The excavations2

1The archaeological excavations in 2007-20093 at the settlement mound were provoked by the fact that the site was violated many times by treasure hunters. The excavations were carried out by a team from the Regional Historical Museum in Ruse and were supported financially by the same institution4. The settlement mound was registered during an archaeological field survey conducted in the Kosharna region in 20065, and entered into the Archaeological Map of Bulgaria database in 2007. It is situated 3 km to the west of the village of Kosharna, in the “Kaynak Dere” locality; it lies on a slope with southeast exposure, next to a small spring. The area is not cultivated at present and is overgrown by acacia trees. The tell, which is roughly conical in shape, is 6 m high; the diameter at the base is 66 m and the circumference is 188 m. A second settlement mound, which is considerably smaller than the one described, is situated ca 25 m to the southeast6.

2In 2007, it was established that 7 trenches were made by treasure hunters in various parts of the site. Our decision to set up our trench at the highest central part of the tell, between some of the treasure-hunt trenches, was based on two reasons: a) to excavate the only spot at the central part of the site which remained undisturbed by treasure hunters’ activity; and b) to connect the two deepest treasure-hunt trenches and explore a larger area, in order to obtain the most detailed information about the thickness and the sequence of the archaeological deposits. A grid of 5 m-squares oriented according to the four cardinal points was laid out, and a datum point (109 m above sea level) was set at the highest point of the site (fig. 1). Four auxiliary datum points were also installed to provide correct measurements on the site. The squares of the grid were marked by letters of the Bulgarian alphabet following alphabetical order along the X axis, and by serial Arabic numerals along the Y axis.

3Trench A was excavated in 2007. It measured 5 x 2 m, with the short sides oriented east-west, and it was situated in-between three of the treasure-hunt trenches (THT I‑III) within the frames of squares O8 and O9. Excavations in Trench A were continued in 2008 and 2009. The trench was extended 8 m to the east, as well as towards THT I to the west, and its complete length reached 16 m. In 2009 three new trenches were opened (Б, B and Г) attached to Trench A, and the entire excavated area reached 94 m².

The remains of later periods

4The mound was situated within the boundaries of a large Roman site7. The top layer dating back to the Roman period was destroyed by agricultural activity. Four pits cut into the Chalcolithic layer were excavated. The most noteworthy are pits 3 and 4.

5Pit 3 cuts into the upper layers of the mound to a depth of 1.80 m. It was documented immediately after removal of the plough soil and extended beyond the western boundary of THT I: it was bell-shaped, 1.40 m at the upper part and 0.95 m in the lower part. The fill of the pit consisted of brown soil mixed with sterile soil, fired daub fragments of various sizes and small stones; a larger stone was found near the bottom. The pit yielded ceramic sherds dating back to three periods – Chalcolithic, Early Iron Age and the Roman period – as well as a fragment of a Roman tegula and a silver denarius of Emperor Heliogabalus (218‑222 AD)8.

6Pit 4 was smaller than pit 3. It was cut into the Chalcolithic layer to a depth of 1.30 m disturbing a layer of grey ash: like pit 3, it was bell-shaped, 0.55 m in the upper part and 0.20 m in the lower part. Its fill consisted of black soil, a few pieces of fired daub and a small number of Chalcolithic and Roman potsherds.

7A few fragments of fired daub were unearthed in a layer 0.20‑0.30 m below the surface (0.77 m from the datum point). The layer was a mixture of top soil and soil from the treasure-hunt trenches. Some of the fired daub pieces were plastered many times, but not all plaster layers were identical: some of them bore traces of organic temper (straw) and others did not. This indicates that the fragments belonged at least to two different features (probably a wall plaster, and plaster of a feature from a building interior).

8The layer yielded numerous potsherds dating back to the Late Chalcolithic, as well as several finds from the same period: an intact clay loom-weight, a fragmented barbotine decorated ceramic potstand, a fragmented undecorated square “cult table”, and four flint blades. A small number of sherds however differ completely from the rest of the ware yielded by this layer. They are red or beige in colour, burnished or polished, and decorated with projections, shallow incisions and grooves. These sherds belong to a different ware and date back to the Early Iron Age. Since an Early Iron Age layer was not documented in any of the trenches opened by us, the pottery presented above probably came from a pit or layer destroyed by the neighboring THT I.

The Chalcolithic Layers

9Four building levels9 dating back to the Late Chalcolithic (the Gumelniţa culture) were excavated10. The thickness of the excavated Late Chalcolithic layers in trenches A and Б was 3.70 m, 1.30 m in trench B and 1 m in trench Г. Virgin soil was not reached in any of the trenches.

Building Level I

10This was a burnt layer documented in all trenches at a depth from 0.77 m to 1.40‑1.90 m (measured from the central datum point); the difference results from the slope toward the east. It has been uncovered over an area of 81 m². The level consists of two separate sub-layers: a) brown layer of debris and charcoal, 0.40‑0.80 m thick; b) a layer consisting of yellow-green clay ca 0.45 m thick.

11The following features were unearthed (fig. 2):

12Building with clay-plastered structure (structure 1). It was documented in trenches A and B. Part of a platform was unearthed; it was 0.25 m thick and the debris covered an area measuring 3.70 x 2.10 m. The platform was replastered several times and was covered by a layer of debris, in which at least two kinds of pieces of fired daub differing in colour and structure can be distinguished. The western end was destroyed by a THT. Feature 1 was constructed on a thick layer of yellow-green clay. It is possible that the platform was part of a feature with an unidentified function, covered by debris from a house destroyed by fire. A concentration of fired pieces of daub and stones, forming a line about 0.60 m wide at the border between squares O8 and O9, in Trench A, marked a section of the eastern wall of the house. The house debris yielded a complete bone figurine, a clay phallus and a small graphite-painted biconical ceramic vessel.

13Oven. A destroyed oven was unearthed in Trench A at a depth of 1.70 m. It was constructed on a thin base made from sand and gravel. The floor was 0.17 m thick and was replastered twice. The first and the second plaster layers were separated by a 0.08 m thick layer of tiny pieces of fired plaster; both oven floors were brown in colour. A concentration of animal bones was unearthed at the southwestern part of the oven. The sediments in the square at this level varied in colour – gray, brown, green – indicating that the oven was situated in an open space. Moreover, it suggests that the oven was not made on a house floor. Further proof of this is the concentration of ash (pit?) documented in the northern section, which probably accumulated as a result of the oven cleaning. The ash concentration and the oven were at approximately the same level and the ash concentration was a continuation of the yellow-green clayey layer to the east.

14Pit 111. It was located in trenches A and Г, square O9, next to structure 1. The pit was slightly bell-shaped, with an oval opening; the upper diameter was 1.80 m and the lower one 1 m. The pit’s top lay at a depth of 1.10‑1.20 m. It cut through building levels I, II and III and the bottom lay at 4.50 m. The walls of the pit were not plastered. The fill consisted of fired house debris, potsherds, animal bones, stones and charred wood. Four depositional events were defined from the bottom to the top (fig. 3):

1. Pieces of fired daub varying in size and shape – from 4.50 to 4.20 m. A charcoal sample for radiocarbon dating taken from the bottom of the pit12 provided a date of 5655 ± 50 BP, i.e. 4597‑4363 cal BC (Lyon-7657/SacA-22618).

2. Brown soil mixed with fragments of fired daub, charcoal and stones – from 4.15 to 3 m. A bone sample taken from a depth between 3.90‑3.75 m provided a date of 5260 ± 50 BP = 4235‑3968 cal BC (Lyon-7659/SacA-22620).
   A complete lamb skeleton was unearthed at a depth of 3.75 m in the northern half of the pit (fig. 4: 3). It was laid on its right side, and the head was twisted to the left and pointed to the northeast; the body was oriented south-north. The bones of the pelvis and the hind limbs were not found in anatomical order, most probably due to the disturbance of rodent burrows. Surprisingly, the sample taken from the skeleton provided a date of 90 ± 45 BP, giving after calibration 1670‑1950 AD (Lyon-7658/SacA-22619). Another skeleton, of a young dog, was found a little higher, at 3.46 m (fig. 4: 2); it measured 0.53 m at the withers. The animal was laid on its right side with the head pointing to the northeast. The bones had the typical brownish colour which is usually results from the decaying of the soft tissues. There was a displacement of the pelvis and the limbs from the lumbar part backwards which resulted in the disturbance of the anatomical order of the skeleton. In addition, the ventral part of the pelvis was turned upside down. The fact that the bones were in anatomical order reveals that the soft tissues had not decayed. All bones belong to a single animal. Most probably the animal was dismembered before it was buried in the pit. This skeleton was not dated.

3. Gray-green clayey soil 0.30‑0.64 m thick – from 3 to 2.50 m. A skeleton of a young roe deer (Capreolus capreolus Linnaeus) was found at 2.70 m (fig. 4: 1). It was laid on its right side and the skull was missing. The skeleton was oriented northwest-southeast. The sample taken from the skeleton for radiocarbon dating provided the date 230 ± 45 BP = 1526‑1950 cal AD (Lyon-7660/SacA-22621).

4. Gray-black soil mixed with small lumps of fired daub and tiny pieces of charcoal – from 2.50 to 1.20 m; three layers of larger fragments of fired daub were visible in it. The layer yielded two fragmented, secondarily fired clay loom-weights, a bone artifact with an opening at the end, a harpoon made from antler, and a bone chisel typical of the Late Chalcolithic.

15The stratigraphic and radiocarbon evi­dence provide reason to believe that the pit was cut into the layer during the final period of the site’s functioning (the second phase of the Gumelniţa culture, possibly represented by the date Lyon-7657) and continued to be used until the end of the Chalcolithic (the third phase of the Gumelniţa culture, represented by the date Lyon-7659). If we consider the radiocarbon dates reliable, it can be suggested that a construction level dating back to the third phase of the Gumelniţa culture had existed once but was probably destroyed by erosion, which explains why it has not been documented archaeologically. Another pit was cut into the layer at the same spot in the 20^th century. Therefore, layer 1 and part of layer 2 mark the Chalcolithic pit, and layers 3 and 4 mark the modern pit.

16The Chalcolithic pit belongs to an assem­blage of features located in the central part of the site. The pit’s dimensions and location do not provide any grounds to regard it as a waste pit, so it was probably connected to cult activities. The fact that at least two layers of fill have been documented suggests that the pit was used for a long period. But the pottery found in it (fig. 5) does not differ in technological or formal characteristics from the pottery yielded by the rest of building level I.

17Pit 2. A smaller pit was located in trenches A and Г, next to pit 1 and structure 1. It was oval in plan, larger at the upper part. Its opening lay at 1.10 m, at the uppermost part of the yellow-green clay, and its bottom at 1.68 m. The maximal width of the pit was 1.60 m and the depth was 0.60 m. The upper part of the fill consisted of black soil with few inclusions of fired daub and potsherds. In the lower part, close to the pit bottom, the fill consisted of a compact layer of fired daub identical in colour and structure to the ones in Feature 1. The pit yielded Late Chalcolithic potsherds and a fragmented bone figurine.

18Pit 6. The pit was located in trench A, square O8 next to the concentration of ash of structure 1. Its opening was documented at 1.60 m and the bottom at 2.86 m; it is oval in plan. Half of the pit was within the excavated area. It was filled with brown soil and tiny lumps of fired daub. The pit yielded a high number of Chalcolithic potsherds and animal bones as well as an adze made from black rock, three flint blades and a limestone ball.

19Ditch. A ditch situated 2 m to the north of structure 1 was located in trench B. It was filled with orange-red coloured fired fragments of daub, stones and Late Chalcolithic potsherds. The ditch was 1.20 m wide, 0.30 m deep and oriented northeast-southwest. The concentration of potsherds was very high, and the sherds were secondarily fired. The ditch was probably empty when the neighbouring structures were destroyed by a fire, and it was filled by debris of the collapsed house.

20A completely preserved antler of a red deer and grinding stones were unearthed to the south of pits 1 and 2 (see fig. 2).

Building Level II

21It was excavated in trenches A and Б over an area of 51 m². The upper part of the level was registered at 1.40‑1.50 m and the lower part at 2.60‑2.70 m, sloping to the east. The level was more than 1 m thick and consisted of 2 layers: a) a layer of light grey ash and stone (maximal thickness of 0.22 m) indicating that the settlements suffered from fire; b) brown clayey soil with inclusions of small lumps of fired daub, animal bones, potsherds and stones. The lower limit of the layer was marked by a concentration of small lumps of fired daub (possibly the lining of a structure?), 2 m long and 0.06 m thick, which continued to the north outside the excavated area. It was documented from 1.60 m (W) to 3.60 m (E) in square O9. A layer sloping to the east (average thickness 0.04 m) consisting of charcoal and ash was unearthed at the borderline between squares O9 and O8.

22Stone concentration (fig. 6). It was unearthed in the easternmost part of trenches A and Б; it started at building level II and continued down to level IV. Single irregularly shaped stones were scattered without any pattern and were documented from 2 m down to 3.95 m. On the borderline between squares O7 and O8, the stones had an average size of 0.15‑0.25 m and occupied an area ca 2.60 m wide; the area was situated transversely to the trench and probably extended outside the trench both to the north and south. In total, 25 stones were unearthed within this area at a depth between 2.05 and 3.30 m, and another 23 stones between 3.70 and 3.95 m. Some of the stones discovered at 2.60‑3 m were on top of each other. A skeleton of a bull was found within the stone concentration at 3.30 m; some of the bones were covered by stones. A copper earring, a small biconical bowl, a bone awl, a copper awl with a bone shaft and a spindle-whorl were found in the same area. Since this is the eastern periphery of the tell, it is possible that it was a sector where a destroyed defensive structure of the settlement existed and/or a refuse area.

23Pit 7. A small pit was located in trench A, square O7. It was kidney-shaped in plan, 0.60 m in diameter and 0.15‑0.20 m deep. The opening of the pit lay at a depth of 3 m. It was filled with a grey-brown clayey soil and potsherds.

24Since no remains of houses were discovered (except for a small spot of tiny lumps of fired daub), it can be suggested that this area was an empty space between the houses.

Building Level III

25It was documented in trenches A and Б over an area of 51 m², from 2.70 down to 3.90 m. It was marked by a brown-greenish clayey soil, small lumps of fired daub, animal bones, potsherds and stones. Parts of two houses were unearthed (fig. 7).

26House 1. It was located in the western part of trench A in square O9; part of the burnt house was located in THT I and was destroyed by it. The excavated debris layer was 1.20 m thick and 4.70 m wide. It consisted of three nicely marked layers which were described from the bottom upwards as follows:

1. Layer of fired daub orange-red in color and 0.40‑0.60 m thick, consisting, in its turn, of two sub-layers: 1.1. a layer of plaster with nicely smoothed surfaces (a floor) slanting to the northwest; 1.2. scattered pieces of daub of various sizes and shapes on top.

2. Small orange-red coloured pieces of daub on top of a level of grey-greenish clay covered by a thin layer of ash and black ashy soil mixed with lumps of fired daub, which marked the end of the construction period.

3. Fragments of plaster 0.15‑0.20 m thick, which were fired and had a two-coloured cross section – black and orange. This layer rested on top of a layer of compact yellow-green clay, probably a floor level, sloping to the east. A fragmented vessel and below it a fragment of a stone adze were found on the floor. An oval pit (pit 5) 0.80 m in diameter and 0.60 m deep was located next to the vessel. The pit was filled with fired daub and potsherds. The house also yielded a fragmented stone axe-hammer, a flint blade and three bone awls.

27A stone concentration consisting of irregularly dressed stones arranged in a row was unearthed at the borderline between squares O8 and O9 at 2.85‑3 m. Lumps of yellow-green clay (brighter in colour), probably used as binding material, were found between the stones. Concentrations of black ashy sediment (burnt daub) were documented in the feature: they probably marked a section of the eastern wall of the house. A small portion of the house floor was discovered – yellow-brown plaster, which can be probably related to the third stage of the house construction. The excavated portion of the house measured 10 m².

28House 2. It was located in the eastern part of trench A in squares O9 and O8, next to house 1, at a depth of 3.60‑3.90 m from the datum point. The house was destroyed by fire and the debris consisted of two layers:

1. Charcoal, charred wood and potsherds overlaid by low-fired orange-red pieces of daub. These debris could be remains from an attic with a plastered floor.

2. Yellow clayey soil on which the remains of charred wooden construction and fragments of orange-red daub (more heavily fired than the ones in layer 1) were discovered. The charred wood on the yellow-greenish floor were probably the remains of burnt beams from the roof construction.

29The short sides of the house were oriented north-south. The house was 5 m wide and the length remains unknown because the building continued beyond the limits of the excavated area. The eastern wall was marked by an irregular row of 10 postholes. It was plastered with a thin layer of green clay on the outside. The western wall was marked by a row of 5 postholes. The house consisted of two rooms at least. They were divided by a screen wall, marked by row of postholes in a narrow ditch filled in with green clay (fig. 8). The diameter of the postholes was 0.10 m and they were 0.15‑0.20 m deep. The floor was plastered at least three times, since three 0.03‑0.05 m thick layers of grey-green clay were documented, divided by two thin layers of red-brown fired clay. Two complete ceramic vessels and three grindstones were discovered in the southern room. A small clay wall was unearthed nearby, which had probably marked an area in the room used for performing household activities (fig. 9). The second room of the house yielded a fragmented ceramic “altar” and three more ceramic vessels, two of which had been broken during the collapse of the house. One of them contained 23 flint blades, two fragments of personal adornments made from a Spondylus shell, and a river shell (fig. 10).

30A charcoal sample yielded by the house produced a radiocarbon date which is relevant to the relative chronology of the feature and the construction level: 5660 ± 40 BP = 4548‑4402 cal BC (Ly-15276).

31The house debris was covered with clayey green soil for leveling the surface and preparing it for the construction of a new structure. A trampled level was unearthed on top. An architectural detail consisting of two walls crossing at right angles was found on the trampled level (fig. 11). The walls were plastered four times with yellow-green clay and the total thickness of the plaster was 0.15 m; the innermost part of the construction was coloured dark brown (a decayed wattle construction?). The preserved length of the wall oriented north-south was 0.92 m and included 4 postholes, whereas the preserved length of the east-west oriented wall was 0.97 m, with 4 postholes. The feature can be interpreted as a sheltered platform.

32Infant burial (fig. 12; 13)13. A skeleton of a baby was unearthed under the floor of house 2, at a depth of 4.22 m. It was not possible to define a grave pit, but the floor of the house was lowered at this spot and there were traces of subsequent levelling. The skeleton was lying 0.32 m below the floor level. A large ceramic vessel was discovered next the skeleton: it has a convex shoulder, a conical lower part, a tall neck and barbotine decoration; it was fragmented by the soil pressure. It seems that the dead body was originally lying on its back in a flexed position, or was lying on the right side and was tilted on its back post-mortem14. The head pointed to the west (240º). The skeleton was unearthed lying on its front with a flexed left leg, the left tibia and femur being in anatomical order. The left thigh bone and the entire right leg were missing. The left leg was twisted 90° to the right from the back to the front, and its front part was lying on top. The upper part of the arm was under the chest, and the lower part of the arm was under the head. The bones of the right arm and both hands were missing, as well as the feet and a part of the pelvis. The skull rested on its right side facing the southeast. The length of the skeleton was 0.44 m. The measurements taken in situ on both humeri, one femur and one tibia suggest that the age of the baby was less than 6 months. A scapula of an animal was unearthed next to the skull and to the left from the chest. A river shell was discovered on the back of the skull, as well as animal bones behind the backbone. Small pieces of charcoal were scattered around the dead body. A concentration of grey ash was documented at a level a little higher than the body.

33The head of the skeleton pointed to the west – the direction in which the extramural cemetery of the tell was situated. But the body position, crouched to the right, and its orientation are not typical for the burials excavated there15. Intramural Late Chalcolithic children’s burials were yielded by other sites as well: Hârşova 1, Borduşani-Popina, Bucşani-La Pod, Căscioarele, Chitila-Ferma, Yunatsite, Ruse. The extramural cemetery at Tell Kosharna has not yielded any skeleton of an infant of similar age up to present. The only infant burial excavated so far yielded the skeleton of a 2‑2.5 year old child. A probable explanation could be that those who died at a very early age had no social identity and had no place in the hierarchy of the prehistoric society, i.e. they were not in fact considered members of this society. Probably for that reason there was a tradition not to bury the dead baby in the “city of dead”, but in the village of the living, in the house where it had lived, in the hope that its soul would protect the house and the family and would return in the body of the next newborn baby in the family.

34Animal bones and shells were also found as grave goods in other children’s burials in the late prehistoric periods. It is not possible to define whether they were deposited by relatives of the dead person at the funeral feast during the burial ritual, or whether they were the remains of ritual food put in the grave pit for the deceased, following beliefs related to afterlife. Animal bones were found near the skeletons in two more graves of children aged below 1 year in the extramural Chalcolithic cemetery Sboryanovo-Demir Baba Teke (burials 1 and 8)16, in many graves in the Durankulak cemetery17, in Slatina cemetery18, and in an intramural grave of a child 8‑9 years old in Samovodene19. Shells in burial contexts were found in a number of Neolithic burials in present-day Bulgaria: at Ezero20, Malak Preslavets21, and in 11 Chalcolithic burials in the Durankulak cemetery22. Since it comes from a creature that lives in water, the shell was probably interpreted as a symbol of a water barrier between life and death, whose later representation was the Ancient Greek’s belief that the souls had to cross the river Styx in order to get to the underworld. The concentration of ash above the skeleton was the result of a fire over the buried body, which was probably part of the burial ritual.

35The infant burial at Tell Kosharna provides additional evidence supporting the suggestion that an intramural burial tradition existed within the Gumelniţa culture area in the Late Chalcolithic. It was a relic of a Neolithic tradition which, although less popular, existed together with the tradition of inhumation burials in extramural cemeteries in the Late Chalcolithic. The reasons for the existence of such a burial tradition have to be sought in unusual circumstances that required alternative decisions for the burial of a small part of the deceased population.

Building Level IV

36The level was documented in trenches A and Б, over an area of 51 m², at a depth of 3.90 m from the datum point. Only a layer of 0.30 m was excavated. It was dark green-brown in colour, and contained concentrations of potsherds in some spots. It also yielded two complete ceramic vessels typical of the Late Chalcolithic and the Gumelniţa culture (phase 1), as well as eight fragmented bracelets made from Spondylus shell. No features were identified.

37The earliest prehistoric settlement which is at the base of settlement mound no. 1 was not dated, as virgin soil was not reached. There is insufficient information about earlier artifacts from the treasure-hunt trenches, but this was not confirmed by our excavations. The fires which destroyed the three latest construction levels provide evidence for turbulent events which took place at the end of the Chalcolithic and which were disastrous for the settlement. The building (or the small excavated part of it) from level I is situated in the central part of the settlement. The thick layer of debris reveals that it was a large building, but it is not possible to define its exact size or interior. These facts, as well as the clay structure excavated inside (which was definitely not an oven), suggest that the building had important functions. The pits situated nearby probably mark a common assemblage of structures which could have had ritual functions. The thick layer of debris accumulated as a result of the destruction of building 1 could be considered an indication that it had more than one storey.

38The formation of a second settlement mound in the proximity of the first one23 was probably a result of demographic changes, which pushed the people to leave their village and to found a new one on the neighbouring hill. It was probably a daughter-settlement of the first one, founded in the 2^nd phase of the Gumelniţa culture, just before life on the first settlement mound ended. Future excavations at the settlement mound no. 2 will confirm or deny this hypothesis.

39The pottery and the small finds have parallels in pottery and artifacts yielded by sites in the territory of present-day Romania and in Tell Pietrele24. These parallels provide grounds to assign them to the Late Chalcolithic Gumelniţa culture. As early as the Late Chalcolithic, it is probable that the Danube played a significant role for the transmission of cultural interaction instead of acting as a territorial and community barrier.