UDK 903(392:234.42-11)"6343"
Documenta Praehistorica XXV (Poročilo o raziskovanju paleolitika, neolitika in eneolitika v Sloveniji XXV)
On the problem of the Anatolian-Balkan relations during the Early Neolithic in Thrace
Tatiana Štefanova
Bulgarian Academy of Science, Prehistory Department, Archaeological Institute with Museum, Sofia
ABSTRACT - Tlie paper concentrates mainly on the typological analysis and distribution ofEarlv Neolithic paintedpotterj assemblages in Thrace and Eastern Rhodope Mountains. It is hypothesised that the Hoca Cesme cultural influence correlates with the processes ofthe setting up and development of Balkan early farming communities in Eastern Rhodope Mountains.
POVZETEK - V članku se osredotočamo predvsem na tipološko analizo in razprostranjenost zgodnje-neolitske slikane keramike v Trakiji in vzhodnih Rodopih. Domnevamo, da je nastanek in razvoj zgodnjih balkanskih kmetovalskih skupnosti v vzhodnih Rodopih povezan z vplivom kulture Hoca Cesme.
This paper is not aimed at presenting a new concept of the Neolithization of the Balkans. It is rather an attempt to add new data to the complicated and stili unclear processes of the setting up and development of Balkan early farming communities through the interpretation of a "special čase". The archaeological excavations at the Neolithic site Hoca Qesme in Eastern Thrace, Turkey and the new evidence has provided the opportunity for a re-interpretation of some old finds from the Neolithic site at Krumov-grad in the Eastern Rhodope Mountains, Bulgaria.
KRUMOVGRAD
The Neolithic site was situated on a low terrace on the left bank of the Krumovitza river, a tributary of the Arda river which now lies under the central part of the modern town of Krumovgrad in the Eastern Rhodope Mountains, Bulgaria. The site was located in 1974 because of some urban construction activi-ty. The materials available were unearthed in a central excavation made for block foundations, cover-ing 560 m2, and two additional small trenches. As almost ali of the deposits were scraped out by ma-chines in the central excavation, trench 1, covering 12 sq. 111, yielded the basic information on stratigra-phy and the site sequence. A mixed layer, containing
pottery from the Middle, Late Neolithic and Early Chalcolithic periods and the Early Bronze age, cov-ered six Early Neolithic construction levels dated to the Karanovo I period; the average thickness of each level being 0.30-0.40 m. They yielded parts of hous-es with wattle and daub construction and plastered floors; hearths and domed ovens (one of them 1.40 m in diameter) made of stamped clay were found in the houses (fomes 1994). The pottery is typical of the Karanovo I period: coarse, semi-coarse and fine ware (Fig. 1). The surface of the coarse and semi-coarse ware is uncoated, smoothed or rusticated; sometimes decorated with plastic or incised orna-ments. The pottery assemblage includes plates, bowls and necked jars with vertical string-hole lugs. The fine pottery is red slipped, brown or grey-black. Ali six construction levels yielded white painted pot-tery. The red or reddish-brown fine ware is white painted: open plates, bowls, tulip-shaped vessels, necked jars on pedestal bases and lids (Fig 1. 7-9). The channelled decoration and plastic knobs are reg-istered on bowls and necked jars on pedestal bases. No 14C dates are available for the site. Certain sha-pes, considered typical of the Karanovo II period, are present in the Krumovgrad pottery assemblage; thus the Early Neolithic layer of Krumovgrad could be referred to the second half of the Early Neolithic period in Bulgaria.
V
H
V
v *

I	/
* \
I I I I +-H
0	5 cm
Fig. 1. Krumovgrad. Pottery types from the Early Neolithic levels.
Map of southeastern Balkan shomng the lo-cation of Krutnovgrad and Hoca Cestne.
HOCA gESME
The Neolithic site is located on a terrace by the estu-ary of the Maritza river, some 5 km inland from the Aegean. The site was excavated from 1990 to 1993 by an Istanbul University team, headed by M. Ozdo-gan. The cultural sequence was divided into four main phases, Phase IV being the earliest (Ozdogan 1993-1997).
Phase IV.
The earliest settlement is small, and heavily fortified by a massive stone wall (Ozdogan 1997.24, Fig. 7). The houses are circular and sit directly on the rocky surface, actually carved into it. The pottery assem-blage is characterised by the total absence of coarse ware. The pottery is fine, thin-walled, with a lustrous red or black surface. Deep bowls, usually with "S" profiles, tubular or crescentic lugs, are common elements in this assemblage. The decoration consists mainly of vertical or curvilinear bands in relief; occa-sionally there are some fine grooved or incised pot-tery (Ozdogan 1997.24-25).
Phase III
It consists of two architectural layers; the buildings are again circular in plan. The massive, enclosing stone wall from the previous phase was stili in use. The pottery assemblage is similar to that of Phase IV, although the wares are slightly coarser and thick-er. A new type of ware, with a thick smeared red
coating on a black surface, is represented by a very few sherds. This distinctive ware increases in quan-tity in Phase II. As for the vessel types, the difference from Phase IV is minimal. Stamped and incised decoration is now slightly more common than in the previous phase. There are some painted pottery (in-cluding white painted ones) from the end of the phase (iOzdogan 1993 Fig. 4; 1997.25-26).
Phase II
The phase consists of three distinct horizons. There is a marked change in the plane and the construc-tion techniques of the buildings; they are rectangu-lar, with plastered walls, similar to the typical Karanovo I period houses. Domed ovens on raised plat-forms, bins, and working platforms represent the new elements of this phase. In spite of changes in the architecture, the massive enclosure wall was stili maintained and used, indicating continuity in the settlement organisation. The red and black wares of the previous phase continue, though in lessening amounts, and the quality of the burnishing is lost and the walls are thicker. There is an increasing amount of reddish-brown and matt-black pottery, the latter occasionally having a smeared red coating. Though minimal, there are some coarse, dully burnished pottery. There are a number of new vessel types in the pottery assemblage. Besides the decoration, typical of the previous phase, there are fine fluting and intentional mottling. Though very few, there are painted sherds: white on red or black, red on cream or black, and black on red (Ozdogan 1993,
Fig. 4). The so-called "pintaderas", bone spatulas and clay figurines are among the common elements of the phase (Ozdogan 1997.26).
Phase I
The last layers of Phase I and Phase II are badly damaged by later intrusions and erosion. Phase 1 consists of three distinct horizons (Ozdogan 1993■ 183-184). Most of the wares of the previous phase have disappeared. Most common for the phase are knobbed handles, footed vessels, plates and bowls with thickened rims, sometimes with channelled de-coration, and triangular vessels with incised and/or encrusted decoration (Ozdogan 1993, Fig. 1, 2, 3). There are some painted sherds, white on red and red on cream, the latter being from the earliest horizon of Phase I (Ozdogan 1993, Fig. 4).
Calibrated l4C dates were published for the site (Ozdogan 1997.27). As the uncalibrated date were also available, Yavor Boyadziev from the Archaeological Institute in Sofia studied the information and proposed his own view (personal communication)1.
M. Ozdogan	Y. Boyadziev
Phase IV 6400-6100 BC	Phase IV 6200-6000 BC
Phase III 6000-5900 BC	Phase III 6000-5800 BC
Phase II 5800-5700 BC	Phase II 5800-5600 BC
On the basis of the Hoca (Jesme evidence, M. Ozdogan suggested the following interpretation of the site: a population from the Aegean part of Anatolia, being in close relations with the Central Anatolian plateau, moved northward and, reaching the estuary of the Maritza river, settled down. Hoca (Jesme "... clearly demonstrates the gradual change and adaptation that an Anatolian type of colony settlement went through in a local environment. It is possible to fol-low not only the roots, but also the stages that led to the emergence of the Karanovo I culture from Phases IV and III of Hoca Cesme" (Ozdogan 1997.27).
The newly published Hoca Cesme evidence is of cru-cial importance for a better understanding of Neoli-thization processes in the Balkans. It provides a new basis for the reconsideration of several sherds from Krumovgrad. Four sherds which differ essen-tially from the rest of the pottery are available in the boxes containing the materials from the Krumovgrad site.
•	A sherd from a jar with a bead rim; brown slipped and burnished surface; the wall is 3-4 mm thick; even brown scatter. The decoration consists of incisions and dots. There are traces of white mat-ter in the dots (Fig. 2. 1). The sherd was found at 2.70 m (construction level IV). Sherds of vessels similar in shape or decoration were found at the end of Hoca £esme Phase III and in Phase II.
•	A sherd from a vessel with a vertical string-hole lug; there are traces of a red wash on the dark brown, very well smoothed surface; the wall is 5-6 mm thick. The decoration consists of two incised lines (Fig. 2. 2). The sherd was found in the scraped soil, therefore lacking a fixed stratigraph-ic position. The peculiar feature is the convex inte-rior part of the lug, an element unconunon for the Krumovgrad pottery assemblage, but existing at the end of Hoca (Jesme Phase III and the beginning of Phase II.
•	A sherd from a necked jar; black burnished surface; the wall is 5-7 mm thick. The decoration consists of a "necklace" of dots and vertical bands of thin incised zigzag lines (Fig. 2. 3). The incisions were made on a semi-dry surface. The sherd was found at 2.70 m (construction level IV). Similar motifs were registered at the end of Hoca Cesme Phase III and in the beginning of Phase II, and similar motifs and incision technique in Ya-rimburgaz phase 4.
•	A sherd from a plate with a slightly thickened rim; gray-black burnished surface. The wall is 6-7 mm thick. There are grooves on the rim and stamped decoration just beneath (Fig. 2. 4). The sherd was found at 1.90 m (construction level II). Sherds of vessels with similar decoration were registered at the end of Hoca (Jesme Phase III and in Phase II. Thickened rims exist in Hoca (Jesme Phase P.
1	I would like to thank Dr. Bojadjiev for the information he shared with me.
2	1 am deeply indebted to Prof. M. Ozdogan for the opportunity he provided to me to work with the pottery from Hoca Qesme, and for the help and the stimulating discussions as well.

The few sherds presented are atypical of the Early Neolithic Krumovgrad pottery assemblage and of the Early Neolithic Karanovo I type of pottery as well. They differ in paste, surface colour, treatment, and decoration, which is why they could be regarded as imports. The only probable exception is the lug sherd; it could be a piece of local production under strong foreign influence.
The fact that sherds similar to the Hoca (Jesme type of pottery were found in Krumovgrad provides grounds for a synchronisation of Krumovgrad con-struction levels IV-II with the end of Hoca Cesme Phase III and Phase II. The white, red or black paint-ed pottery in the layers of the end of Hoca (Jesme Phase III and Phase II support such an assumption. Some of the white on red sherds are identical to the white painted pottery of the Karanovo 1 type of pot-tery. In my opinion, they could be direct imports.
As was already mentioned, the lack of radiocarbon dates from Krumovgrad, makes the fixing of the absolute chronology of the site impossible. During the last decade, archaeological excavations of Early Neolithic deposits in Bulgaria have yielded a consid-erable number of radiocarbon samples. There are over 60 l4C dates calibrated by a computer pro-gramme in the Institute for Prehistory in Frankfurt am Mainz. According to the results obtained, the development of the Karanovo I period was most prob-ably between 5950 and 5600 BC {Nikolov 1989.30).
The studies of Early Neolithic pottery show that the Karanovo I period in Thrace consisted of at least two consecutive stages. The earlier stage is attested in the Mesta valley and the western parts of Thrace -Eleshnitza (construction levels I and II) and Slatina (the lowest levels of the sequence, the "Big House") -and the painted decoration is made with white paint only. The later stage is attested in a larger number of sites in the eastern parts of Thrace and in the lower parts of the Rhodope Mountains - in Karanovo, Simeonovgrad, Krumovgrad, Kardzali, etc, This stage is marked by the appearance of a small number of imported or local vessels with darkly painted or polychrome decoration, indicating that the stage was contemporary with the Starčevo type cultures in the Central Balkan area. At the time this stage developed in the Eastern Rhodope Mountains in sites such as Krumovgrad and Kardzali, the consequence of Karanovo II, II-III and III settlements existed in the eastern parts of Thrace (in Karanovo for example) (Nikolov 1997; 1998a). The distribution of settlements belonging to different stages in certain areas
supports the idea that the routes of the initial distribution of early farming groups followed the Mesta and Struma valleys in a northerly direction. Thrace was later "colonized" and the process took plače from west to east, reaching as far as the lower cours-es of the Maritza and Tundza rivers (Nikolov 1998a).
The presence of vessel types typical of the Karanovo II and II—III periods in the pottery assemblages of Krumovgrad supports the later chronological posi-tion of the site in the Thracian Early Neolithic se-quence. If we accept the dates for Hoca (Jesme pro-posed by Yavor Boyadziev, we see that Hoca Cesme Phase IV and the first half of Phase III should be con-temporary with the stage with white painted decoration in the Central Balkans and the first stage of the Karanovo I period in Eleshnitza and Slatina.
The painted pottery (white on dark brown, white on red, red on black, black on red) at the end of Hoca (Jesme Phase III and those (white on red, white on brown, black on red) in Phase II, refer the relevant phases to the second stage of the Karanovo I period.
Burnished ware decorated with bands of incised zig-zag lines (similar to Fig. 2. 3) was registered in Ya-rimburgaz phase 4, as well, and the pottery assemblage of the phase is correlated with the pottery assemblages at the end of Karanovo I and Karanovo II periods (Nikolov 1998.218).
To return to the "Hoca (Jesme čase", I would like to propose another point of view: an Anatolian popu-lation reached the Maritza estuary, settled down, and established the Hoca (Jesme site. The people protected the village from the potentially hostile or merely unknown environment by a massive stone wall. The enclosing wall suggests that the newcom-ers found the area populated already, othenvise they would not have put so much effort into its erec-tion and maintenance. The settlement developed as a closed community during Phase IV and the greater part of Phase III, and "domesticated" the newly acquired area by maintaining traditions: - living in the same type of houses as in the old homeland, ma-king the same pottery, etc. Contact with the Karanovo I people must have been established earlier, but evidence appears at the end of Phase III - several painted sherds in Hoca Cesme and a few sherds of "foreign" pottery in Krumovgrad. Obviously it was a time of intensifying contact between the two cultural communities. A new house type (much more suit-ed to the local climate and environment) appears in Hoca Cesme Phase II, a significant change, indicating
closer relations with the local people. At the very end of Phase II and in Phase I the similarities to the cultures of the Balkan Middle and Late Neolithic grew in number and the initial cultural identifica-tion of the Anatolian colony changed considerably changed.
Most probably, the roots of Karanovo I culture lie in that type of colony; the migrations of population and the transformation of cultural experience gave birth to a new phenomenon in the Balkans, but there is stili not enough evidence that it happened via the Maritza valley. It does not seeni likely that Hoca (Jesme was a kind of "generator". It was rather a
small colony, established on an area where the exis-tence of the Karanovo I culture was an established fact (though it probably did not cover the lowest Maritza valley entirely). After some tirne, the two cultural phenomena established relations. These are detectable in the two directions from imports or local production under foreign influence. Hoca (Jesme developed independently for some tirne, as is evi-dent from the continuity of house and the pottery types, where Anatolian traditions dominated local trends. According to the available data, the cultural influence of Hoca Cesme was restricted to the relativen small region of Eastern Thrace and the Eastern Rhodope Mountains.
REFERENCES
ICbHHEB K., MOXA4>KMEB M. 1994. HeoAHTHo ceAHiije b KpvMOHrpa/i. In MapHya-m3tok. Apxe0A0-rmecm npoy*tBaHM, II: 13~38.
NIKOLOV V. 1989. Zu einigen Aspecten der Kultur Karanovo I. In Tel Karanovo and das Balkan-Neo-lithikum: 27-41.
1997.	Die neolithische Keramik. In Hiller S. and Nikolov V. (eds.). Karanovo. Die Ausgrabungen im Siidsektor 1984-1992: 105-146.
1998.	ripovhbahha bt>pxy HeoAHTHaTa KepaMmca. KepaMHHHHTe KOMnAeKCH KapaH0B0 II—III, III h III-IV b kohtekcta ha cebepo3ana4ha ahatoahh h I0r0H3T0MHa EBpona.
1998a. The Circumpontic Cultural Zone during the Early Neolithic. Archaeologia Bulgarica 2 (in print).
OZDOGAN M. 1991- An interim report on excava-tions at Yarimburgaz and Toptepe in Eastern Thrace. Anatolica 17:59-121.
1993. Vinča And Anatolia: A new look at a very old problem (or redefining Vinča Culture from the perspective of Near Eastern tradition). Anatolica. Special issue on Anatolia and the Balkans 19: 173-193-
1997. The beginning of the Neolithic economies in Southern Europe: an Anatolian perspective. Journal of European Archaeologv 5/2: 1-33.