Chapter Seven
The Drama microregion

7.1 Gerena flat site

7.1.1 General information and earlier studies

The Gerena flat site was investigated during three archeological seasons and a total of 300m2 was excavated (Lichardus et al. 2001, Figs. 39, 40). The site is partially published and the evidence summarized in the next section is gathered from all the Drama publications and some of the unpublished investigation reports.

The numbers in the brackets are the field numbers of all the features excavated at each of the sites in the Drama microregion and are given here to facilitate references to the original publications.

Archaeological evidence

The earliest occupation at the Gerena flat site and in the Drama microregion as a whole were three pits containing Early Neolithic pottery. There were numerous traces of various cut features but only the three pits were accepted as the initial human occupation of the site. They were followed by some settlement activity, leaving the traces of at least seven dwellings. After their abandonment, there was a period in which the site was used for cultivation rather than for dwelling activity. This claim was based on the presence of a buried soil, indicating a so-called hiatus, between the second and the third occupational levels. The last Neolithic horizon was heavily destroyed by past and present agricultural activity and traces of only two houses were excavated. Two other dwellings were also found but their Neolithic chronology was not very secure. Part of the Neolithic occupational area was overlain by a barrow. The barrow was not apparent at the time of the excavation but was pointed out by the local villagers. On the basis of the scattered surface pottery, the barrow was dated to the Roman period (Lichardus et al. 2001). The evidence for occupation of the Gerena flat site is summarized in Table 7.1.1:  

Table 7.1.1 Archaeological evidence from the Gerena flat site
Stratigraphic position Type of feature Content Traces of burning
I horizon Pit N589 Brown soil mixed with limestone and daub (wall rubble), numerous sherds and animal bones; the bottom was coated by a fine black layer, with a similar one 6cm higher  
I horizon

Pit N590 destroyed by

later pit

Animal bones, sherds, daub and stones; successive dark gray-brown clay layers with layers with more limestone  
I horizon Pit 606 Dark brown soil with few archaeological finds  
II/a horizon

Part of a house (570),

Heavily destroyed by house 571

Oven (485), two fragmented vessels  
II/a horizon Part of a house (579) Oven 568, no | noteworthy finds (Lichardus et al. 2000, p.106)  
II/a horizon Part of a house (580) Oven 569, destroyed by later pits  
II/a horizon Part of a house (595) Oven 567, compact wall rubble  
II/b horizon Part of a house (565) Oven 562, numerous sherds, some of the vessels are restorable, other vessels with parts still missing, a stone pestle  
II/b horizon Part of a house (571) Oven 566, many sherds The soil in the house contains numerous charcoal pieces
II/b horizon Part of a house (560) Oven 561, large quantity of wall rubble  
III horizon Part of a house (564) Pottery, pestles, grinding stones  
III horizon Part of a house (444) Sherds, numerous flakes, antler and flint tools, whole vessels, animal bones scatters, a stone pestle Burnt soil under the wall rubble, mixed with small pieces of burnt daub

Although II/a and II/b are accepted as two different building phases, it was pointed out that such a division is not very certain. Generally, all the houses have beaten floors and a wattle and daub construction. There were single postholes but more often wall rubble with imprints of wattling were found. In the general section pf the report, all the houses were said to be burnt but, in the individual description of each dwelling, evidence for fire was not discussed.

The material found at Gerena is presented in total rather than by its finds context. All together, 187 bone tools, 880 flint artifacts and 40 polished stone tools were found at the site.

The numerous flakes found in horizon II were interpreted as evidence for on-site lithic production in a chipped stone workshop. The presence of 26 microlithics triggered a discussion over their function and origin. It was claimed that they do not belong to some kind of Mesolithic technology since they were found in a context dated a millennium after the first Neolithic settlers at the Balkans. Rather, their application as hunting tools predetermined their shape. Such a functionalist approach will be discussed in a next section (see p.316- 317). The claim for the prevalence of hunting in the subsistence economy was supported by the unpublished animal bone analysis, in which the percentage of domestic animals decreased to less than 50%. The pedological investigations also suggested a limited possibility for cultivation during the Neolithic in Drama (Lichardus et al. 2001:111).

The pottery consisted of 26,000 sherds and numerous ritual objects (number not specified). Among the sherds, there were 40 whole and 60 restorable vessels. Fifteen % of all the sherds were decorated; 23 % of the decorated sherds were rims and 14% body parts. The publication contains illustrations of many whole and restorable vessels, as well as whole and fragmented altars, figurines and other ritual vessels (Lichardus et al. 2001, Figs.41-43, and Tables 26, 27).

7.1.2 The site and its surroundings according to the GIS analysis

The site is located in the flood plain at 113 masl (CDFig.483). It is in a flat area (CDFig.484) with a Northern aspect (CDFig.485). The visibility from the site is mainly along the flood plain 1.2 km to the North West and 1.1 km to the South East (CDFig.486). Also visible are the low hills 1.4 km to the North East of the site, the first terraces of the steep hill 1 km to the South West of the site, as well as some of the gently sloping areas 1.4-2.8 km to the North West. The other two sites are visible from the Gerena location, which means that, if there were contemporary sites on Kajrjaka and Merdzumekja, there was a visual relation between the sites.

The cost distance analysis places Gerena in an area with immediate and easy access to tell Merdzumekja and a more constrained access to the Kajrjaka site (CDFig.487). The short distances, however, between the sites suggest that there was quick and fairly easy inter-accessibility.

The logistical network consists of two paths (CDFig.488). The path to Merdzumekja is 400m long and crosses the river Kalnitsa (CDFig.489). It possible that there was another route between the two sites, if the actual level of the river during the prehistory was very high at this particular point between the sites. The hydrological factor was not included in the GIS analysis due to lack of any relevant data. The visibility from the path coincides very much with the visibility from Merdzumekja (see below) and only a few more visible spots were added at the marginal areas (CDFig.490).

The second path to Kajrjaka is 1.3 km long and winds to the West (CDFig.491). The last segment ascends to the South East; following the gentler path up the hill until the site is reached. The visibility from the path consists of the two viewsheds from the sites together with a few visible spots added to the marginal areas (CDFig.492).

In summary, if during the Neolithic there were contemporary sites on the three places in consideration, then there was a strong visual connection between the sites and they were in an area with easy inter-site access.

Exploitation area

The exploitation areas of the sites in Drama microregion were not studied by the means of GIS, because it was not possible to find a pedological or geological map of the microregion with a scale that could be transferable to the GIS coverage maps. The existing soil maps (see p.124) would produce a huge bias if overlaid on the precise 1:5,000 contour maps. Instead, my own observations made during the targeted field walking in summer 2001 are used here to reconstruct the possible exploitation areas for the sites in the Drama microregion.

The site is located at the confluence of the Kalnitsa river and a small local stream. The soil up to 500m radius from the site is dark grey alluvial clay with huge cracks and a few little stones. The next soil type distributed beyond the 500m limit is smolnitsa. On the margins of the study area, mainly over the hilly areas to the South West, South East and North North East, cinnomonic forest soil and some rendzinas were distributed.

Since the site is located on the right bank of the river, I would suggest that most probably the areas South and East of the site were cultivated during the Neolithic. The combination of the three major soils distributed there - meadow, smolnitsa and cinnomonic forest soil - facilitated mixed farming and crop rotation. The lack of exact data for the population number prevents any estimation of exploitation area. It is likely, however, that the arable land was sufficient to provide the necessary crops for a small hamlet or farmstead.

7.1.3 Summary and discussion

Given the present condition of the data, it may be inferred that human occupation in the Drama microregion started with structured deposition at Gerena, which continued during the following periods together with some settlement activity. Burnt houses are present at the site but whether the burning was deliberate or accidental is difficult to determine. Deliberate fragmentation, however, is a common social practice, as the ratios of the number of sherds: number of whole vessels: number of restorable vessels have shown. There is no feature continuity (e.g. house 571 destroys, rather than overlaying two previous houses); rather there is a pattern for digging into the rubble of the ancestors’ site (e.g. in house 580).

The general summary of the subsistence strategy of the occupants of the Gerena site is consistent within the evidence and interpretative framework applied by its excavators. But there are, however, a few problematic points that should be taken into consideration. After careful cross-reference of all the publications, it becomes clear that the total occupational sequence was not taken into account in discussing subsistence but only the latest Neolithic layer of the flat site (the so-called Gerena C). The attempt to dispute the Mesolithic chronology of the microliths is generally correct but the successive implication that they were used for hunting (as they were allegedly used during the Mesolithic) is a functional determinism (Lichardus et al. 2001, Lichardus et al. 2001a), whose application in any subsistence interpretation could be very misleading. Microliths could be used for hunting or for threshing (Clarke 1975). The problematic layer C contains 42.2% cattle and 53.5% goat/sheep, which in comparison to the first occupational level (17.3% cattle, 76.9% goat/sheep) shows a clear tendency for an increase in cattle-raising, which may have been related to some form of arable cultivation (Beneke, n. d.). Last but not least, the results of the pedological investigation were disputed in general (see p. 120 -125), which together with my own observations make me suggest that the region around Gerena site contains fertile, arable land that may have been used for arable land during each of the Neolithic occupations. Comments on the subsistence patterns of the first two horizons were not made, so any comparison is not possible given the present state of the data.

In summary, I would agree with the Neolithic chronology of the microliths but dispute their necessary hunting function. I would suggest that the Neolithic community at Gerena practiced some kind of mixed subsistence economy in which hunting, gathering and farming were all staple sources.

The type of occupation and the reasons for (re)-settling and abandonment of the site were not discussed. It was pointed out that, after the second horizon, the site was used for cultivation rather than for living but it was not discussed where the people that were cultivating the area were living. The occupational sequence published so far (for Gerena and for the Drama microregion in general) consists of pottery phases rather than inter-related contextual evidence (e.g. houses, other features, archaeological material, osteological material, etc.) and does not allow any reconstruction of contemporary and/or successive sites (settlements, depositional places, etc)

7.2 Merdzumekja tell

7.2.1 General information and earlier studies

Tell Merdzumekja was the main focus of investigation during the long-lasting research project in Drama. The site was almost totally excavated, with documentation provided of occupations from the Neolithic up to the Early Iron Age. The relative chronology followed by the team does not correspond to the commonly-accepted chronology in Bulgarian prehistory (e.g. Karanovo V is termed ECA according to Bulgarian chronology, while in the German version it belongs to the Late Neolithic) but rather uses some individual chronological schemes (e.g. Katincharov’s definition of MBA, which, according to almost all other Bulgarian BA investigators, is termed EBA3: Katincharov 1981). Arguing against such confusing relative chronology is not one of the aims of the current study. The phases mentioned in the current statement follow the original chronology of the Drama team, despite my general disagreement with such relative dating. In some places, the commonly accepted chronology is put in brackets.

Several publications present some of the evidence and materials found on the tell but a detailed monograph on each of the occupational levels is still in preparation. The following section summarizes all the data available so far, including material from some unpublished site reports.

Archaeological evidence

The earliest occupation on Merdzumekja tell dates from the period of Karanovo IV. The evidence from that occupation is very scattered and consists of part of a house (N685) with an oven (N686), a palisade ditch (687) and several pits (Lichardus et al. 2001, Table 25). The ditch is 20m long, 40-55cm wide and 70 cm deep. The postholes are 25cm in diameter and 30 cm from each other. The ditch is filled with brown loamy clay, mixed with fine pieces of daub and numerous charcoal fragments. The pottery found in the ditch has very similar characteristics.

The following occupation was from the Late Neolithic period, or Karanovo V (ECA). An area of 14,000 m2 was excavated, which was generally destroyed by later houses, pits and shallow holes. At least 61 houses were found on the tell - all located within the area bounded by the ditch (N360). The ditch is generally dated to the succeeding Karanovo VI period but its earliest phases (although not found along the whole ditch) date to the Karanovo V period. Also contemporary to the houses were numerous pits, several palisades and some shallow holes (Lichardus et al. 2001, Figs.31).

On the basis of the overlapping of houses, several building phases were claimed for the Karanovo V period. The houses were rectangular to slightly trapezoidal in shape, with one room. Their size varies between 27 m2 and 94m2, and there is a similar variability in orientation. The bases of the houses were dug into the ground and successively this foundation trench was filled with earth a) to serve as insulation and, b) for the leveling of the floor. The earth was overlaid by a wooden floor, in turn covered by a beaten clay layer. The sequence is finished with reed rugs (Lichardus et al. 2001, Figs. 32-34). The postholes found on the tell together with some imprints of woven sticks suggest a wattle and daub construction. The inventory of the houses consisted of ovens, grinding stones – usually located close to the ovens - platforms, shallow holes and ash-pits. Outside the houses, there were numerous pits used mainly for storage or with an unknown function. Once the initial function was over, the pits were turned into rubbish dumps (Lichardus et. al. 2001). Details of pit deposition were given for only two pits (Nos. 67 and 26/33), both of which were interpreted as sacrificial pits (Fol et al. 1989). The first one contained two shepherd’s crooks made from antler. The second one had a compact pottery scatter, over which numerous deliberately fragmented tortoise shells were found.

Several palisades and small ditches were found within the Karanovo V settlement, which the investigators interpreted as features of unknown function.

Very little archaeological material, mainly sherds, was found in the houses in general. This was interpreted as a result of abandonment, after which only the unnecessary or useless things were left over. The artefacts found in the Karanovo V settlement are presented in general and mainly consist of fragments of pithoi, cooking vessels, table vessels, spoons, miniature vessels, vessel imitations, pendants, beads, Spondylus bracelets, buttons and bone applications (Lichardus et al. 2001, Fig. 36 and Table 28). Also found on the tell are figurines, clay plaques, altars and other ritual objects (Lichardus et al. 2001, Tables.19-22). The figurines were divided into two types. The first type was specially made to facilitate deliberate fragmentation. In contrast, the second type was produced in a way, which prevents fairly easy fragmentation (Lichardus et al. 2001, Figs. 37, 38). Both figurine types were found fragmented, which made the investigators conclude that this was some common act of ritual breakage (Lichardus et al. 2001: 94).

A common find were also the perforated circular pieces of pottery, with rounded edges, usually called net weights (the excavators use descriptive characteristics rather than naming them) (Lichardus et al. 2001, Table 24).

Only one case of a foundation deposit was reported from the Karanovo V settlement. Under the floor of house 900, in pit N966 there were two dishes with river shells in each of them (Lichardus et al. 2001, Fig. 35)

The following occupation on the Merdzumekja tell dates to the ECA Karanovo VI period (for Bulgarian and British scholars, LCA). The settlement was totally excavated over an area of more than 10,000m2. At least 25 houses, shallow holes, storage pits and pits with other functions were found. The site was surrounded by a ditch up to 8 m in width (N360) and by a double palisade at the top of the North West slope (Lichardus et al. 2001, Fig.19). At the time of the publication, the link between the ditch and the palisade was not clear. All but two excavated features, however, were within the area bounded by the ditch and the palisade. The exceptions comprised two pits (Nos. 830, 825), interpreted as clay-pits, that lay outside the enclosed area. Traces of house reconstruction (e.g. N224), some overlapping features and dwellings, whose plans were not possible to reconstruct, made investigators infer more than one occupational phase. It was not specified, however, which set of features belonged to the earlier phase.

The construction of the houses was similar to the construction of Karanovo V houses. The only difference was in the rectangular shallow hole dug into the ground and called by the excavators a cellar (Lichardus et al. 2001, Figs.20, 21). The distance between the bottom of the cellar and the dwelling floor varied between 90cm and 1m. The function of these cellars was to isolate the damp and the cold during the winter and for cooling during the summer (Lichardus et al. 2001: 58). Most of the houses had a North West / South East orientation, rectangular shape and their area varied between 20.5m2 - 104m2. Some of the bigger houses had a shed attached to one of the short walls (one exception was in House 137, where the shed was attached to the long wall). All but one (N244) were one-storied houses, with an entrance on the one of the short walls and with no evidence for windows and the type of the roof construction. In most of the dwellings, there were domed ovens and related clay shelves, which were interpreted as holding vessels. In house N380, there was a vessel dug into the clay shelf that contained some stones interpreted as pot-boilers. Also close to the ovens, there were usually big pithoi, strainers, ladles, grinding stones, scrapers and pestles. The vessels were most probably laid on shelves along the walls, since they were found in a row along the walls.

The main source for house furniture is House 244, which, together with the above described features, contained over 200 vessels (Lichardus et al. 2001, Table 4). Some of the vessels were whole and contained other vessels (Lichardus et al. 2001). During a visit to a National Museum of History exhibition about Drama (July 2002), I had the opportunity to see the pottery from house 244. It consisted of mainly whole, well-burnished, fine vessels of different shapes and sizes. According to the excavators, this house was the only one with two storeys; on the second floor, the fine, decorated pottery was kept, while, on the first floor, there were the cooking and storage vessels. There were ovens on both floors, and different types of stone tools were found mainly on the first floor.

Two main types of pit were recognized in the Karanovo VI period. The first type comprises shallow pits of irregular shape, located very close to the houses. The second type includes small, circular to oval pits with different depths, located at some distance from the houses, which were mainly used for storage. Traces of a street were also found, which took the form of a strip covered by small stones and sherds.

All the houses were burnt but the data is spread all over the reports, rather than in a single consistent paragraph detailing the end/abandonment of the houses and/or the settlement.

The later occupational phase of Karanovo VI consisted of a ritual platform and a series of structured deposition places covered by stones and a rectangular building (Lichardus et al. 2001, Fig.16.). The ritual feature (N37) is reconstructed by the excavators as a rectangular platform 3.4 x 4m in size, made from sand, clay and chaff, whose surface was several cms above the ground. On the right and left side of the platform, there were two shallow rectangular pits. Along the North side, a 2m-high wall was built. A raised path 2.2.m long and 0.75m wide was attached to the platform (Lichardus et al. 2001, Fig.17). The feature had traces of a massive fire but excavators had difficulties in deciding whether these were a result of fire during the building of the feature, during its existence or after its active use. It contained sherds, a spoon, a vessel with a round base, two miniature vessels, two clay wheel models, two fragments of clay plaques and a fragment of a zoomorphic figurine. The paucity of clear dwelling traces led to the conclusion that feature 37 should be related to ritual activity.

The building from the later horizon (N206) had two rooms with traces of a massive fire, a hearth, three whole vessels, 130 sherds that belonged to restorable but still not whole vessels, a figurine, a stylized zoomorphic figurine, a wheel model and two rectangular vessels (Lichardus et al. 2001, Fig.18). Bone tools, polished stone tools, grinding stones and many animal bones were also found in the building.

Close to the building, there were two places for structured deposition, each covered by stones, plus one more at some distance; all in all, there was a total of three large (Nos. 371, 241 and 253) and 23 small stone scatters. Generally, they follow a similar pattern of deposition – tools, ritual objects, bones and sherds, overlain by a stone scatter. In some cases, the bones were in anatomical order. Together with the deposition of figurines, fragments of altars, etc. in between the bones, this fact led the investigators to conclude that this resulted from deliberate rather than accidental deposition. Most of the scatters were dug into the earlier Karanovo VI layer (houses 244 and 380 in particular).

The ditch (N360) had at least six re-cuts (Lichardus et al. 2001, Fig.22). The excavations of the 25m wide zone between the ditch and the built settlement area revealed the presence of a bank whose base was fortified with stones. The pottery in the ditch was mainly from the Karanovo VI period, with less material from Karanovo V. The presence of almost whole Karanovo VI vessels and some flint blades was interpreted as an indication of deliberate back-filling of the ditch with house rubble following some kind of ritual activity, after the initial function of the ditch was completed.

The chronology of the six re-cut phases was not yet clear at the time of the publication and a preliminary suggestion was made that it is not impossible for the first three phases to have been filled with material from the Karanovo V settlement. The last (sixth) phase was accepted as belonging to a period post-dating the Karanovo VI occupation of the site. The entrance to the village was accepted as the so-called earth bridge between the North West and South East ends of the ditch (Lichardus et al. 2001, Fig. 23). In that area, a complex of several pits and palisades was excavated, which however, did not receive any interpretation. The data for the fill of the ditch is scattered throughout the site reports and could be summarized as different coloured clay patches, mixed with sherds, bones, charcoal and stones.

The palisade at the North West end of the tell consisted of a double row of postholes. The distance between the rows varies from 160cm to 180cm.

The two pits (825, 830) considered as sources for clay production were filled with settlement rubbish (Lichardus et al. 2001: 65), viz., sherds, charcoal, bones and daub, deposited soon after the final use of the pits.

Each house produced an average of 15,000 sherds, from which up to 200 vessels can be restored (Lichardus et al. 2001, Figs.24, 25). Apart from the vessels and the sets of vessels, there were also lids, ladles, spoons, funnels and strainers. The presence of earlier sherds in a later context received the unlikely interpretation of the storage of building material. Sherds and animal bones were found in the construction of the ovens, floors and walls and it was concluded that these were kept in the houses for future construction work. An alternative explanation concerns the inclusion of older, ancestral material in the materials used for building of new structures, to presence the ancestors (for an example from the Bronze Age of Mataci, in Dalmatia: Chapman et al. 1996).

Very few metal objects were found (Lichardus et al. 2001, Fig.26), which contradicts the numerous finds of slag, globules of metal, a tuyère and smelting pots. These remains of metal production are potentially very significant, since there are few, if any, examples of on-tell evidence for copper smelting. Bone and clay figurines, anthropomorphic vessels, zoomorphic figurines, clay models of wheels and boats, clay horns, stylized zoomorphic figurines, altars, clay plaques, models of ovens and cult buildings complete the variety of finds at the Karanovo VI settlement (Lichardus et al. 2001, Figs. 27-30, Tables 8-16). It was underlined that, despite a careful search, the missing parts of the figurines were not found. On a completely excavated site, this indicates transport of parts of figurines off site (for N E Bulgarian tells such as Ovcharovo, see Chapman 2000).

The 25 houses from the Karanovo VI period were suggested to have been distributed between a few clusters, each consisting of six to eight dwellings. The last settlement was abandoned after the houses were deliberately leveled. The well-preserved pottery in the houses made the investigators infer that the deposition of the vessels and the successive destruction of the houses was a deliberate act. They also suggested that the new settlement moved to the tell Kirchova vodenitsa at 4.5 km to the North West. In this final discussion on the Karanovo VI occupation, it was not specified whether the houses were destroyed by fire, despite the scattered reference to fire in the publication. Possible reasons for the deliberate act of abandonment were also not discussed.

The next occupation on Merdzumekja tell is from the EBA and represented by a paucity of evidence. Two almost whole vessels were found in pit 75 (Lichardus et. al. 2001, Fig.13). The other evidence, mainly sherds in secondary deposits (Lichardus et. al. 2001:41), was considered as post- Karanovo VI but not characteristic enough to be related to Ezero A (EBA1 according to Bulgarian chronology). Since on the neighbouring Gerena flat site, two vessels from the Cernavoda I period were found, it was concluded that the EBA in Drama is represented by the local post-Karanovo VI variant on the tell and Cernavoda I material at the Gerena flat site. The type of the occupation, the paucity of EBA evidence or the differences between the pottery on two adjacent sites, etc. were not discussed.

More secure EBA evidence derives from an area immediately South East of the tell. A settlement from the Cernavoda III period was excavated over an area of 300m2. The cultural layer consisted of a scatter of wall rubble, sherds and numerous pits (Lichardus et al. 2001, Figs.14, 15). A burnt house of wattle and daub construction and a clay-coated wooden floor was found. Ten meters from the building, a pit with pottery, stones, melting pots, fragments of tuyère and metal globules was excavated. This evidence was interpreted as an indication of on-site metallurgy.

The most significant BA presence on the tell is marked by the MBA (EBA3) ditch, one building and a few pits (Lichardus et al. 2001, Figs.5-10). The ditch is located on the North West slope of the tell and encloses an area of 41.50/38.50 m with 10-12 % difference in slope (Fig. 7.2.1). In the Southern part, there is a 3.60m long gap in the ditch, considered to be the entrance to the enclosed area.

Fig. 7.2.1 Middle Bronze Age enclosure and the Karanovo VI settlement

On the basis of the experiments conducted on the tell - a zone along the ditch was left open and, after eight years, it was visible on the surface as a shallow hole – it was inferred that the ditch operated as an open feature for a short time - not more then a generation. Apart from the material that was a result of wall erosion, deliberately deposited material was also recognized, especially fragmented pottery from the Karanovo V and VI periods, deriving from the houses that the ditch construction has destroyed (Fig. 7.2.1). The pottery, which dates the ditch, is from sherds scatters that have produced some restorable vessels (Lichardus et al. 2001, Fig. 11, Tables 1, 2). It is from the MBA (EBA3) and its deliberate deposition was confirmed by the fill of two of the vessels that contained wheat grains. Other vessels were thrown into the ditch, that caused their breakage. The fill of the ditch consisted of stones, wall rubble, loom weights, whorls and animal bones (Lichardus et al. 2001, Table 3). As a general pattern, under the stones. whole or almost whole vessels were placed in single or large scatters. Above the stones and structured depositions, there was burnt rubble. There were no traces of fire on the wall of the ditch, which made investigators to conclude that hot daub was thrown into the ditch and that the actual fire took place in some building close to the ditch. The building was believed to serve some ritual activity. The only candidate for such a building is house N370 situated at 10-12m from the ditch. The reconstruction given by the excavator is that initially the ditch was dug to define the boundaries of an area, to which access was restricted to the South East. During that time, the feature should be regularly cleaned and maintained in order to prevent erosion or unwanted sedimentation. Later the ditch was used for deliberate deposition, after which environmental conditions contributed to the final in-filling of the ditch.

There are two noteworthy facts from the evidence for the fill of the ditch, which were not discussed by the excavators. First, there were pieces of daub in different forms deposited in the ditch. During my exhibition visit, I observed arm- and leg-shaped daub pieces that derive from pre-MBA periods. Therefore, it is likely that the secondary use of daub was an important social practice in Drama microregion. Secondly, the wall construction of some of the rubble in the ditch contained stones of non-local origin (Fol et. al. 1989). Such a pattern may be interpreted as a deliberate incorporation of exotic objects into the social practice of structured deposition.

There was no clear settlement evidence from the MBA, since the only building (N370) was accepted as a ritual feature. It was a rectangular building with probable wattle and daub construction and an entrance from the North West (Lichardus at al. 2001, Fig. 12). There were 13 postholes in the inner space of the building, that were interpreted as a part of the roof construction. The pits excavated from this period were not discussed, only a brief description of storage pits was given – circular to oval in shape, with a broader basal than upper diameter, and often containing stones.

The latest prehistoric occupation on Merdzumekja dates to the LBA but is not discussed since the evidence was scattered and lacking in secure contexts.

The very selectively presented data from the tell contain a little evidence for features that destroyed other features, which is summarized in Table 7.2.1:

Table 7.2.1 Evidence for destruction of lower features by later features
Type of feature Destroyed feature
Ritual platform from Karanovo VI period Earlier houses- 19 from Karanovo VI and 150 from Karanovo V
Stone scatters with structured deposition from Karanovo VI Rubble of houses 244 and 380 from Karanovo VI period
MBA building N370 Ritual platform from Karanovo VI period
BA pit N300 House 206 from Karanovo VI
MBA,EIA and Roman pits MBA ditch
MBA pit 249 MBA building 370

7.2.2 Plant and animal remains

Karanovo V period

The only plant remains published so far from tell Merdzumekja consist of the collective find of the carbonized fruit of Cornus mas (Cornelian cherry). The sample was taken from house 3 and consisted of burnt rubble and several hundred fruits. It was not specified whether the fruits were only Cornelian cherries or whether there were some other species as well. It was inferred that the fruit was used for food either in a fresh or in a dry condition. The find was used to suggest that the shrub was distributed in the low woodland around the tell and was gathered by the ECA population (Kuster 1989).

The detailed osteological analysis is not published yet and only some very coarse-grained general information is published so far. The percentage of the domestic animals is greater than the percentage of the wild animals. Cattle and caprovines were probably equally represented, pigs were around 10%, and dogs were 2% from the total bone assemblage. The wild species consisted of wild boar, hare, fox, fallow deer, red deer, aurochs, wolf, wildcat and brown bear.

Karanovo VI period

Plant remains from the later occupation on the tell were very few despite careful flotation; the materials recovered contained evidence for cereals and some other species.

Animal bones were analyzed during an earlier stage of the investigations that have shown 93% of the bones derived from domestic species and only 7% from wild animals (Bökönyi 1989). Such a unique pattern was explained by either some specific subsistence practices at Drama or by the fact that the bones derived only from dwelling contexts and that bones discarded outside of the houses or in the pits were not taken into account. The later osteological analysis of over 30,000 bones (an average of 1,000-1,500 from a house) is not published yet, thus leaving this big discrepancy in the interpretation of the animal bone evidence. The domestic species are represented mainly by cattle (53%), followed by sheep/goat and pigs. Dog bones amount to no more than 2%. Some traces on the cattle horns were interpreted as evidence for yoking. The wild animals consist of wild boar, auroch, red deer, fallow deer, hare, fox, brown bear, wolf and wildcat.

7.2.3 The site and its surroundings according to GIS analysis

The site is located on low hill in the flood plain of Kalnitsa river at 119 masl (CDFig.483). It is in a flat area (CDFig.484) with a South West aspect (CDFig.485). The visibility from the tell is good over the flood plain 2.4 km to the North West, over the first terraces and the highest areas of the steep hill to the South West, as well as over the low hills 1.3 km to the North East of the site (CDFig.493). The panorama to the South East is limited by a small hill up to 182m high.

The visibility with an additional 3m, which is the maximum height of the tell, shows barely any improvement, and that only in the marginally visible areas (CDFig.494). The remaining two sites are visible from the tell in both viewsheds.

The cost surface analysis results are very similar to the Gerena case study (CDFig.495). The logistical network contains two paths, one of which already commented in the previous case study (CDFig.496). The un-discussed one is the path to Kajrjaka, which is 1.3 km long; 700 m South East of Merdzumekja tell, it joins the path from Gerena to Kajrjaka (CDFig.497). The visibility from the path combines the static viewsheds of the two destination sites and has a few more visible spots toward the edges of the visible areas (CDFig.498).

In summary, during the time of Merdzumekja occupation, the tell was in visual connection with the earlier Gerena site and with the contemporary (?) Kajrjaka site. The former was in immediate vicinity to the tell, the latter was with fairly quick access.

Exploitation area

The distribution of soil types around the Merdzumekja tell is the same as at the Gerena flat site. The actual exploitation area most probably was to the North of the river Kalnitsa, since the site is located on the left bank of the river. The terrain there is less constraining than in the areas South of the river Kalnitsa and has a good cover of arable land. The population of the Merdzumekja tell varied between 125 and 237 (for estimation pattern see Chapters five and six); for this population 131 to 249 ha arable land was needed, that was available to the North and East of the site and was sufficient to maintain a successful long-term agro-pastoral subsistence strategy.

Resources and catchment area

The bone tools found in the Karanovo V houses were awls, polishers, chisels and axes for woodworking (Lichardus et al. 2001, Table 23). The polished stone tools were chisels, axes, pestles and grinding stones – all made from local amphibolite, gabbro and diabase. Flint tools were also found but cores were a rare find.

The bone and horn tools from the Karanovo VI period were highly standardized and were used for the working of wet and dry wood, bark and leather/fur (Sidera 1996). Also for wood processing were used polished axes and chisels made from the same local rocks (Lichardus et al. 2001, Figs. 6, 7).

The flint technology of Karanovo VI period differs from the preceding period in the size of the blades, which are much bigger, as well as in the type of the raw material. According to Dr. Ts. Tsonev (pers. comm.), the chipped stone tools displayed in the Drama exhibition contains both local and Radingrad flint material. Unpublished report of the study of 1,200 flint artifacts from the tell concludes that débitage was made from local sources, while the majority of the tools were considered as imports (Ziesaire n.d.). The studied chipped stone assemblage (n = 157) from the Karanovo VI period derived from 11 houses. The distribution of artefacts in the houses varies from one (house 212) to 44 (house 244). Seven houses have fewer than 11 flint tools. The main types were blades, scrapers, retouched tools and borers. Single residues from cores are the only production waste found in the Karanovo VI settlement. Some of the tools have traces of re-use and most of them suffered some fire/heat influence (Ziesaire n.d.).

Apart from the flints that derive from North East Bulgaria, there is very little published evidence for the catchment area of the settlements at Merdzumekja tell. Spondylus shell (from which the bracelets were made) was believed to derive from the Aegean (but a Black Sea source is also possible). Whether the bracelets were coming as a ready pendant or in the form of raw material is not clear. It is, however, a strong evidence for links with the Mediterranean or Black Sea region. The river shells found as foundation deposit suggest some fishing and gathering activity that may have been in the 5 km agricultural limits. The same is valid for the hunting, which has been practiced by the Merdzumekja population as the presence of the wild animals has shown. The dominant wild species was the wild boar, which may relate to the fact that there were figurines made from wild boar bones. Whether hunting activities were taking place in the adjacent upland areas is difficult to say due to the lack of pollen data, hence evidence for deforestation.

Some non-local rocks have also been brought to the site but the distance from which they derive is difficult to establish. It is important, however, that, despite the presence of rocks around Merdzumekja, some other types of rock were produced or exchanged, and these were considered as an important component of the structured deposition on the site.

7.2.4 Summary and discussion

The long occupational sequence in Merdzumekja shows a pattern of recurrent social practices.

Structured deposition was most probably the commonest as it appears in various forms in each settlement layer. Deliberate fragmentation is the other widely performed activity that is related to both ritual (figurine breakage, deposition of fragmented vessels, etc.) and quotidian (e.g. construction of ovens, floors, etc.) practices.

Various objects were used for deposition that derive from the every-day repertoire. There were, however, some particularities (e.g. burnt rubble, daub features, non-local stones) that imply highly structured practice of interweaving the everyday with the exotic, the ritual with the quotidian.

There was no feature precisely overlaying another earlier structure in terms of deliberate continuity of layout (Bailey 1990, 1996), rather, there was a repeating pattern of cutting into ancestral deposits. It is important to point out that the MBA ditch cuts only three earlier houses in an otherwise densely occupied area (Fig. 7.2.1).

The only strong evidence for accumulation is house 244. There were other fully excavated houses but only this contained such a quantity of material. Taking into account that house 244 was the only two-storied dwelling, I should suggest that the inhabitants of the house have gained some prestige, displayed, and thus authorized, by the large quantity and variety of objects.

Nonetheless, the deliberate abandonment and burning of the houses was not explicitly related, although it may be assumed that this was one of the crucial activities on the Merdzumekja tell, whose major goal was successful social reproduction.

And finally, there is a very clear pattern of structuring the area of the previous settlement. Various types of structured deposits (e.g. pits, platforms, etc.) reveal a complex practice of conceptualizing the space that has specific meaning for the participants and witnesses of such activity. There is evidence for deliberate deposition in pits and ditches in the very first occupational level, as well as throughout the whole occupational sequence during which course the depositional practices diversify (e.g. platforms and stone scatters). Therefore, there is not only a synchronic discourse (exchange of massages) through the way of deposition but also a diachronic discourse and/or continuity of depositional messages. At present state of the data it is difficult to reconstruct the actual sequence and possible meaning of the specific deposition but I should infer that structured deposition within each occupational level, as well as through time was a major means of organizing communication in the social life of the tell Merdzumekja.

7.3. Kajrjaka flat site

7.3.1 General information and earlier studies

The Kajrjaka site was investigated during five archaeological seasons, during which a cultural layer consisting of pottery from Karanovo III, IV, V and VI periods, EBA Ezero A and B periods, MBA, pits and pottery from the EIA and a Roman cemetery was excavated. The major site on Kajrjaka hill was the Roman cemetery and, since the earlier occupation levels were heavily destroyed, the prehistoric evidence was not discussed. The very scanty published data is summarized in the following section.

Archaeological evidence

The Neolithic, Eneolithic and EBA layers on the Kajrjaka site were mentioned as present, a table with EBA sherds was published and further comments on this early evidence were not made at all. More attention was paid to a clay reel found in a secondary context, which had some incised signs interpreted as a Linear A inscription (Lichardus et. al. 2001, Fig. 57). It was pierced, and hence taken to be worn on a necklace (Fol and Schmitt 2000). The probability that the inscriptions were Linear A rather than just incised decoration was discussed in the context of similar finds discovered outside the island of Crete. Different objects (e.g. clay balls, body sherds, etc.) with Linear A inscriptions have been found on some Aegean islands, in the Peloponnese and on the coast of Asia Minor coast; the closest such find to the Drama microregion comes from Samothrace. The relatively coarse execution of the reel made investigators suppose that this was a barbarian imitation of an imported object. Since the reel was found in a secondary context, it was difficult to date it. The suggested chronology was in a period after the LH IIIB/C phase, when Linear A was still in use in unofficial texts (Fol and Schmitt 2000). Many important questions were triggered by this find, such as: – were there any documented trade contacts between the Aegean and the Upper Thrace, were the objects with Linear A inscriptions objects of exotic exchange and why was an imitation of linear text needed? However, these were not explored any further. The important information that this find has revealed is to confirm that prehistoric societies were not in isolation but were participants in regular networks – in particular, the Aegean and Upper Thrace.

Another evidence for the same general direction of contact is the presence of Mycenaean and some Protogeometric sherds. They were related to features from the EIA, which means that either there was an exchange of earlier pottery during the IA, or most probably that the sherds were re-deposited by the EIA population. In both cases, however, there was a trade/exchange of fragments and the deliberate storage of ancestral objects. Nine of the sherds were given for neutron activation analysis (NNA) to trace the possible pottery workshop. There were no conclusive results but four of the sherds were thought to derive from Asia Minor (one almost sure, the other three less), three were possibly related to a Macedonian pottery workshop and the last two shared no similarities with the Aegean world. Despite the relative uncertainty of the data, the long-distance contacts between Drama microregion and the Mediterranean were most probably an important part of socio-economic life in the later prehistory of South East Europe.

Another direction of possible contacts was suggested on the basis of evidence from grave N27. The grave pit was destroyed by the building of one of the EIA features and contained a child crouched on the left side and a cup identified as deriving from the Tei - Monteoru culture to the North of the Danube (Lichardus et. al. 2001, Fig.58). The skull was missing and the anthropological analysis did no show any traces of ritual treatment. Two fragments of a coarse vessel were found near the head and another one in the knee area. Close to the area where the head should have been lay one red deer bone, one cattle bone, one cattle tooth and two caprine bones. The investigators discussed in great detail the chronology and parallels of the cup rather than trying to explain why and how the cup was finally displayed in a grave with some obvious peculiarities. Given the present condition of the data (e.g. it is not clear whether the head went missing after the building of the EIA feature, etc.) I would suggest that the buried child had a specific social status that was underlined through the deposition of prestige grave goods – an exotic pot. The significance of the dead person was reinforced by the fact of the missing skull, if we accept that it was taken to be kept among the living. The lack of ritual treatment could be interpreted that the skull was removed when the flesh has already decayed, hence there were no traces of violent decapitation. The hypothesis that the burial was not a single act but rather a continuous process is indirectly supported by the presence of at least three different animal species in the grave, probably deliberately killed and ritually consumed in a feast during the long decaying process and some of the bones were finally deposited in the grave. It should be recalled that another child burial – this time from the Karaivanovi mogili barrow cemetery – was also furnished with a Tei-style cup (see above, p. 205-209).

Last but not least is the LBA evidence consisting of scattered finds generally related to the Nouă-Sabatinovka-Coslogeni culture. Only one concrete piece of evidence from the Kajrjaka site was presented – a fragment of a so-called sceptre, deposited in a stone cairn (N97). The sceptres are stone bowl-like objects with both practical and symbolic value generally related to the Nouă-Sabatinovka-Coslogeni culture (Lichardus at al. 2001:170). Four more similar artefacts were found in the adjacent areas that, together with some plastic decoration (considered as characteristic for Nouă-Sabatinovka-Coslogeni culture) on both hand- and wheel-made pottery, has suggested the influence of this LBA culture over the population in South East Bulgaria. According to the excavator, it was a result of migration, followed by selective acceptance by the locals (Lichardus et. al. 2001: 169).

Given the present condition of the published data, such an intensive movement of people (Nouă-Sabatinovka-Coslogeni culture) and objects (Mycenaean pottery, Tei – Monteoru pottery) toward the Drama microregion is far from being argumentatively proved and explained. The evidence given to support the North Pontic presence in Bulgaria is selectively chosen (cf. Gaydarska 1998) and few specific parallels have been found amongst Aegean and Anatolian artefacts.

7.3.2 The site and its surrounding according to GIS

The site is located on a high terrace, at 148.8 masl (CDFig.483), in the immediate South East vicinity of the present village of Drama. It is a site with a horizontal stratigraphy, which makes it problematic to make a GIS study based on a point definition. However, it is located generally in an area with a 2o to 5o slope (CDFig.484) with an East North Easterly aspect (CDFig.485). The visibility from the site is very much dependent on the actual point from which the viewshed analysis is performed. The best visibility is achieved while looking from the edge of the terrace, which is one of the possible extremities of the site. Going up the hill to the South West significantly diminishes the visibility, restricting it only to the areas North of the viewing point. The panorama from the edge of the terrace is good over the Kalnitsa valley and the sloping foothills that surround the flood plain to the North and South East (CDFig.499). There is some strip-like visibility to the South West of the site towards the higher parts of the steep hill on which the site is located. The change in the visibility status from different viewing points was confirmed during my visit to the site in summer 2001.

The cost surface analysis puts both Gerena and Merdzumekja sites in the first cost strip (CDFig.500), thus locating them in an area with equal access from the site. The logistical network again consists of two paths already commented in the previous case studies (CDFig.501).

In summary, the earlier and contemporary sites were located in one and the same relative distance from Kajrjaka. They were visible from the edge of the terrace but not from other parts of the site. Most probably, then, intersite visibility is a secondary factor in the location of the Kajrjaka site.

Exploitation area

So far, there is no secure evidence for settlement activity on the Kajrjaka site. However, if there was some kind of occupation, which requires agricultural activity, it should have been very similar to the one discussed for the Gerena flat site. The site is located on a steep hill and, unless some as yet undetected terracing has taken place, it is not suitable for intensive agriculture. There are, however, flat and gently sloping arable areas to the North and East of the site, all situated on the right side of the river Kalnitsa, and these would appear to be the most likely candidates for the exploitation area of the Kajrjaka site.

The resources, land use and catchment area of the Kajrjaka site cannot be discussed because of the paucity of available archaeological evidence.

7.3.3 Summary and discussion

The evidence from the Kajrjaka site points to two possible types of site development. First, the inconsistency of the data in respect of the existence of an occupation layer is maybe due either to erosion process or to later occupations of the site, in which the IA structures and the Roman cemetery have destroyed the earlier layer. Secondly, the lack of clearly interrelated settlement, burial or depositional activities may correspond to some kind of highly formalized structured deposition, from which only scattered evidence is now available due to the later destruction. The hypothesis for the existence of a place whose primary function is structured deposition does not contradict the general evidence from the Drama microregion, in which structured deposition is widely practiced. In such a case, the Kajrjaka site presents evidence for long-lasting depositional continuity, with closely related cultural memory providing another aspect of continuity from the Neolithic up to the LBA.