Location, environment and cultural characteristics of the region3

1The Yagodina Cave4 is situated in the Western Rhodope Mountains in the area of the watershed between the catchments of the tributaries of the Mesta and Maritsa Rivers (41° 37’ 41.80” N, 24° 19’ 43.53” E). It is a typical mountainous area. The high-standing parts are characterized by smooth relief, turning into plateau-shaped formations in the areas between the rivers; they are divided by river valleys cutting deep into the rocks and forming beautiful gorges (fig. 1). This is the location of the Trigrad Karst – one of the largest karstic areas in the Rhodope Mountains with 169 caves and precipices surveyed and mapped there5. The region faces Aegean Thrace to the south; the Vacha River valley provides direct access to the Upper Thracian Plain to the north and is of key importance for the study of the cultural interactions between the people inhabiting the various regions.

2Traces of occupation have been documented in 22 caves and the artifacts discovered date back to the Chalcolithic, the Late Bronze Age, the Early Iron Age, Late Antiquity and the medieval periods. The Trigrad-Yagodina region is known for the high concentration of cave sites yielding Chalcolithic artifacts – Haramiiska Dupka, Gorni Razh I and III, Gorna and Dolna Karanska Caves, etc. (fig. 2). The increase in the number of Chalcolithic sites is also documented in other parts of the Rhodope Mountains, and most of the sites date back to the latest phases of the Chalcolithic. The large number of Late and Final Chalcolithic sites (more than 60) compared to the scarce evidence of earlier occupation suggests that the Rhodope Mountains were “colonized” during this period, including areas situated higher than 1000 m above sea level.

The investigations

3The first investigations in the region were made by V. Mikov in the late 1920s. He published information about the Yagodina Cave including descriptions of pottery and scattered human remains, suggesting that they belonged to the Neolithic and the Bronze Age6. In 1952 P. Detev made a small trench in the anteroom and dated the excavated artifacts back to the Neolithic, the Bronze and the Early Iron Age. Planned archaeological excavations were carried out between 1965‑1966 and supervised by M. Deyanova-Vaklinova. A larger part of the Anteroom was excavated and three occupational levels associated with the Chalcolithic and the Bronze Age were documented7. Between 1983‑1984 M. Avramova carried out additional excavations at the cave entrance. Remains of a pottery kiln were unearthed, and new observations on the sequence were made. Two building levels were distinguished8 and assigned to the Final Chalcolithic and to the transition to the Bronze Age9. The main results from the excavations were presented in a number of articles10 but a considerable amount of the unearthed artifacts, the pottery in particular, is still unpublished.

General information – Sequence

4The Yagodina Cave is situated on the right bank of the Buynovska River, ca 3 km southwest of the modern village of Yagodina. It is a complex, multi-level formation consisting of galleries with a total length of 8501 m. The cave is formed into host marbles with layers of biotite schists and amphibolites. The uppermost storey of the cave was inhabited, due to the relatively low level of humidity and the constant temperature of 13‑14 °C. The entrance is located on a steep slope facing north, 30 m above the level of the river and 1140 m asl. It is arch-shaped, 1.5‑2 m high and 3 m wide at the base.

5A 7 m long tunnel leads to the first chamber named the Anteroom. A narrow gallery in the southeastern end of the room leads to a second small chamber, which gives access to three other chambers. The southwestern one is 47 m long and is named “Grancharnika”, i.e. the Potter’s Workshop (fig. 3: 1).

6The most intensive occupation was documented in the Anteroom. It has an irregular, east-west elongated shape measuring 25 x 35 m. An area of 200 m² was excavated using a 2 x 2 m grid, and the virgin soil was reached almost everywhere in the excavated area (fig. 3: 2). The thickness of the cultural layers varies from 0.20‑0.30 m in the eastern wing to 1.20 m on the southern and western periphery of the Anteroom (fig. 4). It is a result of the varying height of the cave ceiling, the natural slant of the cave “floor” to the west, and activities of the inhabitants meant to level the occupied surface. The stratigraphy is complex and inconsistent in the different excavated parts. The sequence consists of three main layers (from the bottom upwards):

7First occupational level. A preserved layer related to the earliest occupation was documented only in the western and the central part of the Anteroom. Its depth was from 0.70 to 1.20 m from the modern surface. It overlaid the virgin soil, which consisted of yellowish clay mixed with gravel, and was marked by patches of yellowish floor plaster (preserved in the undisturbed areas). The layer consisted of dark soil with traces of firing. Patches and thin layers of charcoal and ash as well as concentrations of charred grain were documented. The concentration of potsherds was high and some of them were secondarily fired. The remains of the first occupational level in the southwestern part of the Anteroom were sealed by rock pieces of various sizes, which had fallen from the cave ceiling, probably as a result of a strong earthquake. The collapse was accompanied by a fire. Similar traces of destruction were also revealed in two of the galleries. In the southwestern gallery – the “Potter’s Workshop” – artifacts related to the first period of occupation were buried under the collapsed rock pieces. Later on, the tunnel leading to the galleries was blocked by a stone wall, and there is no evidence that they were used during the later stages of occupation.

8Second occupational level. Its remains were documented at various depths almost all over the excavated area in the Anteroom. They were either overlying the leveled debris of the first occupational level or lying directly over the bedrock. The thickness of the layer varied between 0.30 and 0.50 m. The occupational level and the related renovations were marked by patches of preserved floor plaster and hearth floors. The layer consisted of dark brown soil, at some places mixed with a large amount of ash, thin layers of charcoal and traces of fire.

9Top soil. The layer was documented over the entire area of the Anteroom. It was 0.05‑0.50 m thick and consisted of reddish-brown soil mixed with gravel, small stones and plant roots. Relatively big pieces of rock collapsed from the cave ceiling were documented in the central part of the chamber. The layer yielded artifacts dating back to the Late Bronze Age, Antiquity and the medieval periods.

Structures and features

10Two ellipsoid hearths measuring ca 1 m in diameter11 were related to occupational level I. The hearths were built directly on top of the yellowish floor plaster. They had a clay border reinforced by vertical slab stones. Grinding stones were found near the hearths, one of them placed in a shallow pit. Remains of a charred mat and fragmented ceramic vessels containing charred grain were discovered in squares 32 and 38 at a depth of 0.90 m. The artifacts unearthed in the Potter’s Workshop and the left (northeastern) gallery were related to the same period of occupation. Round pieces of raw clay prepared for pottery manufacture, lumps of ochre in various colours and cone-shaped pieces of graphite, stone and bone tools, as well as several almost complete and fragmented ceramic vessels, were discovered under and between the collapsed rocks in this gallery. The skeletal remains of an adult male (adultus)12, who might had been the potter and died in the earthquake, were also found there.

11Most of the features unearthed are related to the second period of the occupation of the cave. Nine hearths were excavated; six of them are thought to be connected with the later renovation of the level. Practically, the construction of the hearths does not differ from those of the earlier occupational level. Almost all of them were ellipsoid with a low clay border and encircled by vertically laid flat slab stones. Some of the hearths had several consecutive floors and a relatively solid substructure made from rubble. A special typical feature that can be pointed out is a slanting step of trampled clay that was made in front of the hearth’s entrance (fig. 5: 1‑2). The fireplaces related to the second period of occupation were located on the periphery of the Anteroom, near the cave walls. Grinding stones, charred remains (or imprints) of mats, and complete and fragmented ceramic vessels were usually discovered next to them. Ceramic loom-weights and spindle-whorls were quite common. Lumps of unfired clay and lumps of ochre were discovered directly on the floor in front of one of the fireplaces. A pit which yielded charred grains is also related to the same period. A pottery kiln used only once, situated in the northeastern part of the Anteroom (fig. 5: 3), has to be mentioned among the excavated features associated with occupational level 2. No other features were unearthed in this sector, which is naturally elevated compared to the rest of the area. The floor of the kiln measured 1.90 x 1.75 m and it was made from 3‑4 cm thick uneven plaster. It was laid directly on the yellowish virgin soil. Postholes 0.12‑0.20 m in diameter delineated two ovals. A 6 cm thick layer of fine ash was found on the kiln’s floor; it was overlaid by the remains of the kiln’s dome, consisting of 4 cm thick fragments which were strongly fired and plastered on both sides. Fragments from several low fired ceramic vessels, four spindle-whorls and a fragment of a ceramic figurine were discovered under the remains of the dome.

12The above mentioned screen walls, which were most probably designed as barriers against the draughts in the Anteroom, are probably also related to the second period of occupation. One of the walls was 0.50‑0.60 m thick and was preserved up to a height of 1 m. It was made from crushed stones of medium size and no remains of binding material were detected.

Pottery assemblage13

13The surface of the greater part of the vessels is evenly coloured, and shades of brown are dominant – dark brown, reddish-brown, grayish-brown (more than 70% of the total number of sherds). The colour of the inner and outer surface is usually one and the same, or slightly different in nuance. Uneven colouring is observed more often on semi-coarse ware and is typical for the outer surface. About ¾ of the vessels and the sherds have an additional coating – a fine slip (or self-slip)14, or a rougher and unevenly coated layer of fine clay. The surface of almost half of the sherds is matt or, more rarely, shiny burnished. Many vessels have slightly unevenly smoothed surfaces, on which the irregular traces left by the finishing tool, crossing each other, are clearly visible. A roughly smoothed or rusticated (most often by scratching) pottery surface is one of the typical features of the pottery assemblage. The clay is tempered with a high amount of inorganic inclusions – quartz sand and other mineral inclusions with various sizes. In some cases the clay is tempered with chamotte or organic inclusions. The clay of a distinct type of ware is tempered with a high amount of mica and pyrite particles providing an additional glossy effect to the pots15. It is presumed that the vessels were fired in an oxidizing atmosphere and after reaching the desired temperature they were wrapped in, or covered by, organic materials (leaves, dung) in order to achieve the dark colour of the surface16. Reconstructions of firing facilities and technological processes are proposed based on ethnographic parallels and experiments17.

14Three main groups of ware are defined: fine ware, coarse ware and storage vessels. The proportion between the first two groups is approximately 2:3, and the percentage of the storage vessels is very low – ca 5%. No direct connection can be established between pottery shapes and technological characteristics. In general, the fine (table) ware is carefully made and its surface is nicely finished. The shapes are symmetrical, without any considerable deformations and the wall thickness varies between 0.5‑0.8 cm. At the same time, it is worth mentioning that the vessels are quite heavy; the pots (even the small ones) are massive and they often have thick bases. Thin-walled fine and very fine ware is exceptionally rare.

15The analysis of the Yagodina Cave pottery reveals a change in the proportion between the identified groups of wares typical for the two periods of occupation. There is a well-expressed tendency towards a “decrease” in pottery quality in the later period of occupation18. The prevailing number of vessels have uneven, relatively roughly smoothed or rusticated surfaces; the larger inclusions in the paste, which are clearly visible, create the impression that less care was involved in the process of pottery making. The same tendency has been documented in the pottery assemblages from the same period over vast geographical areas including all of the Balkans and the Aegean19. The data from Thessaly and Southern Greece shows that coarse ware had already increased in frequency in the previous period. During the Final Neolithic and the Chalcolithic, “cooking ware” dominates all pottery assemblages, in some of the sites even exceeding 90%. One of the proposed explanations is that the pragmatic role of pottery in the domestic sphere increases at the expense of its primary symbolic and social role20.

16The Yagodina Cave I‑II pottery assemblage is definitely dominated by shapes without sharp carination. The simple rounded, hemispherical or sphero-conical, shapes with gentle rounded carination are typical for the entire pottery assemblage21. It is worth pointing out the number and the variety of handles, especially in the categories of cups and jars. Almost all of them are vertical loop handles, most often of large size, oval or flat in section. The handles on plates and bowls are usually smaller, placed at the lower part of the vessel. Continuity in the shapes is documented between the two periods of occupation. Since no essential typological differences are recorded, the main shapes and types are presented in a general chart (fig. 6). The changes in the pottery assemblage affect to a greater degree the technology and the quality of the pots, and the technique and style of decoration, rather than the shapes22.

17Plates. The highest typological variety is documented among the plates. Most common are the simple conical or hemispherical plates with a thickened rim (fig. 7: 6, 7, 12). The hemispherical plates are usually not decorated in contrast to the conical ones, 25% of which are graphite-painted and/or crusted. The S-shaped plates with an everted rim are also very typical (fig. 7: 8, 9). The hemispherical plates with an everted rim are more common in the second occupational level. The exterior surface is most often rusticated or smoothed with a tool leaving quite rough strip-shaped traces. In a few cases, traces of a dull whitish graphite paint or crust are visible on the interior. The shallow S-shaped plates with flared rims are very typical for the second period of occupation. Most of them have a pair of vertical loop handles (fig. 7: 17, 19). The main technique of decoration is coating with mineral pigments.

18Bowls. A typical element of the pottery assemblage, especially from the earlier level, is the series of sphero-conical bowls with everted and thickened rims (fig. 8). Some of the bowls have small vertical loop handles on the lower part of the body. The size of the vessels is standard – the diameter of the rim is 20‑24 cm. The decoration is graphite-painted, dull and whitish in most of the cases. The compositions usually consist of groups of narrow and wide parallel oblique lines, but wide curved bands (parts of spirals and arches) are also present. The deep S-shaped bowls with or without a handle on the upper part of the vessel are very typical for both occupational levels. The prevailing number of these bowls belongs to the coarse ware and they are usually undecorated (fig. 9: 3, 5‑7, 10).

19Jars. The spherical or sphero-conical jars with a conical neck and a pair of handles on the shoulders or at the maximal diameter, are also very typical (fig. 10: 4, 7‑10). Their size varies considerably from small jars about 15 cm high to large ones about 60 cm high. The latter were probably used as storage vessels. Less than 10% of the jars are decorated, and most often these are small-sized vessels related to the earlier period of occupation. The decoration covers the upper part of the jars or the entire body, and consists of graphite-painted patterns or combinations of graphite painted and crusted patterns always arranged in vertical, oblique or “waved” bands (fig. 10: 7, 9).

20Cups. Five main shapes are defined (fig. 10: 1‑3, 5). Typical for the earlier occupational level are spherical cups with short cylindrical upper parts and a vertical handle on the shoulders. The size is standard, 8‑10 cm high. Most of them are decorated with graphite paint, usually dull whitish. The patterns consisting of wide vertical bands, or alternating fields and groups of thin or thick lines, cover the entire exterior surface (fig. 10: 1, 2). For the time being, these cups have no exact parallels among ceramic vessels from other sites.

21The rest of the shapes are represented by single vessels.

22The pottery decoration and especially the changes in the decoration techniques and stylistic characteristics, which are easily detectable for the two periods of occupation, deserve special attention. The percentage of the decorated pottery is not very high – ca 20% of the total number of the analyzed samples. The main techniques are graphite painting and crusted decoration. The graphite painting is dominant in the earlier period. Although significantly reduced, it is still present in the 2^nd occupational level. There are two main varieties related to the method of application of the graphite-painted ornamentation. More often it is dull and looks like fine whitish “pseudo graphite painting”23, or low-quality graphite paint24. The occurrence of the latter exclusively in the pottery assemblages from the final stages of the Chalcolithic provides grounds to consider it as the latest manifestation and/or imitation of the graphite painting typical for the classic stages of the Balkan Chalcolithic. It has to be pointed out that there is a stylistic uniformity of the decorative schemes recorded over a vast territory25. The graphite painting on the rest of the sherds is thick and has the typical metallic shine. The traces of a brush used for applying the graphite paint are clearly visible on a number of sherds, although drawing with a graphite cone directly on the surface of some of the ceramic vessels should not be excluded as an option. Two main “styles” are defined in the graphite-painted decoration, provisionally named “linear” and “floral”. The linear style is used more often. The typical motifs consist of alternating groups of thin and wider bands, vertical or crossing each other at various angles; zig-zags and net-patterns occur less often. The floral style consists of more lavish designs – wide graphite painted bands form spirals, concentric circles and festoons26.

23Ceramic vessels with crusted decoration are found in both periods of occupation, but there is a difference regarding the decorative techniques, colours and patterns. The typical crusted decoration on ceramic vessels of the first occupational level consists of a relatively thin layer of mineral pigment, most often red or brick-red in colour, covering ornamental fields, wider bands or elements of more complex compositions, usually arranged vertically; additional colour is added with yellow ochre and linear graphite-painted designs on the upper part of the pots (fig. 11). In the second decorative technique, typical for the later period of occupation, thicker and more compact mineral pigments are used, creating on the surface an impression of “relief” design (fig. 12: 3‑5). The designs are usually positive; the most common ones are simple spirals and curvilinear motifs in the interior of open shapes. The colours are in the softer, pastel nuances of pink-red, various shades of beige, ochre and brown27.

24With regard to shapes, technology and decoration, the Yagodina pottery has a number of distinctive features allowing the definition of a regional West Rhodope pottery style. The hypothesis of a substantive cultural development of the Rhodope Mountains during the final stages of the Chalcolithic is not a recent one, and the main argument is based precisely on the stylistic peculiarities of the pottery28. Many of the details related to the origin of this tradition still need elucidating, but at this point in the research it can be said that the Yagodina pottery assemblage is a mixture of ingenious local characteristics and elements typical of a vast geographical area. The latter can be viewed as a result of an increased intensity of the interactions between the inhabitants of different regions in the final stages of the Chalcolithic. Since this article does not aim for a detailed discussion about the parallels to the Yagodina pottery, our attention will focus on those that are especially important for defining chronological relations and cultural counteractions.

Pottery parallels

25The closest parallels so far to the Yagodina pottery29 come from the Maaras Cave-Angitis situated in the southern foothills of the Falakro Mountain. The site is interpreted by the excavators as a seasonal settlement of a group of transhumant shepherds. The excavations brought to light four hearths surrounded by stones built on two platforms30. The site yielded only a few graphite-painted sherds, but their shape and decorative patterns are completely identical to those found at Yagodina I31. The rest of the published ceramic vessels find amazingly close parallels as well32.

26The latest phases of the Chalcolithic in the Northern Aegean are still not known very well, but the few published sherds provide relatively close parallels to the Yagodina pottery. Among the shapes, the bowls with thickened rims find parallels in the Maaras Cave-Angitis, as well as in the Cyclop’s Cave near Maroneia33 and at Limenaria in Thasos34. Both sites yielded bowl sherds with matt graphite painting and linear patterns. The shape can be defined as typical for the Final Chalcolithic in general and is known from a number of sites in the Balkans – Dolno Dryanovo, Klise Bair, Sozopol, etc.35. The crusted decoration (pink and buff) and the whitish painted sherds yielded by Limenaria and Kastri on Thasos36, dated by the excavators to “the very end of the Neolithic”, can be considered a reliable indicator37.

27The pottery assemblages from Yagodina and from the Eastern Rhodope Mountains share a number of common elements, but the differences demonstrated by certain shapes and decorative patterns might be a result of differences in the chronology, as well as regional specificities38. Such a statement finds proof in the comparison with the materials from the sites of Dolno Dryanovo and Ilinden-Klisura near the town of Gotse Delchev39.

28Further to the south, parallels can be found in the Chalcolithic pottery from Thessaly. Several common features have already been mentioned40. The relatively close parallels in pottery shapes are accompanied by similar technological characteristics, especially those related to surface treatment, the colour range of the pottery and some of the details41. Particular parallels can be specified for plates and S-shaped bowls, bowls with everted thickened rims, jars with two vertical handles, some cups, vessels on a high conical pedestal base, etc.42. The typical Yagodina Cave plates with thickened rims have been found at Sesklo, Rachmani, Magoula Aidiniotiki43, Voulokaliva44 and other sites. The published sherds from Sesklo are thick white and pink crusted, and the technique is completely identical to the one used at Yagodina. The similarities regarding the crusted decoration have been presented in publications as an argument in favour of the connection between Rachmani and the Chalcolithic cultures in the Central Balkans, and particular fragments from Pefkakia and Sesklo are even considered direct imports from the north45.

29This line of comparisons is nicely complemented by a sherd from the shoulder part of a vessel with a handle, which is not typical for the Yagodina pottery46, but finds close parallels in assemblages from Thessaly, South Greece, and the Aegean islands (fig. 12: 1). The specific shape and the details of the handle from Yagodina resemble the Rüsselhenkel or “elephant lugs”47, which are characteristic for the region and find their closest parallels in the earlier building levels at Pefkakia48 and Sesklo49. In contrast to the plain vessels from Thessaly, the pots from Aria near Nafplion50, the Kouveleiki I Cave51 and the Zas Cave on Naxos52 are painted with red ochre at the rim and the handles, combined with pattern burnishing typical for Southern Greece. The parallels for the sherd decorated with a plastic band forming an open spiral also go in the same direction (fig. 12: 2). A relatively close parallel is known from Rebarkovo, in Northwest Bulgaria53, but in general, the decoration is typical for the Thessalian pottery and is represented by numerous examples from Rachmani54, Pefkakia, Tsangli Magoula, Tsani Magoula55 and Pyrasos Magoula56. Similar relief patterns are found in a number of sites in Southern Greece, for instance the Skoteini Cave57 and the Alepotrypa Cave58.

30At the same time, the Yagodina I‑II pottery shows an indisputable relationship with Final Chalcolithic assemblages from the northern parts of the Balkans, the closest parallels coming from the Veliko Tarnovo and Lovech regions – Hotnitsa, Klise Bair, Bezhanovo, Devetaki Cave, etc. Similarities are recorded in the main vessel types, but also and especially in the decorative style – the S-shaped plates, the deep vessels with everted upper parts of the body, the cups with one handle, the bowls with an everted thickened rim59. Vessels from the last category are often decorated with the whitish painted linear patterns typical for the Yagodina Cave60, yet what differentiates them from the Rhodope examples are the wide horizontal grooves carefully executed on the upper part of some of the vessels61. It is worth pointing out that the best analogies of the graphite-painted decorative scheme from the Western Rhodope Mountains come from precisely this region – the Toplya and Devetaki Caves62 as well as Galatin63. Conical potstands (lids?) with a pair of opposite oval openings similar to the one from Yagodina (fig. 10: 12) were found at Klise Bair and Bezhanovo64 and have not been recorded at any other sites.

31On the other hand, there are scores of differences when compared to the pottery assemblages from the Galatin-Sălcuţa IV and Cernavoda I cultures. For example, the diversity of technological methods for pottery making evidenced by the two cultural zones65 is absent at Yagodina. The clay for the pots made at Yagodina was not tempered with shells or a higher amount of organic substances, and the sherds with porous fabric are exceptions. However, similar characteristics are revealed by the pottery from Klise Bair and Bezhanovo. Some of the details of the pottery assemblage provide undeniable proof of the difference in the pottery making tradition: thus the Scheibenhenkel type of handles, especially typical for the western assemblages66, or the Cernavoda I corded decoration, have not been discovered at Yagodina.

32The brief review on the parallels of the Yagodina I‑II pottery marks the north-south direction of the cultural relations. This axis of contact, which is the main one for the Balkans, is detectable during all periods of late prehistory and has been discussed many times by the specialists67.

Small finds

33The chipped and polished stone tools are few, and they are represented mainly by spherical hammer stones made from quartzite and granite, as well as by polishers made from river pebbles. The bone tools include awls/perforators and spatulas or polishers, the latter probably related to pottery making. Small clay finds related to weaving and spinning are worth mentioning. The loom-weights for a warp-weighted loom are two types – disc-shaped and cylinder-shaped with a vertical opening (fig. 13: 10‑12). The spindle whorls are numerous, shaped as low truncated cones with a concave base (fig. 13: 1‑9) and each of them weights about 15 g. It has to be pointed out that the parallels of the finds related to textile production come from sites dating back to the final stages of the Chalcolithic, which allows them to be considered a chronological indicator68.

Relative and absolute chronology

34The chronological position of Yagodina I‑II in the final stages of the Balkan Chalcolithic is based on firm arguments and is beyond doubt69. The clear continuity of the pottery assemblages between the two levels and the lack of a well-defined stratigraphic break between them provide proof for the chronological closeness between the two periods of the cave occupation during the Chalcolithic. The parallels and the comments presented above offer additional details for specifying the relative chronology of the site within the frame of the Final Chalcolithic.

35The two occupational levels can be related to two successive chronological horizons, whose definition is based on the typical features of the pottery and the published ¹⁴C dates from a number of sites. From Yagodina itself, six radiocarbon dates had already been published70. The high values of two of them (Bln-2385 and Bln-2245) are probably due to some objective factor71. The rest of the dates cluster within the time span 5250/5200‑5000 BP and are among the few dates from Bulgaria whose calibrated values are within the first quarter of the 4^th millennium BC.

36The samples analyzed in the Lyon laboratory within the “Balkans 4000” project provided four new ¹⁴C dates (table 1)72:

37As it is seen in Table 1, the new dates fall within the same chronological interval and form a compact cluster around 5150‑5050 BP. One observes that the dates from the first and the second occupational levels show almost identical values. The close chronological position of the two occupational levels can provide a possible explanation of this result, which corresponds well to the smooth transition between the two pottery assemblages. On the other hand, the dates whose values are grouped around 5100 BP can be calibrated either around 3950 or around 3850 cal BC because of the zig-zag jumps in the calibration curve within the time span 4000‑3800 cal BC73, and therefore it can be assumed that they reflect a longer development. The same applies to the dates from Haramiiska Dupka Cave, which have slightly higher values compared to Yagodina74.

38Notwithstanding the difficulties already pointed out, the available information allows a more detailed reconstruction of the chronological positions of the events in the final stages of the Chalcolithic75. In this preliminary scheme, the earliest occupation at Yagodina Cave immediately follows (or is partially contemporary?) the horizon marked by Sozopol-Dolno Dryanovo-Šuplevac. The synchronization with the Haramiiska Dupka Cave is generally not questioned, but the more detailed correlation between the three occupational levels discerned there to the sequence at the Yagodina Cave is still a matter of discussion. At this stage of the research, the earlier occupational level at Yagodina can be related to the materials yielded by the Maaras Cave-Angitis and the Cyclop’s Cave at Maroneia, but the insufficient amount of material and the lack of radiocarbon dates do not allow explicit conclusions to be drawn.

39More reliable reference points are available for the synchronization between Yagodina II and Bezhanovo-Banunya/Klise Bair in Central Northern Bulgaria. The radiocarbon dates from Bezhanovo76 are practically identical to the ones from Yagodina II, and this fact is supported by numerous parallels in the pottery.

40Hotnitsa-Vodopada is related to the next stage (Sălcuţa IV-Galatin), and the dates yielded by the site are clustered around 5000‑4850 BP77. The similar values of the dates from Yagodina II (Bln-2247, Bln-2249, Lyon-6457) and the earliest dates from Hotnitsa-Vodopada (Bln-3682, Bln-3683, Bln-3684), as well as the number of close parallels in the pottery assemblages, provide grounds to suggest a chronological proximity, or even partial synchronicity of the two sites.

41The archaeological sites from the Aegean provide few dates within the 5200‑5000 BP time span. There is a single date from Thassos-Limenaria, which after calibrations provides a value of 3977‑3789 BC78. Dates with similar values are known from Franchthi Cave and Halieis in the Southern Argolid79, as well as from the Kouveleiki B Cave in Laconia80. The published ¹⁴C dates from Pefkakia Magoula81 are considerably earlier (ca 5600/5500 BP) and contradict, to a certain extent, the parallels to the materials from the Yagodina Cave and the contemporary phenomena from the northern parts of the Balkans discussed previously, whose position around and after the end of the 5^th millennium BC or 5300‑5000 BP could hardly be contested. The same applies to the relationship between the Thessalian materials and the ones from Southern Greece – the Attica-Kephala culture, which is well defined on the grounds of typological and stylistic parallels in the pottery82. The discrepancy raises questions about the chronological frame of the Rachmani culture, especially its upper chronological limit, which will be given explicit answers only by a new series of absolute dates and detailed studies of the archaeological materials.

42Within the context of the problems discussed here, attention has to be paid to the increase in the number of caves inhabited during the latest phases of the Chalcolithic – a tendency common both for the Balkans and the Aegean83. The time span between the end of the 5^th millennium BC and the first quarter of the 4^th millennium BC was marked by essential transformations of the settlement patterns, which are well evidenced in both regions. The information about the human occupation of Upper Thrace and Northern Greece in this period is so far fairly insufficient84. At first glance this “depopulation” appears to contrast with the gradual increase in the number of the settlements in the 4^th millennium BC recorded in Southern Greece and the Cyclades. However, there are some common traits, such as the more intensive exploitation of mountainous areas, the prevalence of small sites with a thin layer of deposits, the occupation of naturally protected sites, caves and rock shelters85. These facts indicate a more mobile population, which is logically related to the increased economical significance of stockbreeding in general, and seasonal pastoralism in particular. It is apparent that these changes have an over-regional and complex character, and the differences are due to a great extent to the diversity of the geographical and climatic conditions in the particular areas86.

43At the same time, the analysis of the archaeological data from this period reveals that the entire vast area of the Balkans and the Aegean seemed to be an arena of very active interactions between the separate communities. These interactions are evidenced by common elements in the pottery assemblage, which can also be regarded as a common pottery style87. Separate local phenomena – “cultural groups” – functioned within the boundaries of this vast “pottery koine”, each of them developing its own typical features, specific pottery shapes and decorative techniques. At present we are not able to specify the exact nature and to reconstruct the mechanisms of the contact between the different communities. However, one of the most widespread hypotheses is related to the mobile and semi-mobile stockbreeding practiced by some segments of the population88. The regular and systematic movements of population groups create favourable conditions for communication and exchange of material goods as well as information. Such a model provides a good explanation for the situation recorded by the available archaeological data. However, this is another complex issue directly connected to the problem of the formation of the Early Bronze Age cultures in Southeastern Europe. This is a topic which at the present stage of research raises questions rather than provides answers.