110
Documenta Praehistorica XLVII (2020)
Introduction
While numerous studies have explored how the first
pottery was made at Neolithic sites, relatively little
attention has been given to the demands that the
adoption of this introduced technology would have
made on its new practitioners and their responses to
these challenges (cf. Ingold 2000; Michelaki et al.
2012; 2015). Pottery-making was a signature com-
ponent of the Neolithization process in the region
(see Todorova, Vajsov 1993; Özdogan 2009; 2016;
Çilingiroglu 2009; 2012) and involved the trans-
mission of a new technology adaptive to new set-
tings. Communities would have been tasked with
Understanding diversity in Early Neolithic pottery
production> a study case from Southwest Bulgaria
Tanya Dzhanfezova1,2, Chris Doherty 1, and Malgorzata Grȩbska-Kulow 3
1 School of Archaeology, University of Oxford, Oxford, UK
2 St Cyril and St Methodius University of Veliko Tarnovo, BG
tanya.dzhanfezova@arch.ox.ac.uk
3 Regional Museum of History, Blagoevgrad, BG
ABSTRACT – By recovering and interpreting the hidden technological variability in the first pottery
at Ilindentsi-Massovets, this paper reveals the innovative adaptations to local conditions that the
adoption of pottery production, as a new technology, must have involved. Seventy-one samples were
analysed using low-resolution binocular microscopy and high-resolution petrographic and scanning
electron microscopy. The variety established within each of the major components in pottery pro-
duction at the site is interpreted in the context of the local raw materials (availability) and techno-
logical approaches (decision making), thus reaching beyond the traditional interpretative models
that suggest large-scale uniformity in Early Neolithic pottery production across extensive European
regions.
IZVLE∞EK – S pridobivanjem podatkov in njihovo interpretacijo o skritih tehnolo∏kih spremenljivkah
pri prvih lon≠enih posodah na najdi∏≠u Ilindentsi-Massovets v ≠lanku predstavljamo inovativne pri-
lagoditve na lokalne pogoje, ki jih je moral vklju≠evati prehod na lon≠arstvo kot nove tehnologije.
Analizirali smo 71 vzorcev z binokularnim mikroskopom z nizko lo≠ljivostjo ter s petrografskim in
vrsti≠nim mikroskopom z visoko lo≠ljivostjo. Raznolikost, ki smo jo prepoznali pri vsaki od glavnih
komponent pri izdelavi lon≠enine na tem najdi∏≠u, razlagamo v kontekstu lokalnih surovin (razpo-
lo∫ljivost) in tehnolo∏kih pristopov (odlo≠itve), s tem pa prese∫emo tradicionalne interpretativne mo-
dele, ki nakazujejo na obse∫no enovitost v zgodnjeneolitskem lon≠arstvu na obmo≠ju ∏ir∏e evropske
regije.
KEY WORDS – pottery technology; Southeast Europe; Early Neolithic; raw materials; tradition; inno-
vation
KLJU∞NE BESEDE – tehnologija lon≠arstva; jugovzhodna Evropa; zgodnji neolitik; surovine; tradi-
cija; inovacija
Razumevanje raznolikosti v zgodnjeneolitskem lon;arstvu>
[tudijski primer iz jugozahodne Bolgarije
DOI> 10.4312\dp.47.7
Understanding diversity in Early Neolithic pottery production> a study case from Southwest Bulgaria
111
tlement, located between the river and modern
fields, has been excavated. Two chronological peri-
ods are registered: the Early Neolithic settlement
Ilindentsi I developed in the western zone of the ter-
race, whereas the Middle Neolithic Ilindentsi II en-
larged towards the east and north.
Ilindentsi-Massovets provides an excellent opportu-
nity to test biases such as the broad compositional
uniformity of the first pottery and to examine the re-
lative influence of conformity to traditional approa-
ches versus the introduction of innovative pathways
while spreading the new technology. The studied
region offers a wide range of raw materials that can
potentially be used for preparation of earthenware
bodies and the characteristic Early Neolithic red-and-
white-surface decoration. Understanding the actual
choices allows us to examine the decisions made by
the first Neolithic potters at the site, and what influ-
enced pottery transmission.
The Ilindentsi material culture reveals complete cor-
relation with the later Early Neolithic stages estab-
lished at Kovachevo – Kovachevo Ic-Id (see Lichar-
dus-Itten et al. 2002). The site is considered to have
been founded by Kovachevo groups arriving in the
new region at the time when the largest settlement
of Kovachevo experienced the peak of its develop-
ment. The new dispersal areas towards north were
carefully chosen, taking into account the mineral re-
sources (marble, flint, fluorite, etc.) and topogra-
phic advantages, allowing for natural connection via
the Tsaparevska River Valley with the region of the
Vardar River Valley. The connections with the neigh-
bouring regions include the Vardar River Valley,
Northern Greece, Upper Struma River Valley (the Ga-
labnik group), the Star≠evo culture area to the north
and the Karanovo I culture to the east (Grębska-Ku-
low 2017.250–255).
The data from Ilindentsi suggest much greater com-
plexity of the cultural processes that developed in
the Early Neolithic Struma River Valley, thus ques-
tioning the models only limited to a rapid, linear
and unilateral (from south to north) Neolithization
process. The settlement reveals multiscalar and com-
plex connections maintained in various directions,
from west to east and also from north to south – a
situation analogous to other areas of the Balkan
Peninsula (e.g., Özdogan 2016).
Considered in the context of the settlement’s multi-
lateral connections and the variable abundant and
suitable raw materials, an examination of the tech-
producing stylistically acceptable and functional
forms for the first time by working with the raw
materials present in their local landscapes. In this
paper we consider the significance of fabric diver-
sity shown by the first pottery at a key site for the
transmission of pottery-making during the Neolithi-
zation of Southeast Europe. Specifically, this aims to
establish whether the introduced technology ad-
vanced conservatively with an adherence to tradi-
tional production methods and materials (i.e.
adoption), or whether innovative decisions were
made as adjustments in response to local raw mate-
rial constraints or other, non-technical preferences
(adaptation), (cf. Baldi, Roux 2016; Shott 1996; Bo-
gaard et al. 2017; Maran, Stockhammer 2017).
The site and its pottery in the context of the
Neolithization
In Southeast Europe, the Neolithic way of life start-
ed around 6700/6500 cal BC in the southern areas
the Balkan Peninsula (Reingruber et al. 2017; Urem-
Kotsou et al. 2017) and gradually spread north.
The Struma River Valley has been considered as a
major Neolithization route towards the Central Bal-
kans, which connected the Eastern Mediterranean
and the European hinterland around the end of the
7th and beginning of the 6th millennia BC (Nikolov
1989; Lichardus-Itten 1993; Todorova et al. 2007;
Chohadzhiev 2007; Krauß 2011; Krauß et al. 2017).
Two Early Neolithic sites are known in the Middle
Struma River, south of the Kresna gorge (Fig. 1). The
earliest, Kovachevo, dates to the end of the 7th millen-
nium BC and had continuous development (6120–
5640 cal BC), (Demoule et al. 1994; Lichardus-It-
ten et al. 2002). Thirty-five kilometres to the north,
Ilindentsi-Massovets (6510±60 BP or 5500 cal BC)
was settled some 500 years after the start of the
Early Neolithic period in the Middle Struma River
area at Kovachevo (Ia-Ib), placing it in the so-called
developed stage of the Early Neolithic period in Bul-
garia (5700/5600–5460 cal BC).
The site is located at 246–264m above sea level, on
the western slopes of the Pirin mountain and 4km
east of the Struma River. This open settlement cov-
ers an area of thirty decares, its cultural accumula-
tion reaching between 0.50m in the west and 1m
in the east, with a maximum of 1.80m in the dug
structures and the ditched enclosure (Grębska-Kulo-
va et al. 2011; Grębska-Kulow 2017.250–253; Gręb-
ska-Kulow et al. 2018). Situated on agricultural land,
only an area of 350m2 in the periphery of the set-
Tanya Dzhanfezova, Chris Doherty, and Mal⁄gorzata Grȩbska-Kulow
112
nical approaches used in pottery production can po-
tentially expose any tension between tradition and
innovation. Variation in pottery fabrics can help
trace and interpret the actual decision-making that
took place at progressive technological stages, rang-
ing from the body clay preparation to the final sur-
face treatment.
The pottery
White-on-red pottery is considered a hallmark of
Early Neolithic Balkan production (e.g., Lichardus-
Itten et al. 2002; Nikolov 2002), and dominates the
decorated Ilindentsi assemblage across all archaeo-
logical contexts investigated so far (Fig. 2). Three
main region-specific Early Neolithic decorative styles
are present at the site: (a) the Kovachevo cultural
group (Kovachevo Ic-Id); (b) the Upper Thracian val-
ley Karanovo I style; and (c) the Abstract style, cha-
racteristic of the Upper Struma and the Vardar River
Valleys and the vast area between Albania, Northern
Greece, Southwest Bulgaria and the Western Rho-
dopes.
The Ilindentsi pottery is attributed to the syncretic
Kovachevo cultural group style (Kovachevo Ic-Id).
The latter was formed in the Sandanski-Petrich Val-
ley – the area where the Thracian Karanovo I style
and western Abstract style meet. Generally, the Ka-
Fig. 1. Location of Ilindentsi (1) and Kovachevo (2) and immediate geology of the region. A excavations
at Ilindentsi, Southwest Bulgaria, in the Middle Struma River Valley (looking north-eastwards). Marble
white-paint raw materials outcrop in the low hills (middle distance) and as boulders on site (inset); B
location and immediate geology of the Struma River Valley (Sandanski graben)(after Zagorchev 2001.
Fig. 22); C stratigraphy of the Neogene Sandanski sedimentary infill (after Zagorchev 2001.Fig. 26). Map
of the Balkan Peninsula by Captain Blood~commonswiki – Own work, CC BY-SA 3.0, https://commons.
wikimedia.org/w/index.php?curid=675499
Understanding diversity in Early Neolithic pottery production> a study case from Southwest Bulgaria
113
ranovo I style spreads to the east of the Struma Ri-
ver Valley; it is occasionally present in the Middle,
and almost missing in the Upper Struma River areas.
Its typical rectilinear and reticulate painted motifs
cover the vessels’ rims, bodies and pedestals. The
’Abstract’ style, on the other hand, is unknown to
the east, in the Upper Thracian Valley (the Karano-
vo I style area), and consists of curvilinear abstract
and floral, often ‘positive-negative’ motifs, shaping
compositions that cover wider areas of the bodies
but are missing under the rims (cf. Nikolov 2002;
Lichardus et al. 2002; Sanev 2009).
This meeting of three major Early Neolithic white-
painted pottery styles in the southern Middle Stru-
ma River Valley provides an opportunity to examine
the degree to which the raw material
used for the first pottery at a site (Ilin-
dentsi) was influenced by existing tra-
ditions or involved innovation.
Apart from being emblematic for the
period in the region, the ‘white-on-red-
slip’ surface decoration represents the
possibly longest technological chain,
thus allowing for investigation of nu-
merous technological stages and the
choices of artisans. As this style of pot-
tery spread across the region, potters
at each site would have had to make
three fundamental raw material selec-
tions to make a successful pot.
❶ The first requirement would have been to choose
a body clay including any organic or mineral tem-
per deemed necessary for performance or cultu-
ral purposes. Any inclusions, whether temper or
not, would have had to be prevented from reach-
ing the outer surface to avoid degrading the re-
quired red surface decoration.
❷ There is a series of choices regarding the red sur-
face. For example, this can result from careful
burnishing or from rubbing in a little ochre. Al-
ternatively, a higher quality red surface could be
achieved by applying a red clay slip, again with
the option to enhance this with ochre increments
or other combinations.
❸ Finally, a suitable white pigment would be need-
ed to complete the decoration – i.e. the choice
of white-paint material (considering its quality
to be used and preserved), and how to apply it.
Table 1 illustrates how these possible choices com-
bine to offer a variety of ways of making white-on-
red pottery at any new site in the region. Even this
minimalist breakdown highlights the risks of as-
suming that the first pots at successive Neolithic
sites were always made conservatively – i.e. in the
same way as in the earlier sites in the region or
homeland. In this simple outline there are many
possible ways (64, i.e. 4x4x4) by which body, red
surface and white paint options could be combined
– and the complexity would rapidly expand by con-
sidering the many other variables in the overall pro-
duction sequence (temper sources, temper quanti-
ties, firing schedule, etc.).
Given such a wide range of choices, it seems valid
to ask just what happens when pottery is transmit-
ted into new areas and raw-material landscapes
and/or into new communities? Should it simply be
Fig. 2. Ilindentsi sherds representing the three ma-
jor decorative white-on-red painted styles: A Kara-
novo I; B Kovachevo; C abstract style.
Tab. 1. Possible combinations between body fabrics, red surfaces
and white paints at Ilindentsi.
Tanya Dzhanfezova, Chris Doherty, and Mal⁄gorzata Grȩbska-Kulow
114
assumed that the first pots at the Neolithic settle-
ments would have been made exactly the same way
as in the earlier sites in the region or homeland, or
would there be freedom to innovate, perhaps in res-
ponse to the availability or performance of different
raw materials?
This line of enquiry is justified through indications
from other sites where local conditions required the
first potters in the new region to innovate. For exam-
ple, a pilot study of the earliest Neolithic pottery at
Dzhulyunitsa in Central North Bulgaria (Dzhanfezo-
va et al. 2014; 2015) showed that the bodies were
made from a type of clay (loess-derived) that potters
arriving from the Anatolian heartlands (see Mathie-
son et al. 2018) could not have been familiar with.
It was the relatively high concentrations of very fine-
ly divided potassium mica in the loess that enhanced
surface vitrification during firing, and allowed a sim-
ple burnished surface to appear as if having high-
quality slips, regionally termed ‘engobe’ (see Elenski
2006). Furthermore, although mainly locally derived,
a few white paints have mineralogy that could not
have been sourced within the local Lower Danubian
corridor (Dzhanfezova et al. 2014). The site of Dzhu-
lyunitsa therefore demonstrates the skilful but lar-
gely invisible adaptation to local raw materials by
the potters in their visually successful re-creation of
the authentic pottery style. The perhaps unexpected
evidence of the occasional intra-regional movement
of these early wares also reminds us that technology
transfer in the Neolithic was not necessarily a uni-
directional process. Located on another major Neoli-
thization routeway, the Struma River Valley in South-
west Bulgaria, Ilindentsi provides an ideal opportu-
nity to further test the role of innovation as pottery
making advanced.
The traditional approach to studying prehistoric ce-
ramic assemblages is to establish the production se-
quence, the chaîne opératoire (cf. Darvill 2009).
However, this does little to take us beyond the es-
sential mechanics of how pottery was made. In this
paper we interpret raw material use as expressed
through the resultant fabrics of the sites’ first pottery
to ask why a pot was made the way it was.
Materials and methods
Understanding the decisions involved when making
pottery for the first time in a new region requires
that we reconstruct the full range of choices that
could have been made by the potters. Unless we
have some understanding of the number of options
that could have been selected, it is not possible to
interpret the significance of observed fabric variation
in a pottery assemblage. For example, it is obvious
that a narrow range of fabrics in an area with a di-
verse range of raw materials can mean something
very different than the same in an area of more uni-
form geology.
Interpreting the decision making implicit in early
pottery-making begins with reconstructing the full
range of choices that could have been made to pro-
vide the context within which the actual selection
was made. This involves a multi-stage approach. The
availability of raw materials in the area around Ilin-
dentsi has to be modelled from existing geological
maps and publications (see below), and verified by
site visits and sampling. The earliest Ilindentsi pot-
tery fabrics than have to be analysed to determine
which of these possible raw materials were used and
in what combinations.
Sampling and analysis
Excavations at Ilindentsi have yielded 45 669 cera-
mic fragments (800 569 kilograms) including 941
white-painted pottery sherds (Grębska-Kulow et al.
forthcoming). Seventy-one of these were visually
selected on the basis of key stylistic and technologi-
cal variables, including: macro fabrics, wall thickness,
surface treatment, painted decoration style and the
quality of the white paint. The sampled sherds were
usually too small, or represented wall fragments,
thus hampering the reconstruction of the shape or
dimensions of the vessels. As only the periphery of
the settlement has been investigated, quantitative
correlations are not considered feasible at this stage.
The surfaces and fresh breaks of the sherds (Tab. 2)
were first examined under a stereoscopic microscope
(Wild M400) to determine the general inclusions and
matrix characteristics, the stratigraphy of the deco-
ration, the type and quality of the white pigments.
Correspondingly, 20 selected fragments were pre-
pared as standard vertical thin sections and exam-
ined petrographically, using a Leica DM 2500P pola-
rising microscope. These were supplemented where
necessary by loose grain mounts made by dispersing
detached pigment fragments in immersion oil (Hut-
chison Cuff 1996). A more detailed compositional
and micro-textural analysis of the ceramic bodies,
red colour treatment and white-painted decoration
was performed on 21 polished epoxy resin blocks,
using a Jeol 5910 scanning electron microscope with
an Oxford Instruments INCA 300 energy dispersive
x-ray spectrometer (SEM-EDX). Operating conditions
Understanding diversity in Early Neolithic pottery production> a study case from Southwest Bulgaria
115
were varied to suit the target area and task (spot ana-
lysis, backscatter imaging or X-ray mapping), with
conditions for a typical spot analysis being 15vN,
2nA and a 150 seconds count time.
The raw materials geology of Ilindentsi
The site of Ilindentsi (Fig. 1a) is located on the east
side of a major north-south valley corresponding to
a down-faulted block (the Struma Graben) which is
flanked on its east and west side by low mountain
blocks of older crystalline rock (Zagorchev 2001),
(Fig. 1b). The graben is infilled with relatively young
(Neogene) terrestrial sediments that were deposited
by the north-south flowing proto-Struma river and by
lateral inputs of weathered material from the flank-
ing horsts. Stratigraphically, these sediments are sub-
divided into three formations on the basis of their
relative proportions of conglomerates, sands, silty
clays and less frequent fine clays. There is significant
overlap, but broadly the formations coarsen upwards
to the east (Westaway 2006).
The composition of the Neogene sediments directly
reflects the geology of the confining uplands, being
the breakdown products of predominantly medi-
um-grade regional metamorphic rocks (schists, am-
phibolites, calc-silicates, marble and gneisses, gran-
ites and granitoids) (Zagorchev 2001). The north-
south horst-graben-horst structure is defined by ma-
jor fault zones that give rise to local tectonisation of
the parent rocks (this being represented in the Neo-
gene sediments infilling the graben valley). Along
the axis of the valley, the Struma River has deposited
a thinner cover of Quaternary alluvium
with slope deposits locally developed where
the gradient steepens. Both lie above the
Neogene formations with which they are
compositionally similar, having been de-
rived from the same parent rocks.
The Ilindentsi geology thus provided three
possible clay sources for exploitation by
the early Ilindentsi potters. These are: (1)
the Neogene alluvial and proluvial sedi-
ments that lie at shallow depths at all points
in the Struma Valley and would have been
readily accessible; (2) Quaternary alluvium
associated with the Struma River and its
small tributaries; (3) Quaternary surface se-
diments (colluvium) and soils.
Of these three, the Neogene sediments of
the Sandanski formation are the most ac-
cessible option, as they lie directly beneath
the site. Also, being represented by alter-
nating thin beds of differing sand-silt-clay
ratios (Fig. 1c) they would provide an op-
portunity for the potters to fine-tune the
performance properties of the vessels by
targeting specific beds. Almost all Neogene
clayey sediments would be red-firing, as
they are non-calcareous on account of their
granitoid and schist-dominated parentage.
It is important to note that, being predomi-
nantly colluvial in origin, these potential
pottery clays do not show the morphologi-
cal characteristics (overall angularity, sphe-
ricity, degree of sorting, particle size distri-
bution) typical of alluvial sediments. This is
an important consideration when it comesTab. 2. Studied fragments from Ilindentsi-Massovets.
S# Context Method S# Context Method
ID ID
1 C5-46 OM, TS, SEM 37 C4-8 OM
2 C4-11 OM, TS 38 C5-9 OM
3 C4-7 OM, TS, SEM 39 C5-9 OM
4 C4-7 OM,  SEM 40 A3-4 OM, TS
5 B5-8 OM, TS, SEM 41 B5-6 OM
6 C5-42 OM, TS, SEM 42 C5 OM
7 A5-2 OM, SEM 43 B4-13 OM
8 B4-2 OM 44 C4 OM
9 H5-2 OM 45 B4-1 OM
10 B5-18 OM, TS 46 – OM
11 B5-12 OM 47 A5-7 OM
12 B5-12 OM 48 A5-11 OM
13 B4-11 OM, SEM 49 A5-11 OM, TS
14 A3-4 OM, SEM 50 B4-4 OM, TS
15 A5-11 OM 51 C4-1 OM
16 A5-11 OM 52 C4-7 OM
17 C5-9 OM, TS 53 A3-7 OM
18 H5-2 OM 54 A5-11 OM
19 A3-4 OM, TS 55 A5-22 OM
20 C5-9 OM 56 A5-11 OM
21 C5-15 OM, TS 57 B4-3 OM
22 A3-4 OM 58 C5-9 OM
23 B5-3 OM, TS, SEM 59 B5-22 OM
24 A5-9 OM 60 C4-3 OM, TS
25 C5-45 OM 61 B5-1 OM
26 A4-41 OM, TS, SEM 62 H5-2 OM
27 A3-15 OM 63 A5-11 OM
28 C4-5 OM 64 – OM
29 B5-4 OM, TS, SEM 65 C4-1 OM
30 A3-4 OM, SEM 66 C4-11 OM, TS, SEM
31 C5-15 OM 67 C4-23 OM, TS, SEM
32 A3-4 OM, SEM 68 D4-16 OM, SEM
33 C4-8 OM, TS, SEM 69 C4-5 OM, TS, SEM
34 B5-6 OM, SEM 70 B5-15 OM, SEM
35 C4-3 OM 71 A5-11 OM, SEM
36 H5-9 OM
Tanya Dzhanfezova, Chris Doherty, and Mal⁄gorzata Grȩbska-Kulow
116
to deciding whether Ilindentsi pottery fabrics have
been tempered or not.
Better quality red clay is also present as occasional
fine beds within the Neogene Valley infill, and could
be accessed as the latter was worked for bodies and
red slips. The geological reports indicate that over-
all the Neogene sequence reddens towards its east-
ern border (the Kalimantsi Formation, Fig. 1c), thus
the use of these better red clays may have required
an additional short journey. Ochre sources would
have been available along the fault zones that bor-
der the Struma Valley as hematite veins and encru-
station along fault planes. Equally, any area of local-
ly enhanced weathering of any iron mineral-bearing
rock, such as the amphibolites on the west side of
the Struma, or derived boulders thereof, would rot
down to an iron-rich clay (saprolite) suitable for a
low-grade red ochre.
A generic assumption would probably be that the
white slip or paint would be limestone-based, since
limestones and related lithologies (chalks and lime-
rich clay or ‘marl’) are generally by far the easiest
and most abundant sources of white pigments. How-
ever, for most of the Struma Valley the high quality
(i.e. high whiteness) carbonate sources are the hard,
crystalline marbles mainly restricted to along the
western side of the mountainous Pirin horst. Ilin-
dentsi, however, is an interesting exception, beca-
use it does have a very localized deposit of re-work-
ed white marble at lower elevations close to the site
– the Ilindentsi Formation (Westaway 2006). The
distinctive cliffs around the modern town are resis-
tant outcrops of a very coarse, chaotic marble-dom-
inated conglomerate (olistotrome) that was trans-
ported from upland outcrops by landslides during
the late Neogene (Zagorchev 2001). Both the talc-
bearing marble boulders and the marble-derived
matrix of these ultra-coarse conglomerates would
be possible sources of white paint for the Ilindentsi
potters.
Two other potential sources of non-carbonate white
paint sources are also present, though in very restric-
ted quantities. The first is the highly localized talc
serpentine-magnesium amphibole assemblages asso-
ciated either with calc-silicate members within the
metamorphic horsts, or with the very localized gre-
enstones (meta-ophiolite) bodies. The second, more
widespread though less pure alternative source of
white paint, points to the veins or segregation of
mica and/or altered feldspars with the older crystal-
line rocks that flank the Struma graben.
When the use of these three possible body clays
and three or four possible white paints is consider-
ed with the different ways white-on-red ware can be
constructed (Tab. 1), it is clear that there was poten-
tially a lot of variation for making the first pottery
at Ilindentsi. As such, the possibility for on-site in-
novation would have been very real and a conserva-
tive approach for transmission of fixed pottery-mak-
ing to this location should not be assumed as the
only possible option. Decoding the Ilindentsi pottery
fabrics in this local raw material context provides
the opportunity to identify whether tradition or in-
novation prevailed, especially given the presence of
several stylistic variations.
First results
The body fabrics
Petrographic analysis shows that the Ilindentsi pot-
tery assemblage has a continuous range of mineral
fabrics dominated by medium to high-grade regional
metamorphic facies and granitic material (Fig. 3).
Compositionally, the fabrics are consistent with the
geology of the higher ground that immediately bor-
ders the Struma River valley. The combined inclu-
sions population and microtextures indicate that
these clays are from the finer members of the Neo-
gene sediments which are exposed at shallow depth
at Ilindentsi and elsewhere. Any detailed fabric clas-
sification made on the basis of either composition
of microtexture would be somewhat arbitrary since
there is a complete continuum. In terms of decision
making by the Neolithic potters, these Neogene
sandy clays form essentially one very broad fabric
group. Furthermore, the Struma River valley wide
distribution of the Neogene infill means that there
is no unique provenance signal that would allow us
to determine precisely where the body clays were
being worked from.
The only notable exception here, which does quali-
fy as a separate fabric, is represented by a few rela-
tively fine buff-coloured sherds that stand out from
their coarser and red-firing (when oxidized) coun-
terparts. These represent true alluvial sediments and
are not associated with the modern Struma River
but its Neogene predecessor (Fig. 3c). Their use sig-
nals a definite decision having been taken, since
they would not develop a quality red surface on fir-
ing.
The main points arising from the body fabrics of the
analysed 71 samples are:
Understanding diversity in Early Neolithic pottery production> a study case from Southwest Bulgaria
117
❶ All mineral inclusions are natural. No mineral
temper has been added.
➋ All fabrics are consistent with the locally avail-
able Neogene sediments.
➌ All fabrics have essentially the same inclusion
types and clay type. Proportions differ as a natu-
ral continuum. Most are members of a single fab-
ric group.
❹ Most fabrics are sandy. The inclusion vs. clay ma-
trix ratio is high (estimated above 3:1).
❺ Some, including few paler, buff-coloured fabrics,
have organic inclusions. Quantities vary, but are
nearly always low (1 to 20%).
❻ The organic containing fabrics also have numer-
ous mineral inclusions. In some cases, these are
finer (than those in the fragments without orga-
nic inclusions) and show evidence of alluvial
working. It is justified to consider them as a se-
cond though minor fabric group.
❼ A relatively wide range of clay paste inclusion vs.
matrix ratios were clearly acceptable. This points
to: (1) no attempt to be selective when digging
clay; (2) no attempts to achieving a standard fab-
ric, say for optimized performance. If the clay
worked, that was sufficient.
The red surface treatment
Allowing for post-depositional modifications, four
main different red surface treatments have been ob-
served (cf. Middleton 1987). These are indicative of
different intentions and sequences in preparing the
surface and obtaining the red-colour finish (Fig. 4).
The most fundamental difference pointing towards
dissimilar actions of the artisans is between the ac-
tual red engobe, the main distinction being in the
presence of a general compositional planar bound-
ary.
In Group 1, the red surface was produced simply by
the oxidized outer surface of the body: there being
no difference in chemical composition between the
body and the outer red surface. The composition and
distribution of mineral grains in the red oxidation
zone correspond to those observed immediately be-
low the red layer (Fig. 4b).
Group 2 contains fragments with a burnished sur-
face in which there is a slight change in chemical
composition towards the outer surface, most notably
a small increase in iron. This could either result from
simple densification of the outer surface by the me-
chanical action of burnishing, or represents surface
enhancement by the addition of ochre (Fig. 4f).
Group 3 sherds have a definite red slip (Fig. 4a). Un-
like Group 2 there is a marked chemical difference
between the body and red layer. The latter contains
fewer inclusions, most probably due to simple elutia-
tion, and shows a clear and planar
boundary against the body.
Group 4 shows a red engobe, but in
this case the body outer surface was
worked with a tool beforehand, re-
sulting in a slip/body interface boun-
dary that is sharp but not planar
(Fig. 4c-e). Using the same approach,
the result may vary in thickness, con-
sistency, etc. producing thicker and
wider (Fig. 4c) or thinner and finer
red layers (Fig. 4d). Thus, it may be
a layer proceeded by conspicuous
tooling (i.e. an extra step in the pre-
paration of the surface finish) or a
layer of dense body with sharp boun-
dary which was just pressed on with-
out the tooling.
Despite the possible group bounda-
ries crossover, making only the end
members of the category easily re-
cognizable, this is an interesting ca-
tegory suggesting the use of two dif-
Fig. 3. Typical Ilindentsi fabrics. All inclusions are natural, no ad-
ded mineral temper. Plain polarized light, magnification x50. A
large potassium feldspar and clay pellet in poorly sorted angular
sandy matrix; B detail of tectonized granite inclusion; C sporadic
organic inclusions, low frequency suggests natural, not temper;
D iron-rich variety possibly from fault zone (eastern margin of San-
danski graben).
Tanya Dzhanfezova, Chris Doherty, and Mal⁄gorzata Grȩbska-Kulow
118
ferent raw materials – the body clay and red ochre/
clay to produce the red surface, without the one
being derivative of the other.
White-painted decoration
The low-power microscopy indicates that the white
paint differs considerably in terms of the raw mate-
rials used, its uniformity and the method of applica-
tion. The outer appearance of the paint varies be-
tween lustrous, glittery and matt, with the paint lay-
er either strongly adhering to the red surface or
being more friable and loose. The thinner white
paints are more homogenous and have fewer breaks,
whereas the thicker ones tend to be grainy and less
continuous. Allowing for post-production degrada-
tion, these differences point to variable paint visco-
sities and degrees of dispersion of the
white particles.
The shades of white vary between
very bright or yellowish white to dark
grey (Munsell codes 10YR 8/1 – 7.5
YR 8/2 – 10 YR 6/1). Some fragments
show a clear and high-contrast white
paint-red surface boundary, whereas
on others the white paint edges are
more gradational and blurred. Most
white paints contain a few residual
coarser inclusions (up to 0.5mm) that
have survived processing. Immersion-
oil grain mounts of the latter and thin
sections of the complete sherds exa-
mined using a research-grade petro-
graphic microscope confirmed the
presence of two fundamental white
paint groups at Ilindentsi: (1) marble-
based and (2) mica-based.
The marble-based white paints (Group
1, n=4, Fig. 4b; Fig. 5a-c) have a fine
calcareous matrix and contain a range
of residual angular crystalline calcite
grains (<0.5mm), with the larger
grains exhibiting well-developed clea-
vage traces. Petrographic criteria for
identifying marble are: (1) the com-
plete absence of any sedimentary tex-
tures (clastic, biogenic, cement phase,
etc.), which precludes the well-crys-
talline calcite being a coarse diagenic
sparite; (2) the presence of a variable
magnesium-equilibrated non-carbon-
ate impurity assemblage of talc, mag-
nesium chlorite and serpentine and in some cases
magnesium carbonate (dolomite).
Two further subgroups are recognized. Group 1a
(n=2) is characterized by relatively fine-grained
evenly-distributed calcite (marble) grains with occa-
sional large (<0.3mm) talc laths that show little or
no evidence of mechanical disturbance despite their
low hardness (Fig. 5a). Group 1b (n=2) is characte-
rized by talc-serpentine grains which are both larger
and more abundant than the associated marble. Here
the latter are supported by a calcareous matrix with
finer talc and chlorite/serpentine grains (Fig. 5b).
The Group 2 paints are mica-based (n=28, Fig. 4a,c,
d,e; Fig. 5d-f). Scanning electron microscopy with
energy dispersive analysis provides further details
Fig. 4. Contrasting surface decorations. Plain polarized light, ma-
gnification x100. A Mica-based white paint and red slip derivative
of the body clay; B marble-based white paint and red surface re-
sulting from oxidation; C mica-based white paste and thick red im-
pressed slip with tooling; D mica-based white paste and thin red
impressed slip with possible tooling; E very thin mica-based white
paste and thin impressed red slip, note large iron-rich segregation
in the body; F clay-rich fabric with burnished red surface, body and
surface compositionally continuous.
Understanding diversity in Early Neolithic pottery production> a study case from Southwest Bulgaria
119
for the group, the detail being largely beyond the
resolution of optical microscopy. These are diocta-
hedral potassium micas, but they consistently show
significant alkali loss and concomitant hydroxyla-
tion; they are in the process of being transformed
into hydromica, and occasionally to kaolinite. Two
subgroups can be distinguished on the basis of their
microtextures. In the larger subgroup the potassi-
um micas show partial expansion of their layers
(diagnostic fan-shaped terminations with potassium
loss) as a result of weathering (hydroxylation and
kaolinization) (Fig. 5e,f). A less common variety has
much finer mica grains in which sub-grains are non-
orientated, that is they show less evidence of a me-
tamorphic parentage and less evidence of weather-
ing. Carbonate is completely absent from the Group
2 white paints or is present in very minor amounts
that can be considered as a background.
According to our preliminary observations, the ob-
served technological types do not associate with any
particular ceramic categories, shapes or sizes (see
also Spataro 2006; 2011; 2019). At present, a cor-
respondence between the technological approaches
and the specific decorative style has also not been
observed.
Discussion
The aim of this paper is to understand diversity in
the Ilindentsi pottery production, using the raw ma-
terials to look at the balance between adoption (re-
ceived tradition) and adaptation (innovation). The
body fabrics vary widely, though their differences
are mainly in the proportions and shape characteris-
tics of a common set of mineral and rock fragments
that collectively point to essentially the same local
source rocks. On first appearance
this fabric diversity could suggest
complexity of production. For exam-
ple, it might suggest a period of ex-
perimentation as pots are being
made at the site for the first time; or
it could signal the involvement of se-
veral potters, each using a slightly
different clay; or a single potter
using a variety of sources. These
would be valid interpretations, since
three different white-on-red styles
are represented at Ilindentsi, so a
standard body should not be expe-
cted.
However, having demonstrated that
Ilindentsi is located directly on thick
deposits of Neogene sands, silts and
clay that are present as relatively
thin alternating beds, the body fab-
ric variation can be more easily ex-
plained. The observed fabric varia-
tion is natural and simply reflects the
same action rather than indicating
complex decision making. Because
of the frequently alternating sedi-
ment beds, the single act of digging
a clay pit would be capable of pro-
ducing the observed fabric variation,
as combinations of clay beds were
encountered and batched each time
clay for pottery was dug.
Therefore, apart from the organic-
tempered fabrics (already known in
Fig. 5. Details of variable white paints (SEM backscatter micropho-
tographs, false colour emphasizes compositional contrast). A gra-
nular marble variety (in green) showing large talc flake on sur-
face (in orange); B granular marble variety (in green) with mag-
nesium chlorite flake (in orange-brown); C granular marble vari-
ety, thick slip showing poorly sorted angular marble with talk and
serpentine; D mica-based white paint, low contrast in false colour;
E micaceous white paint showing microgranular texture; F potas-
sium mica undergoing weathering, edges hydrating to give fan
form with incipient kaolinite.
Tanya Dzhanfezova, Chris Doherty, and Mal⁄gorzata Grȩbska-Kulow
120
the earliest Early Neolithic Balkan sites), the potters
appear to have been following a single decision –
that is to collect the finer clays immediately avail-
able at the site below the surface soil and use these
without further mineral tempering. As Ilindentsi is
considered to have been populated by expansion of
the earlier Kovachevo community, and the Kovache-
vo potters used essentially the same Neogene clays
without special tempering (see Lichardus-Itten et al.
2002), there seems to be a direct continuation of
how to make pottery bodies at Ilindentsi and no
temptation to change this by adding mineral temper.
The situation is slightly different with the organic in-
clusions. These are found in about thirty per cent of
the examined 71 Ilindentsi sherds, but mostly in very
small amounts (<5%), usually registered microsco-
pically, and often represented by only a few larger
plant fragments (<2 mm). When found at such low
levels, these non-temper inclusions were either ori-
ginally present in the near-surface clay deposits or
became entrained during pottery making. However,
there are some fabrics with significantly larger quan-
tities of organic inclusions (around 30%), in which
the organic material is finely divided and uniformly
distributed. Representing about 10% of the studied
fragments in the assemblage, these fabrics are inter-
preted as having been intentionally tempered with
plant material. Interestingly, even these tempered
fabrics also have the usual high concentration of
mineral inclusions, which would suggest that the or-
ganic temper was not an entirely functional require-
ment, since the pots made with just the same sandy
clay performed perfectly well.
Organic tempering at Ilindentsi represents a less
common way of making a suitable pottery body, but
it cannot be claimed this was an innovation to work
around technological limitations of the local clay,
since the latter already had natural sand inclusions
as not to need (organic) temper. Furthermore, just
as there is no correspondence between the presence
of organic temper and the fineness of the natural
clay, there is no correlation with the type and qual-
ity of the final surface decoration either. Although
most organic tempered fragments have a plain sur-
face, the organic tempered white-painted and very
high-quality red slipped fragments show that there
are no strict rules for the use of organic tempered
clay for making either fine painted or unpainted
coarse vessels.
The production of the red surface at the site shows
wider decision making, as four different types are
observed. Group 1 represents the simplest, least-ef-
fort approach that involved just a light burnishing
to produce a red surface on oxidation during firing.
The Group 2 surfaces are also burnished but are of
a higher quality, indicating a higher investment in
time to create a better burnish since the underlying
body clays are essentially the same as those of Group
1, and do not naturally develop a higher polish with-
out a greater effort. There is also a tentative sugges-
tion in some cases that small increments of red ochre
may have been rubbed into the surface, which clear-
ly would point to greater investment of time and
perhaps ‘special’ materials, though the evidence is
still equivocal here. To produce the Group 3 red sur-
faces, the potters have committed to refining a red
slip either from the body clay or from redder clays
interbedded with the local sources. The reward is
a much higher quality surface, both in colour and re-
flectance terms, because the slip better hides any
body inclusions that might otherwise disrupt the red
surface. Finally, Group 4 shows attempts to further
improve the basic red slip effect. In these cases, there
is clear evidence that unlike Group 3, here the body
surface had been lightly tooled or scraped, presu-
mably to produce a better adhesion between slip
and body, and so a more durable finish. Whereas
the Group 3 slips show a planar contact against the
body, the Group 4 boundaries are irregular in the in-
fill holes, where the scraping has pulled out inclu-
sions from the outer surface of the body.
The variety of technical approaches for making the
red surfaces at this site could be interpreted in seve-
ral ways: (1) as an indication of the production of
various craftsmen, each associated with different pot-
tery traditions; (2) as a presence of ‘imported’ wares
from other sites nearby – though the mineralogy of
the body inclusions means that this would still be
within the Struma River Valley; (3) as a gradually
changing, very slight chronological progression –
perhaps as the potters became more skilled; (4) as
evidence for experimentation. Work is ongoing to
resolve these possibilities.
While the exact reasons for this diversity of red sur-
face production for Ilindentsi cannot be easily pin-
ned down, we can conclude that several different
approaches were accepted, provided these gave the
desired effect. This suggests that fixed rules on how
the ‘red slip’ should be made were not transferred
to Ilindentsi as part of a pottery-making package. In-
novations, or at least different ways of enhancing
the locally available red-surface raw material, were
acceptable and certainly, there was more experimen-
Understanding diversity in Early Neolithic pottery production> a study case from Southwest Bulgaria
121
tation with the red surfaces than with the body pre-
paration (cf. the more complex body clays proces-
sing in Neolithic Greece in Dimoula et al. 2014; Pen-
tedeka, Kotsakis 2014; Dimoula 2017; Urem-Kot-
sou et al. 2014; 2017; Saridaki et al. 2019). Still,
whether these variable red surfaces involved true
local innovation is difficult to conclude, given the
presumed high mobility of the Ilindentsi population
in all directions and the active multilateral connec-
tions with other sites that also produced red-slipped
wares (see Grębska-Kulova et al. 2017).
The true local innovation is a more convincingly de-
monstrated by the white paints. Previous work on
pottery from Kovachevo and Ilindentsi has identi-
fied the white-paint raw materials as ‘white clay’
(Grębska-Kulova et al. 2011.32, 36; Lichardus-Itten
et al. 2002), though exactly what this term refers to
is not clarified. According to the observations in our
current study, the main white paint at Ilindentsi is
based on altered mica or sericite, which is broadly
equivalent to the term ‘white clay’ paint. We can go
further by asking where this white paint would have
been sourced from and what degree of processing,
if any, would its use have required of the potters.
The simplest possibility would be that the white mica
is made from ground rocks which are themselves
mica-rich, specifically the mica schists and gneisses
that form the bulk of the medium- to high-grade
metamorphic uplands (i.e. the Pirin and Ograzhden
horsts) that flank the Struma River Valley (Sandan-
ski graben) to east and west, respectively. The un-
derstanding that such rock could be ground to pro-
duce a very serviceable white paint would translate
to a readily portable decorative tradition, given the
widespread availability of this parent rock near all
points in the Sandanski graben, and indeed across
the metamorphic terrains of Southwest Bulgaria and
Northeast Greece in general. If so, the use of a mica-
based white paint from ground rocks would not be
an Ilindentsi innovation.
However, our SEM-EDA observations argue against
a fresh rock source. A consistent feature of these
white paint mica grains is their partial loss of potas-
sium and a tendency for the interlayer spaces to in-
crease to give fan-shaped terminations (Fig. 5e, f).
Both features suggest the in situ weathering of pri-
mary dioctahedral mica, which, through the alkali
loss, is gradually being replaced by kaolinite. The de-
licate fan-shaped morphologies point to a complete
lack of transportation, which leads us to suggest that
these white micas were worked from weathered out-
crops of schist or mica-granite or, more likely, given
the lack of iron-staining and the presence of incip-
ient kaolinization, from highly leached sections of
the Neogene Sandanski formation. Such conditions
are found near Ilindentsi, where these loose and po-
rous sandy sediments are exposed and deeply in-
cised by the west-flowing tributaries of the Struma
River, an example of which flows just past the site.
The resulting steep gradients and low water table
provide the ideal conditions for leaching the gran-
ite and schist derived micas in situ, involving the
mechanical transport that would disrupt the delicate
hydromica fans. The petrography therefore suggests
that for the main white paint group the Ilindentsi
potters were making selective use of these highly
leached micaceous beds of the Neogene formations.
This mica-based white paint technique at Ilindentsi
could have been a continuation (adoption) of a Ko-
vachevo technique, as the latter site is earlier than
Ilindentsi and also lies on the same Neogene sedi-
ments.
The review of the white-paint raw material land-
scape around Ilindentsi suggests that the two mar-
ble-based white paints could be interpreted in sev-
eral ways. They could represent the acquisition of
marble and the addition of two different (but relat-
ed) combinations of ‘impurities’. This would imply
a special journey or an extra stage in the overall
chaîne opératoire, since the usual talc-based rocks
(‘greenstones’) are not immediately local and have
a very limited outcrop. A second possibility is that
the talc-based impurities were introduced acciden-
tally, as a by-product of crushing and grinding mar-
ble to produce a white ground. The impurity assem-
blage of talc, serpentine and magnesium chlorite
could also be introduced by using low-grade green-
stone tools to process the marble. However, the larg-
er and very fragile talc laths were observed to show
no evidence of mechanical damage (Fig. 5a). Finally,
these impurities may have been present in the mar-
ble collected and processed by the potters, and the
two groups simply may refer to the opportunistic
use of available material, and thus that the potters
were unaware of the presence of these talc-serpen-
tine-chlorite impurities as they were primarily after
crystalline (sparkly) calcite.
Detailed petrographic analysis supports the latter
possibility and is an example of true local innova-
tion. Our preliminary observations indicate that the
marble-based white slips are present in the preced-
ing site of Kovachevo, with source material being
the in situ marble outcrops that are nearby. Being
Tanya Dzhanfezova, Chris Doherty, and Mal⁄gorzata Grȩbska-Kulow
122
impure, this marble contains talc as a sporadically
distributed impurity, and the working of such mate-
rial would give the Group 1a marble-based paints –
whose use at Ilindentsi is merely a continuity (adop-
tion) of this Kovachevo tradition. As Figure 1a clear-
ly shows, large white boulders of marble are actual-
ly found across the excavation area and would have
been very convenient to use.
However, the Group 1b white paints are an Ilinden-
tsi specialty. The levels of talc and serpentine in
these paints exceed that expected, if marble boul-
ders were simply being worked. Talc could have
been added as a supplement, perhaps to increase
the sparkly appearance, but this may not have been
necessary. As discussed above, the marble source at
Ilindentsi is not an undisturbed primary metamor-
phic rock but an ultra-coarse landslide deposit (the
Ilindentsi Member) that forms easily accessible low
cliffs (Fig. 1), consisting of fractured boulders of
marble set in a powdery matrix of marble-derived
sediment. The latter, which is not present at the ear-
lier Kovachevo, is naturally talc-enriched and seems
to have been targeted as a highly localized source
for the Group 1b white paint raw material.
Conclusion
Conventional chaîne opératoire studies focus on
how the first artifacts were made, but the questions
that need to be asked are why they were made this
way and what this implies in terms of craft transfer.
The approach advocated here for Ilindentsi is based
on the observed variation in pottery production com-
pared within the raw material context, and thus
the choice available to the potter, with a focus on
the three major components of the white-on-red
painted ware – the body, surface and white paint.
This brings us closer to the task of distinguishing the
processes of technology adoption or adaptation, in-
cluding any experimental phase. If pottery-making
was adopted without any local innovation, we can
look at the ways in which bodies, red surface and
white paints were combined to point to the source of
technological transfer. However, if local innovation
was dominant, and traditional approaches discarded,
this would argue for pottery moving involving an
experimental stage. Furthermore, this also opens
new avenues for consideration of the reasons for
such local modifications of traditional ‘recipes’.
This review shows that, from a purely raw-materi-
als basis, the Ilindentsi artisans had a choice of var-
ious raw materials, hence the opportunity for expe-
rimentation to produce their pottery – i.e. they could
have either adopted or adapted the received tech-
nology. The detailed study of Ilindentsi pottery, con-
sidered in the context of the individual settlement
local geology, shows a wide variety of technical ap-
proaches and raw-material choices. This allows us
to assess the interplay between tradition and inno-
vation in pottery production, acknowledging that
the real complexity would have even be greater. For
example, we also consider the concept of ‘quality’ –
whether this is defined by whiteness of the paint or
its ability to stick. Equally, we could relate these ma-
terial choices to distance, such as whether a special
journey was necessary to acquire them.
Further complexity is introduced when the three dif-
ferent decorative styles are considered. There is no
correspondence between the individual technologi-
cal approaches and specific ornamental styles, and
this refers to each of the three major components –
body fabrics, red surfaces and white paints. No ob-
vious correlation was established between the dec-
orative styles and specific body types, nor with spe-
cific red-surface treatment or white-paint raw mate-
rials. Strict combinations between the various com-
ponents or ‘recipes’ are missing, and no obvious cor-
relation is registered between the raw materials used
and the quality of the vessels (if we define this in
terms of vessels’ thickness or surface appearance).
In a broader archaeological context, the results cor-
respond with the assumption that Ilindentsi is clo-
sely related to Kovachevo (the parent site), but is
also widely open in all directions, despite the sur-
rounding mountain ridges. The similarities with Ko-
vachevo lay in the concept of shaping the settlement
space (the system of ditches and the special orga-
nization of the households), the architecture, the
white-painted style, and a series of technologies, in-
cluding the prevailing mica-based white-paint recipe
and the marble working. At the same time, various
‘material culture’ aspects (Grębska-Kulow et al. forth-
coming) reveal that the site maintained active con-
tacts in all directions – northern Greece to the south,
the Thracian valley to the east and the Vardar River
valley to the north-west. The latter, most probably,
also relates to the presence of several decorative
styles and a variety of technological approaches to
make the Early Neolithic pottery at Ilindentsi.
The demonstrated variability of the white-on-red
painted wares (traditionally considered as a techno-
logically consistent group), and the high number of
Understanding diversity in Early Neolithic pottery production> a study case from Southwest Bulgaria
123
possible combinations between the three main com-
ponents of the category, reveal the complexity of
this early technology, characteristic for broad Euro-
pean areas. Exploring the tension between adoption
This project has received funding from the European Union’s Horizon 2020 research and Innovation Programme
under the Marie Skłodowska Curie Grant Agreement No 798143 – the MINERVA project ‘Mapping intentionality:
demonstrating innovation in Neolithic pottery uptake in the Eastern Balkans’ (H2020-MSCA-IF-2017). The first
results were presented at the 25th Annual Meeting of the European Association of Archaeologists in Bern, Switzer-
land, 4–7 September 2019 (Dzhanfezova, Doherty 2019; Dzhanfezova, Grębska-Kulow 2019).
ACKNOWLEDGEMENTS
Baldi J., Roux V. 2016. The innovation of the potter’s
wheel: A comparative perspective between Mesopotamia
and the southern Levant. Levant 48: 236–253.
Bogaard A., Filipovi≤ D., Fairbairn A., Green L., Stroud E.,
Fuller D., and Charles M. 2017. Agricultural innovation
and resilience in a long-lived early farming community:
the 1500-year sequence at Neolithic-early Chalcolithic Ça-
talhöyük, central Anatolia. Anatolian studies 67: 1–28.
https://doi.org/10.1017/S0066154617000072
Chohadzhiev St. 2007. Neolithic and Chalcolithic Cul-
tures in the Struma River Basin. Faber. Veliko Tarnovo.
Çilingiroglu Ç. 2009. Neolithic red slipped and burnished
wares: recognizing their broad distribution. In H. Saglam-
timur, E. Abay, Ü. Aylin, A. Batmaz, F. Dedeoglu, M. Erda-
kıran, M. Bastürk, and E. Konakçı (eds.), Studies in Ho-
nour of Altan Çilingiroglu. A life dedicated to Urartu on
the shores of the Upper Sea. Arkeoloji ve Sanat Yayınları.
Galatasaray: 203–216.
2012. The Neolithic Pottery of Ulucak in Aegean Tur-
key: Organization of production, interregional com-
parisons and relative chronology. British Archaeologi-
cal Reports IS 2426. BAR Publishing. Oxford.
Darvill T. 2009. The Concise Oxford Dictionary of Ar-
chaeology. Oxford University Press. Oxford.
https://doi.org/10.1093/acref/9780199534043.001.0001
Dimoula A., Pentedeka A., and Filis K. 2014. Lete I. The
pottery of a neolithic site in central Macedonia 100 years
after. In E. Stefani, N. Merousis, A. Dimoula (eds.), A cen-
tury of research in prehistoric Macedonia. Internatio-
nal conference proceedings. Archaeological Museum of
Thessaloniki, 22–24 November 2012. Thessaloniki: 491–
504.
Dimoula A. 2017. Early pottery mobility: The case of early
Neolithic Thessaly, Greece. Journal of Archaeological
Science: Reports 12: 209–218.
https://doi.org/10.1016/j.jasrep.2017.01.008
Démoule J.-P., Lichardus-Itten M. 1994. Fouilles franco-bul-
gares du site néolithique ancien de Kova≠evo (Bulgarie
du Sud-Ouest): Rapport préliminaire (campagnes 1986–
1993). Bulletin de correspondance hellénique 118(2):
561–618. https://doi.org/10.3406/bch.1994.6993
Dzhanfezova T., Doherty C. 2019. Starting point or next
stop? Early Neolithic pottery production in Southeast Eu-
rope. 25th EAA Annual Meeting (Bern, 2019) – Abstract
Book. European Association of Archaeologists. Bern. Au-
gust 2019: 585.
Dzhanfezova T., Grębska-Kulow M. 2019. Pottery tradition
in transition? The passage from the Early to the Middle
Neolithic period at Ilindentsi, Southwest Bulgaria. 25th
EAA Annual Meeting (Bern, 2019) – Abstract Book. Euro-
pean Association of Archaeologists. August 2019. Bern:
586.
Dzhanfezova T., Doherty C., and Elenski N. 2014. Shaping
a future of painting: the early Neolithic pottery from Dzhu-
lyunitsa, North Central Bulgaria. Bulgarian e-Journal of
Archaeology: 137–159. https://be-ja.org/index.php/journal/
article/view/be-ja-4-2-2014-137-159
2015. Unpacking the Early Neolithic. Samarskii nau-
chni vestnik 3(12): 63–71.
Elenski N. 2006. Sondazhni prouchvamiya na rannoneo-
litnoto selishte Dzhulyunitsa-Smurdesh, Velikoturnovsko
(predvaritelno suobshtenie). Archeologiya 47: 96–117.
(in Bulgarian)
References
∴
and adaptation, which could be hidden in each of
these three major components, has the potential to
reveal technological transfer and inter-site relations
within Early Neolithic context.
Tanya Dzhanfezova, Chris Doherty, and Mal⁄gorzata Grȩbska-Kulow
124
Grębska-Kulova M., Kulov I., Salanova L., Vieugué J., and
Gomart L. 2011. The Early Neolithic settlement of Ilinden-
tsi, Strumyani Municipality. Excavations 2004–2009. Ar-
cheologiya 1: 24–43.
Grębska-Kulow M. 2017. The Neolithic Period in South-
west Bulgaria. In E. Stojanova Kanzurova (ed.), Dragi∏a
Zdravkovski in memoriam. Cultural Manifestations
during the Neolithic period on the Territory of the Re-
public of Macedonia and Neighboring Regions. Skopje:
243–279.
Grębska-Kulow M., Zidarov P., Vasilev I., Suvandzhiev I.,
Whitford B., Gurova M., Marinova-Wolf E., and De Cupere
B. 2018. Excavations at Ilindentsi-Massovets Neolithic
site in Southwestern Bulgaria. In Archaeological discove-
ries and excavations in 2017. Bulgarian Academy of Sci-
ences, National Archaeological Institute with Museum. So-
fia: 27–29. http://series.naim.bg/index.php/ADE/issue/vi
ew/Archaeological%20Discoveries%20and%20Excavatio
ns%20for%202017
Grębska-Kulow M. et al. forthcoming. The Early Neolithic
site at Ilindentsi, South-West Bulgaria.
Hutchison Cuff Y. 1996. Ceramic technology for potters
and sculptors. University of Pennsylvania Press. Philadel-
phia.
Ingold T. The temporality of the landscape. In T. Ingold
(ed.), The Perception of the Environment: Essays on
Livelihood, Dwelling and Skill. Routledge. London and
New York: 189–208.
Lichardus-Itten M. 1993. La vallée du Strymon – une route
au Néolithique ancien? In V. Nikolov, L. Pernicheva, and
S. Sirakova (eds.), Praistoricheski nahodki i izsledvania.
Sbornik v pamet na prof. G. I. Georgiev. Sofia: 69–72.
Lichardus-Itten M., Demoule J.-P., Perni≠eva L., Grębska-
Kulova M., and Kulov I. 2002. The site of Kova≠evo and
the beginnings of the Neolithic period in southwestern
Bulgaria. In M. Lichardus-Itten, J. Lichardus, and V. Niko-
lov (eds.), Beiträge zu jungsteinzeitlichen Forschungen
in Bulgarien. Saarbrücker Beiträge zur Altertumskunde
74. Habelt. Bonn: 99–158.
Krauß R. 2011. Neolithisation between Northwest Ana-
tolia and the Carpathian Basin – an introduction. In D.
Ciobotaru, B. Horejs, and R. Krauss (eds.), Beginnings –
New Research in the Appearance of the Neolithic Be-
tween Northwest Anatolia and the Carpathian Basin.
Papers of the International Workshop, 8th–9th April 2009,
Istanbul, Organized by Dan Ciobotaru, Barbara Horejs and
Raiko Krauss. Verlag Marie Leidorf. Rahden-Westf.: 1–7.
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
Maran J., Stockhammer P. 2017. Appropriating Innova-
tions: Entangled Knowledge in Eurasia, 5000–1500 BCE.
Oxbow Books. Oxford.
Mathieson I. and 116 co-authors. 2018. The genomic his-
tory of southeastern Europe. Nature: 197–203.
https://doi.org/10.1038/nature25778.
Michelaki K., Hancock R. G. V., and Braun G. V. 2012.
Using provenance data to assess archaeological land-
scapes: an example from Calabria, Italy. Journal of Ar-
chaeological Science 39: 234–246.
https://doi.org/10.1016/j.jas.2011.08.034
Michelaki K., Braun G. V., and Hancock R. G. V. 2015. Lo-
cal Clay Sources as Histories of Human landscape interac-
tions: a ceramic taskscape perspective. Journal of Archa-
eological Method and Theory 22: 783–827.
https://doi.org/10.1007/s10816-014-9204-0
Middleton A. P. 1987. Technological investigation of the
coatings on some ‘haematite-coated’ pottery from South-
ern England. Archaeometry 29: 250–261.
Munsell Soil Color Charts. 1994. Macbeth Division of Kol-
lmorgen Instruments. Corporation. New Windsor, NY.
Nikolov V. 1989. Das Flustal der Struma als Teil der Stra-
se von Anatolien nach Mitteleuropa. In S. Bőkőnyi (ed.),
Neolithic of Southeastern Europe and its Near Eastern
Connections. Varia Archaeologica Hungarica 2. Budapest:
191–199.
2002. Frühneolitische Bemalte Ornamentik. Self-pub-
lishing. Sofia.
Özdogan M. 2009. Earliest use of pottery in Anatolia. In
D. Gheorghiu (ed.), Early Farmers, Late Foragers, and
Ceramic Traditions: On the Beginning of Pottery in the
Near East and Europe. Cambridge Scholars Publishing.
Cambridge: 22–43.
2016. The Earliest Farmers of Europe. Where did they
come from? In K. Bacvarov, R. Gleser (eds.), Southeast
Europe and Anatolia in Prehistory. Essays in honor
of Vassil Nikolov on his 65th anniversary. Universitäts-
forschungen zur prähistorischen Archäologie. Band 293.
Verlag dr. Rudolf Habelt GMBH. Bonn: 25–57.
Pentedeka A., Kotsakis K. 2008. Thin sectioning Neolithic
identities: the red monochrome ware (A1) from Middle
Neolithic Sesklo, Thessaly. In G. Phakorellēs, N. Zacharias,
and K. Polikreti (eds.), Proceedings of the 4th Symposium
of the Hellenic Society for Archaeometry. National Hel-
Understanding diversity in Early Neolithic pottery production> a study case from Southwest Bulgaria
125
lenic Research Foundation, Athens 28–31 May 2003. Bri-
tish Archaeological Reports IS 1746. Archaeopress. Oxford:
305–311.
Reingruber A., Toufexis G., Kyparissi-Apostolika N., Aneta-
kis M., Maniatis Y., and Facorellis Y. 2017. Neolithic Thes-
saly radiocarbon dated periods and phases. Documenta
Praehistorica 44: 34–53. https://doi.org/10.4312/dp.44.3
Sanev V. 2009. Anzabegovo. The Early and Middle Neo-
lithic Settlement in Macedonia. Excavations of the Ma-
cedonian Team. Study of the data from the excavation
was completed in 1977. Zavod za za∏tita na spomenicite
na kulturata i muzej. πtip.
Saridaki N., Kotsakis K, Urem-Kotsou D., Papadakou T.,
and Papaioannou A. 2019. Pottery production at Neolithic
Pieria, Macedonia, Greece. In S. Amicone, P. S. Quinn, M.
Mari≤, N. Mirkovi≤-Mari≤, and M. Radivojevi≤ (eds.), Tra-
cing Pottery Making Recipes in the Prehistoric Balkans
6th–4th Millennia BC. Archaeopress. Oxford: 129–143.
https://doi.org/10.2307/j.ctvndv954
Shott M. J. 1996. Innovation and Selection in Prehistory:
a case study from the American Bottom. In G. Odell (ed.),
Stone Tools: Theoretical Insights into Human Prehistory.
Plenum. New York: 279–309. 
https://doi.org/10.1007/978-1-4899-0173-6_11
Spataro M. 2006. Pottery typology versus technological
choices: an Early Neolithic case study from Banat (Roma-
nia). Analele Banatului. Arheologie (Istorie) XIV–1: 63–
78.
2011. Technological uniformity? Early Neolithic ceramic
vessels, cult objects, net weights and daub production
in Romania. In S. A. Luca, C. Suciu (eds.), The first Neo-
lithic sites in Central/South-East European Transect
vol. II. Early Neolithic Star≠evo-Cris sites on the ter-
ritory of Romania. British Archaeological Reports IS
2188. Archaeopress. Oxford: 63–78.
Spataro M., Katsarov G., Todorova N., Tsurev A., Nikolova
N., Yaneva M., and Bacvarov K. 2019. The chaîne opéra-
toire of 6th millennium BC pottery making in the Maritsa
Valley, Bulgaria: ceramics from Nova Nadezhda. Praehi-
storische Zeitschrift 94(1): 1–30.
Todorova H., Vajsov I. 1993. The Neolithic in Bulgaria.
Nauka i izkustvo. Sofia.
Todorova H., Stefanovich M., and Ivanov G. 2007. Stry-
mon in der Urgeschichte. In H. Todorova, M. Stefano-
vich, and G. Ivanov (eds.), The Struma/Strymon River
Valley in Prehistory. In the Steps of James Harvey Gaul
2. Gerda Henkel Stiftung. Museum of History-Kyustendil.
Sofia: VII-VIII.
Urem-Kotsou D., Papaioannou A., Papadakou T., Saridaki
N., and Intze Z. 2014. Pottery and stylistic boundaries.
Early and middle neolithic pottery in Macedonia. In E. Ste-
fani, N. Merousis, and A. Dimoula (eds.), A century of re-
search in prehistoric Macedonia. International confe-
rence proceedings. Archaeological Museum of Thessalo-
niki 22–24 November 2012. Thessaloniki: 505–518.
Urem-Kotsou D., Dimoula A., Elezi G., Papadakou T., Papa-
ioannou A., Saridaki N., Siamidou I., Silva T., Tzemopou-
lou E., and Kotsakis K. 2017. Patterns in Contemporane-
ous Ceramic Traditions: Inter-Regional Relations between
Thessaly and Macedonia during the Early and Middle Neo-
lithic. In A. Sarris, E. Kalogiropoulou, T. Kalayci, and L. Ka-
rimali (eds.), Communities, Landscapes, and Interaction
in Neolithic Greece. International Monographs in Prehi-
story. Archaeological Series 20. Berghahn Books. New
York: 324–338.
Westaway R. 2006. Late Cenozoic extension in SW Bulga-
ria: a synthesis. Geological Society. Special Publications
260(1): 557–590.
Zagorchev I. 2001. Introduction to the geology of SW Bul-
garia. Geologica Balcanica 31: 3–52.