction of the Neolithic way of life in Southeastern Europe. Changing paradigms of the expansion of the Neolithic way of life. Documen­ Hüreyla Balci, Rana Özbal, Fokke Gerritsen, and Erkan Fidan ta Praehistorica 41: 33–49. https://doi.org/10.4312/dp.41.2 Özdogan M., I. Gatsov. 1998. The Aceramic Neolithic pe­riod in western Turkey and in the Aegean. Anatolica 24: 209–232. Pasinli A., Uzunoglu E., Atakan N., Girgin Ç., and Soysal M. 1994. Pendik Kurtarma Kazisi. Müze Kurtarma Kazilari Semineri IV: 147–163. Popova T., Marinova E. 2007. Paleoethnobotanical data in Southwestern region of Bulgaria. In H. Todorova, M. Ste-fanovich (eds.), The Struma/Strymon River Valley in Prehistory. In The Steps of James Harvey Gaul. Volume 2. Museum of History-Kyustendil. Gerda Henkel Stiftung. Sofia: 523–532. Robb J. 2013. Material Culture, Landscapes of Action, and Emergent Causation: A New Model for the Origins of the European Neolithic. Current Anthropology 54(6): 657– 683. https://doi.org/10.1086/673859 Robert N. 2014. The Climate of Neolithic Anatolia. In M. Özdogan, N. Basgelen, and P. Kuniholm (eds.), The Neoli­thic in Turkey: 10 500–5200 BC Environment Settle­ment, Flora, Fauna, Dating, Symbols of Belief, With Views from North, South, East, and West. Volume 6. Ar­ . chaeology and Art Publications. Istanbul: 67–94. Röhrs M., Herre W. 1961. Zur Frühentwicklung der Haus­tiere: Die Tierreste der neolithischen Siedlung Fikirtepe am Kleinasiatischen Gestade des Bosporus. Zeitschrift für Tierzüchtung und Züchtungsbiologie 75(1–4): 110–127. Roodenberg J. J. 2008. Stratigraphy and Architecture of Phases X and IX. In. J. J. Roodenberg, S. Alpaslan Rooden-berg (eds.), Life and Death in a Prehistoric Settlement in Northwest Anatolia. The Ilipinar Excavations. Volume III. With contributions on Hacilartepe and Mentese. Uit­gaven van het Nederlands Instituut voor het Nabije Oos-ten te Leiden. Nederlands Instituut voor het Nabije Oosten. Leiden: 1–34. Roodenberg J., Van As A., Jacobs L., and Wijnen M. H. 2003. Early Settlement in the Plain of Yenisehir (NW Ana­tolia): The Basal Occupation Layers at Mentese. Anatoli-ca XXIX: 17–59. Sari D., Akyol S. 2019. The Early Neolithic pottery of Ke­çiçayiri and its place in the North-western Anatolian Neo­lithisation process. Documenta Praehistorica 46: 138– 156. https://doi.org/10.4312/dp.46.9 Schroedter T. M., Nelle O. 2015. New insights into Mid-Ho­locene vegetation in the southern Marmara region: Char­coal from the Late Neolithic to Early Chalcolithic settle­ment site Aktopraklik, Northwestern Turkey. Quaternary International 366: 81–95. https://doi.org/10.1016/j.quaint.2014.11.031 Thissen L. 1999. Trajectories towards the Neolithisation of NW Turkey. Documenta Praehistorica 26: 29–39. https://www.dlib.si/stream/URN:NBN:SI:DOC-4K369 HZK/7303e243-bcd7-48b2-b9b1-49e70a9081b1/PDF Thissen L., Özbal H., Türkekul-Biyik A., Gerritsen F., and Özbal R. 2010. The Land of Milk? Approaching Dietary Preferences of Late Neolithic Communities in NW Anato­lia. Leiden Journal of Pottery Studies 26: 157–172. Ulas B. 2020. Reappraisal of the Neolithisation of the Mar­mara Region Through Archaeobotanical Analysis at Pen-dik Höyük and Yenikapi. Arkeoloji ve Sanat 64: 27–40. Yaka R., Mapelli I., Kaptan D., +53 authors, and Somel M. 2021. Variable kinship patterns in Neolithic Anatolia re­vealed by ancient genomes. Current Biology 31(11): 2455–2468. https://doi.org/10.1016/j.cub.2021.03.050 van den Bos E. O. 2021. Living Neolithization: A Multi-Scalar Approach to Houses, Settlements and Habitation Practices in Western Anatolia and Southeastern Europe (c. 7000–5000 BCE). Unpublished PhD thesis. Vrije Uni-versiteit Amsterdam. Amsterdam. https://research.vu.nl/ en/publications/living-neolithization-a-multi-scalar-app roach-to-houses-settlemen van Zeist W., van-Waterbolk R. 1995. Flora Remains From Late-Neolithic Ilipinar. In J. J. Roodenberg (ed.), The Ilipi­nar Excavations I: Five Seasons of Fieldwork in NW Anatolia 1987–91. volNederlands Historisch-Archaeolo­gisch Institut te Istanbul 72. Nederlands Historisch-Archa­eologisch Instituut te Istanbul. Leiden: 159–166. Weninger B., Clare L., Gerritsen F., Horejs B., Krauß R., Özbal R., and Rohling E. 2014. Neolithisation and Rapid Climate Change (6600–6000 cal BC) in the Aegean and Southeast Europe. Documenta Praehistorica 41: 1–31. https://doi.org/10.4312/dp.41.1 Würtenberger D. 2012. Archäozoologische Analysen am Fundmaterial des Barcin Höyüks im Vergleich mit aus­gewählten Fundstellen des 7. und 6. Jt. v. Chr. in Nord-west- und Westanatolien. MA thesis. University of Vienna. Wien. Zeder M. A. 2011. The origins of agriculture in the Near East. Current Anthropology 52(S4): S221–S235. https://doi.org/10.1086/659307 Zvelebil M. 2001. The agricultural transition and the ori­gins of Neolithic society in Europe. Documenta Praehi­storica 28: 1–26. https://doi.org/10.4312/dp.28.1 back to contents Documenta Praehistorica XLIX (2022) What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans 1 Ana — Duri;ic´ ana.djuricic@f.bg.ac.rs Laboratory for Bioarchaeology, Department of Archaeology, Faculty of Philosophy, University of Belgrade, RS ABSTRACT – Food is essential for survival, but how humans obtain and manage it is regulated so­cially. The life of Neolithic and other non-industrial communities depended on environmental varia­tions – temperature patterns and precipitation. For farming communities, even minor changes in those patterns could have led to periods of food scarcity. In order to overcome and prepare for pe­riods of scarcity, non-industrial communities applied different social buffering strategies. In this pa­per, the social buffering strategies Early/Middle Neolithic Star.evo and Late Neolithic Vin.a culture communities applied in overcoming the environmental variability are tested and the most plausible ones are considered. KEY WORDS – Neolithic; Balkan Neolithic; subsistence; mobility; farming Kaj naredimo za hrano – socialne strategije za premagovanje pomanjkanja hrane v neolitiku na srednjem Balkanu IZVLE.EK – Hrana je bistvena za pre.ivetje, in dru.be urejajo, kako jo ljudje pridobivamo in uprav­ljamo. .ivljenje neolitskih in drugih predindustrijskih skupnosti je bilo odvisno od sprememb v oko­lju – temperaturnih in padavinskih vzorcev. .e manj.e spremembe v teh vzorcih bi lahko povzro.i­le pomanjkanja hrane. Da bi to prepre.ile in se pripravile na pomanjkanje, so predindustrijske skup­nosti uporabile razli.ne strategije socialne varnosti. V .lanku preverjamo in predstavljamo najbolj verjetne strategije socialne varnosti, ki sta jih uporabili kulturi Star.evo in Vin.a v zgodnjem, sred­njem in poznem neolitiku za premagovanje posledic spreminjanja okolja. KLJU.NE BESEDE – neolitik; Balkan; strategija pre.ivetja; mobilnost; poljedelstvo Introduction The Neolithic is traditionally perceived as a time leading to the emergence of a new lifestyle in a wi-when farmers lived in permanent settlements, but der area. The Neolithic package is often described as this simplified generalization cannot be universally several factors that differentiate the Neolithic from applied. By introducing the concept of the ‘Neoli-previous lifestyles and it includes agriculture (plant thic package’, archaeologists tried to unify cultural, domestication and cultivation), domestic animal economic, social and ideological innovations which farming (animal domestication), the emergence of occurred in the Near East during the Neolithic. These sedentism, pottery production, polished stone tools innovations spread beyond the original territory, and an ideology compatible with the new lifestyle 1 This paper is a result of the work on the project ‘Humans and Society in Times of Crisis, Archaeology of Crisis’ (Faculty of Philo­sophy, University of Belgrade). In the first phase of the project social buffering strategies applied by Early/Middle Neolithic commu­nities at the Central Balkans were studied (.uri.i. 2021). This paper is a result of a second phase of the project, where this sub­ject is further elaborated and social buffering strategies applied by Late Neolithic communities are added to the study. DOI> 10.4312\dp.49.23 What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans (e.g., Çilingiroglu 2005; Cauvin 2000; .uri.i. 2021), which led to new subsistence strategies and new risks. At the territory of the Central Balkans, the Neolithic lifestyle arrived already formed together with a new population around 6250 cal BC. (Bori., Dimitrijevi. 2007; Bori. 2014; 2016; Cramp et al. 2019; Ivano­va 2020; Stojanovski et al. 2020; .uri.i. 2021). The area these early farming communities of the Early and Middle Neolithic Star.evo culture (6250–5300 cal BC) occupied was predominantly previously non-inhabited, except from the Danube Gorges, where the local Mesolithic population was still thriving (where the local Mesolithic and new Neolithic popu­lations came in contact) (Bori., Dimitrijevi. 2007; Bori. 2014; 2016; Cramp et al. 2019; .uri.i. 2021). The arrival of the Neolithic population coincided with the 8.2-kiloyear event (Por.i. et al. 2021), which affected the temperature and precipitation patterns in the Central Balkan region (Bonsall 2007; Kobashi et al. 2007; Gronenborn 2009). The Late Neolithic Vin.a culture (5350/5300–4500 cal BC) (Por.i. 2020), occupied similar territory as the Star­.evo culture, but their settlements and material cul­tures differed significantly (Fig. 1). Even though the process of cereal and animal domes­tication was a long-lasting one, fully domesticated plants and animals were documented in numerous settlements in the Near East, dating back to the Pre-Pottery Neolithic (having a dominant role in the sub­sistence from the Pre-Pottery Neolithic B) (Kuijt, Go-ring-Morris 2002; Gibbs, Jordan 2016). A sedentary lifestyle forms a base for the intensification of the economy, population growth and increase of social complexity (Neil et al. 2016.1). However, Neolithic innovations were applied differently in different re­gions, resulting in a non-uniform Neolithic lifestyle with diverse sedentism and mobility patterns (see Neil et al. 2016). Agriculture and husbandry usually form the base of Neolithic subsistence. The amount of food in farming communities depends on predic­table seasonal changes and unpredictable weather/ climate factors (temperature, amount of precipita­tion, type of precipitation during certain seasons, droughts, floods, climate change…). Predictable fac­tors (seasonal changes) are something that cannot Fig. 1. Distribution of Star.evo–Körös–Cris and Vin.a cultures. Sites: 1 Blagotin, 2 Star.evo, 3 Medjure., 4 Drenovac, 5 Nosa – Biserna Obala, 6 Donja Branjevina, 7 Golokut – Vizi., 8 Gomolava, 9 Opovo, 10 Se-levac, 11 Vin.a – Belo Brdo, 12 Stubline – Crkvine, 13 Banjica – Usek, 14 Divostin, 15 Grivac, 16 Med-vednjak, 17 Vala., 18 Pavlovac – .ukar, Gumni.te, 19 Lepenski Vir, 20 Zadubravlje, 21 Galovo, 22 Zve.­ka, 23 Vinkovci (map by S. .ivanovi.). Ana —Duri;ic´ be avoided, so the communities apply different buf­fering strategies to obtain a balanced diet through­out the year. But even minor changes in tempera­ture or precipitation patterns can sometimes endan­ger the existence of farming-based non-industrial communities, due to crop failure and lack of fodder for animals. Bad conditions can last one or multi­ple years, forcing communities to employ one or se­veral survival strategies, based on the duration of the unfavourable period and their cultural, social, economic and ideological preferences (Halstead, O’Shea 1989; .uri.i. 2021). The Neolithic lifestyle, being completely different from the previous ones, with the increase in population and reliance on do­mesticates, resulted in new survival strategies. Four social buffering strategies for overcoming food scarcity have been proposed by Paul Halstead and John O’Shea (1989): (1) diversification, (2) storage, (3) exchange and (4) mobility (Halstead, O’Shea 1989; Groot, Lentjes 2013). These buffering strate­gies are not applied only when food shortages occur, but also refer to actions undertaken to prevent the lack of food and prepare for inevitable periods of the year when food is less available in nature (.u­ri.i. 2021). In this paper, the social buffering stra­tegies Early/Middle Neolithic Star.evo and Late Neo­lithic Vin.a culture communities applied in order to prepare for, prevent and overcome periods of food scarcity caused by environmental variability are te­sted and the most plausible ones are considered. The similarities and differences between social buffering strategies practised during the Early/Middle and Late Neolithic of the Central Balkans will also be com­pared. Overcoming the environmental variability Food scarcity is not connected solely to natural disa­sters, and lack of food can occur due to different fac­tors. Solar radiation influences ecosystem dynamics. Temperate climate zones, like the Central Balkan re­gion, have uneven solar radiation, leading to pro­nounced differences between seasons (Rowley-Con­wy, Zvelebil 1989.41; .uri.i. 2021). During spring vegetation grows suddenly, giving the plants limited time for growth, development and reproduction. Summer is the season of abundance with numerous migratory animal and plant species – cereals, fruits, vegetables and berries. A decrease in resources, both local and migratory, is visible during autumn, whilst winter is the most difficult period, with the least amount of available food (Rowley-Conwy, Zvelebil 1989.41). These are seasonal fluctuations or varia­tions, which are predictable and cannot be avoided, so people apply previously established mechanisms for overcoming periods of food scarcity, to obtain a balanced diet throughout the year (Rowley-Conwy, Zvelebil 1989.41; O’Shea 1989.57; Halstead 1989. 71; .uri.i. 2021). Interannual fluctuations or va­riations depend upon numerous unpredictable fac­tors. They can depend on climatological factors (droughts, frosts, storms, hail, excessive rain at the wrong time, floods, and so on), animals, insects, pla­gue, plant and animal diseases or human activity. In farming communities, interannual variations like drought can affect crop yields or animal disease can increase domestic animal mortality rates, forcing people to compensate for the losses in alternative ways. Even though these variations are mainly un­predictable people know how to deal with them, using experience gained from previous similar situ­ations. (O’Shea 1989.58; Halstead 1989.72; .uri.i. 2021). Long-term fluctuations or variations are the results of climate or natural changes and often last for a longer period of time. To survive these, commu­nities have to make considerable adaptations (Row­ley-Conwy, Zvelebil 1989.44–45; .uri.i. 2021). In facing these different types of variations the strate­gies communities apply can vary and change, but they are always in accordance with their specific na­tural environment, social, cultural and economic norms (.uri.i. 2021). Farming communities depend on plant and animal annual reproductive cycles which are controlled by regular seasonal patterns. Even minor shifts in sea­sonal patterns can influence the amount of food available, whether of plant, or animal origin. To overcome a lack of food, either due to predictable or unpredictable factors, communities apply one of the previously mentioned strategies: diversification, sto­rage, exchange and mobility (Halstead, O’Shea 1989.3; Groot, Lentjes 2013.9; .uri.i. 2021). The inclusion of a greater variety of food sources is called diversification. This also includes keeping certain types of food exclusively for ‘rainy days’ and the cul­tivation of different crops on different soils. For far­mers, agriculture and herding are forms of diversi­fication, but it also includes hunting, fishing and ga­thering. Saving food for annual periods of food scar­city is called storage. Exchange includes different social practices – trade, food sharing, obligatory re­ciprocity and negative reciprocity (theft). Relocation of a group of people or the whole community to­wards areas with available food sources is called mobility (Halstead, O’Shea 1989.3–4; Groot, Lent-jes 2013.9–10; .uri.i. 2021). What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans Communities usually mix several buffering strate­gies. Depending on their preferences, certain strate­gies can be practised as prevention, while others could be applied as a last resort. Also, different cul­tures can apply the same strategy differently (.uri-.i. 2021). A typical agricultural/herding community shows some degree of diversification. They cultivate multiple crop types and keep at least two types of livestock, with additional hunting, fishing and ga­thering that provide a balanced diet and year-round food supply. In situations when a plague or a disease attack crops or animals, diversification provides an effective fall-back strategy, ensuring that at least some food will remain for human consumption. Every society has a tendency to accumulate a sur­plus. Early agricultural communities accumulate a surplus as a result of seasonality, as a strategy to overcome colder periods of the year. During bad years they may have switched to hunting and gath­ering, which was also documented in the archaeo­logical record (Groot, Lentjes 2013.9). On the other hand, hunter-gatherers rely on different wild resour­ces, making their diversification strategy dependent on a variety of wild animal species and wild plant taxa that are available during different seasons (Hal-stead, O’Shea 1989; .uri.i. 2021). According to Paul Halstead, every community generates a surplus, but they define it differently. A distinction should be made between direct and indirect storage. The accu­mulation of food for later consumption is direct sto­rage, but food sharing with other members of the community, expecting reciprocity in time of need, is an example of indirect storage – social storage. With indirect storage the risk of food spoilage is reduced, as the food often cannot be stored for a long period of time (Halstead 1989). Evidence for diversification in the Neolithic of the Central Balkans Diversification is a survival strategy applied by both farming and hunting-gathering communities. Hun-ter-gatherers rely on different wild resources, making their diversification strategy dependent on a variety of wild animal species and wild plant taxa that are available during different seasons (Halstead, O’Shea 1989; .uri.i. 2021), while farming communities rely primarily on agriculture and domestic animal herding with the addition of hunting and gathering, which is often accompanied by a sedentary lifestyle (Çilingiroglu 2005; Cauvin 2000). Diversification is suitable for overcoming seasonal, interannual and long-term variations. It not only implies the exploi­tation of alternative resources, but also the cultiva­tion of different types of cereals and pulses of diffe­rent growing patterns and endurance levels on diffe­rent soil types. By applying this strategy, a commu­nity can potentially reduce the possibility of crop fai­lure (Groot, Lentjes 2013; .uri.i. 2021). Subsis­tence based on domesticated plants and animals is also a form of diversification. Domestic animals can convert agricultural waste and plants unsuitable for human consumption into edible food (meat, milk and fat), further reducing the potential for food shortages (O’Shea 1989; .uri.i. 2021). Archaeobotanical analyses conducted on various Star.evo culture sites have revealed a broad spec­trum of plant food used in the human diet, and these can be divided into cultivated crops and wild plants. However, the results of the archaeobotanical analyses were affected by the lack of systematic sampling, which was conducted only at two sites (Blagotin and Drenovac), while at other sites only contexts marked as interesting or important by ar­chaeologists (‘judgement’ sampling) were sampled (Filipovi., Obradovi. 2013; .uri.i. 2021). The greatest plant taxa diversity was recorded at Dreno-vac, but it was also at this site that the most exten­sive sampling was conducted. Archaeobotanical re­mains from the sites of Blagotin, Drenovac, Medju­re., Star.evo and Nosa-Biserna Obala have confirm­ed the cultivation of five crops2 (for distribution see Filipovi., Obradovi. 2013.41): einkorn (Triticum monococcum), emmer (Triticum dicoccum), hulled barley (Hordeum vulgare, hulled), lentil (Lens culi­naris) and pea (Pisum sativum) (Filipovi., Obrado­vi. 2013; Filipovi. 2014; .uri.i. 2021). A total of nine wild taxa has been documented so far (for dis­tribution see Filipovi., Obradovi. 2013.41): corne­lian cherry (Cornus mas), apple (Malus pumila), fruit from the malus genus (Malus sp.), fruit from the pear genus (Pyrus sp.), acorn (Quercus sp.), beech nut (Fagus sp.), blackberry (Rubus fruticosus), un­identified berries (Rubus sp.) and dwarf elder (Sam­bucus ebulus) (Filipovi., Obradovi. 2013; .uri.i. 2021). Over 70% of the examined faunal assemblage be­longed to domestic animal taxa, indicating a reliance 2 Even though broomcorn millet (Panicum miliaceum) was found at Neolithic sites in the Central Balkans, recent studies have dis­puted its cultivation during the Neolithic in Europe, so all millet samples should be treated as an intrusion (Filipovi. et al. 2020), which is why they were excluded from this paper. Ana —Duri;ic´ on domesticates at the majority of Star.evo culture sites (for detailed archaeozoological analysis and di­stribution see Orton 2012) (Greenfield 2008.108; .uri.i. 2021). Cattle is the most dominant domestic animal taxa, with the exception of Donja Branjevina, where goats/sheep dominate (Orton 2012), while pigs were barely represented in the faunal assem­blage (Orton 2012; Ethier et al. 2017; .uri.i. 2021). Dogs were also present, but were probably not con­sumed. However, sites in the Danube Gorges, Golo­kut-Vizi. in Southwestern Ba.ka and Nosa-Biserna Obala in Northern Ba.ka show different patterns. At these sites, a higher percentage of wild than domes­tic animal taxa have been documented (Orton 2012; .uri.i. 2021). Red deer, roe deer, wild boar, fish, birds and smaller mammals, dominated amongst the wild taxa remains (Greenfield 2008; .uri.i. 2021). The analyses of faunal assemblage from the Early/ Middle Neolithic layers from the Danube Gorges sites show the significance of migratory fish for lo­cal subsistence (Dimitrijevi. et al. 2016). Recent pottery lipid analysis has shown that the majority of vessels in the Danube Gorges settlements were used for cooking aquatic resources, contrary to the data from the sites in the rest of the Central Balkans (Cramp et al. 2019), where pottery was used for processing meat, milk/dairy, plant food and storing beeswax (Ethier et al. 2017; Stojanovski et al. 2020; .uri.i. 2021). Stable isotope analyses have shown almost exclusive reliance on terrestrial resources of the Star.evo culture communities with mixed terres­trial and aquatic diet in the Danube Gorges and in certain settlements located in close proximity to ri­vers – Vin.a – Belo Brdo, Sremski Karlovci, Obre. – Ba.tine, and Klisa (Jovanovi. et al. 2019; 2021). Do­mestic animals were not only used as a meat and fat source, as recent lipid analyses undertaken on Star.evo culture pottery confirmed milk and dairy consumption during the Early/Middle Neolithic (Et­hier et al. 2017; Cramp et al. 2019; Stojanovski et al. 2020). Archaeobotanical analyses from Vin.a culture sites have shown a greater variety of domestic crops. Ar-chaeobotanical remains from the sites of Drenovac, Gomolava, Opovo, Selevac and Vin.a – Belo Brdo, have confirmed cultivation of 11 crops (for distri­bution see Filipovi., Obradovi. 2013. 41): einkorn (Triticum monococcum), emmer (Triticum dicoc-cum), hulled barley (Hordeum vulgare, hulled), na­ked barley (Hordeum vulgare, nudum), free-thresh­ing wheat (Triticum aestivum/durum), a new type of wheat (Triticum sp., ‘new type’), lentil (Lens culi­naris), pea (Pisum sativum), bitter vetch (Vicia er-vilia), flax/linseed (Linum usitatissimum), grass pea (Lathyrus sativus/cicera) (Filipovi., Obradovi. 2013; Filipovi. 2014). A total of 17 wild taxa have been documented so far (for distribution see Filipo­vi., Obradovi. 2013.41): cornelian cherry (Cornus mas), hazel (Corylus avellana), wild strawberry (Fragaria vesca), crab apple (Malus sylvestris), com­mon reed (Phragmites communis), Chinese lantern (Physalis alkekengi), fruit from prunus genus (Pru-nus sp.), fruit from pear genus (Pyrus sp.), acorn (Quercus sp.), beech nut (Fagus sp.), blackberry (Rubus fruticosus), dewberry (Rubus caesius), un­identified berries (Rubus sp.) elderberry (Sambucus nigra), dwarf elder (Sambucus ebulus), water chest­nut (Trapa natans), wild grape (Vitis vinifera (ssp. sylvestris)) (Filipovi., Tasi. 2012; Filipovi., Obra­dovi. 2013) and sloe berry (Prunus spinosa) (Fili­povi., Tasi. 2012; Borojevi. et al. 2020). Archaeozoological analyses have confirmed that do­mestic animals (cattle, sheep/goats, and pigs) play­ed a dominant role in the subsistence strategies of Vin.a culture communities. While the most signifi­cant animal was still cattle, the most dramatic shift occurred with pigs, who played an extremely impor­tant role in the diet of the Vin.a culture communi­ties. The results of organic residue analysis conduct­ed on two Vin.a culture sites so far – Drenovac and Motel-Slatina, confirmed the consumption of differ­ent types of meat, milk and dairy products (Kruger et al. 2019). Two different studies have confirmed the presence of beeswax in Vin.a culture pottery, confirming the exploitation of honeybees during the Late Neolithic in this region (Roffet-Salque et al. 2015; Kruger et al. 2019). The majority of the wild taxa remains still belonged to red deer, roe deer, wild boar, auroch and smaller mammals (Orton 2012). At the site of Vin.a – Belo Brdo, apart from previously mentioned wild taxa, the remains of bea­ver, rabbit, otter, badger, vole, fox, wolf, brown bear, fallow deer, birds, tortoise, various Mollusca (mostly unio shells) and fish, have been identified (Dimitri­jevi. 2006). Moreover, 14.5% of the dog bones found at this site between the years 1998 and 2003 show­ed traces of burning, and one vertebra also had but­chering traces, indicating the consumption of dog meat, previously not documented in the Vin.a cul­ture (Dimitrijevi. 2006.252). Plant food can be used in different ways during dif­ferent times of the year. Cereals could have been sown in autumn, as they need a long period of ver­nalization to produce seed and legumes could have been sown in spring, due to their shorter growing What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans season (Filipovi., Tasi. 2012), securing the crop yields in case of unpredicted interannual variations. At the Late Neolithic site of Opovo, autumn/ winter sowing has been proposed for some of the identified cereals. Analysis of weed flora can help in determi­ning crop sowing time, but those analyses have ei­ther not been conducted or the results have been inconclusive (as in the case of the Vin.a – Belo Brdo site) (Filipovi., Tasi. 2012). Information about dif­ferent soil types in the immediate vicinity of settle­ments has been provided for several Vin.a culture sites. The majority of settlements in the Morava Val­ley are surrounded by multiple soil types, some more suitable for autumn/winter, and others for spring-sown crops (Milanovi. 2019). Cereals could have been used in several different ways – they could have been used for bread production or cooked as porridges or gruels. Porridges could have been en­riched with legumes, and legumes could have been added to stews and soups. Acorn is also very versa­tile, and it could have been consumed roasted, dried or ground into flour. Fruit and berries could have been eaten fresh, but also dried or cooked, which prolonged their shelf life (Atalay, Hastorf 2006; .u­ri.i. 2021). The consumption of milk and dairy pro­ducts and their combination with other food catego­ries further increases the variety of food sources in the Neolithic period. During the Early/Middle Neolithic in the Central Bal­kans, diversification as a social buffering strategy was practised through different procurement stra­tegies – agriculture, domestic animal herding, hunt­ing, fishing and gathering. A broad variety of domes­ticated and wild plant and animal resources were used, which enabled Star.evo culture communities to obtain a year-round balanced diet and reduce the potential for food shortages. One definite model for diversification cannot be provided, however, as food procurement strategies could differ from settlement to settlement (Greenfield 2008; Orton 2012; Filipo­vi., Obradovi. 2013; Cramp et al. 2019; .uri.i. 2021). Each settlement should be approached indi­vidually, providing opportunities for further studies on differences and similarities between communi­ties, which surpasses the scope of this paper. Never­theless, certain tendencies can be distinguished, such as higher dependence on domesticates in the human diet at the majority of the sites, both of plant and animal origin (Orton 2012; Filipovi., Obrado­vi. 2013; .uri.i. 2021). Archaeozoological analy­ses have helped in determining the seasonality of occupation of certain settlements. By analysing the seasonal availability of the represented taxa, occu­pation patterns have been determined for several settlements. Archaeozoological analyses of Early Neo­lithic sites in the Danube Gorges show patterns of seasonal habitation, based on the faunal, specifically migratory fish remains (Dimitrijevi. et al. 2016). At the site of Golokut, analyses of faunal remains have indicated that the settlement was occupied sea­sonally between late autumn to late winter (.ivalje­vi. et al. 2017; .uri.i. 2021). Similarly, archaeozo­ological analyses from the site of Blagotin suggest seasonal occupation between late autumn and late spring (Ethier et al. 2017; .uri.i. 2021). Vin.a culture communities used even broader spec­tra of domesticated and wild plants and animals (Di­mitrijevi. 2006; Orton 2012; Filipovi., Obradovi. 2013). The introduction of new types of domesticat­ed cereals (Hordeum vulgare, nudum, Triticum ae-stivum/durum and Triticum sp., ‘new type’) and legumes (Vicia ervilia, Linum usitatissimum and Lathyrus sativus/cicera) could further reduce the possibility of crop failure (Filipovi., Obradovi. 2013; Filipovi. 2014). With regard to domesticated animals, while pigs had negligible significance dur­ing the Early/Middle Neolithic, they were important in the diet of the Late Neolithic communities in the Central Balkans (Orton 2012). Furthermore, the ex­treme diversity of the faunal repertoire at the site of Vin.a – Belo Brdo, including previously not register­ed food sources, specifically in the excavated area of later occupational horizons at this site, raises a que­stion about living conditions at this particular time. It is still not clear whether this broader spectrum of food sources reflects a period of crisis or if this in­clusion of more diverse animal species represents the typical diet of Vin.a settlement residents. After these occupational horizons, the habitation of the settlement continued, but archaeozoological analy­ses of faunal remains of the final occupational phase show typical Vin.a culture dietary choices (except a smaller representation of pigs) (Dimitrijevi. 2006). It is possible that this variety of food sources repre­sents evidence of interannual or even long-term va­riations leading to food scarcity and the application of diversification as a survival strategy, by the intro­duction of an even broader spectrum of animals. Ne­vertheless, one universal diversification pattern can­not be applied to every Vin.a culture settlement. In some settlements, for example, hunting played a more significant role than in others (Orton 2012), indicating differences in procurement strategies throughout the Vin.a culture. Comparison between procurement strategies during the Late Neolithic of the Central Balkans can provide important informa­ Ana —Duri;ic´ tion regarding the Vin.a culture economy, but these analyses surpass the scope of this paper. Evidence for storage in the Neolithic of the Cen­tral Balkans Storage is a common practice, not only among farm­ing communities but also among hunter-gatherers. Even though cereals are the most common type of stored food, fruit, meat and fish can also be stored (Madge 1994; .uri.i. 2021). For seasonal and some­times interannual variations storage is an extremely effective buffering strategy, but it would not be ef­fective if long-term variations affecting the crop yield occur (O’Shea 1989; Halstead 1989; .uri.i. 2021). In those situations, meat, fish or wild plants can still be stored. In the archaeological record, storing features (silos, built-in storage containers, such as storage bins, sto­rage pots or pithoi) or concentrations of carbonized plant remains are good indicators of storage. Other types of food, although stored, are less likely to be preserved in the archaeological record. The proper­ties of the manufacturing material determine the preservation of certain objects or features in the ar­chaeological record, so storage containers from pe­rishable materials – wooden crates, boxes, baskets, sacks or bags made from leather, wood, plant fib-res, branches and cork – would be hard or impossi­ble to identify in the archaeological record (Filipo­vi. et al. 2018.34; .uri.i. 2021). Storage containers in the Star.evo culture are rare­ly found, so this practice is usually determined via concentrations of carbonized grains. At the site of Nosa – Biserna Obala, more than fifty clay-lined pits, resembling silos (with a pear-shaped cross-section), were found. Inside those pits, small amounts of car­bonized grains were detected, together with other fragmented material (Gara.anin 1960.229). The fill of the pits corresponds to the fill of the refuse pits, including carbonized remains, which were charred before they were thrown away. If they were silos then those remains either would not have been car­bonized or would not have been mixed with other material categories. Still, there is a possibility that those pits were primarily silos, but were secondarily used as refuse pits (Filipovi., Obradovi. 2013; Fi­lipovi. et al. 2018). Pithoi, large pots that would in­dicate storage, are scarce at the Star.evo culture si­tes. A large vessel containing barley was found at the site of Bandovi.i (Filipovi. et al. 2018), making it the only reported case so far (.uri.i. 2021). Grains and other plant-based products could have been stored inside smaller vessels, but their volume does not seem sufficient for extensive storage (Tripkovi. 2011). At the site of Drenovac, in the house destruc­tion layer of burnt daub, a concentration of carbo­nized seeds was found (Filipovi. et al. 2018.35; .u­ri.i. 2021). It contained predominantly legumes (75% lentil, 20% peas, small amounts of emmer and einkorn, and several fruits remains). This concentra­tion suggests that these different plant foodstuffs were probably stored separately in organic contai­ners – baskets, bags, or wooden containers with se­veral compartments (Filipovi. et al. 2018.35; .u­ri.i. 2021). At the Early Neolithic site of Tsangli in Greece, indications of baskets lined with clay plaster or dung were documented, showing another possible storing option, which would hardly leave any archa­eological record (Halstead 1989.71). At the sites of the Late Neolithic Vin.a culture, sub­stantially more evidence for storage has been found. Storage containers were predominantly found in­side houses and used for short-term storage. Large storing pots (pithoi) and smaller vessels were found at the majority of the Vin.a culture sites, some of which contained remains of carbonized grains, con­firming their presumed role based on the pottery typology (Jovanovi., Gli.i. 1961; Tripkovi. 2011; Vukovi. 2011; Filipovi. et al. 2018). Pithoi could be found inside separate architectural features – stor­age bins, made from mud plaster, usually located next to an oven (Fig. 2). The walls of those features were about 50cm high. Storage compartments, toge­ther with ovens and querns located in their proximi­ty, form a food processing set (Spasi., .ivanovi. 2015). Storage bins were found in numerous houses: house 01/06 at Vin.a – Belo Brdo (Tasi. et al. 2007; Vukovi. 2011; Filipovi. et al. 2018; Borojevi. et al. 2020), the house at the depth of D 6.73m from 1912 at Vin.a – Belo Brdo (Fig. 3) (Vasi. 1912.94; .uri-.i. 2019), house II/1912 at Vin.a – Belo Brdo (Va­si. 1912.21), house 1 from the trench XIX at Dreno-vac (Peri. 2017), house 1/2010 at Stubline – Crkvi­ne (Crnobrnja 2012; Spasi., .ivanovi. 2015), house 2/79 at Banjica – Usek (Todorovi. 1981; Tripkovi. 2013; Spasi., .ivanovi. 2015), house 2 at Opovo (Tringham et al. 1992), and houses 13 and 17 at Di-vostin (Bogdanovi. 1988). One of the best exam­ples comes from house 01/06, at the Vin.a – Belo Brdo site, where two pithoi filled with carbonized (predominantly) emmer grains were found in one of three storage bins located next to an oven (Vukovi. 2011; Spasi., .ivanovi. 2015; Filipovi. et al. 2018; Borojevi. et al. 2020). Next to them, without a con­ What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans tainer, the remains of emmer, water chestnut and pears were located (Bo­rojevi. et al. 2020). Another type of storage bin has been found in Vin­.a culture houses. Although conne­cted to the house wall, they were smaller and shallower and were not located near ovens or other architec­tural features. Such containers were found in house 1/2008 at Stubline (Spasi., .ivanovi. 2015), house 21 at Grivac (Bogdanovi. 2008), house 1 at Jakovo (Jovanovi., Gli.i. 1961), house 15 at Divostin (Bogdanovi. 1988) and house 1 from the trench XIX at Drenovac (Peri. 2017). The number of these fixed storage bins is certainly higher, but the preserva­tion of these features depends on multiple factors (the most significant being house destruction by fire), as they were built from mud plaster, a material which deteriorates with time (.uri.i. 2020). In house 01/06 at Vin.a – Belo Brdo, multiple caches of cereals without containers were found, but one of them was in association with carbonized wooden planks, which may imply the existence of a wooden crate. In one emmer deposit, relatively large quanti­ties of flax/linseed and bitter vetch were document­ed, probably stored separately in perishable contai­ners (wooden crates with multiple compartments, baskets or bags). Similar contexts were found at the sites of Medvednjak and Vala., suggesting the usage of containers from perishable mate­rials throughout the Vin.a culture. In house 02/06 at Vin.a – Belo Brdo, a cache of well-preserved wild pears was discovered. Those pears could have been stored in bags hung on the house wall, or in pots placed on shelves (Filipovi. et al. 2018). Pears and berries could have been dried and stored for winter consumption. Herbs, certain grass types and weeds could have been dried, stored and used as medicines or spices (Filipo­vi., Tasi. 2012). Storage practices in the Vin.a culture are an important and interesting topic. By following changes in these, broader sets of questions about diachronic changes in social organization, architectural practices and the economy of the Vin.a culture communities can be answered (see Tripkovi. 2013), but detailed analyses of these is­sues surpass the scope of this paper. Raw cereals are prone to spoilage and due to their exposure to insects or rodents, they cannot be stor­ed for a long period of time. Bulgur (cooked and dried cereals) or trahanas (dried fermented cereals) are good options for prolonging their shelf life. Bul­gur is prepared by first cooking, then drying and fi­nally grinding cereals. With this process, the grain gets a hard texture, less prone to spoilage and infes­tation. (Valamonti 2011; Bayram 2000; .uri.i. 2021). Indications for bulgur production were found Ana —Duri;ic´ on emmer grains detected next to a quern in house 01/06 at Vin.a (Borojevi. et al. 2020). Trahanas is similar to bulgur, but it involves cooking either bul­gur, raw grains or flour in milk or soured milk (Va-lamonti 2011; .uri.i. 2021). After cooling, the mass is formed into balls or rectangles, which are dried in the sun and stored. Trahanas can be stored for up to two or three years (Valamonti 2011). Lactic acid fer­mentation from milk also preserves food, protect­ing it from microbes and toxins, making trahanas perfect food for prolonged storage (Daglioglu 2000; .uri.i. 2021). Long-term storage is possible when a year has been particularly good, creating a considerable surplus. One part of the surplus can then be used as an in­vestment for the future – feeding cattle that will be slaughtered during times of food scarcity. This type of storage is intended for feeding the community on an interannual level (Halstead 1989.73). Evidence for exchange in the Neolithic of the Central Balkans Exchange is a broad term for social practices of shar­ing resources between communities, social groups or individuals. It does not have to be literal – goods for goods or goods for services – but also includes an exchange of goods or food between households in social settings: feasts, house hospitality and gifts/ treats (Sahlins 1965; .uri.i. 2021). These actions consolidate social bonds and obligations, forming a base for reciprocity. Food can be exchanged for la-bour or for tokens which symbolize the commitment of a household to return the favour, establishing the grounds of social storage. Social storage and these exchange practices are hard to detect in the archaeo­logical record. An exchange between households within one community can be performed only on seasonal and interannual levels. By living in the same settlement, long-term bad conditions would deprive every household of food supplies, so they would have to ask for help either from neighbours or allies that live in different territories (Halstead 1989; .uri.i. 2021). Possible ways to create allian­ces with outside communities are marriages or trad­ing partnerships. These partnerships create bonds and can prevent the emergence of hostile relation­ships (Sahlins 1972; .uri.i. 2021). Even though ex­change and contact between close settlements is always a possibility, they are difficult to detect in the archaeological record, due to probable similarities in the material culture (.uri.i. 2021). Two categories of exotic materials speak of contacts and long-dis­tance trade between the Star.evo/Vin.a culture com­munities and distant populations – Spondylus/Gly­cymeris shells and obsidian. Objects made from ja­deite/nephrite can also imply contacts with distant populations, as sources of these minerals were not documented in the territory of the Central Balkans, but due to the lack of analyses the provenience of raw materials used for the production of these arte-facts has never been determined (Balaban 2013). Even though these objects cannot provide insight in­to communication routes with distant communities at this moment, they should be taken into considera­tion for further studies, when more data is available. The presence of thermal structures in the open spa­ces within a settlement is considered as an indicator of food-sharing practices between households (Byrd 1994). Fire installations found at Star.evo culture sites were located both inside and outside dwellings. So far, only four hearths (Bogdanovi. 1988; Minich­reiter 2001; Petrovi. 1984–1985; .uri.i. 2019; 2021) and six cooking trenches (Fig. 4) (.uri.i. 2019), previously interpreted as tubular ovens (Mi-nichreiter 1992; Banffy et al. 2010), have been found in open spaces within a settlement (.uri.i. 2021). It should be noted that no cooking trenches were present inside the houses (.uri.i. 2019). So far, all of the ovens (total of 16) and the majority of hearths (total of 11) that have been found were lo­cated inside dwellings (Bogdanovi. 1988; 2008; Mi-nichreiter 1992; 2001; 2007; Banffy et al. 2010; Mari. 2013; .uri.i. 2019; 2021). The location of fire installations in the Star.evo culture settlements implies that at least a portion of cooking activities was conducted in a communal setting. To date, how­ever, it cannot be concluded if these activities were performed on a daily/seasonal basis or on special occasions, nor which part of the community was in­volved in this process (.uri.i. 2019; 2021). Intra-settlement food-sharing was proposed as one of the social buffering strategies practised by the Early Neolithic communities in Thessaly, Greece. Food-sharing activities were practised regularly, not only in times of need, resulting in the consolidation of neighbouring relations. Besides intra-settlement re­lations, they had contact with other settlements. Those contacts were documented through fine pot­tery, suggesting some degree of inter-settlement ex­change (Halstead 1989; .uri.i. 2021). Similar food-sharing practices could have been performed by the Star.evo culture communities (.uri.i. 2021). The presence of artefacts made from materials of non-local provenience suggests contacts with distant What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans populations. In the Early/Middle Neolithic of the Central Balkans, obsidian finds are scarce and most­ly concentrated in the Danube Basin. However, the research on exchange networks has not been exten­sive enough, so definite distribution patterns are yet to be determined (Tripkovi. 2003–2004). Ornaments from Spondylus/Glycymeris shells have been con­firmed at only three Star.evo culture sites with a total of 11 pieces of jewellery. They were probably imported as finished products, as manufacturing de­bris has not been detected. Additionally, ornaments typologically correspond to the examples from the rest of Europe (Vitezovi. 2012; 2016; 2019). These sporadic finds of exotic materials in the Early/Mid­dle Neolithic of the Central Balkans, confirm some degree of contact with distant communities, but not the nature of the contact (trade or maybe dowery upon marriage). Stable isotope analyses can offer more information about the relationships between different settlements and communities established by marriage partnerships. Strontium and oxygen sta­ble isotopic analyses conducted on Star.evo culture individuals from several sites in the Hungarian Transdanubia have shown that there was a higher proportion of non-local and regional females than males. If these data do not point to the migration of the Star.evo culture population to the new territory, they could imply patrilocal marriage arrangements with women from different settlements, within and outside the region (Depaermentier et al. 2020). There are few indications of food-sharing practices in the Vin.a culture. Two indicators for individual food preparation are household storage and indoor thermal structures (Byrd 1994), and both have been documented at the Vin.a culture sites (.uri.i. 2019). Contrary to ovens which were always locat­ ed inside houses, hearths (hearth-pits) were only found at two sites, Pavlovac – .ukar (.uri.i. 2019) and Pavlovac – Gumni.te (eight in total) (Peri. et al. 2016; .uri.i. 2019). These types of hearths indi­ cate either communal (Byrd 1994), seasonal (Rol­ lings 1989) or some specialized activity. Due to the scarcity of these finds, further studies should be conducted in order to determine their practical and social role. Still, even though cooking activities were probably conducted on the household level, this does not exclude other food-sharing practices like feasts, household hospitality or gifts.3 Even though the inter-settlement exchange is hard to determine and identify in the archaeological record, certain pro­ ducts could have been used in negotiations between close settlements. For example, access to pastures in the vicinity of one settlement, by herders and cattle from other settlement, could have been compensat­ ed for with products like milk, butter, cheese, meat or some other trade goods (Gillis et al. 2021). Cop­ per can also be used in determining inter-settlement and regional contacts between Vin.a culture commu­ nities and contemporary non-Vin.a culture sites in modern-day Bulgaria. Copper ores used for the pro­ duction of copper objects found at the Vin.a culture sites of Belovode, Plo.nik, Vin.a – Belo Brdo, Gomo- lava, Selevac and Gornja Tuzla and Ruse and Duran- kulak in the Lower Danube region in modern-day Bulgaria came from deposits located mainly in east­ern Serbia. The social dynamics of these inter-settlement and intercultu­ral contacts have not been determin­ed so far (Radivojevi. et al. 2021). Exotic materials are found more fre­quently at Vin.a culture sites. A num­ber of settlements, especially those located in the Tisa Valley, the South Banat and the Morava Valley, show a considerable amount of obsidian objects. The most abundant collec­tion of obsidian finds comes from the site of Vin.a – Belo Brdo, where this material makes up 69.5% of the chipped stone industry (Tripkovi. 2003–2004; Tripkovi., Mili. 2009). Analyses shows that the obsidian 3 Although there is no evidence of places for gathering and communal feasting in the Vin.a culture settlements, one interesting context found at the Late Neolithic site of Kleitos 1 in Northern Greece can offer some insight how this practice was conduct­ed (Kalogiropoulou, Ziota 2021). Ana —Duri;ic´ came from the source in the Carpathian Basin, im­plying strong and long-lasting connections with Tisza culture communities (Tripkovi., Mili. 2009). Con­nections with these communities are further docu­mented by unique finds from the site of Vin.a – Be-lo Brdo, with pottery and other ceramic objects of Tisza provenience (including the famous Mayres pi-thos) (Ignjatovi. 2008). Spondylus/Glycymeris shell ornaments were more common and present in more varying forms than in the previous period. Their finds were especially abundant at the site of Vin.a – Belo Brdo (Dimitrijevi., Tripkovi. 2002; Vitezovi., Antonovi. 2020). Spondylus/Glycymeris trading routes have not been determined yet, and several options are plausible. Even though it was common­ly assumed that these shells came from the Aegean Sea, the lack of Spondylus/Glycymeris finds along the presumed trading route via the Vardar and Mo-rava River valleys opens the door for new options. The trading network between the communities in the Central Balkans with the communities living on the coast of the Adriatic Sea was established either directly or via intermediaries in what is today cen­tral Bosnia. Spondylus/Glycymeris objects and raw material were found at the sites along this route, suggesting that Vin.a culture communities could have been a part of this exchange network (Dimit­rijevi., Tripkovi. 2006). The number of exotic finds at the site of Vin.a – Belo Brdo distinguishes it from the rest of the Vin.a culture settlements, indicating a prominent role of this community in trade and ex­change, with long-lasting and well-developed conne­ctions throughout south-eastern and central Europe (Vitezovi., Antonovi. 2020). Evidence for mobility in the Neolithic of the Central Balkans Mobility is considered the ‘easiest’ survival strategy, and it is undertaken when a community encounters a lack of resources. This strategy is typical for hunt­ing-gathering and pastoral communities. Certain ar­chaeologists consider mobility as an unfavourable option for agricultural communities, as storage (cha­racteristic of agricultural communities) and mobility are mutually exclusive (Halstead, O’Shea 1989.3–4; .uri.i. 2021). It is thought that forced mobility in agricultural communities occurs when long-term va­riations make arable land unsuitable for farming for a prolonged period of time, and thus these commu­nities would perceive mobility as a last resort (Row­ley-Conwy, Zvelebil 1989.46; .uri.i. 2021). But mobility is not rigid and not all mobile communities have the tendency to become sedentary. Equally it does not mean that sedentary communities cannot become mobile in certain situations. Even in fully se­dentary communities, there are certain social groups with a higher level of mobility (Leary, Kador 2016). Mobility is thus an extremely complex term that can­not can be defined just as relocation but also through movement (a certain period of time, season), mo­tion (mobility pattern), motivation (resources, cul­tural identity, social or economic circumstances) or segment (parts of the population) (Wendrich, Bar­nard 2008.8). Nevertheless, as a social buffering strategy practised by communities to overcome en­vironmental variability, mobility mostly refers to group or settlement relocation towards areas with available food resources (Halstead, O’Shea 1989; Groot, Lentjes 2013; .uri.i. 2021). It is considered that the level of sedentism of a cer­tain community is most accurately determined through architecture. The model for determining mobility/sedentism came from cross-cultural studies conducted on modern non-industrial communities. Those studies have suggested that mobility is imp­lied by settlements with pit-dwelling house types and sedentism by above-ground houses (Greenfield, Jongsma 2006.67; Nicholas 2002.75; .uri.i. 2021). Ethnographic studies show that pit-dwellings are a preferable choice for communities with increased mobility during the colder months (Nicholas 2002. 75). The development of agriculture is followed by the reduction of mobility. Sedentism forms a base for economic intensification, population growth and an increase in social complexity (Neil et al. 2016. 1). This approach, even though suitable for some Neolithic communities, cannot be applied universal­ly, as a number of Neolithic agricultural communi­ties display a certain degree of mobility (Neil et al. 2016; Depaermentier et al. 2020; .uri.i. 2021). Although architecture can be a good indicator of mo­bility, stable isotopic analyses offer the most accu­rate data. Stable isotopic analyses performed on the material from the Early Neolithic sites in this region have been primarily focused on dietary practices, not mobility patterns (Jovanovi. et al. 2019; 2021; Stojanovski et al. 2020), so future studies can offer more clarification on this subject. The most prominent Star.evo culture architectural features are pits, and due to the fact that above­ground structures are rare it is considered that these Early Neolithic populations lived predominantly in pit-dwellings. Although pits at Star.evo culture sites can be very large with large quantities of material, a pit cannot be determined as a dwelling purely on its What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans size. In order to be interpreted as a pit-dwelling, a pit should have thermal structure remains (Gara.a­nin 1949.52–54). Unfortunately, Early Neolithic Cen­tral Balkans sites have not been excavated exten­sively, so the data we have is still fragmentary, espe­cially when architecture is concerned. Pits that meet the criteria for being interpreted as dwellings have been identified at several sites: Divostin (Bogdano­vi. 1988), Lepenski Vir (Srejovi. 1969), Donja Bra-njevina (Karmanski 2005), Drenovac (Peri. 2008), Grivac (Bogdanovi. 2008), Zadubravlje (Minichrei­ter 1992; 2001) and Galovo (Minichreiter 2001; 2007). Their interiors are divided by platforms, ni­ches or differences in floor levels (Petrovi. 2001; .uri.i. 2021) and the presence of a superstructure is indicated by postholes or daub remains (Bogda­novi. 2008; .uri.i. 2021). The roofs were thatched, made from straw or reed (Bogdanovi. 1988.37–39; Petrovi. 2001). Architectural featu­res made from mud plaster, indicat­ing fixed house furnishing, are not present in these dwellings (.uri.i. 2021). Additionally, almost all of the ovens were underground, not built or modelled from mud plaster, and they were easy to make by digging niches (Fig. 5) or chimney-like featu­res (Fig. 6) into the sides of the pit-dwelling. The walls of these ovens were made from burnt soil, due to exposure to fire. The only examples of ovens modelled from mud plas­ter, similar to the ones found at the Late Neolithic sites, were found at the site of Lepenski Vir, but they were also located inside pit-dwel­lings (.uri.i. 2019). These types of houses and ovens, with all of their characteristics, show less investment in the living spaces, suitable for a community with increased mobility (.uri.i. 2019; 2021). Besides pit-dwellings, above-ground houses were also found at Star.evo culture sites, but in a significantly smaller number. They were found at: Divostin (Bogdanovi. 1988), Nosa – Biserna Obala (Brukner 1979), Gri-vac (Bogdanovi. 2008), Zve.ka (To­dorovi. 1966), Zadubravlje (Minich­reiter 2001), Vinkovci (Dizdar, Krz­nari. .krtvanko 2000) and Galovo (Minichreiter 2007). These houses are poorly preserved and are only detected based on the postholes, trenches and poorly preserved daub fragments (Bogdanovi. 1988.34; .uri.i. 2021). Regarding fire installations, a total of four hearths and no ovens were registered inside these houses. Postholes, trenches and daub fragments in­dicate that they were made using the wattle and daub technique and the roof was probably thatched and made from straw or reed (Bogdanovi. 1988; .uri.i. 2021). Spatial organization is unknown, and the determination of potential separate rooms has not been possible (.uri.i. 2021), which is in com­plete opposition to the later Vin.a culture house or­ganization and furnishings. Both strontium and oxygen stable isotope analyses have been conducted on several Star.evo culture sites in Transdanubia in modern-day Hungary. The Ana —Duri;ic´ majority of analysed individuals found at these sites were either mobile or of non-local origin (21 out of 37). However, the non-local individuals may have belonged to the first generation of settlers (12 out of 37) or could point to marriages with partners from different regions. Nonetheless, the number of individuals who led a mobile lifestyle should not be neglected (nine out of 37) (Depaermentier et al. 2020). Studies like these performed on the material from other Star.evo culture sites can enable the re­construction of mobility patterns of these Early Neo­lithic communities. Vin.a culture houses were rectangular in plan, with one or multiple rooms, made using the wattle and daub technique with a thatched roof (Jovanovi., Gli.i. 1961; To­dorovi. 1981; Tripkovi. 2007; 2013; Crnobrnja 2011). Each house had at least one, but usu­ally multiple domed ovens (Fig. 7), with rebuilt floors (up to six at the oven from house 01/2010 at Stubline) (Crnobrnja 2012). Sometimes an oven was inside each room (Tripkovi. 2013; Ta­si. et al. 2007), and on occasion, two ovens could have been lo­cated in the same room (Vasi. 1932) (Fig. 8). Next to ovens, clay bins and composite grinding sto­nes have have been found (To­dorovi. 1981; Bogdanovi. 1988; Tasi. et al. 2007; Tripkovi. 2013; Spasi., .ivanovi. 2015; Peri. 2017; Borojevi. et al. 2020; .u­ri.i. 2019). Symbolic elements, such as bucrania, when present, were fixed to a post next to an oven (Spasi. 2012). Some furni­shing regulations are observed inside Vin.a culture houses, es­pecially in the food processing area, where fixed architectural features (ovens, clay bins, com­posite querns and bucrania) were located. All of the fixed architec­tural features were built from mud plaster, implying time-con­suming activities with intention­al placement of furnishings (.u­ri.i. 2019). Contrary to the Star­.evo culture settlements, Vin.a culture ones are multi-layered, with houses built one on top of the other, during multiple settlement phases (Vasi. 1932; Kraiser, Voytek 1983; Peri. 2008). Settlement duration, ar­chitecture, subsistence and extensive storage, all in­dicate increased sedentism of the Vin.a culture com­munities. Nevertheless, certain social groups could have been more mobile (Leary, Kador 2016). Even some settlements show different occupation pat­terns. A good example for this is the site of Opovo. This settlement was occupied for a shorter period of time, which is implied by its stratigraphy, architec­ture, scarcity of storage features, pottery, tools and a lower percentage of domesticated plants and ani­mals (Tringham et al. 1992). Archaeozoological ana­lysis has shown that 65–70% of faunal material be- What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans longed to wild animal taxa (Orton 2012), indicat­ing that hunting was the dominant procurement strategy (Tringham et al. 1992). Furthermore, cat­tle remains represent 22.6% of the archaeozoologi-cal assemblage, while pigs were not detected at this site (Orton 2012). All of these data have led archa­eologists to propose an interpretation that the site had a short-term or seasonal occupation. Opovo could have been formed as a specialized settlement oriented toward the exchange and procurement of raw materials like venison or antler (Trinhgam et al. 1992.384). So even within fully sedentary com­munities, there are examples of partial, maybe seasonal mobility of certain groups due to some spe­cialized activity they were conducting. Recent stable carbon and oxygen isotopic analyses conducted on cattle, sheep/goats and pigs from the Vin.a – Belo Brdo and Stubline settlements have revealed new herding patterns. During warmer months domestic animals spent some time away from the settlement in different natural environments (Gillis et al. 2021). It is still unknown whether they were kept in the vicinity of the settlements or if Vin.a culture herd­ing strategies required seasonal mobility of animals and herders. Strontium stable isotopic analyses con­ducted on cattle from the Neolithic site of Arbon Bleiche 3 in Switzerland suggest three different herd­ing styles – local herding, the seasonal movement of cattle and non-local herding (Gerling et al. 2017). Similar studies could help in reconstructing the mo­bility patterns of one specialized social group within the Late Neolithic communities of the Central Bal­kans. Nevertheless, these social groups with differ­ent mobility patterns point also to diversification and exchange. With their lifestyle, they obtained and managed different resources and enabled con­tacts between settlements and communities. Discussion Each community applied one or several social buffe­ring strategies in order to overcome environmental variability, whether it is caused by seasonal, inter-annual or long-term fluctuations. Most of the pro­posed buffering mechanisms (diversification, stor­age, exchange and mobility) are complementary, but the ways in which they were implemented was com­munity specific. The difference is how dependent each community was on each of these strategies and which were applied as the last resort. By presenting the Early/Middle and Late Neolithic examples which point to each of the four social buffering strategies, we were able to examine which strategies were pre­dominantly used during these periods in the Central Balkans. Members of both Star.evo and Vin.a cul­ture communities practised agriculture and animal herding, but it seems that they used differing buf­fering strategies in order to prepare for, avoid and overcome periods of food scarcity. Early/Middle Neolithic communities of the Central Balkans practised agriculture and herding, but the evidence for storage is scarce, with caches of plant food indicating storage inside perishable contain­ers (Filipovi. et al. 2018; .uri.i. 2021). The cur­rent data does not allow the determination of the amount of surplus they produced or the extensive­ness of their storage, so it is difficult to assess how reliant they were on this buffering strategy (.uri.i. 2021). They cultivated crops with diverse growing patterns, that could have been sown during different seasons (Atalay, Hastorf 2006), which could pro­vide a balanced food supply on the seasonal level and diminish the risk of crop failure in case of un­predictable interannual variations. In addition to the cultivation of multiple crops, domestic animal herd­ing, hunting and wild plant gathering form part of their diversification strategy. By applying these mu­tually complementary procurement strategies, the community can obtain a balanced diet on the seaso­nal and interannual levels. Even though evidence of long-distance exchange is scarce, there are elements which would point to certain food sharing practices within settlements. While exchange between neigh-bouring settlements is a possibility, the available data still does not allow determination and recons­truction of inter-settlement relations (.uri.i. 2021). Nonetheless, the majority of settlements of these farming communities imply increased mobility (Krai­ser, Voytek 1983; Greenfield, Jongsma 2006.66–67; Whittle 1996.52; .uri.i. 2021). The architecture of Star.evo culture communities with pit-dwellings and above-ground houses, which a lack of thermal struc­tures and fixed architectural features, indicate less investment in house construction and furnishing (.uri.i. 2021). Recent studies offer some insight into potential reasons for this increased mobility of Star.evo culture communities (Ethier et al. 2017; Sto­janovski et al. 2020; Ivanova 2020; .uri.i. 2021). Domesticated plants and animals were brought to the Central Balkans by the new population of Early Neolithic Near Eastern migrants. These plant and animal taxa did not have wild relatives in this re­gion, so their relocation to the temperate climate of the Balkan inland may have created problems for these Mediterranean crops and animals, accustomed to warmer and dryer climate. The climate in the Bal­ Ana —Duri;ic´ kan inland was harsher, with winter frosts and a more precipitation. Acclimatization to new condi­tions was certainly a long-lasting process, so com­munities who previously primarily relied on agricul­ture had to create new subsistence patterns (Ethier et al. 2017; Stojanovski et al. 2020; Ivanova 2020; .uri.i. 2021). The arrival of the Neolithic popula­tion to the territory of the Central Balkans around 6250 cal BC coincided with the 8.2-kiloyear cold event (Bonsall 2007; Kobashi et al. 2007; Gronen-born 2009; Por.i. et al. 2021). This event lasted for roughly 200 years (approx. 6250–6050 BC) (Por-.i. et al. 2021), resulting in 2–3°C lower tempera­tures in the northern hemisphere in comparison to the previous period. In the territory of the Central Balkans this also resulted in an increase in precipi­tation, both in winter and summer months (Bonsall 2007). These conditions could have been unfavour-able for the newly arrived Mediterranean plant and animal taxa, but further studies are necessary for the determination of the effects this event had on the environment. In order to compensate for losses in agriculture due to this acclimatization process, Early/Middle Neolithic communities may have relied more on cattle meat, and dairy products (Stojanov-ski et al. 2020; .uri.i. 2021). Furthermore, goats/ sheep also did not have wild relatives in the Cen­tral Balkans, so they may have had problems ad­justing to the new climate, resulting in changes in their reproductive patterns, which may have led to the increased role of cattle (Ethier et al. 2017; Iva-nova 2020; .uri.i. 2021). In contrast to goats and sheep, pigs had wild relatives in the Central Bal­kans, and this territory is favourable for them, but their significance in the diet of the Star.evo culture population was negligible. Pigs, being less prone to transhumance, are usually a good indicator of seden­tary settlements (Ethier et al. 2017; .uri.i. 2021). Archaeozoological analyses conducted on the mate­rial from the sites of Golokut and Blagotin provided information about the seasonality of these settle­ments (.ivaljevi. et al. 2017; Ethier et al. 2017; .u­ri.i. 2021). These analyses are crucial for the deter­mination of the seasonal occupation of the settle­ments and understanding of the mobility patterns of the Star.evo culture communities. By comparing the data, it is evident that Star.evo culture commu­nities had an increased level of mobility, but the cha­racter of their mobility is still unknown. Sites with both types of dwellings are often interpreted as mul­ti-seasonal settlements (Nicholas 2002.75). Milutin Gara.anin argued that Early Neolithic mobility was cyclic (Gara.anin 1979.138), while Dragoslav Srejo­vi. noted that Star.evo culture communities inha­bited a location while the soil was fertile, abandon­ing it afterword (Srejovi. 1988.15). In contrast to Star.evo culture communities, Vin.a culture ones show less reliance on mobility as a buf­fering strategy but are more dependent on storage. They cultivated an even wider range of crops with diverse growing patterns, possibly sown during dif­ferent seasons (Filipovi., Tasi. 2012; Filipovi., Ob­radovi. 2013). They could rely more on cereals, which enabled storage as a buffering strategy. Sto­rage was well documented on the Vin.a culture sites, with numerous pithoi, food caches and clay bins found inside houses (Vukovi. 2011; Tripkovi. 2013; Spasi., .ivanovi. 2015; Filipovi. et al. 2018; Boro­jevi. et al. 2020; .uri.i. 2019). Even though Vin.a and Star.evo culture communities based their diet on the same domestic animals, cattle and pigs had a more prominent role during the Late Neolithic (Or­ton 2012). Hunting and gathering were also practis-ed as a form of a diversification strategy. Interesting­ly, a wide variety of previously undocumented ani-mas (wild animals and dogs) was consumed during later phases of the Vin.a settlement. Archaeozoolo­gical analysis conducted on the faunal material from the final occupational horizon at this site shows the typical Vin.a culture diet, indicating that the in­clusion of new animal species in previous phases could have represented application of diversification as a buffering strategy during periods of food scar­city. Nevertheless, final conclusion cannot be made before the material from other excavated areas is analysed (Dimitrijevi. 2006.252). The longevity of Vin.a culture settlements, with numerous occupatio­nal horizons and houses built one on top of the other, rectangular buildings with fixed mud plaster architectural features and regulations in house fur­nishings, attest to the sedentary lifestyle of these Late Neolithic communities (Vasi. 1932; Peri. 2008). Pigs, which were well represented in the archaeozo­ological assemblage (Orton 2012), are a good indi­cator of a sedentary lifestyle, as well (Ethier et al. 2017). Even within sedentary communities, certain social groups could have been more mobile (Leary, Kador 2016). Residents of Opovo, a settlement with short-term occupation, could have conducted some specialized activities – the procurement of certain raw materials or exchange (Tringham et al. 1992). The mobility of these social groups could have fur­ther pointed to diversification and exchange, as they performed specialized activities and came in contact with members of different communities. For seden­tary communities, storage and diversification are suitable buffering strategies, especially on the seaso­ What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans nal and interannual level. For them relocation would rent data suggests that these new plant and animal be applied when no other options were available. taxa, accustomed to a warmer and dryer climate, The evidence for food sharing within a settlement had problems acclimating to the harsher conditions is limited, but this practice should not be excluded of the Balkan inland. Star.evo culture farmers were as it is hard to confirm this in the archaeological re-prevented from being fully reliant on agriculture, re­cord. Further studies could point to contacts between sulting in their increased mobility (.uri.i. 2021). settlements and exchange networks within the re-On the other hand, Vin.a culture communities relied gion. Nevertheless, Vin.a culture communities had primarily on diversification and storage. Exchange established connections with communities of other within the settlement and between different Vin.a contemporary cultures (Tripkovi., Mili. 2009), so in culture communities was probable, but we still need times of need they could have contacted their dis-more studies to determine these social practices. tant ‘allies’. Even though this strategy could have They had established relations with communities provided some level of security, the exchange would form other cultures, so in times of need they could not be a long-time solution. have contacted ‘allies’ from distant territories. Vin­ .a culture communities lived in permanent settle-Conclusion ments, so mobility during the Late Neolithic was dif­ ferent from the Early/Middle Neolithic one. Specia-In order to avoid periods of food scarcity, Star.evo lized short-term settlements, like Opovo, could point culture communities relied primarily on diversifi-more to the mobility of certain social groups, diver-cation and mobility. Based on the current archaeo-sification or exchange than to settlement relocation logical evidence, storage seems to have been limit-as a buffering strategy. Vin.a culture communities ed. Even though exchange was practised, as we did not rely on mobility as a buffering strategy on a have indications for food sharing activities within seasonal or interannual level, but they applied it as settlements and confirmations of contacts between a last resort, when no other options were available. communities, it is hard to assess how dependent they were on this strategy. The increased mobility of ACKNOWLEDGEMENTS these farming communities could have been caused This paper is the result of work on the project ‘Hu- by long-term variations. Whether the 8.2-kiloyear mans and Society in Times of Crisis, Archaeology of event had an effect on the adaptation of the Mediter- Crisis’ funded by the Faculty of Philosophy, Univer­ ranean crops and animals (sheep/goats) in the Cen­ sity of Beograd. tral Balkan region has not been established, but cur­ . References Atalay S., Hastorf C. A. 2006. Food, meals, and daily acti­vities: food habitus at Neolithic Çatalhöyük. American Antiquity 71(2): 283–319. https://doi.org/10.2307/40035906 Balaban R. 2013. Predmeti od nefrita i/ili jadeita na neolitskim lokalitetima u Vojvodini. Rad muzeja Vojvo-dine 55: 25–33. Bánffy E., Marton T., and Osztás A. 2010. Early Neolithic settlement and burials at Alsónyék-Bátaszék. In K. Koz-³owski, P. Raczky (eds.), Neolithization of the Carpa­thian Basin: northernmost distribution of the Star.evo/ Körös culture. Polish Academy of Arts and Sciences Kra­ków and Institute of Archaeological Sciences of the Eötvös Loránd University Budapest. Kraków. Budapest: 37–51. Bayram M. 2000. Bulgur around the world. Cereal Foods World 45(2): 80–82. Bogdanovi. M. 1988. Architecture and Structural Features at Divostin. In A. McPheron, D. Srejovi. (eds.), Divostin and the Neolithic of Central Serbia. Ethnology Monogra­phs 10. University of Pittsburgh. Department of Archaeo­logy. Pittsburgh: 35–142. Bogdanovi. M. 2008. Grivac: Settlements of Proto-Star­.evo and Vin.a Culture. Center for Scientific Research of Serbian Academy of Sciences and Arts. Beograd. Bonsall C. 2007. When was the Neolithic transition in the Iron Gates. In M. Spataro, P. Biagi (eds.), A Short Walk through the Balkans: The First Farmers of the Carpa­thian Basin and Adjacent Regions. Proceedings of the Conference held at the Institute of Archaeology UCL on June 20th–22nd 2005. Quaderno-12. Societa per la preis­toria e protostoria della regione Friuli-Venezia Giulia. Tri­este: 53–68. Ana —Duri;ic´ Bori. D. 2014. Lepenski Vir geography and culture. In C. Smith (ed.), Encyclopedia of Global Archaeology. Sprin­ger. New York: 4494–4502. 2016. Deathways at Lepenski Vir: patterns in mortu­ary practice. Srpsko arheolo.ko dru.tvo. Beograd. Bori. D., Dimitrijevi. V. 2007. Apsolutna hronologija i stratigrafija Lepenskog Vira. Starinar LVII: 9–55. 10.2298/STA0757009B 2007. When did the ‘Neolithic package’ reach Lepenski Vir? Radiometric and faunal evidence. Documenta Prae­historica 34: 53–71. https://doi.org/10.4312/dp.34.5 Borojevi. K., Antonovi. D., Vukovi. J., +7 authors, and Tasi. N. N. 2020. Use of Space in a Late Neolithic/Early Eneolithic Building at the Site of Vin.a–Belo Brdo in the Central Balkans. In N. N.Tasi., D. Urem-Kotsou, and M. Bu­ri. (eds.), Making Spaces into Places. The North Aegean, the Balkans and Western Anatolia in the Neolithic. BAR Publishing. Oxford: 157–180. Brukner B. 1979. Körös kultura. In A. Benac (ed.), Prais­torija jugoslavenskih zemalja, II Neolit. Akademija na­uka i umjetnosti BiH. Centar za balkanolo.ka ispitivanja. Svjetlost. Sarajevo: 213–226. Byrd B. F. 1994. Public and private, domestic and corpo­rate: the emergence of the southwest Asian village. Ame­rican antiquity 59(4): 639–666. https://doi.org/10.2307/282338 Cauvin J. 2000. The Birth of the Gods and the Origins of Agriculture. Cambridge University Press. Cambridge. Çilingiroglu Ç. 2005. The concept of “Neolithic package”: considering its meaning and applicability. Documenta Praehistorica 32: 1–13. https://doi.org/10.4312/dp.32.1 Cramp L. J., Ethier J., Urem-Kotsou D., +7 authors, and Iva-nova M. 2019. Regional diversity in subsistence among early farmers in Southeast Europe revealed by archaeolo­gical organic residues. Proceedings of the Royal Society B 286(1894): 20182347. http://dx.doi.org/10.1098/rspb.2018.2347 Crnobrnja A. 2011. Arrangement of Vin.a culture figuri­nes: a study of social structure and organization. Docu­menta Praehistorica 38: 131–147. https://doi.org/10.4312/dp.38.11 2012. Investigations of Late Vin.a House 1/2010 at Crkvine in Stubline. Starinar LXII: 45–64. Daglioglu O. 2000. Tarhana as a traditional Turkish fer­mented cereal food. Its recipe, production and composi­tion. Food/Nahrung 44(2): 85–88. https://doi.org/10.10 02/(SICI)1521-3803(20000301)44:2<85::AID-FOOD85 >3.0.CO;2-H Depaermentier M. L., Kempf M., Bánffy E., and Alt K. W. 2020. Tracing mobility patterns through the 6th–5th mil-lennia BC in the Carpathian Basin with strontium and oxy­gen stable isotope analyses. PloS one 15(12): e0242745. https://doi.org/10.1371/journal.pone.0242745 Dimitrijevi. V. 2006. Vertebrate fauna of Vin.a-Belo Brdo: Excavation campaigns 1998–2003. Starinar 56: 245–269. https://doi.org/10.2298/STA0656245D Dimitrijevi. V., Tripkovi. B. 2002. New Spondylus find­ings at Vin.a-Belo Brdo: 1998–2001 campaigns and regio­nal approach to problem. Starinar 52: 47–62. https://doi.org/10.2298/STA0252047D 2006. Spondylus and Glycymeris bracelets: trade reflec­tions at Neolithic Vin.a-Belo Brdo. Documenta Praehi­storica 33: 237–252. https://doi.org/10.4312/dp.33.21 Dimitrijevi. V., .ivaljevi. I. and Stefanovi. S. 2016. Be­coming sedentary? The seasonality of food resource ex­ploitation in the Mesolithic-Neolithic Danube gorges. Do-cumenta Praehistorica 43: 103–122. https://doi.org/10.4312/dp.43.4 Dizdar M., Krznari. .krivanko M. 1999. Prilog poznavanju arhitekture star.eva.ke kulture u Vinkovcima. Vjesnik ar­heolo.kog muzeja u Zagrebu 32(1): 7–22. .uri.i. A. 2019. Ognji.ta i pe.i u neolitu centralnog Bal­kana – tehnike izrade i upotreba. Unpublished PhD the­sis. Faculty of Philosophy. University of Belgrade. Beo-grad. https://nardus.mpn.gov.rs/handle/123456789/122 28?locale-attribute=sr_RS 2020. Oven daub preservation in the archaeological re­cord – Vin.a culture sites. In S. Vitezovi., K. .ari., and D. Antonovi. (eds.), Current interdisciplinary studies in technology in the archaeology of the South-East Eu­rope. Srpsko arheolo.ko dru.tvo. Beograd: 138–144. 2021. Facing the Environmental Variability in the Early Neolithic of the Central Balkans: Diversification, Sto­rage, Exchange, and Mobility. In S. Babi. (ed.), Archaeo­logy of Crisis. Faculty of Philosophy. University of Bel­grade. Beograd: 75–88. https://nauka.f.bg.ac.rs/wp-co ntent/uploads/2022/02/Archaeology-of-Crisis.pdf Ethier J., Bánffy E., Vukovi. J., +4 authors, and Ivanova M. 2017. Earliest expansion of animal husbandry beyond the Mediterranean zone in the sixth millennium BC. Scienti­fic reports 7(1): 1–10. https://doi.org/10.1038/s41598-017-07427-x What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans Filipovi. D. 2014. Southwest Asian founder-and other crops at Neolithic sites in Serbia. Blgarsko e-Spisanje za Arheologija 4(2): 195–215. Filipovi. D., Tasi. N. N. 2012. Vin.a-Belo Brdo, a Late Neo­lithic Site in Serbia. Consideration of the Macro-Botanical Remains as Indicators of Dietary Habits. Balcanica XLIII: 7–27. https://doi.org/10.2298/BALC1243007F Filipovi. D., Obradovi. .. 2013. Archaeobotany at Neoli­thic sites in Serbia: a critical overview of the methods and results. In N. Miladinovi.-Radmilovi., S. Vitezovi. (eds.), Bioarheologija na Balkanu: bilans i perspective. Srpsko Arheolo.ko Dru.tvo. Beograd: 25–55. Filipovi. D., Obradovi. .., and Tripkovi. B. 2018. Plant storage in Neolithic southeast Europe: synthesis of the ar­chaeological and archaeobotanical evidence from Serbia. Vegetation History and Archaeobotany 27(1): 31–44. https://doi.org/10.1007/s00334-017-0638-7 Filipovi. D., Meadows J., Dal Corso M., +36 authors, and Zerl T. 2020. New AMS 14C dates track the arrival and spread of broomcorn millet cultivation and agricultural change in prehistoric Europe. Scientific reports 10(1): 1– 18. https://doi.org/10.1038/s41598-020-70495-z Gara.anin M. 1949. Naselje i stan prvobitnog .oveka u Sr-biji. Neolitsko doba u Srbiji. Istorijski glasnik 2: 38–67. Gara.anin D. 1959. Nosa-Biserna obala. Arheolo.ki pre­gled 1: 5–11. 1960. Nosa-Biserna obala, praistorijsko naselje. Stari­nar 11: 228–229. Gara.anin M. 1979. Centralnobalkanska zona. In A. Be-nac (ed.), Praistorija jugoslavenskih zemalja. II Neolit. Akademija nauka i umjetnosti BiH. Centar za balkanolo.­ka ispitivanja. Svjetlost. Sarajevo: 79–212. Gerling C., Dopple T., Heyd V., +4 authors, and Schibler J. 2017. High-resolution isotopic evidence of specialised cat­tle herding in the European Neolithic. PloS one 12(7): p.e­0180164. https://doi.org/10.1371/journal.pone.0180164 Gibbs K., Jordan P. 2016. A comparative perspective on the ‘western’and ‘eastern’ Neolithics of Eurasia: Ceramics; agriculture and sedentism. Quaternary International 419: 27–35. https://doi.org/10.1016/j.quaint.2016.01.069 Gillis R. E., Bulatovi. J., Penezi. K., Spasi. M., Tasi. N. N., and Makarewicz C. A. 2021. Of herds and societies – Sea­sonal aspects of Vin.a culture herding and land use prac­tices revealed using sequential stable isotope analysis of animal teeth. Plos one 16(10): e0258230. https://doi.org/10.1371/journal.pone.0258230 Greenfield H. 2008. Faunal assemblages from the Early Neolithic of the central Balkans: methodological issues in the reconstruction of subsistence and land use. In C. Bon­sall, V. Boroneant, and I. Radovanovi. (eds.), The Iron Gates in Prehistory: New perspectives. BAR Internatio­nal Series. Oxford: 103–114. Greenfield H., Jongsma T. 2006. The intrasettlement spa­tial structure of Early Neolithic settlements in temperate Southeastern Europe: a view from Blagotin, Serbia. In E. Robertson, J. Seibert, D. Fernandez, and M. Zender (eds.), Space and Spatial Analysis in Archaeology. University of Calgary Press. Calgary: 69–82. Gronenborn D. 2009. Climate fluctuations and trajecto­ries to complexity in the Neolithic: towards a theory. Do-cumenta Praehistorica 36: 97–110. https://doi.org/10.4312/dp.36.5 Groot M., Lentjes D. 2013. Studying subsistence and sur­plus production. In M. Groot, D. Lentjes, and J. Zeiler (eds.), Barely Surviving or More than Enough? The en­vironmental archaeology of subsistence, specialisation and surplus food production. Sidestone Press. Leiden: 7– 28. Halstead P. 1989. The economy has a normal surplus: eco­nomic stability and social change among early farming communities in Thessaly, Greece. In P. Halstead, J. O’Shea (eds.), Bad year economics: cultural responses to risk and uncertainty. Cambridge University Press. Cambridge: 68–80. Halstead P., O’Shea J. 1989. Introduction: cultural respon­ses to risk and uncertainty. In P. Halstead, J. O’Shea (eds.), Bad year economics: cultural responses to risk and un­certainty. Cambridge University Press. Cambridge: 1–7. Ignjatovi. M. 2008. Katalog. In D. Nikoli. (ed.), Vin.a pra­istorijska metropola. Filozofski fakultet u Beogradu. Na-rodni muzej u Beogradu. Muzej grada Beograda. Srpska akademija nauka i umetnosti. Beograd: 203–280. Ivanova M. 2020. Farmers at the Frontier: A Pan Euro­pean Perspective on Neolithization. In K. J. Gron, L. So­rensen, and P. Rowley-Conwy (eds.), Growing societies: an ecological perspective on the spread of crop cultiva­tion and animal herding in Europe. Oxbow. Oxford: 7–44. Jovanovi. B., Gli.i. J. 1961. Eneolitsko naselje na Korma­dinu kod Jakova. Starinar XI: 113–139. Jovanovi. J., de Becdelievre C., Stefanovi. S., .ivaljevi. I., Dimitrijevi. V., and Goude G. 2019. Last hunters–first far­mers: new insight into subsistence strategies in the Cen­tral Balkans through multi-isotopic analysis. Archaeologi­ Ana —Duri;ic´ cal and Anthropological Sciences 11(7): 3279–3298. https://doi.org/10.1007/s12520-018-0744-1 Jovanovi. J., Blagojevi. T., Markovi. J., +7 authors, and Skelac G. 2021. New radiocarbon dates, stable isotope, and anthropological analysis of prehistoric human bones from the Balkans and Southwestern Carpathian Basin. Documenta Praehistorica 48: 224–251. https://doi.org/10.4312/dp.48.18 Kaiser T., Voytek B. 1983. Sedentism and economic change in the Balkan Neolithic. Journal of Anthropological Ar­chaeology 2(4): 323–353. https://doi.org/10.1016/0278-4165(83)90013-2 Kalogiropoulou E., Ziota C. 2021. Exploring commensali­ty, household and solidarity. Evidence of a medium-scale community gathering place in Neolithic Kleitos 1, north­western Greece. Journal of Material Culture 26(4): 403– 432. https://doi.org/10.1177/13591835211042496 Karamanski S. 2005. Donja Branjevina. A Neolithic set­tlement near Deronje in the Vojvodina (Serbia). Quader-no 10. Societa per la preistoria e protoistoria della regio­ne Friuli-Venezia Giulia. Trieste. Kobashi T., Severinghaus J. P., Brook E. J., Barnola J. M., and Grachev A. M. 2007. Precise timing and characteri­zation of abrupt climate change 8200 years ago from air trapped in polar ice. Quaternary Science Reviews 26(9– 10): 1212–1222. https://doi.org/10.1016/j.quascirev.2007.01.009 Kruger M., Baj.ev O., Whelton H., and Evershed R. 2019. Organic residue analysis and the use of pottery from the Neolithic settlements of Drenovac and Motel-Slatina (Mid­dle Morava Valley, Serbia). In S. Peri. (ed.), The Neolithic in the Middle Morava Valley 3. Institute of Archaeology, Regional Museum Para.in. Beograd, Para.in: 61–76. Kuijt I., Goring-Morris N. 2002. Foraging, farming, and so­cial complexity in the Pre-Pottery Neolithic of the south­ern Levant: a review and synthesis. Journal of World Pre­history 16(4): 361–440. https://doi.org/10.1023/A:1022973114090 Leary J., Kador T. 2016. Moving on in Neolithic studies: Understanding mobile lives. Oxbow Books. Oxford. Madge C. 1994. Collected food and domestic knowledge in The Gambia, West Africa. Geographical Journal 160(3): 280–294. Mari. M. 2013. Za.titna arheolo.ka istra.ivanja na lokali­tetu Jari.i.te 1. In V. Filipovi., R. Arsi., and D. Antonovi. (eds.), Rezultati novih arheolo.kih istra.ivanja u seve­rozapadnoj Srbiji i susednim teritorijama. Srpsko arheo­lo.ko dru.tvo i Zavod za za.titu spomenika kulture Valje­vo. Beograd, Valjevo: 17–31. Milanovi. D. 2019. Insight into the regional distribution and geographic setting of the Vin.a and Bubanj-Sa¢lcuta-Krivodol settlements in the Central Balkans and its impli­cations. Starinar(69): 61–84. https://doi.org/10.2298/STA1969061M Minichreiter K. 1992. Pe.i u star.eva.kom naselju kod Za­dubravlja. Opvscula Archaeologica 16: 37–47. 2001. The architecture of Early and Middle Neolithic settlements of the Star.evo culture in Northern Croa­tia. Documenta Praehistorica 28: 199–214. https://doi.org/10.4312/dp.28.11 2007. Slavonski Brod Galovo. Institut za arheologiju. Zagreb. Neil S., Evans J., Montgomery J., and Scarre C. 2016. Iso­topic evidence for residential mobility of farming com­munities during the transition to agriculture in Britain. Royal Society open science 3(1): 1–14. https://doi.org/10.1098/rsos.150522 Nicholas D. 2002. Basketmaker III: Early Ceramic-Period Villages in the Kayenta Region. In S. Powel, F. Smiley (eds.), Prehistoric Culture Change on the Colorado Pla­teau: Ten Thousand Years on Black Mesa. University of Arizona Press. Tucson: 66–77. Orton D. 2012. Herding, settlement, and chronology in the Balkan Neolithic. European Journal of Archaeology 15(1): 5–40. https://doi.org/10.1179/1461957112Y.0000000003 O’Shea J. 1989. The role of wild resources in small-scale agricultural systems: tales from the Lakes and the Plains. In P. Halstead, J. O’Shea (eds.), Bad year economics: cul­tural responses to risk and uncertainty. Cambridge Uni­versity Press. Cambridge: 57–67. Peri. S. 2008. The oldest cultural horizon of trench XV at Drenovac. Starinar 58: 29–50. https://doi.org/10.2298/STA0858029P 2017. Drenovac: a Neolithic settlement in the Middle Morava Valley, Serbia. Antiquity 91(357): e4. https://doi.org/10.15184/aqy.2017.41 Peri. S., Baj.ev O., Obradovi. .., and Stojanovi. I. 2016. Neolitsko naselje Pavlovac – Gumni.te: Rezultati za.titnih arheolo.kih iskopavanja 2011. godine. In S. Peri., A. Bu­latovi. (eds.), Za.titna arheolo.ka istra.ivanja na auto-putu E75, 2011–2014. godine. Arheolo.ki institut. Beo-grad: 383–395. What we do for food – social strategies for overcoming food scarcity in the Neolithic of the Central Balkans Petrovi. B. 2001. Model neolitske pe.i iz Progara. Godi.­njak grada Beograda XLVII–XLVIII: 11–20. Petrovi. J. 1984–1985. Naselje star.eva.ke kulture na Go-lokutu kod Vizi.a. Rad vojvodjanskih muzeja 29: 9–25. Por.i. M. 2020. Observations on the origin and demogra­phy of the Vin.a culture. Quaternary International 560– 561: 57–64. https://doi.org/10.1016/j.quaint.2020.04.012 Por.i. M., Blagojevi. T., Pendi. J., and Stefanovi. S. 2021. The Neolithic Demographic Transition in the Central Bal­kans: population dynamics reconstruction based on new radiocarbon evidence. Philosophical Transactions of the Royal Society B 376(1816): 20190712. https://doi.org/10.1098/rstb.2019.0712 Radivojevi. M., Rehren T., and Pernicka E. 2021. Metallur­gical knowledge and networks of supply in the 5th millen­nium BC Balkans: Belovode and Plo.nik in their regional context In. M. Radivojevi., B. Roberts, M. Mari., J. Kuz­manovi. Cvetkovi., and T. Rehren (eds.), The Rise of Me­tallurgy in Eurasia-Evolution, Organisation and Con­sumption of Early Metal in the Balkans. Archaeopress. Oxford: 484–527. Roffet-Salque M., Regert M., Evershed R. P., +61 authors, and Pääkkönen M. 2015. Widespread exploitation of the honeybee by early Neolithic farmers. Nature 527(7577): 226–230. https://doi.org/10.1038/nature15757 Rollings W. H. 1989. The Conamche. Chelsea House. New York. Rowley-Conwy P., Zvelebil M. 1989. Saving it for later: storage by prehistoric hunter-gatherers in Europe. In P. Halstead, J. O’Shea (eds.), Bad year economics: cultural responses to risk and uncertainty. Cambridge University Press. Cambridge: 40–56. Sahlins M. 1965. On the sociology of primitive exchange. In M. Banton (ed.), The relevance of models for social anthropology. Rutledge. London: 139–236. 1972. Stone Age Economics. Routledge. New York. Spasi. M. 2012. Cattle to settle – bull to rule: on bovine iconography among Late Neolithic Vin.a culture commu­nities. Documenta Praehistorica 39: 295–308. https://doi.org/10.4312/dp.39.22 Spasi. M., .ivanovi. S. 2015. Foodways architecture: sto­ring, processing and dining structures at the Late Neolithic Vin.a culture site at Stubline. Documenta Praehistorica 42: 219–230. https://doi.org/10.4312/dp.42.15 Srejovi. D. 1969. Lepenski vir: nova praistorijska kul­tura u Podunavlju. Srpska knji.evna zadruga. Beograd. Stojanovski D., .ivaljevi. I., Dimitrijevi. V., +15 authors, and Stefanovi. S. 2020. Living off the land: Terrestrial-based diet and dairying in the farming communities of the Neolithic Balkans. PloS one 15(8): e0237608. https://doi.org/10.1371/journal.pone.0237608 Tasi. N. N., .uri.i. S., Lazarevi. B. 2007. Analiza konstruk­cije .rvnja iz ku.e 01/06 u Vin.i. Glasnik Srpskog arheo­lo.kog dru.tva 23: 211–218. Todorovi. J. 1966. Lug, Zve.ka, Obrenovac – naselje star­.eva.ke culture. Arheolo.ki pregled 8: 10–12. 1981. A recently discovered house in the late neolithic settlement of Banjica in Belgrade. Archaeologia Iugo­slavica 18: 13–16. Tringham R., Brukner B., Kaiser T., +5 authors, and Voy­tek B. 1992. Excavations at Opovo, 1985–1987: socioeco­nomic change in the balkan neolithic. Journal of Field Ar­chaeology 19(3): 351–386. Tripkovi. B. 2003. Obsidian deposits in the central Bal­kans? Tested against archaeological evidence. Starinar 53–54: 163–179. https://doi.org/10.2298/STA0454163T 2007. Doma.instvo i prostor u kasnom neolitu. Srps­ko arheolo.ko dru.tvo. Beograd. 2011. Containers and grains: food storage and symbo­lism in the Central Balkans (Vin.a period). Documenta Praehistorica 38: 159–172. https://doi.org/10.4312/dp.38.13 2013. Doma.instvo i zajednica: ku.ne i naseobinske istorije u kasnom neolitu centralnog Balkana. Filozof-ski fakultet. Univerzitet u Beogradu. Beograd. Tripkovi. B., Mili. M. 2008. The origin and exchange of obsidian from Vin.a-Belo Brdo. Starinar 58: 71–86. https://doi.org/10.2298/STA0858071T Valamoti S. M. 2011. Ground cereal food preparations from Greece: the prehistory and modern survival of tradi­tional Mediterranean ‘fast foods’. Archaeological and An­thropological Sciences 3(1): 19–39. https://doi.org/10.1007/s12520-011-0058-z Vasi. M. 1912. Dnevnik sa iskopavanja u Vin.i. Narodni muzej. Beograd. 1932. Preistorijska Vin.a I. Dr.avna .tamparija. Beo-grad. Vitezovi. S. 2012. The white beauty–Star.evo culture je­wellery. Documenta Praehistorica 39: 215–226. https://doi.org/10.4312/dp.39.15 Ana —Duri;ic´ 2016. Neolithisation of technology: innovation and tra­dition in the Star.evo culture osseous industry. Docu­menta Praehistorica 43: 123–138. https://doi.org/10.4312/dp.43.5 2019. Osseous technology in the early and middle Neo­lithic in the Central Balkans. Cuadernos de Prehistoria y Arqueología de la Universidad de Granada 29: 277– 292. https://doi.org/10.30827/cpag.v29i0.9778 Vitezovi. S., Antonovi. D. 2020. Functional differentia­tion and possible regional specialisation of the Vin.a cul­ture settlements: Viewpoint from osseous and lithic indu­stries. Quaternary International 539: 39–48. back to contents https://doi.org/10.1016/j.quaint.2018.12.029 Vukovi. J. 2011. Tehnike oblikovanja kasnoneolitske grn-.arije: Pitos iz ku.e 01/06 u Vin.i. Zbornik Narodnog muzeja XX(1): 71–82. Wendrich W., Bernard H. 2008. The Archaeology of Mobi­lity: Definitions and Research Approaches. In H. Bernard, W. Wendrich (eds.), The Archaeology of Mobility-Old World and New World Nomadism. Cotsen Institute of Ar­chaeology. University of California. Los Angeles: 1–24. .ivaljevi. I., Dimitrijevi. V., Radmanovi. D., +4 authors, and Stefanovi. S. 2017. Lov, sto.arstvo i simboli.ki zna-.aj .ivotinja na Golokutu: nove analize arheozoolo.kog materijala. Arhailka 5: 1–26. Documenta Praehistorica XLIX (2022) Late Bronze Age food storage in Lower Cerova;ka Cave, Croatia> the archaeobotanical evidence Kelly Reed1, Petra Radakovic´ 2, Sara Essert3, and Dinko Tresic´ Pavi;ic´ 4 kellyreed@hotmail.co.uk 1 School of Architecture, Oxford, UK 2 Primary School Ludbreg, Ludbreg, HR< petra.radakovic1@skole.hr 3 University of Zagreb, Faculty of Science, Department of Biology, Division of Botany, Zagreb, HR< sara.essert@biol.pmf.hr 4 Kaducej d.o.o., Split, HR< dtresic@gmail.com ABSTRACT – This paper presents new archaeobotanical data from the Lower Cerova.ka Cave locat­ed in Dalmatia, Croatia. At the site a high density of carbonized plant remains was recovered, indi­cating the remnants of a burnt crop store dating to the Late Bronze Age. Overall, the assemblage is dominated by lentil (Lens culinaris) and free-threshing wheat (Triticum aestivum/durum), and to a les­ser extent, emmer (Triticum dicoccum), einkorn (Triticum monococcum), spelt (Triticum spelta) and broomcorn millet (Panicum miliaceum). In general, the large botanical collection from Lower Cero­va.ka Cave fits with what is already known about Bronze Age agriculture in Croatia, yet the unique nature of this site brings to the fore questions around storage practices and the use of caves in pre­history. KEY WORDS – crop processing; carbonized chaff and grain; south-east Europe; cave storage Pozno bronastodobna shramba hrane v Spodnji Cerova;ki jami, Hrva[ka> arheobotani;ni dokazi IZVLE.EK – V .lanku predstavljamo nove arheobotani.ne podatke iz Spodnje Cerova.ke jame v Dal-maciji na Hrva.kem. V jami so odkrili mnogo karboniziranih rastlinskih ostankov, ki ka.ejo na ostanke po.gane shrambe pridelkov iz pozne bronaste dobe. Med njimi prevladujejo le.a (Lens culi­naris) in neplevasta .ita (Triticum aestivum/durum), manj pa je dvozrnice (Triticum dicoccum), eno­zrnice (Triticum monococcum), pire (Triticum spelta) in prosa (Panicum miliaceum). Na splo.no se ve­lik botani.ni zbir iz Spodnje Cerova.ke jame ujema s tem, kar .e vemo o bronastodobnem kmetijst­vu na Hrva.kem, vendar izjemno mesto najdbe postavlja v ospredje vpra.anja o praksah shranjeva­nja in uporabi jam v prazgodovini. KLJU.NE BESEDE – predelava pridelkov; karbonizirano pleve in zrnje; jugovzhodna Evropa; shram­ba v jami Introduction The seasonality of the agricultural cycle means that tary overwintering strategies (e.g., Halstead, O’Shea some degree of storage is inevitable. In prehistory, 1989). In addition, large-scale or centralized stor-seasonal and intensive storage of major food re-age has been seen as an indication of social comple-sources for the short, medium or long-term would xity, surplus production, and redistribution, as well have been directly related to coping with seasonal as emphasizing socio-economic inequality (e.g., Bo-variability in agricultural productivity and seden-gaard et al. 2019; Forbes, Foxhall 1995). Surplus DOI> 10.4312\dp.49.22 Late Bronze Age food storage in Lower Cerova;ka Cave, Croatia> the archaeobotanical evidence also links to networks and trade, whereby an indi­vidual or group does not have to store everything themselves but can count on others to provide food at certain times (Angourakis et al. 2015; Hastorf, Foxhall 2017; Winterhalder et al. 2015). Subse­quently, storage has been conceptualized in three different ways (Ingold 1983; Soffer 1989): (1) as in-tra-corporeal, where body fat helps survival through lean times; (2) social storage where formalized ex­change systems and social obligations can be recon­verted into food in times of shortage; (3) and mate­rial or practical storage that involves the processing and accumulation of food resources, and the con­struction of immovable storage features such as sto­rehouses and pits that encourage permanent resi­dence. Recognizing different modes of food storage in prehistory is therefore critical to assessing the roles that the environment, mobility, settlement size, and socioeconomic circumstances play in the development of different storage behaviours. Interpreting the economic and/or social motives for storage facilities in prehistory is challenging. Was their use temporary, seasonal, or long-term? Were they managed by households, networks of extend­ed kin, entire communities, or aspiring or established elites? Were they securing food resources, and/or other goods? In order to help with this interpretation scholars typically look at food storage and prepara­tion facilities, as well as primary deposits of ecofacts, such as in-situ food storage (e.g., Bogaard et al. 2009; Sadori et al. 2006). Here we present unique archaeo-botanical evidence of Late Bronze Age crop storage within Lower Cerova.ka Cave, located in Dalmatia, Croatia. We will examine the use of caves as storage contexts and how this site can expand our under­standing of Bronze Age communities in Dalmatia. The site The Cerova.ke caves are located on the south-east­ern part of Mt. Velebit (Fig. 1), on the steep north­ern slopes of the massif of Crnopac, on the edge of Gra.ac field (elevation 550m). They are represented by three sub-horizontal cave channels, namely the Lower (Donja), Middle (Srednja) and Upper (Gornja) Cerova.ka Cave. The Lower Cave was discovered in 1913 and since then the caves have been a focus for speleologists and other geoscientists (Kure.i. et al. 2021). The first exclusively archaeological excava­tions in the Lower Cave were conducted by Ru.ica Drechsler-Bi.i. in 1966 and 1967. They recovered fragments of ceramic vessels and a few metal arte-facts attributed to the Late Bronze Age (Drechsler­Bi.i. 1970; 1983; 1984). In 2019 new archaeological excavations were con­ducted in the Lower Cave in response to the con­struction of a new visitor’s path running 120m from the entrance (Fig. 2). Excavations covered an area of 173m2, i.e. 120m in length and from 0.5–6m in width. Six phases were identified; Phase 1 was the bedrock, Phase 2 the first human occupation at the beginning of the Late Bronze Age, Phase 3 a further Late Bronze Age layer, Phase 4 the end of the Late Bronze Age, Phase 5 which dated to the Middle Ages (13th century AD), and Phase 6 which dates to the modern era. The Late Bronze Age Phases 2 to 4 were thin layers that together did not exceed 10cm. A large amount (3.5 tons) of Late Bronze Age pot­tery fragments were recovered, as well as several objects made of bronze, amber, bone, ceramics and stone, identified mainly as dress ornaments. Only a small number of animal bones and utilitarian ob­jects (needles, awls, vertebrae) were found, suggest­ing that the cave was not used regularly. Instead, it is suggested that the cave functioned mainly as a storage location during the Late Bronze Age, with periodic episodes of temporary occupation. Three radiocarbon dates (tooth, charcoal and grain) were taken from different locations within Phase 4 and all had similar dates of c. 2870–2910 BP (Tab. 1). Phase 3, dated to c. 2950 BP, and Phase 2 to c. 3090 BP. It is likely that the period of use was relatively short and ended abruptly at the same time in the whole occupied area of the cave, possibly due to fire, resulting in large areas of burnt archaeological features. This was particularly evident in quadrant D21 where a large deposit of carbonized plant ma­terial was found, along with possible remains of a woven basket or other type of receptacle, as well as Kelly Reed, Petra Radakovic´, Sara Essert, and Dinko Tresic´ Pavi;ic´ ceramic fragments (Fig. 2). In the same area three postholes were discovered, which may indicate the presence of a wooden structure, possibly linked to some sort of storage shelf or structure. Clusters of carbonized plant remains were also found in other layers, but in much smaller numbers and are largely isolated cases (Tresi. Pavi.i. 2020). Materials and methods Twenty-three samples were collected for archaeobo­tanical analysis during the 2019 excavation (Tab. 2). Four samples were taken from Phase 6, sixteen from Phase 4, two from Phase 3 and one from Phase 2. Soil samples and hand-picked archaeobotanical re­mains were collected. The samples were taken to the Division of Botany, Department of Biology, Univer­sity of Zagreb. No flotation was conducted due to the high density of plant material in the soil sam­ples, and instead the samples were dry sieved to al­low easy sorting under the microscope (Radakovi. 2021). All samples were 100% sorted, except sam­ples U-130, U-131, U-132 and U-134 (Tab. 2). Due to the high density of the remains, further subsampling was required for samples U-130, U-131, U-132, U-134 and U-135, where the >1mm fraction was fully sort­ed, but only 1/3 of the <1mm fraction was sorted, including only a 1/3 of the chaff remains (Radako­vi. 2021.15). Subsequently, the Supplementary Data contains the multiplied estimates for the plant ma-croremains identified within these samples and not the actual subsampled counts. The carbonized plant remains were sorted and iden­tified under a zoom stereo microscope at a magnifi­cation of 7–45x with the help of reference literature/ seed atlases (Cappers, Neef 2012), as well as the modern carpological collection (under establish­ment) of the Division of Botany. The nomenclature of scientific plant names follows Daniel Zohary and Maria Hopf (2000) for cultivars and the Flora Croa­tica Database (Nikoli. 2018) for wild plants. Whole grains were counted as one, and two longitudinal fragments and embryos of grains were also counted as one. Glume bases were counted as one, while whole spikelet forks were counted as two glume ba­ses. Cereal remains classed as fragments had to be at least 1/4 of the original grain/seed, and anything smaller was not counted. The fruit and weed seeds were counted as one, even when only a fragment was found, except where large seeds were broken and clearly represented the same parts of the same seed (e.g., Quercus sp.). Sample no. Laboratory number Type Phase Conventional ±30 BP Calibrated age (cal BC, 95.4% \ 2s hpd range) 7 Beta-533949 Tooth 4 2890 1133–978 140 Beta-533951 Grain 4 2870 1127–931 152 Beta-533952 Charcoal 4 2910 1209–1011 185 Beta-533953 Charcoal 3 2950 1236–1051 228 Beta-533954 Charcoal 2 3090 1427–1277 Results All 23 samples contained carbonized plant macro-remains, totalling approximately 1 179 000 items (see Supplementary Data). Lumps of broomcorn mil­ let (Panicum miliaceum) were recovered from U­201 (20ml, Phase 3) and U-137 (20ml, Phase 4) and have been estimated to contain up to 4000 grains per sample (Fig. 3a). Overall, preservation was good, especially the plant remains recovered from quad­rant D21. The bulk samples taken from Phase 4 had the highest density of remains that were dominated by lentil (Lens culinaris) and free-threshing wheat (Triticum aestivum/durum) grains, as well as em-mer (Triticum dicoccum), einkorn (Triticum mo-nococcum), spelt (Triticum spelta) grains and chaff and broomcorn millet. The other phases have gene­rally very low quantities of remains, since the plant remains were handpicked during the excavation. For Phase 2 a few acorn fragments (Quercus sp.) were picked out, while four lentil seeds were identified from Phase 3, along with a lump of broomcorn mil­let grains, approximately 20ml (a4000 grains). Phase 6 contained mostly broad beans (Vicia faba) and a few cereal grains and acorn fragments, totalling no more than 76 items. The largest quantity of plant remains were from the burnt area identified in quadrant D21 (Fig. 4). Sam­ples U-130, U-131, U-132, U-134 and U-135, in par­ticular, contained a large quantity of cereal grain and chaff (Fig. 3b), as well as pulses, but only a small proportion of wild/weed type taxa (Fig. 5a). The composition of these samples is relatively sim­ilar except for U-134, a posthole, which has a high­er proportion of broomcorn millet grains and less glume wheat chaff. The propor­tion of crops within Phase 4 is dominated by lentil and free-threshing wheat, while the re­maining crops only represent up to 5% of the assemblage (not in­cluding the cereal chaff, Fig. 5b). The diversity of wild/weed type Tab. 1. Radiocarbon dates from Lower Cerova.ka Cave. taxa identified is extremely low, Late Bronze Age food storage in Lower Cerova;ka Cave, Croatia> the archaeobotanical evidence Fig. 2. Plan of the 2019 excavation in Lower Cerova.ka Cave, focusing on area D21. Kelly Reed, Petra Radakovic´, Sara Essert, and Dinko Tresic´ Pavi;ic´ consisting of grasses, mainly Bromus arvensis and B. secalinus, and cleavers (Galium aparine and G. spurium), which can all be found as weeds in cere­al crops. Measurements were also taken of emmer, einkorn, spelt and free-threshing wheat, lentil, and broad beans recovered from quadrant D21 (Tab. 3). The size of the wheat and pulses correspond with measure­ments taken from Late Bronze Age Kalnik-Igri..e; a site located to the north­east of Lower Cerova.ka Cave in continental Croatia (Radakovi. 2021). Discussion Crop processing The high density of plant re­mains in the five samples from quadrant D21 and the clear evidence of burning in and around the deposit in­dicate that the plant remains were burnt in-situ. Unfortu­nately, the similarities in composition of the plant re­mains recovered from the different areas in D21 pre­vent any assumptions about how or where the different crops were stored. Instead, we can look at the level of crop processing that may have occurred before stor­age. Predictive models have been created to iden­tify which stage of the crop processing sequence an assemblage represents, based on the assumption that each stage produces a characteristically differ­ent ratio of cereal, chaff and weeds within the sam­ple (Hillman 1984; Jones 1984; Van der Veen 1992; Van der Veen, Jones 2006). Here we can examine the ratio of glume bases to glume wheat grains, as Fig. 3. Carbonized (a) lumps of broomcorn millet (Panicum miliaceum) from U-134 and (b) glume wheat (Triticum monococcum/dicoccum/spelta) glume bases from quadrant D21, Lower Cerova.ka Cave. Late Bronze Age food storage in Lower Cerova;ka Cave, Croatia> the archaeobotanical evidence well as the number of weeds to the number of grains to help determine what crop processing has occurred. Unfortunately, the third ratio look­ing at the number of rachis intern-odes to grains is not possible, as these were not recovered for the free-threshing wheat or barley. This could be for several reasons, includ­ing poor preservation, as experi­ments have shown a taphonomic bias against chaff, especially under oxidizing conditions and high tem­peratures (Boardman, Jones 1990). However, the most likely reason for the absence of rachis remains is that the free-threshing wheat and barley were already cleaned by the time they reached the cave. This is because both barley and free-threshing wheat grains easily detach from the chaff during the early stages of crop processing (i.e. threshing, winnowing and coarse sieving), whe­reas glume wheats require a further dehusking stage to remove the glumes from the grains (Van der Veen 1992.81). For emmer and spelt the ear generally contains two grains and two glumes, so the ratio of 2:2 = 1, while einkorn has one grain and two glumes, so the ratio of 1:2 = 0.5. If we apply this to the glume wheats in quadrant D21, we see that nearly every sample has a high to very high ratio of glumes to grain (Fig. 6, based on estimated numbers of remains). This means that there is significantly more chaff than grain in the samples. If we look at the ratio of grains to weed seeds, a ratio of 1:1 = 1, the ratio is extremely low. Of the crop processing stages, this could indicate that the glume wheat grains had been cleaned but not processed through the additional dehusking sta­ges, which would remove the broken spikelet forks (see Hillman 1984; Jones 1984; Stevens 2003). Thus, the grains ducts could be used for a range of purposes, such as a building material, for fuel, or as fodder for live­stock (Van der Veen 1999; Valamoti, Charles 2005). Cereal chaff is also used as temper in pottery, as seen at Bronze Age Monkodonia, Istria (Hellmuth Kram­berger 2017.418), as well as in Eneolithic loom weights found at a Slovenian pile-dwelling sites (To­lar et al. 2016). In Palestine, ethnographic obser­vations noted chaff was laid on top of stored grain before the underground jar-shaped receptacles were sealed with clay (Turkowski 1969.101–112). Thus, there could be several reasons to find chaff in this context. Comparing Lower Cerova.ka Cave with si­milar finds of cereal storage at two Late Bronze Age caves in southern France, glume wheat chaff is strong­ly underrepresented in relation to grains, suggesting that the glume wheats were dehusked before stor­age (Bouby et al. 2005). Multi-cropping and mono-cropping Multi-cropping, or maslins, have been used to de­scribe the growing of more than one crop in a sin- Length (mm) Width (mm) Thickness (mm) could have still been in their glu-Triticum aestivum 5.58 (4.66–6.25) 3.59 (3.11–4.08) 2.98 (2.47–3.4) mes when they reached the cave. Triticum dicoccum 6.3 (5.6–7) 3.14 (2.75–3.57) 3.03 (2.64–3.35) Once at the cave, dehusking could have occurred piece meal, as and when grain was required, and the chaff discarded onto the floor of the cave or kept aside for other purposes. Similarly, two different types of storage could have oc­curred where cleaned grains were stored in containers and the glume bases in another. Cereal by-pro- Triticum monococcum 5.97 (4.56–7.42) 2.53 (1.93–2.88) 3.06 (2.35–3.51) Triticum spelta 6.79 (5.56–7.82) 3.11 (2.45–3.69) 2.42 (2.04–2.85) Vicia faba 7.31 (5.82–10.37) 5.72 (3.94–8.24) 5.64 (4.36–7.96) 2r (mm) Lens culinaris 3.21 (2.55–3.99) Tab. 3. Measurements of free-threshing wheat (Triticum aestivum), emmer (Triticum dicoccum), einkorn (Triticum monococcum), and spelt (Triticum spelta) grains, and broad bead (Vicia faba) and len­til (Lens culinaris) identified from Quadrant D21 at Lower Cerova.­ka Cave. Kelly Reed, Petra Radakovic´, Sara Essert, and Dinko Tresic´ Pavi;ic´ b a Fig. 5. Composition of the carbonized seed assemblage for (a) each sample in quadrant D21, and (b) the proportion of cereal grains and pulses,excluding cereals chaff, in total from Phase 4, Lower Cerova.ka Cave. gle season on the same land (Halstead, Jones 1989; Jones, Halstead 1995; Petrie, Bates 2017). It is sug­gested that a mixed crop could have been more re­liable than a single-grain crop. For example, if the season was colder then rye would flourish, but if the season was hot then wheat would do better. The crops are usually those with similar maturation and/ or crop processing stages, but a recent study by Alex C. McAlvay et al. (2022) highlights other benefits of multi-cropping. This crop practice is distinct from monocropping where only one crop is grown on the same plot for one of more years. Although there are benefits to multi-cropping in terms of reducing risk of total crop failure, what type of grain crop grown would have depended on the local soil and climate, balanced with socio-economic de­mands. Where and when multi-cropping may have occurred in the past is debated, and identification in archaeological contexts can be difficult. Marijke Van der Veen (1995) compared the relative pro­portion of grain types and ana­lysed the weed assemblages in relation to growing conditions in different crops to determine that wheat and rye were probably sown together in medieval west­ern Europe. Weed ecology, such as phytosociology, autecology and FIBS (Functional Identification of Botanical Surveys), have been used to understand cropping practices in the past (e.g., Van der Veen 1992; Stevens 1996; Charles et al. 1997; Bogaard et al. 1999; Jones et al. 2010). Ethnogra­phic observations by Glynis Jones and Paul Halstead (1995) on the Greek island of Amorgos found that sown proportions of up to 80% wheat and 20% bar­ley were considered mixed intercrops by farmers, although this proportion could change depending on the environmental conditions. However, they also highlighted issues of contamination resulting from crops from previous growing cycles becoming incorporated in that season’s crop (Jones, Halstead 1995). Overall, these methods require the archaeo-botanical remains to have enough weeds to study the weed ecologies and be representative of one harvest. However, plant remains that survive in the Late Bronze Age food storage in Lower Cerova;ka Cave, Croatia> the archaeobotanical evidence archaeological record are typically either discarded waste or accidentally preserved remains, which ge­nerally result in contexts where crops from different sources are combined over tens or even hundreds of years, preventing any secure identification of crop husbandry practices (Jones, Halstead 1995; Van der Veen 2007). In prehistoric contexts some suggest that the glume wheats emmer and einkorn were grown as maslins during the Neolithic (Kreuz 2007). Einkorn and ‘new’ glume wheat are also thought to have been cultivated together as a mixed crop during the Neo­lithic and Bronze Age (Jones et al. 2000; Kohler-Schneider 2003). More recently Rebecca A. Fraser et al. (2013) examined the stable isotopes of wheat and barley from an LBK storage deposit at Vaihin-gen in Germany, and found they shared distincti­vely low d13C signatures relative to other samples, suggesting that they grew in similar conditions, pos­sibly in a similar location as a mixed crop. Whether inter-cropping was practiced by the farm­ers who used Lower Cerova.ka Cave to store their crops is hard to determine. There are very few weeds, which prevents the analyses of weed ecolo­gies. Most cereal remains are free-threshing wheat, with only a very small quantity of barley grains pre­sent, far less than 20% if we go with an 80/20 ratio outlined by Jones and Halstead (1995). Thus, the presence of barley could simply indicate contami­nation of the free-threshing wheat crop, maybe from a previous harvest, or could be remnants of a pre­viously stored crop. The glume wheats, emmer, ein­korn and spelt are found in smaller quantities, and again it is unclear from the context whether they were grown together. Instead, they could represent smaller harvests or remnants of previously stored crops. Storage location and containers The utility of each type of storage depends on perish-ability and distribution, the predictability and du­ration of lean periods, as well as the settlement pat­terns and social ethos of the society (Testart 1982). A huge range of food-keeping practices have there­fore evolved. Ethnography, historical documents, and imagery highlight a wide range of storage facili­ties, such as caves, pits, built silos, cellars, and barns, a variety of accompanying equipment used, such as bins, baskets, barrels, sacks, suspension hooks, jars or chests, as well as different preservation methods, such as drying, parboiling, fermenting, etc. (e.g., Pe­na-Chocarro et al. 2015). For cereals and legumes, controlling the humidity is the most important part of maintaining the nutrient quality and usability of the crop (Pa¢un et al. 2021). Before storage, grain must be dry (i.e. have a low moisture content) to minimize infestation by insects and microorgani­sms (bacteria, fungi, etc.), and to prevent germina­tion (Rajendran 2003). The main objective of stor­age systems is to therefore preserve food for an ex­tended period with minimal loss. The recovery of cleaned cereals and pulses in Low­er Cerova.ka Cave, along with other contextual evi­dence, suggests that the remains represent stored crops. The cave itself would have had its own mi-cro-climate, though it is uncertain what the condi­tions would have been in the Bronze Age. Today the dark well-ventilated cave, with low but fluctuating temperatures, could make a practical storage loca­tion (Tresi. Pavi.i. 2020). Yet the cave has a very high humidity (around 90%), which could cause sig­nificant spoilage of the crop, activating sprouting in the surface layer of a store, as well as encourag­ing contamination by micro-organisms, especially mould. This was noted at Baume Layrou, a Late Bronze Age cave situated in southern France, where the high humidity in the cave was thought to have caused germination in the stored grain (Bouby et al. 2005). Yet in Anatolia, caverns in tuffs have been used for food storage in the past and are still regu­larly used today for wine and to extend the shelf life of fruits and vegetables. The caves maintain a relatively constant temperature of around 13°C, with good airflow, and humidity can be as high as 80% in places, although it’s suggested that the tuff rock holds dehumidifying properties making them ideal caves for short-term food storage (Emir, Da­loglu 2012; Aydan, Ulusay 2013). Experiments also show that at low temperatures moisture changes in wheat occur relatively slowly, compared to those stored at higher temperatures (Pixton, Griffiths 1971). At Lower Cerova.ka Cave germinated grains were not identified. This may suggest that the crops were not stored for long periods within the cave, or that these specific grains had not been in the cave for enough time to allow germination before they were carbonized. Storing grain in their chaff is also sug­gested to be a way of protecting glume wheat grains and could have helped preserve the glume wheats discovered in Lower Cerova.ka Cave (e.g., Meurers-Balke, Lüning 1992). As Laurent Bouby et al. (2005) conclude, the conditions within the cave probably suggest occasional short-term storage, possibly Kelly Reed, Petra Radakovic´, Sara Essert, and Dinko Tresic´ Pavi;ic´ through periods of insecurity. They suggest that both caves in southern France are characteristic of refuge caves; being difficult to access, have hidden entrances and lack light. These characteristics are also shared with Lower Cerova.ka Cave. One of the main problems with understanding stor­age practices in prehistory lies in finding direct evi­dence of storage, especially if more perishable items such as woven baskets are used, as well as identi­fying what exactly was being stored. At Lower Ce­rova.ka Cave fragments of ceramic vessels and thin carbonized strips of vegetal material were found within quadrant D21, which could suggest the pres­ence of some sort of wicker basket (Figs. 7 and 8). The carbonized strips have not been identified yet, so it is unclear what type of plant could have been used. Braided plant fibres and basketry are rarely found in archaeological contexts, and mostly in waterlogged contexts, so it is unclear the extent to which these were used in prehistory. The discovery of post holes at Lower Cerova.ka Cave could also suggest the presence of a wooden structure, or shelf, that could have stored items off the ground. Seve­ral methods of storage could thus have been used within the cave that could have allowed short- to long-term storage, under the right conditions. Baskets and textiles are both the result of the inten­tional weaving of fibres. The terms ‘basket’ and ‘tex­tile’ are often definitionally separated, perhaps some­what arbitrarily, by both end-use and construction technique. Baskets generally serve as vessels or other containers or as mats for sitting and sleeping on, floor coverings, in the con­struction of mud-brick archi­tecture, and as burial shrouds and grave liners. Textiles, which are made from softer and more pliable fibres, are used for clothing, bed linens, and to create soft bags or other containers that need to have more flexibility than a basket. A large variety of fi­bres are used in weaving tex­tiles and baskets, including bast fibres from plants and trees as well as hair and wool. Textile fibres generally receive more pre-treatment than the fibres used in basket making. Baskets are created from plant fibres that are generally thick­er and more resilient than textile fibres, and they are often treated with splitting, heating, dying, bend­ing, and bundling. Moreover, baskets are never wo­ven on a loom and generally have a different end-use than textiles (Adovasio 1977.1; Crowfoot 1954. 414; Wendrich 1999.31–35). Tools such as awls and needles are often used in the construction of a bas­ket, and thread or cordage may be used to create a more secure weave or to fasten the end of the weav­ing bundle or the baskets edge. The techniques of basket making are generally classified into three weave types: twining, coiling, and plaiting. Within each of these three classes are many sub-classes; all are mutually exclusive based on technique or “fea­tures of manufacture” (Adovasio 1977.1; Wendrich 1999.41–42). The archaeological visibility of storage methods and stored goods varies widely, making it difficult to de­termine the character, organization and importance of storage within a specific context. For the prehisto­ric Balkans a range of different storage methods have been identified, but usually from indirect evi­dence, such as the discovery of large vessels, clay bins or subterranean features, and are usually inter­preted from ethnographic analogies (Filipovi. et al. 2018; Papaefthymiou-Papanthimou et al. 2013). Observations on construction techniques and meth­ods and materials used for lining and sealing stored crops highlight the wide range of practices that can be used (e.g., Mobolade et al. 2019; Pena-Chocarro et al. 2015). When storing crops, especially cereals, it is important to keep both moisture and tempera­ture levels low if the items are to be stored succes- Late Bronze Age food storage in Lower Cerova;ka Cave, Croatia> the archaeobotanical evidence sfully for long periods (e.g., Reynolds 1979; Currid, Navon 1989). Grain aeration is a technique that is still used today to improve the storability of grain by maintaining a cool, uniform temperature throughout the storage. However, this only works if the aerated air has a relative humidity below the grain’s mois­ture content, otherwise the grain would still slowly absorb water from the air (Jones, Hardin 2017). Sealed, airtight, storage is an alternative method to control moisture, and various methods have been observed where things like dung, clay and straw have been used to help seal containers or pits to keep moisture levels low (e.g., Singh et al. 2017). In Syria, clay lined baskets have been observed, as well as sacks and wooden silos (Al-Azem 1992). While in Palestine burgur and frikkeh were seen stored in cloth sacks or in lined straw baskets with some form of protective cover (Turkowski 1969). Lower Cerova.ka Cave in Bronze Age Croatia The discovery of such a large archaeobotanical col­lection at Lower Cerova.ka Cave is unique in Dalma­tia and Croatia as a whole. Although Bronze Age ma­terial culture has been identified at several cave sites along the Dalmatian coast, only one other site has so far produced botanical remains. Grap.eva .pi­lja, a cave on the island of Hvar in Croatia, yielded only a few plant remains from early and middle Bronze Age occupation horizons, including a few wheat grains (Triticum sp.) and acorns (Quercus sp.; Borojevi. et al. 2008). The cave is thought to have had ritual connotations, but the botanical data is inconclusive and could simply suggest transient occupation. Burials in caves are also seen. A recent study at the Middle/Late Bronze Age (1430–1290 BCE) Bezdanja.a Cave, located slightly inland in the Lika region of Croatia, identified notable quantities of C4 plant consumption, most likely millet, in 16 in­dividuals (Martinoia et al. 2021). At Pupi.ina Cave, located in NE Istria, evidence suggests the use of the site periodically by herders as well as for other, as yet unknown activities from the Neolithic through the Iron Age, though a hiatus is noted from the Late Neolithic to middle Bronze Age (Miracle, Forenba-her 2005). The use of caves as animal stabling is also suggested for four caves in the Trieste Karst, north-eastern Italy (Boschian, Montagnari-Kokelj 2020). Caves were thus utilized in different ways along the Adriatic coast. At present only 17 sites have published archaeobo­tanical evidence from Croatia as a whole, and the quality and quantity vary greatly (Reed et al. 2022a). Along the coast, we see a very limited repertoire of remains, with only the settlement at Monkodonja providing any clear evidence of crop cultivation, in­cluding emmer (Triticum dicoccum), barley (Hor­deum vulgare) and grape pips (Vitis vinifera). In continental Croatia, the Late Bronze Age site of Kal­nik-Igri..e revealed thousands of plant remains with­in a burnt down house, with broomcorn millet, bar­ley, free-threshing wheat (Triticum aestivum) and broad bean predominating (Marekovi. et al. 2015; Reed et al. 2021). Broomcorn millet, barley and free-threshing wheat are also frequently found at other sites in the region by the Late Bronze Age (Reed et al. 2002a). Recent research on the intro­duction and adoption of millet has shown its arrival into Croatia by the middle Bronze Age (Filipovi. et al. 2020; Reed et al. 2022b), but it is not until the Late Bronze Age that we see clear evidence of its cul­tivation as a crop within the Croatian assemblage. Overall, the range of taxa identified from Lower Ce­rova.ka Cave fits well with what is already known about Late Bronze Age agriculture in Croatia. What is absent at Lower Cerova.ka Cave is evidence of the collection of fruits and nuts, such as cornelian cherry (Cornus mas) and Chinese lantern (Physalis alkekengi), from the local environment, which we commonly see at settlement sites during this peri­od in Croatia. We do find a few remains of acorns, although these are largely present in Phases 2 and 6, with only one fragment found in Phase 4. There Kelly Reed, Petra Radakovic´, Sara Essert, and Dinko Tresic´ Pavi;ic´ are many species of acorn, both sweet and bitter and unfortunately here we have not been able to identify to species. Acorns are nutritionally compa­rable to cereals, being a good source of carbohy­drates, fats, proteins, and vitamins, mostly A and C, and have been consumed in the form of bread, soups, porridge, or even as herbal coffee through­out history (Sekeroglu et al. 2017). Acorns have been found at other Bronze Age sites, including Kal­nik-Igri..e (Marekovi. et al. 2015), but it is unclear whether the few acorns recovered here represent deliberate collection. Conclusion At Lower Cerova.ka Cave the unique discovery of a large quantity of burnt plant remains dating to the Late Bronze Age indicate crop storage in the cave. Si­gnificant mixing of the crops prevents any assump­tions about how or where the different crops were stored. Yet the significant quantity of remains indi­cate the storage of lentil (Lens culinaris) and free-threshing wheat (Triticum aestivum/durum), as well as emmer (Triticum dicoccum), einkorn (Triti-cum monococcum), spelt (Triticum spelta) and bro-omcorn millet (Panicum miliaceum). The large quantity of glume wheat glume bases could also sug­gest a multifunctional space where glume wheats were also processed, or alternatively the separate storage of chaff for other purposes. Whether the cave was used for short- or long-term storage is de­batable, as the high humidity could cause crop spoilage. At the Late Bronze Age caves in southern France, it was concluded that the caves were used for short-term storage, owing to the high humidity, and that people took shelter in these ‘refuge caves’ during disturbed times (Bouby et al. 2005). How­ever, if air-tight storage is used then crops can be stored for longer. At Lower Cerova.ka Cave possible evidence of woven containers is present, but it is un­clear whether these were sealed or simply used to contain each of the crops separately. The large quan­tity of other materials found within Phase 4, such as bronze dress ornaments and jewellery, and the short date range (c. 940–960 BC), could indicate a period of instability. . References Adovasio J. M. 1977. Basketry Technology: A Guide to Identification and Analysis. Chicago. Al-Aze A. 1992. Crop Storage in Ancient Syria: A Func­tional Analysis Using Ethnographic Modelling. Unpubli­shed PhD thesis. Institute of Archaeology. University Col­lege London. London. https://discovery.ucl.ac.uk/id/eprint/10117906/ Angourakis A., Ignacio Santos J., Manuel Galán J., and Bal­bo A. L. 2015. Food for all: An agent-based model to ex­plore the emergence and implications of cooperation for food storage. Environmental Archaeology 20(4): 349– 363. https://doi.org/10.1179/1749631414Y.0000000041 Aydan Ö., Ulusay R. 2013. Geomechanical evaluation of Derinkuyu antique underground city and its implications in geoengineering. Rock Mechanics and Rock Engineer­ing 46: 731–754. https://doi.org/10.1007/s00603-012-0301-7 Boardman S., Jones G. 1990. Experiments on the effects of charring on cereal plant-components. Journal of Ar­chaeological Science 17: 1–11. https://doi.org/10.1016/0305-4403(90)90012-T Bogaard A. C., Palmer C., Jones G., Charles M., and Hodg-son J. G. 1999. A FIBS approach to the use of weed ecol­ogy for archaeobotanical recognition of crop rotation. Journal of Archaeological Science 26: 1211–1224. https://doi.org/10.1006/jasc.1998.0364 Bogaard A., Charles M., Twiss K. C., +6 authors, and He-necke J. 2009. Private pantries and celebrated surplus: Storing and sharing food at Neolithic Çatalhöyük, Central Anatolia. Antiquity 83(321): 649–668. https://doi.org/10.1017/S0003598X00098896 Bogaard A., Fochesato M., and Bowles S. 2019. The farm-ing-inequality nexus: new insights from ancient Western Eurasia. Antiquity 93(371): 1129–1143. https://doi.org/10.15184/aqy.2019.105 Borojevi. K., Forenbaher S., Kaiser T., and Berna F. 2008. Plant Use at Grap.eva Cave and in the Eastern Adriatic Neolithic. Journal of Field Archaeology 33(3): 279–303. https://doi.org/10.1179/009346908791071231 Boschian G., Montagnari-Kokelj E. 2000. Prehistoric she­pherds and caves in the Trieste Karst (Northeastern Italy). Geoarchaeology. An International Journal 15(4): 31– Late Bronze Age food storage in Lower Cerova;ka Cave, Croatia> the archaeobotanical evidence 371. https://doi.org/10.1002/(SICI)1520-6548(200004)15 :4<331::AID-GEA3>3.0.CO;2-H Bouby L., Fages G., and Treffort J. M. 2005. Food storage in two Late Bronze Age caves of Southern France: palaeo­ethnobotanical and social implications. Vegetation History and Archaeobotany 14(4): 313–328. https://doi.org/10.1007/s00334-005-0079-6 Capper R. T. J., Neef R. 2012. Handbook of Plant Palaeo-ecology. Groningen Archaeological Studies 19. Barkuis Publishing. Groningen. Charles M., Jones G., and Hodgson J. G. 1997. FIBS in ar­chaeobotany: Functional interpretation of weed floras in relation to husbandry practices. Journal of Archaeologi­cal Science 24: 1151–1161. https://doi.org/10.1006/jasc.1997.0194 Crowfoot G. M. 1954. Prodotti Tessili, Lavori di Intreccio e Stuoie. In C. Singer, E. J. Holmyard, A. R. Hall, and T. I. Williams (eds.), Storia delta tecnologia, 1. La preistoria e gli antichi imperi. Bollati Boringhier. Torino: 420–462. Currid J. D., Navon A. 1989. Iron Age Pits and the Lahav (Tell Halif) Grain Storage Project. Bulletin of the Ameri­can Schools of Oriental Research 273: 67–78. Drechsler-Bi.i. R. 1970. Cerova.ka Donja spilja – Iskopa­vanja 1967. Godine. Vjesnik Arheolo.kog muzeja u Zag-rebu 4(1): 93–110. 1983. Japodska kulturna grupa. In A. Benac (ed.), Pra­istorija jugoslavenskih zemalja, IV Bronzano doba. Centar za balkanolo.ka ispitivanja. Akademija nauka i umjetnosti Bosne i Hercegovine. Svjetlost. Sarajevo: 374–389. 1984. Bron.ano doba u pe.inama Like. In M. Malez (ed.), Deveti jugoslavenski speleolo.ki kongres. Zbor­nik predavanja. Karlovac 17.–20. oktober. Zagreb: 623–639. Emir E., Daloglu G. 2012. Dehumidification property of Eskisehir-Derbent Ciftligi white tuffs in Turkey. Environ­mental Earth Sciences 67: 2101–2108. https://doi.org/10.1007/s12665-012-1649-4 Filipovi. D., Obradovi. .., and Tripkovi. B. 2018. Plant storage in Neolithic southeast Europe: synthesis of the ar­chaeological and archaeobotanical evidence from Serbia. Vegetation History and Archaeobotany 27: 31–44. https://doi.org/10.1007/s00334-017-0638-7 Filipovi. D., Meadows J., Corso M. D., +36 authors, and Zerl T. 2020. New AMS 14C dates track the arrival and spread of broomcorn millet cultivation and agricultural change in prehistoric Europe. Scientific Reports 10: 13698. https://doi.org/10.1038/s41598-020-70495-z Forbes H., Foxhall L. 1995. Ethnoarchaeology and Storage in the Ancient Mediterranean beyond Risk and Survival. In J. Wilkins, D. Harvey, and M. Dobson (eds.), Food in Antiquity. Exeter Studies in History Series. University of Exter Press. Exeter: 69–86. Fraser R. A., Bogaard A., Schäfer M., Abogast R., and Heaton T. H. E. 2013. Integrating botanical, faunal and human stable carbon and nitrogen isotope values to re­construct land use and palaeodiet at LBK Vaihingen an der Enz, Baden-Württemberg. World Archaeology 45(3): 492–517. https://doi.org/10.1080/00438243.2013.820649 Halstead P., Jones G. 1989. Agrarian ecology in the Greek islands: Time stress, scale and risk. Journal of Hellenic Studies 109: 41–55. Halstead P., O’Shea J. M. (eds.). 1989. Bad Year Econo­mics: Cultural Responses to Risk and Uncertainty. Cam­bridge University Press. Cambridge. https://doi.org/10.1017/CBO9780511521218 Hastorf C. A., Foxhall L. 2017. The social and political as­pects of food surplus. World Archaeology 49(1): 26–39. https://doi.org/10.1080/00438243.2016.1252280 Hellmuth Kramberger A. 2017. Monkodonja. Istra.ivanje protourbanog naselja bron.anog doba Istre. Knjiga 2/ 1–2. Keramika s bron.anodobne gradine Monkodonja – Tekst. Forschungen zu einer protourbanen Siedlung der Bronzezeit Istriens. Teil 2/1–2. Die Keramik aus der bronzezeitlichen Gradina Monkodonja – Text. Mo-nografije i katalozi 28/1. Arheolo.ki muzej Istre. Pula. Hillman G. C. 1984. Interpretation of archaeological plant remains: The application of ethnographic models from Turkey. In W. Van Zeist, W. A. Casparie (eds.), Plants and Ancient Man; studies in palaeoethnobotany. Proceedings of the Sixth Symposium of the International Work Group for Palaeoethnobotany. Groningen, 30 May–3 June 1983. A. A. Balkema. Rotterdam: 1–41. Ingold T. 1983. The significance of storage in hunting so­cieties. Man 18(3): 553–571. Jones C., Hardin J. 2017. Aeration and cooling of stored grain. Oklahoma Cooperative Extension Service BAE­1101: 1–4. Jones G. 1984. Interpretation of archaeological plant re­mains: ethnographic models from Greece. In W. Van Zeist, W. A. Casparie (eds.), Plants and Ancient Man; studies in palaeoethnobotany. Proceedings of the Sixth Kelly Reed, Petra Radakovic´, Sara Essert, and Dinko Tresic´ Pavi;ic´ Symposium of the International Work Group for Palaeo­ethnobotany. Groningen, 30 May–3 June 1983. A. A. Bal­kema. Rotterdam: 43–61. Jones G., Halstead, P 1995. Maslins, mixtures and mono-crops: On the interpretation of archaeobotanical crop samples of heterogeneous composition. Journal of Ar­chaeological Science 22: 103–114. https://doi.org/10.1016/S0305-4403(95)80168-5 Jones G., Valamoti S., and Charles M. 2000. Early crop di­versity: a “new” glume wheat from northern Greece. Ve­getation History and Archaeobotany 9: 133–146. https://doi.org/10.1007/BF01299798 Jones G. E. M., Charles M., Bogaard A. C., and Hodgson J. G. 2010. Crops and weeds: The role of weed functional ecology in the identification of crop husbandry methods. Journal of Archaeological Science 37: 70–77. https://doi.org/10.1016/j.jas.2009.08.017 Kohler-Schneider M. 2003. Contents of a storage pit from late Bronze Age Stillfried, Austria: another record of the “new” glume wheat. Vegetation History and Archaeobo­tany 12(2): 105–111. https://doi.org/10.1007/s00334-003-0010-y Kreuz A. 2007. Archaeobotanical perspectives on the be­ginning of agriculture north of the Alps. In S. College, J. Conolly (eds.), The origins and spread of domestic plants in Southwest Asia and Europe. Left Coast Press. Walnut Creek: 259–294. Kure.i. T., Bo.i. N., Wacha L., +7 authors, and Fiebig M. 2021. Changes in Cave Sedimentation Mechanisms During the Late Quaternary: An Example from the Lower Cero­va.ka Cave, Croatia. Frontiers in Earth Science 9: 672229. https://doi.org/10.3389/feart.2021.672229 Marekovi. S., Karavani. S., Kudeli. A., and .o.tari. R. 2015. The botanical macroremains from the prehistoric settlement Kalnik-Igri..e (NW Croatia) in the context of current knowledge about cultivation and plant consump­tion in Croatia and neighbouring countries during the Bronze Age. Acta Societatis Botanicorum Poloniae 84 (2): 227–235. https://doi.org/10.5586/asbp.2015.015 Martinoia V., Benazzi S., Cari. M., and Radovi. S. 2021. Reconstructing the childhood diet of the individuals from the Middle Late Bronze Age Bezdanja.a Cave, Croatia (c. 1430–1290 BCE) using stable C and N isotope analysis of dentin collagen. Journal of Bioanthropology 1(1): 42–56. https://doi.org/10.54062/jb.1.1.3 McAlvay A. C., DiPaola A., D’Andrea A. C., +3 authors, and Power A. G. 2022. Cereal species mixtures: an ancient practice with potential for climate resilience. A review. Agronomy for Sustainable Development 42: 100. https://doi.org/10.1007/s13593-022-00832-1 Meurers-Balke J., Lüning J. 1992. Some aspects and expe­riments concerning the processing of glume wheats. In P. C. Anderson (ed.), Préhistorie de l’agriculture. Nouvelles approches expérimentales et ethnographiques. Mono-graphie du Centre de Recherches Archéologiques 6. Cen­tre de Recherches Archéologiques. Éditions du CNRS. Pa­ris: 341–362. Miracle P. T., Forenbaher S. 2005. Neolithic and Bronze Age Herders of Pupi.ina Cave, Croatia. Journal of Field Archaeology 30(3): 255–581. https://doi.org/10.1179/009346905791072251 Mobolade A. J., Bunindro N., Sahoo D., and Rajashekar Y. 2019. Traditional methods of food grains preservation and storage in Nigeria and India. Annals of Agricultural Sciences 64: 195–205. https://doi.org/10.1016/j.aoas.2019.12.003 Nikoli. T. 2018. Flora Croatica Database. Available from: http://hirc.botanic.hr/fcd. Papaefthymiou-Papanthimou A., Valamot S. M., Tsagkara­ki E., Voulgari E., and Papadopoulou E. 2013. Food stor­age in the context of an Early Bronze Age household eco­nomy: new evidence from Archontiko Giannitson. In S. Voutsaki, S. M. Valamoti (eds.), Diet, Economy and So­ciety in the Ancient Greek World: Towards a Better In­tegration of Archaeology and Science. Proceedings of the International Conference held at the Netherlands Institute at Athens on 22–24 March 2012. Pharos Sup­plement 1. Peeters. Leuven: 103–111. Pa¢un A., Stroescu G., Zaica A., +4 authors, and Ba¢ltatu C. 2021. Storage of grains and technical plants through ac­tive ventilation for the purpose of maintaining the qual­ity of stored products. E3S Web of Conferences 286: 03010. https://www.e3s-conferences.org/articles/e3scon f/abs/2021/62/e3sconf_tererd2021_03010/e3sconf_ter erd2021_03010.html Pena-Chocarro L., Pérez Jorda G., Morales Mateos J., and Zapata L. 2015. Storage in traditional farming communi­ties of the western Mediterranean: Ethnographic, histori­cal and archaeological data. Environmental Archaeology 20(4): 379–389. https://doi.org/10.1179/1749631415Y.0000000004 Petrie C. A., Bates J. 2017. ‘Multi-cropping’, Intercropping and Adaptation to Variable Environments in Indus South Asia. Journal of World Prehistory 30: 81–130. https://doi.org/10.1007/s10963-017-9101-z Late Bronze Age food storage in Lower Cerova;ka Cave, Croatia> the archaeobotanical evidence Pixton S. W., Griffiths H. J. 1971. Diffusion of moisture through grain. Journal of Stored Products Research 7: 133–152. https://doi.org/10.1016/0022-474X(71)90001-4 Radakovi. P. 2021. Arheobotani.ko istra.ivanje prapo­vijesnog lokaliteta Donja Cerova.ka .pilja. MA thesis. Sveu.ili.te u Zagrebu. Prirodoslovno-matemati.ki fakul­tet. Biolo.ki odsjek. Zagreb. https://urn.nsk.hr/urn:nbn:hr:217:777579 Rajendran S. 2003. Grain storage: Perspectives and prob­lems. In A. Chakraverty, A. S. Mujumdar, and H. S. Rama-swamy (eds.), Handbook of Postharvest Technology: Ce­reals, Fruits, Vegetables, Tea and Spices. Marcel Dekker Inc. New York: 183–192. Reed K., Kudeli A., Esser S., Polonij L., and Karavani. S. 2021. House of plenty: Reassessing food and farming in Late Bronze Age Croatia. Environmental Archaeology: ahead-of-print: 1–17. https://doi.org/10.1080/14614103.2021.1979385 Reed K., Hr.ak T., Mihaljevi. M., and Balen J. 2022a. Set­tlements and cemeteries in Bronze Age Croatia: The ar­chaeobotanical evidence. Acta Palaeobotanica 62. in press Reed K., Balen J., Drni. I., +3 authors, and Zavodny E. 2022b. Unearthing millet in Bronze and Iron Age Croatia. In W. Kirleis, M. Dal Corso, and D. Filipovi. (eds.), Millet and what else? The wider context of the adoption of mil­let cultivation in Europe. Scales of Transformation in Prehistoric and Archaic Societies 14. Sidestone Press. Lei­den: 95–106. Reynolds P. J. 1979. A general report of underground grain storage experiments at the Butser ancient farm pro­ject. In M. Gast, F. Sigaut (eds.), Les Techniques de Con­servation des Grains a Long Terme I. Leur rôle dans la dynamique des systémes de cultures et des sociétes. Edi­tions du Centre national de la recherche scientifique. Pa­ris: 70–80. Sadori L., Susanna F., and Persiani C. 2006. Archaeobo­tanical data and crop storage evidence from an early Bronze Age 2 burnt house at Arslantepe, Malatya, Turkey. Vegetation History and Archaeobotany 15(3): 205–215. https://doi.org/10.1007/s00334-005-0029-3 Sekeroglu N., Ozkutlu F., and Kilic E. 2017. Mineral com­position of acorn coffees. Indian Journal of Pharmaceu­tical Education and Research 51: 504–507. DOI: 10.5530/ijper.51.3s.75 Singh V., Verma D. K., Srivastav P. P. 2017. Food Grain Storage Structures: Introduction and Applications. In D. K. Verma, M. R. Goyal (eds.), Engineering Interventions in Foods and Plants. Innovations in Agricultural and Biological Engineering. Apple Academic Press. New York: 247–284. https://doi.org/10.1201/9781315194677 Soffer O. 1989. Storage, sedentism and the Eurasian Pa­laeolithic record. Antiquity 63: 719–732. https://doi.org/10.1017/S0003598X00076857 Stevens C. J. 1996. Iron Age and Roman Agriculture in the Upper Thames Valley: Archaeobotanical and Social Perspectives. Unpublished PhD thesis. University of Cam­bridge. Cambridge. https://doi.org/10.17863/CAM.19155 2003. An investigation of agricultural consumption and production models for prehistoric and Roman Britain. Environmental Archaeology 8(1): 61–76. https://doi.org/10.1179/env.2003.8.1.61 Testart A. 1982. The significance of food storage among hunter-gatherers: residence patters, population densities, and social inequalities. Current Anthropology 23(5): 523– 537. Tolar T., Jacomet S., and Velu..ek A. 2016. Cereal chaff used as temper in loom-weights: new evidence from a Slo­venian Eneolithic pile-dwelling site (c. 3100 cal BC). Ve­getation History and Archaeobotany 25: 291–301. https://doi.org/10.1007/s00334-015-0549-4 Tresi. Pavi.i. D. 2020. Arheolo.ko Istra.ivanje Donje Ce­rova.ke (Turkaljeve) .pilje. Subterranea Croatica 18(29): 60–74. Turkowski L. 1969. Peasant agriculture in the Judaean hills. Palestine Exploration Quarterly 101(1): 21–33. Valamoti S. M., Charles M. 2005. Distinguishing food from fodder through the study of charred plant remains: an ex­perimental approach to dung-derived chaff. Vegetation History and Archaeobotany 14: 528–533. https://doi.org/10.1007/s00334-005-0090-y Van Der Veen M. 1992. Crop Husbandry Regimes: An Ar-chaeobotanical Study of Farming in northern England 1000 BC–AD 500. Sheffield archaeological monographs 3. University Sheffield. Department of Archaeology and Prehistory. J. R. Collis Publications. Sheffield. 1995. The identification of maslin crops. In H. Kroll, R. Pasternak (eds.), Res Archaeobotanica. International Workgroup for Palaeoethnobotany. Proceedings of the 9th Symposium Kiel 1992. Oetker-Voges. Kiel: 335–343. 1999. The economic value of chaff and straw in arid and temperate zones. Vegetation History and Archaeobo­tany 8: 211–224. https://doi.org/10.1007/BF02342721 Kelly Reed, Petra Radakovic´, Sara Essert, and Dinko Tresic´ Pavi;ic´ 2007. Formation processes of desiccated and carbonized plant remains. The identification of routine practice. Journal of Archaeological Science 34(6): 968–990. https://doi.org/10.1016/j.jas.2006.09.007 Van Der Veen M., Jones G. 2006. A re-analysis of agricul­tural production and consumption. Vegetation History and Archaeobotany 15 (3): 217–228. https://doi.org/10.1007/s00334-006-0040-3 Wendrich W. 1999. The World According to Basketry. An Ethno-archaeological Interpretation of Basketry Pro­back to contents duction in Egypt. CNWS publications 83. Research School of Asian, African and Amerindian Studies. Universiteit Lei­den. Leiden. Winterhalder B., Puleston C., and Ross C. 2015. Produc­tion risk, inter-annual food storage by households and po-pulation-level consequences in seasonal prehistoric agra­rian societies. Environmental Archaeology 20(4): 337– 348. https://doi.org/10.1179/1749631415Y.0000000025 Zohary D., Hopf M. 2000. Domestication of Plants in the Old World-The origin and spread of cultivated plants in west Asia, Europe, and the Nile Valley. Oxford University Press. Oxford. Documenta Praehistorica XLIX (2022) Hermetic cereal storage in the Bronze Age> evidence from the Gáva culture settlement at Rotbav, Transylvania Laura Dietrich1, Oliver Dietrich2, and Julia Meister3 1 Deutsches Archäologisches Institut, Berlin, DE< laura.dietrich@dainst.de< corresponding author 2 Freie Universität Berlin, Berlin, DE< o.dietrich@fu-berlin.de 3 Julius-Maximilians-Universität Würzburg, Würzburg, DE< julia.meister@uni-wuerzburg.de ABSTRACT – The present paper explores the possibility to better understand the function of pits through phytolith and starch analysis. A case study from the Late Bronze Age/Early Iron Age settle­ment phase of Rotbav in southeastern Transylvania is discussed in detail. It appears that a large sto­rage vessel originally sealed with a bowl was kept in a pit filled with chaff or straw to preserve its contents. KEY WORDS – Bronze Age; Transylvania; food storage; phytoliths Hermeti;no skladi[;enje /it v bronasti dobi> dokazi iz najdi[;a kulture Gáva v Rotbavu v Transilvaniji IZVLE.EK – V .lanku raziskujemo mo.nost za bolj.e razumevanje namembnosti jam na podlagi ana­liz fitolitov in .kroba. Obravnavamo .tudijski primer iz pozno bronasto- in starej.e .eleznodobne fa­ze naselbine Rotbav v severovzhodni Transilvaniji. Gre za ve.jo shrambeno posodo, ki je bila pokri­ta s skledo in bila postavljena v jamo, zapolnjeno s plevami ali slamo, z namenom konserviranja vse­bine. KLJU.NE BESEDE – bronasta doba; Transilvanija; shranjevanje hrane; fitoliti Introduction Several methods have traditionally been employed to keep humidity, oxygen, and insects away from field crops in order to preserve them, with airtight sealed containers, along with pits lined with straw or chaff, being among the most common (Reynolds 1974; Sigaut 1980; 1988; Fairbairn, Omura 2005, Villers et al. 2006; Diffey et al. 2017; Urem-Kotsou 2017). However, it is difficult to find evidence for such techniques in the archaeological record, parti­cularly regarding regions with earthen architecture and in the absence of carbonized cereals (Monah 2002; Marinova, Valamoti 2014; Hrisrova et al. 2017; Valamoti et al. 2019 for an overview of such evidence from southeastern Europe). So far, the identification of grain storage has mostly been based on the identification of characteristic archaeological features (Sigaut 1988; Fairbairn, Omura 2005 with references). This approach may to some degree be hindered by the frequent and complex reuse of sto­rage pit features (Ivanova et al. 2020). This pilot study sets out to highlight a pit type that with a high probability is connected to a specific cereal storage technique and a methodology that allows its iden­tification by phytolith and starch analysis. Our case study comes from the Bronze Age/Early Iron Age settlement of Rotbav in Transylvania, a region and time for which so far only scarce macrobotanical evidence exists (Cârciumaru 1996; Ciuta¢2012; Ciu­ta¢, Bejinariu 2012; 2019; Ciuta¢, Molnár 2014 with references). DOI> 10.4312\dp.49.1 Hermetic cereal storage in the Bronze Age> evidence from the Gáva culture settlement at Rotbav, Transylvania The Gáva culture settlement of Rotbav The archaeological site of Rotbav-La Pârâut is situ­ated upon a high terrace formation above the River Olt in southeastern Transylvania at 498 m.a.s.l (Fig. 1; 45°83’N/25°56’E). The plateau is delimited by the Valea Ceta¢tii stream to the north and a steep hill to the west (Fig. 2); to the east, the Josephinian sur­vey shows a swampy area with an arm of the River Olt. This landscape was heavily transformed by the creation of two lakes to the north of the site in the 1970s, and today the Olt flows at a distance of rough­ly 500m to the east of the site. The form of the pla­teau was not affected, but the site has been and still is used for cultivation, the plough horizon reaching a thickness of approximately 40cm. The settlement has a size of around 4ha, of which 1800m2 were excavated, and the site was additional­ly investigated by archaeological and geophysical surveys. Rotbav-La Pârâut is thus the most extensi­vely researched site of this period in the region so far, and has been comprehensively published (Diet­rich 2014a). Its importance lies in a long stratigra­phy comprising the timespan from the Middle Bronze Age (in Romanian terminology) to the Bronze Age/ Iron Age transition, being inhabited roughly between 1900/1800 BC and 1200/ 1100 BC, following radio­carbon data (Dietrich 2014b). The stratigraphic se­quence covers six distinct building phases. The first three belong to the early Middle Bronze Age Wieten-berg culture, followed by two of the Late Bronze Age Noua culture. The last building phase belongs to the Gáva culture, which marks the Bronze Age/ Iron Age transition (Dietrich 2012). The pottery from Rot-bav (Dietrich 2012; 2014a. 211–214) places the Gáva set­tlement into a developed phase of the culture, described by Marian Guma¢ as horizon Mahala IV-Somotor II-Medias I-II-Teleac II and dated to Ha B (Guma¢1993.190; cf. Ciugu-dean 2009; 2011). The remains of the Gáva set­tlement phase were not pre­sent in all sections excavated at Rotbav, likely due to ero­sion and ploughing, but they could be excavated on an area of 1372m2 (Dietrich 2014a. 214–217). This allowed us in­sights into the settlement structure, which is char­acterized by large, regularly dispersed, partly subter­ranean constructions (Fig. 2). Fireplaces and pit fea­tures are located between these buildings. The cultu­ral layer was situated immediately below the plough horizon and had a thickness of approximately 15cm, although likely it was originally much thicker. A to­tal of four semi-subterranean houses could be par­tially excavated in the main area of the settlement, situated at distances of four to 15m from each other. Postholes and burnt loam with impressions of wickerwork hint at the superstructures, although the entrances or inner divisions of houses could not be identified. Near the houses, concentrations of pottery and other artefacts suggest that activity areas and pits were regularly associated with the dwel­lings. Most pits were filled with domestic refuse or settlement debris (Dietrich 2014a for an extensive presentation of the features), and thus their original functions could not be determined with security. However, one, labelled feature 4/2008, stood out because of its contents – two nearly complete ves­sels, making an in situ use context highly likely. Pit feature 4/2008 Feature 4/2008 was located approximately 5m to the south of one of the houses (structure 10, Figs. 2– 3) and next to the remains of a fireplace (feature 2/ 2008, Fig. 2) destroyed by ploughing. It first show­ed as an oval-rounded yellowish spot of c. 110cm maximal diameter. The pit filling proved to be very homogenous loamy fine sand. Pottery fragments and burnt loam were observed only in its upper­most part, roughly within the first 10cm. The maxi- Laura Dietrich, Oliver Dietrich, and Julia Meister mum depth of the feature was 54cm, a part of the pit likely being destroyed by the plough. After remo­val of 10cm of sediment, finds became scarce, but the rims of two large vessels became visible. One was a large, bag-shaped vessel decorated with four knobs at the shoulder and a rim drawn towards the exterior (Fig. 3a). The vessel, of which all fragments could be recovered, has a height of 42.7cm, a maxi­mal diameter of 47.6cm and a rim diameter of 35cm. The measurements and formal characteristics – a wide stable bottom and large mouth to enable users to reach the contents – speak in favour of a storage vessel. The second vessel is a so-called ‘Zipfelschüs­sel’, a bowl that originally had four pronounced lo­bes, fluted in the interior (Fig. 3b). This vessel was found broken in several fragments, but unlike the first one not all fragments were present in the pit. The preserved smallest width of the vessel is 41.8cm, and thus it could have served well as a lid for the larger vessel even in a damaged condition. The sediments from the pit and the inside of the ves­sel were separated, sieved through a 1mm mesh and flotated. This produced a number of small bone fragments and <1g of charred wood from the pit fill- Fig. 2. The Gáva settlement phase of Rotbav and the location of the pit feature 4/2008 (graphics by L. Dietrich). Hermetic cereal storage in the Bronze Age> evidence from the Gáva culture settlement at Rotbav, Transylvania ing, but no other charred plant macroremains. As charred grains and plant remains were recovered from other contexts at Rotbav (cf. Dietrich 2014a. Anhang 6), this is not due to preservation condi­tions. The vessel contents were completely decayed, or the vessel had been emptied (which could be indicated by the position of the lid besides the large vessel). No food crusts were observed on either ves­sels’ inner surface. However, neither of the two ves­sels had been placed on the bottom of the pit, indi­cating that the latter had been filled with some kind of material that held the vessels in place (Figs. 4– 5). We suspected that the vessels could have been originally placed in chaff or straw. Accordingly, four sediment samples were taken to check this hypoth­esis. Phytolith evidence Phytolith analyses were conducted on the four soil samples (Tabs. 1–2). RT08-1 is from the upper part of the pit, RT08-2 from inside the pit next to the ves­sels, RT08-4 is from inside the vessel and RT08-3 from the cultural layer outside the pit (sample loca­tions are marked in Fig. 4). Phytolith extraction of the samples followed the pro­cedures outlined by Rosa Maria Albert et al. (1999). To remove carbonates, phosphates, and organic ma­terial, approximately 1g of the air-dried sediment was treated with 3 N HCl, 3 N HNO3, and H2O2. The mineral components of the samples were separat­ed according to their densities using 2.4g/ml sodium polytungstate solution [Na6 (H2W12O40) H2O]. Slides were prepared by weighing out about 1mg of sedi­ment onto a microscope slide, mounting with Entellan New (Merck). The count­ing of about 1300 phytoliths per sample was performed using a KERN OBE-114 microscope at 400x magnification. Un­identifiable phytoliths were counted and recorded as weathered morphotypes. To allow quantitative comparisons between the samples, phytolith numbers per gram of sediment were estimated by re­lating phytolith amounts and weights of the processed sample material to the initial sample weights. Morphological identification of phytoliths was based on standard literature (e.g., Twiss et al. 1969; Brown 1984; Mulholland, Rapp Jr. 1992; Piperno 2006), as well as on modern plant reference collections (Al­bert 2000; Albert, Weiner 2001; Tsartsi­dou et al. 2007; Albert et al. 2011; Por­tillo et al. 2014). The International Code for Phyto­lith Nomenclature was followed where possible (Ma­della et al. 2005). Phytoliths were abundant in all four soil samples examined, ranging from 1.1 to 4.0 million phytoliths per gram of sediment (Fig. 6a; Tab. 1). The highest concentrations were observed in samples RT08-2 and RT08-4, while the lowest concentration was ob­served for sample RT08-3. Phytolith preservation is generally poor, as evidenced by high proportions of weathered phytoliths (mean=21.4%, .=1.5%, n=4; Tab.1) and the absence of multicellular phytoliths, likely in association with a varied range of deposi­tional and post-depositional processes (Alexandre et al. 1997; Cabanes et al. 2011; Madella, Lance-lotti 2012). The morphological analyses show that all samples are similar in their morphotype assemblages (Tab. 1). Grass phytoliths, occurring at a rate of about 56.5% (.=0.6%, n=4), were the most common group identified. According to their short cell morpholo­gies, grasses belong mostly to the C3 Pooideae sub­family that include common cereals, such as wheat and barley. However, the absence of multicellular phytoliths in the samples did not allow for identi­fying the type of grasses and cereals. Grass short cells, commonly produced in leafs, stems and inflo­rescences, were abundant in all samples, averaging 40.4% (.=1.8%, n=4). Epidermal cells from grass leaves and stems, including, for instance, prickles and bulliform cells, show similar values with an ave­rage amount of c. 40.3% (.=2.1%, n=4). Additio- Laura Dietrich, Oliver Dietrich, and Julia Meister Sample ID, description of sample location and phytolith amounts Relative abundances of phytoliths Anatomical origin of grass phytoliths Sample ID Number of phyt. per 1g of sediment Layer\areal\ description Grass phyt. (%) Dicotyledonous leaves (%) Dicotyledonouswood\bark (%) Other phyt. (%) Weatheredmorpho types (%) Leaves and stems(%) Short Cells (%)Inflorescence phyt. (%) RT08-1\MD 5309 2 675 000 inside pit, upper part 56.80 4.18 15.23 0.85 22.95 37.24 42.78 19.98 RT08-2\MD 5310 3 972 000 inside pit next to vessels 56.47 6.38 16.90 0.58 19.66 41.06 38.69 20.25 RT08-3\MD 5311 1 138 000 cultural layer outside pit 55.72 6.97 16.18 0.53 20.60 40.92 39.34 19.75 RT08-4\MD vessel 3 617 500 inside large vessel 57.08 6.27 14.03 0.20 22.42 42.07 40.82 17.11 Tab. 1. Description of samples, phytolith amounts, relative abundances of phytoliths and anatomical ori­gin of grass phytoliths obtained from all sediment samples. nally, grass phytoliths derived from their floral parts (e.g., decorated elongate dendritic and elongate echi­nate cells) account for 19.3% on average (.=1.5%, n=4). Dicotyledonous phytoliths occur at an average rate of 21.5% (.=2.0, n=4; Tab. 1). Parallelepipedal blocky phytoliths, for instance, one of the most common wood/bark morphotypes, account for 12.5% (.= 1.7%, n=4) on average. Other diagnostic dicotyledo­nous morphotypes such as globulars, polyhedrals or jigsaw-shaped phytoliths were not observed. Starch analysis To confirm the original presence of cereals within the large vessel, five subsamples from sample RT08­4 (inside of the vessel) and the control sample RT08­3 (cultural layer outside of the pit) were subjected to microscopic analysis in order to identify possibly preserved starch granules. Sample preparation/mi­crofossil extraction followed the protocol established by Li Liu et al. (2018) with a few modifications. The sediment was mechanically crushed and homoge­nized. One mg of sediment was put into 1.5ml test tubes, dispersed in distilled water and centrifuged for 5 minutes. Microfossil extraction then followed two procedures: (a) EDTA dispersion; after centri­fuge the supernatant was decanted, 0.4ml of EDTA solution was added to each tube. The tubes were left for 2 hours and vortexed each 10 minutes for 30 se­conds to disperse the sediment, then filled with dis­tilled water and centrifuged for 5 minutes at 3000 rpm, and the supernatant was decanted. (b) Heavy liquid separation; 0.4ml of SPT at a specific gravity of 2.35 was added to each tube. The tubes were then centrifuged for 15 minutes at 3000rpm. The top la­yer of organics was removed from each tube by a new pipette and then transferred into a new tube. Distilled water was added, and the samples centri­fuged for 5 minutes at 3000 rpm to concentrate the starch at the bottom of the tube, and the superna­tant was decanted. The process was repeated two more times. The samples were mounted in 50% glycerol and 50% distilled water on glass slides and analysed with polarizing filters at x400 for starch with a Bres­ser Polarisation microscope. Photos were taken with a Bresser Microcam of 12 MP for each slide. The re­ference collection for starch granule types estab­lished by Gismondi et al. (2019) was used for com­ Hermetic cereal storage in the Bronze Age> evidence from the Gáva culture settlement at Rotbav, Transylvania RT08-1\ RT08-2\ RT08-3\ RT08-4\ Phytolith morphotype Bulliform MD 5309 inside pit, upper part 3 MD 5310 inside pit next to vessel 7 MD 5311 cultural layer outside pit 4 MD Vessel inside large vessel 2 Cillindroid psilate 84 121 85 112 Cillindroid scabrate 33 51 45 49 Hair cell 33 65 60 70 Papillae cell 21 22 17 19 Hair cell (prickle) 22 29 32 21 Elongate dendritic 15 20 23 20 Elongate echinate 45 53 44 55 Elongate polylobate 20 24 14 15 Elongate wavy 16 16 15 16 Elongate verrucate 6 8 4 9 Elongate crenate 5 4 2 2 Elongate ruminate 3 5 1 2 Elongate granulate 0 2 3 4 Elongate spilate 3 0 0 3 Elongate corniculate 0 3 2 0 Parallelepipedal blocky psilate square ends 60 86 60 72 Parallelepipedal blocky psilate rounded ends 28 44 29 16 Parallelepipedal blocky scabrate square ends 30 47 31 30 Parallelepipedal blocky scabrate rounded ends 20 11 10 19 Parallelepipedal blocky psilate irregular 11 9 9 8 Parallelepipedal blocky scabrate irregular 3 0 6 6 Parallelepipedal elongate psilate 12 8 17 22 Parallelepipedal elongate scabrate 7 8 13 9 Parallelepipedal elongate facetated 0 0 0 0 Parallelepipedal thin psilate rounded ends 6 11 10 16 Parallelepipedal thin psilate square ends 73 79 55 131 Parallelepipedal thin scabrate rounded ends 0 2 5 4 Parallelepipedal thin scabrate square ends 17 18 21 38 Short cell rondel 187 213 171 226 Short cell tall rondel 2 0 3 5 Short cell trapeziform 68 67 56 95 Short cell saddle 10 8 6 3 Short cell bilobate 19 18 15 15 Short cell cross 11 2 4 5 Trapeziform sinuate 21 25 13 22 Trapeziform polylobate 8 14 13 6 Cylindric sulcate tracheid 8 3 4 5 Weathered morphotype 271 270 234 333 Total number of counted morphotypes per sample 1181 1373 1136 1485 Tab. 2. List of phytolith morphotypes identified and their frequencies (counts) in soil samples and a pot­tery vessel from Rotbav, giving the stratigraphic location and sample information. parison. Starch preservation was overall bad, and well preserved granules were only observed in three subsamples of RT08-4. These allow a tentative deter­mination as Triticum aestivum (common wheat, cf. Fig. 6b and Gismondi et al. 2019.nr. 30a-b). Triti-cum aestivum is not among the species identified for the Wietenberg culture layers at Rotbav, from which macrorests of Triticum monococcum, Triti-cum sp. and Hordeum sp. have been recovered (Dietrich 2014a.Anhang 6). For the Noua culture, evidence is lacking so far. Discussion Phytolith analysis reveals that two samples have particularly high phytolith concentrations. Sample RT08-2 was taken inside the pit, next to the two pot­tery vessels, RT08-4 is from the inside of the large bag-shaped vessel. Another sample, RT08-1, was taken inside the pit filling, but in a stratigraphical position above the two vessels. Here, the phytolith concentration is considerably lower. The lowest va­lue comes from sample RT08-3 which represents a Laura Dietrich, Oliver Dietrich, and Julia Meister control sample from the cultural layer next to the pit. The phytoliths stem mostly from the C3 Pooid subfamily and come largely from stems and leaves. Both concentrations and origin of the phytoliths fit the hypothesis of a pit filled with chaff or straw to protect the contents of the vessel. Starch analyses in­dicate that the vessel contained cereals, likely Triti-cum aestivum. In addition to being covered with the bowl (frag­ment), the vessel could have been sealed airtight with clay. As the upper part of the pit was absent, the possibility that also the pit was sealed in that way cannot be excluded. Sealing in an airtight container would have reduced moisture and kept insects or mice away, conserving the grains for several years (Diffey et al. 2017.1–3). The capacity of the vessel at Rotbav may indicate that it was used to store a part of the provisions for the winter or seed grains. The find context with the bowl next to the large vessel and the vessel itself filled with straw (phytoliths do not form in grains, starch being scarce) makes it highly possible that the vessel was emptied and then left there (damaged during retrieval?) while the straw/chaff decayed and the pit in its upper part was slowly refilled with sediments. Originally there could have been more such vessels stored in the pit. Phytolith or starch analyses have so far not been published for Gáva sites. Organic remains have only sparsely been reported from contexts of the Gáva culture or the Early Iron Age in general. From Sim-leu Silvaniei – ‘Observator’ foxtail millet (Setaria ita-lica), and two wheat species (Triticum monococ-cum and Triticum dicoccum) are mentioned (Ciu­ta¢, Bejinariu 2019). Beatrice Ciuta¢ and Ioan Beji­nariu recently collected the evidence published to date of other finds of cereals from Early Iron Age contexts, and their list contains three more sites (Te-leac: Triticum durum and Hordeum vulgare from a grave; Bernadea: millet; Ta¢sad: mostly Triticum aestivum, but also Triticum monococcum, Triti-cum dicoccum, Triticum spelta, and Panicum mi-liaceum; Ciuta¢, Bejinariu 2019.47). Rotbav now adds to this list, although any sensible discussion of Early Iron Age cereal use still needs much more data. Summing up, our case study proves that combined phytolith and starch analysis are an interesting (and not overly costly) approach to determine the proba­ble use of prehistoric pits in the absence of pre­served macrorests. Fig. 6. a photomicrographs of selected phytolith morphotypes identified in the Rotbav samples. The photo­graphs were taken at 400x magnification: 1 short cell rondel; 2 short cell trapeziform (left), short cell bilobate (right); 3 elongate entire (left), short cell rondel (top view, right); 4 elongate dentritic; 5 elon­gate echinate; 6 prickle (photos made by C. Binder). b photomicrograph of a starch granule (Triticum aestivum), taken at 400x magnification. Hermetic cereal storage in the Bronze Age> evidence from the Gáva culture settlement at Rotbav, Transylvania ACKNOWLEDGEMENTS The excavations at Rotbav were funded by the Romanian Ministry of Culture. Laura Dietrich carried out starch analysis. Julia Meister carried out phytolith analysis. We are grateful to Iris Müller and Christoph Binder from the University of Würzburg for their technical assistance during sample processing and phytolith analysis. . References Albert R. M. 2000. Study of ash Layers through phytolith analyses from the Middle Paleolithic Levels of Kebara and Tabun Caves. Unpublished PhD thesis. Universitat de Barcelona. Barcelona. Albert R. M., Esteve X., Portillo M., +3 authors, and Her­nandez F. 2011. Phytolith CoRe, Phytolith Reference Col­lection. http://phytcore.org/phytolith/index Albert R. M., Lavi O., Estroff L., +3 authors, and Lev-Yadun S. 1999. Mode of occupation of Tabun Cave, Mt Carmel, Israel during the Mousterian Period: a study of the sedi­ments and phytoliths. Journal of Archaeological Science 26: 1249–1260. http://dx.doi.org/10.1006/jasc.1999.0355 Albert R. M., Weiner S. 2001. Study of phytoliths in prehi­storic ash layers using a quantitative approach. In J. D. Meunier, F. Colin (eds.), Phytoliths, applications in earth sciences and human history. A. A. Balkema Publishers. Lisse: 251–266. Alexandre A., Meunier J.-D., Colin F., and Koud J.-M. 1997. Plant impact on the biogeochemical cycle of silicon and related weathering processes. Geochimica et Cosmochi-mica Acta 61: 677–682. http://dx.doi.org/10.1016/S0016-7037(97)00001-X Brown D. A. 1984. Prospects and limits of a phytolith key for grasses in the central United States. Journal of Archa­eological Science 11: 345–368. http://dx.doi.org/10.1016/0305-4403(84)90016-5 Cabanes D., Weiner S., and Shahack-Gross R. 2011. Stabi­lity of phytoliths in the archaeological record: a dissolu­tion study of modern and fossil phytoliths. Journal of Ar­chaeological Science 38: 2480–2490. http://dx.doi.org/10.1016/j.jas.2011.05.020 Cârciumaru M. 1996. Paleoetnobotanica. Editura Glasul Bucovinei. Iasi. Ciugudean H. 2009. Bemerkungen zur Chronologie der befestigten Siedlung von Teleac. Analele Banatului 17: 67–87. 2011. Periodizarea culturii Gáva în Transilvania în lu­mina noilor cerceta¢ri. Apulum XLVIII: 69–102. Ciuta¢B. 2012. Plant species within the diet of Prehisto­ric communities from Transylvania. Editura Mega. Cluj-Napoca. Ciuta¢ B., Bejinariu I. 2012. Date arheobotanice privind continutul unei gropi rituale apartinând grupului cultur­al Ceha¢lut – Hajdúbagos. Terra Sebus 4: 155–167. 2019. Archaeobotanical analyses on charred macrore-mains from Simleu Silvaniei “Observator” site (2015 ar­chaeological campaign). Acta Musei Porolissensis XLI: 43–52. Ciuta¢B., Molnár Z. 2014. Contributii la dieta comunita¢ti­lor Otomani din situl Carei-Bobald. Terra Sebus. Acta Mu-sei Sabesiensis 6: 87–99. Dietrich L. 2012. Die Gáva-Siedlung von Rotbav, Südost­siebenbürgen. Ein kurzer Vorbericht. Satu Mare. Studii si Comunica¢ri XXVIII(I): 101–108. 2014a. Die mittlere und späte Bronzezeit und die äl­tere Eisenzeit in Südostsiebenbürgen aufgrund der Siedlung von Rotbav. Habelt. Bonn. 2014b. Datele radiocarbon din asezarea apartinand epo­cii bronzului de la Rotbav (jud. Brasov, Romania). Stu-dii si Cerceta¢ri de Istorie Veche si Arheologie 65: 59– 70. Diffey C., Neef R., and Bogaard A. 2017. The archaeobo­tany of large-scale hermetic cereal storage at the Hittite capital of Hattusha. In A. Schachner (ed.), Innovation ver­sus Beharrung: Was macht den Unterschied des hethio­tischen Reich sim Anatolien des 2. Jt. v. Chr.? Interna­tionaler Workshop zu Ehren Jürgen Seeher. Ege Yayinla­ri. Istanbul: 185–201. Fairbairn A., Omura S. 2005. Archaeological identification and significance of ÉSAG (agricultural storage pits) at Ka­man-Kalahöyük, Central Anatolia. Anatolian Studies 55: 15–23. Laura Dietrich, Oliver Dietrich, and Julia Meister Gismondi, A., D’Agostino, A., Canuti, L., Di Marco, G., Ba-soli, F., Canini, A. 2019. Starch granules: a data collection of 40 food species. Plant Biosystems 153(2): 273–279. https://doi.org/10.1080/11263504.2018.1473523 Guma¢ M. 1993. Civilizatia primei epoci a fierului în sud-vestul României. Melior Trading S.R.L. Bucharest. Hrisrova I., Atanassova J., and Marinova E. 2017. Plant economy and vegetation of the Iron Age in Bulgaria: ar­chaeobotanical evidence from pit deposits. Archaeologi­cal and Anthropological Sciences 9: 1481–1494. https://doi.org/10.1007/s12520-016-0328-x Ivanova M., Schlütz F., and Benecke N. 2020. Subteranean silos at Vráble, southwestern Slovakia. A case study of geo­chemistry and distribution of finds in fill deposits. In A. Kozubová, E. Makarová, and M. Neumann (eds.), Ultra ve­lum temporis. venované Jozefovi Bátorovi k 70. narode­ninám. Slovenská archeológia. Supplementum 1. Nitra: 257– 271. https://doi.org/10.31577/slovarch.2020.suppl.1.21 Liu L., Wang J., Rosenberg D., Zhao H., Lengyel G., and Na-del D. 2018. Fermented beverage and food storage in 13,000 y-old stone mortars at Raqefet Cave, Israel: Inves­tigating Natufian ritual feasting. Journal of Archaeologi­cal Science. Reports 21: 783–793. https://doi.org/10.1016/j.jasrep.2018.08.008 Madella M., Alexandre A., and Ball T. 2005. International code for phytolith nomenclature 1.0. Annals of Botany 96: 253–260. http://dx.doi.org/10.1093/aob/mci172 Madella M., Lancelotti C. 2012. Taphonomy and phyto­liths: a user manual. Quaternary International 275: 76– 83. http://dx.doi.org/10.1016/j.quaint.2011.09.008 Marinova E., Valamoti S. M. 2014. Crop diversity and choi­ces in the prehistory of SE Europe: the archaeobotanical evidence from Greece and Bulgaria. In A. Chevalier, E. Marinova, and L. Pena-Cocharro (eds.), Plants and peo­ple: choices and diversity through time. Oxbow Books. Oxford: 64–74. Monah D. 2002. Découvertes de pains et de restes d’ali­ments céréaliers en Europe de l’Est et Centrale. Civilisa­tions 49: 67–76. Mulholland S. C., Rapp Jr. G. 1992. A morphological clas­sification of grass silica-bodies. In S. C. Mulholland, G. Rapp Jr. (eds.), Phytolith systematics – emerging issues, advances in archaeological and museum science. Sprin­ger Science & Business Media. New York: 65–89. Piperno D. R. 2006. Phytoliths: A comprehensive guide for archaeologists and paleoecologists. Altamira Press. Lanham. Portillo M., Kadowaki S., Nishiaki Y., and Albert R. M. 2014. Early Neolithic household behavior at Tell Seker al-Aheimar (Upper Khabur, Syria): A comparison to ethno-archaeological study of phytoliths and dung spherulites. Journal of Archaeological Science 42: 107–118. http://dx.doi.org/10.1016/j.jas.2013.10.038 Reynolds P. J. 1974. Experimental Iron Age storage pits: An interim report. Proceedings of the Prehistoric Society 40: 118–131. https://doi.org/10.1017/S0079497X00011348 Sigaut F. 1980. Significance of underground storage in traditional systems of grain production. In J. Shejbal (ed.), Controlled atmosphere storage of grains. Elsevier. Am­sterdam: 3–13. 1988. A method for identifying grain storage techni­ques and its application for European agricultural his­tory. Tools and Tillage VI(1): 3–46. Tsartsidou G., Lev-Yadun S., Albert R.-M., Miller-Rosen A., Efstratiou N., and Weiner S. 2007. The phytolith archae­ological record: strengths and weaknesses evaluated ba­sed on a quantitative modern reference collection from Greece. Journal of Archaeological Science 34: 1262– 1275. http://dx.doi.org/10.1016/j.jas.2006.10.017 Twiss P. C., Suess E., and Smith R. M. 1969. Morphologi­cal classification of grass phytoliths. Soil Science Society of America Journal 33: 109–115. http://dx.doi.org/10. 2136/sssaj1969.03615995003300010030x Urem-Kotsou D. 2017. Storage of food in the Neolithic com­munities of northern Greece. World Archaeology 49(1): 73–89. https://doi.org/10.1080/00438243.2016.1276853 Valamoti S. M., Marinovac E., Heiss A. G., +21 authors, and Kanceva Ruseva T. 2019. Prehistoric cereal foods of southeastern Europe: An archaeobotanical exploration. Journal of Archaeological Science 104: 97–113. https://doi.org/10.1016/j.jas.2018.11.004 Villers P., de Bruin T., and Navarro S. 2006. Development and applications of the hermetic storage technology. In I. Lorini, B. Bacaltchuk, H. Beckel, + 11 editors, and V. M. Scussel (eds.), Proceedings of the 9th International Work­ing Conference on Stored-Product Protection. Brazilian Post-harvest Association. Passo Fundo. Sao Paulo: 719–729. http://spiru.cgahr.ksu.edu/proj/iwcspp/pdf2/9/6119.pdf back to contents Documenta Praehistorica XLIX (2022) Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of Pontokomi-Souloukia Anna Stroulia1, Jérôme Robitaille2, Birgül Ögüt 3, Areti Chondroyianni-Metoki4, and Dimitra Kotsachristou5 astroulia@usi.edu 1 University of Southern Indiana, Evansville, US 2 Laboratory for Traceology, Controlled Experiments (TraCEr), MONREPOS, Neuwied, DE< jerome.robitaille@rgzm.de 3 Deutsches Archäologisches Institut Orient, Orient-Department, Berlin, DE< Birguel.Oeguet@dainst.de 4 Ephorate of Antiquities of Kozani, Kozani, GR< chondrogianniareti@hotmail.com 5 Ephorate of Antiquities of Kozani, Kozani, GR< dkotsach1982@gmail.com ABSTRACT – Despite their widespread presence and potential to shed light on various aspects of prehistoric life, for a long time Neolithic macrolithics attracted little scholarly attention. The situa­tion, however, is rapidly changing as more and more assemblages are being studied and published systematically. The study of the grinding and abrading tools from the earlier Neolithic site of Ponto­komi-Souloukia in northern Greece is part of this recent trend, as it integrates macroscopic exami­nation, use wear, microbotanical and macrobotanical analysis, an experimental program, ethnogra­phic data, as well as contextual analysis. In this article, we present the results of our study and make comparisons with other assemblages, placing the Pontokomi-Souloukia material in its wider Aegean Neolithic context. KEY WORDS – Neolithic; Greece; grinding and abrading tools; use wear analysis; residue analysis Mletje in bru[enje v zgodnjem neolitiku v severni Gr;iji> multi-proksi in komparativni pristop k najdi[;u Pontokomi-Souloukia IZVLE.EK – Kljub raz.irjenosti in sposobnosti osvetliti razli.ne vidike .ivljenja v prazgodovini, so neolitski makroliti pritegnili le malo raziskovalne pozornosti. Ocena se hitro spreminja, saj je siste­mati.no analiziranih in objavljenih vse ve. zbirov. Mednje sodi tudi .tudija orodij za mletje in bru-.enje iz zgodnje neolitskega naselja Pontokomi-Souloukia v severni Gr.iji, saj vklju.uje makroskop­sko analizo, analizo sledov uporabe, mikro in makro botani.ne analize, eksperimentalni program, etnografske podatke in kontekstualno analizo. V .lanku predstavljamo rezultate na.e .tudije in zbir primerjamo z drugimi. Gradivo z najdi..a Pontokomi-Souloukia tako ume..amo v .ir.i egejski neo­litski kontekst. KLJU.NE BESEDE – neolitik; Gr.ija; orodja za mletje in bru.enje; sledi uporabe; analize rezidijev DOI> 10.4312\dp.49.14 Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... Introduction Despite their ubiquitous presence at Aegean Neoli­thic sites, involvement in most (if not all) chaînes opératoires, and potential to illuminate various as­pects of prehistoric life, for a long time macrolithics1 attracted little scholarly attention. It is not an exag­geration to state that traditionally they represented one of the most neglected materials from Neolithic Greece. When relevant information was reported, it often consisted of a few cursory paragraphs in the ‘small finds’ section of a site publication (e.g., Evans 1964.229–231). Some tools, abandoned at the site after the completion of the excavation, were not considered worthy of even such a superficial treat­ment. This is the bad news. The good news is that the situation is rapidly chang­ing. In the last fifteen years or so, the field of Aegean Neolithic macrolithics has witnessed dramatic growth as more and more assemblages are being studied and published systematically. As a result, significant progress has been made in exploring raw materials, manufacturing processes, aspects of use, practices of discard, as well as social and symbolic dimensions (e.g., Almasidou 2019; Bekiaris 2007; 2018; 2020; Bekiaris et al. 2017; 2020; in press; Chadou 2011; Chon­drou 2018; 2020; Chondrou et al. 2018; 2021; Chondrou, Valamoti 2021; Lewis et al. 2009; 2011; Ninou 2006; Stergiou et al. 2022; Stroulia 2002; 2010a; 2010b; 2018a; 2018b; 2020; Stroulia, Chon­drou 2013; Stroulia et al. 2017; 2022; Tsoraki 2008; 2011a; 2011b; 2011c). Despite these developments, the field suffers from two se­rious imbalances. The first is geographic. Much more is known about the macrolithic and Thessaly compared to roughly 10 from the southern part of the country. The second imbalance is chronological. The available information for indu­stries dated to the later part of the Neolithic by far exceeds that for earlier materials; there are roughly twice as many reported Late or Final Neolithic as­semblages as those belonging to earlier phases (see Bekiaris et al. 2020.146–147; Stroulia in press).2 By focusing on the earlier Neolithic material from the site of Pontokomi Souloukia in the Kitrini Lim­ni Basin, in the prefecture of Kozani, west Macedo­nia, this article tackles none of the geographic bias but does address the chronological one. As such, (1) it sheds light on the macrolithic implements of the first agropastoral communities that occupied the Aegean; (2) it contributes to an understanding of the diachronic evolution of the macrolithic industries and related practices in this part of the world. The Site Kitrini Limni was a busy place in the Neolithic. As re­vealed by surface surveys or accidental discoveries related to various development projects, from the 7th to the 4th millennium BCE, this 35km2 basin Fig. 1. Map of prefecture of Kozani with industries of northern Greece locations of Pontokomi-Souloukia and than those from sites farther four other sites in the Kitrini Limni Ba- south; reports (of varying sin mentioned in the text. 1 Pontokomi- length and quality) are avail- Souloukia; 2 Mavropigi-Fillotsairi; 3 Klei- able for almost 30 assembla-tos; 4 Megalo Nisi Galanis; 5 Kremasti-Kilada. Kitrini Limni is indicated ges from Macedonia, Thrace, by dashed line. Graphics by Sofia Vlahopoulou. 1 For a discussion of the term ‘macrolithics’ and its advantages over the traditional term ‘ground stone’, see Adams et al. (2009. 43–44) and Stroulia (2018a.202). 2 The few assemblages that span the Neolithic period were not taken into account in these calculations. Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou served as the homeland for 30 settlements (Chond­royianni-Metoki 2020; in press a; b; c). One of the earliest among them is Pontokomi Souloukia (hence­forth Souloukia) (Fig. 1). Located on the western edge of the basin, Soulou­kia covers c. 1.0–1.2ha and dates to the second half of the 7th millennium and the beginning of the 6th. Roughly half of the site was severely damaged in the past few decades by the construction of a highway, a railroad, and other infrastructure-related projects. The other half (c. 0.4ha) was targeted by salvage ex­cavations in the context of large-scale coal mining operations (Chondroyianni-Metoki in press b; Ka­ramitrou-Mentessidi et al. 2010.39–46; Ziota et al. 2014.77–79). Carried out by the Ephorate of Antiquities of Kozani between 2010 and 2017, the excavations revealed two contiguous but distinct areas. The centre of the site has the form of a low tell with anthropogenic deposits reaching a maximum thickness of 2m. This represents the residential sector as indicated by the remains of successive post-framed buildings. Five (mostly infant) human burials and one animal bur­ial were found inside the buildings, while a concen­tration of 15 or so pits was excavated immediately to the south. A preliminary study of the stratigraphy and pottery of a single trench revealed three Early Neolithic building horizons and one dating to the early Middle Neolithic. Only c. 0.1ha was dug, but the residential area is estimated to have covered be­tween 0.25 and 0.5ha (Chondroyianni-Metoki in press b) (Fig. 2a–d). The surrounding flat area comprises the non-resi­dential sector and includes three types of features: ditches, pits, and clay structures. Two ditches were uncovered. The first – on the western edge of the site – was linear, measuring c. 55m in length, 2m in maximum width, and 2.1m in maximum depth (Chondroyianni-Metoki in press b). The second – on the site’s eastern part – was roughly curvilinear, measuring c. 14.5m in length, 2.4m in maximum width, and 1.4m in maximum depth (Karamitrou-Mentessidi et al. 2010.41–43, 45). The functions of the ditches remain enigmatic, as there is no evidence that they connected with each other or surrounded the site (Fig. 2e). Dispersed around the non-residential area are more than 40 pits of various sizes.3 Their contents consist of pottery, lithics, faunal material, and figurines, but generally speaking, they yielded a small number of finds (Chondroyianni-Metoki in press b; Karamit­rou-Mentessidi et al. 2010.43–44; Ziota et al. 2014. 78). It is tempting to interpret these features as con­tainers for ordinary waste disposal, but their disper­sal over a large area is certainly intriguing. East of the residential area (not far from the eastern ditch), the excavations uncovered a partially pre­served clay structure with ashes – probably the re­mains of a hearth. North of the residential area, a second clay structure was excavated. It was ellipti­cal in plan, with whitish clay coating the interior walls. Another structure may have existed west of the residential sector where masses of clay were un­covered along with, among others, a large number of grinding tools (Fig. 2f). Additional clay structures are vaguely mentioned in the preliminary reports (Chondroyianni-Metoki in press b; Karamitrou-Mentessidi et al. 2010.43; Ziota et al. 2014.78–79). Significantly, the bipartite settlement structure, with a tell-like residential centre and a surrounding flat, extra-residential periphery, recognized at Souloukia, does not characterize Mavropigi-Fillotsairi, the other extensively excavated Early Neolithic site of Kitrini Limni (Karamitrou-Mentessidi et al. 2013; 2015). It has, however, been identified at Kremasti-Kilada, which dates to the Late Neolithic (Chondroyianni-Metoki 2009; 2020). To the best of our knowledge, this settlement layout is not known from other parts of Greece and thus may represent a regional varia­tion. Whether regional or not, this bipartite configu­ration argues against the long-held, simplistic dicho­tomy between tell sites and flat/extended sites and underlines the diversity of the ways in which Aegean people organized their settlements in space during the Neolithic (see also Kotsakis 1999.69–70; Krah­topoulou 2019.77–82; Sarris et al. 2017; Toufexis 2017.23–30, 333–362). Materials and methods The Souloukia excavations uncovered large amounts of pottery, various quantities of stone and bone tools, an unusually high number of figurines, as well as a few rather uncommon artefacts (i.e. two clay house models, a marble vessel, and a bone flute) (Chondroyianni-Metoki in press a; b; Karamitrou-Mentessidi et al. 2010.44–45; Ziota et al. 2014.79). The stone tool inventory includes nearly 400 macro-lithics. Among them are the roughly 170 grinding 3 Intriguingly, a single pit dates to the Final Neolithic (Chondroyianni-Metoki in press b). Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... Fig. 2. Excavation views. Residential sector: a remains of burnt building; b postholes; c pits; d pot burial. Extra-residential sector: e ditch; f area with grinding tools, pieces of raw material, as well as masses of clay possibly from a clay structure. Photos by A. Chondroyianni-Metoki. tools, abrading implements, and related pieces of raw material that make up the focus of this paper. The term ‘grinding tools’ refers to implements used to pulverize or crush foodstuffs and minerals. These artefacts operate in pairs comprising a lower statio­nary component and an upper mobile one. In the li­terature, the first is often referred to as a millstone, quern, metate, grinding slab, etc., the second as a handstone, mano, grinder, rider, roller stone, rubber, etc. (e.g., Adams 2014.142–145; Cappers et al. 2016. 391–392; Elster 2003.186; Evans, Renfrew 1968.71; Hamon, Le Gall 2013.113; Hayden 1987.187; Tso­raki 2008.91, 97; Wright 1992.61; Wright, Baysal 2012.3). Here we use the more neutral terms ‘pas­sive tool’ and ‘active tool’, respectively. The term ‘abrading tools’ refers to implements used without a complementary component for shaping/maintain­ing other artefacts (e.g., celts, bone tools, and orna­ments) through abrasion. We should emphasize that the differentiation between grinding and abrading tools as well as that between passive and active grinding implements serve analytical purposes. As seen below, the Souloukia residents did not always conform to these distinctions. Since the systematic study of both the stratigraphy and pottery is pending, a distinction between Early Neolithic and early Middle Neolithic specimens has not been possible. However, given that three of the four building horizons in the residential area are Early Neolithic, we assume this to be the date of the majority of specimens. Be that as it may, in this ar­ticle all specimens are referred to collectively as ear­lier Neolithic. All but six of the roughly 130 abrading and grinding tools were subjected to use wear analysis with a ste­reoscope (10–80x magnification) and a metallogra­phic microscope (100x and 200x magnification) (Ro­ Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou bitaille). Preservation of microwear was moderate to low. Acetate and polyvinyl siloxane casts were made of used surfaces that were preserved reason­ably well. Use wear analysis was conducted in con­junction with an extensive experimental program that involved a variety of materials: cereals, pulses, nuts, bone, shell, stone, and wood (Robitaille, Strou­lia). Microwear was identified on a total of 63 speci­mens. Residue analysis following a protocol and nomen­clature defined by Rosa M. Albert et al. (1999) and Marco Madella et al. (2005) was carried out on 14 unwashed specimens: 13 grinding tools (both work and dorsal faces were sampled) and one abrading implement (both used surfaces were sampled). Six control samples were analysed as well. The detect­ed phytoliths were examined with a Leica DM 750 microscope at 400x magnification. For each sample, calculations were made for the number of phytoliths per 1g of sediment, the percentages of grass and di­cotyledonous phytoliths, as well as those of weather­ed and multicellular morphotypes. Phytoliths that can be associated to use with a reasonable degree of confidence were identified on only five of the sam­pled artefacts (Ögüt 2018). The Souloukia grinding and abrading tools are among the first from Neolithic Greece to have been subject­ed to use wear and residue analyses.4 In this paper, we present the general results of these analyses, but more detailed data on both, as well as the experi­mental program, will be published elsewhere. While the analysis of microbotanical remains has been completed, the examination of the macrobota­nical material has not. No more than a small sam­ple has so far been analysed (Kotsachristou). Only preliminary observations have therefore been pos­sible regarding the plants processed and consumed at the site. The study of the Souloukia grinding and abrading tools is particularly important. The severe underre­presentation of earlier Neolithic assemblages in the literature (noted above) is one reason. There are another two: these assemblages have been treated superficially or happen to be very small. For exam­ple, over 100 grinding and abrading tools from Achil­leion were presented in fewer than two pages (Winn, Shimabuku 1989.268–272), while the roughly 70 specimens from Nea Nikomedeia were discussed in only seven paragraphs (Pyke 1993.103, 108–109, 111). The macrolithic material from Prodromos, on the other hand, was studied systematically, but in­cludes fewer than 15 grinding and abrading tools (Moundrea 1975.92–99). By integrating macroscopic examination with use wear and residue analysis, experimental, macrobota­nical, and ethnographic data, as well as contextual analysis, our study of the substantial Souloukia as­semblage helps fill this gap and thereby clarify the role these implements played in the lives of the com­munities that made Greece their home in the earlier part of the Neolithic. This paper operates at three levels: (1) It presents the results of our multi-proxy study of the Soulou­kia tools by discussing the raw materials and their acquisition; the choices made in the context of ma­nufacture and the priorities that these reflect; the specimens’ morphometric and technofunctional cha­racteristics; the processed food and non-food sub­stances; as well as the tools’ spatial distribution and processes of discard. (2) It makes references to as­semblages from four sites in Kitrini Limni (Kremasti-Kilada, Kleitos, Megalo Nisi Galanis, and Mavropigi-Fillotsairi) as well as others elsewhere, placing the Souloukia material in both its regional and wider Aegean Neolithic context. (3) It utilizes the limited available information on contemporary industries and makes comparisons to later ones in an attempt to place the Souloukia material in its synchronic and diachronic framework. Abrading tools Only six specimens were securely identified as ab­rading tools. All derive from the residential sector. An additional specimen – found outside the residen­tial area – carries no use wear but may represent raw material intended for an abrading tool. The Souloukia abrading tools share two basic com­monalities: (1) all are a posteriori – the raw mate­rial was put directly to use without modification; (2) all were used passively. These tools, on the other hand, exhibit significant morphological and litholo­gical differences that allow a distinction between two groups. Group 1 comprises four specimens of tabular fine-grained sandstone. This type of raw material is not 4 For other studies, see Danai Chondrou et al. (2021) and Anna Stroulia et al. (2017.3–7). Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... found in the Kitrini Limni Basin, the bottom of which consists of clay marl and lacks stones larger than 5cm (Fotiadis 1988.45; Fotiadis et al. 2019.5–6). It is not found anywhere else in the local landscape either. The closest source is located on the slopes of Mt. Vourinos, close to the village of Agia Paraskevi, about 25km south of the site.5 The sandstone crops out naturally in the form of plaques of varying thick­ness (Stroulia, Dubreuil 2011.3). Only one tool in this group is complete. It is not how­ever intact, as it was found in two pieces (the break­age probably being post-depositional). This specimen is polygonal and measures 14.3x13.8x3.4cm. The other three are fragmentary. None appears to derive from a tool larger than the complete one. On this basis, it is fair to say that all specimens in Group 1 are small and thin. The extent and intensity of use, as well as the ensu­ing morphological changes, vary from tool to tool. The two thickest specimens – one complete, the other fragmentary – were used on both faces. The use was in­tensive enough to create con­cave configurations. The faces of the complete specimen are parallel, but one (A) is deeper than the other (B) (Fig. 3.a). Face A features a central ovate, concave c. 9x6cm area. Due to the presence of concretion, no microwear was identified on this face, but that detected on face B appears compatible with the abrasion of bone. This is the only abrading tool subjected to residue analysis. A relatively high proportion of wood phy­toliths were detected on face A. However, since many of them consist of skeletons rather than weathered morphotypes, they may not be use-related (Ögüt 2018). The faces of the frag­mentary specimen are diagonal (Fig. 3.b). Again, one face is de­eper than the other, but both show an increasing depth to­wards the thinnest edge. We were not able to specify the ma­terial that was processed on these faces, but according to the use wear analysis, it was neither bone nor wood. One of the thinnest specimens was also used on both faces. Judging by the fact that one face is more or less flat, while the other is lightly concave with an increasing depth from one side to the other, the two faces were not used with equal intensity. Use wear analysis points to wood processing, at least for one of the faces (Fig. 3.c). Finally, the fourth and equally thin specimen was used on only one face, resulting in the formation of a slightly concave area. The processed material was most likely bone. This is the least utilized of the Souloukia abrading tools (Fig. 3.d). Group 2 includes two tools. Both are much more massive than those making up Group 1. This is well illustrated by the complete specimen, which mea­sures c. 31x19x11cm (Fig. 4). The raw material con­sists of waterworn boulders, gneiss in one case and sandstone of a coarser variety than that used for 5 This is probably the only primary source of fine-grained sandstone in the prefecture of Kozani. Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou Group 1 in the other. In both tools, one of the faces is lightly convex but has a lightly concave used area measuring c. 15x8cm. In neither case has the proces­sed material been determined. Our general conclusion is that the Souloukia abrad­ing tools were involved in the production (and/or maintenance) of bone objects such as those that were excavated (see Karamitrou-Mentessidi et al. 2010.45) as well as wooden artefacts that were not preserved. It is important to note that bone or wood abrasion has been hypothesized for certain Greek Neolithic macrolithic tools on a macroscopic basis (e.g., Bekiaris 2018.276–277; Chondrou 2018.227– 228; Fotiadis et al. 2019.31; Stroulia 2010a.40– 54, 2018a.211–212; Tsoraki 2008.102–10), but this is the first time that such functions are documented for the Neolithic Aegean. Significantly, no wear related to stone abrasion was detected. This is unexpected in light of the over 100 excavated celts. We find it plausible that celt shap­ing and resharpening took place outside the settle­ment. Such a hypothesis is compatible with the al­most complete absence of unfinished specimens. On the basis of ethnographic evidence, both practical and non-practical considerations may have been be­hind the off-site production and maintenance of celts at Souloukia. Among several Irian Jaya groups, celt grinding is carried out away from the compounds, usually by a river or stream. This activity is invested with a strict prohibition along gender lines as it must take place out of view of women (Pétrequin, Pétre-quin 1993.373). Abrading tools of tabular fine-grained sandstone have been uncovered at three other Kitrini Limni sites: Kremasti-Kilada (Chondrou 2011.101–102; Stroulia, Dubreuil 2011.1), Megalo Nisi Galanis (Fo­tiadis et al. 2019.31), and Kleitos (Chondrou 2018. 200–230). Similar artefacts have also been report­ed from Servia,6 a site in the prefecture of Kozani but not in Kitrini Limni (Mould et al. 2000.155–157). Microscopic analysis of a sample from Kremasti-Kila­da by Laure Dubreuil revealed use wear somewhat compatible to that produced experimentally through stone abrasion and scraping unfired bone-dry clay vessels (Stroulia, Dubreuil 2011.2). Kremasti-Kila­da, Megalo Nisi Galanis, Kleitos, and Servia date to the Middle, Late, or Final Neolithic and are thus later than Souloukia, but the presence of such tools on all five sites points to a certain regional tradition of ex­ 6 They are referred to as ‘palettes’ by the excavators. ploiting fine sandstone tabular pieces from the same source for a variety of abrading purposes. Finally, we should note that farther north and west, but also in Macedonia, the Late Neolithic site of Avgi yield­ed over 60 tabular pieces of fine-grained sandstone. They are of generally larger dimensions than those found at the above sites and were employed in a cooking rather than an abrading context (Bekiaris et al. in press). Grinding Tools Raw material type, procurement The Souloukia excavations yielded a much higher number of grinding than abrading tools. One hun­dred twenty-six specimens were securely identified as grinding tools, while one and possibly up to four specimens represent roughouts (Figs. 5–9). In addi­tion, 31 pieces of gravel were recovered of material similar to that employed for grinding tools but with­out traces of manufacture or use. At least 20 of these are complete or substantially preserved and thus li­kely represent unworked nodules intended for grind­ing tools. Of the remaining specimens, some are very fragmentary, while others have surfaces covered by concretion or altered by fire. Whether these belong to tools or raw nodules is impossible to tell. If tabular sandstone was the preferred material for abrading implements, it was used rarely for grinding Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... tools. For the latter, Soulou­kiotes almost always chose gneiss in the form of cobbles and boulders. In this sense, the grinding tool assemblage is remarkably homogeneous, a reflection of a deeply em­bedded tradition that span­ned several generations. The gneiss used comes in various degrees of coarseness, is of­ten oxidized with a characte­ristic red/brown colour, has a high quartz content and is thus quite hard, typically in­cludes no mica, and its surface is usually anomalous and/or has vesicular areas (Figs. 5– 9). As we found out by expe­rimentally producing a work face through pecking, this type of gneiss is characteriz­ed by high workability. Mo­reover, because of its hard­ness it does not require fre­quent resharpening, nor does it produce much grit during use. Or so we discovered through our grinding and co­oking experiments. This was a good choice. Primary gneiss sources are found on two of the moun­tains surrounding Kitrini Lim­ni: Mt. Askio to the west and Mt. Vermio to the east (Fig. 1). Both sources are extensive, but Mt. Askio is the closest to Souloukia. Although no petrographic analyses have been con­ducted, macroscopic similarities indicate that as a rule Mt. Askio gneiss was utilized by the Souloukio­tes. Nevertheless, with a handful of possible excep­tions procurement did not take place at primary sources. The material is waterworn and thus must have been collected at more proximate secondary locations. A comparison between Souloukia and other Kitrini Limni sites regarding the raw materials of grinding tools revealed one fundamental similarity. In all ca­ses, secondary sources were exploited (Kremasti-Ki­lada: Chondrou 2011.81, 106; Stroulia, Dubreuil 2011.1; Kleitos: Chondrou 2020.291; Megalo Nisi Galanis: Stroulia 2002.576). This comparison also revealed two patterned varia­tions: . The almost exclusive focus on a singular material noted at Souloukia is not paralleled in the later Ne-lithic assemblages of Kremasti-Kilada (Chondrou 2011.80–81; Stroulia et al. 2017.3), Kleitos (Chon­drou 2020.290–291), and Megalo Nisi Galanis (Fo­tiadis et al. 2019.30–31), which are characterized by a variety of materials. Gneiss is part of this vari­ety but never the dominant lithology. On this basis and in a preliminary fashion, we would like to hy­pothesize that in Kitrini Limni an earlier Neolithic focus on a single material was followed by the ex­ploitation of diverse lithologies. What needs and/or opportunities could have led to such a diversifica­tion cannot be systematically discussed with the available data. However, a simplistic and straightfor­ Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou ward equation between the number of raw material types and processed substances does not appear to be the answer. As seen below, the use wear of the Souloukia specimens points to processing a variety of substances. Be that as it may, the nearly exclusive use of a single material for grind­ing tools appears to have been rare in Neolithic Greece. To the best of our knowledge, the only other sites in which this practice has been documented are Dikili Tash (Bekiaris et al. 2020.147) and Alepotrypa Ca­ve (Stroulia 2018a.205), in the northern and southern parts of the country, respecti­vely. Both date to the later part of the Neolithic. . The Mt. Askio gneiss used at Souloukia is macroscopical­ly different from the varieties employed at Kremasti Kilada and Megalo Nisi Galanis (ob­servations by Stroulia) that appear to originate in Mt. Ver-Fig 6. Passive tools: a complete specimen GS277 with concave/concave work face and dorsal face used in the context of recycling; b concave/ mio. Notably, these two sites concave work face and longitudinal profile of fragmentary specimen are located in the eastern part GS255; c concave/concave work face and dorsal face of nearly complete of the basin and thus closer specimen GS324; d concave/concave work face and transverse section to Mt. Vermio than Mt. Askio. of fragmentary specimen GS388. Drawing by T. Gouliafas; photos by A. On this basis, we would argue Stroulia. that western and eastern Kitri­ni Limni communities exploited different regional sources of gneiss. Since all these varieties are of good quality, the distance from the sources may very well have been the determining factor behind these choi­ces. Hopefully, these hypotheses will be tested in the future through comparative petrographic analyses. Manufacture As mentioned above, the Souloukia excavations un­covered at least 20 unmodified gneiss cobbles and boulders that were probably intended for grinding tools. Their presence is indicative of two practices: (1) raw material was brought to the site without prior processing at the sources; (2) larger quantities of raw materials than those immediately needed were periodically collected in anticipation of future needs – the hallmark of curation practices. With the available evidence, it is impossible to tell how the collected raw material was distributed, but the recovery of raw nodules from both the residential and non-residential areas at the very least points to the lack of a single communal spot where raw ma­terial was kept awaiting future use. The fact, more­over, that only one roughout was positively identi­fied (Fig. 5.a) indicates that no designated manu­facturing locus existed either (at least in the exca­vated area). The scarcity of unfinished specimens noted at Sou-loukia matches that known from other Kitrini Lim­ni sites; see Kremasti Kilada (Chondrou 2011.81–82, 134; Stroulia et al. 2017.4), Kleitos (Chondrou 2020. 291), and Megalo Nisi Galanis (observation by Strou­lia). No such match applies to unworked specimens. While by no means high, the number of such speci­mens at Souloukia is rather substantial when com­ Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... pared to Kleitos (Chondrou 2020.291) and Megalo Nisi Galanis (Stroulia 2002.576),7 which are charac­terized by an almost complete absence of unmodi­fied cobbles and boulders. Indeed, the combined paucity of raw nodules and roughouts in the huge assemblage of the almost fully excavated site of Klei­tos led to the hypothesis that tools arrived in a more or less finished state (Chondrou et al. 2018.31; Chondrou 2020.291). A similar scarcity character­izes assemblages from sites beyond Kitrini Limni such as Makriyalos (Tsoraki 2008.81), Alepotrypa Cave (Stroulia 2018a.234), Avgi (Bekiaris 2018. 221), and Platia Magoula Zarkou (Stroulia in press). This pattern deserves systematic investigation, but it appears to suggest a widespread practice of off-site grinding tool production in Neolithic Greece. Two manufacturing techniques were employed for grinding tools at Souloukia: pecking and flaking (Figs. 5.a, 9.a). Both are known from other sites in Kitrini Limni and elsewhere; see Kremasti-Kilada (Chondrou 2011.82–83, 106–107; Stroulia, Dubre­uil 2011), Megalo Nisi Galanis (Stroulia 2002.576), Kleitos (Chondrou 2020.291–293), Avgi (Bekiaris 2020.4), Makriyalos (Tsoraki 2008.114), Koroneia (Almasidou 2019.90, 99–100), Franchthi Cave (Stro­ulia 2010a.35), Alepotrypa Cave (Stroulia 2018a. 206), and Platia Magoula Zarkou (Stroulia in press). The process of manufacture at Souloukia was nei­ther comprehensive nor systematic: work faces were most often created through pecking. Portions not intended for use, on the other hand, were as a rule left in their raw state or received localized treatment in order to facilitate the tool’s gripping/resting or ensure a specific plan (Figs. 5.b, 6.a and c, 9.a–c). This manufacturing approach was aided by a raw material acquisition strategy that favoured cobbles/ boulders with sizes and shapes similar to those of the intended tools or with portions that could be strategically incorporated into the final tool shape. Two examples: the first is a roughout of a passive tool measuring c. 33x23x5.5cm., with one naturally flattish surface that could be converted into a work face with minimal pecking (Fig. 5.a); the second example – an active tool – measures c. 20x19x 17.5cm, pointing to use with two hands. Its dorsal face has a naturally ridged shape that must have fa­cilitated gripping during grinding (Fig. 8.c). The above strategy expedited manufacture, it did result, however, in tools with uneven/anomalous dorsal and peripheral surfaces or asymmetrical plans (Figs. 5.b, 6.a and c, 8.c, 9.c). Clearly, this was not considered a sufficiently serious problem to make the people of Souloukia invest more effort in tool making. Clearly, saving time and energy took prio­rity over appearance. A similar attitude is reflected in the assemblages of other Kitrini Limni sites and 7 The case of Kremasti-Kilada is ambiguous. More than 300 fragmentary specimens without traces of manufacture or use were excavated, but how many represent unworked raw materials and how many consist of unmodified tool portions we cannot tell (Stroulia et al. 2017.23). Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou beyond, which are also characterized by a low ma­nufacturing investment and/or non-comprehensive treatment of the raw materials – see Megalo Nisi Ga-lanis (Stroulia 2002.576), Kleitos (Chondrou 2020. 291), Kremasti-Kilada (Chondrou 2011.82–83, 106– 107; Stroulia et al. 2017.4), and Makriyalos (Tsora­ki 2008.114). Morphometric characteristics Only 30 (24%) of the Souloukia grinding tools are complete or nearly so. The vast majority (n=96) are fragmentary. This imbalance is far from remarkable, since Aegean Neolithic grinding tool assemblages (at least those with available preservation information) are dominated by fragments (see Bekiaris et al. 2020.157–158; Stroulia in press). What is notewor­thy is the stark contrast between the frequencies of complete specimens at Souloukia and other Kitrini Limni sites. As it turns out, at Kremasti-Kilada, Klei­tos, and Megalo Nisi Galanis, such specimens are ex­tremely rare, accounting for <4%, 3%, and <1% of the total, respectively (Chondrou 2011.91–95, 111– 113, 164–165; 2020.290; Fotiadis et al. 2019.31; Stroulia 2002.576; Stroulia, Chondrou 2013.125– 126; Stroulia et al. 2017.3). The hypothesis of de­liberate breakage has been put forward for all three assemblages. The fact that the Souloukia material is earlier than the other three raises the possibility that grinding tools were subjected to different treatments and assigned different dimensions by earlier and la­ter Neolithic communities in Kitrini Limni. On the basis of microwear analysis, morphometric characteristics and/or work face configuration, 63 tools (50%) were identified as passive, 47 (37%) as active. All identifications refer to primary uses. Due to fragmentation or surface alteration, it has been impossible to determine whether the remaining 16 specimens were used passively or actively. The numerical prevalence of passive tools is intri­guing. As known from ethnographic sources, active tools wear out faster and thus have shorter life spans than passive ones (see Delgado Raack, Risch 2016. 129; Hayden 1987.193; Nixon-Darcus, D’Andrea 2017.206; Risch 2008.22; Robitaille 2016. 438). That is why, for example, among the Minyanka of Mali two active tools are produced for each passive one (Hamon, Le Gall 2013.112), while among the Kon-so, Hamar, Mursi, and Dorze of Ethiopia each passive tool is used with two active ones over its life­span (Robitaille 2016.445; 2021. 240, 546). On this basis, one would expect the Souloukia as­semblage to feature precisely the reverse imbalance, i.e., a higher proportion of active than passive tools. What is even more unexpected, similar discrepancies between passive and active tools have been noted at other Greek sites such as Alepotrypa Cave (Strou­lia 2018a.208), Franchthi Cave (Stroulia 2010a.79–94), Ilioto­pos (Chadou 2011.134), and Ma-kriyalos (Tsoraki 2008.Tab. 5.28). They are also known from other periods or countries; see, for example, several Neolithic sites in Serbia (Galdikas 1988. Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... 341; Vu.kovi. 2019. 228), the Greek Bronze Age site of Toumba Thessalonikis (Tsiolaki 2009.61–62), certain Pre-Dynastic and Bronze Age Egyptian sites (Robitaille 2015; Samuel 2010.466), as well as Bronze Age sites in southern Iberia (Delgado Raack, Risch 2009.10; 2016.129; Risch 2002.111–127). We suspect that this is a widespread phenomenon, one that has been severely underappreciated in the lite­rature. Risch and his colleagues attributed the predomi­nance of passive implements in southern Iberian as­semblages to the use of unpreserved wooden active specimens (Delgado Raack, Risch 2009.17; 2016. 139; Menasanch et al. 2002.108; Risch 2002.111– 127). Such a hypothesis does not appear valid for Souloukia. The microtopography of the work faces of passive tools points to active counterparts made of stone. There must be another explanation, and this will be discussed later in this work. The Souloukia passive tools are typically elongated. Their plans are subrectangular, sub-trapeze, or ovate, with occa­sional subsquare or elliptical cases (Figs. 5–7). Active tools are also elongated, with sub-rectangular, ovate, and, more rarely, subtrapeze, subsquare, or subtriangular plans (Figs. 8–9). The eleven complete (or near­ly so) passive specimens range from c. 23 to 45.5cm in le­ngth, from c. 15.5 to 28.5cm in width, and from c. 4.5 to 17.5cm in thickness, averag­ing c. 31cm, 19cm, and 7.7cm, respectively. The 52 frag­ments average c. 23.5cm in length, 17cm in width, and 7.5cm in thickness. The aver­age length of the complete specimens is slightly over the standard limit of 30cm between small and large spe­cimens. This may lead to the general conclusion that the Souloukia passive tools were of moderate size. However, the average length of the many more incomplete specimens is relatively high – an indication that most tools were originally large. This hypothesis is reinforced by the fact that quite a few of the fragments represent half or less of the complete tool, as well as by the high mean width and thickness of the fragmentary specimens in gene­ral. On this basis, we would argue that the Soulou­kia assemblage includes several small (mostly com­plete) passive specimens, but large ones make up the majority (Figs. 5–7). The picture conveyed by the active tools is more straightforward. The 17 complete specimens are clearly large. They range from c. 16 to 30cm in length, from c. 10.5 to 19cm in width, and from c. 4.5 to 9cm in thickness, averaging c. 23cm, 15cm, and 6cm, respectively. With respective averages of c. 15cm, 12cm, and 6cm, the 30 fragments appear to derive from equally big tools. With one possible Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou exception, we consider all active specimens to have been two-handed (Figs. 8–9). The combined metric data of all specimens (passive and active, complete and incomplete) suggest that the grinding toolkits employed by the residents of Souloukia were for the most part large. The active tools could have been used in conjunc­tion with large passive implements such as those found. Active specimens that would have been com­patible with the identified small passive ones, on the other hand, appear to be missing. This is a sec­ond discrepancy between passive and active tools. Given their large sizes, it is tempting to assume that the active implements operated in an overhanging manner. However, both macroscopic and microsco­pic examination suggest that, as a rule, their length was roughly similar to or slightly higher than the width of the associated passive tools. Specimens whose length substantially exceeded the width of their passive counterparts are not common in this assemblage. Overhanging specimens have been re­ported from Kleitos (in Kitrini Limni), Ayios Vlasis, Stavroupoli, and Dikili Tash (Chondrou 2020.293– 294; Chondrou et al. 2021.6–9). It is, however, un­clear whether these tools were slightly or substan­tially longer than the width of the associated passive tools. Given that a number of passive and active speci­mens weigh over 7kg and 4kg, respectively, it is also tempting to assume that at least some of the Souloukia grinding toolkits were fixed in place. How­ever, no grinding installations were uncovered by the excavators. In fact, such an assumption projects to the past our Western modern relationship with heavy objects. The relationship of prehistoric peo­ple with such objects may have been different, as illustrated by ethnographic examples from Ethiopia. Among the Dorze, passive tools weighing 11–25kg are regularly moved from their storage location in­side the house to the yard of the compound where they are used. The Hamar passive tools range in weight from 8 to 60kg. The heaviest among them remain inside the house, but the lighter ones are often moved between indoor and outdoor areas. In most cases, tools are transferred on a daily basis. Other movements are periodical, depending on the season, the substances to be processed, or the con­text of use of the ground product. With certain ex­ceptions, the tools do not leave the boundaries of the compound. The exceptions refer to special occa­sions. Among the Mursi and the Hamar, tools are taken to farther locations for processing large quan­tities of grain in a group context. The processed grain is then used to make beer for weddings or other fes­tive events (Robitaille 2021.181; see also Hamon, Le Gall 2013.117). We close this section with the overall sizes of Neoli­thic Aegean grinding tools. General claims have been made about “the predominance of milling tools of relatively limited dimensions in many…sites” (Chondrou et al. 2018.37; see also Valamoti et al. 2013.171, 184). We disagree with this characteriza­tion and consider it to be a misconception shaped by high rates of fragmentation. Our examination of the sizes of both complete and fragmentary speci­mens has revealed a more nuanced picture. Some Neolithic assemblages include both small and large specimens; see Servia (Mould et al. 2000.146–155), Stavroupoli (Alisoy 2002), and Platia Magoula Zar­kou (Stroulia in press). Others comprise primarily small tools; see Franchthi Cave (Runnels 1981.101; Stroulia 2010a.37–38) and Lerna (Banks 2015.184; Runnels 1981.101). Yet others, like Souloukia, ap­pear to include primarily large specimens; see Kre­masti-Kilada (Stroulia et al. 2017.4; observation by Stroulia) and Alepotrypa Cave (Stroulia 2018a.204, 208–209).8 With its earlier Neolithic date, the Souloukia assem­blage demonstrates that the use of large-sized grind­ing tools was not a later development. Rather, such implements were a part of the material culture of the first sedentary communities established in the Aegean. What is more, the four largest specimens from Souloukia (Fig. 5.b–c)9 are among the most massive known from Neolithic Greece (for examples from Kitrini Limni and elsewhere, see Chondrou 2011.83–84; Mould et al. 2000.150; Stroulia et al. 2017.4; Touloumis 2002.108–109). Not only did early Aegean farmers have grinding tools of large dimensions, they had some of the largest ones yet found. Morphofunctional characteristics The Souloukia tools typically have only one work face. There are nine exceptions to this rule. Two pas­sive (3%), five active (11%), and two indeterminate specimens have two parallel or diagonal work faces 8 See also Bekiaris et al. 2020.154. 9 They measure 45.5x21x17.5cm, 42x28.5x20cm, 41x28.5x9cm, and 40x19.5x7cm. Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... with similar or different configurations. These num­bers do not include tools with localized use wear on the dorsal face (see below). The scarcity of dou­ble work-face specimens indicates that creating two work faces was not a common means for prolonging a tool’s use life at Souloukia. The higher percentage of active tools with two work faces, on the other hand, suggests that these were considered more ap­propriate for use with both faces than passive ones. Regarding other Kitrini Limni assemblages, the si­tuation varies. Double work-face tools are rare at Kleitos (Chondrou 2020.299), but more common at both Kremasti-Kilada (Chondrou 2011.74–113; Stroulia et al. 2017.4) and Megalo Nisi Galanis (Stroulia 2002.576). Since all sites are found in the same region, the variation cannot be attributed to differences in raw material availability and accessi­bility, and instead possibly reflects individual or cul­tural preferences (see also Robitaille 2021.815). Whatever the case, a similar variation characterizes the Aegean as a whole. For example, specimens with two work faces account for roughly a quarter of the assemblage at Makriyalos (Tsoraki 2008.91, Tabs. 5.14 and 5.24), but close to 60% at Alepotrypa Cave (Stroulia 2018a.207, 209). There is a constant in the midst of this variation, however. Among tools with two work faces, active ones always represent the majority. As documented by use wear analysis, the Souloukia grinding tools functioned in a reciprocal fashion. None was used in a circular/elliptical manner, even though such a suspicion was initially raised for a couple of roughly square or elliptical passive tools whose work face is concave along both axes. Use wear analysis, macroscopic examination, and/or morphometric characteristics of specimens from sites, such as Kremasti-Kilada (Chondrou 2011.95– 96; Stroulia et al. 2017.4), Kleitos (Chondrou et al. 2018.31; Chondrou 2020.293), Avgi (Bekiaris 2018. 230, 243), Makri (Bekiaris 2007.45), Makriyalos (Tsoraki 2008.98–100), Koroneia (Almasidou 2010. 100), and Franchthi Cave (Stroulia 2010a.40–46), indicate that use in a back and forth manner was typical not only in Kitrini Limni but Greece in gene­ral (see also Bekiaris et al. 2020.143–144). Due to fragmentation, it has not been possible to se­curely identify the configuration of the work faces of all the Souloukia specimens. That said, a variety of configurations have been identified. Opposed work faces may or may not have the same shape. 10 For some Copper Age Iberial parallels, see Risch 2008.20. Most commonly, the work faces of passive tools are concave both longitudinally and transversally (or concave/concave). At least 30 such specimens (48%) were identified (Figs. 5.b, 6–7). In this respect, the Souloukia assemblage appears to be atypical. No other assemblage from Kitrini Limni exhibits a pre­ponderance of concave/concave passive tools; see Kleitos (Chondrou 2020.293–294), Kremasti-Kilada (Chondrou 2011.85–88; observation by Stroulia), and Megalo Nisi Galanis (observation by Stroulia). With a couple of exceptions, the same is true for as­semblages from other sites, such as Franchthi Cave (Stroulia 2010a), Makriyalos (Tsoraki 2008.99), Ilio­topos (Chadou 2011.73), Platia Magoula Zarkou (Stroulia in press), Dispilio (Ninou 2006.28–56), and Apsalos (Ninou 2006.72–90). The exceptions re­fer to the assemblages of Avgi (Bekiaris 2018.228) and Alepotrypa Cave (Stroulia 2018a.207). Nineteen of the Souloukia passive tools (30%) have a work face that is concave longitudinally but con­vex transversally (or concave/convex) (Fig. 5.c–d). This is the second most frequent configuration among passive specimens. Both concave and con­vex curvatures can be only slight. Concave/convex specimens tend to be larger than concave/concave ones. Regarding other Kitrini Limni sites, con-cave/convex passive tools represent the majority at Kleitos (Chondrou 2020.293–294) but are rare at Megalo Nisi Galanis (observation by Stroulia). Like­wise, in the Aegean in general such tools are com­mon at some sites (see Platia Magoula Zarkou: Strou­lia in press), less common at others (see Avgi: Be-kiaris 2018.228), and nearly absent at others still (see Makriyalos: Tsoraki 2008.Tab. 4.39). Lastly, four of the Souloukia passive tools have work faces that are convex along both axes (or convex/ convex). This is an odd configuration and we can only hypothesize that these specimens were used a posteriori, the convexity representing the natural shape of the raw material.1 0 According to established typologies as well as ethno­graphic and experimental data, passive tools with concave/concave work faces are compatible with active tools whose work faces are convex/convex. Passive tools with concave/convex faces, on the other hand, are compatible with active tools that are also concave/convex (e.g., Delgado Raack, Risch 2009.7; 2016; Lidström Holmberg 2004.213; Risch 2008.20; Robitaille 2016.443; Stroulia et al. 2017. 19). Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou Given the higher ratio of concave/concave vs. con-cave/convex passive tools at Souloukia, one would expect a concomitant higher ratio of convex/con­vex vs. concave/convex active specimens. Yet this is not the case: concave/convex active specimens sur­pass convex/convex ones by a ratio of 2.5 to 1. There are, moreover, three active tools with flat/flat work faces. These would have been compatible with pas­sive flat/flat tools. No such specimens have been identified, however. The discrepancy between pas­sive and active tools regarding work face configura­tions is as intriguing as those mentioned above with respect to numbers and sizes. More about this later, but it is worth noting that a discrepancy regarding work face shapes has also been noted in the assem­blage of Alepotrypa Cave (Stroulia 2018a.209). We close this section with two more morphofunctio­nal features referring to active and passive tools, res­pectively. . A small number of active tools are characterized by a wedge-like transverse section – the result of dif­ferential wear between the proximal and distal sides (Fig. 9.c). A similar configuration has been noted at Kremasti-Kilada (Chondrou 2011.108; Stroulia et al. 2017.4) and Kleitos (Chondrou 2020.295). In the literature, this uneven wear has been mainly inter­preted as the result of application of extra pressure on the proximal side of the tool during grinding (Adams 2014.114; Bartlett 1933.11–16; Stroulia et al. 2017.17–18).1 1 To avoid the negative effect of this unevenness on the tools’ use lives, Hopi grin­ders traditionally employed a specific wear manage­ment strategy as they periodically rotated their ‘ma-nos’ so that the proximal side became the distal one, and vice versa (Bartlett 1933.15–16). Clearly, such a strategy was not used for the wedge-like Souloukia active tools. Yet it may have been popular among the Souloukia grinders, or so is suggested by the lack of a wedge-like configuration among the majority of active tools. . The work face of a handful of passive tools exhi­bits a very strong longitudinal angle (Figs. 5.d, 7). We assume that highly inclined gravels were delibe­rately selected. The rationale behind this choice re­mains elusive, however. We are not aware of such passive specimens from other Greek Neolithic sites, and if they exist, they are rare. Processed materials Before discussing the substances processed with grinding tools at Souloukia, we should note the fol­lowing: . Although almost all specimens were subjected to traceological analysis, microwear was identified on roughly half (n=59 or 47%). The microwear discus­sed in this section resulted from primary uses (for that pertaining to secondary functions, see next sec­tion). . The series of experiments conducted in conjunc­tion with use wear analysis involved: grinding free threshing wheat (both dry and parched); dehusk­ing emmer wheat and subsequently grinding the clean grain; dehusking hulled barley and subse­quently grinding the clean grain; grinding lentils (both dry and soaked); grinding dry and parched chickpeas; grinding acorns. All these experiments were carried out with the same gneiss grinding tool­kit. The last two experiments – abrading a piece of stone and a bovine femur – involved the passive component of the toolkit only. . Thirteen grinding tools (six passive, six active, and one indeterminate) were subjected to residue analysis. In eight cases, very few or no phytoliths were detected (Ögüt 2018). A low amount of phy­toliths may be the result of prehistoric cleaning or a non-plant related use. The first hypothesis is like­ly for five of these specimens, which according to use wear analysis were used for plant processing. The second hypothesis is plausible for two speci­mens with use wear associated with an unspecified abrasive but flexible material. None of these hypo­theses could be evaluated for the eighth specimen, whose use wear remains undetermined. . The excavations yielded a large quantity of char­red macrobotanical remains, only a small sample of which has so far been analysed. The sample is do­minated by cereal remains – a reflection of the im­portance of cereal cultivation, processing, and con­sumption at the site. Four cereal varieties have been identified: einkorn wheat (Triticum monococcum), emmer wheat (Triticum dicoccum), ‘new’ glume wheat type, and hulled barley (Hordeum vulgare). A very high proportion of the assemblage consists of glume wheat chaff, presumably the by-product of de-husking. Whether the chaff was burnt in the context 11 But see Chondrou (2020.295), who considers the wedge-like transverse sections of active tools at Kleitos as the result of specific raw material choices and/or design. Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... Fig. 10. Use wear views of work faces of specific tools at 100x and 200x magnifications (metallographic microscope): a GS252, use wear that appears consistent with cereal grinding; b GS268, use wear that appears consistent with cereal grinding; c GS292, use wear that appears consistent with grinding of wet (possibly soaked) lentils; d GS288, use wear resulting from processing an undetermined flexible but abrasive material; e GS285, use wear that appears consistent with grinding dry dirt or clay. Photos by J. Robitaille and A. Stroulia. of accidents or utilized as fuel is impossible to tell. Pulses are present in low proportions, with lentils (Lens culinaris) being the most common. Fruits, such as cornelian cherry (Cornus mas), are also re­presented in very small amounts, while wild/weed species (Chenopodium sp. and Polygonum avicula-re) occur very sporadically. The cereal dominance noted among macrobotanical remains is mirrored in the results of use wear ana­lysis. Over 75% of tools with identifiable microwear show traces compatible with processing cereals – dry emmer and barley, to be more specific (Fig. 10.a and b). Five of these specimens were subjected to re­sidue analysis. Of these, the phytoliths found on the Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou work faces of four also suggest a cereal-related func­tion (Fig. 11). However, differences in the propor­tions of weathered and multicellular morphotypes lead us to suggest that two were used for grinding, while the other two may have been used for de­husking.1 2 The morphotypes found on the work face of the fifth specimen were too weathered to allow a determination of the processed plants (Ögüt 2018). At any rate, it appears that cereals were processed on passive tools with work faces of all configurations (concave/concave, concave/convex, and convex/con­vex). The active tools used for these tasks are both concave/convex and convex/convex. Clearly, there is no association between cereal grinding and a spe­cific toolkit type. Judging from the results of traceological analysis of specimens from other sites – e.g., Kremasti-Kilada, Kleitos, Ayios Vlasis, Stavroupoli, and Dikili Tash (Chondrou et al. 2021; Stroulia, Dubreuil 2011.3) – cereal processing may have been the dominant function of grinding tools in both Kitrini Limni and the Aegean as a whole. As a rule, clean grain was ground. Only a few specimens from Ayios Vlasis, Stavroupoli, and Dikili Tash carry evidence of hulled grain grinding. However, an association with de-husking was proposed with a reasonable degree of confidence for only a couple of specimens from Ayios Vlasis. In all other cases it was not possible to as­sess whether the end goal of grinding was dehusk­ing or the production of a fibrous meal that may or may not have been later subjected to some kind of cleaning (Chondrou et al. 2021.7–9; see also Proco­piou 2003.23–33). Either way, the available data suggest that in Neolithic Greece grinding tools were not typically used for dehusking. Given the perva­siveness of chaff (Valamoti 2010) and the paucity of (suitable) stone mortars (Bekiaris et al. 2020. 144; Stroulia 2020.5), it can be assumed that this task was carried out by pounding grains on wood­en mortars or with other ethnographically known methods that would leave no archaeological signa­ture under ordinary taphonomic conditions (see Da­vid 1998.25–28; D’Andrea, Mitiku 2002.204; Hil­lman 1984.129–131; Pena-Chocarro, Zapata 2003. 107–110; 2014.230–231; Robitaille 2021.241–242). The use wear of a handful of Souloukia specimens appears compatible with that produced by the ex­perimental grinding of wet (probably soaked) lentils (Fig. 10.c). Four are active with concave/convex or convex/convex work faces, and one is passive with a concave/concave work face, suggesting a lack of differentiation between the toolkits employed for ce­real and pulse processing. While known for some time for Bronze Age Cycla­des (Sarpaki 2001.32), legume grinding was docu­mented only recently for prior time frameworks through use wear analysis of later Neolithic tools (Chondrou et al. 2021.2). The Souloukia findings now extend Aegean pulse flour production farther back, to the earlier part of the Neolithic, as do the new findings from the neighbouring site of Mavro­pigi-Fillotsairi (Ninou forthcoming). Use wear analysis of four tools points to grinding a flexible but abrasive material (Fig. 10.d). Residue analysis carried out on two of these specimens de­tected a minor amount of phytoliths, raising the pos­sibility of a use unrelated to plants (Ögüt 2018). All tools are passive, with work faces that are concave/ concave or concave/convex. What was processed on 12 Regarding dehusking, the results of use wear analysis were inconclusive. The possibility that the same work faces were used for both grinding and dehusking cannot be ruled out. Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... these tools will hopefully be clarified through fur­ther experimentation. No use wear compatible with acorn processing has been detected on the Souloukia tools. Danai Chon­drou et al. (2021.5) reported Greek Neolithic speci­mens with microwear consistent with processing of ‘greasy’ substances, but whether the term refers to nuts, oily seeds, or both has not been clarified. The Souloukia tools yielded no evidence of pigment processing either. In this sense they appear to follow a more general pattern. The association of grinding tools with pigment is relatively uncommon in the Neolithic Aegean. For some reported specimens, see Makri (Bekiaris 2007.45), Theopetra Cave (Kyparis­si-Apostolika 1996.68), Stavroupoli (Alisoy 2002. 573), Dikili Tash (Séfériades 1992.91), Avgi (Bekia­ris 2018.229–232, 244), Makriyalos (Tsoraki 2008. 95, 98), and Drakaina Cave (Bekiaris in prepara­tion). To conclude, the Souloukia grinding tools were pri­marily meant for processing foodstuffs. This is pre­cisely what is expected on the basis of ethnographic data (e.g., Bartlett 1933.3; Hamon, Le Gall 2013.109; Nixon-Darcus, D’Andrea 2017.193; Robitaille 2016. 433; 2021.773; Roux 1985.34–38; Searcy 2011.1). Reuse, recycling, use intensity With one uncertain exception, there is no evidence of redesigning among the Souloukia grinding tools. The assemblage, nevertheless, includes 27 specimens (21%) with traces of reuse and recycling. The term ‘reuse’ refers here to similar uses of different parts of a tool, while ‘recycling’ refers to different uses of the same tool. Only one tool carries firm evidence of reuse. It is active with two work faces, both of which show wear consistent with cereal grinding. A large portion of the tool is missing, but the two faces ap­pear to have different configurations – possibly a re­flection of use with different passive work faces or kinematics. Firm evidence of recycling has been identified on many more tools. Most (n=13) are passive. We have distinguished three varieties of recycling among pas­sive tools. In none did recycling entail an active function. In the first and most common variety, the face which served the primary use was also involved in recycling. There are several combinations. In the most frequent among them, a work face was used first for grinding and later (only locally) for abrad­ing (five instances). Other combinations are rarer, each typically represented by a single specimen: grinding an undetermined substance followed by (a) grinding a flexible but abrasive material; (b) grind­ing an unspecified soft material; (c) grinding dry dirt or clay (Fig. 10.e); (d) percussion. The last case in­volves the largest specimen in the assemblage (weighing over 25kg). Percussion took place in two stages, resulting in a larger, roughly ovate, concave area and an overlapping smaller, deeper, more cir­cular area. Unfortunately, microwear was not pre­served, leaving the precise use/s of these areas un­determined. Given their shallowness and the lack of clear borders, however, cereal dehusking is un­likely. In the second variety, the primary use involv­ed one work face, recycling another. This variety is represented by two specimens: in the first, one of the work faces was used for cereal grinding, the other for processing a flexible but abrasive materi­al; in the second, the work face was used for grind­ing an unspecified substance, the dorsal one for ab­rading. The third variety blends the previous two, with recycling involving both the work and dorsal faces. It is represented by a single specimen: its work face was used first for grinding and then (locally) for abrading, while the dorsal face was used for (localiz­ed) abrading as well (Fig. 6.a). Seven active specimens were recycled. In all cases, recycling consisted of a double conversion of an ac­tive grinding tool to a passive abrading one. In all but one case, both primary and secondary uses in­volved a single face (Fig. 8.b). In the exception, pri­mary and secondary uses were carried out by the work face and dorsal face, respectively (Fig. 8.d). Finally, one of the indeterminate passive or active specimens was first used for cereal grinding and then for abrading with the same face. Determining the abraded material was not possible in all cases. When it was, use wear was associated with bone or wood processing. No evidence of stone abrasion was identified. Generally speaking, the recycling of several grind­ing tools at Souloukia into abrading implements sug­gests a certain flexibility on the part of the users re­garding the two categories. That most cases of recy­cling involve such a conversion suggests that this was considered an appropriate use life trajectory for grinding tools. The fact, moreover, that most of the grinding tools used for abrading are complete and/or relatively thick indicates that recycling was not typi­cally associated with breakage or exhaustion. Rather Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou the choice was made to divert some perfectly good grinding tools to abrading usages. The infrequent se­condary use of active specimens for passive grinding and the complete absence of the reverse type of re­cycling, on the other hand, indicate that passive and active grinding implements were largely conceived as different categories. Finally, that only one grinding tool was used in a percussive manner suggests that grinding and pounding functions were largely per­formed by different tools, and therefore grinding and pounding implements were conceived as distinct categories as well. Not only do grinding tools exhibit a relatively low rate of reuse and recycling, but they were also used less than expected. This is well illustrated in both the thickness range (3.1–20cm) and average (6.9cm).1 3 Schön and Holter (1990.362–363) reported that among the Mahria of Sudan, a passive tool is consi­dered useless “when it is thinner than 1cm, i.e., when a hole appears”. Active tools “become useless when they … get too thin, i.e. become barely 3cm or less in the middle and measure only a few milli­metres at the edges”. Writing on Guatemalan tools, Michael T. Searcy (2011.103) noted that: “It was not uncommon to see manos that were extremely thin (around 3cm), leaving the woman grinding only enough stone on the edges to hold with her finger­tips”. If the Souloukia grinders had similar degrees of tolerance and similar ideas about the limits of their tools as their ethnographic counterparts, then none of the specimens can be considered exhausted. However, the situation is slightly more nuanced, as explained below. Both the thickness range and average of the Soulou­kia specimens mentioned above refer to maximum measurements. However, minimum thickness is in­formative about use intensity and exhaustion, too. A handful of passive tools have a high maximum thickness but a very low minimum one. In fact, these specimens are broken in the area of minimum thick­ness. We consider these to be worn out (Fig. 5.d, 6.d). If thickness offers a way to assess tool exhaustion, it is not the only one. The degree of concave curvature can also serve as an exhaustion indicator, at least in passive tools. According to Roux’s ethnoarchaeolo­gical study in Mauritania, reciprocally operating pas­sive tools with unrestricted work faces are discard­ed when reaching a depth of 4–5cm since they are not comfortable to use (Roux 1985.57). A compa­rable limit has been noticed by Jérôme Robitaille (2021.398–399) among the Hamar of Ethiopia who discard or recycle their passive tools when the work faces become 4–7cm deep. On this basis, two Sou-loukia tools with work faces deeper than 4cm can be considered as exhausted, too (Fig. 6.b). However, another Ethiopian group provides a note of caution: some of the passive implements used by the Konso are 25–30 cm deep. These tools are passed from ge­neration to generation, their use lives reaching up to a hundred years (Robitaille 2021.Appendix 138– 142) (Fig. 12). Be that as it may, the vast majority of grinding tools at Souloukia were abandoned long before the end of their use lives. Why this is so is a question that will be addressed when the study of the stratigra­phy, features, and other finds is completed, and the assemblage is viewed in the context of the site oc­cupation as a whole. What we can say for the mo­ment is that a similar conclusion was reached with respect to the assemblage from Kleitos (Chondrou 2020.300–301). Indeed, as a rule, Greek Neolithic assemblages are not dominated by exhausted tools. For a couple of exceptions, see Dikili Tash and Ayios Vlasis (Chondrou, Valamoti 2021. 68). Spatial distribution, processes of discard All six abrading tools, 45 grinding tools, three pos­sible roughouts, and 15 unmodified pieces of raw material derive from the main residential area of the settlement.1 4 Forty percent of the tools are complete or nearly so (15 grinding and three abrading tools). These were not found in association with the sub­stances and objects they processed or (in the case of grinding implements) as parts of toolkits. The ma­jority were likely not in situ. Both this and the ab­sence of joining fragments suggest a certain post-use or post-breakage1 5 movement of tools around space. This hypothesis is supported by the fact that a cou­ple of fragments were converted into building mate­rials, while another ended its biography in a pit. However, the largest portion of the assemblage (al­most 60%) was excavated outside the main residen­tial area. This material comprises 81 of the grinding 13 These numbers refer to both complete and fragmentary specimens. 14 Whether they come from house interiors or open areas is unclear since the analysis of the excavated features and stratigraphy is pending. 15 As appealing as it may be, the hypothesis of storage of complete grinding tools in-between use episodes is not satisfactory, since it raises the question as to why only one component of the toolkit was stored away. Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... tools (64%), one roughout, as well as 17 gneiss gra­vels and one tabular sandstone piece without traces of manufacture or use – a total of 100 specimens. Notably, fifteen of the grinding tools are complete, representing half of all the complete specimens found at the site. Equally significant, grinding tools make up the only macrolithic type with a higher re­presentation in the non-residential sector. Abrading tools, celts, hammerstones, and so on were exclusi­vely or primarily found in domestic contexts. Even more remarkably, most of the extra-residen­tial grinding tools (n=63), along with several un-worked gneiss cobbles and boulders, derive from an area measuring c. 300m2 and located about 40m away from the house remains. These specimens were found in no particular arrangement, along with sherds, a substantial number of quartz pieces, a few figurine fragments, a concentration of rocks, as well as masses of clay probably from a small structure (Figs. 2.f, 13). Residue analysis of soil samples from this area detected a significantly lower density of phytoliths than those found on the tools (Ögüt 2018), arguing against the in situ use of these specimens. This hypothesis is reinforced by both the extremely high tool density in this area as well as the lack of a match between the work face configurations of passive and active specimens (most of the former are concave/concave, while most of the latter are con-cave/convex). More likely, these implements were transferred to this spot after utilization somewhere else. In fact, general differences in the characteristics of phytoliths identified on the tools and the control soil samples appear to indicate that these tools were amassed not at once, but gradually, over a period of time (Ögüt 2018). Where these specimens originat­ed, we cannot tell for sure. Yet we consider the resi­dential area as the most likely candidate, especially given the aforementioned indirect evidence of post-use and post-breakage movement of tools. We do not know why this material would have been taken out of the residential space. However, the hy­pothesis that this area served as a locus of discard for broken or worn-out grinding implements should be ruled out, since this assemblage includes roughly ten complete specimens, none of which is exhausted. Nor is the provisional storage of tools with the inten­tion of future reuse/recycling in the domestic arena (see Tsoraki 2008.143) a better explanation, since this assemblage includes several recycled specimens. Be that as it may, the movement of material outside the residential area is not new in Aegean Neolithic archaeology. It is known, for example, from Krema­sti-Kilada, a site in the Souloukia neighbourhood where massive quantities of artefacts, animal bones, building material, and so on were found within roughly 460 non-residential pits (Chondrou 2011. 52; Chondroyianni-Metoki 2009.387–389; 2020.54– 56; Stroulia 2010b.63; Stroulia, Chondrou 2013. 109; Stroulia et al. 2017.2–3). It is also known from Makriyalos, a site farther east in Macedonia, where enormous amounts of material were deposited into a huge negative feature known as Pit 212 (Pappa et al. 2004.84; Tsoraki 2008.126, 135). Macrolithics, and grinding tools in particular, feature prominent­ly in both deposited assemblages. Both Kremasti-Ki­lada and Makriyalos are later than Souloukia, but the combined evidence from all three sites possibly suggests that the transfer of material to areas outside the residential space took place at a small scale in the earlier phases of the Neolithic, but intensified in the later ones. Moving perfectly usable tools out of domestic con­texts may have not been enough for the people of Souloukia. The fact that the extra-residential concen­ Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou tration contains 66% of all active specimens but 51% of the passive ones raises the suspicion of an extra layer of manipulation. If our suspicion is correct, ac­tive and passive tools were treated differently, or to be more specific, the former were targeted for trans­fer more often than the latter. Why Souloukiotes would engage in such behaviour is anyone’s guess, but if they did, they may have also moved active tools to other portions of the site that remain unex­cavated. That would explain both the discrepancy between the numbers of passive and active speci­mens and the mismatch between their sizes and work face configurations mentioned above. The differential treatment of passive and active tools is unexpected in light of the practical complemen­tarity of the two implements. As stated by Cecilia Lidström Holmberg (2004.226), grinding toolkits re­present “dual objects with two parts that continu­ously construct each other”. It is also striking in light of ethnographic evidence that underscores the symbolic complementarity of the two grinding com­ponents. Among the Bemba of Zambia and the Hopi of the US Southwest, the relationship between a passive tool and an active one was used as a meta­phor for the relationship between male and female identities and roles (Lidström Holmberg 2004.227), while among the Mursi of Ethiopia (Robitaille 2016. 434), the Minyanka of Mali (Hamon, Le Gall 2011. 27), and the Zapotec of Mexico (Lidström Holmberg 1998.134; 2004.228), it is/was used to convey the close link between a mother and a child. According to the Mursi in parti­cular, “the handstone must rest well on the grinding slab, just as a baby does on its mother’s back” (Robi­taille 2016.434, Fig. 3) (Fig. 14).1 6 Unexpected as a differential treat­ment of passive and active tools may be, unequivocal evidence for such behaviour has been identified at ano­ther Kitrini Limni site. According to Chondrou (2020.302–303), all but one of the specimens found inside pits at Kleitos (I) are passive. Com­parable patterns are known from two sites beyond Kitrini Limni. At Stavroupoli, one of the pits included mostly passive tools, while another featured primarily active ones, indi­ 16 For a different perspective, see Chondrou (2020.303). cating a certain “structuring of how and where ma­terial was deposited” (Alisoy 2002.581–582). At Ma-kriyalos, the passive specimens found within Pit 212 outnumber active ones by a ratio of c. 6:1 (Tsoraki 2008.143, Tab. 6.28).1 7 This phenomenon deserves systematic comparative study, but at this point we would like to point out three differences between Souloukia and these three sites. The first has to do with the state of preserva­tion of the tools subjected to differential treatment. While at the other sites almost all specimens are broken, at Souloukia some are fragmentary, others are not. The second difference has to do with the features from which these tools were recovered. While at the other sites they were found in nega­tive features, at Souloukia they were not. The third difference has to do with time. While the other sites date to the later Neolithic, Souloukia belongs to the earlier Neolithic. The significance of these differences remains to be investigated, but for the moment we would like to suggest that Kitrini Limni was the lo­cus of selective grinding tool deposition in both the earlier and the later part of the Neolithic. Epilogue Despite the recent dramatic growth of Aegean ma-crolithic studies, very little is known about materials from the earlier part of the Neolithic. Our study of the grinding and abrading tools from the site of Pon- 17 For a few non-Greek examples of a differential treatment of passive and active tools, see Lidström Holmberg (2004.222, 229– 230). Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... tokomi Souloukia, in the Kitrini Lim­ni Basin, represents an attempt to address this gap. By integrating systematic macrosco­pic examination, use wear and resi­due analysis, as well as macrobota­nical, experimental, contextual, and ethnographic data, we explored these tools’ raw materials, technomorpho-logical characteristics, and functions, discussed their spatial distribution, and shed light on the ways they were conceived by their producers and users. The assemblage includes a small number of abrading tools. All are a posteriori and mostly of fine-grain­ed sandstone obtained from a regio­nal source. According to traceological analysis, wood and bone are the two materials processed with these tools. No evidence of stone processing was identi­fied, but the several dozen celts excavated from the site hint at the presence of unrecovered abrading implements, perhaps close to sources of water. The number of grinding tools is much higher, com­prising over a hundred specimens. These were made of gneiss (another regional material) in a manner that was far from involved. Most were used in pro­cessing foodstuffs, cereals being the most common. A number were used for abrading purposes in the context of recycling. Generally speaking, grinding tools were not used as intensely as one would ex­pect given the substantial masses of raw material they represent. Our study, moreover, pointed to dis­crepancies between the numbers and configurations of passive and active specimens that possibly result­ed from differential discard processes. Throughout this paper, comparisons were made with other sites in the area and elsewhere in an attempt to place the Souloukia assemblage in both a region­al and a broader Aegean framework. These compa­risons revealed, for example, that: similar abrading tools of similar material were employed at various sites of the Kitrini Limni basin; sites located at diffe­rent parts of the basin used the type of gneiss that could be found in the nearest sources; the high pro­portion of concave/concave passive tools noted at Souloukia is not known from other Kitrini Limni as­semblages and is extremely rare in Neolithic Greece as a whole; and the deposition of substantial amounts of material in non-residential areas at Souloukia pa­rallels that known from a few later Neolithic sites in Kitrini Limni and elsewhere. For the most part, these comparisons refer to later Neolithic assemblages. The one with which we close this paper references earlier assemblages, offering some insights into the roles of grinding tools in the context of the first Aegean agropastoral communi­ties. While sufficient data for a meaningful and sys­tematic assessment of the morphometric, techno-functional, and contextual characteristics of the few known earlier Neolithic assemblages is for the most part missing, basic information about the sizes of some of them is available. It thus appears that at cer­tain sites – e.g., Achilleion (Winn, Shimabuku 1989. 268), Sossandra (Georgiadou 2015.42–43), Revenia Korinos (Besios, Adaktylou 2004.363), and Ayios Vlasis (Bekiaris et al. 2020. 145) – grinding tools were common as they were at Souloukia. At other sites – e.g., Mavropigi-Fillotsairi (Ninou et al. in press), Paliambela Kolindros (Tsartsidou, Kotsakis 2020.11), and Prodromos (Moundrea 1975.92–99) – they were rare. The paucity of grinding imple­ments cannot be considered an artefact of excava­tion biases since substantial areas were investigated at these sites or other macrolithic tools were found in considerable quantities. For example, the exten­sively excavated site of Mavropigi-Fillotsairi yielded only one specimen positively identified as a grinding tool (Ninou et al. in press). Prodromos yielded over 40 celts but only two passive grinding tools and (as far as we can tell) no active ones (Moundrea 1975. 92–99). The Early Neolithic strata of Paliambela Ko­ Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou lindros – one of the most systematically excavated sites in Greece – are characterized by an almost complete absence of grinding tools (Tsartsidou, Kot­sakis 2020.11). This scarcity does not appear to re­flect a different kind of occupation either. None of these sites represents a special use or seasonal set­tlement. Nor, lastly, does it appear to correlate with an absence of cereals and pulses, the types of sub­stances typically processed with grinding tools. Such plant remains were found, for example, at both Mav­ropigi-Fillotsairi (Karamitrou-Mentessidi et al. 2013. 2) and Paliambela Kolindros (Kotzamani, Livarda 2018.86–89; Tsartsidou, Kotsakis 2020.12). Without a thorough study and publication of earlier Neolithic assemblages, this discrepancy cannot be explained systematically. However, we would like in a preliminary fashion to offer two alternative hypo­theses. According to the first hypothesis, the discrepancy re­flects two broad types of cereal and legume prepa­rations. On the one hand, there were recipes that in­volved boiling, toasting, or parching of whole, split, or cracked seeds; e.g., soups, stews, gruels, bulgur, koliva, etc. (for experimental versions of such pre­parations, see Fig. 15.a, Dimoula et al. 2020.Fig 5g).1 8 Grinding had no part in these preparations. On the other hand, there were recipes that involv­ed flour/meal and thus required the use of grinding tools; e.g., regular bread, flat bread (e.g., pita or naan type), falafels, etc. (for experimental versions of two of these preparations, see Fig 15.b–c). In the context of this hypothesis, in some earlier Neolithic Aegean communities there was an emphasis on the first type of recipes. In others, the second type of re­cipes was used, either exclusively or alongside those of the first type. If real, this distinction existed even within the same area: Mavropigi-Fillotsairi is less than 3km from Sou-loukia, while Paliambela Kolindros is not far from Revenia Korinos. Such culinary identities may have had a cultural/ethnic origin, reflecting the ancestral homelands of the different groups that occupied the Greek landscape at the beginning of the Neolithic (see also Valamoti 2017.178–184). According to the second hypothesis, the scarcity of grinding tools at some sites is due to specific practi­ces that involved the removal of tools from residen­tial areas. Such practices would have amounted to more massive versions of the deliberate transfer of specimens identified at Souloukia. We consider the former hypothesis as more likely than the latter, but both (and possibly others) should be tested when we acquire a better understanding of each individual assemblage. Be that as it may, such strong discrepancies in the sizes of grinding tool as­semblages are not visible in the archaeological re­cord of the later part of the Neolithic. During this pe­riod, more or less substantial numbers were the norm. 18 According to Sonya Atalay and Christine A. Hastorf (2006.298–311), these were common preparations for cereals and pulses in the Early Neolithic component of Çatalhöyük. Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... ACKNOWLEDGEMENTS We are grateful to the Mediterranean Archaeological Trust whose two grants (to Stroulia) made the study of the Souloukia material possible. Our thanks also to: Takis Gouliafas for preparing the drawings, Sofia Vlahopoulou for working on the map, Vasilios Melfos for identifying the raw materials, Melfos and Mihalis Fotiadis for helping with the survey of raw material sources, Tasos Bekiaris and Ismini Ninou for sharing unpublished information, Laure Dubreuil for comparing use wear notes, Michael Strezewski for help with the illustrations and editing, the staff of the Archaeological Museum of Aiani for practical support and hospitality, as well as Paschalis Tounas and Vasiliki Koutrafouri for assistance with our project in general. . References Adams J. 2014. Ground Stone Analysis: A Technological Approach. University of Utah Press. Salt Lake City. Adams J., Delgado S., Dubreuil L., Hamon C., Plisson H., and Risch R. 2009. Functional analysis of macro-lithic ar­tefacts. In F. Sternke, L. Costa, and L. Eigeland (eds.), Non-Flint Raw Material Use in Prehistory: Old Prejudices and New Directions. Archaeopress. Oxford: 43–66. Albert R. M., Lavi O., Estroff L., Weiner S., Tsatskin A., Ro­nen A., and Lev-Yadun S. 1999. Mode of occupation of Ta-bun Cave, Mt Carmel, Israel during the Mousterian Pe­riod: A study of the sediments and phytoliths. Journal of Archaeological Science 26(10): 1249–1260. https://doi.org/10.1006/JASC.1999.0355 Alisoy H. A. 2002. Consumption of ground stone tools at Stavroupoli. In D. Grammenos, S. Kotsos (eds.), Sostikes Anaskafes sto Neolithiko Oikismo Stavroupolis Thessa­lonikis. Archaeological Institute of Northern Greece. Thes­saloniki: 561–608. Almasidou E. 2019. Ta Tripta Ergaleia apo to Neolithiko Oikismo sti Limni Koroneia. Unpublished MA thesis. Ari­stotle University of Thessaloniki. Thessaloniki. (in Greek) Atalay S., Hastorf C. A. 2006. Food, meals, and daily acti­vities. Food habitus at Neolithic Çatalhöyük. American Antiquity 71(2): 283–319. https://doi.org/10.2307/40035906 Banks E. C. 2015. Lerna VII: The Neolithic Settlement. American School of Classical Studies at Athens. Princeton. Bartlett K. 1933. Pueblo Milling Stones of the Flagstaff Region and their Relation to Others in the Southwest: A Study in Progressive Efficiency. Museum of Northern Ari­zona Press. Flagstaff. Bekiaris T. 2007. Tripta Ergaleia apo to Neolithiko Oi­kismo tis Makris Evrou. Unpublished MA thesis. Aristot­le University of Thessaloniki. Thessaloniki. (in Greek) 2018. Tehnologies Tripton sti Neolithiki tis Voreias Elladas: O Neolithikos Oikismos tis Avgis Kastorias. Unpublished PhD thesis. Aristotle University of Thessa­loniki. Thessaloniki. (in Greek) 2020. Ground stone technology in context: Consump­tion of grinding tools and social practice at Neolithic Avgi, NW Greece. Journal of Lithic Studies 7(3): 1–24. https://doi.org/10.2218/jls.3078 in preparation. Ground stone tools from Drakaina Cave: Grounds for thought. In G. Stratouli (ed.), Drakaina Cave on Kephalonia Island: A Place of Social Activity during the Neolithic. Bekiaris T., Stergiou C., and Theodoridou S. 2017. Making choices in a Neolithic landscape: Raw materials and ground stone tool technology in Neolithic Avgi, northwestern Greece. In A. Sarris, E. Kalogiropoulou, T. Kalayci, and L. Karimali (eds.), Communities, Landscape, and Interac­tion in Neolithic Greece. Proceedings of the Internatio­nal Conference, Rethymno 29–30 May, 2015. Internatio­nal Monographs in Prehistory. Archaeological Series 20. Ann Arbor: 415–433. Bekiaris A., Chondrou D., Ninou I., and Valamoti S. 2020. Food-processing ground stone tools in the Greek Neoli­thic and Bronze Age: A synthesis of the published data. Journal of Greek Archaeology 5: 135–195. Bekiaris A., Katsikaridis N., Stergiou C. L., and Stratouli G. in press. Cooking on the rocks? An interdisciplinary ap­proach on the use of burnt stone slabs from Middle/Late Neolithic Avgi, Kastoria. In S. M. Valamoti, M. Ntinou, and A. Dimoula (eds.), Plant Foods of Ancient Europe and Beyond. Besios M., Adaktylou F. 2004. Neolithikos oikismos sta “Revenia” Korinou. To Arhaiologiko Ergo sti Makedonia kai Thraki 18: 357–366. (in Greek) Cappers R. T. J., Neef R., Bekker R. M., Fantone F., and Okur Y. 2016. Digital Atlas of Traditional Agricultural Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou Practices and Food Processing. Vol. 1. Barkhuis & Uni­versity of Groningen Library. Groningen. Chadou M. 2011. Ta Lithina Tripta apo ton Iliotopo tou Anatolikou Lagada. MA thesis. Aristotle University of Thessaloniki. Thessaloniki. (in Greek) Chondrou D. 2011. Tripta Ergaleia apo ti Neolithiki The-si tis Toumbas Kremastis Kiladas. Unpublished MA the­sis. Aristotle University of Thessaloniki. Thessaloniki. (in Greek) 2018. I Tripti Lithotehnia apo tous Proistorikous Oi­kismous Kleitos I kai Kleitos II tis Lekanis tis Kitri­nis Limnis Kozanis. Unpublished PhD thesis. Aristotle University of Thessaloniki. Thessaloniki. (in Greek) 2020. Daily practices and special events: Exploring grinding technologies at the two neighbouring settle­ments of Kleitos in Late/Final Neolithic northern Gre­ece. Documenta Praehistorica 47: 286–310. https://doi.org/10.4312\dp.47.16 Chondrou D., Valamoti S. M., Procopiou H., and Papado­poulou L. 2018. Grinding cereals and pulses in the Neo­lithic site of Kleitos: An experimental investigation of mi-croconglomerate grinding equipment, final products and use wear. Journal of Greek Archaeology 3: 23–45. Chondrou D., Bofill M., Procopiou H., Vargiolu R., Zahou­ani H., and Valamoti S. M. 2021. How do you like your cereal? A qualitative and quantitative use-wear analysis on archaeological grinding tools from prehistoric Greek sites. Wear 2021: 1–10. https://doi.org/10.1016/j.wear.2021.203636 Chondrou D., Valamoti S. M. 2021. Mapping life-cycles: Exploring grinding technologies and the use of space at Late/Final Neolithic Kleitos, northern Greece. In P. Peder-sen, A. Jörgensen-Lindahl, M. Sorensen, and T. Richter (eds.), Proceedings of the 3rd Meeting of the Association of Ground Stone Tools Research. Archaeopress. Archaeo-press Access Archaeology. Oxford: 63–81. https://doi.org/10.2307/j.ctv2b07trt Chondroyianni-Metoki A. 2009. Mi Oikistikes Hriseis Ho-rou stous Neolithikous Oikismous: To Paradeigma tis Toumbas Kremastis Kiladas. Unpublished PhD thesis. Aristotle University of Thessaloniki. Thessaloniki. (in Gre­ek) 2020. Outside the residential place at the Neolithic set­tlement of Toumba Kremastis Koiladas, northern Gre­ece. In N. N. Tasi., D. Urem-Kotsou, and M. Buri. (eds.), Making Spaces into Places: The North Aegean, the Balkans and Western Anatolia in the Neolithic. BAR Internationa Series 3001. Oxford: 53–71. 2022. Kosmimata tis neoteris neolithikis periodou apo tin Toumba Kremastis Kiladas. In N. Merousis, M. Niko-laidou, and L. Stefani (eds.), Mirrini: Meletes Aigaiakis Proistorias: Timitikos Tomos gia tin Aikaterini Pa-paefthimiou-Papasteriou. Archaeological Museum of Thessaloniki Publication No. 50. Thessaloniki: 329-349. (in Greek) https://www.amth.gr/research/publictions/ myrrini-meletes-aigaiakis-proistorias in press a. Kitrini Limni Kozanis: Ta haraktiristika tis katoikisis tis periohis ipo to fos ton neon anaskafikon dedomenon apo tin periohi ton lignitorihion tis DEI. Praktika Sinedriou Arhaiologikes Erevnes kai Mega-la Dimosia Erga, 8–9 December 2017. (in Greek) in press b. Kitrini Limni 2016–2017: Ta anaskafika de­domena apo tin periohi ton lignitorihion tis DEI. To Ar-haiologiko Ergo sti Makedonia kai Thraki 31, 2017. (in Greek) in press c. Pontokomi, thesi Souloukia. Arhaiologikon Deltion (Hronika) 72(2017). (in Greek) D’Andrea A. K., Mitiku H. 2002. Traditional emmer pro­cessing in Highland Ethiopia. Journal of Ethnobiology 22: 179–217. David N. 1998. The ethnoarchaeology and field archae­ology of grinding at Sukur, Adamawa State, Nigeria. Afri­can Archaeological Review 15(1): 13–63. Delgado Raack S., Risch R. 2009. Towards a systematic analysis of grain processing techniques. In M. de Araújo Igreja, I. Clemente Conte (eds.), Recent Functional Stu­dies on Non Flint Stone Tools: Methodological Impro­vements and Archaeological Inferences. Lisboa, 23–25 May 2008. LISBOA – Proceedings of the Workshop (CD­ROM). Lisbon: 1–20. 2016. Bronze Age cereal processing in southern Iberia: A material approach to the production and use of grind­ing equipment. Journal of Lithic Studies 3(3): 125– 145. https://doi.org/10.2218/jls.v3i3.1650 Dimoula A., Tsirtsoni Z., Paraskevi Y., +4 authors, and Va-lamoti S. M. 2020. Experimental investigation of ceramic technology and plant food cooking in Neolithic northern Greece. STAR: Science & Technology of Archaeological Research 5(2): 269–286. https://doi.org/10.1080/20548923.2020.1762370 Elster E. S. 2003. Grindstones, polished edge-tools, and other stone artifacts. In C. Renfrew, M. Gimbutas, and E. S. Elster (eds.), Excavations at Sitagroi: A Prehistoric Vil­lage in Northeast Greece Volume 1. Institute of Archaeo­logy University of California. Los Angeles: 175–195. Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... Evans J. D. 1964. Excavations in the Neolithic settlement of Knossos, 1957–60. Part I. The Annual of the British School at Athens 59: 132–240. Evans J. D., Renfrew C. 1968. Excavations at Saliagos near Antiparos. British School at Athens Supplement 5. Thames & Hudson. London. Fotiadis M. 1988. Proistoriki erevna stin Kitrini Limni N. Kozanis: Mia sintomi ekthesi (in Greek). To Arhaiologiko Ergo sti Makedonia kai Thraki 2: 41–54. Fotiadis M., Hondroyanni-Metoki A., Kalogirou A., Mania-tis Y., Stroulia A., Ziota C. 2019. Megalo Nisi Galanis (6300–1800bc): Constructing a cultural sequence for the Neolithic of west Macedonia, Greece. The Annual of the British School at Athens 114: 1–40. https://doi.org/10.1017/S0068245419000145 Galdikas B. 1988. Milling Stones. In A. McPherron, S. Sre­jovi. (eds), Divostin and the Neolithic of Central Serbia. University of Pittsburgh. Pittsburgh: 338–344. Georgiadou A. 2015. The Neolithic House at Sosandra. University Studio Press. Thessaloniki. Hamon C., Le Gall V. 2011. Les meules en pays Minyanka (Mali): étude des carrieres et techniques de production actuelles. In D. Williams, D. Peacock (eds.), Bread for the People: The Archaeology of Mills and Milling. Archaeo-press. Oxford: 19–28. 2013. Millet and sauce: The uses and functions of querns among the Minyanka (Mali). Journal of Anthropologi­cal Archaeology 32: 109–121. https://doi.org/10.1016/J.JAA.2012.12.002 Hayden B. 1987. Past to present uses of stone tools in the Maya Highlands. In B. Hayden (ed.), Lithic Studies Among the Contemporary Highland Maya. University of Arizona Press. Tuscon: 160–234. Hillman G. 1984. Traditional husbandry and processing of Archaic cereals in recent times: The operations, prod­ucts and equipment which might feature in Sumerian texts. Part I: the glume wheats. Bulletin of Sumerian Agri­culture 1: 114–152. Karamitrou-Mentessidi G., Efstratiou N., Kaczanowska M., and Koz³owski J. K. 2015. Early Neolithic settlement of Mavropigi in western Greek Macedonia. Eurasian Prehi­story 12(1–2): 47–116. Karamitrou-Mentessidi G., Efstratiou N., Kozlowski J. K., +5 authors, and Valamoti S. M. 2013. New evidence on the beginning of farming in Greece: The Early Neolithic settlement of Mavropigi in western Macedonia (Greece). Antiquity 87 (336): Project Gallery. http://antiquity.ac.uk/projgall/mentessidi336/ Karamitrou-Mentessidi G., Lokana C., and Anagnostopou­lou K. 2010. Dio theseis tis arhaioteris kai mesis neolithi­kis stin Pontokomi kai Mavropigi Eordaias. To Arhaiolo­giko Ergo sti Makedonia kai Thraki 24: 39–51. (in Greek) Kotsakis K. 1999. What tells can tell: Social space and set­tlement in the Greek Neolithic. In P. Halstead (ed.), Neo­lithic Society in Greece. Sheffield Academic Press. Shef­field: 66–76. Kotzamani G., Livarda A. 2018. People and plant entan­glements at the dawn of agricultural practice in Greece: An analysis of the Mesolithic and early Neolithic archaeo-botanical remains. Quaternary International 496: 80– 101. https://doi.org/10.1016/j.quaint.2018.04.044 Krahtopoulou A. 2019. Current approaches to the Neoli­thic of Thessaly. Archaeological Reports 65: 73–85. https://doi.org/10.1017/S0570608419000048 Kyparissi-Apostolika N. 1996. The Theopetra Cave. In G. A. Papathanassopoulos (ed.), Neolithic Culture in Gre­ece. Nicholas P. Goulandris Foundation, Museum of Cyc­ladic Art. Athens: 67–68. Lewis R., Rahim M., Cripps J., Roubos V., and Tsoraki C. 2009. Wear of stone used to manufacture axes in the Neo­lithic settlement at Makriyalos in Northern Greece. Wear 267: 1325–1332. https://doi.org/10.1016/J.WEAR.2008.12.090 Lewis R., Tsoraki C., Broughton J., Cripps J. C., Afodun S. A., Slatter T., and Roubos V. 2011. Abrasive and impact wear of stone used to manufacture axes in Neolithic Gre­ece. Wear 271: 2549–2560. https://doi.org/10.1016/J.WEAR.2010.12.074 Lidström Holmberg C. 1998. Prehistoric grinding tools as metaphorical traces of the past. Current Swedish Archa­eology 6: 123–142. https://doi.org/10.37718/CSA.1998.10 2004. Saddle querns and gendered dynamics of the Early Neolithic in mid central Sweden. In H. Knuttson (ed.), Arrival: Coast to Coast 10. Wikströms. Uppsala: 199–231. Madella M., Alexandre A., and Ball T. 2005. International code for phytolith nomenclature 1.0. Annals of Botany 96: 253–260. https://doi.org/10.1093/aob/mci172 Menasanch M., Risch R., and Soldevilla J. A. 2002. Las tec­nologías del procesado de cereal en el sudeste de la pen­insula iberica durante el III y el II milenio A.N.E. In H. Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou Procopiou, R. Treuil (eds.), Moudre et broyer I: Méthodes. Memoire Histoire Sc.et Techniq. Comite Des Travaux Hi-storiques Etscientifiques. Paris: 81–110. Mould C. A., Ridley C., and Wardle K. A. 2000. The stone small finds. In C. Ridley, K. A. Wardle, and C. A. Mould, Servia. Vol. 1, Anglo-Hellenic Rescue Excavations 1971– 73. The British School at Athens. London: 112–190. Moundrea H. A. 1975. Le site néolithique de Prodromos (Grece). Outillage lithique et osseux. Position dans le context Thessalien. MA thesis. University of Paris X. Nan-terre. Ninou I. 2006. Analysis of Grinding Stones from Neo­lithic Sites of Northern Greece: Dispilio and Apsalos. MA thesis. University of Southampton. Southampton. forthcoming. Tripti Lithotehnia tis Arhaioteris Neo­lithikis sti Ditiki Makedonia: O Oikismos Mavropigis (Thesi Fillotsairi) – Tehnologia, Oikonomia, Koinoni­kes Diastaseis. Unpublished PhD thesis. Aristotle Uni­versity of Thessaloniki. Thessaloniki. (in Greek) Ninou I., Efstratiou N., and Valamoti S. M. in press. Grind­ing and pounding in Early Neolithic southeastern Europe: Culinary preferences and social dimensions of plantfood processing. In S. M. Valamoti, M. Ntinou, and A. Dimoula (eds.), Plant Foods of Ancient Europe and Beyond. Nixon-Darcus L. A., D’Andrea A. C. 2017. Necessary for life: Studies of ancient and modern grinding stones in Highland Ethiopia. African Archaeological Review 34: 193–223. https://doi.org/10.1007/s10437-017-9252-4 Ögüt B. 2018. Residue Analysis of Ground Stone Tools from the Greek Neolithic Site of Pontokomi-Souloukia. Unpublished report. Pappa M., Halstead P., Kotsakis K., and Urem-Kotsou D. 2004. Evidence for large-scale feasting at Late Neolithic Makriyalos, N. Greece. In P. Halstead, J. C Barrett (eds.), Food, Cuisine, and Society in Prehistoric Greece. Oxbow Books. Oxford: 16–44. Pena-Chocarro L., Zapata L. 2003. Post-harvesting proces­sing of hulled wheats: An ethnoarchaeological approach. In P. C. Anderson, L. S. Cummings, T. S. Schippers, and B. Simonel (eds.), Le traitement des récoltes: un regard sur la diversité, du Néolithique au présent. Actes des XXIIIe rencontres internationales d’archéologie et d’histoire d’Antibes, 17–19 Octobre 2002. Éditions APDC. Antibes: 99–113. 2014. Parching and dehusking hulled wheats. In P. C. Anderson, L. Pena-Chocarro (eds.), Early Agricultural Remnants and Technical Heritage (EARTH): 8,000 Years of Resilience and Innovation, vol. 2. Oxbow Books. Oxford: 226–232. Pétrequin P., Pétrequin A.-M. 1993. Écologie d’un outil. La hache de pierre en Irian Jaya (Indonésie). CRA-Mo­nographie 12. Centre national de la recherche scientifi­que. Paris. Procopiou H. 2003. Les techniques de décorticage dans le monde égéen: Étude ethnoarchéologique dans les Cycla­des. In P. C. Anderson, L. S. Cummings, T. S. Schippers, and B. Simonel (eds.), Le traitement des récoltes: un re­gard sur la diversité, du Néolithique au présent. Actes des XXIIIe rencontres internationales d'archéologie et d’histoire d’Antibes, 17–19 Octobre 2002. Éditions APDC. Antibes: 115–136. Pyke G. 1993. The Stratigraphy, Structures and Small Finds of Nea Nikomedeia, Northern Greece. MA thesis. University of Birmingham. Birmingham. Risch R. 2002. Recursos naturales, medios de produc­ción y explotación social. Un análisis económico de la industria lítica de Fuente Álamo (Almería), 2250–1400 antes de nuestra era. Verlag Philipp von Zabern. Mainz. 2008. Grain processing technologies and economic or­ganization: A case study from the south-east of the Ibe­rian peninsula during the Copper Age. The Arkeotek Journal 2(2): 1–47. Robitaille J. 2015. Analyse fonctionnelle de l’outillage de broyage de Tell el-Iswid: Approche expérimentale et ethnographique. Éditions Universitaires Européennes. Saarbrücken. 2016. The ground stone industry of the Mursi of Maki, Ethiopia: Ethnoarchaeological research on milling and crushing equipment (technique and function). Jour­nal of Lithic Studies 3(3): 429–456. https://doi.org/10.2218/jls.v3i3.1680 2021. Approches éthno (archéo) graphique et mor­photechnique du Macro-outillage de quelques contex­tes éthiopiens. Unpublished PhD thessis. École des Hau­tes Études en Sciences Sociales. Paris. Roux V. 1985. Le matériel de broyage: Étude ethnoar­chéologique a Tichitt, Mauritanie. Éditions Recherche sur les Civilisations. Paris. Runnels C. N. 1981. A Diachronic Study and Economic Analysis of Millstones from the Argolid, Greece. Unpub­lished PhD thesis. Indiana University. Bloomington. https://search.proquest.com/openview/d5a076dc319b1a ba55848f3b1522773f/1?pq-origsite=gscholar&cbl=18 750&diss=y Grinding and abrading activities in the earlier Neolithic of northern Greece> a multi-proxy and comparative approach for the site of ... Samuel D. 2010. Experimental grinding and ancient Egyp­tian flour production. In S. Ikram, A. Dodson (eds.), Be­yond the Horizon: Studies in Egyptian Art, Archaeology and History in Honour of Barry J. Kemp. Publications of the Supreme Council of Antiquities. Cairo: 456–477. Sarpaki A. 2001. Processed cereals and pulses from the Late Bronze Age site of Akrotiri, Thera; Preparations prior to consumption: a preliminary approach to their study. Annual of the British School at Athens 96: 27–40. https://doi.org/10.1017/s0068245400005219 Sarris A., Kalayci T., Simon F.-X., +17 authors, and Stame­lou E. 2017. Opening a new frontier in the study of Neo­lithic settlement patterns of eastern Thessaly, Greece. In A. Sarris, E. Kalogiropoulou, T. Kalayci, and L. Karimali (eds.), Communities, Landscape, and Interaction in Neo­lithic Greece. Proceedings of the International Confe­rence, Rethymno 29–30 May, 2015. International Mono­graphs in Prehistory. Archaeological Series 20. Ann Ar­bor: 27–48. Schön W., Holter U. 1990. Grinding implements from the Neolithic and recent times in desert areas in Egypt and Sudan. Beitrage zur allgemeinen und vergleichenden Archäologie 9–10: 359–379. Searcy M. T. 2011. The Life-Giving Stone. Ethnoarchaeo-logy of Maya Metates. University of Arizona Press. Tucson. Séfériades M. 1992. La pierre polie. In R. Treuil (ed.), Di-kili Tash. Village préhistorique de Macédoine orientale I. Fouilles de Jean Deshayes (1961–1975) volume 1. Bul­letin de Correspondance Hellénique Supplément 24. Pa­ris: 84–99. Stergiou C. L., Bekiaris T., Melfos V., Theodoridou S., and Stratouli G. 2022. Sourcing macrolithics: Mineralogical, geochemical and provenance investigation of stone arte-facts from Neolithic Avgi, north-west Greece. Archaeo­metry 2022: 283–299. https://doi.org/10.1111/arcm.12706 Stroulia A. 2002. Lithina tripta apo tin Kitrini Limni Koza­nis: Proti proseggisi, prota erotimata. To Arhaiologiko Er­go sti Makedonia kai Thraki 17: 571–580. (in Greek) 2010a. Flexible Stones. Ground Stone Tools from Fran-chthi Cave. Indiana University Press. Bloomington and Indianapolis. 2010b. Ergaleia me kopsi apo tin Kremasti Kilada no-mou Kozanis: viografikes paratiriseis. To Arhaiologiko Ergo sti Makedonia kai Thraki 24: 63–71. (in Greek) 2018a. Macrolithics: Ordinary things in an extraordi­nary place. In A. Papathanasiou, W. A. Parkinson, D. J. Pullen, M. L. Galaty, and P. Karkanas (eds.), Neolithic Alepotrypa Cave, in the Mani, Greece. In Honor of George Papathanassopoulos. Oxbow. Philadelphia: 200–241. 2018b. Making tools, reconstructing manufacturing pro­cesses: The celt industry of Varemeni Goulon in north­ern Greece. Journal of Greek Archaeology 3: 47–74. 2020. Mending fragments: Stone vessels from Franchthi and other Greek Neolithic sites. The Annual of the Bri­tish School at Athens 115: 1–58. https://doi.org/10.1017/S0068245420000106 in press. The Platia Magoula Zarkou macrolithics: A Thessalian industry in its Aegean Neolithic Context. In E. Alram-Stern, K. Gallis, and G. Toufexis (eds.), Platia Magoula Zarkou. The Neolithic Period: Environment, Stratigraphy and Architecture, Chronology, Tools, Fi­gurines and Ornaments. Institute for Oriental and Eu­ropean Archaeology. Vienna. Stroulia A., Dubreuil L. 2011. Ground Stone Tools from Kremasti-Kilada, Kitrini Limni, Northern Greece. Unpub­lished grant report submitted to the Institute of Aegean Prehistory. Philadelphia. Stroulia A., Chondrou D. 2013. Destroying the means of production: The case of ground stone tools from Kremasti-Kilada, Greece. In J. Driessen (ed.), Destruction. Archaeo­logical, Philological and Historical Perspectives. Pres­ses universitaires de Louvain. Louvain: 109–131. Stroulia A., Dubreuil L., Robitaille J., and Nelson K. 2017. Salt, sand, and saddles: Exploring an intriguing configura­tion among grinding tools. Ethnoarchaeology 9(2): 1– 27. https://doi.org/10.1080/19442890.2017.1364053 Stroulia A., Tounas P., Robitaille J., Hondroyanni-Metoki A., Kouziakis Z., Wardle K. A., and Theodoropoulou T. 2022. Fishing, weaving, matting: Debating the function of notched cobbles in Neolithic Greece. Journal of Neoli­thic Archaeology 24: 15–46. Toufexis G. 2017. Oikistiki Drastiriotita kai Organosi tou Horou stous Oikismous tis Neoteris Neolithikis sti Thessalia: Paradeigmata apo tous Oikismous ston Pro-fiti Ilia Mandras, Makrihori, Galini kai Rahmani. Un­published PhD thesis. University of Thessaly. Volos. (in Greek) Touloumis K. 2002. I oikonomia enos Neolithikou lim­naiou oikismou (in Greek). In G. Hourmouziadis (ed.), Dispilio 7500 Hronia Meta. University Studio Press. Thes­saloniki: 89–113. (in Greek) Tsartsidou G., Kotsakis K. 2020. Grinding in a hollow? Phytolith evidence for pounding cereals in bedrock mor­tars at Paliambela Kolindros, an Early Neolithic site in Ma­ Anna Stroulia, Jérôme Robitaille, Birgül Ögüt, Areti Chondroyianni-Metoki, and Dimitra Kotsachristou cedonia, north Greece. Archaeological and Anthropolo­gical Sciences 12(8): 1–16. https://doi.org/10.1007/s12520-020-01161-8 Tsiolaki E. 2009. Tripta Lithina Ergaleia tis Mesis kai Is-teris Epohis tou Halkou apo tin Toumba Thessalonikis. MA thesis. Aristotle University of Thessaloniki. Thessalo­niki. (in Greek) Tsoraki C. 2008. Neolithic Society in Northern Greece: The Evidence of Ground Stone Artefacts. Unpublished PhD thessis. University of Sheffield. Sheffield. https://etheses.whiterose.ac.uk/15030/1/637512_Vol1.pdf 2011a. Stone-working traditions in the prehistoric Ae­gean. In V. Davis, M. Edmonds (eds.), Stone Axe Stu­dies III. Oxbow. Oxford: 231–244. 2011b. Disentangling Neolithic networks: Ground stone technology, material engagements and networks of ac­tion. In A. Brysbaert (ed.), Tracing Prehistoric Social Networks through Technology: A Diachronic Perspec­tive on the Aegean. Routledge. New York: 12–29. 2011c. ‘Shiny and colourful:’ Raw material selection and the production of edge tools in Late Neolithic Makriya­los, Greece. In A. Savill (ed.), Flint and Stone in the Neolithic Period. Oxbow. Oxford: 287–303. Valamoti S. M. 2010. Magirevontas ta dimitriaka stin proistoriki Makedonia. In N. Merousis, L. Stefani, and M. Nikolaidou (eds.), IRIS. Studies dedicated to the Memory of Prof. Angeliki Pilali-Papasteriou. Cornelia Sfakiana­ki. Thessaloniki: 35–58. (in Greek) 2017. Culinary landscapes and identity in prehistoric Greece: An archaeobotanical exploration. In M. Gori, M. back to contents Ivanova (eds.), Balkan Dialogues: Negotiating Iden­tity between Prehistory and the Present. Routledge. London and New York: 169–193. Valamoti S. M., Chondrou D., Papadopoulou L. 2013. Plant food processing and ground stone equipment in prehisto­ric Greece. In P. C. Anderson, C. Cheval, and A. Durand (eds.), Regards croisés sur les outils liés au travail des végétaux. Actes des 33e rencontres internationales d’ar-chéologie et d’histoire d’Antibes, 23–25 October 2012. Éditions APDCA. Antibes: 169–187. Vu.kovi. V. 2019. Neolithic Economy and Macro-lithic Tools of the Central Balkans. Unpublished PhD thesis. Universitat Autonoma de Barcelona. Departament d’An­tropologia Social i de Prehistoria. Barcelona. https://www.tdx.cat/handle/10803/669666#page=1 Winn S., Shimabuku D. 1989. Bone and ground stone tools. In M. Gimbutas, S. Winn, and D. Shimabuku (eds.), Achilleion: A Neolithic Settlement in Thessaly, Greece 6400–5600 BC. Monumenta Archaeologica 14. Los Ange­les: 259–272. Wright K. I. 1992. A classification system for ground stone tools from the prehistoric Levant. Paléorient 18(2): 53– 81. Wright K., Baysal A. 2012. Ground stone tools and techno­logies associated with buildings in the BACH Area at Ça­talhöyük. In R. Tringham, M. Stevanovi. (eds.), Last House on the Hill. BACH Area Reports from Çatalhöyük, Tur­key. Cotsen Institute of Archaeology Press at UCLA. Los Angeles: 415–422. Ziota C., Anagnostopoulou K., and Lokana C. 2014. To sos-tiko ergo tis L’ E.P.K.A. stin Pontokomi kai stin Mavropigi Kozanis to 2014. To Arhaiologiko Ergo sti Makedonia kai Thraki 28: 77–88. (in Greek) Documenta Praehistorica XLIX (2022) ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance Nelson J. Almeida1, André Texugo1,2,3, and Ana Catarina Basílio3,4 nelsonjalmeida@gmail.com 1 UNIARQ, Center for Archaeology of the University of Humanities of the University of Lisbon, Lisbon, PT 2 CEG, Centre for Geographical Studies, Institute of Geography and Spatial Planning, University of Lisbon (IGOT), Lisbon, PT< andrelopes@campus.ul.pt 3 Foundation for Science and Technology (FCT), Ministry of Science, Technology and Higher Education, Lisbon, PT 4 Interdisciplinary Center for Archaeology and Evolution of Human Behaviour, Faculty of Human and Social Sciences, University of Algarve, Faro, PT< catarinasbasilio@gmail.com ABSTRACT – This paper presents the results of the excavations carried out in the Chalcolithic con­texts from the walled enclosure of Ota (Alenquer, Portugal). Six new absolute dates allow the dis­cussion of the stratigraphical evidence and chronologically frame the zooarchaeological and tapho­nomical analysis of the faunal assemblage. Domesticated swine, caprine and bovine are prevalent, while wild species, most notably leporids, but also red deer, auroch and wild boar, among others, are less common. Exploitation and management of animals for the acquisition of primary and secon­dary products are inferred. Existing data suggests that the economic intensification that started dur­ing the previous phases was ongoing. KEY WORDS – Chalcolithic; southwestern Iberia; zooarchaeology; taphonomy; walled enclosure ‘?ivalska farma’> favnisti;ni zapis iz halkolitskega najdi[;a Ota (Alenquer, Portugalska) in njegov regionalni pomen IZVLE.EK – V .lanku predstavljamo rezultate izkopavanj halkolitskega konteksta v ogradi v Oti (Alenquer, Portugalska). .est novih absolutnih datumov omogo.a razpravo o stratigrafiji in krono­lo.ko zamejuje arheozoolo.ke in tafonomske analize .ivalskega zbira. V njem prevladujejo udoma-.eni pra.i.i, drobnica in govedo. Manj pogoste so divje vrste, zajci, navadni jelen, tur in divji pra.i.. Domnevamo, da so z .ivalmi upravljali in jih izkori..ali za primarne in sekundarne produkte. Po-datki ka.ejo, da se intenzivna izraba, ki se je za.ela v predhodnih fazah, nadaljuje. KLJU.NE BESEDE – halkolitik; jugozahodni Iberski polotok; arheozoologija; tafonomija; ograde Introduction The third millennium BCE is a key phase of human the Early Bronze Age (Valera 2015) are some of the history in the Iberian Peninsula. It is marked by an important topics that have been a matter of debate important social complexity and economic intensifi-among our colleagues. cation that started during the Late Neolithic (3500/ ~ 3200–3000 cal BCE) and further developed during The beginning of the third millennium BCE marks the Chalcolithic (3000–2000 cal BCE). The appear-the emergence of new domestic architectures in the ance and development of complex societies through-archaeological records, understood as “Walled en-out the Chalcolithic (Cruz Berrocal et al. 2013) and closures” (Jorge 2003). Although not exclusive to the different social perspectives in the transition to this area (Gonçalves 1989; Molina, Cámara 2005; DOI> 10.4312\dp.49.18 ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance Jorge et al. 2006; Mataloto 2010), the Portuguese Estremadura presents a high density of these types of sites. Its Western region concentrates 14 out of the 22 known walled enclosures, but the informa­tion on these is imbalanced – only four have had archaeological excavations in the last decade and a half, and only four have absolute chronologies, with the large site of Zambujal having 25 dates and the remaining three sites only having eight dates. Be­cause they are generally placed on the top of hills that grant them wide visual control of the surround­ing landscape, these sites are characterized by a new way of occupying space. They are recurrently placed near water sources that would function as a source of raw materials and subsistence, but also as a way of communication and connection to other areas and social networks. Other common characteristics are architecture and construction techniques. Struc­tures such as ‘walls’, towers and doors are built with the drystone technique reaching areas with a maxi­mum of 5ha (Kunst 2010) and structures with 200m in length (Texugo 2022). These structures are a no­velty in the region, and continuously go through re­formulations and changes that might bias the preser­vation of the archaeological record. Fauna has also been an important part of the discus­sion regarding economic intensification (Valente, Carvalho 2014; Almeida, Valera 2021; Almeida et al. 2021b). The larger sites from the Portuguese Es-tremadura have historically been used as guidelines for palaeoeconomic debates (e.g., Driesch, Boes­sneck 1976; Cardoso, Detry 2001/2002), but at the same time hindered the understanding of possible variability. The publication of other assemblages from the Estremadura (Correia 2015; Moreno-Gar­cía, Sousa 2015; Detry et al. 2020) and the com­parison with the southern Alentejo region (Arnaud 1993; Davis, Mataloto 2012; Costa 2013; Moreno-García 2013; Almeida, Valera 2021) is only more recently being achieved. The first results for animal diet and mobility have been published (Waterman et al. 2015; .alaité et al. 2018; Wright et al. 2019; Valera et al. 2020b), showing differences in the ma­nagement and exploitation of fauna not only be­tween the Neolithic and Chalcolithic but also with­in the Chalcolithic. In the western region, the 666m Montejunto hill is oriented NE-SW, visually impacting the area while also polarizing the surrounding landscape and cul­tures (Basílio, Texugo 2017; Basílio in press). This hill has several river springs that determine its re­lationship with the surrounding areas. From an ar­chaeological perspective, its natural caves have fu­nerary uses dated at least from the Middle Neolithic (Carvalho et al. 2019) to the Bronze Age (Gonçal­ves 1992). This highlights that there seem to be dif­ferent spatial relationships and meanings that might dictate how the communities were spread through­out the territory and the different cultural influen­ces and expressions (Basílio in press), as is the case of the archaeological site of Ota. In this paper, we present the results of the archaeological excavation of the Ota walled enclosure (Alenquer, Portugal), lo­cated less than 10km from Montejunto hill. We will focus on the unpublished absolute dates and the fau­nal record aiming to contribute to debates on re­source exploitation and management during the Chalcolithic in western Iberia. We will also present an in-depth taphonomical analysis of the assemblage, uncommon for the period and region under analy­sis, which will allow for a clearer understanding of the faunal accumulation and site use. Materials and Methods Materials Ota site is located about 50km north of Lisbon in the heart of the Portuguese Estremadura region, and, like its 22 counterparts, belongs to the walled enclo­sure phenomenon (Fig. 1). It shares, along with the great majority of these, chronological synchrony in its discovery in the first half of the 19th century, by Hipólito Cabaço, when the ground visibility scenario was different, mostly due to efficient forest manage­ment and the existence of communal herds. Current­ly, Ota is densely populated with Quercus coccife­ra, Olea europaea var. sylvestris, Rubus ulmifolius, Pinus pinaster and, in scarce instances, Eucalyptus globulus. This vegetation results from thin soils, al­ready attested in Ota during 2019, 2020 and 2021 fieldwork, reducing the surface visibility and the re­sults of archaeological surveys. In lithological terms, the region is composed of Mesozoic sedimentary rocks and a small area of Cenozoic sediments, mean­ing that past communities were surrounded by lime­stones, sands and clays (Ramos-Pereira et al. 2020), using them as the raw materials for different archi­tectures and artefacts. The geological framework ori­ginates in a landscape marked by small mountains, hills, interior plateaux, plains, and littoral platforms (Ramos-Pereira et al. 2020), with the western re­gion, where Ota is integrated, being particularly re­levant. Regarding climate, paleoenvironmental reconstruc­tions suggest that the first changes in regional vege­tation by anthropic influence, detected in pollen shifts, occurred around 5400 BCE, with the begin­ning of cereal cultivation (Lord et al. 2011). As for the mean sea level and coastline changes, it was observed that they remained stable between 9300 BCE and 2900 BCE (Lord et al. 2011). In this peri­od, periodic floods were identified, a behaviour that changed with the progression of marine transgres­sion (Lord et al. 2011). In sum, the chalcolithic cli­mate seems close to the current Estremadura cli­mate, characterized by a ‘Csa’ hot-summer Mediter­ranean climate (Kottek et al. 2006.Fig. 1) with ave­rage temperatures of 17°C and annual precipitation between 700mm and 900mm (Mora, Vieira 2020). Nelson J. Almeida, André Texugo, and Ana Catarina Basílio The archaeological site of Ota was the target of the archaeological surveys in the framework of a more exhaustive investigation (ROSETTA project) which aims to study the Chalcolithic architectures of the western walled enclosures through remote sensing. In the case of Ota, the LiDAR survey allowed the dis­covery of 21 high potential archaeological anoma­lies through the use of the LiDAR sensor in a system mounted in a UAS (a DJI Matrice 600 Pro with a Phoenix LiDAR Scout-8). To understand, evaluate, test and confirm the results obtained by remote sens­ing, three excavation campaigns were carried out. These resulted in 90 days of intervention between 2019 and 2021. The selected excavation areas relate not only to the anomalies detected but also to potential chalcolithic structures. The strategy adopted thus aimed to deter­mine, in the first place, a vital structure for the defi­nition of the walled enclosures – the wall. Next, and based on the previous knowledge of Wall 1, we un­derstood the potential use of the natural geological platforms to define an occupation made in embank­ments. This corresponds to Structure 9, which ma­terializes the symbiotic relationship between the anthropic architectures and the limestone substra­tum. Finally, the area where more work has been carried out relates to Structure 3, not only to under­stand the structure itself, possibly unique within the Chalcolithic walled enclosure phenomenon, but also to understand the sites’ historiography and Cabaço interventions, of which no records are available. Structure 1 The two surveys carried out on the wall, at different points of its layout, aimed at the typological classi­fication and clarification of its construction rhythms and temporalities. The first one is to the north end (survey 1), and the second survey is in the south ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance end, 150m away from the first one, giving a total of 38m2 of excavated area (Fig. 2). In stratigraphic terms, only two sedimentary depo­sits were intervened. The first, [1007], is also the oldest, lying directly on the limestone base platform. In this one, there were few records of fauna, cera­mic and lithic artefacts. However, there was very considerable chronological homogeneity, with all the elements pointing to prehistoric contexts. This deposit was covered and sealed by a second deposit, [1006], composed of a group of small and medium-sized stones enclosed in compact clayey sediment. The [1006] deposit contains the most significant ar­chaeological materials from Structure 1 – the wall – in which the bone of Ovis/ Capra was found, pro­viding the absolute dating of this structure. Fig. 2. Examples of the structures found in Ota: the final record of Structure 1 with the indication of pos­sible wall alignments in red; different stages of the excavation process of Structure 3; final record and section of Structure 9 with the upper and lower bedrock platforms separated by the structure. Nelson J. Almeida, André Texugo, and Ana Catarina Basílio Both the information relating to the materials and the 14C information (Tab. 1) allowed for the under­standing that this structure was already active dur­ing the third quarter of the third millennium BCE (the regional Middle/Late Chalcolithic), serving as a post-quem indicator. As such, the identified dynamics that were already broadly replicating some of the regio­nal prehistoric behaviours seem to be chalcolithic. Structure 9 The approach to Structure 9 (Fig. 2) arises from two main vectors. On the one hand, following its identi­fication through the data obtained by LiDAR. On the other, the results and information gathered at Wall 1, pointed to an occupation made through mul­tiple platforms. These seemed to result from the na­ture of the base geology and the addition of struc­tures carried out by the first communities that oc­cupied this site. With this in mind, a 2 by 6m survey was set up between the two platforms to characte-rise three apparently distinct spaces: (1) the upper platform; (2) the lower platform; (3) the possible natural or artificial structure between the two plat­forms. The analysis determined that the bedrock was between 5 and 20cm deep and that there were no preserved deposits. However, it was clear that the top and slope of the hill had an intense prehi­storic occupation, since most of the collected mate­rials corresponded to hand-made pottery, with and without the traditional regional chalcolithic motifs, siliceous lithic materials, and even a fragment of a limestone vase. Structure 3 This is the only ‘negative’ structure identified so far at the Ota archaeological site (Fig. 2). It was discovered dur­ing the field surveys, but it was also recorded in the Li-DAR scanning. However, it cor­responds to a structure al­ready intervened by Hipólito Cabaço, although no record or mention of its existence has been recovered. This is a reality of a circular tendency that replicates the construc­tion strategy observed in the other structures of Ota, based on a combination between the ceable in this structure, being partially modelled on the limestone bedrock, complemented by the place­ment of large stone blocks that help in the definition of the apparently intended circular layout. Besides these large blocks, which cover about 40 to 50% of the circumference of the structure, a possible con­tainment structure composed of an agglomerate of medium and small-sized limestones was detected to the west. Its functionality seems to be related to an eventual stabilization and support of the larger blocks (Texugo et al. in preparation). Another aspect of the manipulation of the base geology is the surface inside the structure. Relatively flattened outcrop ben­ches were also detected at other points of the archa­eological site; however, the interior of this structure is perfectly smoothed, suggesting human activity in the formation of a flattened floor. Regarding its excavation, the sediments from this structure were entirely sieved, and there was a dou­ble approach to the fieldwork: interior and exteri­or areas of the circular structure, resulting in an excavated area of 98m2. As for the stratigraphic se­quence, we can state that the deposits around Struc­ture 3 come from last century’s excavations, and that the present materialities were selected with the clear obliteration of materials with greater value – the de­corated ceramics and other goods of greater aesthe­tic value. This situation is opposite to that reflected by the materials collected by Hipólito Cabaço, where no undecorated ceramics are present. After the ex­cavation of these superficial contexts – [1104] to [1110] – it was possible to understand that none of the deposits were preserved, either because they Lab. Sample Context BP date Cal BCE (2s ) Beta-561854 Ovis\Capra Structure 1 .1006] 3960±30 2571–2516 (32.8%) rib 2502–2400 (52.7%) 2383–2347 (10%) Beta-568786 Capra hircus Structure 3 .1103] 3970±30 2575–2444 (87.6%) horn core 2424–2404 (3.3%) 2379–2350 (4.6%) Beta-612398 Sus sp. Structure 3 3860±30 2460–2276 (80.5%) phalange (exterior) .1120] 2256–2206 (15%) Beta-612399 Bos sp. Structure 3 .1115] 2576–2454 (93.4%) phalange 3990±30 2418–2409 (0.9%) 2369–2356 (1.1%) Beta-612400 Sus sp. Structure 3 3980±30 2576–2454 (93.4%) cranium (exterior) .1120] 2418–2409 (0.9%) 2369–2356 (1.1%) Beta-612401 Sus sp. tooth Structure 3 (exterior) .1120] 4000±30 2578–2463 (95.4%) local geology and the addition Tab. 1. Absolute dates obtained for Structures 1 and 3 of Ota. Calibration of constructive elements. This of 14C dates using IntCal20 calibration curve (Reimer et al. 2020) and the symbiosis is particularly noti-OxCal v4.4 program (Bronk Ramsey 2009). ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance were previously excavated or because they were de­cisively affected by intense bioturbation. Nonethe­less, it is possible to conclude that the overwhelm­ing majority of the archaeological materials in these strata belong to late prehistory. Overall, only five stratigraphic units were preserved, with coherent contextual information. Inside Structure 3, [1003] and [1115] correspond to the remains left by Caba­ço, having provided a sample of a Capra hircus horn core with a date fitting into the Middle Chalco­lithic (2574–2350 BCE) and of a Bos sp. phalange with a very similar date – 2577–2459 BCE (Tab. 1). The sediments outside Structure 3 were quite affect­ed and it was only possible to record two preserved contexts in an area with strong anthropic alterations of the base lapis. In addition to this manipulation of space, the presence of several horizontally deposit­ed faunal elements – [1120] – further corroborates the probable preservation of the deposits, with the repeated presence of human action in the organiza­tion of the elements. [1120] was on top of [1121], which, in line with what was noted in the last depo­sit, included various unstructured faunal elements with cohesive materials framed in regional Chalco­lithic social dynamics. Methods The assemblage provenance is diverse and compris­es remains from all the stratigraphical units with absolute or relative chronologies pointing to the Chalcolithic period, independently of the degrees of bioturbation recorded. A larger amount of remains comes from [1104] (n=1450, 24.1%), [1111] (n=959, 16%), [1114] (n=777, 12.9%), [1117] (n=427, 7.1%), [1120] (n=385, 6.4%), and [1113] (n=357, 5.9%). Remaining units have <5% of the NSP: [1000], [1006], [1007], [1008], [1101], [1102], [1103], [1105], [1106], [1107], [1108], [1109], [1110], [1115], [1116], [1118], [1119], and [1121]. Methodologies for the zooarchaeological and tapho­nomical analysis of faunal assemblages were fol­lowed (Lyman 1994; Reitz, Wing 2008). Data are presented according to the number of specimens (NSP), number of identified specimens (NISP), mini­mum number of elements (MNE), and minimum num­ber of individuals (MNI) (Grayson 1984; Lyman 2008). The MNI was calculated by distinguishing stra­tigraphic units (MNIsu) and as a general cumulative estimate not considering differences between strati­graphic units (MNIg). Linear biometrics analysis was carried out by mea­suring bones and teeth with a Lux digital calliper fol­lowing current standards (Driesch 1976; Payne, Bull 1988; Davis 1996; Albarella et al. 2005; Salva­gno, Albarella 2017). The results were compared to regional ‘contemporaneous’ published measure­ments and were considered together with morpho­logy to better characterize bovine, swine, caprine and leporids (Boessneck et al. 1964; Boessneck 1970; Callou 1997; Zeder, Pilaar 2010; Zeder, Lap-ham 2010). Taxonomically indeterminate remains were tentatively classified according to generic weight groups (Brain 1981; Bunn 1983; 1986): in­determinate (WG 0), <20kg (WG 1), 20–100kg (WG 2), 100–300kg (WG 3), and >300kg (WG 4). The age-at-death estimate considered bone (general os­sification and epiphysis fusing) and teeth develop­ment (eruption, replacement, wear) according to data published for the main species documented in the assemblage (Payne 1973; 1987; Bull, Payne 1982; Grant 1982; Jones 2006; Zeder 2006; Lemo­ine et al. 2014; Zeder et al. 2015). The results were grouped in general age groups, namely perinatal, in­fant, juvenile, sub-adult, adult, and senile. The assemblage breakage patterns were assessed fol­lowing Henry T. Bunn (1983) and Paola Villa, and Eric Mahieu (1991). Diaphysis length (<25%, 25– 50%, 50–75%, >75%) and section (<25%, 25–75%, >75%) completeness are presented in relation to the original figures. Breakage planes outline (oblique, longitudinal, transverse), angle (mixed, oblique, right), and surface (jagged, smooth) are considered according to the weight groups mentioned above. The surfaces of the remains were macroscopically and mi­croscopically analysed to record possible BSMs (bone surface modifications) related to processing and con­sumption, but also the sedimentary environment. We searched for anthropogenic breakage (e.g., peeling, percussion impacts, cones, fissures), cutmarks (e.g., incisions, chop marks, scrape marks), tooth marks (e.g., pits, punctures, crenulated edges), other con­sumption indicators (e.g., furrowing, digestion), and thermal alteration (i.e. boiling, burning) (Binford 1978; 1981; Brain 1981; Shipman 1981; White 1992; Stiner et al. 1995; Pickering et al. 2013; So-lari et al. 2015). When present, their location, dispo­sition, relations, typology, morphology, size and in­tensity are described (Almeida 2017). The furrow­ing evaluation follows the proposal of Saladié et al. (2011; 2013) according to the intensity of tissue loss distinguishing between light, moderate, and heavy furrowing. The location of tooth marks was also re­corded considering the type of tissue affected (corti­cal, thin cortical, cancellous – Selvaggio, Wilder 2001; Domínguez-Rodrigo, Piqueras 2003). Nelson J. Almeida, André Texugo, and Ana Catarina Basílio Weathering is presented according to the degrees suggested by Behrensmeyer (1978). Other taphono­mical indicators, such as the presence of vermicula­tions (Lyman 1994), manganese oxide precipitation (López-González et al. 2006), indeterminate chemi­cal corrosion (Fernández-Jalvo et al. 2002), tram­pling (Behrensmeyer et al. 1986; Shipman, Rose 1984), rodent gnawing (Shipman 1981), or concre­tions (Courty et al. 1989) were qualitatively record­ed according to the intensity and surface altered (<25%, 25–50%, 50–75%, >75%) (Almeida 2017). Results Anatomy and taxonomy The assemblage comprises a majority of Mammalia (NISP% 96.5%) remains, with some amount of Bival-via (NISP% 2.4%) and to a lesser extent Amphibia (NISP% 0.5%), Reptilia (NISP% 0.5%), and Aves (NISP% 0.2%) (Tabs. 2, 3, 4). Swine are prevalent, with the majority of them being specifically indeter­minate, and some remains being identified as pos­sible wild boar and pig due to morphology and size. As can be seen in Figure 3, there is some superimpo­sition between Mesolithic wild boar and Late Prehi­story (presumable) pig for the astragalus Greatest Lateral length (GLl) measurement. For the case of Ota and considering the results for other large Chal­colithic assemblages, it seems that some of the astra­gali correspond to small swine, and at least the larg­er one would probably correspond to wild boar. Still, the possible presence of hybrids in the assem- Fig. 3. Histogram of results ob­tained for the Greatest Lateral length (GLl) of swine astraga­lus from the Ota site and other Portuguese and adjacent Spa­nish sites dated to the Pleistoce­ne and Mesolithic (Detry 2007); Neolithic (Almeida 2017; Davis et al. 2018; Encarnaçao, Almei­da 2017; Almeida et al. 2021b); Chalcolithic (Driesch, Boes­sneck 1976; Cardoso, Detry 2001/2002; Castanos 1992; 1997; Rodríguez-Hidalgo, Cabe­zas 2011; Davis, Mataloto 2012; Moreno-García 2013; Correia 2015; Moreno-García, Sousa 2015; Detry et al. 2020; Almei­da et al. 2021a; Almeida, Vale­ra 2021; Cardoso et al. 2021; Pereiro et al. 2021); Iron and Bronze Age sites (Davis 2006; Almeida et al. 2020b). blage must be considered, as pigs were probably bred free. Swine show an important number of iso­lated teeth (n=68). All parts of the skeleton are pre­sent, but vertebrae and ribs are under-represented. A higher number of anterior in comparison to poste­rior long bones is evident according to NISP and MNE values. The obtained MNIs indicate the presence of adult and senile wild boar; juvenile, sub/adult and adult pigs; and infant/juvenile, juvenile, and sub-adult swine that further enlarge the MNIs (Tab. 5). Hence, different age groups are represented in the assemblage. Caprine are well represented, comprising a majority of goat/sheep and similar NISP values for remains identifiable specifically as goat or sheep. Similarly to swine, isolated teeth are also frequent in caprine remains (n=75), with the appendicular skeleton being more abundant than other body parts. The di­stinction between goat and sheep was achieved based on the morphology of horn core fragments, isolated teeth, humerus, radius, metapodial, calca­neus, phalanges, and both morphology and measu­rements of astragali following Simon J. M. Davis (2017b). Sub-adult, adult and adult/senile individu­als of goats, and juvenile, sub-adult and adult indivi­duals of sheep are countable. The goat/sheep follow this pattern even if adding one juvenile individual in the MNIg. Bovine comprises domesticated cattle and wild au­rochs. The latter are scarce in comparison to the ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance NSP % MNE % MNIsu % MNIg % of the GLl, which also show- MAMMALIA ed its small size (Fig. 5). Bo- Artiodactyla 17 1.7 0 0.0 0 0.0 0 0.0 vines have all body parts, but Bos taurus 78 7.6 63 10.1 16 12.5 3 6.5 the higher frequency of pha- Bos cf. primigenius 13 1.3 11 1.8 7 5.5 2 4.3 langes and carpal/tarsal bo- Bos sp. Capra hircus Ovis aries 92 22 16 8.9 2.1 1.6 48 15 15 7.7 2.4 2.4 1 7 8 0.8 5.5 6.3 1 6 2.2 13.0 nes is of interest. Older indi­viduals, i.e. adult or adult/ Ovis\Capra 177 17.2 81 13.0 6 4.7 1 2.2 senile, and several sub-adult/ Cervus elaphus 72 7.0 50 8.0 12 9.4 2 4.3 adult individuals are present. Sus cf. scrofa 7 0.7 4 0.6 5 3.9 2 4.3 Noteworthy is the presence of Sus cf. domesticus 21 2.0 17 2.7 9 7.0 5 10.9 one infant/juvenile cow and Sus sp. cf. Equus sp. 271 1 26.4 0.1 157 1 25.2 0.2 7 1 5.5 0.8 6 1 13.0 2.2 one juvenile/sub-adult auroch. Herbivore Canis sp. 28 3 2.7 0.3 0 3 0.0 0.5 0 2 0.0 1.6 0 1 0.0 2.2 Leporids correspond mostly Felis silvestris 1 0.1 1 0.2 1 0.8 1 2.2 to the European wild rabbit Oryctolagus cuniculus 151 14.7 120 19.2 28 21.9 9 19.6 and also have an important Leporidae 25 2.4 16 2.6 0 0.0 0 0.0 number of remains. Except Sub-total Mammalia 995 96.8 603 96.5 0 0.0 0 0.0 for one sub-adult, all individ- AVES Aves ind. Sub-total Aves 1 1 0.1 0.1 1 1 0.2 0.2 1 0 0.8 0.0 1 0 2.2 0.0 uals are adults. Both the cra­nial and appendicular body AMPHIBIA parts are well represented, Amphibia 2 0.2 1 0.2 1 0.8 0 0.0 mainly mandibles, innomi- Anura 2 0.2 2 0.3 1 0.8 1 2.2 nate, femur and tibia. The Sub-total Amphibia 4 0.4 3 0.5 0 0.0 0 0.0 smaller bones from the extre- REPTILIA cf. Mauremys leprosa Sub-total Reptilia 3 3 0.3 0.3 3 3 0.5 0.5 2 0 1.6 0.0 1 0 2.2 0.0 mities are underrepresented, but it is not clear if this rela- BIVALVIA tes to the selection, survival Bivalvia ind. 7 0.7 4 0.6 3 2.3 0 0.0 or recovery bias since sieving Ruditapes decussatus 4 1.4 7 1.1 6 4.7 1 2.2 was implemented in the area Cerastoderma edule 2 0.2 2 0.3 2 1.6 1 2.2 containing the majority of the Pecten sp. 2 0.2 2 0.3 2 1.612.2 assemblage. Red deer are the Sub-total Bivalvia 25 2.4 15 2.4 0 0.0 00.0 second more numerous spe­ Sub-total identified 1028 100 625 100 128 100 46 100 cies representing wild game, Indeterminate and the different bones, teeth Weight group ind. 2294 46.1 Tab. 2. Absolute and relative val-and antlers seem to corre- WG .20kg 81 1.6 ues obtained for the taxonomical-spond to adults. Canids are re- WG .100kg 294 5.9 ly identified and unidentified fau- presented by two metapodial WG 20-100kg 1421 28.5 nal remains according to the num- WG .300kg 231 4.6 and one phalange, with a pos­ ber of identified specimens (NSP), WG 100-300kg 102 2.0 sible equid phalange and a minimum number of elements WG .100kg 281 5.6 wild cat phalange recovered. (MNE), and minimum number of WG .300kg 276 5.5 individuals by stratigraphic unit Sub-total indeterminate 4980 100 (MNIsu) and minimum number of One must emphasize the re- Total 6008 individuals general (MNIg). covery of some Bivalvia shell fragments (Fig. 6). Although large number of remains identified as Bos taurus some of them are considered taxonomically indeter-and Bos sp., the latter generally also have smaller minate (n=7, 28%), the grooved carpet shells are sizes and probably corresponded to the domesticat-abundant, and the scarce common cockle and scal­ed form as well. The most abundant measurable lop are present. bone used for this distinction are proximal phalan­ges, which in the case of the Ota assemblage are ge-Taphonomy nerally coincident with the sizes considered to cor-The assemblage is composed by a majority of re-respond to cattle, even if a larger specimen falls with-mains with maximum dimensions <5cm (n=5242, in the admitted variability of the much larger auroch 87.3%), others between 5 and 10cm (n=727, 12.1%) (Fig. 4). One astragalus allowed for the measurement and only a few with 10–15cm (n=32, 0.5%) or 15– Tab. 3. Values of the number of identified specimens (NISP) per body part. Legend: AR artiodactyl, BT Bos taurus, BP Bos cf. primigenius, BOS Bos sp., CH Capra hircus, OA Ovis aries, O/C Ovis/Capra, CEE Cervus elaphus, SS Sus cf. scrofa, SD Sus cf. domesticus, SUS Sus sp., EQ cf. Equus sp., HE herbivore, CAN Canis sp., FS Felis silvestris, ORC Oryctolagus cuniculus, LEP Leporidae, AV Aves ind., AMP Amphibia ind., AN Anura, ML cf. Mauremys leprosa, BIV Bivalvia, RD Ruditapes decussatus, CE Cerastoderma edule, PE Pecten sp., IND indeterminate. Nelson J. Almeida, André Texugo, and Ana Catarina Basílio AR BT BP BOS CH OA O\C CEE SS SD SUS EQ HE CAN FS ORC LEP AV AMP AN ML BIV RD CE PE IND Total axial cranial skeletonhorn core\antler 2 7 1 20 18 48 cranium (maxilla) 1 2 1 5 8 11 1 159 188 mandible 3 7 51119 1 18 32 78 isolated incisor 1 9 19 1 1 31 isolated canine 3 14 0 17 isolated pre-molar 2 1 6 3 11 3 10 4 40 isolated molar 7 1 22 1 150 3 3 18 24 8 4 142 isolated tooth 1 2 1 61 65 axial post-cranial skeletonvertebra 6 2 9 55 203230 rib 124 124 sacrum 2 1 2 510 appendicular skeletonscapula 11 1 4 1 412 122 24 62 humerus 2111163 220 81 13 59 radius 111 1 11102 212 6 20 58 ulna 43 62 11 71 236 metacarpal 1 41141 7 1 0 20 ilium 1 01 pelvis 2 2 22 113 22 1357 femur 1 1 11 11 175 1653 patella 2 3 05 tibia 1 3 32 26 145 440 fibula 6 39 tibio-fibula 1 0 1 tarsal-metatarsal 1 0 1 metatarsal 4 21 92 9 1 111 0 40 metapodial 154 4 244 1 15 1 1 6 75 astragalus 3 1 262 4 11 4 2 26 calcaneus 2 1 22 10 9 329 carpal\tarsal 7 5 3 8 5 6 10 44 sesamoid 10 1 11 phalange 1 1181532571 141 11 2 5 67 continue phalange 2 81 1 123 4 14 2 phalange 3 51 5 16 2 10 0 30 phalange 6 other long bone 1292 1292 flat bone 2 1587 1589 indeterminate bone 1360 1360 carapace\plastron 3 0 3 shell 714 2 2 0 25 total 177813922216177727212711283 1151251 2 237142249806008 Tab. 4. Values of the minimum number of elements (MNE) per body part, excluding loose teeth. Legend: BT Bos taurus, BP Bos cf. primigenius, BOS Bos sp., CH Capra hircus, OA Ovis aries, O/C Ovis/Capra, CEE Cervus elaphus, SS Sus cf. scrofa, SD Sus cf. domesticus, SUS Sus sp., EQ cf. Equus sp., HE herbivore, CAN Canis sp., FS Felis silvestris, ORC Oryctolagus cuniculus, LEP Leporidae, AV Aves ind., AMP Amphibia ind., AN Anura, ML cf. Mauremys leprosa, BIV Bivalvia, RD Ruditapes decussatus, CE Cerastoderma edule, PE Pecten sp. BT BP BOS CH OA O\C CEE SS SD SUS EQ CAN FS ORC LEP AV AMP AN ML BIV RD CE PEC Total axial cranial skeleton horn core\antler 2 4 1 6 13 cranium (maxilla) 1 1 1 4 4 10 21 mandible 3 4 4 1117 15 36 axial post-cranial skeleton vertebra 5 2 8 54 24 rib 0 sacrum 11 2 4 appendicular skeleton scapula 111 41 47 112 32 humerus 2111153 29 71 33 radius 11 117 2 210 6 31 ulna 31 629 7 28 metacarpal 1 3 114 1 7 1 19 ilium 1 1 pelvis 2 2 22 110 18 femur 11 11 9 63 patella 2 3 5 ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance tibia 1 3 2224 133 30 fibula 3 tibio-fibula 1 tarsal-metatarsal 1 continue metatarsal 3 11 42 9 1 11 32 metapodial 4 11 31 4 1 24 astragalus 3 1 262 4114 calcaneus 2 1 22 7 9 23 carpal\tarsal 7 5 3 8 5 6 34 sesamoid 10 10 phalange 1 161 4 324 71 131 1 1 2 56 phalange 2 71 123 3 14 31 phalange 3 51 5 16 2 10 30 other flat bone 1 1 carapace\plastron 3 3 shell 4 722 15 total 7313 611916129557201991 3 112116 1 1 2 3 4 722756 Fig. 4. Scatterplot showing the breadth of the proximal end (Bp) and the greatest length of the peripheral half (GLpe) mea­surements obtained for bovine phalange 1 from Portuguese and adjacent Spanish Chalcolithic assemblages (Driesch, Boes­sneck 1976; Castanos 1992; 1997; Rodríguez-Hidalgo, Cabezas 2011; Detry et al. 2020; Almeida, Valera 2021). 18cm (n=7, 0.1%) (Fig. 7). The results obtained for the diaphysis completeness are interesting, with clear differences between long bones of WG 1 and 1/2 and the remaining WGs. In the latter, both WG 2+2/3+3 and WG 3/4+4 show higher frequencies of length and section <25% of the original, while in WG 1+1/2 the length of 25–50% and 50–75%, and the 25–75% section are bet­ter represented. It thus seems that the long bones of smaller ani­mals, which correspond mostly to leporids or animals of similar size, are better preserved in comparison to larger animals, from the size of caprine, swine, cervids and bovids. To better understand this data, we looked at the breakage planes (Fig. 7). Longitudinal outlines are abundant in the different sam­ples, followed by oblique outlines, but these have values between around 28–35%. Angles and edges show a similar pattern, with right degrees and smooth surfaces always above 66%. These pat­terns together with the diaphysis completeness and the indicators of anthropogenic breakage suggest that although fragmentation occurred, an important amount of green breakage exists. The dif­ferent degrees of preservation occurring with the smaller size ani­mals can also relate to the breakage of near to epiphysis portions commonly occurring with human-induced breakage to access mar­row that can result in shaft cylinders. Complete remains (3.2%) comprise mainly 49 teeth (23.4%), 60 carpal and tarsal bones (31.1%), 65 phalanges (33.7%) and 10 metapodia (5.2%). Body parts with greater nutritional value are rarely complete, but this can also relate to density-mediated attri­tion. Indicators of thermal modification are present (Tab. 6). Burning damage (2.9%) is more common than other human-related indi­cators, and is present mainly in taxonomically indeterminate re- ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance Fig. 5. Histogram of the results obtained for the Greatest Late­ral length (GLl) of bovine as-tragalus from the Ota site and other Portuguese and adjacent Spanish sites dated to the Meso­lithic (Detry 2007; Valente 2008; 2013); Neolithic (Davis et al. 2018; Almeida et al. 2021b); Chalcolithic (Driesch, Boes­sneck 1976; Castanos 1997; Ro-dríguez-Hidalgo, Cabezas 2011; Davis, Mataloto 2012; Moreno-García 2013; Correia 2015; Mo-reno-García, Sousa 2015; Aleixo 2018; Davis et al. 2018); Iron and Bronze Age sites (Davis 2006; 2017a). mains, of which we must emphasize WG2 and larger animals. When present in identified remains (n=19), they are mostly in rabbits (n=7) and bovines (n=4). The body parts altered are generally elements from the appendicular (n=93, 54.1%) skeleton or indeter­minate (n=68, 39.5%), although some axial bones (n=19, 5.8%) and a scallop shell were also burnt. The degrees are generally low, with degrees 1 (n=5, 2.9%) and 2 (n=91, 52.9%) being prevalent, but de­grees 3 (n=32, 25%), 4 (n=27, 15.7%) and 5 (n=6, 3.5%) are also present. Double colourations occur in 14.5% (n=25) of the cases. Possible boiling is almost entirely restricted to indeterminate fragments (n= 110, 91.7%), half of which are from 20–100kg ani­mals. The axial skeleton (n=2, 1.7%) has lower fre­quencies in comparison to the appendicular (n=78, 65%) or indeterminate (n=40, 33.3%), and inde­terminate long bones are prevalent (n=68, 56.7%). Butchering practices are recognizable in the collec­tion (Tab. 7), with cutmarks comprising chop marks (n=14), incisions (n=116), zigzag marks (n=2) and complete sectioning (n=2) of bones. All main but­chering phases are present, including dismember­ment, skin removal, evisceration, segmentation of the axial skeleton, disarticulation and filleting. Among anthropogenic breakage, impact points (n=63) are the most abundant type of stigma, followed by cor­tical extractions (n=24) and impact cones (n=18). Other percussion stigmas (n=28) such as fissures, possible counterblows, pitting, anvil abrasions and adhering flakes, are recorded. Considering consumption, the majority of remains with taphonomical indicators were associated with carnivore action (n=127, 77.9%), with possible hu­man tooth marks (n=20, 9.8%) being almost entirely circumscribed to leporid remains (n=18), and some considered indeterminate (n=16, 12.3%). Taxono­mically identified tooth-marked bones (n=78) are mostly from swine (n=30), leporid (n=25) or capri­ne (n=12), whilst indeterminate bones (n=71) are IN\JU JU JU\SU SU SU\AD AD AD\SE SE MNIsu\MNIg Bos taurus 1–1 0–0 0–0 0–0 9–0 6–2 0–0 0–0 16–3 Bos cf. primigenius 0–0 0–0 1–1 0–0 5–0 0–0 1–1 0–0 7–2 Bos sp. 0–0 0–0 1–0 0–0 0–0 0–0 0–0 0–0 1–0 Capra hircus 0–0 0–0 0–0 0–0 1–0 5–0 1–1 0–0 7–1 Ovis aries 0–0 1–1 0–0 0–0 1–0 6–5 0–0 0–0 8–6 Ovis\Capra 0–0 0–1 2–0 1–0 1–0 2–0 0–0 0–0 6–1 Sus cf. scrofa 0–0 0–0 0–0 0–0 0–0 4–1 0–0 1–1 5–2 Sus cf. domesticus 0–0 3–2 0–0 0–0 3–1 3–2 0–0 0–0 9–5 Sus sp. 2–2 1–0 0–0 0–0 4–4 0–0 0–0 0–0 7–6 Oryctolagus cuniculus 0–0 0–0 0–0 1–1 0–0 27–8 0–0 0–0 28–9 Tab. 5. Minimum number of individuals calculated distinguishing between stratigraphic units (MNIsu) and considering the entire assemblage (MNIg). Legend: IN infant, JU juvenile, SU sub-adult, AD adult, SE senile. Nelson J. Almeida, André Texugo, and Ana Catarina Basílio mainly from small (20–100kg) animals, thus very small animals with tooth marks are almost unique­ly leporids. Consumption indicators are present in appendicular (n=117, 71.8%) bones, and compara­tively less in axial (n=12, 7.4%) or indeterminate bones (n=34, 20.9%). Among indicators of consump­tion, pits (n=48, 23–6%), punctures (n=46, 22.7%) and furrowing (n=37, 18.2%) are more frequent than crenulated edges (n=24, 11.8%), digestion (n=21, 10.3%), notches (n=18, 8–9%), scores (n=4, 2%), chipped back-edge (n=3, 1.5%), crushing (n=1, 0.5%) and shaft cylinders (n=1, 0.5%). Furrowing is main­ly in heavy degrees (n=22, 59.5%) compared to mo­derate (n=11, 29.7%) and light (n=4, 10.8%) degrees. The larger number of measurements obtained for pits/punctures are presented in Table 8. Not consi­dering the data from very small animals (WG 1) due ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance Fig. 7. Assessment of breakage per weight group in the assemblage considering: a fragment size, b diaphy- sis completeness, and c fracture planes. to sample size, we present a comparison of the re­maining results with published experimental and actualistic measurements obtained for wolf, dog, wild boar, pig and humans (Delaney-Rivera et al. 2009; Saladié 2009; Andrés et al. 2012; Saladié et al. 2013). Results for cancellous tissues show simila­rities with canids, swine and humans, while cortical/ thin cortical data is better framed within the varia­bilities of canids and swine (Fig. 8). We cannot dis­card the possibility that some smaller and shallow tooth marks could relate to human consumption. At the same time, the absence of the characteristic sco­res made by swine while feeding suggests that they were not preponderant in the modification of these remains. Canids, probably dogs, seem to fit better with the taphonomical patterning recorded and the different types of behavioural consumption indica­tors. Finally, other indicators are present in the faunal as­semblage. Trampling, chemical corrosion and con­cretions are scarce. The higher frequencies of wea­thering (10%), manganese oxide precipitation (10%) and vermiculations (10%) are noteworthy (Fig. 9). Weathering occurs mostly in initial degrees 1 and 2, but this together with the identification of degrees 3 and 4 suggests that while the exposition of the re­mains before sedimentation was small, some remains were subjected to larger exposure or at least some moments of re-exposure. Vermiculations are relat­able with the provenance of remains from low-depth stratigraphy and/or high bioturbation areas, since these normally occur in the top horizons of the stratigraphy. The presence of manganese oxide in degrees 1 and 2 accompanied by some amount of degree 3 and the little degree 4 can relate to humi­dity and decomposition of organic matter. Overall, the low incidence of weathering, vermiculations and scarcity of trampling suggests that although the site shows important bioturbation and disturbance of deposits due to previous historical archaeological works, the faunal remains are fairly preserved. Discussion The faunal record in the Ota site during the Chalcolithic The fauna profile identified so far in the Ota site shows the importance of swine in the economy of these groups during the Chalcolithic. Although it was impossible to further separate the majority of the remains due to the lack of metrical or morpho­logical characteristics, wild boar and pig seem to have been present and, probably, hybrids. The ge­nerally small size of measurable and unmeasurable bone and teeth is suggestive of a small input from larger individuals, presumably wild boars. Caprines are also noteworthy in terms of NISP and MNE, with Nelson J. Almeida, André Texugo, and Ana Catarina Basílio Indicator n % Cutmark 133 2.2 Anthropogenic breakage 126 2.1 Burning damage 172 2.9 Boiling 120 2.0 Tooth marks, digestion 170 2.8 Rodent marks 3 0.0 Vermiculations 624 10.4 Weathering 592 9.9 Trampling 9 0.1 Concretions 9 0.1 Manganese oxide 624 10.4 Chemical corrosion 5 0.1 Complete remains 193 3.2 Recent breakage 954 15.9 Tab. 6. Main taphonomical indicators identified in the Ota faunal assemblage. both goat and sheep present, and a majority of re­mains classified as goat/sheep. The triad of domes­ticated species is completed with the bovine, for which a small number of remains was classified as probable auroch due to their large size. Still, the ma­jority of evidence is from smaller individuals of cat­tle and other bovines that probably correspond to cattle due to their small size. Kill-off patterns among the main domesticated taxa show the prevalence of adult individuals indepen- Burning Boiling Cutmark Breakage Tooth mark Digestion n % n % n % n % n % n % Artiodactyla 0 0.0 0 0 0 0.0 0 0.0 2 2.6 0 0 Bos taurus 1 5.3 0 0 12 17.9 0 0.0 0 0.0 0 0 Bos primigenius 0 0.0 0 0 3 4.5 5 19.2 3 3.8 0 0 Bos sp. 3 15.8 0 0 6 9.0 10 38.5 2 2.6 0 0 Capra hircus 1 5.3 0 0 3 4.5 1 3.8 0 0.0 0 0 Ovis aries 0 0.0 0 0 3 4.5 0 0.0 0 0.0 0 0 Ovis\Capra 1 5.3 1 10 8 11.9 0 0.0 12 15.4 2 20 Cervus elaphus 1 5.3 1 10 7 10.4 3 11.5 4 5.1 0 0 Sus cf. scrofa 0 0.0 0 0 1 1.5 1 3.8 0 0.0 0 0 Sus cf. domesticus 0 0.0 0 0 5 7.5 0 0.0 4 5.1 0 0 Sus sp. 2 10.5 3 30 19 28.4 6 23.1 26 33.3 6 60 Herbivore 2 10.5 0 0 0 0.0 0 0.0 0 0.0 0 0 Oryctolagus cuniculus 7 36.8 4 40 0 0.0 0 0.0 23 29.5 1 10 Leporidae 0 0.0 1 10 0 0.0 0 0.0 2 2.6 1 10 Sub-total identified 19 100 10 100 67 100 26 100 78 100 10 100 Weight group 0 ind. 48 31.4 22 20.0 24 36.4 9 9 12 16.9 6 54.5 WG 1 (.20kg) 2 1.3 0 0.0 0 0.0 0 0 1 1.4 1 9.1 WG 1\2 (.100kg) 7 4.6 8 7.3 1 1.5 1 1 1 1.4 0 0.0 WG 2 (20-100kg) 62 40.5 55 50.0 14 21.2 54 54 37 52.1 4 36.4 WG 2\3 (.300kg) 10 6.5 8 7.3 8 12.1 6 6 8 11.3 0 0.0 WG 3 (100-300kg) 6 3.9 3 2.7 2 3.0 4 4 1 1.4 0 0.0 WG 3\4 (.100kg) 6 3.9 5 4.5 8 12.1 9 9 6 8.5 0 0.0 WG 4 (.300kg) 12 7.8 9 8.2 9 13.6 17 17 5 7.0 0 0.0 Sub-total indeterminate 153 100 110 100 66 100 100 100 71 100 11 52 Total 172 120 133 126 149 21 Tab. 7. Different taphonomical indicators’ absolute and relative values per species and weight groups. ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance Fig. 8. Comparison of results (in mm) obtained for the maximum (length) and minimum (width) axis of pits/punctures recorded in the Ota assemblage according to the type of bone tissue. Values presented are the mean and CI 95%. Legend: *Andrés et al. 2012; “Saladié 2009; /Saladié et al. 2013, )Delaney-Rivera et al. 2009. dently of the MNI calculated. Considering swine, it is not easy to understand age patterns due to the su­perimposition of wild and domesticated specimens. All the larger size individuals identified as possible wild boar are adult or senile, with the possible pigs comprising juvenile, sub-adults/adults and adults, and the wild boar/pig adding infant/juvenile indivi­duals. These omnivorous animals were being slaugh­tered while the meat was tender, but older indivi­duals are also present and they could relate to re­productive purposes. Caprine are represented by dif­ferent individuals, from juvenile to adult/senile in­dividuals, with a clear higher proportion of adults and even one case of one adult/senile goat. More in­formation is needed to better understand the kill-off pattern in caprine, but a focus on older individuals is clear. The large majority of the assemblage comes from contexts that were sieved so we do not expect a bias against small teeth (including deciduous te­eth). An infant/juvenile cattle individual was iden­tified based on the presence of scarce unfused ele­ments and low ossification bones, but teeth that could allow for a better age estimate were not reco­vered. The presence of younger bovine can relate to the acquisition of primary and secondary products, since the slaughter of calves can relate to the need to reduce the amount of fodder needed and no fur­ther need for milk exploitation, or the use of one calf to stimulate several cows (Vigne, Helmer 2007). The slaughter of older cattle after they started to be C.I. C.I. N Mean SD +95% -95% Min Max 95%CI length WG 1 cortical\thin cortical 2 4.89 0,45 5.50 4.27 4.57 5,20 0.62 length WG 2, 2\3 cortical\thin cortical 125 2.46 0,83 2.61 2.31 1.02 5,29 0.15 length WG 3, 4 cortical\thin cortical 24 2.77 0,74 3.06 2.47 1.67 4,27 0.30 width WG 1 cortical\thin cortical 2 2.59 0,60 3.42 1.75 2.16 3,01 0.83 width WG 2, 2\3 cortical\thin cortical 125 1.58 0,55 1.67 1.48 0.51 3,76 0.10 width WG 3,4 cortical\thin cortical 24 1.98 0,48 2.17 1.79 1.24 3,22 0.19 length WG 1 cancellous 7 2.52 0,86 3.15 1.88 1.64 4,25 0.64 length WG 2, 2\3 cancellous 115 2.95 1,21 3.17 2.73 1.31 7,49 0.22 length WG 3,4 cancellous 38 3.48 1,03 3.80 3.15 1.83 5,64 0.33 width WG 1 cancellous 7 1.89 0,48 2.25 1.53 1.32 2,73 0.36 width WG 2, 2\3 cancellous 115 1.96 0,75 2.10 1.82 0.59 4,83 0.14 width WG 3,4 cancellous 38 2.60 0,91 2.89 2.32 0.83 5,07 0.29 Tab. 8. Descriptive statistics for the length and width (in mm) of pits and punctures on cortical/thin cor­tical and cancellous tissues recorded on identified and indeterminate remains identified to weight group (SD standard deviation, CI confidence interval). Nelson J. Almeida, André Texugo, and Ana Catarina Basílio less productive is a common practice, especially if older indi­viduals are not needed for use as traction (Pérez Ripoll 1999). Clear hunting practices are re­corded. Besides the smaller amount of red deer adult indi­viduals, leporids, probably only consisting of the European rab­bit and almost entirely adult in­dividuals, show important NISP, MNE and MNI results. Other taxa comprise a possible equid bone, carnivores, Aves, amphibia ind. and Anura for which the scar­city of available data hinders further discussion. One must emphasize that sieving was implemented in Structure 3, corresponding to 97% of the total as­semblage under study. Finally, some bivalves were recovered; the grooved carpet shell is more frequent, but the common cockle and scallop are present. The assemblage has several indicators of anthropo­genic processing of animal body parts, with cutmarks related to the different stages of butchering and an­thropogenic breakage indicators recognizable. Burn­ed and possible boiled remains are present with si­milar frequencies and can potentially relate to the higher longitudinal outlines and right or mixed an­gles of the breakage planes in the small to large weight groups. The frequency of double colourations and the larger number of degrees 2 and 3 might re­sult from the culinary practices, more than another type of natural or non-culinary anthropogenic action. Overall, the fragmentation as well as modern break­age that affected the collection is also evident. Regarding leporids, they have higher preservation of diaphysis in comparison to other species. None­theless, complete long bones are absent – only iso­lated teeth and bones from the extremities are com­plete, and modern breakage is infrequent (n=9). At the same time, consumption indicators on leporid bones and burning are present. The former compri­ses notches and punctures in appendicular bones (n=16), and crenulated edges (n=12) in innominate (n=8), scapula (n=1), femur (n=2) and mandible (n=1). Morphologically, the majority of these re­mains seem to lack the typical characteristics of car­nivore gnawing tooth marks, showing similarities with human consumption. One must consider the possibility that other carnivores, such as canids, could have had access to the remains. Wolfs, for example, are known to completely consume rabbit remains, thus this would result in a large, digested sample (Lloveras et al. 2020) but this is not found in the assemblage, and the implementation of siev­ing certainly diminished recovery bias. The impor­tance of dogs in Late Prehistory contexts is note­worthy, since they probably ate human-abandoned waste, resulting in a different patterning of body parts, besides distinguishable tooth marks and di­gestion degrees, but data is lacking for better com­parison (Almeida et al. 2022). Other carnivores, as is the case of foxes (Sanchis 2000; Lloveras et al. 2012) and the Iberian-lynx (Lloveras et al. 2008a; Rodríguez-Hidalgo et al. 2013; 2015), have slightly different patterns of consumption, but leave tooth marks that can be distinguishable from human chew­ing. Beak/talon marks created by birds of prey (e.g., Sanchis 2000; Lloveras et al. 2008b; 2009) are ab­sent in the collection. Measurable tooth marks on leporid bones were scarce, because pits are less pre­valent than crenulated edges and tooth notches. Moreover, a lower breakage of leporid bones com­pared to that seen with larger animals was ob­served. The lack of young leporids could relate to the natural dead’s latu senso, or the lack of anato­mical connections that could hint at predation, toge­ther with the abovementioned data, lead us to sug­gest a largely anthropogenic origin for the leporids in the assemblage. Hence, carnivores seem to not have contributed substantially to the faunal accumu­lation of the Ota site for smaller specimens, but we cannot entirely discard secondary access to them, for example, by dogs. The secondary access and mo­dification of remains are well characterized regard­ing larger taxa. In these cases, it is not clear if all consumption damage should be related to carni­vore action, because some smaller and more shallow tooth marks could be produced during human chew­ ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance ing. While swine do not seem to be of relevance in the consumption due to a lack of diagnostic tooth marks, the morphologies, types and size of tooth marks are consistent with a large canid, also identi­fied in the assemblage. The presence of molluscs in Iberian Late Prehistory contexts is common and associated with merely eco­nomic or an ideological and symbolical dimension of their consumption or the use of shells. Regarding scallops, for example, the circulation of their shells is recurrent even in inland sites, while others, such as the grooved carpet shell, could be consumed or circulated in conditions that would allow their con­sumption in farther regions (Valera, André 2016/ 2017). The few bivalves identified do not allow for further considerations, since they were scattered in several stratigraphical units and not related to spe­cific contexts that could further reveal their purpose. On the one hand, they could be consumed and thus represent a small portion of these groups’ diet due to their bromatological value, on the other hand, they could be part of the existing interaction and cir­culation networks where shells are commonly inter­preted (e.g., Coelho 2006; 2008; 2013; Soares 2013; Delicado et al. 2017; Valera, André 2016/2017; Al-meida et al. 2020a; Almeida, Valera 2021). The data from a wider perspective An important amount of Chalcolithic faunal records is published for central and southern Portugal (Fig. 10). Focusing on Estremadura, it is clear that equids are almost absent from the assemblages and, when recovered, their NISP is <1%. Clear hunting strate­gies are represented by the presence of cervids, mainly red deer and the occasional roe deer, but be­sides Columbeira (14%) and Ota (9%), these species have values of <5% in Chibanes, Zambujal, and es­pecially Leceia and Penedo do Lexim. Swine domi­nate the assemblages of Ota (39%) and Chibanes (48%), showing frequencies similar to caprine in Co-lumbeira and Zambujal. Caprine are even more fre­quent than swine in Penedo do Lexim and Leceia. In all cases, bovines are proportionally less common than swine and caprine, although the values ob­tained in the Ota site (24%) are similar to Zambujal (25%) and a bit less to Leceia (20%). The small fre­quency of equids and cervids is maintained during the end of the Chalcolithic, even if Vila Nova da Sao Pedro shows 8% of cervids. This site is the only one where bovines have important values (30%), but are still lower than swine, which are prevalent in all these assemblages, even reaching 58% in the Chiba-nes IC collection. Further results are expected to be published for VNSP that could allow for a better understanding of this pattern and if it is fully repre­sentative of the site. It is not clear how observed differences in propor­tions between the larger faunal groups for these sites relate to environmental and archaeological fac­tors and how this change regarding chronology and context. The lack of contextual and fine-grain chro­nological information, especially for older studied and published collections, hinders the proper discus­sion and assessment of possible patterns that could help explain these differences. These results have some disparities from the ones obtained for the Alentejo region with its peneplains that differ from the Estremadura region. The diffe­rence between these two regions is currently seen not only in the archaeological records themselves but also in the faunal spectra that have been recov­ered (Valente, Carvalho 2014; Almeida, Valera 2021). While equids are slightly more abundant than in the Estremadura, cervids have much higher val­ues, reaching 3–12% during 3000–2400 BCE, with ~ one case of 36% in the S. Pedro site, in Redondo. Later, between 2400–2000 BCE, the two ditched ~ enclosures of Porto Torrao and Perdigoes show that cervids were an important part of these econo­mies with 23% and 29%, respectively. In the Alen­tejo, swine are the most frequent group in all sites with exception of the Perdigoes collection dated to 3000–2400 BCE from where caprine reaches 39%. ~ Caprine has values similar to bovine in S. Pedro, but the latter is less frequent in Monte da Tumba, Perdigoes, Mercador and the older Torrao study by Driesch in Arnaud (1993). During 2400–2200 BCE, ~ the rise in cervids is accompanied by a similarity in abundance between caprine and bovine, always be­tween 14–19% to both groups, with a prevalence of swine. The lower dependency on red deer hunting in the Estremadura compared to the Alentejo might relate to availability and the type of sites. The Estremadu­ra records of larger settlements, more dependent on domesticated animals, could result in a higher an-thropization of the surrounding areas leading to less availability of the large wild game. While the archaeological data (type of sites, material culture) can point in this direction, the scarce archaeobota­nical data does not allow for further discussion of this hypothesis, a human impact on the landscape was already observable during the Chalcolithic (and previously) in several regions (e.g., Lord et al. 2011). Nelson J. Almeida, André Texugo, and Ana Catarina Basílio Equids and mainly cervids are scarce in the Estrema­dura regional records with higher frequencies in the Alentejo, thus similar to other roughly contempora­neous records from Spanish Extremadura, such as Atalaya (Rodríguez-Hidalgo, Cabezas 2011), Horna­chuelos (Castanos 1998), Cerro de la Horca (Casta­nos 1992), Cerro I (Castanos 1991; 1997) and Cer­ro II de Los Castillejos (Castanos 1991). Conclusions With its 170m AMSL and an occupational chronolo­gy ranging from the Neolithic to the Islamic period, Ota is exposed to various climatic actions that re­sult in a generalized lack of sedimentation (Texugo et al. in preparation). This is a reality shared by its congeners in Estremadura, resulting in a cultural phenomenon – the walled enclosure– that is mainly based on palimpsests, very difficult to disentangle and securely study. These problems were taken into consideration during the excavation campaigns that occurred from 2019 up to 2021 at Ota (Texugo et al. in preparation). To reduce their impact and help in the excavation and interpretative process of the iden­tified archaeological units, the taphonomical study of the faunal assemblage of this site was promptly ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance carried out. It allowed us to characterize the con-cessing, with cutmarks derived from butchering, an-sumption patterns of these communities, which in-thropogenic breakage, and thermal alteration with cluded both hunting and local production, and the culinary proposes. The kill-off patterns are sugges-networks to which they might have been connected, tive of secondary product exploitation in caprine due to the presence of exogenous shells. However, and bovine, possibly milk, to produce cheese, mate-it also allowed us to empirically corroborate the exi-rialized in the presence of cheese makers in the ar­stence of paedogenic processes, to which the anthro-chaeological record. The analysis of taphonomical pic influence, through the intense architectonic re-indicators of the sedimentary environment rein-formulations, is added. forced the interpretation based on fieldwork obser­ vations concerning stratigraphies, further support-Nonetheless, and summing up, it seems that Ota re-ed by new absolute dates. This allowed for the pre­plicates the patterns found in neighbouring regio-sentation and discussion of zooarchaeological infor­nal sites, with a fauna profile highlighting the impor-mation from a current perspective, one still uncom­tance of domestic species, namely swine, goat/sheep mon for this region and period. and cattle. Hunting practices are complementary, as is suggested by the less common red deer, auroch and possibly wild boar. Of added interest are the le- ACKNOWLEDGEMENTS porids, mostly composed of rabbits, that after a ta­phonomical assessment seems to relate mainly to The authors are supported by the Foundation for Sci­ence and Technology (FCT) in the framework of the human consumption, even if occasional indicators of project UIDB/00698/2020 (NJA) and through PhD other predators are recorded. As noted earlier, we Grants SFRH/BD/135648/2018 and SFRH/BD136086/ do not discard access to these remains by carnivo­ 2018 (AT and ACB). The authors are thankful to the res, which is also demonstrated in the larger animal editor and the anonymous reviewer for their sugges- remains. However, this would mainly have been se­ tions and comments that helped to improve the ma- condary access to human food refuse. The assem­ nuscript. blage has several indicators of anthropogenic pro­ . References Albarella U., Davis S., Detry C., and Rowley-Conwy P. 2005. Pigs of the “Far West”: the biometry of Sus from archaeological sites in Portugal. Anthropozoologica 40: 27–54. Aleixo P. 2018. Estudo zooarqueológico do sítio do Neo­lítico Final do Barranco do Xacafre, Ferreira do Alen­tejo. MA thesis. University of Algarve. Faro. Almeida N. J. 2017. Zooarqueologia e tafonomia da transiçao para a agro-pastorícia no baixo e médio vale do Tejo. Instituto Terra e Memória. Maçao. Almeida N. J., Basílio A. C., and Valera A. C. 2020a. The faunal record from Santa Vitória (Campo Maior): an ini­tial appraisal based on the remains from 2018 and 2019 excavations. Apontamentos de Arqueologia e Patrimó­nio 14: 9–16. Almeida N. J., Dias I., Detry C., Porfírio E., and Serra M. 2020b. As faunas do final da Idade do Bronze no Sul de Portugal: leituras desde o Outeiro do Circo (Beja). In J. M. Arnaud, C. Neves, and A. Martins (eds.), Arqueologia em Portugal 2020 – Estado da Questao. Associaçao dos Ar-queólogos Portugueses e CITCEM. Lisboa: 1041–1054. Almeida N. J., Valera A. C. 2021. Animal consumption and social change: the vertebrates from Ditch 7 in the context of a diachronic approach to the faunal remains at Perdi­goes enclosure (3400–2000 BC). Archaeofauna 30: 75– 106. https://doi.org/10.15366/archaeofauna2021.30.005 Almeida N. J., Basílio A. C., Silva C., Soares A. M., and Bor­ges A. M. 2021a. Faunal remains manipulation during the Chalcolithic in pits 13, 16 and 54 from Monte das Cabe­ceiras 2 (Beja, Southern Portugal). Zephyrus LXXXVIII: 41–64. https://doi.org/10.14201/zephyrus2021884164 Almeida N. J., Cerrillo-Cuenca E., and Saladié P. 2021b. Framing agricultural intensification in western Iberia du­ring the Late Neolithic: A new insight through the faunal record from Los Barruecos site. Journal of Archaeologi­cal Science: Reports 36: 102815. https://doi.org/10.1016/j.jasrep.2021.102815 Nelson J. Almeida, André Texugo, and Ana Catarina Basílio Almeida N. J., Saladié P., and Cerrillo-Cuenca E. 2022. Rab­bits beyond hunter-gatherers’ diets in Western Europe? The case for leporid accumulations in Neolithic Southwest­ern Iberia. Archaeological and Anthropological Sciences 14: 186. https://doi.org/10.1007/s12520-022-01662-8 Andrés M., Gidna A. O., Yravedra J., Domínguez-Rodrigo M. 2012. A study of dimensional differences of tooth marks (pits and scores) on bones modified by small and large car­nivores. Archaeological and Anthropological Sciences 4: 209–219. https://doi.org/10.1007/s12520-012-0093-4 Antunes M. T. 1987. O povoado fortificado calcolítico do Monte da Tumba. IV – Mamíferos (nota preliminar). Setú­bal Arqueológica VIII: 103–144. Arnaud J. M. 1993. O povoado calcolítico de Porto Torrao (Ferreira do Alentejo): síntese das investigaçoes realiza­das. Vipasca 2: 41–60. Basílio A. C., Texugo A. 2017. O conjunto de Pedra Las-cada da Ota: Questoes tecnológicas e socioeconómicas. In J. Arnaud, A. Martins (eds.), Arqueologia em Portugal 2017 – Estado da Questao. Associaçao dos Arqueólogos Portugueses. Lisboa: 619–630. Basílio A. C. in press. Wearing idol eyes: Bell Beakers and the cultural role of Montejunto Hill (Portugal) during the second half of the 3rd millennium BC. Ophiussa. Behrensmeyer A. K. 1978. Taphonomic and ecologic infor­mation from bone weathering. Paleobiology 4(2): 150–162. Behrensmeyer A. K., Gordon K. D., and Yanagi G. T. 1986. Trampling as a cause of bone surface damage and pseudo-cutmarks. Nature 319 (27): 768–771. https://doi.org/10.1038/319768a0 Binford L. R. 1978. Nunamiut Ethnoarchaeology. New York Academic Press. New York. 1981. Bones. Ancient men and modern myths. New York Academic Press. New York. Boessneck J., Müller H.-H., and Teichert M. 1964. Osteolo­gische unterscheidungmerkmale zwischen schaf (Ovis aries Linné) und zeige (Capra hircus Linné). Kühn-Archiv 78(1–2): 1–129. Boessneck J. 1970. Osteological differences between sheep (Ovis aries Linné) and goats (Capra hircus Linné). In D. Brothwell, E. Higgs (eds.), Science in Archaeology. Prae­ger. New York: 331–358. Brain C. K. 1981. The hunters or the hunted? An intro­duction to African cave taphonomy. Chicago University Press. Chicago. Bronk Ramsey C. 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51(1): 337–360. https://doi.org/10.1017/s0033822200033865 Bull G., Payne S. 1982. Tooth eruption and epiphyseal fu­sion in pigs and wild boar. In B. Wilson, C. Grigson, and S. Payne (eds.), Ageing and sexing animal bones from archaeological sites. BAR British Series 109. Archaeo-press. Oxford: 55–72. Bunn H. T. 1983. Comparative analysis of a modern bone assemblages from a San huntergatherer camp in the Ka­lahari Desert, Botswana, and from a spotted hyena den near Nairobi, Kenya. In J. Vlutton-Brock, C. Grigson (eds.), Animals and Archaeology, Hunters and their Prey. BAR International Series 163. Archaeopress. Oxford: 143– 148. Bunn H. T. 1986. Patterns of skeletal representation and hominid subsistence activities at Olduvai Gorge, Tanza­nia, and Koobi Fora, Kenya. Journal of Human Evolution 15: 673–690. Callou C. 1997. Diagnose différentielle des principaux elements squelettiques du lapin (genre Oryctolagus) et du liévre (genre Lepus) en Europe occidentale. Fiches d’ostéologie animale pour l’archéologie. Série B: Mammi­féres. N.o 8. Association pour la promotion et la diffusion des connaissances archéologiques. Centre de Recherche Archéologiques du CNRS. Valbonne. Cardoso J. L., Detry C. 2001/2002. Estudo arqueozooló­gico dos restos de ungulados do povoado pré-histórico de Leceia (Oeiras). Estudos Arqueológicos de Oeiras 10: 131–182. Cardoso J. L., Silva C. T., Soares J., and Martins F. 2021. A economia alimentar em Chibanes (Palmela) – Horizonte Campaniforme. Ophiussa 5: 103–129. https://doi.org/10.51679/ophiussa.2021.84 Carvalho A. F., Gonçalves D., Díaz-Zorita M., and Valente M. J. 2019. Multi-isotope approaches to the Neolithic ce-metery-cave of Bom Santo (Lisbon): new data and com­parisons with fourth millennium BC populations from central-southern Portugal. Archaeological and Anthropo­logical Sciences. 11: 6141–6159. https://doi.org/10.1007/s12520-019-00908-2 Castanos P. M. 1991. Animales domésticos y salvages em Extremadura. Origen y evolución. Revista de Estudios Ex-tremenos 47(1): 9–66. 1992. Estudio arqueozoologico de la fauna del Cerro de la Horca (Plasenzuela, Cáceres). Archaeofauna 1: 127– 146. ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance 1997. Estudio de la fauna del Cerro I de “Los Castille­jos” (Fuente de Cantos, Badajoz). Revista de Historia 14: 11–45. 1998. Evolució de las faunas protohistóricas em Extre­madura. Extremadura Protohistórica: Paleoambiente, Economía y Poblamiento. Cáceres: 63–72. Coelho M. 2006. A fauna malacológica de Porto Torrao. Os moluscos no Neolítico Final/Calcolítico do Sul de Portugal. MA thesis. University of Lisbon. Lisboa 2008. A fauna malacológica proveniente do Sector I do recinto calcolítico dos Perdigoes. Apontamentos de Ar-queologia e Património 3: 35–40. 2013. Faunas malacológicas do povoado do Mercador. In A. C. Valera (eds.), As comunidades agropastoris na margem esquerda do Guadiana. Memórias d’Odiana 2a Série. Estudos arqueológicos do Alqueva 6. Empresa de desenvolvimentoe infra-estruturas do Alqueva. Direc­çao Regional de Cultura do Alentejo. Alqueva: 365–376. Correia F. R. 2015. O Castro da Columbeira (Bombar­ral): a exploraçao dos recursos faunísticos no Calcolí­tico Estremenho. MA thesis. University of Algarve. Faro. Costa C. 2013. Tafonomia em contexto pré-histórico. A zooarqueologia como recurso para a compreensao das “estruturas em negativo” da Pré-história Recente. Un­published PhD thesis. University of Algarve. Faro. Courty M., Goldberg P., and Macphail R. 1989. Soils and micromorphology in archaeology. Cambridge University Press. Cambridge. Cruz Berrocal M., Sanjuán L. G., and Gilman A. 2013. The Prehistory of Iberia: Debating Early Social Stratification and the State. Routledge. New York. Davis S. 1996. Measurements of a group of adult female Shetland sheep skeletons from a single flock: a baseline for zooarchaeologists. Journal of Archaeological Science 23: 593–612. https://doi.org/10.1006/jasc.1996.0056 2006. Faunal remains from Alcáçova de Santarém, Por­tugal. Trabalhos de Arqueologia 43: 11–144. 2017a. Mammal and bird remains from the Iron Age and Roman periods at Castro Marim, Algarve. Trabal-hos do Centro de Investigaçao em Paleoecologia Huma­na e Arqueociencias 107. Instituto Portugues de Arqueo­logia. Lisbon. 2017b Towards a metrical distinction between sheep and goat astragali. In P. Rowley-Conwy, D. Serjeantson, and P. Halstead (eds.), Economic Zooarchaeology. Stu­dies in hunting, herding and early agriculture. Oxbow Books. Oxford: 50–82. Davis S., Mataloto R. 2012. Animal remains from Chalco­lithic Sao Pedro (Redondo, Alentejo): evidence for a crisis in the Mesolithic. Revista Portuguesa de Arqueologia 15: 47–85. Davis S., Gabriel S., and Simoes T. 2018. Animal remains from Neolithic Lameiras, Sintra: the earliest domesticated sheep, goat, cattle and pigs in Portugal and some notes on their evolution. Archaeofauna 27: 93–172. https://doi.org/10.15366/archaeofauna2018.27.006 Delaney-Rivera C., Plummer T. W., Hodgson J. A., Forrest F., Hertel F., and Oliver J. S. 2009. Pits and pitfalls: taxo­nomic variability and patterning in tooth mark dimensi­ons. Journal of Archaeological Science 36: 2597–2608. https://doi.org/10.1016/j.jas.2009.08.001 Delicado C. S., Santos A. B., Porfírio E., Serra M., and Detry C. 2017. Alto de Brinches 3 (Serpa): estuda da fauna re-cuperada nos contextos do 3o e 2o milénio a.n.e. Cader-nos do GEEvH. Notes in Human Evolution. 6(1): 28–55. Detry C. 2007. Paleoecologia e Paleoeconomia do Baixo Tejo no Mesolítico Final: o contributo do estudo dos ma-míferos dos concheiros de Muge. Unpublished PhD the­sis. University of Salamanca. Autonomous University of Lisbon. Lisboa. Detry C., Francisco A. C., Diniz M., Martins A., Neves C., and Arnaud J. M. 2020. Estudo zooarqueológico das fau­nas do Calcolítico final de Vila Nova de Sao Pedro (Azam­buja, Portugal): Campanhas de 2017 e 2018. In J. M. Ar-naud, C. Neves and A. Martins (eds.), Arqueologia em Portugal/2020 – Estado da Questao. Associaçao dos Ar-queólogos Portugueses. Lisboa: 943–958. Domínguez-Rodrigo M., Piqueras A. 2003. The use of tooth pits to identify carnivore taxa in tooth-marked archaeo-faunas and their relevance to reconstruct hominid carcass processing behaviours. Journal of Archaeological Sci­ence 30: 1385–1391. https://doi.org/10.1016/S0305-4403(03)00027-X Driesch A. 1976. A Guide to the Measurement of Animal Bones from Archaeological Sites (as developed by the Institut für Palaeoanatomie, Domestikationsforschung und Geshchichte der Tiermedizin of the University of Munich). Peabody Museum Bulletin 1. Peabody Museum of Archaeology and Ethnology. Harvard University. Cam­bridge, MA. Driesch A., Boessneck J. 1976. Die Fauna vom Castro do Zambujal (Fundmaterial der Grabungen von 1966 bis 1973 mit Ausnahme der Zwingerfunde). In A. Driesch, J. Nelson J. Almeida, André Texugo, and Ana Catarina Basílio Boessneck (eds.), Studien über frühe Tierknochenfunde von der Iberischen Halbinsel 5. Institut für Palaeoanato­mie, Domestikationsforschung und Geschichte der Tier-medizin der Universität München. Deutsches Archäologi­sches Institut Abteilung Madrid. München: 4–129. Encarnaçao G., Almeida N. J. 2017. O Povoado da Espar-gueira/Serra das Éguas. Trabalhos arqueológicos reali­zados entre 2003 e 2008. Relatórios – 10. Associaçao de Arqueologia da Amadora. Câmara Municipal de Amadora. Amadora. Fernández-Jalvo Y., Sánchez-Chillón B., Andrews P., Fer-nández-López S., and Alcalá Martínez L. 2002. Morpholo­gical taphonomic transformations of fossil bones in con­tinental environments, and repercussions on their chemi­cal composition. Archaeometry 44(3): 353–361. https://doi.org/10.1111/1475-4754.t01-1-00068 Gonçalves V. S. 1989. Megalitismo e metalurgia no Alto Algarve Oriental. Estudos e memórias 2. Instituto Nacio­nal de Investigaçao Científica. Unidade Centro de Arqueo­logia do Centro de História. Lisboa. Gonçalves J. L. 1992. As grutas da Serra de Montejunto (Cadaval). O Arqueólogo Portugues 4: 41–201. Grant A. 1982. The use of tooth wear as guide to the age of domestic ungulates. In B. Wilson, C. Grigson, and S. Payne (eds.), Ageing and sexing animal bones from archaeological sites. BAR British Series 109. Archaeo-press. Oxford: 91–108. Grayson D. 1984. Quantitative Zooarchaeology. London Academic Press. London. Jones E. L. 2006. Prey choice, mass collecting, and the wild European rabbit (Orcyctolagus cuniculus). Journal of Anthropological Archaeology 25: 275–289. https://doi.org/10.1016/j.jaa.2005.11.002 Jorge S. O. 2003. Pensar o espaço da Pré-História recen­te: a propósito dos recintos murados da Península Ibéri-ca. O Passado é Redondo. In Dialogando com os Senti-dos dos Primeiros Recintos Monumentais. Biblioteca de arqueologia 2. Ediçoes Afrontamento. Porto: 169–202. Jorge V. O., Cardoso J. M., Vale A. M., Velho L. G., and Pereira L. S. 2006. Copper age “monumentalized hills” of Iberia: the shift from positivistic ideas to interpretive ones. New perspectives on old techniques of transform­ing place and space as results of a research experience in the NE of Portugal. In V. O. Jorge (ed.), Approaching “pre­historic and protohistoric architectures” of Europe from a “dwelling perspective”: proceeding of the TAG session. Association for the Improvement of Cooperation in Ibe­rian Archeology (ADECAP). Porto: 203–264. Kottek M., Grieser J., Beck C., Rudolf B., and Rubel F. 2006. World Map of the Koppen-Geiger climate classifica­tion updated. Meteorologische Zeitschrift. 15: 259–263. DOI: 10.1127/0941-2948/2006/0130 Kunst M. 2010. Zambujal, a dinâmica da sequencia constru­tiva. In V. S. Gonçalves, A. C. Sousa (eds.), Transformaçao e mudança no Centro e Sul de Portugal: o 4.° e o 3.° mi-lénios a.n.e. Câmara Municipal de Cascais. Centro de Ar-queologia da Universidade de Lisboa (UNIARQ). Cascais: 131–154. Lemoine X., Zeder M. A., Bishop K. J., and Rufolo S. J. 2014. A new system for computing dentition-based age profile in Sus scrofa. Journal of Achaeological Science 47: 179–193. https://doi.org/10.1016/j.jas.2014.04.002 Lloveras L., Moreno-García M., and Nadal J. 2008a. Tapho­nomic analysis of leporid remains obtained from modern Iberian lynx (Lynx pardinus) scats. Journal of Archaeo­logical Sciences 35: 1–13. https://doi.org/10.1016/j.jas. 2007.02.005 2008b. Taphonomic study of leporid remains accumu­lated by the Spanish Imperial Eagle (Aquila adalberti). Geobios 41: 91–100. https://doi.org/10.1016/j. geobios.2006.11.009 2009. The eagle owl (Bubo bubo) as a leporid remains accumulator: taphonomic analysis of modern rabbit re­mains recovered from nests of this predator. Interna­tional Journal of Osteoarchaeology 19: 573–592. https://doi.org/10.1002/oa.995 2012. Feeding the foxes: an experimental study to as­sess their taphonomic signature on leporid remains. In­ternational Journal of Osteoarchaeology 22: 577–590. https://doi.org/10.1002/oa.1280 Lloveras L., Nadal J., and Fullola J. M. 2020. Distinguishing the taphonomic signature of wolves from human and other predators on small prey assemblages. Scientific Reports 10: 8030. https://doi.org/10.1038/s41598-020-64716-8 López-González F., Grandal-d’Anglade A., and Vidal-Roma­ní J. R. 2006. Deciphering bone depositional sequences in caves through the study of manganese coatings. Jour­nal of Archaeological Science 33: 707–717. https://doi.org/10.1016/j.jas.2005.10.006 Lord A., Cabral M. C., Dambeck R., and Kunst M. 2011. Ostracod evidence for the Neolithic environment of Rio Sizandro, Portugal. Palaeobiodiversity and Palaeoenvi­ronments 91: 215–228. https://doi.org/10.1007/s12549-011-0055-3 Lyman R. L. 1994. Vertebrate Taphonomy. Cambridge University Press. Cambridge. ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance 2008. Quantitative Paleozoology. Cambridge Univer­sity Press. Cambridge. Mataloto R. 2010. O 3.°/4.° milénio a.C. no povoado de Sao Pedro (Redondo, Alentejo Central): fortificaçao e po­voamento na planície centro alentejana. In V. S. Gonçal­ves, A. C. Sousa (eds), Transformaçao e mudança no Centro e Sul de Portugal: o 4.o e o 3.omilénios a.n.e. Câ­mara Municipal de Cascais. Centro de Arqueologia da Uni-versidade de Lisboa (UNIARQ). Cascais: 263–296. Molina F., Camara J. A. 2005. Los Millares: Guia del yaci­miento arqueológico. Junta de Andalucia and Consejeria de Cultura. Seville. Mora C., Vieira G. 2020. The Climate of Portugal. In G. Vieira, J. Zezere, and C. Mora (eds.), Landscapes and Landforms of Portugal. World Geomorphological Land­scapes. Springer. Cham: 33–46. https://doi.org/10.1007/978-3-319-03641-0_2 Moreno-García M. 2013. Estudo arqueozoológico dos res-tos faunísticos do povoado calcolítico do Mercador (Mou­rao). In A. C. Valera (ed.), As sociedades agropastoris na margem esquerda do Guadiana (2a metade do IV aos inícios do II milénio AC). Memórias d’Odiana 2a série. Estudos arqueológicos do Alqueva 6. Empresa de Desen­volvimento e Infra-Estrururas do Alqueva (EDIA). Direc­çao Regional de Cultura do Alentejo (DRCALEN). Alque­va: 319–349. Moreno-García M., Sousa A. C. 2015. Para além das mural-has, uma perspetiva dos recursos faunísticos no Calcolí­tico da Estremadura: o conjunto arqueofaunístico do Lo­cus 5 do Penedo do Lexim (Mafra). Revista Portuguesa de Arqueologia 18: 101–124. Payne S. 1973. Kill-off pattern in sheep and goats: the mandibles of Asvan kale. Anatolian Studies 23: 281–303. 1987. Reference codes for wear states in the mandibu­lar cheek tooth of sheep and goats. Journal of Archaeo­logical Science 14: 609–614. https://doi.org/10.1016/0305-4403(87)90079-3 Payne S., Bull G. 1988. Components of variation in mea­surements of pig bones and teeth, and the use of measu­rements to distinguish wild from domestic pig remains. ArchaeoZoologia II(1–2): 27–66. Pereira V. 2016. Repastos alentejanos: dados prelimina-res da fauna de Porto Torrao (Ferreira do Alentejo). In R. Vilaça, M. Serra (eds.), Matar a fome, alimentar a alma, criar sociabilidades. Alimentaçao e comensalidade nas sociedades pré e proto-históricas. Instituto de Arqueolo­gia. Centro de Estudos Pré-Históricos da Beira Alta. Pa-limpsesto, Estudo e Preservaçao do Património Cultural Lda. Coimbra: 39–53. http://estudospre-historicos.weebly. com/matar-a-fome-alimentar-a-alma-criar-sociabilidades­ 2016.html Pereira V., Soares J., and Silva C. 2017. Understanding the First Chalcolithic Communities of Estremadura: Zooarcha­eology of Castro de Chibanes, Portugal. Preliminary Re­sults. Papers from the Institute of Archaeology 27(1): 1–11. Pereiro T., Almeida N., and Valera A. C. 2021. O recinto de fossos calcolítico da Herdade do Álamo (Sao Brissos, Beja). Apontamentos de Arqueologia e Património 15: 28–35. Pérez Ripoll M. 1999. La explotación ganadera durante el III milenio a.C. en la Peninsula Ibérica. In II Congrés del Neolítico a la Península Ibérica. SAGVNTVM-PLAV, Extra 2. Valencia: 95–103. Pickering T. R., Domínguez-Rodrigo M., Heaton J. L., +7 authors, and Diez-Martín F., Mabulla A., Brain C. K. 2013. Taphonomy of ungulate ribs and the consumption of meat and bone by 1.2-million-year-old hominins at Olduvai Gorge, Tanzania. Journal of Archaeological Science 40: 1295–1309. https://doi.orChamg/10.1016/j.jas.2012.09.025 Ramos-Pereira A., Ramos C., Vieira J., Zezere C., and Mo­ra C. 2020. The Southwest Coast of Portugal. In G. Vieira, J. Zezere, and C. Mora (eds.), Landscapes and Landforms of Portugal. World Geomorphological Landscapes. Sprin­ger. Cham: 109–115. https://doi.org/10.1007/978-3-319-03641-0_8 Reimer P. J., Austin W. E. N., Bard E., +33 authors, and Talamo S. 2020. The IntCal20 Northern Hemisphere Ra­diocarbon Age Calibration Curve (0–55 cal kbp). Radio­carbon 62(4): 725–757. https://doi.org/10.1017/RDC.2020.41 Reitz E. J., Wing E. S. 2008. Zooarchaeology. Cambridge University Press. Cambridge. Rodríguez-Hidalgo A. J., Cabezas M. G. 2011. Datos pre­liminaries sobre la fauna del yacimiento calcolitico de la Atalaya (Torrequemada, Cáceres). In P. Bueno Ranirez, E. Cerrillo Cuenca, and A. Gonzalez Cordero (eds.), From the origins: the Prehistory of the inner Tagus region. BAR International Series 2219. Archaeopress. Oxford: 233–240. Rodríguez-Hidalgo A. J., Lloveras L., Moreno-García M., Saladié P., Canals A., and Nadal J. 2013. Feeding behavi-our and taphonomic characterization of non-ingested rab­bit remains produced by the Iberian lynx (Lynx pardi-nus). Journal of Archaeological Science 40: 3031–3045. https://doi.org/10.1016/j.jas.2013.03.006 Nelson J. Almeida, André Texugo, and Ana Catarina Basílio Rodríguez-Hidalgo A., Saladié P., Marín J., and Canals A. 2015. Expansion of the referential framework for the rab­bit fossil accumulations generated by Iberian lynx. Palaeo-geography Palaeoclimatology Palaeoecology 418: 1–11. https://doi.org/10.1016/j.palaeo.2014.11.010 Saladié P. 2009. Mossegades d’omnívors. Aproximació experimental i aplicació zooarqueologica a la Sierra de Atapuerca. Unpublished PhD thesis. University Rovira I Virgili. Tarragona. Saladié P., Huguet R., Díez C., +5 authors, and Carbonell E. 2011. Carcass transport decision in Homo antecessor subsistence strategies. Journal of Human Evolution 61: 425–446. Saladié P., Rodríguez-Hidalgo A., Díez C., Martín-Rodríguez P., and Carbonell E. 2013. Range of bone modifications by human chewing. Journal of Archaeological Science 40: 380–397. http://dx.doi.org/10.1016/j.jas.2012.08.002 Salvagno L., Albarella U. 2017. A morphometric system to distinguish sheep and goat postcranial bones. PLoS ONE 12(6): e0178543. https://doi.org/10.1371/journal.pone.0178543 Sanchis A. 2000. Los restos de Oryctolagus cuniculus em las tafocenosis de Bubo bubo y Vulpes vulpes Vulpes vul­pes y su aplicación a la caracterización del registro fauní­stico arqueológico. Sagvntvm 32: 31–50. Selvaggio M. M., Wilder J. 2001. Identifying the involve­ment of multiple carnivore taxa with archaeological bone assemblages. Journal of Archaeological Science 28: 465– 470. https://doi.org/10.1006/jasc.2000.0557 Shipman P. 1981. Applications of Scanning Electron Mi­croscopy to taphonomic problems. Annals of the New York Academy of Sciences 376(1): 357–385. Shipman P., Rose J. J. 1984. Cutmark mimics on modern and fossil bovid bones. Current Anthropology 25(1): 116– 117. https://www.journals.uchicago.edu/doi/abs/10.1086/ 203091 Soares J. 2013. Transformaçoes sociais durante o 3.o milénio AC no sul de Portugal. O povoado do Porto das Carretas. Empresa de Desenvolvimento e Infra-Estrururas do Alqueva (EDIA). Direcçao Regional de Cultura do Alen­tejo (DRCALEN). Lisboa. Solari A., Olivera D., Gordillo I., Bosch P., Fetter G., Lara V. H., and Novelo O. 2015. Cooked bones? Method and practice for identifying bones treated at low temperature. International Journal of Osteoarchaeology 25: 426–440. https://doi.org/10.1002/oa.2311 Stiner M. C., Kuhn S. L., Weiner S., and Bar-Yosef O. 1995. Differential burning, recrystallization, and fragmentation of archaeological bone. Journal of Archaeological Sci­ence 22: 223–237. Texugo A. 2022. A morte a espreita: a possível estrutura funerária calcolítica da Ota (Alenquer). In V. S. Gonçalves (ed.), Terra e Sal: Das antigas sociedades camponesas ao fim dos tempos modernos. Estudos oferecidos a Car­los Tavares da Silva. Centro de Arqueologia da Universi­dade de Lisboa (UNIARQ). Faculdade de letras da Univer­sidade de Lisboa. Lisboa: 207–218. Texugo A., Basílio A. C., Pina P., Goyanes G., and Vieira G. in preparation. A multi-embankment Chalcolithic walled enclosure: new insights from the usage of remote sensing in archaeological surveys (Ota, Western Portugal). Valente M. J. 2008. As últimas sociedades de caçadores­recolectores no Centro e Sul de Portugal (10.000 – 6.000 anos BP): aproveitamento dos recursos animais. Un­published PhD thesis. University of Algarve. Faro. 2013. Arqueozoologia da Barca do Xerez de Baixo. In A. C. Araújo, F. Almeida (eds.), Barca do Xerez de Baixo – Um testemunho invulgar das últimas comunidades de caçadores-recolectores do Alentejo interior. Memó­ria d’Odiana 2a Série. Estudos Arqueológicos do Al-queva. Alqueva: 243–304. Valente M. J., Carvalho A. F. 2014. Zooarchaeology in the Neolithic and Chalcolithic of Southern Portugal. Environ­mental Archaeology 19(3): 226–240. https://doi.org/ 10.1179/1749631414Y.0000000022 Valera A. C. 2015. Social change in the late 3rd millenni­um BC in Portugal: the twilight of enclosures. In H. Mel-ler, R. Risch, R. Jung, and H. Arz (eds.), 2200 BC – Ein Klimasturz als Ursache für den Zerfall der Alten Welt? 2200 BC – A climatic breakdown as a cause for the col­lapse of the old world. 7th Archaeological Conference of Central Germany October 23–26, 2013 in Halle (Saale). Tagungen des Landesmuseums für Vorgeschichte Halle Band 12/I. Landesamt für Denkmalpflege und Archäolo­gie Sachsen-Anhalt. Landesmuseum für Vorgeschichte. Hal-le (Saale): 409–427. Valera A. C., André L. 2016/2017. Aspectos da interacçao transregional na Pré-história recente do Sudoeste Penin­sular: interrogando as conchas e moluscos nos Perdigoes. Estudos Arqueológicos de Oeiras 23: 189–218. Valera A. C., Basílio A. C., and Almeida N. 2020a. The Frag­ment, The Half, and The Whole: approaching pottery and fauna depositions in Pit 50 of Perdigoes enclosure (first half of the 3rd millennium BC). Estudos Do Quaternário/ ‘Animal farm’> the faunal record from the Chalcolithic Ota site (Alenquer, Portugal) and its regional significance Quaternary Studies 19: 11–33. https://doi.org/10.30893/eq.v0i20.189 Valera A. C., .alaite I., Maurer A. F., +4 authors, and Bar-rocas Dias C. 2020b. Addressing human mobility in Ibe­rian Neolithic and Chalcolithic ditched enclosures: The case of Perdigoes (South Portugal). Journal of Archaeo­logical Science: Rep. 30: 102264. https://doi.org/10.1016/j.jasrep.2020.102264 Vigne J. D., Helmer D. 2007. Was milk a “secondary pro­duct” in the Old World Neolithisation process? Its role in the domestication of cattle, sheep and goats. Anthropo­zoologica 42(2): 9–40. Villa P., Mahieu E. 1991. Breakage patterns of human long bones. Journal of Human Evolution 21: 27–48. Waterman A. J., Tykot R. H., and Silva A. M. 2015. Stable Isotope Analysis of Diet-based Social Differentiation at Late Prehistoric Collective Burials in South-Western Por­tugal. Archaeometry 58(1): 131–151. https://doi.org/10.1111/arcm.12159 White T. D. 1992. Prehistoric cannibalism at Mancos 5MTUMR-2346. Princeton University Press. Princeton. Wright E., Waterman A. J., Peate D. W., Kunst M., Cardo-so J. L., and Detry C. 2019. Animal mobility in Chalcoli­thic Portugal: isotopic analyses of cattle from the sites of Zambujal and Leceia. Journal of Archaeological Sci­ence: Reports 24: 804–814. back to contents https://doi.org/10.1016/j.jasrep.2019.02.005 .alaité I., Maurer A. F., Grimes V., +4 authors, and Vale­ra A. C. 2018. Diet and mobility of fauna from Late Neo­lithic-Chalcolithic site of Perdigóes, Portugal. Journal of Archaeological Science: Reports 19: 674–685. https://doi.org/10.1016/j.jasrep.2018.03.033 Zeder M. A. 2006. Reconciling rates of long bone fusion and tooth eruption and wear in sheep (Ovis) and goat (Capra). In D. Ruscillo (ed.), Recent advances in ageing and sexing animal bones. Proceedings of the 9th confe­rence of the International Council of Archaeozoology, Durham, August 2002. Oxbow Books. Oxford: 87–118. Zeder M. A., Lapham H. A. 2010. Assessing the reliability of criteria used to identify postcranial bones in sheep, Ovis, and goats, Capra. Journal of Archaeological Science 37(11): 2887–2905. https://doi.org/10.1016/j.jas.2010.06.032 Zeder M. A., Pilaar S. E. 2010. Assessing the reliability of criteria used to identify mandibles and mandibular teeth in sheep, Ovis, and goats, Capra. Journal of Archaeologi­cal Science 37: 225–242. https://doi.org/10.1016/j.jas.2009.10.002 Zeder M. A., Lemoine X., and Payne S. 2015. A new sys­tem for computing long-bone fusion age profiles in Sus scrofa. Journal of Archaeological Science 55: 135–150. http://dx.doi.org/10.1016/j.jas.2014.12.017 Documenta Praehistorica XLIX (2022) Colonization dynamics of LBK farmers in Europe under geostatistics test Robin Brigand1,2,3, Jérôme Dubouloz1,3, and Olivier Weller1,3 robinbrigand@yahoo.fr 1 CNRS-University Paris I Panthéon-Sorbonne, UMR 8215 “Trajectoires”, Paris, FR< jerome.dubouloz@univ-paris1.fr< olivier.weller@cnrs.fr 2 Lyon Archaeological Service, FR< 3 Cluster of Excellence “DynamiTe. Territorial and Spatial Dynamics”, University Paris I Panthéon-Sorbonne, Paris,FR ABSTRACT – Exploiting a database developed during a previous research project, this study uses factor analyses, GIS techniques and basic geostatistics to evaluate in detail the agro-ecological deter­minants of the first Neolithic diffusion in continental temperate Europe (the Linearbandkeramik or LBK), as well as its underlying settlement dynamics around half a millennium (5550–4925 BCE). More than 6600 LBK site locations, spread from Moldavia to Normandy, are initially assessed for their informative coherence and ability to offer a unified perspective on the evidence established at more local and regional levels. Most of these data can be used to define the broad geo-pedological options involved in the location of sites across Europe; loess substrate was far from being an exclu­sive settlement choice and a variety of soils, typically of medium moisture, were exploited. LBK far­mers thus had a great capacity to adapt to the different geographical contexts they encountered. With regard to settlement dynamics in Central and Western Europe, the data reveal a systemic interplay between creation, stability and abandonment of sites, supporting the diffusion of the LBK subsistence system. The progressive decline in the number of new sites was compensated by an increase in their stability until the last stage of the expansion process. At this point, abandonments became wide­spread without significant renewal, except in the westernmost regions. The easternmost parts of Europe could not be integrated in the large-scale temporal modelling, since the chronological data available in the database are insufficiently precise. Shedding new light on the systemic variability of the geo-environmental options followed by these early farmers and highlighting some modalities and spatial-temporal limits of the resilience of their agro-sylvo-pastoral system, our overall analysis confirms and somewhat clarifies current interpretations of the LBK phenomenon. KEY WORDS – Neolithic; LBK; Europe; colonization; geo-environment; settlement dynamics Geostatisti;ni test kolonizacijske dinamike LTK kmetov v Evropi IZVLE.EK – S pomo.jo podatkovne baze, vzpostavljene v prej.njem raziskovalnem projektu, v tej .tu­diji predstavljamo rezultate faktorskih analiz, tehnik GIS in osnovne geostatistike, s katerimi natan.­neje vrednotimo agro-ekolo.ke dejavnike med prvo neolitsko difuzijo v Evropo z zmernim celinskim podnebjem (podro.je kulture linearno trakaste keramike ali LTK) in poselitveno dinamiko v pol ti-so.letnem obdobju med 5550–4925 pr. n. .t. Ve. kot 6000 LTK lokacij, ki se raztezajo od Moldavije do Normandije, smo najprej ocenili glede na njihovo informativno koherentnost in sposobnost po­nuditi poenoten pogled na podatke, razvidne na lokalni in regionalni ravni. Ve.ino teh podatkov je mogo.e uporabiti za dolo.anje .irokega nabora geo-pedolo.kih danosti na lokacijah po vsej Evropi; puhlica .e zdale. ni bila izklju.na poselitvena izbira, raznolika, srednje vla.na tla so bila pogosta izbira. LTK kmetje so se bili sposobni prilagoditi razli.nemu okolju, s katerim so se sre.evali. Glede DOI> 10.4312\dp.49.12 Colonization dynamics of LBK farmers in Europe under geostatistics test na poselitveno dinamiko v srednji in zahodni Evropi podatki ka.ejo sistemsko povezanost postavit­ve, stabilnosti in opustitve naselij, skladno z difuzijo LTK pre.ivetvenega sistema. Postopno zmanj.e­vanje .tevila novih naselij je nadomestila njihova stabilnost, ki se je ohranila do zadnje stopnje pro-cesa .irjenja. V njej so opustitve naselij brez obnove postale, razen v najbolj zahodnih podro.jih, ze-lo pogoste. V .asovno modeliranje nismo vklju.ili najvzhodnej.ih delov Evrope, saj so kronolo.ki po­datki, ki so na voljo, premalo natan.ni. Z analizo smo osvetlili sistemsko geo-okoljsko variabilnost in modalitete ter prostorsko-.asovne omejitve odpornosti agro-gozdno-pa.nega sistema prvih kmeto­valcev ter potrdili oz. pojasnili veljavne interpretacije LTK fenomena. KLJU.NE BESEDE – neolitik; LTK; Evropa; kolonizacija; geo-okolje; poselitvene dinamike Introduction This paper discusses the settlement dynamics of the LBK (Linearbandkeramik), from Hungarian Trans-danubia to the shores of the English Channel and the Black Sea, between 5550–5500 and 4950–4900 BCE (Figs. 1.a and 2). The origin of this large cultu­ral complex is to be sought in the peripheral areas of the Painted Pottery cultures of the Carpathian-Balkan regions, most notably the Star.evo/Körös-Cris which extended during the first half of the 6th millennium from Serbia to the lower Danube basin. In its early phase of development, the LBK is charac­terized by the low variability of its ceramics, house types and more generally settlement and food pro­duction systems (Bánffy, Oross 2010; Czekaj-Zas-tawny 2009; Kreuz et al. 2005; Kulczycka-Lecieje­wiczowa 2000; Lenneis 2001; 2003; 2010; Lenneis, Pieler 2016; Lichardus et al. 1985; Lüning 2005; Oross, Bánffy 2009; Quitta 1960; Pavlù 2000; Sa­lavert 2017). Recent bioarchaeological and palaeo-genetic research supports the migration hypothesis developed as early as the 1920s (Childe 1929) and describes a process of direct population movement (Brandt et al. 2013; 2015; Lipson et al. 2017; Ma-thieson et al. 2015), in addition to the internal mo­bility of individuals, particularly women (Price et al. 2001; 2002; Bentley et al. 2012). Current palaeoge­nomic analyses point to an incipient genetic admix­ture from local hunter-gatherers during the LBK pe­riod, further developing through the succeeding cen­turies (Brunel et al. 2020; Rivollat et al. 2020). Prior to these new insights, a pure migration hypothesis formed the basis of a multi-agent modelling ap­proach (ANR-09-CEP-004-01/OBRESOC project), ex­ploring multiple socio-environmental interactions (Bocquet-Appel et al. 2015; in prep.; Dubouloz et al. 2017). Following this experimental approach and based on the archaeological data collected on this occasion (Dubouloz, Gauvry, in prep.), our objec­tive here is to carry out a detailed evaluation of the site database and its capacity to account for settle­ment dynamics. In order to address this issue and refine our understanding of its geography, a plani-metric dataset was built around this archaeological database, inspired by the GIS approaches developed in the ‘ArcheoMedes’ and ‘ArchaeDyn’ programmes (van der Leeuw et al. 2003; Gandini et al. 2012; Gauthier et al. 2022). After a general analysis of the archaeological and en­vironmental data, we will address the question of the dynamics of archaeological settlements based on measures of geographical distribution. An in-depth exploration of the agro-ecological rela­tionships involved in this archaeological distribution will be the subject of another paper developing a space-time model of the LBK expansion process. Fi­nally, to bring this research to a close, a third contri­bution will focus on the role of salt resources in the overall LBK process, in relation to this ecological modelling. The archaeological data The analysis includes 6639 sites, spread over 14 coun­tries and eight major river basins (Danube, Dniestr, Elbe, Oder, Rhine-Meuse, Seine, Vistula, Weser). The database (BD_LBK) is informed by various descrip­tors presented in sections 1.1 and 2.1 and Appendix 6. The initial ‘OBRESOC’ database, with 6566 units, was corrected and reorganized to enable SQL que­ries, and then amended by including data from Ro­manian Moldavia (Robin Brigand, Oliver Weller) and the Republic of Moldova (Madalin-Cornel Va¢leanu, Brigand, Weller). The study window was extended by a buffer zone of 70km around outlying LBK sites. This measurement corresponds to half the observ­ed maximum inter-site distance, which is theoretical­ly possible to cover in two days’ walk. This brings the total study area to 1 211 807km2. For some re­gions, the BD_LBK contains fewer sites than those recorded in the literature. This is principally the case Robin Brigand, Jérôme Dubouloz, and Olivier Weller in the Kraichgau and Middle/Upper Neckar (Baden-Württemberg), where the number of known sites is much higher than in the database (Bogaard et al. 2016). As the latter was designed as an inventory of fully identified sites, many locations on published maps without precise bibliographic and geographi­cal references have been ignored. Additional work to identify all the available sites will be necessary here in order to address these gaps. The question will arise later of the impact of these deficits (5%) ~ on our overall results (see section Brief synthesis and perspectives). Archaeological descriptors Each site is recorded by several descriptors of a strict­ly archaeological nature or linked to a brief geogra­phical description (countries, rivers watershed). They are presented below in increasing order of relevance. . The ‘Country’ variable is coded by two letters (ISO 3166-1 alpha-2). It reflects the general effects of his­tory and research policies in major political regions. . The ‘Basin’ variable is derived from a simplified map of rivers drawn up by the European Commis­sion Environment Agency (Catchments and Rivers Network System, 2012). It shows the apparent struc­turing role of river basins in the geography of LBK diffusion. After verification of its operational value, it will not be used in the multivariate analyses. . The ‘Nature of the intervention’ variable is a means of assessing the primary quality of the archaeologi­cal data, mostly derived from random collection (Collect., 50.7%), organized survey (Survey, 23.9%) and partial excavation (F.Part., 22%); extensive ex­cavations are very much in the minority (F.Ext., <4%). This variable makes it possible to modulate the results according to the uncertainties of the data­base. . The variable ‘Category’ of sites (settlement, en­closure, shelter/cave, cemetery, etc.) is mostly not filled in (67.5%). Therefore, on a European scale, all the entries in the database are integrated under the neutral term ‘site’. . ‘Spatial accuracy’ is of uneven quality: 39.6% of the coordinates of the sites are very well to fairly well positioned (Loc ++ L100m and Loc +– L500m), while 60.4% are more poorly located (Loc –– .1000m). But this range of uncertainty usually coin­cides with the resolution of the geo-environmental data. . The ‘Area’ of a settlement should be a relevant criterion for describing its importance, but the low occurrence of documented instances (18.4%) rules out statistical analysis. . The ‘Dating’ variable (see section Chronology of settlements and Appendix 13 for more details) de­scribes the dating evidence in Boolean coding. Sites dated by major chrono-cultural stage are characteri­zed by a ‘first settlement’ variable (St. x) and a ‘du­ration of occupation’ variable in years of continuous occupation (D. x00); the absence of information is coded ‘nr’ in both cases. The mapping of dating pre­cision (Precise = St. x Loose = St. nr) reveals a very strong dichotomy (Fig. 1.a) between Western Eu­rope (France, Belgium), where sites are more often well dated, and Eastern Europe (Ukraine, Romania, Republic of Moldova). Central Europe (Fig. 1.b) is intermediate in this respect, with some well-docu­mented sectors (Hungary, Austria, Slovakia and Ger­many to a lesser extent) and others that are often more limited (Poland, Czech Republic). . ‘Reliability’ is a synthetic variable which descri­bes the overall quality of the archaeological map. It combines the values of three variables: geography (accuracy of location), chronology (accuracy of dat­ing) and contextual expertise (type of fieldwork). The mapping of this index is carried out through a transformation of the point entity into a raster infor­mation with 5km sides, from the most reliable to the least reliable. This mode of representation offers a general view of the state of research on the data (Fiab.A = 3 positive variables; Fiab.B = 2 positive va­riables including chronology; Fiab.C . 2/3 negative variables = 80.2% of sites). There is a strong hete­rogeneity in the database (Fig. 2), very similar to that produced by the ‘Dating’ variable. The western and southern ‘half’ (Seine, Meuse and Moselle basins, Rhine and Weser basins, Hesse and Württemberg, Danube valley, Austria) is generally well document­ed. The other ‘half’, in the centre and the east of the map (Thuringia/Saxony-Anhalt/Saxony, Bohe­mia, Moravia, Slovakia, Hungary, Poland, Romania, Moldova and the Ukraine) often shows a more limi­ted documentary quality. Multiple Correspondence Analysis (MCA) A multiple correspondence analysis (MCA) was ap­plied to the complete disjunctive table of most of the BD_LBK’s discrete data (Benzecri 1973; Escofier, Pages 1990). This analysis will characterize the do­cumentation in its suitability for addressing the is­sues of the determinants and rhythms of the LBK Colonization dynamics of LBK farmers in Europe under geostatistics test Fig. 1. Upper: LBK settlement map (6639 features) and chronological accuracy (each site – or group of sites – is represented by a pixel of 25km2). Bottom: map of LBK sites in Central Europe (Poland, Slova­kia, Hungary, Czech Republic, Austria, Germany) precisely dated (black dots) and simply dated LBK (white dots). Robin Brigand, Jérôme Dubouloz, and Olivier Weller Fig. 2. LBK settlement map (6639 features) and overall reliability of the archaeological information (each site, or group of sites = 25km2/pixel). Hydronyms are abbreviated: Bu. Bug, Da. Danube, Dn. Dnie­str, Dr. Drava, El. Elbe, Is. Isar, Ma. Main, Ma. Marne (Fr.), Me. Meuse, Mo. Moravia, Mo. Moselle, Od. Oder, Pr. Prut, Rh. Rhine, Sa. San, Se. Seine, Va. Vah, Vi. Vistula, Vl. Vltava, Wa. Warta, We. Weser, Yo. Yonne. agro-sylvo-pastoral system in its continental devel­opment. Six variables are mobilised for 38 active modalities: the variables ‘Category’, ‘Area’ and ‘Ba­sin’ were excluded from the analysis because of the insufficient quality of the available information or because they were too general. A preliminary reading of the relative contributions of the six variables themselves provides a frame­work for the following detailed analysis of their mo­dalities involved in the MCA (Fig. 3). This first approach shows that five variables out of six contribute, through some of their modalities ana­lysed below, to explaining the intrinsic variability of the data recorded in the inventory. The ‘Country’ variable does not seem to play a particular explana­tory role in the overall construction, which is reas­suring. The ‘Reliability’ variable, despite its poly­thetic dimension (cf. 1.1–8), does not sum up the patterning of the data alone: its statistical weight is only significant on the first factor. We shall see be­low that different modalities of its own descriptors are particularly discriminating throughout the ana­lysis. The other four variables do not show very strong correlations: for example, ‘Duration’ is not strictly related to the age of the installation (‘Chro­nology’) nor to the research intensity (‘Fieldwork’); in the same way, geographical precision seems only partially related to ‘Fieldwork’. These general obser­vations suggest that a rather complex dispersion cha­racterizes the variability of the data, which are there­fore not univocally determined. Description of the factors or components (Appen­dices 2–4) The detailed analysis of the first three factors of the MCA, then of the factorial maps of their 38 modal­ities, makes it possible to specify the previous obser­vations (Appendices 2–4, 7 and Figs. 4 and 5). The first factor (original eigenvalue of F1 = 10.69% or 77.79% in Benzecri correction) is detailed in Ap­pendix 2. It contrasts the sites well documented by excavations, dating to different stages of the LBK (especially St. 2 and 4) and of rather limited dura­tion, with the less well-documented and therefore less reliable sites: France and Poland/Czech Repub­lic respectively embody these two opposing poles which, not surprisingly, dominate the structuring of the data. The second factor (F2 = 5.55% or 8.45% Benzecri, Appendix 3) mainly articulates the temporal varia­bility around the differential documentary effects of ‘Survey’ and the ‘Collect’. Colonization dynamics of LBK farmers in Europe under geostatistics test The third factor (F3 = 5.07% or, 5.53% Benzecri, Appendix 4) specifies the variability of the borders of the time sequence, by underlining the opposition of older and long-lived Central European sites, re­sulting from ‘Survey’ prospection (52.1% of the iner­tia), to westernmost sites dating to the later stages of the LBK and correctly located (31.2%). It also shows that some countries linked to ‘Collect’ in F2 (Slovakia, Hungary) are additionally associated with ‘Survey’. Analysis of maps F1/F2, F1/F3 and F2/F3 (Figu­res 4–5 and Appendices 5–7) . Map F1/F2 (Figure 4 and Appendix 5) shows a qualitative and chronological structure in the data and reveals three main groups of modality associ­ations. It contrasts extreme documentary situations, between France/Netherlands (best information) and Poland/Czech Republic (worst information). In addi­tion, it defines for part of Central Europe an inter­mediate group of sites with variable documentation and often ‘average’ quality. Germany and Belgium, at the centre of the scatterplot, are linked to all qua­lities of documentation, but with only a small pro­portion of the best; Austria has a significant rate of good quality information; Hungary and Slovakia are isolated from the other countries mainly because of their stronger ties to the early stages of the LBK. . Map F1/F3 (Figure 5 and Appendix 6) identifies, in addition to the extremes already identified in F1, a central cluster of medium to high level modalities; this indicates a fair to good reliability shared by many sites in central-western Europe belonging to the main part of the LBK sequence and of all dura­tions. It also confirms the better documentation of the western trio (Netherlands, Luxembourg, France), and the more varied documentary quality in Ger­ many, Belgium and Slovakia. The cluster of countries gathered around the least posi­tive modalities confirms the singularity of Poland, linked to ‘Survey’. . Map F2/F3 (Appendix 7) characterizes the database without the influence of poor documentary situations; it shows links masked by the hyper-structuring weight in F1 of the strong correlation of qualitative­ly weaker or undefined modalities. Hence it confirms the existence of the same poor­ly or badly characterized sites in the BD_ LBK (here centrally positioned and there­fore not very significant in describing this additional variability) and their distinction with the very well-characterized group in France and the Netherlands in the later stages of the LBK. In addition, this plot underlines a significant and fa-vourable situation with Poland and its systematic ‘Survey’ practices. It also confirms that, at the time of database completion, Hungary and Slovakia stand out strongly from the other European countries in terms of the relative weight of their older, long-term and partially documented sites. A synthesis of these successive and complementary analyses is quite easy to make, as they provide fairly similar results on the variability of the documenta­tion. Three major poles emerge: first, the group with the best possible combination of documentation, in Western Europe, with sites in France, the Nether­lands and Luxemburg; second, the group with low­est quality information, around the Czech Republic and partly Poland; third, the combination of inter­mediate quality information in Hungary, Slovakia, Germany, and Belgium. The remaining five ‘Country’ modalities are divided between the poor-quality com­binations (Moldova, Romania, Ukraine, Switzerland) and the medium-quality ones (Austria). Apart from the western trio, Poland and Hungary/ Slovakia both present a complex situation. They can be regarded as specific versions of poor and mode­rate documentation. Thus, Poland shares with the western part of Europe a significant proportion of sound documentation (Fig. 4, map F2/F3), thanks to its affinity for high-quality ‘Survey’. In the group of medium quality documentation, Hungary and Slovakia stand apart from others (see Figs. 4–5 and Appendix 7). This may be related to their greater connection to the earlier stages of the LBK. The mid-quality group is also well represented by Germany and Belgium, which possess all the variety of docu­ Robin Brigand, Jérôme Dubouloz, and Olivier Weller mentation in significant proportions. Some is rather poor and broadly similar to the qualitatively under-documented Eastern Europe and Czech Republic, where the huge number of entries in the database statistically overshadows the well-documented sites in Bohemia and Moravia. Finally, Austria’s documen­tation lies between the good and medium quality groups, quite close to that of the western countries. Quality of the BD_LBK and overall heuristic validity. The overall impact of the observed heterogeneity (Appendix 1) in the BD_LBK on the investigation of geo-environmental aspects and LBK settlement dynamics will now be assessed. From the point of view of geography, this diversity is arranged quite clearly along an east-west transect; our information thus gains in quality along this axis and its suitability for interpretation increases at the same time. Apart from language barriers (for Ukraine and Moldova), three logically related constraints con­tribute to this overall European pattern: the number of sites recorded, the temporal depth of the research and the underlying types of archaeological interven­tion. Thus, three countries which are major contri­butors to the database (Germany, Czech Republic, Poland) are characterized by an overall documen­tary situation of average to mediocre quality; and three other countries, lesser contributors to the data­base (France, Netherlands, Austria), present much more favourable situations. This dichotomy, which is somewhat counter-intuitive, is probably due to a mass effect. For the first countries mentioned, the temporal and spatial extent of research has produc­ed a high proportion of sites identified ‘in the past’, while settlements that are well documented by more recent research have in this case relatively less weight than in countries where more recent, more proactive and/or more localized research will have accumulated fewer poorly documented sites. Several intermediate situations are obviously pre­sent. For example, Hungary and Slovakia each illus- Fig. 4. F1/F2 map of the MCA showing the archaeological descriptors. Size and coloured highlighting of the fonts symbolize the level of the modalities’ contribution; radius of the circles shows their represen­tation’s rate; red contours underline the particular distribution of the countries. Colonization dynamics of LBK farmers in Europe under geostatistics test Fig. 5. Graph of the 1/3 factorial map (F3 inertia: 5.07% or, 5.53% in Benzecri correction). Same legend as Figure 4. trate their specificity, and their close positioning seems as much related to the broad trends that structure LBK data as to their own similarity. Thus, their convergence is probably mainly due to the distribution of the chronological variable: these two countries have relatively few well-documented sites and comparable proportions of moderately to poor­ly documented sites, but their share of later LBK sites, lower than in the other countries, may explain their joint isolation in the results of this analysis. Finally, at the geographical margins of the phenome­non under study, the situations are broadly opposed. One, in the east, appears to be mediocre overall, no doubt due to shortcomings in the collection of infor­mation, but also possibly due to a very different hi­story of research and more limited funding. The other situation, in the west, generally appears to be good, probably due to the geographical limitation of the areas concerned (the Netherlands and Luxem­bourg) and/or significant and earlier funding of res­cue archaeology. Thus, as was to be expected, the ‘Country’ variable is not an explanatory descriptor, but is only relevant in highlighting contextual trends. These can be sum­marized by five fairly well-differentiated classes and sub-classes: (i) the class with little information, main­ly comprising eastern sites; (ii) the sub-class ‘Po­land’, which has limited information in general, but is often well documented by ‘Survey’ or extended excavations (‘F.Ext.’); (iii) the class of mixed situa­tions, comprising Belgium, Germany, Slovakia and Hungary, with comparable proportions of poorly and moderately documented sites, together with some large to huge archaeological investigations; (iv) the sub-class ‘Austria’, which has fairly good in­formation overall; (v) finally, the class of the most favourable situations, mainly associated with the Netherlands, Luxembourg and France: Austria parti­cipates to a certain extent in this last group (F1/2). This extensive information variability could have a negative impact on the use of the database for re­search, by restricting the number of comparable sites available in different regions and by blurring the conclusions to a greater or lesser extent, depending on the regional representativeness of the available samples. However, in the broad analysis of the agro-ecological determinants of LBK settlement, it emer­ges that these documentary limitations have only a marginal impact. Indeed, such an approach mainly requires geolocation in line with the resolution used for the geo-environmental space, and this is often enough to provide a good summary of the main cha­racteristics of the LBK point-pattern. Basic geostatis­ Robin Brigand, Jérôme Dubouloz, and Olivier Weller tical techniques such as low-pass filtering will be able to overcome the main documentary weaknes­ses. On the same geo-environmental level, the char­acterization by stages of LBK expansion will be more uncertain, particularly in Eastern Europe, a region that will have to be excluded from detailed analyses. But for the large contributors in Central Europe, with considerable medium- to poor-quality documen­tation, the problem is mitigated by the significant number of available sites and their varied spatial di­stribution. The projected research on settlement dynamics is in turn dependent on the quality of information on the chronology and continuity of occupations. The MCA presented in this section has highlighted the weight of the undetermined modalities in the structuring of the data: thus 60% of the sites in the BD_LBK do not meet the fine chronological aspects. The majority of these sites belong to the eastern regions of Europe, which again cannot be compared with other regions (although there are some pioneering studies such as Dmytro Kiosak (2017) and Thomas Saile (2020)). Among the latter, the three largest contributors to the database (Germany, Czech Republic, Poland) cer­tainly present a significant rate of negative respon­ses (in Thuringia/Saxony/Saxony-Anhalt (Thüringen/ Sachsen/Sachsen-Anhalt in graphs), Bohemia (.echy in graphs), Little Poland (Ma³o-Polska in graphs) and Kuyavia (Kujawy in graphs)), undoubtedly challeng­ing the representativeness of these samples with re­gard to the regional temporal variability of LBK oc­cupation; but the distribution of correctly to very well-informed sites overlaps quite regularly with the general pattern of known sites (Fig. 1), giving reason to hope for good spatial representativeness. Unless we can correct these biases, our forthcoming histo­rical-anthropological interpretations will have to take this uncertainty into account. Topographical and agro-ecological data In this section we will focus on site locations (point coordinates) in order to characterize their geo-con­texts. The set of variables integrated in the analysis is presented in Appendix 8. Each archaeological site is thus documented by two topographic and six agro-ecological variables for a total of 38 descriptors out of 41 (two descriptors and one modality, not repre­sented in the data, were deleted). Data classification . Slopes. The slope map is classified into four map units: SL1 (flat) = from 0 to 3; SL2 (low slope) = from 3 to 9; SL3 (medium slope) = from 9 to 20; SL4 (steep slope) = from 20 to 65; in the analysis, the SL4 modality was grouped with SL3. . Landscape openness. Classified into three sets, this measure is a morphometric indicator that deter­mines the degree of concavity and convexity of the landform (Yokoyama et al. 2002). It considers for each pixel a line of sight over 8 azimuths (r=10km) which may be redundant with the visibility indica­tors (total visibility): OP1 (closed); OP2 (open); OP3 (very open). . Moisture. This is the variable deduced from the low-resolution (1km) surface water saturation map. The map is classified into four map units: HY1 (low); HY2 (medium); HY3 (high); HY4 (very high). . Soil texture. This document, originally classified into seven classes, is simplified into six classes (by combining the and fields): TX0 (n/a); TX1 (coarse, clay<18% and sand>65%); TX2 (medium, clay<35% and sands>15% or clay <18% and 15%60%). . Soil thickness. The original five-class classifica­tion is retained: PR0 (n/a); PR1 (shallow, <40cm); PR2 (moderate, 40–80cm); PR3 (deep, 80–120cm); PR4 (very deep, >120cm). . Loess. The European loess map (Haase et al. 2007), produced at a scale of 1:2 500 000, contains nine classes distinguished according to their thick­ness and degree of alteration; this document has been simplified to the extreme to retain only the exi­stence or not of loess in the substrate: LO1 (pres­ence); LO2 (absence) . Soil group. The 1:1 000 000 Soil Geographical Da­tabase of Europe (SGDBE v4) initially presented in 27 classes according to the 1985 FAO classification has been simplified to 11 classes (European Soil Data Centre (ESDAC), esdac.jrc.ec.europa.eu, European Commission, Joint Research Centre): SO1 (Cambisols = brown soils with little typing or differentiation); SO2 (Chernozems, Greyzems and Phaenozems = soils with a high content of organic matter in their upper horizons, clinohumic soils); SO3 (Fluvisols = alluvial soils); SO4 (no data = glaciers, city, rivers, destroyed soils, n/a); SO5 (Gleysols, Histosols, Marshes, Plano-sols, Vertisols); SO6 (Podzol = soils marked by a true podzolization process); SO7 (Luvisols, Pozoluvisols = Colonization dynamics of LBK farmers in Europe under geostatistics test degraded and leached soils characterized by eluvial upper horizons); SO8 (Lithosols, Regosols, Rankers, Rendzinas, Outcrops = thin soils, limited in depth by a coherent and hard material (bedrock or hardened horizons); SO9 (Arenosols = soils with sandy granu­lometry, not differentiated); SO10 (Salinosols = sa­line soils of Solonetz or Solonchaks type); SO11 (An-dosols = soils of volcanic origin). . Fertility. The fertility map produced as part of the ANR ‘Obresoc’ project (Schwartz et al. in prep.) was used without regrouping or modification. The simplification of the soil cover carried out by the authors is dictated by an attempt to reconstitute Neo­lithic soils. As such, this work differs from classic studies – and from those presented here – which are based on the current state of the soil. Nevertheless, the scale (1:5 000 000) means there is great simpli­fication and often arbitrary cartographic limits: FR0 (n/a); FR1 (very fertile); FR2 (fertile); FR3 (medium fertility); FR4 (low fertility); FR5 (not fertile). Statistical analyses (MCA and HAC) All statistical analyses performed on the LBK set­tlement data are based on the cross-referencing of the point attributes with the set of environmental descriptors operated in matrix form. The characteris­tics of the locations are at pixel level (100m), but include the very close environment since a low-pass filter has been applied to the raster data. The value of each cell is calculated from the eight surrounding pixels, thus obtaining a smoothed raster where the maximum and minimum values of each neighbour are averaged and local variations are reduced. A multiple correspondence analysis (MCA) of the to­pographic and agro-ecological modalities and a hie­rarchical ascending classification (HAC) on the fac­torial coordinates (factor scores) enables a descrip­tion of the settlement choices of LBK populations. Description of the factors (Appendices 9–11) . The strongest contributions to factor 1 (85% of ~ its total variance or inertia) mainly separate sites in poorly defined or rare contexts from all other sites (Appendix 9). Its relevance to our research questions is therefore very limited. . The second factor (Appendix 10) is mainly based (84% of its inertia) on the opposition between ~ leached (Luvisols) or degraded soils, very deep, with a medium-fine texture and good fertility, and brown soils with moderate to deep depth, fine texture, very good or medium fertility, in a closed landscape. This second statistical summary thus distinguishes two kinds of favourable agro-ecological environments. . The third factor (Appendix 11) is mainly based (81% of its inertia) on the predictable opposition ~ between deep and very fertile soils of Chernozem, Greyzem and Phaenozem types and very thin soils or soils with a tendency to waterlogging. Analysis of map F2/F3 (Fig. 6) Because factor 1 is defined by modalities that main­ly describe the absence of information, it is more re­levant to focus on factors 2 and 3 of the MCA (Fig. 6), despite a reduced inertia (total = 14.3%, 35.7% in Benzecri correction). With the results of the HAC conducted on the factor coordinates of the modali­ties, it reveals four main classes (Cl.1 to 4). . Class 1 (Cl.1), with 2,916 sites (44% of the total), is predominately determined (>90%) by sites locat­ed on very deep (PR4), degraded (SO7) or alluvial (SO3) soils; then (> 50%) on medium-fine textured soils (TX3), fertile (FR2), humid and tending to hy­dric saturation (HY3); to a lesser extent (35–50%) by sites located on flat (SL1), open (OP2) to very open (OP3) landforms and medium hydromorphic soils (HY2). Two poles of association of agro-ecolo­gical modalities constitute class 1: the first one is very specific, around degraded, very deep, modera­tely fine-textured soils, and the second one around the most shared modalities, including good fertility, average waterlogging, and flat or very slightly slop­ing land in an open landscape. The presence of loes-sic substrates does not seem to be very discriminat­ing, since their absence also contributes to defining this subgroup of CL.1. . Class 2 (Cl.2) is composed of 132 sites (2%) locat­ed on soils for which the depth (PR0) and nature (SO4) are not given. This very marginal class of un­characterized sites is not very relevant. . Class 3 (Cl.3) concerns 1414 sites (21.3%) locat­ed primarily (>90%) on clinohumic (SO2) or sandy (SO9) soils; then (>50%) on very fertile soils (FR1), or soils marked by a process of true podzolization (SO6) and to a lesser extent (35–50%) on deep soils (PR3), with low surface hydromorphy (HY1) and sometimes with a coarse texture (TX1). Two pre­ferred associations of agro-ecological modalities are also visible in class 3. The first one, very characteris­tic, concerns sites on clinohumic soils, very fertile, and the second one sites on deep soils with low wa­terlogging. Robin Brigand, Jérôme Dubouloz, and Olivier Weller Fig. 6. Plot of the MCA showing topographic and agro-ecological traits (map F2/F3, inertia F2: 7.69% – 22.95 in Benzecri correction; F3: 6.58% – 12.74% in Benzecri correction) and arrangement of the class­es from the HAC. The class boundaries are indicative; they do not exclude other under-represented mo­ dalities. . Class 4 is made up of 2177 sites (32.8%), located primarily (>90%) on moderate depth soils (PR2) or shallow soils (PR1), waterlogged (SO5) or thin soils (SO8) or even saline (SO10); then (>50%) on brown (SO1), moderately fertile (FR3) to poorly fertile (FR4), fine-textured (TX2) soils located in topographic en­vironments marked by a steep slope (SL3) and a closed landscape (OP1); and to a lesser extent (35 to 50%) on slightly sloping (SL2), non-fertile (FR5) and moderately textured (TX3) soils. This very heteroge­neous class is structured around two very significant groups of modalities: the first concerns sites on brown soils of moderate depth, fine texture and, in part, average fertility in a rather closed landscape; the second group, very characteristic, but much less represented, is made up of shallow soils of low ferti­lity, such as Gleysols, Vertisols, Rendzinas and Rego-sols. Intermediate conclusions The broad analysis of the agro-ecological data thus highlights the great variety of contexts in which LBK settlements are found. Rather than, as has often been assumed, a simple constraint by loess (47.6% of the total number of sites are strictly located on a loessic substrate), it is a set of characteristics favourable to sustainable farming that conditions the most fre­quent location of these sites: very good to average fertility, very great to average depth, not too pro­nounced waterlogging, fine to average textures. Many different soil types (SO1, SO2, SO7, even SO3 and SO6), whether on loessic or non-loessic substra­tes and widely present throughout Central Europe, seem to have been acceptable compromises for these early farmers. It is thus clear that they had an ances­tral practical knowledge of the agrological properties of the soils, including their natural vegetation and Colonization dynamics of LBK farmers in Europe under geostatistics test associated fauna. This background will have succes­sfully guided their settlement choices in the Central European landscape. Additional analyses As regards the importance of loess and major soil types in the choices of location, further analysis can be made on a continental scale in order to identify interregional and/or temporal variability. Uncertain­ties related to soil evolution over the last 7000 years will be addressed in the course of the analysis and in the conclusions that can be drawn from the current soil cover. The loess issue Loess substrates make up more than a quarter of the surface area occupied by the LBK as a whole (26.2%). This share is greater if one excludes the mountain ranges, largely avoided by LBK agro-pastoralists. Loess is fairly evenly distributed (see Appendix 12), with a particularly marked presence to the northeast and east of the Carpathians, as well as in Hungary, around the Harz mountains and in Belgium/the northern Paris Basin. It is therefore expected that in most of these regions, loess may have been a major substrate for the establishment of LBK settlement. Two approaches have been followed to specify the archaeological impact of these formations. One is based on the geographical coordinates of the sites and the other on a buffer zone of 2.5km around each site. This second approach partly reduces some of the imprecision in locating the sites and mapping the loess areas. According to the first approach, the general trend between the Black Sea and the English Channel shows (Fig. 7.A) an almost equal number of locations on loess or other substrates (47.5% of occupations and 47.6% of localities). At a more precise temporal scale (Fig. 7.A and B), and therefore at a varying geographical scale, a trend can be seen towards a lesser dependence on Loess (Chi2: df 4, p. 0.001) between the beginning and the end of the LBK (from 55% to 42.4%), well expressed by the deviation from the mean of the occupations. This overall develop­ment thus underlines the importance of loess, but mostly at the beginning of the process, as well as in­creasing adaptation to the varied geo-pedological contexts encountered. This space-time trend, identified on the basis of the present-day distribution of loess across Europe, mo­tivated a second approach and further investigations of the geo-pedological composition of the terroir within a radius of 2.5km around each site, by major settlement region. In order to evaluate loess depen­dence, three surface thresholds in the buffer zone were established: (1) with 80% of loess in the buffer area, the site is considered to be located to maximize the presence of this formation, or has simply bene­fited from its extensive availability in the area; (2) conversely, with between 0 and 20% of loess in the buffer area the site is considered to be located with no particular interest in the presence of this sub­strate or has simply suffered from its low availabi­lity; (3) an intermediate threshold at 40% of the buf­fer area is also tested to define a proximity of the site to loess (<1km), but without excluding other substrates. According to this analysis, approximately 30% of all LBK sites are loess-independent, 34% have predominantly a loessic substrate, and 20% have a broad access to other types of geology. The remaining 16% (between 20 and 40% loess surface) represent sites with little attraction to this loess. At this level of geographic and temporal resolution, the importance of loess in the location of LBK sites is clearly variable. The distribution of this differential access to loess ac­cording to major settlement regions (98% of sites), reveals a number of different patterns of settlement (Fig. 8.A-B). Fig. 7. A proportions of sites punctual location on a loess substratum; B deviations from average Robin Brigand, Jérôme Dubouloz, and Olivier Weller . The opposition of the extreme situations (80% vs 20% of the buffer area) creates four groups (Fig. 8.A): . A first group (green) with 42–64% of sites with over 80% loess and 15–34% of sites loess-indepen­dent (0–20% of their buffer area), occurring in re­gions where the spatial distribution of loess corre­sponds to large, compact sectors (Niederrhein­Westph./Maas (NWM), Wetterau/Main, Franken, Elbe-Saale, Austria (Österreich in graphs), Southern Vistu­la, Romania/ Moldova). . A second group (yellow) with only 12–30% of sites with over 80% loess and 30–48% of sites loess-in­dependent, occurring in regions with smaller and more discontinuous loess sectors (Alsace-Middle Rhine (Alsace-Mittel-Rhein in graphs), Neckar, Low­er Saxony/Fulda-Weser (BSFW), South Bavaria/ Schwabia (Süd-Bayern/Schwaben in graphs (SBS)), Bohemia, Moravia, Slovakia-Hungary and South-East­ern Poland-Ukraine (SPK)). . A third group (orange), far western [Paris Basin/ Hainaut, Franco-Luxembourg Moselle (BPHM)], where only 5–8% of sites have over 80% loess (main­ly the nine sites in Belgian Hainaut) and 67–78% are loess-independent; . The fourth and last ‘group’ (dark red) is made up of the Kuyavia/Chelmno area (KUJ), where 100% of sites are not located on loess. . At the threshold of 40% loess in their buffer area (Fig. 8.B), the sites fall into almost the same four groups as above: . The fourth and third groups (red and orange) are formed by the same regions (KUJ, BPHM), where the attraction of loess concerns only 0% and 11–15% of sites, respectively, except in the small Belgian Hain­aut region. This overall statistic confirms the weak association of the LBK with loess in these regions. . The second group identified above concerns 40– 62% of the sites in the same regions, to which are added the Moldavian-Romanian and Southern Vistu­la regions. By doubling or even tripling the propor­tion of sites concerned, the ‘mixed’ terroirs thus re­present a considerable weight in these regions and reveal a fairly strong link between the LBK and loess (50% ± 10). . The first group retains five of the seven regions with 80% loess, and now includes 72–83% of the sites in NWM, Franken, Wetterau/Main, Austria and Elbe-Saale, which is only a slight increase (1.3 to 1.7 more cases). This weak progression is due to the high spatial availability of loess, which ‘mechanical­ly’ attracts a large number of sites (42–64%). But, by increasing this proportion of sites to 72% and more, the ‘mixed’ sites contribute to the dominant view of the crucial importance of loess for the LBK. This ob­servation remains true for some regions of Germany and Austria, where Eva Lenneis (2001) notes that about three quarters of the settlements are located on loess, from the earliest phase onwards. But in other important regions of Central Europe (Hunga­rian Transdanubia (Magyar-Dunántúl in graphs), Mo-rava, Bohemia, South Bavaria/Schwabia (Südbayern/ Schwaben in graphs), Middle Rhine/Alsace (Mittel­Rhein/Alsace in graphs), the dominance of loess Fig. 8. Geographical distribution of sites according to the proportion of loess area in their 2.5km buffer zone: A <20% vs. >80%; B <20% vs. >40%. Colonization dynamics of LBK farmers in Europe under geostatistics test must be tempered by a significant proportion of ‘mixed’ configurations associating proximity of loess and other substrates. Intermediate conclusions The analyses at supra-regional and regional levels of site position and buffer areas both confirms and moderates the importance of loess in the choice of LBK site location, as already revealed by the MCA (Fig. 6): only 52.3% of the large sample of sites exa­mined (Fig. 8) have over 40% of loess in their buf­fer areas. As the results show many alternatives, the study indicates a certain independence of the LBK system from loess, even in regions where its pres­ence is high, such as the eastern Carpathians or the Transdanubian Hungary. However, there is a slightly stronger link at the beginning of the sequence (St1) than at the end (St4 and 5–6). These figures suggest that starting from a subsidiary exploitation of non-loessic substrates by the earliest farmers (St1), the LBK’s subsistence system has been increasingly adapted to this kind of substrate. LBK and present-day soil groups The geographical distribution of present-day soil groups shows the predominance of Cambisols and Luvi-Podzoluvisols (57% of the total area), far ahead of Clinohumic soils (Chernozems, Phaenozems, Grey-zems), Podzols, Fluvisols, Gleysols and Renzine-Re­gosols (Fig. 9 and Appendix 12). Cambisols develop in large continuous areas in the heart of the LBK territory and particularly in the central and central-western parts. Luvisols (and Pod-zoluvisols) generally border these Cambisols, in ra­ther fragmented patches, and dominate in Poland, as well as in the north-western and western parts of the study area. Clinohumic soils are distributed main­ly in the east and south-central regions, on the peri­phery of the above soils, with some other good pat­ches in the Elbe-Saale region and more modestly in southern Poland and Bohemia. Gleysols occur be­tween the previous groups, mainly in the northern periphery, like the Podzols, and more in connection with Fluvisols. The latter are distributed throughout the study area, particularly in Moldova/Romania, Po­land, Belgium and France. Lastly, Rendzinas essen­tially show the opposite distribution to Podzols, in the southern half of the LBK occupation zone, with important areas along the Paris Basin, the Austrian Danube and in Hungary, Slovakia and Ukraine: the LBK largely avoided these soils. This current distribution of the main soil groups, ex­pressing major topographical, geological and mor­phological factors, appears to be a significant geo­graphical determinant for LBK settlement location. We can nevertheless expect considerable variability, as has been shown for the loess. Available soils/selected soils The distribution of soils at LBK sites shows a good overall correlation with the distribution of soils cur­rently available in the study area (Pearson (0.866) and Spearman (0.7381): p. <0.05). This indicates that the early Neolithic farmers made use of most of the possibilities offered by the environment. But the observation of the deviations from the relative dis­tribution of these soil groups in the study area as a whole and at the sites (Fig. 10) shows a stronger relationship (Chi2, df 7, p. <0.01) with Luvi-Pod­zoluvisols (+51.7%), Clinohumic soils (+45.6 %) and Fluvisols (+28.6%). The best represented group in the study area, Cambisol (33.3%) is a bit under-rep­resented (–20.4%), and Renzinas-Regosols, Gleysols, Arenosols and Podzols (between 2.7 and 8.9%) are clearly under-represented at LBK sites compared to current geographical availability (between –34.5 and –88.8%). The importance of Luvi-Podzoluvisols could be at least partly due to soil developments over the last 7000 years (Lorz, Saile 2011) and could reflect the higher availability of Chernozems (here with Clinohumic soils) at the time of the LBK. This hypothesis can be applied especially around the Harz mountains, as well as in Bohemia and Moravia, judging by the present-day intermingling of Luvisols in the Chernozemic areas and their imme­diate surroundings. However, Car-sten Lorz and Thomas Saile (O.c.) also point out that such a proposal can only be based on detailed pedo­ Robin Brigand, Jérôme Dubouloz, and Olivier Weller logical studies. The question of Cambisols is even more complex, since this notion covers very diverse soils whose main common feature is an incipient pe­dological differentiation, and their age is therefore questionable. We will simply note here that more than 80% of the sites located on these formations, very often close to areas where Chernozems and Lu-visols are concentrated, avoid their most acidic forms. This is the case in the regions noted above, especially along the secondary, or lesser, watercour­ses. Here, the ‘absolute’ quality of the soils would have weighed less than the proximity of densely set­tled areas and site networks. Loess and selected soils Together with these apparent connections, there are significant differences in numbers by substrate for the four main occupied soils (Fig. 11). A Chi-square test confirms (df 3, p. <0.001) that the presence of a loessic substrate at the site coordinates tends to fa-vour Luvi-Podzoluvisols and Clinohumic soils in the establishment of LBK sites, especially in the early LBK; conversely, the absence of loess tends to fa-vour Fluvisols and marginally Cambisols, especially towards the end of the LBK. The weighting of these archaeological observations by the geographical patterns of the three main soil groups attested on loess provides a complementary conclusion (Tab. 1): . Clinohumic soils, preferred on loess more than on other substrates (18.7 vs. 14.6%), have appar­ently been exploited in proportions lower than their availability on loess (18.7 vs. 28%), except in their two main distribution zones east of the Carpathians, as well as in Slovakia and more partially in Moravia and Kuyavia (see Fig. 12). . Luvisols appear to be the main beneficiary of the apparent choices of LBK settlement location, espe­cially on loess (> 42% of sites). . Despite the lower overall availability on loess, Cambisols also seem to be readily chosen, and in higher proportions than their overall availability on loess. This trend may reflect an environmental and/ or social determinant, i.e. a pull of the loessic sub­strate on the location of LBK sites, irrespective of the soil itself, and/or a desire to be close to a site network occupying the best soils. Thus, according to these non-exclusive preferential connections, the locational choices of LBK sites as a whole reveal several major alternatives. The ‘Cher­nozem on loess’ archetype would have been only one of the LBK settlement modes, valid in certain regions and/or at the beginning of the sequence. Elsewhere, according to the gradual spatial variation of environmental availability, LBK settlements seem to have largely found other geo-pedological contexts that were suitable for their farming requirements (Figs. 12 and 13). Regional variability An analysis by major settlement zone (Fig. 12) illu­strates this variety of situations in seven main classes: . Romania-Moldova is distinguished by a preponde­rance of clinohumic soils (> 80% of cases). . Slovakia, Moravia (Chernozem) and Kuyavia/Chel­mno region (Phaenozem) show a preference for both Luvisols and Clinohumic soils (>66% combined). . In a large part of Germany Luvisols dominate Cam-bisols (>70% combined). . In contrast, in the west and southwest of Germany, Cambisols predominate over Luvisols (>70% combin­ed); in Austria, Clinohumic soils take second place (=89% combined), as already observed by Lenneis (2001.Fig. 6) for LBK 1 in Austria, where present-day Brown soils are preferred to present-day Chernoze-mic soils. . Bohemia, Southern Vistula and Silesia show a more balanced distribution of Cambisols, Luvisols and Cli­nohumic or Fluvisol (>67% combined). Colonization dynamics of LBK farmers in Europe under geostatistics test % area Soil groups (%) ters in site location, although such % soil on loess\ soil group area % soil\loess area % soil-sites on loess Clinohumics (11.4) 66 28 18.7 Luvi-Podzoluvisols (23.6) 34 27 42.1 Cambisols (33.3) 14 17 25.4 tential, but also natural plant com­munities. The influence of these se­dimentary characteristics on tree and herbaceous formations implies that the associated vegetation and fauna were most likely important parame- Tab. 1. Proportions of various loess configurations/major soil groups. . On the periphery, Hungary, Mecklenburg-Pomera­nia, south-eastern Poland and the Ukrainian borders show a wide variety of soils, including a type that has been encountered rarely until now (only in Kuya-via): Gleysols, associated with Luvisols and Cam-bisols or Fluvisols (>68% combined). . In the west, except for Belgian Hainaut (nine sites), Fluvisols dominate, associated with Cambi-sols (>75% combined) and a variety of other soils. The cartography of these major classes (Fig. 13 and Appendix 12) shows that, during its spread, the LBK farming system did not focus on a specific geo-pedo-logical niche, but adapted to changing local and re­gional conditions. In view of this general strategy, the Paris Basin sites appear somewhat distinct, since they underuse the soils developed on the large loess patches occurring in this region. Mainly dated here to the end of the LBK sequence but locally pioneer, the occupation initially tied to the main floodplains was extended, during the early 5th millennium (BVSG group), to the loessic plateaus now largely covered by luvisols (Dubouloz et al. 2005; Bostyn et al. 2019). The BVSG farmers seem therefore to have returned to the predilections followed during the European expansion process from which they ultima­tely resulted. Overall weighting of results Beyond issues of loess and soils, the distribution of LBK sites highlights their relationship to highly sig­nificant sedimentary characteristics. The MCA/HAC (Fig. 6) identifies groupings that are largely defined around variations in water saturation, texture and soil depth (Tab. 2). And it should be remembered here that soil quality not only affects agricultural po- factors remain hidden in our strictly geo-pedological analysis. Medium-fine to fine-textured soils thus characterize more than 82% of the sites (Tab. 2.A), and of these about 71% have low to medium water saturation. In the same way, a great to very great thickness of the soils characterizes nearly 80% of the sites, of which nearly 72% have low to medium water saturation. These very high proportions identify the main soil characterization of LBK locations and indicate the skills of already expert farmers. However, a signifi­cant proportion of the occurrences (<30%) combi­nes the same characteristics of depth and texture with high water saturation, thus showing that alter­native choices were made. This option may have been related to a particular interest in the proximity of specific vegetational areas such as riparian zones or marshes. Concerning the other soil types (Tab. 2.B) on which a minority of LBK settlements are lo­cated (L20%), low to medium water saturation also dominates (more than 70% of the cases), thus re­stating the critical value of this characteristic for LBK agriculture. From all these observations it can be inferred that, beyond particular types of soil, LBK farmers first sought and found geo-soil formations that were not water saturated, rather fine-textured and quite deep: in other words, a range of situations favourable to their agricultural and gathering needs and widely re­presented in the territory they explored. These gene­ral characteristics provide a better definition than present-day soil types of LBK farmers’ main criteria for establishing settlements. According to the statis­tical study, we must add to this major trend the com­plementary and less common search for waterlogged soils and their particular vegetation and fauna. Robin Brigand, Jérôme Dubouloz, and Olivier Weller Colonization dynamics of LBK farmers in Europe under geostatistics test These findings are at least partially in line with the tween geo-pedological ensembles, recalling our defi-conclusions of Michael Kempf (2020a), whose re-nition of ‘mixed terroirs’ (section 2.3.1 and Fig. 8). search is not really comparable to that developed Due to the fine resolution of their environmental here, as differences in spatial scale, temporal perspec-data, they ultimately draw the conclusion, unattain­tive, size of the samples examined, resolution of the able at the resolution of our own study, “[...] that basic environmental data and the descriptive system LBK settlers preferred highly productive and light of the geo-pedological units make this comparison soils with easy access to water courses (Sabel rather impractical. However, we note a convergence 1983). These conditions are provided in the dry of our observations with some results of his analy-Loess landscape of the investigated region with a sis, especially the importance of low-medium water distinctly incised, dense drainage network.” saturation in the choice of settlement location. (op. cit. p. 278, underlined by authors). This speci­ fic environmental configuration might be the actual On the other hand, the study by Lorz and Saile source of our observations (section 2.3.2a) concern­(2011) on the genesis of the Chernozems, restricted ing site distribution in the intertwined Chernozems-to southern Lower Saxony, also reaches some con-Luvisols-Cambisols of the loessic region around the clusions compatible with those pro­duced here on a continental scale. A This is the case for the partial dis- Txt. fine to medium-fine connection that can be observed Depth + and depth ++ between site location, loess and the B presence of Chernozems, which has Other textures (very fine, already been proposed by various medium, coarse) researchers cited by these two au- Moderate and shallow depth thors. Thus, their database clearly % of total (N) 82.2 (5455) 79.8 (5200) 17.8 (1184) 20.2 (1317) Low-medium sat. (N) 70.7 (3832) 71.7 (3730) Strong and more sat. (N) 29.8 (1623) 28.3 (1470) 79.5 (941) 20.5 (243) 70.1 (923) 29.9 (394) shows the major importance in LBK Tab. 2. Overall statistics of site distribution by soil water saturation, site location of contact zones be-texture and depth (NB: 1.8% of soils are not characterized by depth). Robin Brigand, Jérôme Dubouloz, and Olivier Weller Harz or in Moravia. As much as the search for a dense and socio-economically supporting neighbour-hood, the need for easy access to water, identified at the meso-regional scale and at higher spatial reso­lution (see also Kon.elová 2005), appears to be a strong complementary determinant of this site distri­bution that is partly disconnected from the best soils. Settlement dynamics Much has already been written on settlement dyna­mics and no attempt is made here to challenge the interpretations and conclusions drawn from previous research. Rather, we will try to show how large-scale geo-spatial statistical analysis can identify and weigh the different factors underlying this space-time phe­nomenon. Chronology of settlements The phasing of the overall extension of the LBK, at least in broad stages, is a necessity for this research project, as it enables continuities and discontinuities across time and space to be considered together. But this ambition is particularly sensitive when the west­ern and central-eastern zones are compared. Despite the predictable discomfort of many specialists, we have nevertheless attempted such a chronological standardization, according to a six-stages (St. 0 to 5) sequence (Appendix 13). The information in the archaeological database can then be used to reconstruct probable settlement dy­namics on a continental scale. This first approach is based on the 2559 locations with LBK occupations that can be dated to a stage (4535 different occupa­tions), excluding both extremities of the sequence (St. 0 and St. 5–6) as the occurrences are very low and/or their definition very specific. Indeed, the ‘5– 6’ and ‘Post CL’ (or ‘6’) stages only appear in the database as marginal information that we did not want to lose at first. The ‘5–6’ stage is related to the Paris Basin where it extends the LBK occupation (early BVSG) and the ‘Post-CL’ stage was taken into consideration only for subsequent occupations on LBK sites: this stage is therefore not really listed as such and cannot be considered here. The total sam­ple thus represents about 39% of the database, but is unevenly distributed among the different regions. In addition to far Eastern Europe, this limitation is likely to be noticeable for Poland, the Czech Repub­lic and Belgium, for example (Appendix 1). Several elaborations of the database are necessary, in the form of simple aggregations by major stage: (1) the total number of occupations; (2) the number of newly created sites, i.e. without immediately pre­ceding occupation; (3) the number of enduring oc­cupations (total number of occupations excluding site creations); and lastly (4) the number of aban­doned sites during stage ‘t’, previously occupied dur­ing stage ‘t -1’. In order to compare settlement trends over similar time periods, these values have been weighted by the approximate duration of the different stages (Appendix 13). Each is split into ‘n’ periods of twen­ty-five years (a human generation), according to the following division (the cut-off effect remains un­avoidable): St. 1 a 7 x 25 y., St. 2 a 6 x 25 y., St. 3 = 2 x 25 years, St. 4 a 4 x 25 y., St. 5 a 4 x 25 y. Lastly, a renewal rate is calculated (ratio of new cre­ations to total numbers) that allows an overall sum­mary of settlement dynamics. The resulting curves (Figs. 14 and 15) reveal singu­larly different dynamics, depending on which of the parameters is considered. Apart from the few settlements of the formative LBK stage (St. 0), the number of creations is similar dur­ing St. 1 and St. 3, around 19% for each stage. How­ever, almost 39% of the dated sites (Fig. 14, red curve: Creation) are established during St. 2. Crea­tions then decline and fall to about 5% during St. 5. At first glance, settlement density is relatively sta­ble between St. 2 and St. 4 (Fig. 14, black curve: To­tal), with the latter appearing as a clear stage of equi­librium, in view of the highest number of enduring sites (Fig. 14, yellow curve: Stability). The process that emerges, already well identified, is one of an ex­panding pioneer front that was particularly mark­ed for stages 1 and 2. However, these trends as perceived through the numbers of occupations per 25-year period offers a different reading (Fig. 15). The curves highlight the St. 3 stage, where the density (black curve) is clearly higher here than in the St. 2 and St. 4 stages, despite a significantly decreasing renewal rate (in green, 42.9 vs. 87.6%) and an increase in abandonments (in light blue, 215 vs. 61 per 25 years). New crea­tions and a great stability of previously existing sites (in red and yellow) explain this counter-intuitive result. At the 25-year generation rate, St. 3 probably represents, at the mid 52nd century BCE, the climax of the expansion process, before a pause at St. 4 which remains at a high level of settlement density. Colonization dynamics of LBK farmers in Europe under geostatistics test At St. 5, all factors considered except ‘abandonment’ reach unprecedentedly low values and indicate the exhaustion of the LBK. As stated earlier in this chapter, this overall reading of the chronological processes is obviously related to the state of the basic information in the database and its representativeness. It is therefore still fragile in terms of detail and should be considered with a margin of error that cannot be easily assessed. Geocentres and standard deviation ellipses First, a method is applied to assess the overall dis­persion of sites and analyse the similarities at each chronological stage: the measurement of centroids (barycentres or geocentres) and deviation ellipses. This approach has long been used in spatial geosta­tistics (Lefever 1926; Pumain, Saint-Julien 2001) and in spatial archaeology (Hodder, Orton 1976; Gauthier 2004; Poirier 2010; Nuninger et al. 2012). It enables the comparison of distributions based on a number of indicators: (i) The geocentre corresponds to the average coordinates of the dispersion by period and the distance between each of these centroids is a way to as­sess the rates of population movement (Tab. 3). (ii) The deviation ellipse en­compasses sites whose loca­tion shifts by no more than one standard deviation from the centroid and therefore al­lows a measurement of dis­persion at each stage of the LBK culture (Fig. 16). For the reasons mentioned in the con­clusions of section I.3, the eastern expansion in Moldova and Romania has not been in­cluded in these analyses. The centroid of the LBK for­mative stage (St. 0, <5500 BCE), located on the shores of Lake Balaton, involves only a very small set of settle­ments. The orientation of the deviation ellipse (in white) shows the dispersion gradi­ent along a SE-NW axis. Then, a shift in polarities is very marked in the next stage St. 1: the roughly 350km distance between the two geocentres (Tab. 3 and Fig. 16) undeniably evokes territorial expansion during this founding stage of the colonization of Central Europe (in yellow). In­deed, between c. 5500 and c. 5325 BCE settlement extended from Transdanubia to the Rhine (nearly 900km of maximum extension) and the upper Vis­tula valley, then towards the end of this stage Kuya-via and northwest Ukraine (nearly 800km). This set­tlement expansion, with standard deviation ellipses increasingly orientated east-west, remains at least until St. 3 (c. 5175–5125 BCE, in blue) very close to that of St. 2 (in purple), with a small extension to the west (lower Rhine/Meuse region, the Moselle and eastern Champagne). The westward expansion restarts with St. 4 (c. 5125– 5025 BCE) during which further parts of the Seine Basin and the Belgian Hainaut are colonized. Never­theless, the geocentre shifts less than a hundred ki- Robin Brigand, Jérôme Dubouloz, and Olivier Weller Central Europe. The situation changes radically in St. 5, at the very end of the 6th millennium and the beginning of the 5th, with the weakening of the Transdanubian and cen-tral-eastern LBK foci. This is clearly underlined by a 308km shift of the barycentre (Tab. 3) and by the change in orientation of the deviation ellipse to­wards a NE-SW axis. It can be assumed that a rather similar process would be observed in Eastern Europe when the dating is better established there, and it will be interesting to compare the rates and average distances of progression there with those measured for the western expansion. Density and transition maps In order to go beyond the point distribution of sites, the ‘kernel density estimation’ tool (KDE) analyses the data in a continuous surface, reflecting the inten­sity of the occupation of the space. This method pro­vides an estimate of the density of sites in a moving window and the resulting values include the situa­tion in the neighbourhood. Thus, an area surround­ed by other densely populated areas will have its own density increased. This method has been well known to geographers since the 1980s (Silverman 1978; 1986; Wand, Jones 1995; Zaninetti 2005), and widely used by archaeologists working at site scale (Baxter, Beardah 1997; Beardah 1999). The results depend on two parameters: (i) h, the va­lue of the radius, fundamental since it determines the degree of smoothness of the data that offers the best representativeness, between the point scatter and a non-discriminating mesh (Silverman 1978); (ii) k, the chosen function, here a quadratic function. In our study, a graphical approach inspired by the ArchaeDyn program (Nuninger et al. 2012) and al­ready proven (Brigand, Weller 2018a) is exploited. This involves plotting the maximum values obtained from a series of calculations according to a given in­terval, here 5000m. The inflection point of the curve corresponds to the best window size, which in our situation is 12.5km, a value between 10 and 15km (Fig. 17) corresponding to about half a day’s walk. At a radius of 12.5km, the KDE density value (in km2) is therefore calculated for each stage and com­pared with another relevant indicator: the average observed distance (AOD) (Fig. 18). These values are considered from stages 1 to 5 (formative St. 0 AOD = 51km and KDE = 0). In the graph combining these two parameters, the values of AOD and KDE are res­pectively very high and very low for stages 1 and 5. It is clear that during these two periods the occupa­tion is the least dense. Conversely, the occupation is more concentrated during stages 2 to 4, with the Colonization dynamics of LBK farmers in Europe under geostatistics test average distance being lowest and the maximum density being higher: the values of stage 4 show a slightly higher AOD, reminiscent of the situation highlighted in Figure 14. In order to evaluate the differences between periods following the density curves and maps (Figs. 19–20), an approach based once again on methodological advances from the ArchaeDyn program (Nuninger et al. 2012) is proposed. This method exploits the re­lative difference in density between two periods through the use of the Normalized Differential Ratio (NDR), defined in the same way as the Normalized Difference Vegetation Index, namely: NDR= (T1–T0)/ (T1+T0). This calculation makes it possible to visua­lize the dynamics of settlement between two chrono­logical sequences, and thus to distinguish the newly created sectors from those more stable or in disuse. A precise example of such a protocol is proposed in Appendix 14. This modus operandi was thus implemented on a European scale (excluding Romania, Moldova and a large part of Ukraine), and a synthesis of the propor­tions and areas relative to the created, stable or abandoned sites by stage is presented in an analyti­cal diagram (Fig. 19) and five maps (Fig. 20). . Between St. 0 and St. 1, the newly colonized areas (Figs. 19 and 20.a, red) are the most extensive, since the LBK leaves its initial focus to reach Moravia, Bo­hemia, Saxony/Thuringia and the Rhine valley in the west, Little Poland, Transcarpathian Ukraine and the upper Bug valley in the north and northeast. . Between St. 1 and St. 2 spatial expansion remains substantial both outside and inside the previous oc­cupied areas (Figs. 19 and 20.b, red or orange). The main features are: expansion or densification of set­tlement in the Rhine-Meuse-Moselle valleys (Fig. 20.b, red-orange), colonization of the hilly areas of the western Carpathians (Fig. 20b, red), and a decline in southern Transdanubia and the Elbe-Saale area (Fig. 20.b, light blue). In Saxony, Saxony-Anhalt and Thu-ringia, for example, the small number of very ‘reli­able’ and well-dated sites (Figs. 1 and 2) relativizes the significance of the results proposed here, espe­cially for the abandonments. . Between St. 2 and St. 3, the dynamics of coloniza­tion decrease and stability dominates (Figs. 19 and 20.c, yellow), in spite of some mobility of settlement in the occupied spaces (orange) and limited expan­sion out of the already colonized zones (red), such as the southern foothills of the Western Beskids up to the confluence of the Vah and the Danube (Slova­kia). This is the beginning of a stability that will be strengthened with the passage to St. 4 (Figs. 19 and 20.d), especially north of Lake Balaton, during which the internal and external ‘conquests’ are weak. . The dynamics between St. 4 and St. 5 (Figs. 19 and 20.e, light blue) have already been observed in sections 3.1 and 3.2 (Figs. 14–16) and indicate a massive decline in Central Europe and a refocusing on western Germany, Benelux and the Paris Basin, whose centre is now occupied. Brief synthesis and perspectives Although many meso- to macro-regional stu­dies have already been carried out on both the settlement dynamics and land-use pat­terns of the LBK, our study is the first to tackle these questions on a continental scale using a large geo-referenced database and GIS. As archaeologists, we are well aware that understanding the outputs of such an ana­lysis depends largely on analytical studies at the lower scales. Due to the higher spa­tial and temporal resolution of the available Robin Brigand, Jérôme Dubouloz, and Olivier Weller information, these more localized stu­dies can often produce a better de­scription and understanding of par­ticular situations. Fortunately, much of this kind of research has already been done so that it is possible to ad­just some of the results of our broa­der approach using these more spe­cific cases and observations (among many others: Bakels 1978; Bogaard 2004; Bogaard et al. 2016; Ebers­bach, Schade 2005; Koncelova 2005; Kreuz 1990; 2008; Lenneis 2001; 2003; 2010; Lenneis, Pieler 2016; Pieler 2010; Saile 2009; 2010; 2020; Tóth et al. 2011; but see also Kempf 2020b). It is also no­table that our results are essentially an estimation of the processes underlying the current distribution maps, as were most of the synthetic narratives al­ready proposed. It is to be hoped that the overall ap­proach and methodological tools applied here will provide a more comprehensive basis for drawing conclusions. One of the main limitations of this ap­proach, which is difficult to quantify globally, lies in the importance of the natural and anthropic phe­nomena which, since the 6th millennium BCE, might have masked and caused the disappearance of nu­merous settlements: direct erosion and colluvial co­ver and burial. These types of processes can appear, when analysed at a local scale, to be of considerable importance and therefore constitute one of the most problematic biases of the available spatial distribu­tions. The overall assessment of this impact on the archaeological record remains to be made through further detailed local research. Nevertheless, at this stage of our general knowledge of the LBK pheno­menon, the analysis proposed in these pages can re­present a satisfactory framework for future improve­ments. Three main topics were thus addressed for testing the overall capacity of archaeological research, as reflected in our database, to deal with some impor­tant issues: (1) the homogeneity, quality and limita­tions of the archaeological data; (2) the definition of agro-ecological determinants, if any, in the expan­sion process of this culture; (3) the highlighting of settlement dynamics that will have affected a signi­ficant part of temperate Europe between mid 6th and early 5th millennia BCE. . The evaluation of the archaeological documenta­tion gathered here delivers a rather complex verdict regarding its potential for description on a continen­tal scale. The easternmost part of LBK Europe can only partially answer the questions raised and can­not be fully incorporated into an overall synthesis, mainly because of the lack of chronological preci­sion on the occupations (see sections Archaeological descriptors and Quality of the BD_LBK and overall heuristic validity). The other settlement areas of the BD_LBK seem to be suitable for joint analysis; some of these are less well-documented, but still have suf­ficiently detailed samples in terms of number and space-time distribution. In this respect, the effect of the information deficit in the Neckar Basin on our results (see section The archaeological data) can be briefly mentioned here. First it can be as­sumed that the analyses are not real­ly affected in their chronological di­mension, since most of the non-se­lected sites are ‘Survey-’ or ‘Collect’­sites probably lacking reliable chro­nological information. Conversely, there is a concern that the regional statistics on agro-ecological determi­nants may be affected. Yet on a con­tinental scale this impact can be as­sessed as marginal, without signifi­ Colonization dynamics of LBK farmers in Europe under geostatistics test cant impacts on the robustness of the overall ob­servations (see below). . The analysis proposed in this article does not seek to describe the exact variability of LBK settle­ment agro-ecology in each region, but rather the do­minant trends that have ruled it across Europe. And the general conclusions of our study appear to be supported by some higher resolution research: the importance of ecological transition zones (ecotones), the only relative relevance of the presence of loess, and the targeting of favourable soil characteristics, with no exclusive focus on their exact soil type. By examining the LBK settlement process in terms of the current geo-pedology, we can show that a major characteristic was the absence of exclusive preferen­ces for particular formations. Thus, the LBK system was adaptive to the changing conditions available across temperate Europe, proof of a farming culture honed by centuries of practice. Here again, evalua- Robin Brigand, Jérôme Dubouloz, and Olivier Weller tion of the documentary bias in the Neckar region suggests that our rejection of many sites there, equi­valent to about 5% of the current database, has only slightly weakened the rate of loessic substrates and of Luvisols and Cambisols in LBK’s settlement sys­tem. This situation does not therefore profoundly affect the validity of our general observations, nor their relevance in describing the overall expansion process. Better still, their presence in the statistics would support the highly variable and adaptive cha­racter of the LBK agro-sylvo-pastoral system as a whole. It therefore appears that the current database, despite its limitations noted in the first part (section 1), is indeed a suitable sample for a reliable analysis of the geo-pedological components of the LBK expan­sion. . In the light of different geo-statistic approaches, the LBK expansion dynamics ultimately appear as a combination of two major quantitative trends distri­buted over time. From stage 1 to stage 4, there is a rapid and steady decrease in the creation of sites be­yond the previous colonization front, at the same time as an ever-stronger growth in the stabilization of places already occupied. The second trend is the fairly regular process of site abandonment, partially balanced by a roughly stabilized rate of creation within already occupied areas. Then, with stage 5, the number of site abandonments soars while the other settlement dynamics further decline, clearly demonstrating the weakening and increasing scar­city of the LBK cultural expression on a continental scale. The analyses also showed that stage 3 of our chronological framework (centred around 5150 BCE) may be considered as the climax of LBK ex­pansion. This composite stage, formed, as noted in section 3.1, by compilation of an identified stage 3 in the west and a reconstructed middle/late transi­tional stage in the east, determines a state of equili­brium in stage 4 (and in the late LBK/post-Noten­kopf in general). Then in stage 5 (or the latest LBK) a gradual dissolution occurs, as if the driving forces of expansion (demography, networks of socio-cultu­ral exchange, sense of common identity) had after four centuries reached the limit of their capacity to maintain basic unity at such a large spatial scale. Thus, in our view, this overall trajectory is firstly and mainly indicative of a socio-cultural process. It might be tempting to relate this trajectory to the ‘boom and bust’ pattern identified by Stephen Shen­nan et al. (2013) and Adrian Timpson et al. (2014) in their analysis of 14C data covering the whole Eu­ropean Neolithic. Yet, for several major reasons, it seems clear that our LBK trajectory cannot be strict­ly linked to the pattern proposed by Shennan and colleagues, and in fact does not really match it at all. The authors of the population model under conside­ration themselves state that their results can only have a good statistical significance on a multi-millen­nial scale, with a time step of more than two centu­ries (Timpson et al. 2014.550). However, the time scale analysed in the two cases is very different, with the duration of the LBK representing a small fraction (10–13%) of the span envisaged for Neolithic West­ern Europe, and occurring very close to its onset. Thus, the LBK trajectory alone (400–500 years, de­pending on the regions considered) cannot corrobo­rate a model built according to other methodologi­cal requirements and with a much more general aim. In fact, it would not be logical to seek in the LBK tra­jectory highlighted by our own work confirmation of a model developed for a process that was taking place over about two millennia. At the same time, it can be observed that the 14C data of the eight European regions involved in the LBK expansion do not clearly illustrate the pattern. Indeed, on closer inspection of the curves proposed by Shennan et al. (2013) and Timpson et al. (2014) in their respective Figure 3, we can see considerable heterogeneity of the results and three different pat­terns. (i) The Paris Basin, Rhineland-Hesse and Mo-ravia-Austria show a statistically valid start of a pro­nounced ‘boom’ approximately concomitant with the onset of LBK. However, this ‘boom’ lasts for about one millennium, i.e. twice the duration of the LBK itself. (ii) South Germany and Kujavia show a short and slight LBK-related ‘boom’, then two bigger ones many centuries later. (iii) The third pattern is rep­resented by central Germany, Little Poland and Bo­hemia, where there is no LBK-related ‘boom’ but a few short ‘booms’ occurring 1500 to 2000 years after the end of the LBK itself. Curiously, these three re­gions have some of the most numerous LBK occupa­tions known, so that these results appear very coun­ter-intuitive. As for the ‘bust’, it appears statistically very significant in only three out of the eight regions, yet around 1700 to 2200 years after the end of an initial LBK-related ‘boom’ (Paris Basin, Rhineland-Hesse, South Germany). A fourth region (Moravia) displays a very short and slight ‘bust’, occurring at the same pace of time. This regressive phenomenon is therefore not directly related to the disappearance of the LBK. Moreover, in central Germany, a ‘bust’ seems to occur about 750 years after the end of a flourishing LBK that did not produce the signal of a ‘boom’, though logically predictable. Lastly, in two Colonization dynamics of LBK farmers in Europe under geostatistics test regions where the strong settlement dynamics of the LBK are very apparent (Bohemia and Little Poland), we observe the absence not only of any ‘boom’, but also of any ‘bust’ over the four millennia considered. In short, the overall heterogeneity of these results, where one would expect a good coherence, does not provide firm support for the population model de­fended by Shennan and colleagues and in our view raises doubts about the real adequacy of the proxy used in their study. Such an approach using a cate­gory of information that has become an expensive standard in scientific protocols is undoubtedly legi­timate for defining new hypotheses, but it is up to researchers to confront them fully with the basic data, which should themselves be critically evaluat­ed. The least that can be concluded, from the point of view of our work, is that the trajectory identified by our approach to the LBK cannot provide solid evi­dence for the population model promoted by Shen­nan et al., at both methodological and factual levels. In this respect, the final LBK dissolution should pro­bably not be directly interpreted as an overall pop­ulation collapse without an assessment at the same time of post-LBK cultural developments. Indeed, the trajectory identified here on the basis of the distri­bution of LBK sites alone cannot say much about the subsequent situations in the same regions (Stich­bandkeramik (Stroke-ornamented ware), Lengyel, Hinkelstein/Grossgartach, BVSG). It is clear that the spatiality of these cultural developments must also be analysed in detail, as closely as possible to what is proposed here for the LBK. This is one of the rea­sons why it seems to us that the attempt to create demographic scenarios on the basis of radiocarbon dates alone is probably only a stopgap measure of very relative reliability. Indeed, such an approach cannot easily escape the heterogeneity of research focuses and their intensity, nor the varying archae­ological visibility of different settlement modes; it cannot therefore guarantee a comparable represen­tation of the different chrono-cultural aspects at stake. Under these conditions, statistical corrections, however rational and sophisticated they may be, can­not in our opinion replace basic archaeological in­formation of the same reliability as that provided by the LBK. Since we have found that even the LBK data may be deficient for some specific analyses and interpretations, the results of radiocarbon analysis alone, disconnected from more or less uniform knowledge of chrono-spatial dynamics, are likely to be somewhat illusory with regard to long-term popu­lation dynamics (see Crombé, Robinson 2014). For reasons of limited space here, we are develop­ing elsewhere a complementary study- about the agro- ecological factors and their related settlement dynamics, through a spatial-temporal modelling (Bri­gand et al., in prep.). In a third contribution, spe­cial attention will be paid to the distribution of salt resources, the importance of which for Neolithic communities is now well identified (Munteanu et al. 2007; Weller, Dumitroaia 2005), especially for the relatively more complex societies of the 5th and 4th millennia (Weller 2015; Brigand, Weller 2018b). While some salt production centres have been stu­died in Eastern Europe (Sordoillet et al. 2018), we know much less about the early LBK farmers in Western and Central Europe. And as already raised in the literature for several geographical areas (Bánffy 2013; Geschwendt 1958; Pyzel 2016; Saile 2012; Weller 2015), to what extent did salt resour­ces contribute to the rapid LBK expansion and the development of its main settlement areas through­out Europe? As we can see, there is still a lot of re­search to be done on this famous LBK culture, both on a local and European scale, in order to under­stand this unprecedented phenomenon. ACKNOWLEDGEMENTS This work has received financial support from the LabEx “DynamiTe. Territorial and Spatial Dyna­mics” (ANR-11-LABX-0046), as part of the French “In­vestissements d’Avenir” program, and from the “Neo-Dyn Project. Dynamiques migratoires des premiers paysans en Europe: Le rôle clé des ressources sali­feres dans la néolithisation” (University Panthéon-Sorbonne, Paris). Robin Brigand, Jérôme Dubouloz, and Olivier Weller References (* for Appendix 13) Bakels C. C. 1978. Four Linearbandkeramik Settlements and their Environment. A Paleoecological Study of Sit-tard, Stein, Elsloo and Hienheim. Analecta Praehistorica Leidensia II. Institute of Prehistory. Leiden University. Lei­den University Press. Leiden. https://www. sidestone.com/ openaccess/9789060214275.pdf Bánffy E., Oross K. 2010. The earliest and earlier phase of the LBK in Transdanubia. In D. Gronenborn, J. Petrasch (eds.), Die Neolithisierung Mitteleuropas. Internationa­le Tagung, Mainz 24. bis 26. Juni 2005. RGZM – Tagun-gen 4. Römisch-Germanischen Zentralmuseum. Forschungs­institut für Vor- und Frühgeschichte. Verlag des Römisch-Germanischen Zentralmuseums. Mainz: 255–272. Bánffy E. 2013. Tracing 6th–5th millennium BC salt explo­itation in the Carpathian Basin. In A. F. Harding, V. Kav­ruk (eds.), Explorations in Salt Archaeology in the Car­pathian Basin. Fieldwork in central and eastern Eu­rope, 2002–2012. Archaeolingua. Budapest: 201–207. Baxter M. J., Beardah C. C. 1997. Some Archaeological Ap­plications of Kernel Density Estimates. Journal of Archa­eological Science 24: 347–354. https://doi.org/10.1006/jasc.1996.0119 Beardah C. 1999. Uses of Multivariate Kernel Density Esti­mates in Archaeology. In L. Dingwall, S. Exon, C. F. Gaf­fney, S. Laflin, and M. van Leusen (eds.), CAA97. Compu­ter Applications and Quantitative Methods in Archaeo­logy. Archaeology in the Age of the Internet. Proceedings of the 25th Anniversary Conference. Archeopress. Oxford (CD-ROM). *Behrens H. 1973. Die Jungsteinzeit im Mittelelbe-Sa­ale-Gebiet. Veröffentlichungen des Landes- museums für Vorgeschichte in Halle Band 27. VEB Deutscher Verlag der Wissenschaften. Berlin. Bentley R., Bickle P., Fibiger L., +11 authors, and Whittle A. 2012. Community differentiation and kinship among Europe’s first farmers. Proceedings of the National Aca­demy of Sciences 109(24): 9326–9330. https://doi.org/10.1073/pnas.1113710109 Benzécri J.-P. 1973. L’analyse des données, Vol. 1 and 2. Dunod. Paris. *Blouet V., Klag T., Petitdidier M.-P., Decker E., Constantin C., and Ilett M. 2013. Le Néolithique ancien en Lorraine, Vol. 1. Étude typochronologique de la céramique. Socié­té préhistorique française. Mémoire 55. Paris. Bocquet-Appel† J.-P., Dubouloz J., Moussa R., Berger J.-F. (eds.), in preparation. Modelling the socio-natural tra­jectory of the LBK farmers. Bocquet-Appel J.-P., Moussa R., Dubouloz J. 2015. Multi-agent Modelling of the Neolithic LBK. In F. Giligny, F. Djin­djian, L. Costa, P. Moscati, and S. Robert (eds.), CAA2014 21st century Archeaology concepts, methods and tools. Proceedings of the 42nd Annual Conference on Compu­ter Applications and Quantitative Methods in Archaeo­logy. Archeopress Archaeology. Oxford: 595–612. Bogaard A. 2004. Neolithic Farming in Central Europe. An Archaeobotanical Study of Crop Husbandry Practi­ces. Routledge. London. Bogaard A., Arbogast R.-M., Ebersbach R., +5 authors, and Krause R. 2016. The Bandkeramik settlement of Vaihin-gen an der Enz, Kreis Ludwigsburg (Baden-Württemberg): an integrated perspective on land use, economy and diet. Germania 94: 1–60. Bostyn F., Charraud F., Denis S. 2019. Variabilités techni­ques, évolutions et aires d'influence des centres de pro­ductions laminaires au sein de la culture de Blicqut/Vil-leneuve-Saint-Germain. In C. Montoya, J.-P. Fagnart, and J. L. Locht (eds.), Préhistoire de l’Europe du Nord-Ouest. Mobilités, climats et identités culturelles. XXVIIIe con-gres préhistorique de France, Amiens 30 mai-4 juin 2016. Société Préhistorique Française. Paris: 43–56. Brandt G., Haak W., Adler C. J., +14 authors, and The Ge-nographic Consortium 2013. Ancient DNA Reveals Key Stages in the Formation of Central European Mitochon­drial Genetic Diversity. Science 342(6155): 257–261. https://doi.org/10.1126/science.1241844 Brandt, G., Szécsényi-Nagy A., Christina Roth C., Werner Alt K., and Haak W. 2015. Human paleogenetics of Eu­rope. The known knowns and the known unknowns. Journal of Human Evolution 79: 73–92. http://dx.doi.org/10.1016/j.jhevol.2014.06.017 Brigand R., Dubouloz J., Weller O. in preparation. Du pro-jet ObReSoc a NéoDyn: une modélisation de la diffusion du Néolithique rubané (LBK) a travers l’Europe. 42e Ren­contres Internationales d’Archéologie et d’Histoire de Nice Côte d’Azur. Dynamique des peuplements, des ter­ritoires et des paysages: bilan et perspective en archéo­logie spatiale. Hommage a Jean-Luc Fiches. 12 – 14 oc­tobre 2022, Nice, France. Brigand R., Weller O. 2018a. Kernel density estimation and transition maps of Moldavian Neolithic and Eneoli­thic settlements. Data in Brief 17: 452–458. doi.org/10. 1016/j.dib.2018.01.051 2018b. Neo-Eneolithic settlement pattern and salt ex­ploitation in Romanian Moldavia. Journal of Archaeo­logical Science. Reports 17: 68–78. doi.org/10.1016/j.jasrep.2017.10.032 Colonization dynamics of LBK farmers in Europe under geostatistics test Brunel S., Bennett E. A., Cardin L., +37 authors, and Pru-vost M. 2020. Ancient genomes from present-day France unveil 7,000 years of its demographic history. Proceed­ings of the National Academy of Sciences 117(23): 12791– 12798. https://doi.org/10.1073/pnas.1918034117 Childe V. G. 1929. The Danube in Prehistory. Clarendon Press. Oxford. Crombé P., Robinson E. 2014. 14C dates as demographic proxies in Neolithisation models of northwestern Europe: a critical assessment using Belgium and northeast France as a case-study. Journal of Archaeological Science 552: 558–566. http://dx.doi.org/10.1016/ j.jas.2014.02.001 Czekaj-Zastawny A. 2009. The First Neolithic Sites in Cen-tral/South-East European Transect. Volume V. Settlement of the Linear Pottery Culture in Southeastern Poland. BAR International Series 2049. Archaeopress. Oxford. *2013. The Structure of Linear Pottery Culture Settle­ment in South-Eastern Poland. In S. Kadrow, P. W³odar­czak (eds.), Environment and subsistence – forty years after Janusz Kruk’s “Settlement studies”. Studien zur Archäologie in Ostmitteleuropa/Studia nad Pradziejami Europy Œrodkowej, Vol.11. Institute of Archaeology Rzeszów University. Verlag Dr. Rudolf Habelt GmbH. Rzeszów, Bonn: 69–84. http://books.arche ologia.rzes-zow.pl/en/71-environment-and-subsistence-forty-ye ars-after-janusz-kruk-s-settlement-studies-.html *.i.máø Z. 1998. Nástin relativní chronologie lineární ke­ramiky na Moravì. Acta Musei Moraviae 83(1/2): 105–139. *Denaire A., Lefranc P., Wahl J., +6 authors, and Whittle A. 2017. The Cultural Project: Formal Chronological Mo-delling of the Early and Middle Neolithic Sequence in Lo­wer Alsace. Journal of Archaeological Method and Theory 24: 1072–1149. https://doi.org/10.1007/s10816-016-9307-x *Dohrn-Ihmig M. 1979. Bandkeramik an Mittel- und Nie­der-rhein. In Beiträge zur Urgeschichte des Rheinlan­des III. Rheinische Ausgrabungen 19. Rheinisches Lan­desmuseum Bonn. Rheinland-Verlag. Köln-Bonn: 191–362. Dubouloz J., Bostyn F., Chartier M., Cottiaux R., and Le Bol­loc’h M. 2005. La recherche archéologique sur le Néolithi­que en Picardie. Revue Archéologique de Picardie 3–4: 63–98. https://www.persee.fr/doc/pica_0752-5656_200 5_num_3_1_2458 Dubouloz J. Gauvry Y. in preparation. An archaeological referential of the settlement system and the domestic unit. In J.-P. Bocquet-Appel†, J. Dubouloz, R. Moussa and J.-F. Berger (eds.), Modelling the socio-natural trajecto­ry of the LBK farmers. Dubouloz J., Moussa R., Bocquet-Appel J.-P. 2017. Modéli­sation, simulation et scénarios d’expérimentation. La co-lonisation LBK de l'Europe tempérée (5550–4950 av. n. e.). In L. Manolakakis, N. Schlanger, and A. Coudart (eds.), European Archaeology. Identities & Migrations. Sidestone Press. Leiden: 315–337. Ebersbach, R., Schade C. 2005. Modelle zur Intensität der bandkeramischen Landnutzung am Beispiel der Altsiedel­landschaft Mörlener Bucht/Wetterau. In J Lüning, C. Frir­dich, and A. Zimmermann (eds.), Die Bandkeramik im 21. Jahrhundert. Symposium in der Abtei Brauweiler bei Köln vom 16.9–19.9.2002. Internationale Archäologie Arbeitsgemeinschaft, Symposium, Tagung, Kongress. Band 7. Marie Leidorf. Rahden/Westf.: 259–274. Escofier B., Pages J., 1990. Analyses factorielles simples et multiples: objectifs, méthodes, interprétation. Dunod. Paris. Gandini C., Favory F., and Nuninger L. (eds.) 2012. Settle­ment patterns, production and trades from the Neoli­thic to the the Middle Ages. ARCHAEDYN Seven Millen­nia of Territorial Dynamics. Final Conference Univer­sity of Burgundy. Dijon, 23–25 June 2008. BAR Interna­tional Series 2370. Archeopress. Oxford. Gauthier E. 2004. L’évolution de la consommation du mé­tal a L’Age du Bronze, en France orientale et en Transda­nubie. Histoire et Mesure 19(3–4): 345–376. https://doi.org/10.4000/histoiremesure.775 Gauthier E., Georges-Leroy M., Poirier N., and Weller O. 2022 (in press). ArchaeDyn. Dynamiques spatiales des territoires de la Préhistoire au Moyen Âge. Vol. 1. PUFC. Cahiers de la MSHE. Besançon. https://hal.archives-ouvertes.fr/hal-02127300 Geschwendt F. 1958. Die Solquellen von Sülbeck, Kr. Ein-beck, in Urzeit und Mittelalter. Die Kunde 9: 53–67. Haase D., Fink J., Haase G., +4 authors, and Jäger K.-D 2007. Loess in Europe. – its spatial distribution based on a European Loess Map, scale 1:2,500,000. Quaternary Sci­ence Reviews 26(9–10): 1301–1312. Hodder I., Orton C. 1976. Spatial analysis in archaeolo­gy. Cambridge University Press. Cambridge. *Kaufmann D. 1987. Linien- und Stichbandkeramik im Eibe-Saale-Gebiet. In T. Wiœlañski (ed.), Neolit i pocz¹tki epoki br¹zu na ziemi Chelmiñskiej. Materia³y z miêdzy­narodowego sympozjum, Toruñ, 11–13 XI 1986. Uni-wersytet Miko³aja Kopernika. Toruñ: 275–301. Kempf M. 2020a. Neolithic land-use, landscape develop­ment, and environmental dynamics in the Carpathian Ba­ Robin Brigand, Jérôme Dubouloz, and Olivier Weller sin. Journal of Archaeological Science. Reports 34: 102637. https://doi.org/10.1016/j.jasrep.2020.102637 2020b. Fables of the past landscape (re-)constructions and the bias in the data. Documenta Praehistorica 47: 476–492. https://doi.org/10.4312/dp.47.27 Kiosak D. 2017. Kamyane-Zavallia, the Easternmost Line­ar Pottery Culture Settlement Ever Excavated. Sprawo­zdania Archeologiczne 69: 253–269. *Kneipp J. 1998. Bandkeramik zwischen Rhein, Weser und Main. Studien zu Stil und Chronologie der Kera­mik. Universitätsforschungen zur prähistorischen Archäo­logie 47. Habelt. Bonn. Kon.elová M. 2005. Struktura osídlení lidu s lineární ke­ramikou ve východních .echách – Linear Pottery settle­ment structure in East Bohemia. Archeologicke rozhledy LVII: 651–706. Kreuz A. 1990. Die ersten Bauern Mitteleuropas – eine archäobotanische Untersuchung zu Umwelt und Land-wirtschaft der Ältesten Bandkeramik. Analecta Praehi­storica Leidensia 23. Publications of the Institute of Pre­history. University of Leiden. Leiden. https://www.sidest one.com/openaccess/9789073368033.pdf 2008. Closed forest or open woodland as natural vege­tation in the sur- roundings of Linearbandkeramik set­tlements? Vegetation History and Archaeobotanyt 17: 51–64. https://doi.org/10.1007/s00334-007-0110-1 Kreuz A., Marinova E., Schäfer E., and Wiethold J. 2005. A comparison of early Neolithic crop and weed assemblages from the Linearbandkeramik and the Bulgarian Neolithic cultures: Differences and similarities. Vegetation History and Archaeobotany 14: 237–258. https://doi.org/10.1007/s00334-005-0080-0 Kulczycka-Leciejewiczowa A. 2000. Early linear pottery communities to the north of the Sudeten and Carpathian Mountains. Recent researches. In I. Pavlù (ed.), In Memo­riam Jan Rulf. Památky archeologické. Supplementum 13. Archeologický ústav Akademie Vìd .eské republiky. Praha: 196–204. Lefever D. 1926. Measuring geographic concentration by means of the standart deviational ellipse. American Jour­nal of Sociology 32: 89–94. https://doi.org/10.1086/214027 *Lefranc P. 2007. La céramique du Rubané en Alsace. Contribution a l’étude des groupes régionaux du Néoli­thique ancien dans la plaine du Rhin supérieur. Mono-graphies d’Archéologie du Grand-Est – Rhin, Meuse, Mosel­le, 2. Université Marc-Bloch. Strasbourg. Lenneis E. 2001. The beginning of the Neolithic in Austria – a report about recent and current investigations. Docu­menta Praehistorica 28: 99–116. https://doi.org/10.4312/dp.28.7 2003. Die frühneolithische Besiedlung Österreichs im Bezug zur natürlichen Umwelt. In J. Eckert, U. Eisenha­uer, and A. Zimmermann (eds.), Archäologische Pers­pektiven. Analysen und Interpretationen im Wandel. Festschrift für Jens Lüning zum 65. Geburtstag. Inter­nationale Archäologie. Studia honoraria 20. Verlag Marie Leidorf. Rahden: 279–292. 2010. Zur Chronologie der älteren Linearbandkeramik in Österreich. In J. .uteková, P. Pavúk, P. Kalábková, and B. Kovár (eds.), PANTA RHEI. Studies on the Chronology and Cultural Developmentof South-East­ern and Central Europe in Earlier Prehistory Present­ed to Juraj Pavúk on the Occasionof his 75th Birth­day. Studia Aechaeologica et Mediaevalia XI. Comme­nius University. Bratislava: 189–200. Lenneis E., Pieler F. 2016. Relative chronologie der Li-nearbandkeramik in Österreich. In J. Kováník (ed.), Cen­tenary of Jaroslav Palliardi’s Neolithic and Aeneolithic relative chronology (1914–2014). University of Hradec Králové. Philosophical Faculty. Hradec Králové: 45–66. Lichardus J., Lichardus-Itten M., and Bailloud G. 1985. La Protohistoire de l’Europe. Nouvelle Clio 1. Presses Uni-versitaires de France. Paris. *Lindig S. 2002. Das Früh- und Mittelneolithikum im Neckarmündungsgebiet. Universitätsforschungen zur prä­historischen Archäologie. Verlag R. Habelt. Bonn. Lipson M., Szécsényi-Nagy A., Mallick S, +53 authors, and Reich D. 2017. Parallel palaeogenomic transects reveal complex genetic history of early European farmers. Na­ture 551: 368–372. https://doi.org/10.1038/nature24476 Lorz C., Saile T. 2011. Anthropogenesis of chernozems in Germany? A critical review. Quaternary International 243: 273–279. https://doi.org/10.1016/j.quaint.2010.11.022 Lüning, J. 2005. Bandkeramische Hofplätze und absolute Chronologie des Bandkeramik. In J. Lüning, Ch. Frirdich, and A. Zimmermann (eds.), Die Bandkeramik im 21. Jahrhundert. Symposium in der Abtei Brauweiler bei Köln vom 16. 9.–19. 9. 2002. Internationale Archäologie. Arbeitsgemeinschaft, Symposium, Tagung, Kongress. Bd. 7. Verlag Marie Leidorf. Rahden: 49–74. Mathieson I., Lazaridis I., Rohland N., +34 authors, and Reich D. 2015. Genome-wide patterns of selection in 230 ancient Eurasians. Nature 528: 499–503. https://doi.org/10.1038/nature16152 Colonization dynamics of LBK farmers in Europe under geostatistics test *Meier-Arendt W. 1966. Die bandkeramische Kultur im Untermaingebiet. Veröffentlichungen des Amtes für Bo-dendenkmalpflege im Regierungsbezirk Darmstadt, Hessen. Vol 3. Verein von Altertumsfreunden im Regie­rungsbezirk Darmstadt. Darmstadt. *1972. Die ältere und mittlere Linienbandkeramik im westlichen Mitteleuropa. Ein Überblick. In H. Schwabe­dissen (ed.), Die Anfänge des Neolithikums vom Orient bis Nordeuropa. Fundamenta A3, Va. Böhlau Verlag. Köln: 66–76. Munteanu R., Garvan D., Nicola D., Preoteasa C., and Du-mitroaia G. 2007. Cucuieti-Slatina Veche (Romania). Pre­historic exploitation of a salt ressource. In D. Monah, G. Dumitroaia, O. Weller, and J. Chapman (eds.), L’exploita­tion du sel a travers le temps. Biblioteca Memoria Anti-quitatis. Piatra Neamt: 57–70. Nuninger L., Saligny L., Ostir K., +5 authors, and Tolle F. 2012. Models and tools for territorial dynamic studies. In C. Gandini, F. Favory, and L. Nuninger (eds.), Settle­ment Patterns, Production and Trades from Neolithic to Middle Ages, Archaedyn, 7 millennia of territorial dynamics. Final Conference. University of Burgundy. Archeopress. Oxford: 23–37. Oross K., Bánffy E. 2009. Three successive waves of Neo­lithisation: LBK development in Transdanubia. Documen­ta Praehistorica 36: 175–189. https://doi.org/10.4312/dp.36.11 Pavlù I. 2000. Life on a Neolithic Site. Bylany. Situatio­nal Analysis of Artefacts. Institute of Archaeology CAS. Praha. http://bylany.com/pdf/BYLANY1_2000LIFE_EN.pdf *Pavúk J. 2004. Early Linear Pottery Culture in Slovakia and the Neolithisation of Central Europe. In A. Lukes, M. Zvelebil (eds.), LBK Dialogues. Studies in the Formation of the Linear Pottery Culture. BAR International Series 1304. Archaeopress. Oxford: 71–82. *2005. Typologische Geschichte der Linearbandkera­mik. In J. Lüning, Ch. Fridrich, and A. Zimmermann (eds.), Die Bandkeramik im 21. Jahrhundert. Sym­posium in der Abtei Brauweiler bei Köln vom 16.9.– 19.9.2002. Internationale Archäologie. Arbeitsgemein­schaft, Symposium, Tagung, Kongress. Bd. 7. Verlag Ma­rie Leidorf. Rahden: 17–39. *Pechtl J., Land A. 2019. Tree rings as a proxy for seaso­nal precipitation variability and Early Neolithic settle­ment dynamics in Bavaria, Germany. PLoS ONE 14(1): e0210438. https://doi. org/10.1371/journal.pone.0210438 Pieler F. 2010. Die Bandkeramik im Horner Becken (Niederösterreich). Studien zur Struktur einer frühne­ olithischen Siedlungskammer. Universitätsforschungen zur prähistorischen Archäologie 182. Verlag Dr. Rudolf Habelt. Bonn. Poirier N. 2010. Un espace rural a la loupe: paysage, peuplement et territoires en Berry de la Préhistoire a nos jours. Presses Universitaires François-Rabelais. Tours. Price T. D., Bentley R. A., Lüning J., Gronenborn D., and Wahl J. 2001. Prehistoric human migration in the Linear-bandkeramik of Central Europe. Antiquity 75(289): 593– 603. https://doi.org/10.1017/S0003598X00088827 Price T. D., Burton J. H., and Bentley R. A. 2002. The cha­racterisation of biologically-available strontium isotopera­tios for investigation of prehistoric migration. Archaeo­metry 44: 117–35. https://doi.org/10.1111/1475-4754.0004 Pumain D., Saint-Julien T. 2001. Les interactions spatia­les. Flux et changements dans l’espace géographique. Armand Colin. Paris. *Pyzel J. 2006. Die Besiedlungsgeschichte der Bandkera­mik in Kujawien. Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz 53: 1–57. *2010. Zofipole/a.kovy/Flomborn. On the problems of the Polish subphase Ib of the Linear Band Pottery Cul­ture. In J. .uteková, P. Pavúk, P. Kalábková, and B. Ko­vár (eds.), PANTA RHEI. Studies in the Chronology and Cultural Development of South-Eastern and Cen­tral Europe in Earlier Prehistory Presented to Juraj Pavúk on the Occasion of his 75th Birthday. Studia Archaeologica et Mediaevalia XI. Commenius Univer­sity. Bratislava: 539–547. 2016. Research perspectives on the exploitation of salt in the Early Neolithic in Kuyavia, In T. Kerig, K. No-wak, and G. Roth (eds.), Alles was zähit… Festschrift für Andreas Zimmermann. Universitätsforschungen zur prähistorischen Archäologie 285. Verlag Dr. Rudolf Habelt GmbH. Bonn: 143–152. Quitta H. 1960. Zur Frage der ältesten Bandkeramik in Mit­teleuropa. Prähistorische Zeitschrift 38(1–2): 153–188. *1970. Zur Lage und Verbreitung der bandkeramischen Siedlungen im Leipziger Land. Zeitschrift für Archäolo­gie 4: 155–176. Rivollat M., Jeong C., Schiffels S., +19 authors, and Haak W. 2020. Ancient genome-wide DNA from France high­lights the complexity of interactions between Mesolithic hunter-gatherers and Neolithic farmers. Science Advan­ces 6(22): eaaz5344. https://doi.org/10.1126/sciadv.aaz5344 Robin Brigand, Jérôme Dubouloz, and Olivier Weller *Sabel K. J. 1983. Die Bedeutung der physisch-geogra­phischen Raumausstattung für das Siedlungsverhalten der frühesten Bandkeramik in der Wetterau (Hessen). Prähi­storische Zeitschrift 58: 158–72. Saile T. 2009. Siedlungsarchäologische Untersuchungen zum Frühneolithikum im südlichen Niedersachsen. In H. J. Beier, E. Claßen, T. Doppler, and B. Ramminger (eds.), Varia neolithica VI. Neolithische Monumente und neo­lithische Gesellschaften. Beiträge der Sitzung der Ar-beitsgemeinschaft Neolithikum während der Jahresta­gung des Nordwestdeutschen Verbandes für Altertums­forschung e. V. in Schleswig 9.–10. Oktober 2007. Bei­träge zur Ur- und Frühgeschichte Mitteleuropas 56. Beier & Beran Archäologische Fachliteratur. Langenweissbach: 43–53. 2010. Aspects of Neolithisation in the Weser-Harz re­gion. In D. Gronenborn, J. Petrasch (eds.), Die Neolithi­sierung Mitteleuropas. Internationale Tagung, Mainz 24. bis 26. Juni 2005. RGZM – Tagungen 4. Römisch-Germanischen Zentralmuseum. Forschungsinstitut für Vor- und Frühgeschichte.Verlag des Römisch-Germani­schen Zentralmuseums. Mainz: 439–447. 2012. Salt in the Neolithic of Central Europe: produc­tion and distribution. In V. Nikolov, K. Bacvarov (eds.), Salt and Gold: The Role of Salt in Prehistoric Europe. Proceedings of the International Symposium (Hum­boldt-Kolleg) in Provadia, Bulgaria, 30 September – 4 October 2010. Provadia, Veliko Tarnovo: 225–238. 2020. On the Bandkeramik to the east of the Vistula Ri­ver: At the limits of the possible. Quaternary Interna­tional 560–561: 208–227. https://doi.org/10.1016/j.quaint.2020.04.036 Salavert A. 2017. Agricultural Dispersals in Mediterranean and Temperate Europe. In H. Shugart (ed.), Oxford Re­search Encyclopedia of Environmental Science. Oxford University Press. Oxford. https://doi.org/10.1093/acrefo re/9780199389414.013.307 Schwartz D., Ertlen D., Davtian G., Dubouloz J., Gauvry Y., Vyslouzilova B., and Berger J.-F. in preparation. Soil fer­tility in the LBK. Identification of methodological prob­lems and distribution of LBK populations in relation to soil fertility. In J.-P. Bocquet-Appel†, J. Dubouloz, R. Mous­sa, and J.-F. Berger (eds.), Modelling the socio-natural trajectory of the LBK farmers. Shennan S., Downey S. S., Timpson A., +4 authors, and Thomas M. G. 2013. Regional population collapse followed initial agriculture booms in mid-Holocene Europe. Nature Communications 4: 2486. https://dx.doi.org/10.1038/ncomms3486 Silverman B. W. 1978. Choosing a window when estimat­ing a density. Biometrika 65: 1–11. 1986. Density Estimation for Statistics and Data Ana­lysis. Chapman and Hall. London. Sordoillet D., Weller O., Rouge N., Buatier M., and Sizun J.-P. 2018. Earliest salt working in the World: From exca­vation to microscopy at the prehistoric sites of Tolici and Lunca (Romania). Journal of Archaeological Science 89: 46–55. https://doi.org/10.1016/j.jas.2017.11.003 *Strien H.-C. 2000. Untersuchungen zur Bandkeramik in Württemberg. Universitätsforschungen zur prähistori­schen Archäologie 19. Verlag Dr. Rudolf Habelt GmbH. Bonn. Timpson A., Colledge S., Crema E., +4 authors, and Shen­nan S. 2014. Reconstructing regional population fluctu­ations in the European Neolithic using radiocarbon dates: a new case-study using an improved method. Journal of Archaeological Science 52: 549–557. https://doi.org/10.1016/j.jas.2014.08.011 Tóth P., Demján P., and Gria.ová K. 2011. Adaptation of settlement strategies to environmental conditions in southern Slovakia in the Neolithic and Eneolithic. Docu­menta Praehistorica 38: 307–321. https://doi.org/10.4312/dp.38.24 Van der Leeuw S., Favory F., and Fiches J.-L. (eds.), 2003. Archéologie et systemes socio-environnementaux. Etu­des multiscalaires sur la vallée du Rhône dans le program-me Archaeomedes. Monographie du CRA 27. CNRS Editi­ons. Paris. Wand M. P., Jones M. C. 1995. Kernel smoothing. Mono­graphs on Statistics and Applied Probability. Chapman and Hall. London. Weller O. 2015. First salt making in Europe: an overview from Neolithic times. Documenta Praehistorica 42: 185– 196. https://doi.org/10.4312/dp.42.12 Weller O., Dumitroaia G. 2005. The earliest salt produc­tion in the world: an early Neolithic exploitation in Poia­na Slatinei-Lunca, Romania. Antiquity 79(306): Project Gallery. http://antiquity.ac.uk/projgall/weller306/ Yokoyama R., Shirasawa M., and Pike R. J. 2002. Visuali­zing topography by openness: a new application of ima­ge processing to digital elevation models. Photogrammet­ric Engineering and Remote Sensing 68: 257–265. Zaninetti J.-M. 2005. Statistiques spatiales, methodes et applications géomatiques. Lavoisier. Paris. Colonization dynamics of LBK farmers in Europe under geostatistics test Appendices Appendix 1. Variability of site data (BD_LBK) across major countries. Robin Brigand, Jérôme Dubouloz, and Olivier Weller Appendix 2. Contributions of the mo­dalities to the first component (F1) of the MCA on the archaeological informa­tion of the database. The vertical dot­ted lines represent the average theore­tical contribution. The first factor (F1) is based on 15 modali­ties at or above the average theoretical con­tribution (81.7% of the total inertia). Five of these (positive, 26.2%) belonging to 5 out of 6 different variables define a weakly documented cluster; it is significantly re­presented in Poland and even in the Czech Republic. Ten other modalities characterise the opposite pole (negative, 55.6%); they belong to all the variables and combine geo­graphical precision, extensive and partial excavations, short/medium duration of oc­cupation, datation at LBK stages 2, 4, 5, and as a whole, good to very good documenta­ry reliability: France contributes more than the other countries to this configuration, due to its strong link with the modalities ‘Loc. ++’, ‘Fiab. A’ and ‘D. 100’. F1 thus reveals a maximum and not very surprising variability of the data around do­cumentary reliability and precision. There is a clear correlation between the most su­perficial types of intervention (prospecting, collecting), imprecise location (Loc ––), un­defined chronology and duration of occu­pation (St.nr and D.nr). Colonization dynamics of LBK farmers in Europe under geostatistics test Appendix 3. Contributions of the mo­dalities to the second component (F2) of the AFCM. The second factor (F2) is based on 11 mo­dalities at or above the average contribu­tion (79.2% of the residual inertia). Of these, six are positive (33.6%) and belong to four out of six different variables. They define a pole representing early and mid­dle LBK sites (Ph.1–3) from surface collec­tions, with short/medium duration, correct­ly located and on the whole limited in re­liability. Two countries contribute particu­larly to this cluster: Slovakia and Hungary. Five other negative modalities (45.5%) cha­racterise the opposite pole; they belong to four of the six variables analysed and com­bine geographical accuracy, surveys and good to very good documentary reliability. Only two countries determine this configu­ration, Poland in particular and France more slightly. This second summary on the remainders of variability not described by F1, therefore reveals the contributions of four new de­scriptors: two modalities of the variable ‘Type of intervention’ (Survey and Collect.) and two of the variable ‘Country’ (Slovakia, Hungary). It contrasts sites mainly North­ern and Western sites with good documen­tary reliability, resulting from ‘Surveys’ and very well located, with a group of early sites in the chronology from southern cen­tral Europe, resulting from more or less ran­dom collections (‘Collect’), but correctly lo­cated. Robin Brigand, Jérôme Dubouloz, and Olivier Weller Appendix 4. Contributions of the mo­dalities to the third component (F3) of the MCA. Finally, the third factor (F3) sumarize a re­sidual fraction of the total inertia of the data, not yet taken into account. In this way, it adds complementary components to the global design of the data variability. It is constructed by 12 modalities at or above the average contribution (83.3% of the re­sidual inertia). Seven of these are positive (52.1%) and belong to 4 out of 6 variables. They define a cluster of central-eastern sites (Hungary, Poland, Slovakia, Austria) result­ing from ‘Survey’, belonging rather to stage 2 of the LBK (Ph.2) and showng a long du­ration (D. 300). Five other negative moda­lities (31.2%) define the opposite pole; they belong to only half of the variables ana­lysed and combine late LBK stages with a correct location: this configuration is found mainly in France and Belgium. This third complementary synthesis thus marginally contrasts early and long-lived Central European sites, not very well locat­ed and resulting from ‘Collect’, with Far-Western sites belonging to the recent sta­ges of the LBK and correctly located. It also tells us that the main countries concerned in F2 by the ‘Collect’ modality are also con­cerned in F3 by the ‘Survey’ modality (Po­land, Slovakia, Hungary). Colonization dynamics of LBK farmers in Europe under geostatistics test Appendix 5 (see Figure 4, Map F1/F2) The analytical data set is displayed in the factorial space around the most discriminating modalities of the first two summaries of the MCA. They oppose, in a ternary system, Poland, Hungary/Slovakia, and France. . To the right of axis 1, are grouped the modalities representing sites which are relatively poorly doc­umented (Fiab. C), both chronologically (St.nr and D. nr) and geographically (Loc ––). Among them are those resulting from the lightest archaeological investigations (Collect. and Survey). Many Polish, Czech, Swiss, Ukrainian, Romanian and Moldovan sites share these characteristics: the first ones par­ticularly linked to the ‘Surveys’ and all the others ra­ther to the ‘Collect’ variable. It is therefore a large group of mostly central-eastern and eastern sites, which are generally less well documented than those in other regions of LBK Europe. . In contrast, to the left of axis 1 and at the bottom of axis 2, are the well documented sites (Fiab. A), with good geographical precision (Loc ++), result­ing from extensive excavations (F. Ext) and dated to the latest stage of the LBK (St.5). The countries concerned by this high-quality documentation are mainly France, the Netherlands and Luxembourg. . As an intermediate to the previous clusters but still to the left of axis 1, meaning not being domi­nated by low quality documentation, the average archaeological information from Germany, Belgium, Slovakia, Hungary and Austria seems to be relati­vely well documented and localised (Fiab. B and Loc +–); it appears to be linked particularly to par­tial excavations (F. Part.), to the main stages of the LBK (St.1 to St.4) and to the most common occu­pation duration modalities (D. 100 to D. 300). This average position applies thus to countries that pre­sent in moderate proportions the whole spectrum of research types and a great variety of documen­tary realities. Austria is rather better documented than Belgium and Germany, where there is a signi­ficant proportion of less well-informed sites. These are therefore similar to those widely encountered further east, notably in the Czech Republic. Finally, Hungary and Slovakia, linked to partial rather than extensive excavations and to the early stages of the LBK, are isolated from the other regions of Europe and particularly from Poland with its particular link to ‘Survey’. Actually, the structuring of the data revealed by this first and most distinctive synthesis of the data, is not surprising at all. Main opposition between po­orly documented groups of sites and others is a kind of truism; similarly, the correlations between loca­tion accuracy, chronological characteristics and type of investigation are also predictable. But it was not taken for granted that all these possible variations should be organized in coherent sequences, evident for almost all the descriptors together. Under these appearances, the ordering of the countries on the factorial map draws an almost perfect east-west pro­gression, which we believe is related to the tempo­ral depth of research on the one hand and its fund­ing on the other. Such a spatial configuration could rapidly change under an unlikely common elevation of the economic and technic standards of archaeolo­gical research in Europe. Poland, Hungary, eastern Lands of Germany, for example, have finally imple­mented effective and promising practices (rigorous survey programs, broad scientific Rescue Archaeo­logy) that show a way forward. But at the same time West Germany has been slowly regressing in these respects, while other regions (Belgium, Slovakia, Ro­mania) have remained behind. We cannot yet be sure whether the recent evolution of practices in France will impact the results upwards or down­wards. Therefore, the harmonisation of data on a continental scale is still a long way off, or could be achieved at a frustrating intermediate standard. Appendix 6 (see Figure 5, Map F1/F3) Following the methodological principles of MCA, map F1/F3 (Fig. 5) modulates the previous conclu­sions. It presents only two groups of opposite mo­dalities: to the right of axis 1, the same set as before is shown again, due to the decisive weight of their modalities in the characterization of the first factor, and confirms the singularity of Poland; but it redis­tributes the other modalities in a single second group around which gravitate the other modalities of ‘coun­tries’ and those, extreme, of the LBK ‘chronology’ (St.0 and 4–5): this map confirms Austria, linked to the long-lived durations (D.300 500), in a positive situation; France and the Netherlands, around late chronological stages and short durations, appear again as more often better informed and the Bel-gium-Germany-Slovakia trio still show their interme­diate quality. Robin Brigand, Jérôme Dubouloz, and Olivier Weller Appendix 7 (map F2/F3) The F2/F3 map characterises the database without the influence of poor documentary situations; it can thus show some secondary links previously masked by the hyper-structuring weight, in F1, of weaker or undefined modality values. The main point of this map is, on the one hand, a significant and favour-able configuration around Poland and its systematic ‘Survey’ practices; on the other hand, a clear high­lighting of the Hungary and Slovakia’s specificity re­lated to the weight of their older and long-lived sites. The group of sites of the BD_LBK with limited or scarce documentation is logically in a central posi­tion and therefore not very significant in this sec­ondary statistical summary. But interestingly, a part of the German sites belongs to this group, as already suggested by the previous diagrams. At last, the western trio (FR, NL, LU) remains linked to the best modality values, while part of the excavated Belgian sites are close to them, probably because of their links with the late stages of the chronology. Colonization dynamics of LBK farmers in Europe under geostatistics test Appendix 8. Topographic and agro-ecological data of the GIS. Robin Brigand, Jérôme Dubouloz, and Olivier Weller Appendix 9. Contributions of the agro-environ­mental modalities to factor 1 of the MCA. Ver­tical dotted red lines show the average theore­tical contribution. The highest contributions to Factor 1 (85% of its ~ inertia) are all negative. They relate to the ‘unde­fined’ modalities PR0 (n.r.), SO4 (no data) and TX0 (n.r.) which almost exclusively construct this first sta­tistical summary (respective contributions of 22.9%, 22%, 27.8%, i.e. 73% of the variability summarised ~ by this factor). The other negative modalities PR1 (7%) and SO5 (5.3%), i.e. shallow soils and Gleysols or Vertisols, contribute only a little. The first factor therefore mainly separates sites in poorly defined or rare contexts from all other sites. Its relevance to our questions is therefore very limited. Colonization dynamics of LBK farmers in Europe under geostatistics test The second factor is mainly built (84% of its iner- ~ tia) upon the opposition between negative PR4/very deep soil (15.7%), SO7/degraded and leached soil (13%) and also, very moderately, TX3/medium-fine texture (6.7%) and FR2/fairly good fertility (3.4%), for a total of 38% of the factor’s inertia; facing the ~ positive modalities PR2/moderate depth (10.3%), SO1/Cambisol (11.9%), and very moderately TX4/ fine texture (4.8%), FR1/very good fertility (4.2%), OP1/closed landscape (3.9%), PR3/deep depth (3.9%), FR3/medium fertility (3.2%) and SO2/clino­humic soil (3.2%) for a total of 40%. ~ That is to say, an overall opposition between set­tlements on leached (Luvisols) or degraded soils, very deep, with medium-fine texture and good fer­tility; and those on brown soils slightly evolved or even clinohumic, with moderate to deep depth, fine texture, very good or medium fertility and in a closed landscape. Robin Brigand, Jérôme Dubouloz, and Olivier Weller The third factor is mainly built (78% of its vari- ~ ance) upon the opposition of the negative modali­ties SO2/clinohumic soil (16.6%), FR1/very good fertility (10.5%), PR3/deep depth (9%), and HY1/low hydric saturation (3.8%), for a total of 40%; against ~ positive modalities PR1/shallow depth (10.9%), PR2/ moderate depth (5.8%), SO8/thin soil (5.2%), SO5/ gleysol-vertisol (5.2%), HY3/high hydric saturation (4.1%), OP1/closed landscape (3.7%), FR3/moderate fertility (3.2%) and FR4/low fertility (2.7%), for a total of 41%. ~ This is an expected contrast between settlements on deep and very fertile soils of Chernozems-Greyzems-Phaenozems types and those on very thin or water­logged soils. Colonization dynamics of LBK farmers in Europe under geostatistics test Robin Brigand, Jérôme Dubouloz, and Olivier Weller Unified phasing of the whole LBK’s extension, at least in large stages, is required for this research pro­ject: continuities and discontinuities across time and space have to be considered at once. However, this ambition is particularly sensitive between the West­ern and Central Eastern chronological systems. Al­though many specialists may not be satisfied, we have nevertheless made such an attempt at chrono­logical standardisation, according to the following approximative six stages sequence. Appendix 13. Chronological stage-system in this paper. Approx. cal BC 5600– 5500 5500– 5325 5350– 5325 5325– 5150 5175– 5125 5125– 5025 5025– 4925 This paper St 0 St 1 St 1\2 St 2 St 3 St 4 St 5 Meier-Arendt 1966< 1972 I II III IV V Dorhn-Ihmig 1979 1a Ib-c Id-IIa IIb-c IId Lorraine-Bassin-Parisien (Blouet et al. 2013) Rubané Ancien (Flomborn) 3–4 – Rub. Moyen 5–7 – Rub. Récent 7–9 Rub. Final\ Termin. – Alsace (Denaire et al. 2017< Lefranc 2007) Rubané Ancien (Flomborn) Rub. Moyen Rub. Récent Rub. Final Mit.\Nied. Rhein (Dohrn-Ihmig 1979) I\II| (Nieder Kass.-Uckend.) Ib-c II- Ältere (Flomborn) Id-IIa III-Mittlere IIb-c IV-Jüngere IId V-Jüngste Würtemberg-Neckar (Strien 2000< Lindig 2002) Älteste 2a | 2b-5 6–7 8 Hi Hessen (Kneipp 1998) Älteste 2 | 3–5 6 7–8 9 Nied. Sachsen-Elbe\ Saale-Sachsen (Behrens 1973< Kaufmann 1987< Quitta 1970) Älteste | Ältere (Flomborn) Mittlere Jüngere Jüngste\ SbK Südbayern\Schwaben (Pechtl, Land 2019) Älteste | Ältere Mittlere Jüngere Jüngste SOB (SbK) :esky´(Pavlu 2000) Ib-c Ältere I\II (Á;kovy´ |) IIa-b-c\ (Á;kovy´) IIIa-b III\IV IVa-b< [arka IV|\SbK Morava (:i/már¡ 1998) Ia-b Ältere LBK 1b2 (Á;kovy´ |) IIa-b Jung LBK IIc NKK\Zseliz III [arka | Niederösterreich\Burgenland (Lenneis, Pieler 2016) Älteste Forma­tive Ältere LBK Übergangsph. I\II II Jung LBK Notenkopf II| Zseliz\ Keszthely III [arka\ Zseliesovce | Centra a západ Slovensko (Pavúk 2004< 2005) Forma­tive| Ältere LBK Bin¡a Milanovce Jung LBK – Notenkopf Zseliesovce Zseliesovce Z. Magyar Dunántúl (Oross, Bánffy 2009) Forma­tive Ältere LBK Bicske-Bin¡a Milanovce Jung LBK Notenkopf\ Keszthely Zseliz\ Keszthely ZI-II Zseliz\ Keszthely ZIII\L. Malopolska (Czekaj-Zastawny 2009< 2013) Ältere LBK Zofipole| Jung LBK Notenkopf Zseliz Zseliz LPC\SbK Kujawy\Chelm. (Pyzel 2006< 2010) Ältere LBK Zofipole| IIa-b Jung LBK III| III LPC\SbK Colonization dynamics of LBK farmers in Europe under geostatistics test Appendix 14. Protocol for the elaboration of transitional maps (example of the Upper-Rhine Valley). As shown in the figure below, the colonization sec­tors for a period t+1 are divided into two groups: the one (orange) that develops (at t+1) inside an area already settled during the previous period (t); and the one (red) that develops outside a previous­ly occupied area. In this way we can better identify true expansion from densification. Left and centre: density maps (r. 12.5km) and 12.5km buffer zone (St. 3 and St. 4); right: transition map. The area of slightly less than 500km2 delimited around sites from a radius of 12.5km is only a study protocol aimed at optimizing the density maps and allowing for a synoptic representation of settlement dynamics. back to contents Documenta Praehistorica XLIX (2022) Neolithic settlement structures in Central Europe> case study of East Bohemia and the Morava River catchment Franti[ek Trampota1,2, Petr Pajdla1,3 trampota@phil.muni.cz< petr.pajdla@protonmail.com 1 Department of Archaeology and Museology, Faculty of Arts, Masaryk University, Brno, CZ 2 Institute of Archaeology of the Czech Academy of Sciences, Prague, CZ 3 Institute of Archaeology, Czech Academy of Sciences, Brno, CZ ABSTRACT – The study examines the degree of similarity of Neolithic settlement structures in two geographically separated regions (eastern half of Bohemia, Morava River Basin) based on the ana­lysis of 11 variables related to the environment and the settlement structures. The period studied corresponds to c. 4900–3400 BC. Although the results of most of the variables analysed using princi­pal component analysis (PCA) do not show significant differences in the preference of settlement locations, the analysis of the individual variables points very clearly to major differences in settle­ment patterns. These are manifested in different settlement dynamics, accessibility to stone raw ma­terials, and the spatial extent of occupation. The general conclusion is that although early agricultu­ral societies are similar in general terms regarding the location of settlements, their individual as­pects are quite different, which must have been reflected in lifestyles during the Neolithic. KEY WORDS – Neolithic; settlement; chronology; central Europe Neolitske poselitvene strukture v Srednji Evropi> primera vzhodne :e[ke in pore;ja Morave IZVLE.EK – V .tudiji, ki temelji na enajstih spremenljivkah, povezanih z okoljem in poselitvenimi strukturami, preu.ujemo stopnjo podobnosti naselbinskih struktur v dveh geografsko lo.enih regijah (vzhodna polovica .e.ke in pore.je Morave). .tudija zajema .as med 4900–3400 pr. n..t.. Analiza glavnih komponent (PCA) pri ve.ini spremenljivk sicer ne ka.e bistvenih razlik v izbiri poselitvenih prostorov, vendar analize posameznih spremenljivk jasno ka.ejo na velike razlike v poselitvenih vzorcih. Te so v razli.nih poselitvenih dinamikah, dostopnosti do surovin za izdelavo kamenih oro­dij in v obsegu poselitve. Splo.na ugotovitev je, da so zgodnje poljedelske dru.be izbirale podobne po­selitvene prostore, .eprav so posamezni vidiki izbire precej razli.ni in so se gotovo odra.ali v .iv­ljenjskem slogu teh dru.b v neolitiku. KLJU.NE BESEDE – neolitik; naselja; kronologija; Srednja Evropa Introduction After the end of the stage of the initial development in this region has been almost exclusively dependent of Neolithic society in central Europe, which is cha-on knowledge of pottery decoration, which serves racterized by the Linear Pottery culture (LBK), the as a basic chronological descriptor. In the traditio-original symbolic style broke down into several re-nal sense, then, symbolism on pottery is seen as a gional factions in terms of the decoration of ceramic key reflection of social affiliation – archaeological vessels. The existing research on Neolithic society culture, as a legacy of Romanticism in the form of DOI> 10.4312\dp.49.15 Neolithic settlement structures in Central Europe> case study of East Bohemia and the Morava River catchment the theoretical approaches of Gustaf Kossina (1911) and Vere Gordon Childe (1925). In contrast to sym­bolic systems, the study of settlement structures was developed under the influence of processual archaeo­logy as a manifestation of strategic adaptation to changes in environmental conditions, which became evident in archaeological research on the Neolithic in Central Europe from the 1980s onwards (Lenneis 1982; Rulf 1983). Despite the influence of processu­al archaeology, however, questions of human adap­tation to the environment were shaped by the in­fluence of cultural historicism, particularly with re­gard to perceptions of the chronological develop­ment of early agricultural societies. In previous research on settlement structures the chronology was determined solely on the basis of symbolic systems, which are reflected in the mor­phology and decoration of pottery and are called ‘archaeological cultures’. However, these lack expla­natory potential for understanding the dynamics of prehistoric societies (most recently Furholt 2021). However, due to their frequent occurrence, Neolithic symbolic systems captured in ceramic production are a useful means for understanding the chronolo­gy of settlements based on knowledge of their tem­poral occurrence using radiocarbon dates (Trampo­ta, Kvìtina 2020). In this study, we focus on characterizing the changes in the settlement organization of human populations that inhabited two separate regions in central Europe: the Morava River Basin and the eastern half of Bo­hemia (Fig. 1). We hypothesise that Neolithic organi­zation of settlements is part of the reflection of life­styles conditioned primarily by the natural environ­ment and by social behaviour patterns. The basic questions we address are: (1) What vari­ables influence the development of settlement pat­terns in two geographically separated regions? (2) Can some variables be regionally distinguished? (3) What weight do the selected variables have for the study of settlement structures? (4) Do changes in ce­ramic style correspond to changes in settlement struc­tures or some of the variables under study? The beginning of the period (c. 4900 BC) included in the study corresponds to the end of the Early Neo­lithic, when there is a decline in human activity in many regions of central Europe, as reflected in the frequency of radiocarbon dates (Shennan 2018), and at the same time the symbolic Linear Pottery style (LBK) comes to an end. In terms of the traditio­nal nomenclature of symbolic styles, the period we study corresponds to the Stroked Pottery (SBK), Lengyel Pottery (LgK), Jordanów/Epi-Lengyel Pot­tery, Michelsberg Pottery (mainly Bohemia) and the Funnelbeakers (TRB), which occur together with Retz-type pottery (mainly Morava River catchment). The first region of interest is the eastern half of Bo­hemia. Here, the settlement area is bounded to the north, east and south by mountains and highlands, Franti[ek Trampota, Petr Pajdla while the eastern part is artificially demarcated by the watercourses of the Vltava and Jizera rivers, which have an approximately north-south course. The number of analysed settlements in the area of c. 9000km2 is 597, which corresponds to an average density of 0.06settlements/km2. The second settlement area is the Morava River Ba­sin, which is mainly located in Moravia, in the north­ern part of Lower Austria and in the westernmost part of Slovakia (Záhorie). This area is bounded by mountains and highlands to the west, north and east, while at the southern border the settlement area continues further. The number of analysed set­tlements in the settlement area of c. 19 000km2 is 1556, which corresponds to an average density of 0.08sites/km2. The two settlement regions are located in similar geographical and climatic areas of the temperate zone of central Europe; the terrain relief is not cate­gorically different, and we consider them compara­ble in terms of the natural environment. Data and methods Chronology In the absence of a dominant successor interpretive model, the model of archaeological culture has been used until now, at least in the sense of the basic clas­sification of material culture, while the change of ce­ramic style is taken into account as a unique chrono­logical factor through whose transformation other changes in the structure of society can be captured. This a priori model is problematic because it does not assume the possibility of significant social chan­ges during the existence of a given symbolic style. On the other hand, we do not currently have a more appropriate chronological identifier that could bet­ter characterize the chronological position of most settlements. In order to understand the chronological evolution of settlement transformations, we turn to three forms of describing individual settlements based on (a) symbolic systems on pottery (archaeological cul­tures – ceramic traditions); (b) basic subdivisions of ceramic typology (phases of archaeological cultures – ceramic groups) based on knowledge of their tem­poral occurrence (following Trampota, Kvìtina 2020); and (c) knowledge of radiocarbon density data that reflect concentrations of human activities. While the chronological distribution is expressed in the aforementioned publication, the determination of the chronology based on the frequency of occur­rence of radiocarbon dates is a sub-subject of this study. A similar analysis has already been performed for both Bohemia and the Moravian-Lower Austrian area (Timpson et al. 2014.553) using the summed probability distribution (SPD). Given the currently significantly higher amount of available radiocarbon dates for both areas, we performed a new SPD in OxCal (Bronk Ramsey 2009) using the IntCal 2020 atmospheric calibration curve (Reimer et al. 2020), which we interpreted using kernel density estima­tion (KDE; Bronk Ramsey 2017). All 14C dates from anthropogenic contexts were in­cluded in the input radiocarbon database (see Sup­ plementary material). All 14C dates related to the period between the beginning of the agricultural pre­history to the end of the TRB, c. 3350 BC, were con­sidered. Only dates with a 1. greater than 100 and dates measured from humic acids were excluded from the dataset. The resulting SPD and KDE plots are validated by the number of known settlements (Fig. 2). In the Morava Basin, the structure of the radiocar­bon dates is divided into three peaks representing three distinct units of population activity. The first peak corresponds to the occurrence of Linear Pot­tery (LBK), and the marked decrease in the frequen­cy of radiocarbon dates at the end of the LBK around 4900 BC is a known phenomenon in other regions in western central Europe (Shennan 2018.104). The second peak is represented by the pottery styles with Stroked Pottery (SBK) and Lengyel pottery (LgK). At the end of this peak, some archaeologists use the term Jordanów culture or Epi-Lengyel. Be­tween c. 4000 and 3800 BC, there is an apparent hiatus in human activity that has not been consid­ered in archaeological research to date, as it cannot be detected outside the context of absolute dates. The last peak of human activity is associated with the Funnelbeakers (TRB) and Retz-type pottery. However, its termination is artificially based on the initial definition of the period under study, which is only associated with TRB, but actually continues in the context of at least Baden pottery. The three identified peaks are a new chronological indicator which we henceforth refer to as Neolithic A, Neolithic B and Neolithic C (Neolithic A is not the focus of this study). In Bohemia, the resulting model is far more proble­matic, as there are significant data biases. At pre­ Neolithic settlement structures in Central Europe> case study of East Bohemia and the Morava River catchment sent, it is not possible to say whether there is a signi­ficant decline in the frequency of 14C dates during the transition from the LBK to the SBK, as the under-dating of contexts associated with Linear Pottery is evident when comparing the number of settlements and the amount of radiocarbon dates. A subsequent significant peak is associated with Stroked Pottery or the study of rondel ditch fills. In this respect, we can speak of a heavy overestimation of the accumu­lation of 14C dates in favour of the one problem ad­dressed. After the SBK period, there is a decline in both the frequency of 14C dates and the number of settlements, which indicates a different development compared to the situation in the Morava Basin. The possibility of a settlement hiatus after 4000 BC is reflected in the modelled dates, but the number of 14C dates from the Early Eneolithic is too low to ob­tain a relevant result. However, similar to the Mora-va Basin, there is a clear increase in the density of 14C dates and the number of settlements around 3800 BC. The resulting den­sity curve of 14C dates is not representative in part, so we provisionally use the same model for Bohemia, which di­vides the Neolithic into peri­ods A, B, and C. For the analysis of settlement characteristics, we use this broad chronological concept in terms of long-term proces­ses. We are interested in whe­ther the settlement during each population phase was chronologically or regionally specific. Settlement data The choice of variables relat­ed to Neolithic settlements re­flects the real possibilities of studying the data structures of the region. Some complex datasets are defined by insti­tutions in three countries and may not be compatible or available. For this reason, pe­dological and geological data, in particular, which may be important for the formation of Neolithic settlements, are not taken into account. The studied variables form two groups – the first is labelled environmental vari­ables, and in general it is comprised of variables de­termined by the environment and terrain surround­ing the settlements; the second group is labelled social variables, and in general these are influenced by processes taking place in the Neolithic society. The database of settlement sites (n=2154) is based on our desire to maximize the information potential with regard to Neolithic settlement sites. In addition to published sites, we acquired data from archaeo­logical archives and some unpublished sources in museum collections. Data were obtained in the form of standard excavations, as well as collections from surface surveys. In its minimalistic form, the Neoli­thic settlement is defined by the repeated presence of ceramics that can be assigned with certainty to at least one of the ceramic traditions (archaeological cultures). Neither isolated lithic finds nor hoards or any cave finds are taken into consideration. The mi­ Franti[ek Trampota, Petr Pajdla nimum distance separating two settlements is c. 200m. The data is described in detail in a dedicated data ar­ticle (Pajdla, Trampota 2021) and is also published in a separate repository under https://zenodo.org/ record/5768049#.YydAhLRBxPY. Data is analysed in the R programming language (R Core Team 2022). To enhance transparency, the compendium containing code is published at the Ze­nodo data repository under https://zenodo.org/recor d/6703463#.YydBi7RBxPY according to principles outlined in Ben Marwick (2017); the original reposi­tory is at https://github.com/petrpajdla/settlements. The most important software packages used in the analysis are the tidyverse family of packages (Wick-ham et al. 2019), sf (Pebesma 2018), and spatstat (Baddeley et al. 2015). We approach the analysis as a multidimensional pro­blem, dimensionality reduction (PCA) is used on the continuous input variables and, consequently, model-based clustering (mclust package, Scrucca et al. 2016) is used to define groups of similar settlements. Analysed environmental variables Altitude is derived manually from online map servers (https://geoportal.cuzk.cz, https://www.geoportal.sk, and https://www.niederoesterreich.at/karte), and al­so from a digital elevation model (DEM) to verify this approach. The values derived from DEM and by hand correlate well (R = 0.97). For the analysis, the values derived by hand are used. The value of slope at the settlement location is de­rived from ASTER GDEM (NASA et al. 2019). ASTER GDEM is used because the territorial scope of the area of interest intersects several state boundaries, namely the Czech Republic, Slovakia and Austria, and obtaining and/or harmonizing individual digital ele­vation models with identical resolutions, etc., for each of the countries was not possible. The slope value is derived as a mean value in a buffer zone around each of the settlement sites with a radius of 300m. The density of the watercourse is the preferred me­thod of assessing the relationship between human occupation and water courses over the distance to a water course. The approach used is based on meth­ods used to study road networks in modern urban areas (Lin et al. 2020). The watercourse density is derived from three individual layers for each of the countries, DIBAVOD (Fojtík et al. 2022) for the Czech Republic and Geofabrik for Slovakia and Lower Au­stria. The density of watercourses is estimated using kernel smoothed intensity from a line segment pat­tern with a kernel size of 2km. The mean value in a buffer zone with a radius of 300m around the set­tlements is recorded. The topographic position index (TPI) is used (toge­ther with the slope) as a proxy for terrain fragmen­tation. It is derived from the digital elevation model using Wilson’s approach (Wilson et al. 2007), i.e. TPI is the difference between the value of a given cell and a mean of eight surrounding cells. TPI effec­tively shows whether the surroundings of a site are flat, or the site is located on a hill (positive values) or in a valley (negative values). As in the case of slope, the TPI for a given settlement site is defined as a mean TPI value in its buffer zone with a radius of 300m. Analysed social variables Settlement density was calculated using kernel den­sity estimation with a kernel size of 4km. Not that in the figures we used a kernel size of 10km. The linear arrangement deals with settlements orga­nized in linear formations along watercourses and on terrain contours, typically in the foothills of up­lands. Settlements have not yet been analysed in this way on a larger scale. Settlement lines were manual­ly defined in the GIS environment based on base layers from ZABAGED and Geofabrik (for Slovakia and Austria) for watercourses and a digital elevation model (OpenStreetMap) for terrain lines. The mini­mum number of settlements per line is three. The location of sources of lithic raw materials in the studied areas (Fig. 3) is defined by Antonín Pøichys­tal (2013). Subsequently, the distance of each settle­ment to each source during the periods when each raw material was exploited was calculated. Resource exploitation was categorized for each period as: none, sporadic, significant. The hierarchical element of settlement structures is considered to be a palisaded or ditched enclosure, which is assumed to have had primarily an econo­mic or fortification purpose. We do not include ron­dels typical for the first half of the 5th millennium BC in this category because of the distinctiveness of rondels from other enclosures – both with regard to the shape of the ditch and the single-phase occur­rence within prehistory. The rondels have a V-shaped Neolithic settlement structures in Central Europe> case study of East Bohemia and the Morava River catchment ditch, whereas the settlement enclosure ditches are flat-bottomed. In the case of hilltop enclosures, then, we observe the remains of a stone wall. Rondels are specific in the context of agricultural prehistory be­cause of their short-term, single-phase occurrence, while settlement enclosures are observed continu­ously. The distance from each settlement to all enclosures has been calculated. The presence of individual ceramic groups at indivi­dual sites was observed, for which we assume chro­nological succession. On the basis of the frequency of successive settlements we observe changes in set­tlement continuity. Results Altitude The character of settlements in relation to altitude differs significantly in Bohemia and in the Morava Basin, both in terms of the evolution of preferences and in terms of data homogeneity. While in Bohemia the values of altitude are very similar for the mean and median, in the Morava Basin the mean and me­dian are significantly different, reflecting the wide range of settlement values; in Bohemia they have a homogeneous character and thus less dispersion of values (Fig. 4). For developmental trends, there is a clear tendency in the Morava Basin to settle sites at higher elevations from the interval 4800–4600 BC (Early Lengyel), which gradually decreases to very low values in the interval 4200–4000 BC (Epi-Len­gyel). A renewed increase in preference for higher altitudes is observed from 3800 BC (TRB) onwards. On the other hand, in Bohemia there is a slight, stea­dy decline in the preference for higher elevations from the beginning of the period under study to 3800 BC. As in the Morava Basin, we observe a pre­ference for higher elevation settlements from 3800 BC onwards in relation to the TRB. The described trends are consistent with the data defined by both ceramic traditions and ceramic groups. Slope The data categorized by ceramic tradition show a si­milar tendency for slope preference as for elevation, which is indirectly related. If we classify the data in more detail according to ceramic groups, there is a greater dynamic (Fig. 5). For both study areas, low Franti[ek Trampota, Petr Pajdla Fig. 4. Violin plot with elevations of settlement sites in individual periods. Data are classified according to pottery groups. settlement slope values (Early SBK) are present at the beginning, followed by an increase from 4800 BC (Early Lengyel, Late SBK) and followed by a gra­dual decrease until 3800 BC. A reversal towards higher values occurs in Bohemia from 3800 BC and in the Morava Basin from 3600 BC. Towards the end of the period under study in Bohemia (Saalzmünde, Boleráz) the settlements are located on flatter sites, while in the Morava Basin (Boleráz) there is a clear preference for more sloping sites. Watercourse density The basic characteristics of the development of the relationship between settlement and watercourse density is the same for data categorized by both ce­ramic tradition and ceramic group. During the peri­od 4900–4600 BC (SBK), settlements in Bohemia are located in areas of high watercourse density, fol­lowed by a sharp decline and a slightly downward trend in the rest of the period. In contrast, the situ­ation in the Morava Basin is negatively correlated with this and the trend is exactly the opposite (Fig. 6). Kernel density (KDE) of settlements The density of settlements in terms of general trends is similar for the data categorized by ceramic tradition and ceramic group, but some changes in the short- Fig. 5. Violin plot with slope values of settlement sites in individual periods. Data are classified accord­ing to pottery groups. Neolithic settlement structures in Central Europe> case study of East Bohemia and the Morava River catchment Fig. 6. Violin plots with KDE of watercourses in individual periods. Data are classified according to pot­tery groups. term trends are significant in the case of the ceramic groups (Fig. 7). In both regions, the beginning (Early SBK) is characterized by low settlement density. This is followed by an increase in settlement density, but in the Morava Basin the settlement density reaches almost twice the values in Bohemia. While in the pe­riod 4600–4200 BC (Late Lengyel) the values in Bo­hemia no longer reach the previous maximum, in the Morava Basin the density continues to increase, followed by a sharp decline in values in the period 4200–4000 BC (Epi-Lengyel). In the Morava Basin a settlement interlude follows in the period 4000– 3800 BC, whereas in Bohemia the settlement hiatus is not yet evident in the radiocarbon dates. The low­est settlement densities, however, are shown by set­tlements in 3800–3600 BC in association with the beginning of the TRB, but this is probably related to the relatively small proportion of TRB settlements whose ceramic typology is more closely subdivided. During the TRB, settlement density increases slightly in both regions. Beyond the statistical results, it is important to con­sider the spatial location of the main settlement clus­ters and their size. We categorize these data only in the context of ceramic typological groups. In Bohe­mia, the settlement core in the Early SBK period is located in the north of the settlement area in the Fig. 7. Violin plot with KDE of settlement sites in individual periods. Data are classified according to pot­tery groups. Franti[ek Trampota, Petr Pajdla broader vicinity of the Ji.ín Upland, an area that is somewhat less populated in other periods, especial­ly in comparison with the Polabí (Elbe) region (Fig. 8). Subsequently, in the context of the Late SBK, set­tlement appears in several distinct settlement clus­ters on the right bank of the Elbe in eastern Bohe­mia and in the vicinity of Chrudim and Kolín. Settle­ment during the Late Lengyel (Fig. 9), albeit in smal­ler numbers, is organized very similarly, but the settlements are more concentrated in the broader vicinity of the Elbe. Settlement structure changes markedly during the Proto-Eneolithic (Fig. 9), when set­tlements are concentrated in the Polabí region between Prague and Kolín, while set­tlements in eastern Bohemia are sporadic. In Bohemia, TRB settlements occupy the area in central Bohemia south of the Elbe River and in eastern Bohemia west of the Elbe Ri­ver. The sub-typochronologi-cal classification of TRB set­tlements is at a low level, but it can be assumed that the characteristics of settlements in 3800–3400 BC are similar to those of the TRB in gene­ral. The conclusion of the TRB (Fig. 10) is interesting in terms of settlement organiza­tion because of the occurrence of one distinct cluster in and around Prague characterized by Saalzmünde pottery, while a separate second smaller clu­ster of settlements is located in the .áslav Basin and is cha­racterized by Boleráz pottery. In the Morava River Basin, the smaller number of settlements in the period characterized by the Early SBK are concentrat­ed in the western half of the settlement area, while the area east of the Morava River shows almost no settlement activity (Fig. 8). Subsequent­ly, in the context of the Early Lengyel there is an increase in the number of settlements concentrated mainly around Brno and in southwest Moravia. The area east of the Morava River is only sporadically settled, and in central Moravia there is also very sparse set­tlement, moreover the ceramic production is charac­terized by Late SBK. In the Late Lengyel period (Fig. 9), the settlement situation reverses: while in the southern half of the basin the density of settlements Neolithic settlement structures in Central Europe> case study of East Bohemia and the Morava River catchment decreases considerably and they are concentrated into more small settlement clusters, in the north­ern half of the basin (central Moravia) a large settle­ment cluster is formed. The area to the east of the Morava River is for the first time since the LBK more heavily populated. During the Epi-Lengyel (4200– 4000 BC), the southern half of the Morava Basin is spar­sely populated, except for the area along the right bank of the lower course of the Mo-rava River. As in the previ­ous period, settlement is con­centrated in central Moravia. Between c. 4000 and 3800 BC, we assume a settlement hiatus, or a form of human presence that left no signifi­cant archaeological traces. In c. 3800–3500 BC (Early TRB, Baalberge, Retz-type), three unequal settlement clusters (Fig. 10) emerge in the vici­nity of present-day Prostìjov, Brno and Znojmo. The south­east of the studied area is sparsely settled, if at all. To­wards the end of the period under study (TRB – Boleráz) we observe an increase in the concentration of settlements in central Moravia, with a small cluster in southwest Mo­ravia and the lower course of the Morava River. The rest of the Morava Basin is not signi­ficantly populated. Linear arrangement – along terrain lines In terms of proportions, set­tlements are located along the terrain lines mostly in the pe­riod 4900–4800 BC (Early SBK), after follows a decline, especially in Bohemia. Here, settlements form slightly more distinctly along terrain lines from 4200 BC onwards (Proto-Eneolithic and TRB). In the Morava River catchment, on the other hand, the forma­tion of settlements along ter­rain lines is a very fundamen­tal phenomenon for the period 3800–3300 BC (TRB), with the main area where settlements are organized in this way being the southern and eastern margins of the Drahany Uplands (Fig. 12). Parallel to these, the linear organization of the hillforts forms a spe­ cific structure during the Boleráz phase. Franti[ek Trampota, Petr Pajdla Linear arrangement – along water lines In terms of frequency, the organization of settle­ments along water lines appears less pronounced than for terrain lines. The only period and region where settlements are significantly located along water lines (Fig. 11) is 4900–4800 BC (Early SBK) in the Morava Basin. There is also significant settle­ment organization along watercourses in the Mora-va Basin during the Late Lengyel, c. 4600–4300 BC, but statistically this disappears in the face of the very high number of settlements during this time. The organization of settlements along watercourses and on terrain lines during the Late Lengyel is also significant (Fig. 12), but this is not statistically signi­ficant in the context of the large number of settle­ments. Hierarchical arrangement The analysis of the relation­ship between settlements and enclosed sites (Fig. 13) was carried out only in the con­text of data described by cera­mic traditions, as the detail­ed chronology of many en­closed areas is unknown. The analysis shows that for most of the periods the distance be­tween settlements and enclo­sures recorded in the Morava Basin is smaller, which is mainly due to the larger num­ber of known enclosures. The opposite situation is in Bohe­mia during 4200–3800 BC (Proto-Eneolithic), when a large number of lowland en­closures are recorded in cen­tral Bohemia (Kri.tuf, Turek 2019). In the Morava River catchment, on the other hand, the distance to enclosed areas decreases significantly in 3800–3300 BC in the context of the TRB, when a number of new hillforts are establi­shed, often with evidence of a stone rampart. The spatial distribution of enclosures in the Morava Basin is interest­ing from the long-term pers­pective, with almost all of these enclosures located in the western half of the occu­pied area. Distance to raw material sources for chipped stone industry A comparison of the data dis­tribution by ceramic tradition and ceramic group shows no Neolithic settlement structures in Central Europe> case study of East Bohemia and the Morava River catchment structural differences. For the settlements in Bohe­mia, the mean and median values are above 100km; their slight diachronic increase is probably not a re­flection of a change in the relationship between sources and settlements in the context of such large distances. In the Morava Basin, the situation is significantly dif­ferent, mainly due to the proximity of the sources of raw materials for chipped industry. In the period 4900–4800 BC (Early SBK), a significant concentra­tion of settlements is evident in the vicinity of chip­ped industry resources (Fig. 14), which is related to the depopulation of the eastern part of the Morava Basin, where resources are less abundant, and the lo­calization of settlements in the western part, where more raw material resources are concentrated. Krum­lovský les chert played a key role in this period. It is completely predominant in the chipped stone in­dustry assemblages within a radius of at least 60km (Trampota 2015.180–183) and is also significantly present at distances above 80km (Olomouc-Slavo­nín; Kazdová et al. 1999), thus quantitatively the most abundant raw material and the one distributed over the greatest distances in the entire studied pe­riod. In subsequent periods, the distance to the re­sources increases in connection with establishing new sites in the whole settlement zone. The highest distance to the sources of raw materials for chipped stone tools is recorded in the period 4200–4000 BC (Epi-Lengyel), when the number of settlements de­creases to the level of the period 4900–4800 BC (Early SBK), but their main concentration is in cen­tral Moravia, in a region relatively distant from the sources. As such, two types of relationships of settle­ments to the sources of the chipped stone industry can be observed in similarly few populations. A distinct phenomenon since 3800 BC (Early TRB, Baalberge, Retz-type) is the mining and local distri­bution of chert from Stránská skála near Brno (Bar-tík et al. 2019), but the overall settlement pattern does not statistically correspond to this activity, al­though a small cluster of settlements forms around the source. Distance to raw material sources for polished stone industry As with the raw materials of chipped industry, the division of data by pottery traditions and pottery groups shows no structural differences. In Bohemia, Jizera Mountain-type metabasite was an important resource during the Early Neolithic (Pøi­chystal 2013.192), whose distribution radius extend­ed beyond central Europe. The use of this raw ma­terial continues in the following period, but only on a regional scale in Bohemia (.ída 2007). In 4900– 4800 BC (Early SBK), the settlement structure in re­lation to the sources of raw materials of polished stone industry is oriented more towards the vicini­ty of the source. In the subsequent period, however, no changes in settlement structures are evident in relation to the distance from the sources of raw ma­terials. In the Morava River Basin, set­tlements are not related to sources of stone raw material at all in 4900–4800 BC (Early SBK) and raw materials are ge­nerally located at a great dis­tance, which reflects the low frequency of tools found. From c. 4800 BC onwards (Lengyel), a marked change follows, with settlements very much located in the vicinity of raw material sources for polished stone tools, which is strongly reflect­ed in the archaeological record. Subsequently, the distance from the sources increases slightly, which is statistically due to the occupation of a wider area. Set­tlements are more strongly ori­ented in relation to the period Franti[ek Trampota, Petr Pajdla 3600–3300 BC (Boleráz phase), but this is not as pronounced in the occurrence of polished stone tools as during the Lengyel period. Settlement continuity The use of data from settlements defined by pottery traditions is not very suitable for the analysis of set­tlement continuity, as this division includes broad time categories and results in the appearance of of­ten significant settlement continuity. The first time period studied is artificially disconti­nuous, as we do not compare the data with the pre­vious period (LBK). In the Thaya River basin (c. half of the entire Morava Basin), continuity between the Late LBK and the Early SBK is found for 37% of set­tlements (Trampota 2015.138). Data classified by pottery groups (Fig. 15) show a higher proportion of continuously occupied sites in Bohemia than in the Morava Basin. In general, a higher proportion of continuously occupied positions can be observed when the number of known settle­ments is low; when population growth and/or settle­ment of new positions occurs, the proportion of con­tinuously occupied positions is low. TPI index The structure of data is the same for pottery tradi­ tions and pottery groups. Values for most settlements in both regions are around zero, reflecting a preference for flat locations (Fig. 16). In the case of concave landforms (hills), a gra­dual transition to higher values is evident, especially in the Mo-rava Basin, during the period of 4800–4400 BC (Lengyel), while in 3800–3400 BC (TRB) hilltop sites form a specific cluster of values (also less pronounced in Bohemia). Here the hillfort si­tes are separated, whereas in the earlier period hilltop settlement is not a defined category. Results of multivariate sta­tistics The similarity of settlement struc­tures based on long chronology (Neolithic B and Neolithic C) were analysed using principal component analysis. All variables except settlement continuity and the hierarchical arrangement of sites were included in the ana­lysis. However, the results did not yield significant differences that could be interpreted accord­ing to the significant factors, nor according to clusters of points (Fig. 17), which are more a ma­nifestation of random arrange­ment. As a result, the settlement patterns defined by long chrono­logy did not differ in principle, and reflected similar preferen­ces for locations suitable for agri­culture. Neolithic settlement structures in Central Europe> case study of East Bohemia and the Morava River catchment Discussion Importance of individual variables We evaluated the weight of each variable using the sum­med factor scores on the axes, which reflect 75% of the to­tal variability. The number of axes varies, so the average value for each variable was calculated. From the results for the individual chronolog­ical classifications (Tab. 1), it is clear that the variables with the largest sum of factor sco­res are those that show great stability over time and do not change much. These are al­ways the topographic position index and the settlements or­ganized along terrain lines in Bohemia. In the Mora-va River catchment, these are also the topographic position index and the estimation of the kernel den­sity of watercourses. These variables are therefore not very suitable for explaining the dynamics of set­tlement structures. In contrast, variables that have the lowest sum of factor scores can be expected to have more varia­bility and thus these variables may be important for understanding the organization of settlement struc­tures at certain time intervals. A second possible ex­planation is then that they are generally insignifi­cant. In Bohemia, these variables are mainly distance to sources of raw materials, especially chipped in­dustry, and an estimate of the kernel density of set­tlements. In the Morava River Basin, distance to sour­ces of raw materials of polished industry and altitude mainly have the lowest sum of factor scores. These two variables are chronologically highly variable in the Morava River Basin, while in Bohemia the vari­ables with the lowest sum of factor scores have low explanatory potential. Difference between settlement structures in two regions Some of the observed variables are different in the two separate settlement regions and are important for understanding the unequal dynamics of the evo­lution of social organization, subsistence or lifestyle. Analysis of the altitude of individual settlements re­veals that areas up to 500m above sea level were commonly settled in in Moravia, whereas in Bohe­mia settlement was common only up to 350m above sea level, and only rarely at higher altitudes. This is not due to different geomorphological conditions, but another factor. A possible demarcation of settle­ment areas between Neolithic and Mesolithic popu­lations comes into consideration, which leave archa­eological traces on a limited scale. Parallel coexis­tence between Late Mesolithic and Neolithic popu­lations in Bohemia can be considered in the context of Bohemia rather than in the Morava River Basin. region variable r B tpi 0.2895 B line_terrain 0.2648 B hydro 0.2548 B altitude 0.2276 B slope 0.2109 B settlements_kde 0.2078 B line_water 0.2053 B rm_dist_polished 0.1888 B rm_dist_chipped 0.1865 M hydro 0.2661 M tpi 0.2368 M settlements_kde 0.2348 M slope 0.2177 M rm_dist_chipped 0.2128 M altitude 0.2029 M line_terrain 0.2010 M rm_dist_polished 0.1934 M line_water 0.1931 Tab. 1. Table with the sum of factor scores of indi­vidual variables. The resulting value is the aver­age value against the number of axes expressing 75% variability. Franti[ek Trampota, Petr Pajdla This is consistent with the relatively large number of known Mesolithic sites in Bohemia (.uláková 2015.Fig. 10.10.3). In contrast, the number of Meso­lithic sites in the Morava River Basin is very small (Svoboda 2008; Oliva 2018) and the chronological coexistence of the two distinct populations is un­likely. As such, in terms of the spatial extent of Neo­lithic settlement, the fact that the Morava River Ba­sin experiences the most extensive settlement dur­ing the Neolithic in agricultural prehistory, which was only surpassed in extent after 1000 AD, is cru­cial. The situation is just the opposite in Bohemia. The Neolithic settlement area here occupies a rela­tively small area, which only increases significant­ly with the onset of the Bronze Age (Demján et al. 2022.Fig. 10). David Graeber and David Wengrow (2021.262–263) explain the rapidity of the Neolithic colonization of central Europe by the absence of human set­tlement. This colonization stopped mainly before the coastal areas, which were strongly occupied by Mesoli­thic populations. Whether we can draw a similar conclusion for the absence of Neolithic settlements in parts of Bohe­mia, especially in the south, is uncertain. In our case we observe the se­condary consequences of this colonization, which do not dif­fer substantially in their extent from the original si­tuation during the LBK. Another important variable is the sources of stone raw materials. Jizera Mountain-type metabasite play­ed an important role at the beginning of the studied period in Bohemia, but after 4800 BC there was a reversal, with the production of polished stone tools dominating quantitatively from sources in south Mo­ravia, around which settlements were heavily con­centrated. In the eastern part of Bohemia, local sources were never used to a great extent for the production of chipped tools, and the vast majority (silicite of glaci-gene sediments) came from Silesia or Saxony, an area separated by mountains. On the other hand, especial- Neolithic settlement structures in Central Europe> case study of East Bohemia and the Morava River catchment ly in south Moravia, the sources of chipped stone in­dustry were mostly part of a densely populated area, which was reflected in the quantity of their exploi­tation. The Epi-Lengyel period, when settlements were generally concentrated at a great distance from the exploited sources, is a significant deviation from this. This may indicate that stone artefacts were of low socio-economic importance at this time. Another significant difference is the dynamics of set­tlement, which took place in different ways in the two regions. In Bohemia, settlement reaches its peak in the context of the Late Stroked Pottery (c. 4800– 4500 BC), whereas in the Morava Basin it is not until the Late Lengyel (c. 4500–4300 BC). It follows that to understand the lifestyle of each society, it is necessary to study each region separately, and the lifestyle of Neolithic people cannot be generalized by similarities in symbolic expression using ‘archaeo­logical cultures’. The relationship between settlement site and slope gradient in the two regions follows a broadly similar pattern according to the expected population dyna­mics, which we estimate based on the frequency of the occurrence of radiocarbon dates combined with the number of settlement sites. In periods when we record low population density, mainly flat sites are occupied, while in periods of the estimated highest population density, sloping sites are also occupied. Pottery style as a social identity indicator? During the period under study, both areas are cha­racterized by one decorative style, but there are ex­ceptions. In central Moravia, there was in 4800– 4600 BC a small cluster of settlements characterized by the Late SBK, while in the area further south there was dense settlement characterized by Early Lengyel pottery (Fig. 8). Towards the end of the studied period, the settle­ment of the .áslav Basin in central Bohemia is cha­racterized by Boleráz pottery, with settlements with Saalzmünde pottery concentrated especially around Prague (Fig. 10). In both cases, the two groups with different symbolic styles are separated by an unin­habited or very sparsely inhabited area. In the re­sults of the statistical analysis of settlement structu­res, apart from spatial location, these groups do not differ. This reflects the similar nature of lifestyle or subsistence, but in the social dimension there is an obvious geographical distance. A distinct form of the relationship between two dif­ferent decorative styles existed especially in the Mo-rava Basin (and other parts of Lower Austria) in 3800–3400 BC, when Funnelbeaker pottery is often found in contexts together with Retz-type pottery (also known as Furchenstichkeramik). These two styles, completely different in vessel morphology and decoration, are not territorially exclusive but di­rectly intertwined. While the Morava Basin is the south-eastern periphery of the distribution of Fun-nelbeakers, pottery of the Retz type is most often found in the western part of the Carpathian Basin. Two different examples of the spatial occurrence of the different pottery styles thus illustrate the ambi­guity of its interpretation. Franti[ek Trampota, Petr Pajdla Evolution of Neolithic settlements The evolution of Neolithic settlements was clearly dynamic. The rate of change of settlement structu­res was greater than the change of symbolic systems – ‘archaeological cultures’. A weakness of the pre­sented study of settlement structures is the fact that we frame the development of settlement structures on the basis of the development of particular pot­tery styles. While the pottery style is the most appro­priate descriptor of settlements from a quantitative point of view, its disadvantage is its interval nature, which may not capture the other possible and even more rapid dynamics of the development of settle­ment structures. Nor can it be used as a prism to consider whether the transformation of settlement structures occurred gradually or in leaps. With the number of settlements analysed, it is not realistic at present to have a series of radiocarbon dates that would define the probabilities of the beginning and end of most settlements. Conclusion In this article, we have analysed the relationships between Neolithic settlement sites and the environ­ment. The research covered a total of 2153 sites from two separate areas, the eastern half of Bohe­mia and the Morava River Basin for the period of c. 4900–3400 BC. The two separate settlement regions studied differ fundamentally in their characteristics. What is strik­ing is the different chronological development of the number of settlements, which in the Morava Ri­ver Basin correlates relatively well with the long­term development of the frequency of radiocarbon dates, whereas in Bohemia only the development of the number of settlements can be relied upon at pre­sent. The two regions differ in the extent of settle­ment, with the Morava Basin being extensively set­tled, while the eastern half of Bohemia is only spar­sely settled compared to later phases of prehistory. The two regions differ markedly in their access to raw material sources for the production of chipped and polished stone tools. In the Morava River Basin, the importance of amphibole diorite and porphyritic microdiorite for the first mass production of axe-hammers is particularly evident in the significant concentration of settlements around the source area. The chronological evolution of most of the variables studied can be seen as proxy information of popula­tion dynamics. The evolution of settlement structures does not correlate with archaeological cultures, but Neolithic settlement structures in Central Europe> case study of East Bohemia and the Morava River catchment greater dynamics can be observed. The information Outer Subcarpathia and the Vienna Basin can be un-found can be generalized to the European level by derstood as Neolithic peripheries. In Bohemia, then, stating that the Neolithic had an insular character, the central areas are the right bank of the Elbe in the dynamics of the settlement development of each the upper reaches and the wider area of the left region is specific, and to understand the lifestyle of bank in central Bohemia. the people it is necessary to study these regions se­parately. ACKNOWLEDGEMENTS While it is not relevant for the Neolithic to work with We would like to thank Ale. Navrátil and Zdenìk Tvr- the question of centrality in terms of individual sites, dý for permission to publish their radiocarbon dates broader regions can be understood in this way. In and Petr Kvìtina for valuable comments to our arti- this respect, these are defined by dense and long- cle. The research described in this paper was accom- term settlement, whereas the periphery is only heav­ plished with support from the project ‘Lifestyle as an ily populated at the population maximum. In the Mo- unintentional identity in the Neolithic’ (Project 19­ rava Basin, the central area can be defined as prima­ 16304S), financed by the Czech Science Foundation. rily the eastern edge of the Bohemian Massif, while . References Baddeley A. Rubak E., and Turner R. 2015. Spatial point patterns. Methodology and applications with R. Chap­man and Hall/CRC Press. London. Bartík J., .krdla P., .ebela L., Pøichystal A., and Nejman L. 2019. Mining and processing of the Stránská skála-type chert during the Late Neolithic and Early Eneolithic peri­ods. Archeologické rozhledy LXXI/2: 373–417. Bronk Ramsey C. 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51(1): 337–360. https://doi.org/10.1017/S0033822200033865 2017. Methods for summarising radiocarbon datasets. Radiocarbon 59: 1809–1833. https://doi.org/10.1017/RDC.2017.108 Childe V. G. 1925. The Dawn of European Civilisation. Kegan Paul. London. .uláková K. 2015. Pøíspìvek k poznání mezolitického osídlení .ech. Unpublished PhD thesis. Faculty of Arts. Charles University. Prague. http://hdl.handle.net/20.500.11956/82605 Demján P., Dreslerová D., Koláø J., +3 authors., and Lieskov­ský T. 2022. Long time-series ecological niche modelling using archaeological settlement data: Tracing the origins of present-day landscape. Applied Geography 141 (10): 102669. https://doi.org/10.1016/j.apgeog.2022.102669 Fojtík T., Ja.íková L., Kurfiøtová J., + 5 authors, and Zbo-øil A. 2022. GIS a kartografie ve VÚV TGM. Vodohospo­dáøské technicko-ekonomické informace 64(1): 47–52. Furholt M. 2021. Mobility and Social Change: Understand­ing the European Neolithic Period after the Archaeogene-tic Revolution. Journal of Archaeological Research 29: 481–535. https://doi.org/10.1007/s10814020-09153-x Graeber D., Wengrow D. 2021. The dawn of everything. A new history of humanity. Farrar, Straus and Giroux. New York. Kazdová E., Pe.ka J., and Mateiciucová I. 1999. Olomouc – Slavonín (I). Sídli.tì s vypíchanou keramikou. Archeo­logiae Regionalis Fontes 2. Vlastivìdné muzeum v Olo­mouci. Olomouc. Kossina G. 1911. Die Herkunft der Germanen. Kabitsch. Leipzig. Kri.tuf P., Turek J. et al. 2019. Arény pøedkù. Posvátno a rituály na po.átku eneolitu. Ancestral Arenas. Cult and Ritual at the beginning of the Eneolithic Period. Katedra archeologie. Západo.eská univerzita v Plzni. Plzeò. Lenneis E. 1982. Die Siedlundsverteilung der Linearband­keramik in Österreich. Archaeologia Austriaca 66: 1–19. Lin Y., Hu X., Lin M., Qiu R., Lin J., and Li B. 2020. Spatial Paradigms in Road Networks and Their Delimitation of Urban Boundaries Based on KDE, ISPRS. International Journal of Geo-Information 9(4): 204. https://doi.org/10.3390/ijgi9040204 Marwick B. 2017. Computational Reproducibility in Ar­chaeological Research: Basic Principles and a Case Study of Their Implementation. Journal of Archaeological Me­ Franti[ek Trampota, Petr Pajdla thod and Theory 24(2): 424–450. https://doi.org/10.1007/s10816-015-9272-9 NASA/METI/AIST/Japan Spacesystems and U.S./Japan AS­TER Science Team (2019). ASTER Global Digital Elevation Model V003 [Data set]. NASA EOSDIS Land Processes DAAC. https://doi.org/10.5067/ASTER/ASTGTM.003 Oliva M. 2018. Mezolit na Moravì ve svìtle nových výz­kumù a poznatkù. Acta Musei Moraviae, Scientiae socia­les CIII(1): 3–33. Pajdla P., Trampota F. 2021. Neolithic Settlements in Cen­tral Europe: Data from the Project ‘Lifestyle as an Unin­tentional Identity in the Neolithic’. Journal of Open Ar­chaeology Data 9(13): 1–6. https://doi.org/10.5334/joad.88 Pebesma E. 2018. Simple Features for R: Standardized Support for Spatial Vector Data. The R Journal 10(1): 439–446. https://doi.org/10.32614/RJ-2018-009 2021. stars: Spatiotemporal Arrays, Raster and Vector Data Cubes. R package version 0.5-3. https://CRAN.R-pr oject.org/package=stars Pøichystal A. 2013. Lithic raw materials in prehistoric times of eastern Central Europe. Masaryk University. Brno. R Core Team 2021. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Com­puting. Vienna. Reimer P., Austin W., Bard E., + 38 authors, and Talamo S. 2020. The IntCal20 Northern Hemisphere radiocarbon age calibration curve (0–55 cal kBP). Radiocarbon 62: 725–757. https://doi.org/10.1017/RDC.2020.41 Rulf J. 1983. Pøírodní prostøedí a kultury .eského neolitu a eneolitu. Památky archeologické LXXIV: 35–95. Shennan S. 2018. The First Farmers of Europe. An Evo­lutionary Perspective. Cambridge University Press. Cam­back to contents bridge. https://doi.org/10.1017/9781108386029 Svoboda J. 2008. The Mesolithic of the Middle Danube and Upper Elbe rivers. In G. Bailey, P. Spikins (eds.), Me­solithic Europe. Cambridge University Press. Cambridge: 221–237. .ída P. 2007. Vyu.ívání kamenné suroviny v mlad.í a pozdní dobì kamenné. Dílenské areály v oblasti horní-ho Pojizeøí. Dissertationes archaeologicae Brunenses/Pra­gensesque 3. Ústav pro pravìk a ranou dobu dìjinnou, Fi­lozofická fakulta, Univerzita Karlova and Ústav archeolo­gie a muzeologie. Filozofická fakulta. Masarykova univer­zita. Praha-Brno. Timpson A., Colledge S., Crema E., + 4 authors, and Shen­nan S. 2014. Reconstructing regional population fluctu­ations in the European Neolithic using radiocarbon dates: a new case-study using an improved method. Journal of Archaeological Science 52: 549–557. https://doi.org/10.1016/j.jas.2014.08.011 Trampota F. 2015. Sídelní a distribu.ní struktury v neo­litu v regionu povodí Dyje. Unpublished PhD thesis. Ma­saryk University. Brno. https://is.muni.cz/th/emcwq/ Trampota F., Kvìtina P. 2020. How do they fit together? A case study of Neolithic pottery typology and radiocar­bon chronology. Archeologické rozhledy LXXII/2: 163– 193. https://doi.org/10.35686/AR.2020.6 Wickham H., Averick M., Bryan J.,+ 20 authors, and Yuta­ni H. 2019. Welcome to the Tidyverse. Journal of Open Source Software 4(43): 1686. https://doi.org/10.21105/joss.01686 Wilson M. F. J., O’Connell B., Brown C., Guinan J. C., and Grehan A. J. 2007. Multiscale Terrain Analysis of Multi-beam Bathymetry Data for Habitat Mapping on the Conti­nental Slope. Marine Geodesy 30(1–2): 3–35. https://doi.org/10.1080/01490410701295962 Documenta Praehistorica XLIX (2022) Disuse of spaces and discard of artefacts during the abandonment of Erimi-Laonin tou Porakou Andrea Villani andrea.villani@uib.eu University of the Balearic Islands, Palma de Mallorca, ES ABSTRACT – The aim of this paper, starting from the analysis of the assemblage and stratigraphy of the unburned rooms, is to analyse the possible discard and disuse processes during the planned and gradual abandonment at Erimi-Laonin tou Porakou (Cyprus). Scholars note how the decision to leave objects when a place is abandoned depends on multiple factors, from functional reasons to ritual practices. At Erimi some markers suggest a possible intentional closure treatment of parts of the site in which it is possible to recognise a mix of functional and symbolic abandonment behaviours. KEY WORDS – abandonment; discard processes; artefacts; Bronze Age; Cyprus Opu[;eni prostori in odvr/eni artefakti med opustitvijo Erimi-Laonin tou Porakou IZVLE.EK – Namen tega prispevka, ki sicer temelji na analizi artefaktnih zbirov in stratigrafiji ne­po.ganih prostorov, je .tudij procesov morebitne neuporabe in zavr.enja med na.rtovano in postop-no opustitvijo naselja Erimi-Laonin tou Porakou (Ciper). Raziskovalci ugotavljajo, da je odlo.itev o tem, katere predmete pustiti, ko kraj zapu..amo, odvisna od razli.nih dejavnikov, tako utilitarnih razlogov kot obrednih. Na Erimiju nekateri ozna.evalci ka.ejo na mogo.e namerno zapiranje delov mesta, v katerih je mogo.e prepoznati me.anje funkcionalnega in simbolnega vedenja med zapustit­vijo. KLJU.NE BESEDE – opustitev; procesi zavr.enja; artefakti; bronasta doba; Ciper Introduction All architectural remains and associated objects are conditioned by abandonment processes (Tomka, Stevenson 1993.191), which affect the related struc­tures, activity areas, whole settlements or regions, and are caused by various factors (Cameron 1991; Cameron, Tomka 1993; Lightfoot 1993). Excluding rapid and unanticipated abandonments, due to natu­ral disasters or enemy attacks, these processes are often planned. Settlements are therefore transform­ed by a sequence of practices and behaviours that entail the decision to leave behind or transport to the new site artefacts and, in some cases, building ma­terials (Schiffer 1972; Schlanger, Wilshusen 1993. 91). Factors that influence this decision are the de­gree of preabandonment planning, the distance from the new settlement, the capability and means of transport, the activities foreseen in the next loca­tion, the season of movement and ritual practices (Cameron 1991.172; Cameron, Tomka 1993; Deal 1985; Lange, Rydberg 1972; Lightfoot 1993.166; Schiffer 1972.160; 1985; Stevenson 1982). In many cases, places return to be lived in cyclically, involv­ing the recovery and reuse of objects and building materials (Schiffer 1985.27). This phenomenon caus­es a significant loss of information related to the pre­vious stages of life and the disuse episodes of an area. DOI> 10.4312\dp.49.9 Disuse of spaces and discard of artefacts during the abandonment of Erimi-Laonin tou Porakou Planned abandonments are often gradual (Cameron 1993.4), and it is fundamental to recognise the pre-abandonment behaviours to investigate the reasons, possible length and socio-political and economic pro­cesses behind the decision to leave a place (Nelson 2000; Schiffer 1976.33; Stevenson 1982) in order to define the moment of change that is always link­ed to an abandonment (Nelson, Hegmon 2001). What is commonly found – when abandonment dy­namics are not distorted by scavenging or resettle­ment – is the result of a choice, influenced by func­tional and ritual practices. In order to define vari­ous type of deposits and their transition from the systemic to the archaeological context, scholars have identified eight different formation processes of the floor assemblage (Schiffer 1985.24–30; see also Hayden 2000). The conditions of objects, their lo­cation in the rooms, the possible practice behind the discard of them and finally the moment in which they were discarded (prior, during or after the aban­donment of the settlement) characterise these dif­ferent processes (Schiffer 1985; Iannone 1990; Hay­den 2000.300). The aim of this paper is to present an analysis on the discard practices and gradual disuse of some units of the workshop complex (area A) during the abandonment phase of Erimi-Laonin tou Porakou at the end of the Middle Bronze Age on Cyprus (c. 1650 BC). The transition from the Middle to the Late Bronze Age (MC III-LC I) has been defined as a mo­ment of change and a first step to the urban society of Cyprus (Crewe 2017.140). This transformative period entails the abandonment of a series of villa­ges left in various conditions and following differ­ent practices. Abandonment studies on prehistoric Cyprus could provide information that would help us to better define the social changes which char-acterise the period, and also fill the related gaps on the study of this phenomenon in prehistoric settle­ments. The site of Erimi-Laonin Tou Porakou Erimi-Laonin tou Porakou (Limassol district) ex­tends over an area of c. 1.20ha on the upper slopes of a natural hill located 250m above sea level. The general topography of the site is characterised by a hilltop surrounded by a series of lower terraces slop­ing southward and eastward (Bombardieri 2017; Amadio et al. 2021.302). The settlement presents two main areas organised in roofed units carved ap­prox. 0.60–0.70m into the bedrock and open spaces (Fig. 1). The workshop complex (named area A) de­velops on the top of the hill and a residential area (areas B, T2-T5) is located on the first lower terrace, while a massive wall divided the settlement from the cemetery (area E) (Christofi et al. 2015). Andrea Villani The position of the productive area (workshop, named area A) in the upper flat plateau and the pe­culiar layout of its rooms permitted the conservation of the majority of objects in their context of prima­ry deposition, saving them from washout and other post-abandonment formation processes. At the same time, the use of the area for cultivation partially de­stroyed the upper layers of stratigraphy. In contrast, the residential area was subjected to considerable erosion and soil movement by natural agents and slope wash. Due to its position, the accumulation of soil partially saved walls and structures from de­struction or massive reuse of building materials in post-abandonment periods (e.g., areas B2 and T2). In addition, frequentation of the area during the Hellenistic and Roman periods could be noted (Bom-bardieri et al. 2009.134–138). Refuse types at Erimi-Laonin Tou Porakou Archaeological investigations suggest a planned and gradual abandonment of the settlement. While some units were discovered empty of objects, most of the still usable artefacts were concentrated in a group of rooms. Finds were recovered sealed on the floors by the collapse of roofs and walls, in some cases accele­rated by a fire. The rapid destruction meant that these rooms were not used for resettlement or mas­sive reuse in post-abandonment periods and, thanks to this possible closure practice, this can provide a precious opportunity to analyse the site’s last phase of life. Abandonment studies stress the recurrence, in var­ious areas of the world, of a deliberate use of fire to voluntarily destroy structures or part of a settle­ment. This action would be performed by those who lived at a place at the end of a sequence of practices linked, in most cases, to funerary rituals (Cameron 1990; Lindskoug 2016; Schiffer 1985.29; Schlan­ger, Wilshusen 1993; Verhoeven 2000; Wilshusen 1986.246). In addition, other cases of destruction by fire are represented by unexpected natural cata­strophes or enemy attacks, and by sites which were affected by closure practices performed by the inha­bitants (Cameron 1991; Chapman 1999; Lightfoot 1993; Stevanovi. 1997). Some evidence seems to suggest that the case of Eri-mi-Laonin tou Porakou belongs to the last group. No human skeletal remains sealed by the collapse of structures as a result of catastrophic disaster have been found, and no signs of scavenging or any at­tempts at rebuilding activities can be detected at the site (Amadio et al. 2021.312–313). Moreover, the concentration of some artefacts in specific units, and the presence of a sequence of functional practices, such as the possible recovery of building materials from disused spaces, as well as possible ritual practi­ces made before the fire, can support this interpre­tation. Analysis in this paper are focused on the as­semblage and stratigraphy of the unburned units with the aim to describe a sequence of functional practices during the abandonment phase at Erimi. In order to define the abandonment behaviours, and the transition of artefacts found in the rooms from the systemic to the archaeological context (fol­lowing Schiffer 1985.24–30; La Motta, Schiffer 1999; Hayden 2000.300), objects have been divided into three different categories: . De facto refuse refers to objects still usable which were left behind at the time of the abandonment. Michael B. Schiffer (1972.160) defines these as “elements which reach archaeological context without the performance of discard activities”. . Primary refuse refers to artefacts left in their lo­cation of use or in proximity to their original posi­tion (Schiffer 1972.161–162). . Secondary refuse refers to objects discarded any­where other than in their location of use (Schif-fer 1972.161), about which Hayden (2000.300) noted: “this is refuse that has been cleaned up and removed from its primary use or manufac­turing context and dumped elsewhere, usually in designated refuse areas”. It is worth specifying that even if they will be used as guidelines in the present work, these categories are not universal but conventional. For example, in the case of the stone tools in the depositional his­tory of stone artefacts at Maki-Alonia, David Frankel and Jennifer M. Webb (2012.480) distinguished ex­pedient tools (de facto refuse), described as still in­tact objects, discarded at or near their location of original use after a short use-life, from curated tools (normal refuse) that they define as objects that spent a long period in the systemic context and were dis­carded “in or near their context of final use when damaged, broken, near exhausted, or exhausted”. Moreover, Arthur A. Joyce and Sissel Johannessen (1993.138) stated that primary and secondary re­fuse disposal, prior or during a gradual process, must be seen as abandonment refuse. As noted by Schiffer (1985), the identification and categorization of these different kinds of refuse, particularly con­cerning pottery, is difficult, but in the case of Erimi-Laonin tou Porakou it is possible to suggest a dis­ Disuse of spaces and discard of artefacts during the abandonment of Erimi-Laonin tou Porakou tinction based on the particular conservation and di­stribution of some assemblages of artefacts. In the workshop complex some de facto refuse, mainly re­presented by ground stone and vessels, was disco­vered. Despite being still usable and in good condi­tion, these items were still left in the buildings, sug­gesting functional motivations, such as difficulties in transport due to the large size and heavy weight of the artefacts, behind their abandonment (Villa­ni, Tripodi, under submission). At the same time, some rooms returned artefacts which it is possible to define as primary or secondary refuse (Figs. 2 and 7). They represent objects at or near the end of their use-life (Joyce, Johannessen 1993.138). The pres­ence of artefacts broken before the collapse of the structures (Webb 2017.187–191; Dionisio 2017.341– 343), modified or reused for a secondary function, such as some ground stone (see Webb 2017.214– 215; McCartney 2017.253–256), repaired (Dionisio 2017.336; see Dooijes, Nieuwenhuyse 2009) or with traces of macro use-wear (Bombardieri et al. 2017. 250) suggests that they were left in the proximity of their original positions or displaced and discard­ed because they were both easily replaceable and almost at the end of their usefulness (Tomka 1993. 22; see also Frankel, Webb 2006.226–232; Webb 2006). As noted by scholars, when secondary refuse had some potential value, it was usually concentrat­ed in provisional discard, until the reuse of the items or their displacement to a permanent dumping loca­tion (Joyce, Johannessen 1993.139; Deal 1985.253; see also Wilson 1994 and Tani 1995.234–235). Fre­quently the inhabitants of the settlement use aban­doned structures for this type of refuse (Seymour, Schiffer 1987.554). To conclude, the presence of primary and secondary refuse – e.g., a broken pithos recovered in unit SA IIa with the upper part in the centre of the room and pieces of the lower part near the north-western corner (Bombardieri 2017.45, Fig. 3.41), or a group of objects (in the small room SA X) which it is difficult to correlate with the place where it was found – suggests a particular function of some rooms during the gradual abandonment of the site. An in-depth description of the assemblages, and the possible identification of some artefacts as de facto, primary or secondary refuse, will be pro­vided in the following paragraphs. Artefacts from the workshop complex (area A) The objects found in the workshop units are terra­cotta vessels (in part destroyed by the collapse of the structures, as accelerated by the fire), ground stone tools and small objects like spindle whorls, worked and unworked ornamental artefacts made of picrolite, loom weights, bronze knives and chipped stones (Bombardieri 2017.219–250). Several studies have shown that the distance from the new settlement, the capability and means of transport, the artefacts’ size and weight and the pre­sence of repairs or damage, tend to influence the choice to leave an object (Schiffer 1976; Schlanger, Wilshusen 1993.91–92). In the case of Erimi-Lao­nin tou Porakou, some of the above motivations can justify the abandonment of the larger vessels and of the pithoi found in situ, often supported by stone emplacements (like in units I, IIa, III, IV and XII). A similar explanation can be applied for the querns (discovered in burned and unburned units) and for the heavy or easily replaceable stone tools in general. At the same time, the presence of a num­ber of transportable objects such as ceramic vessels (which in some cases have few signs of use-wear and do not show repairs or breakages), spindle whorls, bronze blades (two of which have been re­covered, of which one is broken and the other al­most complete), loom weights and worked picrolite artefacts (in particular, two pendants), suggest the possible decision to abandon some still usable arte-facts before setting fire to the area. The units con­taining concentrations of artefacts were totally or partially destroyed by the fire, while the others re­turned few objects and appear to have been partial­ly dismantled and cleaned out (Fig. 3). The ten exca­ Andrea Villani vated units of the workshop complex are characte­rised by a similar layout, since they were carved in the bedrock and share walls with the adjacent buil­dings. The result is an area with contiguous roofs, connect­ed as if they were only one. In this respect, it would be very hard to stop the burning of the roofs to li­mit the fire only to those units where the artefacts were concentrated in a phase of use of the entire area. In addition, stratigraphic analyses show evi­dence for the presence of possible intentional igni­tion points employed to feed the flames in units SA I and SA III (Bombardieri 2017.276), while a speci­fic treatment seems to have been reserved for unit SA V, where the fire started from the roof (Amadio et al. 2021.311–313). The unburned units The outermost units investigated in area A, name­ly units SA VI, SA IX, SA X and SA XI, are those that partially surrounded the rooms affected by the fire. When compared with the burned spaces, these rooms have yielded few artefacts and no pithoi were found on the floor, while in SA X only a spindle whorl was discovered. Scholars have hypothesised that in case of planned and gradual abandonment the areas near living spaces could become discards, and that when people know that a place will be left they tend to reduce the standard of cleaning (Schiffer 1985.25; Stevenson 1982.247–248, 252). Unit SA VI (6.5m x 3m) is the easternmost roofed space in the south­ern wing of the workshop complex and has the same NE–SW orientation as the adjacent units, but it was carved at a lesser depth of 0.50m into the li­mestone bedrock (Bombardieri 2017.52). This room yielded a small number of fragmentary artefacts, two medium-sized Red Polished (RP) jugs, two Drab Polished (DP) juglets, a small Red Polished (RP) bowl, a Coarse Ware (CW) mealing bin, a rubber and four chipped stone tools (Bombardieri 2017.55). Two fragmentary gaming stones, as noted by Luca Bombardieri (2017.56), were discarded in a previ­ous occupation phase and reused as building mate­rial. The access from the north, the presence of few objects, the absence of pithoi or big storage vessels and its multiple phases of construction and occupa­tion, suggest that the use of this space has proba­bly been changed several times (L.c.). In the conti­guous unit SA I, which shares the perimeter walls with SA VI and SA II, three possible ignition points were found. It is therefore arguable that, if a fire oc­curred during the last phase of use of SA VI, its roof would be affected by the fire. However, the absence of a burnt layer – as recovered in the adjacent unit – and the few artefacts found on the floor, make it possible to hypothesise that this space was disman­tled and consequently disused before the deliberate burning of SA I. SA IX is a small annex of SA V. In this unit only fragments of vessels and ground stone tools were found, and no evidence of de facto re­fuse can be observed. The room appears to have been filled with clay, while in the last phase of use a small structure has been built near the eastern limit, on the top of a pre-existent basin. This feature has probably been disused and filled before the burning. Indeed, no ashy deposits have been no­ticed in the filling of the basin, which additionally appears to have not been covered by the layer of plaster which constitutes the floor of SA IX. Worth noting is the case of unit SA X, where the presence of different layers of limestone blocks associated with ceramic sherds could confirm its intentional filling and disuse in an early stage of the abandon­ment phase. In summary, these units returned few artefacts, most­ly interpretable as primary or secondary refuse, and the proximity to the burned rooms suggests the loss of their function at the beginning of the workshop’s dismantling. Having presented the overall picture, the next section analyses in detail the disuse of the unburned unit SA X. Unit SA X Unit SA X is a small space (3.35m x 3.10m) located in the north-western part of the workshop complex, Disuse of spaces and discard of artefacts during the abandonment of Erimi-Laonin tou Porakou lacking a well-defined entrance and interpreted as an annex of SA V. The stratigraphic sequence was characterised by the presence of three different la­yers of limestone blocks (more than 130 of various sizes) associated with ceramic sherds (most of them are bowls with less than 1/3 preserved, and found in all the filling layers) suggesting a deliberate depo­sition, especially if compared with the lower num­ber of limestone blocks discovered in the largest units of the same area (Fig. 4). At the same time, the unit does not show traces of ashy layers, which would have surely been identified if the burning of SA V occurred during the use of SA X. The assem­blage found on the floor was composed of two par­tially worked picrolites, a spindle whorl, two frag­mentary vessels, three chipped stones, a quern (at an early stage of use or not used), a hammer and more than 300 ceramic sherds. It is difficult to estab­lish a priori that all these objects belong to a spe­cific category of refuse, but the small dimensions of the space, the absence of a defined entranceway, the heterogeneity of the assemblage and its casual dispo­sition on the floor, suggest an unspecialized func­tion. It is therefore possible that objects found on the floor are secondary refuse discarded in the unit when it was disused. It is interesting to note that SA X returned one of the highest dimension/sherds ratios of the whole work­shop (c. 35 sherds for m2), but only few attributable to the same vessels. These data and the presence of sherds in all the filling layers suggest that the con­centration could be related to the practice of clean­ing the units still in use, and when people knew that the site would be left and there was no need to re­deposit the refuse elsewhere it accumulated here. This hypothesis could support its final treatment as a discard, particularly for building materials. The burned units At Erimi-Laonin tou Porakou, major concentrations of artefacts were found in the larger units SA I, SA II (in both sub-units, A and B), SA III, SA IV, SA V and SA XII. Only five of the ten rooms investigated in the workshop complex – units SA I, SA III, SA IV, SA V and SA XII – were directly affected by a fire, which partially reached the contiguous unit SA II (Bombar-dieri 2017.46). This first macro evidence suggests a gradual spatial reduction in the use of the workshop, with a pro­gressive transformation of some units into discard rooms destined to be filled with no longer usable artefacts. Therefore, if some objects appear to have been discarded because they had been consumed, exhausted, damaged or were easily replaceable, still usable objects seem to have been intentionally con­centrated in the room whole or partially affected by the fire. These units yielded the highest number of portable artefacts, some of them in good condition, such as the pendants from SA I, SA IIb and SA XII, the spindle whorls recovered from all the units excluding SA IIa, the bronze blade from SA XII, and numerous loom weights and some vessels found in this area. Interestingly, not all these objects can be attributed to the common typologies, as in the case of the decorated RP biconical spindle whorl (Bom- bardieri et al. 2017.219–236) and the RP goat-shaped askos from unit SA III (Bombardie-ri 2017.49). Among the burn­ed units, only one returned a rich assemblage without showing the massive ash la­yer or the presence of possi­ble intentional ignition points observed in some of the other rooms. This peculiar situation is represented by unit SA II, the assemblage and stratigra­phic data of which will be ana­lysed in the following section of this paper, with the aim to define its possible (dis-)use and final treatment during the abandonment phase of the workshop area. Andrea Villani Unit SA II Unit SA II, located in the southern part of the work­shop, was underwent a change in layout during phase A, when the position of the opening which connects the two sub-units (named SA IIa and SA IIb) was modified (Bombardieri 2017.38–39) (Fig. 5). Some easily transportable and still usable arte-facts were found in the smallest SA IIb (Bombardie-ri 2017.43–46). Particularly intriguing for the ana­lysis of the abandonment discard processes is the larger sub-unit SA IIa. The stratigraphic analysis revealed the absence of possible ignition points, in contrast to the contigu­ous SA I and SA III, while an ash layer was record­ed only near the perimeter of the room with a par­ticular concentration along the east and south walls (Bombardieri 2017.46). The assemblage discovered on the floor of SA IIa was composed of a high num­ber of artefacts (35 from phase A; Fig. 6). The state of preservation of most of these objects is funda­mental to define the role that this space played dur­ing the abandonment phase of the settlement. Re­gardless of the preservation of vessels – which in this work includes all the diagnostics fragments without distinction, from almost complete to not restorable – the majority of them can be ascribed to the more common ceramic classes found at the site (see Webb 2017). As noted by scholars, local items are more likely to be left at the time of the abandon­ment (Tomka 1993.22; Joyce, Johannessen 1993. 150). Contemporaneously, they present a bad state of pre­servation: the RP large bowl, for example, is the only almost complete vessel found at the settlement with a repair (Dionisio 2017.336), while the large pithos located near the northern limit of the unit was found crushed on the floor, with the upper part unbroken in the centre of the room (Fig. 7). More­over, the DP spouted jug was broken in antiquity, and presents a small hole of 5mm on the body (Bombardieri 2017.41; Webb 2017.187– 191), the only almost complete vessel – despite a high level of fragmentation (Dionisio 2017.341, Fig. 15.15) – found in the room without repair or breakage is the amphora (Webb 2017.197). Other vessels are com­posed of a few fragments, such as a mealing bin in Coarse Ware or a cooking pot. Residual artefacts are represented by three picrolites (one broken, one un­finished and one with signs of macro use-wear from a post-phase A layer), a broken axe (a third preserv­ed), a broken, very heavy pounder (1.9kg), a broken chipped stone and two reused objects (a broken pes­tle transformed into a pounder with extensive use damage, and a broken perforator made from a re­used gloss crescent) (Bombardieri 2017.43). Fur­thermore, the unit yielded one of the only two bronze knives discovered in the settlement, broken near the junction between the blade and the han­dle (Bombardieri et al. 2017.247). Interestingly, one of the more representative artefacts of Erimi, that is to say spindle whorls, was not found in unit SA IIa. The room returned few partially complete artefacts. Among them, only a complete quern was described as a de facto refuse “in storage or active use into the unit” (Webb 2017.208). Contemporaneously the ground stone shows various traces of use-wear (see Fig. 7; see also Frankel, Webb 2006.227) and, if we Fig. 5. The plan of Unit SA IIa with the location of the majority of the artefacts (readapted from Bom­bardieri 2017.44, Fig. 3.37). Disuse of spaces and discard of artefacts during the abandonment of Erimi-Laonin tou Porakou Artefact Broken Secondary use\ Macro Fragmentary Partially complete\ Inventory unfinished\repaired use-wear complete Amphora x A.423.19b Axe x x A.447.2 Bronze blade x x A.423.1 Cooking pot x A.423.16 Cooking pot x A.423.14 Cooking pot x A.423.11 Core trimming element x A.434.3 DP amphora x x A.423.7 DP amphora x A.423.6 DP hemispherical bowl x A.434.6 DP jug x x A.423.9 DP spouted jug x x A.434.8 Flake\blade bladelet x x A.423.3 Jug or juglet x A.427.16 Jug or juglet x A.423.19 Jug or tankard x A.508.1 Mealing bin x - Medium bowl x A.427.7 Perforator x x x A.427.14 Pestle x x x x A.422.2 Picrolite x x A.434.7 Picrolite x x x A.434.5 Picrolite x x A.342.95 Pounder x x A.423.2 Pyxides x A.427.11 Quern x A.447.1 RP decorated jug x A.516.17 RP hemispherical bowl x A.423.5 RP large bowl x x x A.516.18 RP large pithos x x A.448.11 RP storage jar x A.434.15 Small bowl x A.427.13 Spouted jug or tankard x A.427.12 Spouted jug or tankard x A.427.10 Tool (Piéce esquillee) x x 434.8 Fig. 6. Artefacts discovered inside unit SA IIa. hypothesise that the room lost its function during the gradual abandonment, it could represent prima­ry refuse left in the place where it was used. Almost all the other objects are fragmentary and show da­mage or macro use-wear, while the well-preserved bowl repaired in antiquity found at Erimi was reco­vered in this room. In addition, residual artefacts were found concentrated in the northern part of the unit and in the south-eastern corner behind the door­way. In the workshop complex, the only objects discovered near a threshold belong to unit SA IV (Amadio et al. 2021.316), where four piled and complete vessels were found. In contrast, the ves­sels, the broken bronze knife and the heavy pounder recovered behind the entrance of SA IIa do not ap­pear to be well preserved or placed in an organised manner. Moreover, a bin (Ft. 9; see Fig. 5) was built close to the entry that connects SA IIb and SA IIa during phase A, making it more difficult to move between these spaces. The presence of residual ar­tefacts, the absence of possible ignition points and the construction of the bin seem to suggest the gra­dual disuse and transformation of this unit as a pro­visional discard. Almost all the objects concentrated in SA IIa could be then interpreted as primary and secondary refuse. At the site the reuse of broken or exhausted artefacts is attested by the presence of ground stones and gaming stones in the walls, by the reuse of a quern as a support for a mezzanine in SA III (Ft. 59) (Bombardieri 2017.51, Figs. 3.53, 3.54), and by the modified neck of an RP jug insert­ed in another jug neck and utilised in a double-chambered hearth (Bombardieri 2017.36–37). As stated by Michael Deal (1985), when the abandon­ment of a building complexpeople is foreseen and planned, refuse is accumulated in areas usually kept free and not concentrated into the generally desig­nated areas. At the same time, Marc G. Stevenson Andrea Villani (1982.248) noted that in case of gradual abandonment re­fuse within enclosed living area “will be more abundant and distributed in a more clustered or orderly man­ner”. Some scholars have also defined provisional discard as “the intentional storage of damaged or fragmented items for future disposal” or reuse that had high probabi­lity of being left at sites after abandonment (Deal 1985; Ca­meron, Tomka 1993; Joyce, Johannssen 1993; see also Frankel, Webb 2006.153– 154). The decision to concen­trate these objects in a single unit is thus possibly related to their potential value for re­use. In addition, Deal (1985. 263; see also La Motta, Schiffer 1999.21–22) notes that “if a structure was slowly being dismantled within an ongoing household it might become a dumping location for large inorganic items, and especially pottery”. In the chart, objects defined ‘fragmentary’ are arte-facts for which only as most as a third of their en­tire body has been found (Fig. 8). It is important to note that the majority of these consist of only few fragments. In the category ‘par­tially complete (with breakage or macro use-wear)’ we have included vessels and other objects inter­preted as possibly broken before the collapse of the structures (e.g., the pithos, the axe, one of the pic-rolite items or the bronze knife). It is interesting to note that only 14 of the 35 artefacts are complete or partially complete, of which only three are not cha­racterised by breakage or secondary use. As men­tioned above, in unit SA IIa burned deposits seem to occur mainly on the east and south walls and only along the perimeter of the room (Bombardieri 2017.46). The absence of a layer of ash, indicating the collapse of the roof, and the discovery of the intact upper part of the pithos in the centre of the room seem to indicate a partial dismantling of the roof prior to the fire. In this respect, Swiny inter­preted the presence of collapsed roofs at Sotira-Ka­minoudhia as further evidence of its rapid aban­donment, caused by an earthquake, noting that ser­viceable beams are retrieved when structures are voluntarily abandoned (Swiny 2003.53; see also Horne 1993; Schlanger, Wilshusen 1993.90–95; Ste­venson 1982) (Fig. 9). Moreover, unit SA IIa seems to share disuse proces­ses similar to those seen in the unburned units SA X (dismantled and filled with stones) and SA IX (filled with clay). In SA IIa the abandonment of the quern is possibly related to its weight, as confirmed by the presence of heavy querns and other ground stone tools left or discarded on the floors of other units of the workshop (e.g., SA III, SA IV, SA X and SA XII). The large dimensions and difficulties in transport could also explain the decision to leave the largest RP decorated jug, while the discovery of the pithos deserves a more detailed examination. In contrast to Disuse of spaces and discard of artefacts during the abandonment of Erimi-Laonin tou Porakou what is observed for other units (Bombardieri 2017.53, Fig. 3.57), this vessel does not appear to have been crushed in situ by the collapse of the roof. Its upper part, in fact, was found separated from the body in the middle of the room, suggesting that it was already broken before the unit’s collapse. If, in other units, most of the pithoi represent de facto refuse still usable until the destruction caused by the fire, the pithos from SA IIa could be primary refuse, crushed on the floor and abandoned before the collapse of the unit. As suggested by Muti (2020. 200), pithoi could be moved from their location to get rid of their contents, and the particular breakage and fragmentation of the pithos in SA IIa suggest that it was broken upon such displacement (Bombar-dieri 2017.45, Fig. 3.41). Smaller and easily movable objects would probably have been curated in SA IIa on the basis of different criteria. On the one hand their bad state of preser­vation would make them not useful in the new set­tlement, on the other they would have been still usable in case of extreme necessity. At the same time, the absence of an ashy layer homogeneously distri­buted inside the room can suggest a partial disman­tlement of the roof. The absence of beams, in fact, could force the creation of possible ignition points in SA I and SA III aimed to burn and destroy the two non-contiguous spaces. It is impossible to define if the ignition points were created to burn the rooms or if they represent a concentration of wooden ma­terials which were destroyed by the fire. The ab­sence of wood inside unit SA IIa could be related to the displacement of usable materials in rooms in use until the end of the site occupation. This space provided further elements in support of the planned abandonment without anticipated return. In the case of a planned return, the discard of objects would not have been occurred in the interior of one of the units and, as noted by Stevenson “less refuse would be discarded within enclosed living areas” (Stevenson 1982.260). Finally, it is important to note that SA IIb, the smallest unit (6m2) of the work­shop complex (Bombardieri 2017.38), returned portable artefacts that were still usable (specifically a pendant, six spindle whorls and some small ves­sels) and does not show traces of massive ash la­yer, as well as SA IIa. The concentration of burned debris near the partition wall between SA IIa and SA IIb and the massive presence of artefacts sug­gest that a particular final treatment was reserved for this unit. The absence of an ash layer could be due to the room’s layout, since the unit has partial­ly been carved in the bedrock (approx. 0.30m), pos­sibly facilitating the dispersion of ash. In addition, some of the objects appear to have been destroyed by the collapse of walls and roof. The limited con­centration of ash near the southern limit of the unit (Bombardieri 2017.46), close to the wall that di­vided SA IIa from SA IIb, might be related to the pre­sence of a roof only in the smaller sub-unit IIb at the time of the fire. The residential area Particularly interesting for the abandonment proces­ses is area B, which brought to light a complex se­quence of changes in the layout of some domestic units (Fig. 10). The last sub-phase A1 is characterised by a reduction in terms of use of space, with the dismantling of unit 7 and the construction of a wall to close the passage to unit 6 (Bombardieri 2017.70). As in the case of Marki-Alonia, where Frankel and Webb (2006; 2012. 488) identified nine different phases, it is possible to define changes, renovations and modifications in the layout of the rooms. Since the reduction of di­mensions occurred during the last sub-phase, the re­sidential area seems to have experienced a gradual Andrea Villani and planned disuse, as noted for the Erimi work­shop. At the same time, the excavation of area T2 re­vealed a possible ‘atypical’ disuse practice reserved for two contiguous doorways that give access to two distinct units. During the abandonment of these spa­ces, the thresholds would have been surmounted by monolithic blocks put on the short tight side, pos­sibly testifying to the symbolic isolation of the inter­nal from the external space of the building (see Gar­wood 2011). The peculiar disposition of the mono­lithic blocks suggests an action more related to an abandonment behaviour than a change in the layout of the spaces. The blocks, in fact, are isolated and not included in a possible new structure, while the pas­sage in front of the thresholds was not closed. A si­milar practice appears to be represented by the threshold D3 of unit 3 in area B which, during the passage from phase A2 to A1, changed its use and was transformed into a large square basin. In this case, the decision seems to suggest a change in the layout of the unit and not its final abandonment (Bombardieri 2017.65–70). Abandonment processes at Erimi-Laonin tou Porakou This paper has highlighted the contribution that a cross analysis of the stratigraphy and artefacts’ as­semblages could provide to explain the treatment reserved for the rooms of the workshop area during the gradual abandonment of Erimi-Laonin tou Po-rakou. Ethnoarchaeological studies have shown that when the process is gradual and planned, abandon­ment often occurs through a differential treatment of the spaces (Rothschild et al. 1993). The site was then partially disused and dismantled, and the in­habitants selected and concentrated still usable ar­tefacts in units that were finally set on fire. Studies on discard processes in Bronze Age Cyprus note that part of the households or the alleys were used to concentrate refuse or to discard obsolete objects (Frankel, Webb 2006; Falconer, Fall 2014.174– 176). Due to the particular layout of the workshop, with most rooms not connected by internal passa­ges but only through open spaces, it is possible to suggest that some units lost their productive role and were gradually dismantled, while SA IIa be­came a sort of provisional discard, a place to concen­trate objects not useful but possibly reusable in case of extreme necessity. As noted by Frankel and Webb in the case of Marki-Alonia, vessels with flaws have been “maintained for use in a limited or secon­dary capacity or in provisional discard awaiting mending, reuse or removal” (Frankel, Webb 2006. 153). Unburned units would have been disused and partially dismantled, while in some burned rooms a set of artefacts in still usable condition has been curated. This hypothesis can also be confirmed by the comparison of the ratio between entire vessels and sherds. As noted by some scholars, rooms aban­doned in an earlier phase tend to show the pres­ence of a high number of sherds but few complete artefacts. In contrast, spaces which were disused in a later phase seem to show the opposite (Montgo­mery 1993.157–159; see also Schiffer 1985.23). In this respect, no well-preserved vessels were found in SA X, but it returned one of the highest propor­tions of sherds found in the workshop. In addition, unburned spaces seem to be related to unspecialized functions due to their smaller size and the limited presence of features, an aspect that could justify their final treatment and the exclusion from the pos­sible deliberate burning (see Joyce, Johannessen 1993.151). Disuse of spaces and discard of artefacts during the abandonment of Erimi-Laonin tou Porakou At the workshop of Erimi, the gradual and differen­tiated treatment of spaces could reflect the decision to intentionally seal some rooms as part of the aban­donment strategy. The deposition of some objects, their location and concentration are markers of spe­cific choices, often related to the burial of members of the community or, in other cases, to possible de­tachment rituals before leaving structures or settle­ments. As demonstrated by several ethnoarchaeo-logical studies and experimental archaeologists, a structure could burn for days before collapsing (Ca­meron 1991.162; Gheorghiu 2019.30–47; see also Stevanovi. 1997), and the inhabitants would have had the possibility to recover at least some objects. At Erimi, however, the absence of any attempt to re­trieve artefacts and the presence of possible inten­tional ignition points seem to confirm the decision to voluntarily seal the objects inside the units. At the same time, different studies have highlighted how the distance from the new settlement influences the decision to take or leave an object (Schlanger, Wils­husen 1993.90–91; Schiffer 1985.26–33). Taking into account the empty units of Erimi alone, we could hypothesise that the new site could have been in close proximity, entailing the transport of most of the artefacts to the new settlement. If we exclude possible ritual or symbolic practices, the assemblage of the burned units suggests the opposite scenario, a longer distance and the abandonment of artefacts caused by functional motivations. Deeper analysis on almost complete or fragmentary artefacts – and particularly vessels – will be fundamental to define the real quantity of still usable objects left in the site. Nevertheless, the sequence of the gradual abandon­ment processes, from the reduction and disuse of some spaces to the burning of part of the workshop, seems to confirm that the selection, location and con­centration of some portable and often still usable artefacts depended on specific and voluntary deci­sions, influenced by different motivations. Scholars have often defined abandonment as a strategy fol­lowing practice determined by local populations, as well as local and regional conditions (Schlanger, Wilshusen 1993.85; see also Nelson, Hegmon 2001. 213). The conditions in which the site was aban­doned and the possible foundation of a close, new settlement seems therefore to be ascribed to a wider social phenomenon that involved the island of Cyp­rus at the end of the Middle Bronze Age. Suggesting a gradual and pre-planned abandonment in the case of the Bronze Age site Marki-Alonia, Frankel and Webb (2006.153) stressed how some rooms re­turned complete or restorable vessels while in oth­ers the absence of artefacts could depend on diffe­rent motivations, from the curation to episodes of renovation, which could have affected some rooms. At Kissonerga-Skalia Crewe noted a decrease in the activities that would have produced a slow abandon­ment of the site (Crewe 2017.146), while the pres­ence of intact artefacts and skeletal remains at So-tira-Kaminoudhia has been interpreted as a rapid abandonment caused by a seismic event (Swiny 2003.53–54). Even at Kalopshida, the presence of a burned layer in the house of trench 3 and in the room 7 of site C combined with the rich assemblage found on the floor suggests a rapid abandonment (Aström 1966.139–140; Gjerstad 1926; Webb 2012. 52), while Swiny noted that the absence of certain typologies of objects at Episkopi-Phaneromeni could be related to curating behaviours due to the short distance from the settlement G to A (Swiny 1979. 330). At the same time many artefacts were left in situ when settlement G was abandoned (Swiny 1979.15). Therefore, excluding cases of rapid and not anticipated abandonments, the choice to leave the settlement through a planned and gradual pro­cess could have been based on social and economic motivations. At Erimi, finally, the voluntary destruc­tion of part of the workshop and the particular lo­cation of the partially complete artefacts in some units, seem to reflect the specific way in which the inhabitants decided to detach from their place (see Lamoreux-St-Hilaire, Macrae 2020), performing a voluntary closure (see Adams 2016; Cameron 1990). Conclusions The data presented in this paper, combining an ana­lysis of artefact refuse with the study of the abandon­ment practices performed at the MBA site of Erimi-Laonin tou Porakou, provide interesting informa­tion. Some spaces were cleaned out and the objects were probably progressively put in the compounds still in use and then taken away at the end of a gra­dual abandonment. Some units were disused, parti­ally dismantled and transformed in discard areas; contemporaneously, some artefacts were concen­trated inside the rooms then possibly set on fire. The workshop has likely been impacted by a series of symbolic practices, such as the decision to leave some artefacts inside the units as well as the depo­sition of the piled vessels in front of the entrance of SA IV, the concentration and location of artefacts inside the burned units, and the absence of attempts to recover the smaller objects during the fire. Final­ly, if in the rooms it is possible to distinguish arte­ Andrea Villani ACKNOWLEDGEMENTS facts as de facto, primary and secondary refuse, the differences in preservation, concentration and se­lection of artefacts from some burned units suggest that in some cases objects entered in the archaeolo­gical context possibly through symbolic depositions made before the burn. Elsewhere in Cyprus, the pre­sence of artefacts has been interpreted as the result of a rapid process caused by natural disasters. In contrast, the abandonment practices performed by the Erimi’s inhabitants resulted in the possible depo­sition of some artefacts, clearly suggesting some de-cision-making in this process. I want to particularly thank Dr. Giulia Muti and Giu­lia Albertazzi for their fundamental support and the multiple revisions of the paper. Thanks are also due to the anonymous reviewers for their helpful com­ments. This paper could not have been achieved with­out the support of prof. Luca Bombardieri and of the Erimi Archaeological Project. This paper is part of a PhD research project conducted at the University of Balearic Islands and I am indebted to my supervisor, prof. Manuel Calvo Trias, for his guidance. . References Adams E. C. 2016. Closure and dedication practices in the Homol’ovi settlement cluster, Northeastern Arizona. American Antiquity 81(1): 42–57. https://doi.org/10.7183/0002-7316.81.1.42 Amadio M., Boaretto E., and Bombardieri L. 2021. Aban­donment practices through the microscope lens. Micro-archaeological data from Middle Bronze Age Erimi, Cy­prus. Levant. The Journal of the Council for British Research in the Levant. Volume 52-3: 301–320. https://doi.org/10.1080/00758914.2021.1890400 Aström P. 1966. Excavations at Kalopsidha and Ayios Iakovos in Cyprus. Studies in Mediterranean Archaeology 2. Lund. Bombardieri L. 2017. Erimi Laonin tou Porakou. A Mid­dle Bronze Age Community in Cyprus. Excavations 2009–2014. Studies in Mediterranean Archaeology 145. Aström Förlag. Uppsala. Bombardieri L., Fissore M., and Muti G. 2017. Small finds. In L. Bombardieri (ed.), Erimi Laonin tou Porakou. A Middle Bronze Age Community in Cyprus. Excavations 2009–2014. Studies in Mediterranean Archaeology 145. Aström Förlag. Uppsala: 219–250. Bombardieri L., Menozzi O., Fossataro D., and Jasink A. M. 2009. The Kouris Valley Survey Project: 2008 preliminary report. Report of the Department of Antiquities. Cyprus: 117–129. Cameron C. M. 1990. Pit structure abandonment in the Four Corners Region of the American Southwest: Late Basketmaker III and Pueblo I Periods. Journal of Field Archaeology 17(1): 27–37. https://doi.org/10.1179/009346990791548565 1991. Structures abandonment in villages. In M. B. Schif-fer (ed.), Archaeological Methods and Theory 3. Uni­versity of Arizona Press. Tucson: 155–194. Cameron C. M., Tomka S. A. 1993. The Abandonment of Settlements and Regions. Ethnoarchaeological and Ar­chaeological Approaches. Cambridge University Press. Cambridge. Chapman J. 1999. Deliberate house-burning in the pre­history of Central and Eastern Europe. In A. Gustafsson, H. Karlsson (eds.), Glyfer och arkeologiska rum-en vän­bok till Jarl Nordbland. Gotarc Series A 3. University of Göteborg Press. Göteborg: 113–126. Christofi P., Stefani E., and Bombardieri L. 2015. Bridging the gap: long-term use and re-use of the Bronze Age fune­rary area of Ypsonas-Vounaros and Erimi-Laonin tou Po-rakou. In H. Matthaus, B. Morstadt, and C. Vonhoff (eds.), PoCA (Postgraduate Cypriot Archaeology) 2012. Cam­bridge Scholars Publishing. Cambridge: 130–169. Crewe L. 2017. Interpreting settlement function and scale during MC III–LC IA using old excavations and new: West­ern Cyprus and Kisonerga (Kissonerga) Skalia in context. In D. Pilides, M. Mina (eds.), Four Decades of Hiatus in Archaeological Research in Cyprus: Towards Restoring the Balance. Proceedings of the International One-Day Workshop, Held in Lefkosia (Nicosia) on 24th September 2016, Hosted by the Department of Antiquities, Cyprus. Holzhausen Verlag. Wien: 140–153. Deal M. 1985. Household pottery disposal in the Maya Highlands: an ethnoarchaeological interpretation. Jour­nal of Anthropological Archaeology 4: 243–291. https://doi.org/10.1016/0278-4165(85)90008-X Disuse of spaces and discard of artefacts during the abandonment of Erimi-Laonin tou Porakou Dionisio G. 2017. Conservation and restoration proce­dures for archaeological artefacts: the Erimi conservation laboratory. In L. Bombardieri (ed.), Erimi Laonin tou Po-rakou. A Middle Bronze Age Community in Cyprus. Stu­dies in Mediterranean Archaeology 145. Aström Förlag. Uppsala: 335–345. Dooijes R., Nieuwenhuyse O. P. 2009. Ancient repairs in archaeological research; a Near Eastern perspective. In J. Ambers, C. Higgitt, L. Harrison, and D. Saunders (eds.), Holding it all together, Ancient and Modern Approaches to Joining, Repair and Consolidation. Archetype Publi­cation. London: 8–13. Falconer S. E., Fall P. L. 2014. The meaning of space and place in Middle Cypriot communities. In J. M. Webb (ed.), Structure, Measurement and Meaning: Studies on Pre­historic Cyprus in Honour of David Frankel. Studies in Mediterranean Archaeology 143. Aström Förlag. Uppsala: 175–184. Frankel D., Webb J. M. 2006. Marki Alonia: An Early and Middle Bronze Age Settlement in Cyprus: Excava­tions 1995–2000. Studies in Mediterranean Archaeology 123.2. Aströms Förlag. Sävedalen. 2012. Household continuity and transformation in a prehistoric Cypriot village. In B. J. Parker, C. P. Foster (eds.), New Perspectives in Household Archaeology. Eisenbrauns. Winona Lake: 473–500. Garwood P. 2011. Rites of passage. In T. Insoll (ed.), The Oxford Handbook of the Archaeology of Ritual and Re­ligion. Oxford University Press. Oxford: 262–284. https:// doi.org/10.1093/oxfordhb/9780199232444. 001.0001 Gheorghiu D. 2019. Architectures of Fire: Processes, Space and Agency in Pyrotechnologies. Archaeopress. Oxford. Gjerstad E. 1926. Studies on Prehistoric Cyprus. Univer­sitats Arsskrift. Uppsala. Hayden M. 2000. Site formation processes at Keatley Creek. In B. Hayden (ed.), The Ancient Past of Keatley Creek, Volume I. Archaeology Press Simon Fraser Univer­sity. Burnaby: 299–331. Horne L. 1993. Occupational and locational instability in arid land settlement. In C. M. Cameron, S. A. Tomka (eds.), Abandonment of Settlements and Regions: Ethnoarcha­eological and Archaeological Approaches. Cambridge University Press. Cambridge: 43–53. Iannone G. 1990. The Rise and Fall of Plateau House-pits: Site Formation Processes and Houspit Roof Depo­sits. Unpublished BA dissertation. Archaeology Depart­ment. Simon Fraser University. Burnaby. Joyce A. A., Johannessen S. 1993. Abandonment and the production of archaeological variability at domestic sites. In C. M. Cameron, S. A. Tomka (eds.), Abandonment of Settlements and Regions: Ethnoarchaeological and Ar­chaeological Approaches. Cambridge University Press. Cambridge: 138–153. La Motta V. M., Schiffer M. B., 1999. Formation processes of house floor assemblages. In P. M. Allison (ed.), The Ar­chaeology of Household Activities. Routledge. London: 19–29. Lamoreux-St-Hilaire M., Macrae S. 2020. Detachment from place: beyond an archaeology of settlement aban­donment. University Press of Colorado. Luisville. Lange F. W., Rydberg C. R. 1972. Abandonment and post-abandonment behavior at a rural Central American house-site. American Antiquity 37: 419–432. https://doi.org/10.2307/278440 Lightfoot R. R. 1993. Abandonment processes in prehis­toric Pueblos. In C. M. Cameron, S. A. Tomka (eds.), Aban­donment of Settlements and Regions: Ethnoarchaeolo­gical and Archaeological Approaches. Cambridge Uni­versity Press. Cambridge: 165–177. Lindskoug H. B. 2016. Fire events, violence and aban­donment scenarios in the ancient Andes: The final Stage of the Aguada Culture in the Ambato Valley, Northwest Argentina. Journal of World Prehistory 29: 155–214. https://doi.org/10.1007/s10963-016-9095-y McCartney C. 2017. The chipped stone. In L. Bombardi­eri (ed.), Erimi Laonin tou Porakou. A Middle Bronze Age Community in Cyprus. Studies in Mediterranean Ar­chaeology 145. Aström Förlag. Uppsala: 251–257. Montgomery B. K. 1993. Ceramic analysis as a tool for discovering processes of pueblo abandonment. In C. M. Cameron, S. A. Tomka (eds.), Abandonment of Settle­ments and Regions: Ethnoarchaeological and Archaeo­logical Approaches. Cambridge University Press. Cam­bridge: 157–164. Muti G. 2020. Tracing ancient textiles: production, con­sumption and social uses in Chalcolithic and Bronze Age Cyprus (2800–1450 BC). Unpublished PhD thesis. University of Manchester. Manchester. Nelson M. 2000. Abandonment: conceptualization, repre­sentation and, and social change. In M. B. Schiffer (ed.), Exploration in Social Theory. University of Utah Press. Salt Lake City: 52–62. Nelson M., Hegmon M. 2001. Abandonment is not as it seems: an approach to the relationship between site and Andrea Villani regional abandonment. American Antiquity 66(2): 213– 235. https://doi.org/10.2307/2694606 Rothschild N. A., Mills B. J., Ferguson T. J., and Dublin S. 1993. Abandonment at Zuni farming villages. In C. M. Ca­meron, S. A. Tomka (eds.), Abandonment of Settlements and Regions: Ethnoarchaeological and Archaeological Ap­proaches. Cambridge University Press. Cambridge: 123–137. Schiffer M. B. 1972. Archaeological context and systemic context. American Antiquity 37: 156–165. https://doi.org/10.2307/278203 1976. Behavioral Archaeology. Academic Press. New York. https://doi.org/10.2307/278948 1985. Is there a “Pompeii Premise” in archaeology? Journal of Anthropological Research 41: 18–41. https://doi.org/10.1086/jar.41.1.3630269 Schlanger S. H., Wilshusen R. H. 1993. Local abandon­ments and regional conditions in the North American Southwest. In C. M. Cameron, S. A. Tomka (eds.), Aban­donment of Settlements and Regions: Ethnoarchaeolo­gical and Archaeological Approaches. Cambridge Uni­versity Press. Cambridge: 85–98. Seymour D. J., Schiffer M. B. 1987. A preliminary analy­sis of Pithouse assemblages from Snaketown, Arizona. In S. Kent (ed.), Method and theory for activity area re­search, an ethnoarchaeological approach. Columbia Uni­versity Press. New York: 549–603. Stevanovi. M. 1997. The age of clay: the social dynamics of house destruction. Journal of Anthropological Archa­eology 16: 334–395. https://doi.org/10.1006/jaar.1997.0310 Stevenson M. G. 1982. Toward an understanding of site abandonment behavior: evidence from historical mining camps in the southwest Yukon. Journal of Anthropologi­cal Archaeology 1(3): 237–265. Swiny S. 1979. Southern Cyprus, 2000–1500 B.C. Un­published PhD thesis. University of London. London. https://discovery.ucl.ac.uk/id/eprint/1317750/ 2003. Sotira Kaminoudhia: An Early Bronze Age Site in Cyprus. American Schools of Oriental Research Ar­chaeological Reports 8. The American Schools of Orien­tal Research. Boston. Tani M. 1995. Beyond the identification of formation pro­cesses: behavioral inference based on traces left by cul­tural formation processes. Journal of Archaeological Me­thod and Theory 2(3): 231–252. https://doi.org/10.1007/BF02229008 Tomka S. A. 1993. Site abandonment behavior among transhumant agro-pastoralists: the effects of delayed cu-ration on assemblage composition. In C. M. Cameron, S. A. Tomka (eds.), Abandonment of Settlements and Regi­ons. Ethnoarchaeological and Archaeological Approa­ches. Cambridge University Press. Cambridge: 11–24. Tomka S. A., Stevenson M. G. 1993. Understanding aban­donment processes: summary and remaining concerns. In C. M. Cameron, S. A. Tomka (eds.), Abandonment of Set­tlements and Regions. Ethnoarchaeological and Archa­eological Approaches. Cambridge University Press. Cam­bridge: 191–195. Verhoeven M. 2000. Death, fire and abandonment: ritual practice at late Neolithic Tell Sabi Abyad, Syria. Archaeo­logical Dialogues 7(1): 46–65. https://doi.org/10.1017/S1380203800001598 Villani A., Tripodi P. Under submission. Leaving behind (or not). Abandonment practices and selected assembla­ges at Middle Bronze Age Erimi-Laonin tou Porakou. In C. Peverelli, P. Maillard (eds.), PoCA (Postgraduate Cypriot Archaeology). Basel. Webb J. M. 2006. A material culture and the value of con­text: a case study from Marki, Cyprus. In D. Papaconstan­tinou (ed.), Deconstructing Context: A Critical Approach to Archaeological Practice. Oxbow Books. Oxford: 98– 119. 2012. Kalopshida forty-six years after Studies in Medi­terranean Archaeology volume 2. In J. M. Webb, D. Fran­kel (eds.), Studies in Mediterranean Archaeology: Fifty Years On. Studies in Mediterranean Archaeology 137. Aström Förlag. Uppsala: 49–58. 2017. The pottery. In L. Bombardieri (ed.), Erimi La-onin tou Porakou. A Middle Bronze Age Community in Cyprus. Studies in Mediterranean Archaeology 145. Aström Förlag. Uppsala: 129–205. 2017. The ground stone. In L. Bombardieri (ed.), Erimi Laonin tou Porakou. A Middle Bronze Age Commu­nity in Cyprus. Studies in Mediterranean Archaeology 145. Aström Förlag. Uppsala: 207–218. Wilshusen R. H. 1986. The relationship between aban­donment mode and ritual use in Pueblo I Anasazi Proto­kivas. Journal of Field Archaeology 13: 245–254. https://doi.org/10.1179/jfa.1986.13.2.245 Wilson D. C. 1994. Identification and assessment of sec­ondary refuse aggregates. Journal of Archaeological Me­thod and Theory 1(1): 41–68. https://doi.org/10.1007/BF02229423 back to contents Documenta Praehistorica XLIX (2022) Zoning analysis of Iron Age sites using Analytic Hierarchical Process (AHP) methods in the Middle Atrak River Basin, Northeast of Iran Mohsen Heydari Dastenaei1, Mohsen Dana2 1 Department of Archeology, Shahid Chamran University of Ahvaz, Ahvaz, IR< M.Heydari@scu.ac.ir 2 Ministry of Cultural Heritage, Handicrafts and Tourism, Tehran, IR< mohsendana@gmail.com ABSTRACT – Iron Age settlements in the Middle Atrak Basin in Iran have a particular distribution pattern due to environmental, social, and economic variables, among which geographical factors play an essential role in creating and dispersing settlements. Some of these factors play a more ef­fective and stable role than others. The present study examines and evaluates the role of geogra­phical factors in the distribution of Iron Age sites to determine factors that have a more signifi­cant role than others. Moreover, the zoning map of the Middle Atrak Basin should be presented using four different types of location, grouped in terms of those with a perfectly suitable, relatively suitable, suitable, and unsuitable location. To achieve this goal, seven natural factors, including the distance of sites from the river, altitude, slope, slope direction, distance from communication routes, soil type, and land use, were selected as influential factors in choosing the location of the Iron Age sites. In this study operating maps were prepared digitally using ArcGIS, and then the weight of each index was determined using the AHP model. The results of this study show that 46.7% of the Iron Age set­tlements (or 28 sites) were located in a perfectly suitable environment and geography, 24 sites (29.3%) in a relatively suitable location, seven sites (11.4%) in a suitable place, and one site (1.6%) in a completely unsuitable environment. This last type of location in the region’s landscape indicates the choice of different livelihoods, including agriculture and animal husbandry with both seasonal and permanent methods. KEY WORDS – Iron Age sites zoning; Analytic Hierarchical Process (AHP); Middle Atrak Basin; Northeast of Iran Analiza coniranja /eleznodobnih naselbin z metodami analiti;no hierarhi;nega procesa (AHP) v pore;ju srednjega Atraka, severovzhodni Iran IZVLE.EK – .eleznodobne naselbine v pore.ju srednjega Atraka v Iranu imajo poseben vzorec umes­titve zaradi okoljskih, socialnih in ekonomskih spremenljivk, med katerimi igrajo geografski dejav­niki klju.no vlogo pri nastanku in .iritvi naselij. Nekateri od teh dejavnikov so bolj u.inkoviti in traj­nej.i od drugih. V .tudiji predstavljamo najpomembnej.e geografske dejavnike pri porazdelitvi naj­di... Karto poselitvenih obmo.ij srednjega Atraka predstavljamo s pomo.jo .tirih tipov lokacij, raz­vr..enih glede na popolnoma primerne, manj primerne, primerne in neprimerne pogoje poselitve. Pri tem smo kot klju.ne izbrali sedem naravnih dejavnikov: oddaljenost najdi..a od reke, nadmor-ska vi.ina, strmina, smer pobo.ja, oddaljenost od komunikacijskih poti, vrsta tal in raba tal. Digital-ne karte smo izdelali s pomo.jo programa ArcGIS in obte.ili z indeksi, ki jih dolo.a model AHP. Re-zultati ka.ejo, da je 46,7 % .eleznodobnih naselbin (28 lokacij) ume..enih v povsem primeren, 24 (29,3 %) v manj primeren, sedem (11,4 %) v primeren in ena (1,6 %) v popolnoma neprimeren pro-stor. Zadnji tip lokacije ka.e na izbiro razli.nih vrst pre.ivljanja, tako kmetijstva in .ivinoreje s stal-no ter sezonsko poselitvijo. KLJU.NE BESEDE – coniranje .eleznodobnih najdi..; analiti.ni hierarhi.ni proces (AHP); srednji Atrak; severovzhodni Iran DOI> 10.4312\dp.49.8 Zoning analysis of Iron Age sites using Analytic Hierarchical Process (AHP) methods in the Middle Atrak River Basin, Northeast of Iran Introduction Archaeological findings show that the development and evolution of past human settlements are closely related to the substrate of the natural and social en­vironment. Environmental and natural substrates create the necessary conditions for establishing set­tlements, and some create more stable conditions than others. These natural substrates are each re­gion’s slope, altitude, geological structure, water re­sources, and climate. Each of these factors, both in­dividually and in relation to each other, shows dif­ferences. The existence of such differences causes the characteristics of different regions (Gholami Rad, Wali Shariatpanah 2013.56). Humans have long tried to settle in the natural environment in such a way that makes the best use of it. In other words, human settlements act as the most basic link between man and the Earth, and reflect human inte­ractions with the environment (Zhang et al. 2014). Therefore, ancient societies lived in places that had favourable conditions for life and development – with environmental factors such as rivers, commu­nication routes, and beds of deltas and river terra­ces along with foothills or mineral resources – which provided them with raw materials and the possibi­lity of protection against enemies (Maga. et al. 2021. 21). In addition to these cases, various other factors and forces are involved in the location and forma­tion of rural settlements, which should be consid­ered in any location of settlements. Although the ef­fects of these factors and forces depends to a great extent on the underlying characteristics of the en­vironmental substrate and ecological structures (Zhang et al. 2014.2818), the primary stimulus in this process is the set of motivations that arise to meet basic needs, and the forms of various funda­mental demands among different human groups. As such, different forms and varieties of locations, and the locating of human settlements in certain places because of the demands and motivations they are able to satisfy, are realized in different ways. As a re­sult, settlements are structurally and functionally dif­ferent from one area to another (Rahimi, Hassan-pour 2013.14). For example, settlements formed in hilly areas are more affected by natural factors such as altitude, slope, and slope direction. In contrast, settlements formed in lowland areas are more af­fected by human factors such as communication rou­tes and transportation, surface water networks (hy­drography), and agricultural and cultivated lands (Ma et al. 2017.12). Therefore, these factors affect the texture and body of the settlements and the ways of life of their people. In this study we use several environmental variables and natural criteria, along with the Analytic Hierar­chical Process (AHP) integrated with the Geographic Information System (GIS), for zoning and evaluating the Iron Age sites of the Atrak River Basin to deter­mine: (1) What are the zones in the Middle Atrak? (2) Which of these basins were considered by Iron Age people? (3) What do the Iron Age sites in the different zones reveal? Theoretical foundations AHP is one of the most efficient multi-criteria deci-sion-making techniques, and was developed by the mathematician Thomas L. Saaty (1980) in the late 1970s. One of the advantages of the hierarchical analysis method in the context of this study is that it can deal with the various factors that influence the location of human settlements. It prioritizes re­sidential areas by weighting factors with pairwise comparisons (Abdelouhed 2022.11) and the impact rates of each of them (Liao, Kao 2010.571). AHP is an effective and helpful method for solving multi-criteria problems that use a hierarchical structure to show the problem, and a better way to solve such issues and prioritize different options based on user judgment. In other words, this method is used both in reality and theory in decision-making (Toledo-Aceves et al. 2011.975). To be more specific, the ul­timate goal of the AHP method is to determine the relative weight of each factor in a system (Yao, Zhao 2022.17). By solving decision problems, AHP allows researchers to focus on several criteria simultane­ously and also allows decision-makers to compare quantitative and qualitative criteria (Rodhiah et al. 2021.197). AHP is a multi-objective, multi-criteria decision-making approach that enables the user to reach priorities based on a set of options derived from three principles: parsing, comparative judg­ment, and prioritization (Abdul Rahaman, Arucha-my 2017.3). The AHP method has three basic steps: (1) creating a hierarchy, which is the essential part of the hier­archical analysis process (Cimren 2007.369); (2) de­termining the importance coefficients of variables and criteria using pairwise comparison methods; (3) assessing the consistency of judgments according to the percentage of consistency (Saaty 1980.287). As discussed above, AHP can be used for relative measurements by pairwise comparison of criteria and data, or measurement of data according to cri­teria and variables. Ranking mode and preferences Mohsen Heydari Dastenaei, Mohsen Dana include a pairwise comparison of criteria according to purpose. Ranking levels and preferences – such as excellent, very good, good, average, poor, and very poor – are then determined for each criterion. In the next step, pairwise comparisons are made be­tween the ranking levels of each criterion to obtain a set of priorities (weights) for these levels. For each criterion, scaled weights are considered, and each option is assigned a ranking level and will be scaled (Bahurmoz 2004.6). Materials and methods This research was carried out using a descriptive-analytical method to consider the issue of land suit­ability and its analysis with regard to settlement se­lection. Accordingly, after collecting the required in­formation and also reviewing the status of the Iron Age settlements in the Middle Atrak Basin, using AHP and going through the steps in ArcGIS – including entering variables and criteria, preparing informa- Fig. 1. Maps of the locations of Iron Age sites with regard to environmental factors. 1 distance of sites to water sources; 2 distance of sites to communication routes; 3 degree of slope; 4 height above sea level; 5 location of the site on soil type; 6 land use. Zoning analysis of Iron Age sites using Analytic Hierarchical Process (AHP) methods in the Middle Atrak River Basin, Northeast of Iran tion layers and new maps, classification and evalu­ation of variables, information layers and the com­bination of these layers – suitable locations for set­tlements were identified. Moreover, seven indicators – including distance from ancient sites to communi­cation routes, distance from ancient sites to water sources, land use, altitude, soil type, slope direction, and slope degree – were used to identify suitable areas for settlement (Fig. 1). ArcGIS10 and AHP were used to prepare a database of the layers described above (Fig. 2). Geography and ecology of the Atrak River Basin Although there is no long-standing accurate climate record of the northeastern region of Iran for the Iron Age, studies by researchers have shown that, contrary to popular belief, there were stages of sud­den climate change during the Holocene. There is a consensus that such phenomena are pervasive, and their results can be generalized to different parts of the world (Hejebri Nobari et al. 2021.298). Towards the end of this period, especially from the beginning of the first millennium BC, the tendency of tempera­ture changes tended to be colder (Shaikh Baikloo Islam 2020.40). The same cold and humid climate phenomenon are seen in all parts of the Tibetan Pla­teau (Callegaro et al. 2018) and West Central Asia (Fouache et al. 2020.92). The most recent long-term climate studies have been carried out near the study area of Jazmourian Playa (Vaezi et al. 2019) and Ha-moon Lake in Sistan (Hamzeh et al. 2016), and the coast of Gorgan (Kakroodi et al. 2015). The decrease in temperature caused the inhabitants of arid/semi­arid regions, such as north-central Iran (Shaikh Bai­kloo Islam, Chaychi Amirkhiz 2020.40), to adapt to the cold climate, in addition to agriculture. They also chose a nomadic-herding livelihood system, and a number of the areas covered in this article confirm this. The Atrak basin, one of the largest water basins in northeastern Iran with an area of 33 890km2, origi­nates from the mountains of Hezar Masjed in the north of Quchan. About 26 500km2 of this basin’s area is located in the political area of Iran, and the rest in Turkmenistan (Fig. 3). The Atrak basin is bounded by Turkmenistan in the north, Gorgan and Ka¢lshor basin in the south, the Qaraqum basin in the east, and the Caspian Sea in the west (Noori et al. 2011.160). This basin consists of two parts, plains and mountains. Its climate is barren or continental. Rainfall is less than 200mm in the plains and up to 500mm in the highlands. The maximum altitude of this basin at the site of Taba¢rak River is about 2903m, and a minimum of 22m above sea level is estimated (Sheikhvahed et al. 2011.5). The main waterway of the basin can be divided into three parts: upper, middle, and lower (border) Atrak. After crossing the plains of Quchan, Shirvan, and Bojnourd (Upper At-rak), the river continues its route in Ma¢neh, Ghori Meidan, and Mara¢veh Tappeh, then runs to the bor­der of Iran and Turkmenistan (Middle Atrak). After connecting to the Sumbar branch at the Chat site and forming the Border Atrak (Lower Atrak), it finally flows into the Caspian Sea. The study area includes the middle part of the Atrak River with a length of approximately 150km (the boundary between Reza­abad Gharbi and Sisab villages on the border of Shir-van and Bojnourd cities to Ghazan Ghayeh village Fig. 2. The structure of the hierarchical analysis process used in the research. Mohsen Heydari Dastenaei, Mohsen Dana on the border of Mane and Solmaghan cities with Maraveh Tappeh) (Yamani et al. 2010.4). The Middle Atrak Basin is geographically located be­tween the Hyrkani Plain in the west and the land of Khorasan in the east. The high mountains of Alborz in the west separate it from Hyrkani, and the moun­tains of Kopeh Dagh in the north separate it from the Qarehqom desert. With the Aladagh-Binalood Mountains in the south, the Middle Atrak Basin is safe from the central desert of Iran. The not-so-high altitude set separates this basin from the upper At-rak valley, and such a situation has made the central Atrak basin a relatively independent and closed ba­sin. This feature has a significant impact on the cli­mate of this region, which is something between the humid climate of Hyrkani and the cold and dry­ness of Khorasan. The western parts of this basin, especially in the Solmaghan plain, sometimes find a climate similar to the Gorgan plain, such as in sum­mer. Especially since the Aladagh Mountains, over­looking the Solmaghan plain, have a relatively dense forest cover. However, in higher latitudes (northwest of the central Atrak basin) this part is warmer and very poor in terms of vegetation and water resour­ces, due to the impact of the Turkmen Sahra low­lands in the west on the one hand and the soil of the region on the other. Background of archaeological research The first archaeological activities in the area of the Middle Atrak were the studies and work of Faegh To-hidi, which led to the arena determination of some sites. However, the first scientific excavation in this basin was carried out on the Tape Qaleh Khan, which showed an extended sequence from the Neolithic to the contemporary period (Garazhian 2011; Garaz­hian et al. 2014; Garazhian, Askarpour 2018). Ex­ploration reports on the Tape Ashkhaneh Bimarestan (Dana et al. 2017; Dana, Hejebri Nobari 2018), Ta­pe Ashkhaneh Rivi (Jafari, Thomalsky 2016), and Tape Eshgh Bojnourd (Vahdati 2014), along with efforts to determine the area and boundaries of the tape Kalateh Mostofi Bojnourd (Yazdani 2015), Ta­pe Bruski Ashkhaneh (Adine 2012) and Kohnekand Bojnourd (Dana et al. 2019), have also been pub­lished with regard to this basin. However, studies of the cities of Shirvan (Mirzaei 2008), Bojnourd, Raz and Jirgalan (Rajabi 2013), Mane and Solmaghan (Garazhian 2007; Ataei 2009; Zare 2011) in this area have not been published yet. Iron Age sites of the Middle Atrak Basin In the study and identification work carried out in the Middle Atrak Basin, over 360 archaeological sites Fig. 3. Upper and middle Atrak River Basin. Zoning analysis of Iron Age sites using Analytic Hierarchical Process (AHP) methods in the Middle Atrak River Basin, Northeast of Iran from all periods have been identified. Seventeen sites have been identified in Shirin Darreh springs (a tiny part of Shirvan city), which are located in the middle Atrak area (Mirzaei 2008); 143 sites have been identified in Bojnourd city, which is comple­tely located in the Middle Atrak Basin; 43 sites have been found in Raz and Jarglan city, about half of which is located in the Middle Atrak Basin (Rajabi 2013); and 160 sites have been identified and intro­duced in Mane and Solmaghan city, which are com­pletely located in the Middle Atrak Basin (Garazhian 2007; Ataei 2009; Zare 2011) (see Table 1). Of these, 61 sites were inhabited during the Iron Age. Environmental factors survey Water resources factor Human settlements are usually located where access to surface water is possible, and thus water is an essential factor in the emergence of human habi­tats and the most crucial factor in their growth and development (Heydari Dastenaei, Niknami 2020. 316). The land type and topographic status of each location significantly impacts water storage and flow. Accordingly, villages are established where there is enough water to meet the needs of the inhabitants (Motarjem, Siasar 2017.58). Atrak and its tributa­ries (Fig. 1.1), as a permanent and reliable water source, would be present many attractive locations in this regard. Suitable soil and altitude are also cru­cial for avoiding periodic or seasonal river floods when locating settlements. As shown in Figure 4, 80% of the Iron Age sites are located within 1000m from running water, indicating the connection be­tween the ancient sites and water resources. Communication routes factor Communication routes are another essential variable in the formation of ancient sites, especially in the Bronze Age and beyond, when we see the formation of cities with long-distance and trans-regional trade relations in the Greater Khorasan region. In the past, ancient roads were usually built based on natural paths and systems of valleys and plains (Hejebri No-bari et al. 2021.301), and this region follows this due to its mountainous nature. Com- to travel. Due to the mountainous location and the forested nature of the focal area, the only passable routes were inevitably the same cuts and the lengths of other valleys located between relatively high and steep mountains that were used as paths (Vosogh Babae, Mehrafarin 2018.197). This also applies in historical times, even in adjacent areas such as Dar-gaz, and historic sites have sometimes been formed adjacent to the main communication routes. This communication role is one of the essential factors in securing the economy of the inhabitants of these ci­ties and rural areas (Nami, Mousavinia 2019.239). There are 42 sites (69%) in the range of 0 to 1000m from the communication routes in this area, seven sites (11%) at a distance of 1000 to 2000m, eight si­tes (13%) at a distance of 2000m to 3000m, and four sites (7%) located 3000m or more from the commu­nication routes (Fig. 5). Among these, only one site – Tape Da¢sha¢d (IAMA60) – is located c. 9000m away from the communication routes. More than 70% of the sites are located at the bottom of the valleys, in the middle of the mid-mountain plains, and next to the communication routes (Fig. 1.2). Slope degree factor One of the influential environmental factors in the human settlement distribution system is the height and slope criterion. The slope is one of the essential factors in the transformation of land surface rough­ness (Akbar Aghalli, Velayati 2007.48), and thus it affects human life and activities such as agriculture, keeping livestock, and even some human settlements munication routes in mountainous areas usually pass from the bottom of the valleys. What we have in mind today as a communication route is very different from what existed in the past. Before the creation of mo­dern roads, people used gorges and Middle Atrak Basin Number of Percentage of Percentage of based on city identified sites identified sites Iron Age sites Shirvan 17 5 2 Bohnord 143 39 23 Raz and Jarglan 43 12 3 Mane and Solmaghan 160 44 72 Total 363 100 100 the cuts caused by geological activity Tab. 1. Location of Iron Age Sites based on counties. Mohsen Heydari Dastenaei, Mohsen Dana on the slopes either directly or indirectly (Zomorro­dian 1995.25). The degrees of the slopes in the region were classi­fied into nine separate groups. The lowest slopes (0–5 degrees) were determined as the first group, and the highest slopes were classified as group 9. Since the best slope for establishing human habita­tion is a slope of 0–10 degrees (Anabestani 2011), we examined the location of the sites on the slo­pes. The slope degree of the location of ancient si­tes is an essential factor that affects the area due to its economic impact. Among the sites of this period (Fig. 1.3) 18 (28%) were on slopes of 0–5 degrees, 20 sites (32%) on slopes of 5–10 degrees, 12 sites (18%) on slopes of 10–15 degrees, and 14 sites (22%) on slopes of more than 15 degrees (Fig. 6). Altitude from sea level factor Altitude from sea level can cause climate changes and, consequently, changes in lifestyle and some climatic features (Qazanfarpour et al. 2013.129). In addition, it directly affects ecosystems, vegetation, animals, and livelihood choices (Duckstein et al. 1973.22). The central Atrak region’s sea-level alti­tude varies between 226m and 2962m. The location of the sites in terms of altitude (Fig. 1.4) shows that about 60% of the sites are located at an altitude be­tween 226m to 819m above sea level (Fig. 7). In this region, the average annual rainfall in meteorologi­cal stations is about 250mm, which is suitable for rainfed cultivation. However, it should be noted that despite the appropriate rainfall and altitude, the soil type is also crucial for cultivation. Sufficient rainfall and humidity at altitudes of about 600m above sea level and above allow optimal rainfed cultivation (Kirkby 1979.Tabs. 83–84). However, the annual rainfall is a more critical factor for rainfed cultiva­tion. The minimum annual rainfall suitable for rain-fed cultivation is about 200mm (Adams 1981.12), indicating that this area is suitable for rainfed agri­culture due to having more rainfall. Soil type factor Today, geoarchaeological studies have found a spe­cial place as a helpful tool in archaeological research and explaining ancient Quaternary environments (Maghsoudi et al. 2020.2). Soil is a non-dense or­ganic matter that has been created over many years under the influence of various factors, such as cli­mate, vegetation, and elevation (Salmanpour et al. 2013), and soil type affects the livelihood structure of an area (Estelaji, Ghadiri Masoum 1995.126). As can be seen on the map, large areas of the western parts of the Middle Atrak Basin are geologically cal­careous and unsuitable soils that are also very poor in vegetation. The Iron Age sites of the region are in the category of Incepti soil/Entisoil rocky outcrop soils with a small amount of Incepti soil (Fig. 1.5). In this area, 42 (68%) of the Iron Age sites are lo­cated on Incepti soil, seven sites (12%) are located on Insepti soil with rocky outcrop soil, and 12 sites (20%) are located in areas with enti soil with rocky outcrop soil (Fig. 8). The presence of fine-grained and fertile sediments usually provides suitable ma­terials for agriculture, pottery, and other economic activities and acceptable conditions for developing settlements (Maghsoudi et al. 2020.7). Incepti soils are spread all over the world, and research shows that they are suitable for agricultural and non-agri­cultural uses, and can be widely used for crop culti­ Zoning analysis of Iron Age sites using Analytic Hierarchical Process (AHP) methods in the Middle Atrak River Basin, Northeast of Iran vation, provided that artificial drainage is possible in them (Sohrabi et al. 2013). Land use factor Land use results from a combination of human ac­tivity and the capabilities of a place. Although Land use is the result of population activities, it is also in some ways a reason for the existence of certain ca­pabilities and the possibility of using the capabili­ties of the natural environment (Sadr Mousavi et al. 2018.734). The cultivability of land is one of the fac­tors that is influenced by many important criteria, such as altitude, presence or absence of surface wa­ter, soil type, human manipulation of the environ­ment, and climate. However, this human manipula­tion can also have a decisive role on the erosion rate. Most importantly, the land cultivability and type of vegetation can also be a very determining factor in the type of livelihood of those living in the related settlements. As such the settlement or use of many shelters in an area, especially in association with rais­ing livestock, depends on land cultivability and type of vegetation (Afifi 2018.636). To this end, the pur­pose of land surveying is to determine the land va­lue from the location point of view (Rahimi, Hasan-pour 2011.21). The map of the area based on land use (Fig. 1.6) shows that about half of the sites are located in areas that are currently used for agriculture, whether irrigated or rainfed, and the other half are in areas that are located in a pasture or forest areas (Fig. 9). This local difference in the sites should be consid­ered as related to the livelihood of the residents. This means that in pasture areas the sites indicate that nomads used them for temporary settlement or cemeteries, and that the sites on suitable agricultu­ral land belonged to sedentary farmers. Slope direction factor As a general concept, direction is a well-defined fea­ture for the linear effects of a phenomenon in geo­metry. In the context of this study it also includes other concepts, such as slope and geological slope (Heydari Dastenaei, Niknami 2020.320). Slope di­rection determines the amount of solar energy that the soil receives. This energy determines the tempe­rature of air and soil and the amount of available water in the soil, which are the factors that cause differences in the vegetation of different slopes. In mountainous areas slopes facing the sun seem to be more suitable for settlement, while in tropical areas this is the case for slopes that do not face the sun. In the Middle Atrak Basin the southern slopes are the most important and the northern slopes the least, because the former receive the lowest heat in summer and the most heat in winter. The eastern and western slopes are less important than the southern slopes, and are used in spring and au­tumn (Heydari Dastenaei 2018.7). Surveying the location of Iron Age sites indicates that the northern slopes contain more settlements and the southern slopes are less used (Fig. 10). Accordingly, 12 sites are located on northern slopes, 11 sites on the north­east, two sites on the east, six sites on the southeast, six sites on the south, two sites on the southwest, eight sites on the west, and 12 sites on the west areas. Zoning of Atrak River Basin Potential zones for the placement of Iron Age set­tlements have been determined by weighting and classification of the abovementioned criteria and en­vironmental factors, including communication rou­tes, distance from water resources, land use, slope, slope direction, altitude, and soil type. The layers’ weights, which are the same factors as determined in the model, were evaluated according to the avai­lable options, and finally the total weight of the la­yers was obtained using the calculation formulas. The weight of each layer according to the preference options is shown in Table 2. Mohsen Heydari Dastenaei, Mohsen Dana In AHP, in addition to considering the factors and combining their different levels, the rate of the num­ber of achieved priorities and the accuracy of weight­ing results can be trusted due to the method of cal­culating the comparisons’ compatibility with regard to the studied layers, and determining the overall compatibility rate. In this research, the value of the compatibility rate is calculated as 0.4, which indi­cates the appropriate compatibility of the studied layers. According to Table 3, it can be seen that alti­tude, distance to a water source, distance to a com­munication route, slope, and soil type have the high­est weights, and land use and slope direction have the lowest weights. In Table 4, according to the results obtained from the final weight of the factors affecting the creation of ancient sites, it can be seen that altitude is in the first place with a weight of 0.26, distance to rivers is in the second place with a weight of 0.19, dis­tance to communication routes is third with a weight of 0.15, slope direction is fourth with a weight of 0.14, soil type is fifth with a weight of 0.08, land use is sixth with a weight of 0.08, and in seventh place is slope, with a weight of 0.07. In the next step, GIS is used to prepare layers to se­lect areas with a higher priority. The final map is obtained by stacking the existing layers in terms of weight, as extracted from Table 4. According to the results obtained in this section, those areas with a higher potential for locating settlements have lighter colours, while those with a lower potential for this settlements have darker colours. Slope Slope Distance Landuse direction degree to roads More specifically, those valuable areas that have the most potential for establishing settlements are the relatively limited areas shown in green. These have an area of 2235km2, equivalent to 10% of the total area, with a low slope of about five degrees and a suitable slope direction, land use, suitable vegetation, and rich soil, and are generally suitable for the de­velopment of settlements (Fig. 11). This area has 28 Iron Age sites, accounting for 46.7% of the sites. Moreover, the areas marked in orange with an area of 12 550km2, equivalent to 47% of the total area, also have relatively good values. These areas are geographically hilly, suitable for growing rainfed plants, and also a good place for livestock grazing. This section includes slopes up to 13 degrees, rain-fed agricultural uses with an almost suitable soil type, and a short distance to water sources and com­munication routes. Twenty-four sites (39.3%) are located in this area. The dark brown areas with an area of 7900km2, equivalent to 22% of the Distance Above Type total, have a relatively to rivers sea level of soil low value for settlement. Slope direction Landuse 1 0.33 3.03 1 0.45 4.35 0.45 0.32 0.29 0.45 0.25 0.26 0.83 0.27 These areas are hillsides slope degree 2.21 0.23 1 3.2 0.48 0.48 4 with steep vegetation Distance to roads 2.22 3.12 3.2 1 0.26 0.31 4.55 and steep slopes, and Distance to rivers 3.45 2.22 2.1 3.9 1 0.36 3.33 the land is used only for Above sea level 4 3.8 2.1 3.25 2.8 1 2 pasture. In addition, the Type of soil 1.2 3.7 0.25 0.22 0.3 0.5 1 type of soil and even Total 14.41 17.1 13.45 12.34 5.57 3.15 15.98 the soil depth in these Tab. 2. The binary preference matrix of the components. areas is low, and they are far from permanent Slope direction 0.69 0.177 0.234 0.037 0.052 0.079 0.052 0.071 water sources such as ri- Landuse 0.023 0.058 323 0.026 0.081 0.093 0.017 0.087 vers, as well as commu- Degree of slope 0.153 0.013 0.074 0.259 0.085 0.151 0.25 0.141 nication routes. It is no- Distance to roads Distance to rivers Above sea level 0.154 0.239 0.278 0.182 0.13 0.222 0.238 0.156 0.156 0.081 0.316 0.263 0.046 0.179 0.502 0.098 0.113 0.317 0.248 0.09 0.125 0.155 192 0.266 teworthy that seasonal water springs are usual- Type of soil 0.083 0.216 0.019 0.018 0.054 0.159 0.063 0.087 ly seen in these areas, and this type of area is Tab. 3. Normalized matrix of preferences. used only for livestock Zoning analysis of Iron Age sites using Analytic Hierarchical Process (AHP) methods in the Middle Atrak River Basin, Northeast of Iran grazing. Seven ancient sites in this area can be seen, accounting 11.4% of the sites. Finally, the dark red areas, with an area of 393km2 and equivalent to 15% of the total area, do not have any settlement value. These areas have steep slopes and rocky, non-agricultural lands, with poor range­land vegetation. Thus, due to their high altitude and steep slopes, unsuitable terrain for agriculture, and distance from communication routes, these areas are often unsuitable for settlement. However, an Iron Age site (IAMA 44) (1.6%) is located in this area, which seems to have been a short-term seasonal es­tablishment. Discussion and conclusion The use of GIS and geostatistical techniques can be a useful, practical tool in archaeology. These tools make it possible to apply complex mathematical equations to maps. On the other hand, using the existing interpolation methods in the field of statis­tics, statistical and spatial analysis can be carried out in different places based on the locational and geo­graphical situation of the phenomena. In the present study, we tried to evaluate the effects of environ­mental factors on the formation of Iron Age settle­ments. The results showed that environmental con- Layers Abnormal weight Normalized weight Slope direction 0.0500158 0.071450106 Landuse 0.0611772 0.087396017 Degree of slope 0.0987156 0.14105536 Distance to roads 0.1083421 0.154775398 Distance to rivers 0.1342631 0.191802997 Above sea level 0.1863892 0.266270269 Type of soil 0.061098 0.087282862 Total 0.7 1 Tab. 4. Calculation of final weights based on pref­erences with regard to the environmental factors. ditions in the form of slope characteristics, slope di­rection, altitude, distance from communication rou­tes, access to water resources, soil type, and land use all positively affect the distribution and density of site in the area being studied. In the first step, in order to determine the impor­tance of each layer using the AHP method, indica­tors were compared pairwise with each other, and each indicator was weighted. According to the re­sults of the AHP model, the highest weight is related to the altitude index with a weighted score of 0.26, and the lowest weight is related to the slope direc­tion, with a score of 0.07. In the next step, the lay­ers were standardized into four levels and a zoning Mohsen Heydari Dastenaei, Mohsen Dana map of the Iron Age areas of the Middle Atrak Ba-highland resources and plain agricultural resources. sin was prepared. The results show that one site is Such sites had a combined economy of livestock and in a zone with no settlement value, size sites are in agriculture, and also engaged in trade. Sites in Zone a low importance zone, 24 sites are in a relatively 3 are sites at the foot of the mountains and seaso-appropriate zone and 28 sites are in a very impor-nal sites with livestock. Finally, Zone 1 is not suit-tant zone. able for settlement at all, and for this reason only one site is located in this area, and this site, like The results of analyses show that the locations of those in Zone 3, is a seasonal site. the sites in different zones indicates different kinds of livelihoods were pursued there. Based on zoning ACKNOWLEDGEMENTS analysis, it is determined that the areas that are lo- We are grateful to Norouz Rajabi, Shahram Zare, Mo- cated in Zone 4 (show in dark colours) are villages hamad Taghi Ataei, Omran Garazhian, and Azita that engaged in irrigated and rainfed agriculture. Mirzaei for providing their unpublished data. The areas located in Zone 3 are areas that used both . References Abdelouhed F., Ahmed A., Abdellah A., Yassine B., and Mo­hammed I. 2022. GIS and remote sensing coupled with analytical hierarchy process (AHP) for the selection of ap­propriate sites for landfills: a case study in the province of Ouarzazate, Morocco. Journal of Engineering and Applied Science 69(19): 1–23. https://doi.org/10.1186/s44147-021-00063-3 Abdul Rahaman S., Aruchamy S. 2017. Geoinformatics based landslide vulnerable zonation mapping using ana­lytical hierarchy process (AHP), a study of Kallar river sub watershed, Kallar watershed, Bhavani basin, Tamil Nadu. Journal of Modeling Earth Systems and Environment 3(1): 1–13. https://doi.org/10.1007/s40808-017-0298-8 Adams R. M. 1981. Heartland of Cities: Surveys of An­cient Settlement and Landuse on the Central Floodplain of Euphrates. University of Chicago Press. Chicago. Adine O. 2012. Soundings for determination boundery at Tappe Boruski, Ashkhane, Northern Khorasan. Ab­stract the 11th Annual Symposium of Iranian Archaeology, Tehran. Cultural Heritage Research Institute. Tehran: 53. Afifi M. E. 2018. Analyze the Impact of Natural Factors in the Spatial Distribution of Urban and Rural Settlements of Khonj County. Journal of Studies of Human Settlement Planning 13(3): 629–646. Akbar Aghalli F., Velayati S. 2007. Investigating the posi­tion of natural factors in the establishment of rural settle­ments. International and Scientific Journal of Iranian Geography Association 12–13: 45–66. Anabestani A. A. 2011. The Role of Natural Factors in Sta­bility of Rural Settlements (Case Study: Sabzevar Country). Geography and Environmental Planning 40(4): 89–104. Ataei T. 2009. Report on Archaeological Surveys in Ma­ne & Samalghan County (Northern Khorasan), First Season. Iranian Center for Archaeological Research. Teh­ran. Unpublished. (In Persian) Bahurmoz A. M. A. 2004. The Analytic Hierarchy Process: A Methodology for Win-Win Management. Journal of King Abdulaziz University – Economics and Administra­tion 26(1): 3–16. https://doi.org/10.4197/ECO.20-1.1 Callegaro A., Battistel D., Kehrwald M. N., +5 authors, and Barbante C. 2018. Fire, vegetation, and Holocene climate in a southeastern Tibetan lake: a multi-biomarker recon­struction from Paru Co. Climate of the Past 14(10): 1543–1563. https://doi.org/10.5194/cp-14-1543-2018 Çimren E., Çatay B, and Budak E. 2007. Development of a machine tool selection system using AHP. International Journal of Advanced Manufacturing Technology 35: 363–376. https://doi.org/10.1007/s00170-006-0714-0 Dana M., Hejebri Nobari A. 2018. Site Formation Process as Seen from Tape Bimarestan-e Ashkhane Excavations Data. Journal of Archaeological Studies 10(2): 83–97. https://doi.org/10.22059/jarcs.2019.68527. Dana M., Hejebri Nobari A., and Mousavi Kouhpar M. 2017. Excavation at Tappe Bimarestan Ashkhaneh; an Iron Age Graveyard in the North of Khorasan. Architectu­re and Modern Information Technologies (AMIT) 49: 151–167. Dana M., Hozhabri A., and Rahmati M. 2018. Kohne Kand Bojnord, a Parthian Site with local caracteres in North Khorasan. In M. H. Aziziz Kharanaghi, M. Khanipour, and R. Naseri (eds.), Second International Symposium of Young Archaeologists. Iranology Fondation 2. Tehran: 795–834. Zoning analysis of Iron Age sites using Analytic Hierarchical Process (AHP) methods in the Middle Atrak River Basin, Northeast of Iran Duckstein L., Fogel M. M., and Thames J. L. 1973. Eleva­tion effects on rainfall: A stochastic model. Journal of Hy­drology 18(1): 21–35. https://doi.org/10.1016/0022-1694(73)90023-1 Eselaji A., Ghadiri Masoum M. 1995. Investigating Geo­graphical Factors in the System of Rural Settlements with Emphasis on Quantitative Techniques (Case Study: Wilkij District of Namin County). Journal of Geographical Re­searches 37 (53): 121–136. Fouache E., Cez L., Andrieu-Ponel V., and Rante R. 2020. Environmental changes in Bactria and Sogdiana (Central Asia, Afghanistan, and Uzbekistan) from the Neolithic to the Late Bronze Age: interaction with human occupation. In B. Lyonnet, N. and A. Dubova (eds.), The World of the Oxus Civilization. Routledge Worlds. New York: 82–109. Garazhian O. 2007. Report on Surface Surveys of Sa­malghan Plain, Northern Khorasan. Iranian Center for Archaeological Research. Tehran. Unpublished. (In Persian) 2011. Sunding for Stratigraphy and Documentation of Architectural Remains of Tappe Qal’e Khan, Mane and Samalghan County, Northern Khorasan Province. The 9th Annual Symposium on Iranian Archaeology. Iranian Center for Archaeological Research. Tehran: 145–159. Garazhian E., Askarpur V. 2011. Trends in Evolution of Pottery in Qal’eh Khan, Bojnurd. Journal of Archaeolo­gical Studies 3(1): 107–132. Garazhian O., Jafari J., and Hozhabri A. 2014. Report of Archaeological Researches for Documentation of Archi­tectures at Tappe Qal’e Khan, Khorasan, emphasize on the Historical Period. Modares Archaeological Research 3: 161–199. Garazhian O., Papoli Yazdi L., and Fakhr-e Ghaemi H. 2014. Qaleh Khan a Site in Northern Khorassan and the Neolithic of North Eastern Iranian Plateau. Architecture and Modern Information Technologies (AMIT) 46: 21– 50. Gholami Rad Z., Wali Shariatpanah M. 2013. Investigat­ing the position of natural factors in the establishment of rural settlements in Kermanshah province based on AHP hierarchical analysis process model using GIS. Geo­graphical territory 10(37): 55–76. Hamzeh M. A., Mahmudi-Gharaie M. H., Alizadeh-Lahijani H., Mousavi- Harami R., Djamali M., and Naderi-Beni A. 2016. Paleolimnology of Lake Hamoun (E Iran): Implica­tion for Past Climate Changes and Possible Impacts on the Human Sttlements. PALAIOS 31: 1–14. http://dx.doi.org/10.2110/palo.2016.055 Hejebri Nobari A., Biscione R., and Judy N. 2021. Settle­ment Patterns of the Bronze Age Sites of the Upper and Middle Atrak Basin in the Northeast of Iran. Journal of Archaeological Studies 13(2): 293–317. https://doi.org/10.22059/jarcs.2021.279390.142703 Heydari Dastenaei M. 2018. The effect of environmental factors on the Prehistoric sites in the southern Zayandeh-Rud by using the Pearson correlation. Journal of Iran Pre-Islamic Archaeological Essays 2(1): 1–14. Heydari Dastenaei M., Niknami K. A. 2020. Analysis of the Relationship between the Formation and Continuity of Neolithic Period Settlements with their environment in the Sarfirouz Abad Plain of Kermanshah, West Central Zagross. Physical Geography Research Qurterly 52(2): 313–331. https://doi.org/10.22059/jphgr.2020.285488.1007418 Jafari J., Thomalsky J. 2016. The Iranian-German Tappe Rivi Project (TRP), North-Khorasan: Report on the 2016 and 2017 Fieldworks. Architecture and Modern Infor­mation Technologies (AMIT) 48: 77–120. Kakroodi A. A., Leroy S. A. G., Kroonenberg S. B., Lahijani H. A. K., Ali Mohammadian H., Boomer I., and Goorabi A. 2015. Late Pleistocene and Holocene sea-level change and coastal paleoenvironment evolution along the Iranian Caspian shore. Marine Geology 361: 111–125. https://doi.org/10.1016/j.margeo.2014.12.007 Khosroshahi M., Abbasi H. R., Khashki M. T., and Abtahi M. 2013. Determination of Iran Desert Lands Based on Soil Attributes. Journal of Desert Management 1(1): 27–38. https://doi.org/10.22034/jdmal.2013.17098 Kirkby M. J. 1979. Land and Water Resources of the Deh Luran and Khuzistan Plain. In F. Hole (ed.), Studies in the Archaeological History of the Deh Luran Plain. Univer­sity of Michigan Press, University of Michigan Museum of Anthropological Archaeology. Michigan: 251–288. https://doi.org/10.3998/mpub.11395563 Liao Ch. N., Kao H. P. 2010. Supplier Selection Model Using Taguchi Loss Function, Analytical Hierarchy Process and Multi- Choice Goal Programming. Computers & In­dustrial Engineering 58(4): 571–577. https://doi.org/10.1016/j.cie.2009.12.004 Maga. L., Gajski D., Dziêgielewska-Gajski K. 2021. Spatial multi criteria Approach to the evaluation of Archaeologi­cal Sites. GIS Odyssey Journal 1(1): 21–35. https://doi.org/10.57599/gisoj.2021.1.1.21 Ma L., Guo X., Tian Y., Wang Y., and Chen M. 2017. Mic-ro-Study of the Evolution of Rural Settlement Patterns and Their Spatial Association with Water and Land Re­ Mohsen Heydari Dastenaei, Mohsen Dana sources: A Case Study of Shandan County, China. Sustai­nability 9(12): 1–18. https://doi.org/10.3390/su9122277 Maghsoudi M., Fazeli Nashli H., Azizi Gh., Gillmore G., and Schmidt A. 2020. Geoarchaeology of Alluvial Fans: A Case Study from Jajroud and Hajiarab Alluvial Fans in Iran. Physical Geography Research Qurterly 44(4): 1–22. https://doi.org/10.22059/jphgr.2012.30239 Mirzaei A. 2008. Report on Surface Surveys in Faruj and Shirvan County, Northern Khorasan. Iranian Center for Archaeological Research. Tehran. Unpublished. (In Persian) Motarjem A., Siasar N. 2017. Studying the Changes of Di­stribution Patterns of the Bronze and Iron Ages Settle­ments in Chamchamal Plain, Central Zagros. Pazhuhesh-ha-ye Bastanshenasi Iran 6(11): 75–90. https://doi.org/10.22084/nbsh.2016.1740 Nami H., Mousavinia S. M. 2019. Archaeological Survey of Parthian Sites in Dargaz Plain, North-East of Iran. Pazhu-hesh-ha-ye Bastanshenasi Iran 11: 233–252. https://doi.org/10.22059/jarcs.2019.71116 Noori R., Jafari F., Asgharzadeh F., and Akbarzadeh A. 2011. Offering a Proper Framework to Investigate Water Quality of the Atrak River. Iranian Journal of Health and Environment 4(2): 159–170. Qazanfarpour H., Kamandari M., and Mohammadi Soley­mani M. 2013. The effect of geographical factors on the pattern of rural housing in Kerman province. Zagros Landscape Geography and Urban Planning Quarterly 5(18): 125–142. Rajabi N. 2013. The Final Report on Archaeological Sur­veys in Bojnord and Raz & Jargalan County, Northern Khorasan. Iranian Center for Archaeological Research Tehran. Unpublished (In Persian). Rahimi M., Hasanpour Kh. 2011. Location finding of a new village in Bastak province considering passive de­fense in GIS environment using AHP model. Journal of Hormozgan Caltural Research Review 1(1–2): 12–26. Rodhiah D. M. M., Hidayah N., Adiputra G., and Mukmin T. 2021. Application of the analytic Hierarchy process me­thod in determining a creative Industry strategy: evidence from creative industry in South Tangerang. PalArch’s Jour­nal of Archaeology of Egypt/Egyptology 18(2): 195–205. Saaty T. L. 1980. The Analytical Hierarchy Process. Mc-Graw-Hill McGraw-Hill International Book Company. New York. Sadr Mousavi M., Talebifard R., and Niazy C. 2018. Inve­stigating the Role of Natural Factors in the Geographical Distribution of Rural Settlements (Case Study: Sahneh County). Journal of Studies of Human Settlement Plan­ning 12(4): 731–749. Salmanpour A., Senmar M., and Bakhtiari A. 2013. The role of soil on archaeological analysis and studies. The First Symposium of Archaeology of Iran. Birjad. Shaikh Baikloo Islam B. 2020. Holocene climatic events in Iran. Journal of Climate Change Research 1(4): 35–47. https://doi.org/10.30488/ccr.2020.244327.1017 Shaikh Baikloo Islam B., Chaychi Amirkhiz A. 2020. Adap­tations of the Bronze and Iron Ages Societies of North Cen­tral Iran to the Holocene Climatic Events. Journal of Cli­mate Change Research 1(2): 39–54. https://doi.org/10.30488/ccr.2020.111121 Sheikhvahed B., Bahremand A., and Mushekhian Y. 2011. A Comparison of Trends in Hydrologic Variables in the At-rak River Basin Using Non-Parametric Trend Analysis Tests. Journal of Water and Soil Conservation 18(2): 1–23. Sharifi A., Pourmand A., Canuel E. A., +7 authors, and Lahijani H. A. 2015. Abrupt climate variability since the last deglaciation based on a high-resolution, multi-proxy peat record from NW Iran: The hand that rocked the Cra­dle of Civilization? Quaternary Science Reviews 123: 215– 230. https://doi.org/10.1016/j.quascirev.2015.07.006 Stevens L. R., Ito E., Schwalb A., and Wright Jr. H. E. 2006. Timing of atmospheric precipitation in the Zagros Moun­tains inferred from a multi-proxy record from Lake Mira-bad, Iran. Quaternary research 66(3): 494–500. https://doi.org/10.1016/j.yqres.2006.06.008 Tolga E., Demircan L., and Kahraman C. 2005. Operating system selection using fuzzy replacement analysis and analytic hierarchy process. Journal Production econo­mics 97(1): 89–117. https://doi.org/10.1016/j.ijpe.2004.07.001 Toledo-Aceves T., Meave J. A., González-Espinos M. and Ra-mírez-Marcial N. 2011. Tropical montane cloud forests: Current threats and opportunities for their conservation and sustainable management in Mexico. Journal of Envi­ronmental Management 92: 974–981. https://doi.org/10.1016/j.jenvman.2010.11.007 Vaezi A., Ghazban F., Tavakoli V., +4 authors, and Kylin H. 2019. A Late Pleistocene-Holocene multi-proxy record of climate variability in the Jazmurian playa, southeast­ern Iran. Palaeogeography, Palaeoloclimatology, Palaeo-ecology 514: 754–767. https://doi.org/10.1016/j.palaeo.2018.09.026 Vahdati A. A. 2014. A BMAC Grave from Bojnord, North-Eastern Iran. Iran 52: 19–27. https://doi.org/10.1080/05786967.2014.11834735 Zoning analysis of Iron Age sites using Analytic Hierarchical Process (AHP) methods in the Middle Atrak River Basin, Northeast of Iran Verga Matos P., Cardadeiro E., da Silva J. A., and De Muyl­der C. F. 2018. The use of multi-criteria analysis in the recovery of abandoned mines: a study of intervention in Portugal. RAUSP Management Journal 53: 214–224. https://doi.org/10.1016/j.rauspm.2017.06.005 Vosogh Babae E., Mehrafarin R. 2018. Analysing the Role of Environmen in the Parthian Settlements Distribution: A Case Study in the Chelchay River Drainage, Minodasht, Golestan, Iran. Pazhuhesh-ha-ye Bastanshenasi Iran 16: 183–202. https://doi.org/ 10.22084/nbsh.2018.14420.1635 Yamani M., Dolati J., and Zarei A. 2010. The effect of hy­drogeomorphic factors on temporal and spatial changes in the middle part of the Atrak River. Geographical Re­searches 99: 1–24. Yao X., Zhao F. 2022. A quantitative evaluation based on an analytic hierarchy process for the deterioration degree of the Guangyuan Thousand-Buddha grotto from the Tang Dynasty in Sichuan, China. Heritage Science 10(19): 1–18. https://doi.org/10.1186/s40494-022-00655-z Yazdani A. 2015. Soundings for determination boundery back to contents at Kalate Mostowfi (Bojnourd, Iron Age). Iranian Center for Archaeological Research. Tehran. Unpublished. (In Per­sian) Zare Sh. 2011. Report on Archaeological Surveys in Ma­ne & Samalghan County (Northern Khorasan), Second Season. Iranian Center for Archaeological Research. Teh­ran. Unpublished. (In Persian) Zhang Zh., Xiao R., Shortridge A., and Wu J. 2014. Spatial Point Pattern Analysis of Human Settlements and Geogra­phical Associations in Eastern Coastal China – a Case Study. International Journal of Environmental Research and Public Health 11(3): 2818–2833. https://doi.org/10.3390/ijerph110302818 Zhang R., Jiang D., Zhang L., Cui Y., Li M., and Xiao L. 2014. Distribution of nutrients, heavy metals, and PAHs affected by sediment dredging in the Wujin’gang River basin flowing into Meiliang Bay of Lake Taihu. Environ­mental Science and Pollution Research 21: 2141– 2153. https://doi.org/10.1007/s11356-013-2123-x Zomorrodian M. J. 1995. Application of natural geogra­phy in urban and rural planning. Payam Nur Publica­tions. Tehran. Documenta Praehistorica XLIX (2022) A possible case of ‘accompanying dead’ in the second half of the 6th millennium cal BC at Ug¢ urlu\Gökçeada, Turkey Bas¸ak Boz basakboz@trakya.edu.tr Department of Archaeology, Trakya University, Edirne, TR ABSTRACT – Eleven human skeletons were found in a 2m deep circular pit in an open area dating to 5389–5300 cal BC at Ugurlu/Gökçeada. The pit can be considered as a part of the pit tradition frequently seen in Thracian and Balkan prehistory. Its unique contents, however, are discussed in this paper in the scope of possible motivations. An ‘accompanied dead’ hypothesis is offered as the possible motivation of the case based on the contents and depositional details of bodies within the pit. This type of deposition was practiced throughout Europe starting from the early Neolithic through the Chalcolithic. KEY WORDS – accompanying dead; the second half of the 6th millennium BC; Ugurlu/Gökçeada; Turkey Mogo; primer ‘pridru/enih mrtvih’ v drugi polovici 6. tiso;letja pr. n. [t. na najdi[;u Ug¢ urlu\Gökçeada, Tur;ija IZVLE.EK – Na najdi..u Ugurlu/Gökçeada je bilo v dva metra globoki okrogli jami zunaj naselja, da­tirani med 5389–5300 pr. n. .t., najdenih enajst .love.kih okostij. Jamo lahko razumemo kot del pra­zgodovinske tradicije, ki jo pogosto sre.amo v Trakiji in na Balkanu. Njeno vsebino v .lanku pred­stavljamo v okviru mo.nih praks. Predlagana hipoteza o ‘pridru.enih mrtvih’ kot mo.ni praksi te­melji na vsebini in podrobnostih polaganja trupel v jamo. Tak na.in polaganja se v Evropi za.ne v zgodnjem neolitiku in traja vse do konca halkolitika. KLJU.NE BESEDE – pridru.eni mrtvi; druga polovica 6. tiso.letja pr. n. .t.; Ugurlu/Gökçeada; Tur.ija Introduction Burial customs and ritualistic behaviour patterns are a great source for understanding past belief systems, social structures, and cultural interactions between different groups and regions. Western Anatolia, Thrace, and the Balkans are of special importance in terms of understanding the spread of Neolithic lifeways and the transition to the Chalcolithic. Infor­mation on burial practices and/or any other types of rituals related to human remains is not well repre­sented in this area. A recent discovery of a pit with human remains offers an opportunity to look into the ritual behaviours of people during the 6th mil­lennium BC on the island Gökçeada in Turkey. The settlement of Ugurlu is located in the western part of Gökçeada, in the northern Aegean Sea. The site is on the pathway to Europe from Anatolia, mak­ing it an important site in terms of understanding human cultural interaction (Fig. 1). Six cultural lay­ers have been identified in the excavation at Ugurlu (see Table 1 for the phasing). The earliest phase of the site dates to the pre-pottery Neolithic 6800–6600 DOI> 10.4312\dp.49.13 A possible case of ‘accompanying dead’ in the second half of the 6th millennium cal BC at Ug¢urlu\Gökçeada, Turkey cal BC, and is the earliest evidence of agri­culture and animal husbandry in the area (Erdogu 2014; 2017; 2020). Ugurlu is also one of the few settlements in the region that represents the transi­tion from the Neolithic to the Chalcolithic (5500–4900 cal BC) between western Anatolia and the eastern Aegean islands. This transitional period bears witness to changes in settlement organization, buil- Phasing Period Date VI Pre-pottery Neolithic 6800–6600 cal BC VI-V Transition to Early Neolithic 6600–6500 cal BC V Early Neolithic 6500–5900 cal BC IV Late Neolithic 5900–5500 cal BC IV-III Transition to Late Neolithic to Early Chalcolithic 5500–5300 cal BC III Early Chalcolithic 5300–4900 cal BC Hiatus – 4900–4500 cal BC II Middle Chalcolithic 4500–4300 cal BC I Surface> Early Bronze\Middle Age | ding plans, pottery production, and subsi- Tab. 1. Phasing of Ug urlu settlement (reproduced from Gürçal stence economy (Erdogu 2014; 2017). A 2021.65). particularly distinct feature was the prac­tice of pit use. This new tradition emerged in Phase IV, continued throughout Phase III, and was slowly disappearing in Phase II. To date, more than 37 pits have been found in Phase IV (3 pits) and Phase III (Fig. 2). During Ugurlu Phase III, the settlement appears to be divided into two parts. The western part has a communal building (B4) with a courtyard, while in the eastern part the buildings are multi-roomed dwellings (Erdogu 2020). Most of the pits were in the northwest part of the settlement in the open area around the communal building (Erdogu 2020). The pits differ in terms of size and depth (10–90cm), and with a few exceptions they are all plastered with greenish clay and have similar inclusions. Among the contents of these pits are fragments of pottery and animal bones, clay and marble figurines, bone and flint tools, fragments of shell ornaments (bracelets and rings, etc.), and stone axes (Ka­ramurat et al. 2021). Three of these pits contained a small num­ber of disarticulated human re­mains (seven pieces of human bone in pit 25, and a few others in pits 29 and 104). All pits were deliberately infilled and covered with large stones to end the lives of the pits. Apart from these three pits that have fragments of human remains, one other pit stands out with its remarkable deposition of several individuals (unit 188, feature 88). This paper thus aims to describe pit 188 and place it within larger contextual data, especially in relation to other pits, and to discuss the possible motivations for this deposition of human bodies. The structural and depositional characteristics of the pit 188 The pit is located in the western part of the settle­ment in the courtyard of the communal building (Fig. 2). Two AMS dates were taken from the human bones. The first one from the first skeleton at the bottom of the pit which gave a date of 6380±30 BP (Beta-480187, 5389–5310 BC at 95.4%); the second date comes from a human bone from the upper layer and gave a date of 6340±30 BP (Beta-465445, 5363– 5302 BC at 95.4%. This range is contemporary with the communal building’s earliest use at around 5300–4300 cal BC (Boz, Erdogu 2019.3). Bas¸ak Boz Fig. 2. The distribution of the pits around the communal building (Ug urlu Archive). Having a 1m diameter and c. 2m depth, this pit is one of the largest of 34 pits found in Phase III. The pit was plastered with yellowish-green clay plaster on the side walls and the floor. Fragments of clay were found throughout the infill of the pit, indi­cating that the inner walls were also plastered in a similar way as the other pits. A total of 11 individ­uals1 were found piled up on top of each other with numerous stones in between them. After the last body was interred, the pit was filled with stones for closure at a depth of between 60 and 80cm. In addition to human bodies, the pit contained a large amount of archaeological material, including two large broken grinding stones, a great number of pottery sherds, three pieces of red ochre, a small number of beads, and a worked bone (for details see Karamurat et al. 2021). Partially articulated leg bones and the pelvis bones of two calves were also among these finds, as well as many fragments of ani­mal bones. A sterile grey ash deposit covering a 30cm area and with c. 1.5cm thickness was found on one of the largest stones in the middle of the pit. The pit was filled up with many stones and marked by another stone, as seen in others2 . Human remains Standard osteological analysis was carried out for age and sex estimation. The biological sex of the adults was identified based on morphological fea­ 1 Different skeletal parts of the same individual were recorded under two separate skeleton numbers in the field. The pieces were matched in the lab therefore the total number of the skeletons within this pit is 11 rather than 12, as was indicated in the previ­ous publication (Boz, Erdogu 2019). 2 Pit 187 was directly on top of pit 188 (F88) and has been interpreted as a different pit (Karamurat et al. 2021). However, fur­ther excavation revealed that human bones were extended into the assumed border of pit 187, indicating that this pit appears to be the upper part of pit 188. Therefore, the marking stone on pit 187 was in fact marking pit 188, as in other pits. A possible case of ‘accompanying dead’ in the second half of the 6th millennium cal BC at Ug¢urlu\Gökçeada, Turkey tures of the pelvis and skull (Buikstra, Ubelaker 1994). Changes in pubic symphyses (Brooks, Su­chey 1990) and occlusal tooth wear (Smith 1984) were used to group the adult skeletons into three broad age categories: young adults (20 to 30 years old), mature adults (30s and 40s), and older adults (50+). Dental development (Schour, Massler 1941) was used for subadult ageing. The age range of indi­viduals varies from 3 to 4 years to mature adults, and both biological sexes are represented among the 11 individuals found in the pit (Tab. 2). Taphonomic changes, trauma-related fractures, and other pathological lesions were examined macrosco­pically using osteological methods, as described in Silvia M. Bello and Peter Andrews (2006). The gene­ral condition of bones is poor and fragmentary. Some bones were crushed into a state that they were almost fused as if to form a bony layer. This condition applies especially to some bones of the last two individuals in the sequence (Indv. 1 and 2) whose bones were closer to the surface. Taphono-mic analysis of the better-preserved bones revealed numerous fresh bone fractures as well as dry bone breaks (Fig. 3). Macroscopic analysis of bone surface showed no evidence of cut marks that might be re­lated to traumatic events or sharp force wounds, chopping, or bone peeling. Description of bodies The first body to be interred in the pit was a 5- to 6­year old child (Indv. 11) which was found directly on randomly distributed stones on the floor of the pit (Fig. 4.a). The body was lying on its right side, with the right arm from the elbow and left leg lo­osely open towards to south, the rest of the body was in a flexed position. Different sizes of stones and animal bones were scattered di­rectly on top of this child’s skeleton along with articulated parts of two calves. Two large fragments of grinding stones (not parts of the same one) seemed to have been dumped into the pit. One was leaning perpendicular to the side of the pit, the other was partially covering the child’s head (Fig. 4.b). Above the child, five more bodies were piled up one after another with many dif-ferent-sized stones interred with them. Even though each skeleton was articulated individually, the position of each body was randomly scattered on top of each other Individual number Age group Sex 1 MA Female 2 MA Female 3 Adolescent (~ 18) Female 4 YA Male 5 Juvenile (~ 3–4yrs) Indeterminate 6 Juvenile (~10–11yrs) Indeterminate 7 MA Male 8 MA Female 9 Juvenile (~ 12 yrs) Indeterminate 10 MA Female 11 Juvenile (~ 6–7 yrs) Indeterminate Tab. 2. Age and sex distribution of individuals from grave pit 188. Skeleton numbers represent stratigraphical order from the top to the lowest le­vel. Age categories are: juvenile: 3–12 years; ado­lescent: 12–20; young adult (YA): 20–30; mature adult (MA): 30s–40s. which indicates that they were thrown in rather than placed carefully in the pit. Stratigraphically, the second individual (Indv. 10), a middle-aged adult female, was dumped in with her head on the floor facing down and with the body leaning perpendicu­lar against the wall of the pit. The hips and the legs were leaning on the side at a level that was about 50cm higher than the head. Her arms were twisted towards the back and the hands were found on the back of the body. The position of the arms and hands were unlikely to happen naturally unless they were tied at the back. The two biggest stones in the pit were thrown/ placed in the pit after this female. One of the stones was on top of her head, crushing her skull completely (Fig. 5). Before the other bodies were placed, sterile grey ash was thrown on top of the biggest stone in the middle, which weighed 80kg. Bas¸ak Boz After placing the ash deposit, four more bodies were dumped carelessly (Fig. 6). In contrast to these six people, a young adult male body (Indv. 4) ap­pears to have been intentionally placed on its left side, with the legs flexed tightly against the abdomen, the arms flexed tightly at the elbow on the chest, and the hands under the chin. This individual’s arms and legs might have been tied to the body firmly or wrapped with perish­able material before placement (Fig. 7). After this individual, five more bodies were dumped in the pit with a vast number of stones. Along with the mature adult female (Indv. 10) and young male (Indv. 4) mentioned above, another individu­al, a mature adult male (Indv. 7) also has a body position that suggests binding. This mature male’s body was in full articulation, with the head and the chest facing upwards and the legs up against the side of the pit. The legs were very tightly bent at the knees so that there was almost no space between the femur and lower legs. The arms were opened to the sides and bent tightly at the elbow to­wards the shoulders, the hands were resting on the shoulders. This peculiar position once again is very unlikely to occur naturally and suggests that the hands were tied at the neck and perhaps the legs might have been tied together too (Fig. 8). Except for these three indivi­duals with the possibility of binding, other bodies’ arms and legs were spread out within the pit (i.e. Indv. 3 and 8), and body parts of the dif­ferent individuals were inter­twined while the bones main­tained their anatomical posi­tions (Fig. 6). Apart from one skeleton (Indv. 6), all other individuals’ re­mains were intact with slight displacements of some bones, which appeared to be caused by gravity within the empty spaces of the pit. Some of the bone displacement was caused by the weight of the stones (i.e. the left femoral head of Indivi­dual 8 was removed from the acetabulum by the weight of the stone on top of the left knee, leaving one side of the pelvis and leg bones moved from its original position). Apart from these slight mo­vements of bony elements of some individuals, one child’s (6±2 years) body was separated into two A possible case of ‘accompanying dead’ in the second half of the 6th millennium cal BC at Ug¢urlu\Gökçeada, Turkey parts at the lumbar vertebrae. The child (Indv. 6) was located directly on the abdomen of Individual 7 with no soil in between, lying on its right side. The upper body was in full articulation down to the lum­bar vertebrae, including the hyoid bone and small manual phalanges. The bones of the lower body, on the other hand, were disarticulated and commingled. A few individuals from the upper layers have miss­ing skeletal elements. For example, chronologically the last individual placed in the pit (Indv. 1) was lacking one lower arm bone (right ulna), a young fe­male (Indv. 3) was missing the right fibula, and last­ly a 3 to 4 year-old child (Indv. 5) was missing the right lower leg bones. Discussion Deposition of the bodies Eleven individuals were squeezed into a pit measu­ring one meter wide and two meters deep, along with dozens of large and small stones (Fig. 9). Such an occurrence is rather bizarre and enigmatic. How could eleven bodies possibly fit in this relatively small pit? One possible scenario is that long inter­vals elapsed between the interments of each body, which would allow the previous ones to decay and occupy less space. Throwing/pla­cing stones would also weigh down the bodies. In this scena­rio, the pit must have been cov­ered with a lid to prevent soil ac­cumulation in between intern­ment events. The slight displace­ment of some bones of articulat­ed bodies (i.e. the sacrum of the vertically placed pelvis of Indivi­dual 7 slipped down with gravi­ty) suggests some void spaces within the pit during the decay process indicating that the pit was not filled with soil after the disposal of the bodies. The extre­mities of some individuals were found intertwined with other bo­dies and there was almost no fill between the bones (Fig.10). The only exception to this is the first child’s body at the bottom of the pit, where the body was covered with random stones and animal bones. The bones of none of the above individuals were mingled with or directly touching the bones of this child (Fig. 4). The other possible scenario would be that the bodies were kept unburied for some time before being moved to this death pit, perhaps for a special occa­sion. A ritual ceremony may have been performed, and these 11 bodies may have been moved to this pit to be interred together as a secondary burial. The Bas¸ak Boz only case that is plausible for this is the body of a child (Indv. 6) whose body was separated into two and whose lower body was partially disarticulated. These data could have indicated different stages of decomposition if only the bones of the upper body (even the small finger bones, and the hyoid bone) were not in full articulation. Therefore, it seems un­likely that this body was half defleshed when it was thrown into the pit. Cultural intervention (i.e. dis­memberment) is a likely option for this kind of se­paration of body parts that may have occurred be­fore placing the body into the pit. However, no cut marks were visible on the surviving bones to sup­port the argument of dismemberment. Another plau­sible case for dismemberment is the missing the right lower leg of a 3 to 4 year old child (Indv. 5); how­ever, the presence of animal holes throughout the pit could be another possibility for the loss of the small bones of this child. Along with this child, mis­sing single elements from two other individuals (the right ulna for Indv. 1 and the right fibula for Indv. 3) makes the dismemberment option unlikely for these two cases. Both Individual 1 and the legs of In­dividual 3 were close to the surface, and thus these single bone losses might have been the result of a later intervention (i.e. animals) or poor preservation. Neither scenario is firmly supported by the condi­tion of bones in terms of articulation, skeletal com­pleteness (with a few exceptions), the taphonomic evidence, or the depositional structure of the pit. It is therefore plausible that the bodies were deposit­ed in two different processes. The first one includes two actions; placing/throwing the first child (Indv. 11) as the first event, then the other ten bodies being deposited of as the second event after a certain time. The second option is that all bodies were deposited simultaneously in a single event or within short in­tervals (Fig. 9). However, these tentative attempts at explaining the deposition of 11 bodies in this rela­tively small pit remain highly speculative. Burial customs in western Anatolia in the second half of the 6th millennium cal BC3 Mortuary practices in the region of Thrace, Balkans, Greece, and Anatolia show great diversity during the Neolithic and Chalcolithic periods. However, despite the well-documented mortuary practices of central and south-east Anatolia throughout the Neolithic pe- riod (see for example, Andrews et al. 1995; Biçakçi et al. 2012; Boz, Hager 2013; 2014; Büyükkaraka-ya 2019; Erdal 2015; Özbasaran 2012; Öztan 2012; Lichter 2016; Bori. 2015), western Anatolia is less known. For north-west Anatolia, our information comes mainly from the Marmara region. During the mid­dle of the 7th millennium BC, at Barcin Höyük VI (6600–5900 cal BC), adults were buried in court­yards while infants were buried in abandoned hou­ses (Roodenberg et al. 2013; Özbal, Gerritsen 2019). At Aktopraklik, bodies are flexed and buried under the house floors in the Late Neolithic (6400– 6235 cal BC) layers. The emergence of cemeteries starts in the first half of the 6th millennium at the Early Chalcolithic level (5736–5635 cal BC) at Aktoprak-lik C in the abandoned settlement (Karul, Avci 2013; Lichter 2016.72). Also, at Ilipinar X/IX, most of the Late Neolithic and Early Chalcolithic burials were located in a non-residential area at around the first half of the 6th millennium (Lichter 2016.72; Rooden-berg 2008). In contrast, western Anatolia has revealed very lit­tle information to set a pattern of burial customs and related ritual activities during that period. For example, at Ulucak only a small number of infants have been found buried in the settlement during the 3 The process of Neolithization developed in different regions with different dynamics and different processes. Neolithization and subsequent changes in material culture in the excavated settlements in coastal west and northwest Anatolia have led some resear­chers to propose a different chronology and terminology for this region (for detail see Erdogu, Çevik 2020). Comparative chro­nology between sites and regions is provided in Table 3. A possible case of ‘accompanying dead’ in the second half of the 6th millennium cal BC at Ug¢urlu\Gökçeada, Turkey first half of the 7th millennium (Çevik, Abay 2016. 12; Çevik 2019.223), while sites like Yesilova and Ege Gübre have no information on the ways of dis­posing of human remains in the studied period. During the Early (the first half of the 6th millennium) and Middle Neolithic (the second half of the 6th mil­lennium) in the Balkans, burials are rare and found within settlements (Stratton et al. 2018.1). They are generally grouped into two body positions – flexed burials and extended supine burials (i.e. Cer­nia, Muntenia). According to Du.an Bori. (2015) the latter is a continuation of the local Mesolithic bur­ial customs in the region. The flexed body position was a new way of burying the dead in the region and is considered to originate from the Near East (Bori. 2015). There are also different ways of deal­ing with dead bodies and burials in Balkan prehis­tory. These forms of burials mostly consist of frag­mented bodies, recombination/hybrid burials, or the replacement of body parts with an object and remo­ving body parts and other forms (Chapman 2010; Chapman et al. 2014). Ditch burials are another type of burial, which are off-site burials from the eastern Balkans, such as at Nova Nadezhda (Bacva­rov et al. 2016). During the Neolithic in Greece, single or multiple simple pit burials were common with some variety such as cremations and pot buri­als (Fowler 2004; Chapman et al. 2014). Diversity in mortuary behaviours was also traced in the Neo­lithic of central Europe (for details see Hoffmann, Orschiedt 2014). Towards the late 6th millennium (5400–5000 cal BC) and early 5th millennium, extramural cemeter­ies emerged in the eastern Balkans in settlements such as Durankulak (c. 5000–4500 BC) and Cerni-ca (5355–5215 cal BC) (Stratton et al. 2018.23). This change in the way of disposition of the dead away from the settlements was interpreted as “po­tentially significant new relations between the li­ving and the dead” (Bori. 2015; Nikolov 2011; Stra­tton et al. 2018.2). On the other hand, diversity in the treatment of the dead continued throughout the Neolithic and Chalcolithic in the Balkans (Chapman et al. 2014). This diversity in the treatment of the dead in the 6th millennium extended to another type of ritual acti­vity. Some human remains were found within ritual places related to pits in the eastern Balkans. For in­stance, more than 20 off-settlement ritual areas in eastern Bulgaria have been reinterpreted as pit sanc­tuaries throughout the 6th millennium at sites such as Volaga-Ohaden, Dana Bunar 2, Kapitan Andreevo, and Voden (Nikolov 2011; Bacarov, Gorczyk 2017). These off-settlement places consist of many irregu­lar-shaped dug-out features and pits with filled frag­mented material including grinding stones, animal bones, etc., and closed by a cairn of stones. Among many of these ritual pits, only a small number in­clude human remains such as four single inhuma­ Bas¸ak Boz tions at early Neolithic Volaga-Oha-den (first half of the 6th millenni­um), five inhumations were found in Krum and two isolated skulls placed vertically in pits at Dana Bu-nar 2 and Usae in the second half of the 6th millennium (5400–5000 BC) (Nikolov 2011; Bacarov, Gorczyk 2017). Pit use is interpreted in var­ious ways, for example, Vasil Niko-lov (2011) interpreted 110 small cy­lindric pits around a big pit at Dana Bunar, as a place for ceremonial ac­tivities where sacrificial food along with fire was offered to the Mother-Goddess. John Chapman (2010) sug­gests that pit use could indicate how ritual behaviours produce social meaning or beliefs, whereas Krum Bacvarov and John Gorczyk (2017. 439–440), relate pit use as a way of creating a bond with a socio-culturally important landscape. In terms of the burial practices of Ugurlu, no burials have been found up to this date within the settle­ment, neither in Phase IV nor Phase III. However, this could be a sign of non-residential burial areas, as in western Anatolia and the Balkans, which makes it challenging to place pit 188, whether as part of the burial customs in earlier levels or a part of the changes at the site during the transitional period along with other changes in material culture and the settlement plan. Among these changes, pit con­struction starts to appear at Ugurlu in the first half of the 6th millennium and continue in the second half of the millennium as in the Balkans. Cansu Karamurat and colleagues (2021.8), argue that the emergence of pits at Ugurlu probably start­ed with building closure rituals in Phase IV repre­sented by three pits. During Phase III, pit rituals seem to be repeated more regularly and are concen­trated in the communal area. Based on the findings from the 37 pits with their fragmented material con­tents, the locations and marking stones for each pit seem to have significance and are interpreted as pla­ces for social negotiations (Karamurat et al. 2021. 15). Among these pits, three pits in Phase III in­clude a few pieces of human remains along with other fragmented materials. This behaviour of frag­menting and circulating items is possibly part of the same rituals which served people’s enchainment to the place and perhaps to a particular group. Frag-mentation and circulation of pieces of items and bo­dies are common practices in many societies through­out Near Eastern and Balkan prehistory, and are ge­nerally associated with transforming bodies and so­cial identities, gaining power, claiming places, nego­tiating relations in the societies and a bond between past and present (Chapman et al. 2014; Glencross, Boz 2016; Bloch, Parry 1982; Metcalf, Huntington 1991; Kuijt 2008; Karamurat 2013). Pit 188 was constructed during the second half of the 6th millennium (5389–5310 cal BC) which in many ways is similar to other pits at Ugurlu yet dif­fers somewhat by its differentiating properties, es­pecially the presence of articulated whole human bodies. Can we place this pit within the general pit rituals at the site, or could other motivations ex­plain the contents of the pit and the reason for its construction? Even though it is difficult to give a straight answer given the nature of the archaeologi­cal record, this pit with 11 human bodies could have been more than just another variation of human dis­posal within a region with a wide variety in terms of mortuary practice. In the following parts of the text, the other possibi­lities of the motivations for this pit will be discussed. A consequence of an epidemic The depositional structure of the pit contents seems to represent a single, if not two, depositional episo­des. General carelessness in depositing the bodies raises the question of whether these people died of A possible case of ‘accompanying dead’ in the second half of the 6th millennium cal BC at Ug¢urlu\Gökçeada, Turkey Date (cal BC) Northwestern-West Anatolia Sites Greece Balkans Sites c. 7000–6800 Epi-paleolithic Final Mesolithic c. 6800–6500 Initial Neolithic Barcin, Ulucak Ug¢urlu Initial Neolithic c. 6500–6000\5900 Early Neolithic Barcin, Aktopraklik Ulucak, Ug¢urlu Ege Gübre, Yes¸ilova Early Neolithic c. 6000–5700\5500 Late Neolithic Ulucak, Ug¢urlu Ege Gübre, Yes¸ilova Middle Neolithic Volaga-Ohaden c. 5700–5400\4900 Early Chalcolithic Aktopraklik Ilipinar, Ug¢urlu Middle Neolithic Early Neolithic (Star;evo) Volaga-Ohaden c. 5400–4600\4200 Middle Chalcolithic Ug¢urlu Late Neolithic Middle – Late Neolithic (Vin;a A, B) Durankulak Cernica, Dana Bunar 2,Usae Krum c. 4600-4200 Final Neolithic Chalcolithic (Vin;a C, D) Durankulak Tab. 3. Absolute and relative chronologies between the region and sites used for comparison. The compa­rative framework is mainly based on Çevik and Abay (2016), Erdog u and Çevik (2020), Tomkins (2007), Demoule and Perles (1999), Radivojevi. and Roberts (2021), Sampson (2018), Tsirtsoni (2018), Krauss (2011), and Blum et al. (2014). an acute illness. The expected mortality profile for the rapid death from disease should be similar to the living population, with a representation of all age groups with a higher proportion of infants, chil­dren, and the elderly (Margerison, Knüsel 2002). The presence of different age groups, including young children, adolescents, and adults, and the re­presentation of both sexes in the Ugurlu pit make it plausible for a deadly disease to be an alterna­tive cause of this deposition. The paleopathological analysis did not show any lesions on the bones, in­dicating that none of the individuals suffered from long-term disease, but this in no way diminishes the possibility of acute illness being the cause of death due to the fact that acute diseases do not show any traces on the surface of bones in a short period. Nevertheless, the indications of possible bindings of at least three individuals raises doubts about the acute disease scenario, since such binding requires close contact with sick, dead bodies. Why would they tie up the sick bodies instead of quickly getting rid of them? Moreover, the presence of some arte-facts (i.e. animal parts, the layer of ash, and grind­ing stones) indicates this was not a case of the quick deposition of sick, dead bodies, as they were most likely to be related to rituals. Victims of murder or ritual sacrifice The composition of the pit, what seems to be the ca­reless body positions of some individuals, and the indication of tied hands and legs of some other in­dividuals, makes the case intriguing. Could the mo­tivation for the construction of the pit and its depo­sition be a result of some sort of sacrificial killing or other types of destruction of one group within the community or outsiders? The existence of aggressive behaviours has been documented from various pre­historic sites spanning the Early Neolithic to Chalco­lithic. Some of the cases were interpreted as victims of prehistoric warfare, as in Halberstadt dated to c. 5600–4900 cal BC (Meyer et al. 2018), or the mas­sacre of local people by invaders, as in the case of Talheim in the early Neolithic (Wahl, Trautmann 2012). Here we can also note Asparn-Schlets in Au­stria (Teshler-Nicola et al. 1999), Potocani in Croa­tia (Novak et al. 2021), and Els Traocks in the Spa­nish Pyrenees, where Early Neolithic farmers were massacred, possibly over resources (Alt et al. 2020). In the case of Ugurlu, the lack of weaponry within the pit and the demographic composition of the group differs from that of war victims, where young and middle-aged males are predominant, although females and children are present in cases of attacks on a settlement (Wahl, Trautmann 2012). Analysis of the bones in pit 188 showed no fractures related to blunt force trauma or cut marks on the skull bones or the other bones of the skeletons. On the other hand, fresh bone breaks were detected on some bones, especially on the long bones of certain indi­viduals (Fig. 3). The presence of heavy stones di­rectly on top of some of these fractures indicates that Bas¸ak Boz these particular fractures may have been the result of the weight of the stones, and they were thrown/ placed when these bodies were still fleshed. The ability of the bones to retain the flexible collagen matrix after a certain time and the general frag­mented condition of bones make it difficult to as­sess if any of these fresh bone fractures were peri­mortem or early post-mortem taphonomic fractures (i.e. after death but still fleshed). The fractures that occur at the time in between the fresh bone and complete dry bone tend to show varying degrees of perimortem fracture characteristics (Wieberg et al. 2008). For this reason, it is not possible to use these fractures to argue for either the presence or absence of violence. Besides, the difficulty of finding the in­tent of observed injuries is a generally accepted issue (Schulting, Fibiger 2012.12). On the other hand, typical battle-related injuries such as skull (includ­ing face) trauma, since this is often the main target for interpersonal/intergroup violence, traumas on ribs and scapulae, or defence injuries to the fore­arms and hands were not observed on the bones. Another possible explanation for the human depo­sition is a ritual sacrifice, however recognizing a sa­crificial event and distinguishing it from other forms of ritualized killing in the archaeological record can be very difficult (Swartz 2017.224). Swartz summa­rizes possible indications of sacrifice in an archaeo­logical context. Analysis of architectural and strati­graphic context, repetition of the behaviour, devia­tions from normative human and animal deposi­tions, and signs of violent death are some of the in­dications for sacrifice (Swarzt 2017.228). Localities can be also important signs for sacrificial purposes if the bodies were placed in a different location than usual burial places. If the case is a retainer sacrifice then it can be identified through the positions of bo­dies, as the person who the others were sacrificed for will have a specific body position and sophisti­cated grave goods, while the others will have dis­respectful modes of disposition (Chenal et al. 2015). In the case of Ugurlu, pit 188 was located in the courtyard of the communal building along with other pits. This location seems to be special, al­though this does not entirely differentiate this pit from the others. This is also true for the pit con­struction and some of the inclusions. The specific structure of the pit in terms of lining with plaster, closing and sealing by stones, placing ‘marking sto­nes’, the inclusion of fragmented artefacts, and the ubiquitous presence of animal bones are the shared features of almost all pits. Therefore, based on con­struction specifications and contents other than the human remains, pit 188 seems to represent some of the common ritual behaviours related to pits. The presence of a distinct artefact could be a sign of a sacrificial act such as an altar, a specific ornament, or animal parts that might have been used during ritual behaviour (Swartz 2017). For example, an im­ported pink stone slab was found in the Skull Buil­ding at Çayönü Tepesi (southeast Turkey) which contained human and animal blood (Özdogan 2007; Croucher 2010). Pit 188 contains some fleshed parts of animals, specifically the upper legs and arms of two young calves, which is the only case within 37 pits at Ugurlu. Three small lumps of red ochre, a thin layer of sterile grey ash deposit, and broken grinding stones are also distinct depositions that could be remnants of offerings in pit 188. In terms of specific ornaments, none of the individuals had grave goods directly associated with the bodies, in­cluding the adult male (Indv. 4) in the centre, whose body was in a tightly flexed position as opposed to the other individuals with aberrant positions. It is unclear if the objects found in the fill were part of ritual activities and placed intentionally, or if they arrived there through accidental intrusion, as none of the artefacts were directly associated with the bod­ies. Philippe Lefranc and colleagues (2010; see also Schmitt, Déderix 2018), in a literature review, ag­reed that ‘accompanying dead’ seem to be wide­spread in the Late Neolithic Upper Rhine region (see also Testart 2010). However, it is difficult to con­clude that this is a common or ordinary way of dis­posing of the dead, since this type of burial has only been found in small numbers. Because of the vari­ety of different modes of disposal of the dead, in­cluding ‘non-normative’ body positions accompa­nied by grave goods, Lefranc and colleagues (2010) indicate that some of these disposals might have had other social implications than sacrifices for a high-status individual. An example from the late Neolithic site Bergheim in Colmar/France reveals a case where at least 14 individuals’ body parts were scattered within the pit and showed various modifi­cations, i.e. amputation, cuttings some limbs, and some traces of violence (Chenal et al. 2015). Fanny Chenal and colleagues (O.c.1322), suggest that these individuals were sacrificed to accompany a privi­leged individual. Another example is from the Diden­heim region in Alsace, where a man was accompa­nied by one adult and two children (Testart et al. 2010). A possible case of ‘accompanying dead’ in the second half of the 6th millennium cal BC at Ug¢urlu\Gökçeada, Turkey Traces of violence on human and/or animal bones are other indications of a sacrificial event if confirm­ed that they were not caused by warfare or murder (Swartz 2017.229). As explained earlier, there are no observable fractures that could be directly relat­ed to violence on the Ugurlu human remains (i.e. cranial blunt force trauma, cut marks on bones). Nonetheless, killing a person or an animal does not have to leave marks on bones. For example, heart extraction, throat-slitting, and decapitation were modes of sacrifice for the victims in Mesoamerica and Andes (Swartz 2017.232). Despite the lack of observable evidence of violence on bones, the presence of an adult male tightly bound in the middle (Indv. 4) whose body was ap­parently carefully placed as opposed to the other ten bodies, along with some special deposition (i.e. a layer of ash, ochre lumps) raises the possibility that the others might have been sacrificed for the individual in the centre. Conclusion The first proper human body deposition (except for a few fragments of human bone) has been found at Ugurlu, and has the potential to give insight into the ritual behaviours of the inhabitants of the is­land. Taking all the evidence together, pit 188 stands out as part of the pit ritual practice. However, the pit deviates from others by its inclusion of people with a variety of positions (side, at the back, and face down), including bodies with haphazard body posi­tions, one individual separated into half and three bound bodies, which makes interpretation a chal­lenge. The nature of this deposition within the wider pit tradition context at Ugurlu indicates a symbolic act, but the impetus and motives behind this specif­ic pit are difficult to interpret with certainty. Consi­dering the possibilities based on contextual evi­dence and information from other sites, ritual killing is certainly one possibility and may have been a re­gular practice at Ugurlu among the inhabitants dur­ing the 6th millennium, as also seen in the Balkans (see Chanel et al. 2015; Testart et al. 2010). If these ten people were killed to accompany the individual in the middle whose body was laid out intentional­ly, as Alain Testart and colleagues (O.c.) suggest, it may indicate a hierarchy within the society and its continuity to the world of the dead. However, Te-start offers another possibility of social relegation of the status of these people by burying them in uncon­ventional positions. Nonetheless, some of the accept­ed indications of ritualistic purpose can change meaning in different circumstances and in different communities. For some, dedication to underground or above ground forces, for a rebirth/renewal of anything, seasonal transitions, rain, fertility, power relations or to eliminate something bad or improve what exists, may all have been factors, but overall, they seem to have been for the same greater pur­pose of community cohesion. The phenomena were practiced in central and east Europe at around 4500 BC and going south at around 3500 BC (Testart et al. 2010), which sug­gests that this cultural behaviour travelled from the northeast to west and south (Chanel et al. 2010; Te-start et al. 2010). Nonetheless, Ugurlu’s earlier exam­ple with the date 5300 cal BC raises another ques­tion of the origins of this type of practice of human disposal. Further findings will provide more insight into the burial practices and ritualistic behaviours of the inhabitants of Ugurlu and their relation to other settlements around the region. ACKNOWLEDGEMENTS I am grateful to Prof. Dr. Burçin Erdogu and Assoc. Prof. Dr. Çigdem Atakuman for inviting me to study the material. I would like to thank Begona Rodriquez for her beautiful and painstaking illustrations, and also to Cansu Karamurat for kindly creating Table 3. I would also like to thank the anonymous reviewers for their valuable comments on the text. The author has no conflicts of interest with regard to this work. Bas¸ak Boz References Alt K. W., Tejedor Rodríguez C., and Nicklisch N. 2020. A massacre of early Neolithic farmers in the high Pyrenees at Els Trocs, Spain. Scientific Reports 10 (2131): 1–10. https://doi.org/10.1038/s41598-020-58483-9. Andrews P., Molleson T., and Boz B. 2005. The human burials at Çatalhöyük. In I. Hodder (ed.), Inhabiting Ça­talhöyük: reports from the 1995–99. McDonald Insti­tute Monographs. The British Institute of Archaeology at Ankara Monograph Series 38. Cambridge: 261–278. Bacvarov K., Gorczyk J. 2017. The Ritual Package at The Neolithic Pit Field of Sarnevo, South-Central Bulgaria. In M. Margarit, A. Boroneant (eds.), From Hunter-Gatherers to Farmers. Human Adaptations at The End of The Ple­istocene and The First Part of The Holocene. Papers in Honour of Clive Bonsall. Cetatea de Scaun. Romania: 439–451. Bacvarov K., Todorova N., Katsarov G., Petrova V., and McSweeney K. 2016. The Dead and The Nested Pots: An Early Neolithic Ditch Burial at Nova Nadezhda, Bulgarian Thrace. In K. Bacvarov, R. Gleser (eds.), Southeast Europe and Anatolia in Prehistory. Essays in honor of Vassil Nikolov on his 65th anniversary. Universitätsforschun-gen zur Prähistorischen Archäologie. Band 293. Verlag Dr. Rudolf Habelt GmbH. Bonn: 149–158. Bello S., Andrews P. 2006. The Intrinsic Pattern of Preser­vation of Human Skeletons and Its Influence on the In­terpretation of Funerary Behaviours. In R. Gowland, C. Knüsel (eds.), The Social Archaeology of Funerary Re­mains. Oxbow. Oxford: 1–13. Biçakci E., Godon M., and Çakan Y. K. 2012. Tepecik-Çift­lik. In M. Özdogan, N. Basgelen, and P. Kuniholm (eds.), The Neolithic in Turkey 3. New Excavations and New Research. Central Anatolia. Arkeoloji ve Sanat Yayinlari. · Istanbul: 89–134. Bloch M., Parry J. 1982. Death and the Regeneration of Life. Cambridge University Press. Cambridge. Blum S. W., Horejs B., and Mehofer M. 2014. The Middle Chalcolithic Cultural Sequence of the Troad (Northwest Anatolia): Chronological and Interregional Assessment. In B. Horejs, M. Mehofer (eds.), Western Anatolia before Troy: Proto-Urbanisation in the 4th Millennium BC. Pro­ceedings of the International Symposium Held at the Kunsthistorisces Museum Wien Vienna, Austria, 21-24 November, 2012. Austrian Academia of Sciences Press. Vienna: 125–155. http://www.jstor.org/stable/j.ctv5vdd0s Bori. D. 2015. Mortuary Practices, Bodies and Persons in the Neolithic and Early–Middle Copper Age of Southeast Europe. In C. Fowler, J. Harding, and D. Hofmann (eds.), The Oxford Handbook of Neolithic Europe. Oxford Uni­versity Press. Oxford: 927–957. Bori. D., Griffiths S. 2015. The Living and The Dead, Me­mory and Transition: Bayesian Modelling of Mesolithic and Neolithic Deposits from Vlasac, the Danube Gorges. Oxford Journal of Archaeology 34: 343–364. https://doi.org/10.1111/ojoa.12063 Boz B., Hager L. D. 2013. Living Above the Dead: Intra­mural Burial Practices at Çatalhöyük. In I. Hodder (ed.), Humans and Landscapes of Çatalhöyük: Reports from the 2000–2008 Seasons. Vol. 8. Cotsen Institute of Archa­eology Press. Los Angeles: 413–440. 2014. Making Sense of Social Behavior from Disturbed and Commingled Skeletons: A Case Study From Çatal­höyük, Turkey. In A. Osterholtz, K. Baustian, and D. Martin (eds.), Commingled and Disarticulated Human Remains: Working Toward Improved Theory, Method and Data. Springer. New York: 17–33. Boz B., Erdogu B. 2019. A sixth-millennium BC burial pit at Ugurlu on the island of Gökçeada. Antiquity 93(369): 1–5. https://doi.org/10.15184/aqy.2019.54 Brooks S., Suchey J. M. 1990. Skeletal age determination based on the os pubis: A comparison of the Acsádi-Nemes­kéri and Suchey-Brooks methods. Human Evolution 5: 227–238. https://doi.org/10.1007/BF02437238 Buikstra J. E., Ubelaker D. H. 1994. Standards For Data Collection from Human Skeletal Remains. In Proceedings of a Seminar at the Field Museum of Natural History. Arkansas Archaeological Survey Research Series, No. 44. Fayetteville. Büyükkarakaya A. M. 2019. Ölüm Uygulamalari ve Davra-nisin Ekolojisi. In A. M. Büyükkarakaya, E. B. Aksoy (eds.), Momento Mori: Ölüm ve Ölüm Uygulamalari. Ege Yayin­ · lari. Istanbul. Chapman J. 2010. Deviant Burials in the Neolithic and Chalcolithic of Central and Southeast Europe. In K. Re-bay-Salisbury, K. Sorensen, and M. L. Hughes (eds.), Body Parts and Bodies Whole. Changing Relations and Mean­ings. Oxbow Monographs. Oxford: 30–45. Chapman J., Wallduck R., and Triantaphyllou S. 2014. Dis­articulated Human Bone Disposal During the Mesolithic, Neolithic and Chalcolithic in The Balkans and Greece. An-nales Universitatis Apulensis. Series Historica 18(2): 11–45. A possible case of ‘accompanying dead’ in the second half of the 6th millennium cal BC at Ug¢urlu\Gökçeada, Turkey Chenal F., Perrin B., Barrand-Emam H., and Boulestin B. 2015. A farewell to arms: a deposit of human limbs and bodies at Bergheim, France, c. 4000 BC. Antiquity 89(348): 1313–1330. https://doi.org/10.15184/aqy.2015.180 Croucher K. 2012. Death and Dying in the Neolithic Near East. Oxford University Press. Oxford. Çevik Ö. 2019. Changing Ideologies in Community Making through the Neolithic Period at Ulucak. In A. Marciniak (ed.), Concluding the Neolithic. The Near East in the Se­cond Half of the Seventh Millenium BC. Lockwood Press. Atlanta: 287–305. Çevik Ö., Abay E. 2016. Neolithization in Aegean Turkey: Towards a More Realistic Reading. In Ü. Yalçin (ed.), Ana­tolia Metal 7: An Anatolian und seine Nachbarn vor 10.000 Jahren/Anatolia and Neigborus 10.000 Year Ago. Deutschen Bergbau-Museum Bochum. Bochum: 187–197. Demoule J. P., Perles C. 1993. The Greek Neolithic: a new review. Journal of World Prehistory 7(4): 355–416. Erdal Y. S. 2015. Bone or Flesh: Defleshing and Post-De-positional Treatments at Körtik Tepe (Southeastern Ana­tolia PPNA period). European Journal of Archaeology 18(1): 4–32. https://doi.org/10.1179/1461957114Y.0000000072 Erdogu B. 2014. Gökçeada Ugurlu Archaeological Project: A Preliminary Report from the 2011–2013 Field Seasons. Anatolica XI: 157–78. 2017. A preliminary report on the earliest Neolithic levels at Ugurlu on the island of Gökçeada. Anatolica XLIII: 71–82. 2020. Architecture and Organization of Early Chalco­lithic Settlement of Ugurlu (5500–4900 cal BC) on the Island of Gökçeada. In W. Stephan, E. Blum, T. Efe, T. L. Kienlin, and E. Pernicka (eds.), From the Past and Present: Studies in Memory of Manfred O. Korfmann. Verlag Dr. Rudolf Habelt GmbH. Bonn: 75–80. Erdogu B., Çevik Ö. 2020. Kiyi Bati Anadolu’nun kro­nolojisi ve terminolojisinin yeniden degerlendirilmesi. Ege Üniversitesi Arkeoloji Dergisi XXV: 45–66. Erim-Özdogan A. 2007. Çayönü. In M. Özdogan, N. Basge­len (eds.), Neolithic in Turkey. Arkeoloji ve Sanat Yayin­ · lari. Istanbul: 57–97. Fowler K. D. 2004. Neolithic Mortuary Practices in Gre­ece. BAR International Series 1314 Archaeopress. Oxford. Glencross B., Boz B. 2014. Representing Violence in Ana­tolia and the Near East during the Transition to Agricul­ture: Readings from Contextualized Human Skeletal Re­mains. In C. Knüsel, M. J. Smith (eds.), The Routledge Handbook of the Bioarchaeology of Human Conflict. Routledge. London: 90–108. Gürçal E. 2021. Gökçeada Ugurlu-Zeytinlik yerlesmesi­nin Kalkolitik Dönem Mimarisi ve Düzeni. Unpublish­ed PhD thesis. Trakya University. Sosyal Bilimler Ensti­tüsü. Arkeoloji Anabilim dali. Edirne. Hofmann D., Orschiedt J. 2015. Mortuary practices, bod­ies and persons in Central Europe. In C. Fowler, J. Hard­ing, and D. Hofmann (eds.), The Oxford Handbook of Neolithic Europe. Oxford University Press. Oxford: 987– 1004. Karamurat C., Atakuman Ç., and Erdogu B. 2021. Digging Pits and Making Places at Ugurlu During the 6th Millen­nium BC. Oxford Journal of Archaeology 40(1): 23–42. https://doi.org/10.1111/ojoa.12209 Karul N., Avci M. B. 2013. Aktopraklik. In M. Özdogan, N. Basgelen, and P. Kuniholm (eds.), The Neolithic in Tur­key. New Excavations and New Research. Northwestern · Anatolia, Arkeoloji ve Sanat Yayinlari. Istanbul: 45–68. Krauß R. 2011. Neolithization between Northwest Anato­lia and the Carpathian Basin – an introduction. In R. Krauß (ed.), Beginnings – New Research in the Appea­rance of the Neolithic between Northwest Anatolia and the Carpathian Basin. Menschen – Kulturen – Traditionen. Studien aus den Forschungsclustern des Deutschen Archä­ologischen Instituts 1. Verlag Marie Leidorf. Rahden: 1–7. Kuijt I. 2008. The regeneration of life: Neolithic structu­res of symbolic remembering and forgetting. Current An­thropology 49(2): 171–197. https://doi.org/10.1086/526097 Lefranc P., Denaire A., Chenal F., and Arbogast R. M. 2010. Les inhumations et les dépôts d’animaux en fosses circu­laires du Néolithique récent du sud de la plaine du Rhin supérieur. Gallia Préhistoire 52(1): 61–116. https://doi.org/10.3406/GALIP.2010.2471 Lihter C. 2016. Burial customs of the Neolithic in Anato­lia: An overview. In Ü. Yalçin (ed.), Anatolian Metal VII Anatolien und seine Nachbarn vor 10.000 Jahren Ana­tolia and Neighbours 10.000 years ago. Veröffentlichun-gen aus dem Deutschen Bergbau-Museum. Bochum: 71–83. Margerison B. J., Knüsel C. J. 2002. Paleodemographic comparison of a catastrophic and an attritional death as­semblage. American Journal of Physical Anthropology 119(2): 134–43. https://doi.org/10.1002/ajpa.10082 Metcalf P., Huntington R. 1991. Celebration of Death. Cambridge University Press. Cambridge. Bas¸ak Boz Nikolov V. 2011. A reinterpretation of Neolithic complex­es with dug-out features: pit sanctuaries. Studia Praehi­storica 14: 91–119. Novak M., Olalde I., Ringbauer H., Rohland N., Ahern J., and Balen J. 2021. Genome-wide analysis of nearly all the victims of a 6200 year old massacre. PLOS ONE 16(3): e0247332. https://doi.org/10.1371/journal.pone.0247332 Özbal R., Gerritsen F. 2019. Barcin Höyük in Interregio­nal Perpective. In A. Marciniak (ed.), Concluding the Neo­lithic. The Near East in the Second Half of the Seventh Millenium BC. Lockwood Press. Atlanta: 287–305. Özbasaran M. 2012. Asikli. In M. Özdogan, N. Basgelen, and P. Kuniholm (eds.), The Neolithic in Turkey. New Excavations and New Research. Central Anatolia. Arkeo- · loji ve Sanat Yayinlari. Istanbul: 135–158. Öztan A. 2012. A Neolithic settlement in Nigde-Bor Pla­teau. In M. Özdogan, N. Basgelen, and P. Kuniholm (eds.), The Neolithic in Turkey: New Excavation and New Re­search. Central Turkey. Archaeology and Art Publications. · Istanbul: 31–70. Radivojevi. M., Roberts B. W. 2021. Early Balkan metal­lurgy: origins, evolution and society, 6200–3700 BC. Journal of World Prehistory 34(2): 195–278. https://doi.org/10.1007/s10963-021-09155-7 Roodenberg A. S. 2008. The Neolithic cemetery – The an­thropological view. In J. Roodenberg, S. Alpaslan Rooden-berg (eds.), Life and Death in a prehistoric settlement in Northwest Anatolia. The Ilipinar Excavations III. Ne-therland Instituut voor het Nabije Oosten. Leiden: 35–68. Roodenberg A. S. A., Gerritsen F., and Özbal R. 2013. Neo­lithic burials from Barcin Höyük. The 2007–2012 Excava­tion seasons. Anatolica XXXIX: 93–111. Sampson A. 2018. The Mesolithic Hunter-Gatherers in the Southeastern Mediterranean and their Contributions in the Neolithisation of the Aegean. Archaeology and Cul­ture 1(1): 11–37. https://doi.org/10.22158/ac.v1n1p11 Schmitt A., Déderix S. 2018. What Defines a Collective Grave? In A. Schmitt, S. Déderix, and I. Crevecoeur (eds.), Gathered in Death. Presses universitaires de Louvain. Louvain: 195–214. Schour I., Massler M. 1941. Development of human denti­tion. Journal of American Dental Association 20: 379– 427. Schulting R. J., Fibiger L. 2012. Sticks, Stones, and Bro­ken Bones. Neolithic Violence in a European Perspec­tive. Oxford University Press. Oxford. Schwartz G. M. 2016. The Archaeological Study of Sacri­fice. Annual Review of Anthropology. 46(1): 223–240. https://doi.org/10.1146/Annurev-Anthro-102116-041434 Stratton S., Griffiths S., Kogalniceanu R., + 9 authors, and Whittle A. 2018. The Emergence of Extramural Cemete­ries in Neolithic Southeast Europe: A Formally Modeled Chronology for Cernica, Romania. Radicarbon 61(1): 319–346. https://doi.org/10.1017/RDC.2018.34 Teschler-Nicola M., Gerold F., Bujatti-Narbeshuber M., Pro-haska T., Latkoczy C., and Stingeder G. 1999. Evidence of genocide 7000 BP-Neolithic paradigm and geoclimatic re­ality. Colleqium Antropologium 23: 437–450. Testart A., Jeunesse C., Baray L., and Boulestin B. 2010. Les esclaves des tombes néolithiques. Pour la Science 396: 74–80. Tomkins P. 2007. Neolithic: Strata IX–VII, VII–VIB, VIA– V, IV, IIIB, IIIA, IIB, IIA and IC Groups. In N. Momigliano (ed.), Knossos Pottery Handbook: Neolithic And Bronze Age (Minoan). British School at Athens Studies 14. Bri­tish School at Athens. London: 9–2. Tsirtsoni Z. 2016. The chronological framework in Gre­ece and Bulgaria between the Late 6th and the Early 3rd millennium BC and the “Balkans 4000” project. In Z. Tsirtsoni (ed.), The human face of radiocarbon: reasses­sing chronology in prehistoric Greece and Bulgaria, 5000–3000 cal BC. Publications de la Maison de l’Orient et de la Méditerranée. Lyon: 13–39. Wahl J., Trautmann I. 2012. The Neolithic massacre at Talheim: A pivotal find in conflict archaeology. In R. Schulting, L. Fibiger (eds.), Sticks, stones, and broken bones: Neolithic violence in a European perspective. Oxford University Press. Oxford: 77–100. Wieberg D. A., Wescott D. J. 2008. Estimating the Timing of Long Bone Fractures: Correlation Between the Post­mortem Interval, Bone Moisture Content, and Blunt Force Trauma Fracture Characteristics. Journal of Forensic Science 53(5): 1028–1034. https://doi.org/10.1111/j.1556-4029.2008.00801.x back to contents Documenta Praehistorica XLIX (2022) Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe Noémi Beljak Pa/inová1, Alena Bistáková2 nbpazinova@ukf.sk 1 Constantine the Philosopher University in Nitra, Department of Archaeology, Nitra, SK 2 Institute of Archaeology of the Slovak Academy of Sciences, Nitra, SK ABSTRACT – Cremation was one of the permissible burial practices in the Linear Pottery Culture (LPC). This treatment has largely remained unexplained by previous research in Central Europe due to its descriptive nature. As a result, we present several thoughts on some key issues here. First, we discuss the current state of research in Central Europe, as well as some specifics with regard to the LPC and cremation. Second, we focus on two graveyards with exclusively or primarily cremation graves. Based on the current state of research, our goal is to evaluate the results obtained using a quantitative data analysis method, as well as an evaluation of the interpretation of cremation with­in the LPC population. KEY WORDS – Central Europe; Linear Pottery culture; cremation; state of research; customary mode of disposal Mrtvi in kremirani> o kremiranih pokopih v kulturi Linearne keramike v srednji Evropi IZVLE.EK – Upepelitev je bila ena od sprejemljivih pogrebnih praks v kulturi Linearne keramike (LPC). Ta praksa je v prej.njih raziskavah v srednji Evropi ostala nepojasnjena. Zato v .lanku pred­stavljamo nekaj razmislekov o nekaterih klju.nih vpra.anjih. Predstavljamo trenutno stanje razis­kav v srednji Evropi in nekaj posebnosti kremiranja v LPC. Osredoto.amo se na grobi..i z izklju.no ali predvsem .arnimi grobovi. Na podlagi trenutnega stanja raziskav .elimo ovrednotiti rezultate, ki smo jih pridobili z metodo kvantitativne analize podatkov in oceniti pojasnitve upepeljevanja v LPC populaciji. KLJU.NE BESEDE – srednja Evropa; kultura Linearne keramike; kremacija; stanje raziskav; obi.ajen na.in pokopa Introduction Linear Pottery culture (hereinafter LPC) played a number of LPC graves found in the burial context. key role in the Neolithization of Central Europe. It The situation was changed by the discovery of the is a period of the early Neolithic farming communi-biritual burial grounds in Kleinhadersdorf, Austria ties, and is dated between the middle 6th and the (Neugebauer-Maresch, Lenneis 2015) and Kralice second half of the 5th millennium BC (Manning et na Hané, Moravia (.míd 2006; 2012), and of the al. 2014). cremation cemetery in Mod³niczka, Little Poland (Czekaj-Zastawny, Przyby³a 2012). Not much attention has been paid to the topic of cremation burials of the LPC in Central Europe so The last person to deal with the analysis of the be- far, perhaps except for the study by Edith Hoffmann ginnings of cremation from the Mesolithic to the (1973). The main reason was the previously small Early Neolithic with a focus on the LPC culture with- DOI> 10.4312\dp.49.24 Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe in the wide territory of Europe was Agnieszka Gil-Drozd (2011), a decade ago. Therefore, the present study attempts to fill in this gap and utilize new data and information which have been added to the study of the topic. The concept of cremation within the presented archaeological material is used to indicate the burial rite, i.e. method of burial of a dead individual. According to Howard Williams (2004.Tab. 1) and Lise Harvig (2015) cremation can be defined as the use of fire to dispose of a body (burying a dead individual’s body), i.e. it is a process in which the dead body was partly or completely burned. For the analysis, we selected two examples (burial grounds), which in the current state of research seem to be the most suitable for using the method of correspondence analysis1 (CA). CA (Greenacre 2007) works on a presence/absence basis. It does not consider the number of items in any individual grave. To qualify for the analysis, each object and variable must be represented twice (each burial must contain two or more grave good types; each grave good type must occur in two or more buri­als). For both selected sites a catalogue of burials and grave goods was available. Besides aspects about the body (sex and age), we recorded the grave goods associated with the burial (number and type of grave goods in each burial). We were looking for po­tential patterns between the grave goods themsel­ves, and by adding in further information such as age and sex, we were able to see if these variables explain the variation seen in the dataset. We intend to find out whether the obtained terrain information or the state of cemeteries itself, is suit­able for this type of analysis and what results can be expected. We will study the place in which the cremation burials are situated, who was buried in them and what grave goods were added to the indi­viduals’ graves. We want to know what it says about the buried individual, and if the burial method re­flects the deceased person’s social identity. Historical context of cremation in hunter-ga­therer society Fire has been part of prehistoric burial rituals since the Mesolithic, with its cleansing but also destructive power (Gray Jones 2017; Larsson, Nilsson Stutz 2014). The number of known Mesolithic cremation burials has increased significantly in the last two decades. In Europe approximately 22 sites are re­corded, which seems to be a small number in com­parison to more than 200 sites of Mesolithic inhuma­tion graves with more than 2100 individuals (Grün-berg 2000.I, 51–54, 170, 171; 2008.40, 54; Gray Jo­nes 2017.Fig. 2.1; Little et al. 2017.Tab. 2). Specific features of burial rituals associated with Mesolithic cremations include, for instance, collecting or clean­ing of bones before the burial (e.g., Hammelev, Gon­gehusvej 7 – grave A) or, on the other hand, the cu­stom when bones were not completely collected from the funeral pyre (e.g., Hermitage – pits A and B, Gongehusvej 7 – grave N, Oirshchit V, Coswig, Rotterdam, Heffingen?, la Chaussée-Tirancourt; see Gray Jones 2017.Tab. 2.1). The research is focusing on various methods of depositing the bodily re­mains in the ground (Fahlander 2012; Gray Jones 2017.41–45) as well as on interpretation of the va­riability in the number and typological composition of grave goods to confirm their active role in the bu­rial ritual (Little et al. 2017.235–236). Within Cen­tral Europe, cremation in the Mesolithic was dis­covered in Poland and in the territory of the great Danube Bend on the border between Serbia and Ro­mania (the region of the Iron Gate). In Poland, this includes the cremation of an adult male(?) in a shal­low pit on the Wieliszew VII site and the cremation of several individuals (at least one child and other individuals) in a dwelling at the Pomorsko 1 settle­ment (Sulgostowska 2006.196). Intentionally burn­ed human skeletons were recorded in four Mesoli­thic graves at Mszano 14, which the author consid­ers to be evidence of a combination of cremation and inhumation (Marciniak 1993.7). In the Iron Gate region at the site of Vlasac, 18 con­texts with burned human remains were examined between 1970 and 1971, and the number of docu­mented cremations increased further between 2006 and 2007, when seven more cremation graves and at least six secondary cremations were found in ske­letal graves (Bori. et al. 2009.247–282; 2014.14, 20, Tab. 3). The phenomenon of secondary crema­tion consisted of disarticulation of some parts of the skeleton and their tanning/burning, while other parts of the skeleton remained without cremation (Bori. et al. 2009.257). The secondary cremation on the site was characteristic of the Late Mesolithic and had the purification aim to prepare the spot for a new grave (Bori. et al. 2009.272, 273). This habit 1 Correspondence Analyses was carried out using the PAST software (version 4.05 online on: https://www.nhm.uio.no/english/rese arch/infrastructure/past/) (Hammer et al. 2001). Noémi Beljak Pa/inová, Alena Bistáková persisted until the transitional period of Mesolithic-Neolithic dating in the region between 6200–5900 BC (Bori. et al. 2014.20, 25). All in all, the Mesolithic cremations in Europe pro­vide clear evidence that hunter-gatherers knew and used cremation as a method of disposal of human bodies. It is also obvious that cremation of the dead as well as depositing utility artefacts/tools on the pyre or in the grave were important aspects of the ritual behaviour of communities (Lenneis 2007). Some actions were undoubtedly associated with co­oking or a funeral feast. We can conclude that the ri­tual depositing of cremated remains together with grave goods at a previously decided place was a cha­racteristic feature of prehistoric life before the start of the Neolithic. It is also true that cremation was not the only burial practice at Mesolithic burial grounds (sites), but was closely associated with inhumation or another form of burial customs (Bori. et al. 2014, 20; Gray Jones 2017; Little et al. 2017). Evidence of biological interaction between incoming farmers and local hunter-gatherers during the ear­liest stages of the arrival of farming in Neolithic Cen­tral Europe (the formative LPC phase) has also re­cently been proved based on bioarchaeological ana­lysis of the remains of the interred at Brunn 2 site of the Brunn am Gebirge-Wolfholz complex, one of the oldest LPC sites (Nikitin et al. 2019). Further­more, prior studies also found a limited presence of hunters-gatherers in the DNA of the LPC population (e.g., Shennan 2018; Lipson et al. 2017; Szécsényi-Nagy et al. 2015; Haak et al. 2010). On the other hand, cremation was not often prac­tised in the Starevo-Körös-Cris culture, which was the likely ancestor of the LPC in Europe (Bánffy et al. 2007). The oldest incineration grave is M7 from the settlement Gura Baciului (Star.evo-Cris culture) (Vlassa 1968.371–379). Until now, it is the only cer­tain discovery for the Carpathian-Danubian Early and Middle Neolithic (Gligor, Ba¢cuet-Crisan 2014. 50). The find is unique not only for the peculiarity of the cremation grave in the given timeframe, but also for its position between the fragments of the house debris (complex P24), which relates to rituals com­mitted to the priests of the time. The finding of a mo­numental stone head (stela) covering the burial em-phasises the cult’s supremacy (Lazarovici, Lazarovi­ci 2006.107, Fig. 6). Another discovery that demon­strates the Körös culture’s use of cremation is like­wise cultic in nature. It was discovered in a second­ary location on the outskirts of the tell Hódmezvásár­hely-Gorzsa I (Kovács-tanya). The lower part of an anthropomorphic container called the ‘Venus of Gor­zsa’ contained evidence of charred skull fragments and grain (Gazdapusztai 1957.12, Tabs. I.3, II). Cal­cined bone fragments belonged to the skull of a 60­year-old male (Paluch 2012.182; Farkas 2005.13). Linear Pottery culture burial practices The LPC demonstrates the diversity of funeral cus­toms. Generally left-sided crouched inhumations do­minate, right-sided crouched positions, cremations, double graves, prone and supine positions are recur­rent; partial burials, disarticulated bodies, cenotaphs, and post-mortem manipulations are also document­ed. In addition, cemeteries and burials in settlements coexist in most areas (Pechtl, Hofmann 2013; Bic-kle, Whittle 2013.46–48, 57–58, 107–114, 168–170; Hofmann, Orschiedt 2015; Zeeb-Lanz, Haack 2016). With regard to documented burials in Central Eu­rope, inhumations with crouched bodies lying main­ly on their left sides prevailed (Lichter 2003.139; Ka­licz, Makkay 1977; Oross, Marton 2012. 282–291). Burial districts/areas are first recorded in the Late LPC, e.g., in Austria, Kleinhadersdorf (Neugebauer-Maresch, Lenneis 2015) and in Slovakia, Nitra-Prie­myslová ulica (Pavúk 1972). The proximity of bur­ial grounds to settled areas, like in western Europe, e.g., in Elsloo, the Netherlands (Moddermann 1970) or in Arnoldsweiler, Germany (Cziesla, Ibeling 2014. 125–150) is also confirmed. In Vedrovice, Moravia, dated to LPC I and LPC II, graves were located at a separate burial ground as well as in the residential area (Podborský 2002.9–21; Pettitt, Hedges 2008). The biritual burial ground Kralice na Hané, Moravia, was also situated near the contemporary settlement (.míd 2012.14–15). The dead buried in the residential area near houses were either in graves or settlement pits, e.g., in Slo­vakia – .túrovo (Pavúk 1994.94–99); Hurbanovo-Bohatá (Bøezinová, Pa.inová 2011.168–171), in Moravia – .ádovice (.i.máø, Geislerová 1997); Br-no-Bohunice (Do.kalová, .i.máø 2008.43); in Hun­gary – Balatonszárszó-Kis-erdei-dõlõ (Marton 2008. 197–198; Kreiter et al. 2017.113); Harta-Gátõrház (Kustár et al. 2014.31–33); Füzesabony-Gubakút (Al-föld/eastern LPC; Kalicz, Koós 1997; Bickle, Whit­tle 2013.64–66); Mezõkövesd-Mocsolyás (Szatmár II group/Alföld LPC; Raczky et al. 1997.28–33; Ka­licz, Koós 2014.65–69); in Poland – Miechowice 4, Brzeœ. Kujawski 3 (Grygiel 2004); Ludwinowo 7 (Czerniak, Kabaciñski 2004.154–155); Stary Za­ Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe mek (Kulczycka-Leciejewiczowa, Romanow 1985. 45–47). A burial directly connected with a house (foundations) has been confirmed in Little Poland (Kulczycka-Leciejewiczowa 2008.176) and in Slo­vakia (Bátora 1999). Daniela Hoffman (2009.230–232) concluded that in both such cases it is a regular burial act, and thus that burials within settlements do not mean any ex­ceptional burial tradition (Sonderbestattung in Ger­man), and this also undoubtedly applies to Central Europe. For instance, in Slovakia, the study of Mi­chaela Niklová (2014) revealed that LPC burials at settlements became a fixed feature of funeral cus­toms and are one of the hallmarks of this period. In Germany, mass burials probably caused by vio­lent conflicts (Talheim near Stuttgart, Wahl, Traut­mann 2012; Schöcken-Killianstädten near Frank­furt, Meyer et al. 2015) and cases of post-mortem manipulation (e.g., Herxheim in southwestern Ger­many; Zeeb-Lanz 2019) have been discovered. In Central Europe, we can find some parallels too. Clear evidence for severe inter-personal violence was documented in Lower Austria (Asparn/Schletz; Tesch­ler-Nicola 2012). Secondary manipulation with re­mains has been discovered at the site of Bicske-Ga­lagonyás (Makkay et al. 1996.20, 21, Fig. 6) in Hun­gary, where re-depositing of the upper left limb (fe­mur, tibia, fibula) in the anatomic position beyond the head (without fingers) shortly after placing the dead in grave 1 occurred (end of LPC/Sopot-Bicske). Post-mortem manipulation was also confirmed by a newly discovered LPC site in south-west Slovakia – Vráble, where ‘regular’ burials with the deceased in a crouched position were placed on either side of the large outer ditch within the settlement (Müller-Scheeßel et al. 2021.66–69, Fig. 2). Besides that, the second category of mortuary treatment (‘irregular’ burial type) was represented here by three headless individuals (Müller-Scheeßel et al. 2021.69, 70; Fig. 3.1, 2, 5), who were each placed in an extended po­sition at the bottom of the outer ditch. These buried individuals show clear signs of post-mortem manipu­lation, the skulls were removed during the decompo­sition process. Possible open-air staging of the dead for a certain period presents a burial ritual not pre­viously taken into consideration for the LPC. Such a practice could explain seemingly disturbed burials found also elsewhere (Müller-Scheeßel et al. 2021. 74, 80). As for the grave goods of the buried LPC individuals in Central Europe, older graves usually did not con­tain any goods (e.g., LPC graves at the burial ground in Balatonszárszó – Kis-erdei-dõlõ: Oross, Marton 2012.263), while the Late LPC graves or the .elie­zovce group graves included burials with above-average grave goods: e.g., in Budapest-Nagytétény (Gallus 1936), Budakeszi-Szõlõskert (Czene, Otto-manyi 2007; Czene 2008) and Baj.-Medzi kanálmi (Cheben 2000.72, Abb. 11.1, 12, 13). In general, we can state that graves without goods make up more than half of the known burials (Veit 1996.Tab. 4). We can use the cemetery in Nitra-Priemyslová ulica (Slovakia) as an example, where burials without grave goods made up as much as 62.2% of all gra­ves (Peschel 1992.176). Cremation in the LPC Finds from the LPC contributed profoundly to the study of cremation graves and the beginnings of the cremation burial rite (Gil-Drozd 2011.11). As early as the beginning of the 20th century, the first biri­tual burial ground of the LPC was discovered in Arn­stadt, Germany (Hoffmann 1973.71). The twelve cremations discovered here, besides ten inhuma­tions, represented the first case of a previously un­known method of depositing burned remains in small heaps and then covering them with vessels placed with their bottoms up. Other continuously appearing rather large biritual burial grounds of the LPC in the territory of Germany (Aiterhofen-Öd­mühle: 69 cremations and 159 inhumations; Wan-dersleben-Gotha: 132 cremations and 179 inhuma­tions; Stephansposching: 31 cremations and 10 inhu­mations) and the Netherlands (Elsloo: 47 cremations and 66 inhumations) allowed wider considerations of the burial rites of the population. Christine Pe­schel (1992), for instance, found out that cremation burials at most biritual burial grounds contain few­er grave goods. However, differences between indi­viduals of different genders or ages were not noti­ceable. Based on the analysis of the burials, Peschel (1992.199) formed a hypothesis that individuals with lower social status were cremated, while the richer social class had full-body burials. Norbert Nieszery (1995.17, 18, 43, 44) was among the first who suggested that cremation was the do­minant form of burial in the Neolithic. He compar­ed the number of uncovered graves with the cove­rage of LPC settlements and concluded that only around a fifth of the population is buried in the bu­rial grounds, as most of the shallower cremation bu­rials were eroded and destroyed by farming activi­ties. Noémi Beljak Pa/inová, Alena Bistáková Iris Trautmann (2006.183, 185) presented some important insights with regard to the genesis and development of cremation in Europe. Based on the analysis of burnt bone remains of the LPC in Ger­many and their comparison with skeletal remains, Trautmann concluded that practising various funer­ary rites is a result of affiliation with different groups of the population, i.e. Mesolithic hunters and gathe­rers would often unite with a Neolithic population that had newly arrived in an area, still retaining their funerary traditions, in addition to the new customs practiced by Neolithic communities. Overall, we can conclude that biritual cemeteries were once especially frequent in Central Germany, Bavaria, and the Netherlands (Jeunesse 1997). From a funerary perspective, while inhumations dominate, they coexist side by side with recurrent cremations in most areas. For example, the percentage of crema­tion burials is the following at selected sites: Elsloo 42%, Wandersleben 42.44%, Niedermerz 9.73%, Schwetzingen 4.45%, Fellbach-Oeffingen 7.27%, Ai­terhofen 30.13%, Stephansposching 75.61%, Arn­stadt 54.55%, Niederdorla 23.07% and so on (Mod­derman 1970; Nieszery 1995; Jeunesse 1997; Cze­kaj-Zastawny, Przyby³a 2012.49, Fig. 31; Gerling 2012). Moving to Central Europe (Fig. 1) the finds include mostly separate cremation burials. In Bohemia, a cre­mation burial of a five-to-six-year-old child, probably a male, was uncovered in a construction pit of the Late LPC near a longhouse in Litice near Plzeò (Braun 2001; Pavlù, Zápotocká 2007. 83). Other cremation burials are known from the Moravia region. In the south-west part of the settled area in Brno-Starý Lískovec, a cremation grave together with burned animal bones was document­ed (Do.kalová, .i.máø 2007. 34). In Prostìjov-.echùvky, three cremation burials were uncovered at the Kopaniny site, probably from a larger burial ground (.míd 2011. 8). Cremated human remains were also documented at the LPC cemetery Vedrovice-.iro­ká u lesa (Skutil 1941.28, 29). Nevertheless, their condition did not allow further analyses. In Poland, a cremation burial of probably the East­ern LPC was discovered in Gródek Nadbu¿ny in Lit­tle Poland, where the grave itself was documented by remains of burned bones, secondarily burned pottery fragments (probably from two globular ves­sels) and pieces of charcoal, arranged in and east-west direction (Kempisty 1962.284, 285). A possible LPC cremation burial (without finds, but near to a LPC clay extraction pit) is mentioned at Zwiêczyca 3 in south-eastern Poland (Debiec et al. 2014.107, 142), where pit No. 409 (31x30cm) contained the burned remains of an adult individual (maturus, 35– 50 years old), probably a male. In Slovakia, at the only LPC cemetery in Nitra–Pri­emyslová ulica, in addition to 72 inhumation buri­als of the Late LPC/.eliezovce group, at least eight groups of burned human bones of adults as well as children without preserved grave pits or associated grave goods have been reported (Pavúk 1972.39, 69). Reliable evidence of cremation comes from the LPC cemetery in Kleinhadersdorf in Lower Austria, where up to 100 burials are estimated. Among them, there were 57 inhumations, 26 symbolic (empty) graves and at least four cremations (Neugebauer-Maresch 1992.5–6; Neugebauer-Maresch, Lenneis 2015). Al­together, seven graves with traces of burned bones were found at the site. However, only four of them (graves 37, 46, 54 and 82) were identified as crema­tion burials (Neugebauer-Maresch, Lenneis 2015. Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe Tab. 5). Most of the cremations were in the north­ern (older) part of the cemetery, and one of them (grave 37) in the southern (younger) part. Their dat­ing falls into the Late LPC phase IIa (grave 54), or .eliezovce group – LPC III (grave 37). An inhuma­tion burial (grave 37a) was documented very close to grave 37. Similarly, inhumation burial 55 was located near to grave 54. The sizes and shapes of grave pits were difficult to identify, but it is proba­ble that they were approximately circular, with dia­meters between 40 and 80cm. Grave 54 has larger dimensions – 139x116cm – and thus exceeds this range (Neugebauer-Maresch, Lenneis 2015.Tab. 5). Grave goods were found only in two graves (37 and 54) and did not differ from grave goods in inhuma­tion burials (Neugebauer-Maresch, Lenneis 2015. Tab. 13). They consisted mainly of pottery, flakes, adzes (grave 37) or a bone tool (grave 54). No other pyre cremation remains (e.g., ashes, charcoals) have been found. The remains of burned grave goods were not stated. The discovery of the biritual burial ground at the site of Kralice na Hané (.míd 2012) in Prostìjov di­strict, Moravia, brings a significant change in our opinions on cremation in the LPC in Central Europe. Of equal importance is the cremation burial ground (39 graves) Mod³niczka, site 2 in Little Poland (Cze­kaj-Zastawny et al. 2011.53; Czekaj-Zastawny, Pr­zyby³a 2012). We deal with both cemeteries more in detail below. Example 1. Kralice na Hané (Moravia) – biritu­al cemetery of the Late LPC Kralice na Hané is located on the alluvial flat (214m a.s.l.) of the Rom.e flood plain in central Moravia, in the region of Olomouc (.míd 2012.10). During the construction of the industrial zone of Prostìjov in Kralice na Hané, a large multi-phase settlement was excavated in 2002–2012, including a LPC settle­ment. On the northern edge of the LPC settlement, a contemporary biritual cemetery (Fig. 2) was disco­vered in 2005–2006. The results of the excavation were published with detailed descriptions of find contexts, a catalogue of graves and grave goods, an­thropological analysis, etc. (.míd 2012; Stránská 2012). The burial ground was situated along the northern edge of the LPC settlement. The cremation burials were only 0.3–0.4m below the current terrain. The estimated size of the cemetery is 1.8ha (dimensions: 300x60m). However, a considerable (central) part was destroyed in the second half of the 15th cen­tury because a pond was built, followed by the sub­sequent modern cultivation of soil (.míd 2012.15). The excavated area of the cemetery makes up only one-tenth of the burial site’s area. Seventy-eight gra­ves in total were uncovered – there were 69 crema­tions, eight inhumations and one empty grave pit, as well as one burial of a child without the outlines of the grave pit (.míd 2012.9, 72, Obr. 5). Dating of the cemetery The older phase (stage LPC I–early II) of the ceme­tery was characterized mainly by inhumation buri­als of crouched individuals in rectangular to oval grave pits. From the total of 30 cremations with datable pottery, only three burials (graves 15/05; 20/05; 28/05) belonged to the older phase (LPC Ib, I/II). A change in the funerary rite occurred in stage LPC II (cremation was used exclusively from LPC IIb on) – as many as 26 burials belonged in this pe­riod (graves 1/05; HOPI 1/06; 6/05; 7/05; 9/05; 23/ 05; 26/05; 30/05; 31/05; 32/05; 1/06; 4/06; 5/06; 6/ 06; 7/06; 8/06; 11/06; 17/06; 18/06; 21/06; 22/06; 23/06; 24/06; 25/06; 26/06; 28/06). One grave (9/ 06) was dated as the youngest (LPC IIc/III). Another 39 cremations were not suitable for dating (.míd 2012.72–77). Cremation graves2 Graves were detected on the level of topsoil and subsoil (approx. 0.3m deep). Grave pits were made up of clusters of burned small bones which were mainly situated in bowl-shaped depressions whose preserved depth varied from 2cm to max. 30cm. The grave pits were mainly circular in shape, with a dia­meter from 0.3m (13 examples), 0.4–0.5m (16 exam­ples) or 0.52–0.7m (nine examples) to 0.74–0.9m (seven examples). In 19 cases, the shape of the grave pits was oval with the dimensions ranging from of 0.3x0.4m to 1.05x0.65m. Any association between age, gender and size of the grave pit has not been confirmed (.míd 2012.108), e.g., in small pits with a diameter of 0.3m children (graves 16/05; 22/05; 11/06) as well as adults (graves 4/05; 11/05; 28/ 05; 26/06) were buried. Even a double burial (child and adult) was identified in a pit with a smaller dia­meter (grave 17/05). However, a young child (0.5–6 years old) in grave 18/06 was buried in a pit of 0.92x 2 In the book (.míd 2012), we noticed slight differences between the catalogue and the analytical parts. Thus, we follow exclusively the catalogue part when processing graves, their goods, dating and anthropological identification (.míd 2012.19–60). Where our results are identical and we adopt the previously discovered facts, we give the exact citation of the analytical part of this earlier study. Noémi Beljak Pa/inová, Alena Bistáková 0.8m and an adult (grave 21/06) was deposited in a pit of 1x0.6m. Almost half of the remains weighed less than 10g per burial (Stránská 2012.126). The surface colours of the burnt bones were grey to a blue-grey, the ave­rage temperature of burning reached approx. 550°C. The examined remains suggest uneven cremation depending mainly on the access of oxygen to indivi­dual body parts (Stránská 2012.126, Tab. 2). A fu­neral pyre was not discovered in the cemetery or its immediate vicinity (.míd 2012. 109). Immature individuals were buried in 16 graves (In-fans I-III, or Infans III-juvenis). A child together with an adult was buried in two graves (3/05 and 17/05). In 26 cases, adults were buried. Two in juvenis­adultus age (grave 5/05 and 41/05), another two were identified as adultus I–II (grave 25/05 and 42/ 05), two others were older than 30 (grave 3/06 and 8/06), and the oldest individuals included one adult in adultus II-maturus I age (grave 27/05) and male adult in maturus I–II age (grave 30/05). Gender was identified only in three cremation burials. A woman was buried in grave 19/05, probably a man was de­posited in grave 32/05, and the oldest individual was a man in grave 30/05. Two superpositions were also found in the ceme­tery. In one case, inhumation burial 33/05 was in su­perposition with cremation grave 30/05 (.míd 2008. 251). Another superposition was detected in grave Fig. 2. Kralice na Hané, Moravia. Biritual cemetery. Location of LPC graves: a cremation, b inhumation (after .míd 2012.Obr. 6). Colour-coded graves suitable for correspondence analysis (child – green; adult – black). Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe 8/06, where a vessel dated to stage LPC IIb was in the upper part of the ‘fill’ of the grave pit with a ves­sel with archaic decoration at the bottom (LPC IIa), i.e. it could be a disturbance of an older cremation grave with a younger cremation grave (.míd 2012. 105). Grave goods Fifteen (10%) of cremation burials did not contain any grave goods. Nevertheless, many graves were disturbed or very shallow (just below the topsoil/ subsoil), and thus it is very probable that grave goods were richer and the ‘empty’ graves originally contained something. Twenty (14%) cremations con­tained exclusively pottery preserved mainly in frag­ments. Vessels or their fragments were present in 45 graves (31%). Complete vessels were uncovered in grave 8/06, where two vessels deposited with their bottoms up were documented (.míd 2012.49, Obr. 47), in grave 9/06, there were three vessels with their bottoms up (.míd 2012.51, Obr. 48). One com­plete vessel (lying) occurred also in grave 21/06 (.míd 2012.55, Obr. 53). Traces of fire were record­ed also on most grave goods and vessels with tra­ces of burning occurred also in inhumation burials (grave 33/05 and 46/05). Nineteen graves contained polished stone industry. In 18 of them, children’s or adults’ adzes or their parts were found. One exemplar was also discover­ed in the double burial of a child and an adult 3/05. In one grave of an adult individual (grave 27/05), a complete stone hoe was found. Only seven adzes were not damaged (in graves 4/5, 9/05, 24/05, 38/ 05, 3/06, 21/06, 23/06), four others were perforat­ed (in graves 1/05, 6/05, 45/05, 28/06) and other­wise damaged – it is not clear whether they bear tra­ces of intentionally damaging or a result of internal tension caused by the high temperature of a pyre (.míd 2012.83). Chipped stone industry (mainly silex blades and fla­kes) was present in 10 burials in form of miniature burned fragments of former tools. They were pre­sent in children’s burials (grave 15/05, 1/06 HOPI), but dominated mostly with adults or old men (gra­ves 27/05, 30/05, 32/05, 4/06, 6/06). It was also found in the grave 3/05 where a child was buried together with an adult individual. In the other two graves (26/05 and 9/06), it was not possible to iden­tify the age or gender of the buried. As for the num­ber of flakes in individual graves, most often there was one example, with a maximum of three. Burial 30/05 with an older man (maturus I-II) was an ex­ception to this, as 12 fragments of burned silex bla­des were discovered among the cremation remains (.míd 2012.32, Obr. 26). Grinding stones (without visible traces of modifica­tion) were found in two graves (15/05 and 9/06); their central (symbolic) position was obvious, as other grave goods were arranged around them (.míd 2012.82). In three graves (12/05, 15/05, 31/05) the remains of bone (species: sheep/goat) pointed tools were preserved. It was even possible to restore the bone burin from grave 31/05 (.míd 2012.Obr. 27.3). In the other two cremation graves the remains of meaty food (medial phalanx of sheep/goat with the child in grave 1/06 HOPI; radial bone of a rabbit with the adult in grave 27/05) were discovered (.míd 2012. Tab. 8). Red pigment (hematite) was recorded in nine (6%) cremation burials (graves 7/05, 8/05, 15/05, 23/5, 32/05, 41/05, 42/05, 1/06 HOPI, 6/06). It was found in the form of lumps of several millimetres up to 3.5mm, and its occurrence in the cemetery was not dependent on the age or gender of the buried indivi­dual (.míd 2012.84). In a single child grave 15/05, a lump of raw graphite was confirmed (35mm long). Correspondence analysis Twenty-eight cremation burials with grave goods (Fig. 2) met the relevant CA criteria. The plot (Fig. 3) does not show significant deviations, and no notable hidden data are visible. However, the graves with mineral pigment show an interesting position. In to­tal, lumps of red pigment were discovered in nine cremation burials – eight of them (apart from grave 8/05) were analysed. Four burials of probably adult individuals (graves 7/05, 23/05, 41/05, 42/05) with hematite constitute a separate group, outside the main cluster of burials. Nearby, burial 6/06 of ano­ther adult individual with lumps of red pigment is located. It appears that the inclusion of ochre in a grave was based on a variable distinct to other choi­ces of grave goods. Of course, it is worth emphasi-sing that the CA, in this case, is based on a very small sample number, but the lack of correspon­dence between ochre covered burials and those with­out is interesting, both in terms of those with grave goods and without. The remarkable proximity (and thus good correspon­dence) of children’s graves 15/05, 31/05, 1/06 HOPI and 18/06 is also interesting. There is also a double Noémi Beljak Pa/inová, Alena Bistáková burial of a child with an adult (grave 3/05) nearby. In this case, we can assume that the grave goods of children corresponded with the standard tradition and did not exceed the conventions very much. The last not very isolated group consisted of graves (probably adult men) containing adzes together with pottery (graves 4/05, 6/05, 9/05, 20/05, 3/06, 7/06, 21/06, 23/06, 25/06, 28/06). This combination was probably commonly used as well, although due to the small size of the sample, we cannot speak of a universal rule now. Nevertheless, certain artefacts might be associated with or point to specific identi­ties and their combination, with other grave goods could provide information on the personality (sta­tus) of the individual. A comparison of cremation and inhumation burials from the biritual cemetery using the CA in Kralice was not successful, since – except for one inhuma­tion burial – none of them met the criteria of the CA. In the deepest grave 40/05, a 25–40-year-old woman was buried, crouched on her left side, with pottery and a grinding stone. As it was the only in-humation burial suitable for CA, its information va­lue was very low. No significant similarities were re­corded in the basic comparison with cremation bu­rials which also contained grinding stones, as their grave goods were richer and more varied. Example 2. Mod³niczka 2 (Little Poland) – cre­mation cemetery of the LPC Site 2 in Mod³niczka (238 m a.s.l.) is located in the valley of the Wedonka River (a small tributary of the Rudawa) in western Little Poland (Czekaj-Zastawny et al. 2011.53). The cremation cemetery (39 buri­als) was in a wet, seasonally watered part of the val­ley (Czekaj-Zastawny, Przyby³a 2012.Fig. 16; Cze­kaj-Zastawny 2008). The burials were grouped on about 70 acres in the eastern part of the site (Fig. 4), earlier occupied by a settlement of the Zofipole phase. The graves did not form any bigger concen­tration, but were rather clustered in several places (Czekaj-Zastawny, Przyby³a 2012.Fig. 2, Plan 1). A few graves were dug into the ditch linked with the early settlement. In 26 cases outlines of cremation pits were visible, and the others were marked by concentrations of artefacts and burnt bones. Chronological context of the cemetery Due to the relatively small amount of pottery in gra­ves, the chronological position of the burial ground within the LPC is not very evident. Almost all pot­tery fragments found in graves were decorated with Music Note motives. Only grave 1109 contained frag­ments with engraved lines and single .eliezovce not­ches. Considering this, graves equipped with stone artefacts only or not equipped at all may also come Fig. 3. Correspondence analysis plot of Kralice na Hané cremation burials, on 1. and 3. principal axes (child – green; adult – black). Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe Fig. 4. Mod³niczka, Site 2, Little Poland. Cremation cemetery. Location of LPC graves (after Czekaj-Za­stawny, Przyby³a 2012.Fig. 2). Colour-coded graves suitable for correspondence analysis (child – green; adult – black; male, adult – red). from the Music Note phase (Czekaj-Zastawny, Przy­by³a 2012.42–44). The cremation cemetery in the eastern part of Site 2 in Modlniczka was possibly linked with the settle­ment dated to the Music Note phase and the turn of Music Note/.eliezovce phase located a few hundred metres apart (Modlnica, Site 5). It was situated on a loess elevation on a slope of the low river terrace, and two contemporary inhumation burials were dis­covered there, too (Czerniak 2010). There are no vi­sible relations between the cremation cemetery with the nearer settlement Modlniczka 2 from the Zofi-pole phase (Czekaj-Zastawny, Przyby³a 2012.31). Graves3 All cremation burials in Modlniczka are pit graves. The original shapes of the grave pits are very poorly preserved (Czekaj-Zastawny, Przyby³a 2012.Fig. 17–18), although various outlines have been found: oval with dimensions 100–185x60–120cm, circular with diameters 40–90cm, or rectangular with dimen­sions 60–130x45–110cm. The preserved parts were up to 30cm deep. The cross-section of pits was in all cases basin-like. They were filled with grey, grey-brown, or black-brown sand with burnt human bo­nes and remains of the funerary equipment. In bet-ter-preserved pits two fill layers were visible – slight­ly lighter sand with scattered bone fragments and very small charcoal pieces. Pieces suitable for palaeo-botanical analyses were exceptional, including one oak (Quercus sp.) fragment from grave 1354. Origi­nal grave pits had small depressions into which the remains of the pyre had been placed, possibly in or­ganic containers (small sacks?). Oval and quasi rectan­gular pits were in three cases orientated N-S, in other three cases E-W, in two cases NW-SE, and in one case NE-SW (Czekaj-Zastawny, Przyby³a 2012.56–62). Among 39 graves (Szczepanek 2012.305–313), the sex of buried individuals has been determined in six cases (one woman – grave 1205; five men – graves 1087; 1213; 1423; 4688; 7496) and age in seven ca­ses (woman Maturus, two men Adultus/Maturus, man Maturus, man Maturus/Senilis, one Adultus of undetermined sex, and one Infans). The remains from 14 graves have been determined as belonging to adults. For 17 graves no information on the bu­ried persons is available. No links between age, gender and the size of the grave pit have been established. However, only one child was identified at the burial ground (grave 1249), and thus this claim is not generally valid. The 3 In the publication (Czekaj-Zastawny, Przyby³a 2012), we noticed slight differences between the catalogue and the analytical parts. Thus, we follow exclusively the catalogue part when processing graves, their goods, dating and anthropological identification (Cze­kaj-Zastawny, Przyby³a 2012.96–129). Noémi Beljak Pa/inová, Alena Bistáková adult woman (grave 1205) and the three grown men (graves 1087, 1213, 4688) were buried in smaller pits than the child. The largest grave pits were for adult individuals (graves 1061, 7427/7428) or adult men (grave 1423 and 7496). No superpositions were documented at the burial ground. All burials were single graves, hence no grave included more than one individual (Szczepa­nek 2012.305). Grave goods Sixteen burials contained grave goods (Czekaj-Za­stawny, Przyby³a 2012.96–129). The other 23 gra­ves were probably also equipped, but the goods were destroyed and poorly preserved. Generally, grave assemblages included adzes (13 graves), pot­tery (nine graves), and flint artefacts (nine graves). Altogether, 78 pottery fragments were recovered from grave pits (Czekaj-Zastawny, Przyby³a 2012. 38). Nine graves were equipped with pottery (23%). In four cases (graves 1145, 1354, 7422, 7496) pot­sherds came from one vessel, in other cases from up to 11 vessels (grave 1109). In grave 7427/7428 were altogether 35 potsherds (of which 14 frag­ments were from one vessel, 15 fragments from ano­ther vessel, four fragments from a third). All frag­ments found in the graves were either thin or me-dium-thick. For the most part, they were decorated. Whole vessels have not been documented. No traces of fire have been observed on potsherds. Fourteen burials were equipped with stone tools. There were 17 adzes, four adze fragments, and 3 hoes (Czekaj-Zastawny, Przyby³a 2012.41, 42). Most of these were made of amphibolite, probably from the Sudety Mts. Only two tools from graves 1218 and 1354 are of feldspar (Tr¹bska 2012). Almost all adzes are burnt, usually on two-thirds of the length (from the blade). The grave inventory usually con­tained one or two adzes (e.g., grave 1423, determin­ed as a man), or one adze and a pot (e.g., grave 7496, determined as a man). The male grave of Adul­tus/Maturus (feature 1087) contained three items: an adze, a hoe, and a trapeze – functionally an ar­rowhead. The richest and most diverse equipment was found in grave 1061 (adult of undetermined sex): three whole adzes and one shoe-last tool frag­ment, one hoe, three flint artefacts, one obsidian ar­tefact and 9 potsherds (Czekaj-Zastawny, Przyby³a 2012.96–98, Pls. 29–31). Nine graves were equipped with flint or obsidian (one case) artefacts (Trela-Kieferling, Zaj¹c in press), ranging from one to eight of these (in grave 1061). Twelve artefacts were of the Cracow Jurassic flint, one of obsidian, and one of undetermined burnt rock. The obsidian artefact (from feature 1061) is a small single-platformed core for blades. The assem­blage of tools includes a combined tool (burin + per-forator) and a retouched blade from feature 1109, a trapeze (arrowhead) from feature 1087, a double truncated blade with sickle gloss from feature 1218, and a burin on truncation from feature 1249. The last mentioned example is a child grave, and the bu-rin was in the grave together with an adze. In grave 4688, identified as the burial of male aged Maturus/Senilis, apart from a hoe, four flint arte-facts, and four potsherds, there was a unique perfo­rated stone fragment of an oblong polished object made of white aplite, oval in cross-section (Czekaj-Zastawny, Przyby³a 2012.105, 124, Pl. 42.2). It has a small opening (diameter approx. 5mm) in half of its length, bored from both sides. Correspondence analysis Eleven graves with grave goods met the CA criteria (Fig. 4). Since this is a small ensemble with unique graves, several divergences could be tracked (Fig. 5). In any combination of main axes, the grave 4688, which has several distinctive attributes, was set apart from the others. It is the burial of the only male aged Maturus/Senilis, and apart from other grave goods he had a unique perforated stone arte-fact in the whole cemetery – a pendant(?). A sepa­rate position is also noted for the male Adultus/Ma­turus grave 1087, with a trapeze arrowhead com­bined with an adze and a hoe. Nearby is the richest grave 1061 of another adult individual, which, in ad­dition to a combination of adzes, hoes, obsidian core and flint tools also contains potsherds. Smaller clusters form graves without potsherds con­taining a combination of adzes and flint tools such as a burin and blades (child grave 1249; adult grave 1218; adult grave 1040). Another cluster is made of possible adult male graves 1109, 1354, 4619. These have in common a combi­nation of potsherds, adzes and flint tools and a lar­ger grave pit. Discussion The LPC cremation cemeteries presented in this study were certainly used for regular burials of peo­ Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe Fig. 5. Correspondence analysis plot of Modlllniczka, Site 2 cremation burials, on 1. and 3. principal axes (child – green; adult – black; male, adult – red). ple from contemporary settlements. The cemetery in Kralice na Hané was in the immediate vicinity of a settlement. The neighbourhood whose inhabitants probably buried their dead in the cemetery in Mod³­niczka, Site 2, was only a few hundred metre from their last resting place. The cemeteries were used at least from phase LPC II to LPC II/III. During this pe­riod, biritual burials were used. In Kralice na Hané, human skeletons were recorded not only at the biri­tual burial ground, but also in pits right at the near­by settlement (.míd 2012.110). In Mod³niczka, Site 2, only cremation burials were uncovered. Neverthe­less, two contemporary inhumation burials were found at the nearby settlement of Mod³niczka, Site 5 (Czerniak 2010). This funerary tradition documents the considerably diverse spiritual world of the LPC communities. Biritual burying in Kralice na Hané was probably not used until the end of the cemetery’s existence, and it seems that cremation prevailed here in the later period as well. In both cases, however, we must bear in mind that only a small part of the original sacred area has been studied. Despite this, the burial grounds maintained their status as impor­tant places in the landscape for a long time, and they were used repeatedly. The progression from high quantities of inhumations in the east to their decline in the west in favour of cremations was also seen as chronological development at Aiterhofen-Ödmühle (Lower Bavaria). Cremations are here likewise be­lieved to be a younger phenomenon (Nieszery 1995. 89; Hahnekamp 2021.982, 983). The fact that the burials found at both the above analysed sites were mostly undisturbed or not de­stroyed in the LPC period is important. Therefore, we can assume a certain form of indication of graves or long remaining memory of the buried individuals (see Chapman 2000.46). Naturally, we cannot be sure about the relations between the graves or be­tween the burying and the buried without an exact chronology of the place (cemetery and the related settlement). Previous sequences of ceramic decora­tion are insufficient for revealing chronological de­velopments, and thus more radiocarbon and genetic studies are suggested. Superpositions were recorded only in two cases in Kralice na Hané. The reuse of old funerary spots maintains community ties and may represent a link between generations. The use of the same location for cremation remains (grave 30/05) and inhuma­tion (grave 33/05) in Kralice na Hané (.míd 2008. 251) and traces of fire detected on the vessel depo­sited next to the skeleton might indicate the tradi­tional use of fire at the funerary rite. More informa­tion in relation to this hypothesis could be brought by detailed DNA analyses of both dead people, which has so far not been possible. With regard to burial practices, there are no distinct differences between burials of adult women or men and children. The presented burial grounds reflect the demography of the population. The right and Noémi Beljak Pa/inová, Alena Bistáková possibility of being buried at the burial ground were not denied to neither sex nor any age category. Ne­vertheless, there are also some exceptions. In some cases, whole groups of graves are excluded from the burial ground, and sometimes there are no female, male, or child graves within a given site (Chapman 2000; Wunn 2001). Such localization of graves can be observed in the Vedrovice – .iroká u lesa LPC ce­metery (Podborský 2002). Selected clusters can also be divided by sex, with one group of graves being populated exclusively by females and the other one by men, children, and anthropologically undefined individuals (Podborský 2002.301, Fig. 2). Grave goods in cremations do not show any signifi­cant differences at both analysed sites. They were placed in graves of adult men, women, and children. Grave goods specific for sex (e.g., occurrence of ad­zes and hoes with men) are an exception. Numerous experts have pointed out that LPC adzes appear to have conveyed social, or even status, differences also in relation to the LPC inhumation burials (Jeu­nesse 1997; Nieszery 1995; John 2005; Mandák Niklová, Mandák 2020). For example, in Vedrovice wealthy burials including prestige items such as adzes were only associated with men (the only ex­ception was a child grave 39) (Kvìtina 2004.385). Regarding social hierarchy exceptionally wealthy male burials, comprising older individuals (matures, senilis) were therefore attributed to chieftains, big men or elders (Kvìtina 2004.387). The increased frequency of polished stones in cremations at Aiter­hofen-Odmühle also indicate a predominating prese­nce of men, with half of the provided ones confined to adzes as the only gift (Hahnekamp 2021.979, 982). The difference in grave goods equipment be­tween inhumations and cremations is thus seen here as potentially signifying gender distribution inequa­lities. Moreover, adult men buried with polished adzes in certain sites (Aiterhofen-Ödmühle, Vedrovi­ce, and Nitra) have a higher protein intake of animal origin (15N) so it seems that this social group has certain privileges such as a richer nutrition and a stable residence (Augerau 2021.959). In burial contexts, no symbolic elements associated with the female gender are found. It seems that wo­men have no gendered tools, and their adornments (if they have any) are not specific. However, there might be a preservation issue, therefore we must remember that some grave goods are possible not so visible in the archaeological record. Members of the Linear Pottery population who cremated their dead may have considered the act of burning itself to be significant, reducing the need of specific as­semblages to identify the deceased’s social standing or sex and instead focused on fewer but meaning­ful burial goods. Alternatively, a variety of organic material might have been burnt at the funeral, re­sulting in their absence in the grave-pit. Stone tools and remains of former – often unidentifiable – tools (fragments of chipped industry) represent an impor­tant tradition in grave goods and point to the every­day work of the inhabitants. Three bone tools (awls) and grinding stones can be considered representati­ves of human activities in Kralice na Hané, although two of the grinding stones found in cremations were without traces of any activity whatsoever. Pottery was represented by the highest number of artefacts (as for fragments) which occurred in crema­tion burials. It was preserved mainly in fragments or as damaged vessels. The shape in which they were deposited in the graves, however, can reflect a diffe­rent form of burial traditions. From the aspect of the rite of passage, for instance, intentional damaging of things (e.g., breaking a vessel) might symbolize the dying of things so that they can further be used by the dead person (Lutovský 1998.256). Unlike Mod³niczka, Site 2, three graves in Kralice na Hané contained complete vessels. It could suggest a different function of pots and fragments deposited in graves. All this pottery is probably for the serving and to a lesser extent storing food and drink. Whe­ther they were part of a funerary ritual involving the mourners eating and drinking at the grave, or whether they were placed in the grave for the de­ceased’s use in the afterlife is open to debate. In this context, the fact that the presence of animal bones in cremation graves was minimum is very impor­tant. Unique burned bones (rabbit and sheep/goat) that might indicate adding meat to graves were iden­tified only in two cremations in Kralice na Hané. We must not overlook the significant fact that there is generally no indication of urn burials in the LPC (as well as in the studied sites). Vessels are placed deposits (grave goods) to the burial or used to cover cremated remains. In this context, it would be appro­priate to regard the vessel as an upside-down urn? A funeral pyre has not been recorded at any of the burial grounds. They were possibly at some distance from the burial ground. Transporting pyre remains into graves probably followed specific rules. In Mo-d³niczka, Site 2, was observed that human bones, often with flint artefacts, were placed first, then – on Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe the same level or slightly above – adzes and pottery. Many stone and flint artefacts are burnt, but no tra­ces of fire have been documented on potsherds. It suggests that broken vessels might have been plac­ed directly into the pit (Czekaj-Zastawny, Przyby³a 2012.38–42). Certain differences and variations occurring among graves are a result of other factors – hidden in ar­chaeological records – which could point to the sta­tus of an individual and/or a family (?), to wealth, and they could have been inherited, as children’s graves rank among the richest ones from the aspect of typological diversity of grave goods. They include e.g., grave 15/05 and 1/06 HOPI in Kralice na Hané and grave 1249 in Mod³niczka, Site 2. A certain change in the number of grave goods can be noticed when comparing older and newer graves. The youngest grave 9/06 (LPC phase IIc/III) belong­ed to the richest graves in the cemetery in Kralice na Hané. In the grave, three vessels, chipped lithic industry and a sandstone grinding stone were found. The youngest grave 1109 (.eliezovce phase) in Mo-d³niczka, Site 2 also ranks among the richest ones discovered at the site with its inventory. It was noticed in the past that cremation played va­rious roles in various territories. In some regions, it was almost as common as inhumations, regardless of a chronological phase, in others much less fre­quent (Czekaj-Zastawny, Przyby³a 2012.49, Fig. 31). We can also add – based on the above-presented facts – that it depends mainly on the state of re­search, not the regional tradition. Conclusion When considered alongside the small sample size, the potential of CA did not seem promising. Never­theless, the use of CA in this contribution was in­tended to be exploratory and, in this case, it served its purpose. As we see the CA results brought not hidden, previously unrecognized data, but rather highlighted certain facts. We intended to find out what cremation says about the buried individual and if the burial method reflects the deceased per­son’s social identity. In the first place, we can confirm that any attempt to extrapolate a relationship between age and the use of cremation is frustrated by the low rate of suc­cess in identifying the cremations. Secondly, there seems to have been no association between sex or age and the type of burial used. Naturally, as the bo­dies were burnt, age and sex determinations are missing for the most part. There are also several as­pects of burial practice that seem to have been ‘cross-regional’. For example, the grave goods in crema­tion did not differ significantly from grave goods in inhumation burials, including exceptionally furni­shed burials of wealthy adult men or the presence of ochre in burials of adult individuals. These are LPC practices that occur across Europe and may point to shared understandings or values. It is es­pecially relevant in addressing the mainstream ideo­logy in the LPC society (Augereau 2021). However, this pattern does not characterise the whole LPC burial practice. Indeed, in the case of placing commo­dities at the funeral pyre together with the dead, unification does not apply, thus the ideology of cre­mation in the LPC culture was probably different or expressed in a different way, not perceptible in the funerary data. Thirdly, a key issue in understanding the LPC sites is time. Unfortunately, all the sites with cremations in Central Europe are lacking a refined chronology, and the burials at the cemeteries have not been ra­diocarbon dated. Chronologies of the cemeteries dis­cussed above were only broad phases based on ty­pologies. And as the analysis of Franti.ek Trampota and Petr Kvìtina (2020) showed changes in the ori­ginal material culture do not necessarily occur on a time axis. Hence more refined chronologies, such as whether a burial group/row was added to in a parti­cular direction, or whether one area was finished before another came into use, do not yet exist. It is therefore impossible to follow the burial practice, in our case cremation, more closely in time. However, summarizing the current facts on cremation burials in Central Europe, their beginnings and gradual es­tablishment in the Neolithic, we can state that their origin no longer needs to be sought in distant and more developed regions. In the LPC, cremation fu­nerary rite becomes ‘domesticated’. The rite itself cannot be associated with ethnicity and probably not with the social status of individuals either. How­ever, cremation can be considered a common burial custom spreading simultaneously with inhumation since the Neolithic. This fact is documented also by recent excavations at biritual cemeteries in Central Europe. They gradually open the way to a better un­derstanding of this funeral tradition and objective evaluation of the situation in a wider territory. At the beginning of the millennium, we learned about cremation in LPC based on data from west­ Noémi Beljak Pa/inová, Alena Bistáková ern Europe. New excavations and published data in the studied territory point to the fact that cremation was represented at cemeteries to a much greater ex­tent as assumed. Therefore, we can state that crema­tion in the LPC was not limited territorially, chrono­logically, or demographically and it was used paral­lelly with inhumation also in Central Europe. Al­though cremations are still far less numerous than inhumations, they occurred in significant numbers depending on region and site, and thus must have played an important role in Linear Pottery funerary rites and might have reflected changes in notions of the afterlife. For example, at Vedrovice, Kleinha­dersdorf and Nitra, cremations are scarce to absent, while they predominated in sites such as Kralice na Hané and Mod³niczka, Site 2. All of these features point to dynamic and adaptable funeral practices and norms rather than fixed systems. The mode of body disposal could be affected by social organiza­tion, hierarchies, family, economic factors, religious beliefs, and by the individual preferences of the de­ceased or their survivors, and other factors not ap­parent in today’s ideology. The act of cremation can be associated with the pu­rifying power of fire. The ritual aspect of the crema­tion process was a strong spiritual experience for the people attending a funeral (Schlentner 1960; Ucko 1969). The practical aspect of cremation is pre­sented by the effort of the survivors at the creation of a funeral pyre, as well as reduced requirements for the deposition space. All these facts might have played equally important roles in deciding on the funerary rite. We also agree with the idea that the funerary treatment represents a chronological act expressing idealized social identities (more details in Fowler 2013), i.e. the archaeological remains are not an a priori reflection of a current social organi­zation – they present a deeper view of the world of the contemporary community (family of the deceas­ed person). Since only a fraction of the living population was being buried in cemeteries, there will always be doubts about what could be considered a mortuary norm and exception (cf. Boyad.iev 2009; Perles 2001.274). Nevertheless, the known data provide in­formation on the religious, ritual, and social tradi­tion, but these aspects can constantly change and vary in everyday practice. The presented examples confirm that the burial customs of individual com­munities primarily reflect domestic traditions and rules. A good example in cremation can be seen in the various compositions and ways of depositing grave goods – either on a pyre or in a grave inter­mingled with the burnt remains on variable sites. As for the future of excavations of cremations and cremation cemeteries of the LPC, there are several determining factors, including localization of ceme­teries in inundation areas or floodplains where set­tlement and associated activities are not expected or have not been searched by survey activities so far. The second determinant is the high proportion of destroyed graves which is associated with their smal­ler and shallower grave pits being disturbed or completely damaged by (not only) recent activities. Bearing this fact in mind, we must pay more atten­tion to targeted searches and exact documentation of find contexts, which are unique testimonies of the LPC funerary rite. ACKNOWLEDGEMENTS This research was funded by the Ministry of Educa­tion, Science, Research and Sport of the Slovak Re­public, grant VEGA No. 1/0261/20 (Archaeology of death. Burial practices in the Late Prehistory to Pro-tohistory in the territory of Slovakia) and VEGA No. 2/0075/21 (Settlement agglomeration of Linear cul­ture in the .itava valley). Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe References Augereau A. 2021. Funerary Practices as a Testimony of Ideology in Western Linearbandkeramik Culture. In F. Borrell, I. Clemente, M. Cubas, +5 editors, and X. Terra-das (eds.), The Early Neolithic of Europe. Open Archaeo­logy (Special Issue) 7: 953–971. https://doi.org/10.1515/opar-2020-0184 Bánffy E., Eichmann W. J., and Marton T. 2007. Mesolithic foragers and the spread of agriculture in western Hungary. In J. K. Koz³owski, M. Nowak (eds.), Proc. XV UISPP World Congress vol. 6, Lisbon, 4–9 September 2006. BAR International Series 1726. Archeopress. Oxford: 53–62. Bátora J. 1999. Nález kostier v základovom .l’abe domu .eliezovskej skupiny v Jel.ovciach. In I. Kuzma (ed.), Otá­zky neolitu a eneolitu na.ich krajín – 1998. Zborník re-ferátov zo 17. pracovného stretnutia bádatel’ov pre výs­kum neolitu a eneolitu .iech, Moravy a Slovenska (Dudi­nce 22.–24.9.1998). Archeologický ústav Slovenskej aka-démie vied. Nitra: 11–14. Bickle P., Whittle A. 2013. The first farmers of Central Europe: diversity in LBK lifeways. Oxbow. Oxford. Bori. D., French C. A. I., Stefanovi. S., +5 authors, and Fi­lipovi. D. 2014. Late Mesolithic lifeways and deathways at Vlasac (Serbia). Journal of Field Archaeology 39(1): 4–31. https://doi.org/10.1179/0093469013Z.00000000070 Bori. D., Rai.evi. J. and Stefanovi. S. 2009. Mesolithic cre­mations as elements of secondary mortuary rites at Vla-sac (Serbia). Documenta Praehistorica 36: 247–282. https://doi.org/10.4312/dp.36.16 Boyadzhiev K. 2009. About the Function of Stone Ham-mer-axes from the Cholcolithic Period in Bulgaria. In I. Gacov (ed.), Saxa loquuntur. Sbornik v .est’ na 65-go­di.ninata na Nikolaj Sirakov. Sofia: 237–243. Braun P. 2001. K osídlení kultury s lineární keramikou u Litic, okr. Plzeò-mìsto. In M. Metli.ka (ed.), Otázky neoli­tu a eneolitu na.ích zemí 2000. Západo.eské múzeum v Plzni. Plzeò: 102–108. Brìzinová G., Pa.inová, N. 2011. Neolitická osada Hurba­novo-Bohatá. Archeologický ústav Slovenskej akadémie vied. Nitra. Chapman J. 2000. Tensions at funerals. Micro-Tradition Analysis in Later Hungarian Prehistory. Archaeolingua Alapítvány. Budapest. Cheben I. 2000. Baj. – eine Siedlung der .eliezovce-Gruppe. Entwicklungsende der .eliezovce-Gruppe und Anfänge der Lengyel-Kultur. Universitätsforschungen zur prähistorischen Archäologie. Band 68. Verlag Dr. Rudolf Habelt GMBH. Bonn. Czekaj-Zastawny A. 2008. Osadnictwo spo³ecznoœci kul­tury ceramiki wstêgowej rytej w dorzeczu górnej Wis³y. Instytut Archeologii i Etnologii Polskiej Akademii Nauk. Kraków. Czekaj-Zastawny A., Przyby³a M. 2012. Modlniczka 2, po­wiat krakowski, cmentarzysko kultury ceramiki wstêgo­wej rytej i osady neolityczne z analizami specjalistycz­nymi Damiana Stefañskiego; Anity Szczepanek i Joanny Tr¹bskiej (Via Archaeologica. ród³a z badañ wykopali­skowych na trasie autostrady A4 w Ma³opolsce). Instytut Archeologii i Etnologii Polskiej Akademii Nauk. Kraków. Czekaj-Zastawny A., Przyby³a M., and Trela-Kieferling E. 2011. Cia³opalne cmentarzysko kultury ceramiki wstêgo­wej rytej na stan. 2 w Modlniczce, pow. Kraków. In M. Po-pelka, R. .midtová (eds.), Otázky neolitu a eneolitu na-.ich zemí. Sborník referátù z 28. zasedání badatelù pro výzkum neolitu a eneolitu (nejen) .ech, Moravy a Slo­venska. Mìlník 28.9.–1.10.2009. Vìnováno prof. PhDr. Jiøímu Slámovi. Praehistorica 29. Univerzita Karlova. Praha: 53–67. Czene A. 2008. Újkõkori falu Budakeszin (Neolithic set­tlement at Budakesziò. In M. Gyöngyössy, K. Ottományi (eds.), Képek a múltból. Az elmúlt évek ásatásaiból. Pest megyében. Megyei Múzeumok Igazgatósága. Szentendre: 14–15. Czene A., Ottomanyi K. 2007. Budakeszi, Szõlõskert– Tan-gazdaság (MR T 7. k. 4/8 lh.). In J. Kisfaludi (ed.), Régé­szeti kutatások Magyarországon 2006 – Achaeological Investigations in Hungary 2006. Magyar Nemzeti Múze-um. Budapest: 162. Czerniak L. 2010. Osady kultury ceramiki wstêgowej ry­tej. Modlnica, gm. Wielka Wieœ, woj. ma³opolskie, stano­wisko 5 (AUT 4). Krakowski Zespó³ do Badañ Autostrad. Kraków. Czerniak L., Kabaciñski J. 2004. Archeologiczne badania ratownicze Instytutu Archeologii i Etnologii PAN w Poz­naniu na trasie autostrad A-1 i A-2 w Wielkopolsce i na Kujawach w roku 2001 i 2001. In Z. Bukowski (ed.), Ra-port 2001–2002 (Zeszyty Oœrodka Ochrony Dziedzictwa Archeologicznego, seria B. Materia³y Archeologiczne). DiG. Warszawa: 141–155. Cziesla E., Ibeling T. (eds.). 2014. Autobahn 4. Fundplatz der Extraklasse. Archäologie unter der neuen Bundes-autobahn bei Arnoldsweiler. Archäologische fachlitera­tur. Beier & Beran. Langenweissenbach. Noémi Beljak Pa/inová, Alena Bistáková .i.máø Z., Geislerová K. 1997. Pohøby v jamách na sídli-.ti kultury s lineární keramikou v .ádovicích, okr. Hodo­nín. Pravìk 7: 39–64. Debiec M., Becker V., Dêbiec M., +5 authors, and Szczepa­nek A. 2014. Zwiêczyca 3. Eine bandkeramische Sied-lung am Wis³ok. Oficyna Wydawnicza ZIMOWIT. Rzeszów. Do.kalová M., .i.máø Z. 2007. Neolithic Children Burials at Moravian Settlements in the Czech Republic. Anthropo­logie. International Journal of Human Diversity and Evolution 45: 31–59. http://puvodni.mzm.cz/Anthropologie/article.php?ID=42 2008. Neolithic Settlement Burials of Adult and Juve­nile Individuals in Moravia, Czech Republic. Anthropo­logie. International Journal of Human Diversity and Evolution 46: 37–76. http://puvodni.mzm.cz/Anthropo logie/article.php?ID =43 Fahlander F. 2012. Mesolithic Childhoods: Changing Life-Courses of Young Hunter-Fishers in the Stone Age of Southern Scandinavia. Childhood in the Past 5: 20–34. https://doi.org/10.1179/cip.2012.5.1.20 Farkas L. Gy. 2005. Szakvélemény a Hódmezovásárhely-Gorzsa Kovács tanya lelohelyen Gazdapusztai Gyula ré­gész által 1955-ben feltárt hamvasztott csontokról. Zalai Múzeum 14: 13. Fowler C. 2013. Identities in transformation: identities, funerary rites and the mortuary process. In S. Tarlow, L. Nilsson Stutz (eds.), The Oxford handbook of the archa­eology of death and burial. Oxford University Press. Ox­ford: 511–525. Gallus S. 1936. A nagytétényi neolitikus sír. Archeológiai Értesítõ 49: 85–88. Gazdapusztai G. 1957. A Körös kultúra lakótelepe Hódme-zõvásárhely-Gorzán. Archeológiai Értesítõ 84: 3–13. Gerling C. 2012. Das linearbandkeramische Gräberfeld von Schwetzingen, Rhein-Neckar-Kreis. Fundberichte aus Baden-Württemberg 32: 7–263. Gil-Drozd A. 2011. The Origins of Cremation in Europe. Analecta Archaeologica Ressoviensia 5: 9–94. Gligor M., Ba¢cuet-Crisan S. 2014. Inhumation versus cre­mation in Transylvanian Neolithic and Eneolithic. Studia Antiqua et Archaeologica 20: 37–67. Gray Jones A. 2017. Cremation and the Use of Fire in Me­solithic Mortuary Practices in North-West Europe. In J. I. Cerezo-Román, A. Wessman, and H. Williams (eds.), Cre­mation and the Archaeology of Death. University Press. Oxford: 27–51. Greenacre M. 2007. Correspondence analysis in prac­tice. Chapman and Hall/CRC. London. Grünberg J. M. 2000. Mesolithische Bestattungen in Eu­ropa. Ein Beitrag zur vergleichenden Gräberkunde. In­ternationale Archäologie 40. VML Verlag Marie Leidorf. Rahden. 2008. Aufrecht ins Jenseits: Die sitzende Haltung von Verstorbenen im Mesolithikum. Die Kunde N. F. 59: 39–90. Grygiel R. 2004. Wczesny neolit kultura ceramiki wstê­gowej rytej. Neolit i pocz¹tki epoki br¹zu w rejonie Br­zeœcia Kujawskiego i Os³onek I. Fundacja Badañ Archeo­logicznych Imienia Profesora Konrada Ja¿d¿ewskiego. £odŸ. Haak W., Balanovsky O., Sanchez J. J., +13 authors, and Cooper A. 2010. Ancient DNA from European early Neoli­thic farmers reveals their Near Eastern affinities. PLoS Biology 8: e1000536. https://doi.org/10.1371/journal.pbio.1000536 Hahnekamp Y. 2021. A Quantitative Study of the Linear Pottery Culture Cemetery “Aiterhofen-Ödmühle”. Open Archaeology 7: 972–985. https://doi.org/10.1515/opar-2020-0161 Hammer O., Harper D. A. T., and Paul D. R. 2001. Past: Pa-leontological Statistics Software Package for Education and Data Analysis. Palaeontologia Electronica 4(1): 1–9. http://palaeo-electronica.org/2001_1/past/issue1_01. htm Harvig L. 2015. Fast cremation practices from a bioarcheo-logical perspective: How new methods and techniques re­vealed conceptual changes in cremation practices during the late Bronze Age and early Iron Age in Denmark. The archaeology of cremation: 43–62. Hofmann D. 2009. Cemetery and settlement burial in the Lower Bavarian LBK. In D. Hofmann, P. Bickle (eds.), Cre­ating communities. New advances in central European Neolithic research. Oxbow. Oxford: 220-234. Hoffmann E. 1973. Zur Problematik der bandkeramis­chen Brandbestattungen in Mitteleuropa. Jahresschrift für mitteldeutsche Vorgeschichte 57: 71–103. Hofmann D., Orschiedt J. 2015. Mortuary practices, bod­ies and persons in central Europe. In C. Fowler, J. Harding, and D. Hofmann (eds.), The Oxford Handbook of Neoli­thic Europe. Oxford University Press. Oxford: 987–1004. Jeunesse C. 1997. Pratiques Funéraires au Néolithique Ancien: Sépultures et Nécropoles Danubiennes 5500 – 4900 av. J. C. Éditions Errance. Paris. Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe John J. 2005. Pøíspìvek ke struktuøe pohøební výbavy v kultuøe s volutovou keramikou. In E. Neustupný, J. John (eds.), Pøíspìvky k archeologii 2. Ale. .enìk. Plzeò: 11–19. Kalicz N., Koós J. 1997. Mezõkövesd-Mocsolyás. Újkõkori telep és temetkezések a Kr. e. VI. évezredbõl. In P. Raczky, T. Kovács, and A. Anders (eds.), Utak a múltba. Az M3-as autópálya régészeti leletmentései. Paths into the Past. Rescue excavations on the M3 motorway. Magyar Nem­zeti Múzeum. Eötvös Loránd Tudományegyetem Régészet­tudományi Intézet. Budapest: 28–33. 2014. Mezõkövesd-Mocsolyás. A neolitikus Szatmárc­soport (AVK I) települése és temetõje a Kr. e. 6. évez-red második felébõl. Mezõkövesd-Mocsolyás. Borsod­Abaúj-Zemplén megye régészeti emlékei 9. Herman Ot­tó múzeum. Miskolc. Kalicz N., Makkay J. 1977. Die Linienbandkeramik in der Grossen Ungarischen Tiefebene. Akadémiai Kiadó. Buda­pest. Kempisty E. 1962. Pierwszy grób kultury wstêgowej cera­miki rytej na LubelszczyŸnie w Gródku Nadbu¿nym, pow. Hrubieszów. Wiadomoœci Archeologiczne 28: 284–285. Kreiter A., Marton T., Gomart L., Oross K., and Pánczél P. 2017. Looking into houses: Analysis of LBK ceramic tech­nological change on a household level. In L. Burnez-La­notte (ed.), Matieres a Penser: Raw materials Acquisi­tion and Processing in Early Neolithic Pottery Produc­tions. Proceedings of the Workshop of Namur (Bel­gium) 29 and 30 May 2015. Société Préhistorique Fran-çaise. Paris: 111–132. Kulczycka-Leciejewiczowa A. 2008. Samborzec. Studium przemian kultury ceramiki wstêgowej rytej. Instytut Ar-cheologii i Etnologii Polskiej Akademii Nauk. Wroc³aw. Kulczycka-Leciejewiczowa A., Romanow J. 1985. Wczes­noneolityczne osiedla w Gniechowicach i Starym Zamku. Silesia Antiqa 27: 9–68. Kustár R., Lantos A., Hajnalová M., and Sümegi P. 2014. Výskum neolitického sídliska na lokalite Harta-Gátõrház. Studia Archaeologica Brunensia 19(1): 23–43. Kvìtina P. 2004. Mocní mu.i a sociální identita jednotliv­cù – prostorová analýza pohøebi.tì LnK ve Vedrovicích. Archeologické rozhledy 56(2): 383–392. Larsson A. M., Nilsson Stutz L. 2014. Reconcilable differ­ences: cremation, fragmentation and inhumation. Mesoli­thic and Neolithic Sweden. In I. Kuijit, C. Quinn, and G. Cooney (eds.), Transformation by fire. The Archaeology of Cremation in Cultural Context. University of Arizona Press. Tucson: 47–66. Lazarovici G. C., Lazarovici M. C. 2006. A Home altar at Gura Baciului. Analele Banatului, S. N., Arheologie – Is-torie 14: 103–111. Lenneis E. 2007. Mesolithic heritage in early Neolithic bu­rial rituals and personal adornments. Documenta Praehi­storica 34: 129–137. https://doi.org/10.4312/dp.34.10 Lichter C. 2003. Continuity and Change in Burial Cus­toms: Examples. In L. Nikolova (ed.), Early Symbolic Sy­stems for Communication in Southeast Europe. Archaeo-press. Oxford: 135–152. Lipson M., Szécsényi-Nagy A., Mallick S., +53 authors, and Reich D. 2017. Parallel palaeogenomic transects reveal complex genetic history of early European farmers. Na­ture 551: 368–372. https://doi.org/10.1038/nature24476 Little A., van Gijn A., Collins T., +4 authors, and Warren G. 2017. Stone dead: Uncovering Early Mesolithic mortu­ary rites, Hermitage, Ireland. Cambridge Archaeological Journal 27(2): 223–243. Lutovský M. 1998. Poslední cesta: K úloze komunikace v rituálech spojených se smrtí. Archaeologica historica 23: 253–258. Makkay J., Starnini E., and Tulok M. 1996. Excavations at Bicske Galagonyás III: The Notenkopf and Sopot-Bi­cske cultural phases. Edizioni Svevo. Trieste. Mandák Niklová M., Mandák P. 2020. A gender archaeo­logy approach to selected features of Neolithic burials. In A. Bistáková, G. Bøezinová, and P. C. Ramsl (eds.), Multi­ple identities in prehistory, early history and presence. Archeologický ústav Slovenskej akadémie vied. Nitra: 49– 63. Manning K., Timpson A., Colledge S., Crema E., Edinbo-rough K., Kerig T., and Shennan S. 2014. The Chrono­logy of Culture: A Comparative Assessment of European Neolithic Dating Approaches. Antiquity 88: 1065–1080. https://doi.org/10.1017/S0003598X00115327 Marciniak M. 1993. Mesolithic burial and dwelling struc­ture from the boreal period excavated at Mszano site 14, Toruñ district, Poland. Preliminary report. Mesolithic Mi­scellany 14(1–2): 95–123. Marton T. 2008. Development of Pottery Style on the LBK Settlement of Balatonszárszó – Kis-Erdei-Dûlõ in Hungary. Acta Terrae Septemcastrensis 7: 197–216. Meyer C. C., Lohr Ch., Gronenborn D., and Alt K. W. 2015. The massacre mass grave of Schöneck-Kilianstädten re­veals new insights into collective violence in Early Neoli­thic Central Europe. Proceedings of the National Acade­ Noémi Beljak Pa/inová, Alena Bistáková my of Sciences of the USA 112: 11217–11222. https://doi.org/10.1073/pnas.1504365112 Moddermann P. J. R. 1970. Linearbandkeramik aus El-sloo und Stein (Analecta Praehistorica 3). University Press. Leiden. Müller-Scheeßel N., Hukel’ová Z., Meadows J., Cheben I., Müller J., and Furholt M. 2021. New burial rites at the end of the Linearbandkeramik in South-west Slovakia. Anti­quity 95(379): 65–84. https://doi.org/10.15184/aqy.2020.103 Neugebauer-Maresch Ch. 1992. Der bandkeramische Fried-hof von Kleinhadersdorf bei Poysdorf, NÖ. Archäologie Österreichs 3(1): 5–11. Neugebauer-Maresch Ch., Lenneis E. 2015. Das Linear-bandkeramische Gräberfeld von Kleinhadersdorf. Mit­teilungen der Prähistorischen Kommission 82. Austrian Academy of Sciences Press. Wien. Nieszery N. 1995. Linearbandkeramische Gräberfelder in Bayern. Internationale Archäologie 16. Verlag Marie L. Leidorf. Espelkamp. Nikitin A. G., Stadler P., Kotova N., +7 authors, and Reich D. 2019. Interactions between earliest Linearbandkeramik farmers and central European hunter gatherers at the dawn of European Neolithization. Scientific Reports 9(1): 19544. https://doi.org/10.1038/s41598-019-56029-2 Niklová M. 2014. Pohreby a pohrebné zvyky kultúry s li­neárnou keramikou na území juhozápadného Slovenska. Studia Historica Nitriensia 18: 213–249. Oross K., Marton T. 2012. Neolithic burials of the Linear-bandkeramik Settlement at Balatonszárszó and their Euro­pean context. Acta Archaeologica Academiae Scientia-rum Hungaricae 63(2): 257–299. Paluch T. 2012. Graves in the Körös culture distribution area in Hungary. In A. Anders, Zs. Siklósi (eds.), The First Neolithic Sites in Central/South-East European Transect Volume III: The Körös Culture in Eastern Hungary. BAR Internationa Series 2334. Archeopress. Oxford: 179–185. Pavlù I., Zápotocká M. 2007. Archeologie pravìkých .ech. Neolit. Archeologický ústav AV .R. Praha. Pavúk J. 1972. Neolithisches Gräberfeld in Nitra. Sloven-ská archeológia 20: 5–105. 1994. .túrovo – ein Siedlungsplatz der Kultur mit Li-nearkeramik und der .eliezovce-Gruppe. Archäologi­sches Institut der Slowakischen Akademie der Wissen­schaften. Nitra. Pechtl J., Hofmann D. 2013. Irregular Burials in the LBK – All or None? In N. Müller-Scheeßel (ed.), “Irreguläre” Best-attungen in der Urgeschichte: Norm, Ritual, Strafe …? Akten der Internationalen Tagung in Frankfurt a. M. vom 3. bis 5. Februar 2012. Römisch-Germanische Kom-mission des Deutschen Archäologischen Instituts. Kollo­quien zur Vor- und Frühgeschichte 19. Bonn: 123–138. Perles C. 2001. The Early Neolithic in Greece. Cambridge World Archaeology. Cambridge. Peschel Ch. 1992. Regel und Ausnahme. Linearband­keramische Bestattungssitten in Deutschland und an-grenzenden Gebieten, unter besonderer Berücksichti­gung der Sonderbestatungen. Internationale Archäologie 9. VML Verlag Marie Leidorf. Erlbach. Pettitt P., Hedges R. 2008. The age of the Vedrovice ce­metery: the AMS radiocarbon dating programme. Anthro­pologie. International Journal of Human Diversity and Evolution 46(2–3): 125–134. http://puvodni.mzm.cz/Anthropologie/article.php?ID=145 Podborský V. 2002. Dvì pohøebistì neolitického lidu s lineární keramikou ve Vedrovicích na Moravì. Masary­kova univerzita. Brno. Raczky P., Anders A., Nagy E., Kurucz K., Hajdú Zs., and Meier-Arendt W. 1997. Polgár-Csõszhalom-dûlõ. Újkõkor végi telep és sírok a Kr. e. V. évezredbõl – Late Neolithic settlement and graves from the 5th millennium B.C. In P. Raczky, T. Kovács, and A. Anders (eds.), Paths into the past. Rescue excavations on the M3 motorway. Hunga­rian National Museum. Archaeological Institute. Eötvös Loránd University. Budapest: 34–43, 168–174. Schlentner U. 1960. Brandbestattung und Seelenglau­ben. Verbreitung und Ursachen der Leichenverbren­nung bei außereuropäischen Völkern. VEB Deutscher Verlag der Wissenschaften. Berlin. Shennan S. 2018. The First Farmers of Europe. An Evo­lutionary Perspective. Cambridge World Archaeology. Cambridge University Press. Cambridge. Skutil J. 1941. Moravské prehistorické výkopy a nálezy. Acta musei Moraviae. Zemské múzeum. Brno: 45–134. Stránská P. 2012. Antropologie populace lidu s kulturou lineární keramiky. In M. .míd (ed.), Kostrové a .árové pohøebi.te kultury s lineárnou keramikou v Kralicích na Hané. Pravìk Supplementum 25. Ústav archeologi­cké památkové pé.e Brno. Brno: 123–135. Sulgostowska Z. 2006. Mesolithic mobility and contacts on areas of the Baltic Sea watershed, the Sudety, and Car­pathian Mountains. Journal of Anthropological Archaeo­ Dead and cremated> on cremation burials in the Linear Pottery culture in Central Europe logy 25:193–203. https://doi.org/10.1016/j.jaa.2005.11.005 Szczepanek A. 2012. Analiza antropologiczna koœci przepa­lonych z cmentarzyska kultury ceramiki wstêgowej rytej w Modlniczce, Stan. 2, Gm. Wielka Wieœ. In A. Czekaj-Za­stawny, M. Przyby³a (eds.), Modlniczka 2, powiat kra­kowski, cmentarzysko kultury ceramiki wstêgowej rytej i osady neolityczne z analizami specjalistycznymi Da­miana Stefañskiego; Anity Szczepanek i Joanny Tr¹b­skiej (Via Archaeologica. ród³a z badañ wykopalisko­wych na trasie autostrady A4 w Ma³opolsce). Instytut Ar-cheologii i Etnologii Polskiej Akademii Nauk. Kraków: 305 –314. Szécsényi-Nagy A., Brandt G., Haak W., +24 authors, and Alt K. W. 2015. Tracing the genetic origin of Europe’s first farmers reveals insights into their social organization. Proceedings of the Royal Society B 282: 20150339. http://dx.doi.org/10.1098/rspb.2015.0339 .míd M. 2008. Kralice na Hané. In Z. .i.máø (ed.), .ivot a smrt v mlad.í dobì kamenné, katalog výstavy. Ústav archeologické památkové pé.e. Brno-Znojmo: 60–62. 2011. Eneolitická pohøebi.te z Prostìjova–.echùvek. Sborník prací Filozofické fakulty brnìnské univerz­ity M16: 60–62. 2012. Kostrové a .árové pohøebi.te kultury s lineár­nou keramikou v Kralicích na Hané. Pravìk Supple-mentum 25. Ústav archeologické památkové pé.e Brno. Brno. Teschler-Nicola M. 2012. The Early Neolithic site Asparn/ Schletz (Lower Austria). In R. J. Schulting, L. Fibiger (eds.), Sticks, stones, and broken bones: Neolithic violence in a European perspective. Oxford University Press. Ox­ford: 101–120. https://doi.org/10.1093/acprof:osobl/978 0199573066.003.0006 Tr¹bska J. 2012. Badania makroskopowe i mikroskopowe neolitycznych zabytkow kamiennych ze stan. 2 w Modlni­czce, gm. Wielka Wieœ. In A. Czekaj-Zastawny, M. Przyby³a (eds.), Modlniczka 2, powiat krakowski, cmentarzysko kultury ceramiki wstêgowej rytej i osady neolityczne z analizami specjalistycznymi Damiana Stefañskiego; Anity Szczepanek i Joanny Tr¹bskiej (Via Archaeologi-ca. ród³a z badañ wykopaliskowych na trasie autostrady A4 w Ma³opolsce). Instytut Archeologii i Etnologii Polskiej Akademii Nauk. Kraków: 315–318. Trampota F., Kvìtina P. 2020. How do they fit together? A case study of Neolithic pottery typology and radiocar­bon chronology. Archeologické rozhledy 72(2): 163–193. https://doi.org/10.35686/AR.2020.6 Trautmann I. 2006. The Significance of Cremation in Early Neolithic Communities in Central Europe. Unpub­lished PhD thesis. University of Tübingen. Tübingen. https://publikationen.uni-tuebingen.de/xmlui/handle/ 10900/49077 Trela-Kieferling E., Zaj¹c M., in press. Analiza zabytków krzemiennych ze stanowiska 2 w Modlniczce, pow. kra­kowski (Via Archaeologica. ród³a z badañ wykopalisko­wych na trasie autostrady A4 w Ma³opolsce). Instytut Ar-cheologii i Etnologii Polskiej Akademii Nauk. Kraków. Ucko P. J. 1969. Ethnography and Archaeological Interpre­tation of Funerary Remains. World Archaeology 1: 262– 280. https://doi.org/10.1080/00438243.1969.9979444 Veit U. 1996. Studien zum Problem der Siedlungsbestat­tungen im europäischen Neolithikum. Tübinger Schrif-ten zur Ur- und Frühgeschichtlichen Archäologie 1. Wax-mann Verlag GmbH. Münster, New York. Vlassa N. 1968. Sondajul de salvare de la “Gura Baciului”. Acta Musei Napocensis 5: 371–379. Wahl J., Trautmann I. 2012. The Neolithic massacre at Tal­heim. In R. J. Schulting, L. Fibiger (eds.), Sticks, stones, and broken bones: Neolithic violence in a European perspective. Oxford University Press. Oxford: 77–100. https://doi.org/10.1093/acprof:osobl/9780199573066. 003.0005 Williams H. 2004. Death Warmed up: The Agency of Bo­dies and Bones in Early Anglo-Saxon Cremation Rites. Journal of Material Culture 9(3): 263–291. https://doi.org/10.1177/1359183504046894 Wunn I. 2001. Götter, Mütter, Ahnenkult. Religionsent­wicklung in der Jungsteinzeit. VML Verlag Marie Leidorf. Rahden. Zeeb-Lanz A. 2019. The Herxheim ritual enclosure: a syn­thesis of results and interpretation approaches. In A. Zeeb-Lanz (ed.), Ritualised destruction in the Early Neolithic: the exceptional site of Herxheim (Palatinate, Germany) 2. Generaldirektion Kulturelles Erbe. Direktion Landesar­chäologie. Speyer: 423–482. Zeeb-Lanz A., Haack F. 2016. Diversity in ritual practice at the end of the LBK. In L. Amkreutz, F. Haack, D. Hofmann, and I. van Wijk (eds.), Something out of the ordinary? Interpreting diversity in the Early Neolithic Linearband­keramik and beyond. Cambridge Scholars Publishing. Cambridge: 247–279. back to contents Documenta Praehistorica XLIX (2022) Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary Judit Regenye1, Krisztián Oross2, Eszter Bánffy3,2, Elaine Dunbar4, Ronny Friedrich5, Alex Bayliss 6, Nancy Beavan7, Bisserka Gaydarska 8, and Alasdair Whittle 9 whittle@cardiff.ac.uk 1 Laczkó Dezso½ Museum, Veszprém, HU 2 Institute of Archaeology, Research Centre for the Humanities, Eötvös Loránd Research Network, Centre of Excellence of the Hungarian Academy of Sciences, Budapest, HU 3 Römisch-Germanische Kommission des Deutschen Archäologischen Instituts, Frankfurt\Main, DE 4 SUERC Radiocarbon Dating Laboratory, East Kilbride, UK 5 Curt-Engelhorn-Zentrum Archäometrie GmbH, Mannheim, DE 6 Historic England, London, UK 7 Institute of Environmental Science and Research Limited (ESR), Kenepuru Science Centre, Porirua, NZ 8 Manchester Metropolitan University, Manchester, UK 9 Department of Archaeology and Conservation, Cardiff University, Cardiff, UK ABSTRACT – A handful of new radiocarbon dates from three Balaton-Lasinja culture graves at the site of Veszprém-Jutasi út in western Hungary form the starting point for formal models for late Lengyel and post-Lengyel chronology in that region. The graves date to the later fifth millennium cal BC. They provide the opportunity to put the earlier Copper Age Balaton-Lasinja culture of Trans-danubia into its regional and wider context, and to highlight both gradually improving understand­ing of its character and remaining problems of chronology and classification. The Balaton-Lasinja culture was part of a whole series of regional shifts in settlement and society connected to the end of the Neolithic and the demise of major settlement aggregations which had dominated lifestyles in previous centuries. This study indicates how much further detailed research continues to be needed to get fully to grips with this set of important changes, which run on into the Copper Age. Contrasts are drawn between western and eastern Hungary, and the uncertainties surrounding the chronol­ogy of the fourth millennium cal BC, including for the Furchenstich pottery style, are emphasised. KEY WORDS – western Hungary; Late Neolithic and Copper Age; post-Vin.a development; Balaton- Lasinja culture; radiocarbon dating; Bayesian chronological modelling Nekaj grobov kulture Balaton-Lasinja iz najdi[;a Veszprem-Jutasi út in oris kronologije starej[e bakrene dobe na zahodnem Mad/arskem IZVLE.EK – Novi radiokarbonski datumi iz treh grobov kulture Balaton-Lasinja v Veszprém-Jutasi út na zahodu Mad.arske predstavljajo izhodi..e za izdelavo formalnih pozno lengyelskih in postleng­yelskih kronolo.kih modelov v regiji. Grobovi so datirani v pozno peto tiso.letje pr. n. .t. Omogo.a­jo umestitev zgodnje bakrenodobne Balaton-Lasinja kulture v Transdanubiji v regionalni in .ir.i kon­tekst in pomagajo pri bolj.em razumevanju njenega zna.aja, kronolo.kih te.av in klasifikacije. Kul­tura Balaton-Lasinja je del vrste regionalnih poselitvenih in dru.benih premikov na koncu neoliti­ka, povezanih z razpadom ve.jih poselitvenih obmo.ij, ki so prevladovala v prej.njih stoletjih. .tudi­ja ka.e koliko detajlnih raziskav je .e potrebnih, da bi razumeli spremembe, ki so se nadaljevala DOI> 10.4312\dp.49.4 Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary v bakreno dobo. Predstavljamo kontraste med zahodno in vzhodno Mad.arsko in opozarjamo na ne­zanesljivo kronologijo .etrtega tiso.letja pr. n. .t., tudi lon.arskega sloga brazdastega vreza. KLJU.NE BESEDE – zahodna Mad.arska; pozni neolitik in eneolitik; post-Vin.a razvoj; lasinjska kultura; radiokarbonsko datiranje; Bayesovo kronolo.ko modeliranje Introduction: aims and contexts This paper presents a handful of new radiocarbon dates from three Balaton-Lasinja culture graves at the site of Veszprém-Jutasi út in western Hungary, as well as a series of formal models for late Lengyel and post-Lengyel chronology of that region based on the small number of existing dates. The graves date to the later fifth millennium cal BC. They pro­vide the opportunity to put the middle Copper Age Balaton-Lasinja culture of Transdanubia into its re­gional and wider context, and to highlight both gradually improving understanding of its character and remaining problems of chronology and classi­fication. The Balaton-Lasinja culture was part of a whole series of regional shifts in settlement and society connected to the end and aftermath of the Neolithic and to the demise of major settlement ag­gregations, including tells, which had dominated lifestyles in previous centuries. This study indicates how much further detailed research continues to be needed to get fully to grips with this set of impor­tant changes, which run on into the Copper Age. The immediate focus here is on western Hungary, but there are implications for other neighbouring re­gions as well. From the Late Neolithic to the Copper Age in western Hungary and beyond There has long been an imbalance in studies of the Neolithic and the Copper Age in Hungary between its eastern and western parts (Kalicz 1991; Bánffy 1994). This applies not least to the end of the Neo­lithic, broadly in the middle of the fifth millennium cal BC. This period is represented by a series of changes in eastern Hungary that are archaeologi­cally well documented, like the abandonment of tell settlements and the appearance of new pottery styles. A significant shift in subsistence strategy has also been noted, with more reliance on stockbreed­ing and pastoralism in emerging Early Copper Age communities (Bánffy 1995). The Copper Age of east­ern Hungary was already well-investigated by the mid-late 1950s (Kutzián 1955; Banner, Bognár-Ku­tzián 1961), and even underpinned by some reliable stratigraphic observations (Kalicz 1958). Its pre­sumed cultural and chronological sequence (from Tiszapolgár to Bodrogkeresztúr to Baden), was estab­lished in the early 1960s, but has been severely que­stioned more recently (Raczky, Siklósi 2013; Sikló­si, Szilágyi 2021; and see below), allowing consid­erable overlap between Tiszapolgár and Bodrogke­resztúr. In contrast, research on the same times in Transda­nubia in western Hungary has historically lagged behind. One obvious gap here was between the Late Neolithic Lengyel culture and the Late Copper Age Baden culture. From the 1970s onwards (Raczky 1974), the late phase of the long-enduring Lengyel culture, Lengyel III, was seen to persist in the peri­od that was traditionally called the Early Copper Age in the central Carpathian basin, contemporary with Tiszapolgár in the previously accepted cultural scheme for eastern Hungary (Raczky, Siklósi 2013; Siklósi, Szilágyi 2021). Just as there was historically uneven knowledge of Lengyel settlement in Transdanubia – whilst large settlements and cemeteries were found and exca­vated in east Transdanubia, including the epony­mous site (Wosinsky 1889; 1891) and a recent site with more than 2000 burials at Alsónyék (Osztás et al. 2016), data from the western Lengyel complex is characterised by smaller, dispersed settlements, with almost entirely ‘Sonderbestattungen’ or unusu­al graves – very little was known about late Lengyel settlement before 1974 (Raczky 1974) and more robust settlement data followed only in the 1990s (Bánffy 1995), with subsequent insights into chan­ges in subsistence strategy in favour of stockbre­eding (Bánffy 1994; 1995; Barna et al. 2019). Ac­cording to current knowledge, the Lengyel III hori­zon can be divided into two typo-chronological pha­ses, IIIa and IIIb. The latter was taken to represent the gradual adoption of more elements common in the subsequent Balaton-Lasinja culture (Kalicz 1991; Bánffy 1994). As part of a larger cultural formation already described in Croatia, Slovenia and eastern Austria, the Balaton-Lasinja culture was initially sup­posed to be the result of a massive migration into southern Transdanubia (Kalicz 1969), based on the work of Josip Koro.ec (1958) who summarised its J. Regenye, K. Oross, E. Bánffy, E. Dunbar, R. Friedrich, A. Bayliss, N. Beavan, B. Gaydarska, and A. Whittle principal characteristics in the late 1950s, and on 25 sites of the Lasinja culture published by Stojan Dimi­trijevi. (1961), who considered them to be part of the suite of post-Vin.a phenomena. The Balaton-Lasinja culture One of the earliest identifications of the Balaton-La­sinja culture resulted from systematic work on the archaeological topography of Hungary in the early 1960s. Volumes 1 and 2 were published on results from the Balaton Uplands in County Veszprém, where surface collections were complemented by small excavations (Bakay et al. 1966; Éri et al. 1969). The similarity of some assemblages to re­cently published Slovenian and Croatian material led to the definition of the Balaton group in 1969 by Nándor Kalicz, who stressed the dominance of southern traditions in its early phase of its development, while also emphasising cen­tral European ties both in the ceramic assemblages and in the metallurgy of its later phase (Kalicz 1969). Soon after, ceramics were already being classified as Balaton-Lasinja I, II and III types (Ka­licz 1973). They became Ba-laton-Lasinja I and Balaton-Lasinja II-III cultures by the early 1980s in a further com­prehensive survey, in which Kalicz discussed the dynam­ics, the cultural connections and the metallurgy of the pe­riod (Kalicz 1982). The label of ‘Furchenstich’ was also pro­posed for the more recent pot­tery assemblages in this pro­posed typo-chronological se­quence. Related cultural groups are also known with many different names in ad­jacent regions (such as Retz-Gajary in Croatia, Baj.-Retz in Slovakia and Mondsee in Au­stria). set to be contemporaneous with Tiszapolgár, and Balaton-Lasinja with Bodrogkeresztúr, in eastern Hungary, and with Ludanice in the north-central part of the Carpathian basin; Furchenstich was seen as parallel to the Hunyadihalom group (Bodrogke­resztúr B). Protoboleráz already marked the emer­gence of Late Copper Age cultural groupings, such as Boleráz and Baden. Although meticulous studies of typo-chronology were the principal focus for decades, architecture, settle­ment structure, burials, subsistence strategy and me­tallurgy also received attention. Two major types of houses of the Lengyel III and Balaton-Lasinja peri­ods can be distinguished. Bipartite houses, with a larger northern and a smaller southern room, pro­ved to be very characteristic for the period. No tra- Kalicz went on to distinguish Balaton-Lasinja III as the Pro-toboleráz horizon, with some sites in eastern Hungary as well (Kalicz 1991; 2001). In this system, Lengyel III was Fig. 1. Radiocarbon dated earlier Copper Age sites in western Hungary: 1 Balatonõszöd-Temetõi-dûlõ; 2 Balatonszárszó-Kis-erdei-dûlõ; 3 Gyõr-Szabadrétdomb; 4 Keszthely-Fenékpuszta, Halászrét-Nádgazdaság agyag­bányája; 5 Keszthely-Fenékpuszta, Vasúti õrház; 6 Mosonszentmiklós-Pál-major; 7 Nagykanizsa-Sánc; 8 Szentgál-Teleki dûlõ; 9 Szombathely-Metro; 10 Veszprém-Jutasi út; 11 Zalaszentbalázs-Szõlõhegyi mezõ; 12 Zalavár-Mekenye. Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary ces of internal division and just a single row of postholes in the position of the ridge purlin were observed in other buildings, which were extremely variable in shape and size. Some further construc­tions are more like Neolithic Lengyel houses with a single cross-row of heavy posts in their interior. Bedding trenches usually mark the walls of the hou­ses. The exceptional visibility of the house plans, comparable only with those of LBK buildings, is due to these features. Different house types were dated to both the late Lengyel and Balaton-Lasinja peri­ods, and no strict architectural sequence could be distinguished (Virág 2003; 2005; Virág, Figler 2007; Oross et al. 2010). Balaton-Lasinja settlement layouts are clearly looser than those of the densely built-up extended Neoli­thic sites of the LBK and the Lengyel periods. Small clusters of houses and a limited number of buildings have been discovered at sites such as at Zalavár-Basa­sziget and Balatonszárszó-Kis-erdei-dûlõ (Virág 2003; Oross et al. 2010). Archaeozoological and especially archaeobotanical evidence has remained extremely scarce. In sharp contrast to the extended extramural cemeteries of the earlier phases of the Copper Age in eastern Hungary, only single burials and small grave groups have so far been discovered on con­temporaneous Transdanubian sites. Metallurgy in western Hungary does not present the abundance of large copper axes and the series of other different artefacts known from eastern Hun­gary that reveal close ties with the Balkans. Attri­buted to a central European tradition, hooked cop­per spirals, so-called glass-shaped spirals, and dif­ferent kinds of copper wire belong to repertoires re­covered in the western Carpathian basin. However, the so-called Csáford-Stollhof type gold discs form a very spectacular and characteristic group of objects that have their close counterparts in the adjacent central European regions, sometimes also made from copper, and rarely from silver (Korek 1960; Bóna 1963/1964.33–37; Makkay 1976; Kalicz 1982. 10–16). The Balaton-Lasinja graves at Veszprém-Jutasi út During the excavation of the Neolithic and Copper Age site of Veszprém-Jutasi út in 2003, some Bala­ton-Lasinja graves in the southern and eastern parts of the excavated area were recorded. Settlement fea­tures related to the graves could not be observed on the site, although Balaton-Lasinja pottery occurred in the Lengyel pits, into which Copper Age graves were cut. Balaton-Lasinja pottery was found in a well-defined small area around the graves, and the situ­ation can be interpreted as relating to burial ritual. The scattering of broken vessels into graves is a phe­nomenon known in other Balaton-Lasinja graves. Veszprém-Jutasi út (earlier called Felszabadulás út) is a well-known site of the Lengyel culture, entering the literature as the first known settlement of the late Lengyel period (Raczky 1974). The western edge of the site was investigated during the exca­vation in 2003, and mainly the early stage in the set­tlement’s life (Regenye 2004; 2006; 2007; Regenye, Biró 2014; 2019). Based on previous research, we know that the Lengyel site includes an extended Lengyel III settlement in the eastern direction as well; the centre of the site shifted eastward over time. The Balaton-Lasinja graves were located in a group on the westernmost edge of this large Lengyel III settlement. The graves excavated in 2003 are mostly dated to the Lengyel culture. Eight burials can be listed here. These graves are located in a group. Four more gra­ves (9, 13, 14, and 15), about 90m away from this group, were interpreted as burials of the Balaton-La­sinja culture (Regenye 2006). One of these, grave 15, an unfurnished burial, was subsequently shown by radiocarbon dating to be of Lengyel date (SUERC­54643; Regenye et al. 2020.Tab. 2). So only three graves of the Balaton-Lasinja culture are securely recorded. Grave 9 is that of a child (of Infans II age) lying on its right side in a contracted position, the skeleton oriented east–west. Grave 13 Fig. 2. Veszprém-Jutasi út, grave 9. J. Regenye, K. Oross, E. Bánffy, E. Dunbar, R. Friedrich, A. Bayliss, N. Beavan, B. Gaydarska, and A. Whittle is that of an adult man, lying on his right side in a contracted position, his skeleton oriented southwest–northeast. Grave 14 is that of an adult woman (of Maturus-Senilis age, Köhler 2006), her upper body lying in supine position, the legs bending to the left, and skeleton oriented south-southeast– north-northwest. The grave pit could be observed only in the case of grave 9, which was roughly round, 1.2m in diameter. The other two graves were buried into a Lengyel III pit; the form of the graves could not be determined. All three burials are inhumations, but their orientation varies. Grave 9 has an east– west layout; the orientation of graves 13 and the 14 varies slightly from south–north. Two of the skeletons (9, 13) lie on their right side and the third turns to the left. The location of the two graves in a pit roughly parallel to each other may suggest a linked burial. Grave 14 was covered with irregular large limestone blocks which extended partly to grave 13. There was a possibility of post-depositional manip­ulation or disturbance in the case of grave 14, the left arm being separated between the two skeletons and covered by the stones. Two of the three graves had grave goods. Ceramics, ornaments and several stone implements are in­cluded in the child’s grave, but only one of the adult graves had a single vessel. Both the composition of the grave goods and their placement in the grave fol­low Lengyel traditions in the case of the child’s grave. One vessel was placed behind the head, two in front of it. At the neck were pieces of shell ornaments, in the form of rectangular flat beads (two intact, with three fragments, with two drilled holes at the top) made from river shells, and an unworked shell. The stone tools were placed around the head: a core in front of the neck, a core-flake behind the head, and a polished basalt axe preform above the head. Three out of four vessels were in grave 9 (that of the child) and the fourth in one of the two adults, in grave 13. In that grave fill there were two more pots, incomplete pieces of a pedestalled vessel and a jug. These are not included in the grave furniture. Together with the fragments of different pots found in the grave fill beneath the stone deposit and un­der the legs of grave 14, they may have been placed in the grave as part of the funerary rite. Both the graves had a pedestalled bowl with a bell-shaped base; in grave 9 there were also a biconical bowl and a small cup. The vessels are typical forms of the Transdanubian Copper Age (Kalicz 1995b. 75–76; 2003.14); both the biconical bowl and the bell-shaped pedestal are basic forms, just like the pe­destalled bowl in grave 13 (cf. Kalicz 1995a.Fig. 3, 12). The biconical bowl with a short upper part is a frequent type deriving from the Lengyel culture, but the decoration of the upper part, typical of the Balaton-Lasinja culture, is missing here. The bowl on a pedestal ornamented with four drop-shape knobs fits well into the repertoire of both the Balaton-Lasi­nja (Horváth, Simon 2003.Figs. 21.9,24.3) and the Ludanice culture (Virág 1995.Abb. 2, 10). The little biconical cup is not characteristic of the Balaton-La­sinja culture, lacking decoration or handle. The four pots found in the two graves show strong Lengyel traditions. The vessel forms are typical of the Balaton-Lasinja culture, but the characteristic channelled decoration is missing. Based on these characteristic features, these ceramics can be relat­ed to the material of the north Transdanubian sites. Radiocarbon dating New radiocarbon measurements were obtained on three human skeletons from the Balaton-Lasinja gra­ves at Veszprém-Jutasi út as part of the Times of Their Lives project (see Acknowledgements, and Ta­ble 1). Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary The samples dated at the Curt-Engelhorn-Zentrum Archäometrie (CEZA), Mannheim (sample identifiers ‘MAMS’), were prepared by gelatinisation and ultra-filtration (Brown et al. 1988), combusted in an ele­mental analyser, graphitised and dated by Accelera­tor Mass Spectrometry (AMS) (Kromer et al. 2013). The samples dated at the Scottish Universities En­vironmental Research Centre (sample identifiers ‘SUERC’), East Kilbride, were also gelatinised, ultra-filtered and then graphitised and dated by AMS (Dunbar et al. 2016). These measurements are conventio­nal radiocarbon ages (Stuiver, Po-lach 1977). At CEZA they have been corrected for fractionation using .13C values measured by AMS. These val­ues can include an element of frac­tionation introduced during the pre­paration and measurement of the samples in addition to the natural isotopic composition of the sample, and so they are not suitable for die­tary analysis. For this reason, where sufficient material was available, sub­samples of the dated gelatin pre­pared at MAMS- were analysed for .13C and .15N using IRMS at the Iso- trace Facility, University of Otago Chemistry Depart­ment, using methods outlined by Beavan Athfield et al. (2008.3). At SUERC, .13C and .15N samples were prepared and analysed from sub-samples of the dated gelatin as described by Kerry Louise Sayle et al. (2014), and these .13C values were used for age calculation. Groups of replicate radiocarbon measurements are available on two skeletons, both of which are statisti­cally inconsistent at the 5% significance level (Ward, Wilson 1978) (Tab. 1). Those for Ve-9 are statisti­cally consistent at the 1% significance level and have been combined before inclusion in the chrono­logical models, but those from Ve-13 are significant­ly divergent. This degree of replication arises from attempts to resolve some differences between repli­cate measurements reported in August 2014 by Man­nheim (MAMS-21328–41) and East Kilbride (SUERC­54631–4, SUERC-54638–44 and SUERC-54648–9) and is discussed further by Judit Regenye et al. (2020.24). Although the .13CAMS value for MAMS­14829 is unusually depleted, the remaining three measurements are still significantly divergent (T’= 22.0, T’(1%)=11.3, .=2) and so, in the absence of contextual information suggesting which results may be in error, we have again incorporated a weighted mean of the results in our modelling. Details of existing radiocarbon measurements from settlement features associated with ceramics of the Balaton-Lasinja phase are listed in Table 2, those of the preceding Lengyel III ceramic phase in Table 3, and those of the succeeding, related, Furchenstich phase in Table 4. J. Regenye, K. Oross, E. Bánffy, E. Dunbar, R. Friedrich, A. Bayliss, N. Beavan, B. Gaydarska, and A. Whittle Laboratory number Grave\Sample number Material and context Radiocar­bon Age (BP) . 13 C (‰) .AMS] . 13 C (‰) .IRMS] . 15 N (‰) C>N ratio MAMS-21338 Ve-9 Human bone, left femur from 8–9-year-old child inhumation in grave 9 with three diagnostic pottery vessels 5307±33 –18.4 MAMS-23171 Ve-9 Replicate of MAMS-21338 5406±27 –19.6 MAMS-14828 Ve-9 Replicate of MAMS-21338 5418±29 –24.7 14 C age> 5384±17 BP, T'=7.4, T'(5%)=6.0, . =2Human bone, right femur from 23–40-year-old male inhumationSUERC-54641 Ve-13(i) in grave 13 with one diagnostic pottery vessel and cut into feature 5431±33 –19.2±0.2 10.1±0.3 3.2 348, a settlement pit containing Lengyel III ceramics MAMS-21339-1 Ve-13(ii) Replicate of SUERC-54641 5251±21 –19.6 –19.6±0.11 10.0±0.13 3.3 MAMS-21339-2 Ve-13 (ii) Replicate of SUERC-54641 5279±26 MAMS-14829 Ve-13 Replicate of SUERC-54641 5213±31 –30.4 14 C age> 5281±14 BP, T'=27.9, T'(5%)=7.8, . =3< . 13 C> –19.5±0.1‰, T'=3.1, T'(5%)=3.8, . =1< . 15 N> +10.0±0.12‰, T'=0.1, T'(5%)=3.8, . =1 Human bone from grave 14 with diagnostic sherds in the fill, lying SUERC-54642 Ve-14 parallel with grave Ve-13 and cut into feature 348, a settlement pit 5384±33 –19.6±0.2 10.0±0.3 3.3 containing Lengyel III ceramics Tab. 1. Radiocarbon and stable isotopic measurements for Balaton-Lasinja graves from Veszprém-Jutasi út. Radiocarbon calibration and reservoir correc­tions Radiocarbon ages from fully terrestrial samples have been calibrated using IntCal20 (Reimer et al. 2020). The measurement on mussel shell from Balaton­õszöd-Temetõi-dûlõ (KI-16690) probably has a fresh­water reservoir effect derived from the nearby Lake Balaton. Unfortunately, no data are currently avail­able for the freshwater reservoir in the lake, and so we use a value of 540±70 BP as calculated for Sche-la Cladovei on the Danube at the Iron Gates (Cook et al. 2001). Human bone can also exhibit a reservoir age, if peo­ple ate foods that did not derive entirely from the terrestrial biosphere. Accurate calibration in this case requires the proportions of different diet-sources in each individual to be estimated. This allows a mixed-source calibration curve to be constructed for each person, which incorporates the aquatic reservoir in the appropriate proportion for that individual. For this reason, source-proportional dietary modelling was undertaken for the dated human skeletons in this study. Existing pairs of radiocarbon measure­ments on human and animal bone from Neolithic and Copper Age graves in this region, however, are statistically consistent, suggesting that consumption of non-terrestrial foods by the population was prob­ably negligible (Bayliss et al. 2016.Tab. 1; Jakucs et al. 2016.Tab. 1; Raczky, Siklósi 2013. Tab. 1). This does not mean, however, that particular individuals might not have consumed a larger component of freshwater resources. Proportional dietary analysis for the Balaton-Lasinja skeletons from Veszprém-Jutasi út The .13C and .15N isotopic values for burials Ve-13 and Ve-14 are from replicate measurements, and the weighted means (Ward, Wilson 1978) (Tab. 1) have been used in the analysis. As no stable isotopic val­ues have been measured from Ve-9, the proportio­nal diet estimates have been derived from the mean FRUITS (Food Reconstruction Using Isotopic Trans­ferred Signals; Fernandes et al. 2014) estimates for Veszprém burials for which there are stable isotope values (Regenye et al. 2020.Tab. 4a). The individual mixed-source calibration curve for each of the Veszprém burials incorporates the aquat­ic reservoir in the proportion suggested by the die­tary estimates provided by the Bayesian mixing mo­ Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary Laboratory number Site Material and context Radiocar­bon Age (BP) . 13 C (‰) .AMS] . 13 C (‰) .IRMS] . 15 N (‰) C>N ratio References OxA-13784 Balatonszárszó­Kis-erdei-du½lo½ Disarticulated Sus scrofa domesticus, right ulna, from feature B-5713, internal posthole of house A43 5356±34 –19.8±0.3 9.95±0.4 3.2 Oross et al. 2010.392 deb-13379 Balatono½szöd-Temeto½i-du½lo½ “Pit 1099. Well 1> Grave 70, human (Boleráz)skeletons in fill of the well of the Balaton-Lasinja culture” 4480±70 –20.6 Horváth et al. 2008. Tab. 1 VERA-4806 Balatono½szöd-Temeto½i-du½lo½ Unidentified animal bone from pit B-432 5000±40 –26.8±0.9 Horváth et al. 2014a. Tab. 2 deb-2171 Gyo½r-Szabadrétdomb Unidentified animal bone from feature 29 5160±60 –20.25 Figler et al. 1997.Tab 2 Bln-1206 Nagykanizsa-Sánc Sample material and context unknown 4940±|80* Kalicz 1995a.41 Bln-1207 Nagykanizsa-Sánc Sample material and context unknown 4990±80 Kalicz 1995a, 41 MAMS-44913 Mosonszent­miklós-Pálmajor Cattle tibia, found in a closed assemblage includinga clay horn which can be played as a musical instru­ment, several intact vessels and an intact cattle femur squeezed into the horn 5387±25 –21.1 3.2 Bánffy, Egry forth­coming * Bln-1206–7 were published by Kalicz (1995.41), following information from a letter by Hans Quitta, with neither standard deviations, details of the material dated nor contextual infor­mation. Kalicz (1982.10) writes that radiocarbon dates for the Balaton-Lasinja culture are between 3040–2980 uncal bc ±80 y. Footnote 38 makes clear that this is based on the datesreported from Nagykanizsa-Sánc in Quitta’s letter. We infer from this information that the quoted error on these measurements may have been ±80 BP. Tab. 2. Radiocarbon and stable isotopic measurements associated with Balaton-Lasinja settlement features. del FRUITS vß2.0. FRUITS produces esti­mates of the mean percentage (and stan­dard deviation) for each of the possible food sources making up the diet for each consumer. To estimate the relationship between an individual’s isotopic profile and the food sources that were likely available, we first create a FRUITS model starting with the ’baseline’ isotopic values for foods from the isotopic averages of each likely food source. The FRUITS proportional dietary estimates were modelled on two diet pro­xies (.13C and .15N) for the average iso­topic data and its associated mean error for each of three general food sources – cereals, terrestrial herbivores and omni­vores (cattle, sheep, and pigs), and fresh­water fish. The cereals baseline used car­bon and nitrogen values for archaeobo­tanical samples of wheat (n=12) and bar­ley (n=6) from Nives Ogrinc and Mihael Budja (2005), and emmer wheat (n=1) and barley (n=3) from Amy Bogaard et al. (2013), producing mean cereal values and errors of –24.6±0.3‰ (.13C) and +5.0± 0.4‰ (.15N). The food baseline data are particularly robust for animal protein sour­ces, as the data are drawn from sites with­in approximately 1000km of Veszprém-Ju­tasi út. Baseline values for terrestrial ani­mals (pig, sheep, cow, n=89; .13C –20.3± 0.2‰ and .15N +6.9±0.2‰) are from fau­nal materials in the Star.evo, Sopot, and Lengyel sites at Alsónyék-Bátaszék (includ­ing 27 sets of analyses on terrestrial fauna provided by the Bioarchaeology Work-group Mainz; Bayliss et al. 2016). These values for terrestrial fauna complement mean isotope ranges cited from sites with­in 250km of our study area (cf. Gamarra et al. 2018; McClure et al. 2020), extend­ing the potential geographical relevance of the baseline terrestrial fauna isotope values in this study. Isotopic values for archaeological freshwater fish were drawn from Olaf Nehlich et al. (2010; n= 3), Du­.an Bori. et al. (2004; n=12), and Ale­xandra Bayliss et al. (2016; n=4) and were further supplemented with six sets of carbon and nitrogen values on fish from Alsónyék-Bátaszék, also provided by the Bioarchaeology Workgroup Mainz. This J. Regenye, K. Oross, E. Bánffy, E. Dunbar, R. Friedrich, A. Bayliss, N. Beavan, B. Gaydarska, and A. Whittle Laboratory number Site Material and context Radiocarbon Age (BP) . 13C (‰) .IRMS] References deb-10274 Szengál-Teleki du½lo½ Unidentified animal bone from pit 1, Lengyel III 5530±60 –20.3 Regenye 2011.43 deb-3365 Zalaszentbalázs-Szo½lo½hegyi mezo½ Unidentified charcoal from feature 5\2, –45cm 5728±58 –25.3 Bánffy 1995 deb-3378 –\\– Unidentified charcoal from feature 2, –80cm 5767±70 –25.8 Bánffy 1995 deb-3385 –\\– Unidentified charcoal from feature 4\2, –100cm 5720±71 –24.9 Bánffy 1995 deb-3379 –\\– Cattle tibia from feature 4, –85–90cm 5682±57 –21.8 Bánffy 1995 deb-3380 –\\– Cattle from feature 4, –90cm 5614±70 –20.8 Bánffy 1995 deb-8408 Szombathely-Metro Unidentified animal bone, from foundation trench of house 1 5520±60 Ilon 2004.27, Fig. 26 deb-8486 –\\– Unidentified animal bone, feature 35 5590±60 Ilon 2004.Fig. 26 deb-8518 –\\– Unidentified animal bone, feature 19 5450±90 Ilon 2004.Fig. 26 Tab. 3. Radiocarbon and stable isotopic measurements associated with Lengyel III ceramics. group of samples provided a final freshwater fish baseline value of .13C of –21.4±0.2‰ and .15N of +8.7±0.2‰. The FRUITS proportional diet model also uses the metabolic enrichment of stable isotope values that occur in the course of building consumer tissue. This is known as an ‘isotopic offset’ between diet and consumer. We used an isotopic offset in the FRUITS model of 4.8±0.2‰ for .13C (Fernandes et al. 2014), and for 6.0±0.5‰ for .15N (O’Connell et al. 2012). The FRUITS dietary model also allows for further constraints on the calculations from a priori obser­vations in the archaeological record and logical con­siderations. We had observed in a previous dietary analysis of populations at Alsónyék-Bátaszék (Bay­liss et al. 2016.40–46) that there was a possibility for the consumption of freshwater resources. We ran tests of the FRUITS model first using no prior infor­mation, and then with prior information weighting terrestrial protein over fish. The final version of the FRUITS model was modified, incorporating prior in­formation that the proportion of terrestrial protein was greater than that of cereals which added weight Lab. number Site Material and context Radiocarbon Age (BP) . 13C (‰) .IRMS] References Bln-500 Keszthely-Fenék- Charcoal (Quercus sp.), from pit Objekt 2, 4780±80 Quitta, Kohl 1969. puszta, Vasúti o½rház at 1.5m depth. Finds include lots of sherds of Gajary-type assigned by Kalicz (1982.10, fn39) to Balaton-Lasinja II–III group 241 Bln-501 Keszthely-Fenék­puszta, Halászrét-Nádgazdaság agyagbányája Charcoal (Quercus sp.), from big, multi-part pit complex. Sample at 1m depth in pit sector 2. Lower part of pit currently below water table. Finds include lots of sherds of Gajary-type assigned by Kalicz (1982.10, fn39) to Balaton-Lasinja II–III group 4890±80 Quitta, Kohl 1969. 242 Bln-502 Zalavár-Mekenye Charcoal (Abies cf. alba Mill.), from settle­ment pit 13, lower down in fill at 1m below surface. Finds include white-painted late Lengyel sherds and undecorated sherds and associated copper smelting debris. 5400±80 Quitta, Kohl 1969. 247 KI-16690 Balatono½szöd-Temeto½i-dilo½ Mussel from pit B-1984 5210±40 –10.3±0.3 Horváth et al. 2014a.Tab. 2 deb-2196 Gyo½r-Szabadrétdomb Unidentified animal bone from feature 251 4650±60 –20.89 Figler et al. 1997. Tab. 2 deb-2194 –\\– Unidentified animal bone from feature 251 4850±60 –19.49 –\\– deb-2178 –\\– Unidentified animal bone from feature 510 4730±60 –20.81 –\\– deb-2213 –\\– Unidentified charcoal from feature 510 4630±60 –26.27 –\\– deb-2095 –\\– Unidentified animal bone from feature 589 4820±60 –20.11 –\\– Tab. 4. Radiocarbon and stable isotopic measurements associated with Furchenstich features and ceram­ics. Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary Grave Sex Age . 13C (‰) . 15N (‰) Cereals Terrestrial protein Freshwater fish Ve-9 Unknown 8–9 years – – 48.3±1.1% 50±1.2% 1.7±1.7% Ve-13 Male 23–40 years –19.5±0.1* +10±0.21* 48.4±1.1% 49.9±1.2% 1.7±1.7% Ve-14 Female over 40 years –19.6±0.2* +10±0.3* 47.9±1.5% 49.9±1.5% 2.2±2.2% Tab. 5. Proportional dietary estimates for the dated burials from Veszprém-Jutasi út, derived from the FRUITS analysis (*mean of values from Table 1). to the higher .15N contribution of terrestrial meat in the diet. The results of the FRUITS analysis for the individu­als from Veszprém-Jutasi út indicates the population all had diets that were made up almost entirely of cereals and terrestrial protein, and that freshwater fish are a negligible contribution to the diet (Tab. 5). Chronological modelling The Bayesian chronological modelling has been un­dertaken using the program OxCal v4.4 (Bronk Ram­sey 2009). The algorithms used are defined exactly by the brackets and OxCal keywords on the left-hand side of Figures 6 and 8 (http://c14.arch.ox.ac.uk/). The posterior density estimates output by the mod­els are shown in black, with the unconstrained cali­brated radiocarbon dates shown in outline. The other distributions correspond to aspects of the mo­del. For example, the distribution “start Veszprém Lengyel” (Fig. 6) is the posterior density estimate for the start of Lengyel burial at Veszprém. In the text and tables, the Highest Posterior Density intervals of the posterior density estimates are given in italics. Furnished inhumation burials from Veszprém-Jutasi út The first model examines the dating of furnished in-humation burials at Veszprém (Fig. 6). Five furnish­ed Lengyel graves have been dated (Regenye et al. 2020.Tab. 2), four of which have been included in the correspondence analysis of Lengyel funerary ce­ramics (Regenye et al. 2020.Fig. 2). Ve-5 is allocated to phase 2 of that seriation, and ve-2, ve-3, and ve-7 to phase 3. One further, dated but unfurnished, grave appears to belong to this period of activity (Regenye et al. 2020.Fig 5). The model illustrated in Figure 6 suggests that Len-gyel burial at Veszprém began in 4950–4710 cal BC (95% probability; start Veszprém Lengyel) (Fig. 6), probably in 4830–4725 cal BC (68% probability). It ended in 4695–4485 cal BC (95% probability; end Veszprém Lengyel) (Fig. 6), probably in 4670–4575 cal BC (68% probability), having been used for a period of 30–245 years (95% probability; duration Veszprém Lengyel) (Fig. 7), probably for 65–180 years (68% probability). There was then an extend­ed gap with no furnished burial on the site, which endured for a period of 20–490 years (95% proba­bility; gap Veszprém Lengyel/Balaton-Lasinja (Fig. 7), probably for 200–435 years (68% probability). Only three Balaton-Lasinja graves from Veszprém have radiocarbon dates, two of which are furnished with diagnostic pottery. The other has Balaton-Lasi­nja sherds in the grave fill, and is parallel to one of the furnished graves. The model shown in Figure 6 suggests that this period of burial began in 4565– 4160 cal BC (93% probability; start Veszprém Bala­ton-Lasinja) (Fig. 6) or 4150–4115 cal BC (2% pro­bability), probably in 4370–4180 cal BC (68% pro­bability). It ended in 4235–3575 cal BC (95% pro­bability; end Veszprém Balaton-Lasinja) (Fig. 6), probably in 4215–4010 cal BC (68% probability), having been used for a period of 1–215 years (93% probability; duration Veszprém Balaton-Lasinja (Fig. 7) or 230–260 years (2% probability), probably for a period of 10–130 years (67% probability) or 155–165 years (1% probability). The earlier Copper Age in western Hungary The second model explores the context of the Vesz­prém cemeteries within the ceramic sequence of the earlier Copper Age in western Hungary (Figs. 8–11). First, we consider the relationship between the end of Lengyel furnished burial (which is largely restrict­ed to the eastern part of this region) and the start of Lengyel III ceramics, which do not appear in gra­ves and are restricted to settlement contexts. Regenye et al. (2020.Figs. 11–12) present a model for a three-phase seriation of Lengyel graves, which we have re­calculated using IntCal20 (Reimer et al. 2020). This suggests that Lengyel furnished burial started to de­cline in 4770–4560 cal BC (95% probability; start end Lengyel 3) (Fig. 11), probably 4750– 4635 cal BC (56% probability) or 4610–4580 cal BC (12% probability), and finally ended by 4605– 4460 cal BC (95% probability; end end Lengyel 3) (Fig. 11), probably 4575–4505 cal BC (68% probability). J. Regenye, K. Oross, E. Bánffy, E. Dunbar, R. Friedrich, A. Bayliss, N. Beavan, B. Gaydarska, and A. Whittle Fig. 6. Probability distributions of dates from inhumation burials at Veszprém-Jutasi út. Each distribu­tion represents the relative probability that an event occurred at a particular time. For each of the dates two distributions have been plotted, one in outline which is the result produced by the scientific evidence alone, and a solid one which is based on the chronological model used. The other distributions corre­spond to aspects of the model. For example, “start Veszprém Lengyel” is the estimated date when the Lengyel cemetery was initiated. The large square brackets down the left-hand side of the figure along with the OxCal keywords define the overall model exactly. There may then have been a short gap of –55 –210 years (95% probability; gap Lengyel furnished graves/Lengyel III; Fig. 10), probably of 30–155 years (68% probability)1 . The model shown in Figure 8 suggests that Lengyel III ceramics appeared in western Hungary in 4570– 4360 cal BC (95% probability; start Lengyel III; Fig. 8), probably in 4490–4390 cal BC (68% prob­ability). They were last used in 4445–4300 cal BC (95% probability; end Lengyel III; Fig. 8), probably in 4415–4335 cal BC (68% probability), having been used for a period of 1–155 years (95% proba­bility; duration Lengyel III) (Fig. 9), probably for 1–70 years (68% probability). This apparently rela­tively short phase was succeeded by Balaton-Lasinja ceramics, probably with little or no gap: 1–165 years (95% probability; gap Lengyel III/Balaton-Lasinja) (Fig. 10), probably 1–85 years (68% probability). The model suggests that Balaton-Lasinja pottery first appeared in 4400–4190 cal BC (95% probability; start Balaton-Lasinja) (Fig. 8), probably in 4360– 4255 cal BC (68% probability). It ended in 3895– 3585 cal BC (95% probability; end Balaton-La­sinja) (Fig. 8), probably in 3790–3625 cal BC (68% probability), having been used for a period of 280– 610 years (95% probability; duration Balaton-La­sinja; Fig. 9), probably for 355–540 years (68% pro­bability). This extremely long tradition was succe­eded by a related pottery style, the Furchenstich. Again, there was probably little or no gap: 1–275 years (95% probability; gap Balaton-Lasinja/Fur­chenstich) (Fig. 10), probably 1–140 years (68% probability). The Furchenstich style first appeared in 3770–3520 cal BC (92% probability; start Furchenstich; Fig. 8) or 3460–3405 cal BC (3% probability), probably Fig. 7. Probability distributions of the durations of the cemeteries at Veszprém and the interval between them, derived from the models defined in Figure 6. 1 The negative part of this range reflects the 10% probability that start Lengyel III was before end Lengyel 3, and thus that there was a period of overlap rather than a gap between these ceramic phases. It should be noted that our sample of radiocarbon dates for the Lengyel III phase (Tab. 3) is pitifully small and that these derive from only three sites. In these circumstances, the potential gap may be an artefact of our current dataset. Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary Fig. 8. Probability distributions of dates from the earlier Copper Age ceramic sequence in western Hun­gary. The format is identical to that of Figure 6. The large square brackets down the left-hand side of the figure along with the OxCal keywords define the overall model exactly. in 3670–3545 cal BC (68% probability). It ended in 3545–3125 cal BC (95% probability; end Furchen­stich) (Fig. 8), probably in 3485–3320 cal BC (68% probability). Overall Furchenstich pottery was used over a period of 1–265 years (95% probability; dura­tion Furchenstich) (Fig. 9), probably for 60–210 years (68% probability). Given the small number of radiocarbon dates currently available associated with the Furchenstich style, this ending may fall with­in the earlier part of this date range and is then com­patible with the suggestion that this ceramic style was replaced by Boleráz pottery c. 3500 cal BC (Fur­holt 2008). Figure 11 illustrates the outline chronology for the earlier Copper Age ceramics in western Hungary. These traditions span more than a thousand years, which brings into sharp relief the small number of radiocarbon dates on which this chronology is based. Nonetheless, it is striking that the Balaton-Lasinja tradition appears to have endured for much longer than the other ceramic styles. J. Regenye, K. Oross, E. Bánffy, E. Dunbar, R. Friedrich, A. Bayliss, N. Beavan, B. Gaydarska, and A. Whittle Discussion Great changes occurred across big swathes of Europe in the mid-fifth millennium cal BC. In the Carpa­thian basin, the previous system of tells and large agglomerated flat settlements came to an end. In the Vin.a orbit, this process had probably begun around 4700 cal BC, as seen for example in the se­quence of the Uivar tell in western Romania (Draso­vean et al. 2017; Drasovean, Schier 2020; Bayliss et al. 2020). The great tell of Vin.a-Belo Brdo itself was probably the latest tell in its area to be aban­doned, around 4500 cal BC (Tasi. et al. 2016; Whit­tle et al. 2016). In the Lengyel sphere of Transdanu­bia and beyond, the major settlement aggregations with substantial houses and large grave groups had also largely ended by the middle of the fifth millen­nium cal BC (Bánffy et al. 2016; Regenye et al. 2020). The largest known Lengyel aggregation, at Alsónyék-Bátaszék, had reached its peak size around 4700 cal BC, declining steadily from then until probably the 45th century cal BC (Osztás et al. 2016; Bánffy et al. 2016). In the broad, so-called Danubian distribution of the LBK and post-LBK, longhouses and longhouse settlements also largely came to an end in the mid-fifth millennium cal BC. One regional study, in Low­er Alsace in the upper Rhine valley, suggests that the last longhouses there belonged to the Rössen phase, dating probably to the 46th–45th centuries cal BC (Denaire et al. 2017). In another region, that of the Polish lowlands, the last longhouses were in use till the turn of the millennium (Czerniak et al. 2017). These worlds have traditionally been studied by se­parate research communities, and as a result, while there are plenty of hypotheses about the causes of such changes, there has been little discussion about the convergent timings. In the aftermath of these changes, a different life­style came into existence. This can be broadly charac­terised by a more dispersed settlement system, the existence of generally smaller settlements (in mod­ern terminology hamlets rather than villages), often smaller groups of graves or cemeteries, less obvious Fig. 9. Probability distributions of the durations of the earlier Copper Age ceramic traditions of western Hungary, derived from the model defined in Figure 8. Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary social differentiation, and finally an often less osten­tatious material culture (Parkinson 2006; Yerkes et al. 2009; Siklósi, Szilágyi 2021.585). There was, how­ever, considerable variability across east-central, cen­tral and western Europe, perhaps hardly surprising given the size of the area. In some regions, houses remained a visible and frequently detected feature of settlements, while in others a more elusive pat­tern of pits and other remains is met with, for exam­ple across many parts of the TRB distribution. On the Great Hungarian Plain, cemeteries became a notable feature, seemingly supplanting the graves previous intermingled in tells and flat settlements; the grave ground of Tiszapolgár-Basatanya remains the largest (Siklósi, Szilágyi 2021), perhaps serving as some kind of regional centre (Chapman 2020). There are signs of differentiation within the grave goods at that cemetery (Chapman 2020; Sofaer De-revenski 2000), and further north there has been a long debate about the character of megalithic and related constructions. While a shift to plainer, or less exuberantly decorated, pottery characterises many regional sequences, the increasing appearance of copper and gold should be noted (Siklósi, Szilágyi 2021.585), and the production and distribution of jadeitite continued in the second half of the fifth mil­lennium cal BC in western Europe (Pétrequin et al. 2012). From this deliberately broad perspective, the results presented here for Veszprém-Jutasi út and the post-Lengyel sequence in western Hungary contribute in the first place to the better establishment of more robust local and regional detail, mirroring the pro­gress towards the same end being made on the Great Hungarian Plain (Yerkes et al. 2009; Siklósi, Szilágyi 2021). It is, however, important to stress the varied quality of the dating samples currently available. The date estimates from Veszprém-Jutasi út and for the regional west Hungarian post-Lengyel sequence (Figs. 6–11) may suggest gradual change from the peak of earlier Lengyel activity (again, as seen at Alsónyék-Bátaszék), and then apparent con­siderable longevity for the Balaton-Lasinja pheno­menon; the estimate for its duration, lasting several centuries (Fig. 9), is particularly striking. Although clearly incomplete and in need of much further re­search, the settlement record suggests much small­er settlement units than in the Lengyel heyday in Transdanubia, especially in its southeastern part. In­terestingly, individual houses may not have been substantially smaller than in Lengyel times, and are archaeologically recognisable and visible. It is hard to quantify settlement density, but Balaton-Lasinja sites have been recurrent finds in both surveys and rescue excavations and appear to have existed in significant numbers. Finally, the contents of Balaton-Lasinja graves do not obviously suggest major so­cial differentiation, though the presence of gold is again to be noted. By way of contrast, the situation on the Great Hun­garian Plain was different. According to recent ana­lyses (Raczky, Siklósi 2013; Siklósi, Szilágyi 2021), the Tiszapolgár and Bodrogkeresztúr ceramic tradi­tions probably overlapped in time, rather than being successive as previously thought; there may also be regional variation within the Plain (Siklósi, Szilágyi 2021.622). Associated material, such as heavy cop­per items and gold ornaments in Bodrogkeresztúr contexts (Siklósi, Szilágyi 2021.585), should there­fore have been unevenly available or accessible across the communities of the Plain. The cemeteries themselves, though the majority are clearly much smaller than that of Tiszapolgár-Basatanya, are more prominent than those in the Balaton-Lasinja orbit, and might speak for local social differentiation, in the form say of established places for the dead of prominent social status. Date estimates currently available suggest that some of the smaller cemete­ries at least were much shorter-lived than the long span originally attributed to Tiszapolgár-Basatanya (Siklósi, Szilágyi 2021.586). In these ways poten­tially less stable than the Balaton-Lasinja phenome­non, the emergent situation on the Plain may have also been shorter-lived than the Balaton-Lasinja sphere, with the Tiszapolgár-Bodrogkeresztúr phe­nomenon ending by the turn of the millennium (Sik­lósi, Szilágyi 2021.619). A possible contrast of this kind, however, will depend on the chronology of Transdanubia being confirmed and refined in the future. The existing absolute chronological evidence rein­forces earlier assumptions about the use of the Fur-chenstich pottery style as a phenomenon succeed­ing the Balaton-Lasinja era in western Hungary. That is, moreover, at a surprisingly late period, virtually in the second third of the fourth millennium cal BC. This conclusion must be tentative, however, until further radiocarbon dates are obtained on short-lived samples associated with Furchenstich pottery from a larger number of sites. Another significant implication is the apparent partial contemporaneity of Furchenstich with the Boleráz-Baden complex, over several centuries. The emergence of the Bole-ráz-Baden complex has been placed c. 3650 cal BC and the expansion of the Boleráz style c. 3500 cal J. Regenye, K. Oross, E. Bánffy, E. Dunbar, R. Friedrich, A. Bayliss, N. Beavan, B. Gaydarska, and A. Whittle BC in the comprehensive study of Martin Furholt (2008). It must be emphasised here, however, that literally no scientific research targeting problems associated with communities using Furchenstich and related pottery styles has been carried out in Trans-danubia in the past two decades. The only exception is probably the long-lasting settlement at Balaton­õszöd-Temetõi-dûlõ where an abundance of features and finds represented the Baden complex, while the earlier Copper Age occupation was only a limited part of the prehistoric assemblage (Horváth et al. 2014a). As a consequence, reliable radiocarbon dates associated with this kind of occupation are scarce. Therefore, only a very tentative picture can be pre­sented of the development and the dynamics of early and mid-fourth millennium cal BC communities in the region. By contrast, large cemeteries of the Bole-ráz-Baden complex such as from Budakalász-Luppa­csárda (Bondár, Raczky 2009) and Pilismarót-Basa­harc (Bondár 2015) and extended sites like the afore-mentioned Balatonõszöd-Temetõi-dûlõ (Hor­váth et al. 2014b) have been published, including statistically modelled series of radiocarbon dates. Further research needs therefore to continue to pick away at these local and regional differences within the post-Lengyel and post-Vin.a worlds. In the fu­ture, we need to expand our focus to neighbouring regions as well. A lot of new data have been obtain­ed during rescue and other excavations in Croatia and Slovenia in the last decade. These have contri­buted greatly to the refinement of the Neolithic-Cop­per Age sequence in this region (e.g., Balen, Drni. 2014; .ataj 2014; 2016; 2020; Kramberger 2014; 2020; McClure et al. 2020). Finally, it is worth considering whether the emerg­ing conditions of the later fifth millennium cal BC, seen here through the particular lens of Veszprém-Jutasi út, contribute to a better understanding of the circumstances in which the previous system of ma­jor settlement aggregation came to an end. That pro­cess has of course been much debated (among a host of others, see for example, Tringham, Krsti. 1990; Chapman 2000; 2020; Tripkovi. 2010; Por-.i. 2011; Crnobrnja 2011; Müller et al. 2013; Bo­ri. 2015; Bánffy et al. 2016; Whittle et al. 2016), and many factors have been suggested. This is not the place to re-air all the issues, but it is an opportu­nity to reflect briefly on how what came after may enable further insight into some of the processes perhaps at work; a small-scale situation like Vesz­prém-Jutasi út may be valuable in this quest. Clearly, it is unlikely that the system of major aggregations just petered out, though it is the case that a gradual decline in numbers at Alsónyék-Bátaszék has been carefully documented (Osztás et al. 2016; Bánffy et al. 2016). It is much more likely that there was something unstable, with high social costs, making the maintenance of prominent places and large ag­gregations unsustainable in the long run. A favou-rite idea has been the rise of the individual house­hold at odds with the wider community (for exam­ple, Tringham, Krsti. 1990; Bori. 2015), though the overt evidence for difference is generally hard to find, not helped by the lack of cemeteries in the Vin.a culture or by the apparent paucity of signs of violence on the more than 2000 skeletons from Len-gyel Alsónyék-Bátaszék (Osztás et al. 2016; Bánffy et al. 2016); successive burnings at the top of Vin.a-Belo Brdo (Tasi. et al. 2015), or through the se­quence at Uivar (Drasovean, Schier 2020), however, should be remembered. Perhaps situations like Vesz­prém-Jutasi út in Transdanubia help to support the idea of the importance of the single (if not neces­sarily autonomous) household, showing more indi­vidual social units in smaller and potentially simpler social settings. The apparent durability of the Bala­ton-Lasinja phenomenon may also indicate the suc­cess of this resolution of putative former social ten­sions. The situation on the Great Hungarian Plain, however, as described briefly above, may suggest that competition and local or regional rivalries had not entirely disappeared. ACKNOWLEDGEMENTS Dating of the graves at Veszprém-Jutasi út was car­ried out within the project The Times of Their Lives, funded by an Advanced Investigator Grant from the European Research Council (2012–2017; 295412), and led by Alasdair Whittle and Alex Bayliss. Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary References Bakay K., Kalicz N., and Sági K. 1966. Magyarország Ré­gészeti Topográfiája 1. Veszprém megye régészeti topo­gráfiája. A keszthelyi és tapolcai járás. Akadémiai Kia­dó. Budapest. Balen J., Drni. I. 2014. Archaeological excavations at Bar-barsko – A new contribution to understanding of the Mid­dle Copper Age in northern Croatia. Vjesnik Arheolo.kog muzeja u Zagrebu 47: 39–76. Bánffy E. 1994. Transdanubia and Eastern Hungary in the Early Copper Age. A Nyíregyházi Jósa András Múzeum Évkönyve 36: 291–296. 1995. South-west Transdanubia as a mediating area. On the cultural history of the Early and Middle Chalcoli­thic. Antaeus 22: 157–196. Bánffy E., Egry I. forthcoming. Feasting with music? A mu­sical instrument and its context from the later fifth mil­lennium BC Hungary. Germania 99. 2022. Bánffy E., Osztás A., Oross K., + 6 authors, and Whittle A. 2016. The Alsónyék story: towards the history of a persis­tent place. Bericht der Römisch-Germanischen Kommis­sion 94: 283–318. Banner J., Bognár-Kutzián I. 1961. Beiträge zur Chrono­logie der Kupferzeit des Karpatenbeckens. Acta Archaeo­logica Academiae Scientiarum Hungaricae 13: 1–32. Barna J. P., Serlegi G., Fullár, Z., and Bánffy E. 2019. A cir­cular enclosure and settlement from the mid-fifth millen­nium BC at Balatonmagyaród-Hídvégpuszta. In E. Bánffy, J. P. Barna (eds.), Trans Lacum Pelsonem. Prähistorische Forschungen in Südwestungarn (5500–500 v. Chr.) – Prehistoric Research in South-Western Hungary (5500– 500 BC). Castellum Pannonicum Pelsonense 7. Verlag Ma­rie Leidorf Publisher. Rahden: 117–160. Bayliss A., Beavan N., Hamilton D., + 13 authors, and Whittle A. 2016. Peopling the past: creating a site biog­raphy in the Hungarian Neolithic. Bericht der Römisch-Germanischen Kommission 94: 23–91. Bayliss A., Gaydarska B., Whittle A., Drasovean F., and Schier W. 2020. Scientific dating and chronological mod-elling. In F. Drasovean, W. Schier (eds.), Uivar “Gomila”: a prehistoric tell settlement in the Romanian Banat. Vol. I: Site, Architecture, Stratigraphy and Dating. Verlag Marie Leidorf Publisher. Rahden: 491–548. Beavan Athfield N., Green R. C., Craig J., McFadgen B., and Bickler S. 2008. Influence of marine sources on 14C ages: isotopic data from Watom Island, Papua New Gui­nea inhumations and pig teeth in light of new dietary standards. Journal of the Royal Society of New Zealand 38: 1–23. Bogaard A., Fraser R., Heaton T. H. E., + 16 authors, and Stephan E. 2013. Crop manuring and intensive land ma­nagement by Europe’s first farmers. Proceedings of the National Academy of Sciences 110: 12589–12594. https://doi.org/10.1073/pnas.1305918110 Bóna I. 1963–4 [1965]. The peoples of southern origin of the Early Bronze Age in Hungary I-II. Alba Regia 4–5: 17–63. Bondár M. 2015. The Copper Age cemetery at Pilisma-rót-Basaharc. István Torma’s excavations (1967, 1969– 1972). Institute of Archaeology. Research Centre for the Humanities. Hungarian Academy of Sciences. Budapest. Bondár M., Raczky P. (eds.) 2009. The Copper Age ceme­tery of Budakalász. Pytheas. Budapest. Bori. D. 2015. The end of the Vin.a world: modelling the Late Neolithic to Copper Age transition and the notion of archaeological culture. In S. Hansen, P. Raczky, A. An­ders, and A. Reingruber (eds.), Neolithic and Copper Age between the Carpathians and the Aegean Sea: chrono­logies and technologies from the 6th to the 4th millen­nium BCE. International Workshop Budapest 2012. Ar-chäologie in Eurasien 31. Deutsches Archäologisches In-stitut. Eurasien-Abteilung. Habelt. Bonn: 157–217. Bori. D., Grupe G., Peters J., and Miki. .. 2004. Is the Me-solithic-Neolithic subsistence dichotomy real? New stable isotope evidence from the Danube Gorges. European Journal of Archaeology 7: 221–248. https://doi.org/10.1177/1461957104056500 Bronk Ramsey C. 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51: 37–60. https://doi.org/10.1017/S0033822200033865 Brown T. A., Nelson D. E., Vogel J. S., and Southon J. R. 1988. Improved collagen extraction by modified Longin method. Radiocarbon 30: 171–177. https://doi.org/10.1017/S0033822200044118 .ataj L. 2014. Middle Eneolithic Lasinja and Retz-Gajary cultures in northern Croatia – development of chronol­ogy. In W. Schier, F. Drasovean (eds.), The Neolithic and Eneolithic in southeast Europe: new approaches to dat­ing and cultural dynamics in the 6th to 4th millennium BC. Prähistorische Archäologie in Südosteuropa 28. Ver­lag Marie Leidorf GmbH. Rahden/Westf.: 397–408. J. Regenye, K. Oross, E. Bánffy, E. Dunbar, R. Friedrich, A. Bayliss, N. Beavan, B. Gaydarska, and A. Whittle 2016. Lasinja, Retz-Gajary and Boleraz? Radiocarbon dates and the sequence of Copper Age cultures in cen­tral Croatia. In J. Kovárik (ed.), Centenary of Jaroslav Palliardi’s Neolithic and Aeneolithic relative chro­nology (1914–2014). International symposium Cen­tenary of Jaroslav Palliardi’s Neolithic and Aeneoli­thic relative chronology (2014; Moravské Budìjovi­ce). University of Hradec Králové. Philosophical Faculty. Hradec Králové: 181–192. 2020. Mali Komor-Vrci: the beginning of the Eneolithic in Zagorje. In M. Dizdar, K. Boti., and H. Kalafati. (eds.), Homo universalis: zbornik radova posve.en Zorku Markovi.u povodom 65. obljetnice .ivota. Zbornik In-stituta za arheologiju 15. Institut za arheologiju. Zag­reb: 173–186. Chapman J. 2000. Fragmentation in archaeology: peo­ple, places and broken objects in the prehistory of south eastern Europe. Routledge. London. 2020. Forging identities in the prehistory of Old Eu­rope: dividuals, individuals and communities, 7000– 3000 BC. Sidestone Press. Leiden. Cook G. T., Bonsall C., Hedges R. E. M., McSweeney K., Boroneant V., and Pettitt P. B. 2001. A freshwater diet-de­rived 14C reservoir effect at the stone age sites in the Iron Gates gorge. Radiocarbon 43: 453–460. https://doi.org/10.1017/S0033822200038327 Crnobrnja A. 2011. Arrangement of Vin.a culture figuri­nes: a study of social structure and organization. Docu­menta Praehistorica 38: 131–147. https://doi.org/10.4312/dp.38.11 Czerniak L., Marciniak A., Bronk Ramsey C., + 4 authors, and Whittle A. 2017. House time: Neolithic settlement de­velopment at Racot during the fifth millennium cal BC in the Polish lowlands. Journal of Field Archaeology 41: 618–640. https://doi.org/10.1080/00934690.2016.1215723 Denaire A., Lefranc P., Wahl J., + 6 authors, and Whittle A. 2017. The cultural project: formal chronological model-ling of the Early and Middle Neolithic sequence in Lower Alsace. Journal of Archaeological Method and Theory 24: 1072–1149. https://doi.org/10.1007/s10816-016-9307-x Dimitrijevi. S. 1961. Problem neolita i eneolita u sjevero­zapadnoj Jugoslaviji (Problem des Neolithikums und Ae­neolithikums in Nordwestjugoslawien). Opuscula Archa­eologica 5: 5–85. Drasovean F., Schier W. (eds.) 2020. Uivar “Gomila”: a prehistoric tell settlement in the Romanian Banat. Vol. I: Site, Architecture, Stratigraphy and Dating. Verlag Ma­rie Leidorf Publisher. Rahden. Drasovean F., Schier W., Bayliss A., Gaydarska B., and Whittle A. 2017. The lives of houses: duration, context and history at Neolithic Uivar. European Journal of Ar­chaeology 20(4): 636–662. https://doi.org/10.1017/eaa.2017.37 Dunbar E., Cook G. T., Naysmith P., Tripney B. G., and Xu S. 2016. AMS 14C dating at the Scottish Universities En­vironmental Research Centre (SUERC) radiocarbon dating laboratory. Radiocarbon 58: 9–23. https://doi.org/10.1017/RDC.2015.2 Éri I., Kelemen M., Németh P., and Torma I. 1969. Magyar-ország Régészeti Topográfiája 2. Veszprém megye régé­szeti topográfiája. A veszprémi járás. Akadémiai Kiadó. Budapest. Fernandes R., Millard A. R., Brabec M., Nadeau M.-J., and Grootes P. 2014. Food reconstruction using isotopic trans­ferred signals (FRUITS): a Bayesian model for diet recon­struction. PLoS ONE 9(2): e87436. http://doi:10.1371/journal.pone.0087436 Figler A., Bartosiewicz L., Füleky Gy,. and Hertelend E. 1997. Copper Age settlement and the Danube water sys­tem: a case study from north western Hungary. In J. Chap­man, P. Dolukhanov (eds.), Landscape in flux: central and eastern Europe in Antiquity. Oxbow Books. Oxford: 209–230. Furholt M. 2008. Pottery, cultures, people? The European Baden material re-examined. Antiquity 82: 617–628. https://doi.org/10.1017/S0003598X0009726X Gamarra B., Howcroft R., McCall A., + 11 authors, and Pinhasi R. 2018. 5000 years of dietary variations of pre­historic farmers in the Great Hungarian Plain. PLoS ONE 13(5): https://doi.org/10.1371/journal.pone.0197214 Horváth L. A., Simon H. K. 2003. Das Neolithikum und die Kupferzeit in Südwesttransdanubien. Inventaria Praehistorica Hungariae 9. Magyar Nemzeti Múzeum. Bu­dapest. Horváth T., Svingor É. S., and Molnár, M. 2008. New ra­diocarbon dates for the Baden culture. Radiocarbon 50: 447–458. https://doi.org/10.1017/S0033822200053546 Horváth T., Svingor É. S., and Sipos G. 2014a. Absolute and relative chronology. In T. Horváth (ed.), The prehi­storic settlement at Balatonõszöd-Temetõi-dûlõ, The Middle Copper Age, Late Copper Age and Early Bronze Age occupation. Archaeolingua. Budapest: 39–42. 2014b. Absolute and relative chronology. In T. Horváth (ed.), The prehistoric settlement at Balatonõszöd­Temetõi-dûlõ, The Middle Copper Age, Late Copper Age Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary and Early Bronze Age Occupation. Archaeolingua. Budapest: 567–604. Ilon G. 2004. Szombathely õskori településtörténetének vázlata (Outline of the pre-historic settlement of Szom­bathely). Õskorunk 2. Vas Megyei Múzeumok Igazgató­sága. Szombathely. Jakucs J., Bánffy E., Oross K., + 7 authors, and Whittle A. 2016. Between the Vin.a and Linearbandkeramik worlds: the diversity of practices and identities in the 54th–53rd centuries cal BC in south-west Hungary and beyond. Jour­nal of World Prehistory 29: 267–336. https://doi.org/10.1007/s10963-016-9096-x Kalicz N. 1958. Rézkori sztratigráfia Székely község ha-tárában. Archaeologiai Értesítõ 58: 3–6. 1969. A rézkori balatoni csoport Veszprém megyében. (Die Kupferzeitliche Balaton-Gruppe im Komitat Vesz­prém). Veszprém Megyei Múzeumok Közleményei 8: 83–90. 1973. Über die chronologische Stellung der Balaton-Gruppe in Ungarn. In B. Chropovský (ed.), Symposium über die Entstehung und Chronologie der Badener Kultur. Verlag der Slowakischen Akademie der Wissen­schaften. Bratislava: 131–165. 1982. A Balaton-Lasinja kultúra történeti kérdései és fémleletei. Archaeologiai Értesítõ 109: 3–17. 1991. Beiträge zur Kenntnis der Kupferzeit im ungari­schen Transdanubien. In J. Lichardus (ed.), Kupferzeit als historische Epoche. Symposium Saarbrücken und Otzenhausen, 6.11.–13.11.1988. Saarbrücker Beiträge zur Altertumskunde. Habelt. Bonn: 347–387. 1995a. Die Balaton-Lasinja-Kultur in der Kupferzeit Süd­ost- und Mitteleuropas. In Neuere Daten zur Siedlungs­geschichte und Chronologie der Kupferzeit des Karpa­tenbeckens. Inventaria Praehistorica Hungariae 7: 37–49. 1995b. Letenye-Szentkeresztdomb: ein Siedlungsplatz der Balaton-Lasinja-Kultur. In Neuere Daten zur Sied­lungsgeschichte und Chronologie der Kupferzeit des Karpatenbeckens. Inventaria Praehistorica Hungariae 7: 61–106. 2001. Die Protoboleráz-Phase an der Grenze von zwei Epochen. In P. Roman, S. Diamandi (eds.), Cernavoda¢ III–Boleráz. Ein Vorgeschichtliches Phänomen zwi­schen dem Oberrhein und der Unteren Donau. Man-galia/Naptun 18.–24. Oktober 1999: 385–436. Institu­tul Român de Tracologie. Bucuresti. 2003. Az újkõkorvégi és a rézkori megtelepedés ma-radványai a nagykanizsai Inkey-kápolna mellett (Kr. e. 5. évezred elsõ harmadától a 3. évezred elsõ feléig). Endneolithische und kupferzeitliche Besiedlung bei Nagykanizsa (Inkey-Kapelle). Zalai Múzeum 12: 7–47. Köhler K. 2006. A lengyeli és a Balaton-Lasinja kultúra embertani leletei Veszprémbõl (Anthropological finds of the Lengyel and Balaton-Lasinja culture from Veszprém). Veszprém Megyei Múzeumok Közleményei 24: 37–48. Korek J. 1960. Die Goldscheiben von Csáford. Folia Ar-cheologica 12: 27–33. Koro.ec J. 1958. Eine neue Kulturgruppe des späten Neo­lithikums in Nordwestjugoslawien. Acta Archaeologica Academiae Scientiarum Hungaricae 9: 83–93. Kramberger B. 2014. The Neolithic-Eneolithic sequence and pottery assemblages in the fifth millennium BC in north-eastern Slovenia. Documenta Praehistorica 41: 237–282. https://doi.org/10.4312/dp.41.13 Kramberger B. 2020. Zur relativen und absoluten Chro­nologie des späten Neolithikum und frühen Äneolithi­kums im kontinentalen Teil Sloweniens. In Ch. Gutjahr, G. Tiefengraber (eds.), Beiträge zur Kupferzeit am Ran-de der Südostalpen. Akten des 4. Wildoner Fachgesprächs am 16. und 17. Juni 2016 in Wildon/Steiermark (Öster­reich). Verlag Marie Leidorf. Rahden: 53–89. Kromer B., Lindauer S., Synal H.-A., and Wacker L. 2013. MAMS – A new AMS facility at the Curt-Engelhorn-Centre for Archaeometry, Mannheim, Germany. Nuclear Instru­ments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 294: 11–13. https://doi.org/10.1016/j.nimb.2012.01.015 Kutzián I. 1955. Die Ausgrabungen in Tiszapolgár-Basa­tanya. In K. Fejér (ed.), Conference archéologique de l’Academie Hongroise des Sciences. Magyar Nemzeti Mú­zeum. Budapest: 69–87. Makkay J. 1976. Problems concerning Copper Age chro­nology in the Carpathian Basin. Acta Archaeologica Aca­demiae Scientiarum Hungaricae 28: 250–300. McClure S. B., Zavodny E., Novak M., + 3 authors, and Kennett D. J. 2020. Paleodiet and health in a mass burial population: the stable carbon and nitrogen isotopes from Poto.ani, a 6,200-year-old massacre site in Croatia. Inter­national Journal of Osteoarchaeology 30(4): 507–518. https://doi.org/10.1002/oa.2878 Müller J., Rassmann K., and Hofmann R. (eds.) 2013. Oko­li.te 1 – Untersuchungen einer spätneolithischen Sied­lungskammer in Zentralbosnien. Universitätsforschun­ J. Regenye, K. Oross, E. Bánffy, E. Dunbar, R. Friedrich, A. Bayliss, N. Beavan, B. Gaydarska, and A. Whittle gen zur prähistorischen Archäologie 228. In Kommission bei Dr. Rudolf Habelt GmbH. Bonn Nehlich O., Bori. D., Stefanovi. S., and Richards M. P. 2010. Sulphur isotope evidence for freshwater fish con­sumption: a case study from the Danube Gorges, SE Eu­rope. Journal of Archaeological Science 37: 1131– 1139. https://doi.org/10.1016/j.jas.2009.12.013 O’Connell T. C., Kneale C. J., Tasevska N., and Kuhnle G. G. C. 2012. The diet-body offset in human nitrogen isoto­pic values: a controlled dietary study. American Journal of Physical Anthropology 149: 426–434. https://doi.org/10.1002/ajpa.22140 Ogrinc N., Budja M. 2005. Paleodietary reconstruction of a Neolithic population in Slovenia: a stable isotope app­roach. Chemical Geology 218: 103–116. https://doi.org/10.1016/j.chemgeo.2005.01.023 Oross K., Marton T., Whittle A., Hedges R. E. M., and Cramp L. J. E. 2010. Die Siedlung der Balaton-Lasinja-Kul­tur in Balatonszárszó-Kis-erdei-dûlõ. In J. .uteková, P. Pa-vúk, P. Kalábková, and B. Kovár (eds.), Panta Rhei. Stu­dies on the chronology and cultural development of south-eastern end central Europe in earlier prehistory, presented to Juraj Pavúk on the occasion of his 75th birthday. Comenius University in Bratislava. Archaeologi­cal Centre. Bratislava, Olomouc: 379–405. Osztás A., Zalai-Gaál I., Bánffy E., + 11 authors, and Whit­tle A. 2016. Coalescent community at Alsónyék: the tim­ings and duration of Lengyel burials and settlement. Be-richt der Römisch-Germanischen Kommission 94: 179– 282. Parkinson W. A. 2006. The social organisation of Early Copper Age tribes on the Great Hungarian Plain. British Archaeological Reports. Oxford. Pétrequin P., Cassen S., Errera M., Klassen L., Sheridan A., and Pétrequin A.-M. (eds.), 2012. Jade. Grandes haches alpines du Néolithique européen. Ve et IVe millénaires av. J.-C. Presses Universitaires de Franche-Comté. Besan­çon; Centre de Recherche Archéologique de la Vallée de l’Ain. Gray. Por.i. M. 2011. An exercise in archaeological demogra­phy: estimating the population of Late Neolithic settle­ments in the Central Balkans. Documenta Praehistorica 38: 323–332. https://doi.org/10.4312/dp.38.25 Quitta H., Kohl G. 1969. Neue Radiocarbondaten zum Neo­lithikum und zur frühen Bronzezeit Südosteuropas und der Sowjetunion. Zeitschrift für Archäologie 3: 223–255. Raczky P. 1974. A lengyeli kultúra legkésõbbi szakaszának leletei a Dunántúlon (Funde der spätesten Phase der Len- gyel-Kultur in Westungarn). Archaeológiai Értesítõ 101: 185–210. Raczky P., Siklósi Zs. 2013. Reconsideration of the Copper Age chronology of the eastern Carpathian Basin: a Baye­sian approach. Antiquity 87: 555–573. https://doi.org/10.1017/S0003598X00049127 Regenye J. 2004. Háztípusok és településszerkezet a ké-sõi lengyeli kultúrában veszprémi és szentgáli példák alapján (House types and settlement structure in the Late Lengyel culture, based on the examples at Veszprém and Szentgál). Veszprém Megyei Múzeumok Közleményei 23: 25–47. 2006. Temetkezések Veszprém, Jutasi u. lelõhelyen (Lengyeli kultúra, Balaton-Lasinja kultúra) (Burials at the site of Veszprém, Jutasi Street). Veszprém Megyei Múzeumok Közleményei 24: 7–35. 2007. The Late Lengyel culture in Hungary as reflected by the excavation at Veszprém. In J. K. Koz³owski, P. Raczky (eds.), The Lengyel, Polgár and related cultu­res in the Middle/Late Neolithic in central Europe. Po­lish Academy of Arts and Sciences, and Eötvös Loránd University. Institute of Archaeological Sciences. Kra­ków, Budapest: 381–396. 2011. Kõ és agyag. Település és életmód az újkõkor­rézkor fordulóján a Dunántúlon Stone and clay. Set­tlement and way of life at the Neolithic/Copper Age transition in Transdanubia. Veszprém Megyei Múzeu-mi Igazgatóság. Veszprém. Regenye J., Bánffy E., Demján P., + 8 authors, and Whittle A. 2020. Narratives for Lengyel funerary practice. Bericht der Römisch-Germanischen Kommission 97: 5–80. Regenye J., Biró K. T. 2014. Veszprém, Jutasi út neoli­tikus település leletanyaga I. Kerámia, kõ (Finds from the Neolithic settlement Veszprém, Jutasi street I. Cera­mics and lithics). Laczkó Dezsõ Múzeum Közleményei 28: 29–73. 2019. Veszprém, Jutasi út neolitikus település leletanya­ga II. Kerámia, kõ (Finds from the Neolithic settlement Veszprém, Jutasi street II. Ceramics and lithics). Lacz­kó Dezsõ Múzeum Közleményei 29: 7–70. Reimer P. J., Austin W. E. N., Bard E. + 37 authors, and Talamo S. 2020. The IntCal20 Northern Hemispheric ra­diocarbon calibration curve (0–55 kcal BP). Radiocarbon 62: 725–757. https://doi.org/10.1017/RDC.2020.41 Sayle K. L., Cook G. T., Ascough P. L., Gestsdóttir H., Ha­milton W. D., and McGovern T. H. 2014. Utilization of .13C, .15N, and .34S analyses to understand 14C dating anomalies within a Late Viking Age community in North­ Some Balaton-Lasinja graves from Veszprém-Jutasi út and an outline chronology for the earlier Copper Age in western Hungary east Iceland. Radiocarbon 56: 811–821. https://doi.org/10.2458/56.17770 Siklósi Zs., Szilágyi M. 2021. Culture, period or style? Re­consideration of Early and Middle Copper Age chronolo­gy of the Great Hungarian Plain. Radiocarbon 63: 585– 646. https://doi.org/10.1017/RDC.2020.115 Sofaer Derevenski J. 2000. Rings of life: the role of early metalwork in mediating the gendered life course. World Archaeology 31: 389–406. https://doi.org/10.1080/00438240009696928 Stuiver M., Polach H. A. 1977. Reporting of 14C data. Ra­diocarbon 19: 355–363. https://doi.org/10.1017/S0033822200003672 Tasi. N., Mari., M., Filipovi. D., + 6 authors, and Whittle A. 2016. Interwoven strands for refining the chronology of the Neolithic tell of Vin.a-Belo Brdo, Serbia. Radiocar­bon 58: 795–831. https://doi.org/10.1017/RDC.2016.56 Tasi. N., Mari. M., Penezi. K., + 8 authors, and Whittle A. 2015. The end of the affair: formal chronological model-ling for the top of the Neolithic tell of Vin.a-Belo Brdo. Antiquity 89: 1064–1082. https://doi.org/10.15184/aqy.2015.101 Tringham R., Krsti. D. 1990. Selevac and the transforma­tion of southeast European prehistory. In R. Tringham, D. Krsti. (eds.), Selevac: a Neolithic village in Yugoslavia. University of California Press. Los Angeles: 567–616. Tripkovi. B. 2010. House(hold) continuities in the central Balkans, 5300–4600 BC. Opuscula Archaeologica 33: 7–28. https://hrcak.srce.hr/60615 Virág Zs. M. 1995. Die Hochkupferzeit in der Umgebung von Budapest und in NO-Transdanubien (Das Ludanice-Problem). Acta Archaeologica Academiae Scientiarum Hungaricae 47: 61–94. 2003. Settlement historical research in Transdanubia back to contents in the first half of the Middle Copper Age. In E. Jerem, P. Raczky (eds.), Morgenrot der Kulturen. Frühe Etap-pen der Menschheitsgeschichte in Mittel- und Südost­europa. Festschrift für Nándor Kalicz zum 75. Geburts-tag. Archaeolingua. Budapest: 375–400. 2005[2007]. Középsõ rézkori épületek rekonstrukciós lehetõségeirõl (Reconstruction of a Middle Copper Age house). Õsrégészeti Levelek – Prehistoric Newsletter 7: 60–72. Virág Zs. M., Figler A. 2007. Data on the settlement his­tory of the Late Lengyel period of Transdanubia on the basis of two sites from the Kisalföld (Small Hungarian Plain). A preliminary evaluation of the sites Gyõr-Szabad­rétdomb and Mosonszentmiklós-Pálmajor. In J. K. Koz-³owski, P. Raczky (eds.), The Lengyel, Polgár and relat­ed cultures in the Middle/Late Neolithic in Central Eu­rope. The Polish Academy of Arts and Sciences and Eötvös Loránd University. Institute of Archaeological Sciences. Kraków and Budapest: 345–364. Ward G. K., Wilson S. R. 1978. Procedures for comparing and combining radiocarbon age determinations: a criti­que. Archaeometry 20: 19–31. https://doi.org/10.1111/j.1475-4754.1978.tb00208.x Whittle A., Bayliss A., Barclay A., + 9 authors, and Vander Linden M. 2016. A Vin.a potscape: formal chronological models for Neolithic cultural development in southeast Europe. Documenta Praehistorica 43: 1–60. https://doi.org/10.4312/dp.43.1 Wosinsky M. 1889. Lengyeli ásatások 1888-ban (Die Aus­grabungen in Lengyel im Jahre 1888). Archaeologiai Ér­esítõ 9: 331–335. 1891. Das Prähistorische Schanzwerk von Lengyel. Seine Erbauer und Bewohner I–III. F. Kilian. Budapest. Yerkes R. W., Gyucha A., and Parkinson W. 2009. A multi-scalar approach to modeling the end of the Neolithic on the Great Hungarian Plain using calibrated radiocarbon dates. Radiocarbon 51: 1071–1109. https://doi.org/10.1017/S0033822200034123 Documenta Praehistorica XLIX (2022) Prehistoric stone disks from entrances and cemeteries of north-eastern Adriatic hillforts Federico Bernardini1,2, Giacomo Vinci3,2, Vanja Macovaz4, Andrea Baucon5, Angelo De Min3, Stefano Furlani3, and Sne/ana Smolic´ 6 federico.bernardini@unive.it 1 Department of Humanities, Ca’ Foscari University of Venice, Venice, IT 2 Multidisciplinary Laboratory, The Abdus Salam International Centre for Theoretical Physics, Trieste, IT< fbernard@ictp.it< giacomo8vinci@gmail.com 3 Department of Mathematics and Geosciences, University of Trieste, Trieste, IT< giacomo8vinci@gmail.com< demin@units.it< sfurlani@units.it 4 Department of History, Archaeology, Geography, Fine and Performing Arts (SAGAS), University of Firenze, Firenze, IT< macovazvanja@gmail.com 5 Geology Office, Naturtejo UNESCO Global Geopark, Idanha-a-Nova, PT< andrea@tracemaker.com 6 Public Institution Brijuni National Park, HR< snezana.smolic@np-brijuni.hr ABSTRACT – The paper presents a group of four, approximately 0.5m large, stone disks from entran­ces or cemeteries of two protohistoric hillforts of north-eastern Adriatic. The disks, having a sparse chronology with the exception of one dated to the Middle Bronze Age, show flat and plain surfaces or covered with sub-circular depressions. One disk shows two larger cup-marks at the centre of both faces. They are interpreted as ritual artefacts based on the association with sacred settlement loca­tions and comparisons with similar coeval stones found mainly close to citadel entrances, burials and thresholds in the Aegean area and Anatolia. KEY WORDS – north-eastern Adriatic; hillforts; stone disks; cup-marks; SfM photogrammetry; Bronze Age religion Prazgodovinski kamniti diski z vhodov in grobi[; na gradi[;ih na severovzhodnem Jadranu IZVLE.EK – V prispevku predstavljamo .tiri pribli.no 0,5 m velike kamnite diske z vhodov oziroma grobi.. na dveh prazgodovinskih gradi..ih na severovzhodnem Jadranu. Diski so, razen enega, ki sodi v srednjo bronasto dobo, slabo datirani. Vsi so plo..ati z gladkimi povr.inami ali prekriti s pol-kro.nimi vdolbinami. Na enem sta v sredini plo..atih delov kroglasti vdolbini. Interpretirani so kot obredni predmeti, povezani s svetim mestom naselbine. Primerjamo jih s so.asnimi kamni, najdeni-mi blizu vhodov v citadele, grobove in prehode v Egejskem prostoru in v Anatoliji. KLJU.NE BESEDE – severovzhodni Jadran; gradi..a; kamniti diski; kroglaste vdolbine; SfM fotogra­metrija; bronastodobna religija DOI> 10.4312\dp.49.7 Prehistoric stone disks from entrances and cemeteries of north-eastern Adriatic hillforts Introduction The Karst plateau and Istrian peninsula at the north­eastern shore of the Adriatic Sea (Fig. 1) are marked by the presence of hundreds of protohistoric settle­ments, generally located on hilltops. These sites, protected by dry-stone walls, locally called castellie­ri, gradine or gradi..a, featured clear originality and cultural unity in pottery production, architec­tural models, defensive systems and funerary prac­tices. They were settled for a very long time, span­ning from the late Early Bronze Age (EBA), approxi­mately between 1800 and 1650 BC, to the late Iron Age (IA; Mihovili. 2013; Borgna et al. 2018). The formation and rising of castellieri chronologically corresponds to the EBA II in the Italian relative chro­nological system (Cardarelli 2009) and to the BZ A2 in the Central-Europe Reinecke’s system (Hänsel 2009). Their origin is still debated. The complex fortified entrances, the inner-space subdivision and settlement layout of some of them, e.g., Monkodo­nja/Moncodogno (hereafter Monkodonja) hillfort in Istria, together with some peculiar ceramic and me­tal artefacts, such as pottery tripods and bronze knives with two rivet holes, suggest that the first Istrian settlements had direct contacts with the east­ern Mediterranean (Hänsel et al. 2015). However, other archaeological artefacts, such as the so-called enigmatic tablets, defined also as Brotlaibidole, and some types of pottery vessels, show that Istria was also connected to the Pannonian-Carpathian area (Hänsel et al. 2015; Borgna et al. 2018). Other pot­tery materials suggest connections with northern Italy and Apulia, too (Hänsel et al. 2015). Considering the funerary practices, small cemeteries close to or within the hillfort fortifications and bur­ial mounds are the main funerary contexts at least from the EBA to the later Bronze Age phases. In more detail, single or small groups of burials have been identified next to the gates of settlements and/ or within the ramparts. Recent radiocarbon dates obtained from two individuals buried in a small mo­numental cemetery found next to the main forti­fied entrance at Vr.in/Mt. Orcino (hereafter Vr.in) in southern Istria (the so-called sepolcreto gentili­zio, consisting of about 20 cist tombs) point to a time span approximately between about 1700 and 1200 BC (Battaglia 1958; Cupito et al. 2018). Cist tombs included within stone platforms very similar to those from Vr.in are known from the Gradina hillfort on Veliki Brijun/Brioni Maggiore island (hereafter Veli­ki Brijun; Vitasovi. 2002; 2005; Bur.i.-Matija.i., .eri. 2013). At the Monkodonja hillfort, two cist tombs of similar chronology have been identified next to the western entrance (Hänsel et al. 2015). The radiocarbon dating of some of the human re­mains from Monkodonja covers approximately the period between 1900 and 1600 BC (Hänsel et al. 2015). A similar cist tomb has been discovered at the Gradac-Turan hillfort located along the eastern Istrian coast (Mihovili. 1997; Bur.i.-Matija.i., .e­ri. 2013). Burial mounds have been found in the Karst and Istria in the surroundings of several hillforts, such as the tumuli of Mu.ego near the settlement of Monko­donja (Mihovili. et al. 2012), those close to the Vr-.in hillfort (Battaglia 1958) or the one excavated close to Barbariga (Codacci-Terlevi. 2012). General­ly, as for burial mounds excavated in other part of Istria, they feature one or more depositions within stone cists covered by earth and stone caps. In this contribution, using structure from motion (SfM) photogrammetric techniques, we present a group of four stone disks associated with entrances or Bronze Age cemeteries of two hillforts, Rupinpic­colo and Gradina, on Veliki Brijun, located in the Tri­este Karst (Italy) and in the Brijuni/Brioni islands fac­ing the southern Istria (Croatia), respectively (Fig. 1). Rupinpiccolo/Repni. hillfort The Rupinpiccolo/Repni. hillfort is located in the southern side of the central Karst ridge, which deve­lops through the Karst plateau with a Dinaric orien­tation and approximately marks the border between Italy and Slovenia (Fig. 1). The area belongs to the Cenomanian to Turonian Repen Formation consist­ing of bedded and massive, partly re-crystallized, limestone containing chert and with displaced, lo­cally broken and rounded rudist shells (Jurkov.ek et al. 2016). The ruins of its massive defensive structures were already noticed at the beginning of last century (Marchesetti 1903). The first archaeological inves­tigations were carried out several decades later in 1965, 1970–1974, 1986 and 1988 (Cannarella 1970; 1975; Maselli Scotti 1988), making it possible to un­cover the entire surviving fortification system. The site, built along the slope of a modest hill, shows a sub-rectangular plan but its western side has been destroyed by a modern quarry. It was defended by a massive rampart, up to 7m tall, built with two main external stone alignments with the intervening space filled with smaller stones and partially supported Federico Bernardini, Giacomo Vinci, Vanja Macovaz, Andrea Baucon, Angelo De Min, Stefano Furlani, and Sne/ana Smolic´ Fig. 1. The Rupinpiccolo and Gradina on Veliki Brijun hillforts. a Position of Rupinpiccolo and Gradina on Veliki Brijun hillforts and other main sites mentioned in the text. b Plan of the Bronze Age cemetery (t1-t.2, tombs 1-2; i.t., infant tomb) of Gradina on Veliki Brijun hillfort with the Gradina disk 2 in light red. c Drone-derived orthophoto of the Rupinpiccolo hillfort with the position of the stone disks (1–2), an­cient quarrying sites (3–4) and a block of the rampart with a wedge hole (5). d-e south-eastern and north­eastern gates of Rupinpiccolo, respectively. by lower stone reinforcements. Two entrances have been discovered approximately at the north-eastern and south-eastern corners. The first one is located on the hilltop and is about 2m tall, while the south­ern one is larger, about 3.2 m, externally delimited by a corridor-like structure and walled up in ancient times. Remains of four inner transversal terraces, probably built to support the dwellings, have been identified. According to available data, the stones used to build the fortification, including massive blocks up to about 3m wide, were taken from the local limestone out­crop which is broken in many parts, just outside the rampart. In addition, chisel marks and a wedge hole were identified a few meters east of the north­eastern corner of the fortification (Priuli 1977). The use of such huge blocks is not reported in other sites of the Trieste Karst, and makes Rupinpiccolo a uni­que context. Based on the typology of ceramic finds, Rupinpicco-lo was firstly dated to the IA (Cannarella 1975), then to a period spanning from the Late Bronze Age (LBA) to the advanced IA (Maselli Scotti 1982; 1983). How­ever, Bronze Age materials were only partially pub­lished, and our revision of pottery finds has allow­ed to identify pottery shards dating back to the EBA/ initial Middle Bronze Age (MBA; see infra). No tra­ces of a later Roman occupation of the site have been identified (Cannarella 1975). Archaeological survey (Bernardini 2012) and airborne laser scanner in­vestigations (Bernardini et al. 2013; Vinci, Bernar­dini 2017) have revealed that Rupinpiccolo is not isolated, but is part of a small-scale cluster of four hillforts. Gradina on Veliki Brijun/Brioni Maggiore The Gradina hillfort, also known as Monte Castellier, is located on the eastern side of the Veliki Brijun is­ Prehistoric stone disks from entrances and cemeteries of north-eastern Adriatic hillforts land in front of southern Istria (Fig. 1). Veliki Brijun is part of the Upper Albian to Middle Cenomanian carbonate succession of southern Istria (Ti.ljar et al. 1998). The hillfort is composed of three concentric ram­parts extending over a surface of more than seven hectares (Vitasovi. 2002; 2005). Already recognized more than a century ago (Puschi 1898; Marchesetti 1903), it was investigated by Anton Gnirs (1925), who identified an entrance in the north-western part of the upper rampart and, later, by Boris Ba.i. and Anton Vitasovi. (2002; 2005). These last resear­chers excavated the entrance previously identified, revealing a complex structure designed to defend the access to the site. They also investigated other structures and a small cemetery next to the external side of the upper rampart, composed of a few cist tombs within platforms divided by low walls (Fig 1B; Vitasovi. 2002; 2005; Bur.i.-Matija.i., .eri. 2013). The Gradina entrance finds significant com­parisons in Bronze Age hillforts of Istria, such as Monkodonja (Hänsel et al. 2015), while the small cemetery is very similar to the Vr.in necropolis, re­cently radiocarbon dated to between about 1700 and 1200 BC (Cupito et al. 2018). The available data suggest that the Gradina on Veliki Brijun was set­tled for a long time span, at least from the late EBA and the IA (Vitasovi. 2002; 2005). Materials and methods Identification of the stone disks and related archaeological evidence The stone disks 1 and 2 from Rupinpiccolo were un­covered during the excavations carried out in the 1970s close to the inner side of the south-eastern en­trance (Cannarella 1970; 1975), but they were not recognized as valuable artefacts and were left on site (Fig. 1). Since the identification of the disks by Federico Bernardini, all the area has been carefully and repeatedly surveyed, in order to identify pos­sible chisel marks and wedge holes on blocks of the rampart, similar to those identified on the disks (see infra) and on the karst outcrops just next to it. This research confirmed the presence of chisel marks and a wedge hole on the hilltop (Priuli 1977) (Fig. 1 and supporting Fig. 1A) and led to the identifica­tion of wedge hole remains on the karst outcrop next to the southern wall of the corridor-like struc­ture in front of the south-eastern entrance (Fig. 1 and supporting Fig. 1B) and on a block of the ram­part (Fig. 1 and supporting Fig. 2). The outcropping rocks with traces of wedge holes and chisel marks on the hilltop and close to the entrance were buried under the ruins of the rampart and brought to light by the archaeological excavations (Priuli 1977). Stone disk 1 from Gradina on Veliki Brijun was dis­covered during the excavation of the entrance of the upper rampart and was preliminarily described and recognized as an artefact with a probable ritual value (Vitasovi. 2005). Stone disk 2 was identified during a visit to the small Bronze Age cemetery of the hillfort in 2019 by Federico Bernardini. It lies on the eastern side of the low wall that delimits a tomb (Fig. 1B). Drone structure from motion photogrammetry A drone survey of the whole Rupinpiccolo site was performed in 2017 in order to produce orthophotos and plans of the site, taking advantage of the low and sparse vegetation (Eltner et al. 2016). Two se­parate flights, with the camera aligned perpendicu­lar to the flight path and with a tilt angle of 45 de­grees, respectively, were planned with FlightPlanner software (AeroScientific, Blackwood, Australia) to maintain a constant ground sample resolution (GSR) and optimize the area coverage. Drone pictures were taken using a DJI Mavic drone (DJI, Nanchan District, Shenzen, China) capable of providing 12Mp files with a zoom lens equivalent to 24–48mm. Working with a medium-long focal lens instead of a wide angle allowed us to perform higher flights above the vegetation and maintain a high ground resolution (12.7mm/pixel) over an area of 46 953m2. In order to reduce the flight time, jpeg files were saved, cloudy days were preferred and almost no shadows were registered. A total of 414 orthogonal images and another series of 94 images with the same focal length but different angulations and ground resolutions were taken and processed using Agisoft Metashape (Agisoft LLC, St. Petersburg, Russia, 2019). The images were aligned, and a sparse point cloud generated using high quality settings. Thickening the cloud was done using the Dense Cloud algorithm, which was run at high resolution to provide a large number of points, suitable for DEM generation. Such an approach was not applied at Gradina on Ve­liki Brijun because the area is covered by dense evergreen vegetation. Terrestrial structure from motion photogram­metry The terrestrial structure from motion (SfM) approach was applied to produce a plan of the Gradina on Ve­liki Brijun cemetery and high-resolution 3D models Federico Bernardini, Giacomo Vinci, Vanja Macovaz, Andrea Baucon, Angelo De Min, Stefano Furlani, and Sne/ana Smolic´ of the stone disks (Verma et al. 2019; Porter et al. 2016), a Rupinpiccolo rampart block with wedge-hole remains, the Rupinpiccolo quarrying sites and a block from the Bronze Age Vr.in cemetery in Is-tria (Croatia). A full frame 21 megapixel camera was used with a 17mm lens, and raw format images were taken in order to produce files with low contrast ca­pable of generating point clouds of even dark spots. A total of 56 images of the Gradina on Veliki Brijun burial site, 270 images of the Rupinpiccolo quarry­ing site east of the north-eastern gate, 91 images of the Rupinpiccolo rampart block with two symmet­rical half-wedge holes and 65 images of the block from Vr.in cemetery were acquired. For the stone disks, two series of pictures represent­ing both sides of the artefacts and their edges were taken to ensure the correct alignment. When neces­sary, a flashlight was used to achieve better visibili­ty in the shadows and avoid uneven lighting (Men­na et al. 2016). Finally, a colour reference was used to calibrate colours during postproduction and to scale the model. A total of 98 and 96 images of Ru-pinpiccolo stone disk 1 faces, 104 and 125 images of Rupinpiccolo stone disk 2 faces, 113 and 134 ima­ges of Gradina on Veliki Brijun stone disk 1 faces and 38 and 82 images of Gradina on Veliki Brijun stone disk 2 faces were taken. All the acquired images were processed using Agi-soft Metashape (Agisoft LLC, St. Petersburg, Russia, 2019), as described for the drone photogrammetry of Rupinpiccolo. In addition, to remove outliers and improve the overall quality of the models, after sparse cloud generation the gradual selection tool was applied to remove non-correctly aligned pictu­res. After separating the sparse and dense cloud ge­neration of the stone disk faces, manual alignment of the two clouds selecting homolog points along the stone edges was performed just before the mesh generation and texturing. The large number of pixels and the high dimensio­nality of the CCD sensor used provided high resolu­tion models: Gradina on Veliki Brijun cemetery: 10.7 mm/pixel; Rupinpiccolo rampart block: 0.254 mm/ pixel; quarrying site: 12mm/pixel; Vr.in block: 1.68 Fig. 2. Rupinpiccolo stone disk 1. a Radiance scaling and lattice visualizations of the disk with chisel marks highlighted in black (full circles correspond to the bottom of the chisel marks). b Digital elevation mo­del of the frontal face of the disk. c Curvature map of the disk with chisel marks indicated by black cir­cles and numbers. d Enlarged view of chisel marks 10–15, 18–19 and 23. The other chisel marks are clear­ly visible in Fig. 3. Prehistoric stone disks from entrances and cemeteries of north-eastern Adriatic hillforts mm/pixel; Rupinpiccolo stone disks 1 and 2: 0.114 and 0.104mm/pixel, respectively; Gradina on Veliki Brijun stone disks 1 and 2: 0.111 and 0.627mm/ pixel, respectively. Elaboration of 3D models The obtained 3D models were visualized and ren­dered by using MeshLab 2020.03 (Cignoni et al. 2008) in order to enhance artificial features, such as chisel marks and wedge traces, and bioerosional evidence detected on the frontal face of the disk 1 from the Gradina on Veliki Brijun hillfort. Surface features have been enhanced applying radiance scaling, and/or lattice plugins. The radiance scaling technique makes it possible to enhance shape de­tails such as convexities and concavities (e.g., Ver­gne et al. 2012), while the lattice shader makes it possible to detect surface details by removing the colour and regulating the light direction (e.g., Cas-sen et al. 2014). In order to enhance the bottom of the chisel marks of stone disk 1 from Rupinpiccolo, the plugin Colorize curvature (APSS) was applied using the default parameters, with the exception of the filter scale which was set at 4. Typological analysis of Bronze Age pottery and initial occupation of Rupinpiccolo Bronze Age materials from Rupinpiccolo have only been partially published and, for this reason, we have reviewed the pottery finds from the site kept in the Soprintendenza Archeologia Belle Arti e Pae­saggio del Friuli Venezia Giulia. Results Rupinpiccolo stone disks Two disks, carved from the local massive limestone belonging to the Repen Formation (Jurkov.ek et al. 2016), have been identified next to the south-east­ern gate of the Rupinpiccolo hillfort (Trieste Karst, Italy) (Fig. 1). Rupinpiccolo stone disk 1 The Rupinpiccolo disk 1, about 50cm wide and 20cm thick, was carved by using a metal point chisel and a hammer. Producing a rounded disk from very mas­sive and compact limestone is not a simple task, since advanced stone-working skills are required. The stone has a relatively flat upper face showing numerous chisel marks arranged in patterns (Figs. 2–3). The depth of the chisel marks ranges from about 20 to 1mm in those more severely affected by weathering processes. The diameter of the chisel marks ranges from about 10 to 5mm with most of them being about 7mm. This suggests that the orig­inal point of the chisel was about 6–7mm in diam­eter. The relatively flat morphology of the frontal sur­face of disk 1 was probably achieved by centripetal flaking using a precursor. The position of the chis­el marks corresponds to limited portions of the disk upper surface, suggesting they were not due to surface flattening processes. This is also shown by the digital elevation model of the frontal surface of the disk (Fig. 2B), where some marks were pro­duced on the highest portion of the surface but without making it flatter, and the others, located at the lowest portion of the stone, have produced an irregular depression. Moreover, the very small dis­tance between some marks (i.e. less than 2cm) fur­ther supports the idea that they were intentionally created to reproduce a pattern. The chisel marks are highlighted in the curvature map of the disk due to their hemispherical shape, which separate them well from the other surfaces of the disk (Figs. 2C and 3). They are iso-oriented and were likely created by a right-handed person holding an oblique metal chisel with the left hand and a ham­mer with the right one. Considering the north-eastern Adriatic region, the use of bronze tools to work limestone is document­ed in the cemetery of Vr.in hillfort in Istria covering approximately a time span between 1700 and 1200 BC (Battaglia 1958; Cupito et al. 2018; see infra). Most of the cist tombs are included within platforms delimited by low walls made of well-shaped blocks which show clear marks produced by a metal point chisel (Battaglia 1958; supporting Fig. 3), similar to those found at Rupinpiccolo. Rupinpiccolo stone disk 2 The Rupinpiccolo stone disk 2 was found close to stone disk 1 (Fig. 4). They share the same raw ma­terial, size (about 50cm large and 30cm thick) and shape, but its frontal surface is very flat and without any chisel marks (Fig. 4). To obtain such a flat sur­face a rock with a very regular bedding plane was selected, extracted and carefully worked. Unlike the other stone, disk 2 still has the remains of two diffe­rent half-wedge holes that made it possible to shape the artefact. One of them, originally part of a com­plete hole with a triangular cross-section, is about 12cm wide and 6cm deep and still preserves faint parallel and oblique chisel marks on its surface (Fig. 4, black arrows and lines). It develops on the lateral Federico Bernardini, Giacomo Vinci, Vanja Macovaz, Andrea Baucon, Angelo De Min, Stefano Furlani, and Sne/ana Smolic´ Fig. 3. Curvature map of the frontal face of the Rupinpiccolo stone disk 1 and enlarged views of the chisel marks divided in three groups (I-III). Chisel marks of Group I are indicated by black circles, those of group II (with the exception of one mark which is not shown) by blue circles and those of group III by red circles. surface of the disk from the frontal surface towards the bottom of the disk. It was likely opened up to se­parate the block from the outcrop. The other half-wedge hole, about 15cm wide and 7cm deep, shares a similar cross-section and chisel marks but it is lo­cated on the bottom surface (Fig. 4, grey arrows and lines). It was opened to reduce the thickness of the disk. The presence of such wedge holes is significant, as it makes it possible to connect the disks with the quarrying sites used to build the rampart. After the production of the rough stone, it was carefully re­touched by using a hammer to make the frontal sur­face round. Typological analysis of Bronze Age pottery and initial occupation of Rupinpiccolo Based on our review of the pottery assemblage from Rupinpiccolo, the initial occupation of the site is likely to be placed during the late EBA/MBA. In particular, one elbow (a piastra) shaped handle with straight upper profile (Fig. 5.1), one fragment of handle with a large central impression (Fig. 5.2) and another a lingua pointing up handle (Fig. 5.3) all refer to typical castellieri types well attested in the Istrian peninsula, the Trieste Karst and Friuli be­tween late EBA-initial MBA (e.g., Urban 1993. Pl. 1. 2–3; Mihovili. 1997.Pl. 1.1 and 10; Hellmuth Kram­berger 2017.Pl. 100.8–10). The base of a bowl decorated with rectilinear fur­rows on the external side (Fig. 5.4) points to a simi­lar chronological span. This decoration refers to a cruciform-like set of motifs, probably recalling the solar symbolism associated with pottery finds (ge­nerally to the external base of bowls and cups) wide­spread in several coeval contexts from the North Ad­riatic (Trieste Karst, Istria and Kvarner), the Po plain and the adjacent peri-Alpine zone and the Panno­nian-Danubian region (Hellmuth Kramberger 2017. 161–168). The best comparisons for the Rupinpic­colo base are found at the hillfort Gradac-Turan in Istria (Mihovili. 1997.Pl. 1.3). A generic chronology pointing to the MBA can be proposed for some relatively small globular bowls and jars with everted rims, sometimes decorated with a circular knob under the rim (Fig. 5.5–7), well attested in the castellieri assemblages (Cardarelli 1983.Pl. 17, type 71 and Pl. 18, type 6) and very common at several sites in the Trieste Karst and Is-tria (e.g., Hellmuth Kramberger 2017.Pl. 32.2). Finally, a carinated bowl (Fig. 5.8), a jar with a thick crown-like rim (Fig. 5.9) and a handle with a deep circular impression (Fig. 5.10) find good compar­isons with types in use between the final MBA and the first centuries of the LBA (Cardarelli 1983.Pl. 18, type 97a, Pl. 19, type 14; Hellmuth Kramberger 2017.Pl. 20.1). Prehistoric stone disks from entrances and cemeteries of north-eastern Adriatic hillforts Chronology of Rupinpiccolo disks The pottery assemblage from Rupinpiccolo shows that the hillfort was probably in use from about 1800/1650 to 400 BC. The disks can only be safely referred to this long time-span (the original finding context was not recorded in detail during excava­tions), but several data suggest they can be lined to an early building phase of the rampart. The half-wedge holes on disk 2 and chisel marks on disk 1 can be associated with identical stone-cutting marks identified next to the external face of the rampart (Fig. 1.C, locations 3–4) and in one large block of the rampart itself (Fig. 1.C, location 5). These stone-cut­ting marks do seem to be related to the building of the rampart itself. . Location 3 of Fig. 1.C. East of the north-eastern gate of the Rupinpiccolo hillfort, an area covering about 16 square meters, that before excavations was buried under the ruins of the rampart, is characte­rized by outcropping rocks showing numerous frac­tures produced by the extraction of blocks to build the rampart (Fig. 1.C,E, location 3). Chisel marks, interpreted as a representation of a halberd (Priuli 1977), and a wedge hole were already identified on the face of a large outcropping rock in the 1970s (supporting Fig. 1.A). An accurate re-examination of the rock and the elaboration of its 3D model have made it possible to identify a line of additional chisel marks parallel to the handle of the putative halberd and many other previously unreported marks on the adjacent eastern rock face. The wedge-hole is complete and is identical to those identified on stone disk 2, showing a triangular cross-section and similar dimen­sions (about 20cm wide and 7cm deep). The chisel marks present on both wedge-holes and the other surfaces of the rock are comparable to those identified on the stone disks, likely produced by a point chisel. The position of chisel marks and wedge-hole sug­gest that all of them are relat­ed to an aborted attempt to extract a stone block after a first block was successfully obtained. The putative hal­berd, located opposite to the wedge-hole, seems more likely a quarry mark relat­ed to the extraction activity. . Location 4 of Fig. 1.C. An additional half-wedge hole has been identified on an outcrop next to the southern wall of the corridor-like structure in front of the south-eastern entrance (Fig 1.C,D, location 4 and supporting Fig. 1.B). It is also very similar to those present on stone disk 2. Originally part of a complete hole with a triangular cross-section, it is larger (about 40cm wide and 20cm deep) and still preserves faint parallel and oblique chisel marks on its surface. This area was completely buried un­der the ruin of the entrance, testifying to its ancient origin. . Location 5 of Fig. 1.C. Finally, a wedge-hole di­vided in two symmetrical parts has been identified on a large block broken in three big pieces, which belongs to the inner stone alignment of the rampart of Rupinpiccolo (Fig 1.C, location 5 and supporting Fig. 2). Three chisel marks and half-wedge hole have been recognized on the southern fragment of the block (supporting Fig. 2, black arrows and lines). The chisel marks are probably related to the pre­liminary operations aimed at the production of the hole. Just in front of this first half-wedge hole, ano­ther one originally belonging to the same hole has Federico Bernardini, Giacomo Vinci, Vanja Macovaz, Andrea Baucon, Angelo De Min, Stefano Furlani, and Sne/ana Smolic´ been identified on the north­ern fragment of the block (supporting Fig. 2, white ar­rows and lines). The wedge-hole is about 18cm wide and 10cm deep and shows a tri­angular section but no paral­lel chisel marks, similar to those visible on the half-wedge holes of disk 2 and the outcropping rock close to the south-eastern entrance, are visible. This suggests that the block suffered higher le­vels of dissolution probably because it was not completely buried under the ruins of the rampart such as the stone disks and the outcropping rocks with quarrying traces. At the macroscopic level, the original surfaces of the block and the wedge-hole faces appear much smo­other than the fracture surfaces, suggesting the block was put in place after an aborted attempt to split it and broke up in recent times. The ancient origin of the quarrying traces was ascer­tained based on the fact that they were covered by the massive ruins of the rampart and are located next to it. A possible later quarrying activity is un­likely, considering that the ruins of the rampart could easily provide abundant already extracted lime­stone blocks of different size. The identification of one of the newly identified wedge holes on one block of the rampart would further con­firm such a hypothesis. Considering that the surviving rampart has been entirely excavated, and only one main building phase was recognized with some possible minor additions (i.e. rampart stone reinforcements and the corridor-like structure con­nected to the south-eastern gate; Cannarella 1975), its construction could have taken place during an early phase of the hillfort. Most of the large blocks used to build the Rupinpiccolo rampart were proba­bly extracted levering along natural fractures in the local limestone out­crop, but, when necessary, wedge-holes produced by metal chisels were used to split the rock in absence of natural discontinuities by using wooden levers or wooden/stone wedges. In Egypt, wedging was generally believed to be a quarry technique developed during the mid-1st mil­lennium BC with the introduction of iron tools/wed­ges (e.g., Harrel, Storemyr 2009), but the identifi­cation of u-shaped holes in greywacke quarries of the Wadi Hammamat (eastern desert of Egypt) and Prehistoric stone disks from entrances and cemeteries of north-eastern Adriatic hillforts stone wedges of the same rock type suggest that such technique was in use much earlier, probably since the Old Kingdom (mid-3rd millennium BC; Bloxam 2015). Stone disks from Gradina on Veliki Brijun/Bri­oni Maggiore The other pair of disks are from the Gradina hillfort (Veliki Brijun island, Croatia; Fig. 1), in use at least from the late EBA to IA. Stone disks 1–2 from Gradina on Veliki Brijun Gradina 1 is about 50cm wide and 10cm thick and was found next to the entrance of the upper ram­part (Vitasovi. 2005). Both faces show a central ar­tificial cup mark, 6cm wide, but only one face is co­vered with shallow hemispherical depressions, like­ly produced by marine bivalves (Fig. 6 and supple­mentary text), while the opposite one is flat (Fig. 6). The other disk, Gradina 2, has been identified in the Bronze Age cemetery next to the upper rampart (Figs. 1.B and 7), composed of a few cist tombs with­in platforms divided by low walls. Despite its slight­ly smaller size, with a diameter of about 44cm and without wedge holes, it closely resembles the Rupin-piccolo disk 2. Its frontal surface, partially damaged on one side, is very flat and without any chisel marks. It was obtained by selecting a stone slab with a regular bedding plane and through a careful retou­ching of the disk edges by using a hammer precur­sor. Chronology of disks from Gradina on Veliki Brijun Disk 2 gives very important clues to date the north-eastern Adriatic disks, since it is the only one belonging to a secure Bronze Age context. The disk comes from the small cemetery next to the upper rampart (Fig. 1.B), composed of three burials within platforms divided by low walls. The infant burial and burial 1 of Figure 1.B are characterized by a stone cist, not present in burial 2. Significant Bronze Age pottery was found in tombs 1 and 2. Both a hemispheri­cal bowl characterized by a straight rim with an inwardly slanted edge (burial 2; Vitasovi. 2002. Pl. 3.3) and a cup with a raised triangular handle with a rounded end plate (bu­rial 1; Vitasovi. 2002.Pl. 5.1) point to a MBA I-II chronology according to the Italian relative chronological system (Cardarelli 2009; 1983.91, Pl. 17, type 15; 93, Pl. 18, type 65) corre­sponding to the Bz III phase of the Istrian chrono­logy (Hänsel et al. 2015). They can be compared to similar artefacts from the recent phase of Monkodo­nja (roughly 1600–1450 BC; Hellmuth Kramber­ger 2017.139, Fig. 110; 144, Fig. 115, variant a1; 324, Fig. 255 for the bowl; 84, Fig. 57; 89, 323, Fig. 254). Considering the stratigraphic relations, the platform of burial 2 was the first one to be built. The disk was found on the top of the low wall of the platform of burial 2, likely being visible when the burial was in use. The cemetery was uncovered by excavations which brought to light only Bronze Age materials. Disk 1 can be generally attributed to protohistory and probably to the Bronze Age on the basis of com­parisons with Gradina disk 2. Discussion and conclusions Structure from motion (SfM) photogrammetry has been proved to be a low-cost, effective and accurate method to document and represent in 3D the stone disks and related quarrying areas presented in this paper. Gradina disk 2 can be safely dated to the Bronze Age and likely to the MBA on the basis of pottery find- Federico Bernardini, Giacomo Vinci, Vanja Macovaz, Andrea Baucon, Angelo De Min, Stefano Furlani, and Sne/ana Smolic´ Fig. 8. Gradina disk 1 (a) compared to the Bronze Age offering table from Chrysolakkos (Crete) (b). Photo top left from de Pierpont 1987, the other photo and the drawing from Demargne 1932 (image modified from Arca 2015). ings; Rupinpiccolo disks could be related to an early building phase of the settlement that was likely es­tablished during the late EBA/initial MBA, while no precise contextual chronological data are available for Gradina disk 1, and a general protohistoric attri­bution can be proposed. However, considering their typological homogeneity and similar finding con­texts – i.e., entrances or cemeteries – all probably belong to a similar chronological horizon (i.e., an early phase of the castellieri culture). The disks show flat and plain surfaces or are cov­ered with sub-circular depressions and without any macroscopic use-wear traces. Gradina disk 1 shows two larger cup-marks at the centre of both faces. It is worth mentioning that simple slabs and blocks with cup-marks are reported from IA cremation ce­meteries of Istria (i.e., Nezakcij/Nesazio, Ka.tel/Cas­telvenere and Limska gradina/Gradina di Leme), but at least some of them could originally belong to Bronze Age tombs (Mladin 1964; Mihovili. 1996; 2014). Cup-marks are, indeed, reported from a block belonging to tomb 2 of the Bronze Age cemetery of Gradina on Veliki Brijun itself (Vitasovi. 2002.14), where the Gradina disk 2 has been identified. We propose to interpret the stone disks as ritual ar­ tefacts and/or possible cult representations based on the association of all disks with significant and sacred settlement locations (they could be related to some type of cult designated to protect the hillforts and/or connected to the dead and the afterlife), the absence of functional traces and comparisons with cup-marks from Istrian protohistoric cemeteries (see above), and similar Bronze Age stones covered with circular depressions found mainly close to citadel en­ trances in Anatolia and burial sites, entrances and thresholds in the Aegean area. a In Anatolia blocks with cup-marks have been found in front of the gates of Troy, pro­bably reflecting Anatolian reli­gious elements before the end of the 2nd millennium BC (Korfmann 1998), in the city gates or funeral contexts of Bo­gazköy-Hattu.a (Neve 1977-78), ¢ close to the north-west gate of Kusakli-Sarissa (Mielke 2018. Fig. 6.6) and in some 2nd-mil- b lennium BC citadels in the Marmara Lake basin of the Ge-diz Valley (Luke, Roosevelt 2017). They are generally in­terpreted as evidence of liba­tion rituals (Neve 1977–78; Luke, Roosevelt 2017). Other significant comparisons are cylindrical stones of simi­lar size with a central large cup-mark but surrounded by circularly arranged circular de­pressions, defined as offering tables, known from protopala­tial Crete, more precisely from the New Palace of Mallia (ro­ughly 1700–1500 BC; Chapou­thier 1928; Cucuzza 2010.Fig. 1; Arca 2015) and Chrysolak­kos (Demargne 1932; de Pier­ Prehistoric stone disks from entrances and cemeteries of north-eastern Adriatic hillforts pont 1987; Arca 2015). Similar circularly arranged cup-marks, mainly carved into steps and pavements and belonging to the same chronological horizon, are common in Phaistos and in other sites in Crete, where they have been variously interpreted as evi­dence of rituals or games, or both at the same time. Among other identifications, they are variably de­fined as stone kernoi, Minoan cup-holes, and stone slabs with depressions (Whittaker 2002; Cucuzza, Ferrari 2004; Hillbom 2005; Cucuzza 2010; Arca 2015). Gradina disk 1 is quite similar to the Crete offering tables from Mallia and Chrysolakkos due to the size, shape and presence of the central, quite large cup-marks (Fig. 8), but it differs because the smaller cir­cular depressions around the central cup-marks are irregularly arranged and probably the result of ma­rine bio-erosion. Some chisel marks of Rupinpiccolo disk 1 show a sub-circular distribution which is, anyway, more irre­gular than in the kernoi from the Aegean area. More­over, their dimeter is relatively small (5–10mm) compared to that of most kernoi (Cucuzza, Ferrari 2004). This leaves open the question as to whether the chisel marks of Rupinpiccolo disk 1 could have a different meaning and function. Interestingly enough, disks with a plain face and no cup-marks similar to Gradina 2 and Rupinpiccolo 2 are so far unknown in the eastern Mediterranean. Considering their possible ritual meaning, stone disks with a plain face could perhaps be interpreted as a representation of the Sun. So­lar motifs are in fact known from EBA/MBA pottery from the north-eastern Adriatic (Hänsel et al. 2015), includ­ing Rupinpiccolo itself. Icono-graphic elements referring to the Sun and its cyclic move­ments spread out in Europe at least from the late Copper Age, probably in connection with the large-scale migra­tions from the Pontic steppe region to western Europe du­ring the 3rd millennium BC (Kaul 1998; West 2007; Kri­stiansen 2010; 2012; Allen-toft et al. 2015; Haak et al. 2015; Kristiansen et al. 2017). The bronze and gold chariot from Trundholm, Den­mark, carries an elaborate image of the Sun of the 14th century BC (Kaul 1998; West 2007). Similar ornate gold disks have come to light in various parts of Europe and probably are deity representations. They were produced over a long time, between the late Copper Age and late Bronze Age (Cahill 2015). The occurrence of two circular faces on the same ar­tefact, such as at Veliki Brijun, or on two associated but separate objects, such as at Rupinpiccolo, is re­miniscent of solar symbolic imagery of the European Bronze Age related to the representation of the Sun’s daily journey from day to night (Kaul 1998; West 2007; Pásztor, Roslund 2007; Kristiansen 2010; 2012). If this interpretation is correct, the plain faces could be a representation of the Sun, while those covered with chisel marks of the night sky. Bronze Age representations of the night sky in the European Bronze Age are rare, and the Nebra disk from Germany is the most famous example (about 1600 BC; Meller 2002; Schlosser 2002; Pásztor, Ros-lund 2007; Kristiansen 2010; Pásztor 2015; Per-nicka et al. 2020). In conclusion, we propose interpreting the disks from north-eastern Adriatic hillforts as Bronze Age ritual artefacts reflecting the position of north-east­ern Adriatic regions between the Mediterranean and central Europe. They seem to be related to the solar symbolic imagery of European Bronze Age, but at the same time show strong connections with the eastern Mediterranean during an early phase of the Federico Bernardini, Giacomo Vinci, Vanja Macovaz, Andrea Baucon, Angelo De Min, Stefano Furlani, and Sne/ana Smolic´ castellieri culture (Fig. 9). This is in line with the hypothesis – mainly based on specific types of pot­tery, metal artefacts and architectural models – that the first Istrian settlements had direct contacts pre­cisely with this area (Hänsel et al. 2015; Hellmuth Kramberger 2017). The stone disks would show that the inhabitants of north-eastern Adriatic regions not only shared some aspects of the material culture, but also common religious habits. ACKNOWLEDGEMENTS The project was supported by the Monton MMS company in the framework of the ICTP project “Interdisciplinary study of the ancient landscape of north-eastern Adriatic regions”. Author contributions: F. B. identified the Rupinpiccolo stone disks and related quarrying traces, designed, initi­ated and led the study; F. B., G. V., V. M. produced photogrammetric data; G. V. produced plans of investigated sites; F. B., G. V. performed archaeological research; F. B., G. V., A. D. M., S. F. performed geomorphologic re­search; A. B. performed ichnological research; F. B. wrote the manuscript with significant input from G. V., A. B. and with contributions from all authors; all authors contributed to final interpretation of data. . References Allentoft M. E., Sikora M., Sjögren K., +61 authors, and Willerslev E. 2015. Population genomics of Bronze Age Eurasia. Nature 522: 167–172. https://doi.org/10.1038/nature14507 Arca A. 2015. Phaistos, Crete, cup-marks and other signs. TRACCE Online Rock Art Bulletin 36. https://www.rupestre.net/tracce/?p=9924 Battaglia R. 1958. I castellieri della Venezia Giulia. In F. Franco, F. Reggiori (eds.), Le meraviglie del passato. Mondadori Editore. Milano: 419–434. Bernardini F. 2012. Nuove strutture forti.cate proto­storiche a Rupinpiccolo (Carso triestino). Atti del Museo Civico di Storia Naturale di Trieste 55: 3–11. Bernardini F., Sgambati A., Montagnari Kokelj M., +6 au­thors, and De Min A. 2013. Airborne LiDAR application to karstic areas: the example of Trieste province (north-east­ern Italy) from prehistoric sites to Roman forts. Journal of Archaeological Science 40: 2152–2160. https://doi.org/10.1016/j.jas.2012.12.029 Bloxam E. 2015. ‘A Place Full of Whispers’: Socializing the Quarry Landscape of the Wadi Hammamat. Cambridge Archaeological Journal 25: 789–814. https://doi.org/10.1017/S0959774315000426 Borgna E., Cassola Guida P., Mihovili. K., Tasca G., and Ter-.an B. 2018. Bronzo Antico-Bronzo Recente. In E. Borgna, P. Cassola Guida, and S. Corazza (eds.), Preistoria e Proto­storia del ‘Caput Adriae’. Studi di Preistoria e Protostoria dell’Istituto Italiano di Preistoria e Protostoria nr. 5. Isti­tuto Italiano di Preistoria e Protostoria. Firenze: 75–96. Bur.i.-Matija.i. K., .eri. H. 2013. Pogrebni obredi i uko-pi na istarskim gradinama u bron.ano doba. Tabula. .a-sopis Odjela za humanisti.ke znanosti. Sveu.ili.te Jur­ja Dobrile u Puli 11: 67–92. Cahill M. 2015. ‘Here comes the sun....’ solar symbolism in Early Bronze Age Ireland. Archaeology Ireland 29: 26– 33. Cannarella D. 1970. Nota preliminare su alcune ricerche effettuate in castellieri carsici. Annali Gruppo Grotte 11 dell’Associazione XXX: 37–43. 1975. Il castelliere di Rupinpiccolo sul Carso triestino, relazione preliminare. Atti della Societa per la Preisto­ria e la Protostoria della Regione Friuli-Venezia Giu­lia 2: 117–127. Cardarelli A. 1983. Castellieri nel Carso e nell’Istria: cro­nologia degli insediamenti fra media eta del bronzo e prima eta del ferro. Preistoria del Caput Adriae. Istituto per L’enciclopedia del Friuli Venezia Giulia. Udine: 116– 124. 2009. The collapse of the Terramare culture and growth of new economic and social systems during the Late Bronze Age in Italy. Scienze dell’Antichita 15: 449–520. Cassen S., Lescop L., Grimaud V., and Robin G. 2014. Com­plementarity of acquisition techniques for the documen­tation of Neolithic engravings: lasergrammetric and pho­tographic recording in Gavrinis passage tomb (Brittany, France). Journal of Archaeological Science 45: 126– 140. https://doi.org/10.1016/j.jas.2014.02.019 Prehistoric stone disks from entrances and cemeteries of north-eastern Adriatic hillforts Chapouthier F. 1928. Une table a offrandes au palais de Mallia. Bulletin de correspondance hellénique 52: 292– 323. Cignoni P., Callieri M., Corsini M., Dellepiane M., Ganovelli F., and Ranzuglia G. 2008. Meshlab: an open-source mesh processing tool. In V. Scarano, R. De Chiara, and U. Erra (eds.), Sixth Eurographics Italian Chapter Conference. Eurographics Association. Geneve: 129–136. Codacci-Terlevi. G. 2012. La scoperta della sepoltura a tumulo di Barbariga (Istria, Croazia). Novita e riscontri in relazione ai tumuli dell’eta del bronzo rinvenuti nella pen-isola istriana e nell’Italia nord-orientale. In E. Borgna, S. Müller-Celka (eds.), Ancestral landscapes: burial mounds in the Copper and Bronze Ages (Central and Eastern Europe – Balkans – Adriatic – Aegean, 4th–2nd millen­nium BC). Maison de l’Orient et de la Méditerranée. Lyon: 153–162. Cucuzza N. 2010. Game boards or offering tables? Some remarks on the Minoan ‘pierres a cupules’. Kernos 23: 133–144. Cucuzza N., Ferrari C. 2004. I cosiddetti kernoi di Festos. Creta antica. Rivista annuale di studi archeologici, sto­rici ed epigrafici 5: 43–52. Cupito M., Tasca G., Pulcini M. L., Carrara N., and Lotto D. 2018. Il “sepolcreto gentilizio” del castelliere di Monte Orcino/Vr.in – Scavi Battaglia-Tamaro 1925–1928. Rilet­tura del contesto archeologico e risultati del riesame bio-archeologico dei resti umani. In E. Borgna, P. Cassola Gui-da, and S. Corazza (eds.), Preistoria e Protostoria del ‘Ca-put Adriae’. Studi di Preistoria e Protostoria dell’Istituto Italiano di Preistoria e Protostoria nr. 5. Istituto Italiano di Preistoria e Protostoria. Firenze: 509–526. Demargne J. 1932. Culte funéraire et foyer domestique dans la Crete minoenne. Bulletin de correspondance hel­lénique 56: 60–88. de Pierpont G. 1987. Réflexions sur la destination des édifices de Chrysolakkos. In R. Laffineur (ed.), Thanatos. Les coutumes funéraires en Egée a l’âge du Bronze. Aegaeum: annales d’archéologie égéene de l’Université de Liege 1. Université de l’Etat a Liege. Liege: 79–94. Eltner A., Kaiser A., Castillo C., Rock G., Neugirg F., and Abellán A. 2016. Image-based surface reconstruction in geomorphometry – merits, limits and developments. Earth Surface Dynamics 4: 359–389. https://doi.org/10.5194/esurf-4-359-2016 Gnirs A. 1925. Istria Praeromana. Verlag von Walther Heinisch. Karlsbad. Haak W., Lazaridis I., Patterson N., +35 authors, and Reich D. 2015. Massive migration from the steppe was a source for Indo-European languages in Europe. Nature 522: 207– 211. https://doi.org/10.1038/nature14317 Hänsel B. 2009. Die Bronzezeit 2200–800 v. Chr. In S. Von Schnurbein (ed.), Atlas der Vorgeschichte. Europa von den ersten Menschen bis Christi Geburt. Theiss. Stuttgart: 106–149. Hänsel B., Mihovili. K., and Ter.an B. 2015. Monkodonja. Istra.ivanja protourbanog naselja bron.anog doba Is-tre. Knjiga 1. Iskopavanje i nalazi gra.evina/Monkodo­nja. Forschungen zu einer protourbanen Siedlung der Bronzezeit Istriens. Teil 1. Die Grabung und der Baube-fund. Monografije i katalozi/Monographien und Kataloge 25. Arheolo.ki muzej Istre/Archäologisches Museum Is-triens. Pula. Harrell J. A., Storemyr P. 2009. Ancient Egyptian quar­ries – an illustrated overview. In N. Abu-Jaber, E. G. Blo­xam, P. Degryse, and T. Heldal (eds.), QuarryScapes: An­cient stone quarry landscapes in the eastern Mediter­ranean. NGU Special Publication 12. Geological Survey of Norway. Trondheim: 7–50. http://www.eeescience.utole do.edu/faculty/harrell/egypt/Harrell_Storemyr_Ancient EgyptianQuarries_IllustratedOverview.pdf Hellmuth Kramberger A. 2017. Monkodonja. Istra.ivanje protourbanog naselja bron.anog doba Istre, Knjiga 2. Keramika s bron.anodobne gradine Monkodonja/Mon­kodonja. Forschungen zu einer protourbanen Siedlung der Bronzezeit Istriens. Teil 2 – Die Keramik aus der bronzezeitlichen Gradina Monkodonja. Monografije i katalozi/Monographien und Kataloge 28. Arheolo.ki mu-zej Istre/Archäologisches Museum Istriens. Pula. Hillbom N. 2005. Minoan Games and Game Boards. Lund University. Lund. Jurkov.ek B., Biolchi S., Furlani S., +5 authors, and Cuc-chi F. 2016. Geological map of the Classical Karst Region (SW Slovenia–NE Italy). Journal of Maps 12: 352–362. https://doi.org/10.1080/17445647.2016.1215941 Kaul F. 1998. Ships on bronzes: a study in Bronze Age religion. Publications from the National Museum Studies in Archaeology & History. Copenhagen. Korfman M. 1998. Troia, an Ancient Anatolian Palatial and Trading Center: Archaeological Evidence for the Pe­riod of Troia VI/VII. Classical world 91: 369–385. Kristiansen K. 2010. The Nebra find and early Indo-Euro­pean religion. In H. Meller, Bertemes F. (eds.), Der Griff nach den Sternen. Wie Europas Eliten zu Macht und Reichtum kamen, Tagungen des Landesmuseum für Federico Bernardini, Giacomo Vinci, Vanja Macovaz, Andrea Baucon, Angelo De Min, Stefano Furlani, and Sne/ana Smolic´ Vorgeschichte Halle (Saale), Band 5. Verlag Beier & Be-ran. Langenweißbach: 431–439. 2012. Rock art and religion — the sun journey in Indo-European mythology and Bronze Age rock art. Adoran-ten: 69–86. Kristiansen K., Allentoft M. E., Frei K. M., +8 authors, and Willerslev E. 2017. Re-theorising mobility and the forma­tion of culture and language among the Corded Ware. An­tiquity 91(356): 334–347. https://doi.org/10.15184/aqy.2017.17 Luke C., Roosevelt C. H. 2017. Cup-Marks and Citadels: Evidence for Libation in 2nd Millennium B.C.E. Western Anatolia. Bulletin of the American Schools of Oriental Research 378: 1–23. https://doi.org/10.5615/bullamerschoorie.378.0001 Marchesetti C. 1903. I Castellieri preistorici di Trieste e della regione Giulia. Il Museo civico di storia naturale, Trieste. Maselli Scotti F. 1982. Cattinara e i castellieri triestini (sca-vi 1977–79). In Ritrovamenti archeologici recenti e re-centissimi nel Friuli Venezia Giulia, Relazioni della So-printendenza per i Beni Ambientali, Architettonici, Ar-cheologici, Artistici e Storici del Friuli-Venezia Giulia 2. Lint. Trieste: 31–35. 1983. Le strutture dei castellieri di Monrupino e Rupin-piccolo (Trieste). In Preistoria del Caput Adriae. Istitu-to per L’enciclopedia del Friuli Venezia Giulia. Udine: 214–216. 1988. Il castelliere di Rupinpiccolo. Atti e memorie della Societa Istriana di Archeologia e Storia Patria 36: 215–219. Meller H. 2002. Die Himmelsscheibe von Nebra – ein früh­bronzezeitlicher Fund von aussergewöhnlicher Bedeutung. Archäologie in Sachsen-Anhalt 1: 7–20. Menna F., Nocerino E., Remondino F., Dellepiane M., Cal-lieri M., and Scopigno R. 2016. 3D digitization of an heri­tage masterpiece – a critical analysis on quality assess­ment. XXIII ISPRS Congress 41: 675–683. Mielke D. P. 2018. Hittite Fortifications Between Function and Symbolism. In M. Fernández-Götz, D. P. Mielke, and A. Ballmer (eds.), Understanding ancient fortifications: between regionality and connectivity. Oxbow Books. Oxford: 63–82. Mihovili. K. 1996. Nezakcij-nalaz grobnice 1981. godine. Monografije i katalozi 6. Arheolo.ki muzej Istre. Pula. 1997. Fortifikacija gradine Gradac-Turan iznad Koro-ma.na/The fortification of Gradac or Turan hill-fort above Koroma.no. Arheolo.ka istra.ivanja u Istri: Iz­danja Hrvatskog arheolo.kog dru.tva 18: 39–59. 2013. Castellieri-Gradine of the Northern Adriatic. In A. Harding, H. Fokkens (eds.), The Oxford Handbook of the European Bronze Age. Oxford University Press. Oxford: 864–487. 2014. Histri u istri. .eljezno doba Istre. Monografije i katalozi 23. Arheolo.ki muzej Istre. Pula. Mihovili. K., Hänsel B., Mato.evi. D., and Ter.an B. 2012. Burial mounds of the bronze age at Mu.ego near Monko­donja. Results of the excavations 2006–2007. In E. Borg-na, S. Müller-Celka (eds.), Ancestral landscapes: burial mounds in the Copper and Bronze Ages (Central and Eastern Europe – Balkans – Adriatic – Aegean, 4th–2nd millennium BC). Maison de l’Orient et de la Méditerra-née. Lyon: 367–374. Mladin J. 1964. Umjetni.ki spomenici prahistorijskog Nezakcija. Arheolo.ki muzej Istre. Pula. Neve P. 1977–1978. Schalensteine und Schalenfelsen in Bogazkoy-Hattusa. Istanbuler Mitteilungen des Deutschen Archäologischen Instituts 27/28: 61–72. Pásztor E. 2015. Nebra Disk. In C. L. N. Ruggles (ed.), Handbook of Archaeoastronomy and Ethnoastronomy. Springer-Verlag. New York: 1349–1356. Pásztor E., Roslund C. 2007. An interpretation of the Ne-bra Disc. Antiquity 81: 267–278. https://doi.org/10.1017/S0003598X00095168 Pernicka E., Adam J., Borg G., +9 authors, and Reichenber­ger A. 2020. Why the Nebra Sky Disc Dates to the Early Bronze Age. An Overview of the Interdisciplinary Results. Archaeologia Austriaca 104: 89–122. Porter S. T., Huber N., Hoyer C., and Floss H. 2016. Por­table and low-cost solutions to the imaging of Paleolithic art objects: A comparison of photogrammetry and reflec­tance transformation imaging. Journal of Archaeolo­gical Science: Reports 10: 859–863. https://doi.org/10.1016/j.jasrep.2016.07.013 Priuli A. 1977. Le incisioni rupestri carsiche. In L. Ruaro Loseri (ed.), Preistoria del Caput Adriae: ricerche, primi risultati, nuove scoperte, restauri. Comune di Trieste. Trieste: 21–23. Puschi A. 1898. Le Pullari. Atti e Memorie della Societa Istriana di Archeologia e Storia Patria 14: 540–549. Prehistoric stone disks from entrances and cemeteries of north-eastern Adriatic hillforts Schlosser W. 2002. Zur astronomischen Deutung der Him-melsscheibe von Nebra. Archäologie in Sachen-Anhalt 1: 21–23. Ti.ljar J., Vlahovi. I., Veli. I., Mati.ec D., and Robson J. 1998. Carbonate Facies Evolution from the Late Albian to Middle Cenomanian in Southern Istria (Croatia): Influence of Synsedimentary Tectonics and Extensive Organic Car­bonate Production. Facies 38: 137–152. Urban T. 1993. Studien zur Mittleren Bronzezeit in Nord­italien. Universitätsforschungen zur prähistorischen Ar-chäologie 14. Universität Berlin. Bonn. Vergne R., Pacanowski R., Barla P., Granier X., and Reuter P. 2012. Enhancing surface features with the Radiance Scaling Meshlab Plugin. In Proceedings of the Computer Applications and Quantitative Methods in Archaeology (CAA) 2012. Amsterdam University Press. Amsterdam: 417–421. Verma A., Bourke M. C. 2019. A method based on struc­ture-from-motion photogrammetry to generate sub-milli­metre-resolution digital elevation models for investigating rock breakdown features. Earth Surface Dynamics 7: 45– 66. https://doi.org/10.5194/esurf-7-45-2019 Vinci G., Bernardini F. 2017. Reconstructing the protohi­storic landscape of Trieste Karst (north-eastern Italy) through airborne LiDAR remote sensing. Journal of Ar­chaeological Science: Reports 12: 591–600. https://doi.org/10.1016/j.jasrep.2017.03.005 Vitasovi. A. 2002. Gradina. Histria Archaeologica 31: 5–60. 2005. La cultura dei castellieri sulle isole Brioni. In G. Bandelli, E. Montagnari Kokelj (eds.), Carlo Marcheset-ti e i castellieri, 1903–2003. Editreg. Trieste: 301–324. West M. 2007. Indo-European Poetry and Myth. Oxford University Press. Oxford. Whittaker H. 2002. Minoan Board Games: The Function and Meaning of Stones with Depressions (So-called Kernoi) from Bronze Age Crete. Aegean Archaeology 6: 73–87. . Appendix Supporting information: Gradina stone disk 1 ported by the morphological correspondence with the boring Gastrochaenolites, which is produced by Both frontal and opposite faces show at their centre marine bivalves (Donovan, Hensley 2006). In fact, an artificial cup mark about 6cm wide but only the Gastrochaenolites is a clavate boring that is pre-frontal one is covered with a network of slender pits served as a hemispherical depression when truncat-(width: about 1–3mm) arranged in sub-linear sets ed by physical and/or biological erosion (Domenech (length: about 1–3cm; Fig. 6). This morphology is et al. 2001). In the Mediterranean Sea, the bivalve characteristic of the sponge boring Entobia cateni-Rocellaria (Gastrochaena) dubia produces incipi­formis, consisting of cylindrical chambers arranged ent Gastrochaenolites, whereas the date mussel Li-in coalescing sub-linear chains (Farber et al. 2016). thophaga lithophaga is another important bivalve Clionid sponges produce incipient Entobia in the Ad-bioeroder (Casolia et al. 2016). riatic Sea (Bromley, D’Alessandro 1989). Entobia is a proxy for marine rocky substrates (Bromley, D’Ales-The association between Gastrochaenolites and En-sandro 1989; Domenech et al. 2001). tobia typically arises from long-term bioerosion, such as occurs on sediment-free submarine cliffs (Entobia On Gradina stone disk 1, Entobia crosscuts larger, ichnofacies sensu Bromley and Asgaard 1993). The shallow hemispherical depressions (diameter: about Gradina borings are unfilled and therefore might 0.5–3cm). Anton Vitasovi. (2005) considered these have been produced few decades before being col-hemispherical depressions as artificial cup marks, lected by humans. This hypothesis is supported by but some of them are cut by the central cup mark or the relatively fast bioerosion rates observed in the interrupted at the edge of the stone, implying they Mediterranean Sea, i.e. Entobia cateniformis can be were already present when the disk was shaped. In produced in two years whereas Gastrochaenolites addition, the observed cross-cutting relationships requires longer (Domenech et al. 2001; Farber et suggest that they were produced in marine environ-al. 2016). The depth of penetration (tiering) of En-ments as well as Entobia. This hypothesis is sup-tobia cateniformis is usually restricted to the first Federico Bernardini, Giacomo Vinci, Vanja Macovaz, Andrea Baucon, Angelo De Min, Stefano Furlani, and Sne/ana Smolic´ centimetre within the substrate (Farber et al. 2016). A similar explanation is provided by Rosa Dome-Therefore, the Entobia from Gradina could have nech et al. (2001) for Entobia-Gastrochaenolites as-been produced a short time after Gastrochaenolites. semblages of Spain. . References Bromley R. G., D’Alessandro A. 1989. Ichnological study of shallow marine endolithic sponges from the Italian Coast. Rivista Italiana di Paleontologia e Stratigrafia 95: 279–314. Bromley R. G., Asgaard U. 1993. Two bioerosion ichnofa­cies produced by early and late burial associated with sea-level change. Geologische Rundschau 82: 276–280. Casolia E., Ricci S., Antonelli F., Belluscio C. S. P. A., and Ardizzone G. 2016. Impact and colonization dynamics of the bivalve Rocellaria dubia on limestone experimental panels in the submerged Roman city of Baiae (Naples, Italy). International Biodeterioration and Biodegrada­tion 108: 9–15. https://doi.org/10.1016/j.ibiod.2015.11.026 Domenech R., De Gibert J. M., and Martinell J. 2001. Ichno-logical features of a marine transgression: Middle Miocene rocky shores of Tarragona, Spain. Geobios 34: 99–107. https://doi.org/10.1016/S0016-6995(01)80051-6 Donovan S. K., Hensley C. 2006. Gastrochaenolites Ley­merie in the Cenozoic of the Antillean Region. Ichnos 13: 11–19. https://doi.org/10.1080/10420940500511629 Farber C. et al. 2016. Long-term macrobioerosion in the Mediterranean Sea assessed by micro-computed tomog­raphy. Biogeosciences 13: 3461–3474. https://doi.org/10.5194/bg-13-3461-2016 Supporting Fig. 1. Quarrying sites next to the Rupinpiccolo rampart. a Quarrying site east of the north­eastern gate of the rampart with short and linear chisel marks (black lines) and a wedge hole (dotted black lines). b Quarrying site next to the corridor-like structure in front of the south-eastern entrance. Prehistoric stone disks from entrances and cemeteries of north-eastern Adriatic hillforts Supporting Fig. 2. Rupinpiccolo rampart block with two symmetrical half-wedge holes and a few chisel marks. The southern half-wedge hole and close chisel marks are shown by black arrows and lines; the northern half-wedge hole is shown by white arrows and lines. Supporting Fig. 3. Block belonging to one of the platforms of the Bronze Age Vr.in cemetery in Istria with clear chisel marks. a Position of the block (in light red) in the cemetery; plan taken from Battaglia (1958). b Radiance scaling and lattice visualizations of the block. c Lattice visualization of the block with chisel marks indicated by black arrows. back to contents Documenta Praehistorica XLIX (2022) Analysis of Neolithic pottery technology along the Iranian Zagros foothills Natalia Petrova1, Hojjat Darabi 2 petrovany@iaran.ru 1 Institute of Archaeology, Russian Academy of Science, Moscow, RU 2 Department of Archaeology and Research Center for Art and Architecture, Razi University, Kermanshah, IR< h.darabi@razi.ac.ir ABSTRACT – The article presents the results of a technological analysis of the ceramic samples from Neolithic settlements of Ali Kosh, Mahtaj and Guran (the 7th mill. BC). The use of sheep and goat dung in the paste prevailed. While two-layer slabs were applied as the main construction method across the region, a few samples from Guran show the appearance of coil construction around the middle of the 7th millennium BC. First an overall coating with the same clay and red colouring ap­peared, and later a new type of red slip emerged – a mixture of clay with red pigment. KEY WORDS – Zagros; Neolithic; pottery technology; paste composition; construction methods; slip Analiza tehnologije neolitske lon;enine iz predgorja Zagrosa v Iranu IZVLE.EK – V .lanku predstavljamo rezultate tehnolo.ke analize kerami.nih vzorcev iz neolitskih naselbin Ali Kosh, Mahtaj in Guran (7. tiso.letje pr. n. .t). V lon.arski masi prevladujejo ov.ji in koz­ji iztrebki. V regiji pri izdelovanju posod prevladuje metoda dvojnih glinenih trakov, vendar se v Gu­ranu sredi 7. tiso.letja pr. n. .t. uporablja tudi tehnika svaljkov. Posode so bile na za.etku v celoti premazane z istim premazom rde.e barve. Kasneje se je pojavila nova vrsta premaza – me.anica gline in rde.ega pigmenta. KLJU.NE BESEDE – Zagros; neolitik; lon.arska tehnologija; sestava lon.arske mase; oblikovanje po-sod; premaz Introduction The emergence of pottery has long been seen as an large storage containers, sometimes attached to the important technological innovation in human life. wall of buildings, and samples from much smaller Across west Asia, the overall archaeological evi-ones. They were mostly found in the burnt deposits dence suggests that the first unfired clay vessels known as layer D at the site (Smith 1974; 1990). appeared at the turn of the 9th to 8th millennia BC, This highlights the fact that Neolithic communities as shown by the Ganj Dareh materials (Darabi et were long dealing with such unfired clay contain-al. 2019) while the earliest pottery vessels appear-ers during the pre-pottery period as a direct pre-ed at around 7000 BC (see Le Miere 2017; Le Miere, decessor to the fired pottery vessels in the 7th mil-Picon 1998; Nieuwenhuyse, Campbell 2017; Tsu-lennium BC. In addition, this important technolog­neki 2017). In the Zagros region, excavations at Ganj ical innovation should have been influenced by Dareh yielded two types of clay vessels, including some other preceding items, such as stone vessels, DOI> 10.4312\dp.49.11 Analysis of Neolithic pottery technology along the Iranian Zagros foothills white ware, and waterproofed mat containers, as well as pyro-technological experiments with making clay objects. In this regard, we also assume a tech­nological correlation between the construction of cob walls and subsequent pottery vessels. Along the Zagros foothills and intermountain valleys, the ear­liest available pottery samples are dated to the turn of the 8th to 7th millennia BC (Darabi 2018). How­ever, they were regionally variated, although later inter-regional interactions led to some stylistic simi­larity or uniformity. When it comes to Neolithic pot­tery in the Zagros (Bernbeck 2017; Hole 2018; Mat­thews, Fazeli Nashli 2022.89), the majority of scho­larship has dealt with stylistic consideration of var­ious types, in particular their form and decorative elements, while the composition of the pottery paste, construction methods, and firing of the early ceramics remain poorly understood. Previously, exa­minations by Frederick R. Matson (1960) and Pame­la Vandiver (1987) presented some information on the Iraqi and Iranian Zagros, respectively. The Zag-ros region is formed of high intermountain valleys and plains or foothills at lower altitudes. This spec­tacular geomorphological feature has always played a major role in human life in the area, enabling the coexistence of local, regional and interregional cul­tural facts resulting from socio-economic interac­tions. Frank Hole (2018) recently pointed out the ‘diversity’ and local development of various types of Neolithic ceramics across the Zagros piedmont, in Deh Luran, Susiana, Hulailan, Mahidasht Fars. He thus refers to these internal ceramic trajectories as ‘creative centuries’. According to Hole, due to the for­midable Zagros heights the nearby lowlands, such as the Deh Luran Plain, and intermountain valleys or plains, such as Hulailan, show distinct ceramic trajectories during the Neolithic period. This claim can be assessed through investigating of a large body of various artefacts, including ceramics. This article presents a comprehensive analysis of Neolithic pottery technology across the Iranian Zag-ros, with a focus on the samples recovered from three Neolithic sites of Ali Kosh, Mahtaj and Guran. The first two sites are located in the two corners of the lowlands of southwestern Iran, while the last lies at a small, closed intermountain valley, Hulailan, in the central Zagros. Such distinct natural settings may provide us with a better comparison of the Neo­lithic ceramics in the light of technological, not sty­listic, matters. Moreover, both Ali Kosh and Guran represent the most common Neolithic ceramic types, which are ubiquitous on the lowlands and high­lands, respectively (Fig. 1). Materials and their chronology As noted above, the pottery assemblages examined and presented here are some selected samples from the Neolithic sites of Ali Kosh and Guran, as well as few sherds from Mahtaj. All three sites contain both pre-pottery and pottery levels, though the last one lacks any in situ samples due to anthropogenic de­struction (see below). As a result of the stratigraphic excavation in 2017, a total of 227 pottery sherds were found from the upper levels of Ali Kosh, Deh Luran Plain (32°33’ 28.14”N, 47°19’29.82”E, about 6km to the north­west of Mousian town, Ilam Province) (Darabi 2018; Darabi et al. 2017). Following the early 1960s ex­cavations, these levels were previously defined as the Mohammad Ja’far Phase in which three pottery types were recovered: Ja’far Plain, Ja’far Painted, Khazineh Red (Hole et al. 1969.113). Accordingly, these are mostly chaff-tempered and burnished. Judging from the cracking and peeling of surface, the two former types (Ja’far Plain and Ja’far Painted) seem to have been covered with a “wash of the same clay” (Hole et al. 1969.115,117). In relation to Khazineh Red, both ‘slip’ (“as a fine solution of well-cleaned clay”) and ‘self-slip’ (formed during wet smoothing of the vessels) were applied. Only Ja’far Painted was decorated in geometric designs, such as chevrons, zig-zags and checkerboard (for fur­ther information see Hole et al. 1969.113–124). Technological analysis was carried out on 31 frag­ments of ceramics (for basic information on the sam­ples see Table 1 – the number and other informa­tion of the fragments associated with their figures is given by site in the corresponding tables), including 13 samples of the Ja’far Plain type, 12 samples of the Ja’far Painted type, and six samples of the Kha­zineh Red type. It’s noteworthy that the last is not uniform in appearance, with a variation in the red colour on the outer surface. The Ja’far Plain type has a wall thickness between 0.6–1.2cm, predominantly 1.0–1.1cm. The thickness of the walls of Ja’far Paint­ed and Khazineh Red is mostly less than 1cm (0.7– 0.9cm). The diameter of the vessel rims of all types ranges from 15 to 25cm. Chronologically, Hole (1987) attributed the pottery layers of Ali Kosh to 6300–6000 BC and the preced­ing pre-pottery layers (the Boz Mordeh and Ali Kosh phases) to c. 7500–6300 BC. Melinda A. Zeder (1999; 2008) dated the entire sequence to c. 7500–7000 BC. However, recent determinations placed the site within c. 7500–6500 BC, and proposed that the pot­ Natalia Petrova, Hojjat Darabi tery emerged from roughly 7000 BC onwards (Da-rabi 2018). During the sounding at Tapeh Mahtaj, Behbahan Plain (30°38’7.64”N, 50°12’15.33”E, about 3km to the west of Behbahan city, Khuzestan Province), a few pottery fragments were recovered from the site. However, they were all intrusions from upper, de­stroyed levels. The site presents traces of some sub­sequent seasonal occupations spanning from the late 8th to early 7th millennia BC (Darabi et al. 2017; 2021). If the upper levels had not been destroyed, the site could have helped with better understand­ing of the transition from the pre-pottery to pottery Neolithic in southwestern Iran. However, the two samples that were analysed for this research should be attributed to the early 7th millennium BC. They represent fragments of a vessel base, and provide adequate technological information (for basic infor­mation on the samples, see Table 2). Our third assemblage comes from surface of Tapeh Guran (33°45’23.83”N, 47°05’ 51.90”E, about 4km to the west/ northwest of Tohid town in the Hulailan valley, Ilam Province). Diachronically, the site present­ed various types of Neolithic pottery in the central Zagros: Greyish-brown ware (level S), Buff ware (levels R-D), Archaic Painted (levels R-F), Standard Painted ware (levels O-D), Sa­rab Geometric (level L-D)1 , and Red-slipped ware (Mortensen 1972; 2014; Meldgaard et al. 1964.116–117) (Fig. 2). Greyish-brown ware is the ear­liest type made of untempered pure clay, but later tiny pieces of chaff and husk appeared as inclusions. It is also wet-smooth­ed or slightly burnished, with incised lines and crescentic im­pressions from fingernails or the end of a bone tool. Variated grey and brown surface colours are the result of ‘poor firing’ (Mortensen 2014.50). The oc­currence of such pottery is con­troversial, as it was also reported from Ganj Dareh (see Smith 1976), a site that has recently been dated to the pre-pottery Neolithic period (see Zeder 2008; Meiklejohn et al. 2017; Darabi et al. 2019). As the most common type at the site, Buff ware contains some limestone and sand, as natural inclusions of the clay, and tempered with tiny fragments of chaff in medium or large quantities. Some samples were reported to have contained dung as temper. The sur­face of the Buff ware is wet smoothed or slipped, often slightly burnished and ranges from buff to orange-buff. Standard Painted ware and Sarab Geo­metric style are tempered with tiny pieces of chaff, grits of limestone and small particles of sand, some­times naturally added to the clay. They are also slipped and usually burnished, varying from buff to orange buff to reddish colour. In this regard, the in­ner layer is buff; the outer surface is slipped (light orange to red in colour) and usually burnished. The decoration is also painted in red ochre; though the Standard Painted has bobbled lines (tadpoles) and Sarab Geometric style is decorated with elements 1 This type was later re-classified by the site excavator as a sub-type (group d) of the Standard Painted (see Mortensen 2014.59), though they are obviously distinct in terms of their decorative elements. Analysis of Neolithic pottery technology along the Iranian Zagros foothills such as chevrons, zig-zags and bands. Archaic Painted ware is usually burnished and heavily tempered with chaff and decorated with groups of vertical, oblique or horizontal lines. Red-slipped ware is also medium-heavily tempered with chaff and small particles of sand. Both outer and inner surfaces of the ware are slipped and burnished. The outer, and sometimes inner, surface is covered with a medium-dark red to orange-buff slip. In most cases, a grey or black layer is seen inside the fragments, which results from ‘in­sufficient firing’ (Mortensen 2014.50–66). Peder Mortensen (2014.17) placed the site within c. 6700–5500 BC while Zeder (2008) suggested a time spanning c. 7300–6000 BC. Judging from a regional perspective, the latter seems to be more realistic. With regard to the current research, 22 potsherds were analysed. The available assemblage consisted of various types, with Standard Painted ware including the Charmo (Jarmo) style (one sample) and Guran style (one sample), Sarab Geometric (one sample), Buff ware (four samples) and Red-slipped ware (15 samples), while no samples associated with the Gre­yish-brown ware and Archaic Painted ware are in­vestigated, as they were stratigraphically limited to the lower levels and hardly found on the surface (for basic information on the samples, see Table 3). Methods The ceramic technological analysis included the ana­lysis of ceramics in terms of raw materials, pottery paste, and methods of construction, surface treat­ment and firing to study the stages of pottery tech­nology. The technique applied is based on a binocu­lar microscopy examination of technological traces on the surfaces and in fresh cross-sections of cera­mic fragments,2 as well as experimental modelling of individual elements of pottery technology to ve­rify issues that arose in the microscopic analysis (Bobrinsky 1978; 1999). Raw materials In order to reconstruct the knowledge of the potter about the kind of clay to be used for making a de­sired pottery, the clay ferrugination, qualitative com­position of natural inclusions, their dimensions and concentrations were recorded. The clay ferrugina­tion was determined by re-firing samples in a muf­fle furnace at a standard temperature of 850°C. The concentration of natural inclusions was also mea­sured in comparison with the special tables, previ­ously obtained as a result of the data of numerous experiments (Bobrinsky 1999.35–40). Based on the presence of natural sandy inclusions, the used clay can be divided into low sandy, medium sandy and high sandy groups. Low sandy clay contains single grains that are mostly fine (0.1–0.25mm) and some­times medium-grained sand (0.25–0.5mm) in a con­centration of 1:10 (sand:clay). High sandy clay con­tains very fine (0.05–0.1mm) and fine sand grains (0.1–0.2mm) in a concentration 1:5–1:1 (sand:clay). The larger sand grain inclusions are usually rare (Lo-patina, Kazdym 2010). Fig. 2. Archaeological layers at Tapeh Guran showing their related pottery styles (modified by H. Dara-bi after Meldgaard et al. 1963.109, Fig. 9). 2 Carl Zeiss 2000-C stereo microscope. Natalia Petrova, Hojjat Darabi Pottery paste The qualitative composition of intentionally added inclusions, their dimensions and concentrations were determined. Archaeological samples were compared with experimental samples containing various types of organic inclusions: fresh grass, hay, straw and dung of cattle, sheep and goats (Bobrinsky 1999.18– 19,32–33,41–44,86; London 1981; Rice 1987.82; Tsetlin 2003). The dung in pottery paste is repre­sented by prints of small plants and sometimes the remaining residues, 0.2–1mm wide and up to seve­ral centimetres long, with pointed or rounded tips. In particular, there are a lot of highly degraded par­ticles (0.1–0.2mm wide, less than 1mm long) in the dung of sheep and goats. In case of using fresh dung a strong curvature of small plants is seen in the pot­tery paste. Experiments show that an admixture of straw, crushed by any type of grinding or by impact, has uneven stepped edges (along the fibres), along with the splitting of the stems along the fibres and the falling of individual plant fibres. Only when cut­ting is an even cut fixed without splitting into indi­vidual fibres. Plant residues (imprints) are not bent and do not break at the bend. The concentration of organic inclusions was examined in comparison with experimental samples (Petrova 2012.78; 2019; in preparation). Construction methods The construction methods were examined on the basis of studying the vertical and horizontal cross-sections of ceramic fragments, which would attest to the presence of junctures at the places of joints of sequential “clay building elements” and the orienta­tion of pores (Bobrinsky 1978.174–184; Albero San-tacreu 2014.78; Roux,Courty 2019.164– 166; Rue 1981; Shepard 1956.184; Vandiver 1987.30–31). Analysis of the shape, size and direction of the junc­tures in the sherd allows us to find out from which sequential elements (slabs, coils, bands) the vessel was made. There are two known methods for gluing slabs: unsystematically and along circular horizon­tal zones. It is possible to assess the construction me­thod by the presence of traces of some action that occured during the forming sequence that were left unsmoothed, changes in the relief and thickness of the vessel walls at the places of joints of sequential elements on the outer and inner surfaces of the cera­mic fragments, and the presence of a mould con­nected with slabs (Bobrinsky 1978; Rice 1987.125; Vasil’eva, Salugina 2010.72–87). With regard to the slab construction, the vertical and horizontal cross-section is divided by junctures into many separate parts. The coil construction can be detected by the extended horizontal line of juncture at the horizon­tal cross-section of the vessel wall. In a vertical cross-section, in the case of coils, the wall is divided into many separate parts by horizontal or oblique junc­tures (Fig. 3). Surface treatment The surface treatment is assessed through analysing the micro-traces left on the surface. To verify the presence of a clay covering, we conducted experi­ments using different clays to make a basic paste and other types with the addition of various pig­ments. A full study of all the features of the clay coating is still ongoing. At present, however, it can be noted that at least in the case of applying an ad­ditional clay coating (including the slip) before fir­ing, characteristic rounded (micro) cracks and chips appeared on the surface of the vessels, as noted in other studies (Rue 1981.41,54; Shepard 1956.67). This resulted from uneven shrinkage of the clay that was used in the basic paste composition and coating. With regard to the clay covering, we need to clarify the concepts used in our research. We divided the concept of ‘slip’ into two types: “coating with the same clay” – a type of very thin or watery clay simi­lar to the main raw material from which the vessel was made without using additional admixtures; and ‘slip’ – the application of clay composed of the same or different clay material mixed with a pigment. This is necessary to show the development of idea of the ‘slip covering’, which will be shown below. In addition, the traces left by the tools used for smo­othing the surface of the vessels were studied and compared with our experimental observations. Firing The firing regime, its duration and temperature are determined based on the thickness of the oxidized and un-oxidized layers and the quality type of the transition of margin between them (sharp or gradu­al) in the cross-section, as well as changes in the form of intentionally added or naturally occurred inclu­sions (Bobrinsky 1999.93–95; Rue 1981.118; Vol-kova, Tsetlin 2016). Analyses Clay raw materials With regard to the Ali Kosh ceramics, a wet clay with varying degrees of ferrugination was used: medium (19 fragments), high (12 fragments), and low (one fragment). In all the three pottery types recovered Analysis of Neolithic pottery technology along the Iranian Zagros foothills from the site, the clay contained an insignificant concentration of sand (no more than 1:10), which is a natural rounded admixture of sand (with a par­ticle size of 0.2–0.5mm) and a fine admixture of limestone (0.2–0.5mm) in an even lower concentra­tion. The samples of Mahtaj are made of low-ferrugi­nated, low sandy clay (sand particle size 0.2–0.5mm, concentration less than 1:10). At Guran, medium (14 samples) or low ferruginous (eight samples) and low sandy clays are seen. Red-slipped ware, in the overwhelming majority of cases, shows the use of medium ferruginous clay (13 samples), while the Buff ware was only made of low ferruginous clay. In all the samples the clay contains very fine natural sand (0.2–0.5mm) at a concentration of less than 1:10, and there are sporadic larger grains of sand. Limestone was also found in only two fragments in insignificant amounts. Pottery paste An examination of the paste of the Ali Kosh ceram­ics showed different scales of organic prints and re­sidues: very small plants (0.1–0.2mm wide, L1mm long) remaining from sheep and goat dung with pointed (needle shape) (Fig. 4.a.23) or rounded ends (Fig. 4.a.2) in wet condition, as evidenced by curved long (Fig. 4.a.15) and very small compressed plant prints (Fig. 4.a.10); coarse plant residues (L 0.5mm wide, often L1mm long) in a dry state with straight (cut) ends characteristic of me­chanical crushing (Fig. 4.b.12). The pre­sence of grain husk residues is also pos­sible (Fig. 4.b.6). It is not clear whether coarse plant admixture was added in­tentionally or was associated with the dung pellets (like the remnants of un­digested fodder, or occurred accidently when the dung was collected or picked up from the ground). However, we may assume that in low concentrations (5– 10%) large plant inclusions resulted from dung, while in higher amounts (.30%) they were added intentionally. The presence of husks can be indicative of adding chaff to the pottery paste, but it is not clear as not enough evidence is available yet. Ja’far Plain ceramics present the high­est concentration of organic admixture – approximately 50% of the volume of the pottery paste – which was recorded in two fragments recovered from the lowest layer of the phase (in one case only coarse plants, and in the other a mixture of dung and larger plants, probably added separately from the dung). A fragment containing only dung was also found in the same layer. Upper layers yield­ed samples that show only dung or plant inclusions. However, the concentration of organic impurities decreased over time. In terms of Ja’far Painted, ex­cept for one fragment with only coarse plant inclu­sions all the ceramics presented a pure admixture of dung. In general, the concentration of organic admixture in this type of ceramic is less than that in the Ja’far Plain samples. It seems that its amount remained stable, as represented by two values: 10 and 30%. In the Khazineh Red samples the pres­ence of dung is ubiquitous. Both Khazineh Red and Ja’far Painted types show a significant amount – up to 30% – in the lower layers, while only an impuri­ty of dung in a small concentration – up to 10% – is seen in the upper layers. In the remaining two pie­ces of Khazineh Red, dung forms approximately one-third of the total volume of the pottery paste. In general, we recorded organic admixtures in the cera­mics of Ali Kosh which may have resulted from the use of sheep and goats dung and especially crush­ed, coarse plants (possibly chaff). The highest densi­ty of organic impurities (50%) is seen in the Ja’far ~ Plain samples that were recovered from the lowest layer, where dung was also deployed as temper, whether added to the plant inclusions or specifical­ Natalia Petrova, Hojjat Darabi Fig. 4. The composition of pottery paste at Ali Kosh, Mahtaj and Guran (photos made by N. Petrova). a Ali Kosh, plant prints, dung-related: 23 general view of dung (pointed ends of prints), 15 curved plant prints, 10 compressed plant prints, 2 rounded end of plant print; b Ali Kosh, plant prints, not related to dung: 6 probably husk imprint; 12 straw imprint with straight (cut) ends; c Mahtaj, plant prints, dung-related. 1a,b general view; d Guran, plant prints, dung-related. 6,20a,b general view, 20c shell chip, presumably dung-related. The number of fragments here and below are given by number in the corresponding (by sites) tables (see Tabs. 1–3). Analysis of Neolithic pottery technology along the Iranian Zagros foothills ly used. In the later layers, ceramic fragments with only dung or large plants or a mixture of both com­ponents in the pottery paste are common. The lat­ter two types of admixtures dominated the Ja’far Plain ceramics. There is a general tendency toward decreasing the concentration of organic temper over time: from 30 to 50% in the lower layers to 10–30% in the upper ones. In the composition of the samples selected from Mahtaj both dung in a wet and dried state and coarse plant imprints and residues were recorded in ap­proximately the same concentrations. In general, the organic admixture to the clay makes up to at least 50% of the total volume of the pottery paste (Fig. 4.c.1a,b). Regarding Guran, the addition of dung as temper is seen in all the ceramic types (Fig. 4.d.20a,b). The admixture of dung is represented by the predomi­nance of very small (<0.1–0.5mm wide) plant im­prints with pointed (needle) ends in the cross-sec­tions (Fig. 4.d.6). In addition, a fragment of a mol­lusc shell (0.3mm long) was encountered, which is also most likely connected with dung. The shell of the mollusc is highly thinned and transparent. On its surface, there is absolutely no surface pattern ty­pical for the outer layer of the shell, possibly indi­cating only the inner, pearlescent layer. In this re­gard, it can be assumed that it had gone through a process of digestion (Fig. 4.d.20c). The dung was added both in lightly wet (Fig. 4.d. 20a) and dry (Fig. 4.d.6) condition, as shown by the form of plant prints (lightly curved or straight). Sometimes larger plant prints (>1mm wide) are also seen, but in low concentration. Their presence is also most likely connected with dung. In most cases the dung of goats or sheep was added (Fig. 4.d. 20a,b). The concentration of dung varies between 10–50%. In the earlier sub-type of Standard Zagros – Jarmo style – dung consisted of around 50% of the total pottery paste, while in the Guran and Sarab Geometric styles this decreased to about 30%. In the Buff ware, dung mostly presented in a concen­tration of about 30% (four samples), but in one case it was 10–15%. In the Red-slipped ware it varied much more, from 10 to 50% (Fig. 4). Forming techniques The features of the sequential construction of clay building elements were recorded in the cross-section of the Ali Kosh samples. Owing to the very small size of the available fragments, there is not enough in­formation to distinguish the details by ceramic type. In most cases, slab construction was used in making vessels (Fig. 5). The clay slabs were joined sequen­tially along a horizontal circular path. This can be seen in only one large example of Ja’far Painted type (Fig. 5.a.13a) showing horizontal zones bounded by deepened lines. In the vertical and horizontal cross-sections joints of these zones are visible. They divide this horizontal zone into elements (slabs) (Fig. 5.a.13b,c). Layering of slabs divided horizon­tal and vertical cross-sections of sherds into two parts, inside of which joins are located at a short distance from each other (Fig. 5.a,b,c). In all cases, the clay slabs were elongated (approx. 2 x 3cm in size) and placed horizontally. In most of the ceramic fragments, there is a two-layer sequential slab con­struction. Two-layer slab construction is also visible in the cross-sections of the two bases from Mahtaj in the way that slightly-deformed and elongated slabs are evident (1.2cm height, 3cm wide, 4.2cm long) (Fig. 6). Sequential slab construction is also present in all the analysed samples from Guran (Fig. 7). The Jarmo style ceramics are made of two-layer slabs and, pos­sibly, the same can be assigned to the fragment of the Guran style vessel (Fig. 7.a.1). In most of the Red-slipped samples two-layer slabs were also used (Fig. 7.a.6,13,14). In the horizontal and vertical cross-sections of the vessel walls, the junctions are located a short distance from each other and at a large angle to the walls of the vessel, creating lay­ering. The slabs have a horizontal elongated shape. However, in two cases we can assume the construc­tion of vessels from coils – in horizontal cross-sec­tion – due to the extended horizontal line of junc­ture, parallel to the vessel walls, in vertical cross-sec­tion, and the division into many separate parts by junctures (Fig. 7.b.4,10). Such a construction method is also seen in three fragments of Buff ware (Fig. 7. b.19). Surface treatment and decoration With regard to the Ja’far Painted (Fig. 8) and Ja’far Plain (Fig. 9) types at Ali Kosh, the outer and inner surfaces in all cases are covered with an additional dense layer, most likely of the same clay as the main raw material but without any organic temper, as suggested by the smooth surface of most fragments without plant imprints. Almost all ceramics with pre­served coating have some loss of the upper layer (for example see Figs. 8.5a,b,6,7b, 8,9; 9.a.16b,18b,20) and cracks (Fig. 9.c.6c), which can be associated Natalia Petrova, Hojjat Darabi with uneven shrinkage of the coating clay and the basic composition of the pottery paste. However, on some ceramic fragments with a destroyed surface this slightly ferruginous clay is visible in large amounts under a layer of clay covering (Fig. 9.a. 14,15,19). In some cases, over the clay, the surface of the vessels with varying degrees of wetness has been smoothed with a hard tool, probably made of bone or pebble (Figs. 8.a.3b; 9.b.25a). The geometric designs, presumably painted with red ochre, are seen on the outer surface of all the Ja’far Painted vessels. Hole also mentioned the use of a spe­cular hematite in one case (see Hole et al. 1969.117). Under magnification, the paint showed a grainy tex­ture, and it was applied unevenly (Fig. 8.a.11b). Fragments of Ja’far Plain are usually covered with the same slightly ferruginous clay (Fig. 9a), but some­times with highly ferruginous clay, as evidenced by characteristic cracks on the surface (Fig. 9b,c). This last case is interesting: when smoothing over a well-dried surface and then firing it in an oxidized at­mosphere, the effect of a ‘reddish colour’ appeared. In one case the vessel was polished with a hard tool. As the result, both surfaces of the vessel acquired a bright red colour (Fig. 9.c.26a-d). In case of Khazineh Red (Fig. 10), two sub-types of surface treatment can be distinguished. First, the surface was covered with an additional clay coat­ ing from the same clay on both surfaces of the ves­ sel. This is usually accompanied with chips of the upper layer (Fig. 10.a.27b,29) and characteristic cracks (Fig. 10.a.30). In two cases, burnishing with a hard tool is recorded on a not completely dried surface (the so-called ‘leather-hard’ condition, when barely visible grooves left by the smoothing tool remain on the surface) (Fig. 10.a.29, 30). In general, this tradition of sur­face treatment is also characteristic of the Ja’far Plain and Ja’far Painted types. Regarding Khazineh Red spe­cifically, however, the outer surface of fragments is completely red or sometimes plum coloured (Fig. 10. a.30). Second, the vessel has a layer of red slip (clay mixed with some red pigment) on both surfaces. The slip is indicated both by cracks on the outer surface (Fig. 10.b.31b) and by a layer, 0.1mm thick (Fig. 10.b. 31c), that is easily distinguishable under a microscope in the cross-sec­tion. Interestingly, in one case while the outer surface of the fragment was only covered with red slip, the inner red-slipped surface was overlaid with plum colouring (Fig. 10.b.31a,d,c), but the outer surface only had a la­yer of coating with the same clay (Fig. 10.b.31d,e). In the case of the samples from Mah­taj, an additional coating with the same clay is distinguishable on the outer and inner surfaces of the ves­sels. Cracks are also visible on both sides. Analysis of Neolithic pottery technology along the Iranian Zagros foothills At Guran, the Standard Painted ceramics, including Charmo (Fig. 11.a.1) and Guran styles (Fig. 11.a.2), Sarab Geometric (Fig. 11.b) and buff wares (Fig. 11.c) were slipped with a clay similar to their paste before firing, as evidenced by characteristic cracks on the surface (Fig. 11. b.3c,19c) and delamination of the coating layer, under which the surface is visible with a high density of organic inclu­sions (Fig. 11.c.22). The buff ware was also burnished in all cases. The ceramics types of the Charmo style, Guran style and Sarab Geometric were de­corated with geometric designs painted in red ochre. Here, Red-slipped ware can also be divid­ed into two sub-types. In most of the available sam­ples, both outer and inner sur­faces (or sometimes just the outer) of the Red-slipped ware (Fig. 12), based on the colour uncharacteristic of natural clay, are covered with a mixture of clay and red pigment before firing (Fig. 12.a). In addition, a variation of brown slip is found (Fig. 12.a.14). The thickness of the slip layer varies (for a comparison see Fig. 12.a.5d – 0.15mm, and Fig. 12.c.17c – 0.05mm) and in all cases it is applied to the unfired sur­face, as evidenced by characteristic cracks (Fig. 12.a.6c) and by surface losses – i.e. rounded chips inside these cracks (Fig. 12.a.5a,b,c,6a, 8b). The slipped surface is usually slightly or highly burnished. From time to time, one can find prints of textile indicating its application either in smoothing or slipping with a textile (Fig. 12.a.11c). In three cases the surface of vessels was cov­ered with the same clay and then fully over­laid by colouring (Fig. 12.b,c,d). Two samples represent a reddish-brown colouring just on the outer surface (Fig. 12.b.15a,16a). In ano­ther case, brown colouring was applied to the outer surface (Fig. 7.d.17a,c,d), while the in­ner surface contains red colouring (Fig. 12.d. 17b,e). In one interesting case, brown paint covered a layer of red slip applied to the outer sur­face of the vessel (Fig. 12.c.18a,c,d), but the inner Natalia Petrova, Hojjat Darabi surface was covered by red slip only. Different va­riants of brown colour can be reached by using iron oxide pigments with firing at higher temperatures (Hole et al. 1969.113), for example magnetite, he­matite and limonite in a different oxidative and re­duction firing atmosphere (Nieuwehuyse et al. 200.158–160). Firing All the Ali Kosh vessels were fired using an oxidiz­ing atmosphere. Two fragments of Khazineh Red pottery are fully oxidized (Fig. 5.d). Others have a dark or light grey core. In half of the cases, the margin of the transitional layer between the outer orange layer and dark core is sharp (this mainly re­fers to the Ja’far Plain ceramics (Fig. 5.c.9a, 16c, 19b,20b), which indicates that the vessel did not cool down in the firing device and was abruptly re­moved after a short exposure to high temperatures. In other cases – mostly Ja’far Painted – the margin of the dark core is gradual and the oxidized layer is much thicker, which may indicate a longer stay in the firing device (Fig. 5.a,b). Judging from the thickness of the oxidized layers and the presence of calcined organic residues in some fragments, the ceramic vessels of Ali Kosh seem to have been fired at a temperature of about 650–700°C (cf. Bobrinsky 1999.99; Rue 1981.118). Our analysis showed that the bases recovered from Mahtaj are completely oxi­dized (see Fig. 6). All the Guran samples were also fired in an oxidizing environment at a temperature not lower than 700–750°C, and the lack of organic inclusions suggests their long exposure to the heat. Thin-walled fragments of Jarmo, Guran and Sarab Geometric styles are all oxidized (Fig. 7.a.1), while just half of other types had the same feature (Fig. 7.a.6,b19), and the rest have a light grey core with gradient margins, although complete oxidation is also recorded in different parts of the fragment (Fig. 6.a.13,14, b4,10). Of the analysed samples only one piece of Red-slipped ware contains a very thin oxidized layer, with a dark grey core and a sharp margin between oxidized and core layers of its cross-section. Discussion: development of Neolithic pottery technology in Zagros The ceramic assemblages analysed for this research were sampled from Neoli­thic sites that are located in both the highlands and lowlands. At the same time, they, and specifically Ali Kosh and Guran, represent the most common Neo­lithic ceramic types in the region. This in­spired us to deploy a deductive approach in order to reach some ‘generalization’, though local criteria should also be given attention. In the light of available evi­dence and based on the current research, however, we may discuss the develop­ment of early pottery technology of the Zagros region with regard to the analy­ses outlined above. Addressing the emergence of pottery tech­nology, it is necessary to draw on Vandi­ver’s assumption highlighting its connec­tion with building technologies (Vandi­ver 1987). In this regard, general clay architectural remains from the pre-pot­tery Neolithic, and specifically those from Ganj Dareh, indicated the presence of a large plant admixture added during the Analysis of Neolithic pottery technology along the Iranian Zagros foothills construction of adobe buildings. At Ganj Dareh, the clay building construction elements are found in two types: chineh (strips) and mud bricks (Smith 1990. 328–332). It is noteworthy that such features were omnipresent at early Neolithic sites such as Ali Kosh, Sheikhi Abad, Mahtaj and Abdul Hosein. Along with the production of clay vessels, stone vessels, baskets and white ware, as well as clay objects, such pise walls reveal both long experimentation and knowl­edge before the appearance of true pottery. From an architectural perspective, pise building and plas­tering or coating walls either with clay or red ochre merits specific attention as a key precursor to pot-tery-making and decorating. The small and large clay containers from Ganj Dareh evidently support this idea. Here, large unbaked storage vessels were in some cases fixed to the floor or attached to the wall. These were large, round, up to one meter in height, storage vessels, and semicircular storage compartments that had one common wall with the dwelling (Smith 1990. 332). Vandiver (1987.25) noted that montmorillonite clays are mainly used in the settlements of the region. Our observations of earlier ceramics of Ali Kosh, Mahtaj and Guran demonstrat­ed the careful and intentional selec­tion of raw materials, and that the clay was used without large mineral impu­rities, including limestone seen in an insignificant concentration and small size. This shows differences compared to Ganj Dareh, where some vessels were made of clay with a large mine­ral admixture (cf. Smith 1990.332). As for intentionally added impurities, the earliest clay vessels of Ganj Dareh are of two types: small samples with­out any visible, specially added tem­per, and oversized items (larger pot­tery, storage vessels, and basins) where plant impurities are recorded (Vandiver 1987.17; 1985.194–195). When it comes to the pottery paste, the putative organic temper is con­sidered for the Neolithic assemblages though various terms like ‘vegetal’, ‘straw’ or ‘chaff’ are also applied. How­ever, the presence of dung was already reported from some sites, such as Charmo, Gird Ali Agha and Tell el-Khan (Adams 1983; Matson 1960. 68), Shimshara (Tauber 1970.143) and Guran (Mortensen 2014.50; also see above). With regard to the samples from Ali Kosh and Mah­taj, the organic temper from the very beginning has a different and multicomponent composition: dung and crushed coarse plant inclusions, both separate­ly and in combination. It can be assumed that the tradition of adding the dung of sheep and goats to the pottery paste spread during the 7th millennium BC. It is noteworthy that dung had also been used as fuel resource in the Zagros region during the pre-pottery Neolithic, as evidenced by micro-morpholo­gical analysis (Matthews et al. 2013; 2016; 2020; Fatui Dilanchi et al. 2020). We assume that dung along with crushed plant admixtures was used by local potters, though a full understanding of the or­ganic material (threshing waste, chaff or some other type of crushed straw) requires further considera­tion. The technological characteristics of the later stage at Guran (from level O onwards) showed the Natalia Petrova, Hojjat Darabi predominance of dung added to the pottery paste. The Neolithic ceramics in the studied region were mostly produced using clay slabs (Fig. 13). Even their precur­sor, namely clay containers from Ganj Dareh, had a simi­lar method of production (Vandiver 1987.18). Slab con­struction in the Zagros region has been reported from a number of sites, such as Sa­rab (Vandiver 1987.18), Cho­gha Sefid (in layers of later Neolithic – Sefid and Surkh phase) (Hole, Tonokie 2021), Sabz (Hole et al. 1969.111–112), Chogha Mish (De-lougaz, Kantor 1984.228), Hajji Firuz (Vandiver 1987.18–19; Voigt 1983.149–152) and in Northern Mesopotamia – Yarim Tepe I, Sotto, Umm Dabaghi-ya of Proto-Hassuna and Hassuna periods (both Ar­chaic and Standard) (Petrova 2019; 2021). Based on our analysis, the ceramics were built up in two-layer slabs in Ali Kosh and Mahtaj. This method is also attributable to the later Guran ceramics from level O onwards, spe­cifically in the cases of tadpole (Char-mo and Guran styles) and Red-slip­ped wares. In connection with the use of se­quential slab construction, the ques­tion of using the mould to which these slabs were stuck arises. On the one hand, the connection with buil­ding technologies suggests free mo-delling, and that the first storage ves­sels were most likely built in this way. On the other hand, some of the vessels made of two-layer slab construction are small and thin-wal­led, which is difficult to do without using a mould and paddling. It is noteworthy that sticking clay slabs onto a base mould is still being used by some women who produce pot­tery vessels in Baneh, Iranian Kurdi­stan (Sedighian, Mahjour 2010.83). According to archaeological materi-56–58). The prints of weaving, which possibly re­mained from a liner that covered the mould (to help separate the future vessel from mould), were found on the ceramics of the later site of Hajji Firuz, north­western Iran, where they are seen inside the bases of the vessels and on the outer surface under a layer of coating (see Voigt 1983.149, Pl. 25). In addition, some samples from Chogha Mish showed the appli- Fig. 11. Guran. Coating with the same clay. als, the use of a mould was previo- a Standard Painted ceramics: 1 Charmo usly supposed for the vessels from style, 2 Guran style; b Sarab Geometric; c Ganj Dareh (Smith, Crepeau 1983. Buff ware. Analysis of Neolithic pottery technology along the Iranian Zagros foothills cation of the same method on their inner sides (Delou­gaz, Kantor 1996.228). The use of a mould in making pot­tery vessels has not yet been reported from other Neolithic sites in Zagros. Hole et al. (1969.111) noted the absence of features of paddling in Ali Kosh, and although such fea­tures are also seen in our as­semblage from the site, they are comparatively smaller in size. However, as indicated by some fragments from Hajji Fi­ruz, a mould or link might have been applied during the Neolithic, though their prints were hidden by subsequent clay coating. At Guran, some samples of Buff and Red-slipped wares were produced using coil con­struction. However, it is dif­ficult at present to conclude whether this method was used from the beginning of ceramic levels of the site, or it was a later tradition that had been brought from out­side. This issue requires fur­ther data. The use of coils was previously recorded from the sites of Hajji Firuz and Ali Agha, both dated to the late 7th and early 6th millennia BC, in northern and north­western/western Zagros, res­pectively (Voigt 1983.149– 152) and in Northern Meso­potamia – Yarim Tepe I, Sotto, Umm Dabaghiya of Proto-Hassuna and Archaic Hassuna periods (Petro­va 2019; 2021). According to ethnographic data, different technolo­gical stages react differently to innovations (Bob­rinsky 1978; Schiffer, Skibo 1987; Stark 1999). At all stages of pottery technology, it is very important to distinguish the construction methods deployed for making the vessels. This technological stage is least susceptible to outside influence, and may indi­cate not only cultural interactions but also the com­position of the population and its changes, in con-trast to the methods of making the pottery paste and surface treatment (Bobrinsky 1978.244; Fow­ler 2017.14; Gosselain 1992.582; 2000.192). The surface of the ceramics from our three sites share a feature, the presence of an additional clay coating (covering by the same clay or a slip). This is also reminiscent of an earlier building construc­tion technique when the walls of the houses were covered with additional clay coating in pre-pottery Neolithic settlements. Clay coatings are noted on the ceramics recovered from various sites such as Ali Kosh (Hole et al. 1969.110), Guran (Mortensen Natalia Petrova, Hojjat Darabi 2014.50–66), Chogha Bonut (Alizadeh 2003.46, 56), Chog-ha Mish (Delougaz, Kanto 1984.227), Chogha Sefid (Ho­le 1977), and Sarab (McDo­nald 1979). In most cases, the coating is similar to the tex­ture of the vessels. Smoothing and burnishing of the surface is ubiquitous dur­ing the Neolithic period. We identified various methods of surface treatment in associa­tion with Ali Kosh and Guran: smoothing the surface hav­ing varying degrees of dry­ness with a hard tool (seen in all ceramic types of Ali Kosh) up to the appearance of the effect of slightly burnishing (in Ja’far Plain) and full bur­nishing (characteristic of Buff ware and Red-slipped ware at Guran). In addition, the sur­face was also smoothed with textile at Guran. Stamped and carved ornaments, specifically punc­tate impressions like carved lines and crescent not­ches produced by a fingernail or the end part of a bone, are seen on the earliest ceramics of Guran and also those from the earlier site of Ganj Dareh (Mor-tensen 2014.50; Smith 1974.207). The nature of such early decorated samples is still ambiguous, as they are documented as vessels made of untemper­ed clay and it is still arguable to what extent they can be taken as the incipient clay/pottery contain­ers. However, this type of ornament was then re­placed by the paint decoration in the region. Soon after, the surface decoration is associated exclusive­ly with the use of pigments such as red ochre, which were either applied for making geometric designs or completely covered the surface. As mentioned above, at Ali Kosh and Guran such pigment was ap­plied over a surface previously covered with a red slip. This suggests an attitude by the early potters to treat slip as a way of covering, but not decorat­ing, the surface. Our analysis showed various meth­ods of achieving the ‘red surface’ of Khazineh Red and Red-slipped ware at both sites. This can be di­vided and diachronically traced. Firstly, a full cov­erage of the surface with only a colouring pigment such as ochre on a covering of the same clay was common. Secondly, it seems that during the process of treatment a red-slipped surface appeared when the covering with the same clay was mixed with red pigment. This method came about over time, as seen from uppermost levels at Ali Kosh, where fully cov­erage of the outer surface with of plum-coloured paint is synchronously seen. All the fragments we studied were fired using an oxidative firing. The vessels from Ali Kosh show the existence of various heat treatments: short or long exposure in the heating zone, depending on the type of ceramics. In this regard, the Ja’far Plain samples were usually fired for a relatively short time, while the Ja’far Painted and Khazineh red ones underwent longer exposure. However, we have some fragments from Ali Kosh of different ceramic types that are fully calcined. At Guran an increase in firing dura­tion for all types is seen over time: in half of the cases the items they are fully calcined; in some other cases (Red-slipped ware and Buff ware) the vessels were fired with long exposure to the heat zone. In general, this may indicate possible improvements in the process of ceramic firing. According to some researchers, dung was an out­standing fuel resource in western Asia from the early Neolithic onwards (see Hesse 1984; Matthews Analysis of Neolithic pottery technology along the Iranian Zagros foothills et al. 2020; Miller 1984; Miller, Marston 2012.97; Spengler 2019). However, in ethnographic research it is cattle dung that has been mostly given atten­tion (Mahjour et al. 2014.25; Matson 1974.345; Pet-rova 2011.135; Salimi 2014.589), while the earli­est evidence for cattle domestication is attributed to the early 6th millennium BC in Zagros (see Arbuc­kle et al. 2016). Thus, the use of animal dung at the earlier settlements such as Ali Kosh and Guran would have been restricted to that of goats and sheep. Ac­cording to Wendy Matthews (2016.116–117) the type of fuel used varied depending on local condi­tions, and dung was widely used as a fuel. In this regard, woods (and specifically oak trees) were more often burned in the highlands (in particular the landscape surrounding Ganj Dareh), while in the foothill-steppe region and on the plains, in par­ticular the Deh Luran Plain where the settlement of Ali Kosh is located, grasses and reeds were used for this purpose (Helbaek 1969.387; Miller 1996.521– 525). However, the presence of dung in the pottery assemblage from Ali Kosh suggests the multi-pur­pose usage of dung. As mentioned before, micromor­phological analysis has also shown presence of ani­mal dung, most likely goat, at the pre-pottery sites of Sheikhi Abad (Matthews et al. 2013) and Mahtaj (Fotuhi Dilanchi et al. 2020) prior to its use in as­sociation with early pottery production, whether as temper or fuel. Our knowledge about firing devices still remains at the level of assumption: firing was carried out either in bonfires, or in hearths coated with clay and built up with mud bricks, or in ovens (Schmandt-Besserat 1974.15; Hole et al. 1969.40, 42). However, excavations at Yarim Tepe I (level 10) have shown the remains of a two-stage pottery kiln dated to the Archaic Hassuna period in Northern Me­sopotamia) (Munchaev, Merpert 1981.75). The pre­sence of such a developed device at the end of the 7th millennium BC allows us to assume the existence of simpler devices at an earlier time, coinciding with the emergence of pottery in the Iranian Zagros foot­hills. Concluding remarks The technological analysis conducted on the ceramic assemblages from the three Neolithic sites shed new light on the nature of early developments of pottery production along the Iranian Zagros. To conclude the current research, we may highlight both simi­larities and differences in terms of construction me­thods, paste, raw materials, firing and surface treat­ments that were applied by the early potters across highlands and lowlands. Of the samples from Ali Kosh, the Ja’far Plain type is somewhat different from the other two types with regard to the techno­logy. The walls are thicker and firing time was shor­ter. Sometimes it contained an abundant admixture of coarse plant residues in the pottery paste, occasio­nally combined with dung. It was previously be­lieved that Ja’far Plain was identical to Ja’far Paint­ed unless the latter was decorated (cf. Hole et al. 1969.117). However, our analysis indicated that the Ja’far Painted type shared some similarities with Khazineh Red, as they both show the predominant use of dung as a temper and a longer firing. The Khazineh Red type represents three variants: the earlier indicated by the samples bearing a red paint fully overlaid with clay slip, and the later represent­ed by red-slipped vessels showing a mixture of clay slip and reddish pigments. However, the third group includes some different cases showing red slip on the outer surface overlaid with full colouring, and on the inner surface with a layer of the same clay covering. We assume that the technology of the Ali Kosh ceramics had definitely passed some stages of development when it emerged at the site (cf. Hole et al. 1969. 352). However, its predecessor still re­mains a controversial issue. The clay was specially selected without any large natural mineral inclu­sions. The pottery paste also has a different and mul­ticomponent composition: crushed plants (possibly, threshing waste – chaff) and the dung of sheep and goats. In the construction stage, two-layer elongated slabs were built up along a circular horizontal path. During surface treatment, there is always a clay coat­ing on the vessels. Various smoothing techniques were applied at the site using a hard tool, with vary­ing degrees of dryness on the surface to be slightly or fully burnished. With regard to firing, there are various approaches: short or long exposure in the high temperature zone, depending on the type of vessel being made. The main technological characteristics of a few pot­tery fragments from Mahtaj are close to those of Ali Kosh in the way that a combination of dung and coarser plant impurities are seen in the paste. More­over, the sequential two-layer slab construction of vessels and coating the surfaces with wash are also notable. Although various ceramic types are present at Gu-ran, they share some technological characteristics: selection of clay without any large amount of natu­ral mineral inclusions, intentional adding of dung into the pottery paste, two-layer slab construction, an intentional covering layer on the vessel (wash or Natalia Petrova, Hojjat Darabi slip) and the presence of significant amount of com­pletely calcined sherds during firing. In addition to the ubiquitous method of slab construction, clay coils were attested for some Red-slipped and Buff ware fragments. However, the emergence of the coil construction method at Guran is as yet unknown either as a locally developed or an imported tradi­tion. To better understand this a study of stratified early materials from the site seems important. As in case of Ali Kosh, at Guran three variants of ‘red surface’ can also be determined in Red-slipped ware: first, a fully red painted surface overlaid with a layer of clay similar to the paste; second, red slip­ped (clay mixed with pigment) surface, which is present in the majority of cases. The third case is re­presented by two very interesting ceramic fragments recovered from Ali Kosh and Guran. The pigment was applied over a surface previously coated with a red slip. The presence of similar items in both set­tlements could synchronize the upper layer(s) of the Ali Kosh settlement and the corresponding layer(s) at Guran, if the ceramics came from stratified exca­vations. However, the existence of this unusual treat­ment at both sites may indicate cultural interactions between highlands and lowlands. Such close inter­actions are mirrored in a majority of comparable ar­chaeological finds, from lithic types to architectural layouts and to pottery style sand construction me­thods, as well as in the spread of obsidian across Zagros. This may undermine the idea that in the Neolithic period contacts between the foothills of the Zagros (Ali Kosh) and its central regions (Ganj Dareh, Guran) were unlikely, because they were se­parated by the ‘formidable’ Kabir Kuh – the longest of the Zagros Mountain ranges (Hole 2018.178). Although such natural barriers could have affected the path and rate of interactions, the role played by the transhumant people who were tracking a verti­cal movement and river valleys, such as Seimarreh, should not be overlooked. In case of Ali Kosh, the emergence of pottery seems to have happened abruptly at the site, an issue sug­gesting that some earlier steps may have been taken somewhere else. In this regard, Abbas Alizadeh (2003) assumed that the so-called phase of ‘Susiana Formation’ at Chogha Bonut was a precursor of the Mohammad Ja’far phase at Ali Kosh. However, this hypothesis is still open, as both sites experienced different pottery styles. Therefore, one may see the pottery Neolithic (c. 7000–6000 BC) as a millenni­um which saw a combination of local creativity and inter-connectedness with surrounding areas. With regard to the initial construction methods, we suppose both free construction of vessels and mak­ing them on a mould. The former is seemingly con­nected with building technologies, while the latter suggests another experience, possibly influenced by coating pits with clay. However, the early ceramics were commonly produced using a slab construction method, though clay coils were sometimes also ap­plied. The origin of the coil construction method in the Zagros region is not yet known. Nevertheless, one may see it as one of the variants in the develop­ment of building techniques or elongated slabs. This idea, of course, requires further investigation. In this study we have brought to light some general information about the technology of the earliest ce­ramics along the Iranian Zagros. However, placing such analyses within a larger region will allow us to better understand the Early Neolithic pottery technology in the Eastern Fertile Crescent. ACKNOWLEDGEMENTS This research benefited from helps and comments of a number of people. In this regard, we wish to ex­press out thanks to Dr. Yuriy Tsetlin, Dr. Sajjad Ali-baigi, Prof. Frank Hole, Hafez Ghaderi, Minoo Sali-mi, Elham Fatuhi, Dr. Shokouh Khusravi, Fereshteh Sharifi and Dr. Hossein Sedighian. Hafez Ghaderi was very helpful with the photography of some sam­ples. We also thank Razi University, Kermanshah, and State Historical Museum, Moscow for their colla­boration and support for this research. Analysis of Neolithic pottery technology along the Iranian Zagros foothills References Adams R. Mc. C. 1983. The Jarmo stone and pottery ves­sel industries. In L. S. Braidwood, R. J. Braidwood, B. Ho­we, Ch. A. Reed, and P. J. Watson (eds.), Prehistoric ar­chaeology along the Zagros flanks. Oriental Institute Publications. Vol 105. University od Chicago Press. Chi­cago: 209–232. Albero Santacreu D. 2014. Materiality, Techniques and Society in Pottery Production: The Technological Study of Archaeological Ceramics through Paste Analysis. De Gruyter Open Poland. Warsaw. https://doi.org/10.2478/9783110410204 Alizadeh A. 2003. Excavations at the prehistoric mound of Chogha Bonut, Khuzestan, Iran. Seasons 1976/77, 1977/78, and 1996. Oriental Institute Publications. Vo­lume 120. Oriental Institute of the University of Chicago. Chicago. Arbuckle B. S., Price M. D., Hongo H., and Oksüz B. 2016. Documenting the initial appearance of domestic cattle in the Eastern Fertile Crescent (northern Iraq and western Iran). Journal of Archaeological Science 72: 1–9. https://doi.org/10.1016/j.jas.2016.05.008 Bernbeck R. 2017. Merging clay and fire: earliest evidence from the Zagros mountains. In A. Tsuneki, O. Nieuwen­huyse, and S. Campbel (eds.), The Emergence of pottery in West Asia. Oxbow Books. Oxford: 97–118. Bobrinsky A. A. 1978. Pottery of Eastern Europe. Sources and methods of studying. Nauka. Moscow. (in Russian) 1999. Pottery technology as an object of historical-and-cultural study. In A. A. Bobrinsky (eds.), Actual prob­lems of ancient pottery investigation. Nauchnaya bib-lioteka Samgpu. Samara: 5–109. (in Russian) Darabi H. 2018. Revisiting stratigraphy of Ali Kosh. Paz­huhesh-hay-e Bastanshenasi Iran. Archaeological Re­search of Iran 16: 27–42. (in Farsi) Darabi H., Aghajari M., Nikzad M., and Bahramiyan S. 2017a. In search of Neolithic Appearance along the North­ern Shorelines of the Persian Gulf: a report on the exca­vation at the pre-pottery Neolithic site of tapeh Mahtaj, Behbahan plain. International Journal of the Society of Iranian Archaeologists 3(5): 13–22. Darabi H., Bahramiyan S., Mostafapour S., and Khademi Bami M. 2017b. Re-excavation at tapeh Ali Kosh, Deh Lu-ran plain, Iran. Neo-Lithics. 2: 15–18. Darabi H., Bansgaard P., Arranz-Otaegui A., Ahadi G., and Olsen J. 2021. Investigating early Neolithic occupation of the lowlands in southwestern Iran: New Evidence from Tapeh Mahtaj. Antiquity 95(379): 27–44. https://doi.org/10.15184/aqy.2020.215 Darabi H., Richter T., and Mortensen P. 2019. Neolithiza­tion process in the central Zagros: Asiab and Ganj Dareh revisited. Documenta Praehistorica 46: 44–56. https://doi.org/10.4312/dp.46.3 Delougaz P., Kantor H. J. 1996. Chogha Mish. Volume I. The first five seasons of excavations 1961–1971. Orien­tal Institute Publications 101. The Oriental Institute. Chi­cago. Fotuhi Dilanchi E., Darabi H., and Heydari Guran S. 2020. A micromorphological analysis of the Neolithic site of Mahtaj, Behbahan Plain. Journal of Research on Archa­eometry 6(1): 81–96. (in Farsi) http://jra-tabriziau.ir/article-1-218-en.html Fowler K. 2017. Ethnography. In A. Hunt (ed.), Oxford Handbook of Archaeological Ceramic Analysis. Oxford University Press. Oxford: 469–486. https://doi.org/10.10 93/oxfordhb/9780199681532.002.0003 Gosselain O. 1992. Technology and Style: Potters and Pot­tery among Bafia of Cameroon. Man (N.S.) 27: 559– 586. 2000. Materializing identities: an African perspective. Journal of Archaeological Method and Theory 7: 187– 217. https://doi.org/10.1023/A:1026558503986 Helbaek H. 1969. Plant collecting, dry-farming and irriga­tion in prehistoric Deh Luran. In F. Hole, K. V. Flannery, and J. A. Neeley (eds.), Prehistory and Human Ecology of the Deh Luran Plain. An Early Village Sequence from Khuzistan, Iran. Memoirs 1. Museum of Anthropology at the University of Michigan. University of Michigan Press. Ann Arbor: 383–426. Hesse B. 1984. These are our goats: The origin of herding in West Central Iran. In B. J. Clutton, C. Grigson (eds.), Animals and Archaeology 3: Early Herders and Their Flocks. BAR International Series 202. British Archaeolo­gical Reports. Oxford: 243–264. Hole F. 1977. Studies in the archeological history of the Deh Luran plain. The Excavations of Chagha Sefid. Me­moirs 9. Museum of Anthropology at the University of Michigan. University of Michigan Press. Ann Arbor. 1987. The Archaeology of Western Iran. Settlement and Society from Prehistory to the Islamic Conquest. Smithsonian series in Archaeological Inquiry. Smithso­nian Institution Press. Washington. Natalia Petrova, Hojjat Darabi 2018. The creative centuries: diversity and innovation in Iranian Neolithic ceramics. In A. Gomez-Bach, J. Bec­ker, and M. Molist. (eds.), II Workshop on Late Neoli­thic ceramics in Ancient Mesopotamia: pottery in context. Museu d’Arqueologia de Catalunya. Barcelona: 173–184. Hole F., Flannery K. V. and Neely J. A. 1969. Prehistory and human ecology of the Deh Luran plain. An early village sequence Khuzistan, Iran. Memoirs 1. Museum of Anthropology at the University of Michigan. University of Michigan Press. Ann Arbor. Hole F., Tonokie Yu. 2021. Preforms for Sequential Slab Manufacture? In R. Özbal, M. Erdalkiran, and Y. Tonoike (eds.), Neolithic pottery from the Near East. Production, Distribution and Use. Third International Workshop on Ceramics from the Late Neolithic Near East, 7–9 March, 2019 – Antalya: proceedings. Koç University press. Antal-ya: 69–76. Le Miere M. 2017. The earliest pottery of West Asia: que­stions concerning causes and consequences. In A. Tsune­ki, O. Nieuwenhuyse, and S. Campbel (eds.), The Emer­gence of pottery in West Asia. Oxbow Books. Oxford: 9–16. Le Miere M., Picon M. 1998. Les debuts de la ceramique, Paleorient 24(2): 5–26. London G. 1981. Dung-tempered clay. Journal of Field Archaeology 8(2): 189–195. https://doi.org/10.1179/009346981791505049 Lopatina O. A., Kazdym A. A. 2010. The natural sand in­clusions in ancient ceramics (the discussion of the issue). In Yu. B. Tsetlin, N. P. Salugina, and I. N. Vasil’eva (eds.), Ancient pottery production: results and prospects of in­vestigation. Institute of archaeology of Russian Academy of Science. Moscow: 46–57. Mahjour F., Ebrahiminia M., and Sedighian H. 2014. Com­parative of Homemade pottery production technology at Bikash and Showy villages, Baneh, Kurdistan. Zagros Re­search Quarterly 2–3: 18–28. (in Farsi) Matson F. R. 1960. Specialized ceramic studies and radio­active-carbon techniques. In R. J. Braidwood, B. Howe (eds.), Prehistoric investigations in Iraqi Kurdistan. Stu­dies in Oriental Civilizations 31. The Oriental Institute of the University of Chicago. The University of Chicago Press. Chicago: 63–70. Matthews W. 2016. Humans and fire: changing relations in early agricultural and built environment in the Zagros, Iran, Iraq. The Anthropocene Review 3(2): 107–139. https://doi.org/10.1177/2053019616636134 Matthews R., Matthews W., and Mahammadifar Y. 2013. The Earliest Neolithic of Iran. 2008 Excavations at Tapper Sheikh-e Abad and Tappeh Jani: Central Zagros Archaeological Project. Oxbow Books. Oxford. Matthews R., Matthews W., Rasheed R. K., and Richard­son A. 2020. The Early Neolithic of the Eastern Fertile Crescent: Excavations at Bestansur and Shimshara, Iraqi Kurdistan. Oxbow Books. Oxford. Matthews R., Fazeli Nashli H. 2022. The archaeology of Iran from the Paleolithic to the Achaemenid Empire. Routledge World Archaeology. New York. McDonald M. M. A. 1979. An examination of mid-Holo­cene settlement patterns in the Central Zagros region of western Iran. Unpublished PhD thesis. Department of Anthropology. University of Toronto. Toronto. Meiklejohn C., Merrett D. C., Reich D., and Pinhasi R. 2017. Direct dating of human skeletal material from Ganj Dareh, Early Neolithic of the Iranian Zagros. Journal of Archaeological Science. Reports 12: 165–172. https://doi.org/10.1016/j.jasrep.2017.01.036 Meldgaard J., Mortensen P., and Thrane H. 1964. Excava­tions at tepe Guran. Preliminary report of the Danish ar­chaeological expedition to Iran, 1963. Acta Archaeolo­gica 34: 97–133. Miller N. F. 1984. The use of dung as fuel: an ethnogra­phic example and an archaeological application. Paleori­ent 10(2): 71–78. 1996. Seed eaters of the Ancient Near East: human or herbivore. Current Anthropology 37(3): 521–528. https://doi.org/10.1086/204514 Miller N. F., Marston J. M. 2012. Archaeological fuel re­mains as indicators of ancient west Asian agropastoral and land-use system. Journal of arid environments 86: 97–103. https://doi.org/10.1016/j.jaridenv.2011.11.021 Mortensen P. 1972. Seasonal Camps and Early Villages in the Zagros. In P. J. Ucko, R. Tringham, and G. W. Dimb­leby (eds.), Man, Settlement and Urbanism. Gerald Duck-worth & CO LTD. London: 293–297. 2014. Excavations at Tepe Guran. The Neolithic pe­riod. Peeters. Leuven-Paris-Walpole. Munchaev R. M., Merpert N. Ya. 1981. The Early Agricul­tural Settlements of Northern Mesopotamia. Nauka. Mos­cow. (in Russian) Nieuwenhuyse O. P., Campbell S. 2017. Synthesis: The emergence of pottery in West Asia. In A. Tsuneki, O. Nieu­ Analysis of Neolithic pottery technology along the Iranian Zagros foothills wenhuyse, and S. Campbel (eds.), The Emergence of pot­tery in West Asia. Oxbow Books. Oxford: 167–192. Petrova N. Yu. 2011. Pottery making in the village of Bal­har, Dagestan. Archeology, ethnography and anthropo­logy of Eurasia 4(48): 130–135. https://doi.org/10.1016/j.aeae.2012.02.015 2012. A technological study of Hassuna culture ceram­ics (Yarim Tepe I settlement). Documenta Praehisto­rica 39: 75–81. https://doi.org/10.4312/dp.39.5 2019. The development of Neolithic pottery technology in Eastern Jazira and the Zagros Mountains. Documen­ta Praehistorica. 46: 128–136. https://doi.org/10.4312/dp.46.8 Petrova N. 2021. Neolithic pottery technology of Sinjar Valley, Northern Iraq (Proto-Hassuna and Archaic Hassu­na periods). In R. Özbal, M. Erdalkiran, and Y. Tonoike (eds.), Neolithic pottery from the Near East. Production, Distribution and Use. Third International Workshop on Ceramics from the Late Neolithic Near East, 7–9 March, 2019 – Antalya: proceedings. Koç University press. Antalya: 213–228. In preparation. The Dung among others temper in Zagros and Mesopotamia Neolithic pottery. Rice P. M. 1987. Pottery Analysis. A sourcebook. Univer­sity of Chicago press. Chicago and London. Roux V., Courty M. A. 2019. Ceramics and Society. A technological approach to archaeological assemblages. Sprigler. Cham. Rue O. S. 1981. Pottery technology. Principles and re­construction. Australian National University. Washington, Taraxacum. Salimi M. 2014. Huh Vorgar. Ethno-archaeological Study on Female Potters at Kurdistan. International Congress of Young Archaeologist. University of Tehran. Tehran: 587–596. Schiffer M. B., Skibo J. M. 1987. Theory and Experiment in the Study of Technological Change. Current Anthropo­logy 28(5): 595–622. https://doi.org/10.1086/203601 Schmandt-Besserat D. 1974. The use of clay before pot­tery in the Zagros. Expedition 16(2): 11–17. Sedighian H., Mahjour. F. 2010. Evolution of domestic pottery production in Bikash village, Baneh, Kurdistan Province. Iranian Journal of Anthropology 12: 78–95. (in Farsi) Shepard A. O. 1956. Ceramics for the Archaeologist. Car­negie institution of Washington. Washington. Smith P. E. L. 1974. Ganj Dareh tepe. Paleorient 2(1): 207–209. 1976. Reflection on Four Seasons of Excavations at Tap-peh Ganj Dareh. In F. Bagherzadeh (ed.), Proceedings of the 4th Annual Symposium on Archaeological Re­search in Iran, 3rd–8th november 1975. Iranian centre for archaeolgical research. Tehran: 11–22. 1990. Architectural innovation and experimentation at Ganj Dareh, Iran. Word Archaeology 21(3): 323–335. https://doi.org/10.1080/00438243.1990.9980111 Smith P., Crepeau R. 1983. Fabrication experimentale de repliques d’un vase neolithique du site de Ganj Dareh, Iran. Paleorient 9(2): 55–62. Spengler R. N. 2019. Dung burning in the archaeobota­nical record of West Asia: where we now? Vegetation Hi­story and Archeobotany 28: 215–227. https://doi.org/10.1007/s00334-018-0669-8 Tauber H. 1970. Radiocarbon dating of potsherds from Tell Shimshara In P. Mortensen (ed.), Tell Shimshara. The Hassuna period. Det Kongelige Danske Videnskaber­nes Selskab Historisk-Filosofiske Skrifter 5, 2. Munksgaard. Kobenhavn: 143–144. Tsetlin Yu. B. 2003. Organic tempers in ancient ceramics. In S. Di Pierro (ed.), Proceedings of the 6th European Meeting on Ancient Ceramics, Fribourg, Switzerland, 3–6 October 2001. Department of Geosciences, Minera­logy and Petrography, University of Fribourg. Friburg: 289–310. 2012. Ancient pottery. Theory and methods of the hi­storical and cultural approach. Institute of Archaeo­logy. Russian Academy of Sciences. Moscow. (in Russian) Tsuneki A. 2017. The significance of research on the emer­gence of pottery in West Asia. In A. Tsuneki, O. Nieuwen­huyse, and S. Campbel (eds.), The emergence of pottery in West Asia. Oxbow Books. Oxford: 2–8. Vandiver P. 1985. Sequential slab construction: a Near Eastern pottery production technology, 8000–3000. The PhD thesis. Massachusetts Institute of Technology. Massachusetts. 1987. Sequential slab construction: a conservative Southwest Asiatic ceramic tradition, ca. 7000–3000 B.C. Paleorient 13(2): 9–35. Natalia Petrova, Hojjat Darabi Vasil’eva I. N., Salugina N. P. 2010. Slab constructions. In Tsetlin Yu. B. (ed.), Ancient pottery. Results and pros­pects of study. Institute of Archaeology. Russian Academy of Sciences. Moscow: 72–87. (in Russian) Voigt M. M. 1983. Hajji Firuz Tepe, Iran. The Neolithic settlement. University of Pennsylvania Museum of Archa­eology and Anthropology. Hasanlu Excavations Reports 1. University Museum Monograph 50. Philadelphia. Volkova E. V., Tsetlin Yu. B. 2016. To the methodology for studying the firing modes of ancient ceramics. In Tra­ditions and innovations in the study of ancient ceram­ics. Materials of the International Conference, May, 24–27, 2016, St. Petersburg, Russia. Institute of History of Material culture of Russian Academy of Science. The State Hermitage Museum. Samara State Academy of Social Sciences and Humanities. UMR 8215 – Trajectoires CNRS-Université Paris 1. St. Petersburg: 76. (in Russian). Zeder M. A. 2008. Animal domestication in the Zagros. An update and direction for future research. In E. Vila, L. Gourichon, A. M. Choyke, and Buitenhuis H. (eds.), Ac-haeozoology of the Near East VIII. Proceedings of the eighth international Symposium on the Archaeozoology of southwestern Asia and adjacent areas. Tome II. Tra­vau de la Maison De L’orient et de la Méditerranée 49. Archéorient, Maison de l’Orient et de la Méditerranée. Lyon: 243–277. Appendix Tab. 1. Technological characteristics of Tapeh Ali Kosh ceramics. Firing Oxidized Painting Outer Inner Surface treatment Outer Inner Constructi­on method Added inclusions % Dung Large plant Clay ferru­gination Thickness (mm) D of rim (cm) Part of vessel No. Context Type Vessel No. Ja’far Painted (covering with the same clay + paint ornament) 1 118 180 spout 8-9 high 20 the same clay+ bone\pebble the same clay +fingers red ornament partial, sharp boarders 2 Ja’far Painted 118 181 wall 7 high 30 the same clay+ bone\pebble the same clay +bone\pebble red ornament partial, gradual boarders 3 Ja’far Painted 116 175 wall 10 high 20 10 slabs the same clay+ bone\pebble the same clay +bone\pebble red ornament partial, gradual boarders 4 Ja’far Painted 115 161 rim 7 medium 10 slabs the same clay+ bone\pebble the same clay +bone\pebble red ornament partial, gradual boarders 5 Ja’far Painted 115 162 rim 23 7-11 medium 20 10 slabs the same clay+ bone\pebble the same clay +bone\pebble red ornament partial, gradual boarders 6 Ja’far Painted 115 164 rim 10 medium 10 the same clay+ bone\pebble the same clay red ornament partial, gradual boarders 7 Ja’far Painted 113 138 wall 8-12 high 20 10 slabs the same clay the same clay +bone\pebble red ornament full 8 Ja’far Painted 112 135 rim 8-10 medium 10 slabs the same clay+ bone\pebble+ slightlyburnishing the same clay +bone\pebble red ornament partial, gradual boarders 9 Ja’far Painted 112 116 wall 10 medium 20 10 slabs the same clay+ bone\pebble the same clay +bone\pebble red ornament partial, gradual boarders 12 Ja’far Painted 106 049 wall 10 medium 20 10 the same clay the same clay red ornament full 10 Ja’far Painted 111 111 rim 17 7-9 medium 10 slabs the same clay the same clay red ornament partial, gradual boarders 13 Ja’far Painted 102 024 wall 8-12 low 10 20 the same clay bone\pebble the same clay red ornament full 11 Ja’far 108 083 rim 15 6-13 medium 30 5 slabs the same clay the same clay +bone\pebble red ornament partial, gradual boarders Analysis of Neolithic pottery technology along the Iranian Zagros foothills Natalia Petrova, Hojjat Darabi Continuation Ja’far Plain (covering with the same slightly ferruginous clay)15 127 198 wall 14 medium 50 the same clay the same clay partial, sharp boarders 16 Ja’far Plain 127 196 wall .11 high 30 30 destroyed surface destroyed surface full 14 Ja’far Plain 127 202 rim 25 7 medium 30 the same clay+ bone\pebble + slightlyburnishing the same clay partial, sharp boarders 18 Ja’far Plain 115 148 wall 11 medium 30 5 slabs the same clay+ bone\pebble the same clay +bone\pebble partial, sharp boarders 17 Ja’far Plain 107 077 rim 10 medium 15 5 slabs the same clay the same clay +bone\pebble full 19 Ja’far Plain 108 102 wall 11 medium 30 slabs the same clay destroyed surface partial, sharp boarders 20 Ja’far Plain 105 028 wall 12 medium 15 5 the same clay the same clay partial, sharp boarders 21 Ja’far Plain 108 097 wall 10 medium 30 slabs the same clay the same clay partial, sharp boarders 22 Ja’far Plain 124 190 wall 15 high 30 slabs the same clay+ bone\pebble the same clay partial, sharp boarders 23 Ja’far Plain 117 178 wall 12 high 30 the same clay+ the same clay partial, sharp boarders bone\pebbleJa’far Plain (covering with the same highly ferruginous clay + burnishing)24 Ja’far Plain 107 083 wall 12 high 15 the same clay the same clay full +bone\pebble+slightlyburnishing 25 108 017 wall 13 high 20 10 the same clay+ bone\pebble+ burnishing the same clay +bone\pebbleburnishing partial, gradual boarders 26 Ja’far Plain 116 176 wall 10-14 high 30 5 slabs the same clay+ the same clay partial, sharp boardersbone\pebble +bone\pebbleKhazineh Red (covered with the same clay + full pigment coloring)27 Khazineh 116 177 rim 7 medium 30 slabs the same clay+ the same clay red full partial, gradual boardersRed bone\pebble +bone\pebble coloring 28 Khazineh Red 116 168 wall 10 medium 10 20 slabs the same clay the same clay red full coloring partial, gradual boarders 29 Khazineh Red 105 043 rim 10 medium 10 slabs the same clay+ bone\pebble+ slightly burnishing the same clay +bone\pebble red full coloring red partial coloring partial, gradual boarders Continuation 30 Khazineh 102 014 wall 8 medium 10 the same clay+ the same clay full plum partial, gradual boarders Red bone\pebble Khazineh Red (covered with red slip (pigment + clay)) 31 Khazineh 105 027 wall 8-12 medium 30 5 slabs red slip red slip+ plum full Red the same clay full cover Tab. 2. Technological characteristics of Tapeh Mahtaj ceramics. No. Part of vessel Thickness (mm) Clay ferrugination Added inclusions % Contruction method Outer and inner surface treatment Firing Dung Large plant Oxidized 1 base\wall 19\9 medium 20 30 slabs soft full 2 base\wall 20 medium 20 30 slabs soft full Tab. 3. Technological characteristics of Tapeh Guran ceramics. Firing full full full Painting Inner red ornament Outer red ornament red ornament Surface treatment Inner the same clay the same clay the same clay Outer the same clay+ slightly burnishing the same clay+ textile| the same clay Constructi­on method slabs slabs Added inclusions % Large plant Dung 30 50 20 Clay ferru­gination medium low low Thickness (mm) 7 5 7 D of rim (cm) 20 Part of vessel wall wall rim Type Jarmo style Guran style Sarab Geometric Number 1 2 3 Red-slipped ware (both surfaces covered with slip) 4 Red-slipped ware rim 41 10 medium 20 20 coils red slip+ burnishing red slip partial, gradual boarders 5 Red-slipped ware rim 28 8 medium 50 red slip+ burnishing red slip+ burnishing full 6 Red-slipped ware wall 10 medium 50 slabs red slip+ burnishing red slip+ burnishing full Analysis of Neolithic pottery technology along the Iranian Zagros foothills Natalia Petrova, Hojjat Darabi Continuation 7 Red-slipped ware rim 24 9 medium 50 red slip+fingers +burnishing red slip full 8 Red-slipped ware rim+wall 8-11 medium 20-30 5 slabs red slip+ burnishing red slip full 9 Red-slipped ware rim 27 8 medium 20-30 slabs red slip+textile +burnishing red slip partial, gradual boarders 10 Red-slipped ware rim 14 5 medium 15 coils red slip+textile +slightlyburnishing the same clay partial, gradual boarders 11 Red-slipped ware rim 6 medium 20-30 slabs red slip+ burnishing red slip partial, gradual boarders 12 Red-slipped ware rim 8 medium 20-30 slabs red slip the same clay +burnishing partial, sharp boarders 13 Red-slipped ware rim 10 7 medium 15 5 slabs red slip+textile red slip partial, gradual boarders 14 Red-slipped ware wall 7 medium 30 5 brown slip the same clay full +fingers Red-slipped ware (different variants)15 Red-slipped ware rim 6 medium 10-15 5 slabs the same clay+ the same clay red-brown partial, gradual boardersslightly full burnishing coloring 16 Red-slipped ware wall 7 low 50 5 slabs the same clay the same clay brown full coloring red full coloring partial, gradual boarders 17 Red-slipped ware rim 7 low 10-15 red slip+ slightly burnishing red slip brown full coloring partial, gradual boarders 18 Red-slipped ware wall 8 medium 30 the same clay+ burnishing the same clay red-brown fullcoloring full 19 Buff ware wall 5-9 low 30 5 coils the same clay+ burnishing the same clay full 20 Buff ware wall 13-15 low 30 5 coils the same clay+ burnishing the same clay +burnishing partial, gradual boarders 21 Buff ware wall 12 low 10-15 the same clay+ burnishing the same clay partial, gradual boarders 22 Buff ware rim 7 low 30 5 coils the same clay+ burnishing the same clay partial, gradual boarders back to contents Documenta Praehistorica XLIX (2022) The end of the earliest ceramic traditions> Dnieper-Dvina region became part of the Circum-Baltic space at the turn of the 6th to 5th millennium BC Andrey Mazurkevich 1, Ekaterina Dolbunova 1, Yolaine Maigrot 2, and Veronika Filippova3 1 The State Hermitage Museum, RU< a-mazurkevich@mail.ru, katjer@mail.ru 2 UMR 8215 Trajectoires, FR< yolaine.maigrot@cnrs.fr 3 St. Petersburg State University, RU< vera.filippova.97@mail.ru ABSTRACT – The Dnieper-Dvina area is one of the regions in Eastern Europe which was part of a wider network of the earliest ceramic traditions, spread in the first half to the middle of the 6th mil­lennium BC. After the collapse of this network new ceramic complexes appeared here, called the Rudnya culture, and at the end of the 6th millennium BC this manifested in changes in the directions of cultural connections. This region became part of the cultural space of the Circum-Baltic area. Se­veral complexes within the Rudnya culture originated in different groups of Narva pottery, and are dated to c. 5400–4400 cal BC. KEY WORDS – early pottery; hunter-gatherers; Narva culture; Circum-Baltic space; bone industry Konec najzgodnej[ih kerami;nih tradicij> regija Dneper-Dvina postane na prehodu iz 6. v 5. tiso;letje pr. n. [t. del cirkumbaltskega prostora IZVLE.EK – Podro.je Dneper-Dvina je ena od vzhodnoevropskih regij, ume..ena v .iroko mre.o naj­zgodnej.ih kerami.nih tradicij, ki so se raz.irile v prvi polovici 6. tiso.letja pr. n. .t. Po razpadu te mre.e so se pojavili novi kerami.ni kompleksi, imenovani kultura Rudnya, ki ob koncu 6. tiso.let­ja pr. n. .t. ka.ejo spremembe v povezavah kultur. Regija je tako postala del kulturnega prostora cirkum-Baltika. Ve. kompleksov znotraj kulture Rudnya namre. izvira iz razli.nih skupin kulture Narva keramike, datirane ok. 5400–4400 pr. n. .t. KLJU.NE BESEDE – zgodnja keramika; lovci in nabiralci; kultura Narva; podro.je cirkum-Baltik; predmeti iz kosti Introduction The hunter-fisher-gatherer economic and cultural change their adaptation mechanisms – through the system, formed as a result of the adaptation of the development of new habitats, the adoption of inno-Final Palaeolithic population to natural changes, ap-vation, the formation of new social and economic peared to be very flexible and existed until at least systems and networks (Burger, Fristoe 2018). The the Middle Neolithic in Eastern Europe (Zhilin 2004). emergence and spread of ancient pottery in Eastern The rich and varied ecological niches of the East Europe at the end of the 7th to the first half of the European Plain allowed ancient communities to 6th millennium BC can be seen as part of these pro-maintain a hunter-gatherer way of life for a long cesses (Mazurkevich, Dolbunova 2015; Andreev, period of time (Dolukhanov 1997; 2008; Zvelebil Vybornov 2020). The chaînes opératoires of pot­2008). The instability of ecological niches due to cli-tery manufacture are suggested to be embedded in matic and/or anthropogenic factors and the varia-social trajectories and social identity (Gosselain bility of biodiversity may have forced societies to 2002; Livingstone-Smith 2001; Pétrequin, Pétre- DOI> 10.4312\dp.49.21 The end of the earliest ceramic traditions> Dnieper-Dvina region became part of the Circum-Baltic space ... quin 2006; Gallay 1991), constituting ceramic tra­ditions unique to each group which all makes it a powerful proxy for social reconstructions. By the middle of the 6th millennium BC ceramic traditions spread over most of the territory of Eastern Europe, following mainly the south-north direction, along river systems. Regional ceramic traditions were formed on their basis, preserved their initial tech­nological, morphological and decorative features, and continued to be part of these earliest ceramic traditions (Mazurkevich, Dolbunova 2015). Later, by the end of the 6th millennium BC, this early cera­mic network, partly overlapping the pre-existing Me­solithic network (Dolbunova, Mazurkevich, submit­ted; Timofeev 1998a) collapsed. Ceramic styles changed greatly, new regional traditions appeared, as well as new directions of contacts, and a new su­pra-cultural network was formed. This complete change in all the constituents of chaînes opérato-ires may be regarded as the end of the earliest ce­ramic traditions on the territory of Eastern Europe which occurred at the end of the 6th millennium BC. They were replaced by communities with comple­tely different ceramic complexes (e.g., Lyalovo cul­ture in the Upper Volga (Lozovskaya et al. 2016), Eneolithic cultures in the Don (Skorobogatov et al. 2015) and Lower Volga region (Vybornov et al. 2016)). In turn, the first pottery spread in the Cir-cum-Baltic space only from 5200–5000 cal BC (Lo­ze 1988; Hartz, Lübke 2006; Raemaekers 2011; Povlsen 2013; Piezonka 2015; Kriiska et al. 2017) (Fig. 1). The Dnieper-Dvina area (Fig. 2) is one of the regions in Eastern Europe where the oldest ceramic tradi­tions penetrated in the first half to the middle of the Fig. 1. Earliest ceramic complexes in the Circum-Baltic space (Ertebolle, Narva, Neman culture, D¹bki site), sites with Narva culture materials in the eastern part and Dnieper-Dvina basin (based on the data from Courel et al. 2020; Kotula et al. 2015; Hartz, Lübke 2006; Povlsen 2013; Tkachou 2018; Wawrusie­wicz et al. 2017). Andrey Mazurkevich, Ekaterina Dolbunova, Yolaine Maigrot, and Veronika Filippova 6th millennium BC (Mazurkevich, Dolbunova 2015). Later, after their disappearance, it became part of the Circum-Baltic cultural space manifested through ceramic complexes which have direct analogies in Narva culture. In this study we examine to what ex­tent this tradition is similar to the preceding initial ceramic complexes, or if it should be regarded as an independent phenomenon; how this area was em­bedded into the Circum-Baltic space at the end of the 6th to 5th millennium BC; and why there was no long evolutionary development of the earliest cera­mic traditions that originated on the territory of fo­rest-steppe and steppe zone of Eastern Europe. Dnieper-Dvina region. Paleoenvironmental set­ting The Western Dvina Lakeland (western part of the East European Plain) is close to the European wa­tershed of three catchments: the Baltic Sea, Black Sea and Caspian Sea. The main traits of the geology and terrain relief of this area were formed during the Valdai (Weichselian) Glaciation and later trans­formed in the Holocene. This region is characterized by developed lacustrine landscapes widely chosen for inhabitation by ancient hunter-gatherers. The en­vironmental conditions of these basins seem to have Fig. 2. Rudnya culture sites distribution in Dnieper-Dvina basin. The end of the earliest ceramic traditions> Dnieper-Dvina region became part of the Circum-Baltic space ... been very attractive for hunter-gatherer communi­ties – due to a high level of landscape geo- and bio­diversity, with small lakes, rivers and fertile hydro-genic and semihydrogenic soils. Microregional studies were concentrated within the Serteyka River basin (Serteya microregion), left tri­butary of the Western Dvina (Daugava) River (Fig. 3). The river was presumably draining subsequent palaeolake water bodies during the Holocene. A few palaeolake basins were documented within the low­er section of the present-day valley. These basins are 100–600 metres wide, 100–2000 metres long and are connected by narrow erosive segments of the ri­ver valley. They are filled with organic deposits of lacustrine and swamp origin up to 8 metres thick co­vered with sandy-silty overbank alluvia (Kulkova et al. 2001; Kittel et al. 2018). There are several of Fig. 3. Rudnya culture sites distribution in Serteysky microregion of Dnieper-Dvina basin (ceramics of phases ‘c-1’, ‘d’, ‘d-1’, ‘e’). Andrey Mazurkevich, Ekaterina Dolbunova, Yolaine Maigrot, and Veronika Filippova them noted – the Great Serteya Palaeolake Basin (GSPB), Nivniky Basin, and Rudnya Basin, on the shores of which sites of the Rudnya culture were found. Seasonal and centennial climatic fluctuations might have influenced ancient communities. Detailed multi-proxy palaeoecological analyses in the Serteya micro-region indicate weak cooling oscillations at the 7.0– 6.8, 6.2, 5.9 and 4.2 ka cal BP and an increase of continentality during these periods, manifested in a lower winter temperature and lower annual preci­pitation (Kittel et al. 2020; Mroczkowska et al. 2020). Rudnya archaeological culture Ceramic complex The Rudnya archaeological culture was distingui­shed based on materials of the stratified Rudnya Ser­teyskaya site, excavated in the Serteya microregion (Fig. 3) in 1983–1987 (Dolukhanov et al. 1989; Ma-zurkevich, Miklyaev 1998). Several ceramic phases were distinguished named ‘c-1’, ‘c-2’, ‘d’, ‘d-1’, and ‘e’. Ceramic phase ‘c-1’ (Fig. 4. 1, 6, 7). The vessels were made from fat paste tempered with organics and shells, judging by the traces of burned-out re­sidue. The vessels were made of short coils/slabs, elongated during vessel construction, 1.7–3.5cm in height with a very sharp horizontal and vertical cut. A separate group is represented by vessels made from coils 1–1.5cm in height, with N-juncture (phase ‘c-2’). The technique of beating is testified by flat­tened roundish areas on the surface of fragments. Traces of scratches on both surfaces were recorded, left after the application of a thin liquid clay layer. They were polished afterwards, which can be seen on a number of fragments with a well-preserved sur­face, which might have been common for all the vessels of this phase. The shapes of the vessels are mostly closed with a pointed or rounded rim. Cera­mics was not decorated. Ceramic phase ‘d’ (Fig. 5; 6.1–4, 6, 9). Pottery was made from paste tempered by a large amount of shell and organics. The vessels were manufactured from stretched coils/slabs with an S-juncture. The extreme fragmentation of the vessels complicates ac­curate determination the size and type of coils/slabs and the presence of the beating technique. The ves­sels are 0.4–0.7cm thick. On both sides there are tra­ces left by a comb tool. The external side of some of the vessels was polished. The vessels were poorly fired, only thermally dried at low temperatures (Ma­zurkevich, Miklyaev 1998). The rims are flattened, and the vessels are slightly profiled. One conical bot­tom with an added clay lump at the extremity and a roundish bottom were found. The vessels were deco­rated with small pits, notches, small triangular marks and impressions of a thin, curved comb stamp, which differed from the decoration of the vessels of the previous Serteya culture. They were decorated by one or several horizontal rows or a combination of horizontal and vertical ones. Only two vessels were decorated by a combination of pits and notches, pits and impressions of a comb stamp. Only the upper part of the vessels was decorated, the rest was often covered by scratches, which were deliberately left and not smoothed over. Ceramic phase ‘d-1’. Vessels were made from paste tempered by shell and organics (Fig. 7; 6.5, 7, 8, 10). The vessels were made from stretched coils with an N-juncture. On both sides there are traces left by a comb tool. The surface was polished afterwards in major cases. The beating technique was widespread. On the surface of the vessels there was a cracking mesh due to drying clay, which was subjected only to temperature drying. The rims are bevelled in­wards, flattened, straight or inclined inwards. The bottoms are rounded. The vessels are 0.7–0.9cm thick. The pots were not decorated or decorated with a single line of holes under the rim or a net made from scratches. One of the vessels is decorated with triangular marks, arranged in horizontal lines. This group also includes a series of small bowls with a C-shaped profile and pointed rim. Ceramic phase ‘e’. Vessels were made from fat paste tempered with organics, judging by the burnt-out remains, from short coils/slabs in the S-technique (Fig. 4.2–5, 8–11). In a few cases, the use of U-junc­ture was noted which was applied for vessels con­struction from the slabs. The surface of the vessels was smoothed, occasionally there are traces of scrap­ing, smoothed afterwards in a number of cases. The rims of the vessels are flattened and straight. The pottery is decorated with small pinholes, a grid of dashed lines and small oval imprints. A comparison of the ceramic manufacturing techno­logy of the Rudnya culture and preceding Serteya culture using correspondence analysis of features, which constitute the chaîne opératoire (Fig. 8.1; see description of features – Fig. 8), indicates two com­pletely different technological groups. Major differ­ences are also observed in decoration and morpho­logy (see the description of the Serteya culture in The end of the earliest ceramic traditions> Dnieper-Dvina region became part of the Circum-Baltic space ... Mazurkevich and Dolbunova (2015)), which may in­dicate a complete change in the ceramic traditions and, probably, local population that existed in the Dnieper-Dvina Basin in the first half to the middle of the 6th millennium BC and later – at the end of the 6th to the first half of 5th millennium BC. A comparison with the Narva complexes shows par­ticular similarities both in the technological sphere (see the correspondence analysis of technological features in Fig. 8.2) and decorative, morphological one. Ceramics of ‘d-1’ appear to be close to some types of ceramics of the Zvidze site and sites in Es­tonia (Kääpa complex) (Kriiska 1996; Kriiska et al. 2017) and in Belarus (Charniauski 2017). Pottery of phase ‘d’ seems to be closer to the ceramics found at sites of Lubana Lake, and specifically the Zvidze and Osa sites (Loze 1988; Zagorskis 1973). Pottery of phases ‘c-1’ and ‘c-2’ is similar to that found at the coastal river estuaries and coastal lagoon group of sites in Estonia (Kriiska 1996; Kriiska et al. 2017). Flint and bone industry, the remains of con­structions A few flint and bone items were found nearby the pottery of the phases ‘d’ and ‘d-1’ at the Rudnya Ser- Fig. 4. Pottery fragments of phase ‘c-1’ (1 Serteya XII; 6, 7 Uzmen); ‘e’ (2–5, 8–11 Uzmen). Andrey Mazurkevich, Ekaterina Dolbunova, Yolaine Maigrot, and Veronika Filippova teyskaya site (Mazurkevich, Miklyaev 1998), and single items were found at Serteya II (II-2) and Ser­teya X (Figs. 9, 10). At the Rudnya Serteyskaya site, the flint industry includes a rhomboid-shaped arrow­head with bifacial retouch, end scraper, roundish scrapers with a rounded blade, oval axes with an asymmetrical lenticular cross-section and polished areas on the blade. Two single-platform edge-faceted cores and two double-platform nuclei were also found. Bone tools include knives, which are characterized by a lateral cutting edge (Fig. 10.1, 6). They were produced by scraping from various anatomical sup­ports, always from elk. It should be noted that the knives known for the Middle-Late Neolithic Usvyaty and Zhizhitsa cultures were made exclusively from elk ulna. Early Neolithic bone knives are known in the Baltic area (Vankina 1999), and also in Belaru­sian sites (Charniauski 2007). A few ornaments are also recorded. The Serteya X site provided a pendant made from an incisor with a grooved suspension (Fig. 10.4). Two other boar incisors with grooved suspension were found at the Serteya II-2 site. Several elk tooth grooved at their proximal part were also found in Early Neoli­thic layer at the Zamostje 2 site (Lozovskaya 2018). Teeth pendants remained fairly common through­out the Neolithic period, but their attachment sys­tem changed, and perforation began to be used in the Middle Neolithic. Most of the Early Neolithic bone projectile points are cha­racterized by bulging and of­ten biconical heads, mostly with a tapered tip (Fig. 9.2, 4, 5), one of them was decorat­ed with short incisions (Fig. 9.4). Their shaft, quite short, is often pointed and could be flattened or rounded in cross-section. This type is al­so quite common in Upper Volga (Lozovsky, Lozovskaya 2010; Lozovskaya 2019; Zhi­lin et al. 2002), and in the Narva culture (Vankina 1999; Loze 1988). A particular type of a biconical flattened arrow­head with symmetrical wings and a short, pointed shaft was found at Rudnya Serteyskaya (Fig. 9.3) and the Serteya II­2 site (Fig. 9.1). These projec­tile points found in the Dnie­per-Dvina Basin constitute di­scriminating elements of the Early Neolithic which will no longer exist in the following periods. Rudnya Serteyskaya had a rather particular bone projectile point with a very long shaft and a head deco­rated with short transverse incisions (Fig. 9.8). A wood­en projectile point imitating bone items was found at the Rudnya Serteyskaya site (Fig. 10.2). The end of the earliest ceramic traditions> Dnieper-Dvina region became part of the Circum-Baltic space ... Two harpoons (one is fragmented) were found (Fig. 10.6, 7). The whole piece consists of a three-barbed harpoon with shouldered proximal part. It might have been made from a bone blank extract­ed from an elk metapodium. Most of the harpoons from the Upper Volga culture feature a tight succes­sion of barbs and a pointed proximal part (Lozov­skaya, Lozovsky 2013), which does not correlate with Early Neolithic pieces from the Dniper-Dvina basin. Neither of the analogies can be traced within the Baltic complex (Vankina 1999). In these two re­gions, shouldered proximal parts of the harpoons seem to arrive later (Middle or Late Neolithic). The remains of stake structures of rectangular or circular shape with ground hearths were recorded at the Serteya X and XIV sites (Mazurkevich et al. 2003). The remains of a fishing trap at the Serteya XIV and Rudnya Serteyskaya sites and strongly erod­ed wooden objects (on Serteya II-2 site) can also be attributed to different stages of the Rudnya culture (Tab. 1). Fig. 6. Pottery fragments of phase ‘d’ (1–4, 6, 9), and ‘d-1’ (5, 7, 8, 10) on Rudnya Serteyskaya site with indication of coils/slabs juncture (2, 7, 10). Andrey Mazurkevich, Ekaterina Dolbunova, Yolaine Maigrot, and Veronika Filippova Rudnya culture chronology The archaeological layers with materials of the Rud­nya culture lie above the layers or are separated spa­tially from the materials of the Serteya culture, which made it possible to create the sequence of these cultures (Mazurkevich et al. 2017), supple­mented by typological analysis. Ceramics of phases ‘c-1’/‘c-2’, and ‘d-1’ are among the earliest within the sequence of Rudnya culture, following by materials of ceramics of the phase ‘d’ and later phase ‘e’. The absolute chronology is based on a series of radiocar­bon dating by associated terrestrial material (wood and charcoal), deposited together with finds of cera­mics of the Rudnya culture at the Rudnya Serteyska-ya, Serteya II and Serteya XIV sites (Tab. 1). The dif­ficulty of using direct dating of ancient hunter-gathe­rer ceramics is connected with the reservoir effect arising at dating of food crust formed after proces­sing of mainly aquatic products in vessels (Courel et al. 2021.SI). The Rudnya Serteyskaya site is located on a terrace that was buried under wetland deposits, on a small promontory that juts into the Rudnya lake basin. The uncovered area of the site was 146m2 (Mazur­kevich, Miklyaev 1998). Archaeological layer A, with finds of Early Neolithic ceramics of the Serteya cul­ture dated to the 6th millennium BC, was deposited in a layer of fine sand at the base of bluish gyttja. The overlying layer B, where fragments of Rudnya culture pottery were found (phases ‘d’ and ‘d-1’), Fig. 7. Pottery fragments of phase ‘d-1’ (1, 3, 4–6 Serteya XIV; 2 Usviaty II) with indication of traces left during surface treatment (1, 6) and coils juncture (1, 2, 5, 6). The end of the earliest ceramic traditions> Dnieper-Dvina region became part of the Circum-Baltic space ... Fig. 8. Correspondence analysis of Rudnya and Serteya culture pottery (1); Rudnya, Serteya and Narva culture pottery (Zvidze and Narva Joaorg sites) (2). Andrey Mazurkevich, Ekaterina Dolbunova, Yolaine Maigrot, and Veronika Filippova was deposited in a layer of sandy gyttja with char­coal inclusions, located between bluish and brown­ish gyttja. Vessel fragments of phases ‘d’ and ‘d-1’ were recorded in different parts of the site. The ra­diocarbon dates obtained probably reflect several stages of site occupation. Fragments of worked wood, deposited in close proximity to the ceramics of the Rudnya culture, are dated to 5210– 4952 cal BC (first group) (Mazurkevich, Miklyaev 1998). The charcoal from the base of the oval ground structure of Serteya XIV site can also be dated to this time (5332–4944 cal BC). Wood fragments overlying the layer B are dated to c. 4932–4608 cal BC – the time when gyttja accumu­lated over the Rudnya culture re­mains during the transgression phase. The second group of dates made on wooden items (includ­ing the remains of a fishing trap) (c. 4727–4497 cal BC) possibly correlates with the appearance of sites on mineral shores with materials of the ‘e’ phase. Ceramic fragments of the Rudnya culture on the Serteya II site were found in its western part (II-2 area), which is a multilayer com­plex with several succeeding pha­ses of occupation (Mazurkevich et al. 2020; Kittel et al. 2018). A very dynamic local hydrological system and palaeolake shore con­ditions had a considerable influ­ence on the formation of archaeo­logical layers and distribution of finds. The oldest ceramics attri­buted to the Serteya culture were found in the lowermost sandy la­yer, attesting to a coastal activity zone during the Early Neolithic. Rudnya culture fragments were found in a dark brownish-grey sandy layer with organic remains and in brown gyttja (see Kittel et al. 2020). The fragments were eroded which could indicate the redeposition of the archaeologi­cal layer and organic matter in the lake shore zone. They could have been washed away from higher parts of the site. The bone arrowheads attributed to the Rud­nya culture deposited in the dark brownish-grey sand and black gyttja layers may be evidence of hunt­ing. Two heavily eroded wooden items (one of them probably a paddle) are dated to c. 5208–4849 cal BC. A wooden stake – evidence of coastal activities – was dated to 5373–5213 cal BC. The results from the macrobotanical study show a gradual shallowing of the lake level after 6900 cal BC. During this time, the palaeolake already seems to be rich in faunal and floral aquatic resources, and thus attractive for exploitation. In the further course of time, a transgression phase occurred between c. 5550 and 3600 cal BC (Wieckowska-Lüth et al. 2021). Natural accumulation of minerogenic and or- The end of the earliest ceramic traditions> Dnieper-Dvina region became part of the Circum-Baltic space ... ganic material in the transition zone between the lake and land, fluctuations in lake water level (in­flow of minerogenic material, wave erosion), and periodic drying was recorded for this time period (Kittel et al. 2020). Distribution of Rudnya culture sites in the Dnie­per-Dvina Basin Upper Western Dvina basin. The vessels of phase ‘d-1’ are known on the Rudnya Serteyskaya, Serteya II-2, X, XIV, XXXIV sites (Fig. 3) located at low hyp­sometric levels, and archaeological layers are buried under wetland deposits. Vessels of the Rudnya culture are found much less frequently at the sites located on drylands. These are mostly vessels of the ‘c-1’, ‘c-2’ and ‘e’ phases. There are a few fragments of ‘d-1’ and ‘d’ phases, found in the sandy deposits at the Uzmen and Us-vyaty II sites, Shugailovo, and Mochary sites. Vessels of phase ‘e’ are found within sandy sediments at the Date (BP) Date (cal BC) Index Material Site Context Attribution (culture) 5770±60 4730-4488 Le-2570 wood Rudnya Serteyskaya peaty gyttja with wooden remains and algae, wood from fishing construction Rudnya cul­ture 5780±40 4723-4535 Le-7182 charcoal Serteya XIV sq. A\14, yellow sand –\\– 5780±50 4727-4497 Le-2577 wood Rudnya Serteyskaya peaty gyttja with wooden remains and algae, wood from fishing construction Rudnya cul­ture 5850±150 5057-4362 SPb-1197 food crust Serteya XXXIV ceramic belly fragment undecorated (No. 1726-1728) Rudnya cul­ture 5890±60 4935-4605 Le-2586 wood Rudnya Serteyskaya wood from the horizon above layer B Rudnya cul­ture 5900±40 4850-4686 Le-7173 charcoal Serteya XIV sq. 6\3, dwelling, base layer –\\– 6090±40 5081-4896 Poz-146276 wood Sertyeya II-2 treated and eroded wood (w405), lying in horizontal position, on whitish sand Rudnya cul­ture 6090±40 5081-4896 Poz-146882 wood Sertyeya II-2 treated and eroded wood (paddle|) (9814) Rudnya cul­ture 5940±130 5137-4537 Le-2566 wood Rudnya Serteyskaya sandy gyttja with charcoals, layer B, sq. D\1 Rudnya cul­ture 5940±130 5137-4537 Le-4101 wood Rudnya Serteyskaya sandy gyttja with charcoals, layer B, sq. B\9 Rudnya cul­ture 6050±40 5053-4837 Le-9764 wood Serteya XXXIV sq. D\2, low part of ferruginized sand Rudnya cul­ture 6130±30 5210-4988 Le-9763 wood Serteya XXXIV sq. D\2, No. 2254, low part of ferruginized sand Rudnya cul­ture 6130±40 5210-4952 Le-2579 wood Rudnya Serteyskaya sandy gyttja with charcoals, layer B, sq. B\4 Rudnya cul­ture 6130±50 5215-4935 Le-7175 charcoal Serteya XIV yellow sand with charcoal, sq. b\13 –\\– 6180±40 5218-5003 Le-2569 wood Rudnya Serteyskaya sandy gyttja with charcoals, layer B, sq. D\1 Rudnya cul­ture 6320±40 5373-5213 Poz-146296 wood Sertyeya II (part 2) sharpened stake (W454) in vertical position Rudnya cul­ture 6210±80 5332-4944 Le-7176 charcoal Serteya XIV dwelling’s floor, yellow sand –\\– 6230±40 5306-5054 Le-2568 wood Rudnya Serteyskaya sandy gyttja with charcoals, layer B, sq. D\1 Rudnya cul­ture 6240±60 5326-5028 Le-3054 wood Rudnya Serteyskaya sandy gyttja with charcoals, layer B, sq. B\3 Rudnya cul­ture 6388±38 5417-5306 Le-7174 charcoal Serteya XIV sq. b\13, fish-trap –\\– 6640±110 5738-5372 SPb-750 burnt bones Serteya XXII located nearby early neolithic vessels (phase ‘b-1’\‘b-5’) Serteya cul­ture 6792±120 5913-5482 SPb-748 burnt bones Serteya XXVII sq. M-L\4, located nearby early neolithic vessel (phase ‘b-5’) Serteya cul­ture 7300±120 6419-5983 SPb-749 burnt bones Serteya XXIV sq. B\1, located nearby early neolithic vessel (phase ‘b-4’) Serteya cul­ture 7350±180 6571-5885 Le-5260 wood Serteya X bluish sandy gyttja layer with shells, cultural remains of early neolithic Serteya culture Serteya cul­ture Tab. 1. Radiocarbon dates of materials attributed to Serteya and Rudnya culture. Andrey Mazurkevich, Ekaterina Dolbunova, Yolaine Maigrot, and Veronika Filippova Uzmen, Usvyaty II, Serteya IIa sites, Shugailovo, and Mochary sites. The pottery of phase ‘c-1’ was found on sites situated on mineral terraces of palaeolakes in the southern (Serteya X, Serteya XII) and north­ern lake basin (Serteya XIV, XXVII), on shores of pa-laeolakes in the northern basin of the Serteya micro-region (Serteya XXI, XXII, XXIV, 3–3, and 3–2). The pottery of phase ‘c-2’ was found on sites situated in mineral terraces of palaeolakes (field above Rudnya Serteyskaya No. 3 (PRS3), Serteya XIV) and also on the shore of palaeolake in the northern lake basin (Serteya 3–3, XXXVI, XXXVII, XLIV). Upper Dnieper Basin. Pottery of phase ‘c-1’ was found at the Katyn 9 site, and of phase ‘e’ at the Katyn 17 and Kozichino sites. These are the most south-eastern sites of the Rudnya culture. Vessels of the Rudnya culture were found on the sites lo­cated in the immediate vicin­ity of the paleolacustrine sho­reline, which were most like­ly seasonal one (testified by avifauna at the Rudnya Ser­teyskaya site (Sablin et al. 2011)). The supposed settle­ment system differs from the previous time, when various types of sites were recorded: summer and winter camps, long-term and specialized hunting or fishing sites (Ma­zurkevich, Dolbunova 2009). Vessels of the Rudnya culture are accompanied by finds of single bone arrowheads, the remains of fishing traps, strongly eroded wood items with traces of processing (one of which may be a paddle), and wooden stakes, testifying to household activity on the ancient shoreline (Serteya II and XIV sites). Discussion The cultural space, formed in the late 6th to 5th millennium BC in the Circum-Baltic re­gion, includes the Ertebolle culture and the D¹bki site in the west, Narva and Neman cultures in the east, and Sperrings and Sär 1 in the north (Gurina 1967; Rimantiene 1992; Timofeev 1998a; Loze 1988; Kriiska 1996; Kriiska et al. 2017; Charniauski 1979; 2017; Pili.iauskas 2002; German 2018; Torvinen 2000; Kotula et al. 2015). A number of regional and regional-chronological groups have also been distinguished within these (Kriiska 1996; Kriiska et al. 2017; Vankina et al. 1973; Rimantiene 1973; Miksaite 2005; Piezonka 2015; Tkachou 2018; Wawruciewicz 2013). The hunter-gatherer-fisherman cultures of the West­ern and Eastern Baltic existed at the same time as agricultural communities to the south. Thus, the ori­gin of pottery in the Ertebolle culture has been sug­gested to be a local innovation, the influence of hun­ter-gatherer communities from the East (Gronen- The end of the earliest ceramic traditions> Dnieper-Dvina region became part of the Circum-Baltic space ... born 2011), or of Neolithic farming groups (Povlsen 2013). The emergence of pottery among hunter-ga­therer tribes of the Swifterbant culture c. 5200/5000 cal BC was explained by the influence of neighbour­ing agricultural groups (Raemaekers 2011). The influence of local farming communities on the for­mation of hunter-gatherer pottery complexes in the 5th and 4th millennia BC was also noted for Central Europe (Nowak 2017; Guminski 2020). The origin of the Narva culture early ceramic complex is still under discussion. The asynchrony of the processes of ceramics acquisition in different regions by hun­ter-gatherer communities is evidenced by the diffe­rences in time of the pottery’s appearance and signi­ficant variations in the ceramic technology. The dif­ferences in the chaînes opératoires of ceramic ma­nufacture testify to major differences between vari­ous cultural traditions of hunter-gatherer communi­ties of the Circum-Baltic world (Dumpe et al. 2011; Glykou 2010). The widespread S-profiling of the vessels in early complexes forced researchers to look for the eastern origin of these in the Elshanskaya culture, where S-profile vessels are known (Timofeev 1998; Gronen-born 2011; Andreev, Vybornov 2021). However, the differences in technology, chronology and mor­phology (Courel et al. 2021.Fig. S2) indicate the more different nature of these complexes. Studies into ceramics contents and their use (through traces of use) point to different functional patterns of vessel use among foraging communities (Courel et al. 2020; 2021; Papakosta et al. 2019; Pääkkö­nen et al. 2016). A predominance of vessels used for aquatic products was found typical for Narva (in­cluding vessels from the sites located in the Serteya microregion (Courel et al. 2020)). Vessels of the Rudnya culture were found on sites with rather spe­cific contexts – oriented towards the use of water re­sources – with finds of fishing traps and paddles lo­cated in the shoreline zone, associated with fishing grounds or household activity in the shoreline area. Specific technological, morphological and ornamen­tal features of ceramic phases (‘c-1’, ‘c-2’, ‘d’, ‘d-1’ and ‘e’) and the particularities of their deposition within archaeological layers allow us to identify se­veral groups within Rudnya culture. These may re­flect penetration of various traditions from different territories, which are also testified by comparison with different Narva complexes. The flint complex of the Rudnya culture does not have any similarities with the preceding stone industry, bone and antler items find analogies both in the Late Mesolithic and Early Neolithic materials of Eastern European forest zone sites, and in the Lubana basin (Loze 1988; Van-kina 1999). The chronological timeframes of the Rudnya culture correlate with chronological periods identified for the Zvidze site, where the early Narva pottery was found in several archaeological layers, dated to c. 5409–4944 cal BC, 5211–4835 cal BC and 4850– 4582 cal BC; the lower border of Narva culture here was attributed to 4446–4157 cal BC (Loze 1988.73– 74). The appearance of this ceramic complex around 5500–5300 cal BC is confirmed by stratigraphic ob­servations, dating of accompanying materials (wood) (Loze 1988) and direct dating of charcoal found in the vessel fragments (Courel et al. 2020). A compa­rison of the radiocarbon chronology of the Lubana sites with those of the Rudnya culture shows that the latter appeared later in the east in the upper rea­ches of the Western Dvina River. The similarity of the technological, ornamental, mor­phological and functional characteristics of the ves­sels of the Rudnya culture with the groups of the Narva culture makes it possible to consider the trans­fer of ceramics as one package resulting in the Rud­nya culture formation: along with chaînes opérato-ires, ornamental and morphological traditions, the functional pattern was transferred. All of these may testify to the migration of people from different re­gions of the Eastern Baltic to the south-east. There are single and rare evidences of Narva culture materials on other sites in northwestern Russia, but all these complexes are extremely sparse and could instead indicate single penetrations of individual groups to the east: Veksa 3 (Nedomolkina et al. 2015; Piezonka 2015), Kuzemkino 1–6, Galik 3–4, 6–7, 10 (Holkina 2019), Sjaberskaya III (Timofeev 1993), and in the Upper Dnieper basin (Fig. 1, 2). Conclusion The emergence of pottery in hunter-gatherer-fisher communities in continental Europe may have follow­ed different spatial and cultural trajectories, respond­ing to different economic or cultural challenges. The emergence of pottery was accompanied by extensive development of the Eastern European territories, overlapping only in part with the preceding Mesoli­thic network. Destruction of this initial network is manifested through the disintegration of the oldest ceramic traditions that originated in the steppe and Andrey Mazurkevich, Ekaterina Dolbunova, Yolaine Maigrot, and Veronika Filippova forest-steppe zones. The role of pottery might have out the Neolithic (e.g., knives made from elk long been different – as an important adaptive mecha-bones) and other particular for only the Early Neoli­nism and innovation in some regions, and perhaps thic – pendants with a grooved suspension and spe-a non-utilitarian element in others (Courel et al. cific types of bone arrowheads. The latter fits into 2020). Societies that practiced pottery manufacture the evolutionary scheme of arrowhead development might have existed independently among Mesolithic from the Mesolithic to Early Neolithic following the societies on the territory of Eastern Europe, occupy-common trajectory of the size getting smaller. The ing free ecological niches, for a rather short period shortening of bone projectile points could reflect the of time, and thus did not make any significant con-appearance of a new hunting strategy as the role of tributions to later ceramic traditions. The change of bows increased, and may indicate the increasing im-ceramic types, groups and cultures did not lead to portance of fur hunting. Analogies in both the Up-the formation of hybrid ceramic types. The tradition per Volga area and the Baltic may reflect this com-of pottery making seems to be not of a widely used mon trend. practice, which is evidenced in the small number of the earliest vessels found on a number of sites (Ma-Different areas of origin evidenced for the Serteya zurkevich, Dolbunova 2015). The emergence of pot-and Rudnya cultures testify to changes in the direc­tery outside the East European Plain in hunter-gathe-tions of cultural interactions at the end of the 6th rer communities at a later time in Central Europe millennium BC from south-north to west-east. This may refer to a completely different process – the could be a marker of the destruction of the estab-borrowing of pottery-making skills from agricultural lished network that existed before, during the late groups (Nowak 2017; Guminski 2020). 7th and 6th millennia BC. Such a significant change in material culture was due not only to the cultural Ceramic complexes of the Rudnya culture are not the impulse, but possibly the penetration of a new popu-most ancient or first in the territory of the Dnieper-lation from the western territories, which likely es-Dvina Basin, unlike in the Eastern Baltic. This cultu-tablished a new system of cultural and social rela­ral group follows the earlier ceramic traditions (Ma-tions. zurkevich, Dolbunova 2015), which originated in the Upper and Middle Don, Desna River, Bug-Dnies­ter Basin. The closest analogies to the Rudnya cul- ACKNOWLEDGEMENTS ture ceramic complex can be found in the lower The research was supported by a grant from the Rus- course of the Western Dvina River in the Lubana Ba- sian Science Foundation (Project No: 22-18-00086). sin. Thus, the Rudnya culture can represent an inde- Dates Poz-146276, Poz-146296 were funded by the pendent phenomenon within a larger cultural entity French Ministry of Foreign Affairs (MEAE Mission of the Narva culture. The bone industry of the Dnie­ 2NOR), date Poz-146882 – by CNRS (IRP No. 293933). per-Dvina region shows some items existed through­ . References Andreev K. M., Vybornov A. A. 2020. Migration and Dif­fusion in the Neolithization of Volga Region. Stratum 2: 15–30. 2021. Ceramic traditions in the forest-steppe zone of Eastern Europe. Open archaeology 7(N1): 705–717. https://doi.org/10.1515/opar-2020-0169 Burger R. J., Fristoe S. T. 2018. Hunter-gatherer popula­tions inform modern ecology. Proceedings of the Natio­nal Academy of Sciences of the United States of Ameri­ca 115(6): 1137–1139. https://doi.org/10.1073/pnas.1721726115 Charniauski M. M. 1979. Nealit Belaruskaga Paniamon­nia. Minsk. (in Russian) 2007. Kascjanyja i ragavyja vyraby na paselishchah Kryvinskaga tarfjaniku. Minsk. (in Russian) Charniauski Max M. 2017. Pamjatniki narvskoj kul’tury na territorii Belarusi: sostojanie issledovanija i perspektivy. In A. A. Vybornov, A. N. Mazurkevich, E. V. Dolbunova, and Tkach E. S. (eds.), Kul’turnye processy v cirkumbal­tijskom prostranstve v rannem i srednem golocene. Do-klady mezhdunarodnoj nauchnoj konferencii, posvja­shhennoj 70-letiju so dnja rozhdenija V. I. Timofeeva. Sankt-Petersburg: 247–251. (in Russian) The end of the earliest ceramic traditions> Dnieper-Dvina region became part of the Circum-Baltic space ... Courel B., Robson H. K., Lucquin A., +23 authors, and Craig O. E. 2020. Organic residue analysis shows sub-re­gional patterns in the use of pottery by Northern Eurpean hunter-gatherers. Royal society open science 7(4): 192016. https://doi.org/10.1098/rsos.192016 Courel B., Meadows J., Lucquin A., +10 authors, and He­ron C. 2021. The use of early pottery by hunter-gatherers of the Eastern European forest-steppe. Quaternary Sci­ence Reviews 269: 107–143. https://doi.org/10.1016/j.quascirev.2021.107143 Dolbunova E., Mazurkevich A., submitted. Mesolithic of Eastern Europe. In Nilsson Stutz L., Peyroteo Stjerna R., and Torv M. (eds.), The Oxford handbook on Mesolithic Europe. Dolukhanov P. M. 1997. The Pleistocene-Holocene transi­tion in Northern Eurasia: environmental changes and hu­man adaptations. Quaternary International 41/42: 181– 191. https://doi.org/10.1016/S1040-6182(96)00051-1 2008. The Mesolithic of European Russia, Belarus, and the Ukraine. In Bailey G., Spikins P. (eds.), Mesolithic Eu­rope. Cambridge University Press. Cambridge: 280–301. Dolukhanov P. M., Gey N. A., Miklyayev A. M., and Mazur­kiewicz A. N. 1989. Rudnya-Serteya, a stratified dwelling-site in the upper Duna basin (a multidisciplinary research). Fennoscandia archaeological VI: 23–27. Dumpe B., Berzin. V., and Stilborg O. 2011. A dialogue across the Baltic on Narva and Ertebolle pottery. In S. Hartz, F. Luth, and T. Terberger (eds.), Early pottery in the Baltic – Dating, Origin and Social Context. Interna­tional Workshop at Schleswig from 20th to 21st October 2006. Bericht der Römisch-Germanische Kommission. Band 89, 2008. Römisch-Germanische Kommission des deutschen aachäologischen instituts. Philipp von Zabern. Frankfurt a. M.: 89–110. Gallay A. 1991. Itinéraires ethnoarchaeologiques I. Do­cument du Département d’Anthropoogie et d’Ecologie de l’Université de Geneve 18. Geneve. German K. J. 2018. Kul’tura sperrings (sovremennoe so-stojanie izuchenija). Samarskiy nauchnyj vestnik 7(3): 225–230. (in Russian) Glykou A. 2010. Technological and typological analysis of Ertebolle and early Funnel Beaker pottery from Neustadt LA 156 and contemporary sites in northern Germany. In B. Vanmontfort, Louwe Koojimans, L. Amkreutz, and L. Verhart (eds.), Pots, Farmers and Foragers. Pottery tra­ditions and social interaction in the earliest Neolithic of the lower Rhine Area. Archaeological Studies Leiden Uni­versity 20. Leiden University Press. Leiden: 177–188. Gosselain O. 2002. Poteries du cameroun meridional. Styles techniques et rapports a l’identité. Monographies du CRA 26. CNRS Editions. Paris. Gronenborn D. 2011. Early pottery in Afroeurasia – ori­gins and possible routes of dispersal. In S. Hartz, F. Lüth, and T. Terberger (eds.), Early pottery in the Baltic – dat­ing, origin and social context. International Workshop at Schleswig from 20th to 21st October 2006. Bericht der Römisch-Germanische Kommission. Band 89, 2008. Rö­misch-Germanische Kommission des deutschen archäolo­gischen instituts. Philipp von Zabern. Frankfurt a. M.: 59– 88. Guminski W. 2020. The oldest pottery of the Para-Neoli­thic Zedmar culture at the site Szczepanki, Masuria, NE-Po­land. Documenta Praehistorica 47: 126–154. https://doi.org/10.4312/dp.47.8 Gurina N. N. 1967. Iz istorii drevnih plemen zapadnyh oblastej SSSR. Materiali i Issledovaniya po Arkheologii SSSR 144. Leningrad. (in Russian) Hartz S., Lübke H. 2006. New Evidence for a Chronostra­tigraphic Division of the Ertebolle Culture and the Earli­est Funnel Beaker Culture on the Southern Mecklenburg Bay. In C.-J. Kind (ed.), After the Ice Age – Settlements, subsistence and social development in the Mesolithic of Central Europe. Proceedings of the International Confe­rence, 9th to 12th of September 2003. Rottenburg/Nec­kar, Baden-Württemberg. Materialhefte zur Archäologie in Baden-Württemberg 78. Theiss. Stuttgart: 59–74. Holkina M. A. 2019. Keramicheskie tradicii Narvsko-Luzhskogo mezhdurech’ja. Pamjatniki kamennogo ve­ka rossijskoj chasti Narvsko-Luzhskogo mezhdurech’ja. Sankt-Petersburg: 129–140. (in Russian) Kittel P., Mazurkevich A., Wieckowska-Lüth M., +11 au­thors, and S³owiñski M. 2020. On the border between land and water: the environmental conditions of the Neo­lithic occupation from 4.3 until 1.6 ka BC at Serteya, West­ern Russia. Geoarchaeology an Interdisciplinary Journal: 1–30. https://doi.org/10.1002/gea.21824 Kittel P., Mazurkevich A., Dolbunova E., +7authors, and Wieckowska-Lüth M. 2018. Palaeoenvironmental recon­structions for the Neolithic pile-dwelling Serteya II site case study, Western Russia. Acta Geographica Lodziensia 107: 191–213. Kotula A., Czekaj-Zastawny A., Kabaciñski J., and Terber­ger T. 2015. Find distribution, taphonomy and chronol­ogy of the D¹bki site. In J. Kabaciñski, S. Hartz, D. C. M. Raemaekers, and T. Terberger (eds.), The D¹bki Site in Pomerania and the Neolithisation of the North Euro­pean Lowlands (c. 5000–3000 calBC). Archäologie und Andrey Mazurkevich, Ekaterina Dolbunova, Yolaine Maigrot, and Veronika Filippova Geschichte im Ostseeraum 8. Verlag Marie Leidorf GmbH. Leidorf: 113–137. Kriiska A., Oras E., Lougas L., Meadows J., Lucquin A., and Craig O. 2017. Late Mesolithic Narva stage on the territo­ry of Estonia: pottery, settlement types and chronology. Estonian Journal of Archaeology 21(1): 52–86. Kriiska A. 1996. The Neolithic pottery manufacturing technique of the Lower course of the Narva River. In T. Hackens, S. Hicks, and V. Lang (eds.), Coastal Estonia: Re­cent Advances in Environmental and Cultural History. European Study Group on Physical, Chemical, and Mathe­matical Techniques Applied to Archaeology (PACT). Vo­lume 51. Conseil de l’Europe. Strasbourg: 373–384. Kulkova M. A., Mazurkevich A. N., and Dolukhanov P. M. 2001. Chronology and palaeoclimate of prehistoric sites in Western Dvina-Lovat area of North-western Russia. Geo-chronometria 20: 87–94. Livingstone-Smith A. 2001. Chaînes opératoires de la po­terie. Références ethnographiques, analyse et reconsti­tution. PhD thesis. The Université libre de Bruxelles. Fa-culté de Philosophie et Lettres. Section d’Histoire de l’Art et Archéologie. Bruxelles. https://doi.org/10.4000/nda.955 Loze I. L. 1988. Poselenija kamennogo veka Lubanskoj niziny. Mezolit, rannij i srednij neolit. Riga. (in Russian) Lozovskaya O., Lozovski V. 2013. Barbed points from the site of Zamostje 2. In V. Lozovsky, O. Lozovskaya, and C. I. Clemente Conte (eds.), Zamostje 2 – Lake settlement of the Mesolithic and Neolithic fisherman in Upper Vol­ga. Institute for the history of material culture, Russian Academy of Science. Sankt-Petersburg: 76–109. Lozovskaya O. V., Lozovski V. M., Zaitseva G. I., +3 au­thors, and Rishko S. A. 2016. Neoliticheskiye drevnosti stoyanki Zamost’ye 2. Katalog radiouglerodnykh dat. In A. N. Mazurkevich, M. A. Kulkova, and E. V. Dolbunova (eds.), Radiouglerodnaya hronologiya epohi neolita Vo­stochnoi Evropy v VII–III tys. do n.e. Gosudarstveniy Er-mitaz. Institut materialnoy kulturi, Rossiskaya akademiya nauk. Samarskiy gasudarstveniy socialno-pedagogicheskiy universitet. Svitok. Smolensk: 182–202. (in Russian) Lozovskaya O. V. 2018. Osnovnye formy izobrazitel’noj dejatel’nosti v pozdnem mezolite i rannem neolite Volgo-Okskogo mezhdurech’ja po materialam stojanki Zamost’e 2. Problemy istorii, filologii, kul’tury 2: 208–226. (in Russian) 2019. Razvitie kostjanyh nakonechnikov strel v VII–V tys. cal BC v Verhnevolzhskoj nizmennosti po materia-lam mnogoslojnoj stojanki Zamost’e 2. Kratkiye soob­shcheniya Instituta arkheologii 255: 71–90. (in Rus­sian) http://ksia.iaran.ru/wp-content/uploads/2019/07/ ksia-255-redu.pdf#page=71 Lozovsky V. M., Lozovskaya O. V. 2010. Izdelija iz kosti i roga ranneneoliticheskih sloev stojanki Zamost’e 2. In I. S. Kamenetsky, A. N. Sorokin (eds.), Chelovek i drevno­sti. Pamyati Aleksandra Aleksandrovicha Formozova (1928–2009). Grif & Co. Moskva: 237–252. (in Russian) Mazurkevich A. N., Mikljaev A. M. 1998. O rannem neoli­te mezhdurech’ja Lovati i Zapadnoj Dviny. Arheologiche­skij sbornik Gosudarstvennogo Jermitazha 33: 7–31. (in Russian) Mazurkevich A. N., Kul’kova M. A., Polkovnikova M. J., and Savel’eva L. A. 2003. Ranneneoliticheskie pamjatniki Lo-vatsko-Dvinskogo mezhdurech’ja. In V. I. Timofeev (ed.), Neolit-Eneolit Juga i Neolit Severa Vostochnoi Evropy. Institut istorii material’noi kul’tury, Rossiyskaya Akademi-ya nauk. Sankt-Petersburg: 260–268 (in Russian) Mazurkevich A., Dolbunova E. 2009. Reconstruction of the Early and Middle Neolithic settlement systems in the Upper Dvina region (NW Russia). In J. W. H. Verhagen, A. G. Po-sluschny, and A. Danielisova (eds.), Go Your Own Least Cost Path. Spatial Technology and Archaeological Inter­pretation. Proceedings EAA 2009. Riva del Garda. BAR International Series 2284. Archaeopress.Oxford: 25–32. 2015. The oldest pottery in hunter-gatherer communi­ties and models of Neolithisation of Eastern Europe. Documenta Praehistorica 42: 13–66. https://doi.org/10.4312/dp.42.2 Mazurkevich A. N., Dolbunova E. V., Zaitseva G. I., and Kulkova M. A. 2017. Chronological timeframes of cultu­ral changes in the Dnepr-Dvina region (7th to 3rd millen­nium BC). Documenta Praehistorica 44: 162–175. https://doi.org/10.4312/dp.44.10 Mazurkevich A., Kittel P., Maigrot Y., +3 authors, and Piech W. 2020. Natural and anthropogenic impact on deposits’ formation in the wetland shore area: case study from the Serteya site, Western Russia. Acta Geographica Lodzien­sia 110: 81–102. Miksaite S. 2005. Production of ceramics of Narva culture (reconstructions based on experimental archaeology). Es­tonian Journal of Archaeology 9(1): 60–72. Mroczkowska A., P³óciennik M., Paw³owski D., +10 au­thors, and Kittel P. 2020. Northgrippian climate oscilla­tions recorded at the Western Dvina Lakeland (Serteyka Valley). Water 13(11): 1611. Nedomolkina N. G., Piezonka H., Meadows J., Craig O., and Lorenz S. 2015. Neolithic complexes of the Veksa The end of the earliest ceramic traditions> Dnieper-Dvina region became part of the Circum-Baltic space ... sites in the Upper Sukhona basin, north-western Russia: New natural-scientific research. In V. M. Lozovski, O. V. Lo-zovskaya, A. A. Vybornov (eds.), Neolithic Cultures of Eastern Europe: Chronology, Paleoecology and Cultural Traditions. Materials of the international conference dedicated to the 75th anniversary of Victor Petrovich Tretyakov, May, 12–16, 2015. St. Petersburg. Institut isto­rii materialnoy kulturi. Rossiyskaya akademiya nauk. Go-sudarstvenniy Ermitazh. Povolzhaya gosudarstvennaya socialno-gumanitarnaya akadeiya. Muzei antropologii i et-nografii im. Petra Velikogo (Kunstkamera). Sankt-Peter-burg: 151–158. Nowak M. 2017. The second stage of neolithisation and para-neolithic in the southern Baltic. Samarskiy nauchnyi vestnik 6(4): 116–123. Pääkkönen M., Bläuer A., Evershed R. P., and Asplund H. 2016. Reconstructing food procurement and processing in early Comb ware period through organic residues in early comb and Jäkärlä ware pottery. Fennoscandia ar­chaeologica XXXIII: 57–75. Papakosta V., Oras E., and Isaksson S. 2019. Early pottery use across the Baltic – A comparative lipid residue study on Ertebolle and Narva ceramics from coastal hunter-ga­therer sites in southern Scandinavia, northern Germany and Estonia. Journal of Archaeological Science: Reports 24: 142–151. https://doi.org/10.1016/j.jasrep.2019.01.003 Pétrequin A-M., Pétrequin P. (eds.). 2006. Objets de pou­voir en Nouvelle-Guinée. Approche ethnoarchéologique d’un systeme de signes sociaux. Catalogue de la dona­tion Anne-Marie et Pierre Pétrequin. Réunion des musé-es nationaux. Paris. Piezonka H. 2015. Jäger, Fischer, Töpfer. Wildbeuter mit früher Keramik in Nordosteuropa im 6. und 5. Jahrtau-send v. Chr. Archäologie in Eurasien 30. Deutsches Archä­ologisches Institut. Eurasien-Abteilung. Habelt-Verlag. Bonn. Pili.iauskas G. 2002. Dubi.iu tipo gyvenvietes ir neolitine Nemuno kultura pietu Lietuvoje. Lietuvos Archeologija 23: 107–136. Povlsen K. 2013 The introduction of ceramics in the Erte­bolle culture. Danish Journal Archaeology 2: 146–163. https://doi.org/10.1080/21662282.2013.904127 Raemaekers D. C. M. 2011. Early Swifterbant pottery (5000–4600 cal BC): research history, age, characteristics and the introduction of pottery. In S. Hartz, F. Luth, and T. Terberger (eds.), Early pottery in the Baltic – Dating, Origin and Social Context. International Workshop at Schleswig from 20th to 21st October 2006. Bericht der Rö­misch-Germanische Kommission. Band 89, 2008. Römisch- Germanische Kommission des deutschen aachäologischen instituts. Philipp von Zabern: 89–110. Rimantiene R. K. 1973. Neolit Litvy i Kaliningradskoj ob-lasti. Jetnokul’turnye oblasti lesnoj i lesostepnoj zony ev­ropejskoj chasti SSSR v jepohu neolita. Materiali i Issledo­vaniya po Arkheologii SSSR 172: 218–225. (in Russian) 1992. The Neolithic of the Eastern Baltic. Journal of World Prehistory 6(1): 97–143. https://doi.org/10.1007/BF00997586 Sablin M. V., Panteleev A. V., and Syromjatnikova E. V. 2011. Arheozoologicheskij analiz osteologicheskogo ma-teriala iz neoliticheskih svajnyh poselenij Podvin’ja: Hoz­jajstvo i jekologija. Trudy Zoologicheskogo instituta Ros-siyskoy akademii nauk 315(2): 143–153. (in Russian) Skorobogatov A. M., Smolyaninov R. V., Surkov A. V., Oino­nen M., and Possnert G. 2016. Khronologiya neolitiches­kikh pamyatnikov v lesostepnoy zone Dona. In A. N. Ma-zurkevich, M. A. Kulkova, and E. V. Dolbunova (eds.), Ra-diouglerodnaya hronologiya epohi neolita Vostochnoi Evropy v VII–III tys. do n.e. Gosudarstveniy Ermitaz. In-stitut materialnoy kulturi, Rosiyskaya akademiya nauk. Samarskiy gasudarstveniy socialno-pedagogicheskiy uni-versitet. Svitok. Smolensk: 244–260. (in Russian) Timofeev V. I. 1993. Pamjatniki mezolita i neolita regiona Peterburga i ih mesto v sisteme kul’tur kamennogo veka Baltijskogo regiona. In Drevnosti severo-zapada Rossii (slavjano-finno-ugorskoe vzaimodejstvie, russkie goroda Baltiki). Saint-Petersburg: 8–34. (in Russian) 1998. The east-west relations in the Late Mesolithic and Neolithic in the Baltic region. Baltic-Pontic studies 5: 44–58. 1998a. The beginning of the Neolithic in the Eastern Baltic. In M. Zvelebil, L. Domañska, and R. Dennell (eds.), Harvesting the sea, farming the forest. The emergence of Neolithic societies in the Baltic region. Academic Press. Sheffield: 225–236. Tkachou A. 2018. Early Neolithic pottery from western Belarus. Archaeologia Baltica 25: 82–99. Torvinen M. 2000. Säräisniemi 1 Ware. Fennoscandia Archaeologica XVII: 3–35. Vankina L. V., Zagorskis F. A., and Loze I. A. 1973. Neoli­ticheskie plemena Latvii. Jetnokul'turnye oblasti lesnoj i lesostepnoj zony evropejskoj chasti SSSR v jepohu neoli­ta. Materiali i Issledovaniya po Arkheologii SSSR 172: 210–217. (in Russian) Andrey Mazurkevich, Ekaterina Dolbunova, Yolaine Maigrot, and Veronika Filippova Vankina L. 1999. The collection of stone age bone and antler artefacts from lake Lubana. Latvijas Vestures Mu-zeja raksti 4. N.I.M.S. Riga. Vybornov A. A., Yudin A. I., Kulkova M. A., Goslar T., Pos­snert G., and Philippsen B. 2016. Radiouglerodnyye dan­nyye dlya neoliticheskoy khronologii basseyna Nizhney Volgi. In A. N. Mazurkevich, M. A. Kulkova, and E. V. Dol­bunova (eds.), Radiouglerodnaya hronologiya epohi neo­lita Vostochnoi Evropy v VII–III tys. do n.e. Gosudarst­veniy Ermitaz. Institut materialnoy kulturi, Rossiyskaya akademiya nauk. Samarskiy gasudarstveniy socialno-pe­dagogicheskiy universitet. Svitok. Smolensk: 62–73. (in Russian) Wawrusiewicz A., Kalicki T., PrzeŸdziecki M., Fr¹czek M., and Manasterski D. 2017. Gr¹dy-Woniecko. Ostatni ³owcy­zbieracze znad œrodkowej Narwi. Bia³ystok Wawrusiewicz A. 2013. Ceramika typu Soko³ówek na Pod-lasiu i jej znaczenie w rozwoju spolecznoœci subneolitycz­nych Polski Pó³nocno-Wschodniej. Studia i materialy do badañ nad neolitem i wczesn¹ epok¹ br¹zu na Mazow­back to contents szu i Podlasiu III: 5–24. Wieckowska-Lüth M., Gauthier E., Thiebaut E., +6 authors, and Kittel P. 2021. The palaeoenvironment and settlement history of a lakeshore setting: An interdisciplinary study from the multi-layered archaeological site of Serteya II, Western Russia. Journal of Archaeological Science: Re­ports 40 Part B: 103219: 1–17. https://doi.org/10.1016/j.jasrep.2021.103219 Zagorskis F. 1973. Agrais neolita laikmets Latvijas aus­trumdala. Latvijas PSR Zinatnu Akademijas Vestis 4: 56– 69. Zhilin M. G., Kostyleva E. L., Utkin A. V., and Engovatova A. V. 2002. Mezoliticheskie i neoliticheskie kul’tury Verh­nego Povolzh’ja (po materialam stojanki Ivanovskoe VII). Moskva. (in Russian) Zhilin M. 2004. Prirodnaya sreda i hozyajstvo mezoliti­cheskogo naseleniya centra i severo-zapada lesnoj zony Vostochnoj Evropy. Academia. Moskva. (in Russian) Zvelebil M. 2008. Innovating hunter-gatherers: the Mesoli­thic in the Baltic. In G. Bailey, P. Spikins (eds.), Mesolithic Europe. Cambridge University Press. Cambridge: 18–59. Documenta Praehistorica XLIX (2022) A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida Nina Kyparissi-Apostolika1, Evita Kalogiropoulou2, Dimitris Roussos3, Niki Saridaki 4, Odysseas Metaxas5, Georgia Kotzamani 6, Katerina Trantalidou 7, and Yorgos Facorellis8 nkyparissi@hotmail.com 1. Emerita director, Ephorate of Palaeoanthropology and Speleology, Hellenic Ministry of Culture and Sports, Athens, GR 2. University of Crete, Rethymnon, GR< ekalogir79@gmail.com 3. National and Kapodistrian University of Athens, Faculty of Geology and Geoenvironment, Athens, GR< droussos@geol.uoa.gr 4. Aristotle University of Thessaloniki, Faculty of Philosophy, School of History and Archaeology, Thessaloniki, GR< n.saridaki@yahoo.gr 5. Independent researcher, GR< odymetaxas@hotmail.com 6. Independent researcher, GR< g.kotzamani@gmail.com 7. Hellenic Ministry of Culture and Sports, Athens, GR< ktrantalidou@yahoo.gr 8. Department of Antiquities and Works of Art Conservation, School of Applied Arts and Culture, University of West Attica, Athens, GR< yfacorel@uniwa.gr ABSTRACT – This paper presents the first known and systematically excavated Middle Neolithic pot­tery workshop in southwestern Thessaly at Imvrou Pigadi. The excavations and in situ finds, along with the pronounced kiln structures, their typological classification and pyrotechnological operation, suggests considerable expertise in pottery manufacture. The pottery itself, together with the chipped stone industry and other small finds, as well as the fauna and archaeobotanical assemblages are pre­sented. The results of the 14C dating programme point to use of the workshop at the beginning of the 6th millennium. All this evidence suggests an active settlement where pottery production was carried out, which was then circulated within the wider region. KEY WORDS – Neolithic; Thessaly; workshops; pyrotechnology; specialised production; kilns; pottery; exchange networks Srednjeneolitska lon;arska delavnica v magoula Imvrou Pigadi na sti;i[;u vzhodne in zahodne Tesalije in Ftiotide IZVLE.EK – V .lanku predstavljamo prvo znano in sistemati.no izkopano lon.arsko delavnico iz srednjega neolitika v jugozahodni Tesaliji pri Imvrou Pigadi. Izkopavanja in najdbe in situ skupaj z deli lon.arske pe.i, njihovo tipolo.ko razvrstitvijo in pirotehnolo.kim delovanjem, ka.ejo na precej.­nje poznavanje lon.arskih tehnik. Predstavljamo lon.enino, odbitkovna kamena orodja, druge drob­ne najdbe ter .ivalske in arheobotani.ne zbire. 14C datiranje ka.e na delovanje delavnice na za.et­ku 6. tiso.letja pr. n. .t. Vse to pri.a o aktivni naselbini, v kateri se je izdelovala lon.enina, ki se je nato .irila v regiji. KLJU.NE BESEDE – neolitik; Tesalija; delavnice; pirotehnologija; specializirana proizvodnja; pe.i; lon.arstvo; izmenjevalna omre.ja DOI> 10.4312\dp.49.16 A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida Introduction Until recently, western Thessaly was widely believed to have been less intensively occupied than the east­ern part of the Thessalian plain during the Neolithic. This perception was due to limited research and the large-scale land-reform project carried out in the lowlands of Karditsa during the 1960s, leading to more intensive farming and irrigation which flat­tened the plain and buried several tells (magoula). In fact, the area has many Neolithic sites, which form the predominant settlement type. In no way could we have imagined then that Imvrou Pigadi would produce the first complex of pottery kilns ever found in either Thessaly, or Greece as a whole, dated to as early as the Middle Neolithic. As indicat­ed by the geophysical survey, during which rectan­gular anomalies representing prehistoric buildings were identified, the tell itself is very extensive, with the rest of the deposits probably belonging to a typi­cal settlement (Tsokas et al. 2009). If the excava­tions had only been carried out within the more central part of the tell, it is almost certain that the pottery workshops would not have been discovered. The workshop complex at Imvrou Pigadi was first referred to during 2o AETHSE (Kyparissi-Apostoli­ka 2006) and then later in Documenta Praehistori-ca as indications of the presence of relevant activity in the area (Kyparissi-Apostolika 2012). This excep­tional discovery was actually presented in a preli­minary report, about which we were initially rather sceptical as there were no similar examples to com­pare it with. In the present paper some points which were less clearly described in the previous publica­tions are now clarified. Additional research at the site has revealed an extensive pottery workshop spread over two different but adjacent levels, with a height difference between the two of approximately 2m, covering around 50m2, as far as we currently know. A total of 23 structures have been identified, although more are believed to be located under the adjacent unexcavated deposits. Our investigations included excavation, geophysical survey and several different analytical approaches to the datasets. In this paper we present some of the results of the ex­cavations and the study of the material recovered, while focusing in parallel on the cultural context of the settlement and its special role in the area. The site and the excavations Based on a surface survey, Imvrou Pigadi was first referred to as a prehistoric site by Sofia Dimaki (1994). Located at an elevation of 165masl on the southwestern edge of the Thessalian Plain (Fig. 1), the site is administratively located in Phthiotida, Neo Monastiri, in the municipality of Domokos. It lies at the crossroads between eastern and western Thessaly and Phthiotida, and is also very close to the N. Kyparissi-Apostolika, E. Kalogiropoulou, D. Roussos, N. Saridaki, O. Metaxas, G. Kotzamani, K. Trantalidou, and Y. Facorellis railway linking Athens and Thessaloniki, just over a kilometre north of the Domokos station. Other Neolithic tells are visible in the area, amongst which Koutroulou stands out (Kyparissi-Apostolika 2003; Hamilakis, Kyparissi 2012; Hamilakis et al. 2017), while many other Neolithic settlements are located in the adjacent western part of the Thessa­lian Plain in Karditsa (Kyparissi-Apostolika 2019; Orengo et al. 2015; Krachtopoulou et al. 2020). Imv­rou Pigadi is one of the most outstanding tells in the area because of its significant extent and height, at over 4m above the road. Based on the geophysi­cal survey, the site extends to at least 14 000 square metres (Tsokas et al. 2009), although since part of its western side was cut during construction of the local road and railway, it may well have been even larger, 22 326 square metres according to Dimaki (1990). The tell is located in a private field where cotton has been regularly cultivated. The excavations were begun on the western upper edge of the site in 2002, and were concluded in 2019.1 They consisted of four trenches (1, 2, 3 and a/2016), with pottery kilns in all except for trench 3 (Fig. 2). Trench 1 was 5m by 5m square and be­gan with a deposit ranging from 1.03m to 1.5m in thickness (phase A). This produced evidence for pos­sible domestic activities, as defined by the pottery (see below, in the section on the pottery evidence), along with some figurines and obsidian and flint tools. At this depth, a yellowish-red clay surface was identified, beginning in the northeastern corner of the trench and sloping down towards the south. It Fig. 2. Trenches 1, 2 and a/2016 showing the location of all structures (drawn by Chryssoula Founda of the Ephorate of Antiquities of Phtiotida, with digital processing by Jamie Donati). 1 The excavator is thankful to the archaeologists Nana Almatzi, Eleni Froussou, Sophia Ntafou, Stavroula Gaga, Evita Kalogiropoulou and Niki Saridaki, who as trench supervisors worked very carefully, recording invaluable information that helped in document­ing the special character of the site. A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida formed all of the eastern side and most probably continues beyond the edges of the trench towards the east and north. In the rest of the trench, parts of this original hardened clay surface were also found. Beneath it, damaged clay structures were unco­vered (phase B1) (Kyparissi-Aposto­lika 2006.Fig. 2), at depths ranging from 1.5m to 2.58m. Amongst these, significant quantities of sling bullets were found, as well as brick frag­ments with surfaces and corners at­tributed to the structures. Nine struc­tures in total were identified in this trench (Nos. 8–16), along with two clay boxes and one clay basin, while more structures were seen along the sides of the trench in unexcavated deposits (see the detailed description below, in the section on clay-based structures). The pottery recovered was mainly monochrome, although examples typical of Middle Neolithic dec­orated wares were also present (see below, in the section on the pottery evidence). Between two struc­tures (Nos. 8 and 10) a painted fruit stand was re­covered, possibly a foundation offering (Fig. 3). The excavations in trench 1 were halted at a depth of 2.5m in order to clarify the nature of the structures. Trench 2 measured 7.8m by 2.6m and was opened to the west of trench 1, by the edge of the local road. It was around 2m below the excavated level of trench 1 and also below the local road, about 4.27m from the top of the tell, extending into the eastern draining ditch of the road. The structures in trench 2 were iden­tified as kilns and located within an area of approximately 15 square metres. These were bounded to the north, east and south by solid argil­laceous walls, with the eastern leg only revealed in 2018, so not re­ferred to in the earlier publications (see Fig. 2). The walls were light yellowish brown to brownish yel­low (Munsell 10YR 6/4 – 5/4 and 6/6 respectively) in colour, with the largest and strongest along with northern side, with a visible length of 4.3m. At its eastern end the wall turned south, forming the eastern branch. This continued for 1.37m before disappearing under a thick section of unexca­vated deposits for 4.5m, before meeting the south­ern branch, so an estimated total length of 5.87m. The south wall measured around 2m in length, with its eastern end abutting the eastern wall, the latter estimated to be around 0.9m long, but also covered by the unexcavated deposits. The western branch of the wall was completely buried beneath the local road and therefore unexcavated. For the moment the full lengths of the walls remain unknown, as well as where the western wall is located within the probable rectangular shaped enclosure. For this rea­son, the total extent of the kiln structures will un­doubtedly be larger than that observed during the excavations. The width of the walls differed consi­derably, with the northern branch measuring from 0.5m to 0.68m at its western and eastern ends res- N. Kyparissi-Apostolika, E. Kalogiropoulou, D. Roussos, N. Saridaki, O. Metaxas, G. Kotzamani, K. Trantalidou, and Y. Facorellis pectively. The western part of the northern wall must have been con­structed in two phases, one on top of the other, where it reached a thickness of 0.4m, while its eastern part measured only around half that thickness. In this latter area, the stones seen beneath the argillaceous wall (Fig. 4) at a depth of 4.71m may represent an earlier phase, or they may have been placed in order to provide a foundation. The width of the eastern wall where it was visible was 0.6m, while the southern wall was 0.5m to 0.6m wide. The northern wall is at the elevated level of 4.23m compared to the southern (5.36m), with a dif­ference of about 1.18m. Between 0.2m and 0.26m outside of the southern wall was another feature of intense bright red colour (Fig. 5). Apart from some random pieces of pottery and a few bones, it was covered only by ashy deposits. It was subsequently determined that this red fea­ture was actually the edge of a ditch, consisting of tiny pieces of pottery and hardened by ash, rather than another similar complex of kilns extending to the south of those seen in trench 2, as originally sug­gested (Kyparissi-Apostolika 2012.435). Instead, it consisted of ash thrown from the kiln complex to the north, as can be seen in the sloping ashy de­posits in the profile. The ditch was excavated to a depth of 5.83m, producing nothing apart from ash and charcoal, and it probably con­tinues further towards the south be­yond the edge of trench 2. Enclosed within the argillaceous walls were eight structures (Phase B2) (Nos. 1–7 and 20) and four clay boxes (see below, in the section on clay-based structures), while more structures were also visible within the east and the west profiles of the trench, extending into unexcavated deposits. The structures were cov­ered by soil of about 0.5m thick, with colours ranging from red to dark red, while in areas with combu­stion residues, it was reddish black. The red material in this deposit was friable and dry, and included pieces of broken clay with reed imprints and other flat sur­faces, giving the impression of destruction by fire. At a depth of between 5.12m and 5.20m the top surfaces of several structures were uncovered. The first strong indication of such a kiln structure was found within the corner between the northern and eastern walls, just a few centimetres below the up­per surface of the wall. We excavated kiln No. 20 (Fig. 6) and found two vertically plastered clay walls of relatively low height forming a corner, together with a flat clay surface at the base, as previously de­scribed (Kyparissi-Apostolika 2012.434, Figs. 2a,b). At the time, all of this newly discovered burnt mate­rial was attributed to the interior of a house, rather than a kiln. This conclusion was based on the limit- A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida ed amount of evidence we had at the time, as the excavations had only just begun and had not yet been extend into trench 2. This structure continued to the south beyond the edge of the trench, where it is still visible today in the profile of the unexca­vated deposits. Pottery from amongst the structu­res was very dense and almost all was monochrome, with some from complete small and large vessels (Kyparissi-Apostolika 2012.Figs. 4, 5). Several rec­tangular clay containers measuring 22cm by 23cm by 26cm high were also found, probably used as storage boxes (Kyparissi-Apostolika 2012.Fig. 3b). Close to the northwestern corner of the enclosed complex and close to structure No 5 at a depth 4.92m, a four legged kiln figurine with two openings was found, with one of the legs missing (Fig. 7). Also recovered from within the kilns were some deer antlers and a few pieces of bone, possibly used as fuel. Some obsidian and flint tools were found, along with a few additional figurines. An exceptional example was structure No 2 where deer antlers of various sizes were found, not by chance but purpo­sefully and carefully placed (Fig. 8). At the bottom of the structure a fire had been set, with traces of burning on some of the antlers. The burnt layer of intense red colour found outside of the northeastern corner of the walled enclosure, along with the slightly darker ones from within, both point to temperatures ranging between 600 and 700°C (Roussos, Kyparrisi 2019). This range is si­milar to those that have been proposed for the kiln structures within the enclosure (Kalogiropoulou et al. in press), further suggesting that the burnt lay­ers may be discarded burnt sediments from the kilns mixed with ashes. In order to investigate the relationship between the two pottery workshop areas in trenches 1 and 2, which differed in height by around 2m, trench a/ 2016 was excavated in the gap between, correspond­ing almost exactly with an earlier stratigraphic clea­rance area of 2.2m by 2.1m square (Kyparissi-Apo­stolika 2012.433; 2006.Fig. 8). The thickness of the deposits was 2.75m at the eastern edge of trench 1 at a depth of 2.58m, where the excavation was halt­ed, while to the west the new trench abutted the argillaceous wall of trench 2 at a depth of 4.22m. The linking trench was begun in 2016 and complet­ed in 2018. Pottery finds included mainly mono­chrome sherds, along with a few flint and obsidian tools, as well as some figurines and bones. Another argillaceous wall was found at a depth of 3.05m, crossing trench a/2016 from north to south. Al­though not as robust as those seen in trench 2, it ranged from 0.64m to 0.79m in width (see Fig. 2). Another clay kiln structure, or possibly the remains of a failed one, was found at the east­ern end of the trench, along with another which disappeared into the profile of the northeastern corner. The argillaceous walls and these two structures were at an intermediate elevation compared to the workshop areas in trenches 1 and 2, hereafter referred to as phases B1 and B2, res­pectively. So the intermediate work­shop area was 0.5m below B1 and 1.2m above B2, representing either an additional phase, or possibly the initial working surface of B1. Trench 3 was opened in 2007, almost 20m to the northeast of trench 1. N. Kyparissi-Apostolika, E. Kalogiropoulou, D. Roussos, N. Saridaki, O. Metaxas, G. Kotzamani, K. Trantalidou, and Y. Facorellis Measuring 5m by 5m square, its purpose was to check for possible similarities or differences in the use of space, as the geophysical survey had high­lighted the presence of several rectangular anoma­lies in this area (Tsokas et al. 2009.1261–1267). At a depth of approximately 1m, a ditch of 0.5m to 0.7m wide was found crossing the trench from north-east to south-west. Filled with loose soil, the ditch is recent as modern material was recovered, and according to local knowledge the area was in use during World War 2. Due to the presence of amorphous compact layers attributed to the effects of water in the southern part of the trench, the ex­cavations were continued only in the north, where they were halted at a depth of 1.9m. The most dis­tinctive elements of this deposit were the abundance of charcoal, the presence of mostly monochrome pottery, along with decorated wares typical of the Middle and early Late Neolithic (see the section on the pottery evidence). Figurines and ornaments, as well as several obsidian and flint tools, were also re­covered. The excavations in trench 3 have been tem­porarily halted, with no structures like those seen in the other three trenches found so far. Although located close to the pottery workshops, this area ap­pears to have had a more domestic function, possi­bly supplementary to or supporting the activities carried out in the workshops. Chronology During fourteen years of excavations from 2003 until 2017, numerous pieces of charcoal were collected. A total of eight samples were selected for radiocar­bon dating, three by the radiocarbon unit of the La­boratory of Archaeometry of the Institute of Nano-science and Nanotechnology of the National Centre for Scientific Research in Athens (Demokritos), using the gas counting technique (Facorellis et al. 1997). The other five were dated by AMS at the Accelerator Mass Spectrometry Laboratory of the University of Arizona in Tucson (Donahue et al. 1990a; 1990b). The conventional 14C ages were calibrated using the latest version of the IntCal20 calibration curve (Rei-mer et al. 2020) and OxCal v4.4.4 software (Bronk Ramsey 2009). The results are presented in Table 1, along with the probability distribution plots of the calibrated ages (Fig. 9). The calendar ages of the samples from trench 1 (DEM-1435 and AA112953) ranged from 5888–5726 to 5985–5836 BC (within 2s). They were derived from almost the same area from square A1 in the southeastern corner of trench 1, although with a dif­ference in depth of approximately 0.4m. Trench 2 produced three dates within 2s (DEM-1434, AA­112955 and DEM-1402). The first was from a char­coal sample within the southern leg of the argilla­ceous wall, with a calendar age of 5968–5747 BC. The second was derived from the area south of the clay structures, close to clay box 2 and at almost the same depth as the first sample, producing a calen­dar age of 5987–5786 BC. The third sample was de­rived from a deeper level (6.2–6.32m) just outside of the northern leg of the argillaceous wall, produc­ing a calendar age of 5618–5474 BC. This probably Fig. 9. Probability distribution plots for the calibrated dates from Imvrou Pigadi, sorted by increasing age. A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida Laboratory code Sample ID and location Collection date Sample type d 13C (‰) 14C age (yr B.P.) Calendar age (yr B.C.) DE-1402 Samples 34, 35, 36, Trench 2 at the side of the road, depth 6.20–6.32 15.10.2003 Charcoal –25 6567±40 5553–5479 (1s) 5618–5474 (2s) DE-1435 Sample 1, Trench 1, P11, A1, Depth 1.66 29.9.2003 Charcoal –25 6923±36 5835–5745 (1s) 5888–5726 (2s) AA112954 Trench 3, Square B4, Pass 7, East> 1.38, North> 3.35, depth 1.48 29.11.2011 Charcoal –24,5 6928±24 5836–5755 (1s) 5883–5733 (2s) DE-1434 Sample 24, Trench 2 at the side of the road, Depth 4.7 29.9.2003 Charcoal –25 6962±25 5887–5795 (1s) 5968–5747 (2s) AA112957 Sample 3, Trench a\2016, Pass 47, North> 0.92, East> 0.75, depth 4.27 14.9.2018 Charcoal –25,9 6993±90 5982–5782 (1s) 6056–5716 (2s) AA112955 Trench 2 (b\2016), East of clay box 2 8.6.2016 Charcoal –25,5 7008±40 5978–5842 (1s) 5987–5786 (2s) AA112953 Trench 1, Square A1, South> 0.8, West> 0.3, depth 2.02 5.12.2012 Charcoal –24,5 7016±25 5977–5848 (1s) 5985–5836 (2s) AA112956 Sample 3, Trench a\2016, Pass 36, depth 3.4 11.10.2017 Charcoal –25,3 7069±35 5992–5901 (1s) 6021–5847 (2s) Tab. 1. Radiocarbon dates from Imvrou Pigadi Neolithic tell, sorted by increasing age. suggests a different activity which occurred around 250 years later than those activities seen within the enclosing walls. Two other charcoal samples (AA112956 and AA­112957) with a depth difference of 0.87m from trench a/2016 produced overlapping calendar ages within 2s, of 6021–5847 BC (depth 3.4m), and within 2s, of 6056–5716 BC (depth 4.27m). They were consistent with the two samples from trench 2 (DEM 1434 and AA112955), suggesting overlap­ping use of the areas uncovered in trenches 1, 2 and a/2016. A final sample from trench 3 (AA112954) produced a calendar age within 2s, of 5883–5733 BC, which coincides with the dates from the workshops, again suggesting contemporary use, although without the kiln structures. This area may have played a suppor­tive role for the workshops and/or domestic activi­ties. Coring Coring was carried out during 2012 in order to trace the outer edge of the tell, to explore the underlying stratigraphy, to measure the depth of the original paleosoil of the valley, and if possible, to trace any deeper deposits that might be linked with the pyro­technical processes. Four cores were drilled down on a north south axis through trench 2, the first 18m to the north and a further two 10m and 16m to the south. The fourth was drilled at the deepest point in the northeastern corner of trench 2. All of the cores within the walls and those to the south reached the paleosoil at depths of between 7.76m and 8.54m from the top of the tell. The north­ernmost core reached a maximum depth of 7.35m without hitting the paleosoil. Anthropogenic mate­rial including charcoal, ash and pottery fragments were observed in all cores from top to bottom, indi­cating continuous use of the tell throughout the Mid­dle Neolithic and possibly extending the chronology of the site even further back in time. Deposits that could be linked with pyrotechnical activities were located in the two outer cores only, north and south of the excavated area, at depths similar to those seen in the excavated kiln area, while the internal core did not reveal any deeper deposits. This suggests that similar workshop areas may exist towards the north and south, but the kiln area in trench 2 is pro­bably the earliest workshop phase uncovered so far. The natural fine-grained deposits found mixed with fine anthropogenic material at around 5.5m and 6.5m may suggest two small-scale flooding epi­sodes. The clay-based structures, their typology and pyrotechnological characteristics A total of 23 clay structures were recorded, predo­minantly in the walled areas on the western edge of the settlement, with most attributed to the Middle Neolithic cultural horizon identified in that area. They were grouped into two general categories, the first including structures dedicated to the use of fire, like kilns and hearths made of clay. The second cat­egory included large containers like clay boxes and basins (Tab. 2). N. Kyparissi-Apostolika, E. Kalogiropoulou, D. Roussos, N. Saridaki, O. Metaxas, G. Kotzamani, K. Trantalidou, and Y. Facorellis ID* Trench Type Shape Size (m) ** Location of the entrance *** Construction elements TS 1 Trench 2 Hearth Rectangular 1 x 0.75 x 0.12 East Single heating surface TS 2 Trench 2 Kiln Horseshoe 0.85 x 0.76 5 heating surfaces TS 4 Trench 2 Kiln Horseshoe 1.26 x 0.98 x 0.42 West 2 heating surfaces, 3 wall plasters, outer plastic leaf-like decoration TS 5 Trench 2 Kiln Horseshoe 1.3 x 0.76 x 0.55 2 wall plasters, Apron TS 6 Trench 2 Kiln Horseshoe 1.05 x0.9 x 0.12 East 2 heating surfaces TS 7 Trench 2 Kiln Horseshoe 0.78 x 0.82 x 0.47 West, partly excavated Wall plasters, Apron TS 8 Trench 1 Kiln Horseshoe 1.08 x 0.85 x 0.38 Northeast 4 heating surfaces, plas­tered outer dome sur­faces, Apron TS 9 Trench 1 Kiln Horseshoe 1.33 x 0.9 x 0.10 North 3 heating floors TS 10 Trench 1 Kiln Horseshoe 0.85 x 0.74 x 0.27 Northeast 4 heating surfaces, 2 wall plasters TS 11 Trench 1 Kiln Horseshoe 0.54 x 0.5 x 0.10 West, partly excavated 3 heating surfaces, 5 wall plasters TS 12 Trench 1 Kiln Horseshoe 0.82 x 0.71 x 0.40 West, partly excavated 3 heating surfaces, Apron TS 13 Trench 1 Kiln Irregular 0.58 x 0.55 x 0.43 North, partly excavated 8 heating surfaces, 6 wall plasters TS 14 Trench 1 Clay basin Vertical walls 1 x 0.82 x 0.05 East Low sidewalls TS 15 Trench 1 Kiln Irregular Partly excavated TS 16 Trench 1 Hearth Horseshoe 0.8 x 0.78 x 0.35 partly excavated Northwest, partly excavated Low side walls, Apron TS 18 Trench 2 Hearth Irregular 1.12 dia, height 0.09 Unexcavated Single heating surface TS 19 Trench 2 Kiln Unexcavated CB 1 Trench 2 Clay box Rectangular 0.47 x 0.36 x 0.22 North CB 2 Trench 2 Clay box Rectangular 0.49 x 0.42 x 0.18 North CB 3 Trench 2 Clay box Rectangular 0.51 x 0.4 x 0.20 West CB 4 Trench 2 Clay box Square 0.32 x 0.32 x 0.13 CB 5 Trench 1 Clay box Rectangular 0.44 x 0.4 x 0.09 East CB 6 Trench 1 Clay box Rectangular 0.29 x 0.2 x 0.25 Southwest 3 plastered wall surfaces Tab. 2. The 23 clay-based structures identified at Neolithic Imvrou Pigadi (*TS=Thermal structure, CB= Clay box). Kilns were the most numerous and impressive group, with 13 in total, while also present were three hearths, six clay boxes and a single clay basin. The preservation of the structures ranged from very good to heavily eroded, while the entire assemblage reflects a major investment of time and labour and a high degree of structural uniformity and sophisti­cation. The kilns were studied both macroscopical­ly and through several geoarchaeological approa­ches, in order to investigate their construction and use (Germain-Vallée, Prévost-Dermarkar, Lespez 2011; Mentzer 2014; Mallol, Mentzer, Miller 2017; Roussos, Kyparissi-Apostolika 2019). The kilns consisted of above ground horseshoe-shaped structures with a solid vaulted compartment and front-loading entrance (Prévost-Dermarkar 2002; Kalogiropoulou 2013). On average, they were 1m in length by 0.9m in width, with a typical open­ing of 0.8m wide. They were made of local clay en­riched with straw binding and the solid vaulted com­partments were around 0.4m in height, with the dome built using the coiling technique or with a wooden mesh inner frame. The apron formed a pro­minent construction feature, preserved in four cases and consisting of a flat working bench beneath the entrance. One of the characteristic aspects of the kilns was that both the floors and walls of the domes were subject to multiple repairs. There were up to four phases of replastering of the smooth heating floors, with a consistent thickness of between 2cm to 3cm, set upon a substantially thicker initial foun­dation layer of up to 6cm thick. Micromorphological analysis revealed a striking absence of charcoal or any other combustion by-products between the indi­vidual plaster layers of the floor, suggesting that the structures were thoroughly cleaned before being re­paired. The walls followed a similar pattern of re­ A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida pair in terms of frequency. Apart from TS4 that pre­served a relief leaf-like decoration on its sidewall, the kilns were probably not ornamented. All the above observations point to careful and labour-in­tensive construction, maintenance and repair of the kilns, suggesting a significant element of expertise and standardisation. Another characteristic of the kilns is that they were all fired in oxidizing condi­tions, as indicated by their consistent bright red co-lour. That this uniformity in combustion is unrelat­ed to the thickness of the walls and floors suggests high burning temperatures and probably repeated use of the structures. The analytical methods used indicated that the majority of the kilns were re­peatedly heated to temperatures ranging between 700 and 1000°C (Roussos, Kyparissi-Apostolika 2019; Conati Barbaro et al. 2021). These observa­tions suggest a controlled burning environment and a sophisticated le­vel of pyrotechnology for the Mid­dle Neolithic. In the same way as the kilns, hearths were also above ground features with smooth heat­ing surfaces. They were irregular or horseshoe-shaped, with an average diameter of around 1.1m. Five suc­cessive replastering phases of the heating surface was noted in hearth 16, which were also of a unique form, with low walls, an apron and an entrance opening. Clay boxes were common at Imvrou Pigadi, directly linked with the kilns both spatially and functionally (Fig. 10). They were small rectangular structures with an average length of 0.45m and width of 0.4m, and low walls with an average height of 0.2m. Their floors were laid directly on the ground surface and consisted of a single smooth clay layer that co­vered the lower surface of the structure. Their small size and the material found within them, including lithic artefacts, clay sling bullets, horns, pottery sherds and ash, suggest their use as storage or dis­posal features close to the kilns. Formed of two ver­tical low clay walls of 1m and 0.82m, a single clay basin (structure 14) was uncovered in the northeast­ern part of trench 1 (Fig. 11). It contained pure clay masses, along with two ground stones tools, and was roughly constructed with minimal investment of labour or building expertise. N. Kyparissi-Apostolika, E. Kalogiropoulou, D. Roussos, N. Saridaki, O. Metaxas, G. Kotzamani, K. Trantalidou, and Y. Facorellis The pottery evidence Trenches 1 and 3, the domestic character of the assemblage The ceramic material from trench 3 and the upper layers of trench 1 is dated to the Middle Neolithic, while some sherds from the upper layers and sur­faces of the trenches are Late Neolithic. The pottery included large quantities of red monochrome wares, along with substantial amounts of various styles of painted wares (Fig. 13). The most common shapes were open (40%), fol­lowed by closed vessels (30%), while a large propor­tion of sherds were without any diagnostic traits (30%). Most vessels were spherical, hemispherical and conical bowls, as well as basins and cups of small and medium size. The rims were usually straight, in­wardly curving or S-profile. The necks of closed ves­sels were cylindrical or concave. Most bases were ring-shaped, often retaining traces of coil construc­tion, with fewer flat and concave bases. Monochrome vessels were predominant and in par­ticular red slipped, followed by brown burnished or slipped, cloud burnished and then red burnished wa­res. The most common decorated ware was painted, followed by rare impresso and incised sherds. The well-known Thessalian red on white painted wares were also prevalent (Wace, Thompson 1912; Theo-charis 1973; Kotsakis 1983). Decorated motifs were mostly linear, lines and zig-zags, along with solid tri­angles and zones, all indicative of an advanced phase of the Middle Neoli­thic (Kotsakis 1983). White on red and scraped wares were less com­mon. Vessels were mainly used for display and consumption, for transport of li­quid or solid contents, and less fre­quently for storage. Preliminary asses­sment of the assemblage also noted the presence of cooking vessels. The sherds had traces of organic inclusi­ons, while a considerable proportion had clouds (16%), although it was not always clear whether these de­rived from the initial firing of the ves­sels or whether they were the result of later use during cooking (Fig. 12). The pottery from trench 3 and the upper layers of trench 1 highlighted the domestic character of the assemblage, as it in­cludes all of the types of vessels that a Neolithic hou­sehold would require for consumption, display, sto­rage, and cooking. The site is dated primarily to the Middle Neolithic, as characterised by the classic Thes­salian red on white wares. The characteristic Late Neolithic sherds from the upper and surface layers of the trenches included black and grey burnished, polychrome painted (light red-orange and black paint on a white background), and incised wares (see also Fig. 13). These ceramic categories are dated to an early phase of the Late Neolithic, specifically Late Neolithic I, although their continuation into Late Neolithic II cannot be excluded. Trenches 1 and 2, the pottery workshop Small and medium-sized vessels continued to be the predominant categories in trenches 1 and 2. Vessel shapes included open (28%) and closed (21%) forms, with the most common types being bowls, cups, jars, A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida and less frequently basins. The predominant deco­rative category was monochrome, red-slipped bur­nished, followed by brown burnished, brown slip­ped and red burnished wares, while the main manu­facturing technique was coiling. The surfaces were highly burnished to a polished finish, while a few were smoothed, with shapes and surface treatment pointing to a high degree of standardisation and re­petitive production. The completed vessels recovered from around the pottery kilns in trench 2 were all monochrome red slipped bowls (Fig. 14). The vessels shared similar morphological characteristics with differences only in size. A significant proportion of the sherds were recorded as uncertain, and could represent unfini­shed vessels, such as three closed jars with unfini­shed smoothed exterior surfaces and visible tool tra­ces. These jars preserve part of the manufacturing process, with unfinished surfaces prior to the appli­cation of the red slip. The detailed macroscopic analysis of the pottery from trenches 1 and 2 indicated that over 70% of the assemblage consisted of monochrome red slip­ped wares of all shapes and sizes, allowing us to conclude that this was the predominant output, again pointing towards the presence of specialised pottery workshops at Imvrou Pigadi (Fig. 15). Alongside the monochrome burnished wares, smal­ler quantities of decorated pottery were recovered from both trenches. These included painted wares, those with relief decoration, and rare impressed and incised motifs. The predominant painted style was the well-known Thessalian red on white (Wace, Thompson 1912; Theocharis 1973; Kotsakis 1983). Predominant decorative themes included linear li­nes, zigzags, combinations of bundles of rectilinear or wavy lines, along with solid inverted triangles and flames. Less common types included stepped Tzani, abacus and hatched motifs. Other rare painted types included white on a red background and scraped wares. Buttons were a distinctive relief decoration, along with very rare grains. Vessels with the former were decorated around the main body of the vessel. The monochrome red burnished slipped open bowls with relief button decoration shared similar morpholo­gical characteristics (open hemispherical bowls) and were differentiated only in terms of their size, with heights of 13cm to 22cm. These vessels may also have been produced at Imvrou Pigadi (Fig. 16). The vessels recovered from trenches 1 and 2 were main­ly for display and consumption, or the transport of liquid or solid contents. Conversely, storage and co­oking vessels were rare. The ceramic assemblage was characterised by a va­riety of painted decorative styles, with the dominant red on white wares and rare white on red. These painted vessels are common throughout Thessaly during the Middle Neolithic, pointing to their wide circulation (Wace, Thompson 1912; Kotsakis 1983; Gallis 1996; Rondiri 2009; Hamilakis et al. 2017). This was also confirmed by two similar open bowls with red on white decoration, one from Imvrou Pi-gadi and the second from Astritsa in Karditsa (Ky­parissi-Apostolika 2006.Fig. 5; 2019.Fig. 9; Hatziag­gelakis 2006.Fig. 4). In addition, the small numbers N. Kyparissi-Apostolika, E. Kalogiropoulou, D. Roussos, N. Saridaki, O. Metaxas, G. Kotzamani, K. Trantalidou, and Y. Facorellis of scraped ware sherds may suggest links with the area of Rizava, where a regional production centre for such pottery is reported (Krachtopoulou et al. 2018; 2020). The small number of buff burnished sherds found at the site may also suggest exchange with the Sesklo area (Schneideretal 1994.64; Di-moula 2014), with a programme of petrographic analysis currently being carried out on the material by Areti Pentedeka. The chipped stone assemblage The excavations at Imvrou Pigadi produced 653 chipped stone artifacts. Obsidian was the predomi­nant raw material in use (n=411, 62.9%), followed by flint (n=234, 35.8%) and quartz (n=8, 1.2%). The obsidian was exclusively from Melos, while the sili­cious material was derived from a range of sources. The most abundant variety was the so-called choco­late radiolarite. This was imported from the Pindos mountains (Kourtessi-Philippakis 2009) and local­ly knapped, although the introduction of partially worked tools and blanks cannot be ruled out. The provenance of the so-called honey flint is less clear, but it was certainly being brought in from distant sources.2 Tools made from this raw material were relatively rare, consisting only of retouched blades. Evidence for onsite knapping was present for both obsidian and flint (Fig. 17). Obsidian knapping ap­pears to have been less intense, with only five cores recovered, along with four crested blades and three core rejuvenation tablets. Even though six cortical flakes could indicate that obsidian nodules were occasionally reaching the site, it is more likely that they were derived from a core that retained some cortex after the initial preparation and forming of the crests. Four cores were used for the production of blades/bladelets, while the remaining one had flake scars only. The evidence suggests that obsidian knapping was mainly geared towards the production of blades/bladelets using the pressure technique. The evidence for the working of flint, in the form of 14 cores and 15 cortical flakes, suggests that although blades/bladelets were being produced, this material was more commonly used for the expe­dient production of flakes. This was also evident in the blade/bladelet versus flake ratio for each type of raw material. In the case of obsidian, blades/bladelets were over­whelmingly predominant (195 blades/bladelets ver­sus 80 flakes), while in the case of flint the picture was different (19 blades/bladelets versus 64 flakes). Note that if the blanks used for formal tools were counted, the difference becomes somewhat less pro­nounced due to the tendency towards applying re­touch on flint blades/bladelets. Preparation of core striking platform overhangs by trimming was occa­sionally evident on both flint and obsidian blades/ bladelets. Although flint knapping was to a large extent geared towards the production of flakes, it is noteworthy that obsidian rather than flint flakes were more fre­quently used for the manufacture of retouched tools. But flint and obsidian retouched blades/bladelets form the predominant formal tool category (n=50 and 37 respectively) (Tabs. 3–9). Retouched flakes and notches (Fig. 18), the latter always made on obsidian blanks, were less numerous, while side and end scrapers were rare. Other tool categories were represented by only a few examples, including trun­cated blades/bladelets, borers and burins. Although 2 Raw materials referred to as honey flint could be distinct. For possible provenance see Catherine Perles (1990), Georgia Kour­tessi-Philippakis (2009) and Odysseas Kakavakis (2014). A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida Retouched tool types Obsidian Flint Quartz Total phic evidence for a change in raw Retouched blades\bladelets 35 56 91 material utilisation over time at the Retouched flakes 16 10 26 site. This suggests the existence of a Splintered pieces 9 4 13 well-established regional trend in the Notches 14 14 far southwestern Thessalian Plain. Truncations Arrowheads Burins 5 1 3 2 3 1 7 4 4 The location of the site, close to rou­tes leading from both the Spercheios End scrapers 1 2 3 Valley and the coastal area of Volos Borers 11 2 towards the western Thessalian Side scrapers 1 1 Backed blades 1 1 Technical piece, core rejuvenation tablet 1 1 Retouched indeterminate 4 1 5 Total 86 85 1 172 Tab. 3. Retouched tool categories. Trenches 1 and 2 (Phases B1 and B2) Obsidian Flint Quartz Total Blades\bladelets 54 2 56 Flakes 27 10 37 Cores 13 4 Technical pieces 1 1 Plain, was certainly conducive to the formation of this regional pattern. While obsidian was preferable for the production of blades/bladelets because of its flaking properties, sturdier flint blades were also local­ly made and probably, or certainly in the case of honey flint, imported through networks connecting the site with the rest of Thessaly and areas beyond. Tools 18 16 34 Other (debris, indeterminate, Imvrou Pigadi appears to be in a 6 18 24 natural pieces) transitional position due to its en- Total (%) 107 (69) 49 (31) 156 gagement within spatially overlap­ping networks. It was simultaneous- Tab. 4. Trenches 1 and 2 (phases B1 and B2) basic assemblage ly part of the supply zone for the ob- structure. sidian trade, but also had unhinder-rare, projectile points (n=4) may be helpful in terms ed access to flint raw materials and finished tools, of chronological differences. A transverse arrowhead traded through networks prevalent in areas to the is probably Middle Neolithic in date, while the two west and north. As a result of its connection with-tanged points probably date to after the Middle to in a variety of networks, the site appears to have Late Neolithic transition. Flint retouched blades been able to exercise a flexible strategy of raw ma-were frequently used for cutting siliceous plants, as in 24 cases there was gloss along one or both edges. In addition to retouched tools, the assemblage also included splintered pieces and several unretouched pieces with macroscopic wear traces. The chipped stone assemblage from Imvrou Pigadi is of additional interest due to the location of the site in the vicinity of the largely contemporary Kou­troulou Magoula settlement. The latter has a raw material utilisation pattern that sets it apart from what is known from Thessalian sites further to the north, due to the high proportions of obsidian.3 The evidence from Imvrou Pigadi is therefore useful in reconstructing these distinct regional patterns.4 Imv­rou Pigadi suggests that the high proportions of ob­sidian seen at Koutroulou Magoula are not unique, since at Imvrou Pigadi it was even more prevalent. It should also be noted that there is no stratigra- 3 Yannis Hamilakis et al. (2017.92). 4 For an evaluation of lithic networks in mainland Greece during the Neolithic, see Katherine Perles (1990; 2009). Regarding the area of Thessaly, see Evagelia Karimali (2009). N. Kyparissi-Apostolika, E. Kalogiropoulou, D. Roussos, N. Saridaki, O. Metaxas, G. Kotzamani, K. Trantalidou, and Y. Facorellis terial and tool procurement. Locat- Trenches 1 and 2 (phases A1 and A2) Obsidian Flint Quartz Total ed on the edge of Thessaly, Imvrou Blades\bladelets 80 11 91 Pigadi or other neighbouring sites Flakes 33 36 69 may have served as important nodes Cores 1 4 5 in these flint and obsidian trading Technical pieces 2 2 networks, which may have a bearing on the social perception and integra- Tools 43 43 Other (debris, indeterminate, 25 22 3 natural pieces) 86 50 tion of lithic production and trade. 184 116 3 Total (%) 30 Other small finds (60.7) (38.3) (1) Small finds were not very common, Tab. 5. Trenches 1 and 2 (phases A1 and A2) basic asse mblage reinforcing the interpretation of this structure. part of the site as a workshop rather Trench a\2016 Obsidian Flint Quartz Total than a domestic area. Figurines were Blades\bladelets 10 1 11 less abundant compared to other ad- Flakes 8 4 12 jacent sites, just 65 from all years Cores 2 2 and all trenches. Amongst these were Technical pieces 1 1 20 small items with three- or four- Tools 5 5 sided bases, a type common at the Other (debris, indeterminate, natural pieces) neighbouring Koutroulou Magoula Total (%) 26 (83.9) 5 (16.1) 31 (Kyparissi-Apostolika 2003.Fig. 4; Tab. 6. Trench a/2016 basic assemblage structure. Hamilakis et al. 2017.Colour pla­ tes, Fig. 8) and other sites in western Trench 3 Obsidian Flint Quartz Total Thessaly, such as Platia Magoula Zar­ Blades\bladelets 48 6 54 kou (Alram-Stern in press) and Sy- Flakes 11 14 25 keon (Alexiou 2015). The remaining Cores 1 7 8 45 figurines included several types, some of which are naturalistic and Technical pieces 3 Tools 19 18 1 Other (debris, indeterminate, natural pieces) 9 16 4 3 38 29 in other cases they had ornaments Total (%) 91 (58) 61 (38.8) 5 (3.2) 157 or paint on the body (Fig. 19) (Ky­ parissi-Apostolika forthcoming). Tab. 7. Trench 3 basic assemblage structure. Twelve figurine heads were also re­ covered, some of which were relati- Surface finds Obsidian Flint Quartz Total vely large, up to 6cm (for example, ME 65/2009 and ME 9/2007), obvi­ Blades\bladelets 2 Flakes 1 Cores 2 1 ously from bigger figurines. Some Technical pieces larger examples of legs were also re- Tools 3 3 covered (for example, ME 64/2017 Other (debris, indeterminate, natural pieces) and ME 35/2007). Most of the figu- Total (%) 3 (50) 3 (50) 6 rines were found in trench 3, fol­ lowed by a/2016, with fewer in tren- Tab. 8. Surface finds basic assemblage structure. ches 1 and 2. Apart from the figurines, 20 orna- Unretouched debitage, cores and Obsidian Flint technical pieces Blade\bladelet cores 4 4 Total 8 ments were also found, mostly beads Flake cores 1 8 9 made of stone or bone, along with Blade\flake cores 2 2 Total 293 112 405 some from clay. Most were found in Crested blades 4 trench 3, with less in trenches 2 and a/2016. Another impressive catego­ry were the numerous sling bullets, with 73 in total, most of which were intact, as if just baked. They were found close to the kiln structures, also suggesting that they had been Core rejuvenation tablets 3 Cortical flakes 6 15 21 Blades\bladelets 195 19 214 Flakes 80 64 144 Tab. 9. Unretouched debitage, cores and technical pieces from the complete assemblage. A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida fired therein. The majority (n=55) were found in phase B1 of trench 1 and then in trench a/2016, while only rarely in trenches 2 and 3. The clay ob­jects also included a few spindle whorls and spin­ning and weaving implements, as well as a few cubes with incised dots (n=29). Ground stone tools and millstones were abundant, with at least 132 recovered. Additionally, there were 41 bone tools and worked horns in the assemblage of small finds. The bone tools came from all trenches and were di­rectly associated with pottery production, including perforators, spatulae and needles, although the ana­lysis of these objects is still in progress. They prob­ably reflect the specific activities carried out in the workshops, but if or when excavations are carried out in areas of a more domestic character, this pic­ture may change. mer wheat (Triticum dicoccum), barley (Hordeum vulgare) and possibly oats (cf Avena sp.). Wheat was mainly present in the form of seeds, while the occurrence of glume bases, either separate or as whole spikelets, was more limited. The assemblage also contained smaller quantities of various pulses, including lentils (Lensculinaris), bitter vetch (Vici­aervilia) and probably the common pea (cf Pisum sp.) and chickpea (cf Cicer sp.). Only the fig (Ficus carica) has been identified so far in the category of fruit and nuts. A much wider range of wild plants were identified in the assemblage, including genera such as Fuma­ria sp., Chenopodium sp., Galium/Asperula sp., Scirpus/Cyperus sp., Carex sp., Lolium sp., Bromus sp., and several more as yet unidentified species of Leguminosae, Polygonaceae and Gramnineae (Tab. 10). All of these could have been present amongst the natural vegetation surrounding the site, or may have been included in the assemblage as weeds of cultivation or as evidence for a much broader vari­ety of, as yet unknown uses within the context of daily life. The unique character of Imvrou Pigadi as a site where specialised pottery production was being un­dertaken poses an interesting challenge for archaeo-botany, in that it is associated with evidence for the use of plants in non-domestic contexts, in contrast to Latin name Comon name Triticum monococcum Einkorn The archaeobotanical evidence Archaeobotanical investigation was incorporated within the programme of research at Imvrou Pigadi from the very beginning of the excavations. Based on a systematic soil sampling strategy, the aim was to gather as much evidence as possible on the utili­ sation of plants and practices of cultivation, har­ vesting and consumption. The location of the site on the southwestern edge of the Thessalian plain, and its chronology within the Middle Neolithic, provide us with a unique opportunity to complement the fragmentary archaeobotanical evidence from this part of Greece and this phase of prehistory. The study of the plant remains is still in progress, and Fruit Wild flora Pulses Cereals and nuts Triticum dicoccum Emer Hordeum vulgare Barley cf Avena sp. Oat Vicia ervilia Bitter vetch Lens culinaris Lentil cf Pisum sativum Common pea cf Cicer sp. Chickpea Ficus carica Fig Polygonaceae Knotweed family Chenopodium sp. Goosefoot Fumaria sp. Fumitory\Fumewort Leguminosae Legume family Galium\Asperula sp. Bedstraw\Woodruff Lolium sp. Ryegrass Bromus sp. Brome grass only generic and preliminary observations are pre­ sented. Graminae Grass family Scirpus\Cyperus sp. Club-rush\Sedge Carex sp. True sedge Carbonisation was the only mode of preservation encountered in the assemblage. Cereals were repre- Tab. 10. Preliminary list of plant species/genera/ sented by einkorn (Triticum monococcum) and em-families identified at Imvrou Pigadi. N. Kyparissi-Apostolika, E. Kalogiropoulou, D. Roussos, N. Saridaki, O. Metaxas, G. Kotzamani, K. Trantalidou, and Y. Facorellis more mainstream archaeobotanical studies. The in­creased presence of seeds of the Gramineae family may relate to the preparation of the clay. Such seeds were probably attached to plant stems used as clay admixtures for crafting pottery or other clay objects. A comparative study of the plant imprints on the kilns and other structures may help to further clari­fy these questions. The analysis of the archaeobotanical assemblage is ongoing and includes the identification of the plant species present in the assemblage, and an under­standing of their taphonomic history and role in nu­trition, as tools or as construction materials. Compa­rison of the range of plant remains within the four trenches may help to further clarify issues related to the use of space, as well as providing a more ho­listic view of the contribution of plants in the econo­mic and socio-cultural life of the site. Comparison of the Imvrou Pigadi archaeobotanical evidence with that from the nearby contemporary settlement of Koutroulou Magoula will significantly broaden our current understanding of the vegetational environ­ment, along with its role within the Neolithic of Thessaly and Phthiotida. Taphonomy of the faunal remains Trenches 1, 2 and 2016/a produced a total of 6284 faunal remains, including six human bones.5 Al­though the four leporid, 15 avian and 109 cervid (red deer) bones may have derived from wild spe­cies that were hunted (one amphibian and eight tor­toise bones, possibly from intrusive vertebrates were also recovered), the bulk of the assemblage from all three phases was dominated by domestic animals (Fig. 20). According to the NISP counts,6 domestic species make up 98% of the assemblage, including sheep/goat, pig, cattle and dog, while a donkey first phalanx is a later intrusion. The species composition points to an agrarian way of life during the Neolithic in western Thessaly, and a pastoral economy based on sheep and goat herd­ing, which was common in the region. The 1338 ani­mal bones recovered from trench 3 were assigned to the same species. Sheep and goat bones accounted for 81% of the com­plete assemblage. Although cluster analysis based on the limited morphological features does not clearly separate sheep and goats,7 it appears that sheep were more common, with twice as many fragments as those of goats. Table 11 provides an overview of where the bones were found within the trenches, and how their dis­tribution is related to other features of the manu­facturing complex. Bones recovered from features such as kilns, hearths, clay boxes and pots, that were attributed to cultural phase B, accounted for 50.15% (n=3152) of the faunal assemblage examined so far. From the sediments covering the kiln features and the deposits underlying the fire installations and at­tributed to cultural phase C, a total of 2838 bones and 294 fragments were recovered. In neither of these assemblages were any of the bones articulated. Evidence regarding the state of preservation (degree of fragmentation, erosion, gnawing, burning) has been taken into account at both the contextual and individual levels, and has been assessed for all do­mestic species and red deer. It should be noted that the stages of burning visible on bones were classified by colour. Three categories were established to assess the impact of fire and the damage caused.8 These 5 The recording per species, anatomical unit, fragment size, degree of preservation, age and sex (as well as the biometry in an elec­tronic database) was carried out by Dimitris Filioglou and Kostas Nikolaou, supervised by Katerina Trantalidou and supported by the Institute for Aegean Prehistory. 6 Methods of quantification for the Imvrou Pigadi assemblage included Number of Identified Specimens (NISP), whole bone equiva­lents, Minimum Number of Individuals (MNI) and Minimum Number of Elements (MNE). But for the purpose of this paper, main­ly to do with the taphonomy of the samples from the clay structures, only NISP is considered. 7 The distinction of sheep and goat was established based on the criteria of Joachim Boessneck (1970), Wietske Prumell and Hans-Jörg Frisch (1986), Melinda A. Zeder and Heather A. Lapham (2010) and Melinda A. Zeder and Suzanne E. Pilaar (2010). 8 The heating stages in bone have been described by Pat Shipman et al. (1984), Gilles Grévin et al. (1990), Paloma Vidal-Matura-no et al. (2019), all with previous bibliography. A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida Tab. 11. Spatial distribution of 6284 bone fragments recovered from the three trenches (By Filioglou and Nikolaou using NISP). included moderate alteration (brown or intense orange), carbonisation at temperatures between 200 and 400°C (black) and calcination at temperatures greater than 600 degrees (grey or white). For exam­ple, red deer antlers tended to be highly altered in the area of the kilns and clay boxes, whereas light alteration was more common amongst suid bones than other species. The following are some general comments regarding the assemblage: . The skeletal part representation within the bone assemblage was almost complete, even though phalanges were rare (Figs. 21 and 22). When the animals were slaughtered, the phalanges may have remained attached to the hide and discard­ed later during its processing; . There is little evidence for any significant varia­tion in the pattern of body part representation between phases for any of the domestic species; . There is spatial variation in the preservation of the assemblages across the site, with bones in trenches 1 and 2 relatively more affected by fire than those in trench 3. In contrast, the bones in trench 3 were more coated by sediment than those found in the firing areas; . Unburned bones were recovered from the layers above the clay structures, but also from the lay­ers incorporating the structures; . High degrees of fragmentation and pieces of less than 3cm were common. Approximately 15% of the Caprinae sub-family could not be identified to the level of major limb bone, humerus, radius, fe­mur, tibia or metapodials; . There was a wide range of variability in colour, related not only to burning time but also the hi­stological structure of each bone; . Hyperfragmentation (<1–2cm) and total thermo-alteration, as seen in heavily burnt bones from altars and special deposit pits of Classical and Hellenistic antiquity (Trantalidou 2013.Fig. 10; in press), or where the bone itself was burnt (Vi-dal-Maturano et al. 2019.Fig. 11), has not been full attested at Imvrou Pigadi; . A working hypotheses is that much of the bone was used as fuel. Those fragments that do not ex­hibit visible traces of burning were either due to be used in the kilns or were in the sediments that intentionally covered the structures when they were no longer in use. More detailed spatial ana­lysis using GIS is required; . Bone burn time in those features was either limi­ted or the level of technology in the Middle Neo­lithic did not permit them to reach much higher temperatures. Therefore, the destruction of all bones and the traces of combustion are not visi­ble. Woody taxa should also be examined. N. Kyparissi-Apostolika, E. Kalogiropoulou, D. Roussos, N. Saridaki, O. Metaxas, G. Kotzamani, K. Trantalidou, and Y. Facorellis Discussion and conclusions Excavations since the 1950s in western and east­ern Thessaly, as well as elsewhere, never produced any evidence for the existence of early organised kiln workshops, with the firing of vessel in open fires widely accepted until recently (Rondiri, Ka­logianni 2018). Several experimental programmes were initiated in order to investigate how pottery was fired (Vitelli 1994; Rondiri, Kalogianni 2018). Based on the observation of sherds, Karen D. Vitel­li (1997) proposed several hypothetical designs for kilns that could have been used during the 6th mil­lennium BC, which are quite close to those identi­fied at Imvrou Pigadi. Her suggestion that such in­stallations should probably be located far from the settlement is also reasonable. The excavations at Imvrou Pigadi and the structures themselves have convinced us regarding the function of the site as a kiln workshop. When some years later the site of Rizava was discovered a few kilometres to the west on the plain of Karditsa (Krachtopoulou et al. 2018), as well as more recently the neighbouring site of Koutroulou (unpublished, Kyparissi et al. 2022 forthcoming), our suggestion regarding the use of pottery kilns since the beginning of the Middle Neo­lithic in Thessaly was confirmed. The complex of ce­ramic kilns at Imvrou Pigadi appears to have been in operation almost from the beginning of the use of the western edge of the site, although the con­struction of the provincial road has obscured its ori­ginal boundaries. When the kilns in trench 2 ceased operation (phase B2), the area appears to have been buried beneath a layer of burnt deposits, upon which, on the basis of the finds, activities continued although with a possible hiatus. The same appears to be the case in trench 1, where the deposits above the kiln structures were again covered with a layer of clay, before being reused for domestic activities rather than as a workshop. The workshop areas, al­beit at different elevations, appear to have been used in parallel, as their dates are similar. Since they were not spatially overlapping, they could have been ope­rated simultaneously, possibly for the firing of dif­ferent objects or controlled by different groups of craftsmen. Coring as well as the excavations in trenches 1 and 2 have shown that the workshop area would have extended even further towards the north and south of trench 2. This was also indicated by the geophy­sical survey, especially in the southwestern area of site and towards the west, now under the public road. Such an extensive workshop in operation at one of the largest magoula sites in the region sug­gests that it probably catered for the pottery needs of other settlements as well. This is supported by the lack of other similar workshop complexes at sites nearby, although others would be expected at great­er distances. It is also possible that different types of vessels were being fired in the various settle­ments, which were then redistributed amongst them. Such a scenario may be implied by the newly disco­vered complex at the site of Rizava, located around 30km to the west on the plain of Karditsa. But we also need to be mindful of the recent discovery of a smaller complex of kilns at nearby Koutroulou Ma-goula, just 3km away, which could potentially wea­ken the suggested model of longer distance ex­change. The situation will hopefully become clearer when the excavations at Koutroulou are completed and we can compare the kiln complexes from the two sites, both chronologically and in terms of the finds. But it is also possible that due to its size and importance, Koutroulou may have been self-suffi­cient for at least some of its pottery needs with its own kilns. The abundance of obsidian at Imvrou Pigadi is ano­ther striking aspect of the site, where the material accounted for 63% of all chipped stone recovered, contradicting the established model for western Thessaly, where chocolate flint from the Pindos is A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida predominant. However, given that similar propor­tions of obsidian have also been reported from near­by Koutroulou Magoula, it would appear that both sites were located along the route between eastern and western Thessaly on the one hand, but also to­wards Phthiotida to the south. On the other hand (Perles 2022 in press), it would appear that the ma­terials that arrived here, such as obsidian, were not widespread in western Thessaly. According to Kari-mali (2022.195), these two positions are characte­rised as atypical. The pottery from Imvrou Pigadi presents the same decorative elements as the rest of Thessaly, indicat­ing a wide network of communication and exchange. The presence of figurines, jewellery, stone grinding tools and millstones, as well as animal bones and fruit, is probably related to the wider use of the area around the magoula, which likely constituted a reg­ular settlement, something that may be established in the future if the excavations are extended. The faunal remains were dominated by sheep and goat bones, and reflect an agrarian way of life in Neolithic western Thessaly and a pastoral economy which is common in the geographi­cal region. The picture in trench 3 was similar, while the bones in tren­ches 1 and 2 were relatively more affected by fire, probably related to differences in the use of space, lead­ing to the working hypothesis that much of the bone recovered was used as fuel. The presence of cereals and pulses in the archaeobotanical evidence points towards cultivation, while the broad range of wild plants may re­late to the preparation of clay. The increased presence of seeds belong­ing to the Gramineae family were probably attached to plant stems used as clay admixtures for crafting pottery, other objects and the clay ovens themselves. Because of its size, but mainly the systematic use of kilns, it would ap­pear that the pottery workshops at Imvrou Pigadi operated as a produ­ction hub during the Middle Neoli­thic on the route between eastern and western Thessaly on the one hand, and Phtiotida via the Sperchios Valley and southern Greece on the other. It appears to have played a decisive role in the circulation of trade goods and the exchange of ceramic vessels and other aesthetic objects, which because of its position were promoted throughout western Thessaly, but also in the eastern part of the region and in settle­ments further south. AUTHORS’ CONTRIBUTON Nina Kyparissi-Apostolika: Introduction, The site and the excavations, Chronology, Other small finds, Dis­cussion and conclusions; Yorgos Facorellis, Chrono­logy; Dimitris Roussos: Coring; Evita Kalogiropoulou and Dimitris Roussos: The clay-based structures, their typology and pyrotechnological characteristics; Niki Saridaki: The pottery evidence; Odysseas Metaxas: The chipped stone assemblage; Georgia Kotzamani: The archaeobotanical evidence; Katerina Trantalidou: Taphonomy of the faunal remains. N. Kyparissi-Apostolika, E. Kalogiropoulou, D. Roussos, N. Saridaki, O. Metaxas, G. Kotzamani, K. Trantalidou, and Y. Facorellis ACKNOWLEDGEMENTS The excavations at Imvrou Pigadi were funded by the Greek Ministry of Culture, INSTAP and the Psycha Foun­dation. The Ephorate of Antiquities of Phthiotida and Evrytania facilitated our works in several ways all the years. We are indebted to them all. Figure (2) was produced by Dr. Jamie Donati, Post-doctoral Fellow on the project “Beyond Oikos: Outdoor spaces, daily life and sociality in Neolithic Greece” (BONDS), funded by the Hel­lenic Foundation for Research and Innovation (H.F.R.I.) under the 2nd call for H.F.R.I. research projects to support post-doctoral researchers (Project number: 00229). Dr. Gilbert Marshall undertook the editing of the texts, and we are grateful to him for his consistency and helpful advice. . References Alexiou N. 2015. Neolithika eidolia apo tin dytiki Thes­salia. Ta eidolia tis thesis Sykeon. In A. Mazarakis Ainian (ed.), Praktika epistimonikis synantisis, Volos 26.2.– 1.3.2015. Archaiologiko Ergo Thessalias kai Stereas Ella-das 5. Vol. I. Thessalia: 141–154. (in Greek) Alram-Stern E. 2022. Figurines, house model and ritual ves­sels. In E. Alram-Stern, K. Gallis, and G. Toufexis (eds.), Pla­tia Magoula Zarkou. Volume 1. The Neolithic period. En­vironment, stratigraphy and architecture, chronology, tools, figurines and ornament. Oriental and European Ar­chaeology. Österreichische Akademie der Wissenschaften. Wien. Boessneck J. 1970. Osteological differences between sheep (Ovis aries Linné) and goats (Capra hircus Linné). In D. Brothwell, E. Higgs (eds.), Science in archaeology. A sur­vey of progress and research. Praeger. New York: 331–358. Bronk Ramsey C. 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51(1): 337–360. https://doi.org/10.1017/S0033822200033865 Chatziaggelakis L. 2006. To erevnitiko ergo tis LD Efore­ias Prohistorikon kai Klassikon Archaiotiton. In A. Maza­rakis Ainian (ed.), Praktika epistimonikis synantisis, Vo­los 2009. Archaiologiko Ergo Thessalias kai Stereas Ella-das 2, Vol. I. Thessalia: 771–793. (in Greek) Conati Barbaro C., Forte V., Muntoni I. M., and Eramo G. 2021. A multidisciplinary approach to the study of Early Neolithic pyrotechnological structures. The case study of Portonovo (Marche, Italy). Open Archaeology 7: 1160– 1175. https://doi.org/10.1515/opar-2020-0198 Dimaki S. 1994. Prohistorikoi oikismoi sti voreia Phthio­tida. In La Thessalie, Quinze années de recherches ar­chéologiques, 1975–1990, Bilans et Perspectives. Actes du Colloque International, Lyon, 17–22 Avril 1990. Mini-stere Grec de la Culture. Editions Kapon. Athenes: 91–102. Dimoula A. 2017. Early pottery mobility: The case of early Neolithic Thessaly, Greece. Journal of Archaeological Science. Reports 12: 209–218. http://dx.doi.org/10.1016/j.jasrep.2017.01.008 Donahue D. J., Jull A. J. T., and Toolin L. J. 1990a. Radio­carbon measurements at the University of Arizona AMS facility. Nuclear Instruments and Methods in Physics Re­search Section B: Beam Interactions with Materials and Atoms 52(3–4): 224–228. https://doi.org/10.1016/0168-583X(90)90410-V Donahue D. J., Linick T. W., and Jull A. J. T. 1990b. Isoto­pe-ratio and background corrections for accelerator mass spectrometry radiocarbon measurements. Radiocarbon 32(2): 135–142. https://doi.org/10.1017/S0033822200040121 Facorellis Y., Maniatis Y., and Kromer B. 1997. Study of the parameters affecting the correlation of background versus cosmic radiation in CO2 counters: Reliability of dat­ing results. Radiocarbon 39(3): 225–238. https://doi.org/10.1017/S0033822200053224 Gallis K. 1996. Kentriki kai dytiki Thessalia. In G. Papa-thanassopoulos (ed.), Neolithic Culture in Greece. Nicho­las P. Goulandris Foundation. Museum of Cycladic Art. Athens: 61–66. (in Greek) Germain-Vallée C., Prévost-Dermarkar S., and Lespez L. 2011. Stratégies de prélevement et de mise en oeuvre de la »terre a bâtir« des structures de combustion Néolithi­ques du site de Dikili Tash (Grece) a partir d’une étude micromorphologique. ArchéoSciences 35: 41–63. Grévin G., Baud C. A., and Susini A. 1990. Etude anthropo­logique et paléopathologique d’un adulte inhumé puis in-cinéré provenant du site de Pincevent (Seine-et-Marne). Bulletin et Mémoires de la Société d’Anthropologie de Paris 2: 77–88. Hamilakis Y., Kyparissi-Apostolika N. 2012. Koutroulou Magoula in Central Greece: From the Neolithic to the pre­sent. Antiquity 86 (333). Project Gallery. http://antiquity.ac.uk/projgall/hamilakis333/ A Middle Neolithic pottery workshop at Magoula Imvrou Pigadi, at the crossroads of eastern-western Thessaly and Phtiotida Hamilakis Y., Kyparissi-Apostolika N., Loughlin Th., +7 au­thors, and Zorzin N. 2018. Koutroulou Magoula in Phthio­tida, Central Greece: A Middle Neolithic Tell Site in Con­text. In A. Sarris, E. Kalogiropoulou, T. Kalayci, and L. Ka­rimali (eds.), Communities, Landscapes, and Interaction in Neolithic Greece. International Monographs in Prehi­story. Archaeological Series 20. Ann Arbor. Michigan: 81– 96. Kakavakis O. 2014. In search of social networks during the Neolithic period in Macedonia, Northern Greece. The case of chipped stone industries. In E. Stefani, N. Merous-sis, and A. Dimoula (eds.), Ekato chronia erevnas stin prohistoriki Macedonia, 1912–2012. Proceedings of the international conference. Archaeological Museum of Thessaloniki, 22–24 November 2012. Thessaloniki: 599– 606. (in Greek) Kalogiropoulou E. 2013. Cooking, space and the forma­tion of social identities in Neolithic Northern Greece: Evidence of thermal structure assemblages from Avgi and Dispilio in Kastoria. Unpublished PhD thesis. School of History, Archaeology and Religion. University of Car­diff. Cardiff. Kalogiropoulou E., Roussos D., Saridaki N., and Kyparissi-Apostolika N. in press. A Middle Neolithic pottery work­shop at the site of Imvrou Pigadi, Thessaly, Greece: Crafts­manship, operation, and the configuration of social space. Journal of Field Archaeology. Karimali, L. 2009. Katanomi lithinon proton ylon sti Neo­lithiki Thessalia: Mia sygritiki exetassi. In A. Mazarakis Ainian (ed.), Archaiologiko Ergo Thessalias kai Stereas Elladas 2, Volos 16.3–19.3.2006. Vol. I. Thessalia: 17–29. (in Greek) Karimali L., Papadopoulou S. 2022. Viewing Neolithic ob­sidian networks from the western side of the Aegean Sea (Greece): Distribution parameters and data reconsidered. In A. Baysal (ed.), The Lithic Studies from Anatolia and Beyond. Oxford: Archaeopress. Oxford: 164–212. Kotsakis K. 1983. Keramiki technologia kai keramiki diaforopoiissi. Unpublished PhD thesis. Aristotle Univer­sity of Thessaloniki. Thessaloniki. (in Greek) Kourtessi-Philippakis G. 2009. Lithics in the Neolithic of Northern Greece: Territorial perspectives from an off ob­sidian area. Documenta Praehistorica 36: 305–312. https://doi.org/10.4312/dp.36.19 Krachtopoulou A., Dimoula A., Livarda A., and Saridaki N. 2018. The discovery of the earliest specialised Middle Neolithic pottery workshop in western Thessaly, central Greece. Antiquity 92 (362): Project Gallery. https://doi.org/10.15184/aqy.2018.54 Krachtopoulou A., Frederick C., Orengo H. A., +4 authors, and Garcia-Molsosa A. 2020. Re-discovering the Neolithic landscapes of western Thessaly, central Greece. In A. Blan­co-González, T. L. Kienlin (eds.), Current approaches to tells in the prehistoric old world. Oxbow Books. Oxford, Philadelphia: 25–40. Kyparissi-Apostolika N. 2006. Magoula Imvrou Pigadi sto Neo Monastiri Phtiotidas: Akomi mia egatastassi tis Mes-sis Neolithikis sta voreia dioikitika oria tou nomou Phthio­tidas. In A. Mazarakis Ainian (ed.), Archaiologiko Ergo Thessalias kai Stereas Elladas 2, Volos 16.3–19.3.2006, Vol. II. Phtiotida: 839–851. (in Greek) 2012. Indications of the presence of Middle Neolithic pottery kilns at Magoula Imvrou Pigadi, SW Thessaly, Greece. Documenta Praehistorica 39: 433–442. https://doi.org/10.4312/dp.39.31 2019. Xana-diavazontas to Neolithiko topio tis dytikis Thessalias: sheseis, allilepidraseis, omoiotites kai diafo-res. In K. Kotsakis (ed.), I antipera ochthi. Koinonikos choros kai ideologia stis prohistorikes koinotites. Ari­stotle University of Thessaloniki. Thessaloniki: 71–89. (in Greek) forthcoming 2023. Adornment of human population in the Middle Neolithic Magoula Imvrou Pigadi: what the figurines could tell us? In S. Nanoglou, F. Mavridis (eds.), Figurine-making in the Neolithic Aegean. Kyparissi-Apostolika N., Hamilakis Y., Tsamis V., Loughlin T., and Zorzin N. 2022. forthcoming. Koutroulou Magou-la: mia anthoussa egatastasi tis Messis Neolithikis se ena kombiko tristrato (2018–2021). In A. Mazarakis Ainian (ed.), Archaiologiko Ergo Thessalias kai Stereas Elladas 7, Volos 3–5.6.2022. Thessalia. (in Greek) Mallol C., Mentzer S. M., and Miller C. E. 2017. Combustion Features. In C. Nicosia, G. Stoops (eds.), Archaeological Soil and Sediment Micromorphology. Wiley. Hoboken: 299–326. Mentzer S. 2014. Micro archaeological approaches to the identification and interpretation of combustion features in prehistoric archaeological sites. Journal of Archaeolo­gical Method and Theory 21(3): 616–668. https://doi.org/10.1007/s10816-012-9163-2 Orengo H. A., Krachtopoulou A., Garcia-Molsosa A., Palaio­choritis V., and Stamati A. 2015. Photogrammetric re-dis­covery of the hidden long-term landscapes of Western Thessaly, central Greece. Journal of Archaeological Sci­ence 64: 100–109. https://doi.org/10.1016/j.jas.2015.10.008 Perles C. 1990. L’Outillage de Pierre Taillée Néolithique en Grece: Approvisionnement et Exploitation des Matie­ N. Kyparissi-Apostolika, E. Kalogiropoulou, D. Roussos, N. Saridaki, O. Metaxas, G. Kotzamani, K. Trantalidou, and Y. Facorellis res Premieres. Bulletin de Correspondence Hellénique 114(1): 1–42. 2009. Les industries lithiques Néolithiques: Logiques techniques et logiquessociales. In De Méditerranée et d’ailleurs... Mélanges offerts a Jean Guilaine. Archi­ves d’Écologie Préhistorique. Toulouse: 557–571. Perles C., Papagiannaki L. 2022. The Flaked stone assem­blages. In E. Alram-Stern, K. Gallis, and G. Toufexis (eds.), Platia Magoula Zarkou. The Neolithic Period. Environ­ment, Stratigraphy and Architecture, Chronology, Tools, Figurines and Ornaments. Oriental and European Archa­eology. Österreichische Akademie der Wissenschaften. Wien: 197–309. https://austriaca.at/?arp=0x003dbb33 Prévost-Dermarkar S. 2002. Les foyers et les fours domes-tiques en Égée au Neolithique et a l’Âge du Bronze. Civi­lisations 49: 223–237. Prummel W., Frisch H. J. 1986. A guide for the distinc­tion of species, sex and body side in bones of sheep and goat. Journal of Archaeological Science 13: 567–577. https://doi.org/10.1016/0305-4403(86)90041-5 Reimer P., Austin W., Bard E., +38 authors, and Talamo S. 2020. The IntCal20 Northern Hemisphere Radiocarbon Age Calibration Curve (0–55 cal kBP). Radiocarbon 62 (4): 725–757. https://doi.org/10.1017/RDC.2020.41 Rondiri V. 2009. Thessaliki Neolithiki Keramiki: Tech-nologia kai katanomi ston choro. Unpublished PhD thesis. Aristotle University of Thessaloniki. Thessaloniki. (in Greek) Rondiri V., Kalogianni A. 2018. Pro-histories fotias. In A. Mazarakis Ainian (ed.), Archaiologiko Ergo Thessalias kai Stereas Elladas 6. Vol I. Thessalia: 679–688. (in Greek) Roussos D. N., Kyparissi-Apostolika N. 2019. An integrat­ed geoarchaeological methodology to reveal site forma­tion processes linked to building practices and pyrotech­nology at Middle Neolithic Imvrou Pigadi, Greece. Geo-archaeology 34: 200–220. https://doi.org/10.1002/gea.21711 Schneider G., Knoll H., Gallis K., and Démoule J.-P. 1994. Production and circulation of Neolithic Thessalian pottery: Chemical and mineralogical analyses. In La Thessalie, Quinze années de recherché archéologiques, 1975– 1990, Bilans et Perspoectives, Actes du Colloque Inter­national, Lyon, 17–22 Avril 1990. Ministere de la Cul­ture. Athenes: 61–70. Shipman P., Fosterb G., and Schoeninger M. 1984. Burnt bones and teeth: An experimental study of color, morpho­logy, crystal structure and shrinkage. Journal of Archa­eological Science 11: 301–325. https://doi.org/10.1016/0305-4403(84)90013-X Theocharis D. R. 1981. Neolithikos Politismos: syntomi episkopissi tis Neolithikis ston elladiko choro. Morpho­tiko Idryma Ethnikis Trapezis. Athens. (in Greek) Trantalidou K. 2013. Dans l’ombre du rite: vestiges d’an­imaux et pratiques sacrificielles en Grece antique. Note sur la diversité des contexts et les difficultés de recher­che recontrées. In J. Ekroth, J. Wallensten (eds.), Bones, behaviour and belief. The zooarchaeological evidence as a source for Greek ritual practice. Skrifter utgivna av Svenska institutet i Athen. 4o;55. Svenska Institutet i Athen. Athens: 55–80. in press. Social traditions and customary things at the temple of Apollo on the west coast of the Pagasetic Gulf: The archaeozoological evidence. In Sanctuaries and Cults in the Aegean from the Early Historic Times to the End of the Late Antiquity. International Con­gress, Lemnos, 11–15.09.2019. Tsokas G. N., Vargemezis G., Stambolidis A., Kyparissi-Apostolika N., Tassis G., and Tokmakidis K. 2009. Geophy­siki diaskopisi sti Magoula Imvrou Pigadi sto Neo Mona-stiri (Nomarchia Phthiotidas). Archaiologiko Ergo Thes­salias kai Stereas Elladas 3. Volos 12–15.3.2009. Volos: 1261–1267. (in Greek) Vidal-Maturano P., Alberto-Barroso V., Marrero E., García J. C., Pou S., and Arnay de la Rosa M. 2019. Vitrified wood charcoal and burnt bones from the pre-Hispanic site of Chasogo (Tenerife, Canary Islands, Spain). Journal of Archaeological Science 28: 102005. https://doi.org/10.1016/j.jasrep.2019.102005 Vitelli K. D. 1994. Experimental approaches to Thessalian Neolithic ceramics: Gray ware and ceramic colour. In La Thessalie, Quinze ans de rechérches archeologiques 1975–1990. Greek Ministry of Culture. Athens: 143–148. 1997. Inferring firing procedures from sherds: Early Greek Neolithic kilns. In P. Rice, D. Kingery (eds.), Pre­history and history of ceramic kilns. Ceramics and Civilization. Volume VII. American Ceramic Society. Westerville: 21–40. Wace A. J. B., Thompson M. S. 1912. Prehistoric Thes­saly. Cambridge University Press. Cambridge. Zeder M. A., Lapham H. A. 2010. Assessing the reliabili­ty of criteria used to post-cranial bones in sheep, Ovis, and goats, Capra. Journal of Archaeological Science 37: 2887–2905. https://doi.org/10.1016/j.jas.2010.06.032 Zeder M. A., Pilaar S. E. 2010. Assessing the reliability of criteria used to identify mandibles and mandibular teeth in sheep, Ovis, and goats, Capra. Journal of Archaeologi­cal Science 37(2): 225–242. https://doi.org/10.1016/j.jas.2009.10.002 back to contents Documenta Praehistorica XLIX (2022) Pottery firing in the Early Iron Age in western Slovenia Manca Vinazza1, Matej Dolenec2 manca.vinazza@ff.uni-lj.si< matej.dolenec@ntf.uni-lj.si 1 University of Ljubljana, Faculty of Arts, Department of Archaeology, SI 2 University of Ljubljana, Faculty of Natural Sciences and Engineering, Department of Geology, SI ABSTRACT – The article discusses the possible use of kilns for the firing of pottery in western Slove­nia during the Early Iron Age. In the absence of archaeologically attested kilns, their use in this area is studied based on indirect factors, i.e. the analysis of the vessel firing technique, and with the help of experiments from the field of experimental archaeology. The article strives to determine the rea­sons for the poor state of preservation of the kilns in the area in question. Samples from archaeo­logical experiments and archaeological pottery were subjected to AMS measurements, petrographic and mineralogical analyses (X-ray diffraction), which revealed the importance of considering the soaking time as a criterion for observing the firing processes and use of single-chamber kilns for the firing of pottery, even if they have not yet been discovered. KEY WORDS – pottery; firing process; experimental archaeology; pottery kiln; ceramic technology; apparent magnetic susceptibility AMS; X-ray diffraction XRD; Early Iron Age; Kras; Slovenia ?ganje lon;enine v starej[i /elezni dobi v zahodni Sloveniji IZVLE.EK – V .lanku obravnavamo potencialno rabo pe.i za .ganje keramike v zahodni Sloveniji v starej.i .elezni dobi. Ker arheolo.ko .e niso bile izkopane, je njihova uporaba predvidena na pod-lagi posrednih kazalnikov, kot so na.in .ganja in poskusi s podro.ja arheologije poskusov. Namen raziskave je prepoznati razloge za slabo ohranjenost pe.i na tem obmo.ju. Na vzorcih, ki smo jih pridobili iz arheolo.kih poskusov in na arheolo.ki keramiki, smo merili navidezno magnetno suscep­tibilnost, izvedli petrografske in mineralo.ke (rentgenska difrakcija) analize. Rezultati teh analiz so pokazali na pomembnost upo.tevanja .asa .ganja kot merila pri opazovanju .galnega procesa in uporabi enoprostornih pe.i za .ganje keramike, .eprav .e niso bile odkrite. KLJU.NE BESEDE – keramika; proces .ganja; arheologija poskusov; pe. za .ganje keramike; tehno­logija; navidezna magnetna susceptibilnost; rentgenska difrakcija XRD, starej.a .elezna doba; Kras; Slovenija Introduction How was pottery fired in the Early Iron Age in what pear no sooner than in the Late Iron Age. We assume is today western Slovenia? Can we assume the use that the preservation of such structures in the dis-of kilns despite their absence? On the territory of cussed areas depends on various factors, so we will present-day Italy, two-part updraft kilns with a sepa-show the possible reasons for the poor preservation rate fireplace and firing chamber were known in this of these based on the results of experimental archa-period, while on the territory of Slovenia they ap-eology. Through macro- and microscopic analyses of DOI> 10.4312\dp.49.20 Pottery firing in the Early Iron Age in western Slovenia the pottery and samples from archaeological experi­ments, we will try to reveal the features of the pot­tery associated with the firing process, which will in­directly help us identify structures for the firing of pottery. Firing is one of the most important steps in pottery production, as it involves the transformation of clay into ceramics. Due to the complexity of the produc­tion process of pottery vessels, from the raw mate­rial to the final product, the concept of the ceramic chaîne opératoire has recently developed in cera­mic analysis (Lemonnier 1993; Roux 2016.104– 107). Based on such observations, we try to identi­fy technological traditions and patterns of certain technical traits (cf. Roux 2016.104, 112). By includ­ing the ceramic chaîne opératoire approach, cera­mic experimental archaeology has gained a more solid methodology (Jeffra 2015. 141), but only if when the principle of so-called controlled compar­ison is considered (Roux 2016. 7). Until the advent of experimental archaeology and scientific analyses, the process of pottery firing was actually the least known technological process in the ceramic chaîne opératoire (Rado 1988.92). We will attempt to an­swer the question of what type of structures were used for firing pottery in western Slovenia in the Early Iron Age by integrating data from the macro­scopic and microscopic analyses of the pottery mass and firing technology. The data will be acquired from experimental archaeology, measurements of appar­ent magnetic susceptibility (AMS), and with the re­sults of mineralogical analyses (X-ray diffraction). The firing process In the past, people had to rely on personal experi­ence with firing, which in practice probably meant conducting numerous successful and unsuccessful experiments, as evidenced by the considerable over-fired vessel waste at archaeological sites (Cuomo di Caprio 1985.130). Today, laboratory and archaeolo­gical experiments are carried out, adding signifi­cantly to the knowledge and, above all, to the un­derstanding of technological processes, which are usually different under controlled laboratory condi­tions (cf. Thér 2014.96). After clay transforms into ceramics during the firing process (Cuomo di Caprio 1985.125), a series of chemical and physical reactions occur affecting the hardness, permeability, porosity, and mineral com­position of the final product. The products become impermeable, change colour, and lustre (Heimann 1978–1979.79, 82), and become hard and resistant to decay (Rice 2005.55, 80). The colour change is re­lated to the presence or absence of iron minerals (chlorites, micas, Fe-oxides/hydroxides, sulphides) in the clay. In the are discussed in this study the clays are very rich in iron, which usually causes the vessels to turn red (oxidation atmosphere) or grey and black (reduction atmosphere) (cf. Maritan 2018. 206). The colour can also be affected by the pres­ence of organic material, which converts to carbon and oxidizes into CO2 in the presence of sufficient oxygen. This change occurs at a temperature of about 800°C and can be recognized by a change in colour (grey/dark grey when carbon is present and cream/ reddish when carbon is oxidized) (Gliozzo 2020.26). During firing, structural changes also occur in the minerals in the clay (Cuomo di Caprio 1985.130; Levi 2010.112). The thermal stability of the mineral phase and the changes induced by heat depend on numerous factors, such as grain size, the mineralogi­cal and chemical composition of inclusions and tem­per, presence of aplastic inclusions, presence of or­ganic material, position in the vessel, the position of the vessel in the kiln, the soaking time, and cool­ing (Gliozzo 2020.5). What changes occur during the firing process and which of them are relevant for the pottery of the dis­cussed area? Hydration begins at different tempera­tures, depending on the heating and the type of clay minerals (e.g., illite 300–600°C), but the mixing of clays can lower it (Rice 2005.87–88). Carbon oxi­dation of carbon starts at 200°C and burns out com­pletely between 600 and 750°C or at least just be­low 800°C (Cuomo di Caprio 1985.131; Levi 2010. 121; Gliozzo 2020.26). Experiments showed that the main loss of organic material (the use of barley straw) in daub and kilns occurs between 200 and 300°C, with the final loss at about 400°C (Macphail, Goldberg 2018.235). Between 675 and 870°C, cal­cite decomposes completely into calcium oxide, with cell volume decreasing and crystal size increasing (Gliozzo 2020.6). Above 800°C, complex aluminosi­licates form and the phase of sintering begins (Levi 2010.121), which is lower for carbonate clay (around 800°C). Hence, we cannot develop a unified phase diagram for the firing process (Gliozzo 2020.5). The presence of calcium carbonate contributes to a low­er sintering temperature because lime acts as a flux, causing vessels with admixed calcium carbonate to sinter faster (Maggetti et al. 1984; Shoval 2016.12). In highly processed deposits, a fired mineral is pre­sent as rubefied mineral inclusions ranging in size Manca Vinazza, Matej Dolenec from silt to sand-size. The presence of these materi­als provides increased magnetic susceptibility, whe­reas iron-free minerals are not rubefied and there­fore magnetic susceptibility is naturally high (Mac­phail, Goldberg 2018.235). Iron oxides were frequently used for coatings, espe­cially for the so-called ceramica zonata, which is typical for the Early Iron Age in a large area, rang­ing from northern Italy (Este, Padua) to Slovenia. Examples of so-called ceramic situlae from Slovenia are mostly considered imports (cf. Grahek 2018. 315). Red slips are also typical of other vessel types (e.g., Dolenjska region in south-eastern Slovenia; Dular 1982.90). Iron oxide haematite (Fe2O3) pro­vides a red or reddish colour, while magnetite (Fe3O4) provides grey, blue, green, and grey-brown colours, and in the reduction, black (Heimann 1978–1979. 86; Rice 2005.334–336). The firing atmosphere controls the partial pressure of oxygen, which is higher in the oxidizing atmos­phere. By leading the firing, we change the firing phases (Heimann 1978–1979.86). In an oxidizing atmosphere, complete firing occurs. We need dry firewood and an air supply to achieve the combus­tion of organic matter and the decomposition of sul­fides if the latter is present. In a reduction atmo­sphere, incomplete firing happens. We close the air supply and pile organic material into the kiln (e.g., horse hooves, straw, etc.), which may be slightly moist (Cuomo di Caprio 1985.126, 131). In prehistory, different structures for firing pottery were known. The basic division is into firing in the open (bonfire) and firing in a kiln. In a bonfire, the maximum temperature is reached quickly (approx. 10–50 minutes, usually 20–30 minutes), while firing in a kiln takes longer (approx. 60 minutes to 11–12 hours). In a bonfire, the soaking time at maximum temperature is shorter (a few minutes) than in a kiln (up to 30 minutes). Oxidative and reductive atmo­spheres can be achieved in both, yet the latter is much more controlled in two-chamber kilns with a perforated floor, especially the exchange. Cooling takes less time in a bonfire (a few minutes to 1 hour), while it is slower in the kiln (1–4 days), and firing in a bonfire also takes less time than firing in the kiln (Gliozzo 2020.2–3). Any natural clay type can be fired at low tempera­tures, i.e. below 800°C, but the lowest possible tem­perature for firing pottery is 500°C (Rye 1981. 16, 96). The data shows that prehistoric pottery was mostly fired between 550 and 650°C or at the most up to 750–850°C. Analyses from the field of experi­mental archaeology generally reveal that tempera­tures of 950°C and even 1100°C could be reached in the kilns known in the Neolithic (Kovárník 1999. 315–317). Nevertheless, it was found that the soak­ing time is of greater importance than the maximum firing temperature (Gosselain 1992.244, Fig. 1). Fur­thermore, it was also found that the temperature of the core of the vessel burned in a bonfire is not unified and that a temperature difference occurs be­tween the outer surface and the fracture up to 220°C. The latter was also confirmed in the exam­ple of vessels fired in a kiln (Maggetti et al. 2011; Gliozzo 2020.4). Consequently, it is necessary to examine the question of which part of the pottery production is associated with a particular firing structure (Thér 2014.78) or which forms or types of vessels were fired in which structure (Gliozzo 2020. 27). Based on 72 archaeological experiments, Ri­chard Thér (2014) showed that there are no diffe­rences between firing structures at temperatures up to 1050°C when thermal profiles are observed and that the firing method is more important than the firing structure (Thér 2014.79–80, 93). Later, he tried to find out if it was possible to distinguish be­tween products fired in a bonfire and those fired in a kiln by observing the thermal gradient of maxi­mum temperature (XRD analyses) in the core of the vessel and in the outer and inner surfaces. He disco­vered that the difference between the maximum temperature between the outer surface and the core was 100–200°C when fired in a bonfire, and be­tween 0 and 50°C when fired in a kiln (Thér et al. 2018.1144–1145, 1169). This means that we have finally found a way to distinguish firing in bonfire from firing in a kiln. Archaeological background In general, only a few pottery firing structures from the Bronze and Iron Ages have been found in the area of present-day Slovenia. Until the end of the Early Iron Age, only single-chamber kilns are known (e.g., Oloris near Dolnji Lako. from the Late Bronze Age, Dobrava near Oto.ec (Horvat .avel 1988–1989. 130–131; Dular et al. 2002.37, T. 24–25; Josipovi. et al. 2015.16, Figs. 11–12). While a two-chamber kiln with a perforated floor was found at the Late Iron Age site Hajdina at Ptuj (Tomani. Jevremov, Gu.tin 1996.271, Fig. 4). Generally, two-chamber kilns were used much earlier, as they appear indivi­dually in Italy no later than in the Middle Bronze Age (Bronzo recente), while their use increases in Pottery firing in the Early Iron Age in western Slovenia the Early Iron Age1 (Levi 2010.117). To date, no fir­ing structures have been found in western Slovenia. All the consumptions in the literature are still based on indirect data, on the macroscopic analysis of the technology of pottery firing from three hillforts in the Kras, Tabor near Vrab.e, Tomaj, and .tanjel (Fig. 1). In addition to stratigraphic data, also radiocar­bon dates are also available (Vinazza 2021.430– 433, Fig. 5). The Tabor site near Vrab.e belongs to the transition from Late Bronze to Early Iron Age (Phase 1; 11.–10. cent. BC) and Early Iron Age (Phase 2; 8.–7. cent. BC). The Tomaj site belongs to the Early Iron Age (6.–5. cent. BC) and .tanjel be­longs to the end of Early Iron Age (6th and 5th cent. BC). The macroscopic analysis indicated that at the end of the Late Bronze Age (Ha A2/B1) the major­ity of pottery was fired in a reduction atmosphere (e.g., Tabor near Vrab.e 38.2%). In the 8th and 7th cent. BC (Ha C0–C2) the ratio begins to change as less and less reductive firing take place, only 4.8% at Tabor near Vrab.e and 25% at Tomaj. This trend continues until the end of the Early Iron Age, as shown by the analysis of pottery from .tanjel (7.1– 23.9%). It should be emphasized that incomplete oxidation firing predominates throughout at all the sites, with the proportion of oxidation firing in­creasing only at the end of the Early Iron Age in the case of .tanjel (38.7%) (ibid. Tab. 3). Changes are also observed in the preparation of the pottery paste. The macroscopic analysis of pottery mass confirms the findings in the wider area of north-eastern Italy and western Slovenia, as a temper of calcite prevails at the end of the Early Iron Age, while pottery is more frequently fired in oxidation atmosphere (Sa­racino 2014.104–122, 131–132; Grahek 2018.311; cf. Vinazza 2021.433, Fig. 1). Certain sites in Kras and in the Poso.je region2 re­vealed individual examples of the so-called red-black painted pottery (Este style, Ita. ceramica zonata). Case analyses from Poso.je region, from Most na So­.i, showed the presence of iron and manganese to achieve the red and black colour of coating on the pottery (Grahek 2018.313–314). This final colour effect is the result of the so-called three-stage fir­ing (ORO), in which oxidation and reduction firing Fig. 1. Sites mentioned in the text (© Google Earth Pro). 1 Ponte San Marco (Poggiani Keller 1994.76), Forcello di Bagnolo S. Vito (Rapi et al. 2019.107), Montedoro di Scapezzano, Mate-lica (Macerata), Marche, Cesena, Foro Annonario (Gasparini, Miari 2017.24), Padova, Ex Brolo (Iaia, Moroni Lanfredini 2009.65, 68, 70). 2 Repentabor (7th and 6th cent. BC) (Maselli Scotti 1978–1981.Fig. 9.1); Repni. from the 7th and 6th cent. BC (Maselli Scotti 1983.T. 54, 214), .tanjel (Vinazza 2011.T. 9.103), and Most na So.i (Grahek, Ko.ir 2018.315). Manca Vinazza, Matej Dolenec atmospheres exchange (cf. Aloupi-Siotis 2020.3, 5). This is made possible by the two-chamber kilns with a separated fireplace and firing parts, which have been already mentioned several times before. One of the most important parts of such kilns is the per­forated floor. When determining this type of kilns based on the perforated floor, we need to be careful since such perforated floors were also used in, for example, salt extraction in the area under study in this work.3 Such a ceramic situla was also found at the .tanjel site (Vinazza 2011.T. 9. 103), and is part of the present study. Methods and sampling For the present study, which is first of this kind in the studied area, we analysed 18 samples. The ar­chaeological pottery comes from two hillforts, Tabor near Vrab.e and .tanjel (for more details, see Vinaz­za 2021.422–425). Other samples were produced during archaeological experiments. Besides experi­mental archaeology, we also carried out additional analysis, such as AMS measurements, ceramic petro­graphy analysis, and X-ray diffraction analysis (see Tab. 1). We conducted experiments on the construction and use of a single-chamber kiln in order to better un­derstand why the remains of kilns are so poorly pre­served in the archaeological record in the studied area, and to get samples for observing changes in material under different temperatures. The kiln was built on the model of a Late Bronze Age kiln from the Oloris site near Dolnji Lako. (Horvat .avel 1988– 1989.130–131). We were thus able to observe its manufacture, material, construction, and use during firing, as well as its decay. Clay,4 straw, and hazel branches were used for the construction of the kiln (16.6.2020). We prepared the mixture of clay, water, and straw (40%). The bottom of the 10cm deep pit was first covered with a clay mixture. After that, a construction from hazel branches, which was covered on the outside with clay strips of 20x15x5cm, was built. Seven people built the kiln in 5 hours (the preparation of the ma­terial and the construction). After 4 days, the kiln was dried by burning spruce chips (we used up 12kg), which took 6 hours and 30 minutes. After that, the kiln was ready for pottery firing. The firing of the pottery took place after two months (28.8.2020). The kiln was loaded with 57 vessels that we formed from different local clays.5 Two ther­moelements6 were installed onto the kiln, one along the kiln wall and the other in the centre, just below the vessels. We wanted to understand if there was any change in temperature in the kiln during firing. To obtain different types of data and their possible application to the archaeological remains, we mea­sured the AMS of the kiln.7 The magnetic susceptibility of the ground8 or sed­iment is determined by the amount of magnetic mi­nerals present. During burning, the magnetic suscep­tibility increases because iron minerals are bound. If the ground or the sediment does not contain iron minerals, the magnetic susceptibility is not high. The iron content depends on the geological background (Goldberg, Macphail 2006.350–351; see also Mu.i. 1999.363). If magnetite is present in the clay, mag­netic susceptibility is naturally high (Macphail, Gold­berg 2018.236). Measurements were taken in the laboratory on soaked clay (the Ren.e clay) (mixed with water and straw: 0.850) from which the kiln was built, on samples from the kiln after drying, and on samples from the kiln after firing (Fig. 3). The va­lues given are average values.9 Clay from Ren.e that was used for building the kiln was also fired in a controlled atmosphere. We have 3 A perforated floor from the Ellerji hillfort has been interpreted several times before as the remains of salt production (Lonza 1981. T. 44–45; Zendron 2018), while the remains of a perforated floor from the Monkodonja site in Istria were among the earliest in the wider area (Early and Middle Bronze Age) (Mihovili. 2020.36–39, Fig. 31). This means that they appear significantly earlier than in the entire Italic peninsula, which indicates the supra-regional role of Monkodonja. 4 The Ren.e deposit: GKY 396454, GKX 83339. 5 Gri.e: GKY 417619, GKX 69497; Veliki Dul: GKY 411812, GKX 70871; Lukovica: GKY 476504, GKX 112322, and Ren.e: GKY 396454, GKX 83339. 6 Thermoelement type MTC500 with a Ni-Kr-Ni tip. 7 The Kappameter KT-7 (GF Instruments) instrument was used. 8 Values of AMS were measured on various samples of clays and present the results that do not enable simplified conclusions, since the values range from 0.1 to 8·10–3SI. Location near Tupel.e (GKY 407902, GKX 73627): clay to 6–8·10–3SI; Vrab.e (GKY 409555, GKX 77491): clay 0.063–0.207·10–3SI; Veliki Dul (GKY 409555, GKX 77491): 0.532–0.666·10–3SI; Ostri vrh (GKY 409555, GKX 77491): 0.766–0.966·10–3SI. 9 Three measurements were taken for every point and the average value was calculated. Pottery firing in the Early Iron Age in western Slovenia Sample ID Petrolab ID Source Description Clay Add informations Manipulation Firing Method Sample 1 2020-6 Archaeological pottery, site {tanjel Neck of the pot Unknown US 52 0 0 Macroscopic technological analysis1%). Sharp calcite edges (rhombohedral cleavage) indicate intentional crushing (Fig. 6.2–3). Calcite also predominates as a temper in the final stage of the Early Iron Age, in the 6th and 5th cent. BC (40%), as pottery from .ta-njel shows (Tab. 1.1). The difference is visible in the size of the calcite temper. The grains are bigger in pastes from the 8th and 7th cent. BC than in those from the 6th and 5th cent. BC. In the later period there is a finer temper of calcite. The temper is still poorly sorted, but in comparison with older materi­al there are no voids. The kiln wall pottery thin section (Fig. 5) corre­sponds to the basic characteristics of such objects found at archaeological sites. Here we have in mind the main micromorphological features of mudbricks, such as the presence of organic elongated fragments and randomly oriented channels and voids, reflect­ing the addition of straw into the otherwise very compact structure during the preparation (cf. Frie­sem et al. 2018.99–100). In our case (Sample 4) the are many planar voids and channels, indicating that straw is the only temper used for kiln paste. Other features are inclusions, such as Fe-oxides, clay pel­lets, quartz grains, and some other opaque minerals. In order to determine the origin of the classic situ-la from .tanjel and thus the possibility of the pres­ence of the two-chamber kilns, we conducted a com­parative study of the pottery masses using pottery thin sections. We took samples of local clay (Fig. 7B; Tab. 1.16), silos1 0 (Fig. 7.D; Tab. 1.15), and pottery from the .tanjel site. One from the local form (Fig. 7.A; Tab. 1.1) and one from a presumably imported ceramic situla (Fig. 7.C; Tab. 1.14). Silos, such as ce­ramic rings, loam weights, and house plaster, are ge­nerally made from the clay closest to the site, mak­ing them a good comparison for determining local/ imported products. Even a quick look at the pottery thin section of a silo (Fig. 2.4) and the ceramic sit-ula (Fig. 7.3) shows that we are dealing with diffe­rent clays. The silo contains muscovite/illite (up to 20%) and polycrystalline and monocrystalline quartz (25%), while the muscovite/illite is not present in the ceramic situla. The latter also did not include clasts of trachyte, which is typical of such forms from the Euganean area (see Saracino 2014.120, 144). The clay matrixes of the local vessels, the silo, and the local clay, sampled near .tanjel are very close in composition, while the ceramic situla stands out. 10 A silo petrographic thin section (Fig. 2.4) indicates the presence of certain carbonates, which are unchanged, meaning that firing took place at a temperature from 675 to 870°C. The same is true for the pottery from Ren.e (Sample 2), fired at 670°C. The silos as such is also solid (7 according to the Mohs scale), which reflects firing at a high enough temperature and at the same time changes the idea that such pottery forms were fired at low temperatures (Vinazza 2016.7). Manca Vinazza, Matej Dolenec The XRD analyses were performed on 12 samples (Tab. 1.1–13). The objective of the analysis was to determine the comparison of the XRD analysis re­sults with the firing temperature in the kiln, which was measured with thermoelements during firing. We also tried to show that the firing temperatures of individual kiln parts do not reflect the tempera­ture of the pottery firing, which consequently cannot be applied to the archaeological material. Third, fol­lowing the lead of Ther (2020), we sought to determine whether we could detect differences between the results of XRD analysis at the fracture and on the outer surface of the pottery, and thus determine the use of a bonfire and/or kiln at the site. The analyses were carried out on the samples of the kiln from the Ren.e clay, which is of the illite-chlorite type (Rokavec 2014.35), and on the pottery from the .tanjel site, which belongs to the end of the Early Iron Age (Tab. 1; Fig. 8). Fig. 3. The AMS measurement points were chosen on various parts of the kiln (in the centre) and of the individual parts (kiln wall (A–C), chimney (D–F), and kiln floor (G–H)). Ren.e clay (Samples 2–8, 12) Clay from Ren.e (Sample 8) has a higher amount of kaolinite (Fig. 8) which is not present in the other samples (Samples 2–5), meaning that the latter were fired at over 550°C. Comparing the parts of the kiln (floor/walls/chimney), most kaolinite is found in the kiln walls (Sample 4), less in the floor (Sample 3), while no kaolinite is present in the sample from the chimney (Sample 5). It is therefore understandable that most of it is in the wall where the temperature in the kiln was the lowest. Sample 6 (firing at 600°C) contains very little kaolinite, while Sample 7 (firing at 800°C) and Sample 12 (firing at 670°C) contain no kaolinite. Illite, which begins to decompose at 900°C, is present in all samples, while only Sample 7 contains less because it was fired at 800°C. Quartz is also present in all samples. Calcite is also present in Samples 6 and 8, which we attribute to its natural occurrence in the clay. Sample 12 (Fig. 8) has an ele­ Pottery firing in the Early Iron Age in western Slovenia vated calcite value, which we attribute to the inten­tional addition of the temper of calcite to the clay, which is also confirmed from pottery thin sections. It is no longer present in Sample 7, as it begins to decompose above 670°C. Dolomite is found in Sam­ples 6 and 8, but in small amounts and is no longer present in Sample 12. .tanjel pottery (Samples 1, 9–11) Calcite is present in all samples indicating that the pottery was fired at temperatures below 870°C, at which calcite decomposes completely. Sample 1 (Fig. 8) has an increased value of calcite, which we attri­bute to the intentional admixture of the temper of the calcite to the clay (Fig. 8), which is also confirm­ed by pottery thin sections. At the same time, kaoli­nite is no longer present, indicating that the pottery was fired at over 550°C. The samples from .tanjel contain very little illite, which is due to the minera­logical composition of the clay. A comparison of the calcite in the vessel’s core (Sample 10) and in the outer (Sample 9) and inner surfaces (Sample 11) shows that the vessel’s core contains more calcite. Discussion Archaeological finds that would indicate the firing of the pottery in Kras and the Poso.je region in the Bronze and Iron Ages are not known for either a bo­nfire or a kiln. There are at least two possible rea­sons for this. First, slow sedimentation at Kras is very problematic from a stratigraphical point of view. In most cases, the sites have very thin archa­eological layers and only rarely do we discover a longer stratigraphic sequence (cf. Monkodonja in Is-tria; Hänsel et al. 2015.75). The soil at Kras is char­acterized by the bedrock, various limestones and do­lomites, their decomposition and dissolution, and the leaching of debris into relief depressions. On karstified hills and in higher areas there is less soil, while in depressions, e.g., dolinas (Habi. 1979.150), there are uniform and thicker layers. The discussed pottery originates from the hillfort sites of Tabor near Vrab.e and .tanjel (Vinazza 2021), where there is in both cases less soil. The second reason is connected with the firing struc­tures and the question of how to recognize them in order to understand the firing process. Structures such as bonfires do not leave any significant traces behind, and are thus difficult to discern. If we con­sider a burned layer of soil, a large pile of plant char­coal, wooden charcoal, burned-through soil, and burned lumps of soil as the key indicators of the re­mains of a bonfire for firing pottery (Guo 2017. 184), then some of the structures found at several sites from Eneolithic to the Early Iron Age in cen­tral and north-eastern Slovenia could be interpret­ed as bonfires.1 1 We are still missing this kind of data for the Kras area. 11 The Eneolithic: Kalinovjek, SE 171, 174, 176, 178, 257, 259 (Kerman 2013.58, 59, 62); the Early Bronze Age; Nova tabla, PO 29, PZ 24 (Gu.tin et al. 2017.112, 115); the Middle and Late Bronze Ages: Nedelica pri Turni..u SU 344/343, 372, 381 (.avel, San­kovi. 2013.78), Svetje, SU 41/42 (Leghissa 2011.86); the Late Bronze Age: Pod Kotom – sever pri Krogu, SU 347 (Kerman 2011. 71); Orehova vas, SU 160M, 191A, 81R (Grahek 2015.53, 59, 88)); the Early Iron Age: Nova tabla, PO 223 (Gu.tin et al. 2017. 127); Hotinja vas near Maribor, SU 271/272 (Gerbec 2015.43). Manca Vinazza, Matej Dolenec The situation is a different matter with kilns. They decay in a certain phase, but their remains depend on different situations. Today’s climate in the dis­cussed area is too dry, and poorly fired structures decompose into the matrix of the archaeological de­posit (Amicone et al. 2020.522), and it is also pos­sible that the space is reused at a later date. We believe one of the reasons why these kilns have not been preserved is also due to the use of these structures. Our archaeological experiments have shown that different parts of kilns are exposed to different temperatures, which is consistent with the results of the AMS. The most exposed parts were the kiln floor and the chimney. Higher values of the AMS mean that these parts were fired better, which affects the degree of preservation of these parts of the kiln. In the exposure to different temperatures, we see the reason for the poor preservation of the kilns, which can be applied to the wider area.1 2 From this, we can say that we can mistake the remains of the kiln floor with, for example, a hearth. Hence we believe that we should focus our attention on pos­sible remains of an interlacement that could have belonged to the former dome,1 3 which will signifi­cantly contribute to the final interpretation of whe­ther it is a kiln or a hearth. Consequently, we believe that the firing tempera­tures of vessels cannot be determined by analysing the temperature of the parts of kilns. As already mentioned, different values of AMS indicated diffe­rent temperatures of the firing of different kiln parts. Since kiln parts are usually randomly preserved, it means we do not necessarily obtain the best-fired part when sampling. Moreover, in the case of kilns the soaking time plays an important role, since bricks, for example, are fired for several days be­fore they are properly fired. When macroscopically observing the core of cubes made of Ren.e clay (App. 1: Samples 6 and 7) it can be seen that the core is still grey at 600°C, while at 800°C the grey part shrinks. Evidence of this is the results of the XRD analysis of vessels and kiln parts from the Kra.nja site in Slove­nia, which showed that the vessels were fired at about 800°C, while samples of the wall and bottom of the kiln indicate a temperature of no more than 500°C (.ibrat Ga.pari. et al. 2014.232, 234). We can also apply these results to pottery firing. As Thér et al. (2018) have already shown, the soaking time needs to be considered when comparing firing in a bonfire with firing in a kiln. 12 Here we always need to compare AMS measurements of the local clays, since the values may depend on the natural composition and the presence of, for example, magnetite. 13 We have found five pottery kilns at the roman site Otok pri Metliki in southeastern Slovenia. Above the perforated floor there was a red layer full of small pieces of burned clay. There were no visible marks of the construction made with branches (see Udov., Vinazza 2018.147–149). Pottery firing in the Early Iron Age in western Slovenia Fig. 6. Changes in the pottery recipes (petrographic thin sections and site phases (1 Late Bronze Age; 2 Early Iron Age; 3 end of Early Iron Age)) in combination with radiocarbon dates from the sites of Tabor near Vrab.e and .tanjel. Photos taken under plain polarized and cross polarized light. Manca Vinazza, Matej Dolenec There are no differences in the mineralogical com­position of the .tanjel pottery when comparing the vessel’s core and inner and outer surfaces (Samples 9–11). We take the lack of variation in these values as evidence that the pottery was fired in a kiln. The calcite values have not changed, which means that the decomposition of the calcite has not started at a temperature that has not exceeded 850°C. Here, we would like to point out that the increase of calcite that was already shown by macroscopic tech­nological analysis (Vinazza 2021) and confirmed with pottery thin sections in this study, in pottery from sites Tabor near Vrab.e and .tanjel is not linked to a particular vessel type, but is noted in va­rious forms, such as pots, dishes, and lids. This means the reason for the addition of calcite is not only related to the functional properties, which in­creases the resistance to thermal shock (Bronitsky, Hamer 1986.95–99), but also to the firing process, since calcite acts as a flux that allows carbonate clays to be fired at lower temperatures (Shoval 2016.12). Since calcite decomposes at a temperature of up to 870°C and since the XRD results for .tanjel pottery show that the temperature did not exceed 870°C, we assume that the use of kilns makes it easier to control the temperature and thus the use of such an amount of calcite.1 4 Finally, we analysed ceramic situla from .tanjel in order to find confirmation of ORO firing of this type of vessel and consequently the potential use of two-chamber kilns with a perforated floor in the Kras area. As the ceramic situla from .tanjel was not made from the same clay, as the other samples show, we see it as an imported vessel. Samples from Most na So.i from Poso.je (Grahek, Ko.ir 2018.309–311, 314–315; sample MNS D or Most 4) suggest the pos­ 14 This material is ubiquitous in the discussed area (see Jurkov.ek 2013). Pottery firing in the Early Iron Age in western Slovenia Fig. 8. Mineralogy distribution (XRD results). sibility that some of the pieces were imported from workshops in the Este area, while others are prob­ably the product of local workshops. In our case, no pieces of trachyte were found in the pottery mass, which is typical for the Euganean area (see Saraci-no 2014.144). We believe that the area of origin of the ceramic situla from .tanjel is elsewhere and the possibility of local workshops is still open to discus­sion, since the XRD results also show no or a very small amount of muscovite in pottery from .tanjel (Fig. 8.samples 9, 10). Conclusion With the help of various research methods and sci­entific analyses, we have tried to determine whether the use of kilns for pottery firing can be expected in the Early Iron Age in western Slovenia. The pottery-manufacture technology of this period suggests this possibility, and the same is true for the results of XRD analysis, which do not reveal major tempera­ture deviations between the outer surface and core of the vessel, which means that the soaking time was long enough to allow gradual and uniform fir­ing of the vessels. The XRD analyses of the pottery from .tanjel show that the temperature did not ex­ceed 870°C, while the addition of calcite as a tem­per, which did not decompose, suggests that the fir­ing took place under controlled conditions, which can be better controlled in a kiln (e.g., no sudden temperature rise due to the wind blowing). The pre­valence of oxidative firing, which is much more con­trolled in a kiln, is also supported by the macrosco­pic analysis of pottery in western Slovenia at the end of the Early Iron Age. Based on the above, we as­sume that in western Slovenia at the end of the Early Iron Age (the 6th and 5th cent. BC) only single-chamber kilns for the firing of pottery were known, even though archaeological excavations have not (yet) brought them to light. We need to point out that a single-chamber kiln from the Early Iron Age was found in the Dolenjska region (site Dobrava near Oto.ec), as mentioned above, but for the area of Friuli Plain in Italy we still have no evidence. In Manca Vinazza, Matej Dolenec the future, we will have to pay more attention to the lity) of the area and not only in terms of the level excavations of such structures, and in the Kras area of technological development, as is often the case in there is a lack of scientific research. Only with such archaeological studies. an approach will we be able to understand more about pottery technological practices in the Early Iron Age. ACKNOWLEDGEMENTS The article was written within the Archaeology Re- However, the non-local origin of the ceramic situla search Programme at the Scientific Institute of the from .tanjel does not suggest the use of two-chamber Faculty of Arts of the University of Ljubljana (P6­ kilns with a perforated floor for the firing of the pot­ 0247) and the Geoenvironment and Geomaterials tery at the end of the Early Iron Age in the Kras area. Programme Group at the Faculty of Natural Sciences Since the clay for ceramic situla does not originate and Engineering (P1-0195), which were co-financed from the Eugaeum area, we still need to find a closer by the Slovenian Research Agency from the state bud-production area for this type of vessel. Some addi-get. The authors acknowledge the financial support tional local clay sampling in a broader area close to from the Slovenian Research Agency (research core key Early Iron Age sites (e.g., Tomaj, Most na So.i, funding No. P6-0247 and P1-0195). Gradisca di Spilimbergo in Italy) thus needs to be We would like to thank all the students of the Depart­ ment of Archaeology, Faculty of Arts, University of done. Ljubljana who helped build the kiln, Paola Koro.ec for help during the firing process, and Dr. Jaka Bur- Finally, we would like to point out that the level of ja from The Institute of Metals and Technology (IMT) technological knowledge was also determined by the for firing samples in the electrically operated kiln. properties of the raw material available in a certain We would also like to thank the company Gori.ke area. Thus, the final results must also be under- opekarne d.o.o. for the donation of the clay. stood in light of the natural resources (e.g., clay qua­ . References Aloupi-Siotis E. 2020. Ceramic technology: how to charac­terise black Fe-based glass-ceramic coatings. Archaeologi­cal and Anthropological Sciences 12: 191. http://doi.org/10.1007/s125210-020-01134-x Amicone S., Croce E., Castellano L., and Vezzoli G. 2020. Building Forcello: etruscan wattle-and-daub technique in the Po plain (Bagnolo San Vito, Manua, Notrhern Italy). Archaeometry 62–3: 521–537. https://doi.org/10.1111/arcm.12535 Cultrone G., Rodriguez-Navarro C., Sebastian E., Cazalla O., and De La Torre M. J. 2001. Carbonate and silicate phase reactions during ceramic firing. European Journal of Mineralogy 13: 621–634. Cuomo di Caprio N. 1985. La ceramica in archeologia. La ceramica in archeologia: antiche tecniche di lavora­zione e moderni metodi d’indagine. L’Erma di Bretschnei­de. Rome. Dular J. 1982. Hal.tatska keramika v Sloveniji. Dela SAZU 23. Slovenska akademija znanosti in umetnosti. Ljubljana. Dular J., Tecco Hvala S., and Horvat .avel I. 2002. Brona­stodobno naselje Oloris pri Dolnjem Lako.u. Opera Insti­tuti Archaeologici Sloveniae 5. Zalo.ba ZRC. Ljubljana. Gasparini D., Miari M. 2017. Lo scavo del Foro Annonario di Cesena: indagini di un sito dell eta del Bronzo in un contesto pluristaficato urbano. In F. Rubat Borel, M. Cupi­to, C. Delpino, A. Guidi, and M. Miari (eds.), Le eta del Bronzo e del Ferro in Italia: contesti protostorici in sca-vi urbani. Incontri annuali di preistoria e protostoria. II Incontro Annuale di Preistoria e Protostoria IAPP. Isti­tuto italiano di preistoria e protostoria. Firenze: 23–26. Gerbec T. 2015. Hotinja vas pri Mariboru. Zbirka Arheo­logija na avtocestah Slovenije 45. Zavod za varstvo kultur­ne dedi..ine Slovenije. Ljubljana. Gliozzo E. 2020. Ceramic technology. How to reconstruct the firing process. Archaeological and Anthropological Sciences 12: 260. https://doi.org/10.1007/s12520-020-01133-y Goldberg P., Macphail R. I. 2006. Practical and theoretical geoarchaeology. Wiley-Blackwell. Malden, Oxford, Victoria. Gosselain O. P. 1992. Bonfire of the Enquiries. Pottery Firing Temperature in Archaeology: What for? Journal of Archaeological Science 19: 243–259. Grahek L. 2015. Orehova vas. Zbirka Arheologija na avto­cestah Slovenije 46. Zavod za varstvo kulturne dedi..ine Slovenije. Ljubljana. Pottery firing in the Early Iron Age in western Slovenia 2018. Naselbinska keramika z Mosta na So.i/Pottery from the settlement at Most na So.i. In J. Dular, S. Tec-co Hvala (eds.), .eleznodobno naselje Most na So.i. Razprave/The Iron Age settlement at Most na So.i. Treatises. Opera Instituti Archaeologici Sloveniae 34. Ljubljana: 249–307. Grahek L., Ko.ir A. 2018. Analiza naselbinske keramike z Mosta na So.i z vrsti.nim elektronskim mikroskopom/ Scanning electron microscopy analysis of the pottery from the settlement at Most na So.i. In J. Dular, S. Tecco Hva-la (eds.), .eleznodobno naselje Most na So.i. Razprave/ The Iron Age settlement at Most na So.i. Treatises. Ope­ra Instituti Archaeologici Sloveniae 34. Ljubljana: 307–321. Guo M. 2017. Variability in pottery firing technology. Choice or technical development? Chinese Archaeology 17: 179–186. https://doi.org/10.1515/char-2017-0015 Gu.tin M., Tiefengraber G., Pavlovi. D., and Zorko M. 2017. Nova tabla pri Murski Soboti. Zbirka Arheologija na avtocestah Slovenije 52/1. Zavod za varstvo kulturne dedi..ine Slovenije. Ljubljana. Habi. P. 1979. Problematika geografskega vrednotenja Krasa. Geografski vestnik LI: 147–158. Hänsel B., Mihovili. K., and Ter.an B. 2015. Monkodonja. Istra.ivanja protourbanog naselja bron.anog doba Is-tre. Knjiga 1. Iskopavanje i nalazi gra.evina / Monko­donja. Forschungen zu einer protourbanen Siedlung der Bronzezeit Istriens. Teil 1. Die Grabung und der Bau­befund. Monografije i katalozi/Monographien und Katalo­ge 25. Arheolo.ki muzej Istre/Archäologisches Museum Is-triens. Pula / Pulj. Heimann R. B. 1978–1979. Mineralogische Vorgänge beim Brennen von Keramik und Archäothermometrie. Acta Praehistorica et Archaeologica 9–10: 79–102. Horvat .avel I. 1988–1989. Bronastodobna naselbina Oloris pri Dolnjem Lako.u. Arheolo.ki vestnik 39–40: 127–145. Iaia C., Moroni Lanfredini A. (eds.). 2019. L’eta del ferro a Sansepolcro. Attivita produttive e ambiente nel sito di Trebbio. Aboca Edizioni. Peruggia. Jeffra C. D. 2015. Experimental approaches to archaeolo­gical ceramics: unifying disparate methodologies with the chaíne opératoire. Archaeological Anthropological Sci­ence 7: 141–149. Josipovi. D., Turk M., Rupnik J., Brezigar B., To.kan B., and Kova. M. 2015. Kon.no strokovno poro.ilo o arheo­lo.ki raziskavi ob gradnji. Arheolo.ko izkopavanje v Dobravi – Oto.cu (parc. .t. 121, k. o. 1460 .entpeter). Unpublished archaeological report. Institute for the Pro­tection of Cultural Heritage of Slovenia, Nova Gorica. Id-rija. Jurkov.ek B. 2013. Geolo.ka karta Krasa/Geological map of Kras (Slovenia). Kerman B. 2011. Pod Kotom – sever pri Krogu. Zbirka Arheologija na avtocestah Slovenije 24. Ljubljana. 2013. Kalinovnjek pri Turni..u. Zbirka Arheologija na avtocestah Slovenije 33. Ljubljana. Kovárník J. 1999. Die Technologie der vorgeschichtlichen Keramik mit Rücksicht auf ihr Brennen. In E. Jerem, I. Po-roszlai (eds.), Archaeology of the Bronze and Iron Age. Experimental Archaeology. Archaeolingua. Budapest: 315–327. Leghissa E. 2011. Bronastodobna naselbina Medvode – Svetje. Obdelava arheolo.kih ostalin in najdb iz izkopa­vanj 2007. Diploma thesis. University of Ljubljana. Faculty of Arts. Ljubljana. Lemonnier P. (ed.). 1993. Technological choices. Trans­formation in material cultures since the Neolithic. Ma­terial cultures. Routhledge. London & New York. Levi S. T. 2010. Dal coccio al vasaio. Manifattura, tecno­logia e classificazione della ceramica. Zanichelli. Bo­logna. Lonza B. 1981. La ceramica dei castelliere degli Elleri. Societa per la preistoria e protostora della regione Friuli-Venezia Giulia, Quaderno n. 4. Edizioni “Italo Svevo”. Tri­ste. Macphail R., Goldberg P. 2018. Applied Soils and Micro-morphology in Archaeology. Cambridge Manuals in ar­chaeology. Cambridge. Maggetti M, Neururer C., and Ramseyer D. 2011. Tempe­rature evolution inside a pot during experimental surface (bonfire) firing. Applied Clay Science 53(3): 500–508. https://doi.org/10.1016/j.clay.2010.09.013 Maritan L. 2018. Ceramic material. In C. Nicosia, G. Stopps (eds.), Archaeological soil and sediment micromorpfo-logy. John Wiley & Sons. Oxford: 205–212. Maselli Scotti F. 1978–1981. Primi risultati sullo scavo di Cattinara ed i castellieri triestino nel’eta del ferro. Atti della Societa per la Preistoria e Protostoria della regio­ne Friuli-Venezia Giulia 4: 281–309. 1983. Le strutture dei castellieri di Monrupino e Rupin-piccolo. Preistoria del Caput Adriae. Videm: 214–215. Manca Vinazza, Matej Dolenec Mihovili. K. 2020. Zidni glineni dijelovi gra.evina i opre-me iz bron.anodobne gradine Monkodonja. In B. Hän­sel, K. Mihovili., and B. Ter.an (eds.), Monkodonja. Istra-.ivanje protourbanog naselja bron.anog doba Istre. Knjiga 3. Nalazi od metala, gline, kosti i kamena te ljud­skih i .ivotinjskih kostiju/Forschungen zu einer proto­urbanen Siedlung der Bronzezeit Istriens. Teil 3. Die Fun-de aus Metall, Ton, Knochen und Stein sowie die mensch-lichen und tierischen Knochen. Pula/Pulj: 13–109. Mu.i. B. 1999. Geophysical prospecting in Slovenia: an overview with some observatons related to the natural environment. Arheolo.ki vestnik 50: 349–405. Poggiani Keller R. 1994. L’ultima fase di vita dell’insedia­mento. Due fornaci per ceramica della media eta del fer­ro. In R. Poggiani Keller (ed.), Il villaggio preistorico e le fornaci di Ponte S. Marco. Calcinato. Ikonos: 75–89. Quinn P. S. 2015. Ceramic Petrography. The Interpreta­tion of Archaeological Pottery & Related Artefacts in Thin Section. Archaeopress. Oxford. 2022. Thin section Petrography, Geochemistry and Scanning Electron Microscopy of Archaeological Ce­ramics. Archaeopress. Oxford. https://doi.org/10.32028/9781803272702 Quinn P. S., Benzolli A. 2018. XRD and Materials Analysis. DOI: 10.1002/9781119188230.saseas0619 Rado P. 1988. An introduction to the technology of pot­tery. The Institute of ceramics. Pergamon press. Oxford. Rapi M., Quirino T., Castellano L., + 3 authors, and Bus-nelli S. 2019. Per scalfare, per cuocere e per produrre. Le strutture da fuoco dell’abitato etrusco del Forcello di Bag-nolo S. Vito: aspetti tipologici e funzionali. In A. Peinetti, M. Cattani, and F. Debandi (eds.), Focolari, forni e forna­ci tra neolitico ed eta del ferro. Comprendere le attivi­ta domestiche e artigianali attraverso lo studio delle in-stallazioni pirotechnologiche e dei residui di combus­tione. Sesto incontro annuale di preistoria e protostoria. Universita di Bologna. Bologna: 107–109. Rice P. 1987 (2005). Pottery analysis. A sourcebook. The University of Chicago Press. Chicago, London. Rokavec D. 2014. Gline v Sloveniji. Geolo.ki zavod Slove­nije. Ljubljana. Roux V. 2016. Ceramic Manufacture: The chaîne opérato-ire Approach. In A. Hunt (ed.), The Oxford Handbook of Archaeological Ceramic Analysis. Oxford University Press. Oxford: 101–117. Rye O. S. 1981. Pottery technology. Principles and re­construction. Taraxacum. Washington, D.C. Saracino M. 2014. Dalla terra al fuoco. La tecnologia ce­ramica degli antichi Veneti. Scienze e Lettere. Roma. Shoval S. 2016. FTIR in Archaeological Ceramic Analysis. In A. Hunt (ed.), The Oxford Handbook of Archaeologi­cal Ceramic Analysis. Oxford University Press. Oxford: 509–530. .avel I., Sankovi. S. 2013. Nedelica pri Turni..u. Zbirka Arheologija na avtocestah Slovenije 39. Zavod za varstvo kulturne dedi..ine Slovenije. Ljubljana. Thér R. 2014. Identification of pottery firing structures using the thermal characteristics of firing. Archaeometry 56: 78–99. https://doi.org/10.1111/arcm.12052 Thér R., Kallistová A., Svoboda Z., +3 authors, and Bajer A. 2018. How was Neolithic Pottery fired? An Exploration of the Effects of Firing Dynamics on Ceramic Products. Journal of Archaeological Methodology and Theory 26: 1143–1175. https://doi.org/10.1007/s10816-018-9407-x Tomani. Jevremov M., Gu.tin M. 1996. Keltska lon.arska pe. s Spodnje Hajdine pri Ptuju. Arheolo.ki vestnik 47: 267–278. Udov. K., Vinazza M. 2015. Rimski lon.arski obrat na Otoku pri Metliki. In M. Jane.i., B. Nadbath, T. Mulh, and I. .i.ek (eds.), Nova odkritja med Alpami in .rnim mor­jem: rezultati raziskav rimskodobnih najdi.. v obdobju med leti 2005 in 2015. In memoriam Iva Mikl Curk. Za­vod za varstvo kulturne dedi..ine Slovenije. Center za preventivno arheologijo. Pokrajinski muzej Ptuj Ormo.. Ljubljana, Ptuj: 143–158. Vinazza M. 2011. Prazgodovinski .tanjel na Krasu. Izko­pavanja 2010. Diploma thesis. University of Ljubljana. Fa­culty of Arts. Ljubljana. 2016. Silosi – posebne kerami.ne oblike: prispevek k poznavanju gospodinjstev v starej.i .elezni dobi na Kra­su/Silos – special ceramic forms: contribution to the knowledge of the Early Iron households in the Karst region. Arheo 33: 7–23. 2021. Naselbinska keramika starej.e .elezne dobe na Krasu/Settlement pottery from the Early Iron Age in Kras. Arheolo.ki vestnik 72: 419–452. https://doi.org/10.3986/AV.72.14 .ibrat Ga.pari. A., Horvat M., and Mirti. B. 2014. Ceramic petrography, mineralogy and typology of Eneolithic pot­tery from Kra.nja, Slovenia. Documenta Praehistorica 41: 225–236. https://doi.org/10.4312/dp.41.12 Pottery firing in the Early Iron Age in western Slovenia Appendix 1. .tanjel, red slip coating on the situla, outer sur­face. 8x magnification Leica Stereomicroscope EZ4. 3. A thin section of a vessel fired in a kiln at 670°C. 8x magnification Leica Stereomicroscope EZ4. 5. Ren.e clay, fired at 600°C, Sample 6. 8x magni­fication Leica Stereomicroscope EZ4. 2. .tanjel, coating thin section, a fracture (orange colour). 40x magnification. PPL with Zeiss Axio-cam 305 color. 4. A thin section of Sample 2. XPL with Zeiss Axio-cam 305 color. 6. Ren.e clay, fired at 800°C, Sample 7. 8x magni­fication Leica Stereomicroscope EZ4. 7. .tanjel, Samples 9 (outer surface), 10 (fracture), 8. Ren.e clay with admixed calcite, Sample 12. 8x 11 (inner surface). 8x magnification with Leica magnification Leica Stereomicroscope EZ4. Stereomicroscope EZ4. back to contents Documenta Praehistorica XLIX (2022) Between object and subject> multiple approaches to a prehistoric human-shaped pot from Romania Vasile Opris¸1, Bogdan Manea2, Mircea Lechintan3, Roxana Bugoi3, Florin Constantin3, Theodor Ignat1, and Catalin Lazar 2 catalin.lazar@icub.unibuc.ro 1 Bucharest Municipality Museum, Bucharest, RO 2 The Research Institute of the University of Bucharest, Division of ArchaeoSciences, Bucharest, RO 3 Horia Hulubei National Institute for Nuclear Physics and Engineering, Ma¢gurele, RO ABSTRACT – The current paper aims to reveal the potential of combining multiple approaches (tech­no-functional analysis, experimental archaeology, and X-ray Computed Tomography) when it comes to studying unique earthenware artefacts, such as the prehistoric human-shaped pot discovered with­in the tell settlement from Sultana-Malu Rosu (Romania), that belongs to the Kodjadermen-Gumel­nita-Karanovo VI civilization (KGK VI) which thrived during the 5th millennium BC. This human-shaped pot, also known as ‘The Goddess of Sultana’, is an emblematic artefact that fascinates with its shape, gestures, and decoration. It was apparently made from a standard clay paste recipe and using basic forming techniques, with little care for the internal surface. This vessel also has several hidden cracks and some manipulation traces on its backside. In order to explore its relevance, our approach to this particular human-shaped pot included the use of archaeological data in correlation with other techniques in order to decipher the manufacturing process for such vessels, the possible way of using them, but also the meanings that they might have had for past human communities. KEY WORDS – Balkans; Eneolithic; anthropomorphic pot; CT scans; technological analysis; experi­mental archaeology Med objektom in subjektom> razli;ni pristopi k preu;evanju prazgodovinske antropomorfne posode iz Romunije IZVLE.EK – V .lanku predstavljamo potencial zdru.evanja razli.nih pristopov (tehni.no-funkcional­na analiza, eksperimentalna arheologija in ra.unalni.ka tomografija) pri .tudijah unikatnih gline­nih predmetov, kot so prazgodovinske antropomorfne posode, kakr.ne so bile odkrite na tell-nasel­bini Sultana-Malu Rosu (Romunija) iz stopnje Kodjadermen-Gumelnita-Karanovo VI v 5. tiso.letju pr. n. .t. Antropomorfna posoda, znana tudi kot boginja Sultane, je zna.ilna najdba, ki izstopa po obliki, potezah in okrasu. Oblikovana je bila iz obi.ajne lon.arske mase z uporabo osnovnih tehnik ob-likovanja in dodelavo zunanje povr.ine. Na notranji strani so vidne .tevilne razpoke in sledovi iz­delave. V raziskavo smo vklju.ili arheolo.ke podatke in jih povezali s tehni.no-funkcionalnimi ana­lizami, eksperimentalno arheologijo in ra.unalni.ko tomografijo z namenom prikaza procesa izde-lave tovrstnih posod, mo.nega na.ina uporabe ter njihovega pomena za pretekle skupnosti. KLJU.NE BESEDE – Balkan; eneolitik; antropomorfna posoda; rentgenska ra.unalni.ka tomografija; tehnolo.ka analiza; eksperimentalna arheologija DOI> 10.4312\dp.49.3 Between object and subject> multiple approaches to a prehistoric human-shape pot from Romania Introduction The emergence of fired clay human-shaped pots is undoubtedly a technological breakthrough and a clear evolution of the artistic conception of human communities and that of the social development of humankind. This type of object is generally consid­ered part of the spectrum of figural art in prehisto­ric communities (Schwarzberg, Becker 2017). It re­presents a particular category of vessels, undoubt­edly different from ‘everyday pottery’ (Opris et al. 2017). Currently, it is evident that the meaning of these anthropomorphic vessels, alongside all human fig­urines documented in the Neolithic and Eneolithic (Chalcolithic or Copper Age) in different parts of the world is related to corporeal identities, and the way that people from the past understood existence and perceived the human body (Bailey 2013; 2015). The old assumptions regarding their religious or mytho­logical meanings (e.g., cult items, representations of divinities, Mother-Goddesses, etc.) are no longer in use, as proved by critical approaches postulated over time (e.g., Meskell 1995; Biehl 2006; Bailey 2013; 2015; 2017). The current paper will explore the technological background of an emblematic human-shaped pot for the prehistory of the Balkans, known as ‘The God­dess of Sultana’, in order to trace its meanings and how it actually was used by people in the past. Our approach on this particular human-shaped pot from Sultana-Malu Rosu will include using archae­ological data in correlation with techno-functional analysis and use-wear analysis, along with experi­mental archaeology and X-ray Computed Tomogra­phy (XCT). The main goal is to decipher the man­ufacturing process of such vessels, the possible way of using them, and their implied meanings. The use of multiple approaches is very effective, as demon­strated by our previously published studies (Ignat et al. 2017; 2018; 2019; Manea et al. 2019). Archaeological background The tell settlement from Sultana-Malu Rosu is located in southeastern Romania (N 44°15’41.5853”, E26°52’15.3378”), near the Danube River and the border with Bulgaria (Fig. 1.a). Archaeologists have known the site since 1923, when the first excava­tions began. The tell settlement is located on the high terrace of the Mostistea River, and it was used Fig. 1. a location of the tell settlement from Sultana-Malu Ros u in the area of the Kodjadermen-Gumel­nit a-Karanovo VI civilization; b the research algorithm used in our paper. Vasile Opris¸, Bogdan Manea, Mircea Lechintan, Roxana Bugoi, Florin Constantin, Theodor Ignat, and Catalin Lazar for more than 500 years by the Kodjadermen-Gu­melnita-Karanovo VI (KGK VI) communities in the second half of the 5th millennium BC (Lazar et al. 2016). In 1965 a human-shaped vessel known as ’The God­dess of Sultana’ (Fig. 2.a) was discovered by chance in the area of the tell settlement (Marinescu-Bîlcu, Ionescu 1967). The name given to this artefact re­flects the Gimbutasian interpretation of the archae­ologist who discovered this vessel and attributed a meaning that is currently considered out of date, as suggested by various criticisms formulated by dif­ferent scholars (Bailey 2013; 2015; 2017). Unfortunately, there are not many details about the archaeological context of this discovery. Most pro­bably, as we previously noted (Opris et al. 2017), the secondary burning traces observed on the exte­rior of this vessel represent strong indicators that it was found in a burned house. Unfortunately, any other data regarding the association of this human-shaped vessel with other features or artefacts are missing. However, from a typological point of view, it fits well in the humanized pottery series characte­ristic of KGK VI communities (Voinea 2005). The absence of any data regarding the stratigraphic con­text made impossible its assignment to a specific phase (e.g., A1, A2, or B1) of that civilization, but based on available radiocarbon data, we could set the possible timeframe between 4546–3973 cal BC (2. calibrated) (Lazar et al. 2016; 2018). This particular human-shaped vessel is internation­ally renowned and has been included in numerous catalogues of collections and exhibitions (Marines-cu-Bîlcu, Ionescu 1967; Wielen-van Ommeren et al. 2008; Anthony, Chi 2010; Lazar 2015) but also in synthetic works dedicated to Neolithic and Chalco­lithic anthropomorphic representations (Andreescu 2002; Ignat, Opris 2015; Opris et al. 2017). Currently, the pot is part of the National History Mu­seum of Romania collection in Bucharest, Romania (inv. nr. 102236). Materials and methods The multiple approaches used on the human-shaped pot from Sultana-Malu Rosu were based on the re­search algorithm illustrated in Figure 1.b. Generally, this kind of interdisciplinary investigation has the potential to reveal different relevant infor­mation regarding the technological background of the artefacts that, in correlation with other data types (e.g., archaeological context, functional and use-wear traces, etc.), could offer an integrative in­terpretation about past human’s material culture and its multiple meanings. The archaeological investigation of the anthropomor­phic pot was made by following the recommenda­tions regarding ceramic analysis available in various studies published over time (Rye 1981; Rice 1987; Skibo et al. 1989; Skibo 2013; Orton, Hughes 2013; Hunt 2016), and considered a technological-typolo­gical study, but also the examination of use-wear traces (Skibo 2013). This approach involved measu­ring the vessel’s dimensions (length, height, diam­eter, etc.), weight, and volume (Tab. 1). A binocular magnifier was used for use-wear investigation, along with photos taken with a camera and a macro lens. Prehistoric archaeological discoveries represent a significant challenge compared to other chronolog­ical periods (e.g., antiquity, medieval age, etc.) due to the lack of complementary sources of information (e.g., written sources, oral information, inscriptions, Fig 2. The human-shaped vessel known as ‘The Goddess of Sultana’: a the original (left – frontal view; right – lateral view); b the experimental replica (left – frontal view; right – lateral view). Between object and subject> multiple approaches to a prehistoric human-shape pot from Romania etc.). Under these circumstances, the data provided by archaeological investigations remain the only ones that can help us understand these prehistoric communities and their material creations. Unfortu­nately, these are also limited, far from encompass­ing all the information needed to fully understand the behaviour of past individuals and their decision-making regarding different aspects of daily life (e.g., food procurement, production of material goods, raw material exploitation, basic choices, etc.). Expe­rimental archaeology aims to verify the techniques, procedures, and processes involved in obtaining cer­tain objects or structures and then assess the theo­ries and hypotheses based on archaeological data, thus facilitating the possibility of providing new contributions to knowledge and understanding of the past (Lazar 2015). For the experimental replication of the anthropo­morphic pot from Sultana-Malu Rosu, we applied the methods and protocols developed by our team and previously published (Ignat et al. 2017; 2018; 2019). Generally, this kind of approach could offer us essen­tial data about the invisible elements of the past, re­garding the technological process (e.g., chaîne opé­ratoire employed in making these prehistoric clay items), which are not identified in the archaeologi­cal excavation (Lazar 2015). Therefore, the primary aim was to verify the manufacturing process, with all technological segments (e.g., gathering raw ma­terials, paste preparation, the drying and firing me­thods) involved. When it comes to XCT investigations, during the last few decades this kind of imaging method has been increasingly applied in archaeological research (Kahl, Ramminger 2012; Thér 2016; Kozatsas et al. 2018; Ross et al. 2018; Park et al. 2019). This type of ana­lysis can provide details about the internal structure of the artefacts that cannot be otherwise disclosed but through destructive investigations. Naturally, the more that is known about the inner structure of an object then the more speculations about the way it was manufactured or about the roles it might have played can be made. In particular, the application of XCT for the study of prehistoric clay artefacts proved to be an excellent non-invasive approach, being especially relevant to investigate the internal structure and/or hidden details of unbroken/intact archaeological objects. The XCT scans reported in this paper were per­formed with a Nikon XT H 225 device that contains a micro-focus X-ray source (225kV maximum volt­age, 1mA maximum current, focal spot size: 3µm below 7W and up to 225µm at 225W) that provides a conical X-ray beam, a Varian 2520 flat panel de­tector (127µm pixel size; 1900x1516 active pixels), and an accurate 5 axes positioning system, that al­lows the movements of the sample in the X, Y, and Z direction, as well as its rotation with 360° and tilt­ing by ±30°. The average voxel size is roughly 1.5x 105µm3. This equipment allows the inspection of re­latively large volume objects with high image reso­lution and an ultrafast CT reconstruction. The X-ray tube working parameters were optimized for the tomography of clay artefacts, namely a tube voltage of 100kV and a current intensity of 45µA. The entire volume of the objects was scanned in 360 steps, re­presenting the full rotation of the object around its central axis. The acquisition time for the full tomo­graphic scan was roughly 6 minutes, while the image reconstruction performed with VG Studio Max 3.0 software took several hours. The XCT image analy­sis was carried out to inspect the reconstructed ima­ges visually. This approach was followed because VGStudio Max 3.0 software is a general-purpose tool for 3D reconstruction. The analysed voids result from a visual inspection of the slices, and therefore not all the identified voids are perpendicularly ori­ented concerning the orthoview planes. The interpretation of the XCT images was made ac­cording to the methodology developed in the latest studies that involved the analysis of archaeological clay artefacts by XCT-images (Kahl, Ramminger 2012; Green et al. 2017; Ignat et al. 2017; 2018; Ko­zatsas et al. 2018; Manea et al. 2019; Park et al. 2019). Results Archaeological data From the preservation point of view, the human-shaped vessel from Sultana was discovered in a complete state, without visible cracks, with only an ear missing (the right one, now restored with plas­ter) and some secondary burning traces on the exte­rior surface (Opris et al. 2017). This particular human-shaped pot represents a per­son in a standing position. The bottom is dispropor­tionate compared to the top. The thighs, buttocks, pelvis, and hips regions are highlighted, showing clear elements of steatopygia. The upper limbs are stylized, placed very high, and stuck to the body. The elbows are bent, the right arm resting on the left, which in turn has placed its fingers under the Vasile Opris¸, Bogdan Manea, Mircea Lechintan, Roxana Bugoi, Florin Constantin, Theodor Ignat, and Catalin Lazar chin. The facial area is well represented. An arched projection renders the eyes. The nose is outlined, and the mouth is shown by a triangular notch, un­derlined with holes. Two large ears are added on the sides, pierced with four holes. On top of the figu­rine, the round neck of the vessel is attached, with a threshold for supporting a lid (missing). The ves­sel was fully painted with white decoration (e.g., spi­rals, circles, stripes, palmettes, and a triangle for the genital area). The attitude of this character, repre­sented as a standing human with one hand brought to the mouth, seems a meditative one, which re­minds us of the ’Thinker of Cernavoda’ (Marinescu-Bîlcu, Ionescu 1967). However, according to a new interpretation, it can represent a human in a ‘dead position’ based on gestures and wear-trace analysis (Opris et al. 2017). If we consider the anatomical elements presented, the humanized pot can be considered realistic only to a small extent because it does not reflect a com­plete representation of the human body. Certainly, the general shape of the pot suggests a human in a standing position with most anatomical features pre­sented (e.g., head, body, hands, legs). Some of them are designed very schematically (e.g., hands) to a scale that does not correspond to a natural body and is disproportionate to the other anatomical elements. The other ones (e.g., head, legs) are represented on a proportional scale and are much more realistic. Re­garding the facial attributes of the vessel from Sulta­na-Malu Rosu, it displays sufficient elements, high­lighting the most relevant facial physiognomy (eyes, nose, mouth, eyebrows, ears), placed anatomically correctly. The eyes and mouth are depicted as in­cised lines, and the hands, nose and ears are em­bossed. There are four holes in the left ear and ano­ther ten under the mouth, possibly signs of some form of body piercing (Opris et al. 2017). Although it was previously discussed that the repre­sented person would be a woman (Marinescu-Bîlcu, Ionescu 1967), we have already shown that the piece does not present sufficient anatomical elements to assign it to this gender because some relevant body parts for this category (e.g., breasts, vulva) are miss­ing (Opris et al. 2017). Nevertheless, the analysed vessel presents a series of features that make it a true masterpiece of prehisto­ric art. Techno-typological analysis Strong evidence about the technique used to make this vessel is missing since the inner surface was smoothed and the external surface was polished. The macroscopic analysis of the surfaces allowed us to observe that the paste was tempered with grog and contained natural non-plastics like rare, round­ed and sub-rounded, calcareous inclusions (up to 5mm) and fine white mica (Opris et al. 2017). This paste recipe is widespread for the Eneolithic pottery discovered within the Sultana-Malu Rosu tell settle­ment (Ignat et al. 2013). Moreover, since the vessel Fig. 3. The use-wear analysis: the abrasion traces on buttocks (a-b) and heels (c-d) areas of the human-shaped pot. Without scale (after Opris¸ et al. 2017). Between object and subject> multiple approaches to a prehistoric human-shape pot from Romania is intact, unbroken, and represents an artefact cat-alogued in the collection of national heritage, it is impossible to take samples for a different kind of archaeometrical analysis. The non-destructive XCT scans performed on this item thus offered supple­mentary data (see the dedicated section below). All external surfaces were painted with specific mo­tifs in white dye before the initial firing, but the se­condary firing partially damaged it. There is a hypo­thesis that the painted motifs represent some form of tattoo art (Marinescu-Bîlcu, Ionescu 1967). From a typological perspective, the investigated pot falls into human-shaped vessels that represent a standing person. The item is similar to other huma­nized pots discovered within the tell settlement from Sultana-Malu Rosu (Opris et al. 2017) or other con­temporary sites belonging to KGK VI communities (Andreescu 2002; Voinea 2005). Functional and use-wear analysis One of the main functional features is represented by the storage capacity of the recipient, in our case 2 litres, an element that places it in the type of me­dium storage vessel. According to the morphometric analysis and experimental replication, it can be used very well as a container, especially for liquids or ’flowing solids’ (like cereal grains) (Opris et al. 2017). The use-wear analysis (Schiffer, Skibo 1989; Skibo 2013) made on this humanised pot indicates that it was used repeatedly, this being proved by the pres­ence of some abrasion traces on different sides of the pot. Interestingly, even though we are dealing with a vessel that, according to its shape, should be used in a vertical position, most abrasion traces are not on the base (soles) but mainly on the back of the heels and buttocks (Fig. 3). This fact demonstrates that this vessel was used most of the time on its back position. Moreover, the lack of some specific marks on the soles could be evidence that while complete, the vessel was not moved (Opris et al. 2017). The experimental replication In order to identify the invisible elements of the past technological process and follow the features of the humanized vessel from Sultana-Malu Rosu, we tried to make an experimental replica (Fig. 2.b) using a local source of clay and prehistoric techno­logies. The experimental replica was made prior to the XCT analysis of the original pot, using only available macroscopic data and the potter’s experi­ence and intuition. The previous archaeometrical analyses performed on both Eneolithic pottery from the tell settlement and local clay sources around the site (Ignat et al. 2019) indicated a local source for the clay used for making pottery. The source that had the most com­mon features with the pottery from the site is locat­ed on the shore of the Mostistea Lake, about 300 meters north-west of the tell settlement. This source was denoted as Source nr. 8. From there, we col­lected two types of clay: one brown with sparse (<5%) carbonate concretions and one greyish-white in colour also with sparse (<5%) carbonate concre­tions in composition (Ignat et al. 2019). The paste recipe used for modelling the experimen­tal human-shaped pot was a mix made out of grey­ish-white clay (75%) and grog (25%, with granules <5mm). The vessel’s shape was modelled by hand by an archaeologist with limited experience in pottery making. The forming sequence was performed out­doors, in a shaded place during a day with moder­ate wind. According to the primary forming typol­ogy developed by Owen Rye (1981), coiling com­bined with pinching were the methods used. Thus, the bases of the feet were the first to be modelled, the process performed by hand and with the help of a bone spatula used for scraping. Afterwards, the whole body was built by using small coils of clay. After every two or three coils were added, the out­er part of the vessel in that area was scraped with a bone spatula and then smoothed with a dried reed stem and water. The thickness of the walls was main­tained at about 0.5cm. The anthropomorphic featu­res such as arms, mouth, nose, eyes, eyebrows and ears were added and modelled right after the ves­sel was shaped. The whole vessel was built in a sin­gle sequence giving that the lower constructed parts had been gradually drying during the forming pro­cess that took six hours. The final height of the ves­sel was 37cm. After one day of drying, a crack ap­peared in the area where the legs join, visible from both the outside and inside of the pot. The finishing method used was polishing the exter­nal surface with a fine river pebble after the vessel was dried indoors for two days. This process lasted one hour and a half (90min). A wood stick with human hair bound at an end was used as a brush to decorate the vessel. The paint was made of crushed carbonate concretions (c. 75%), collected from the shore of Mostistea Lake, mixed Vasile Opris¸, Bogdan Manea, Mircea Lechintan, Roxana Bugoi, Florin Constantin, Theodor Ignat, and Catalin Lazar with egg white (c. 25%). Paint­ing the decoration took al­most three hours (170 minu­tes). However, the process was not very complicated, in­volving only some basic artis­tic skills and an experimental brush to replicate the painted motifs documented on the ori­ginal item. After one year, the vessel was fired in a big kiln with two chambers that replicated an Eneolithic pottery kiln disco­vered in Ukraine (Tencariu 2015). The fire was powered with 85kg of dry wood, and the maximum temperature obtained was 786°C in the centre of the kiln after three hours of the firing. The maximum temperature measured on the God­dess replica was 623°C. The firing temperatures re­corded on the experimental human-shaped vessel during the combustion process are shown in Figure 4. After three hours of firing, the access door in the back of the kiln was opened, and no more wood was added. The grill under the vessel was crushed by the end of the firing, but the replica of ‘The Goddess of Sultana’ remained unbroken. Imaging investigations The XCT approach involved scanning the original archaeological object, as well as the experimental replica. In the latter case, for correct and complete control of the recorded data, following the objective to trace the technological elements related to the production of the humanized vessels, we made sev­eral XCT-scans both on the object after drying and also after its firing. In this way, we tried to record how the object transformed during the two major processes involved in making prehistoric ceramic vessels. Moreover, the imaging investigations per­formed on the experimental replica offer the great advantage of analysing an object made under con-trolled conditions, for which all the production se­quences and quantities of raw materials used have been recorded, thus allowing a reasonable interpre­tation of the resulting imaging data and their corre­lation with the structural and composite data of the analysed item. Moreover, this type of information becomes an accurate reference index that can be used in the case of archaeological artefacts for which this kind of data is missing. Due to the large dimensions of both the original pot and experimental replica compared to the measure­ment capacity of the X-ray detector, the imaging in­vestigations were made in three stages for every pot (lower, middle and upper parts, applied on original pot, dried replica and fired replica, respectively). Generally, considering the potential of the imaging investigations to answer different questions regard­ing pottery technology (Berg 2008; Kahl, Rammin­ger 2012; Ignat et al. 2017; 2018; Kozatsas et al. 2018; Manea et al. 2019), our approaches aimed to identify the following aspects: (i) characteristics of ceramic paste; (ii) primary forming techniques; (iii) joins of the anatomical elements on the pot; (iv) sur­face finishing; (v) cracks or repairs. Characteristics of the ceramic Goddess of Sultana pot Weight Capacity Max. pot Height (g) (l) diameter (cm) (cm) Original 1873 2.0 19.8 32.3 Replica (after modelling) 3335 n\a 23.6 37.0 Replica (after drying) 2790 n\a 22.5 35.2 Replica (after firing) 2540 3.3 22.0 34.5 paste The most visible non-plastics in the clay recipe of the original pot and experimental replica were the car­bonate concretions inclusions (Fig. 5). Their density, shape, size and di- Tab. 1. The size and weight of the experimental human-shaped ves-stribution (Tab. 2) indicate their pre­sel at different stages of production compared to the original pot. sence in the clay matrix as natural in­ Between object and subject> multiple approaches to a prehistoric human-shape pot from Romania Characteristics of ceramic paste Original human-shaped vessel Experimental replica (after drying process) Experimental replica (after firing process) Natural inclusions Carbonate concretions> . whitish or light grey in colour . rare density (1–2%) . size 1–5mm . high sphericity . well-rounded, rounded or sub-rounded Organic matter> . black in colour (voids) . rare density (.1%) . elongated, ovoid or spherical Carbonate concretions> . whitish or light grey in colour . sparse density (2–5%) . size 1–10mm . high and low sphericity . sub-rounded, sub-angular or angular Carbonate concretions> . whitish or light grey in colour . sparse density (2–5%) . size 1–10mm . high and low sphericity . sub-rounded, sub- angular or angular Tempers Grog> . hardly noticeable . light grey or ash grey in colour . undeterminable density . dimensions 1–2mm . low sphericity . angular Grog> . hardly noticeable . light grey or ash grey in colour . undeterminable density . dimensions 1–2mm . low sphericity . angular Grog> . hardly noticeable . light grey or ash grey in colour . undeterminable density . dimensions 1–2mm . low sphericity . angular Clay matrix . no particularities could be distinguished . no particularities could be distinguished . no particularities could be distinguished Voids . black in colour . sparse density (3–5%) . size 1–15mm . irregular, elongated, ovoid or spherical shapes . vertical, diagonal or horizontal orientation . black in colour . rare density (1–2%) . size 1–15mm . irregular, elongated or ovoid shapes . vertical or mostly random orientation . black in colour . rare density (1–2%) . size 1–15mm . irregular, elongated or ovoid shapes . vertical or mostly random orientation Tab. 2. Characteristics of ceramic paste resulted from the X-ray CT performed both on the original hu­man-shaped vessel and experimental replica (dry and fired). clusions specific for the clay sources. There are slight differences in density, roundness, and shape be­tween the original pot and the experimental replica. The high sphericity and the rounded shapes of the carbonate concretions show that the clay source used for the original pot was a redeposited sediment gradually washed and eroded by the alluvial waters. The clay used for the experimental replica was col­lected from the base of the loess deposit near the site (from a depth of about 4m). The movement of the carbonate concretions from this layer was slow­er over time, and thus their roundness and shape are of both high and low sphericity, and many of them have sub-angular and angular shapes. Never­theless, the presence and characteristics of the car­bonate concretions in the original pot point out to a local clay source exploited in the vicinity of the Sul-tana-Malu Rosu tell settlement. The presence of grog in the original pot was observ­ed when both surfaces were analysed in detail with the naked eye and the digital reconstruction of the scraped interior areas (Fig. 7.A1). As was already stat­ed, the clay recipe used for the experimental repli­ca contained 25% grog. Knowing this information, it was interesting that grog was hardly noticeable as non-plastic inclusions in all the tangential sections of the analysed pots (Fig. 5), even though the digi­tal images were processed in various light intensities in order to highlight the non-plastics from the clay matrix. The difficulty of identifying grog inclusions in pottery through XCT analysis has also been re­ported by Wolf-Achim Kahl and Britta Ramminger (2012). According to them, this is a consequence of the similarity between the attenuation coefficients of the grog and fired clay matrix. In the case of the tangential sections of the unfired replica, the obser­vations were identical (Fig. 5.B1-3) and showed no noticeable differences when unfired and fired clay was analysed using XCT images. The fine clay matrix had no different particularities in the radiographic images of the analysed vessels. The voids were observed in radial sections (Fig. 6) and are marked by black patches inside the vessel’s walls. Their sources can differ depending on the tech­nological variation in the production chain (Kahl, Ramminger 2012). The ones in the original pot have a sparse density and various sizes and shapes (Tab. Vasile Opris¸, Bogdan Manea, Mircea Lechintan, Roxana Bugoi, Florin Constantin, Theodor Ignat, and Catalin Lazar 2). The ovoid and spherical ones (Fig. 6.A1-2) could have been organic grains (possible seeds) that van­ished during the firing process. Some of the elon­gated voids from the feet walls seem to be formed due to the disappearance of organic matter during combustion (Fig. 6.A3). The sparse frequency shows that the organic inclusions were natural or acciden­tal in the clay and not intentionally added as temper. Other elongated and irregular voids were formed during the modelling of the pot and will be discussed in the section assigned to primary forming techni­ques. The void network in the experimental replica remained virtually the same after the firing process (Tab. 2; Fig. 6.B1-3,C1-3). Firstly, this situation is due to the absence of orga­nic inclusions in the clay used for the experimental replica. Secondly, this is a solid clue that no signifi­cant changes in the clay fabrics occurred during the firing. Consequently, the primary void network in the experimental replica was formed during the mo-delling and drying process. Primary forming techniques (Tab.3) The identification of primary forming techniques (Rye 1981; Rice 1987; Thér 2016) has been mainly based on a visual analysis of the inclusions and void orientations observed in the radial sections (Koza­tsas et al. 2018). The tangential and horizontal sec­tions were also analysed, but their examination brought a minor contribution to the final interpreta­tion of the forming techniques. The observations were made on three different parts of the vessels (lower, middle and upper) according to the XCT-ima­ges obtained for each part (Fig. 6). The elongated and irregular voids in the core walls of the lower part of the original pot (the feet) predo­minantly have a vertical orientation or more rarely a diagonal one (Fig. 6.A3). Their morphology and orientation can be recognized as traces left by coil­ing combined with sequential pinching, a technique that uses coils to build the walls followed by pinch­ing for modelling the shape and to control the thick­ness of the walls (Thér 2016). Both horizontal and vertical voids are detectable in the middle part of the original pot (Fig. 6.A2). The horizontal ones attest to the coiling technique, while the vertical ones indicate pinching after adding the coils. A flattened patch was used to reinforce the back of the vessel (Fig. 6.A2, down left). In the case of the experimental replica, the coiling technique is more noticeable in the XCT-images of the middle part, considering the presence of horizontal cracks and voids. The upper part of the original pot has voids that are relatively vertically oriented (Fig. 6.A3), similar to those identified in the lower part of the vessel. The most probable technique used for the primary form­ing of this part of the vessel was coiling and pinch- Fig. 5. CT-images of tangential sections of the original pot (A1-3), dried experimental replica (B1-3) and fired experimental replica (C1-3). Between object and subject> multiple approaches to a prehistoric human-shape pot from Romania Fig. 6. CT-images of radial sections of the original pot (A1-3), dried experimental replica (B1-3) and fired experimental replica (C1-3). ing. The coiling used to build the upper part of the experimental replica was detectable only in specific areas of the XCT-images. The application of the rim as a separate coil is the most visible, while the rarely encountered horizontal voids can be interpreted as traces left by the use of the coiling technique. Joins of the anatomical elements on the pot (Tab. 3) The embossed anatomical elements on the original pot (hands, nose, ears) could have been made in three ways: (a) by dragging clay from the vessel’s walls, (b) separately modelled and then attached to the vessel or (c) gradually formed by attaching small pieces of clay modelled in place. Right after the ves­sel was built, the anatomical elements of the exper­imental replica were attached as pieces of clay that were subsequently modelled into the desired shape. Prior to joining the anatomical elements, the vessel’s surface was prepared by scraping and moistening with water. The radial section of the upper part of the original pot (Fig. 6.A3) shows elongated voids in the joining areas of the hands and ears, indica­tive of a wet join performed while both the vessel’s walls and anatomical elements were wet and moist (Kozatsas et al. 2018). In the radial section of the experimental replica (Fig. 6.B3,C3), the same pat­tern is visible for the joining of the ears, but there are no noticeable traces to indicate the joining of the hands. The joining of the nose was not visible in the XCT-images, neither for the original pot nor the experimental replica. Surface finishing (Tab. 3) The XCT-images for the internal surface of the origi­nal pot (Fig. 7.A1-3) showed multiple traces of scrap­ing and smoothing in vertical and diagonal paral­leled stripes that can be interpreted as traces left by the potter’s hand. In contrast, the external surface was well polished (no tool traces left) and then de­corated by painting (Fig. 8.A1-3). The interior of the experimental replica was constantly scraped with a bone spatula and then smoothed with a reed stem during the vessel’s building process. Conse­quently, the traces left and identified on the XCT-images consist of small lines of horizontal or diag­onal orientation (Fig. 7.B1-3,C1-3). Inside the pot, traces of fingers can only be spotted in the middle area (Fig. 7.B2,C2). The external surface of the rep­lica vessel was well polished using a river pebble, and the traces left are not detectable in the XCT-images (Fig. 8.B1-3). Vasile Opris¸, Bogdan Manea, Mircea Lechintan, Roxana Bugoi, Florin Constantin, Theodor Ignat, and Catalin Lazar Repairs or cracks (Tab.3) The original pot has no visible cracks when exam­ined by the naked eye. However, the investigation of the images obtained by XCT revealed several cracks that appear mainly in the joining area of the feet but also on the middle part of the pot (Figs. 6. A2-3; 7.A2). The first ones are related to the pres­sure exerted on the joining area of the feet while the upper part was built, and the standing pot was slight­ly moved during the work. The same cracks appear­ed on the experimental replica during the building process (Figs. 6.B2,C2; 7.B2,C2). The horizontal crack from the middle part (Fig. 7.A2) of the original pot was most probably formed due to a deficiency in joining the coils used for the primary forming. Other diagonal cracks in the same area could result from the secondary firing of the pot. The horizontal cracks in the experimental replica also appear in the mid­dle part (joining of the coils) and at the joining of the rim with the vessel’s body (Fig. 6.B1,C1). Discussions The multiple approaches used to examine the inside and the outside of a human-shaped pot from Sulta­na-Malu Rosu allowed us to obtain a specific data set regarding the characteristics of the humanized pot­tery produced by these prehistoric communities. Our goal was to focus on the manufacturing process that past potters followed to make such humanized vessels. This approach allows us to record some ex­citing data and identify the invisible elements about the chaîne opératoire undocumented in the archa­eological record. Firstly, it should be noted that de­spite the complex appearance of the vessels with hu­man morphological attributes, the manufacturing process would not be a great challenge for a potter with regular experience in the field. As we previ­ously showed, various stages of collection and pre­paration of raw materials alongside the modelling process fit well in the pottery production standards manufactured by hand. Moreover, the anatomical components are easy to achieve and do not require special skills. The painting decoration process is also included in the production standards previously ob­served (Ignat et al. 2012; 2013; 2017; 2018; Ignat, Opris 2015; Opris et al. 2017). The time required is imposed not by the shape of the ceramic vessel but by the complexity of the painted motifs, for which some artistic skills are needed. However, beyond these general observations there was a series of changes in how the experimental container behaved in different stages of production. Thus, according to the data presented in Table 1, we observe that the object’s size and weight changed in Forming features Original human-shaped vessel Experimental replica (after drying process) Experimental replica (after firing process) Primary forming . coiling + pinching (|), . no particularities could be . no particularities could be techniques flattened patches in the lower part of the feet . horizontal coiling and flattened patches for the middle part of the pot . coiling + pinching (|) for the upper part distinguished for the lower part (feet) . hardly noticeable horizontal coiling for the middle part . horizontal coiling for the rim distinguished for the lower part (feet) . hardly noticeable horizontal coiling for the middle part . horizontal coiling for the rim Joins of the anato­mical elements on the pot . ears and hands added after the primary forming of the pot . ears clearly added after the primary forming of the pot . ears clearly added after the primary forming of the pot Surface finishing Internal> . vertical scraping with the fingers on the feet . diagonal and vertical scraping and smoothing with the fingers in the middle and upper part External> . polished with a hard tool Internal> . horizontal and diagonal scraping and smoothing with a hard tool . fingers prints in the middle part External> . polished with a hard tool Internal> . horizontal and diagonal scraping and smoothing with a hard tool . fingers prints in the middle part External> . polished with a hard tool Cracks or repairs . crack in the feet joining area . horizontal crack in the back of the middle part . no traces of repairs . multiple cracks in the feet joining area . no traces of repairs . multiple cracks in the feet joining area . no traces of repairs Tab. 3. Forming features resulted from the X-ray CT performed both on the original human-shaped ves­sel and experimental replica (dry and fired). Between object and subject> multiple approaches to a prehistoric human-shape pot from Romania the different stages of the manufacturing process. All the recorded changes are related to the chemical-physical characteristics of the paste used for the pot making process, and its evolution during the drying and firing stages. More valuable data regarding the compositional and structural aspects of the original pot were provided by the XCT investigation and comparison with the experimental replica. The XCT-images have allowed for the first time to view inside the walls of the ori­ginal pot and consequently helped us characterize its ceramic paste. The presence of natural non-plas­tic inclusions such as rare spherical carbonate con­cretions and sparse organic matter, followed by the comparison with the inclusions from the experimen­tal replica that was made with clay collected near the site, are indicative of a raw clay collected from a local deposit of alluvial nature. The hypothesis that the original pot was manufactured locally is strengthened by the presence of grog used as tem­per, a long-term tradition in making ceramic paste recipes observed for most of the Eneolithic pottery found within the tell settlement from Sultana-Malu Rosu. The shape, size and orientation of the voids in­dicate that the original pot was primarily formed by coiling followed by pinching, and in some cases, the flattened patches were used to reinforce the walls. However, some questions on this topic still remain, mainly because when these primary forming techni­ques (e.g., coiling and pinching) were applied on the experimental replica, their specific traces were only partially revealed by the XCT analysis. More clearly, results were obtained regarding the joining of the anatomical elements. The arms and ears of the ori­ginal pot were attached after the building of the ves­sel’s walls, while the nose, mouth and eyebrows did not bear any traces indicative of joining to the pot. The internal surface of the original pot was only scraped and smoothed by fingers. This little care for the internal surface made it unusable for liquid con­tents, even if its shape recommends it for this kind of use (Skibo 2013). In contrast, the external surface was well polished and decorated by painting. The XCT-images also revealed hidden cracks in the join­ing area of the feet and the middle part of the origi­nal pot. The high similarity with the cracks observ­ed in the experimental replica can be a solid argu­ment that both vessels were built using comparable steps and construction techniques. When it comes to the possible functional attributions that the analysed human-shaped pot could have ful­filled, we were able to identify some features that might suggest certain functions. However, a clear in­terpretation is difficult to achieve based only on the structural and typological particularities of the archa­eological artefact and the experimental replica. As mentioned earlier, certain elements (such as the vo­lume or the shape) may suggest a distinct function for this particular vessel, for example, its use as a re­cipient for short-term storage of liquids or for Fig. 7. 3D CT-images of the internal surfaces of the original pot (A1-3), dried experimental replica (B1-3) and fired experimental replica (C1-3). Vasile Opris¸, Bogdan Manea, Mircea Lechintan, Roxana Bugoi, Florin Constantin, Theodor Ignat, and Catalin Lazar gathering and transferring them to another container. However, other characteristics (such as the little care for the internal surface or the multiple cracks inside the walls) make it imprac­tical for liquid-related uses and indicate other functions. At the same time, its usage as a storage container for differ­ent types of ’flowing solids’ (Opris et al. 2017), even if it is a more viable option, at some point may also become questionable due to the wear traces that suggest the use of this vessel for gathering rather than storage. Therefore, while several features may suggest one or more functions, the presence of other elements may dis­mantle the previous assumptions, thus creating a vicious circle in interpreting the function of such a pot. In this con­text, for an interpretation as accurate as possible regarding the function of this particular human-shaped vessel, further experiments must be conducted by using several replicas of it under dif­ferent conditions (for example, using them for gathering different types of goods, transporting the vessels in dif­ferent ways while empty or/and full, etc.). This approach, in correlation with the functional and use-wear information provided by the current study, may ultimately lead to a better understand­ing of how this vessel was used in the distant past. Conclusion The present study revealed the potential of combin­ing multiple approaches in investigating a unique Eneolithic pot discovered at Sultana-Malu Rosu. Thus, with the help of archaeological data and experimen­tal archaeology, coupled with X-ray Computed To­mography, we have been able to identify multiple invisible aspects regarding the manufacturing pro­cess of the human-shaped pot known as ‘The God­dess from Sultana’. By carefully following and ana­lysing all the operational sequences (harvesting and preparation of raw materials, modelling and deco­ration of the pot, the drying and afterwards the fir­ing process) in terms of the technology, time and skills required for the different stages of production, we managed to shed some light upon the methods used to obtain this final product and on what the entire process implied for the potter. Furthermore, the XCT investigation facilitated a detailed exami-nation of the archaeological pot’s internal structure, as well as that of the experimental replica, thus of­fering the possibility of identifying technological ele­ments (characteristics of ceramic fabric, primary forming techniques, surface finishing, repairs or cracks) related to the chain of production. Moreover, the techno-functional analysis provided a small but interesting set of data, which allowed us to set up some new research goals for the near future to in­vestigate how this human-shaped vessel – or similar ones – was used in the past. Unfortunately, the correlation of all typological-sty­listic, experimental, XCT-scan and techno-functional data about this unique humanized pot suggests a controversial function that at the moment can be framed neither as special nor as ordinary. However, this particular vessel, ’The Goddess from Sultana’, in­directly reflects the ideology, beliefs, wishes, desires, visions about the world, humans, nature, daily life cycles and differing technological, organizational, and social approaches of past peoples. The tenden­cy of human communities to represent the familiar silhouette of the human body or give objects a hu­man (or quasi-human) form is a natural process of expressing the visual identity perceived by these Between object and subject> multiple approaches to a prehistoric human-shape pot from Romania people in the past. The act of adding typically hu­man anatomical elements (e.g., eyes, ears, hands, etc.) to inanimate objects is well documented in KGK VI communities. This seemingly technological act transforms inanimate objects into living objects, and thus active tools that are part of, live with and contribute to the daily setup of these people’s lives. Some particular elements found in this study (e.g., the use of the vessel in a horizontal position, not in a vertical one (which is the anatomical orientation of the figure), prove the complexity of the vision and imagination of prehistoric people, but also the high amplitude of the abstract dimension of the past minds that created and enlivened these artefacts. The way these abstract ideas are manifested through the material culture that these peoples created led (indirectly) to drawing inferences about the factors that governed the daily lives of those communities, and reflected their collective and individual identi­ties. Moreover, humans do not behave under the in­fluence of their senses alone but also through their individual and collective past experiences such as their upbringing correlated with the technological level of expression, beliefs, traditions, ideas, aims, fears, desires, symbols, and myths. These experien­ces contribute to each individual’s unique view of the world, and in this way human groups that live together tend to develop a shared view of the world, which in turn influences their group material culture (Henley et al. 2020). ACKNOWLEDGEMENTS This work was supported by a grant from the Mini­stry of Education and Research, CNCS – UEFISCDI, project number PN-III-P4-ID-PCE-2020-2369, within PNCDI III. We would also like to thank the anony­mous reviewers for their positive comments that helped us to improve the manuscript. AUTHORS’ CONTRIBUTIONS CL supervised the study. VO, BM and TI performed the technological analysis of the human-shaped pot. ML, FC, and RB performed the X-ray Computed To­mography scans. BM, VO, and ML analysed and/or helped to interpret the XCT-scan data. VO made the experimental replicas, and quantification of the data. VO, BM, and CL wrote the manuscript with input from all co-authors. All authors read and approved the fi­nal manuscript. . References Andreescu R. R. 2002. Plastica antropomorfa¢ gumelni­teana¢. Analiza¢ primara¢. Daim. Bucuresti. Anthony D. W., Chi J. Y. (eds.) 2010. The Lost World of Old Europe. The Danube Valley, 5000–3500 BC. Prince­ton University Press. New York. Bailey D. W. 2013. Figurines, Corporeality, and the Ori­gins of the Gendered Body. In D. Bolger (ed.), A Compa­nion to Gender Prehistory, 1st ed. Wiley-Blackwell. Ox­ford: 244–264. 2015. Figurines: Europe. In P. Whelehan, A. Bolin (eds.), The International Encyclopedia of Human Sexua­lity. Wiley-Blackwell. Oxford: 339–401. https://doi.org/10.1002/9781118896877.wbiehs155 2017. Southeast European Neolithic figurines: beyond context, interpretation, and meaning. In T. Insoll (ed.), The Oxford Handbook of Prehistoric Figurines. Ox­ford University Press. Oxford: 823–850. DOI: 10.1093/oxfordhb/9780199675616.013.040 Biehl P. F. 2006. Figurines in action: methods and theo­ries in figurine research. In R. Layton, S. Shennan, and P. Stone (eds.), Festschrift Peter Ucko: A Future for Archa­eology – the Past as the Present. UCL Press. London: 199–215. Green A. F., Hartley C. W., Doumani Dupuy P. N., and Chi-nander M. 2017. The digital radiography of archaeologi­cal pottery: Program and protocols for the analysis of pro­duction. Journal of Archaeological Science 78: 120– 133. https://doi.org/10.1016/J.JAS.2016.11.007 Henley T. B., Rossano M. J., and Kardas E. P. (eds.) 2020. Handbook of Cognitive Archaeology Psychology in Pre­history. New York. Routledge. https://doi.org/10.4324/9780429488818 Hunt A. M. W. (ed.) 2016. The Oxford Handbook of Ar­chaeological Ceramic Analysis. Oxford University Press. Oxford. https://doi.org/10.1093/oxfordhb/97801996815 32.001.0001 Vasile Opris¸, Bogdan Manea, Mircea Lechintan, Roxana Bugoi, Florin Constantin, Theodor Ignat, and Catalin Lazar Ignat T., Opris V., Voicu M., Andreescu R. R., and Lazar C. 2012. Ceramica din locuinta nr. 5 de la Sultana ‘Malu Ro­su’. Analiza¢primara¢(I). Buletinul Muzeului Judetean Te-leorman 4: 71–132. Ignat T., Opris V., and Lazar C. 2013. Ceramica din locuin­ta nr. 5 de la Sultana ‘Malu Rosu’. Analiza¢primara¢(II). Bu-letinul Muzeului Judetean Teleorman 5: 155–172. Ignat T., Opris V. 2015. Ceramica. In C. Lazar (ed.), Ra-diografia unei lumi dispa¢rute. Privind înapoi spre tre-cut: Sultana-Malu Rosu, o asezare preistorica¢de acum 6000 de ani. Muzeul National de Istorie a României. Bu-curesti: 47–49. DOI: 10.13140/RG.2.1.2557.3203/1 Ignat T., Constantin F., Bugoi R., Parnic V., and Lazar C. 2017. Investigatii imagistice asupra unor piese de lut eneo­litice. Buletinul Muzeului Judetean Teleorman 9: 65–73. Ignat T., Bugoi R., Constantin F., Parnic V., and Lazar C. 2018. Identifying the chaîne opératoire of prehistoric clay figurines using experimental archaeology and imaging me­thods. In A. Erickson, A. Bernstein, and M. Hamm (eds.), Proceedings of the 2017 International Conference on In Applications of Nuclear Techniques (CRETE17) Crete, Greece, June 11–17, 2017. International Journal of Mo­dern Physics: Conference Series Vol. 48: 1860107. https://doi.org/10.1142/S2010194518601072 Ignat T., Luca A., Dimofte D., Lazar C., Constantin F., and Bugoi R. 2019. Multidisciplinary study on prehistoric pot­tery from South East Romania. ArcheoSciences. Revue d’Archéométry 43(2): 165–185. https://doi.org/10.4000/archeosciences.6592 Kahl W. A., Rammiger B. 2012. Non-destructive fabric ana­lysis of prehistoric pottery using high-resolution X-ray mi-crotomography: a pilot study on the late Mesolithic to Neo­lithic site Hamburg-Boberg. Journal of Archaeological Science 39: 2206–2219. https://doi.org/10.1016/j.jas.2012.02.029 Kozatsas J., Kotsakis K., Sagris D., and David K. 2018. In­side out: Assessing pottery forming techniques with micro-CT scanning. An example from Middle Neolithic Thes­saly. Journal of Archaeological Science 100: 102–119. https://doi.org/10.1016/j.jas.2018.10.007 Lazar C. 2015. Proiectul Archaeodrom în contextul arhe­ologiei experimentale din România. Buletinul Muzeului Judetean Teleorman 7: 193–199. Lazar C., Margarit M., and Balasescu A. 2016. Dogs, jaws, and other stories: Two symbolic objects made of dog man­dibles from Southeastern Europe. Journal of Field Archa­eology 41(1): 101–117. https://doi.org/10.1080/00934690.2015.1114850 Lazar C., Margarit M., Radu V. 2018. Evidence of produ­ction and use of Lithoglyphus naticoides beads in the Ho­locene from Europe: the case of Sultana-Malu Rosu site (Romania). Quaternary International 472(A): 84–96. https://doi.org/10.1016/j.quaint.2017.10.033 Manea B., Lechintan M., Popescu G., +6 authors, and La­zar C. 2019. Looking beyond appearances: a multi-analy­tical approach on the prehistoric clay weights. Heritage Science 7: 88. https://doi.org/10.1186/s40494-019-0326-2 Marinescu-Bîlcu S, Ionescu B. 1967. Catalogul sculpturi­lor eneolitice din Muzeul Raionului Oltenita. Sibiu. Meskell L. 1995. Goddesses, Gimbutas and “new age” ar­chaeology. Antiquity 69: 74–86. https://doi.org/10.1017/S0003598X00064310 Opris V., Ignat T., and Lazar C. 2017. Human-shaped pot­tery from the tell settlement of Sultana-Malu Rosu. In H. Schwarzberg, V. Becker (eds.), Bodies of clay: on prehi­storic humanised pottery. Proceedings of the session at the 19th EAA Annual Meeting at Pilsen. 5th September 2013. Oxbow. Oxford: 191–212. https://doi.org/10.4000/acost.2004 Orton C., Hughes M. 2013. Pottery in Archaeology. 2nd ed. Cambridge University Press. Cambridge. https://doi.org/10.1017/CBO9780511920066 Park K. S., Milke R., Rybacki E., and Reinhold S. 2019. Ap­plication of Image Analysis for the Identification of Pre­historic Ceramic Production Technologies in the North Cau­casus (Russia, Bronze/Iron Age). Heritage 2: 23272342. https://doi.org/10.3390/heritage2030143 Rice P. M. 1987. Pottery analysis: a sourcebook. The Uni­versity of Chicago Press. Chicago. Rye O. 1981. Pottery technology. Principles and recon­struction. Manuals on archaeology 4. Taraxacum. Wa­shington DC. Ross J., Fowler K. D., Shai I., Haskel J., Greenfield J., and Maeir A. M. 2018. A scanning method for the identifica­tion of pottery forming techniques at the mesoscopic scale: A pilot study in the manufacture of Early Bronze Age III holemouth jars and platters from Tell es-Safi/ Gath. Journal of Archaeological Science: Reports 18: 551– 561. https://doi.org/10.1016/j.jasrep.2018.01.036 Schwarzberg H., Becker V. 2017. Preface. In H. Schwarz-berg, V. Becker (eds.), Bodies of clay: on prehistoric hu­manised pottery. Proceedings of the session at the 19th EAA Annual Meeting at Pilsen. 5th September 2013. Ox­bow. Oxford: vii. https://doi.org/10.4000/acost.2004 Between object and subject> multiple approaches to a prehistoric human-shape pot from Romania Skibo J., Schiffer M. B., and Kenneth C. R. 1989. Organic-Tempered Pottery: An Experimental Study. American Antiquity 54(1): 122–146. https://doi.org/10.2307/281335 Skibo J. 2013. Understanding pottery function. Springer. New York. https://doi.org/10.1007/978-1-4614-4199-1 Tencariu F. A. 2015. Instalatii de ardere a ceramicii în civilizatiile pre- si protoistorice de pe teritoriul Româ­niei. Editura Universita¢tii A.I. Cuza. Iasi. Thér R. 2016. Identification of pottery-forming techni­ques using quantitative analysis of the orientation of in­clusions and voids in thin sections. Archaeometry 58(2): 222–238. https://doi.org/10.1111/arcm.12166 Voinea V. 2005 Ceramica complexului cultural Gumel­nita – Karanovo VI. Fazele A1 si A2. Ex Ponto. Constanta. Wielen-van Ommeren F., van der Chamay J., Wullschleger M., Alamoreanu A., Popovici D., and Drasovean F. (eds.) 2008. Neolithic art in Romania. Catalogue for the ex­hibition held at the Historisches Museum Olten. Switzer­back to contents land. June 3-Oct. 5 2008. Arte-M. Naples. Documenta Praehistorica XLIX (2022) The Neolithic worked bone assemblage from Ulucak Höyük, Western Anatolia Jarrad W. Paul 1, Cos¸kun Sivil2, and Özlem Çevik2 1 Classics and Archaeology Department, School of Historical and Philosophical Studies, University of Melbourne, Parkville, AU< jpaul@trinity.unimelb.edu 2 Faculty of Letters, Department of Archaeology, Trakya University, Edirne, TR< coskunsivil@trakya.edu.tr< ozlemcevik@trakya.edu.tr ABSTRACT – In this investigation we detail the results of a systematic analysis of worked animal bone from Ulucak, one of the earliest Neolithic sites in western Anatolia. The collection exhibits a wide range of types, including points, needles, spatulas, bevelled tools, perforated objects, and other unique objects. A study of the raw material shows a preference for sheep and goat long bones, while large-sized animal rib bones were also utilized. Manufacturing techniques employed included split­ting, grooving, and abrasion, while a contextual analysis of the material underscores an intricate connection with other objects made from stone and clay. Items found in buildings relate to textile, leather, and ceramic production, while personal ornaments may have played a part in abandonment rituals. Examination of this assemblage is understood as a common set of regional tool types with some localized variations. KEY WORDS – Early Neolithic; Western Anatolia; animal bone artefacts Neolitski zbir ko[;enih izdelkov iz Ulucak Höyüka, zahodna Anatolija IZVLE.EK – V razpravi predstavljamo rezultate sistemati.ne analize obdelanih .ivalskih kosti iz Ulucaka, enega najzgodnej.ih neolitskih najdi.. v zahodni Anatoliji. Zbir predstavlja .iroko paleto orodij, konic, igel, lopatic, prirezanih orodij, prebodenih predmetov in drugih unikatnih izdelkov. Analiza surovin ka.e, da prevladujejo ov.je in kozje dolge kosti. Uporabljali so tudi rebra velikih .ivali. Tehnike izdelave vklju.ujejo razcepljanje, .lebljenje in bru.enje. Kontekstualna analiza je po­kazala na povezave z drugimi predmeti, narejenimi iz kamna in gline. Ko..eni predmeti, ki so bili najdeni v zgradbah, so povezani s tkanjem, izdelavo usnjenih izdelkov in lon.enine. Ko..en osebni nakit je bil morda uporabljen v obredih opu..anja bivali..a. Predmete razumemo kot regionalen zbir orodij, ki vklju.ujejo tudi lokalne razli.ice. KLJU.NE BESEDE – zgodnji neolitik; zahodna Anatolija; izdelki iz .ivalskih kosti Introduction The systematic study of worked animal bones at Neo-and spatial analysis (Samei, Alizadeh 2020) can add lithic and Chalcolithic sites in Anatolia has increased to our understanding of sedentary communities. in the last two decades (Russell 2016). No longer a However, some long-standing excavations from Ana-neglected sub-discipline, studies have shown how tolia are yet to undergo the first element of systema-new avenues of scientific testing (Bradfield et al. tic analysis: the establishment of a typology, raw ma­2019), use-wear analysis (Campana, Crabtree 2018) terial identification, and technological, contextual, DOI> 10.4312\dp.49.10 The Neolithic worked bone assemblage from Ulucak Höyük, Western Anatolia and comparative investigations of their collections to establish a foundation for these further studies. One of these sites is Ulucak, located 25km east of Iz­mir. Continuous archaeological excavations since 1995 have uncovered a vast amount of material cul­ture (Çevik, Erdogu 2020; Çevik 2019; Çilingirog­lu et al. 2004; 2012). Uninterrupted occupation at the site occurred between Level VI (6850/6830– 6500 cal BC), Level V (6500–6000 cal BC) and Level IV (6000–5700 cal BC), with habitation in Level III (5600–5460 cal BC) occurring after a brief cultural break. Later levels belong to the Early Bronze Age (Level II) and Middle Bronze Age (Level I), with evi­dence of Late Roman/Byzantine remains on the sur­face. The earliest occupation at Ulucak, Level VI, has been exposed in trenches L13 and partly L12 and K13. This period is known from two adjacent buildings (Buildings 42 and 43), which has a possible commu­nal function, lime plastered and red painted floors and open spaces with fire installations around the buildings (Fig. 1). Level V has five sub-phases (Va-e) and is exposed in trenches L13, L12 and K13. This level consists of rectangular single-roomed domestic buildings with either mud-slab or post-framed walls. After 6000 cal BC, in Level IV, distinctive changes oc­curred both in the construction technique and size of the buildings. Of the 10 sub-phases of Level IV, only IVb has been exposed in a wider area of about 1000m2. In Level IVb one- or two-roomed substan­tial domestic buildings with sun dried mudbrick walls on stone foundations were divided by narrow streets. Level IVc is known from a specialized cera­mic workshop with six rooms built of post-framed walls (Çevik 2016). The inventory of the workshop consists of several clay loaves, hematite lumps, un­finished vessels, bone tools and grinding stones pos­sibly used for powdering hematite, which indicates the whole sequence of pottery manufacturing (Fig. 2). Pottery has been attested starting from Level V on­wards, while the earliest occupation at Ulucak is de­void of pottery and any other clay objects. Ceramics of the earliest periods include cream, brown and grey burnished wares, with an increase in red-slip­ped burnished wares in Level IV (Çilingiroglu 2012; Çevik, Vuruskan 2020). Other items recovered at the site include stone tools (with obsidian mostly originating from Melos), figurines, stamp seals, spin­dle whorls and loom weights (Sevindik 2018). The faunal assemblage (Pilaar Birch et al. 2019; Çakir­lar 2012) consists mainly of domestic sheep and goat, with cattle and pig frequently recorded. Deer (most commonly fallow), small mammal (such as hare, fox, and tortoise), bird and fish remains are documented in lower numbers. Fig. 1. Architectural remains Levels VI, V, and IV. Fig. 2. Level IVc ceramic workshop. The worked animal bone collection constitutes a sizeable part of the small finds at the site. Analysis of this material has been limited in previously pub­lished accounts (Çilingiroglu A. et al. 2004.50; Çilin­giroglu Ç. 2012.21), with reports outlining common types such as awls, spatulas, and perforated items. Subsequent systematic analysis of the material has been conducted in more recent unpublished stud­ies (Paul 2016; Sivil 2017). An initial assessment of the worked animal bone was made by Jarrad W. Paul during the 2014 and 2015 excavation seasons. The objects analysed were reco­vered from the 1997–2015 excavation seasons. An emphasis during the initial assessment was on re­cording all items stored onsite, establishing a typo­logy, and identifying raw material. A study based on this initial recording of items, which number 268 from Levels VI–III, is included in a comparative ana­lysis of worked animal bone from sites in the north Aegean (Paul forthcoming). Coskun Sivil has fur­ther conducted a contextual analysis of the mater­ial, a vital step in placing the material within its set­ting. Sivil’s research was based solely on worked bone items from the Neolithic, Levels VI–IV. In his assessment 549 items were recorded, including those stored in the local museum. In this paper, Neolithic material will again be the focus of investigation from the uninterrupted Levels VI–IV, uniting the work of both Paul and Sivil to illustrate a complete picture of the Neolithic worked bone assemblage from Ulucak. The typology below is informed by Paul’s initial ana­lysis combined with Sivil’s typological categoriza­tion, with a description accompanying each type and subtype. Identification of manufacturing techniques was conducted on-site by Paul using both macro (viewing distinct markers on the object or tool, for instance colouration and breakage patterns) and micro (using a x3 magnifying eye-piece to detect striations left on the object or tool) techniques. The analysis of raw material was also conducted on-site by Paul, with the assistance of Evangelia Piskin. The contextual analysis is based on research conducted by Sivil. Discussion and interpretation in the present study is conducted by Paul, Sivil, and Çevik. Items are inclusive of those uncovered from 1995–2017. Typology A total of 554 worked animal bone items have been placed in Levels VI–IV. The typology created for this assessment was informed by previous research in the region, especially the work of Nerissa Russell (2016), Alexandra Legrand and Isabelle Sidera (2007) and Marcella Marinelli (1995). Objects have been sepa­ The Neolithic worked bone assemblage from Ulucak Höyük, Western Anatolia rated based on their surmised functional attributes (for example, tools used for pier­cing). Raw material was then used to fur­ther separate objects within types (referred here as sub-types) when necessary. The 554 items identified within this typology do not include any preforms, items too fragmentary to identify, or manufacturing waste as a by-product of worked animal bone manufacture found at the site. Al­though these items are found on site and mentioned later in this paper (see Con­text) they are not included in this assess­ment as the typology only includes items that have been positively identified. The typology includes six types and 23 sub­types (see Tab. 1). 1. Points The most common type in the assemblage are points (n=319). Points are character­ized by their pointed tips (Fig. 3). Their primary function was as a piercing tool. Type Subtype Number 1A. Metapodial points 115 1B. Ulna points 7 1C. Other long bone points 62 1. Point 1D. Oval-bodied points 35 1E. Square-bodied points 54 1F. Fragments 40 1G. Flat bone points 6 2A. Perforated needle 24 2. Needle 2B. Notched needle 2 3A. Flat spatula 92 3. Spatula 3B. Spatula-spoon 15 4A. Smoother 26 4. Bevelled tool 4B. Chisel 39 5A. Flat bone perforated objects 2 5. Perforated object 5B. Long bone perforated objects 6 6A. Comb 6 6B. Bipoint 2 6C. Bone handle 10 6D. Antler handle 5 6. Other 6E. Worked antler 3 6F. Fastener 1 6G. Hook 1 6H. Arrow\spearhead 1 Total 554 They are likely to have been used in textile Tab. 1. Ulucak Höyük animal worked bone and antler object typology from Levels VI–IV. manufacture, sewing animal hide, scrap­ ing ceramic surfaces and in basket weaving. They are separated into seven subtypes. 1A. Metapodial points These points are made from metapodial (metacar­pal and metatarsus) bones and have mostly polished surfaces. Manufactured by splitting along a whole bone, the base of these objects is often left intact. They are the most numerous subtype point (n=115) in the collection. 1B. Ulna points Points made from ulna bones are uncommon, with only seven examples recorded. Their natural form lends itself to use as a perforator, with an area to grip to the tool at its base. They would have been a suitable tool for pro­cessing soft material. 1C. Other long bone points These points are made from undetermin­ed long bones. The base of these tools is often rounded and smoothed, making further identification difficult. They are the second most frequent subtype (n=62). 1D. Oval-bodied points Points in this subtype have their base missing and are defined by the form of their shaft, in this case oval-shaped (n= 35). They are made of long bone frag­ments. 1E. Square-bodied points As with subtype 1D, square-bodied points have their base missing and are defined by the shape of their shaft (n=54). They are also made of long bone fragments. Jarrad W. Paul, Cos¸kun Sivil, and Özlem Çevik 1F. Fragments Points in this subtype were made hastily, most with asymmetrical shafts, with little polishing on their surface (n=40). The construction of these points included minor morphological changes to the nat­ural structure of the bone fragment. Based on their manufacturing techniques and less elaborated sha­pes, they may have met immediate needs. 1G. Flat bone points Flat bone points are made from rib bones. They are the least common subtype, with only six examples, and were used intensively, as indicated by their bur­nished surfaces. They may have been better suited to working on soft materials, such as hide and tex­tiles, due to their form. They share similar proper­ties with spatulas and may be in fact more akin to the pointed spatula type used for ceramic shaping (Ma¢rga¢rit 2017). 2. Needles Needles are made from both long and flat bones and include two subtypes: those that are perforated (2A) and those that are notched (2B) (Fig. 4). Perforated needles were drilled either on one side or both and are the preferred subtype, with 24 examples re­corded. They are mostly made from rib bones. Notched examples are less frequent, with only two objects noted. The notches for these objects appear at the base. Needles would have been used to com­bine materials together, with wider and flatter examples associated with basket weaving. 3. Spatulas Spatulas are also separated into two subtypes: those made from flat bones (3A) and those with a handle (3B) (Fig. 5). Those made from flat bones, rib bones in this case, are the second most frequent subtype in the collection (n=92). They would have been used for pottery moulding, stripping any excess materi­al. Less frequent are spatulas with a handle (n=15), also known as spatula-spoons (Paul, Erdogu 2017). They are made on long bones. All subtype 3B spat­ulas at Ulucak were located within domestic struc­tures and open spaces. Further use-wear and trace analysis needs to be conducted to determine the function of these tools. 4. Bevelled tools Tools in this type are sturdy and recovered mostly intact. They are grouped together here for their scraping function and are further separated into two subtypes. The first (4A) are smoothers made from tibia bones (n=26). They have a bevelled tip, un-worked base, and a hollow shaft. They are associat­ed with leather manufacture; in particular, the pre­paration of animal hide. The second subtype (4B) are chisels made from long bones (n=39). They are strong and robust tools, with a thick cortex and be-velled tip (Fig. 6). These tools are associated with wood working activities, ideal for carving and chip­ping. 5. Perforated bone objects This type includes any object of worked bone that has been perforated and not considered a needle. They are not common in the collection, with only eight examples (Fig. 7). They are separated into two subtypes based on their raw material: those made Fig. 5. Type 3. Spatulas. The Neolithic worked bone assemblage from Ulucak Höyük, Western Anatolia from flat bones (5A) and those made from long bones (5B). For those made from flat bones (n=2), they may be considered a type of pendant, but due to their fragmented condition it is difficult to assign function. Perforations are drilled from either side of the object. For the long bone examples, they may have been related to weaving activities due to their size (n=6). Further use-wear analysis is needed to determine function. 6. Other Included in this type are eight subtypes that do not fit into the categories above. They are often unique tools or objects for specific purposes. 6A. Comb Six objects contain serrated edges, giving the ap­pearance of a comb (Fig. 8). All are recovered from open areas in Late V and IV levels. They are made from long bone fragments and may have been used in textile manufacture to separate fibres, or possi­bly in pottery decoration, to incise ceramics. How­ever, incised ceramics from the site have deeper and wider lines and dots than bone comb tips. Additio­nally, the teeth of these combs are often squared and flattened at their ends. Their use in textile pro­duction is thus more plausible. 6B. Bipoint There are two items defined by their flat surface and double active tips. They are found in levels VI and V (Fig. 9). 6C. Bone handle Also known as a shaft/sheath, these items were used to protect the user’s hand during use. Inside the handle would fit an additional tool, such as a sharp stone tool. They are made from long bones (n=10) (Fig. 9). 6D. Antler handle Similar to the objects above, antler handles, or shafts/ sheaths, where made from antler and used to hold an additional tool, most likely a stone tool (n=5) (Fig. 9). Fig. 6. Type 4. Scrapers. 6E. Antler tools Three additional tools were manufactured from deer antler. Further use-wear analysis is needed to ascer­tain the function of these variously shaped tools (Fig. 9). 6F. Fastener This object was recovered on the floor of Building 43 in Level VI. This object has been expertly con­structed, delicate, and is fragmented at one end. Its suggested use is as a clothes fastener due to its size and shape. This bone object may have been delib­erately placed to the building’s floor together with a scapula as part of closing ritual since the building appears to have been left clean (Fig. 9). 6G. Hook This item was uncovered in Level Vc and could have been used as a hook, owing to its shape and dimensions. It has a perforated tip which curves into a wide and flat shank, akin to fishhooks found in similar Neolithic contexts in the region (Powell 1996). However, its functional status is not entirely certain, as a similar item found at Çatalhöyük has been interpreted as a belt hook (Russell, Griffitts 2013) (Fig. 9). 6H. Arrow/spearhead A single arrow or spearhead be­longs to Level IVb. It has an oval-bodied shaft, with a pointed and flat tip (Fig. 9). Not included in this assessment are an additional nine decorated worked animal bones, which contain several pat­terned incisions, including zigzags, dots, and chev­rons. These items are currently being examined in detail in a separate study. Jarrad W. Paul, Cos¸kun Sivil, and Özlem Çevik Raw material Analysis of the Ulucak raw material occurred on site. It is often difficult to identify species of work­ed bone items, especially if the objects are exten­sively worked. As a result, the number of positively identified species is substantially less than the over­all number of worked bone items. Moreover, the number of species identified can then be limited when conducting an initial identification in the field. In this case, 95 items were positively attributed to a species from Levels VI–IV (Tab. 2). The results below are summarised from this sample. Results from this sample show that medium-sized animals (n=61) are favoured for bone tool construc­tion, with large-sized animals also often utilised (n= 32). Bones from small-sized animals (n=2) are sel­dom used. Sheep/goat and sheep-sized animals are most prominent for their use in constructing points, the most frequent tool type. Bones of cattle were also used repeatedly (n=32), while pig, hare, bird, and dog/wolf bones are rare. For the deer (n=14), roe deer are most common, with possible fallow and red deer examples in the collection. In terms of ele­ment selection, from the sample (n=95) of tibia bones were used most frequently for bone tool and object creation. Metapodials, including both meta­tarsal and metacarpal bones, were also used often, along with antler and rib bones. The use of ribs was also not constrained to a certain type; used to con­struct points, spatulas, and perforated objects. Deer antler was also used in far greater numbers than deer bones. Animal species selection for bone tool and object construction are similar when viewing the zooar­chaeological record for Levels VI through IV (see Pilaar Birch et al. 2019). That is, the most common species in the zooarchaeological record, sheep/goat, is also favoured for tool manufacture. Cattle is also frequent, although to a lesser extent, in both the worked and unworked bone groups. Suzanne E. Pila­ar Birch and colleagues also note an increase in deer bone frequency over time at Ulucak, which mirrors species selected tool manufacture. From the n=14 items made from deer antler and bone investigat­ed in this sample, n=12 are contextually placed in Level IV, with one in Level V, and one in Level VI. Manufacturing techniques Craftspeople at Ulucak used a range of manufacturing tech­niques to construct individual tool types. The most common technique for point manufac­ture was the splitting and gro­oving of metapodial bones, before shaping the tip into a point (subtype 1A). Separa­tion of the bone was via bi­partition as a result of percus­sion followed by grooving. The creation of ulna points (subtype 1B) was less labour The Neolithic worked bone assemblage from Ulucak Höyük, Western Anatolia intensive with abrasion of the tip to create a point with the base left unworked. Other bone points in the assemblage (1C, 1D, 1E) were created by split­ting a bone into irregular splinters and then through the process of grooving and abrasion creating a pointed object. The practice of splitting was also used to create flat spatulas (3A), where a rib bone was split in two and the inner spongy bone was smoothed via abrasion. Tibia bones were used for smoother (4A) items, with the base and shaft of the bone left unworked, while the tip was fractured via percussion then bevelled through abrasion. Objects that were perforated are also mostly drilled from both sides of the tool, evident in the slanted areas around the drill hole, although this may also be an indicator of extensive use-wear. While most tools contain the usual amounts of soil staining, some show evidence of intense burning, turning some objects black, grey, white, and blue. Bone colour can be an indicator of the degree of burning, with white and blue colouration suggestive of intense heating (Gilchrist, Mytum 1986.32). Tools with traces of burning are associated with contex­tual units that were destroyed by fire, also evident on damaged clay objects. One example, a pointed tool, shows evidence of controlled burning at the tip of the object, a technique used by the Ulucak crafts­people to strengthen the bone. This procedure re­quires specialized knowledge, as too much exposure to heat will result in damage. For use-wear, due to time constraints liner striations were only observed under x3 magnification, and this produced limited results. Due to this, we focus here on two subtypes: metapodial points (1A) and smoothers (4A). For metapodial points, the direc­tion of the striations (when positioning the object with the base at the bottom) ran transversely across the body in 21 examples. Nine points showed evi­dence of longitudinal striations, with less occur­rences of diagonal striations (n=6) and no discer­nible direction (n=5). The remaining points con­tained no evidence of striation pattern or direction under x3 magnification. For smoothers, most had a collection of random directional striations (n=10) with some exhibiting either longitudinal (n=2) or transverse patterns (n=2). The remaining smoothers showed no evidence of striation pattern of direction under x3 magnification. Further microscopic analy­sis is needed to ascertain whether striations where the result of use or manufacture wear, and the types of techniques involved in the process, such as scrap­ing or abrasion. Context Level VI (6850/6830–6500 cal BC) Worked bone objects from this period are found in Buildings 42 and 43 and in open areas surrounded by hearths. Low numbers of worked animal bone objects are attributed to Buildings 42 and 43, both of which have red painted lime floors and walls. The buildings are believed to have been deliberately emptied during an abandonment ritual (Çevik 2019). This ritual act is supported through evidence of ob­ject placement, particularly grinding stones, posi­tioned directly above the location in the previously built structure. As a result, any worked animal bone objects found after clean­ing may have been associated with this abandonment ritual. In Building 42 these objects include two points (1A, 1D), and a spatula (3A), while in Buil­ding 43 a bone fastener (6F) was found with a scapula. Due to its find context, the bone fastener may therefore be considered an important personal or­nament, and due to its connection with the burial of the building, possibly a communal building, may have played a part in wider burial traditions. Far more worked animal bones are at­tributed to open areas in this level, found in connection with several hearths and ovens. Points are most fre­quently found here, with spatulas and Jarrad W. Paul, Cos¸kun Sivil, and Özlem Çevik bevelled tools also numerous (Fig. 10). The level of skill used to produce spatulas and bevelled tools is lower when compared to examples in later levels. Si­milarly, their used surfaces have less abrasion and deformation. The types of tools suggest a mixed production area of leather and textile manufacture. The presence of a high number of animal bones around fire installations may suggest the processing of animals (butchering and hide processing) also took place in the same area. As such, an immediate need for an item may have been met with these op­portunistically made tools. Level V (6500–6000 cal BC) Ten buildings from this period (22, 23, 27, 30, 33, 40, 47, 51, 54 and 59) contain evidence of worked animal bone tools. Buildings 40 and 59 represent the earliest buildings from this period (Level Ve). Both buildings con­tained a low number of processed bone items, fol­lowing the trend in Level VI. Three items were re­covered from Building 40: points of subtypes 1A, 1C and 1E. Associated with Building 59, to its south, is a shallow lime-covered pit, that is believed to be another deposit connected to ritual abandonment. A single bone point (subtype 1G) was found in the pit alongside chipped stone tools, animal bones, and a ceramic sherd. Use-wear on the point suggests its function as a possible scraper involved in ceramic production. Building 54 (Level Vd) shows an in­crease in worked bone, and material more general­ly, with the inclusion of four metapodial points (1A), one oval-bodied point (1D), one perforated needle (2A) and one spatula (3A). Among the finds in this building was a higher than usual collection of spin­dle whorls (a total of 20). The appearance of the needle and spindle whorls together suggests textile production was one important activity carried out in Building 54. The frequency of worked bone items found in buildings increases to­wards to end of this period. In Le­vel Vb, Building 30 includes five worked bone pieces (three subtype 3A spatulas and two subtype 4A be-velled tools) and may have been linked to ceramic production given the types of tools recorded and ab­sence of other textile-related items such as bone points and spindle whorls. Building 33 contains three worked bone items (subtype 1A point, subtype 1E point and subtype 3A spatula), while in Building 47 four points are found: one made from a metapodial bone and three from other long bones. Building 51 contained a high density of bone tools (n=17), a standout for this period due to its breadth of types: seven subtype 1A points, one subtype 1E point, a perforated needle (2A), five sub­type 3A spatulas, two bone handles (6C) and one antler tool (6E). A stone chisel found in the building was able to be inserted into one of the bone han­dles, providing a direct link between the two mate­rial groups. In addition, an unworked metapodial bone and unworked rib bone were found at the south of the building, perhaps stored for future tool production. The end of Level V (Va) is seen in Buil­dings 23, 23 and 27. Building 23 contained a single perforated bone needle among 21 spindle whorls and 31 stone slingshots, while in Building 22, a me-tapodial point, two spatula-spoons, and an antler tool were uncovered. A single bone point was found in Building 27. Level IV (6000–5700 cal BC) Buildings in this period are placed in two categories: the specialized ceramic production of Level IVc and residential buildings of Level IVb. Regarding structures in Level IVc, Building 55 con­tains the most evidence of worked animal bone. Eighteen items include: two metapodial points (1A), a point made from a long bone (1C), four square-bo­died points (1E), a perforated needle (2B), six spat­ulas (3A), two chisels (4B), a perforated object (5B), and an antler tool (6E). Most items were made from the bones of large-sized animals. One of the points, made from a deer long bone, contained red pigment on its tip. This tool was therefore used for a diffe­rent purpose than the other bone points in the as­semblage, and suggests that at least some tools may have been used haphazardly for mixing/diluting The Neolithic worked bone assemblage from Ulucak Höyük, Western Anatolia paint, not necessarily its originally intended pur­pose. Worked bone preforms and manufacture waste in relation to wider on-site tool creation are the fo­cus of an additional ongoing study, but it is worth mentioning that Building 55 contained a variety of preformed worked bone tools including a tibia that had been divided into equal parts by splitting but was discarded, and a split long bone with its epi­physis cut. Other buildings in this category contain fewer work­ed bone objects. Building 56 contains a single point made from a rib bone (1G), while in Building 61 a metapodial point (1A), an oval-bodied point (1D), a square-bodied point (1E), and two spatulas (2A) were recorded. For Building 62 the types of tools are va­ried: two metapodial points (1A), a long bone point (1C), a fragmented point (1F), two perforated nee­dles (2A), a spatula (3A), a chisel (4A), and a perfo­rated object (5B). The worked bone objects in the residential buildings of Level IVb offer an interesting insight into daily tool types. Building 6, for example, contains four worked bone tool objects (a metapodial point, two spatulas, and a perforated needle) among other small finds, such as figurines, pendants, and spindle whorls. Building 13 also records figurines and pen­dants with the inclusion of two metapodial points and a spatula. Points are the only type found in Building 12, with four made from long bones (1C) and two that are square-bodied (1E), with other ob­jects including a stone tool, a pendant, and 13 beads. A figurine, stone tools, pestles, a grinding stone, and a spindle whorl are seen in Building 52, along with eight worked animal bone items: two metapodial points (1A), a square-bodied point (1E), a perforat­ed needle (2A), two spatulas (3A), a chisel (4B), and a perforated bone object (5A). Four worked bone items were also found in the street between Building 52 and Building 12, although they are too fragment­ed to be identified. Regional parallels From a typological standpoint, the collection at Ulu­cak is consistent with other collections found in western Anatolia from the Neolithic. For instance, pointed tools are also common at Yesilova, where deer antler was also used to create tools, including antler handles (or sheaths) (Derin 2012.180–182). Points dominate the Ege Gübre collection (Saglam­timur 2012.200), while similar types of objects are found at Çukuriçi Höyük, including points, spatu­las, smoothers, and spoons, common after 6500 cal BC (Horejs et al. 2015.304). This trend – the domi­nance of pointed tools, usually made from the long bones of sheep/goats – is also present in collections in the northwest, at sites such as Ugurlu, Ilipinar, Barcin, and Aktopralik (Paul, Erdogu 2017), and more broadly throughout Anatolia (Russell 2016). However, there are variations in the Ulucak collec­tion that are unique to the site. For instance, while medium-sized animals were favoured for making bone tools, large-sized animals were also utilized in greater numbers when compared to other sites in western and northwestern Anatolia (Paul 2016). Likewise, the under-representation of small-sized animal bones is regionally uncommon. The presence of unique items in the collection, such as the arrow-head/spear, comb, and the intricate fastener, are also rarely seen in the wider region. The number of items in the assemblage is also worth noting, as it is higher than the average for the region (Paul forth­coming). Overall, the Ulucak worked bone assemblage is con­sistent with other collections in the region. How­ever, when the typology is examined closely on a type-by-type basis and compared with other collec­tions in the region, certain unique trends emerge that illustrate some localization within a regionally established toolkit. Conclusion Results of this systematic investigation (including a typology, raw material and contextual analysis) un­derscore the prevalence and significance of worked animal bone at Ulucak. In the earliest levels of the settlement (Level VI) low numbers and a restricted diversity of types characterize a largely utilitarian toolkit, with production occurring rapidly for short periods of use. As the settlement grew, so did the number and range of tools produced, peaking in Le­vels V and IV. These tools were also used in conjunc­tion with an increasing number of items made from stone and clay. The tools at Ulucak were produced primarily from the metapodial and tibia bones of me-dium-sized animals (most commonly sheep/goat), with the rib bones of large-sized animals (such as cat­tle) also used. A variety of techniques were employ­ed, most notably grooving and splitting of long bo­nes, splitting of rib bones, and shaping via abrasion. Contextual analysis also provides clues as to the function of these items. Textile manufacture is seen Jarrad W. Paul, Cos¸kun Sivil, and Özlem Çevik as the primary role for many of the items in the as-ors. This may help in understanding the status of semblage. Bone points, needles, and combs in par-worked bone more generally at the site, with scarce ticular formed part of a wider textile toolkit at Ulu-evidence of tools found discarded in the streets be-cak. This is especially clear in Levels V–IV, with evi-tween buildings after Level IV, perhaps indicative dence of spindle whorls and loom weights (num-of a possible secondary symbolic importance. bering more than 300) at the site being one for the largest collections in western Anatolia for this peri-Worked animal bone tools were therefore an impor­od (Sevindik 2018). However, tool types do not al-tant aspect of the social and economic life at Ulu-ways seem to be restricted to set functions. For in-cak for over 1000 years of initial occupation. This stance, the bone point with the painted tip in Level study has laid the foundation for future worked bone IVc is indicative of an item with an intended func-research at the site, with additional analysis needed, tion (boring or sewing) that may have been hapha-such as detailed use-wear analysis, to confirm the re-zardly used for another purpose, in this case mix-sults of current interpretations. ing/diluting paint. Leather processing is also sug­gested to have been conducted using scraping tools, while spatulas could have aided ceramic production. ACKNOWLEDGEMENTS Evidence of preformed items also suggests on-site The Ulucak project is supported by the Turkish Mi- manufacture. nistry of Culture and Tourism. Jarrad W. Paul would also like to acknowledge support from the 2014 Jes- Tools in use during the Neolithic are also seen to be sie Webb Scholarship and the 2015 Lizette Bentwich part of complex abandonment rituals at the site, with Scholarship. items repeatedly placed deliberately on cleaned flo- . References Bradfield J., Forssman T., Spindler L., and Antonites A. R. 2019. Identifying the animal species used to manufacture bone arrowheads in South Africa. Archaeological and Anthropological Sciences 11: 2419–2434. https://doi.org/10.1007/s12520-018-0688-5 Çakirlar C. 2012. The evolution of animal husbandry in Neolithic central-west Anatolia: the zooarchaeological re­cord from Ulucak Höyük (c. 7040–5660 cal. BC, Izmir, Turkey). Anatolian Studies 62: 1–33. Campana D. V., Crabtree P. J. 2018. Bone implements from Chalcolithic Tepecik-Çiftlik: Traces of manufacture and wear on two classes of bone objects recovered from the 2013 excavation season. Quaternary International 472: 75–83. https://doi.org/10.1016/j.quaint.2017.11.042 Çevik Ö., Abay E. 2016. Neolithisation in Aegean Turkey. Towards a More Realistic Reading. In Ü. Yalçin (ed.), Ana­tolia and neighbours 10.000 years ago. Anatolian Me­tal VII, Der Anschnitt 31. Der Anschnitt. Zeitschrift für Kunst und Kultur. Beiheft 31. Veröffentlichungen aus dem Deutschen Bergbau-Museum 214. Deutsches Bergbau-Mu­seum. Druck SRS GmbH. Bochum: 187–197. Çevik Ö., Vuruskan O., Göz B., and Özturk C. 2020. Ulu­cak Höyük 2017–2018 Yili Kazi Çalismalari. Kazi So­nuçlari Toplantisi 41(1): 1–16. Çevik Ö. 2019. Changing ideologies in community-mak­ing through the Neolithic period at Ulucak. In A. Marci-niak (ed.), Concluding the Neolithic: The Near East in the Second Half of the Seventh Millennium BC. Lock­wood Press. Atlanta: 219–239. Çevik Ö., Erdogu B. 2020. Absolute Chronology of Cultu­ral Continuity, Change and Break in Western Anatolia Be­tween 6850–5460 Cal. BC: The Ulucak Höyük Case. Medi­terranean Archaeology and Archaeometry 20(1): 77– 92. DOI: 10.5281/zenodo.3605670 Çilingiroglu A., Derin Z., Abay E., Saglamtimur H., and Ka­yan I. 2004. Ulucak Höyük Excavations Conducted Be­tween 1995 and 2002. Peeters. Louvain. Çilingiroglu A., Çevik Ö., and Çilingiroglu Ç. 2012. Ulu­cak Höyük: Towards Understanding the Early Farming Communities of Middle West Anatolia: The Contribution of Ulucak. In M. Özdogan, N. Basgelen, and P. Kuniholm (eds.), The Neolithic In Turkey. Vol. 4. Western Turkey. Archaeology and Art Publications. Istanbul: 139–175. Çilingiroglu Ç. 2012. The Neolithic Pottery of Ulucak in Aegean Turkey: Organisation of production, interregio­nal comparisons and relative chronology. BAR Interna­tional Series 2426. Archaeopress. Oxford. The Neolithic worked bone assemblage from Ulucak Höyük, Western Anatolia Derin Z. 2012. Yesilova Höyük. In M. Özdogan, N. Basge­len, and P. Kuniholm (eds.), The Neolithic In Turkey. Vol. 4. Western Turkey. Archaeology and Art Publications. Istanbul: 177–195. Gilchrist R., Mytum H. C. 1986. Experimental archaeolo­gy and burnt animal bone from archaeological sites. Cir-caea: Bulletin of the Association for Environmental Ar­chaeology 4.1: 29–38. Horejs B., Mili. B., Ostmann F., Thanheiser U., Weninger B., and Galik A. 2015. The Aegean in the Early 7th Millen­nium BC: Maritime Networks and Colonisation. Journal of World Prehistory 28: 289–330. https://doi.org/10.1007/s10963-015-9090-8 Legrand A., Sidéra I. 2007. Methods, Means, and Results when Studying European Bone Industries. In C. Gates St-Pierre and R. Walker (eds.), Bones as Tools: Current Me­thods and Intepretations in Worked Bone Studies. BAR International Series 1622. Archaeopress. Oxford: 67–80. Ma¢rga¢rit M. 2017. Spatulas and abraded astragalus: Two types of tools used to process ceramics? Examples from the Romanian prehistory. Quaternary International 438: 201–211. https://doi.org/10.1016/j.quaint.2015.07.057 Marinelli M. 1995. The bone artifacts of Ilipinar. In J. Ro­odenberg (ed.), The Ilipinar Excavation I. Five Seasons of Fieldwork in Northwestern Anatolia, 1987–91. Neder-lands Historisch-Archaeologisch Instituut te Istanbul. Lei­den: 121–142. Paul J. W. 2016. The Prehistoric Worked Bone and Antler Assemblage at Ugurlu, Gökçeada: A Local and Regional Perspective. Unpublished PhD thesis. Univer­sity of Melbourne. Melbourne. forthcoming. Worked Animal Bone of the Neolithic North Aegean. Ancient Near Eastern Supplement Se­ries. Peeters. Leuven. Paul J. W., Erdogu B. 2017. An examination of the work­back to contents ed bone and antler assemblage at Ugurlu (Gökçeada, Tur­key). Documenta Praehistorica 44: 368–385. https://doi.org/10.4312/dp.44.23 Pilaar Birch S. E., Scheu A., Buckley M., and Çakirlar C. 2019. Combined osteomorphological, isotopic, aDNA, and ZooMS analyses of sheep and goat remains from Neolithic Ulucak, Turkey. Archaeological and Anthropological Sci­ence 11: 1669–1681. https://doi.org/10.1007/s12520-018-0624-8 Powell J. 1996. Fishing in the Prehistoric Aegean. Paul Atröms Förlag. Jonsered. Russell N. 2016. Anatolian Neolithic Bone Tools. In Ü. Yal-çin (ed.), Anatolia and neighbours 10.000 years ago. Anatolian Metal VII, Der Anschnitt 31. Der Anschnitt. Zeitschrift für Kunst und Kultur. Beiheft 31. Veröffentlich­ungen aus dem Deutschen Bergbau-Museum 214. Deut­sches Bergbau-Museum. Druck SRS GmbH. Bochum: 125– 134. Russell N., Griffitts J. L. 2013. Çatalhöyük worked bone: South and 4040 Areas. In I. Hodder (ed.), Substantive Technologies at Çatalhöyük: Reports from the 2000– 2008 Seasons. British Institute at Ankara Monograph. Vo­lume 48. British Institute at Ankara. Los Angeles: 277–306. Saglamtimur H. 2012. The Neolithic Settlement of Ege Gübre. In M. Özdogan, N. Basgelen, and P. Kuniholm (eds.), The Neolithic In Turkey. Vol. 4. Western Turkey. Archaeology and Art Publications. Istanbul: 197–225. Samei S., Alizadeh K. 2020. The spatial organization of craft production at the Kura-Araxes settlement of Köhne Shahar in northwestern Iran: A zooarchaeological ap­proach. PLoS One 15(3): e0229339. https://doi.org/10.1371/journal.pone.0229339 Sevindik K. 2018. Ulucak Höyük Dönem Tekstil Üretimi. Unpublished MA thesis. Trakya University. Edirne. Sivil C. 2017. Ulucak Höyük Neolitik Dönem Kemik Aletleri’nin Tabakalar ve Mekanlar Arasi Degerlendiril­mesi. Unpublished MA thesis. Trakya University. Edirne. Documenta Praehistorica XLIX (2022) Pattern and variation in jewellery production sequences> analysis of 4th millennium BC amber assemblages from the Latvian coast Valdis Be—rzin,[1, Agnese :akare2 1 Institute of Latvian History, University of Latvia, Ri -ga, LV< valdis-b@latnet.lv 2 Institute of Latvian History, University of Latvia, Ri -ga, LV< agnese.cakare@inbox.lv ABSTRACT – This study considers the production sequences of amber jewellery from Sarnate and Silinupe, at the coast of present-day Latvia. Differences between the two sites in terms of the relative frequency of items discarded in various production stages may be related to the degree of integra­tion into exchange networks. Within-assemblage variation in terms of the point within the process­ing sequence when perforation was performed indicates a strong element of heterodoxy with respect to amber processing within the communities, congruent with a domestic setting of production, even though the output consisted of a rather standardized range of forms. KEY WORDS – amber; Neolithic; jewellery; chaîne opératoire; Latvia Vzorci in variacije pri zaporedjih izdelave nakita> analiza zbira jantarja iz 4. tiso;letja pr. n. [t. z obmo;ja latvijske obale IZVLE.EK – V .tudiji obravnavamo postopek izdelave jantarnega nakita iz najdi.. Sarnate in Sili-nupe na obali dana.nje Latvije. Razlike med najdi..ema so z vidika pogostnosti predmetov, ki so bili odlo.eni v razli.nih fazah izdelave, povezane s stopnjo integracije v menjalni krog. Kljub vidni stan­dardizaciji oblik, ugotavljamo pri obravnavanih jantarnih predmetih mo.na odstopanja v izdelavi, in sicer v fazi prebadanja, kar je skladno z doma.o izdelavo. KLJU.NE BESEDE – jantar; neolitik; nakit; operacijska sekvenca; Latvija Introduction Amber has been a major focus of research into the concluding epoch of the Stone Age at the south­eastern and eastern shores of the Baltic Sea. Amber nodules washed up on the beaches were being widely collected and processed to make jewellery, and distributed via exchange networks far into the continental interior as well as northwards on the eastern side of the Baltic (Vankina 1970.Fig. 145; Zagorska 2003; Zhulnikov 2008; Núnez, Franzén 2011; Loze 2001; 2003; 2008). While the role of amber within the social milieu of the time is, as yet, poorly understood, it is evident that the processing and exchange of amber was an important factor in coast-inland contacts and interaction starting from about 4000 cal BC, when this material makes its appearance in archaeological collections. This article considers, from a comparative perspec­tive, amber assemblages dated to 3600–2900 cal BC from two sites at the Latvian coast, seeking to characterize and interpret processing sequences through the analysis of semi-manufactured orna­ments. Following the nomenclature widely utilized in this region of Europe, our sites fall within the DOI> 10.4312\dp.49.5 Pattern and variation in jewellery production sequences> analysis of 4th millennium BC amber assemblages from the Latvian coast Neolithic, as defined by the presence of pottery, even though paleaoeconomic data indicate a pre­dominant reliance on wild foods, especially fish and marine mammals in the case of these coastal settle­ments (Zagorska 2000; Berzin. 2008.293–330; Ber­zin. et al. submitted). The sites and amber assemblages The Sarnate and Silinupe sites (Fig. 1), providing the amber assemblages discussed here (both held at the National History Museum of Latvia), are coastland habitations yielding abundant evidence for the ex­ploitation of a range of wild food resources, includ­ing terrestrial and marine mammals, fish and birds. Both were situated so as to provide good access to freshwater as well as marine environments. Importantly, amber was a locally available raw ma­terial at both sites, since lumps of amber, transport­ed by longshore sediment drift from the Sambian Peninsula, could be gathered on nearby beaches (Vankina 1970.350; Berzin. et al. submitted). This article analyses a subset of the amber assem­blages from the two sites, namely the semi-manu­factured and finished pieces belonging to the three main jewellery classes: button-shaped beads, tubu­lar beads and pendants (i.e. disregarding compara­tively rare forms, such as discs, rings and unusual forms of beads, and a small number of figurines and other atypical pieces). The quantities of unworked lumps and flakes are also considered, but without further analysis. Sarnate The amber assemblage from Sarnate, located near the north-western coast of the Kurzeme Peninsula, was reco­vered in the course of exca­vations directed by Eduards .turms in 1938–1940 and Lu­cija Vankina in 1949 and 1953–1959 (Vankina 1970; Berzin. 2008). It may be no­ted that this is a peat-bog site with excellent wetland pre­servation, yielding a diverse array of organic artefact finds. More relevant in the context of amber studies is the clus­tered pattern of artefact dis­tribution on this extensive site. Distinct clusters of artefacts are in most cases associated with structural remains surrounding a hearth, and on this basis the collection has been sub­divided into smaller assemblages relating to indivi­dual household units or ‘dwellings’. In view of the range of seasonally exploitable food resources, the substantial character of the living structures and the presence of fragile and bulky equipment, Sarnate is thought to have been occupied year round (Berzin. 2008.381–383). Sarnate has given a very large collection of amber finds, which includes 638 semi-manufactured and finished ornaments along with unworked lumps and a great quantity of debitage (Vankina 1970.Tab. 6). Here, however, for the purpose of direct compa­rison with the Silinupe assemblage, we shall con­sider only that part of the Sarnate amber collection which was recovered from dwellings with pottery classed as Early Sarnate Ware, representing one of three phases of occupation on this site. Importantly, the Early Sarnate Ware phase at Sar­nate is also taken to include dwelling no. 2. This particular dwelling is quite exceptional, in that it produced a vastly greater quantity of amber finds than all of the rest. The dwelling consisted of an artefact cluster associated with a spread of sand measuring 4x8m, which seems to approximately re­flect the size of the living structure (Berzin. 2008. 297). The ceramic assemblage from this dwelling was a very small one. Nevertheless, the distinct preva­lence of porpoise tooth stamp (Berzin., Dumpe 2016; ‘tooth stamp’ in Berzin. 2008.Tab. 10) among the various kinds of stamped pottery decoration serves to place it in the Early Sarnate Ware phase, Valdis Be—rzin,[, Agnese :akare so that we may usefully compare its amber assem­blage with the much smaller assemblages from the other dwellings of this phase. Five samples from the Early Sarnate Ware dwellings at Sarnate have been dated. If we exclude two dates obtained from residue on pottery (Ua-33828 and Ua­15984), considered to be subject to the reservoir ef­fect, we are left with three dates from charcoal that give an overall date range of 3516–2880 cal BC for the Early Sarnate Ware phase at Sarnate. The amber ornaments from Sarnate considered here, namely the pieces identifiable as semi-manufactured and finished button-shaped beads, tubular beads and pendants from dwellings of the Early Sarnate Ware phase, number 152 in total. Dwelling 2 alone provides 117 of these, and the remaining 35 come from several other dwellings and one artefact cluster of this phase (dwellings D, E, MZA/MZR/MD, Pa, RZ/RD, IZ/ID, W and the artefact cluster recorded as ‘hearth 16’). In addition, seven unworked amber nodules and 2071 flakes were recovered from the dwellings of this phase, as shown in Table 2. The typology of amber jewellery from Sarnate has been treated comprehensively by Vankina (1970. 105–114), which also offers a general discussion of amber processing and exchange. A new analysis was undertaken by Valdis Berzin. (2003), focusing spe­cifically on production stages and intrasite spatial analysis, and evaluating the evidence for craft spe­cialization. The current article utilizes part of the data compiled in Berzin.’s study, reconsidered in the light of recent experimental work and theoret­ical developments, and viewed from a comparative Sample description Lab. no. 14C age BP Calibrated date, Notes cal BC, 2. Sa—rnate, Early Sa—rnate Ware dwellings Residue on potsherd with porpoise tooth impressions (A 11416>42), dwelling ID Ua-33828 5480±40 4443–4249 (95.4%) Previously published in Be—rzin, [ 2008.Tab. 2. Subject to freshwater and\or marine reservoir effect from ves­sel contents| (Not adjusted) Residue on potsherd with porpoise tooth impressions (A 11417>313), dwelling MD Ua-15984 5065±75 4036–3655 (95.4%) Charcoal, dwelling RZ\RD Tln-2918 4570±65 3516–3036 (95.4%) Charcoal, dwelling M FTMC­UM96-5 4484±29 3342–3090 (92.3%) 3054–3034 (3.2%) Charcoal, dwelling D FTMC­UM96-4 4300±31 3011–2880 (95.4%) Silin,upe, Early Sa—rnate Ware phase Charcoal (deciduous wood) extracted from pottery sherd no. 84 (with porpoise tooth decoration), area 7, level 3, 1989 Poz-133269 5330±50 4326–4287 (8.6%) 4266–4043 (85.2%) 4012–3998 (1.7%) Subject to freshwater and\or marine reservoir effect from ves­sel contents| (Not adjusted) Wood charcoal, area 7, central part, level 4, squares 8–11\f-g-h, 1988 FTMC­UM96-3 4791±31 3640–3524 (95.4%) Wood charcoal, area 5, level 3, hearth, 1988 FTMC­UM96-2 4743±31 3633–3500 (75.8%) 3432–3380 (19.6%) Bark roll – net float| (Betula sp.), inv. no. 292\539, area 7, S part, level 4, depth 1.1–1.3m, 25.07.1989 Poz-123665 4690±40 3624–3580 (8.8%) 3532–3368 (86.6%) Wood charcoal, area 5, level 3, hearth, 1988 FTMC­UM96-1 4545±30 3370–3306 (32.1%) 3300–3282 (2.5%) 3276–3266 (1.2%) 3243–3102 (59.7%) Bone (Bos taurus, centrotarsale), area 10, level 3, ~0.5–0.8m, 1989 Poz-137614 4490±35 3351–3089 (91.8%) 3057–3033 (3.6%) Bone (Alces Alces, phalanx), area 7 (S part), level 3, ~0.6–1.0m, 1989 Poz-137613 4450±35 3337–3210 (40.0%) 3194–3010 (51.9%) 2981–2961 (2.0%) 2951–2935 (1.6%) Tab. 1. Radiocarbon datings for the Early Sarnate Ware phase at Silinupe and Sarnate. Conventional 14C ages have been calibrated using OxCal v4.4 (Bronk Ramsey 2009) and the IntCal20 atmospheric curve (Reimer et al. 2020). Tln-2918 is a conventional radiocarbon dating; all the rest are AMS datings. Pattern and variation in jewellery production sequences> analysis of 4th millennium BC amber assemblages from the Latvian coast perspective, setting it against newly obtained data for the Silinupe assemblage. Silinupe The Silinupe site, at the south-western shore of the Gulf of Riga in the present-day village of Lapme.­ciems, was excavated by Vankina in 1954 and by Ilga Zagorska in 1988–1989 (Zagorska 2000; 2003; Berzin. et al. submitted). The site occupies the slo­pes of a former beach ridge and a dune ridge run­ning parallel to it, as well as a wet hollow between the ridges. Several occupation phases may be distin­guished on the basis of ceramic wares, but the great majority of the pottery assemblage may be classed as Early Sarnate Ware. As with Sarnate, considera­tion of the seasonality of the food resources used at the Silinupe site suggests year-round occupation (Berzin. et al. submitted). Six samples from strata at Silinupe containing Early Sarnate Ware have been radiocarbon-dated. If we exclude sample Poz-133269, thought to be subject to freshwater and/or marine reservoir effect, then the remaining five datings give an overall range of 3640–2935 cal BC for the Early Sarnate Ware occu­pation phase, in good agreement with the date range for the equivalent phase at the Sarnate site, as given above. A total of 316 semi-manufactured and finished orna­ments have been found at Silinupe, of which 104 pieces identifiable as button-shaped or tubular beads or pendants are considered here. In addition, 288 unworked nodules and 229 debitage pieces were recovered (Tab. 2). A brief general treatment of the Silinupe amber assemblage within the regional context of amber-working and exchange has previously been given by Zagorska (2003). In the present study, the mate­rial has been re-analysed from a technical perspec­tive, determining the form, dimensions, degree of surface treatment and state of perforation, in order to classify the ornaments into specific production stages (.akare 2020), following the approach pre­viously applied to the Sarnate collection. Jewellery forms Most frequent in the amber jewellery assemblage from the dwellings at Sarnate belonging to the Early Sarnate Ware phase are button-shaped beads, totall­ing 80 (Fig. 2.1–5). These beads normally have a V-shaped perforation visible only from the back, al­though pieces were often repaired by making a sim­ple front-to-back perforation (Fig. 2.3,5). They are generally circular in plan, with the occasional rec­tangular or oval example, and the cross-section is bi­convex or, less commonly, plano-convex. The dwellings of this phase have yielded 15 tubular beads, almost all of them cylindrical (Fig. 2.6–8). These have a long, straight perforation, which could be drilled by different methods, from one or both ends. There are 57 whole and fragmentary pendants from the dwellings of this phase (Fig. 2.9–14), the most common forms being trapezoidal and droplet-shaped. A variety of idiosyncratic pendant forms are also re­presented, as well as fragments of indeterminate form. The pendants have a short, drilled perforation, usually placed in the thinnest part of the margin. The perforation could be made from one or from both faces, thus resulting in a conical or biconical opening. At Silinupe, by contrast, amber processing was con­centrated mainly on the production of pendants (Fig. 3.6–13), which number 71, constituting two-thirds of the amber jewellery from the site. Most characte­ristic of this site are trapezoidal and irregular pen­dants. The high frequency of irregular pendants may have to do with the straightforward production pro­cess: they were made from unworked pieces of am­ber, without any surface processing, simply drilling a hole where the margin of the piece is thinnest (Fig. 3.9). Pendants made in this way predominate among the pendants from sites of this period along the whole of the western shore of the Gulf of Riga (Za-gorska 2003). The other amber pendants from Silinupe were given a specific form during the initial processing stage. Trapezoidal pendants (Fig. 3.8,11,12) are common­est, generally with a straight lower margin, this be­ing a characteristic form of amber jewellery from gra­ves in the Eastern European forest zone, in Latvia as well as further to the east. Trapezoidal pendants with a concave lower margin also occur. Present in smal­ler numbers are elongated and triangular pendants, as well as rarer pieces: rhombic pendants, pendants with one straight and one convex margin, and idio­syncratic forms that are not typologically classifi­able. At Silinupe, pendants drilled from both faces predominate among the finished, unbroken items. Relatively large numbers of button-shaped beads (Fig. 3.1–3) were also produced, with a total of 26 Valdis Be—rzin,[, Agnese :akare Fig. 2. Amber ornaments from Sarnate discarded in various stages of production (photos by V. Berzin.). Button-shaped beads: 1 unperforated, flaked/ ground (A11415:542); 2 un­perforated, ground all over (A11417:100); 3 broken at completed V-shaped perfora­tion, with secondary straight perforation, ground all over (A11415:515); 4 broken at complete V-shaped perforation, polished all over (A11415: 512); 5 broken at completed V-shaped perforation, polished all over (A11415:243). Tubu­lar beads: 6 unperforated, ground/polished (A11415: 544); 7 broken at incomplete perforation, flaked/ground (A11417:55); 8 broken at com­pleted perforation, flaked (A11417:20). Pendants: 9 un­perforated, flaked/ground (A11421:48); 10 unperforat­ed, ground all over (A11417: 17); 11 broken at incomplete perforation, flaked all over (A11415:218); 12 broken at completed perforation, front polished, back ground (A11421:27); 13 broken at completed perforation, polished all over (A11421:26); 14 broken at completed perforation, polished all over (A11415:539). finds. Commonest are circular beads with a V-shaped perforation, although oval beads were also found. They are most frequently plano-convex in cross sec­tion, less often biconvex. The class of tubular beads is represented at Silinupe by seven fragmentary cylindrical beads, all of them in a partially processed state (Fig. 3.4–5). Production methods and stages Our understanding of the processes applied in amber jewellery production and the assignment of the pie­ces to production stages is based on examination of the artefacts by naked eye, guided by the findings of previous analyses of amber-working by Vankina (1970), Ryszard F. Mazurowski (1984), Annelou van Gijn (2006; 2014) and Erik van Drenth (2013), and by the results of experiments performed by Eryk Popkiewicz (2012). First, it was important to select the most suitable raw material for amber ornaments. Only hard and clear amber would be chosen, as it is less likely to fracture and spoil the work. If the natural cortex of the am­ber lump was to be removed, this could be done using a knife or grinding stone in the form of a sand­stone or limestone slab, as attested by archaeologi­cal and ethnographic studies. There is ethnographic evidence of pre-processing heat treatment to im­prove the structure of amber (Popkiewicz 2012). The process of shaping and surface treatment can be divided into several stages: subdivision of the no­dule, flaking, grinding, and polishing. In the course of initial processing, a blank of suitable size would be obtained. Before further working, larger pieces of amber would be cut or divided into smaller ones, utilizing tools of bone, antler, flint or other stone (Mazurowski 1984). It has been shown experimen­tally that, when using a miniature axe, by control­ling the point where the flint blade will strike the amber, it is possible to determine the size of the amber fragment that will detach (Popkiewicz 2012). Amber lumps can also be split by indirect percus­sion, using an antler punch to apply strong point pressure, or they can be subdivided by cutting with a wet thread (van Gijn 2006; Popkiewicz 2012). Pattern and variation in jewellery production sequences> analysis of 4th millennium BC amber assemblages from the Latvian coast Fig. 3. Amber ornaments from Silinupe discarded in various stages of production (photos by M. Kalnin. and V. Berzin.). But-ton-shaped beads: 1 perforation complete, front polished, back ground (VI292); 2 perforation complete, polished all over (A11399:1); 3 broken at incom­plete V-shaped perforation, with completed secondary straight perforation, front polished, back ground/polished (A11399: 52). Tubular beads: 4 broken at incomplete perforation, flaked/ ground (VI292); 5 broken at in­complete perforation, flaked (VI292). Pendants: 6 unperfo­rated, flaked (A11399:34); 7 broken at complete perforation, flaked (A11399:27); 8 broken at complete perforation, ground/ polished (VI292); 9 completed perforation, natural surface (A11399:112); 10 completed per­foration, natural surface, expos­ed to heat? (A11399:76); 11 bro­ken at complete perforation, ground/polished (11399:6); 12 completed perforation, partial­ly ground/polished (VI292); 13 completed perforation, partial­ly polished (VI292). Once a suitable-sized piece of amber was obtained, it would be shaped into the desired form. Various kinds of blanks may be distinguished: nodules, blocks, and flakes (van Gijn 2014). Unshaped am­ber lumps could be used for making pendants, but the blanks for more elaborate pendants and beads were shaped after splitting. Amber flaking involved similar techniques and instruments to those used for flint-knapping, with the difference that amber-working requires smaller tools that permit greater precision (Popkiewicz 2012). Flaking could be per­formed with a flint flake or miniature axe. This leaves traces of retouch on the amber surface, where small fragments of amber are detached by the pres­sure of the tool edge (Figs. 2.1,7,8,11; 3.4–6; Popkie­wicz 2012). This technique is comparable to re­touching on flint (Vankina 1970.112–114). The size of the detached fragments depends on the pressure exerted on the amber and on the angle of the tool edge to the surface of the amber (Popkiewicz 2012). Grinding was employed to smooth the surface and round the sharp edges. Grinding stones with diffe­rent grain sizes could be used for this, leaving regu­lar striations running in various directions (Fig. 2.2, 3,6,10). Subsequently, the surface would be polished using soft leather or fabric (Popkiewicz 2012). Drilling was the most difficult operation in amber processing, and was a frequent cause of breakage. The piece could be drilled at any stage of process­ing. Different kinds of drilled perforations may be dis­tinguished, relating to the form of the ornament: pendants most commonly have short perforations; button-shaped beads have V-shaped perforations; and tubular beads have long, straight perforations. A different technique was used for each kind of per­foration, and they differ in the degree of difficulty. Pendants were simplest to perforate, as with such a short perforation, there was less chance of the or­nament fracturing. A flint drill could be used for this purpose, drilling from one face or both and obtain­ing a conical or biconical perforation, respectively (Popkiewicz 2012). Drilling from both faces was the most common practice, possibly in order to reduce Valdis Be—rzin,[, Agnese :akare the likelihood of breakage (Drenth 2013; van Gijn 2014). It was also the most common in cases of re­peated perforation, undertaken in repairing items at Silinupe. Often, however, the placement of the two conical perforations was incorrectly judged, and so it was necessary to drill obliquely, which could lead to breakage (Popkiewicz 2012; van Gijn 2014). It has been confirmed experimentally that drilling may also be performed with a pointed piece of wood or antler, using a fine slurry, with this process leav­ing fine, regular, circular scratches (van Gijn 2014). V-shaped perforations were provided for the button-shaped beads, drilling obliquely from two points on the same side, so that the perforations met to form the V shape. The tubular beads required a long, straight perfora­tion. Moreover, it was important for the perforation to be symmetrical in relation to the ornament, which was difficult to achieve. In order to obtain a cor­rectly aligned perforation, the first step is to mark the drilling site, so that the drill does not slide across the surface of the amber, which can be done using the sharp edge of a flint tool (Popkiewicz 2012). For drilling the perforation itself, a longer instrument was required. In cases where a straight, cylindrical perforation is observable, a hollow drill could have been used, perhaps a bird bone, leaving a smooth perforation (van Gijn 2006; Drenth 2013). These perforations were drilled from one or both ends. A flint drill attached to a wooden shaft could also have been used for such longer, more difficult perforations (Popkiewicz 2012). The perforations of tubular beads broke or failed more commonly than those of other ornaments (Figs. 2.7; 3.4,5), re­flecting the higher degree of difficulty. The amber ornaments were classified into five pro­duction stages, taking into consideration both the character of the surface finish and the state of the perforation (Berzin. 2003). The following degrees of surface finish were distin­guished: . natural cortex all over; . partially covered in natural cortex, partially flaked; . surface entirely flaked; . partially flaked, partially ground; . surface entirely ground; . partially ground, partially polished; . surface entirely polished; . front polished, reverse ground (mainly seen on button-shaped beads); . partially natural cortex, partially polished (irreg­ular pieces not shaped by flaking, only polished). The following perforation states were distinguished: . unperforated (piece intact); . partly perforated (piece intact); . perforation complete (piece intact); . broken at the perforation, with perforation clearly incomplete; . broken at the perforation, with perforation pos­sibly or definitely complete; . fragmentary piece not showing a perforation (e.g., the lower part of a pendant or the margin of a but-ton-shaped bead). The pieces were assigned to one of five production stages as follows: 1. Unperforated pieces: these are pieces abandoned either without completing pre-perforation working, or else leaving the piece ready for perforation with­out actually starting this operation. Such finds are taken to indicate shaping/surface treatment before perforation or the stocking of unperforated blanks. 2. Pieces with an unfinished perforation: either intact or broken at the perforation. These indicate failed or interrupted drilling. In the case of some broken pieces, it is not evident whether the perfora­tion was complete at the time of breakage. For the purpose of assigning such pieces to a production stage, pieces which have a flaked but not ground surface are included in this stage, taking into ac­count that flaking most commonly preceded perfo­ration (as shown in Table 3). On the other hand, if such a broken piece has a ground surface, the state of surface treatment does not give any indication as to whether it broke during perforation or during subsequent finishing, and such pieces are classed as “production stage indeterminate”. 3. Pieces with a finished perforation, but unfin­ished surface treatment: i.e. not all of the surface (or the front in the case of button-shaped beads) has been polished. Such pieces indicate that post-perforation finishing was being conducted, and that work on this piece was abandoned before this operation was completed. 4. Finished products: pieces with a finished perfo­ration, where the whole surface (or the front in the case of button-shaped beads) has been polished. 5. Pieces with a secondary perforation: these are cases of repair after breakage at the perforation. Pattern and variation in jewellery production sequences> analysis of 4th millennium BC amber assemblages from the Latvian coast Tab. 2. Amber finds belonging to the main ornament categories from the Early Sarnate Ware phase at Sarnate and from Silinupe, classified according toproduction stage. Valdis Be—rzin,[, Agnese :akare Table 2 and Figure 4 give the numbers of button-shaped and tubular beads and pendants assigned to each of the production stages in the assemblages from Sarnate dwelling 2, from the other dwellings belonging to the Early Sarnate Ware phase at Sarna­te, and from Silinupe. Out of a total 256 ornaments, 209 (82%) could be assigned to a specific production stage. The chaîne opératoire and its variation The basic classification according to production sta­ges is a logical starting point for elaborating the chaîne opératoire. Analysis of the semi-manufac­tured pieces reveals, however, that the production steps do not follow a fixed sequence. While shap-ing/surface treatment proceeds from flaking to grind­ing to polishing, there is considerable variation in terms of the point within this sequence at which the piece was perforated. This is apparent if we con­sider the state of surface treatment only for those semi-manufactured ornaments, 41 in total, which clearly display an incomplete perforation (Tab. 3), almost all of them fragmentary pieces that evident­ly broke during perforation.1 Table 3 shows that among the pieces exhibiting a failed perforation, seven had been flaked all over the surface (with no grinding traces) before they were perforated; eight had been flaked and only partially ground; and the greatest number – 12 – had already been ground all over. There are also small­er numbers of incompletely perforated pieces in va­rious other states of shaping/surface treatment, rang­ing from completely unshaped (with natural cortex all over) to completely polished. All three ornament classes – both kinds of beads as well as pendants – show considerable variation in the degree of shaping/surface treatment carried out in advance of perforation. And such variation occurs within all three assemblages. Even the assemblage from Sarnate dwelling 2, which must be regarded as representing the output of a relatively restricted human group (a household or small workshop), shows an element of variation in this respect. Accordingly, the generalized chaîne opératoire for amber ornament production on these sites cannot be viewed as a sequence of steps performed in a set order, of the kind often visualized in schematic form in studies on processing sequences for lithic materi­als (e.g., Inizan 2008.Figs. 8–10), bone (e.g., David 2006.Fig. 2) or ceramics (e.g., Perry 2016.Fig. 11). The kind of chaîne opératoire indicated by the am­ber assemblages, where a particular operation may come at various points in the processing sequence, is challenging to represent schematically. We have chosen to view it as consisting of two independent sub-chaînes: (1) the sequence of shaping/surface treatments, and (2) perforation (Fig. 4). The same kind of variation in the position of perfo­ration within the order of production steps is noted by van Gijn (2014) in her study of a Neolithic amber assemblage from Zeewijk in the Netherlands. It is likewise observed in Stone Age assemblages in Po­land and in collections from medieval amber work­shops in the Baltic Sea region (Popkiewicz 2012 and reference therein to Wojtasik 1990.147–148). Proceeding from his amber processing experiments, Popkiewicz (2012) highlights the ergonomic and labour-saving aspects of this process: (1) a piece still in the initial stage of processing, and hence larger, is easier to hold or fix in position for drilling the perforation; and (2) because drilling is the most dif­ficult part of the work, presenting the greatest risk of failure, less time/labour will have been wasted in the case of breakage while drilling, if this operation is performed on a little-worked piece than on a piece already in an advanced stage of completion. This begs the question: why did amber-workers ne­vertheless often leave drilling until after grinding, or even after polishing? There are evidently techni­cal issues at play here, presumably relating to the particular drilling technique employed and the means of fixing or holding the piece for drilling, as well as the form of the unworked lump, the kind of ornament it was to become, and the kind of perfora­tion required for the particular ornament. All of these aspects deserve attention in future experiment­al work. Leaving aside these practicalities, what does the oc­currence of variation in the sequence of processing tell us? Following the theoretical approach taken by Gwendolyn O. Kelly (2016) in her study of stone bead production in Early Historic South India, we are witnessing an element of heterodoxy within the 1 Unlike in the assignment of production stages in Table 2, pieces where the perforation is assumed to be incomplete because the surface is flaked but not ground are excluded from this analysis to avoid any ambiguity. Pattern and variation in jewellery production sequences> analysis of 4th millennium BC amber assemblages from the Latvian coast communities producing the ornaments. Thus, there was an acceptance of alternative practices, in this case with respect to the order in which the opera­tions of producing an amber ornament were carried out – in contrast to an orthodox attitude, where only one set of practices is acceptable. In her discussion of the Zeewijk assemblage, Van Gijn (2014) proposes that the variability observed in the exact production sequence of amber beads supports the assumed domestic production of the ornaments, with different people having slightly different techniques, whereas if the beads had been produced in workshops a more standardized pro­duction sequence would be expected. Such a conclusion may also be valid with respect to amber processing in the Neolithic of the Eastern Bal­tic. Thus, in addition to the evidence for intensive amber-working from Sarnate dwelling 2 (not just the amber assemblage itself but also grinding stones and certain lithic tool types that probably constitute amber-working gear), this dwelling also yielded a number of flint spear- and arrowheads, two pebble net sinkers and a significant amount of pottery (Van­kina 1970.75–76, Tab. 3, Fig. 133; Berzin. 2008.Tabs. 10, 18), providing evidence of engagement in subsistence ac­tivities. In other words, am­ber processing appears to be taking place in what may be described as a domestic con­text. Lithic debitage was also abundant, showing that not just amber but also lithics were being intensively pro­cessed. It appears that the domestic production of amber orna­ments provided a context in which heterodoxy in proces­sing practices could thrive, which is an important obser­vation, if we consider that the output, in terms of the dominant classes of amber ornaments, was actually ra­ther standardized. Thus, the makers of amber ornaments, apparently dispersed across a great many household pro­duction settings, were follow­ ing somewhat different methods, even though they were aiming to produce standard ornament forms of recognized value for participation in a supra-re­gional exchange network. Intensity of amber-working and production of different jewellery classes The absolute quantities of amber recovered (Tab. 2, ‘Total finished/semi-manufactured items’, ‘Un-worked nodules’, ‘Flakes’) and the number of finds per square metre give at least some indication of the relative intensity of amber-working. In this respect, Sarnate dwelling 2, with an excavated area of 86m2, is in a class of its own, with its 126 (1.47/m2) semi-manufactured and finished items and 2027 (23.57/m2) pieces of debitage (but no unworked nodules recorded!). The other Early Sarnate Ware phase dwellings at Sarnate, with a combined exca­vation area of 524m2, are relatively very impover­ished in terms of amber finds, together yielding only 43 (0.08/m2) finished/semi-manufactured items, se­ven (0.01/m2) unworked nodules and a mere 44 (0.08/m2) debitage pieces. Silinupe lies in between these extremes: here, an excavated area of 336m2 Valdis Be—rzin,[, Agnese :akare Tab. 3. Degree of surface treatment on ornaments with an unfinished perforation. Note: the pieces considered here are only those where the perforation isclearly observed to be incomplete, unlike in Table 2, where broken pieces with a perforation not clearly identifiable as incomplete are also classified in pro­duction stage 3 if they have been flaked but not ground. has given 316 (0.94/m2) finished/semi-manufac­tured items, 288 (0.86/m2) unworked nodules and 229 (0.68/m2) debitage pieces. It should be added that in the case of Silinupe (but not Sarnate), the natural strata at the base of the archaeological sequence also contain amber lumps, transported by longshore drift in the sea and depo­sited along with other sediment (Berzin. et al. sub­mitted), so we cannot be certain that all of the re­covered unworked amber on this site was actually collected by the inhabitants. The general situation is clear enough: dwelling 2 at Sarnate was a locus of very intensive amber-working activity, starkly contrasting with a very low level of amber-working in the other dwellings of this phase, while Silinupe falls in between. Because the Sarna­te collection can be subdivided into separate dwel­ling assemblages which we can equate with small-scale social groups, a pattern is revealed where one such group within a larger community – that asso­ciated with dwelling 2 – was engaging much more intensively in amber-working than others. Accor­dingly, dwelling 2 was previously regarded as an amber workshop (Vankina 1970.114), and has sub­sequently been considered through the theoretical lens of craft specialization (Berzin. 2003). We can­not exclude the possibility that there was a similar concentration of amber-working in the hands of a subset of the community at Silinupe, too. However, on this site the artefactual remains cannot be split into separate dwelling assemblages, and thus such patterns are not clearly detectable. We may next compare our three assemblages in terms of the proportional representation of the three classes of ornaments among all the semi-manufac­tured pieces (Tab. 2, “Total semi-manufactured pie­ces (stages 1–3)”), which should, at least to some degree, reflect how much of the amber production effort was in each case being devoted to the mak­ing of particular classes of ornaments. There are also salient differences in this respect. Thus, at Sarnate dwelling 2, button-shaped beads are the most com­mon among semi-manufactured pieces, numbering 27, or 53% of the total number of semi-manufactured pieces in the assemblage belonging to the three or­nament classes; tubular beads are somewhat less common, represented by 10 pieces (20%), while pendants number 14 pieces (27%). The situation is broadly similar for the other Early Sarnate Ware dwellings at Sarnate: 10 semi-manu­ Pattern and variation in jewellery production sequences> analysis of 4th millennium BC amber assemblages from the Latvian coast factured button-shaped beads (56%), just one tubu­lar bead (6%) and seven pendants (39%). The main difference from dwelling 2 is the virtual absence of semi-manufactured tubular beads (and no finished examples of such beads occur, either). From Silinupe we have 20 semi-manufactured but-ton-shaped beads (27%), six tubular beads (8%) and 47 pendants (64%). Thus, in contrast to both of the Sarnate assemblages, pendant production appears to have dominated at Silinupe. These patterns may be considered in relation to the level of difficulty involved in the different kinds of perforation required for the three ornament classes, as indicated by the experimental work of Popkie­wicz (2012) and by the occurrence of characteristic broken pieces. Pendants would have been the sim­plest items to perforate: only a straight, relatively short perforation was needed. The V-shaped perfo­ration of button-shaped beads is harder to achieve, while most difficult is the drilling of the long per­foration of the tubular beads. In all three assemblages, the proportion of semi-manufactured tubular beads is lowest among the three ornament classes, and on this basis they may be considered rather ‘exclusive’, presumably being made only from very good quality amber nodules and only by individuals with advanced drilling skills. The virtual absence of such pieces from the Early Sarnate Ware dwellings other than dwelling 2 sug­gests that within this community the skills (and per­haps also equipment) needed to make these pieces were only possessed by the people working in this dwelling. This difficulty of producing tubular beads also co-lours our view of the sets of amber ornaments pro­vided as grave goods in this period. Thus, for exam­ple, the set of 12 tubular beads provided for child burial 194 at the Zvejnieki cemetery (Zagorska 2001; Zagorskis 2004.Pl. 17) represents a particu­larly valuable item of jewellery in terms of the amount of highly skilled labour invested in it. Like­wise very valuable from this perspective was a set of tubular beads found on the Abora site in eastern Latvia (Loze 1975.Fig. 10; 2008.125). Representation of different production stages If we now proceed to examine the production-stage data for our three main ornament classes (Fig. 5; Tab. 2, “Total of 3 ornament classes”), then we find that each assemblage shows a somewhat different picture in this respect as well. Sarnate dwelling 2 is distinguished by a large number of pieces with a completed perforation but unfinished shaping/sur­face treatment (stage 3). The other Sarnate dwel­lings have a high proportion of unperforated (stage 1) pieces. Meanwhile, Silinupe has a very large share of incompletely perforated pieces (stage 2); pieces in this stage are also fairly common in the other Sar­nate dwellings, but comprise only a small percent­age in dwelling 2. Silinupe stands out in having a much lower proportion of finished ornaments (stage 4) than dwelling 2 or the other Sarnate dwellings. When considering the button-shaped beads specifi­cally, the pattern is similar but not quite the same: many perforated but incompletely finished pieces (stage 3) from Sarnate dwelling 2; a high propor­tion of finished pieces (stage 4) from dwelling 2 as well as the other dwellings at Sarnate; and a very marked predominance of incompletely perforated beads (stage 2) from Silinupe. In this particular or­nament class, the proportion of incompletely per­forated pieces is also high in the other Sarnate dwel­lings. As noted above, tubular beads are much less abun­dant in general, and are virtually absent from the other Sarnate dwellings. Although the absolute num­bers are small, we still see a similar difference be­tween Sarnate dwelling 2 and Silinupe to that ob­served in the case of button-shaped beads: namely, at Sarnate dwelling 2 completely perforated but in­completely finished pieces (stage 3) predominate, while Silinupe has produced almost exclusively pie­ces with an incomplete perforation (stage 2), which seems to reflect the difficulty of perforating tubular beads – a high proportion were evidently ruined at this stage. In the case of pendants, within the relatively small assemblage from the other Sarnate dwellings there is a large share of unperforated (stage 1) pieces. Si-linupe has many incompletely perforated (stage 2) pendants. The proportion of pendants with a com­pleted perforation (stage 3) is higher from Sarnate dwelling 2 and Silinupe than in the case of the other Sarnate dwellings. Finished pendants (stage 4), on the other hand, are commonly represented from the other Sarnate dwellings, while Silinupe stands out as the only assemblage with pendants re­paired by secondary perforation (stage 5). If we con­sider the proportions of the different production stages for the pendants in relation to the propor­ Valdis Be—rzin,[, Agnese :akare tions for the other two ornament classes, then the most salient common characteristic to emerge is that Silinupe has a high proportion of incompletely per­forated (stage 2) pieces. We are not in a position to attempt a comprehensive interpretation of all the patterns identifiable in these statistics. More detailed technical analysis would no doubt permit a clearer understanding of the factors behind the representation of the different produc­tion stages. However, the current understanding of Neolithic amber-working and our knowledge of the sites from which the assemblages originate does per­mit us to try to account for at least some of the va­riation between them. The previous treatment of amber-working at Sarna­te by Berzin. (2003) noted a degree of complemen­tarity in the data for dwelling 2 and for the other Early Sarnate Ware phase dwellings: thus, the rela­tively high proportion of pieces in the early stages of processing (stages 1 and 2) in the material from the latter suggested that perforated and unperfo­rated blanks may have been supplied from these dwellings for further processing (by specialists) in dwelling 2. The absence of unworked nodules in dwelling 2 also points to such a scenario. This is still seen as a valid hypothesis – although for a proper appreciation of such patterns we would really need to gain a clearer insight into the small-scale social re­lationships linking household groups within these communities. It is not, unfortunately, possible to as­certain whether dwelling 2 was inhabited simulta­neously with other excavated dwellings of this occu­pation phase. The most salient overall difference between the as­semblages, applying to all three ornament classes, is that Silinupe has a much higher proportion of pieces abandoned during the perforation process (stage 2), i.e. failed perforations, than either Sarna­te dwelling 2 or the other Sarnate dwellings, while both of the Sarnate assemblages are dominated by finished ornaments (along with perforated but in­completely finished pieces in the case of dwelling 2). In the knowledge that amber ornaments were impor­tant as exchange items, distributed from the coast along waterways far into the continental interior, we might indeed expect a high proportion of finished items to have been removed from our production sites, leaving mainly pieces that broke during pro­cessing or were considered in some way defective. The pattern seen at Silinupe appears to accord well with such a scenario: here, it seems that if the perfo­ration succeeded, then the ornament was almost always brought to completion and taken away. But this was apparently not quite so at Sarnate, espe­cially in the case of the intensive amber-working ac­tivity in dwelling 2. Here, a high proportion of suc­cessfully perforated pieces were also being retained. In many cases post-perforation finishing was left in­complete, but many finished pieces were also kept. In seeking to explain this difference, we may consi­der the geographical positions of the two sites in re­lation to exchange networks in the region east of the Baltic Sea. From such a perspective, Silinupe ap­pears to be much more centrally located, since it lies close to the mouths of two major waterways, the Daugava and the Lielupe. The River Daugava, in particular, along with its tributaries, has been viewed as a major route for amber distribution – to the Lake Lubans basin in eastern Latvia as well as much further eastwards into present-day Belarus and Russia (Loze 2001; 2003; Charniauski 2001). The Lielupe and its tributaries connect with present-day northern Lithuania. By comparison, Sarnate can be seen as occupying a rather peripheral location: it is fairly close to the River Venta, but this river has a much smaller drainage basin than the Daugava, and is, moreover, oriented southwards, connecting with a region not so distant from the coast of pre-sent-day Lithuania, where amber was likewise avai­lable locally. We would expect, then, that the people at Silinupe had much greater opportunities to engage in the eastwards-oriented exchange networks than the com­munity at Sarnate. Hence, at Silinupe there would have been a stronger stimulus to maximize amber ornament production and supply the great majori­ty of the finished pieces (perhaps even those consi­dered second rate or slightly defective) to the ex­change network. It might also have been an incen­tive towards the utilization of less-than-perfect am­ber nodules and to the involvement of a wider circle of individuals in the working of amber, even those with inferior skills – both of which would have in­creased the frequency of failure during perforation, as reflected in the assemblage. The relatively high proportion of semi-manufactured pendants at Silinupe (noted in the previous section) might, once again, be linked to this community’s in­tensive involvement in exchange, in which context it was perhaps advantageous to focus on the simpler-to-make forms. Pattern and variation in jewellery production sequences> analysis of 4th millennium BC amber assemblages from the Latvian coast Meanwhile at Sarnate, with its disadvantageous lo­cation for participation in the exchange network, there could have been a tendency for finished pie­ces to accumulate on the site, perhaps only those considered the best being selected for exchange, and in such a situation we might expect there to have been less incentive to bring to completion all of the successfully perforated pieces by undertaking the laborious process of polishing. This would explain the high percentages of perforated but incompletely finished (stage 3) as well as finished (stage 4) pieces from Sarnate dwelling 2. There is another important difference between the sites, which must have affected patterns of amber-working and may also partially account for the above-discussed differences be­tween the assemblages – namely access to various lithic raw mate­rials. Thus, recent studies by Mar­cis Kalnin. reveal that the inhabi­tants of the Early Sarnate Ware phase dwellings at Sarnate, includ­ing dwelling 2, were largely reli­ant on Silurian flint, which could be collected from the beaches along that stretch of the coast; at Silinu­pe, by contrast, the dominant lith­ic raw material is Cretaceous flint, which came from present-day sout­hern Lithuania or Belarus, at least 200km away, while the surround­ing area apparently lacked good lo­cally available lithic raw materials (Berg-Hansen et al. 2019; Kalnin., submitted; Berzin. et al. submit­ted). The pattern of lithic raw material use thus not only provides additio­nal evidence that the Silinupe com­munity was more intensively par­ticipating in long-distance exchange than the community at Sarnate. It was also much more dependent on exchange, because the exchange networks were providing the lith­ic material used for toolmaking. And moreover, as described above, flint tools had an important role in amber-working itself. Zagorska (2003), in her initial treat­ment of the Silinupe amber assem­blage, emphasizes the importance of amber ex­change with respect to this site, noting the preva­lence of broken semi-manufactured items in the as­semblage, and our comparison with Sarnate brings this into sharper relief. One further remark needs to be made here, namely that considerations of geo­graphical location and material flows are in them­selves inadequate for a proper understanding of am­ber exchange (and long-distance exchange in this re­gion of Europe in general). We also require a bet­ter grasp of the social context in which exchange was embedded, and this is an important task for future research (see Zhulnikov 2008 for an attempt to characterize the regional flow of amber ornaments in terms of prestige item exchange serving to estab­lish and strengthen social ties). Fig. 5. The representation of different ornament production stages in the amber assemblages from Sarnate dwelling 2, the other dwel­lings at Sarnate with Early Sarnate Ware and the Silinupe site (data from Table 2): 1 unperforated; 2 incomplete perforation; 3 complet­ed perforation, unfinished surface treatment; 4 finished artefact; 5 artefact repaired by secondary perforation; 6 production stage inde­terminate. Valdis Be—rzin,[, Agnese :akare Conclusions Our approach to amber assemblages from produc­tion sites, looking at the relative intensity of pro­duction of different jewellery forms, the representa­tion of different production stages among the pieces remaining on the site and the sequence in which the processing operations of shaping/surface finishing and perforation were performed, has revealed major differences in the overall character of the amber as­semblages from the two sites as well as intra-site (i.e. intra-community) variation in working practices. The differences in representation of the various pro­duction stages of amber jewellery at Sarnate and Si-linupe would appear to be at least partially explica­ble in terms of the Silinupe community’s closer inte­gration into amber exchange networks, owing to the advantageous geographical location for participation in long-distance exchange as well as the high depen­dency upon lithic raw materials obtainable through the exchange network. Meanwhile, the variation within the site assemblages in terms of the point within the processing sequence when perforation is performed indicates a strong element of heterodoxy with respect to amber pro­cessing practices within the communities engaged in this activity, congruent with a domestic setting of production – even though the actual output, in terms of the major jewellery classes, was rather standar­dized. These questions deserve further attention in future research on the technical as well as the social as­pects of ancient amber-working and exchange. ACKNOWLEDGEMENTS The preparation of this article has been funded by the Latvian Council of Science, project “People in a dynamic landscape: tracing the biography of Latvia’s sandy coastal belt”, lzp-2018/1-0171. Study of the Si-linupe amber assemblage has been undertaken by Agnese .akare in the frame of her undergraduate and master’s degree studies at the Faculty of History and Philosophy, University of Latvia. The authors are most grateful to Ildze Milgrave for assisting in data collection, and to Ilga Zagorska for valuable com­ments on a draft of this article. . References Berg-Hansen I. M, Damlien H., Kalnin. M., Zagorska I., Schülke A., and Berzin. V. 2019. Long-term variation in lithic technological traditions and social interaction: the Stone Age of the Eastern Baltic (Latvia), 10500–2900 cal BC. Fennoscandia Archaeologica 26: 6–32. http://www.sarks.fi/fa/PDF/FA36_6.pdf Berzin. V. 2003. Amberworking as a specialist occupa­tion at Sarnate Neolithic site, Latvia. In C. W. Beck, I. B. Loze, and J. M Todd (eds.), Amber in Archaeology. Pro­ceedings of the Fourth International Conference on Am­ber in Archaeology. Talsi 2001. Institute of the History of Latvia Publishers. Riga: 34–46. 2008. Sarnate: Living by a coastal lake during the East Baltic Neolithic. Acta Universitatis Ouluensis. B. Humaniora 86. University of Oulu. Oulu. http://jultika.oulu.fi/files/isbn9789514289415.pdf Berzin. V., .akare A., Kalnin. M., Lougas L., Milgrave I., and Zagorska I. Amber wind and porpoise jaw: Resource use at Silinupe (4th mill. BC), on the Baltic’s Gulf of Riga coast. Submitted. Berzin. V., Dumpe B. 2016. Cukdelfinu zobu iespiedumi neolita keramikas rotajuma. Latvijas Vestures Instituta .urnals 1(98): 5–27. https://www.lvi.lu.lv/lv/LVIZ_2016 _files/1%20numurs/V_Berzins_B_Dumpe_Cukdelfinu_L VIZ_2016_1.pdf Bronk Ramsey C. 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51: 337–360. https://doi.org/10.1017/S0033822200033865 Charniauski M. M. 2001. Amber on archaeological sites of Belarus. In A. Butrimas (ed.), Baltic Amber. Proce­edings of the International Interdisciplinary Conference “Baltic Amber in Natural Sciences, Archaeology and Ap­plied Arts”, 13–18 September 2001, Vilnius, Palanga, Ni-da. Vilnius Academy of Fine Arts. Vilnius: 141–148. .akare A. 2020. Dzintara apstrade Silinupes neolita ap­metne. Unpublished BA degree paper. Faculty of History and Philosophy. University of Latvia. Riga. David E. 2006. Contributions of the bone and antler in­dustry for characterizing the Early Mesolithic in Europe. In C. Kind (ed.), After the Ice Age. Settlements, subsis­tence and social development in the Mesolithic of Cen­tral Europe. Proceedings of the International Conference 9th to 12th of September 2003, Rottenburg/Neckar, Baden-Württemberg, Germany. Materialhefte zur Archäologie in Baden-Württemberg. Theiss. Stuttgart: 135–145. Pattern and variation in jewellery production sequences> analysis of 4th millennium BC amber assemblages from the Latvian coast Drenth von E. 2013. Stone Age amber from the Nether­lands – an outline. In F. Both (ed.), Die Kunde: Zeitschrift für niedersächsische Archäologie 65. Niedersächsischer Landesverein für Urgeschichte e.V. Isensee: 205–242. Inizan M. L., Reduron-Ballinger M., Roche H., and Tixier J. 1999. Technology and Terminology of Knapped Stone: followed by a multilingual vocabulary. Arabic, English, French, German, Greek, Italian, Portuguese, Spanish. Cercle de Recherches et d’Etudes Préhistoriques. Nanterre. Kalnin. M. Silurian flint on Neolithic (5300–1800 BC) sites in the central part of the East Baltic (Latvia). Submitted. Kelly G. O. 2016. Heterodoxy, orthodoxy and communi­ties of practice: Stone bead and ornament production in Early Historic South India (c. 400 BCE–400 CE). Archaeo­logical Research in Asia 5: 30–50. http://dx.doi.org/10.1016/j.ara.2016.03.001 Loze I. 1975. Neolithic amber ornaments in the Eastern part of Latvia. Przegl¹d archeologiczny 23: 49–82. 2001. Some aspects of research on Middle Neolithic amber in the Lake Lubans depression. In A. Butrimas (ed.), Baltic Amber. Proceedings of the International Interdisciplinary Conference “Baltic Amber in Natural Sciences, Archaeology and Applied Arts”, 13–18 Sep­tember 2001, Vilnius, Palanga, Nida. Vilnius Academy of Fine Arts. Vilnius: 125–133. 2003. Middle Neolithic amber workshops in the Lake Lubans depression. In C. W. Beck, I. B. Loze, and J. M Todd (eds.), Amber in Archaeology. Proceedings of the Fourth International Conference on Amber in Archaeo­logy. Talsi, 2001. Institute of the History of Latvia Pub­lishers. Riga: 72–89. 2008. Lubana ezera mitraja neolita dzintars. Latvijas vestures instituta apgads. Riga. Mazurowski R. F. 1984. Amber treatment workshops of the Rzucewo culture in .ulawy. Przegl¹d Archeologiczny 32: 5–60. Núnez M., Franzén P. 2011. Implications of Baltic amber finds in northern Finland 4000–2000 BC. Archaeologia Lituana 12: 10–24. https://doi.org/10.15388/ArchLit.2011.12.5128 Perry G. J. 2016. Pottery production in Anglo-Scandinavian Torksey (Lincolnshire): Reconstructing and contextualising the chaîne opératoire. Medieval Archaeology 60(1): 72– 114. https://doi.org/10.1080/00766097.2016.1147788 Popkiewicz E. 2012. Rekonstrukcje narzêdzi technik i technologii obróbki bursztynu. Prace Muzeum Ziemi 50: 91–101. Reimer P., Austin W., Bard E., + 38 authors, and Talamo S. 2020. The IntCal20 Northern Hemisphere radiocarbon age calibration curve (0–55 cal kBP). Radiocarbon 62: 725–757. https://doi.org/10.1017/RDC.2020.41 Van Gijn A. L. 2006. Ornaments of jet, amber and bone. In L. P. Louwe Kooijmans, P. F. B. Jongste (eds.), Schiplui-den: a Neolithic settlement on the Dutch North Sea coast c. 3500 cal BC. Analecta Praehistorica Leidensia. The Fa­culty of Archaeology. Leiden: 195–205. 2014. Beads and pendants of amber and jet. In E. M. Theunissen, O. Brinkkemper, R. C. G. M. Lauwerier, B. I. Smit, and I. M. M. van der Jagt (eds.), A mosaic of ha­bitation at Zeewijk (the Netherlands): Late Neolithic behavioural variability in a dynamic landscape. Ne-derlandse Archeologische Rapporten. Cultural Heritage Agency of the Netherlands. Amersfoort: 119–128. Vankina L. V. 1970. Torfyanikovaya stoyanka Sarnate. Zinatne. Riga. Wojtasik J. 1990. Materia³y bursztynowe ze Srebrnego Wz­górza w Wolinie. Materialy Zachodnio-Pomorskie 32: 139–152. Zagorska I. 2000. Sea mammal hunting strategy in the Eastern Baltic. Lietuvos Archeologija 19: 275–285. https://talpykla.istorija.lt/bitstream/99999/2419/1/LA_ 19_275-285.pdf 2001. Amber graves of Zvejnieki burial ground. In A. Butrimas (ed.), Baltic Amber. Proceedings of the Inter­national Interdisciplinary Conference “Baltic Amber in Natural Sciences, Archaeology and Applied Arts”, 13– 18 September 2001, Vilnius, Palanga, Nida. Vilnius Aca­demy of Fine Arts. Vilnius: 109–124. 2003. The ‘Gold Coast’ of the Gulf of Riga. In C. W. Beck, I. B. Loze, and J. M. Todd (eds.), Amber in Archaeology. Proceedings of the Fourth International Conference on Amber in Archaeology. Talsi, 2001. Institute of the Hi­story of Latvia Publishers. Riga: 106–115. Zagorskis F. 2004. Zvejnieki (Northern Latvia) Stone Age cemetery. BAR International Series 1292. Archaeo-press. Oxford. Zhulnikov A. 2008. Exchange of amber in Northern Eu­rope in the III millennium BC as a factor of social interac­tions. Estonian Journal of Archaeology 12(1): 3–15. https://doi.org/10.3176/arch.2008.1.01 back to contents Documenta Praehistorica XLIX (2022) The emergence of metal use in Greek Eastern Macedonia during the Neolithic period (late 6th–5th millennia BC) Dimitra Malamidou1, Zoi Tsirtsoni2, and Markos Vaxevanopoulos3 1 Ephorate of Antiquities of Serres, Serres, GR< dmalamidou@culture.gr 2 CNRS, UMR 7041 Archéologies et Sciences de l’Antiquité, Nanterre, FR< zoi.tsirtsoni@cnrs.fr 3 École Normale Supérieure de Lyon, CNRS, and Université de Lyon, FR< vaxevanopoulos@gmail.com ABSTRACT – Copper, gold, and silver artefacts, together with evidence of metallurgical activities, have been retrieved from Late Neolithic strata in several settlements in Greek Eastern Macedonia. Recent excavations at Dikili Tash revealed that gold was further used in paints for the decoration of pottery. It appears that the area’s inhabitants had a great familiarity with different metals and the distinct stages of the production-elaboration processes, including those interfering with other chaînes opéra­toires. Considering also the results from geological research, we propose a reflection on the socio-eco­nomic role of metal production and consumption for these societies, in their broader Balkan context. KEY WORDS – metalworking; Eastern Macedonia; Greece; Late Neolithic Za;etek uporabe kovin v gr[ki vzhodni Makedoniji v obdobju neolitika (pozno 6.–5. tiso;letje pr. n. [t.) IZVLE.EK – V ve. naseljih v gr.ki vzhodni Makedoniji so bili v pozno neolitskih plasteh skupaj z do-kazi o metalur.kih dejavnostih odkriti bakreni, zlati in srebrni artefakti. Nedavna izkopavanja v Di-kili Tashu so pokazala, da se je zlato uporabljalo v pripravi barv za okra.evanje keramike. Zdi se, da so prebivalci tega obmo.ja dobro poznali razli.ne kovine in faze procesov priprave in izdelave, tudi tiste v drugih operacijskih sekvencah. Ob upo.tevanju rezultatov geolo.kih raziskav, predlaga-mo razmislek o dru.beno-ekonomski vlogi proizvodnje in uporabi kovin v .ir.em balkanskem kon­tekstu. KLJU.NE BESEDE – obdelava kovin; vzhodna Makedonija; Gr.ija; pozni neolitik Introduction Northern Greece has been diachronically associated lack direct evidence about their exploitation in pre-with the exploitation of its mineral wealth. Greek historic and protohistoric times (with the remark-Eastern Macedonia in particular, i.e. the area be-able exception of the Palaeolithic ochre mine at Tzi-tween the Strymon and Nestos river valleys includ-nes, Thasos, around 20 000 years ago; Koukouli­ing the nearby Thasos island (Fig. 1), is known for Chryssanthaki, Weisgerber 1999), we have good its numerous polymetallic resources, which have reasons to believe that many of the gold, silver and been intensively exploited from Antiquity until mo-copper artefacts retrieved in recent decades from a dern times, with gold, silver, copper, lead and iron number of Neolithic, Chalcolithic and Bronze Age being the main metals extracted. Although we still sites in the area were produced from local resour- DOI> 10.4312\dp.49.6 The emergence of metal use in Greek Eastern Macedonia during the Neolithic period (late 6th–5th millennia BC) ces. Moreover, recent analytical studies in artefacts and metallurgical remains support the use of local ores on Thasos island (Bassiakos et al. 2019). The geological setting of the area, its metallic resources From a geotectonic point of view, Greek Eastern Macedonia belongs to the Rhodope Massif, which possesses several small- to large-scale magmatic-hy­drothermal ore mineralizations (Melfos et al. 2002; Melfos, Voudouris 2017). Most of these occurren­ces present traces of ancient mining (Wagner, Weis­gerber 1988; Koukouli-Chrysanthaki 1990; Vaveli-dis et al. 1995; 1996a; 1996b; Chiotis et al. 1996; Va-xevanopoulos 2017a; Vaxevanopoulos et al. 2022). Precious metal mineralizations also appear in the Serbo-Macedonian Massif at its borders with Rho-dope Massif in the Strymon valley. The Strymon ri­ver is well known for its placer gold from ancient writers. Pseudo-Aristotle describes the gold nuggets found in the riversides after heavy rains in the Paeo­nian territory (De Mirabilibus Auscultationibus 45). The Pangaeon mountain is often mentioned by an­cient writers, and ores rich in gold, silver and cop­per were extracted in many mining areas during An­tiquity (Vaxevanopoulos 2017a; 2017b). Copper rich mineralizations and metallurgical areas with copper being one of the main extracted metals, es­pecially in Roman times, have also been recorded (Vaxevanopoulos et al. 2018). Placer gold has been identified at several spots in the surroundings of the mountain. Alluvial gold is further reported in the streams crossing the region between Pangaeon and Symvolon (Baker et al. 1992), and limonitic veins with gold occurrences have been located in Alistrati, at the feet of the Menoikion moun­tain (Vavelidis et al. 1995). Fur­ther north, on the Angistron and Orvilos mountains, several gold and silver deposits have been recorded and gold-bearing veins in marble have also been exploited (Chiotis et al. 1996). Several mineralizations rich in gold and silver have also been identified in the region of Palea Kavala, on the Lekani Mounta­ins, which are considered as the ancient Skapti Yli described by Thucydides (Photos et al. 1989; Koukouli-Chrysanthaki 1990). More than 150 occurrences rich in Fe, Cu, Ag and Au have been recorded, whereas underground galleries are widespread in this area and can be dated from the 6th century BC until the Ottoman period (Vave­lidis et al. 1996a; 1996b). Finally, Thasos island, which as previously mentio­ned offers the earliest evidence for exploitation of metallic minerals in the Aegean, also has a number of gold, silver-lead and copper deposits. Gold depo­sits are located mainly on the eastern coast of the island where ancient galleries have been studied (Wagner et al. 1979; 1981; Vavelidis et al. 1988a; 1988b). Ancient exploitation of silver has been stud­ied in the Marlou-Kourlou, Koumaria and Vouves mining areas, in the opposite southwestern part of the island, with numerous underground galleries, shafts and surface extraction (Hauptmann et al. 1988; Pernicka, Wagner 1988; Pernicka et al. 1992; Wagner, Weisgerber 1988; Sanidas et al. 2018). Brief history of archaeological research in the area The first Neolithic settlements were established in the area (Fig. 2) in the second half of the 7th millen­nium BC (Lespez et al. 2013; Maniatis 2014.207). The plains of Serres and Drama were ideal for hu­man occupation as they offered abundant arable land and extended pastures, with sizable rivers or perennial sources of water present (Fotiadis 1985; Andreou et al. 1996). Relative proximity to the wo­oded slopes of the surrounding metalliferous moun­tains as well as to the Northern Aegean coastline must have added considerably to the subsistence/ economic potential of the area. Dimitra Malamidou, Zoi Tsirtsoni, and Markos Vaxevanopoulos Although both plains comprise many important Neo­lithic (late 7th–mid 4th millennium BC) sites (Gram-menos, Fotiadis 1980; Grammenos 1991.120–126; Koukouli-Chryssanthaki et al. 2008), only a few have been properly investigated. Promachon-Topolnitsa, on the west bank of Strymon on the Greek-Bulgarian border, is the only one excavated systematically to a large extent in the Serres basin (Koukouli-Chryssan­thaki et al. 2007). Smaller scale excavations have been conducted in the cave Orpheas-Alistrati (Kon-taxi et al. 2004) and the open-air sites at Agio Pnev-ma (Tolia-Christakou, Siopi 2008), Amphipolis-Hill 133 (Lazaridis 1964; 1965), Dimitra (Grammenos 1991; 1997) and Kryoneri (Malamidou 2007; 2016). In the plain of Drama, important systematic excava­tions have been conducted at the tell sites of Sitagroi (Renfrew et al. 1986; Elster, Renfrew 2003) and Di-kili Tash (Treuil 1992; Koukouli, Romiopoulou 1992; Darcque et al. 2020 with references; and www.dikili-tash.fr), whereas smaller investigations took place at Polystylo (Mylonas, Bakalakis 1938. 109–111), Arkadikos-Drama (Touloumis, Peristeri 1991; Peristeri 2002; 2004) and in the cave of Ma-aras at the Angitis sources (Trantalidou et al. 2005). In the small valley of Pieria, between the Pangaeon and the Symvolon mountains, only the site of Akro­potamos has undergone some investigation (Mylo­nas 1941). A small-scale excavation was further con­ducted at the site of Paradeisos, in the Nestos river valley (Hellström 1987). The sites of Kastri-Theolo-gos (Koukouli-Chryssanthaki 1974; Koukouli-Chrys­santhaki, Papadopoulos 2016.342–349) and Lime-naria (Papadopoulos, Malamidou 2012) are the only two Neolithic sites excavated so far on Thasos island. Limited remains of the final stages of the pe­riod have also been found in the neighbouring site of Agios Antonios (Maniatis et al. 2015.810–811). Most of these sites have provided evidence in con­nection with metals or metallurgy, in the form of fi- Fig. 2. Map of the area with the main Neolithic and Copper Age sites: 1 Paradeisos, 2 Kastri, 3 Limenaria, 4 Polystylo, 5 Dikili Tash, 6 Eleftheroupoli, 7 Kalamonas, 8 Kalambaki, 9 Doxato, 10 Kefalari, 11 Adriani, 12 Kallifytos, 13 Arkadikos-Drama, 14 Xeropotamos, 15 Mylopotamos, 16 Petroussa, 17 Maaras-Angitis Sources, 18 Kali Vrysi, 19 Megalokambos, 20 Sitagroi, 21 Mavrolefki, 22 Symvoli, 23 Orpheas-Alistrati Cave, 24 Angista R.S.-Paliokostra, 25 Nea Bafra, 26 Airi Bairi, 27 Dimitra, 28 Fidokoryfi, 29 Mikro Souli, 30 Moustheni, 31 Podochori, 32 Loutra Eleftheron, 33 Akropotamos, 34 Kokkinochori, 35 Galepsos, 36 Ofrynio, 37 Amphipolis-Hill 133, 38 Kryoneri, 39 Kastanochori, 40 Zervochori, 41 Tholos, 42 Toumba, 43 Pentapoli, 44 Agio Pnevma, 45 Fakistra Chryssou, 46 Chrysso R.S., 47 Vergi, 48 Strymoniko, 49 Proma­chon-Topolnica, 50 Katarraktes-Sidirokastro, 51 Asprovalta-Agia Lydia, 52 Asprovalta-Platoma, 53 Are-thoussa, 54 Mikri Volvi, 55 Nea Apollonia, 56 Profitis. Sites no. 51 to 56 belong administratively to the region of central Macedonia. Red dots indicate sites with evidence of metal objects and/or metallurgy. The emergence of metal use in Greek Eastern Macedonia during the Neolithic period (late 6th–5th millennia BC) nished objects, metal-processing installations, instru­ments or by-products. Part of the evidence relates with later occupation layers (Early or Late Bronze Age), frequently present at the same sites, but much of it comes from secure stratified Late Neolithic de­posits and even, exceptionally, from in situ contexts (Aslanis, Tzahili 1990). Archaeometric examination of the related finds has produced valuable informa­tion on the technology of early copper, gold and sil­ver exploitation in the area (Mirtsou et al. 1997; Koukouli-Chrysanthaki, Bassiakos 2002; Renfrew, Slater 2003; Koukouli-Chrysanthaki, Papadopou­los 2009.2; Papadopoulos 2008.67; Bassiakos 2012; Nerantzis, Papadopoulos 2013; Bassiakos et al. 2019). Aim of the paper Our aim is to go through this available disparate in­formation in order to produce a comprehensive syn­thesis of our present knowledge on metal produc­tion and use in Greek Eastern Macedonia during the late 6th and 5th millennia BC, which is an important formative period not only here but across the entire Balkan peninsula (Pernicka 1993; Mazanova 2004; Jovanovi. 2009; .ljivar 2006; Hansen 2013; Kun­ze, Pernicka 2020a; Radivojevi., Roberts 2021) and the Aegean (Zachos 1996a; Papadatos et al. 2004; Bassiakos, Catapotis 2006; Kakavogianni et al. 2008; Zachos 2010), as well as in Anatolia and the Near East (Efe 2002; Maddin et al. 1999; Özdo­gan, Parzinger 2000; Rothenberg, Merkel 1998; Thornton 2001; Yalçin 2000; Yener 2000). More re­cent and still unpublished data are also included, mainly from the recent works at the tell settlement of Dikili Tash. We focus on technological informa­tion about the chaîne opératoire of the different metals, from the acquisition of ores and minerals from local sources and extractive metallurgy through smelting, to the production of finished objects. We combine material and earth sciences with well-es­tablished archaeological facts from individual sites, thus hoping to identify patterns of human behavi-our in connection with other activities (e.g., other chaînes opératoires), in order to better understand the functioning of Neolithic metallurgy and its so­cial background (cf. Storberg 2002.469; Ottaway, Roberts 2008; Bartelheim et al. 2015). By privile­ging well-dated and secure excavation contexts we try to gather pertinent evidence about metal re­source exploitation and consumer choices, taking into consideration the overall practices of local po­pulations, as well as possible networks of contacts and exchange at various geographical scales during the aforementioned period. Data presentation Copper Promachon-Topolnitsa Copper artefacts and the remains of copper-proces­sing were found in levels of phase Promachon III in the Greek sector, corresponding roughly to the first half of the 5th millennium BC. Excavations yielded several malachite beads (Koukouli-Chrysanthaki et al. 2007.51, Fig. 8), one copper pendant, copper fragments and slags. One heavily burned clay cru­cible containing traces of copper smelting was found at the bottom of a small pit (ibid. 48–51, Fig. 7.1, 2,4). The surrounding area also revealed traces of copper on the floor. In trench B, a series of hollows was discovered in the floor with successive layers of burnt clay in the interior (Fig. 3), which offered clear evidence for copper extraction. These features re­semble similar constructions at Dikili Tash (see be­low). According to preliminary analyses (optical mi­croscope, XRF and SEM), a special feature of copper production at Promachon-Topolnica is the use of a secondary copper ore (carbonate) of high pureness, consisting solely of malachite with >95% CuO. Such a pure copper ore, with no alumina-calcium-silica and iron admixtures, allows the production of cop­per at temperatures lower than 1000°C without slag being produced (Koukouli-Chrysanthaki, Bassiakos 2002; Bassiakos et al. 2019). Although relevant ana­lysis of the aforementioned pure copper ores is still ongoing, it is clear that the metallurgical activity in Promachon-Topolnica is closely related to technolo­gical traditions of copper production in the Balkan hinterland (Pernicka et al. 1997; Gale et al. 2003; Bori. 2009; Radivojevi. et al. 2010; Radivojevi., Rehren 2016; Rehren et al. 2020; Radivojevi., Ro­berts 2021). Dimitra Metal finds from Dimitra are proportionally numer­ous (Grammenos 1997.Pl. II) as all the soil from the two trenches investigated was sieved. Neolithic stra­ta yielded 38 copper objects – 21 beads, one fish­hook (ibid. 51, Pl. 36.2), five parts of pins or wires, and 11 undefined fragments. The majority come from 5th millennium BC layers, but some certainly date to the second half of the 6th millennium (i.e. five beads from ‘Middle Neolithic’ layers in trench II). Fourteen copper artefacts have been analysed with several methods, including emission spectro­scopy for the estimation of their qualitative compo­sition, proton induced x-ray emission (PIXE) to deter­mine their quantitative elemental composition, and Dimitra Malamidou, Zoi Tsirtsoni, and Markos Vaxevanopoulos x-ray radiography in order to confirm the internal metallic structure of the objects (Mirtsou et al. 1997). One copper artefact was sectioned for metallogra­phic examination to obtain information about the manufacturing procedure. All samples had different compositions, with arsenic, zinc and bismuth present in various proportions. Copper beads were produced by cold hammering and annealing at about 600°C. Apart from the metal objects, small amounts of poly-metallic ore (rich in copper, iron, lead) have been collected from both trenches, but they are not yet analysed. No melting or smelting installations have been identified in the excavated areas. Consequently, there is still no direct evidence for on-site metallurgi­cal activity, and we are not aware which metal source was exploited. Kryoneri Three copper pins/awls, two fragments of copper rings and some ore fragments were found in Late Neolithic deposits (mid-5th millennium BC), from both domestic context and waste pits (Malamidou 1997.518, Fig. 11; 2007.302, Fig. 8; 2016.312, Fig. 26). Preliminary examination of two pins under an electron-scanning microscope (SEM) by Yannis Ma-niatis at the Laboratory of Archaeometry of the NCSR Demokritos showed that they are made of pure cop­per. No further analysis has yet been conducted. Sitagroi The finds of phases Sitagroi II and III (late 6th and 5th millennia BC) include a copper pin, an awl, five beads and a few copper fragments or lumps (Ren­frew, Slater 2003.305, 319–320, Fig. 8.1.a-d,f-g). In addition, a marked concentration of sherds with copper incrustation was detected in square MM la­yer 20 and adjacent levels (phase Sitagroi III). Four sherds with copper remains from square MM, lay­ers 61 and 60, probably belong to an earlier con­text. Given the nature of the ceramics (coarse, with rough surfaces and curved walls) and the distribu­tion of the remains (mostly inside and on the bro­ken edges), they are interpreted as parts of crucibles or as sherds used to remove or to hold back dross during metal pouring (Fig. 4) (Renfrew, Slater 2003. 303, 312, Fig. 8.4). One of the best-preserved is a fragment of an oval-shaped crucible (dimensions c. 9x6cm, wall thickness up to 1.7cm) (ibid. 306, Fig. 8.4.I, Pl. 8.10). No moulds were recognized. Sixty-eight samples from Neolithic phases, either me­tallic or non-metallic but presumably related to me­tal technology, were sectioned and analysed on a Microscan V Microprobe Analyser (ibid. 301–302). It should be kept in mind that these analyses were conducted in the late 1960s using a first-generation microprobe analyser that had limited levels of preci­sion and sensitivity. The five non-metallic samples from the Sitagroi I levels provided no evidence of metalworking. Four out of the 12 analysed samples from the phase II levels were made of copper with no intentional alloying, while the other eight were non-metallic, showing no evidence of on-site smelt­ing. By contrast, phase III finds (11 copper objects and 40 sherds with copper deposits) show clear evi­dence of smelting and casting. The very low concen­trations of impurities in some samples may indicate that smaller pieces of native copper were melted to­gether in order to form an object, whereas objects with higher impurity levels may well be the product of smelting from ores. Other specimens show evi­dence of cold working and annealing. Three objects contained tin, but not in percentages as high as those normally associated with deliberate alloying, and two contained lead in small amounts. The question arises of where the activities of melt­ing or smelting, casting, cold working and annealing took place. Five separate contexts yielded sherds with adherent copper deposit. The richest one is, as already mentioned, square MM level 20, which is a rich fill including pottery, figurine fragments, bone, clay and stone objects, as well as two finished cop­per objects (the roll-headed pin and the awl or fish­hook). All fragments with copper deposit were con­centrated around an installation, consisting of a ba­ The emergence of metal use in Greek Eastern Macedonia during the Neolithic period (late 6th–5th millennia BC) ked clay surface surrounded on three sides by a pac­ked clay wall and open to the east (Renfrew et al. 1986.212, Fig. 8.19, Pl. XXXVII. 2). This may be in­terpreted as a pyrotechnological installation. Per­haps crucible melting was carried out in such places, although the smelting process was probably under­taken at or near the mining areas from which the copper ore was obtained. Dikili Tash A relatively small number of copper artefacts were found during the first excavation program, conduct­ed by Jean Deshayes and Dimitrios Theocharis be­tween 1961 and 1975. A small copper bead (Séféria­des 1992.115, Pl. 146.a) was the only artefact re­trieved from a phase I context (end of 6th millen­nium BC), roughly contemporary with Sitagroi II. The levels of Dikili Tash phase II (contemporary with Sitagroi III) in the sectors excavated by the French team yielded nine copper awls or pins (ibid. 115–116, Pl. 146.b, 199.e,f), but no ornaments. Two unstratified copper roll-headed pins from the same sectors might also come from levels of this period (ibid. 118–119, Pl. 147.b,c; 200.b). Five pins from phase II have been analysed by ato­mic absorption spectroscopy (AAS). They all contain minor percentages of lead, iron, silver, and nickel, and in two cases tin (ibid. 114, Tab. 12). All these elements can be present in small percentages in cop­per oxides such as malachite. A few occurrences of copper oxides and copper frag­ments are reported in the archives of Jean Deshayes in contexts of both phase I (square X29-niv. 13) and II (squares W29-niv. 9 and W30-sol 6). Interesting­ly, the last of these contexts shows heavy traces of combustion around a clay-plastered pit, which has been interpreted as a possible hearth (Treuil 1992. 23). A heavily burnt sherd with copper traces is also recorded in phase II (square X30-niv. 7). Sectors I and II, excavated by Theocharis in 1961 and 1967, respectively, have yielded some 10 cop­per artefacts and a sherd with copper incrustation, most probably a crucible fragment, and these are all unpublished. At least two of them (one pin and a flat sheet of copper) found in sector II/1967 could date from the Late Neolithic. The second and third excavation programs (1986– 2016) added further copper objects to the site’s in­ventory. Parts of four pins or awls were retrieved from late-5th millennium contexts in sector 6 (which is an extension of Theocharis’ sector II), although none in an entirely secure position. A fifth one was stuck on the outer surface of a sherd found in a slightly earlier level in sector 5 (level V/East/1, c. 4700–4500 BC). A copper bead (Fig. 5) was found near a hearth or oven in House 2, again in sector 6. The best-contextualized copper artefacts are however two other beads and an awl (Fig. 6) found in the neighbouring House 1, whose destruction has been dated to the years 4340–4260 cal BC. They come from two distinct groups of ornaments and raw ma­terials found on the house floor, less than 2m apart from each other. The first group, from which the beads come, contained some 270 items in total, mainly clay, shell and stone beads and a few pieces of unworked shell, whereas the second group con­taining the awl comprised no less than 1000 beads of stone, a dozen of spondylus bracelets, hundreds of perforated plaques and pieces of boar tusks, gra­phite cones, and also a few gold ornaments (Darc­que et al. 2012–2013.753–758; 2014.605–608; 2015; 2020.256–265, Fig. 6–39 to 6–48, 274–275, Fig. 6– 64). One of the two copper beads (actually a sheet of copper folded in the shape of a cylinder) was caught in the hole of a stone bead. Analysis with a portable XRF conducted in 2013 by Sariel Shalev (Haifa University) confirmed that both beads were made from pure copper (the awl was not analysed). A few more copper artefacts were retrieved from sector 2 at the southern periphery of the tell, but their chronology is not safe as they all come from colluvia with later components (Darcque et al. 2020. 86–87, Fig. 3–52.b-d). The only exception is a small copper bead, which comes from a colluvium secure­ly dated to the 5th millennium BC. Dimitra Malamidou, Zoi Tsirtsoni, and Markos Vaxevanopoulos From the same layer of sector 2 also comes a frag­mented clay mould (Fig. 7), the first discovered at Dikili Tash (ibid. 75–76, Fig. 3–39.e). It is made from well-refined clay and has fine incisions on all sides. The shape of the cavity on its upper surface suggests that it was probably used for the casting of chisels – a type of object common in assemblages of this period in both the Aegean and Balkans (Cher­nykh 1978.Pls. 10, 12; Zachos 2010.86, Fig. 6–6). XRF-analysis detected copper traces in the cavity’s walls. No crucible fragments have been found so far. One vitrified sherd with a copper deposit on its internal surface was found in the area to the southwest of House 3 in sector 6, but very near the surface. The existence of melting or smelting installations in Dikili Tash is debated. A series of small cavities whose inner surfaces were covered with successive layers of burned clay plaster, alternating with ash layers and bearing traces of intense fire, have been interpreted as possible metallurgical structures. All of them were found in layers of phase Dikili Tash I (three in level XIV of square W30, one in level XIII of square W30 and two in level 11 of square X30) (Séfériades 1983.647; Treuil 1992.21–23, Pls. 13.B, 15, 32.A,B; Séfériades 1992.115). However, unlike Promachon-Topolnitsa (see above), no trace of slags or other elements related to metallurgical activity have been detected within or nearby these struc­tures. Paradeisos The Late Neolithic levels at Paradeisos yielded two copper needles, slightly bent, 7 and 7.2cm in length respectively (Fig. 8) (Hellström 1987.85, Fig. 48.18, 19). Some other shapeless copper fragments were collected. No other indications of metallurgical acti­vity were recorded. Kastri Theologos-Thasos Two copper pins were retrieved from Late Neolithic levels at Kastri (mid-5th millennium BC). Analysis c a has demonstrated the existence of arsenical copper, probably of Thasian origin, although this could not be safely attributed to any of the known extraction sites on the island (Koukouli 1992.677). Limenaria-Thasos Excavations of Neolithic strata at Limenaria brought to light a significant number of finds associated with copper related metallurgical activities. The recovered evidence includes copper artefacts, slags, and nume­rous fragments of hematite/limonite lumps, some containing secondary cupriferous minerals, mainly malachite and azurite. A malachite bead was reco­vered from an upper layer in the Konstantinidis plot, dating from the end of the 6th or the beginning of the 5th millennium BC (Fig. 9) (Papadopoulos 2008.64, Fig. 2, p. 67; Papadopoulos, Malamidou 2012.41). It is a pierced piece of malachite, made simply by percussion and polishing. Twenty samples from slags and/or ore lumps, com­ing from various layers of the excavated areas, were studied with optical microscopy (OM), scanning elec­tron microscopy (SEM- EDX), x-ray diffraction spec­troscopy (XRD), and neutron activation analysis (NAA) (Bassiakos 2012; Bassiakos et al. 2019). Ana­lyses showed that copper slag samples derive from reduction smelting to obtain copper. A notable hete­rogeneity in their texture, either micro-morphologi­cal or chemical, seems to have resulted from ineffi­cient reducing conditions, suggesting a certain insuf­ 5mm 5mm 5cm Fig. 6. Dikili Tash, sector 6, copper beads (a-b) and copper awl (c) from House 1 (© Dikili Tash Project-EFA). The emergence of metal use in Greek Eastern Macedonia during the Neolithic period (late 6th–5th millennia BC) ficient control of furnace operation with some loss in the gaining of metallic copper. Despite that, these slags offer solid proof of a successful metallurgical process for efficient production of metallic copper. Until recently a similar heterogeneity in smelting slags from copper extraction was only noted at Aegean sites dated to the late 4th and 3rd millennia BC, such as Kephala-Petras, Crete (Catapotis et al. 2011; Papadatos et al. 2007) and Kythnos, Cyclades (Bassiakos, Philaniotou 2007), but this is now also reported from the late 5th millennium BC site of Ak­ladi Cheiri in South-East Bulgaria (Rehren et al. 2020). One sample represents a case of slagged lining de­riving from the furnace’s internal surface. Slag frag­ments like this are common within or very close to the smelting furnace from numerous other sites of early copper pyrometallurgy in the Aegean (Geor-gakopoulou 2005). Therefore, locating finds of this kind at Limenaria would suggest that the main smelt­ing site, from which the slags under study derive, was not far from the excavated sectors. The numer­ous slagged ceramic sherds found in a pit at Akladi Cheiri have also been attributed to the lining of a nearby smelting hearth (Rehren et al. 2020; Radivo­jevi., Roberts 2021.221). As for the numerous lumps or fragments of iron ores from the same excavated layers, they most probably derive from local sources and could be explained as waste products of a final enrichment stage of a cu-priferous iron ore (cf. Bassiakos, Catapotis 2006), 5cm Fig. 8. Paradeisos, copper needles (© Ephorate of Antiquities of Kavala). although they could also have been waste products of a different enrich­ment process for the extraction of a red or yellow pigment. Taken altogether, the analytical re­ sults strongly suggest that the metal­ workers exclusively utilized the local mineral resources for copper produc­ tion. They demonstrated significant skill and succeeded in extracting me­tallic copper from the locally available polymetallic (and poor in cupriferous secondary ores) mineral re­sources. Silver – lead Limenaria An unexpectedly early metal find, a small fragment of a silver pin, was discovered in a late Middle/early Late Neolithic horizon (second half of the 6th millen­nium BC) at the Lioudas plot (Fig. 10) (Papadopou­los, Malamidou 1997.836–837; Papadopoulos 2008. 65, Fig. 3). Compositional data and microstructural examination have shown that the pin was shaped by hammering of a lump of silver. Its internal mi-cromorphology reveals that the silver is heteroge­neous and consists of pure metal in which semi-translucent, minute pieces of slag are embedded. It is therefore presumed that the silver used to make the pin is not native, but the product of an extrac­tion from argentiferous lead ore (Bassiakos 2012. 210–211). No silver finds are recorded from the 5th millenni­um, which is altogether less well documented on the site. However, there is firm evidence for silver extraction practice from a context of the early 4th millennium BC (‘Final Neolithic’). Three litharge fragments coming from Markoulis plot belong to a piece of the ‘shallow bowl’ type, which derives from lead-silver extraction process (Fig. 11) (Papadopou­los 2008.66, Fig. 5.a-b, 68–69; Papadopoulos, Mala-midou 2008.431). Compositional analysis and mi-crostructural examination of these fragments sug­gest extraction and treatment of local Pb/Zn/Ag ores Fig. 9. Limenaria, mala­chite bead (© Ephorate of Antiquities of Kava-la). 1cm Dimitra Malamidou, Zoi Tsirtsoni, and Markos Vaxevanopoulos for the production of silver artefacts (Bassiakos 2012.208–210; Bassiakos et al. 2019). The circular, discoid morphology and the chemical composition of this type of litharge fragments do not relate to the platy or tubular litharge deriving from the well-known cupellation process of historic times (Bassia­kos et al. 2013). The process of early silver produc­tion that leaves behind as a by-product a shallow bowl of litharge has not been fully clarified yet, but similar examples are known from a number of sites from the end of the Final Neolithic/Early Bronze Age I (second half of the 4th millennium BC) in Me-sogeia, Attica (Kakavogianni et al. 2006.79; 2008; 2016.446–447, Fig. 15), or neighbouring Eastern Mediterranean regions (e.g., East Anatolia – Hess et al. 1998; Syria – Pernicka et al. 1998). The dating of the litharge from Limenaria is supported by both the relative and absolute chronology (calibrated date between 3977 and 3789 BC), which makes it one of the earliest, if not the earliest piece, of firm evidence for the process of silver production in the Aegean. Gold Dimitra Three golden artefacts were collected during the ex­cavation in Dimitra, all from Late Neolithic contexts, more or less securely dated to the advanced 5th mil­lennium BC. Two of them are beads: one collected from the upper part of the Neolithic deposits in trench I, disturbed by the LBA wall (Grammenos 1997.49, Pl. 35.8), the other from a lower level in the same trench (ibid. Pl. II). The third artefact, also from trench I, is described as a fishhook (Fig. 12) (ibid. 51, Pl. 36.1). The two beads have been exam­ined with different methods (Mirtsou et al. 1997; see above). Analysis showed that they were made from gold (both contained 10% silver and 0.05% copper), which was cold hammered and annealed to produce ring-pearls with a wall thickness of 0.4mm and an inner diameter of 2 and 5mm (ibid. 93). The polished but unetched structure of both gold sam­ples exhibited a number of quartz inclusions, so the artefacts were most probably made from native gold without melting. The etching with aqua regia solu­tion revealed a single-phase structure with few twin­ned crystals, indicating hammering followed by a reheating process. Sitagroi One gold bead was retrieved from the sieving of se­diments in a small sounding (ZB) adjacent to the main stratigraphic sounding ZA, and it is assigned to phase III, i.e. the 5th millennium BC (Renfrew, Slater 2003.319–320, Fig. 8.1.e, Pl. 8.4). It belongs to the narrow cylindrical type. Examination under a microscope showed that the bead was made by beating a flat piece of metal around an inner core (which was not preserved). The ends were cut dia­gonally and had tooling marks near their edges, in­dicating that an attempt was made to weld them to­gether. No further analysis has been conducted. Dikili Tash A small gold bead was found in the sector excavat­ed by Theocharis in 1961 on the east slope of the tell (sector I), but was not included in the relevant preliminary report (Koukouli, Rhomiopoulou 1992). Comparison with artefacts from other sites and from the more recent excavations at Dikili Tash itself allows its dating to the late 5th millennium BC (Tsirtsoni 2018.1276–1279, Fig. 2). The piece of gold M206 kept at the Philippi Museum is probably the one mentioned by Michel L. Séfériades (1992. 113) as coming from a coeval context from the ex­cavations by Deshayes. During the 2012–2013 excavations at Dikili Tash, four new gold objects were recovered, but this time from a perfectly secure context: a rich group of fin­ished and half-finished ornaments, tools and raw ma­terials found in the northern part of House 1, in sec­tor 6 (see description and chronology above). Two of the gold ornaments, conventionally described as cylindrical beads, were flat strips of gold rolled around smaller beads from stone, while the third, which was smaller in size, was made with the same technique but contained no beads from a different material. The fourth was a twisted band forming a ring at one end (Fig. 13). The last type had not been attested yet in Northern Greece, but is known from a number of sites (mostly cemeteries) in the Balkans (e.g., at Varna, Le premier or de l’humanité 118– 119 (grave 4), 136–137 (grave 41), 149 (grave 97); see also Tsirtsoni 2018.1280). An SEM Examination by Michael Vaveli-dis (Department of Geology, Aristotle University of Thessa- The emergence of metal use in Greek Eastern Macedonia during the Neolithic period (late 6th–5th millennia BC) made of gold with a low percentage of silver (4–9%). This points to native gold as a raw material. The sur­faces show traces of hammering and rubbing (Fig. 14). Although no direct connection has been made yet between the traces observed on the gold ornaments, or the other associated artefacts, and the tools found near them (one copper and one bone awl, one stone hammer/polisher, and several chipped-stone tools), it is clear that we see here an on-going process of manufacturing of jewels and decorative elements. Other craft activities were taking place in House 1 (e.g., decorating of pottery) together with more – and less – ordinary activities, such as storage, cook­ing and wine-making (Darcque et al. 2012–2013. 752–758; 2014.607–610; 2020.298–303). The same house yielded another unexpected find connected with gold: a complete clay vessel, a bowl decorated with brown-on-cream painted decoration, which proved to display gold ‘stains’ at several spots of its exterior surface and in the inner part of the rim (Tsirtsoni 2018. 1313, Fig. 8; originally illustrat­ed in Darcque et al. 2015.408, Fig. 7; Tsirtsoni 2016.285, Fig. 14) (Fig. 15). The bowl was standing on the floor a few meters to the south of the group of ornaments discussed above, and at short distance from the other group of ornaments that contained the copped beads (see above). The gold material seems to have been added after the original firing of the vessel, and indeed one wonders if this was done intentionally, as part of a restoration, or accidental­ Fig. 12. Dimitra, gold ‘fish­hook’ (© Ephorate of Antiqui­ties of Serres). ly, when someone working with gold paint on another object would have touched the vessel with his/her gold-stained fingers or tools. However, intentional gold-painting on pot­tery is also attested at Dikili Tash. Evidence comes from a unique sherd found in sector 2, at the southern periphery of the tell: it is a fragment from the rim and handle of a small amphora, decorated with parallel ob­lique lines (Fig. 16). Its stratigraphical posi­tion is not secure, but according to its typological characteristics (shape, fabric, firing, surface finish­ing and decorative motifs) it can be dated with cer­tainty to the years between 4800–4200 BC, and most probably after 4500 BC. This vessel type is indeed very common for this period both at Dikili Tash and the wider area, with decoration typically executed with graphite paint. Here, graphite has been replac­ed by gold, applied before firing (Tsirtsoni 2018. 1285–1288). Examples of such substitution of an or­dinary colouring material (graphite) for a less ordi­nary, and presumably precious one (gold) are known so far with certainty only from the Varna necropolis (Le premier or de l’humanité 118–125: two vessels from grave 4). Two sherds from the site of Bubanj in Serbia (Stoji., Joci. 2006.154–155, Pl. 41.b; Bu­ latovi. et al. 2020.46, 53, Pl. 6/1) and two more from Krivodol and Chirpan in Bulgaria (personal communication of our colleagues Nadezhda Todoro­va and Petur Leshtakov) could be the only other can­didates, if their dating was confirmed. Discussion Metal objects, albeit in small numbers, are attested in most sites investigated to some extent (Tab. 1). They are all either small tools (awls, pins, fishhooks, wires) or ornaments (beads, twisted bands, pen­ Dimitra Malamidou, Zoi Tsirtsoni, and Markos Vaxevanopoulos dants), and no objects with any substantial thick­ness of metal (like flat axes, shaft-hole axes or axe-adzes) have been found. Few of the metal finds came from a recognisable in situ context of use, with the exception of Dikili Tash, where a copper awl was found in House 1, sector 6, together with golden beads and a group of other finished and semi-fini­shed elements of jewellery. Although a complete metallurgical chain of produc­tion (chaîne opératoire), beginning with the raw material and ending with the finished product, was not successfully proven for any of the excavated sites, extraction from ores was attested for copper and sil­ver. Metallurgical installations are not well docu­mented in general either in the Balkans or Aegean. The recovery of a significant number of sherds or crucible fragments and associated copper deposits from phase III levels at Sitagroi is of considerable in­terest. They offer evidence of, at the least, copper melting on the site. Smelting in small shallow pit-shaped hearths opened in the clay-rich ground is probably also attested at Dikili Tash. Copper production at Promachon-Topolnitsa is among the earliest in the Balkans, dating from the end of the 6th and the early 5th millennium BC, and is documented both by stratigraphy and absolute dating. There, the production is based on simple smelting of very pure malachite, producing no slag (Koukouli, Bassiakos 2002), a method consistent with what we know of the main metal-producing techniques of the Balkan hinterland, described as ‘slagless’ or ‘nearly slagless’ (Pernicka et al. 1997; Radivojevi., Rehren 2016; Rehren et al. 2020; Ra­divojevi., Roberts 2021). The finds from Limenaria provide a more diversified picture. Items like the malachite bead show that the properties of the ores, which were collected from sources not far from the settlement, were already understood with regard to their use as pigments and The emergence of metal use in Greek Eastern Macedonia during the Neolithic period (late 6th–5th millennia BC) raw materials for other purposes. Given the long tra­dition among prehistoric communities on Thasos in the exploration of local minerals, it is not surprising that as early as the 5th millennium BC the inhabi­tants of Limenaria were trying to extract copper from the Thasian polymetallic ores (Nerantzis, Papado­poulos 2013; Bassiakos et al. 2019). The gaining of copper from such polymetallic raw materials re­quired, apart from enriching the ore with copper minerals, a considerable level of skills and experi­ence, which recalls pyro-metallurgical practices that are also attested later in the Aegean (Bassiakos, Ca-tapotis 2006). On the basis of this evidence, Thasos seemed to be part of a technological tradition with ties to the Aegean and Asia Minor, rather than to the Balkans (Bassiakos et al. 2019.2754). The recent discovery of similar remains at the late 5th millen­nium BC Akladi Cheiri, in South-East Bulgaria (Re­hren et al. 2020), invites us to reconsider the evolu­tion patterns of early small-scale metallurgy in the wider area. The fact that Limenaria, Thasos, provided firm evi­dence for extraction of silver in the early 4th mil­lennium, and possibly earlier, is of great interest. Silver objects are not known in the Balkans before the end of 4th millennium (Nikolova 1999.303–308; Alexandrov 2009; 2018), whereas they appear ear­lier and are more common in the Aegean and south­ern Greece (e.g., Cyclades, Attica, Dimakopoulou 1998.64–65; Zachos 2010.89; see also Tsirtsoni 2014.295). Silver is rarely found as a native mineral element. Unlike gold, it is rarely found in significant amounts in placer deposits. The principal sources of silver are the ores of copper, copper-nickel, lead, and lead-zinc. Lead and silver extraction was attested through compositional data and microstructural exa­mination of fragments of litharge and the silver pin found at Limenaria. The availability of resources lo­cally was an important factor for the early exploita­tion of silver ores by Neolithic people. Nevertheless, given the fact that Thasos is open to southern influ­ences in many regards, sharing of Aegean technolo­gical traditions must also have played an important role. Gold objects are found in mainland sites of Eastern Macedonia only as beads in the form of hammered sheets rolled in a cylindrical shape, while hammered sheets of other shapes and ring-shaped pendants are absent, perhaps because objects of this type are more common in tombs (e.g., Aravissos, central Macedo­nia, Zachos 2010.89; 1996b.167, 339–340; Thessaly, ibid. 167, 339). All the above have close typological similarities with the ones from southeast Europe, which were found in large quantities in tombs, main­ly in the extensive cemeteries of Bulgaria. Gold paints in Dikili Tash show that the Late Neoli­thic inhabitants of the site were familiar with this metal. They were using it not only as precious mate­rial for jewellery making, but also in the chaîne opé­ratoire of decorating pottery, in the same way as graphite and other minerals. This fact provides evi­dence for close relations with the pyrotechnology traditions of the Balkans. Evidence for the extraction of metals from mining sites is lacking at the moment. Placer gold areas and alluvial gold washeries are difficult to distinguish and locate after centuries of erosional processes. Never­theless, bearing in mind the abundance of minerali­zation in the area, one can assume that some prehi­storic hard rock mining locations, which survived later mining activities, must exist. Therefore, and de­spite difficulties, locating traces of prehistoric min­ing and identifying the origin of the various raw ma­terials and their relationship to the finished objects should be targeted, and ongoing research (Vaxeva­nopoulos 2017a; Vaxevanopoulos et al. 2021) is in engaged in this. Conclusion Despite the relatively modest nature of the indivi­dual metal objects retrieved from Eastern Macedo­nian settlements, their widespread presence and co­herent picture add significant data about the devel­opment of metal production in the North Aegean and more broadly in the Balkans. The fact that they come from secure stratigraphic contexts allows their dating to be firmly established, whereas their inte­gration in considerably long sequences such as those Settlement Copper Artefacts Silver Gold Artefacts Ore and Metallurgical Dating References Artefacts Findings Promachon-malachite beads, 1 crucible, hollows with remains of First half of the Koukouli-Chrysanthaki et al. 2007 Topolnitsa 1 pendant, fragments burnt clay, slags 5th mill. BC Dimitra 21 beads, 1 fishhook, 2 beads, 1 fishhook polymetalic ore fragments Second half of 6th Grammenos 1997< Mirtsou et al. 1997 5 parts of pins or wires, to 5th mill. BC 11 undefined fragments Kryoneri 3 pins\awls, copper ore fragments Mid-5th mill. BC Malamidou 1997< 2007< 20162 fragments of rings Sitagroi 1 roll-headed pin, 1 bead oval crucible, copper fragments of lumps, Late 6th to 5th Renfrew, Slater 20031 awl, five beads sherds with copper incrustation mill. BC Dikili Tash 4 beads, 10 awls or pins, 4 beads, 1 twisted copper oxides and copper fragments, End of 6th to late Séfériades 1992< Koukouli, Rhomiopou­2 unstratified roll-headed band, 1 unidentified, 1 burnt sherd with copper traces, 5th mill. BC lou 1992< Darcque et al. 2015< 2020< pins, parts of 5 more pins 1 ceramic vessel with 1 sherd with copper incrustation, Tsirtsoni 2016< 2018or awls, 1 flat sheet gold stains, 1 gold-1 clay mould, 1 clay-plastered pit painted sherd Paradeisos 2 needles and fragments Late 5th mill. BC Hellström 1987 Kastri Theolo-2 pins Mid-5th mill. BC Koukouli 1992gos (Thasos) Limenaria 1 malachite bead 1 pin 20 slags and fragments of End of 6th to early Papadopoulos, Malamidou 1997< 2008< (Thasos) hematite\limonite lumps 4th mill. BC Papadopoulos 2008< 2012 a case study on a hoard from Dezmir, Romania Oliver Dietrich1, Laura Dietrich2, and Botond Rezi3 o.dietrich@fu-berlin.de 1 Landesamt für Denkmalpflege und Archäologie Sachsen-Anhalt, Halle (Saale), DE 2 Österreichisches Archäologisches Institut, Vienna, AT< laura.dietrich@oeaw.ac.at 3 Muzeul Judet‚ean Mures,, Târgu Mures,, RO< reziboti@yahoo.com ABSTRACT – Large ‘scrap’ metal hoards have so far been seen as a characteristic of the older Urn-field Period in the eastern Carpathian Basin. Using the hoard from Dezmir, Romania, as a starting point, this paper describes a group of hoards with chronologically and spatially diverse components. These hoards were hidden during the younger Urnfield Period and represent smaller variants of the older, large ‘scrap’ hoards. Earlier research has used them to define a chronological horizon (‘Phase 3-Jupalnic-Turia-Ha A2’). Based on archaeological observations, spatial data and a correspondence analysis, we argue that they represent a specific hoarding practice instead, characteristic of Phase 4, i.e. the younger Urnfield Period. KEY WORDS – Carpathian Basin; Bronze Age; hoards; chronology Depoji ‘kovinskih odpadkov’ iz pozne kulture /arnih grobi[; v Karpatski kotlini> [tudija depoja iz Dezmirja v Romuniji IZVLE.EK – Doslej je veljalo, da so veliki depoji ‘kovinskih odpadkov’ zna.ilnost starej.e kulture .ar­nih grobi.. v Karpatski kotlini. S pomo.jo depoja v Dezmirju v Romuniji predstavljamo skupino kro­nolo.ko in prostorsko razli.nih depojev. Depoji so bili skriti v mlaj.i kulturi .arnih grobi.. in pred­stavljajo manj.e razli.ice starej.ih depojev ‘odpadkov’. Z njimi so v preteklih raziskavah opredelili kronolo.ki horizont ‘faza 3-Jupalnic-Turia-Ha A2’. Na osnovi arheolo.kih opazovanj, prostorskih podatkov in koresponden.nih analiz trdimo, da predstavljajo depojsko prakso, zna.ilno za 4. fazo, torej mlaj.e obdobje kulture .arnih grobi... KLJU.NE BESEDE – Karpatska kotlina; bronasta doba; depoji; kronologija The hoard Dezmir (com. Apahida, jud. Cluj) is situated approx. sor’ are mentioned (Crisan et al. 1992.183–186). 2km ESE of Cluj-Napoca at the edge of a valley The latter saw some small-scale excavation in the formed by the Za¢podie, a small tributary of the So-early 1960s (Popescu 1964.556–557, Nr. 44). What mesul Mic. The repertory of archaeological finds is entirely missing in this repertory though is infor-from the district of Cluj lists single finds of stone mation on a hoard (Fig. 1) that was discovered at axes, an Eneolithic settlement, another one of the roughly the same time. Cotofeni Culture, Scythian and Celtic graves as well as Roman buildings; for the Bronze Age, a single The hoard was first published by Mircea Rusu in find of pottery and a settlement in a plot called ‘Ta¢u-the frame of the ‘Inventaria Archaeologica’ series DOI> 10.4312\dp.49.17 ‘Scrap metal hoards’ of the Later Urnfield Period in the Carpathian Basin> a case study on a hoard from Dezmir, Romania (Rusu 1977a). There he mentions that the hoard was accidentally discovered on 29th August 1964, while deep ploughing a plot called ‘Bocomaia’ some 400m to the west of the Somes depression. A short time after the discovery, Rusu and Viorica Pintea conducted excavations at the findspot, leading to the discovery of several more bronze artefacts and ‘Hal­lstatt pottery’.1 About 50m to the west of the hoard, a pit house dating to the same period was found. Whether the hoard was originally placed within a pottery vessel remains unclear, as the publication does not indicate the exact findspot of the illustrat­ed pottery fragments. Rusu further mentions that the hoard was discovered next to the wall of a Ro­man villa rustica, which could indicate that the ori­ginal context was disturbed to some degree. The finds are kept today in the Muzeul National de Isto­rie a Transilvaniei (MNIT) Cluj-Napoca. Given the excavations at the findspot, it seems pro­bable that all preserved items of the find were reco­vered. The hoard was hidden next to or in a (small?) lowland settlement, within a settlement cluster that also included larger sites in dominant positions. Dez­mir-Ta¢usor, excavated in 1963, is located on a pro­montory (Popescu 1964.556–557, Nr. 44). A closer examination of this Bronze Age microregional sys­tem of different settlement types and hoards is un­fortunately not possible at the moment. The finds from Dezmir-Ta¢usor have never been published, and the plot ‘Bocomaia’ cannot be exactly located based on the information provided by Rusu. The Somesul Mic flows to the north of Dezmir. Following Rusus’s description (“400m to the west of the Somes depres­sion”), this would mean that the findspot was locat­ed to the northwest of the village, as the neighbour­ing village Sânnicoara¢ is located immediately to the northeast. However, within this area no plot of land called ‘Bocomaia’ could be identified. The bronze finds Thirty-six bronzes and one lead artefact could be do­cumented at the MNIT Cluj-Napoca (Fig. 1).2 All ob­jects underwent conservation procedures and the ones made of bronze now have a brownish to olive-green surface with patches/remains of light blue pa­tina. 1. Beaked socketed axe with a nearly symmetrical mouth and a loop placed slightly below the rim. Both broadsides are uneven, likely due to gas bub­bles trapped in the mould during casting; one has a hole caused by a casting fault. Excess metal was left in the area of the loop and casting seams are visible on the lateral sides. Traces of hammering are visible on the blade; the cutting edge is blunt and has small notches. Registration number MNIT Cluj-Napoca P59.547. Length (l.) 12.5cm, depth of socket (d/s.) 8.7cm, width of blade (w/b.) 5.25cm, weight (w.) 322g (Fig. 2.1). 2. Beaked socketed axe with a loop. Traces of the casting jet are visible on the upper part of the loop. Casting seams are detectable on the upper part of the lateral sides, in the blade area they were level-led through hammering. The cut­ting edge is blunt and has large nicks. The surface has adhesions of earth. Registration number MNIT Cluj-Napoca P 59.548. L. 11.6cm, d/s. 8.3cm, w/b. 4.6cm, w. 256g (Fig. 2.2). 3. Socketed axe with a sharp-edged mouth, loop and curved body. Below the rim there are two horizontal parallel ribs; one broadside has a casting fault (a large hole), the other one has a hole below the rim. There is unre-moved excess metal on the loop, Fig. 1. The hoard from Dezmir (photo B. Rezi). traces of ground casting seams 1 In Romanian terminology, meaning Ha A-B. 2 The artefacts were documented by Botond Rezi. Oliver Dietrich, Laura Dietrich, and Botond Rezi are detectable on the upper part of the lateral sides, while below they have been removed through hammering and use-wear. The cut­ting edge has notches and the cor­ners have broken away. The sur­face is uneven due to gas bubbles trapped in the mould during cast­ing. Registration number MNIT Cluj-Napoca P 59.549. L. 9.4cm, d/s. 5.2cm, w/b. 4.4cm, w. 116g (Fig. 2.3). 4. Beaked socketed axe with a loop and curved body; the trape­zoid blade is separated from the body by a low ridge. The raw cast has not been overhauled, the two halves of the moulds have acci­dentally moved slightly in the ver­tical direction during casting. One broadside is uneven due to gas bubbles trapped in the mould and has a casting defect (an irregular hole). The casting jet was remov­ed, the casting seams are clearly vi­sible. Registration number MNIT Cluj-Napoca P 59.550. L. 13.2cm, d/s. 7.5cm, w/b. 4.8cm, w. 261g (Fig. 2.4). 5. Beaked socketed axe with a curved body, the loop mostly bro­ken off. The broadsides, particu­larly in the blade area, have nume­rous casting defects. The casting jet was carefully removed from the loop, the blade area was intensive­ly hammered, leading to the for­mation of a ridge between blade and body. The cut­ting edge is sharp and has small notches. Registra­tion number MNIT Cluj-Napoca P. 59.551. L. 12cm, d/s. 7.2cm, w/b. 4.5cm, w. 230g (Fig. 3.5). 6. Beaked socketed axe with a loop and curved body; a large portion of the socket area, including the beak, has broken off. The casting jet has been removed from the loop and the contact point ham­mered. The casting seams are visible on the upper two-thirds of the narrow sides; one broadside is un­even due to gas bubbles trapped in the mould. The cutting edge is uneven, with small notches. Registra­tion number MNIT Cluj-Napoca P 59.552. L. 12.8cm, d/s. 8.5cm, w/b. 4.8cm, w. 262g (Fig. 3.6) 7. Beaked socketed axe with a loop and pronoun­cedly curved body. The contact point of the casting jet is visible on the top of the loop, the casting seams are visible on the upper two-thirds of the nar­row sides; below they have been removed. A large part of the cutting edge is missing in the centre. Re­gistration number MNIT Cluj-Napoca P 59.553. L. 13.3cm, d/s. 8.4cm, w/b. 6.1cm, w. 381g (Fig. 3.7) 8. Socketed axe with a thickened rim, without a loop. The parallel narrow sides broaden only in the lower quarter of the body into a narrow blade. On one broadside an ornament composed of four chev­rons hanging from the rim is preserved, the other broadside has only faint traces of a similar orna­ ‘Scrap metal hoards’ of the Later Urnfield Period in the Carpathian Basin> a case study on a hoard from Dezmir, Romania ment. No casting seams are detectable on the nar­row sides, the cutting edge is rounded and has seve­ral notches. The surface is rough. Registration num­ber MNIT Cluj-Napoca P 59.554. L. 12.5cm, d/s. 8.3cm, w/b. 4.6cm, w. 321g (Fig. 3.8) 9. Socketed axe with a thickened facetted rim and loop, the body is slightly curved. Both broadsides have three parallel horizontal ribs, one curved rib and below that a Y-shaped ornament, with a flexed rib on each side. The two halves of the moulds accidentally mov­ed slightly horizontally during casting; the casting seams are vi­sible on the upper two-thirds of the narrow sides; below they have been removed. A ridge di­vides the socket and blade, the cutting edge is jolted and has a large section with old damage. Registration number MNIT Cluj-Napoca P 59.555. L. 14.8cm, d/s. 9cm, w/b. 5.6cm, w. 518g (Fig. 4.9). 10. Socketed axe, due to a cast­ing defect a part of the socket, in­cluding the loop, is missing. The parallel narrow sides evolve in­to a trapezoid blade. Two hori­zontal parallel ribs are located below the mouth of the socket, followed by pseudo-wings and a ridge. Traces of the casting seams are visible on the upper two-thirds of the narrow sides; below they have been removed. The surface is smooth, the cutting edge is sharp with only small notches and dents. Registration number MNIT Cluj-Napoca L. 9.2cm, remaining d/s. 5.1cm, w/b. 4.5cm, w. 115g (Fig. 4.10). 11. Socketed axe with a thick­ened rim, the loop broken off, and the body slightly curved. A remainder of the casting jet is vi­sible on the socket mouth above the remains of the loop. One broadside has a ridge between the socket area and blade. The casting seams are visible on the narrow sides, the cutting edge is blunt with notches, and one edge is missing. The surface is rough; the blade in particular has numer­ous defects from gas bubbles trapped during casting. Registration number MNIT Cluj-Napoca P 59.567. L. 9.5cm, d/s. 5.1cm, w./b. 3.7cm, w. 96g (Fig. 4.11). 12. Lower half of a socketed axe. The breaking edges and the broadsides show traces of hammer blows. Faint traces of a decoration comprising of a ‘Y’ flanked by two flexed ribs on each side (similar to Oliver Dietrich, Laura Dietrich, and Botond Rezi Nr. 9) are visible. Traces of the casting seams are detectable on the narrow sides; the cutting edge is asymmetrical. The sur­face is very uneven due to gas bubbles trapped during casting. Registration number MNIT Cluj-Napoca P 59.565. L. 7.6cm, d/s. (preserved) 3.1cm, w/b. 4.95cm, w. 187g (Fig. 4.12). 13. Blade of a socketed axe, tra­ces of hammer blows on the broadside at the break. During casting, the core was out of cen­tre and the two halves of the mould shifted slightly horizon­tally. The narrow sides preserve the casting seams, the cutting edge is uneven. The surface is slightly uneven due to gas bub­bles trapped during casting. Re­gistration number MNIT Cluj-Na­poca P 59.568. L. 6.4cm, d/s. (preserved) 2.5cm, w/b. 4.8cm, w. 148g (Fig. 4.13). 14. Fragment of an axe, lead. The shape and proportions indicate a socketed axe. However, as only the lower two-thirds are preserv­ed and there is no indication of a socket, the object likely was massive. Traces of the casting seams are detectable on the nar­row sides; the dark grey surface is uneven and shows white deposits. Registration number P 59.569. L. 8.4cm, body diameter 4x1.4cm, w/b. 5cm, w. 358g (Fig. 4.14). 15. Fragment of a knob sickle (“Knopfsichel”). The hafting area and a large portion of the sharply curv­ed blade are preserved. The base of the hafting area is slightly bent inwards and features a conical knob with a round cross section. Two oblique ribs follow­ed by an antithetic rib pair are only faintly visible. The thickened rim of the blade is followed by a rib. The cutting edge has deep notches and nicks. Regis­tration number P 59.556. L. 10.9cm, width 2.8cm (Fig. 5.15). 16. Fragment of a hook sickle (“Hakensichel”) with a thickened rim. A casting defect (hole) did not pre­vent use, as the damage on the cutting edge shows. Registration number P 59.557 (MNIT), 59.558 (Ru­su). L.7.4cm, width 2.4cm (Fig. 5.16). 17. Knob sickle, the point is missing. The base of the hafting area is slightly bent inwards and features a conical knob with a round cross section. The thick­ened rim of the blade is followed by two ribs. The cutting edge is heavily notched. Registration number P 59.560 (MNIT) 59.563 (Rusu). L. 14.1cm, width 2.2cm (Fig. 5.17). 18. Blade of a knob sickle with heavily thickened rim, which is followed by two ribs. The casting jet is still preserved, the cutting edge has several large notches. Registration number P 59.558 (MNIT), 59.557 (Rusu). L.12.6cm, width 2.5cm (Fig. 5.18). 19–20. Two fitting blade fragments of a hook sickle with a thickened rim. Registration number P 59.559. ‘Scrap metal hoards’ of the Later Urnfield Period in the Carpathian Basin> a case study on a hoard from Dezmir, Romania Fig. 5. Sickle frag­ments (15–23), fragments of bra­celets (25–26) and a plano-con­vex ingot (36) from Dezmir (photo B. Rezi). ribs are added. The cutting edge shows several and quite regular notches. Registration number P 59.562 (MNIT), P. 59.560 (Rusu). L. 6.3cm, width 2.2cm (Fig. 5.24). 25. Fragment (approximately half) of a bracelet with a round cross section and pointed ends. Towards the end the surface is damaged and shows traces of ornamentation with regularly dis­posed vertical lines. Registration number P 59.563. Max. diameter 7cm, w. 29g (Fig. 5.25). 26. Fragment (approximately half) of a bracelet with rhomboid cross section and pointed ends. Towards the end the surface is damaged. Registration number P 59.564. Max. diameter 9.7cm, w. 58g (Fig. 5.26). 27. Irregular fragment of a plano­convex ingot (PCI),3 bronze. Re­gistration number P 59.570. L. 2.7cm, thickness 1.6cm, w. 106g (Fig. 6.27). 28. Fragment of a flat/flat PCI, bronze. Registration number P 59.571. L. 4.7cm, width 2.8cm, L. 5.5cm, width 3.4cm. Registration number P 59. 562b. L. 3.2cm, width 3.8cm (Fig. 5.19–20). 21–22. Two fitting fragments of a knob sickle. The base of the hafting area is rounded and features a conical knob. The rib runs along the edge, followed by a smaller, parallel one. The preserved part of the blade is straight, traces of several hammer blows are visible on the cutting edge. Registration number P 59.561, P 59.562a. L. 6.2cm, width 2.5cm. L. 3.3cm, width 2.1cm (Fig. 5.21–22). 23. Fragment of a knob sickle, the curved middle part is preserved. Registration number P 59.562c. L. 6.5cm, width 2.4cm (Fig. 5.23). 24. Hafting area of a knob sickle with an inwards bent base and small knob. Below the thickened rim runs a small rib; at the base, three parallel inclined w. 447g (Fig. 6.28). 29. Irregular flat/flat PCI. Registration number P 59.573. L. 7.1cm, thickness 0.9cm, w. 130g (Fig. 6.29). 30. Irregular fragment of a flat/flat PCI, bronze. Re­gistration number P 59.574. L. 5.8cm, thickness 3.3cm, w. 813g (Fig. 7.30). 31. Fragment of a presumably sub-rectangular, flat/ convex PCI with a very large cavity. Registration number P 59.575. L. 6.4cm, width 2.8cm, w. 348g (Fig. 7.31). 32. Fragment of a likely drop-shaped convex/con­vex PCI, bronze. Registration number P 59.576. L. 5.8cm, width 3.9cm, w. 628g (Fig. 7.32). 3 The description of the PCIs follows the terminology proposed by Modl (2019.380–381, Figs. 6–7). Oliver Dietrich, Laura Dietrich, and Botond Rezi 33. Complete round flat/convex PCI. Determined as copper after the colour of the metal. Registra­tion number P. 59.577 (MNIT) 59.580 (Rusu). Diameter 8.9cm, thickness 2.5cm, w. 3950g (Fig. 6.33). 34. Fragment of a concave/convex PCI, bronze. Registration num­ber P 59.578. L. 5.3cm, thickness 5.9cm, w. 866g (Fig. 7.34). 35. Fragment of a presumably ir­regular convex/convex PCI with a very large cavity, determined as copper after the colour of the me­tal. Rusu published a concave cross section drawing, the re-do­cumentation shows clearly that the object is hollow and roughly half of it is preserved. Registration number P 59.579. L. 10.7cm, thick­ness 5.8cm, w. 1008g (Fig. 7.35). 36. Complete slightly ovoid flat/ convex ingot. Registration number P 59.580 (MNIT), P 59.577 (Rusu). Diameter 19.6cm, thickness 3.4cm, w. 740g (Fig. 5.36). 37. Bronze ingot, bar/rod-shaped with a triangular cross section. Excess bronze on both narrow sides hints at casting in an open mould. Registration number P 59.572. L. 11.6cm, cross section 2.6x1.5cm, w. 133g (Fig. 6.37). A total of 14 socketed axes, 10 sickle fragments, two fragments of bracelets, 10 plano-convex ingots and one bar-shaped ingot can be recorded for the hoard from Dezmir. These numbers differ slightly from earlier publications. Rusu published 14 socketed axes, nine sickle fragments (seven illustrated), two bracelets, one bar-shaped and ten plano-convex in­gots (Rusu 1977a).4 Mircea Petrescu-Dîmbovita (1977.122–123) lists, without explanation, 14 soc­keted axes, nine sickle fragments, two bracelets, a ‘casting residue’ (“un rest de turnat/Gussrest”: the 4 The find was first discussed in Rusu’s (1972.553–554, Nr. 84) doctoral thesis, which remained unpublished. The illustrations are the same while some of the information differs. Rusu states that the findspot was a slight slope and that the plough would have reached a depth of 70–80cm. The find was discovered in a depth of 60cm and two pit houses were discovered at a distance of 50m from the findspot. The content of the hoard is also listed differently: “6 celturi cu plisc, un celt masiv cu decor, un celt fa¢ra¢ toarta¢ si decor, 2 celturi zvelte, decorate cu cîte doua¢ linii paralele reliefate, un celt zvelt cu aripioare false, 2 fragmente de celturi, dintre care unul cu decor; un fragment de celt din plumb; 10 fragmente de seceri dintre care 4 cu butoni si 6 de ti-puri neprecizate, 2 fragmente de bra¢ta¢ri nedecorate cu sectiunea rotunda¢ sau ovala¢, o bra¢tara¢ cu sectiunea plan-convexa¢, 14 turte de bronz si mai multe fragmente de vase de lut hallstattiene.” As the thesis clearly has the character of a manuscript, it seems likely that Rusu corrected this account for publication. However, some discrepancies remain, and are hard to explain as the published account states that Rusu personally knew the findspot. ‘Scrap metal hoards’ of the Later Urnfield Period in the Carpathian Basin> a case study on a hoard from Dezmir, Romania term is open to interpretation) and 13 ingots.5 He illustrates five of the sickles already published by Rusu; the ingots and the ‘casting residue’ are not illustrated (O.c.122–123, Pl. 288.11–17, 289.1–7). In the German variant of his monograph, the num­ber of ingots is corrected to ten, and all nine sick­le fragments are illustrated (Petrescu-Dîmbovita 1978.137–138, Nr. 198, Pl. 217E, 218; sickles Pl. 218.15–23). The identification of the bar-shaped object nr. 37 as an ingot or a casting residue is a matter of inter­pretation, although we share Rusu’s opinion and see it as a bar-shaped ingot. The dis­crepancies regarding the number of sickles are harder to explain. In Rusu’s first publication of the hoard, several captions for sick­les and bracelets on page R65b got mixed up. Rusu states that four fragments of one sickle (in­ventory numbers 59.561–59.562 a,b,c) were part of the hoard, but only one fragment (P. 59.561) is illustrated. Research on the orig­inals showed that only the frag­ments P 59561 (Nr. 21) and P 59562a (Nr. 22) are from the same sickle, while P 59562c (Nr. 23) is at least not a directly fit­ting piece, and P59562b (Nr. 20) fits with P 59559 (Nr. 19). In his 1978 monograph, Petrescu-Dîm­bovita illustrated nine sickle fragments. Research at the MNIT Cluj-Napoca produced one more fragment, making a total of 10 fragments from at least seven, possibly eight sickles (if Nr. 23 is not from the same artefact as Nr. 21–22). There is also a contradiction in the total number of hoarded ob­jects. In 1977, Rusu mentioned a total of 43 artefacts with a weight of 13.114kg and published 34 bronzes with illustrations. A sim­ple explanation in the sense of transposed digits is doubtful, as he further lists the mentioned sickle fragments with­out illustrations. Additionally, there is an unexplain­ed gap in the inventory numbers between Nr. 59.564 and 59.571. Of course, this is not proof that there are missing objects. The 37 objects listed above are all that are stored at the MNIT. If the sickle frag­ments are assembled, the total number of artefacts would be 34. On the level of particular objects, the renewed ana­lyses of the artefacts reveals a few differences com­pared to the published record. The socketed axe Nr. 8 has to date been described as plain (Rusu 1977a. 5 Why Petrescu-Dîmbovita mentions 13 ingots remains unclear. However, in his doctoral thesis Rusu mentions 14 ingots, illustrat­ing 10. If Petrescu-Dîmbovita used that source (which he does not cite) or other information from Rusu, the number could be explained by subtracting the ‘casting rest’ from the number of ingots. Oliver Dietrich, Laura Dietrich, and Botond Rezi Pl. R65a.8; Petrescu-Dîmbovita 1977.Pl. 288/16). However, it has traces of chevron-shaped ornaments on one broadside. Socketed axe Nr. 2 has a clear ridge between the blade and socket area in the pub­lished drawings (Rusu 1977a.Pl. R 65a.2; Petrescu-Dîmbovita 1977.Pl. 288.12), which does not exist on the original object. The bracelet Nr. 25 has been published as undecorated (Rusu 1977a.R65b.12; Petrescu-Dîmbovita 1998.62, Nr. 487, Pl. 48.487), but there are indeed traces of ornaments. The plano­convex ingot Nr. 35 has been drawn and described as ‘massive’ (cf. Rusu 1977a.Pl. R65c.34; Petrescu-Dîmbovita 1978.Pl. 218/36), although it is clearly hollow. In summary, most of the artefacts are fragmentary or damaged. The socketed axes have a striking num­ber of casting defects. However, as notches, break­ages and deformations of the cutting edges show, all of them have been used except for the unfinished cast Nr. 4. The degree of use (ABN), a measure for the intensity of the use of socketed axes (Dietrich 2021a.155–156, 162–164, Fig. 7/14), points in the same direction. Five degrees of use can be deter­mined using the ratio between blade and socket length between ABN 0 (raw casts) and ABN 4, mean­ing an artefact whose cutting edge has been moved nearly up to the socket by continuous use, render­ing the axe unusable.6 Applying this scheme to the axes from Dezmir, it becomes clear that most of them could still have been used.7 This is different for the sickles, which are heavily damaged or in fragments. Several fragments have heavily damaged cutting edges, and for Nr. 24 in particular the unifor­mity of the notches could hint at intentional damag­ing. In addition to two complete ingots, the hoard contains eight fragments. Using the terminology es­tablished by Bianka Nessel (2014.404–405) for in­got segments, they are mostly amorphous, some nearly square. All can be characterized as small seg­ments of large plano-convex ingots, none of the frag­ments fit together. Of particular interest are the two hollow fragments Nr. 31 and 35. Following Daniel Modl (2019.381), these cavities are too big to be caused by gas distention during casting of the in­gots. One plano-convex ingot from the hoard of Mi-ljana, op.. Klanjec, Croatia (Dörfler et al. 1969.69– 72, Pl. I.1-2) has a lead core instead of the cavity. Modl assumes that this could also have been the case for the quite rare ‘hollow’ plano-convex ingots (the same idea already in Hansen 1994.231). The reason for these poly-metallic ingots remains un­clear. However, this practice could be interpreted as an attempted fraud within the trade for raw bronze, or as a way to manufacture less valuable votive items (Hansen 1994.230–231). In any case the fraud would have been easily detectable when the ingots were partitioned. The fragment of a lead axe The socketed axe made of lead (Nr. 14, Fig. 4.14), preserved fragmentarily, cannot have been intend­ed for use, judging from the soft material and the missing socket. While lead is known from several finds in the Carpathian Basin (Boroffka et al. 2016), the lead axe is unique so far. The only distant com­parison is a socketed axe from the first hoard from Várvölgy, Zala/Veszprém megye, Hungary, with a lead content of 60.53% (Müller 2006.15, Fig. 7). The axe from Dezmir has been described as an axe-shaped ingot (Boroffka et al. 2016.409). The shapes of ingots in the Carpathian area are quite standar­dized, though (Modl 2019; Nessel 2019.185–204), and axes are not among them. Lead socketed axes are however known in larger numbers from Western Europe, and their find con­texts offer another possible interpretation. A bronze casting mould from Cambridge-New Street contained a fragment of a lead socketed axe (Lawson 1979. 178; Needham, Hook 1988.274, App. 2.1); another find is known from Southall-Brickfield (Gowland 1901.368, Nr. 2; Britton 1960.Pl. GB51.1a-b). A casting mould from a hoard discovered on the Isle of Harty, Kent, had lead adhesions (Evans 1881. 441–442), and Needham and Hook have published a total of seven finds of lead in casting moulds, in­cluding the above mentioned axes (Needham, Hook 1988.App. 2). There are further instances of lead axes appearing without associated casting moulds (see list in Dietrich 2011), but this association was the basis for their interpretation. Ronald F. Tylecote (1962.125–128) saw the lead axes as cores that took the role of wax cores in a procedure similar to lost wax casting. Ernst Foltz (1980) has argued against the feasibility of this tech­ 6 Degree of use (“Grad der Abnutzung”, ABN). 0-raw cast, unused. 1-light: Blade socket ratio only slightly reduced; blade is more than 1/3 of the whole axe. 2-medium: the blade is reduced to nearly 1/3 of the whole axe. 3-heavy: the blade is reduced to 1/4 or less of the whole axe. 4-extremely heavy: the blade is reduced up to the socket area. The axe is unusable. 7 As far as determinable: Nr. 1: ABN 3; Nr. 2: ABN 2; Nr. 3: ABN 1; Nr. 4: ABN 0; Nr. 5 ABN 2; Nr. 6: ABN 2–3; Nr. 7: ABN 1; Nr. 8: ABN 2; Nr. 9: ABN 1; Nr. 10: ABN 1; Nr. 11: ABN 1. ‘Scrap metal hoards’ of the Later Urnfield Period in the Carpathian Basin> a case study on a hoard from Dezmir, Romania nological process, pointing out that lead lumps would remain in the mould when the molten bronze was poured in. Additionally, socketed axes were not produced by lost wax casting, as the casting seams on all axes known from Eastern and Western Europe demonstrate. Although they accept these arguments, Needham and Hook still do not completely exclude the possibility of lead cores (Needham, Hook 1988. 265–268). An alternative interpretation has been proposed though, starting from numerous lead artefacts of the first millennium AD (for a summary see Bergen 2005.26–37): a multiphase technological process to produce clay or sand casting moulds from lead mo­dels (Drescher 1978.97–98; Foltz 1980). Klaus Gold-mann (1981) has argued that casting in sand moulds may have been the most important technique of Bronze Age casters, as there are only very few stone or bronze moulds preserved. Although this seems debatable, there is good evidence for the use of bronze models in mould production for southeast­ern Europe (Dietrich 2011). There­fore, a similar role seems possible for the lead axe from Dezmir. The pottery As (preserved/recovered) ceramic vessels associated with hoards are relatively scarce in the Carpathian Basin, a closer look at the fragments published by Rusu8 is of interest (Rusu 1977a.Pl. R65d, Nr. 35–46). The following 12 fragments were published (the numbers correspond with Fig. 8): 1. Rim sherd, undiagnostic; 2. Rim sherd of a coarse, bellied ves­sel with flaring rim, decorated with a (damaged?) knob; 3. Rim sherd of a fluted shallow bowl with fluted rim. The sherd is positioned incorrectly in Rusu’s drawing, suggesting a large ves­sel; 4. Body sherd with handle, decorat­ed with a ridge with finger im­prints; 5. Body sherd, undecorated, likely part of a large vessel with cylin­drical neck; 6. Undiagnostic bottom sherd; 7. Rim fragment of a globular fluted bowl with slight­ly incurved rim (“Einzugsschale”: Pankau 2004. 62, Typentafel IV, kalottenförmige Schalen Va-riante 3); 8. Globular body of a large bowl, the flaring neck is lost, undecorated; 9. Fragment of a broad handle (“Bandhenkel”); 10. Fragment of a fluted bowl (?); 11. Undiagnostic bottom sherd; 12. Neck and globular body of a vessel with coni­cal neck (“Kegelhalsgefäß”). The neck has hori­zontal, parallel fluting, the body arcuate fluting (“konzentrisch übereinandergelegte Halbkrei­se, die sich an spitz nach oben gerichteten Um-bruchstellen zu einer Bogenreihe schließen”: Pankau 2004.71, Pl. 5.A2a). Three of the twelve pottery fragments are diagnos­tic for typo-chronological questions (Nr. 3, 7 and 12). The remaining nine are too heavily fragment­ed (Nr. 1, 4, 6, 9, 10, 11) or are part of types wide- 8 The pottery was not available for study at the MNIT Cluj-Napoca at the moment the bronzes were documented. Oliver Dietrich, Laura Dietrich, and Botond Rezi ly distributed in time and space (Nr. 2, 5, 8). The vessel with a conical neck Nr. 12 has analogies in the so-called Medias-Reci Group of the Gáva Culture, for example in the settlements of Medias-Cetate, jud. Sibiu (Layers I and II: Zaharia 1965.Fig. 9.2, 6, Fig. 10.1, 6; Pankau 2004.Pls. 2.1, 12.6, 15.2) and Re-ci, jud. Covasna (Zaharia 1965.Fig. 11.10, 12). There are further analogies from the so-called Gra¢nicesti Group of the Gáva Culture, for example from Gra¢-nicesti, jud. Suceava (László 1994.Fig. 25. 1–4, Fig. 26, Pl. II.3A and Pl. XII.16, with ‘closed’ bows). Layer II of the multi-stratified settlement of Teleac, jud. Alba, has more developed variants of the same type (Vasiliev et al. 1991.Fig. 32.1, 9, Fig. 41/ 3, 5). The vessels Nr. 3 and particularly Nr. 7 also have good comparisons in Teleac (Layers II and III; Vasiliev et al. 1991.Figs. 31.4, 34.3,6,10,13, 35.6,8,11). The dates for Layers I and II at Medias and Reci lie between Ha B1-B3 (Zaharia 1965.102; Vasiliev et al. 1991.128) A. László has proposed a start date for Gra¢nicesti already in Ha A (László 1994.93–94); la­yer II of Teleac was dated to Ha B2-C, layer III to Ha D by Vasiliev (Vasiliev et al. 1991.118–129; but see also the modifications by Ciugudean 2012). In short, the best chronological indicator for the pottery is provided by vessel Nr. 12, which has good analogies in Medias and dates broadly to Ha B. The impor­tance of this chronological hint for the hoard is les­sened by the fact that Rusu did not clearly state which fragments were found directly associated with the hoard, or in the house at a distance of about 50m. A framework for the spatial-chronological ana­lysis of the hoard from Dezmir Rusu placed the hoard of Dezmir at the border of the hoarding horizons Jupalnic-Turia (Ha A2) and Moigrad-Ta¢uteu (Ha B1), while Petrescu-Dîmbovita dated the find clearly into the Jupalnic-Turia horizon (Rusu 1977a, Pl. R65d.7; Petrescu-Dîmbovita 1977. 122–123). Meanwhile, doubts have been raised about the existence of an equivalent of the central European phase Ha A2 within the Carpathian Basin (Hansen 1996; Tarbay 2015; Dietrich 2021a). Ad­ditionally, general methodological issues regarding the fine dating of hoard finds have been highlighted. Many hoards represent collections of metalwork ac­cumulated over a longer period of time (e.g., Vachta 2016; Hansen 2019), and thus the depth of time re­presented by the artefacts can be extensive. There are, for instance, Middle Bronze Age shaft hole axes from hoards like Dipsa (Ciugudean et al. 2006.37. 22, Nr. 185, Pl. XXXVIII.2, XLVII.2), Gusterita II (Reissenberger 1872.14–15, Pl. II.119 ) and Spa¢lna-ca II (Petrescu-Dîmbovita 1978.Pl. 140B.31 0 ), hoards that have conventionally been dated to Ha A1 (hoard horizon Cincu-Suseni). Spa¢lnaca II has additionally produced Middle Bronze Age flanged axes (Petrescu-Dîmbovita 1978.Pl. 140B.1–2). Moreover, hoards like Gusterita II, Spa¢lnaca II, or Uioara de Sus include artefacts dating to the younger Urnfield Period, i.e. to a time after their conventional dating (Dietrich 2021b). Svend Hansen (2019) has recently argued that these new insights into the formation processes of hoards should lead to a rethinking of the ‘closed find’ as a core concept of archaeology. These long-term col­lections of metalwork are closed finds, but they are not useful for the construction of chronologies, as they may span several hundred years. Hansen used the hoard from Moosbruckschrofen/Piller as a case study, whose artefacts were collected over roughly 300 years. For Romania, Soroceanu et al. (2017) have described this phenomenon for the first time for the hoard from Band. Whether the long-term ac­cumulation of hoards reflects holy places continual­ly used for depositions or sanctuaries where accu­mulated votive offerings were buried after being on display for some time, remains open at the moment and would be a topic for a proper study. With re­gard to the present paper, the chronological implica­tions are of importance. The only chronological system developed for hoard finds of the Carpathian Basin that takes their chro­nological fuzziness into account (although not iden­tifying the formation processes as its reason) was developed by Wilhelm A. v. Brunn (1968). He diffe­rentiated four combination groups of bronze arte-facts, which he equated with the phases defined by Herman Müller-Karpe (1959; v. Brunn 1968.29) for central Europe: Phase 1 – Bz D – Uriu-Doma¢nesti; Phase 2 – Ha A1 – Kisapáti-Lengyeltóti; Phase 3 – Ha A2 – Jaszkarajenö-Uszavölgy; Phase 4 – Ha B1 – Ro­hod-Szentes. However, v. Brunn explicitly stressed the problems of defining clear phases (v. Brunn 1968.29). Combinations of types from different pha­ 9 Two axe fragments belong to the hoard, but not the one illustrated by Petrescu-Dîmbovita (1978.Pl. 106/68), which has been identified as a single find from the same location already by Alexandru Vulpe (1970.45, Nr. 148). 10 Vulpe (1970.65, Nr. 284) interprets the artefact as a fragment of a hammer axe. ‘Scrap metal hoards’ of the Later Urnfield Period in the Carpathian Basin> a case study on a hoard from Dezmir, Romania ses were common, and many finds were not clearly attributable to one phase. Only the contents of the earliest and latest phases were mutually exclusive, which led him to his well-known scheme of an older and younger group of types (“ältere und jüngere Typengesellschaft”). He particularly pointed out that the third phase (Jaszkarajenö-Uszavölgy) was detec­table only by the gradual appearance of new types and the slow disappearance of older ones, i.e. a mix­ture of older and younger artefacts in closed finds (v. Brunn 1968.46). The hoard of Dezmir should be analysed within this theoretical framework, first establishing the chronological position of each of its components. Chronological analysis The fragments Nr. 19–20 and 16 belong to hook sick­les judging from their strong lateral ribs and plain blades, but cannot be determined typologically beyond this point. Hook sickles are characteristic of the older Urnfield Period/ältere Typengesellschaft, they are rare in younger hoards (Petrescu-Dîmbovi­ta 1978.69–70). Most of the sickles from Dezmir are knob sickles. The well-preserved Nrs. 15 and 17 were classified as belonging to his type Panticeu by Petrescu-Dîmbovita, which is mostly characteristic of the same older phase (Petrescu-Dîmbovita 1978. 14, Nr. 59–60, 24). Fragment Nr. 18 was listed un­der the type Cenadu Mare-Spa¢lnaca I, which is dated to Jupalnic-Turia und Moigrad-Ta¢uteu (Petrescu-Dîm­bovita 1978.24.18, Nr. 169). The remaining frag­ments are too small for classification. Bracelets with pointed ends like Nr. 25 appear throughout the later Bronze Age (Petrescu-Dîmbovita 1998.71). For bra­celets with pointed ends and a rhomboid section (Nr. 26), one Bz D grave context from Cruceni, jud. Timis, Romania, has been published (Petrescu-Dîm­bovita 1998.124, Nr. 1374). Plano-convex ingots have been discussed repeatedly in recent decades, but so far no chronological relevance of certain types has been detected (Mozsolics 1984; Czajlik 1996; Nessel 2014; 2019.185–204, 272–274; Modl 2019). The chronological position of Dezmir is thus large­ly dependent on the socketed axes. The Romanian socketed axes have been analysed and published recently (Dietrich 2021a), and brief remarks on the types present at Dezmir may suffice here. Two large groups of socketed axes are present in the hoard: beaked axes and axes decorated with ribs. The beaked axe Nr. 1, with a nearly symmetri­cal mouth and rounded cross section of the body, represents a type characteristic of Phase 4 in north­western Romania and eastern Hungary (Dietrich 2021a.290, Nr. 0399, 291–292; type A11). Nr. 2, a socketed axe with a short, vertical beak and nearly straight rim belongs to a chronologically indifferent variant, which is attested in the older and younger Urnfield period over a large area within the Carpa­thian Basin (Dietrich 2021a.277, Nr. 0356, 282; va­riant A8h). The long and slender axe Nr. 4 with a rounded cross section and a long vertical beak is one of the most common variants of Phase 4 in Ro­mania, where 57 finds concentrate along the middle Mures and the Somes-region; further finds are known from Slovakia, northeastern Hungary, and from the Transcarpathian regions of Ukraine (Dietrich 2021a. 264, Nr. 0279, 280, variant A8a). Axe Nr. 5, a ra­ther squat and massive axe with a short, steep beak and a rounded cross section, belongs again to a va­riant mostly characteristic of Phase 4 (Dietrich 2021a.249, Nr. 0211, 254; variant A6e). Socketed axes like Nr. 7, with curved mouths, short beaks and rounded cross sections, are mainly known from Phase 4 contexts but also appear earlier (Dietrich 2021a.233, Nr. 0130, 236, variant A5b). A round cross section, a feature determined as being in gene­ral characteristic of beaked socketed axes of the younger Urnfield period (Petrescu-Dîmbovita 1944– 1948) is the only indication for the age of fragment Nr. 6 (Dietrich 2021a, 297, Nr. 0453). The remaining socketed axes belong to variants with rib decoration. Axe Nr. 3, with a massive, curved body and broadening blade area has two horizon­tal ribs running below the mouth area. There are a few axes of this variant from contexts of the older Urnfield period, while most finds belong in Phase 4 (Dietrich 2021a.672, Nr. 2731, 682–683; variant D5w). On the other hand, long and slender axes without loops decorated with chevrons like Nr. 8 are mostly known from Phase 2 contexts (Dietrich 2021a.542, Nr. 1927, 548–549; variant D1g). The long and slender trapezoid axe Nr. 9 has a Y-shap­ed ornament below horizontal ribs; the rounded upper part of the Y is repeated above it by a bow – this variant is almost always associated with Phase 4 hoards (Dietrich 2021a.596, Nr. 2264, 614; vari­ant D3m). The massive, curved axe Nr. 10, which is further characterized by an abruptly and signifi­cantly broadening blade and an hourglass-shaped decoration, belongs to a variant of the so-called Pas-sau Type (Blajer 2018), dated securely in Phase 4 (Dietrich 2021a.666, Nr. 2688, 681; variant D5m). The same is true for the undecorated slender and trapezoidal axe Nr. 11 (Dietrich 2021a.604, Nr. 2323, 618; variant D3y). Fragment Nr. 14 has a Y­ Oliver Dietrich, Laura Dietrich, and Botond Rezi shaped motif, which substantiates a Phase 4-date (Dietrich 2021.606, Nr. 2343). Finally, fragments 13–14 cannot be classified (Dietrich 2021a.688, Nr. 2796–2797). Summing up the dating of all hoard components (Fig. 9), the general impression is of a chronologi­cally diverse assemblage. Some types are clearly or partly characteristic of the older Urnfield Period, while others have a tendency towards or a clear date in the younger Urnfield Period. The youngest artefacts define the moment the hoard was hidden, Phase 4. There is no reason to date the hoard of Dezmir to Phase 3 – Ha A2 – Jupalnic-Turia. Spatial analysis The hoard is not only chronologically, but also geo­graphically diverse. Knob sickles of the Panticeu type concentrate between the middle Mures and the So-mes region, in the upper Tisza area and again in Cen­tral Germany (Petrescu-Dîmbovita 1978.Pl. 291A).1 1 The Cenadu Mare type of knob sickles concentrates in the Upper Tisza region and the Somes area (Pe­trescu-Dîmbovita 1978.Pl. 291B). As there still is no general overview on knob sickles except for central Germany (Sommerfeld 1994), these observations are not much more than tendencies though. Again, the socketed axes can provide a clearer image. Between the types and variants present in the hoard, beaked axes of variant A8a (Fig. 10) and the axes with chevron motifs of variant D1g (Fig. 11) are the most frequent types in the Carpathian Basin and adjacent areas. Variant A8a has a very dense distri­bution in the Upper Tisza region, and a less dense one to the east of the Apuseni mountains in the So-mes region. Socketed axes of variant D1g on the other hand concentrate in Transdanubia and in the Drava-Sava region; the eastern Carpathian Basin re­presents the periphery of their distribution, where larger numbers of these axes appear only in excep­tionally large hoards like Uioara de Sus or Gusteri­ta II (Dietrich 2021b). They are, indeed, foreign to this region. These two geographical areas – the Up­per Tisza region on one hand and Transdanubia/the Drava-Sava region on the other – represent the main geographical components within the hoard of Dezmir. The remaining types/variants of beaked axes in the hoard follow the same trend, with the notable exception of variant A5b, which is clearly indigenous to the Somes area and the Crisana, with few finds further to the west (Fig. 12). Between the axes with plastic ribs, variant D5w is characteristic of the eastern Carpathian Basin, variant D5m con­centrates in the Somes region and the Crisana (Fig. 13), while variant D3y is found throughout the Car­pathian Basin north of the Mures River and variant D3m again is mostly known from the Drava-Sava re­gion, with a few finds reaching the Crisana and even the middle Mures (Fig. 13). The best approach to better determine the interac­tion zone the hoard from Dezmir is part of is map- Fig. 9. Chronological overview of the components of the hoard from Dezmir (white: absent; light grey: sporadic appearance; dark grey: main time of use). 11 Petrescu-Dîmbovita’s map partly reflects the degree to which the archaeological material in certain regions had been analysed and published at the time he finished his monograph. The high concentration of Central German finds is partly due to an in-depth study published on the sickle hoards of this area (v. Brunn 1958). ‘Scrap metal hoards’ of the Later Urnfield Period in the Carpathian Basin> a case study on a hoard from Dezmir, Romania ping hoards with a similar combination of socketed axe types (Fig. 14). It comes as no surprise that hoards which have three or more types of socketed axes in common with Dezmir concentrate in the So-mes region. The closest match is a hoard discovered at Dârja, a few kilometres to the north of Dezmir (Rusu 1977b). Another concentration of hoards with similar assemblages of socketed axes is located in the Upper Tisza region. In Transdanubia or in the Drava-Sava area, however, similar combinations are nearly completely missing. Discussion: Dezmir and the ‘scrap hoards’ of the younger Urnfield Period Based on the analysis carried out above, it is possible to de­termine the characteristics of the hoard from Dezmir as fol­lows: 1. A high frequency of dam­aged artefacts or fragments; 2. Different metals (bronze, copper, lead); 3. Raw material/ingots; 4. Chronologically diverse ar­tefacts, the youngest, how­ever, belong in Phase 4 af­ter v. Brunn; 5. A relatively large catchment area from which the arte-facts derive. These characteristics are high­ly reminiscent of the large ‘scrap metal hoards’, which usually represent long-term collections but have a major accumulation phase in the old­er Urnfield Period (Dietrich 2014; 2021b). For Dezmir, a date in a horizon Jupalnic-Tu­ria or Phase 3 after v. Brunn does not hold, as shown abo­ve. This raises the question as to whether the mixture of older and younger artefacts is a general characteristic of the hoards once used to de­fine this chronological phase. What was once believed to be a chronological grouping could in reality be a certain hoarding pattern. To test this hypothesis, all hoards once attributed to the Phases Jupalnic-Turia/Moigrad-Ta¢uteu were tested by a cor­respondence analysis (CA). Hoards are multivariate datasets, differentiated by the presence or absence of chronologically relevant types (“Leittypen”). The appearance, main time of use, and slow disappearance of these types should fol­low a unimodal model for hoards dating to a short, clearly determined period. The CA has been used suc­cessfully for similarly structured datasets in archaeo­logy (Müller, Zimmermann 1997; Müller-Scheeßel 2008; Siegmund 2015; Greenacre 2017). Oliver Dietrich, Laura Dietrich, and Botond Rezi A CA calculates the similarity of multivariate datasets ba­sed on a contingency table that lists the presence/ab­sence or abundancy of attrib­utes (in this case the types of artefacts in the hoards). First, scores are calculated for the columns and rows of the tab­le, whose similarity is then checked by chi-squared dis­tances. Then, the inertia of the columns and rows is cal­culated (i.e. their deviation from the mean values of the columns/rows). This results in the overall inertia of the analysis. If the inertia is very low, the data are too similar for a statistically valid result. The results of a CA are multidimensional. Thus, there are several solutions for every CA, which in the re­duced visualization of the results take the form of dif­ferent axes. The first one (axis 1) is the dominant re­sult. The other solutions are independent of the first axis, and all represent a ratio of the inertia. This ra­tio is expressed through the eigenvalue of each axis and through its percentage of the total inertia. In ar­chaeology, the first three solutions are usually check­ed for statistical meaning. An ideal solution would have a high eigenvalue and the datapoints would be distributed in a parabolic or u-shaped plot in the dia­gram. Deviations from this distribution hint at prob­lems within the dataset, e.g., types used for a long time distorting the solution. All hoards defined by Petrescu-Dîmbovita (1978) as characteristic of the phases Jupalnic-Turia and Moi­grad-Ta¢uteu were used for the CA, with some addi­tions of more recent finds (Fig. 15). The list was re­duced to secure find contexts with at least two rel­evant types (“Leittypen”) that appear in at least two hoards. The types were chosen based on the list pro­vided by v. Brunn (1968.Fig. 1–4, Tab. 1), with some additions regarding the socketed axes based on Diet­rich (2021a).1 2 The software package PAST 5.0 (Hammer et al. 2005) was used for the analyses. If the resulting axes 1 and 2 are mapped in a coordi­nate plane, the distribution takes on the form of a parabola, although the area to the left of the y-axis has fewer finds and types (Fig. 16). The axes rep­resent 11.6 and 10.1% of the total inertia, respec­tively. The hoards to the left of the Y-axis are char­acterized by types 12 (‘Transylvanian’ socketed axes), 20 (richly decorated belts), 43 (hook sickles), and 45 (saw blades). A combination of all these types causes the single outlier, the hoard of Zimandu Nou. They are also the key to an explanation of this re­sult. All are characteristic for the older Urnfield Pe­riod. The result of the CA therefore reflects chrono­logy – some of the finds have older artefacts in their composition, and therefore move away from the main cluster of finds. Dezmir is within the group of chronologically diverse finds, as are most of the hoards proposed earlier as characteristic of Phase 3-Jupalnic-Turia-Ha A2. This becomes clearer if the finds are arranged in a combinatory table according to their position in the CA (Fig. 16). The table lists their main components: tools, weapons, ornaments, raw material, bronze vessels, wagon parts, and horse gear. The general tendency is that the chronological­ly diverse finds are also heterogenous in their com­ 12 The following types characteristic to Phases 2–4 are present in the hoards and were chosen for analysis (numbering after v. Brunn (1968), with additions): 12 socketed axes of the ‘Transylvanian’ type; 16 early winged axes with wings in the upper part of the axe; 20 richly decorated belts; 30 socketed axes with flexed rib decoration; tongue sickles, variant 2; 32 massive winged axes, the wings are located in the upper part and there is no ridge between blade and body; 33 socketed axes of the ‘Passau Type’; 34 socketed axes decorated with vertical curved bundles of ribs; 35 socketed axes decorated with chevrons and a pronounced ridge between blade and body; 36: bracelets/legrings decorated with hatched chevrons; 37 tongue sickles of type 3; passementerie fibulas with small spirals at the bow and large terminal ones; 43 hook sickles; 44 flange hilted swords of the Reutlingen type; 45 saw blades; 49 beaked socketed axes with round diameter and steep beak; 51 kettles of type B1; 53 cups of the Kirkendrup type. ‘Scrap metal hoards’ of the Later Urnfield Period in the Carpathian Basin> a case study on a hoard from Dezmir, Romania position, while the finds clear­ly datable to Phase 4 have fewer components and are ge­nerally smaller. Result A close analysis of the hoard from Dezmir and comparable finds reveals good evidence against the formerly propos­ed chronological hoarding ho­rizon ‘Phase 3-Jupalnic-Turia-Ha A2’. In the Eastern Carpa­thian Basin, this group com­prises of hoards with chrono­logically diverse artefacts hid­den in Phase 4. Their struc­ture, here described and ana­lysed using Dezmir as a case study, is highly reminiscent of the large so-called scrap metal hoards like Uioara de Sus, which were accumulated mostly during the older Urn-field Period. These hoards comprise a large number of chronologically diverse, often damaged or fragmented arte-facts and a high percentage of raw material, while some of the objects are foreign to the region they were deposit­ed in (Dietrich 2014; 2021b). Contrary to earlier beliefs, this hoarding scheme does not disappear in the younger Urnfield Period, but survives in a quantitatively reduced form. The hoard from Dezmir is a characteristic example of this group of hoards. ACKNOWLEDGEMENTS We want to thank Dr. Tudor Soroceanu, Berlin, for the idea to work on the hoard of Dezmir, as well as numer­ous suggestions along the way. We are also grateful to the MNIT Cluj-Napoca for the permission to re-document the hoard. Work at the Museum was carried out by Botond Rezi. Oliver Dietrich, Laura Dietrich, and Botond Rezi ‘Scrap metal hoards’ of the Later Urnfield Period in the Carpathian Basin> a case study on a hoard from Dezmir, Romania Oliver Dietrich, Laura Dietrich, and Botond Rezi References Bergen C. 2005. Technologische und kulturhistorische Studien zu Bleifunden im 1. Jahrtausend. Universitäts­forschungen zur Prähistorischen Archäologie 125. Verlag Dr. Rudolf Habelt GmbH. Bonn. Blajer W. 2018. Some remarks on socketed axes of the Passau type. Recherches Archéologiques Nouvelle Serie 10: 7–19. https://doi.org/10.33547/rechacrac.ns10.01 Boroffka N., Nessel B., Prange M., Ciugudean H., and Ta-kács M. 2016. Neues Licht auf alte Fragen – Einige beson­dere Metallobjekte aus dem Depotfund von Aiud, Kreis Alba, Rumänien. In M. Bartelheim, B. Horejs, and R. Krauss (eds.), Von Baden bis Troia. Resourcennutzung, Metallur­gie und Wissenstransfer. Eine Jubiläumsschrift für Ernst Pernicka. Oriental and European Archaeology 3. Verlag Marie Leidorf GmbH. Leidorf: Rahden/Westf.: 339– 421. Britton D. 1960. Great Britain. Inventaria Archaeologia 8: 48–54. von Brunn W. A. 1958. Der Schatz von Frankleben und die mitteldeutschen Sichelfunde. Prähistorische Zeitschrift 36: 1–70. https://doi.org/10.1515/prhz.1958.36.1.1 1968. Mitteldeutsche Hortfunde der jüngeren Bronze-zeit. Römisch-Germanische Forschungen 29. Walter de Gruyter. Berlin. Ciugudean H., Luca S. A., and Georgescu A. 2006. Depozi­tul de bronzuri de la Dipsa. Bibliotheca Brukenthal V. Altip. Alba Iulia. Ciugudean H. 2012. The chronology of the Gáva culture in Transylvania. In W. Blajer (ed.), Peregrinationes Ar-chaeologicae in Asia et Europa Joanni Chochorowski Dedicatae. Instytut Archeologii Uniwersytetu Jagielloñ­skiego Wydawnictwo. Profil-Archeo. Kraków: 229–243. Crisan I. H., Ba¢rbulescu M., Chirila¢ E., Vasiliev V., and Winkler I. 1992. Repertoriul arheologic al judetului Cluj. Bibliotheca Musei Napocensis 5. Institutul de Istorie si Arheologie Cluj-Napoca. Cluj-Napoca. Czajlik Z. 1996. Ein spätbronzezeitliches Halbfertigpro­dukt: Der Gußkuchen. Eine Untersuchung anhand von Funden aus Westungarn. Archaeologia Austriaca 80: 165–180. Dietrich O. 2011. Zentralisierte Produktionsstrukturen? Überlegungen zur räumlichen Beziehung von bronzezeit-lichen Gussformen und Fertigprodukten in Südosteuropa am Beispiel der rumänischen Tüllenbeile. Marisia XXXI: 77–91. 2014. Learning from ‘scrap’ about Late Bronze Age hoarding practices. A biographical approach to indivi­dual acts of dedication in large metal hoards. European Journal of Archaeology 17(3): 468–486. https://doi.org/10.1179/1461957114Y.0000000061 2021a. Bronzezeitliche Tüllenbeile aus Rumänien. Studien zu Chronologie, Chorologie und Hortung. PhD thesis. Dissertationen FU. Refubium Freie Universi­tät Berlin. Berlin. http://dx.doi.org/10.17169/refubium-30761 2021b. in press. Aiud – Gusterita II – Spa¢lnaca II – Uioara de Sus. Gedanken zu den Formationsprozessen und der Interpretation von bronzezeitlichen Großhor-ten des südlichen Karpatenbeckens. Dacia N. S. LXV. Dörfler G., Neuninger H., Pittioni R., and Siegl W. 1969. Zur Frage des Bleierz-Bergbaus während der jüngeren Ur-nenfelderkultur in den Ostalpen. Archaeologia Austriaca 46: 68–98. Drescher H. 1978. Untersuchungen und Versuche zum Blei- und Zinnguß in Formen aus Stein, Lehm, Holz, Ge-weih und Metall. Frühmittelalterliche Studien 12: 84–115. https://doi.org/10.1515/9783110242119.84 Evans J. 1881. Ancient bronze implements of Great Bri­tain and Ireland. Longmans, Green & Co. London. Foltz E. 1980. Guss in verlorener Form mit Bleimodellen? Archäologisches Korrespondenzblatt 10: 345–349. Goldmann K. 1981. Guss in verlorener Sandform – Das Hauptverfahren alteuropäischer Bronzegiesser? Archäolo­gisches Korrespondenzblatt 11: 109–116. Gowland W. 1901. The early Metallurgy of Silver and Lead. Archaeologia 57(2): 359–422. https://doi.org/10.1017/S0261340900014211 Greenacre M. 2017. Correspondence analysis in prac­tice. Routledge. Boca Raton, London, New York. Hammer O., Harper D. A. T., and Ryan P. D. 2001. PAST: Palaeontological Statistics Software Package for Education and data analysis. Palaeontologia Electronica 4(1): 1–9. https://palaeo-electronica.org/2001_1/past/past.pdf Hansen S. 1994. Studien zu den Metalldeponierungen während der älteren Urnenfelderzeit zwischen Rhône­tal und Karpatenbecken. Universitätsforschungen zur Prähistorischen Archäologie 5. Dr. Rudolf Habelt. Bonn 1994. ‘Scrap metal hoards’ of the Later Urnfield Period in the Carpathian Basin> a case study on a hoard from Dezmir, Romania 1996. Bemerkungen zur zeitlichen Stellung der Hort-funde des Typus Gyermely. Archäologisches Korres­pondenzblatt 26: 433–441. 2019. Metalldeponierungen in Eurasien. Ein Phänomen der Langen Dauer, der Konjunkturen und der Ereignis­se. In S. Hye, U. Töchterle (eds.), UPIKU: TAUKE. Fest-schrift für Gerhard Tomedi zum 65. Geburtstag. Uni-versitätsforschungen zur Prähistorischen Archäologie 339. Dr. Rudolf Habelt GmbH. Bonn: 201–217. László A. 1994. Începuturile epocii fierului la est de Car-pati. Culturile Gáva-Holihrady si Corla¢teni-Chisina¢u pe teritoriul Moldovei. Bibliotheca Thracologica VI. Institu­tul Român de Tracologie. Bucharest. Lawson A. J. 1979. A late bronze age hoard from West Caister, Norfolk. In C. Burgess, D. Coombs (eds.), Bronze Age hoards. Some finds old and new. BAR British Se­ries 67. BAR Publishing. Oxford: 173–179. Modl D. 2019. Recording plano-convex ingots (Gussku­chen) from Late Bronze Age Styria and Upper Austria – A short manual for the documentation of morphological and technological features from production and partition. In R. Turck, T. Stöllner, and G. Goldenberg (eds.), Alpine Copper II – Alpenkupfer II. New results and perspecti­ves on prehistoric copper production. Der Anschnitt. Bei-heft 42. Deutsches Bergbau-Museum Bochum. Bochum: 373–398. Mozsolics A. 1984. Ein Beitrag zum Metallhandwerk der ungarischen Bronzezeit. Berichte der Römisch-Germa­nischen Kommission 65: 19–72. Müller R. 2006. Investigation of a hill settlement from the Late Bronze Age at Várvölgy, Nagyláz-hegy (2003–2006). Régészeti Kutatások Magyarországon 2006: 5–26. https:// archeodatabase.hnm.hu/hu/download/2607/file/e85e5 06d-015d-41fd-a775-fc4a2d32d7c9?type=pdf Müller J., Zimmermann A. (eds.) 1997. Archäologie und Korrespondenzanalyse. Beispiele, Fragen, Perspektiven. Internationale Archäologie 23. Verlag Marie Leidorf. Es-pelkamp. Müller-Karpe H. 1959. Beiträge zur Chronologie der Urnenfelderzeit nördlich und südlich der Alpen. Römisch-Germanische Forschungen 22. Walter de Gruyter. Berlin. Müller-Scheeßel N. 2008. Korrespondenzanalyse und ver­wandte Verfahren. In M. K. H. Eggert (ed.), Prähistori­sche Archäologie: Konzepte und Methoden. Narr Francke Attempto Verlag. Tübingen, Basel: 219–239. Needham S. P., Hook D. R. 1988. Lead and lead alloys in the Bronze Age. In E. A. Slater, J. O. Tate (eds.), Science and Archaeology. Symposium Glasgow 1987. BAR Bri­tish Series 195. BAR Publishing. Oxford: 259–274. Nessel B. 2014. Bronze Age portioning of raw metal – con­cepts, patterns and meaning of casting cakes. Apulum 51: 401–416. 2019. Der bronzezeitliche Metallhandwerker im Spie­gel der archäologischen Quellen. Universitätsforschun-gen zur Prähistorischen Archäologie 344. Dr. Rudolf Habelt GmbH. Bonn. Pankau C. 2004. Die älterhallstattzeitliche Keramik aus Medias/Siebenbürgen. Universitätsforschungen zur Prähi­storischen Archäologie 109. Dr. Rudolf Habelt GmbH. Bonn. Petrescu-Dîmbovita M. 1944–1948. Depozitul de bronzu­ri de la Bîrsana (Maramures). Anuarul Institutului de Studii Clasice 5: 264–281. 1977. Depozitele de bronzuri din România. Editura Academiei Bucharest. 1978. Die Sicheln in Rumänien mit Corpus der jung­und spätbronzezeitlichen Horte Rumäniens. Prähisto­rische Bronzefunde XVIII, 1. C. H. Beck Verlagsbuch­handlung. München. 1988. Der Arm- und Beinschmuck in Rumänien. Prä­historische Bronzefunde X, 4. Steiner. Stuttgart. Popescu D. 1964. Sa¢pa¢turile arheologice din Republica Populara¢ Romîna¢ în anul 1963. Studii si Cerceta¢ri de Istorie Veche 15(4): 551–580. Reissenberger L. 1872. Der neueste archäologische Fund bei Hammersdorf. Archiv des Vereins für Siebenbürgi­sche Landeskunde 10: 8–37. Rusu M. 1972. Metalurgia bronzului din Transilvania la începutul Hallstattului. Unpublished PhD thesis. Uni-versitatea “Al. I. Cuza”. Iasi. 1977a. Dezmir, Kr. Cluj, Rumänien. Inventaria Archa­eologica Rumänien 10: R 65a-d. 1977b. Dîrja, Kr. Cluj, Rumänien. Inventaria Archaeo­logica Rumänien 10: R 66a-d. Siegmund F. 2015. Gewußt wie: Praxisleitfaden Seria­tion und Korrespondenzanalyse in der Archäologie. Ei­genverlag. Norderstedt. Sommerfeld C. 1994. Gerätegeld Sichel. Studien zur mo-netären Struktur bronzezeitlicher Horte im nördlichen Mitteleuropa. Vorgeschichtliche Forschungen 19. Walter de Gruyter. Berlin, New York. Oliver Dietrich, Laura Dietrich, and Botond Rezi Soroceanu T., Rezi B., and Németh R. E. 2017. Der Bronzedepotfund von Bandul de Câmpie, jud. Mures. Universitätsforschungen zur Prähistorischen Archäologie 307. Dr. Rudolf Habelt GmbH. Bonn. Tarbay J. G. 2015. The reanalysis of the eponymous hoard from Gyermely-Szomor and the Ha A2 period in the ter­ritory of Hungary. In R. E. Németh, B. Rezi (eds.), Bronze Age Chronology in the Carpathian Basin. Editura Mega. Târgu Mures: 311–371. Tylecote R. F. 1962. Metallurgy in archaeology. A prehis­tory of metallurgy in the British Isles. Taylor & Francis. London. Vachta T. 2016. Thesaurierungsprozesse bronzezeitlich-back to contents er Hortfunde anhand ihrer Kompositionselemente. In S. Hansen, D. Neumann, and T. Vachta (eds.), Raum, Gabe und Erinnerung: Weihgaben und Heiligtümer in prähi­storischen und antiken Gesellschaften. Edition Topoi. Berlin: 93–117. Vasiliev V., Aldea I. A., and Ciugudean H. 1991. Civilisa­tia dacica¢ timpurie în aria intracarpatica¢ a României. Contributii arheologice: Asezarea fortificata¢ de la Te-leac. Editura Dacia. Cluj-Napoca. Vulpe A. 1970. Die Äxte und Beile in Rumänien I. Prä­historische Bronzefunde IX, 2. C. H. Beck’sche Verlags­buchhandlung. München. Zaharia E. 1965. Remarques sur le Hallstatt ancien de Transylvanie. Fouilles et trouvailles de Medias 1958. Da­cia N.S. IX: 83–104.