UDK 903.2:7.045(234.421 )"633/634"__
Documenta Praehistorica XXV (Poročilo o raziskovanju paleolitika, neolitika in eneolitika v Sloveniji XXV)
Fish, faces and fingers: presences and symbolic identities in the Mesolithic-Neolithic transition in the Carpathian basin
Alasdair VVhittle
School of History and Archaeology, Cardiff University, whittle@cardiff.ac.uk
ABSTRACT - There are many neglected difficulties ivitli a colonisation model for south-east Europe at the start of the Neolithic, though some kind of sloiv and fragmented process may hold good for the southern Balkans. Thispaper concentrates on the northern Balkans, and especially the Carpathian basin east of the Danube, ivhere the character of the early Neolithic lifestyle raises the possibility of indigenous acculturation. Varied Mesolithic presences, mobilities and regional systems in south-east Europe are discussed, and compared tvith Carpathian basin early Neolithic distributions and life-styles. In seeking possible indigenous continuities, particular attention is given to symbolism and identity, via material culture, includingpotterj andfigurines, and burials. A comparison is made betuieen the symbolic system of the Starčevo-Koros culture and contemporaneous developments in the Danube Gorges. The tivo ideologies may have overlapped in many ways, and the many-sidedper-sonal identities of the Starčevo-Koros population may themselves have had a long local history. New concepts focus on ancestral beginnings and marked tirne, the human form and a more conscious difference betuieen people and animals, and participation by the living in broadpatterns of social interaction; the potential complexity of their derivation must now be recognised.
POVZETEK - Težave z modelom kolonizacije jugovzhodne Evrope na začetku neolitika ostajajo, čeprav velja ocena, da lahko dogajanje na južnem Balkanu morda vendarle označimo kot del nekakšnega počasnega procesa. V razpravi se ukvarjamo s severnim Balkanom in s Karpatsko kotlino vzhodno od Donave, kjer je zgodnje neolitski način življenja mogoče navezati na staroselsko akultu-racijo. A nalizirali smo različne mezolitske zapise, mobilnost ter regionalne sisteme v jugovzhodni Evropi in jih primerjali z zgodnjeneolitsko distribucijo in načinom življenja v Karpatski kotlini. Pri iskanju domnevne staroselske kontinuitete je bila s pomočjo lončenine, figurin in pokopov, posebna pozornost namenjena identiteti in simbolizmu. Primerjali smo simbolna sistema kulture Starčevo-Koros in sočasnega razvoja v Džerdapu, Ideologiji sta se najbrž v mnogočem prekrivali, saj identiteta Starčevo-Koros populacije gotovo temelji na dolgi lokalni zgodovini. Potrebujemo nov konceptualni pogled na začetke naših prednikov in časa, ki so ga zaznamovali, na človekove navade in na zavestno ločevanje med ljudmi in živalmi, na participiranje živih v obširnih vzorcih socialne interakcije in na potencialno kompleksnost njihovega izvora.
COLONISATION MODELS
How did the Neolithic begin in south-east Europe, and what did this Neolithic consist of ? Answers to the two questions have been closely intertwined in the long dominant model of colonisation. The Neolithic has often been seen as the arrival of a new population, from Anatolia and points east, with a new subsistence economy based on domestication of plants and animals and a concomitant sedentary life-style. Since Neolithic expansion from the Levant can be traced westwards (e. g. Cauvin 1994), and since the Mesolithic or Epipalaeolithic presence in south-east Europe has long seemed both patchy and thin (e. g. Tringham 1971), debate within the colonisa-
tion model has concentrated not on challenging ba-sic assumptions or considering possible alternatives, but rather on investigating details of dates and rou-tes (e. g. Kaiser and Voytek 1983; Perles 1990; Hansen 1991). There has been some recognition of the possibility of filtered or fragmented colonisation by sea, for example in the 'boat people' model of Chapman and Miiller (1990), but this has hardly been connected with a wider review of the supposed colonisation phenomenon as a whole.
That colonisation did take plače, and by sea, under conditions presumably more difficult than on land,
is amply documented by what happened on Cyprus and Crete (Cherry 1990; Broodbank and Strasser 1991), and indeed on other islands in the central and west Mediterranean (Patton 1996). On the other hand, probably both Cyprus and Crete may have been empty of resident population at the start of the Neolithic, and their intake was not therefore neces-sarily typical of wider processes. While the strengths of the colonisation model have often been empha-sised, its weaknesses are less often debated. I have set out these arguments elsewhere (Whittle 1996, chapter 3; cf. Zvelebil 1995; Zvelebil and Lillie forthcoming; Chapman 1994a), and need only briefly allude to them here to set the scene for spe-cific discussion of the northern Balkans and the Car-pathian basin in particular.
The distribution and density of the early Neolithic in western Anatolia remain to be established (e. g. Cau-vin 1994; Ozdogan 1989; Ozdogan 1995; Ozdogan 1997). At the present tirne, it is far from clear that western Anatolia was sufficiently well populated to have generated significant budding-off on the scale required for full-scale colonisation, though of course that does not exclude more episodic or opportunis-tic fission. Expansion into western Anatolia might itself only date to the sixth millennium BC (Yakar 1996.6); recent finds in the Marmara area (Ozdogan 1997) have not so far been matched further south. Pottery was a recent innovation in Anatolia itself, and the possibility of an aceramic phase remains in Greece; one of the supposed principal material sig-natures of a new, intrusive population may in fact have been characteristic of neither alleged source population nor alleged first incomers. By contrast, the presence of obsidian in early Neolithic Thessaly (.Perles 1992) relates to the continued exploitation of a source known to indigenous population since the Palaeolithic and in regular use in the Mesolithic (.Perles 1990). Above ali, the establishment of what we regard as the typical elements of the early Neolithic may have been a long and slow process (Whittle 1996, ch. 3). The important investigations at Platia Magoula Zarkou in northern Thessaly, for example, show that a teli began in an unstable and periodi-cally inundated creek/floodplain environment (van Andel et al. 1995), making permanent settlement impossible. The character of early levels at Argissa, Sesklo and elsewhere in Thessaly (Milojčič 1960; Gimbutas et al. 1989; Wijnen 1982) shows that early occupations were not continuous (though that does not exclude the possibility at some of them of year-on-year residence) and did not include sub-stantial built above-ground structures. Tells are any-
way something that came into being through the later and continued histories of chosen places (cf. Chapman 1997a), and 'open' sites have begun to be recognised in north-east and northern Greece, in Ma-cedonia and Thrace (Andreou et al, 1996). For ali the past excavations of tells in central-southern Bulgaria (e. g. Todorova 1995), we lack detailed infor-mation on early levels, and a regional contrast is also apparent in the different character of early Neolithic settlement in north-east and north-west Bulgaria (Todorova 1995). And so on.
It is possible therefore to envisage that the begin-nings of the Neolithic in the southern Balkans were at the least both slower and more regionally varied than commonly supposed in vulgar versions of the colonisation model. This raises also the possibility of transformation involving more centrally the indigenous population. To resolve this question will re-quire much more research, including - apart from excavation and locally-oriented studies (Miracle 1997) - more radiocarbon dating, survey (including in western Anatolia) and if possible DNA analysis of ancient human bone, animal bone and plant material (cf. Heun et al. 1997). My first aim has been to show that even in the southern Balkans the model of fullscale colonisation rests on less secure grounds than commonly supposed. This does not exclude the possibility of episodic or filtered movement of new population.
In the northern Balkans the čase for fullscale colonisation is weaker stili. It has long been noted that the early Neolithic Starčevo-Koros lifestyle looks different from that of the supposedly typical areas of teli settlement to the south (e. g. Tringham 1971; Trog-mayer 1968.18-19; cf. Banner 1937). There are scat-tered sites and occasional clusters; occupation levels are thin, generally without significant stratigraphic build-up, which strongly implies residential mobility, on a spatial and temporal scale stili to be established (cf. Whittle 1997); material culture is in some ways (especially as seen in pottery) simpler; and a wide range of resources was exploited, including wild game, fish, birds and shellfish alongside domesticat-ed animals and cereals. Within the subsistence econ-omy the balance of resources is unclear. The scale of cereal cultivation may have been quite restricted in the 'island' pattern of Koros waterside occupations (Kosse 1979; Sherratt 1982a; cf. Willis and Bennett 1994), and the dominance of sheep and goats in such a setting (B6k6nyi 1974) has always seemed more than a little odd. If these are reasonable doubts about the plausibility of continued incoming popu-
lation, can we envisage in more detail the processes by which a regional indigenous population could have changed, to become what we increasingly inad-equately call Neolithic? To answer that question, ra-rely formulated in any specific fashion for south-east Europe (but see Chapman 1994a), we must hirther consider aspects of identity and lifestyle. But first, there is the issue of Mesolithic presences and distributions.
INDIGENOUS PRESENCES
It was noted above that the apparent lack of Mesolithic distributions in south-east Europe has often been taken as a further support for the colonisation model. This now requires the closest examination (cf. Zvelebil 1995). First, there is the matter of research history and coverage (Chapman 1989). The Mesolithic or Epipalaeolithic has been a poor relation in the development of most parts of south-east Europe. After ali, no one anticipated the discovery of the spectacular finds in the Danube Gorges before inves-tigations began in 1965 (Srejovič 1972). Finds there remain restricted to the bottom of the Gorges, and despite the existence of a wide range of terrestrial resources in Gorges-bottom sites including pig and deer which could hardly have shared the same nar-row water-edge areas as people, no survey has yet been carried out of the varied hinterland terrain on either side of the Gorges; Baile Herculane on the Ro-manian side, though probably very early in the Holocene sequence (Nicoldescu-Plop§or and Pdunes-cu 1961; Dinan 1996), indicates what might be ex-pected in side valleys and plateaus. Repeated obser-vations in the main part of the Great Hungarian plain have so far failed to locate signs of Mesolithic presence (Makkay 1996.41), but knowledge of local collections combined with careful survey and exca-vation have begun to produce evidence on the northern edge of the Plain for an early Holocene presence, just beyond the Koros culture distribution (Kertesz 1996). Against this, there are some exam-ples of areas where systematic survey has not produced or has not been able to recognise evidence for a Mesolithic or Epipalaeolithic presence, for example along the Peneios in northern Thessaly and in in-land Epiros in north-west Greece (Runnels 1988; G. Bailey 1998).
These cases constitute only partial or anecdotal evidence until much wider and more systematic as well as detailed local surveys have been carried out. But even in the present state of evidence it is possible
to consider the overall nature of Mesolithic distributions, to compare them with the evidence for the also non-continuous distributions of the early Neolithic, and to begin to model variation in Mesolithic regional systems. Recognising that there may not have been a single kind of Mesolithic presence, just as with the early Neolithic, may be an important first step to further progress.
Mesolithic populations can in fact be documented over a wide area of south-east Europe as a whole. The general situation has been well mapped by Zvelebil {1995, fig. 5), though with brief accompanying detail. There are sites and/or concentrations: in the north-eastern Peloponnese at the Franchthi Cave (Hansen 1991; Perles 1990); at the Theopetra cave on the northern edge of the Thessalian plain (Kypa-rissi-Apostolika 1995); in the Dinaric chain from Montenegro to Slovenia (Srejovič 1989; Srejovič 1996; Budja 1993); on the northern side of the Great Hungarian Plain in the Jaszsag region north of Szolnok (Kertesz 1996), and then further north in Slovakia and Moravia (e. g. Kozloivski 1982; Matei-ciucovd forthcoming); in the Danube Gorges (Srejovič 1972; Radovanovič 1996); in the Southern Bug and Dniestr valleys east of the Carpathians (Marke-vitch 1994; Zvelebil 1995; Zvelebil and Dolukha-nov 1991); and in eastern Bulgaria (Gatsov 1989) and easternmost Thrace (Gatsov and Ozdogan 1994).
Absences have therefore probably been much exag-gerated, just as differences to early Neolithic distributions may have been overdrawn. For the early Neolithic, it is normal and understandable practice to present maps with cross-hatched or othenvise gen-eralised distributions (e. g. Tringham 1971, fig. 10; Gimbutas 1991, fig. 2-14). These can conceal the va-riations in early Neolithic settlement type and dura-tion already noted, just as they can also mask areas with stili surprisingly low Neolithic presence, for example the Vardar valley compared with the Stru-ma, and Yugoslav Macedonia and southern Serbia in general (Garašanin 1982; Tasič 1997). Koros distributions in southern Hungary are in places strong-ly clustered, with micro-regional distributions evident in the area of the Double and Triple Koros rivers, for example around Szarvas, Devavanya and Gyomaendrod (MRT1989; cf. Kalicz 1990.83-8); it is also possible that there are less dense distributions, in the area of the Maros-Tisza confluence (Trogmayer 1968; Horvdth 1989), on the Danube itself (Kalicz 1990) and on the north-west fringe of the overall distribution around Szolnok in the Tisza
valley (Raczky 1976). Likewise, there is a wide scat-ter of Starčevo sites in the Vojvodina, but it is not yet clear whether these form the dense riverine clus-ters characteristic of parts of the Koros distribution. Perhaps by way of contrast, the range of Starčevo locations in northern Serbia is rather broad (e. g. Chapman 1990).
Bevond the mere question of presence and absence there is the issue of the nature of regional systems. It seems both short-sighted and unhelpful to sup-pose that ali Mesolithic regional settlement systems were uniform throughout south-east Europe. Varia-tion is already apparent, even in the current state of research, and may be both a diachronic and spadal feature.
Evidence from Franchthi Cave shows two dominant, perhaps related features. The deposits themselves represent a long continuity of occupation from late Pleistocene into the Holocene. The intensity of occupation seems to have varied, though it was regular-ly more intense in the early Holocene than earlier; the period of Mesolithic-Neolithic transition is miss-ing, however, due to erosional hiatus (Perles 1990; Hansen 1991). The presence of graves reinforces the importance given to this chosen plače. Secondly, there was a broad-spectrum subsistence economy, elements of which would have taken people far afield. It is not clear exactly how far to sea in the Aegean the catching of large tunny would have taken people, but it is possible that the distances cov-ered were considerable (vari Andel and Runnels 1987). The regular bringing of obsidian from Melos back to the cave reinforces this possibility. It can be stressed that in the Mesolithic the cave itself was close to rather than on the coast itself (Cnrtis and Runnels 1987), and thus safe (for archaeological purposes) from subsequent sealevel rises. To the west, in Sicily, the Grotta deli' Uzzo provides a rather similar sort of situation, again in a location a lit-tle above the sea (Costantini 1989). Given the range of the Franchthi exploitation system, it would re-cjuire only a couple more such sites to have existed in the Aegean, physically closer to early Holocene water levels, say in Euboia or southern Thessaly and in south-west Turkey (compare the Okuzini cave inland: Otte et al, 1995), for the Mesolithic of the Aegean as a whole immediately to look more busy.
The Danube Gorges are the obvious next example, and in discussing them I follow the chronology of Radovanovič (1996), according to which some sites are pre-Neolithic but others, including most of the
Lepenski Vir sequence, run parallel to Starčevo-Ko-ros elsewhere in the region. In the Gorges people exploited fish from the river. Isotopic evidence from Vlasac and Schela Cladovei indicates that some parts of the population may have been heavily dependent on fish (Bonsall et al. 1997), although the largest anadromous fish, Acipenser huso or beluga, appears not to have been exploited in later periods (Radovanovič 1997). Use of fish may have bound some people closely to the river, in differing parts of the Gorges. But there were also numerous finds of ter-restrial animals, notably red deer, which also had symbolic significance in mortuary rituals. Hunting or othervvise exploiting such animals must have taken people further afield, away from the Gorges. The movement of raw material also shows wider move-ment, to bring flint, obsidian, basalt and igneous ročk from the north and west and pre-Balkan platform' flint and graphite from north Bulgaria (Chapman 1989; Kozloivski 1982). It remains a moot point (and see further below) whether the sites are to be regarded as merely settlements or whether some or several can be characterised as special pla-ces or shrines, especially those in the upper Gorges including Lepenski Vir itself (Radovanovič 1996; Wlrittle 1996; for detailed maps see Radojčič and Vasič 1997); this may have been a feature especial-ly of the period of Neolithic contact. The important implication here is that sites and/or shrines in the upper Gorges may have served a much vvider population, at least partially mobile by land or by river over varving but sometimes considerable distances.
In other cases, Mesolithic systems may have been more limited. Hypothetically, sites up and down the Dinaric chain (Srejovič 1996; Miiller 1994; Budja 1993; Chapman et al. 1997) could have been part of a system of seasonal movement, which involved summer occupations in the high hills and winter stays in the narrow coastal lowlands. Likewise the Southern Bug-Dniestr sites may have been based on a combination of local river fishing and forest-step-pe hunting.
Different kinds of radius and mobility are evident. In at least two cases, though each was different, the combination of local activity with long-range mo-bility may be the key to understanding the distribution of people and sites. Were areas like Thessaly, therefore, which was so important in the Neolithic from the early Neolithic onwards, literally empty in the Mesolithic? Despite the general continuing non-recognition of Mesolithic sites, there is a document-ed presence now in the Theopetra cave (Kjparissi-
Apostolika 1995), and this could indicate - albeit unclearly at this stage - something of the same kind of regional system. There is also the matter of where some early Neolithic sites were placed. Early sites in-clude many examples away from the most fertile lo-cations suitable for easy permanent occupation, including Achilleion dose up to the southern hills fring-ing the Thessalian plain, and Sesklo set in its strik-ing natural amphitheatre of hills (Kostas Kotzakis, pers. comm.; Mills 1997). It is as though there was already knowledge of where to go.
The Neolithic pattern of settlement could therefore have been based on what went before, but equally it does not represent a direct continuation of this. At a regional scale there was infill and perhaps a shift in the range of mobilities (though note the contin-ued importance of Melian obsidian, brought to Thes-saly, and of pre-Balkan platform flint, taken to Starčevo sites). Importantly, however, in the perspective suggested here, such infill and shifts were relative. A 'clean slate' or 'empty niche' model of colonisation of the Balkans can hardly any longer be supported. In the past such expansion, whatever predse form it took, has been seen chiefly as the outcome of the operation of new ways of getting fed. The rest of this paper is concerned with the significance for this question of matters of identity.
EARLY NEOLITHIC LIFESTYLE IN THE NORTHERN BALKANS
If the Neolithic phenomenon in the northern if not also the southern Balkans was not simply a matter of changing resource procurement and diets, what other changes were fundamental?
We have already noted above that there were sub-sistence changes, notably the appearance of domes-ticated animals including sheep and goats and the be-ginnings of cultivation of non-indigenous cereals. These new elements became very widely distributed, including within the Danube Gorges, where isotopic evidence indicates a less aquatic diet in the contact phase (Bonsall et al. 1997). What, however, was their importance? To answer this, much basic research remains to be done, especially now at local scales (cf. Miracle 1997). It has long been clear (cf. Banner's brilliant initial 'ethnology' of the Koros culture: Banner 1937) that a very varied range of resources was exploited in the Koros context. Game, fish, birds and shellfish are documented, and the suc-cession of deposits in pits in Maros-Tisza confluence
sites could show patterns of resource exploitation changing by the season (Tringham 1971.92; Trog-mayer 1968). Fine sieving, cementum increment studies (cf. Lieberman et al. 1990; Burke 1993; Burke and Castanet 1995) and detailed micromor-phology of feature fills are among approaches that need to be applied, to refine our understanding of seasonality and seasonal variation in resource use. From Starčevo itself comes a long list of game, fish and birds which were exploited (Clason 1980), a range which seems to be matched on Koros sites CB6konyi 1974; Bokonyi 1992; Takdcs 1992). Starčevo itself is on the edge of the Danube floodplain (Barker 1975)] the extent and duration of annual flooding there remain to be established. Further north in the Koros river system, the extent and duration of backswamp flooding both seem likely to have been greater (Kosse 1979; Sherratt 1982a,-Sherratt 1982b), though again this remains to be established in much more detail. People of the Koros culture may have lived much of their lives in a fragmented pattern of islands. If so, it seems unlike-ly that either limited cereal cultivation or the hus-bandry of sheep and goats could have constituted the critical key resources which enabled the intake or infill (if such it really was) of this environment from the early Neolithic onwards. It is possible that future research into river history could indicate changes in natural conditions which allowed easier exploitation of this zone than in the very early Ho-locene (there might be an issue of malaria in wet loivlands; Andreiv Sherratt, pers. comm,; and Sherratt 1997.21). When occupation came, levee cultivation of cereals is plausible enough (cf Sherratt 1980; van Ande/ et al. 1995), but the scale and regularity may have varied. Flotation at the short-lived, perhaps seasonal Cri§ occupation site of Foeni-Salas in western Romania produced no cereal remains (Gre-enfield and Drasovean 1994). The keeping of sheep and goats might even appear somewhat perverse in this kind of setting. The motive for possession of these animals could rather have been novelty or their connection with new beliefs and identities.
As already noted, Starčevo-Koros sites characteristi-cally have thin levels, and in the current state of research built structures are relatively rare. That built structures did exist is well enough shown by exam-ples like Divostin and Tiszajeno (McPherron and Srejovič 1988; Selmeczi 1969; Raczkv 1976; cf. Trogmayer 1966), and suggested elsewhere by sur-face finds of burnt daub (e. g. Sherratt 1983), and the only slightly later example of new discoveries of longhouses in the northern Linear Pottery cultu-
re of the Hungarian Plain at Fiizesabony (.Bombo-roczki 1997) shows how dependent such observa-tions can be on the scale of excavation possible; before the motorway rescue excavations, AVK long-houses could only be documented episodically front the Szakalhat phase onwards. There is also an enor-inous amount to be done to understand the possible rhythms of occupation of Koros waterside sites (cf. Sherratt 1982b). But even in the current state of research, it seents likely that there was corning and going in the Koros lifestyle, and given that Starčevo sites include also waterside ones and caves in the hills, it is plausible that the generalisation holds good over a wider area, and not just in the Koros river system itself.
Mobility in the Starčevo-Koros lifestyle could be con-sidered at seasonal, annual and lifetime scales (cf. Whittle 1997; Chapman 1997b; Zvelebil 1993)■ We do not know whether or to what extent there was year-on-year occupation of single locations; seasonal mobility looks a likely and recurrent feature, and the wider scale of lifetime mobility may also be important. Given this possible, if stili largely hypothet-ical diversity, and compared to the varied pre-Neoli-thic situations or systems sketched above, there is plenty of scope for adjustment of existing practices. To have moved from pre-Neolithic systems of mobil-ity to Starčevo-Koros systems of mobility may not have required major adaptation.
If the Neolithic was not a matter only of nutrition, and if its patterns of settlement could have been de-scended from pre-existing regional practice, what can we say about the beliefs and senses of identity which could have served both to change and define a new world?
SYMBOLIC IDENTITIES
This dimension can be approached in two ways: through material culture, especially pottery and fig-urines, and mortuary rites. Each can be taken in turn. This will then lead to comparison with indigenous traditions including that seen in the Danube Gorges sequence.
Material culture: pottery and figurines
Starčevo and Koros sites are rich in pottery, poor in stone. The quantities of lithic waste and tools are li-mited. There are stone axes, but these are recur-rently quite small and never abundant. In the Koros
phase, one has the impression that flint and similar materials were scarce; their availability varied re-gionally (Kertesz 1996). At Endrod 39, one cache of 101 flints had been put in a pot which was deliber-ately placed in a pit cut through a soil over a pre-existing house. The flints, consisting of various pre-paration flakes, including for platform preparation, probably cante from three nodules of flint from the western Banat (so to the south-east), suggesting both long-range procurement and careful hoarding (Kac-zanoicska et al. 1981). Some other lithic remains were recovered from the site. The abundant material on Starčevo-Koros sites is pottery. Numbers of sherds can ran into the thousands from single featu-res; up to 30 000 were recorded from Pit 1 at Roszke-Ludvar (Trogmayer 1968; John Chapman, pers. comrn). Contexts are known in which pottery has been found in houses or structures (e.g. Tiszajeno: Raczky 1976), but it is also clear that much greater quantities are to be found in the spaces in between, including in pits and other features (Trogmayer 1968.12; Makkay 1992). While there is much to do in the future in terms of residue analysis as a guide to function and breakage/erosion analysis as a guide to deposition, three aspects of pottery can be consi-dered here: the significance of style boundaries, dec-orative motifs and deposition as sherds rather than whole pots.
The traditional culture history approach, with its un-derstandable concern for chronology, has given us a familiar vocabulary of separation into cultures or groups within cultural complexes: Starčevo, Koros, Cri§, and so on. This has rarely been challenged, except by Nandris (1970) and more recently by Mak-kay (1996.36-8). That there are stylistic differences between the pottery of, say, the Koros rivers area of the Hungarian Plain and the southern part of the Vojvodina is not really in doubt. Techniques of roughening and decorating the surface of coarse pot-tery varied and the quantities of the rarer fine wares, including those with painted decoration, seem nor-mally to be greater in Starčevo than in Koros con-texts. What this may have meant in terms of human recognition and social interaction is quite another matter. Most maps of the phenomenon present bor-ders and boundaries, within the normal style of the culture history approach, with little or no overlap (e. g. Dimitrijevič 1974, fig. 1; Garašanin 1979, map 2; Tringham 1971, fig. 10; Kalicz 1990, Taf. 1.2). Really only Brukner (1966, fig. 1; cf. Garašanin 1982.111) has mapped a more subtle picture of overlap in the northern Vojvodina, with areas of 'Starčevo-Koros' distribution between 'Koros' and
'Starčevo'. Individual sites within this area like Do-nja Branjevina may show varying styles from stage to stage in their sequence (cf. Ružič and Pavlovič 1988).
This may indicate a picture of continuum rather than sharp boundaries in ceramic style. Pottery may have been a medium through which convergence and cohesion rather than ethnic difference were expressed, as the culture model has so often, if im-plicitly, implied. Pottery then becomes a symbol of participation rather than badge of separation. It is hard to envisage a closed ethnic unit over the total range of the Starčevo-Koros phenomenon, any more than over the total area of the distribution of early Neolithic white-painted wares, but both could indicate areas of shared practice. Pottery was a new material medium in this area, and if the population using it were indigenous, some of the abundance of pottery might be explained by the novelty of a new medium being used to express versions of existing material practice (cf Stevanovič 1997). The general similarities between, say, indigenous lithic projectile distributions (e. g. Kozlowski 1982) and early Neolithic ceramic distributions might be considerable. The next step will be to examine more closely the manufacture and use of such pottery. It appears to have been easily made, including fine wares. There are some very large vessels in Koros contexts, which may have been used for storage (cf. Banner 1937. 37), but it is possible that many pots were made with a very short use-life in mind. That is certainly one way to explain the abundance of pottery, which could represent as disposable a material in its way as flint in other circumstances.
Pottery was a new medium for visual display. Sur-faces of fine wares were smoothed and/or burnished, and some painted, with generally simple motifs. Sur-faces of 'coarse' wares were also treated, either by roughening or applications of clay and frequently by finger-tip and fingernail impressions. In Koros con-texts there are relief representations of both animals and human or human-like figures (e. g. Banner 1937; Kalicz 1970). The human figures are charac-teristically very stylised, with virtually no sign of in-dividualism in terms of face or expression (Pollock 1995), and recurrent gestures such as bent arms, which might represent particular meanings, actions or contexts (Kalicz 1970; Banner 1937.41 suggested stylised representation of dancing). The animals are in part more recognisable, such as the stag from Csepa or the probable goats (with strongly curved horns) from Hodnezovasarhely-Kotacpart (Kalicz
1970, pls. 6-8)-, others, though said to be species-spe-cific, such as the claimed deer on the vessel from H6dnez6vasarhely-Hamszarito are more ambiguous (Kalicz 1970, pl. 9). Human-like figures and animals occur together on the same large Koros vessels, and the combination must surely be significant; it is not yet clear whether they can also occur separately. This kind of representation seems in general much rarer in Starčevo contexts, though there are inter-esting examples from Donja Branjevina (Garašanin 1979, fig. XXXIX). These are made by incision, and represent animals whose identity is quite unclear; some have projections from their heads which could be either antlers or horns.
The tactility and immediacy of 'coarse ware' decora-tion have been neglected. This decoration is very common, but it seems shortsighted to relegate it to unconscious practice simply because it occurs on so-called coarse pottery. Roughening and finger-tipping bring the human hand into direct contact with the clay. This is a kind of signing of the pots, just as in other contexts and times ročk art can be thought of as signing the land (Bradley 1997). It is possible that particular individual potters or decorators can be distinguished by variations on nail size and shape (Eszter Bdnffv,pers. comm.), but the fact that these 'signatures' are superficially so similar may be the real point, expressing both participation and a merg-ing of individualism in collective practice. This would be ali the more significant if the manufacture and use of pots were episodic, based on either seasonal movement or a rhythm of cyclical gatherings and feasts. These humble sherds, on which so much dust accumulates in the museums of the region, may stili loudly be proclaiming a central and important ethic of participation and communality.
Until very recently, the fact that so much of the pot-tery is represented by broken sherds has gone large-ly unremarked (Makkay 1992.149; Chapman 1996; Chapman forthcoming). It is likely that the signifi-cance of pots was carried over into the practices sur-rounding their deposition. Pots may have been de-liberately broken after use in particular events, gatherings or feasts: another way of explaining the great quantities involved. It can be argued that sherds stood metonymically, as part for whole, for past social interaction, and carried something of their past history into the ground in chosen places, as people consciously selected and deposited them. There is enormous scope in future fieldwork for more de-tailed study of variation in such depositional practice (cf. Last 1996).
Figurines may present both overlaps with and con-trasts to what may be represented in pottery. Starče-vo-Koros figurines are overwhelmingly of human form. Two unique four-footed and double-horned pieces from Szolnok-Szanda may be a rare, if rather abstract, representation of buli imagery (Kalicz and Raczky 1981)] some four-footed lamps may also have schematic animal heads (.Kalicz 1970, fig. 13)-Given the more frequent representation of animals on Koros pots and as figurines in subsequent phases of the sequence, for example from the AVK on the Hungarian Plain (e. g. Domboroczki 1997) or from the Vinča culture further south (Gimbutas 1991), this absence may be significant. It may suggest claims for the centrality of the human form and human identity, although in other contexts these were treat-ed in combination with those of animals.
Traditionally, figurines have been seen as some kind of representation of spirits or ancestral figures (e. g. Gimbutas 1991, and a vast literature). It has also been suggested that figurines in some contexts may represent individuals or 'acting human beings' (e. g. Bailey 1994; Biehl 1996). For the purposes of this discussion (and without wishing to reduce a highly complex issue), it is neither possible nor desirable to settle upon a single meaning. The apparent anonymi-ty of Starčevo-Koros figurines may speak against their representing specific individuals as such. They do not seem to occur in Starčevo-Koros burials, whe-re pots are perhaps the most recurrent (but stili in-frequent) grave good (e. g. Galovič 1964; Trogma-yer 1969). A more typical sort of context is represented by one context at Endrod 39, in which parts of four figurines, already broken, were deposited close together at the base of a substantial pit, with animal bones, sherds and bone tools above and near-by (Makkay 1980.210). A possible inference is that figurines were something held in common, akin to the signings on pots suggested above, and circulated widely among the living until (deliberately) broken and deposited. Nor were figurines necessarily the only token of concepts of ancestry, if this was indeed part of their field of reference. So-called sacrificial pits in Koros contexts held carefully deposited lay-ers of material and finds including pottery, animal bones, fish bone and snails (e. g. Makkay 1992).
Superficially, the overwhelming representation in the figurines is of the mature female form, with va-rying emphasis on heads, breasts, genitalia and but-tocks; limbs seem less important (a contrast which can again be heightened by comparison with pottery and with later figurines). Heads and necks are elon-
gated (and see below); there is some treatment of eyes as schematic slits, and the occasional sugges-tion or representation of nose and mouth. There are some suggestions of hair. Generally faces appear to our eyes abstract, expressionless and anonymous. This may be the combination again of individual and collective. Breasts and genitalia are separately mod-elled or indicated on the bodies of most figurines. They are not normally further emphasised, though occasionally there is a kind of startling realism, as in the Szajol figurine (Raczky 1980). Buttocks and thighs are normally disproportionately large.
As well as the superficial emphasis on the female form, and the apparent anonymity of faces, there is another neglected feature of these figurines: their ambiguity in terms of gender or sexual representation. Is it fanciful to suppose that elongated heads and necks are in fact also a representation or a suggesti-on of erect male genitalia? The same suggestion has been made, independently, for Greek material (Kok-kinidou and Nikolaidou 1997). Many of the Starče-vo-Koros figurines in fact offer quite striking images of the head of the erect penis. One of the most sug-gestive examples is from a Starčevo context at Glad-nice (Garašanin 1979, fig. XXIV), well to the south, and others also occur further south, including in Gre-ece (Kokkinidou and Nikolaidou 1997), but these objects are widespread including within the Koros distribution (see for example Gyomaendr6d 119: Mak-kay 1992; and Szajol: Raczky 1980). The whole figurine may also be regarded as in part a representation of erect male genitalia, in which buttocks beco-me transformed into testicles. There is no need to in-sist on either interpretation to the exclusion of the other. What seems most interesting is the potential ambiguity created, in a medium - fired clay - which itself presents the theme of transformation (Talalay 1993). There is thus in these apparently simple figurines a possibly complex set of beliefs. The human form is emphasised separately from animals. Female form is emphasised, with overt attention to reproductive or sexual parts. Heads and necks are important, but faces are more anonymous. At the same time there is some kind of concern for the combination of female and male gender and/or sexuality. It is a striking pre-sentation of a particular kind of self-consciousness, once again a merging of perhaps several different identities. I will consider below possible differences and continuities with the indigenous system of representation of identity as seen in the Danube Gorges; the concern for reproduction and fertility may be old, while the heightened awareness of several di-mensions of a separate human identity may be new.
Mortuary rites
Starčevo-Koros mortuary rites were simple but var-ied. The principal visible element of such rites seems to have been in settlements or occupations. Not ali occupations contain burials or human remains, and it is hard in the present state of evidence to distin-guish whether burials occur only on particular kinds of site. Gyomaendrod 119, for example, apparently a quite small occupation, has a number of burials, while the larger area opened at Divostin had only one shallow burial of an adult woman, uncertainly attributed to the Starčevo phase (McPherron and Srejovič 1988). From the indications of sequence at Gyomaendrod 119 (Makkay 1992), it seems likely that the rate of deposition was slow: perhaps only one burial every few years at the most. There do not appear so far, in the current state of excavation, to have been cemeteries or burial grounds, so much as episodic accumulations or small concentrations in places chosen and re-chosen for occupation. It has been suggested that a sense of pollution in the Ko-ros culture could have caused site abandonments and short-distance relocations (Chapman 1994b), but this may be too extreme an explanation for spe-cific instances like Gyomaendrod 119. The further obvious implication is that much of the population is not represented in the evidence excavated so far, which could reinforce the sense of fluidity and mo-bility that characterises other aspects of the settlement record and the lifestyle as a whole. The dead may have been used to reinforce the attachment of the living to particular places, but that attachment itself was a broad one.
The diversity of rites is striking. These have been de-scribed often enough before (e. g. Garašanin 1982; Borič 1996; Trogmayer 1969; Chapman 1983; Chapman 1994b), but will bear brief rehearsal in order to contribute to the discussion of lifestyle, relations between individual and collective, and comparison with pre-Neolithic rites; analysis of context-related variation has so far not been systematic enough. Women, men and children are represented in the mortuary record; so far, women might be in the ma-jority (Chapman 1983-8; Zoffmann 1986, for Hun-gary; Borič 1996, table 1 for the Srem region in northern Yugoslavia). The dominant mode was in-humation of fleshed corpses, either contracted or sometimes extended with some flexing of the legs. Single burials are recurrent, though double burials also occur, and small collective deposits are found in both Starčevo contexts, as at Vinča (Garašanin 1982; Letica 1968; the context conld be very early Vinča
culture), and Koros contexts, as at Hodnezovasar-hely-Kotacpart-Vata tanya (Trogmayer 1969; Zof fmann 1986). There are also in Koros contexts par-tial inhumed remains, skull deposits and even rare cremation deposits (Chapman 1994b).
Single burials normally occur either in their own grave pits or in larger, presumably abandoned featu-res normally interpreted as pits or pit-dwellings. It is not yet clear whether there is any structured differ-ence between the remains and their treatment in such differing contexts. Burials have been found in-side structures, as at Szajol and Szanda near Szol-nok, and it is possible that these were deliberately fired following deaths of occupants or 'household' members (Raczky 1982-3; Chapman 1994b; cf. Ste-vanovič and Tringham 1997; Stevanovič 1997). A related example could be the collective deposit at Vinča in a supposed former pit-dwelling (Garašanin 1982). The orientation of the body seems to have varied in Starčevo contexts as a whole (Garašanin 1982)-, a recent discussion of the Srem region evidence suggests greater variation for left-side inhu-mations (Borič 1996, fig. 3a). Less variation is clai-med in Koros contexts (Trogmayer 1969.13). There has been no context-related examination of orientation, to consider body position in relation, for exam-ple, to natural features. It has been suggested that details of the position of heads and upper limbs, as at Zlatara A, could be related to personal identity or position (Borič 1996.74).
Many burials were not accompanied by grave goods. There are early reports of Koros burials with red ochre around the skull (Trogmayer 1969), echoing practices in the Danube Gorges (Radovanovič 1996; Bradley 1998), but ochre does not seem to be an element of Starčevo rites. In various cases whole pots and sherds were deposited with the dead. At Golokut in Srem an adult woman was interred be-low the skull of an aurochs (Borič 1996, and pers. comm,), while there were red deer antlers with a woman at Zlatara B (Borič 1996).
It was formerly suggested that complete inhuma-tions in these contexts might represent more social-ly prominent persons than the partial remains incor-porated into refuse deposits (Chapman 1983.10). It has also been suggested that Starčevo communities emphasised 'certain communal rights' through their burials (Borič 1996.75). I would prefer to empha-sise diversity and fluidity. Diversity and mobility do not seem easily compatible with rigidly fixed social positions. Some of the dead may have been buried
or exposed elsewhere before eventual deposition, or even moved around the landscape before final inter-ment. The contrast then would be between those buried after death and those selected for ancestral veneration. The apparent numerical dominance of women is significant. It was formerly linked to the hypothetically central role of women in hoe agriculture (Chapman 1983-10), but this is to assume that hoe agriculture had a central role in Starčevo-Koros subsistence. It may have more to do with other gen-der-based division of labour or gender-based varia-tion in lifetime mobility. It is tempting to see a link with the superficial dominance of the female form in figurines. Identities and social roles were perhaps much more open than we are accustomed to think of or experience. Burials may have reinforced a sense of plače, but there were many places so reinforced. People were perhaps more attached to regions or landscapes than to particular places alone, and the fluidity of social relations may have allowed the individual or groups to move and to merge freely with others. Burials recurrently present the individual, but the individual is also subsumed in the collective. Once again there is ambiguity (I have dis-cussed the co ncept of the indi vidual more widely elseivhere: Whittle forthcoming).
DESCENTS: COLONISATION, ACCULTURATION AND INDIGENOUS CHANGE
So far, I have east doubt on the applicability of the colonisation hypothesis for the northern Balkans, while leaving the matter open for the southern Balkans. I have indicated that at a broad regional scale there were widely distributed Mesolithic populations in south-east Europe as a whole, which had varying patterns of lifestyle, mobility and subsistence. I have suggested that the early Neolithic northern Balkan lifestyle was based on mobility of varying kinds and a very broad subsistence spectrum; some elements represented, such as sheep and goats in wet Koros contexts, may have had more to do with novelty than practical reason. Identities may also have been open, fluid and ambiguous. Material culture pattern-ing, for example as seen in pottery, looks weak, and we need to break away from the traditional assump-tions of differentiation implicit in the culture model approach. Decoration of pots and their frequent deposition as broken sherds may have served to sub-merge the individual in a wider collective. Burials also celebrate the individual, but without clear em-phasis on particular persons or their social position.
The dead populatecl the whole landscape in varying guises, again merging individual and collective. If the colonisation hypothesis is unreliable, how can we plausibly derive this situation from the indige-nous setting? It is my aim here to suggest refine-ments to existing acculturation models (see also Zve-lebil 1998a; Zvelebil 1998b).
A straightfonvard acculturation model would accept the existence of more or less widely distributed Mesolithic populations, and suggest that under the in-fluence of innovations to the south there followed a series of changes in the northern Balkans, including the adoption of cereal cultivation and animal hus-bandry, including the use of sheep and goats, the adoption of pottery and figurines, built struetures and so on. Such changes might be seen as extensive, driven above ali by change from the outside. While not denying the importance of changes in the situation from the outside, what I wish to explore is the possibility of something more complex.
Indigenous traditions: generalities
Taken again at a broad scale, it is possible to use the south-east European Mesolithic evidence to suggest many elements of continuity of lifestyle. Mesolithic people were regularly mobile, though to varying de-grees, and the possibility of restricted mobility, for example in the Danube Gorges or in the Southern Bug and Dniestr valleys cannot be excluded. Particular places were emphasised by repetition of occupa-tion, from obvious examples like Franchthi Cave and locations in the Danube Gorges to spectacular inland Montenegran caves like Crvena Stijena (Srejovič 1989). A broad spectrum subsistence economy was praetised, and there was long-distance movement of raw materials. Burials reinforced the importance of plače, with examples at Franchthi, Theopetra, and in the Danube Gorges (Jacobsen and Cul/en 1981; Ky-parissi-Apostolika 1995; Radovanovič 1996). Indi-viduals in this world too may have moved freely from group to group; the patterning in material culture is also broad and not sharply differentiated.
In this perspeetive, the scale of early Neolithic changes could actually appear relatively restricted, to the extension of zones of settlement, the limited take-up of some cultivation and husbandry, and the exuberant use of fired clay for pottery and figurines. It is not so much the material conditions of existence that may be at stake, important though those obvi-ously are, as shifts in the sense of identity of individual and collective. Can that further be explored?
Indigenous traditions:
the čase of the Danube Gorges
My discussion wdl principally concern the Danube Gorges. The major features of the phenomenon are well known and need no re-description here (Srejovič 1972; Radovanovič 1996). The chronology of de-velopments in the Gorges is central. There is a large body of opinion which attributes the significance of the Gorges phenomenon principally to its pre-dating the Neolithic (e. g. Srejovič 1972; Srejovič 1989; Bo-roneant 1989; and many others). The more likely sequence, however, is that while some sites in the Gorges can indeed be dated to before the Neolithic in the wider region as represented archaeologically by Starčevo-Koros material, the apogee of the Gorges developments was contemporary with early Neolithic culture elsewhere in the wider region (Whittle 1985.115-8; Radovanovič 1996; Whittle 1996.24-9). From this it follows that the belief system or ideol-ogy seen in its most developed form at Lepenski Vir itself could in some sense have been a resistance to or variation on early Neolithic belief and ideology (Whittle 1985.118; Chapman 1993; Radovanovič 1996; Whittle 1996.44-6). It is not therefore a pre-cursor, but, even more interestingly, a foil to early Neolithic ideology. The Lepenski Vir system is not necessarily completely opposed to that of the early Neolithic, but its major features may serve further to highlight what is new about the early Neolithic sense of identity and belief.
Srejovič himself insisted that there were mythic di-mensions to the symbolism of Lepenski Vir I and II:
... the existence of a specific fish-like deity came into being relatively late in the Lepenski Vir culture. It probably descended from the belief that ali men ivere children of the river, or tlie descendants of mermen, or perhaps from a myth in ivhich ivater, stone, the boul-ders,fish, deer and human heads held the most importantplaces (Srejovič 1972.122).
This kind of interpretation was curiously neglected for a long time, including by this writer. Renewed attention was given to the symbolism of Lepenski Vir by Hodder (1990), but that brief analysis concen-trated on simple binary oppositions between hearth and burials, life and death, and so on. Handsman (1991; cf. Chapman 1993) took note of the carved boulders, but principally as representations of lin-eage ancestors, in a discussion of the development of social relations along presumed lineage divisions.
More recently stili, Bradley (1998) has drawn attention to the unifving features of the materials and prac-tices drawn upon in Lepenski Vir, to suggest a world-view more in harmony with its natural surroundings.
It is possible to go stili further, and the most suc-cessful detailed attempt to develop Srejovič's view has been made by Radovanovič (1996; 1997). This account accepts that Vlasac, only a little downstream in the Upper Gorges, is earlier than Lepenski Vir. The burials there may be of two phases. As else-where in the Gorges, ochre was scattered in an earlier phase on the bodies of the dead (on men, wo-men and children). In its later phase, ochre is scattered only on women, in the pelvic area, becoming perhaps a symbol not just of life but also of birth. Ochre was not a feature of Lepenski Vir burials. There is continued interest in fertility, for example in the combination of female mandibles and hearths, and one might add in the form of red deer antlers near the hearths of phase II (Srejovič 1972.123). An earlier burial in phase le had an aurochs skull by the deceased s shoulder, a red deer skull by one hand and antlers nearby (Srejovič 1972.120, pl 61; grave 7, house 21). Birth symbolism shifts into the houses or shrines in the form of sculptures with vul-vae, for example in Lepenski Vir II house XLIV, thus being transformed from something associated with individuals and becoming intenvoven into a com-plex set of other symbols belonging to a collective heritage. The collective heritage acted as a myth, even as a dogma...' (Radovanovič 1997.88). Other features are important. The heads of the dead at Lepenski Vir (children often under the house or shrine floors, with adults in the spaces in between) were oriented downstream. Sculptures from an early part of Lepenski Vir I onwards present fish-like faces, which become both larger and more accentuated in Lepenski Vir II. These can be seen to represent the massive anadromous beluga, Acipenser huso, though that was largely absent from fish remains them-selves in later levels. In a rather different way to Hodder, Radovanovič comes to a duality between life and death, with the river itself of critical and central importance as the conduit for the passage upstream of the ancestors (as beluga) and the depar-ture downstream of the dead, and as a metaphor for death and endings on the one hand and life and re-turn on the other (Radovanovič 1997.89).
One could add two emphases, both to do with the dynamic development of the sequence. The early burials of Lepenski Vir appear to be very varied in nature, and include partial remains, heads only and
jaws only (Srejovič 1972.117-8). The later burials seem therefore to represent a relatively greater for-malisation of mortuary rites, and perhaps therefore a consolidation also of collective identity, especially if, as I have argued elsewhere (Whittle 1996) the houses were in fact shrines and the whole site a special sanctuary serving a wider area and population.
The other point to stress is once again the wider con-text. These spectacular developments at Lepenski Vir took plače on the chronology advocated here at a time of Neolithic contact. They emphasised a special plače and a special area with a long history. By the apogee of Lepenski Vir II, there were major ideas to do with belonging, the merging of the individual into a wider collective, origins, ancestral return and the destination of the dead, which had developed, am-plified or made explicit earlier ideas to do with the centrality of fertility, reproduction and unity with nature.
It would be naive to suppose that the belief-system represented in the Danube Gorges should reveal that of the whole of Mesolithic south-east Europe. But its major elements may help also to define what was different about early Neolithic ideology, and therefore give further insight into what was involved in the conceptual shifts of an indigenous transition. Ideolo-gies need not necessarily have been completely op-posed. This is not the only likely čase of delay and resistance. The Ertebolle čase springs to mind (Whit-tle 1996, clis 6 and 7, and references), and in that čase some of the long process of stasis may have been conditioned by convergence as much as by dif-ference. The early Neolithic belief-system as sketched earlier was in varying ways to do with belonging, origins and ancestral figures, fertility and reproduction. There were therefore perhaps considerable elements in common at one sort of level. Belonging and iden-tity may have been more ambiguous and fluid in the early Neolithic situation, as discussed above. Perhaps it was so also in the Mesolithic, and the apogee of Lepenski Vir could be seen as an attempt to fix be-haviour into a particular mode. The interest in ances-tors in the early Neolithic seems to have been bound up with a greater interest in the human form and human body, as expressed in the form of figurines and in their often ambiguous gender. There was an interest in animals as separate beings, perhaps a con-cern for human relationships with animals created by the new practices associated with domestication.
Both sets of people, if such a crude distinction can be made, thought about where they came from and
to what they belonged. In the Gorges, this was fo-cused on concepts of the natural world and ances-tors who took natural form, on a cycle of life, reproduction and death. In a wider world, and undoubt-edly affected by developments to the south, other people focused on concepts of a human world, the importance of belonging to a broad community, of tracing descent from ancestors in human form, and of a more conscious difference between people and animals. The human dead were hardly neglected, but their treatment suggests that they were not a central focus in the same way as in the Gorges. I have deliberately tried to avoid simplistic opposition between a Mesolithic and an early Neolithic belief system, nor do I suggest that these would have been uniform; the domus concept (Hodder 1990) runs both risks. But it is as though, as well as the over-laps, there were fundamental divergences: on the one hand, an emphasis on cyclicity, the merging of time, and the importance of death, and on the other, an emphasis on ancestral beginnings, marked time, and participation by the living in social life.
I am trying to avoid both simplistic or universalising models and excessive opposition between putative worldviews. The elements sketched here, however, do recall the contrasts made by several authors be-tween one worldview, associated with at least some recent hunter-gatherers or foragers, in which nature is perceived as a partner, if it is actually conceptual-ly distinguished at ali, and another worldview, thought to be more characteristic of cultivators and others, in which 'nature' is both separated and ap-propriated (lngold 1986; Ingold 1992; Ingold 1993; Bird-David 1990; Bradley 1998). The contrast here, if valid, might best be summed up in the differences in the representation of faces: in the Danube Gorges context a composite image which draws on both fish and humans, but in Starčevo-Koros contexts an image based on human features alone.
People in a process of transition could have drawn on both sets of ideas. There is no need to suppose instant or wholesale change. The Starčevo burials from Golokut and Zlatara B, with their animal remains, strongly echo certain of the deposits at Lepenski Vir, and the diversity of Starčevo-Koros mor-tuary rites also recalls Gorges practices before they became more formalised. On the other hand, new ideas filtering from the south may have spread the quicker or more easily because they were not whol-ly dissimilar to existing ones. The potentially com-plex set of interactions is thus poorly conveyed in the term 'acculturation'. Just as Srejovič emphasised
the importance of myth in the Danube Gorges, so I suppose that mind-sets were changed by myths and stories, by new tellings of the beginnings of the world, of the nature of human social life and of human relationships with the natural world (cf. VVhittle 1996; I ivill discuss these ideas further elseivhe-re). I presume that these would have spread more quickly than anything else, and could have encour-aged people to dwell in parts of south-east Europe previously little used or swiftly passed through.
A final example is the neglected upper level III at Lepenski Vir. The plače was stili used, but much changed (Srejovič 1972). Structures were of irregu-lar shape and earth-sunk, and a small number of burials were set in deep graves next to these. Among other new material culture, extraordinarily abundant pottery replaces the old symbolisms. The motif on one large globular pot from level lila is particularly
telling: an outstretched human hand (Srejovič 1972, pl. VIII).
ACKNOWLEDGEMENTS
I ant particularly grateful to: The British Acadenty and The Hungarian Academy of Sciences for ntaking possible an exchange visit to Hungary in 1997, and to Dr Eszter Banffy for help with my programme; Dušan Borič, Ivana Radovanovič and others, including the Rotary Club, Belgrade, for rnaking possible a visit to Yugoslavia in 1998; and Mihael Budja and the University of Ljubljana for their invitation and sub-sequent hospitality. I arn very grateful also to Dou-glass Bailey, Richard Bradley, Mihael Budja, John Chapman, Vicky Cummings, Detlef Gronenborn, Andrew Sherratt and Marek Zvelebil for comment-ing on draft versions of the manuscript.
REFERENCES
ANDREOU S, FOTIADIS M. and KOTSAKIS K. 1996. Review of Aegean prehistory V: the Neolithic and Bronze Age of northern Greece. American fournal of Archaeologv 100: 537-91.
BAILEY D. 1994. Reading prehistoric figurines as in-dividuals. World Archaeologv 25: 321-31-
BAILEY G. (ed.) 1998. Klithi: Palaeolithic settlement and Quaternary landscapes in northivest Greece.
BANNER J. 1937. Die Ethnologie der Koros-Kultur.
Dolgozatok 13: 32-49.
BARKER G. 1975. Early Neolithic land use in Yugosla-via. Proceedings of the Prehistoric Society 41: 85-104.
BIEHL P. F. 1996. Symbolic communication systems: symbols on anthropomorphic figurines of the Neolithic and Chalcolithic from south-eastern Europe. fournal of European Archaeologv 4: 153-76.
BIRD-DAVID N. 1990. The giving environment: another perspective on the economic system of hun-ter-gatherers. Current Anthropologv 31: 183-96.
BOKONYI S. 1974. Historv of domestic mammals in central and eastern Europe.
1992. The Early Neolithic vertebrate fauna of Endrod 119- In S. Bokonyi (ed.), Cultural and landscape changes on south-east Hungary, 1: re-ports on the Gyomaendr6d project: 195-299. Budapest: Archaeolingua.
BONSALL C, LENNON R, McSWEENEY K, STEWART C., HARKNESS D, BORONEANT V, BARTOSIEWICZ L., PAYTON R. and CHAPMAN J. 1997. Mesolithic and Early Neolithic in the Iron Gates: a palaeodietary perspective,Journal of European Archaeologv 5:50-92.
BORIČ D. 1996. Social dimensions of mortuary prac-tices in the Neolithic: a čase study. Starinar 47: 67-83.
BORONEANT V. 1989- Thoughts on the chronologi-cal relations between the Epi-Palaeolithic and the Neolithic of the Low Danube. In C. Bonsall (ed.), The Mesolithic in Europe: 475-80.
BRADLEY R. 1997. Ročk art and the prehistory of Atlantic Europe: signing the land.
1998. The significance of monuments: on the shap ing of h u ma n experience in Neolith ic a nd Bronze Age Europe.
BROODBANK C. and STRASSER T. F. 1991. Migrant farmers and the colonization of Crete. Antiquity 65: 233-45.
BRUKNER B. 1966. Einige Fragen iiber die Verhalt-nisse der Starčevo und Koros-Gruppe. Acta Antiqua et Archaeologica 10: 7-10.
BUDJA M. 1993- The Neolitization of Europe: Slove-nian aspect. Poročilo o raziskovanju paleolitika, neolitika in eneolitika v Sloveniji 21: 163-93■
BURKE A. 1993. Observation of incremental growth structures in dental cementum using the scanning electron microscope. Archaeozoologia 5: 41-54.
BURKE A. and CASTANET J. 1995. Histological obser-vations of cementum growth in horse teeth and their application to archaeology .Journal of Archaeological Science 22: 479-93■
CAUVIN J. 1994. Naissance des divinites, naissance de T agriculture: la r evolution des symboles au Neolithique.
CHAPMAN J. 1983. Meaning and illusion in the study of burial in Balkan prehistory. In A. Poulter (ed.), Ancient Bulgaria, Volume 1:1-45.
1989.	Demographic trends in neothermal south-east Europe. In C. Bonsall (ed.), The Mesolithic in Europe: 500-15.
1990.	The Neolithic in the Morava-Danube con-fluence area: a regional assessment of settlement pattern. In R. Tringham and D. Krstic, Selevac: a Neolithic village in Yugoslavia: 13-43■
1993. Social power in the Iron Gates Mesolithic. In J. C. Chapman and P. Dolukhanov (eds), Cultural transformations and interactions in eastern Europe: 71-121.
1994a. The origins of farming in south east Europe. Prehistoire Europeenne 6:133-56.
1994b. The living, the dead and the ancestors: tirne, life cycles and the mortuary domain in later European prehistory. In J. Davies (ed.), Ritual
and remembrance: responses to death in human societies: 40-85-
1996. Enchainment, commodification and gen-der in the Balkan Copper Ag z. Journal of European Archaeologr 4: 203-42.
1997a. The origins of tells in eastern Hungary. In P. Topping (ed.), Neolithic landscapes: 139-64.
1997b. Places as timemarks - the social construc-tion of prehistoric landscapes in eastern Hun-gary. In G. Nash (ed.), Semiotics oflandscape: archaeology of mind: 31-45- British Archaeological Reports.
forthcoming. The fractality of human relations in the Mesolithic and Neolithic of southeast Europe. In R. Kertesz and J. Makkay (eds), Szolnok Con-ference 1996.
CHAPMAN J. and MULLER J. 1990. Early farmers in the Mediterranean basin: the Dalmatian evidence. Antiquity 64:127-34.
CHAPMAN J., SHIEL R. and BATOVIČ S. 1997. The changing face of Dalmatia.
CHERRY J. F. 1990. The first colonisation of the Mediterranean islands: a review of recent research. Journal of Mediterranean Archaeologv 3:145-221.
CLASON A. 1980. Padina and Starčevo: game, fish and cattle. Palaeoliistoria 22: 141-73-
COSTANTINI L. 1989- Plant exploitation at Grotta deli' Uzzo, Sicily: new evidence for the transition from Mesolithic to Neolithic subsistence in southern Europe. In D. R. Harris and G. C. Hillman (eds), For-aging and farming. The evolution of plant ex-ploitation: 197-206.
DIMITRIJEVIC S. 1974. Das Problem der Gliederung der Starčevo-Kultur mit besonderer Rucksicht auf den Beitrag der siidpannonischen Fundstellen zur Losung dieses Problems. Materijah 10: 93-U5-
DINAN E. 1996. A preliminary report on the lithic assemblage from the early Holocene level at the Iron Gates site of Baile Herculane. Mesolithic Misce-llany 17 (2): 15-24.
DOMBOROCZKI L. 1997. Fiizesabony-Gubakut. In P. Raczky, T. Kovacz and A. Anders (eds), Utak a mult-
ha: az M3-as autopalya regeszeti leletmentesei: 19-27.
GALOVIČ R. 1964. Neue Funde der Starčevo-Kultur in Mittelserbien und Makedonien. Bericht der Ro-misch-Germanischen Kommission 43-44 (1962-1963): 1-29.
GARAŠANIN M. 1979- Centralbalkanska zona. In M. Garašanin (ed.), Praistorija jugoslovenskih zema-Ija, II, Neolitisko doba: 79-212.
1982. The Stone Age in the central Balkan area. In J. Boardman, I. E. S. Edwards, N. G. L. Ham-mond and E. Sollberger (eds), The Cambridge Ancient History, second edition, Vohune III, Part I, The Prehistory of the Balkans: 75-135.
GATSOV I. 1989. Early Holocene flint assemblages from the Bulgarian Black Sea coast. In C. Bonsall (ed.), The Mesolithic in Europe: 471-4.
GATSOV I. and OZDOGAN M. 1994. Some epi-paleo-lithic sites from NW Turkey. Anatolica 20:97-120.
GIMBUTAS M. 1991. The civilization ofthe god-dess.
GIMBUTAS M, WINN S. and SHIMABUKU D. 1989-Achilleion. A Neolithic settlement in Thessaly, Gre-ece, 6400-5600 BC.
GREENFIELD H. J. and DRA§OVEAN F. 1994. Preli-minary report on the 1992 excavations at Foeni-Sala§: an Early Neolithic Starčevo-Cri§ settlement in the Romanian Banat. Analele Banatului 3: 45-85.
HANDSMAN R. G. 1991. Whose art was found at Le-penski Vir? Gender relations and power in archae-ology. In J. M. Gero and M. W. Conkey (eds), Engen-dering archaeology: ivomen and prehistory: 132-59.
HANSENJ. M. 1991. Thepalaeoethnobotany ofFran-chthi Cave. Blooinington and Indianapolis.
HEUN M, SCHAFER-PREGL R, KLAWAN D., CASTAG-NA R., ACCERBI M, BORGHI B. and SALAMINI F. 1997. Site of einkorn wheat domestication identified by DNA fingerprinting. Science 278: 1312-4.
HODDER I. 1990. The domestication of Europe. Oxford.
HORVATH F. 1989- A survey on the development of Neolithic settlement pattern and house types in the Tisza region. In S. Bokonyi (ed.), Neolithic of south-eastern Europe and its Near Eastern connections: 85-101.
INGOLD T. 1986. The appropriation of nature: essays on human ecology and social relations. Man-chester.
1992. Culture and the perception of the environ-ment. In E. Croll and D. Parkin (eds), Bush base: forest farni: 39-56.
1993- The temporality of the landscape. World Archaeologv 25: 152-74.
JACOBSEN T. W. and CULLEN T. 1981. A considera-tion of mortuary practices in Neolithic Greece: buri-als from Franchthi Cave. In S. C. Humphreys and H. King (eds), Mortality and immortality: the anthro-pology and archaeology of death: 79-101.
KACZANOWSKA M., KOZtOWSKI J. K. and MAKKAYJ. 1981. Flint hoard from Endrod, site 39, Hungary (Koros culture). Acta Archaeologica Carpathica 21: 105-117.
KAISER T. and V0YTEK B. 1983. Sedentism and economic change in the Balkan Neolithic. Journal of Anthropological Archaeologv 2: 323-53.
KALICZ N. 1970. Claygods: the Neolithic period and Copper Age in Hungary. Budapest.
1990. Friihneolithische Siedlungsfunde aus Suduiestungarn. Budapest.
KALICZ N. and RACZKY P. 1981. The precursors to the "horns of consecration" in the southeast European Neolithic. Acta Archaeologica Aacademiae Scientiarum Hungaricae 33: 5-20.
KERTESZ R. 1996. The Mesolithic in the Great Hun-garian Plain. In L. Talas (ed.), At the fringes of three ivorlds: hunter-gatherers andfarmers in the middle Tisza valley: 5-34.
KOKKINIDOU D. and NIKOLAIDOU M. 1997. Body imagery in the Aegean Neolithic: ideological impli-cations of anthropomorphic figurines. In J. Moore and E. Scott (eds), Invisible people andprocesses: uriting gender and childhood into European archaeologv: 88-112.
KOSSE K. 1979- Settlement ecology of the early and Middle Neolithic Koros and Linear Pottery cultures in Hungary. British Archaeological Reports.
KOZtOWSKI J. K. (ed.) 1982. Origin of the ch ipped stone Industries of the early farming cultures in the Balkans.
KYPARISSI-APOSTOLIKA N. 1995. Prehistoric inhabi-tation in Theopetra cave, Thessaly. In J-C. Decourt, B. Helly and K. Gallis (eds), La Thessalie, Colloque in-ternational d'archeologie: 15 annees de recherches (1975-1990), bilans et perspectives, Lyon, 1990: 103-8.
LAST J. 1996. Neolithic houses - a central European perspective. In T. Darvill and J. Thomas (eds), Neolithic houses in north-ivest Europe and beyond: 27-40.
LETICA Z. 1968. Starčevo and Koros at Vinča. Ar-chaeologia Iugoslavica 9:11-18.
LIEBERMAN D. E., DEACON T. W. and MEADOW R. H. 1990. Computer image enhancement and analysis of cementum increments as applied to teeth of Gazella gazella. Journal of Archaeological Science 17:519-
33-
MAKKAY J. 1980. Endrod, Szujokereszt (Komitat Bekeš, Kreis Szarvas). Siedlung der Koros Kultur, Sar-matenzeitliches, Landnahme- und Friiharpadenzeitli-ches Graberfeld; 1976. Mitteilungen des Archdologi-schen Instituts der Ungarischen Akademie der Wis-senscliaften 8/9 (1978/79): 209-13-
1992. Excavations at the Koros culture settlement of Endrod-Oregszolok 119 in 1986-1989- In S. Bokonyi (ed.), Cultural and landscape changes in south-east Hungary, I: reports on the Gyo-maendrod project: 121-93■ Archaeolingua.
1996. Theories about the origin, the distribution and the end of the Koros culture. In L. Ta-
las (ed.), At the fringes of three ivorlds: hunter-gatherers and farmers in the middle Tisza val-ley: 35-53■
MARKEVITCH V. I. 1974. Bugo-Dnestrovskya kultura na territorii Moldavii. Kishinev.
MATEICUICOVA I. forthcoming. Silexindustrie in der altesten LBK Kultur in Mahren und Niederosterreich auf der Basis der Silexindustrie des Lokalmesolithi-
kums. In R. Kertesz and J. Makkay (eds), Szolnok Conference 1996.
McPHERRON A. and SREJOVIČ D. (eds) 1988. Divo-stin and the Neolithic of central Serbia. Pittsburgh.
MILLS S. F 1997. Towards a systematic approach to the study of visual experiences in prehistory: inves-tigations at Neolithic Sesklo. Unpublished MA disser-tation, University of Wales Cardiff.
MILOJČIČ, V. 1960. Hauptergebnisse der deutschen Ausgrabungen in Thessalien 1953-1958. Bonn.
MIRACLE P. 1997. Earlv Holocene foragers in the karst of northern Istria. Poročilo o raziskovanju pa-leolitika, neolitika in eneolitika v Sloveniji, 24: 43-61.
MULLER J. 1994. Das ostadriatische Fruhneolithi-kum: die Impresso-Kultur und die Neolithisierung des Adriaraumes. Berlin.
MRT 1989. Magvarorszdg Regeszeti Topogrdfidja 8. Bekeš Megye Regeszeti Topografiaja. Budapest.
NANDRIS J. 1970. The development and relation-ships of the earlier Greek Neolithic. Man 5:191-213-
NICOLAESCU-PLOP^OR C. S. and PAUNESCU A. 1961. Azilianul de la Baile Herculane in lumina noi-lor cercetari. Studii si Cercetari de Istorie Veche si Arheologie 12: 203-13-
OTTE M, YALCINKAYA I., LEOTARD J-M., KARTAL M., BAR-YOSEF 0., KOZLOWSKI J., BAY0N I. L. and MARSCHACK A. 1995. The Epi-Palaeolithic of Okiizi-ni cave (SW Anatolia) and its mobiliary art. Anti-quity 69: 931-44.
OZDOGAN M. 1989. Neolithic cultures of north-western Turkey. A general appraisal of the evidence and some considerations. In S. Bokonyi (ed.), The Neolithic of southeastern Europe and its Near Eastern connections: 201-15- Varia Archaeologi-ca Hungarica.
1995. Neolithization of Europe: a view front Anatolia. Poročilo o raziskovanju paleolitika, neolitika in eneolitika v Sloveniji 22: 25-61.
1997. The beginning of Neolithic economies in southeastern Europe: an Anatolian perspective. Journal of European Archaeology 5 (2): 1-33-
PATTON M. 1996. Islands in tirne: island sociogeo-grdphy and Mediterranean prehistory. London.
PERLES C. 1990. Les industries lithicjues taillees de Franchthi (Argolide, Grece). II. Les industries clu Me-solithique et du Neolithique initial. Bloomington and Indianapolis.
1992. Systems of exchange and organisation of production in Neolithic Greece. Journal of Mediterranean Archaeologv 5: 115-64.
POLLOCK D. 1995. Masks and the semiotics of iden-tity.Journal ofthe Rova/Anthropological Institute 1:581-97.
RACZKY P. 1976. A Koros kultura leletei Tiszajenon. Archaeologiai Ertesito 103: 171-89.
1980. New figural representations of the Koros culture from the middle Tisza region and their historical connexions. Szolnok Megvei Muzeumi Evkonyve 1979-1980: 5-33■
1982-3. Origins of the custom of burying the dead inside houses in south-east Europe. Szolnok Megvei Muzeumi Evkonyve 1: 5-10.
RADOJČIČ N. and VASIC V. 1997. Archaeological journey in the Iron Gates. Belgrade.
RADOVANOVIČ I. 1996. The Iron Gates Mesolithic. Ann Arbor: International Monographs in Prehistory.
1997. The Lepenski Vir culture: a contribution to interpretation of its ideological aspects. In M. La-zič (ed.), ANTIAC2PON Dragoslavo Srejovič: 87-93•
RUNNELS C. N. 1988. A prehistoric survey of Thes-saly: new light on the Greek Middle Paiaeolithic. Journal of Field Archaeology 15: 277-90.
RUŽIČ M. and PAVLOVIČ N. 1988. Neolithic sites in Serbia explored in the period 1968-1988. In D. Srejovič (ed.), The Neolithic of Serbia: 69-124.
SELMECZI L. 1969. Das Wohnhaus der Koros-Grup-pe von Tiszajeno. Neuere Haustypen des Friihneolithi-kums. /1 Mora Ferenc Muzeum Evkonyve 2: 17-22.
SHERRATT A. 1980. Water, soil and seasonality in early cereal cultivation. World Archaeologv 11:313-30.
1982 a. Mobile resources: settlement and exchan-ge in early agricultural Europe. In C. Renfrew and S. Shennan (eds), Ranking, resource and exchange: 13-26.
1982b. The development of Neolithic and Cop-per Age settlement in the Great Hungarian Plain Part 1: the regional setting. Oxford fournal of Archaeologv T 287-316.
1983. Early agrarian settlement in the Koros region of the Great Hungarian Plain. Acta Archaeo-logica Academiae Scientiarum Hungaricae 35: 155-69.
1997. Changing perspectives in European prehi-story. In A. Sherratt, Economy and society in prehistoric Europe: changing perspectives: 1-34.
SREJOVIČ D. 1972. Lepenski Vir. London.
1989. The Mesolithic of Serbia and Montenegro. In C. Bonsall (ed.), The Mesolithic in Europe: 481-91.
(ed.) 1996. Prehistoric settlements in caves and rock-shelters of Serbia and Montenegro. Fasci-cule 1.
STEVANOVIČ M. 1997. The age of clay: the social dynamics of house destruction. Journal of Anthropological Archaeologv 16:334-95.
STEVANOVIČ M. and TRINGHAM R. 1997. The sig-nificance of houses in the archaeological record of southeast Europe. In M. Lazič (ed.), ANTIAC2PON Dragoslavo Srejovič: 193-207-
TAKACS I. 1992. Fish remains from the Early Neolithic site of Endrood 119- In S. Bokonyi (ed.), Cultural and landscape changes on south-east Hungary, 1: reports on the Gyomaendr6dproject, Archaeo-lingua: 301-11.
TALALAY L. 1993. Deities, dolls and devices: Neolithic figurines from Franchthi Cave, Greece. Bloomington and Indianapolis.
TASIČ N. N. 1997. Hronologija starčevačka kulture. Unpublished PhD thesis, Faculty of Philosophy, Uni-versity of Belgrade.
TODOROVA H. 1995. The Neolithic, Eneolithic and Transitional Period in Bulgarian prehistory. In D.
Bailey and I. Panayotov (eds), Prehistoric Bulgaria: 79-98.
TRINGHAM R. 1971. Hunters, fishers and farmers of eastern Europe, 6000-3000 BC. London.
TROGMAYER 0. 1966. A Koros-csoport lakohazardl. Ukori hazmodell-toredek Roszkevol. Archaeologiai Ertesito 93: 235-40.
1968.	Die Hauptfragen des Neolithikums der un-garischen Siidtiefebene. A Mora Ferenc Muzeum Evkonyve 1: 11-19.
1969.	Die Bestattungen der Kdros-Gruppe. A Mora Ferenc Muzeum Evkonyve 2: 5-15.
VAN ANDEL T. H., GALLIS K. and T0UFEXIS G. 1995. Early Neolithic farming in a Thessalian river landsca-pe, Greece. In J. Lewin, M. G. Macklin and J. C. Wood-ward (eds), Mediterranean Quaternary river envi-ronments: 131-43•
VAN ANDEL T. H. and RUNNELS C. 1987. Beyond the Acropolis: a Greek rural past. Stanford.
WHITTLE A. 1985. Neolithic Europe: a survey. Cam-bridge: Cambridge.
1996.	Europe in the Neolithic: the creation of neiv ivorlds. Cambridge.
1997.	Moving on and moving around: Neolithic settlement mobility. In P. Topping (ed.), Neolithic landscapes: 15-22.
forthcoming. Beziehungen zwischen Individuum und Gruppe: Fragen zur Identitat im Neolithikum der ungarischen Tiefebene. Ethnographisch-Ar-chaologische Zeitschrift.
WIJNEN M. 1982. The early Neolithic I settlement at Sesklo: an early farming community in Thessaly, Greece. Leiden.
WILLIS K. J. and BENNETT K. D. 1994. The Neolithic transition - fact or fiction? Palaeoecological evidence from the Balkans. The Holocene 4: 326-30.
YAKAR J. 1996. The Neolithic transformation in the Near East and Anatolia's role in the Neolithization of southeastern Europe. Poročilo o raziskovanju pale-olitika, neolitika in eneolitika v Sloveniji 23:1-13.
ZOFFMANN Z. K. 1986. Neue anthropologische Funde der neolithischen Koros- und Theiss-Kultur aus Ostungarn. A Mora Ferenc Muzeum Evkonyve 1984/1985: 39-64.
ZVELEBIL M. 1993. Hunters or farmers? The Neolithic and Bronze Age societies of north-east Europe. In J. Chapman and P. Dolukhanov (eds), Cultural transformations and interactions in eastern Europe: 146-62.
1995. Neolithization in eastern Europe: a view from the frontier. Poročilo o raziskovanju pa-leolitika, neolitika in eneolitika v Sloveniji 22: 107-51.
1998a. What's in a name: the Mesolithic, the Neolithic, and social change at the Mesolithic-Neolithic transition. In M. Edmonds and C. Richards
(eds), Understanding the Neolithic of north-ivestern Europe: 1-36.
1998b. Agricultural frontiers, Neolithic origins, and the transition to farming in the Baltic basin. In M. Zvelebil, L. Doman/ska and R. Dennell
(eds), Harvesting the sea, farming the forest: 9-27.
ZVELEBIL M. and DOLUKHANOV P. 1991. Transition to farming in eastern and northern Europe. Journal of World Prehistory 5: 233-78.
ZVELEBIL M. and LILLIE M. forthcoming. Transition to agriculture in eastern Europe. In T. D. Priče (ed.), The first farmers of prehistoric Europe.