Chapter Five
The Sokolitsa Microregion

In this chapter, the sites and monuments located in the Sokolitsa valley are discussed in turn as regards the general information about their excavation, the state of publication, the site contexts and their material culture and chronology and any remains of plants or animals, as well as a GIS analysis of their location, visibility, route network and site territory. The chapter starts with tell Galabovo, which is the Westernmost site in the Maritsa Iztok study area (Fig. 5.1.1). The presentation sequence of the sites is from West to East. An overall synthesis of the landscape, material culture and society is deferred to Chapter 8. Here, the dates of settlement occupations and burials are tabulated by site for Maritsa Iztok study area (Table 5.1.1).

5.1 Tell Galabovo

5.1.1. General information and earlier studies

Tell Galabovo has been investigated for four archaeological seasons, as part of the long-term research scheme of the Maritsa Iztok Expedition. The Western periphery of the tell was heavily damaged prior to the excavations by road construction, that removed a substantial part of the archaeological deposits. In addition, two channels for electric cables have destroyed the upper part of the tell. The nearby Briquette Factory produces coal dust as industrial waste that has coated the tell with a thick layer of hard, black, carbonaceous deposit.

The site was excavated by stratigraphic trenches and a network of sondages at the top of the tell. The trenches are oriented North-South along number 4 of the total 5 x 5 grid of the excavations. The central profiles follow the four cardinal points and are along gridline L7. The sondages in the upper and South West part of the tell started in squares O5-O7 and B4/C4 and were subsequently enlarged in accordance with the contexts discovered, the aims of the investigation and the current financial status of the Expedition (AFig. 5.1.1). The total excavated area of some of the occupational levels is given in table 5.1.1.

So far, fourteen building horizons have been dated to the Bronze Age1 and three horizons to the Late Copper Age (AFig. 5.1.2). Sherds from the Late Neolithic and very final Copper Age were found in a disturbed stratigraphic context. Settlement occupations from the Iron Age, Roman and Medieval times were also documented on the tell. The current height of the tell is no more than 7m and the surviving dimensions of its base measure 125m North-South and 100m East-West.

Table 5.1.2 Excavated area of the BA horizons of tell Galabovo
BA horizons Horizontally investigated area
VIII 100m2
IX 150m2
X 175m2
XI 600m2
XII 500m2
XIII 650m2
XIV 75m2

Palaeo-botanical and archaeo-zoological studies have been made, as well as some lithic and petrological analyses (Popova 1991, Ribarov n.d., Gatsov n.d).

The results of the excavations and some major archaeological interpretations of the Galabovo data were published in a series of articles and monographs (Panayotov et al. 1991, Leshtakov 1993, 1995, 1996, Leshtakov et al. 2001, Leštakov 1993, 2000).

Archaeological evidence

Before turning to the occupational sequence of the tell, an important point should be mentioned. The main research priority in Galabovo investigations was the vertical stratigraphy of the tell. Special attention was paid to the horizontal stratigraphy and plans in the publications of only some of the Late Chalcolithic and MBA layers (X-XIV). The limited excavation area and grid-oriented documentation impedes the horizontal correlation of the features. Horizontal juxtaposition of archaeological features for each subsequent layer has not been undertaken, apart from in the above-mentioned Late Chalcolithic and MBA publications (Panayotov et al. 1991, Leštakov 1993, 2000). The correlations presented in the following pages are based on my own work with the field documentation, partially supported by site illustrations.

Copper Age

The three building horizons from the Chalcolithic were claimed on the basis of identified dwelling floors, usually made of beaten clay. The excavated area is 50 m2 in the south periphery of the tell (AFig. 5.1.1a). Archaeological evidence is summarised in Tables 5.1.2-5.1.

Table 5.1.3 Copper Age dwellings from tell Galabovo
Horizon/ square Type Pithos Ovens/ hearths Inventory Traces of burning Comments
I horizon/C4 Part of a house (AFig. 5.1.3c) No An oven and feature for keeping the warm air 2 whole and 4 fragmented vessels, a billet stone hammer-axe, a grinding stone, a stone adze, 4 flint tools. In one of the whole vessels there were 2 flint tools and some flint flakes considered as a production waste layer of burnt house rubble, ash and charcoal
II horizon/ Part of a house many pottery sherds, animal bones, four whole vessels, a collective find of stone tools no
III horizon/ C4 Part of a house (AFig. 5.1.3a) Dug into - filled with black soil mixed with charcoal, many sherds and two bone tools (one imitating a long blade (AFig. 5.1.7H)) many sherds, a pot with missing upper part, animal bones, a bone figurine with traces of red ochre, copper slag; some of the sherds with red ochre. Published as burnt house Probably to the same house belong the burnt house rubble overlaid by fragmented flint tool and a stone axe found in the same square C4; the red ochre present more often in burial context
III horizon/ B4 Part of a house a row of four stones under which fragments of two vessels were laid Burnt rubbles Most probably the house is related to the dwelling part in C4
Table 5.1.4 Copper Age features from tell Galabovo
Location Type of feature Content/description of the feature  
II horizon/ C4 in a house Pit (5-7 cm deep) with periphery coated by sherds (AFig. 5.1.3b) 700 snail shells and a complete bivalve, pottery fragments cover the soil above the snails; broken animal bones around the pit
II horizon Pit (5-7 cm deep) yellow soil mixed with charcoal, animal bones, big amount of sherds, a whole vessel and a bone hammer.
II horizon Pit (20 cm deep) with carefully clay-coated walls light grey soil mixed with charcoal, sherds, animal bones, two flint tools and a fragment of a cup.
III horizon Pit dug into the two previous horizons dark-grey earth mixed with charcoal, ash, animal bones and big amount of sherds (several vessels were possible to be reconstructed, none of which, however, was completely whole). A bottom and big fragments of the lower part of a pithos and a rim and fragments from the walls from another pithos. Next to the pithos bottom some big stones were discovered. To the east of the pit there were fragments of a big vessel
III horizon Pit – not fully excavated reddish clay soil, big amount of sherds and animal bones
Unknown BA horizon Pit dug into the three Chalcolithic horizons light grey soil, mixed with charcoal, a small amount of sherds and a flint spearhead.

The published Eneolithic pottery from Galabovo tell is typical for the Karanovo VI ceramic assemblage (Late Copper Age in Bulgaria) (AFig. 5.1.4-6). A characteristic range of Copper Age artefacts and raw materials is shown in AFig. 5.1.7.

In addition to the data from the tables, it should be mentioned that ECA pottery was found in one of the LCA pits. Although I was not able to reconstruct the context of discovery, the presence of Early Chalcolithic sherd (AFig. 5.1.7N) in a later pit is an important indicator for maintaining the past in the Copper Age settlement of the Galabovo tell.

Bronze Age

The Bronze Age occupation on Galabovo tell is believed to start after a period of abandonment characterised by a sterile soil (hiatus) between the Chalcolithic and Bronze Age layers. The Bronze Age occupation sequence is defined through stratigraphic observations in the trench profile of squares E4 and D4 (AFig. 5.1.1b, c). Nine building horizons were identified on the basis of successive beaten clay surfaces. The vertical stratigraphy of the tell showed the presence of pebbles in the first Bronze Age horizon and layers of ash and charcoal in the second and third building horizons. The last two were claimed to be successive occupational layers, during which this part of the tell was not built on. The area was occupied again during the fourth Bronze Age horizon, whose inhabitants levelled the region before building. The first, fifth, sixth, seventh and eighth horizons were additionally marked by dwelling floors and in horizons 5, 6 and 8 also by oven floors. The floor of the dwelling in horizon 6 is built immediately over the house rubble of the preceding horizon without the traditional clay levelling. The evidence for dwellings and other archaeological features from the 7th to the 13th BA horizons is summarised in Tables 5.1.5-5.1.6.

Table 5.1.5 Bronze Age dwellings from tell Galabovo
Horizon/square Type Structure Pithos Ovens/ hearths Inventory Traces of burning Comments
VII horizon/ E4 Part of a house (AFig. 5.1.1c) Massive burnt house rubble- 20 cm thick
VII horizon/ F4 Part of a house Dug into hearth Two (bull) horns, a horseshoe movable hearth, overlaying similar feature Under the first movable hearth a layer of ash and charcoal, among which big amount of river shells Probably part of the above described dwelling
VIII horizon/ F4 Part of a house A part of dug into Probably an oven N/A Burnt house rubble pithos’s base coated with clay mixed with straw
VIII horizon/ H4 Part of a house Two; one only with its base preserved oven A cup and a whorl in the pithos’s base Layer of ash and charcoal, overlaid by burnt house rubble
VIII horizon/ G4 Heavily destroyed house Sherds, animal bones, upper and lower grinding stones, around them spread Vicia ervilia Layer of ash and charcoal Dwelling floor of beaten clay
VIII horizon E4/E5 Heavily destroyed house Scatters of fragments of big vessels Layer of ash and charcoal
IX horizon F3/F4 Part of a house Eight postholes oven 10 loom weights, fragment of a big dish very close to the wall Spot of burnt rubbles over the oven E-W wall preserved at 2.30m
IX horizon/I4 Heavily destroyed house Pise- the wall is 50 cm long, 7.5 cm thick Pottery scatter, two middle size broken stones, one antler, one flint and nine stone tools; among the rubbles many big fragments of vessels Spot of ash and charcoal, spot of burnt house rubbles Beaten clay floor 3.30m N-S, 1.90m E-W; the antler and 8 stone tools were under the clay base of the wall, the other two above it
IX horizon/H4 Part of a house Layer of ash and charcoal above the clay, the soil contained pieces of burnt rubble/daub, sherds and animal bones a beaten clay level
IX horizon /N5 Part of a house scatter of sherds, small broken stones and animal bones; a whorl, a fragment of a whorl, a flint scraper, antler hoe Among the scatter a big fragment of a vessel has covered a tortoise carapace
IX horizon /O7 Part of a house oven plasters sherds and animal bones
IX or X horizon/ F4/E4 Part of a house Oven two heavily fragmented vessels in situ. Above the floor layer of ash and charcoal, overlaid by burnt wall rubble preserved floor 570 cm long and 245 cm wide
X horizon/ N5/N6(1) Heavily destroyed house (AFig. 5.1.10) rectangular; Northeast/ southwest orientation; 7 postholes big fragmented jug, an urn-like vessel, a loom weight 10cm layer of ash and charcoal cover the floor; spots of massive burnt rubble south wall preserved; between this and the next house 80 cm wide street was identified
X horizon/ N5/N6(2) Heavily destroyed house (AFig. 5.1.10) apsidal; Northeast/ southwest orientation; 8 postholes Two build pithoi, belonging to two occupational stages fragmented jug and two urn-like vessels; within the same area but in not very clear context a flint tool was found layer of ash and charcoal 60 cm to the west of one of the pithos north wall preserved 380 cm long , 25cm wide; wall base supposed to be coated by stones
X horizon/ F3/G3 Part of a house (Afig. 5.1.9a) Unknown Heavily destroyed base of an oven or a hearth under the rubble and weights; another base of a rectangular oven to the east of the weights 15 loom weights, some of them had lain under the fallen walls, others above them; around one of the ovens animal bones and very fragmented pottery; whole vessel, pottery scatter; in situ grinding stone (Afig. 5.1.9b) burnt house rubble The floor is immediately above the destruction of the preceding horizon even without the usual levelling; the depositional pattern of the weights made investigators conclude that more then one wall has collapsed
X horizon/ G3 Part of a house four postholes wall rubbles
X horizon /O5 Part of a house (AFig. 5.1.10) build pithos Rectangular oven preserved floor 340 cm long N-S and 200 cm wide E-W
X horizon/O7 Heavily destroyed house (AFig. 5.1.10) oven In the square a flint spear and a whorl were found
XI horizon/ J3-5/K3-5/L3-5 Part of a house (AFig. 5.1.11a) Northeast/Southwest orientation; eastern wall - 7 postholes; western and south each - 8 An oven, a second oven or a hearth, a feature with oxbow shape (oven or a hearth) Over 30 whole and fragmented vessels (AFig. 5.1.12), eleven of them form a depot (AFig. 5.1.11b); in the base of the west wall a big stone axe the soil was full of charcoal; burnt rubble investigated floor area 10 m to 5.75 m; the floor is directly upon rubble of a dwelling from the tenth building horizon without the usual clay levelling
XI horizon/ L7/K7 Part of a house 7 postholes Sherds and animal bones big pieces of burnt house rubble Preserved floor 5 m long,120cm wide
XI horizon/ J8 Part of a house 4 postholes Many broken stones; tools, vessels, a fragment of an altar derive from the same square thick layer of burnt house rubble
XI horizon/ O7 Part of a house Oven A whorl, a fragment of a whorl, small fragmented jug found in the square Under the floor an urn with infant (child under 1) burial. The body was in crouched position
XI horizon/ O5 Parts of two houses Two ovens, one in each house fragments of a vessel next to one of the ovens; a whorl, a fragment of a whorl, antler tool found in the square
XI horizon/ N6 Part of a house build pithos oven two whorls, a stone polisher, a dish
XI horizon/ M6 Heavily destroyed house house debris
XI horizon/ M4 Heavily destroyed house sherds Burnt house rubble
XI horizon/ K4/5 Part of a house oven or a hearth many sherds; middle size broken stones; 3 vessels in situ; big spots of burnt house rubble
XI horizon/ K5 Part of a house pithos oven or a hearth Scattered broken stones; pottery scatters burnt rubbles
XI horizon/ K6 Part of a house sherds and broken stones spread all over the square burnt rubble
XII horizon/ M4/5-L4/5 heavily destroyed house (AFig. 5.1.14a) N-S orientation three build pithoi -one whole and two fragmented oven and a hearth 10 loom weights; fragmented lower grinding stone in situ; 18 different finds around the pithoi and the grinding stone; three restored vessels with parts still missing; among the remaining sherds matching pieces but not restorable vessels (AFig. 5.1.14b – 5.1.15) Massive burnt rubble according to the field documentation the finds are 26, rather than 18
XII horizon/ J5/6-K5/6-L5/6 Almost complete house with two rooms (AFig. 5.1.16a) Northeast/ Southwest orientation; east wall 14 postholes, west – 11, south – six postholes big dug into pithos Two ovens – one round, the other with oxbow shape Two stone polishers, one fragmented stone tool, three whorls, four whole and 20 (AFig. 5.1.16b – 5.1.17A-Q) fragmented vessels; north of one of the ovens Bronze awl (AFig. 5.1.16G) claimed to be destroyed by fire Build immediately above the dwelling of 11 occupational level; preserved size 6 m to 8.5m.
XII horizon/ O7 Part of a house broken stones; clay walls big stone axe by the wall; pottery scatters; middle size broken stones; a dish, a cup, a whorl, stone hoe, fragment of stone tool layer of ash and charcoal, burnt rubbles eastern wall preserved 220 cm long and 20cm wide
XII horizon/ M7/N7 house with two rooms Rectangular, N-S orientation Scattered fragments of four pithoi an oven/hearth, probable movable hearth burnt broken stones; pottery scatters burnt rubble
XII horizon/ N6 heavily destroyed house Pithos Base of an oven, movable hearth Broken stones, animal bones, sherds, fragmented vessel in situ, two whorls, bone handle, clay ball, clay weight Ash and charcoal, spots of burnt rubble
XII horizon/ N5 heavily destroyed house hearth 4 whorls, an altar, 5 flint tools, a loom weight; a stone smoother
XII horizon/ O5 heavily destroyed house oven fragment of an altar; fragmented vessel Burnt house rubble
XII horizon/ M6 Part of a house destroyed by Medieval pit base of an oven pottery and stone scatters burnt house rubble
XIII horizon/ N7/8-M7/8 heavily destroyed house with two rooms (AFig. 5.1.18) probably rectangular and with North-South orientation build pithos hearth a cup and a dish under the floor in the south room; fragments of 14 vessels all together, among which fragment of a Trojan cup spots of heavily burnt fine clay soil, as a result of a long fire activity but with no connection to any heating feature
XIII horizon/ J6 heavily destroyed house fragments of big vessels, animal bones among the rubble; big broken stone layer of burnt rubble
XIII horizon/ Q6 Part of a house pottery scatter
XIII horizon/ M6 Part of a house an oven or a hearth with 2 loom weights in its base pottery scatter, stone scatter burnt house rubble next to the stone scatter
XIII horizon/ N5 Part of a house (AFig. 5.1.19a) small broken stones in a semicircle shape; two broken vessels; part of an upper grinding stone next to a stone polisher The stones are over a level of beaten clay
Table 5.1.6 Bronze Age features from tell Galabovo
Location Type of feature Content/description of the feature Discuss
VIII horizon /G4 in a dwelling Pit Ash, charcoal and a whole vessel 15 cm deep, dug into the dwelling floor
X horizon/ N5/N6 in dwelling (1) two shallow holes ash and charcoal
X horizon/ N5/N6 in dwelling (2) A pit and a hole
X horizon/ F3/G3 in a dwelling A pit reached and penetrated the destruction of the ninth building horizon
X horizon/ N5 few shallow holes grey/grey-blackish soil Depth up to 8 cm
X horizon/ N5 A pit, 55 cm deep, cut the previous two horizons grey/blackish soil, mixed with ash and charcoal; animal bones and sherds rubbish pit, related to the adjacent house in N5/N6
XI horizon/ O7 A pit soil darker than the surrounding soil, sherds and animal bones The pit has destroyed the oven of the lower dwelling
XI horizon/ O7 A hole soil darker than the surrounding soil, sherds and animal bones The hole has destroyed the previous horizon
XI horizon/ O7 A hole soil darker than the surrounding soil, sherds and animal bones The hole has destroyed the previous horizon
XI horizon/ O6 four ovens with traces of several reconstructions Burnt clay level, covered by layer of ash and charcoal, within which fragments of a vessel; similar spot of burnt floor, ash, charcoal and vessel fragments was also identified in the square; an antler hoe and antler awl derive from the same area no burnt rubble mentioned to present
XI horizon/ N5 heavily destroyed build pithos scatter of small broken stones and sherds around it not clear whether the feature is connected to the adjacent house in N6 or is an evidence of some outdoor activity
XI horizon/ L6/K6 feature of beaten clay Over it two fired spots of scattered sherds and bones possible relation to any dwelling/s in L6 and L7/K7 is not clear
XII horizon/ O7 round feature made of small and middle broken stones small amount of fragmented pottery and bones between the stones; a whorl, a fragment of a whorl, a cup, a fragment of a stone tool, a dish and a stone hoe found in the square, fragment of lower grinding stone Suggests some kind of outdoor activities
XII horizon/K5 spot of pottery scatter and a spot of burnt house rubble The two spots do not overlay each other
XII horizon/K4 Beaten clay level Few sherds, animal bones and bottom of a vessel on the clay
XII horizon/O6 Stone scatter Suggests some kind of outdoor activities related to houses in O7 or O5
XII horizon/O4 A base of oven/hearth Suggests some kind of outdoor activities related to houses in O7 or O5
XII horizon/J8 a spot of burnt house rubble animal bones, a dish, a piece of slag; a few sherds
XIII horizon/ O6 or O7 A pit ash, charcoal and sherds; imported pilgrim flask and bronze dagger (AFig. 5.1.19A, F) prior to its excavation the feature looked like a spot of burnt soil and ash and charcoal
XIII horizon/M8 spot of dark-brown soil mixed with burnt daub small stones and little charcoal and animal bones Among the sherds a rim and a neck of wheel made amphora, and a clay model of wheel

In addition to the information in Tables 5.1.5 – 5.1.6, a few more points should be made.

There are traces of burning visible in the profile of E4 (7th BA horizon) (AFig. 5.1.1c) while, in the description of the dwelling in the adjacent F4, no evidence for such activity is mentioned at all. If the stratigraphic correlation is correct, the features in E4/F4 (part of one or two houses) were treated differently in terms of ending the house(s)’ life cycles.

The dwelling in F4 (8th BA horizon) overlays the burnt house in E4 from the preceding horizon. Its pithos was dug into the rubble of a dwelling from the 7th BA horizon but whether there was a beaten clay levelling between the two horizons was not specified.

Given the present condition of the data, it is difficult to conclude whether or not the houses from the 8th BA horizon are four separate or fewer, larger houses. However, they all share a common feature - traces of fire. And, what is more important, they show traces of different kinds of use of fire (see below, p. 167). In the centre of the dwelling in M5 (12th BA horizon) (AFig. 5.1.14), the burnt rubble lay under fragments of big cooking vessels that made investigators suggest the presence of a two-storey building. There were sherds among the rubble as well. Under the burnt debris, there was no layer of ash and charcoal - another argument for a second floor, since the usual evidence of a burnt thatched roof should have been a layer of ash and charcoal. There were no traces of expected floor levelling as well. Instead there was a layer of soil, ash, stones and small pieces of daub.

The other almost complete burnt house from 12th BA horizon in squares J5/6-K5/6-L5/6 was built immediately above the dwelling in the 11th occupational level. The house has two rooms, as the party walls do not match in the middle, thus suggesting some kind of formalised access (AFig. 5.1.16). According to some members of the excavation team, the dwelling had two entrances – on the East and the South walls. The number of tools, whole and fragmented vessels made the investigators conclude that the house was deliberately emptied before the fire. However, I should dispute such a claim on the basis of the number and especially the type of the inventory found in the dwelling (Table 5.1.8).

The information available for squares K4 and J4 is very contradictory and it was not possible to come up with a final consistent description of the archaeological evidence. The data from these squares should not be omitted as they contain an important claim that was not supported by the field documentation. Squares K4 and J4 were accepted as an unburnt house in the context of massive burning of the remainder of the 12th BA occupational level. The dwelling had a North - South orientation and contained vessels in situ some of which whole, as well as some tools. According to the field documentation for K4, there was a part of dwelling floor and a bottom of a big vessel, while, in J4, no structures but some whole vessels were found, as well as a small spot of a clay that covered the burnt debris from the previous horizon. However, the inventory book contains information on vessels found on an unburnt dwelling floor in J4 that contradicts the claim for the lack of a house feature. The square was heavily destroyed by past and present intrusions, which perhaps caused the obvious confusion in the description of the data. To summarize, it is important that there were no traces of fire in these squares but whether or not the identified dwelling activities were contemporary with the 12th building horizon is difficult to justify given the present condition of the data. Even in one of the publications whose main topic is the chronology of the tell based on pottery typology, the dwelling in J4 that contains one of the discussed vessels- a teapot- is dated to the 12th-11th building horizon.

Some other features were identified as belonging to the 12th building horizon as well. In K5, a small pottery scatter and a spot of burnt house rubble were identified. The two spots do not overlay each other. The data from K5 is important since it shows that it is possible the evidence for fire is not in situ but rather an indication of a pattern of the re-use of burnt rubble. In the case of the 12th BA horizon, such a fact has special meaning since there is a house with traces of massive fire in J5/6-K5/6-L5/6 that borders on two areas (K4/J4) with no traces of fire at all.

The last feature to be mentioned from the twelfth occupational layer is a pit from M7. The data for this pit is very problematic as it has at least three different descriptions. In the first one, it contained burnt animal bones from a goat, pig, sheep and two cattle species. In the uppermost level of the pit, a gold hair spiral was found. In the second description, the pits contained a (human!?) cremation and a golden ring. And finally, in one of the Galabovo publications, this pit appears as belonging to the dwelling in M6/M7. It had almost vertical walls and its North wall goes under the hearth of the dwelling. The fill consisted of grey-black soil mixed with burnt organic material and a few sherds. At different levels of the fill of the pit, there were small burnt pieces of bones. Close to the base of the pit, there was a gold spiral ring.

In spite of these different descriptions of the fill, I would suggest that this is one and the same pit, whose content and interpretation has passed through several transformations from the field documentation to its publication. Undoubtedly, it is crucial to have precise information for a certain feature in order to present coherent discussion. However, for the pit from M7, it is not possible to reconstruct its initial content and context. The reason to take this pit into consideration here is that, even with such contradictory data, there is important evidence of associations. Features common to all the description are the bones (human or animal) and the gold (spiral ring or hair ornament). If the bones were human, that constitutes a burial within a tell. If the bones were animal, that questions the assumption that they were thrown away as rubbish, since gold is usually not connected with refuse activity. In both cases, these are important social practices that will be discussed in section 5.1.5.

The 13th BA occupational level is heavily destroyed by past and present human activities and there is very little evidence in situ. In O7, a level of beaten clay and disordered medium-size broken stones were discovered. In L7/J7, a pottery scatter, disordered medium-size broken stones and burnt house rubble were found. There was burnt house debris in P5/O5, together with a few, highly fragmented vessels. In P6, above a beaten level, there were four pottery scatters mixed with broken stones of different sizes. This evidence together with the data in Tables 5.1.5 - 5.1.6 make it likely that this is a destroyed horizon, which may contain burnt houses and secondary use of burnt daub.

The last (14th) occupational layer has very uncertain vertical and horizontal stratigraphy and was identified during the last working season in 1995. Above the dwelling debris in N7/M7 (AFig. 5.1.18) from what at that time was accepted to be the last occupation, another area of burnt rubble and traces of a hearth base were discovered. This fact made the excavators reconsider the existing stratigraphic sequence and a new building horizon (No 0 in the publication and field documentation) was added. On the floor of the newly identified house in N7 from the 14th building horizon, there was a pottery scatter, a hearth, part of a built-in pithos and fragments of other large vessels. In M7, a pit that reaches the 12th occupational layer but was dug from the level of 13th or 14th building horizon was found. It had step-line walls, coated with white clay and filled with grey-brown soil mixed with ash and sherds. The burnt house debris and few sherds and bones in M8 were assigned to the 14th occupation as well. On this basis, it was concluded that the final MBA settlement was killed by fire.

Burnt houses

For the purposes of this study, the archaeological evidence from Galabovo tell is summarised in Tables 5.1.7 - 5.1.10 with regard to the research issues outlined in chapter 3.2. Table 5.1.7 presents the evidence for traces of fire within each occupational layer. Such evidence is documented in two out of the three Copper Age building horizons and in 10 out of 14 Bronze Age building horizons (IX/X is not considered as separate horizon). This relatively high number raises serious doubts about the explanation of hostile invasions that may have caused the fires. Such a claim has not been made for the Galabovo tell in particular but this is one of the explanations given for burnt houses/horizons in Bulgarian prehistory (e.g. Yunatsite tell: cf. Chapman 1999). The other possible explanation is accidental fire that was not discussed but simply taken for granted in one case in Galabovo – the house in J5/6-K5/6-L5/6 (XII BA horizon).

In order to evaluate the nature of the Galabovo burnt horizons, the evidence for burning was summarised in Table 5.1.7, where in the first column are the squares with no traces of fire mentioned at all (the information derives mainly from the field documentation). In the second column are the squares/features that have some traces of fire but where the in situ situation cannot be clarified (e.g., because of strong past or present destruction) and the burnt remains are not closed complexes (e.g. non-overlaying spots of rubble and pottery scatters). Some of the squares contain fire products as ash, charcoal, burnt clay floors but not burnt house debris. On this basis, I should assume that these squares present evidence for control over fire or fire products. The scattered burnt rubble in some squares (e.g. K5, J8 XII horizon, etc.) and in the pithos from the XI occupational level suggests secondary use of daub from contemporary or earlier houses. Claims for the controlled use of fire are additionally supported by the data from the houses in F3/G3 from the X building horizon and in M7/N7 from the XII building horizon, where parts of the houses in F3 and M7 contain traces of fire in situ (floor overlaid by layer of ash and charcoal and finally sealed by burnt rubble), while parts of other houses, respectively in G3 and N7, have some traces of fire but not massive burnt daub wall discards. It is possible that the rubble was destroyed by later intrusions. However, that should result in the fragmentation and consequent spread of burnt debris rather than in the disappearance of burnt daub. The last column summarises the evidence for intensive fire documented by the above-mentioned in situ sequence. In most cases, these concern almost entire houses (e.g. in M4/5 – XII building horizon) but there are parts of houses as well (e.g. O7- XII building horizon).

Table 5.1.7 Spatial distribution of evidence for fire by level in Tell Galabovo
Burnt horizons Squares/features with no traces of fire Squares/features with some traces of fire Squares/features with traces of massive fire
Copper Age I     C4/B4
III B4   C4
Bronze Age II   D4  
III   D4  
VII   F4 E4
VIII   G4, F4 H4
IX O7, N5 H4 I4, F4
IX/X     F4/E4
X O5, N5 O7 N5/N6, F3/G3,
XI N5, N6, O7, M6 O5, O6, K6/L6, L7/K7, M4 J4/5-K4/5-L4/5, L6, J8
XII N5, K4, J4 N6, O5, J8 M4/M5, J5/6-K5/6-L5/6, O7, M7/N7,K5
XIII N5, Q6 M7/8-N7/8, M6 J6
XIV   M7/8-N7-8  

Table 5.1.7 shows a trend towards the diversification of use of fire with the increase of the excavated area. Whether such a trend reflects the situation in which management of fire has started from the time of IX building horizon or such practice of deliberate and controlled fire was known since the Chalcolithic occupation is not possible to establish according to present data. It is also not possible to reconstruct the exact process of intensive or less devastating fire due to the paucity of consistent details of the sequence of burnt remains in the publications and field documentation. However, some activities that most probably have taken place before the actual fire, such as emptying the house of artefacts or, conversely, the deposition of objects as a house set could be explored on the basis of the summary in Table 5.1.8

Table 5.1.8 Inventories of burnt houses, Galabovo tell
Horizon House inventory
Copper age I B4/C4 – a grinding stone, 2 whole and 4 fragmented vessels, 11 flint tools (excluding the flints in one of the vessels), 5 stone tools, 2 fragments of stone tools, fragment of a whorl, a bone pendant
III C4 – 9 flint tools, 2 vessels, a clay spoon, 2 stone tools, an antler tool, a horn tool, a piece of copper slag, a bone figurine, fragment of a bone figurine, fragment of a whorl, fragment of a loom weight
Bronze Age VIII H4 – 4 vessels, 2 whorls, a lid
IX

I4 – 2 whorls, many fragmented vessels (number not known)

F4 – fragments of a big dish, 10 loom weights

IX-X F4/E4 – 2 heavily fragmented vessels
X

N5/6 – house 1- 2 vessels, a loom weight

N5/6 – house 2 – no finds

F3/G3- 10 vessels, a clay spade, a whorl, 10 loom weights, 2 grinding stones

XI

J4/5-K4/5-L4/5 – over 30 vessels, a fragment of a stone adze

L6- fragment of a dish, 3 whorls, a horn tool, a stone adze

J8 – an altar, 2 flint tools, a bone awl, a lead

XII

M4/5 – a grinding stone, 18 finding (unspecified bone, stone and pottery tools), whole and fragmented vessels (number not known)

J5/6-K5/6-L5/6 – two stone tools, 3 whorls, a bronze awl, fragment of a stone tool, at least 4 whole and 20 fragmented vessels

O7- 2 vessels, a stone hoe, a fragment of a stone tool, fragment of a whorl, fragmented vessels

M7/N7 – 6 whole and some fragmented vessels

The table contains information on the type and number of objects found in the houses under burnt rubble. Although it is important to relate the size of the excavated structure to the number and variety of the discovered objects, in this particular case it is problematic. There are parts of houses that are empty (N5/6 – house 2 - X building horizon) and parts of houses that contain over 20 objects (C4 – III Chalcolithic building horizon). Bearing in mind that some of objects may well have been located in the destroyed/ unexcavated parts of the house, I should suggest that the patterns documented in table 5.1.2 are to some extent conditional. According to the present data, it is evident that there were no massive fires which burned an entire settlement horizon. Accidental fires are not to be excluded but the presence of some objects in the burnt houses (e.g. the bronze awl in J5/6-K5/6-L5/6 – XIII building horizon) that could easily have been removed in a dangerous situation suggests that the inventory of the burnt houses was not chance occurrences. Therefore, it is very likely that the fire was not accidental but, rather, that the burning of houses was an intentional social practice. Looking at the burnt house inventories, two patterns are apparent:- houses that are empty(0-10 objects all together) and houses that contain, if not a full, a fairly sufficient set of household objects. I should assume that these two patterns serve different purposes within one and the same social practice of successful reproduction. Brück has argued (1999) that the life-cycle of a house is related to the life-cycle of its inhabitants. Renovation of the floor, building/ digging a pit for a new pithos or for other purposes may well have denoted an important event in a household or community life in addition to any practical benefits. Deliberate burning of houses seems irrational only from our modern concept of a house. If a house is a way of mediating the lives of the inhabitants with the world as a whole, as well as a place to live in, burning the dwelling could mark the entry of a new member into the household, which should be followed by a) negotiating his/her social status, and b) enlarging the living space of the house. Other socio-economic events (birth, death, a good harvest or a successful long-distance journey) may be memorized, celebrated or disputed on the constrained tell area through the act of burning, using the burnt materials and arranging the pre-fire house inventory. Unfortunately, the present condition of Galabovo data does not allow such a complex approach for investigation of the possible social events and their material expression in archaeological evidence.

The only case in which a possible symbolic burial (perhaps of an important member of the community) can be assumed is the burnt house from the last Chalcolithic occupation. It contains a bone figurine with traces of red ochre, as well as some sherds with red ochre. Red ochre is always found in a burial context in Maritsa Iztok, in graves usually connected with the Pit Grave culture. The earliest burial in Maritsa Iztok (see p. 250- 254 for details) is the primary grave of Gonova mogila, dated to the end of the IV mil. BC (Kunchev 1991). The deceased was covered by red ochre and grave goods consisted of an obsidian blade, copper beads and shells. Apart from red ochre, two more objects from the burnt house inventory are reminiscent of the grave set of the earliest grave in Maritsa Iztok - a bone imitation of a flint superblade, deposited in a pithos (AFig. 5.1.7H) and a piece of copper slag. These general similarities between the grave set and the burnt house inventory give grounds for interpreting the house inventory as a symbolic burial, with the bone figurine as the dead persona.

Structured deposition

Structured deposition features (mainly pits) were excavated in almost every building horizon. The content of the pits varies enormously but, in general, all of them contain soil, ash, charcoal, sherds and animal bones. Foundation deposits, as well as some unusual findings (e.g. a bead in an imported Syrian bottle) were also discovered. The evidence for structured deposition and foundation deposits is summarised in Table 5.1.9.

In addition to the data from Table 5.1.9, there are four more pits which cut through several earlier levels:– a pit that has destroyed the three Chalcolithic layers; a pit penetrating the X and IX building horizons; a hole that has broken the pithos in house N7/M7 from the 13th occupational level; and finally a Medieval pit that has destroyed at least three upper BA layers. Therefore, it may be concluded that pit digging was not a rare practice on the Galabovo tell. In the reports, the function of pits was considered as unknown or, in one case, as a rubbish dump. I should rather suggest that pit digging was a purposeful process that has no unified explanation. It was argued that pits on tell are connected with a targeted interrelation with the ancestors (Chapman 2000c). This may well be the case with the pits in square B4 from the last Chalcolithic horizon, that contain parts of two different pithoi and vessels with missing parts (see p. 171). The evidence from O7 in the 11th building horizon, with three pits and an infant burial, could be interpreted as some kind of deliberate depositional pattern, in which the symbolic relation between the ancestors, the newly dead and the living was crucial.

Table 5.1.9 Evidence for structured deposition and foundation deposits
Horizons Structured deposition Foundation deposits
Copper Age I A pot with flint tools and flakes  
II 3 pits  
III 2 pits, fragments of two vessels under row of stones, pithos filled with soil, charcoal, sherds and two bone tools  
Bronze Age VIII A pit, base of a pithos containing a cup and a whorl A pit beneath the floor that contains a whole vessel
IX   2 whole and 6 fragmented stone tools, a fragment of a antler tool and 1 flint and one stone tools
X 2 pits and several shallow holes  
XI Child burial, 3 pits, a vessel with a bead, a vessel in a tortoise shell, pithos with big pieces of daub A stone axe
XII A pit and a few shallow holes A stone axe
XIII A pit A cup and a dish
XIV A pit (not sure stratigraphic data – either XIII or XIV horizon)  

Such an intentional depositional practice may have taken place during the next occupation in square M7, if the bones in the pit were human. If the bones were not human, I should suggest that this was a symbolic burial as spiral gold pendants are found in three barrow burials in Maritsa Iztok, while so far gold objects have not been excavated within any other settlement context. Deposition of animal bones in pits is fairly common on the Galabovo tell. It may be connected with feasts or fertility cult or the memorial rites of feeding the ancestors. I would suggest that the gold spiral pendant was deliberately put to emphasise the burial element of digging the pit. In a moment of social tension when there was no dead human body, (burnt) animal bones together with the gold pendant recalling/evoking real burials may well have framed the stage for the re-negotiation of important social issues.

Alternatives to the ancestral hypothesis of pit-digging are also likely. Burying over 700 snail shells may have been a memorialisation of a communal feast, when the inhabitants from the second Chalcolithic horizon re-settled the area above the burnt first settlement. The careful construction of the pit itself and the position of the shells - ordered, not simply thrown - reveal some act of deliberate deposition rather than rubbish dumping. The distribution of animal bones suggests that the place may have become a place for recurring feasts or memorials.

Fragmentation

Fragmented objects are widely spread all over the 17 occupational layers in a variety of contexts – on the floor, in pits, under ovens, etc. Pottery sherds are the most numerous findings at Galabovo tell. The possibility of accidental and/or breakage has already been discussed (Chapman 2000) and the emphasis here is on the post-breaking treatment of the pottery. In several cases (e.g. XII BA horizon – dwelling in M4/5-L4/5) some vessels from the pottery scatters could be restored, while others could not. An attempt to find the missing parts in other places/features within the tell has not been made, apart from the targeted search for missing fragments of imported vessels and large pithoi. More evident deliberate fragmentation practice is to be found in structured deposition features. The base and the lower part of a pithos and a rim and walls from another pithos were deposited in a pit from the third Chalcolithic building horizon. I should assume that the lower and the upper part of two different vessels were intended to evoke the image of a new entity that interrelates previously separated objects and in the same time links them to the ancestors, as they are deposited in a pit that penetrates the two horizons. The pit also contains some vessels with missing parts. I should suggest that the content of the pit reveals an act of lineage/household enchainment (the pithoi fragments) and personal enchainment (the vessels with missing parts), memorialised through burial in ancestral soil.

There are some groups of objects that support the idea of deliberate fragmentation . These are objects that are hard to break accidentally and useless after breakage. Their type and distribution is shown in Table 5.1.10.

Table 5.1.10 Fragmented objects from Tell Galabovo
Horizons Whorls Loom weights Stone tools Bone/horn objects Net weights others
Copper Age I 1   2      
II     2 1   1 lid
III 1 1   1   2 altar
Bronze Age VIII 1     2    
IX 1          
X 1 1     1 Clay mould for awl
XI 4   1 1 1  
XII 2 4 5 1 3 1 lid
XIII 5   3 1 1 2 - clay models of a wheel,1 lid
XIV       3 1 1 spoon
Unstratified 2 4 8 7   1 altar, 1 figurine

The last archaeological evidence to be discussed, concern the notion of the vertical continuity of features as an indicator of diachronic continuity in social relations (Bailey 1990). The data from Galabovo is summarized in Table 5.1

Table 5.1.11 Evidence for feature continuity and discontinuity
BA Horizon Square Type of feature
VI over V    
VIII over VII E4/F4 Pithos dug into destruction of the previous dwelling
X over IX F3/4 Floor overlaying destruction of the previous dwelling, without the usual clay levelling but there is change either of the direction or the plan of the house, since the two floors do not match
X N5/6 Destroyed pithos, the new one moved
XI over X K3/6-J3/6 Floor overlaying destruction of the previous dwelling without the usual clay levelling
XII over XI J5/6-K5/6-L5/6 Floor overlaying destruction of the previous dwelling without the usual clay levelling
XII over XI O7 An oven moved 80cm to the south of the preceding
XIII over XII M7-N7 Overlaying ovens/hearths

As Table 5.1.11 shows, the pattern of direct overlaying of features is not consistent. It is possible that those dwellings which are built immediately above the preceding belong to members of the society whose presence on the tell required material reinforcement. However, given the present state of the data, conclusive claims cannot be made.

5.1.2 Plant and animal remains

Several thousands animal bones were found during the first three seasons of the Galabovo excavations2. Among them, 5,033 could be assigned to species level.

The three Eneolithic layers contained 1,289 identifiable pieces. The wild mammals were represented by 47.7% (208 bones) of the mammal bones, while the domestic animals were 52.3%. Another 755 individuals (among which 741 snail shells) complete the list of the wild species. These general figures led to the conclusion that hunting was a major subsistence practice during the Chalcolithic. The main prey species were red deer (26.7%), wild boar (24.4%), roe deer (15.6%), aurochs (11.1%) and fallow deer (8.9%). The domestic sample shows the following species distribution - cattle (31.7%), caprovines (33.3%), pigs (18.3%) and dogs (16.7%). More than half of the cattle were killed at an adult age (52.6%) - that was interpreted as an indicator for cattle traction – ploughs and carts - and as a milk source. The slaughter pattern of 31.6% of the cattle at a juvenile age and 15.8% at infant age suggests that beef consumption was also practised. The main meat source, however, was the pig. Only 9.1% of this species reached adult age; 27.3% were killed at infant age and 63.6% at juvenile age. A source of milk, wool and hair and also meat was the flock of caprines, as 65% of al the individuals in the sample reached adult age, 30% were killed at juvenile age and only 5% at infant age. The high percentage of dogs suggested to the investigator an important role in protecting both house and herd. But, on the basis of some evidence for roasting, it was suggested that dogs were also used for eating or for some ritual purposes.

The Bronze Age sample contains 3,797 bones, of which 2876 are from domestic species (76.2%) and 921 (23.8%) are bones from wild animals. The former were cattle, pig, sheep, goat, horse and dog. The latter were 25 different species, among which red deer (34.9%), wild boar (24.6%), roe deer (16.6%), aurochs (10.9%), fox (4%) and fallow deer (1.7%). The cattle were the commonest species among the domestic animals (28.3%), followed by pigs, sheep and, lastly, goats. At least two breeds of dogs were bred and, as in the Eneolithic, traces of roasting dog meat were found. There is a relatively high percentage of horse bones – 4.1% at the beginning of the Bronze Age (1-7 building horizons), 1.3% in the middle horizons (8-10) and 2% during the MBA (11-13 building horizons). Two bone cheek-pieces were found in I3 (XI BA horizon), which completes the evidence for horse domestication at tell Galabovo.

The age at death of the cattle shows 73% adults, 21.6% juveniles and just 5% infants, on which basis it was concluded that the species was mainly used for traction, for milk and skin and to lesser extent for meat. Again the main meat source was the pig – 43.6% were killed at infant age, 46.1% at juvenile and just 10.3% reached adult age. The ratio of sheep/goat flock was 83.3%/16.7%. The age at death of both species was similar – 65.7% of the sheep and 71.4% of the goats reached adult age. So it was concluded that sheep and goats were mainly used for milk, wool and hair and less for consumption.

The development of domestic species through time shows some interesting trends. The frequency of cattle is gradually decreasing with time – 34.7% in the first horizons (1-7), 31.6% in horizons 8 to 10 and 28.1% in the last BA horizons (11-13), while, at the same time, the incidence of pigs is increasing, respectively 16.3%, 25% and 25.7%. The number of sheep, goats and dogs is relatively stable with little fluctuation through time.

The general conclusion of the faunal analysis was for an increase in agricultural activities during the BA as the number of domestic individuals has increased in comparison with the Copper Age. Some climatic changes have also been assumed. The presence of fallow deer and a specific snail species (Cepaea vindobonensis) during the Chalcolithic were indicators for a somewhat drier and warmer environment, in comparison with the BA, with a falling percentage of fallow deer and an increased number of shade-tolerant garden snails (Helix pomatia) (Ribarov n.d.).

The distribution of the number of bones by horizon has not been presented, which prevents estimation of possible flock size. Therefore, it was not possible to calculate the range of necessary resources for pasture.

Palaeo-ethnobotanical investigations have been undertaken for both the Chalcolithic (36 samples) and Bronze Age (36 samples) layers, as well as for the hiatus (1 sample)3. The samples were extracted by flotation from different archaeological features. Seventeen samples did not contain any plant remains, including the sample from the buried soil between the Copper Age and Bronze Age occupations (the so-called hiatus).

Copper Age

Samples were taken from a pit and from vessels in the house in B4. There were no plant remains in the pit, while the soil from the dwelling contained grains of Triticum monococcum, Triticum dicoccum, Lens culinaris, Vicia ervilia and Hordeum vulgare. The same species were present in the sample taken from a vessel in C4. Imprints from burnt daub (floors and walls) have also been analysed. Out of 192 studied fragments, 129 had traces of grain impressions. They contained the same suite of cereals as in the flotation data (Table 5.1.12).

Table 5.1.12 Plant evidence Tell Galabovo
Species Chalcolithic grains/flotation Chalcolithic grains/impressions Bronze Age
T. monococcum + + +
T. dicoccum + + +
T. spelta + + +
T. compactum + + +
Hordeum sp. + + -
Hordeum vulgare vulgare + + +
Hordeum vulgare nudum + + +
Vicia ervilia + - +
Lens culinaris + - +
Cornus mas - - +
Ficus carica - - +
Rumex asetosa + - -
Rumex acetosella + - -
Bromus secalinus - - +
Chenopodium album - - +
Polygonum lapatifolium - - +
Argostema githago - - +
Secale cereale - + -
Lathyrus p. - + -
Carpinus betulus - + -

Bronze age

There are only two samples from the layer that immediately follows the hiatus and they contain single grains of T. dicoccum and V. ervilia. The next samples are from the tenth building horizon and consisted of Triticum monococcum, Triticum dicoccum, Lens culinaris and Vicia ervilia. Some contextual information is available for the plant remains from this layer. In the house in square F4, the soil around the group of loom weights contained grains of einkorn, emmer, vetch and lentils. More interesting is the find of 500 g of vetch spread around the grinding stone in square G4. This evidence raises the question of vetch preparation and consumption. It is generally accepted that soaking of vetch to remove toxic elements us necessary before its use like other legumes. The Galabovo find suggests that it may also have been ground like cereal grains.

Samples from the next building horizon (11th) are poor in general, with single grains of wheat, barley, lentils and vetch. Interesting finds are a charred fruit of fig and a cornelian cherry stone. The twelfth occupational level has the same distribution of main cereal and legume species but provides some more contextual data. Within a pottery scatter in K4, single grains of barley and more than 100 grains of vetch were found. Around the oven in square K6, a large quantity of lentils mixed with vetch was recovered. About 50 g of vetch was found in the soil of an amphora-like vessel in the dwelling in M5. A similar amount of lentils was extracted from the profiles in L5/6.

The weeds presented in BA Galabovo are typical mainly for spring-sown crops. Rumex acetosella, Bromus secalinus, Chenopodium album, Polygonum lapatifolium and Argostema githago are usually spread across meadows and fallow lands and some of them are suitable food for both humans and animals. Rumex acetosella indicates a dry acid soil, more common in winter cereal fields than in summer cereal fields, also typical for fallow lands and dry pastures. The seeds are surely not edible, but the leaves are nutrient-rich and vitamin-rich and can be eaten as sorrel sauce or soup. Bromus secalinus is an indicator of dry meadows, cereal fields, suitable as animal fodder (usually grazed by cows and sheep) and could be used for human famine food. Chenopodium album is an almost ubiquitous weed but especially in cereal fields and on trampled ground (by tracks). It is definitely edible, since the seeds were often harvested and cooked, producing a mush. Polygonum lapatifolium is a typical component of wet weed communities but is also natural in channel beds. It was probably not used for human consumption. Agrostemma githago is found in both dry and wet cereal fields but is surely not edible (Eniko Magyari pers. comm).

Samples of carbonised wood have also been analysed from the Galabovo tell. The fifteen samples contain 327 fragments that showed a dominant presence of oak, and less hornbeam, maple and hazel. There is also a very high percentage of coniferous species, especially in squares N7 and J8 of the last MBA layer. The wood had suffered some specific deformation, that could be a result either of high pressure or of a very old age. There were two types of such torsion:- a) fragments with typical traces of remaining in water; and b) fragments with a very hard shiny surface. The wood taxa ware juniper, fir-tree and cypress, that are not typical for the region now and during the BA as well. This made the investigator conclude that most probably these pieces of wood derived from some kind of coal seams. Surface coal seams are often met in Maritsa Iztok and it was suggested that they were already used by prehistoric communities.

Plant remains at Galabovo tell are typical for the agricultural societies of the Balkans. Most of the grains were found in contexts of food storage (e.g., pots) or food preparation (e.g., ovens). The charcoal remains found at the tell indicate tree species which may have been used for both building and fuel. An additional fuel supply is possible from the abundant surface coal. The degree of human impact on the natural vegetation (deforestation and cultivation) is not yet possible to reconstruct.

5.1.3 The site and its surroundings according to the GIS analysis

The tell of Galabovo is located in the Westernmost part of the Maritsa Iztok study area. It is on a 1-2° gradient terrace of the river Sazliika (CDFig.1) but the initial distance from the river is not possible to reconstruct. At present, almost no elements of the original natural environment have been left intact. The maps used for the terrain reconstruction are from the early 1970s, when most of the tell surroundings were flat or with a 1-2O slope. About 450-800 m to the North East, there are terraces that reach up to a 4-5° gradient. The tell has a Western aspect (CDFig.2) and relatively low visibility. No archaeological sites are visible from the tell, apart from the (much earlier) Neolithic settlement near the village of Obrutchishte (CDFig.3). Since the latter has an uncertain location (see section 5.2.2), intervisibity between Galabovo and Obrutchishte is possible but not sure. The publications of the Galabovo tell claim an altitude of 111 masl, while the maps used in the current study show 106 masl (CDFig.4). A second visibility analyses was undertaken with an additional 8 m tell height. In comparison with the previous viewshed, only the immediate vicinity had better visibility, yet with no evidence for any intervisibilily with other sites (CDFig.5). In both viewshed analyses, the maximum visible area from the tell was 3km. There are some visible spots 5 km to the South East towards Obrutchishte and 10 km to the North East, as well as towards the site of Mednikarovo.

The cost distance analysis from Galabovo was made on the basis of slope. The result was a set of 10 zones differentiated according to the accumulated cost needed to reach any point within the landscape from the tell (CDFig.6).

The distribution of sites within the 10 cost strips is summarised in Table 5.1.13.

Table 5.1.13 Site distribution around Galabovo tell
N of cost strip Sites located in the cost strip
2 Obrutchishte flat site, Iskritsa dwelling site, Iskritsa pit site, Atanasivanova mogila, Tcherniova mogila – all locations, Goliama Detelina flat site,
3 MIBC, Goliamata, Malkata and Ovchartsi barrows, Taniokoleva mogila 1, Manchova mogila, Klisselika, Gudgova and Mednikarovo tells
4 Taniokoleva mogila 2-4, KMBC, Kurdova mogila, Barrow4
5 Aldinova, Polski Gradets tell
6 Ovcharitsa I and II, Gonova barrow, Polski Gradets pit site

In summary, when Galabovo was founded, it was relatively far from possibly contemporary Eneolithic sites. The same pattern was observed during the Bronze Age re-occupation but, at this time, there was greater site diversity, as at least 10 barrows were located in the second and third cost strips.

On the basis of cost distance maps, least cost paths were derived between Galabovo and all the remaining archaeological sites (CDFig.7). There are three major paths that start from the tell. One of them follows the river Sokolitsa valley in the Southern part of the study area, the other follows the river Ovcharitsa valley in the Northern part of the region. Between them is the third route that follows the hilltops to the Mednikarovo/Iskritsa barrow cemetery (MIBC). From the two main routes along the valleys, there are separate tracks to each of the sites (e.g. the route form Galabovo to another BA site – Gonova mogila - has several branches that lead to Ovcharitsa I and II, to Polski Gradets tell, etc).

The main South route from Galabovo connects the tell with Gudgova tell (CDFig.8). The first segment of the path connects Galabovo with Atanasivanova mogila (CDFig.9), followed by the two Iskritsa sites (CDFig.10), (CDFig.11), finally reaching the Klisselika (CDFig.12) and Gudgova tells. The path gradually ascends the Sokolitsa valley, following the gentlest slopes. There are two branches from the path – to the Obrutchishte flat site and Mednikarovo tell. The first one bifurcates from the main route roughly 4.5 km from Galabovo, ascending South East for another 1.5 km to the Neolithic flat site. The second one forks to the right 8 km from Galabovo, ascending for 3 km to the South East until it reaches Mednikarovo tell (CDFig.14).

The general visibility from the South route is 2 km to the North and 1-2 km to the South along the Sokolitsa valley (CDFig.13). There are also scattered views over the hilly areas South of the flood-plain. All sites in the valley (Atanasivanova mogila, both Iskritsa sites, Klisselika tell) are on the main route and share similar views – no visibility North and South of the valley and intervisibility between the sites from the flood-plain. It should be mentioned, however, that the distance between the sites varies from 100 m to 5 km. On a clear winter day, the sight link between the most remote sites is possible but this is less likely in full-leaf spring, summer and autumn with natural woodland vegetation. Moving Eastwards from site to site and adding new segments to the main route, the visibility is increasing mainly in a Southeasterly direction to the mountain foothills.

The main North route connects Galabovo tell with Gonova mogila. In contrast to the South route, none of the sites is located on the path. There is a separate track to each of the 15 sites in the North and North East parts of the study area (CDFig.15).

The general visibility of the North path is mainly over the flood-plain of the Ovcharitsa River and not that wide as the visibility along the South route (CDFig.16). The low hills South East of the path are visible or not depending on the different branches that lead to a particular site. In that sense, the best visibility over the hills is provided while walking along the path from Galabovo to the Polski Gradets tell. Moving North East towards the most remote sites, the visibility increases in length but not much in width, remaining mainly within the limits of the valley and occasionally over the hills. The Southern part of the study region – the Sokolitsa valley and the Sakar foothills - are hardly visible at all, apart from some small spots to the South West at the very edge of the coverage.

The last route - to MIBC – crossed the areas affected by the mines and its tracks could not be clearly established (CDFig.17). Due to its uncertainty, the visibility from this path is not taken in consideration.

Since the tracks of the two main routes match the tracks derived from the cost surface of the two destination points – Gonova mogila to the North East and Gudgova mogila to the South East - the visibility from the routes matches as well. However, the change of direction of movement imposes a different sequence of views until the final panorama is achieved. These sequences are discussed in the Gonova mogila and Gudgova tell case studies.

Due to word limit, the landscape visibility from paths to the sites situated on these main routes is not discussed in the relevant case study, since such paths and their visibility will be discussed later (e.g. Atanasivanova mogila is between Galabovo and Gudgova tells and the paths and their visibility from/to the barrow and from/to the tells will be discussed in the case studies of the other two sites). Site visibility, however, from the segments that form the main routes is taken into consideration and discussed in each case study.

Visibility from paths

The visibility from the first segment of the South route – to Atanasivanova mogila (for Obrutchishte see p. 189.) - is very good over the Sokolitsa valley up to the point of the Iskritsa site and subsequently towards the North part of the valley (CDFig.18). The sites visible from that path are Obrutchishte, the Iskritsa pit site, Klisselika and Gudgova tells. In published reports, the site of Iskritsa is implicitly accepted as having two phases, hence being, in fact, two sites – the pit site and the dwelling site. In my opinion, such a division is not very well supported and I should argue that the site was only one but with a long horizontal occupational sequence (details in section 5.6). In the current GIS coverage, the site is presented by two points (two different parts of the same site) and hence the positive visibility for one of them and the invisibility for the other. Assuming that Iskritsa was a single site, it means that some parts of it were visible, while others were not.

Despite the distance – over 5.5 km - between the last point of the path and the two tells in the Southeasternmost part of the study area, it is important to point out the possibility for visual contact with earlier and contemporary sites, while walking between Galabovo tell and Atanasivanova mogila.

The paths to both Iskritsa sites and their visibility are very similar and will be discussed together. Both paths share visibility over the same sites (Atanasivanova mogila, Klisselika and Gudgova tells) and landscape (very good over the valley, and East of the Iskritsa sites over the Northern part of the valley) (CDFig.19). There is only one, very important difference between the two viewsheds. From the path to the Iskritsa pit site, the Iskritsa dwelling site is not visible, although the sites are 500m apart. This confirms the particular visual status of the dwelling site, which appears to be restricted from the West. The path to the dwelling site (CDFig.11) passes by the pit site, which assures a view over the Iskritsa pit site (CDFig.20).

The paths to Klisselika (CDFig.21) and Gudgova tells (CDFig.13) share similar visibilities but the views from the latter are denser towards the Easternmost parts of the study area. The general visibility is very good over the valley and with scattered spots towards the areas South of the valley. All the sites in the valley are visible from the paths.

The first segment of the North route is the path Galabovo – Tcherniova mogila (CDFig.22). The path follows the lowest area in the study region along the valley of the Ovcharitsa and has very good visibility over the area through which it passes (CDFig.23). There are single scattered spots and strips visible South East of the path towards the hilly central area of the study region. Six barrows, Goliama Detelina site and Polski Gradets tell are visible from the path.

Roughly 1.6 km before Tcherniova mogila, the path bifurcates, heading South East for 3 km when it reaches Kurdova mogila (CDFig.24). The path shares the same visibility over the valley as in the Tcherniova mogila case but it is better over the central and Eastern parts of the study area (CDFig.25). Again, six barrows (4 with certain location and two with uncertain location) and Polski Gradets tell are visible from the path.

The next segment of the North route connects Galabovo with Manchova mogila (CDFig.26). The path turns right from the main route 13.6 km from Galabovo, ascending South South East for 1 km. It has a good overview of the valley up to the point of Goliamata and Malkata barrows (CDFig.27). It also has consistently visible areas in the Northeasternmost, central and Eastern part of the study area, as well as three tiny visible strips to the South of the path itself. Six barrows (surely five and one (Taniokoleva) with only one of its possible locations) and the Goliama Detelina flat site are visible from this path.

Following the same main North route 800m before Manchova mogila, the path bifurcates, reaching the EBA settlement of Goliama Detelina after 300m (CDFig.28). The visibility from that path is generally the same as from the path to Manchova mogila but it has less consistent and dense views over the North Eastern, Eastern and central part of the study area (CDFig.29). Seven barrows are visible from the path.

The branch that turns right from the main route and from which starts the segments to Goliama Detelina and Manchova mogila is the path to Taniokoleva mogila (CDFig.30). The route ascends for 2 km to the South East and shares the general landscape visibility of the two previously discussed paths (CDFig.31). The same high site visibility is characteristics for this path as well - six barrows and the EBA Goliama Detelina site.

The next two sites – Goliamata (CDFig.32) and Malkata mogila (CDFig.33) - have a common branch from the main route 16 km after Galabovo tell. It ascends for 700m to the South East, when it forks and, after 300m to the South, each segment reaches its respective barrow. These almost identical tracks result in very similar path viewsheds (CDFig.34). The path to Goliamata mogila has a slightly more consistent general view (CDFig.35). In addition to the previous paths’ good view over the valley, the more Easterly position of the two barrows assures visibility over the relatively low-lying parts to the North East and the Eastern hilly areas of the study area. There is also a single visible strip South East of the path towards the central part of the region. One and the same sites are visible from both paths - seven barrows (surely six and one (Taniokoleva) with only one of its possible locations), the Goliama Detelina flat site and Polski Gradets tell.

Continuing along the North route, the next fork is for the path towards Barrow IV, accepted as belonging to the same barrow cemetery as Goliamata and Malkata mogila, although it is more than 1 km to the South East (CDFig.36). The path ascends to the South South East for 2 km and has good visibility along the valley up to the point of Ovchartsi barrow and scattered visible spots to the South East, East and North East (CDFig.37). Seven barrows (surely five, one with 3 out its 4 possible location, and one (Taniokoleva) with only one of its possible locations) and the Goliama Detelina flat site could be seen from the path. It is possible that Aldinova mogila and the Ovcharitsa II enclosure in particular have also been visible from the path, since before their migration to the centre of the cell, they were very close to the edge of a visible/invisible cell.

The path to the next barrow of Ovchartsi is one of the shortest branches in the North part of the study region (CDFig.38). It is 700 m long to theEast South East. It shares the visibility of the previously discussed paths and in addition there is good visibility towards the hilly areas South East of the path, as well as to the North East parts of the study region (CDFig.39). Six barrows (surely five and one (Taniokoleva) with only one of its possible locations) and the Goliama Detelina flat site are visible from this path. Three other sites – Aldinova mogila, Barrow 4 and the Ovcharitsa II enclosure - are very close to visible cells and it is possible that they may have been also visible from the path.

As already mentioned, the route with the best visibility over the landscape in the Northern part of the study region is Galabovo tell – Polski Gradets tell. Roughly 20 km after Galabovo tell, the north route bifurcates and the South branch ascends East South Eastwards towards Polski Gradets tell for 5 km (CDFig.40). The path has a very good panorama over the Ovcharitsa valley up to the end of the GIS coverage, a relatively good view over the Eastern part of the study area, as well as scattered visible strips in the central part of the region (CDFig.41). The same high visibility is valid for sites as well. Nine barrows are visible from the path (surely eight and one (Taniokoleva) with only one of its possible locations) and three flat sites. Barrow 4 has the same uncertain visibility status as mentioned above. It is worth discussing the visibility over the neighbouring site Polski Gradets pit site (for site location, see section 5.10.3). Three of the possible corners of the site (hence the area they enclose) are visible, and only one is not. Cell size and site location within a cell have already been discussed (p.94) so here it is only to be mentioned that most probably, while walking along the path, some parts or whole the site were visible.

The path to the Polski Gradets pit site forks from the branch to Polski Gradets tell and heads 2.5 km to the East (CDFig.42). The general visible areas from the path are as in Polski Gradets tell viewshed but less dense in the Eastern path of the study area (CDFig.43). The number of visible sites is also fewer - eight barrows are visible from the path (surely six and two (Taniokoleva and Tcherniova) with only one of their possible locations) and two flat sites. Barrow 4 is again close to the edge of visible /invisible cells.

The visibility from the paths to the last four sites is very similar and will be discussed in general. Only the tracks of each path will be discussed in some detail.

Three km after the Polski Gradets crossroad is the branch leading to the Ovcharitsa II enclosure (CDFig.44). It is 1km long and heads to the East. The next site – Aldinova mogila (CDFig.45) – is the only site in the Northern part of the study region that is located almost on the main North route. The segment leading to the barrow is less than 300m long.

The last fork of the main North route is 2.7 km after Aldinova mogila. One of the branches continues to the East for 500m before turning right and ascending 800m until it reaches Gonova mogila (CDFig.15). The other branch heads due South for 800m before ascending to the East for 300m, where it reaches Ovcharitsa I (CDFig.46).

The visibility from the last four paths is one and the same in both landscape and site terms up to the point of the Ovchartsi barrow (CDFig.16), (CDFig.47), (CDFig.48), (CDFig.49)– very good along the valley and with scattered strips and spots to the South East of the path. All but three of all sites in the North part of the study region are located in the visible areas. It is likely that the three sites were also visible, since in the case of the Tcherniova and Taniokoleva barrows, one out of four possible locations are visible from the paths, while Barrow IV is close to the edge of a visible /invisible cell.

There are differences in the landscape and site visibility of the last paths due to the particular location of the sites and the segments of the North route that lead to each of them. The differences concern only the Northeasternmost part of the study area and follow a pattern in which the Southwesternmost site (Ovcharitsa II) (CDFig.47) has the least visibility, while the Northeasternmost site (Gonova mogila) has the best visibility (CDFig.16).

Aldinova and Gonova barrows and Ovcharitsa II are contemporary sites, (dating to the EBA1-2 phase, and from the paths to each of them the other two sites are visible. There is only one exception – Gonova mogila cannot be seen from the path to Ovcharitsa II but instead Polski Gradets pit site is visible (CDFig.47). Most probably, parts of the same site were also visible from the path to Gonova mogila (CDFig.16). Ovcharitsa I is the latest site in the Northeasternmost part of the study region and, from its side-path, all the three earlier sites are visible (CDFig.49).

In summary, the majority of sites along the two main routes are located in an area in which they can be seen from the paths to the different barrows and flat sites. The number of visible sites increases as one moves East North East along the North route and, respectively, East South East along the South route.

Resources and land use

The site catchment analysis for the Galabovo tell was applied using a circle of 5 km radius (1 hour walking) from the site, the commonly accepted subsistence area limit for agricultural societies (Chisholm 1968; Higgs & Vita-Finzi 1970: see above, p.84 -86). Table 5.1.14 shows the distribution of soils around Galabovo in 10 successive rings, each of 500 m radius.

It is obvious that the present status of soil distribution around the tell suffers a huge human impact, seen as a removal or replacement of 1,185ha of natural soil (CDFig.50). A further 4,541ha do not contain any information for soil distribution since they are either occupied by contemporary mining constructions and settlements (the town of Galabovo and the village of Obrutchishte) or fall outside the study area for which relevant data was not available. Given the devastated condition of the region, a traditional application of SCA was not possible, so an alternative approach was used to explore possible resources and land use. First, estimations were done for subsistence potential according to the contemporary conditions. Secondly, an interpolation was made for possible soil distribution that provides different estimates for subsistence potential. Finally, it is suggested that the Galabovo exploitation potential lies between the two estimates.

Table 5.1.14 Soil distribution around Tell Galabovo
Distance from site Meadow Cinnomonic Smolnitsa Initial pedogenesis Artificial soil Without soil
0-500m 17ha - - - - -
500-1000m 59ha 1ha - - - -
1000-1500m 107ha 49ha 11ha - - -
1500-2000m 159ha 59ha 17ha 41ha 7ha -
2000-2500m 165ha 2ha 47ha 113ha 30ha -
2500-3000m 171ha 19ha 65ha 146ha 64ha -
3000-3500m 57ha 2ha 154ha 177ha 21ha -
3500-4000m 83ha - 169ha 164ha 27ha -
4000-4500m 96ha 8ha 248ha 127ha 29ha 13ha
4500-5000m 58ha 12ha 289ha 47ha 77ha 102ha

Exploitation area

A starting point in this case study of site catchment is the reconstruction of the population number of the tell. Since the site was not fully excavated, only indirect data from other prehistoric sites in Bulgaria was used. It is accepted that one and the same number of people have inhabited the tell through the whole occupational sequence, bearing in mind, however, that this is a mean figure and fluctuations and deviations were highly probable. Two sources for demographic analyses were used –Todorova’s estimations for the Eneolithic tell Ovcharovo in North East Bulgaria (Torodova et.al 1983) and Russell’s calculations for Near Eastern tells (Russell 1958, cited by Dennell and Webley 1975). According to Todorova, the average number of occupants of each building horizon is 48 (Torodova et.al 1983). The area of the Ovcharovo tell is 2,826 m2, i.e. a quarter of the Galabovo tell area of 12,500 m2. If we assume that the number of people inhabiting the Galabovo site were 4 times more than the number at Ovcharovo, this gives a figure of 192 persons. Russell’s estimations are for 125 persons per ha, which for Galabovo case result as 150 persons. The average4 value of the figures is 171 and that is the number of people accepted in the current study as a starting point of the SCA.

Dennell and Webley (1975, citing Clark and Haswell (1967)) have claimed that 210kg of grain per person per year is the minimum amount of cereal that would provide the necessary calories and protein for a population entirely relying on cereal consumption. They have also argued that a yield of 400kg per ha is an appropriate crop for prehistoric agriculture (Dennell and Webley 1975,106). If we reduce the amount by 50% taking into account spillage, disease, rotting and seed for the next year, that will give a figure of 200kg/ha. So, for a population of 171 persons an annual yield of 35,910 kg grain is needed. For a crop of 200kg/ha, that indicates an arable land requirement of approximately 180ha.

The second point is a further development of the idea discussed in Gaffney et.al (1985) (see p.86), in which domestic animals and pasture are included in the estimations of exploitation area.The very first circle around the site is assumed to be mostly used for animal pasture for three reasons:- a) protection of stock from predators and b) the availability of good grazing on riverine soils (Dennell and Webley 1975), and c) the preservation of crops from domestic animals.

Following the two major points, calculations of the possible resources around Galabovo tell were made. According to the data from Table 5.1.14, only 17 ha of meadow soil is available within the first 1-km ring. The land that covers the distance between 500 and 1500 m around the site (1-3 km in diameter) contains the necessary amount of 180ha area suitable for agriculture. Considering the fact that some of the land might have been used for fallow and/or browse, another 500m were added in order to delineate the possible exploitation area. The total area of 2 km in radius from the site contains enough potential to sustain a mixed agro-pastoral subsistence for a community of 171 persons.

The second estimation of the potential Galabovo exploitation area is made on the basis of interpolation. The pattern of soil distribution as given in Table 5.1.14, despite the obvious gaps, shows a dominance of meadow soil around the site. More substantial soil diversification appears in the area beyond the first km around the site. Such a pattern could be anticipated, bearing in mind that the river Sazliika was in the near vicinity of the site and that the tell lies near the confluence of the rivers Sokolitsa and Sazliika5. The assumption that the area of 1 km radius around the site is covered by only meadow soil gives a figure of 314ha – easily sufficient for the necessary 180 ha arable land, plus areas for pasture and fallow.

If this is an extreme situation of the distribution of only one type of soil - and there is no certain evidence to support this - I should rather suggest that the exploitation area for the Galabovo tell population was between the two figures – a 1 km radius if totally dependent on meadow soil and a 2km radius given the current soil distribution. Most likely, the active exploitation area did not exceed 1.5 km in radius. This is not to say that the land beyond that point was not in use. Fruit and herb collection and some form of herding may have taken place within the area of 1 hour’s walking. Hunting most probably was at a greater distance in the natural forests but not beyond the area bounded within 10km in radius from the site - a limit accepted for hunter-gatherers (Chisholm 1968; Higgs & Vita-Finzi 1970). Even in the present devastated condition of the study area, there is patchy woodland 10 km from the tell (Fig. 4.3.2), not to mention the distribution of the cinnomonic forest soil as an indicator for possible woodland South East of the tell Galabovo. The issue of initial forest clearance to free space for cultivation cannot be discussed in the absence of any pollen data from the region. Only an assumption could be made that it was a gradual process that started towards the end of the Copper Age. In the first place, wood was cut for house construction and fuel. The cleared area was then expanded and some cultivation may have started. The area was gradually enlarged until the exploitation area was more widely utilised.

A different scenario depends on the fact that, at the end of the Eneolithic, the Upper Thracian Plain was relatively densely settled by agricultural communities. If the first Galabovo occupants moved to the place as a result of some demographic or social process and they came as agriculturists, then a more intensive and target-oriented forest clearance should have taken place. However, until palynological data is available, the second hypothesis seems more likely as the material culture of Eneolithic Galabovo bears close similarities to what is known from contemporary Thrace; the first settlers in Galabovo were part of the extensive agricultural Copper Age network called KGK VI.

An expansion beyond the 2-km exploitation area could either maintain/sustain a larger population or contribute to social storage as surplus and/or for trade (Halstead & O’Shea 1989). The presence of three pithoi in a house in the 12th building horizon and two such pithoi in a 8th building horizon house supports the idea that some households had produced and stored more grain/flour than necessary for their daily/yearly consumption. There are two possible explanations for an increase of crop production –a) decrease of population given the diminishing occupational area in each subsequent horizon but with cultivation of the same area of arable land; and b) cultivation of new areas that incorporated the smolnitsa soil, distributed beyond the 1.5 km exploitation area. The 12th building-horizon house with the three pithoi dates to the MBA. No other settlements are known from that time in Maritsa Iztok study area and, in comparison with the EBA settlement distribution, there is an obvious overall population decline. Population decline, however, means diminishing of the number of workers, hence the ability to process the whole exploitation area (1.5 km in radius). Although the possibility for demographic change should not be excluded, it is worth considering the second possibility as well. So far there is no certain evidence from the Galabovo tell for tools that could facilitate the processing of the heavy smolnitsa soil. There are some indirect data from the osteological analysis, if we make the assumption that adult cattle were required for plough cultivation (Ribarov, n.d.). The number of the adult cattle individuals during the BA occupations is 73% - 20% more than during the Chalcolithic. But, within the BA sequence, this percentage diminishes through time, leaving the MBA horizons with fewer adult cattle individuals in the total sample. Therefore, it is possible to assume that the increased grain quantity along with smaller cattle herds implies the cultivation of very fertile soil. Indeed, it is also possible that the increased production of grain is due to some form of manuring but so far there is no secure evidence. Moreover, after centuries of cultivation, the meadow soil in the initial exploitation area (1.5 km in radius) probably suffered some exhaustion and new areas may well have been incorporated. Within a 1-hour walking distance, there was enough arable land for cultivation and it is equally likely that the pattern of land use was segmental as that it was concentric. The circular shape of an exploitation area was already discussed above (see p. 84 -85). It is used here because that allows the estimation of resources lying at equal distance from the site. In terms of cost, it is beyond the 2 km ring where more efforts are needed to reach a certain place. Since the terrain around Galabovo is not very uneven, the increased cost beyond the second km is due to the longer distance rather than the demands of a hilly landscape. In summary, within one hour’s walk from the Galabovo tell, and not necessarily in concentric areas, there was enough arable and pasture to sustain long-term site occupation. As long as the resources around the tell share, if not an even, at least a consistent distribution, it is likely that cultivation was carried out in segments around the site rather than as a continuous round strip.

Catchment area

Outlining the catchment area of the inhabitants of the Galabovo tell is based on the presence of excavated organic and non-organic remains.

The charred fruit of Ficus carica marks the longest possible distance for the connections of the Galabovo inhabitants. The fig is not a native species of Bulgaria but Middle Holocene environmental conditions in the Thracian valley tolerate its secondary development in the study area. So far, figs are found on Neolithic and Bronze Age settlements in Greece (Lisitsina and Filipovitch 1980), which outlines one possible direction of the Galabovo catchement area. The opportunity for local cultivation of Ficus carica is also possible, as remains from that fruit have been reported from the Neolithic layers of the Karanovo tell, some 50 km North of Galabovo (Marinova 2002). In the present state of investigations, it is difficult to determine the origin of the Ficus carica found at Galabovo. However, until more evidence for local cultivation is provided, the possibility of short-distance (Karanovo) and/or long-distance (Greece) exchange or trade of organic products (seeds in the case of the Greek sites and fruits in the case of Karanovo) should not be excluded.

Much more certain are the contacts with the Black Sea area. In the BA layers, one example of the marine shell Pecten was found, which suggests occasional trips between Galabovo and the Black Sea area during the Bronze Age. It is noteworthy that one of the possible routes to the Black Sea passes by the Drama microregion (Fig. 1.1.2). Another evidence for marine contacts either with the Black Sea or the Mediterranean Sea is the Spondylus ornament found in M6.

The polished stone tool assemblage (n = 81) contains a wide variety of raw materials (Table 5.1.15). Apart from the tools and materials given in the Table, a haematite ball and a mould of weakly consolidated sandstone were also found.

Table 5.1.15 Rocks used for polished stone tools at Galabovo tell. Source: Matchev, n.d.
Raw material Axes Adzes∗ Polishers and pestles Whetstones∗ figurines∗
Andesite 6 + 2   +
Limestone   +      
Amphibolite   + 2    
Marl   +      
Chlor-sericite schists   +      
Quartz (white, pink, reddish)     22    
Sandstone 1        
Small-grain sandstone     5 +  
Microgabbro 1        
Argillite 1        
Aplite         +

Matchev concluded that the tool usage determined the type of raw material. However, Table 5.1.15 shows that there is no clear tendency to produce a certain type of tool from a particular raw material. All of the rocks mentioned in Table 5.1.15 are spread around Galabovo within 10 to 50 km. There are natural exposures in the Sakar foothills, the Svetiiliiski vazvishenia (St. Ilia hills), the Manastirski vazvishenia (Monastery hills) and the Sredna Gora Mountain. The most distant are the andesites, that derive from the Sredna Gora range, about 50 km North of Galabovo. It was suggested that the andesite items were transported by river (Machev n.d., citing earlier sources). River transport has been claimed for some of the quartz tools that had traces of river rounding rather human processing.

The chipped stone tool assemblage shows a much larger catchment area than that of the polished tools. Ten types of raw materials (43 tools) have been identified in the Chalcolithic assemblage (Gatsov, n.d.). No cores were found but the flakes were derived from cores in an advanced stage of exploitation. It was suggested that the flint production process had happened outside the settlement (Gatsov, n.d.). Where this may have taken place and the possible raw material source were not pointed out. However, in a study of the flint assemblage of a site at 35-40 km South West of Galabovo, a local source of raw material was tentatively suggested (Gatsov, 1997) that was specified on the basis of mineralogical analysis of the tools to be in the Eastern Rhodopes (Kurchatov and Stanimirova 1997). Such statement is a general breakthrough in the late prehistory of lithic studies in Bulgaria. So far there are few special investigations on later prehistoric lithic assemblages and their possible sources (exceptions are Sirakov & Tsonev 1995, 2001) and it is a common practice to relate finished tools to raw material sources in North East Bulgaria or to an unknown source.

A similar uncertainty is also the case for the BA flint assemblage. Twelve types of raw material have been identified (n = 93 tools) that, on the basis of parallels with other sites, were claimed to derive from North Bulgaria or from unknown sources. However, some more targeted suggestions were made for the possible local raw material sources. So far, more than 30 exposures of Si2O are known in the Eastern Rhodopes Palaeogene depression. Although not very abundant in quantity, they show a wide diversity of quality and types of raw material, mainly jasper, jet, chalcedony and quartzite. An exposure of jasper was also found in the Sredna Gora range. Among the 49 blades, 10 flakes, 32 retouched tools, one blade in preparation and one small chip, there were specimens made of material with not very good technological properties that make the finished tools rougher than the specimens produced from the high-quality North Bulgarian flint. It was suggested that, despite the low technological properties of the local raw material, it was used because of its relatively easy access. It was also suggested on the basis of the low presence of production waste (one blade in preparation and one small chip) that the main flint processing took place off-tell (Zlateva-Uzunova, 2003).

In summary, the flint source catchment covers a huge area of 150-180 km to the North, crossing the Stara Planina Mountain range and about 100 km to the South. The Galabovo evidence supports the idea of long-distance contacts that most probably involved long-distance specialists, especially for crossing the Stara Planina mountain range. A general discussion of the distribution of flints in the study area will be made in Chapter 8.

5.1.4 Summary and discussion

The data from Galabovo tell are too inconsistent for a precise contextual, socio-economic and material culture study. The variety of evidence and some repeating patterns, however, provide a good basis for a general reconstruction of prehistoric life on one of the tells in the study region.

Tell Galabovo was located in a fertile area with a variety of natural resources (raw materials, minerals, vegetation cover, etc.). The long-lasting occupation (LCA-MBA) suggests that cultivation and exploitation of these resources was not devastating to the local environment and that there were enough organic and non-organic supplies to support a balanced agro-pastoral economy aided by some hunting and gathering activity.

The tell was relatively far from the other sites in the study region and not visible from any of them. It was, however, on the interfluve of two valley routes and contained strong evidence for short- and long-distance supply through exchange and/or trade.

Not all of the evidence on the tell represents deliberate activity, i.e. accidental fires or pottery breakage have probably taken place. Some outdoor activities and natural processes (e.g. house M7/8-N7/8 from the 13th building horizon) may have contributed to the depositional pattern discovered on the Galabovo tell. However, a striking continuity of social practices was observed throughout the whole occupational sequence of the Galabovo tell. Even in the present limited investigated area (in relation to the total tell area), there is repeated evidence for burnt houses, structured deposition in pits and foundation offerings, as well as personal and lineage enchainment through fragmentation. All of these social practices are implicit for the general concept of living on a tell, in which the link with the ancestors is a major motivation of social reproduction. In some cases, the possible search for ancestral identity is reinforced by building new houses directly over the destruction deposits of the preceding dwellings.

5.2. Obrutchishte flat site

5.2.1 Earlier studies and present condition of the data

The flat Neolithic settlement near Obrutchishte was excavated in the early 1970s but has never been properly published. The site is briefly described in a general article on the character of the Karanovo IV culture (Dimitrov 1971). Recently, the site was included in the study of the Maritsa-Iztok settlement pattern but without the provision of any new data (Leshtakov et al. 2001). During my museum study period, it was not possible to access the archaeological material from Obrutchishte. However, the site is considered in the current research, as it provides important evidence for human occupation at the end of the Neolithic. Although the archaeological features at Obrutchishte cannot be discussed, the landscape aspects of the site can be investigated. The evidence for Late Neolithic occupation near Obrutchishte should not be omitted, because, despite the paucity of data, the site is an important indicator of settlement diversity. In a landscape of growing tells (Klisselika and Mednikarovo), the foundation of a new settlement raises the question of its possible relation to the earlier and/or contemporary sites within the study region. Also a crucial point is the abandonment of the site and, in particular, why it did not developed into a tell. According to the present condition of the data, the answers to these questions can be explored only from the landscape perspective of the site.

5.2.2 The site and its surroundings according to the GIS analyses  

According to the publication, the settlement was located on a high terrace 1.5km South East of the village of Obrutchishte in the locality Selishteto (the settlement). Precise co-ordinates have not been given, leading to the imposition of a random choice of possible site locations within the Selishteto area. A point has been chosen roughly in the middle of the Selishteto locality, whose area is not more than 30 ha. The maximum size of the site is 1ha (Dimitrov 1976). A second point 500m South East of the first one (at the very edge of the locality) was considered as another possible site location. A parallel set of GIS analyses was performed for both of the site locations and the results do not differ significantly. The analyses presented here are valid for the first site location and comments are made when they do not match the results from the second site location.

The flat site of Obrutchishte is located on a high terrace (140-164 masl) of the river Sokolitsa (CDFig.51). It is on a 1-2° slope (CDFig.52) with a Southwestern aspect (CDFig.53). The same elevation is shared by the second place but it is on a flat surface with a Northeastern aspect. The general visibility from the site is mainly over the Sokolitsa valley (CDFig.54). The sloping terraces North of the valley are visible, with a less patchy view from the second location. Of all the Neolithic sites, only Klisselika tell could be seen from both possible locations. The distance between the two sites is over 11 km. Such a long-distance landscape and site visibility from a static point appears in other cases as well (e.g. Gudgova mogila) but its feasibility may be questioned. Generally speaking, such long-distance visibility is possible, as shown by my own field-walking experience in other research projects. However, in the case of Maritsa Iztok, the pattern cannot be tested due to the degree of present landscape devastation.

The cost surface analysis based on slope shows that, despite the longer distance to tell Klisselika, in terms of cost, similar efforts are needed to reach both Klisselika and Mednikarovo – the two Neolithic sites in the South part of the study area (CDFig.55). The route network derived from cost surface analysis consists of three paths (CDFig.56). The South routes to Klisselika (CDFig.57) and Mednikarovo tell (CDFig.58) join the major South route from Galabovo tell to Gudgova tell, discussed in detail in section 5.1.2. Since the former is earlier then the latter and both routes derive from a different cost surface source, I should suggest that this repetition of road tracks is an important evidence for the existence of such a route in the later prehistory of Maritsa Iztok. The same track match is valid for the major North route that connects Galabovo with Ovcharitsa II. The Neolithic route (CDFig.59) from Obrutchishte to Ovcharitsa II (the only Neolithic site in the Northern part of the study area) crosses the study area from South to North, then turns North East and roughly 4 km North East of Galabovo tell joins the main North route.

Visibility from the path between Obrutchishte and Mednikarovo tell is over the Sokolitsa valley, while the route is along the valley itself (CDFig.60). Most of the areas North of the valley, including the Klisselika tell, are also visible, as are the hills South of Mednikarovo.

The visibility from the path Obrutchishte – Klisselika tell should not be discussed again, as in fact it forms part of the South route discussed above (see p. 178 -179) and hence shares the same visibility. The route to Ovcharitsa II also is not going to be discussed further, since it crosses the contemporary mining areas and should produce a biased viewshed. The visibility from the point in which the path joins the North route has been discussed in section 5.1.2 as well. The paths and their visibility from/to Obrutchishte and the remaining 24 sites is not discussed in the text because a) the site is with uncertain location and, b) there is no evidence that the site was visible after its abandonment (i.e. during the Chalcolithic and Bronze Age).

Resources and land use

The site catchment analysis for Obrutchishte settlement was applied using a circle of 5 km radius. The soil distribution in Table 5.2.1 is given for 10 successive rings of 500m radius around the first site location. Since the second site location remains within the first circle, any soil distribution analysis will move by 500m to the South East. Despite some quantitative differences, significant differences in the soil variety and distribution are very unlikely. Therefore, soil distribution analysis was made just for the site location in the middle of the locality Selishteto (CDFig.61). Table 5.2.1 does not include areas that are now contemporary settlements, as well as areas that fall outside the study region. The last three columns show the recent anthropogenic impact on the natural soil distribution. As long as the latter is not possible to reconstruct, the figures in the last three columns were not taken into consideration in SCA estimations. Although their resources were not investigated, these areas should provide some opportunities for additional resources and land use diversification.

Table 5.2.1 Soil distribution around the Obrutchishte flat site
Distance from site Smolnitsa Meadow Cinnomonic Artificial soil Initial pedogenesis No soil
0-500m 48ha 12ha 20ha - -  
500-1000m 114ha 63ha 19ha 5ha 5ha  
1000-1500m 185ha 45ha 24ha 49ha 21ha  
1500-2000m 214ha 68ha 72ha 75ha 53ha  
2000-2500m 259ha 75ha 109ha 53ha 155ha  
2500-3000m 285ha 108ha 151ha 143ha 145ha  
3000-3500m 318ha 105ha 148ha 73ha 180ha 2ha
3500-4000m 103ha 163ha 188ha 10ha 13ha 100ha
4000-4500m 85ha 138ha 108ha 6ha 122ha 189ha
4500-5000m 87ha 170ha 136ha 15ha 78ha 256ha

Exploitation area

Following the population estimation algorithm performed for the Galabovo tell, the results for the Obrutchishte settlement fall within the range of 125 (according to Russell) and 168 (according to Todorova). The annual crop needed to feed such a number of people is 26,250 –35,280kg. For an annual yield of 200kg/ha, that gives a figure of 131 – 176 ha of arable land.

As in the Galabovo case, the very first circle around the site is assumed to be used mostly for animal pasture.

The soil distribution pattern around Obrutchishte shown in Table 5.2.1 follows a trend of successive increase and dominance of the heavy smolnitsa soil up to 3,500m in radius from the site. There is a tendency for an increase in the quantity of available meadow and cinnomonic forest soil as well but their extent falls well below that of the smolnitsa spread. Two models of possible land use were explored. In the first one, it is assumed that the smolnitsa was too heavy to cultivate in the Neolithic (cf. Dennell & Webley 1975) and hence the areas with smolnitsa distribution were excluded from the estimation of exploitation areas. The second one assumes the existence of high-effort but high-yield cultivation for which the right conditions for successful breaking of the smolnitsa are vital, viz., the ground is softened after rain in the previous day (Chapman, pers. comm.).

In the first model, only meadow and cinnomonic forest soils are considered as possible arable land. According to the data from Table 5.2.1, the area from 500m to 2000m in radius from the site contains the necessary amount of arable land, if the areas with the cinnomonic forest soil were fully deforested and one third of all the possible cultivation area was left for fallow. To avoid such an extreme claim for deforestation –given the complete lack of pollen data- another 500m was added as a possible exploitation area. Within an area of 500-2,500m radius from the site, there were enough resources for a dynamic agro-pastoral strategy of fallow/arable land rotation and some forest clearance for both arable and browse land.

In the second model, such a balanced land use with broadly equal proportions of arable, fallow and browse without any severe deforestation is possible within an area 500-1,500m radius from the site. In the first model, a successful subsistence strategy was possible while exploring areas relatively far from the site - up to 2500m but still within 1 hour’s walk. In the second model, the available resources were much closer to the site – up to 1,500m from the site but only in the case of successful labour management of arable production.

5.2.3 Summary and discussion

It would be unhelpful to give definitive interpretations of the abandonment of the site based on an unsuccessful subsistence strategy before checking similar situations with the other sites. It is hard to see that any over-reliance on the high-effort cultivation of smonitsas could have led to anything but a transfer of attention to the lower-risk cinnomonic forest soils and meadow soils, rather than site abandonment. It is suggested that social rather than direct pedological reasons (e.g., lineage fission, high dispute levels) may have been responsible for site abandonment prior to tell formation.

5.3 Atanasivanova mogila (barrow) site

5.3.1 Earlier studies and present condition of the data

Atanasivanova mogila is located 2.5 km North East of the village of Mednikarovo. The toponym of the site is a self-evident argument for the general perception and concept of these mound-like landscape features. Such a common-sense acceptance that this is a burial monument initiated the excavation of the site in 1987 (Borisov n.d.b). The feature was 72m long along the North-South axis and probably of similar length along the East-West axis (the latter was destroyed by construction works). It was at least 8 m high (the height was not specified) and contained just one burial. The grave was on the North East edge of the mound, 32m from its centre. The deceased was buried in a pit no more than 80cm from the present surface of the mound. The lack of any grave goods prevented any chronological attribution for this burial. The unsuspected paucity of graves in such a big mound (there is a barrow in Maritsa Iztok that is smaller than Atanasivanova mogila and contains 36 graves: (see p. 277-278) raised the question about the character of this feature. After a consultation with the mining geologist, it became apparent that Atanasivanova mogila is, in fact, a mud volcano (see p. 113 and Fig.4.1.4).

A grave in a prominent landscape feature immediately provokes the question of the perception of the feature at the time of the burial. Was it considered as a natural (a low hill) or a cultural (ancestral mound) feature? Did the participants in the burial process make the mistake of the modern farmers and experienced archaeologists – the former calling the hill mogila, the latter excavating it, misled by their professional background ? The answers to these questions is the reason to include Atanasivanova mogila in the present study, despite its uncertain chronology. Performing a set of landscape GIS analysis, I would argue that it is highly probable that the burial in Atanasivanova mogila can be dated to the beginning, or during the course, of the EBA.

5.3.2 The site and its surroundings according to the GIS analysis

Atanasivanova mogila is located on a river terrace at 115 - 140 masl (CDFig.62). It is on a 12° slope (CDFig.63) with a North Eastern aspect (CDFig.64). The visibility from the site is mainly to the North and South East, with restricted views to the South West (CDFig.65). The Iskritsa pit site and Gudgova tell are visible from the site. Ten meters were added to the actual terrain model surface in order to reconstruct the height of the barrow, as well as some additional height that may have been swept away by past or present human activity. The visibility (CDFig.66) from the site with the additional 10 m is about 4 km West along the valley of the river Sokolitsa; roughly 2 km around the site; and a denser view to the North of the valley in comparison to the previous patchy Northern views In addition to the previously visible sites, one can now see the Klisselika tell. In both viewsheds, the special visual status of the Iskritsa dwelling site is confirmed here as well.

The cost surface analysis (CDFig.67) resulted in the following sites distribution (Table 5.3.1):

Table 5.3.1 Site distribution around Atanasivanova mogila
N of cost strip Sites located in the cost strip
0 Iskritsa dwelling site, Iskritsa pit site, Klisselika, Gudgova and Mednikarovo tells
1 Obrutchishte flat site, MIBC, KMBC
2 Taniokoleva mogila 2-4, Kurdova mogila, Galabovo tell
3 Tcherniova mogila – all locations, Goliama Detelina flat site, Taniokoleva mogila 1, Manchova mogila,
4 Goliamata, Malkata and Ovchartsi barrows, Barrow4
5 Polski Gradets tell
6 Ovcharitsa I and II, Gonova and Aldinova barrows, Polski Gradets pit site

In summary, the least effort is needed to reach the sites along the valley located to the East, as well as Mednikarovo tell, situated South of the Sokolitsa valley. These sites are also some of the earliest in the study area. It is interesting that Galabovo tell is in the second group of cost strips, although it is also along the same Sokolitsa valley.

The logistics network derived from the cost surface analysis matches the path system of Galabovo in the layout of the two main routes (CDFig.68). Atanasivanova mogila is situated on the main South route of the Maritsa Iztok study region and the differences with the previous route network are in the tracks to Mednikarovo tell, MIBC, Manchova, Kurdova and Taniokoleva barrows and Barrow 4. Viewshed analyses were run only for the new paths.

The path to Mednikarovo follows the valley for 800m to the West, then turns South South West for 2 km before ascending South East for less than 1 km (CDFig.69). The visibility from this route is very good over the areas along the path – up to 3 km - and towards the North part of the Sokolitsa valley (CDFig.70). The Iskritsa pit site and Gudgova tell are also visible.

An interesting pattern is observed for the routes to the MIBC. One main route is followed to barrows 1,2 and 4, (CDFig.71), (CDFig.72), (CDFig.73) which bifurcates to three paths some meters before the immediate access to each barrow. The general visibility from the routes is good over the valley and the hills to the South and patchy towards the North part of the valley. The panorama becomes denser in the last two viewsheds for the paths to MIBC2 and 4 (CDFig.74), (CDFig.75). All the sites in the Eastern part of the valley are visible from the path but it is very interesting that the final segments after the bifurcation has a crucial impact on intervisibility between the barrows of the cemetery (CDFig.76). The first segment that leads to MIBC1 has no visibility over the other three barrows of the cemetery. The second segment to MIBC2 that is 350m long in a Northeasterly direction provides the visibility over barrows 1 and 3. The last segment that leads to MIBC4 and continues North of the previous segment has the best visibility, as all four barrows can be seen. The route from Atanasivanova mogila to MIBC3 follows a different track that, in general, is parallel to the path to the other three barrows (CDFig.77). The availability of this second route means that, if a special journey was intended to barrow 3, related to some common/communal pilgrimage activity (e. g. a family memorial rite for a person buried in barrow 3), there was an opportunity for direct access to this particular barrow. It is interesting to point out that, in comparison with the visibility from the path to the other three barrows, Gudgova mogila is not visible (CDFig.78). The partial intervisibility between the barrows is also confirmed, as only MIBC2 and 4 are visible from the path.

The paths to the sites located North of Atanasivanova mogila have shown the advantages of GIS applications in archaeological studies. There are eight barrows and one flat site relatively evenly distributed North of Atanasivanova mogila. An oversimplified and schematic site distribution map would plot the sites at an absolute distance from each other in a straight Northwesterly direction. Most probably one would assume that the link between the sites was such a hypothetical straight route from South to North (if the movement was from Atanasivanova mogila to the northern sites, and respectively from north to the south if the travel was in the other direction). In fact, the GIS cost path shows a completely different pattern, in which the paths to four of the barrows and the flat site follow the main South route Westwrads before crossing the study area from South to North West through the modern mining areas and joining the main North route 4km North East of Galabovo Despite the probable bias of the actual outlines of this route as it crosses an area of destroyed landscape, the direction and the track of the path are against the common-sense assumption that paths between sites always follow the principle of least distance. In terms of efficiency, it is sometimes better to walk along longer distance but on relatively even terrain, rather then along short distance but hiking over steep hills. The same pattern appears in other case studies as well (e.g., Klisselika tell – Ovcharitsa II), which confirms the validity of such a pattern.

The paths to the other four barrows indeed cross the study region from South to North, thus confirming that the neglect of a direct route can also be inappropriate. Rather, the logistics between sites should be based on both the landscape and archaeological data.

The route to Kurdova mogila ascends to the North for 2.7 km and then follows the ridge of the hill to the North West for 3.5 km (CDFig.79). The hilly central area along the path and the hills in the Eastern path of the study area and South of the Sokolitsa valley are seen from the path (CDFig.80). The valley is also visible, hence all the sites along the Sokolitsa as well. The scattered visible spot to the North of the valley assures the view over MIBC1 and 2. Another four barrows located in the Northern part of the study area are also visible.

The route to Manchova and Taniokoleva mogila follows one and the same track for 6.5 km. It starts Eastwards, following the valley for 1.5 km, then turns Northwards, ascending for 2.8 km and finally descends for another 2.5 km. At that point, the route bifurcates, with the West wing descending to the North and North East for 2.2 km until it reaches Taniokoleva mogila, in its path passing the all four possible locations of the barrow (CDFig.81). The East wing descends and ascends again for total of 3.1km to the North East up to Manchova mogila (CDFig.82). The visibility from the two paths is similar – good over the Sokolitsa valley and the hills in the Eastern part of the region and South of the valley. There is a consistent strip-like visibility over the central part of the study area and a patchy view to the Northern parts of the region. The visibility from the path to/from Manchova mogila assures better visibility over the Northern study area (CDFig.83). All the sites in the Sokolitsa valley - MIBC2, Polski Gradets tell, five barrows and the Goliama Detelina site - are visible from the paths. The only difference is that Goliamata mogila is not visible from the path to Taniokoleva mogila (CDFig.84).

The last path to be discussed here is to Barrow 4 (CDFig.85). It forks from the path to Manchova and Taniokoleva mogila 2.4 km after Atanasivanova mogila, ascends for 2.5 km to the North/North East, descends and ascends again for 1.4 km to the North East and finally follows the ridge of the hill to the North West for 3 km. The visibility pattern from the path is patchy and scattered across the central, Eastern and Northern parts of the study area (CDFig.86). It is interesting to point out that, despite the lack of consistent landscape view, there is a very high site visibility – six or seven (Aldinova is on the edge visible/invisible area) barrows, two tells and two flat sites. The visibility pattern of Atanasivanova mogila - Barrow 4 path will be discussed in section 6.6.2.

The paths to both Iskritsa sites are discussed in section 5.6.2 because of the specific visibility status of the Iskritsa dwelling site.

The path to Klisselika tell is a part of the main South route and shares the visibility Galabovo tell - Klisselika tell from the location of Atanasivanova mogila Eastwards.

In summary, the paths to the three barrows that cross the study region from South to North present a visibility pattern in which the movement from the Sokolitsa valley to the North revealed all the earlier and contemporary sites, as well as a large number of barrows located in the Northern part of the study region. If Atanasivanova mogila was formalised as a mortuary place at the end of the Copper Age/ or the beginning of the Bronze Age, it is likely that some of the barrow locations may have been chosen in respect to their visual status from the path to/from Atanasivanova mogila. A direction of movement from North to South would have happened after the beginning of the Bronze Age; in that case, the visibility over the earlier sites and only one contemporary site (Gudgova tell) in the Sokolitsa valley may have had some importance.

The pattern of site visibility is very similar from the paths to the other sites in the Northern part of the study area, although there is a huge difference in the routes chosen and their landscape visibility6. This confirms the possibility of deliberate barrow location in respect to their visibility from the paths, as well as the importance of visual contact with earlier sites while walking across the landscape.

5.3.3 Summary and discussion

According to the results of the GIS analyses, Atanasivanova mogila was located along one of the main routes used in the later prehistory of the Maritsa Iztok study area. The volcano was there prior to any human occupation but its mound-like shape gained some specific cultural meaning, most probably after the end of the Chalcolithic. At the beginning of the BA, the Maritsa Iztok study area consisted of one mature tell (Klisselika) and four growing tells (Galabovo, Gudgova mogila, Polski Gradets and Mednikarovo). There were also at least two barrows – Gonova mogila and Goliamata mogila. In a landscape of mound-like cultural features and with an already established concept of a formal mortuary domain, a prominent hill silhouette that dominates the local landscape provides an excellent opportunity for incorporating the feature in the system of landscape communication. The act of burial within a feature that is strongly reminiscent of an ancestral tell place, while at the same time resembling a formalized funerary arena could be seen as a deliberate act of relating the newly dead to the (potential) ancestors and in the same time emphasizing the status of the deceased, which gives him/her the right to be buried under a barrow. The appearance of barrows in Maritsa Iztok will be discussed in Chapter 8; here, it is noteworthy that, although not a real barrow, the Atanasivanova mogila site is a burial place that very much resembles a barrow visually and therefore it is considered as such in the current study.

It is also likely that Atanasivanova mogila burial was made during the course of the EBA, when even more barrows had appeared. In such a case, the idea of imitating barrow burial is additionally supported by the presence of conceptualized mortuary places that are standing as powerful social landmarks, and thus stimulating the negotiation of social re-production between the living inhabitants of the landscape. The position of the body - back extended with slightly contracted legs, gathered feet and knees - does not contradict an early date within the EBA for this burial. Parallels for this body position in the Maritsa Iztok area are known from Goliamata mogila. The lack of grave goods is also not an exception in the study area (e.g. Aldinova mogila). Recently, a date was suggested for the Atanasivanova mogila burial in the late Roman period (Borisov and Ivanova, in prep.). Given the current condition of the data, the authors, however, are cautious in insisting on such a late date.

In summary, Atanasivanova mogila site is a natural place that contains the remains of a cultural practice. The modern perception of the site as a cultural feature was later opposed by its natural character but the evidence suggests that such division was most probably not valid at the time of the burial. Whether the mound was natural or not was not particularly significant, as long as the mound containing the burial served its role in the social re-definition of the landscape. Being at an important place within the logistics network of late prehistoric Maritsa Iztok, Atanasivanova mogila was fairly easily incorporated into the social landscape.

5.4 Mednikarovo tell

5.4.1 General information and earlier studies

The site of Mednikarovo entered the archaeological record in 1987, after the autumn field survey of the Maritsa Iztok Expedition and was then assigned as a tell. At that time, the Eastern part of the site was destroyed by a channel and its Western periphery by road construction (AFig. 5.4.1A, B). Six years later, a Bulgarian-American team undertook sondage excavations of the site, leading to an alternative view of the site. At present, the site is considered as a flat settlement (Nikolov 1998). It is situated on a small 5m-high hilltop that, after years of cultivation, resembled a tell-like settlement mound. The rare reference to the site (investigation results are not published yet) refer to the height of the site as both 2m and 3 m. Setting Mednikarovo within its contemporary Neolithic landscape, I should argue that it is a tell-like site, which did not develop into a mature tell but which, at the same time, was not an exceptional settlement type at the time of its existence.

The excavation consisted of a step-like trench consisting of 5 successive sondage units. Each sondage was oriented North - South and measured 12m long and 2m wide. Between the sondages, 25cm-wide control profiles were left. An additional trench, measuring 3.60 to 2.90m, was excavated in the South West part of the hill that is currently the edge of a high terrace overlooking the Karapelitska stream. The type and size of the trenches were chosen in consideration of contemporary agriculture ploughing of the whole site and according to the aims of the investigation: to establish the stratigraphy and chronology of the tell. (AFig. 5.4.1B, AFig. 5.4.2). In the excavated area of 130m2, five successive layers were recognised on the basis of changes in the soil colour and texture (AFig. 5.4.1C, D). Four different niveaux were distinguished in layer 4 and two niveaux in layer 3. Within layers 1 (arable land) and 5, archaeological features were not found and layer 2 contained five pits. The general stratigraphy of the tell and the relationship between contemporary features are not available for Mednikarovo, as the site is still unpublished. On the basis of the pottery found during the excavations, two Neolithic occupational phases were recognised – the final stages of the Early Neolithic (Proto-Karanovo III) and the final stages of the Middle Neolithic (Karanovo III-IV) (Nikolov 1998). Chalcolithic and BA sherds and vessels were also found but without any clear stratigraphic context (Leshtakov et al.2001).

Archaeological evidence

During my museum study in 2000, I was given full access to the excavated archaeological material and all the available site documentation in Bulgarian. I was able to look at 1/10th of all the material - mostly pottery. Unfortunately, it was impossible to relate the information from the storage unit records written by the American team to the stratigraphic units recorded in the Bulgarian field documentation. For example, the former contained many more archaeological contexts than were summarised in the final site report. The following description of the archaeological sequence and features contains only data that can be validated by at least two sources (e.g. the site report and the site documentation or the site documentation and the storage units).

Sondages 2 and 3 were the only two zones with in situ remains. The earliest occupation of Mednikarovo was identified in sondage 3, where house rubble was excavated (AFig. 5.4.2H). Details of the construction and plan of the dwelling were not given. Early Neolithic sherds were the only finds within the restricted excavated area. The next occupation was identified in the same sondage after a levelling layer of light-brown soil, mixed with burnt daub/house rubble and sherds. A dwelling floor of beaten clay, 1.25m in length and 5-7cm in thickness, and a posthole were excavated. Not many sherds were discovered and none of them had any chronological significance.

The following occupation was securely dated to the Late Neolithic and was marked by a burnt dwelling floor. The latter was made of beaten clay and measured 2.27m in length, 15cm in width and 5cm in thickness. The floor was disturbed by a pit from a later occupation. Most of the dwelling floor and the pit fell within an un-excavated area. The excavated part of the pit contained single sherds and dark-brown soil, very similar to the surrounding layer 2. Within the same Late Neolithic layer (Karanovo III-IV) but 71cm above the dwelling in sondage 3, another area of house debris was found in sondage 2. A dwelling floor of beaten clay 5-7cm in thickness and with a preserved size of 3.5m by 2m was excavated. No sherds and traces of fire were mentioned to be present. The base of a rectangular oven was also found. The floor and the oven were covered by a layer of burnt house debris. Sondage 2 contained 4 more pits, dug from different depths within layer 2 (AFig. 5.4.1D). According to AFig. 5.4.1D, the earliest of the four pits was pit N4. It is 40cm in depth and contains single non-characteristic sherds and soil similar to the surrounding matrix from layer 2. Most probably, the next pit to be dug was N5, which was filled with dark-brown soil, burnt daub/house rubble and few sherds. Pits 1 and 2 are 50cm apart and the latter is earlier as its mouth is below the mouth of the former (AFig. 5.4.2D). Pit 2 contains dark-brown soil, similar to the surrounding matrix in layer 2, a few uncharacteristic sherds and medium-sized broken stones. Pit 1 has the same characteristics as pit 2 but contains some bones and sherds (AFig. 5.4.4 Q-Ff) as well as five almost complete vessels with a secure Neolithic date (Karanovo III-IV) (AFig. 5.4.3K-M, AFig. 5.4.6L, M). During the course of the excavations, a human skull was noticed in the profile of a pit, that prompted an expansion of the excavated area. Two human skulls and numerous disarticulated and heavily broken human bones were found. The burials were dated to the Late Mediaeval /Pre-Modern time, as local peasants confirmed the presence of a cemetery from that period. The poor condition of the skeletons was assigned to the modern cultivation techniques but how the burials related to pit 1 remained unclear.

Layer 2 as a whole was dated to both the Neolithic (AFig.5.4.4 A-P) and BA, according to the associated sherds. The exact find spots of these datable sherds were not provided. On the basis of the pit fill, I should assume that digging pits into the Neolithic layers during the Copper (single sherds found) and BA caused the sherds from these earlier occupations to become spread over the contemporary Copper/Bronze Age surfaces. Pits were filled with the surrounding soil, explaining the similarity between the pit fill and the soil matrix of the layer from which they were dug. Other finds from layer 2 comprised: 4 flint tools, a spindle-whorl, a figurine head, a fragment of bone awl, fired clay sling bullet and two net weight that according to the excavators were made from body sherds (AFig. 5.4.6F-I).

The pit digging practice was confirmed by evidence from sondage 6, where several inter-cutting pits were found. They contained bones, stones and sherds from the Neolithic, Copper and BA. There were rims, body parts and bases in both fine and coarse ware, fragments decorated with different techniques and patterns, whole and fragmented handles, 2 fragments of altars (AFig. 5.4.6J), 2 bone awls, 9 flint blades and one almost whole BA vessel. Given the present state of the data, it is not possible to reconstruct whether the pit-digging practice has started in the Neolithic or the Neolithic sherds derive from disturbed Neolithic layers. Chalcolithic and BA pit digging on Mednikarovo tell, however, must have been a recurrent practice.

Apart from the artefacts mentioned so far, 6 flint blades were found in layer 1 and two flint blades and three stone tools were un-stratified.

Although most of the ceramic material consists of sherds, there is a great typological diversity of shapes, pointing to some kind of intensive dwelling activity (AFig. 5.4.3-5.4.6).

During my museum study, I came upon a number of stone tools that I was not able to relate to any part of the archaeological sequence described above. However, their presence should not be omitted. They are nine stones of different shapes and sizes, which, according to the excavators, were grinding stones. All of them were made of quartz and had at least one smooth/polished side. Two were visibly fragments of bigger tools. It is interesting that all of them derive from one and the same sondage but from different levels.

The uppermost layer 1 (the arable soil) contained numerous sherds with traces of heavy wear and erosion, indicating long-term surface exposure. Among the 5 kg of pottery examined in 2000, there were sherds from the Neolithic, Copper and Bronze Age and a fragment of Early Iron Age ware but most of them were very uncharacteristic. They derived mainly from sondage 1 but the chronological incoherence of layer 1 was also confirmed by the data from the site report. The sherds contained both fine and coarse ware, with rims, bases and body sherds all present. Very similar were the characteristics of the finds from sondages 4 and 5, located in the highest zone of the site. In addition, there were some animal bones, as well as fragments from wheel-made pottery, among which there were sherds from very Late Medieval/Pre-modern times. On the base of this evidence, I should suggest that, after the last Neolithic occupation, there were some short/temporary settlement activities or most likely some structured deposition, such as burials and pit-digging that, after years of intensive cultivation, were totally destroyed. Not only have the in situ contexts of the later occupations been destroyed but the Neolithic layers have also suffered past and present anthropogenic intrusion.

5.4.2 Plant remains

Pollen samples have been taken from a drill core, every 20 cm up to 180 cm in depth. The number of pollen grains was 5-6 in a sample, while to be reliable they have to be 300 in a sample. However, pollen from Chenopodiaceae, Poaceae, Compositae and Alnus glutinosa was found. There is also some evidence for cereals but it was not possible to identify these to species level (Popova 2001).

5.4.3 The site and its surroundings according to the GIS analyses

Mednikarovo is located on a high terrace of the river Karapelitska, at 140-164 masl (CDFig.87). It is on a 2-3° slope (CDFig.88) with a South West aspect (CDFig.89). Visibility from the tell is very restricted – mainly to the areas South East and North West around the site itself and patchy spots 5 – 7 km to the North West (CDFig.90), (CDFig.90a). No sites are visible at all. The restricted visibility status of Mednikarovo appears as a trend in the viewshed analysis of the other sites and more importantly in the viewshed analysis of the paths between sites. Among the routes within the logistics network of the study area, Mednikarovo is seen only from the path Klisselika- Ovcharitsa II.

The results of the cost distance analysis (CDFig.91) and the distribution of sites within the 10 cost strips are summarised in Table 5.4.1:

Table 5.4.1 Site distribution around Mednikarovo tell
N of cost strip Sites located in the cost strip
0 Atanasivanova mogila, both Iskritsa sites
1 Obrutchishte, KMBC, Klisselika and Gudgova tells, MIBC1
2 Galabovo tell, MIBC 2-4
3 Kurdova mogila, Taniokoleva mogila–all locations, Tcherniova mogila – all locations, Goliama Detelina flat site
4 Manchova, Goliamata, Malkata and Ovchartsi barrows, Barrow 4
6 Ovcharitsa I and II, Gonova and Aldinova barrows, Polski Gradets tell and Polski Gradets pit site

In summary, during the Neolithic, Mednikarovo tell was not in immediate access to any of its possible contemporary sites. Quick and easy access to contemporary sites would have gained some importance from the beginning of the Copper Age onwards, when there was increased site density around the tell. During the Neolithic and Copper Age, there was a tendency for sites to be more easily reached than in the BA, when site accessibility was lower in cost.

The logistics network based on the Mednikarovo cost surface (CDFig.92) matches the main valley routes and the South-North routes, as discussed in the path analyses of Atanasivanova mogila (see above, p. 193 - 195). There are two segments starting from the tell – one towards the Sokolitsa valley, the other to the KMBC. The latter will be discussed in section 5.5.2. The former ascends to the North North East for 2.5 km, when it joins the main South route. The segment is a part of the path Mednikarovo tell - Atanasivanova mogila and hence shares similar visibility. Therefore, the visibility from the Mednikarovo tell paths are not going to be discussed here, since they combine the viewshed from the path Mednikarovo tell - Atanasivanova mogila and the visibility from the Atanasivanova mogila network.

It is noteworthy that, if the main South route was in use during the Neolithic, connecting Mednikarovo tell with Klisselika tell, it may have affected the establishment/foundation of the later sites (Atanasivanova mogila, both Iskritsa sites and MIBC2), since all of them are visible from the Copper/Bronze Age route Mednikarovo tell – Gudgova tell. This means that all the places on which later sites have emerged are visible from the Neolithic route and that may have played a role in the choice of their location. From the path that connects Mednikarovo and Ovcharitsa II during the Neolithic, Klisselika tell could be seen. The comparison of several route tracks and their visibility have confirmed that the point from which the tell was visible is South of the present mining area; hence, it is likely that such a sight-line has existed during the Neolithic. However, such a claim was not confirmed for the later sites and visibility from the path that crosses the study area through the contemporary mining area is not taken into consideration.

Two of the paths that cross the study area from South to North (to Barrow 4 and to Taniokoleva mogila) have the same tracks as the paths from Atanasivanova mogila. Therefore, they share the same site visibility and very similar landscape visibility (the viewsheds from Mednikarovo are more consistent around the tell itself) (CDFig.93), (CDFig.94).

The only difference between the Atanasivanova mogila logistical network and the Mednikarovo tell logistical network lies in the paths to MIBC and Kurdova mogila, as well as the path to Manchova mogila (CDFig.95), which does not cross the study region but follows the main valley routes.

When the path from Mednikarovo tell reaches the Sokolitsa valley, it bifurcates, one branch joining the main South route, the other heading North towards Kurdova mogila. The former splits from the main route after 600m and ascends to the North East for 2.3 – 3 km to the different barrows of MIBC (CDFig.96), (CDFig.97), (CDFig.98), (CDFig.99). The four tracks and their visibility are very similar. There is very good visibility over the Southern part of the Sokolitsa valley and the hilly areas to the South. The view over the North parts of the valley and the Eastern hills is patchy. There is a consistent panorama to the East of the paths. All the sites in the valley are visible from the paths. The only difference lies in the number of visible barrows from the cemetery itself when it is approached from the South East. From the path to MIBC1 (CDFig.100), only this barrow is visible; continuing to the East toward MIBC2 (CDFig.101) – apart from the already passed barrow 1- MIBC3 is also visible. The paths to MIBC3 (CDFig.102) and 4 (CDFig.103) – 300m North of MIBC2 - assure full intervisibility between the barrows in the cemetery. A similar pattern of visibility while approaching the MIBC was observed in the previous case study as well. Final comments on the patterns are going to be made in Chapter 8.

The path to Kurdova mogila ascends to the North for 2.5 km, turns right for 1.1 km and then follows the hill ridge North West for 2.7 km (CDFig.104). There is a consistent view from 1 - 5 km East and West of the path (CDFig.105). The panorama over the North part of the Sokolitsa valley is good, as well as towards the hilly areas South of the valley. Scattered spots and strips are visible in the central, Eastern and Northern parts of the study area. Most of the sites in the Sokolitsa valley are visible from the path, as well as six barrows (surely five and one (Taniokoleva) with two out of four possible locations) and Polski Gradets tell.

In summary, the paths and their visual pattern from/to Mednikarovo tell repeat or confirm the observations made in the previous case studies.

Resources and land use

The size of Mednikarovo is recorded with different values in the available sources. As site size is a basic figure in building the SCA and especially the exploitation area, all the mentioned values were taken into consideration. The range of figures is summarised in Table 5.4.2.

Table 5.4.2 Estimation of exploitation area according to different site size estimates
Site area Population number Necessary crop Necessary arable land
1ha 125-168 26,250 –35,280kg 131 – 176 ha
1.8ha 225-264 47,250 – 55,440kg 236 - 277 ha
2.4ha 300-336 63,000 – 70,560kg 315 - 353 ha

Exploitation area

The distribution of soils around Mednikarovo, given in Table 5.4.3, shows a clear pattern of the patchy distribution of both smolnitsa and, especially, meadow soil within an area of 1500m from the site (CDFig.106). Considering this fact, two estimations of possible exploitation area were made – one using only cinnomonic forest soil as an arable resource and one using both cinnomonic soil and smolnitsa. Meadow soil was excluded as being rare within the 500-1,500m zone.

Table 5.4.3 Soil distribution around tell Mednikarovo
Distance from site Meadow soil Cinnomonic forest soil Smolnitsa Artificial soil Initial pedogenesis No soil
0-500m 23ha 54ha 3ha -    
500-1000m 13ha 115ha 84ha      
1000-1500m 9ha 282ha 72ha      
1500-2000m 79ha 294ha 61ha      
2000-2500m 197ha 258ha 35ha      
2500-3000m 237ha 298ha 40ha      
3000-3500m 130ha 284ha 114ha 33ha 30ha  
3500-4000m 68ha 338ha 146ha 53ha 57ha 4ha
4000-4500m 115ha 339ha 171ha 81ha 43ha 73ha
4500-5000m 80ha 311ha 259ha 90ha 32ha 124ha

The data in Table 5.4.3 indicates that, if the site size was 1ha, the area within 500 to 1500m from the site contains all necessary arable land. This area could also provide sufficient arable if the site size was 1.8ha but assuming total deforestation. However, since this scale of deforestation is unlikely, the exploitation area was probably up to 2000m from the site. The area within 500 - 2000 m has sufficient cinnomonic forest soil to sustain a population over 300 persons (2.4 ha site size) and still not suffer from severe deforestation. Within this area, the percentage of meadow soil increases because as the main valley of the Sokolitsa falls into the site catchment.

The model incorporating smolnitsa cultivation shows that this should have had some impact on the exploitation area only with a population size of 236-277 (1.8ha). If smolnitsa was cultivated, this would reduce the exploitation area to 1500m from the site, instead of the 2000m required for solely cinnomonic forest soil cultivation. If that was the case, this involves some high-effort agriculture because of the already discussed particularities of smolnitsa cultivation. The area 500 - 1500m contains sufficient arable land to sustain the higher population of 315 - 353 persons but assuming total deforestation. There is no evidence for such a severe process, which means that the exploitation area may well have been expanded to 2000m from the site. In practice, this is the same as if only cinnomonic forest soil was cultivated, which makes the issue of possible Neolithic cultivation of smolnitsa difficult to evaluate.

In summary, the area up to 2000m from Mednikarovo facilitates long-term and sustained mixed farming subsistence for a wide range of possible inhabitants – from 131 to 353.

5.4.3 Summary and discussion

The model of investigation applied to tell Mednikarovo does not allow conclusive claims to be made but some general comments on the deposition patterns at the site are possible.

Evidence for house burning is scattered but still gives some support for the idea of the emergence of this possible deliberate social practice. The late Neolithic house floor in sondage 3 had traces of fire but was not overlaid by burnt house rubble. If this inconsistency is a result of the type of the investigation, we can assume that there was fire in situ and this was a typical case of house burning. On the contrary, if the excavation situation represents a de facto deposit, this should mean that the burnt rubble was removed for some kind of subsequent use. One possible secondary implication of the burnt daub is for surface levelling – a case that has been documented on Mednikarovo itself. Thus, a practical issue to make the new building surface may even have integrated the social issue of successful social reproduction in which the link with the ancestors is seen through possession of a fragment of their house. Support for such intentional use of burnt daub comes from pit 5, which contains pieces of burnt house rubble. The very intriguing situation in sondage 3 demonstrates an unburnt dwelling floor overlaid by burnt house rubble. It is possible that the floor was artificially covered by house debris in order to imitate a real house fire or that there was some specific house burning technique in which only the walls were affected. Unfortunately, given the present state of the data, no further comments are possible on the pattern of house burning at Mednikarovo. It is evident, however, that fire has played an important role at the site and, although accidental fires should not to be excluded, the use of secondary fire products points to some deliberate and managed social practice.

Another common practice on the tell is structured deposition by pit digging. There is no secure evidence that such activity has started in the Neolithic but the five Middle/Late Neolithic vessels found in pit 1 suggest that, maybe at the end of the Neolithic occupation at Mednikarovo, structured deposition in pits had already become a deliberate social practice. It continued during the Copper and BA, as confirmed by the data from sondage 6. Whether pit digging was within a settlement context or the site was some kind of a pit-field is difficult to establish. In any case, however, exchange with the ancestors was present – Neolithic sherds found on the BA surface and BA soil and objects in features dug into Neolithic layers. This is a rare example where exchange of objects between different periods can be securely demonstrated.

Deliberate fragmentation is very difficult to document on such a heavily cultivated site but there are indirect evidence suggesting pottery breakage, which was not the result of intensive ploughing. First, the vessels from pit 1 were restorable but not whole. Secondly, the surface and the edges of the majority of the sherds were so heavily worn that I should doubt this was a result of 50 years’ modern ploughing. Deliberate fragmentation and re-distribution of pottery is known as trizna in Bulgarian archaeology (see p. 70) and it is usually connected with burial and memorial rites. If some whole vessels were able to survive after a devastating cultivation (e.g. the vessels in Pit one, that is 52 cm below the present surface, and the vessel in Sondage 6), this could mean that the sherds we find today may have been deposited as fragments in the first place and that modern ploughing has contributed to their additional fragmentation and wear. This is not to say that past and present human activity did not damage possible whole objects deposited on the tell; rather, it is to assess the evidence for possible initial deposition of fragments on the site. Apart from the sherds, direct evidence for fragmentation practice is sparse. There were just a few fragmented objects- two altars, two bone awls, two grinding stones and a weight. Although not numerous, the presence of these useless-once-broken objects suggests if not deliberate fragmentation then the deliberate keeping of fragments. As long as it is assumed that Mednikarovo was not a settlement during the Chalcolithic and BA (see below), it seems probable that the fragments were deliberately brought and deposited on the tell as part of a social practice of personal enchainment (Chapman 2000).

The type(s) of Copper and BA occupation on Mednikarovo should be envisaged in the context of tell formation. As stated above (p.71), the settlement dynamics and site-formation processes of Bulgarian tells have received little attention. However, general observations on the scattered data on tell stratigraphies reveal that a high proportion of multi-layer sites became mature tells during the Copper and/or Bronze Ages. At the time of the latest Neolithic occupation at Mednikarovo, there were very few tells higher than 4m in the whole of the Upper Thracian Plain (three examples are Karanovo, Klisselika and Kapitan Dimitrievo). Some of the sites during the late Neolithic shared the height of Mednikarovo (e.g. Ezero, with 2.60m), others were in the initial stage of possible tell development (e.g. Komunalni uslugi and Hlebozavoda near the town of Nova Zagora, both with 1-1.10m-thick Neolithic layers: Kunchev & Kuncheva 1988) yet never developed into a tell. Thus, the site of Mednikarovo was not an exceptional settlement type for its time and did not develop into a mature tell because of the lack of subsequent Copper and Bronze Age occupational layers. It is also possible that the height of the site was reduced by later severe destruction. Indirect evidence for such damage is the lack of any visible sign and/or features of the Late Medieval/Pre-modern cemetery. In addition, the digging of the grave pits would have contributed to the destruction of the late occupational levels of the site. The current condition of the data, however, does not support a process of the widespread removal of settlement layers. Instead, I would suggest that Mednikarovo is an adolescent tell that has become the focus for other types of human activities during its post-Neolithic biography. Pit-digging and/or surface deposition of pottery have taken place on Mednikarovo, thus including the site within Chalcolithic and BA social networks as a place for possible ancestral rites and rituals.

5.5 Karaivanovi mogili barrow cemetery (KMBC)

5.5.1 Earlier studies and present condition of the data

Karaivanovi mogili barrow cemetery is located cca 2.5 km South East of the present village of Mednikarovo. It consisted of three barrows but archaeological investigations were undertaken at just one of them. Despite the co-ordinated efforts of archaeologists and mining managers, two of the three barrows were destroyed by the mining work prior to their planned excavations. Half of the third barrow was also destroyed. In 1974, the remaining part of the barrow was excavated. The site was not published and the current data derives from the written investigation report that contains no illustrations at all.

According to the report, the three barrows were 15 - 20m apart, situated in a line on the high right bank of the Karapelitska stream. South of them, on the very bank of the stream were traces of a large Classical/Roman settlement. On the left bank of the stream there were two significantly larger barrows. The concentration of sites – 5 barrows and a settlement - led the investigators to conclude that these were interrelated Roman sites belonging to one and the same complex. However, under the 4 secondary Roman cremations dug into the mound, there were two earlier graves. The latter were in the centre of the barrow and one of them was a child burial. Both skeletons were found in crouched position on their left side, with the head pointing to the West. The depth of the graves was 30 cm. from the present surface, although whether the deceased were placed in a pit or on the surface was not specified. The grave goods of the child burial comprised a spindle whorl and two fragmented jugs. The other grave contained fragments of one vessel. Stylistic parallels for the decoration of one of the jugs – incised net-like ornaments, with triangles, spirals and concentric circles - have dated the graves to the Late Bronze Age. General similarities for the pottery were found in the areas of Central and North West Bulgaria, as well as in the Tei culture of Muntenia, South East Romania.

5.5.2 The site and its surroundings according to the GIS analyses

Karaivanovi mogili barrow cemetery is located on a hill at 189-213 masl (CDFig.107). It is on a 2-3° slope (CDFig.108) with a South Western aspect (CDFig.109). Visibility from the site is very low, mainly to the West and South of the barrow itself (CDFig.110). There are some visible spots at 7 km and 10 km to the North West. No sites were visible at all.

A second viewshed was run with an additional 5m, as the actual height of the barrow is not known (CDFig.111). There was just one visible site – barrow 2 from the Mednikarovo-Iskritsa barrow cemetery – but the overall view has significantly increased. In addition to the previously visible areas, now there was an almost continuous panorama of a 1.7-km-wide zone North of the Sokolitsa valley.

So if the barrow was 5m high, that should have provided a good view over the landscape rather than the sites. Apart from MIBC2, a LBA site north of Gudgova tell may have also been seen, since the area North of the tell is visible. However, the presence of such a settlement is not confirmed and further comments cannot be made due to the paucity of the data.

The cost surface analysis (CDFig.112) is summarised in Table 5.5.1:

Table 5.5.1 Site distribution around KMBC
N of cost strip Sites located in the cost strip
1 Atanasivanova mogila, both Iskritsa sites, Mednikarovo, Klisselika and Gudgova tells
2 Obrutchishte, MIBC
3 Galabovo tell, Kurdova mogila, Taniokoleva mogila 2-4
4 Taniokoleva mogila 1, Tcherniova mogila – all locations, Goliama Detelina flat site, Manchova, Goliamata, Malkata and Ovchartsi barrows, Barrow 4
6 Aldinova barrow, Polski Gradets tell, Ovcharitsa I
7 Ovcharitsa II, Gonova barrow, Polski Gradets pit site

According to the cost surface analysis, KMBC and its possible contemporary sites are located in areas that need substantial efforts (a day’s walk at most) to be reached. On this basis and together with the complete lack of visibility from the barrow, I should assume that KMBC was related to the areas along the valley of the Sokolitsa and the foothills of Sakar Mountain immediately to the South – a zone that falls out of the present study region. Recently, several LBA pottery scatters were reported from that area (Expeditsia Maritsa Iztok n.d.).

The route network from Karaivanovi mogili barrow cemetery is an important justification of the previous claim of a consistent repetition of routes (CDFig.113). In the case of KMBC, the routes start from the Southernmost part of the study area, leading to the main direction of movement along the North-South axis, rather then to the West-East axis, as in the most previous cases. It is apparent from CDFig.113 that once the route descents into the Sokolitsa valley, it follows the already existing paths to the East and West. The routes to the North coincide with the tracks from Atanasivanova mogila. This is an important confirmation of the presence of a long-lasting network of tracks, in which the main routes follow the valleys and there are individual paths to/from the sites situated in some distance from the valleys. Usually but not always, the traffic trend is along the main routes. In the cases when pairs of sites - both away from the valley - should be connected (e.g. KMBC- Manchova mogila), the path between them crosses the main route and follows only the individual tracks to/from the site, respectively to/from the valley. As several cost surface case studies have shown, these individual paths share recurrent outlines despite the different initial staring and/or destination points. Therefore, they could be considered as secondary routes that may have lost their importance once the destination site was abandoned. But they may also have been used as pilgrimage routes to earlier/ancestral sites by the later inhabitants of the landscape. In the case of the Karaivanovi mogili barrow cemetery, the route network shown on CDFig. 113 make sense only in the context of such sacred trips to earlier sites, since all the contemporary sites are in the Northern part of the study region and access to them is along the main North route. In order to reach the contemporary LBA sites, once in the valley one should turn left and head West and later North West to cross the contemporary mining area and to join the route 4 km to the North East of Galabovo. From this route, there are individual paths to each LBA site, whose track and visibility is discussed in the Galabovo case study.

The KMBC logistical network has one new path – to Mednikarovo tell - and two paths with minor differences from the Atanasivanova mogila logistical network – the path to Atanasivanova mogila itself and the path to both Iskritsa sites.

The path to Mednikarovo is not entirely new, since it appeared in the Mednikarovo logistical network but should be discussed here because its possible use may have started after the emergence of the KMBC during the LBA.

The path follows the ridge of the hill to the North West for one km and then turns left, descending to the West for 1.1 km (CDFig.114). The areas North of the Sokolitsa valley are visible from the path, as well as the region South of the path itself. Atanasivanova mogila, MIBC2 and Iskritsa pit site are visible from the path (CDFig.115). It is important to point out that the possible paths to the sites situated in the same North West direction (Obrutchishte and Galabovo) do not pass by Mednikarovo tell but rather use the main South route. It is a confirmation of previously discussed advantages of GIS, in which the presumed least-distance routes are not as effective as the least-cost routes. It is also worth noting that least cost is not only an economical category but also has meaning in terms of bodily experience- e.g. a pilgrimage in which the load should not be broken or spilt; or on a matchmaking trip, the participants should stay in proper/ specific condition.

Such an absence of a chain of sites along one and the same route is confirmed by the path to Atanasivanova mogila. Although the mud volcano is situated 300m from the main South route, there is no branch from it that leads to the site. Rather there is a separate path from KMBC to Atanasivanova mogila, although it runs parallel to the path leading to the Sokolitsa valley (and the main South route) (CDFig.116). There is only one path from KMBC that descends due South and then bifurcates after 1.6 km. The left branch is for Atanasivanova mogila, and it continues to descend to the North West for another 1.6.km, the straight branch leads to the main South route, while the right branch descends to the North East for 1.2 km until it reaches the Iskritsa pit site (CDFig.117). The latter is the last minor difference in the track network from the Atanasivanova network.

The segment to the Iskritsa dwelling site forks from the path to the Iskritsa pit site 500 before its destination and heads North East for 500m (CDFig.118). The landscape visibility from the paths to Atanasivanova mogila (CDFig.119) and both Iskritsa sites is very similar (CDFig.120), (CDFig.121) – a good panorama over the Sokolitsa valley and its Northern areas and scattered visible area towards the hilly Southern regions. The view from the paths to both Iskritsa sites is slightly better towards the Eastern parts of the valley that assures better site visibility. Only MIBC2 and the Iskritsa pit site could be seen from the path to Atanasivanova mogila, while all the sites in the valley plus MIBC2 are visible from the paths to both Iskritsa sites. It is important to point out that the nearby Mednikarovo is not visible at all: even more – the site is in the middle of an invisible island area.

The specific visibility status of Iskritsa is confirmed in the KMBC viewshed analyses. These results contribute to the visual restrictions of the site - this time from the West. Final comments on the site visibility will be made in section 5.6.2.

The only track and its visibility that has not been discussed in any of the previous case study is KMBC - Galabovo tell. The path is, in fact, a combination of the segment that goes from the cemetery to the valley and the main South route following the East - West direction (CDFig.122). The panorama from this track is relatively good as apart from the whole valley and the sites in it (apart from the Iskritsa dwelling site), some spots South and North of the plain are also seen (CDFig.123).

5.5.3 Summary and discussion

KMBC is the only example of the creatiuon of new barrows in the LBA. This may be interpreted as a deliberate attempt at the monumentalization of the place, after successful colonisation of the landscape. The site is the Southernmost of all and as the recent investigations have shown, the areas South of the study region (Sakar foothills) were mainly settled during the LBA and the IA.

Another peculiarity of this barrow is the child burial and the Tei pottery found in the grave, which resembles the evidence from Grave 27 in the Drama microregion (see below, p. 329), which also contained a child buried with a Tei jug. It is possible that the two children were related to each other and/or to a third person. But it is also possible that a specific burial practice was followed, in which children with certain social status should be buried with exactly this type of exotic pottery. Given the present condition of the data, conclusive claims are difficult to make but such similarity constitutes strong evidence that, during the LBA, the people living in the study microregions were in contact with each other, as well as with more remote areas.

5.6 Iskritsa flat site

5.6.1 General information and earlier studies

In 1988, during the spring field survey of the Maritsa-Iztok Expedition, scattered prehistoric pottery was found over an area of 0.15ha near the contemporary village of Iskritsa. The site was located on the left bank of the river Sokolitsa, on three neighbouring low hills (AFig. 5.6.1a). Later in 1988, four sondages were excavated on the Easternmost hill, which had the densest Medieval and prehistoric pottery spread (Borisov 1989). In the next year, excavations were continued on the other two hills, where a Medieval settlement, fortress and cemetery were found. The place was also occupied during the Iron Age (Sheileva 1994). Investigations on the Eastern hill were renewed in 1992, when three new trenches were laid out near the previous sondages. Burnt rubble and two floor levels were found during the excavation that made investigators interpret the feature as a house (Leshtakov n.d.b). On the Westernmost hill, among the Medieval graves, at least 10 prehistoric pits were excavated (Sheileva 1994). Thus, the current interpretation of the prehistoric site near Iskritsa is that it consists of two sites - an Early Chalcolithic pit site (Iskritsa I) and a Late Chalcolithic settlement site (Iskritsa II) (Leshtakov et al. 2001). The end of the settlement was connected to the eruption of a mud volcano. On the basis of the results of my own research on the data from Iskritsa, I would dispute both of these claims.

Archaeological evidence

The following site description summarises the prehistoric data from all excavations at the site, as well as the results from my museum study in 2000.

At the so-called Iskritsa II site, two pits and a burnt house were excavated. The surrounding general cultural layer consisted of sand, gravel, clay, burnt house rubble, charcoal and pieces of daub. Two fragments of cult vessels, 14 flint tools – three from the surface, 11 from different depths and locations in the trenches (AFig. 5.6.8), a small adze(AFig. 5.6.5 I), a fragment of a bone needle and a complete small dish (AFig. 5.6.5 K) were discovered during the excavations. Sherds (AFig. 5.6.6-7), a bovine skull, fragments and whole animal bones, such as long bones, ribs, vertebrae and jaws, complete the contents of the cultural layer.

The dwelling contained two occupational levels, each marked by beaten clay floors. Three postholes were also found. Burnt house rubble was spread all over the area of the sondages. The stratigraphy of the burnt feature was not coherent. In the Eastern part of the structure, the two floors and the rubble were relatively intact, having sunk into a fault and were covered by clay and gravel (AFig. 5.6.2a). The West side of the feature was severely folded and, all around it, there were traces of long-lasting surface exposure. Some of the house rubble into the fault was not fully fired.

Two almost simultaneous activities were given as an explanation for this unusual stratigraphy. Together, or soon after the burning of the house, the mud-volcano erupted and opened a fault into which the East side of the dwelling had sunk, while the West part left on the surface and was subsequently folded. The clay and gravel from the eruption sealed the floors and the plasters in the fault, thus preventing them from complete combustion (Leshtakov n.d.b.).

Two pits were found in the vicinity of the house. The first one contained two bovine skulls, one on the bottom, and the other 10cm from the top of the pit. The lower jaw was missing from the latter, which had a large piece of charcoal placed on the forehead. The pit was filled with crumbly black soil, mixed with sherds and a few animal bones.

The second pit was filled with reddish sand and gravel, without any finds.

The occupation of the Iskritsa II site in the Late Copper Age was claimed on the basis of the sherds found on and above the dwelling floor (Leshtakov et al. 2000:18)(AFig. 5.6.2; 5.6.5a).

During my museum study, I looked at 3 out of the 13 kg of pottery from the burnt house. It contained more Late Chalcolithic than Early Chalcolithic sherds of both fine and coarse ware. There were two vessels that had more than 20 fragments of their rim and body but were still not complete.

The Early Chalcolithic Iskritsa I site was 200 m to the West, on the Westernmost hill. Among the Mediaeval graves, there were up to 10 pits with prehistoric material, mainly concentrated in the North part of the hill (AFig. 5.6.1b).

Pit N10 was disturbed by a Medieval grave. It was 90 cm in diameter and 35 cm in depth. The bottom of the pit contained a thick, compact clay soil, mixed with lots of charcoal and very few sherds. This layer was covered by 10 cm of black gray crumbly soil, mixed with charcoal and decayed wall daub or ceramics, that have coloured the earth with scattered red spots. The small amount of sherds from the pit was claimed to be uncharacteristic but generally assigned to the Copper Age.

Pit N12 was 1m in depth and with a diameter of 1.80/1.70m. It was filled with black-gray compact soil, mixed with small pebbles, small pieces of burnt daub, sherds and animal bones.

The bottoms and sides of pits N 15 and 18 are carefully plastered with clay. On the bottom of pit N 15, there was a thin strip of ash and charcoal over which there were spread broken vessels. They were covered by a brown-yellow soil mixed with lots of sand. The latter was overlaid with a 1-mm strip of ash and charcoal, with prehistoric sherds on top of them. The sequence finishes with a 0.30 m-thick grey-whitish soil with both tiny and large pieces of burnt clay. The pit is cylindrical in shape with an upper diameter of 0.67 m and a depth of 0.55 m.

Pit N 18 was piriform with an upper diameter of 0.83m and a basal diameter of 0.90m. Its depth is 0.65 m in the West part and 0.60m in the rest of the pit because of the displacement of the terrain. The sequence in the pit started with yellow clay containing lots of charcoal. This was covered by a 0.10m-thick layer of sandy soil. The uppermost 0.22m-thick layer was grey –whitish in colour, with lots of tiny charcoal fragments and pieces of burnt clay. Among the sherds, two restorable but incomplete vessels were found. Two fragments of flint tools were also excavated.

The oily black-grey clay layer with lots of charcoal and single sherds at the bottom of the pit N 11 was interpreted as a possible pit plastering. Above it, there was a 0.15- 0.28m-thick black-grey crumbly layer, containing numerous animal bones and sherds. The pit was cylindrical in shape, with an upper diameter of 1.20m and a depth of 0.36m. An adze, a fragment of a flint tool and sherds of three restorable vessels were found.

Pits N 20 and 21 were oval in plan and with uneven bottoms due to the displacement of the terrain. Pit N 20 had an upper diameter of 1.30 to 1.90m and was filled with red-brown compact soil with burnt daub and charcoal. Several boulders were found at different levels in the pit, as well as upper and lower parts of grinding stones. There were also a few sherds and animal bones.

The fill of pit N 21 (diameter - 1.35/1.22 m) was the same as pit N 20, but some pebbles were present, too. A few sherds and animal bones were found as well.

Pit N4 contained sherds of pottery that belong to one technological group (N. Todorova, n.d.). The fabric was very sandy with three kinds of filler - lots of mica, rare fine organic material and tiny pieces of ochre. The pottery was not very well fired and had sporadic traces of self-slip. Long-distance parallels in the Cucuteni-Tripolye area, the Aegean and Anatolia were claimed for the vessels and they were dated to the very end of the Late Copper Age (Todorova, N. n.d.).

These observations are important since they are not valid for the content of the other pits. Pit N4 appears to be an exception, as all the remaining pits contain both coarse and fine ware and a variety of decoration patterns and shapes. This suggests the deposition of similar materials in several pits, with special, fine and different pottery in others.

An important note in Todorova’s study of pit N4 concerned the surface of the sherds. They confirmed my own observations on the material from pit Nos. 15, 16, 18 and 21. The sherds from these pits were very heavily worn on both their outer and inner sides, as well as on the cross-section. Todorova suggested that this was maybe due to the post–excavation washing. Since, I found unworn (but not unwashed !) Medieval sherds among the extremely worn fragments from pit N 16, I would rather conclude that the prehistoric sherds were exposed to the open air for a long time and then deliberately re-used as a component of the pit fill.

The content of the other pits (Nos 15, 18 and 21) did not conform to such a hypothesis, since there were many heavily worn non-characteristic fragments and just one or two sherds from each pit with secure evidence for prehistoric date. This was the case with pit N15 that was claimed to contain many sherds dating to the Early Copper Age. When I studied the material, I could find just one sherd of clearly Early Chalcolithic date (AFig. 5.6.3P); the other 74 small pieces of fine ware and 10 rims were absolutely unsuitable for dating.

The content of the pits was not published by context and their ECA chronology is based on the very few published vessels and sherds with typical Early Chalcolithic (viz., Maritsa) decoration (AFig. 5.6.3).

The stratigraphic sequence as described in the excavation diary, which lacks plans and sections, does not show severe disturbance of the pits and I should suggest that some of them were dug and filled during the Copper Age (e.g., N4 and N15). Others date from the Medieval occupation, when earlier pottery was dug out during the digging of grave pits (e.g., N16).

5.6.2 Plant remains

The archaeo- botanical study of 178 plant impressions on burnt house rubble has identified 94 samples of einkorn, 45 samples of emmer, 38 of hulled barley and one sample of vetch (Popova 1994).

5.6.3 The site and its surroundings according to GIS analysis

Both parts of the Iskritsa site are located on a terrace with a 1-2° (CDFig.124), at 115-140 masl (CDFig.125). The pit zone has a North Western aspect, the dwelling zone a Northern aspect (CDFig.126). The viewshed analyses of both zones show different patterns. They share a similar patchy view over the Northern parts of the valley but, from the pit zone, the area South of the zone is visible as well (CDFig.127). All the sites in the valley are visible from the pit zone except for the dwelling zone. Both tells located in the Eastern part of the Sokolitsa valley are visible from the dwelling zone (CDFig.128).

The cost surface analyses of both Iskritsa sites were almost identical (CDFig.129), (CDFig.130) and are united in Table 5.6.1:

Table 5.6.1 Site distribution around Iskritsa
N of cost strip Sites located in the cost strip
0 Atanasivanova mogila, both Iskritsa sites, Mednikarovo, Klisselika and Gudgova tells
1 Obrutchishte, MIBC , KMBC
2 Galabovo tell, Kurdova mogila, Taniokoleva mogila 2-4
3 Taniokoleva mogila 1, Tcherniova mogila – all locations, Goliama Detelina flat site, Manchova barrow,
4 Goliamata, Malkata and Ovchartsi barrows, Barrow 4
6 Aldinova barrow, Polski Gradets tell, Ovcharitsa I and II, Gonova barrow
7 Polski Gradets pit site

In summary, the site was located in an area with easy access to the earlier Neolithic sites and in addition, one of them (Klisselika tell) was even visible from both Iskritsa sites. The interrelation of the Copper Age sites along the Sokolitsa valley, in terms of cost, was also fairly quick accessibility. There was one exception – tell Galabovo needed more efforts to be reached. Tell Polski Gradets in the Northern part of the study region required a half-day trip to access the site.

Logistic networks from both sites coincide due to the similar cost surfaces (CDFig.131), (CDFig.132). There is a minor difference in the paths to Goliamata and Malkata barrows but they will be discussed in section 6.7.3. In general, the route network of both Iskritsa sites matches the KMBC logistical network. There are, however, some differences – the path to Mednikarovo tell is the same as from Atanasivanova mogila, there is one path to MIBC1/ 2 and another to MIBC3/4 (see section 6.14.2), the above mentioned path to Goliamata mogila, the path to Kurdova mogila (see section 6.13.2) but the most significant difference is the path to Polski Gradets tell.

The path to Polski Gradets tell goes Eastwards along the main South route for 3.2 km, then turns to the left crossing (ascending and descending) the hilly area Northwards for 6.9 km following the gentlest sloping hills and finally ascending and descending to the East North East for five more km (CDFig.133). The visibility from the path is very good over the hills in the Eastern part of the study region and to the South over the foothills of the Sakar mountain (CDFig.134). The view over the Sokolitsa valley is patchy, as well as over the areas West of the path. There is consistent visibility to the Northernmost and North Western parts of the study region. If the path was initially used during the Copper Age, all the earlier and contemporary sites in Sokolitsa valley were visible from the path. However, seven Bronze Age barrows were also visible from the path. It is likely that, while walking along and across the landscape, the people have noticed visible places in the landscape, which were later used for the location of sites whose visual aspect was crucial.

Both Iskritsa sites are on the main South route, so the paths between the sites in the valley and their visibility are discussed in Galabovo and Gudgova tell case studies. The recurrent invisibility of the Iskritsa dwelling site from most of the paths and/or sites followed a certain pattern - not visible from paths leading to the Southern (Mednikarovo, KMBC) and the Western (Galabovo, Atanasivanova mogila) areas. The dwelling zone was only visible from the path to/from Klisselika and Gudgova tells. Summing up the evidence from the paths along the valley (already discussed in the previous sections or forthcoming), it is possible to say that a part of the site was visible from the West and South only when approached at close distance, while there were no visibility restrictions from the East and the North.

5.6.4 Summary and discussion

Summarizing the above evidence, it is likely that feasting, the breaking and deposition of pottery (trizna) and structured deposition in pits was a common social practice at Iskritsa, just as at Mednikarovo. Pit deposition most probably started during the Early Copper Age and the consumption and/or deposition of ritual food may have accompanied the event. The same activity was continued during the following centuries. In addition, the surface deposition of pottery fragments was practised and a building for deposition was constructed. One possible reason for the emergence of the building may be the deliberate monumentalization of the place, in which cultural inscription onto the landscape is accomplished through the erection of a positive feature in contrast to the negative features distributed on the site (the pits). Thus a specific entity is created in which the ancestors (the pits), the present occupants (the surface deposition) and the descendants (the building remains survive even the death of its builders) are harmonized in the eternal landscape.

The place on which the building was constructed was specially chosen to be visible only for people in the close vicinity of the site. I would agree with the late Chalcolithic chronology of the building, as long as two floor renovations and a 1m-thick cultural layer are more likely to be a result of a few hundred years of human activity (within the Late Chalcolithic), rather than a millennia (during the whole Copper Age). The presence of Early Copper Age pottery (AFig. 5.6.4) in the burnt rubble suggests a long-lasting ancestor cult, in which personal, household or communal enchainment with the previous inhabitants of the landscape was crucial for successful social reproduction. It is likely that ECA sherds were deposited on the surface and/or in pits below or under the place where the building was erected, which later were deposited in the ready building. But it is also possible that the Early Chalcolithic sherds were kept at settlement sites and deliberately brought and deposited at Iskritsa during the Late Copper Age. In both cases the link with the ancestors appears to be an important issue during the Late Chalcolithic.

The end of the building was not a result of devastating natural process but rather an intentional and managed burning of the feature. The presence of unburnt together with burnt rubble in one and the same in situ context is strong evidence for managed fire. I should also suggest that the house was deliberately burnt as part of a rite of passage, in which killing (burning the old house) is followed by re-birth (the construction of a new house). Indirect evidence for such a cycle is the renovation of the floors of the burnt feature. Given the present state of the data, it not possible to explore the character of this internal transition of the building. After the managed fire event, the building was not re-built because of the eruption of the mud volcano. The latter was not necessarily a rapid and devastating process (see above p. 114) and therefore probably did not cause the house destruction. What it prevented, however, was the subsequent occupation of the site. The next traces of human activity are from the end of the Bronze Age onwards.

The evidence from the pit zone has revealed that the latest inhabitants (AD IX-XIIth centuries) treated the earlier material carefully and with consideration. In spite of the controversial data, it could be inferred that there were intact prehistoric pits (N4), prehistoric pits with subsequent disturbance (N10, 18) and post-prehistoric pits containing prehistoric material (N16).

The long-lasting occupation and site formation of Iskritsa is oversimplified by following a currently favourable continuity explanation (see above, p. 56). As already discussed, continuity is both the reason for, and an explanation of, recurrent long-term site occupation. In addition, the lack of formal or commonly agreed definition for prehistoric settlement led to the interpretation of the evidence from Iskritsa as the result of settlement activities on the basis of one single burnt house. Observations on the pottery sherds from Iskritsa during my museum study and in particular, on deposition patterns at both Iskritsa sites, were crucial for the reconstruction of the site dynamics of foundation, abandonment and re-occupation.

During all the investigation seasons (1988-1994), a total of almost 8ha was excavated. A single prehistoric house was found on the Eastern hill only, as the area around the house was surveyed but not excavated. The lack of any other prehistoric buildings was taken to reflect limited excavation and/or later destruction (Leshtakov n.d.b). I would challenge both conclusions. First, within the 8 ha investigated area traces of prehistoric occupation were found, which has the following implications – a) if there were prehistoric house rubble, it should be noted during field-walking as well as excavation; and b) pits should not be separated from the social practices leading to deposition on the Eastern hill. Secondly, if a 1m thick layer can survive a mud-volcanic eruption and subsequent Medieval and Modern destruction, then any other prehistoric settlement activity (presumably 1m thick), if present at all, should have left similar traces. Therefore, I would assume that the prehistoric site at Iskritsa consisted of one building and several pits. Such a combination of features is not considered to be typical for Bulgarian prehistoric sites and I would suggest that Iskritsa was a place with special meaning, for the enactment of significant social practices.

Both Iskritsa sites contain evidence for such practices, which are usually named as non-utilitarian or sacred. According to their understanding in current studies (Brück 2000; Brück and Goodman 1999), these are elements of contemporary habitus in which the very act of fragment deposition, pit digging or house burning emphasises some current social issue(s) but at the same time is indivisible from the long-term attitude of reverence for their place and their ancestors. Return journeys to the place where once the ancestors have started the practice of surface and pit deposition add value to the place. In turn, the place constitutes additional specific meanings for any activity held on it, thus providing an area for (re) negotiation of social issues, for pilgrimage, worship and devotion. The reason for the initial choice of this particular place is difficult to reconstruct. However, an assumption for the possible attraction of the place could be made on the basis of past and present environmental phenomena in Maritsa Iztok. The river Sokolitsa is well known for the coal seam in the profile of its banks. Some of them were still visible around Iskritsa even a few years ago. A characteristic feature of the coal in Maritsa Iztok is their spontaneous bursting at the very moment of the first surface exposure when they come into contact with oxygen. This is not a devastating process, usually producing with slow-burning embers and smoke (pers. comm. P. Karacholov). So it is likely such spontaneous mini-eruptions took place near Iskritsa when communities have already inhabited the landscape along the Sokolitsa valley. Indeed, the toponym Iskritsa is a diminutive form of Iskra, which means sparkle. The illumination effects and the smoke may have attracted people’s attention and, after the active process has stopped, the place where the natural phenomenon had happened became a sacred place. As with Atanasivanova mogila (see above, p. 192 -197), the visual properties which attracted people to this place were transformed into a cultural statement.

5.7 Klisselika tell

5.7.1. Earlier studies and present condition of the data

The tell of Klisselika is located immediately North of the modern village of Mudrets (AFig. 5.7.1a). It was firstly investigated in the early 1970s, when M. Dimitrov made some soundings/trenches in order to establish the stratigraphy and chronology of the site. The results of these excavations have never been published and the archaeological material that was found has restricted public access. Prior to the Maritsa-Iztok expedition, it was known that the site was founded in the Early Neolithic, most probably was abandoned some time in the Early Copper Age and re-used during the Medieval times when it was turned into a cemetery. In the late 1950s, the South end of the tell was cut by agricultural amelioration work. As a result, the present bed of the river Sokolitsa passes through the site, thus forming a natural profile of the archaeological sequence. In 1998, the investigation of the tell was renewed and its aim was to clarify the stratigraphy and chronology of the tell using the earlier archaeological profiles, as well as the site exposure left after the amelioration work. Adverse weather conditions and restricted funding prevented the team from finding and documenting in situ remains and only the chronological aim was partly met. It was confirmed that the site occupation has started during the Early Neolithic (Karanovo I); four occupational levels were claimed to be present at the approximate depths of 2.50-2.70m from the top of the tell. The total height of the tell is not mentioned in any of the reports or publication of the site. My own observations (without any surveying equipment) made me conclude that the levels of pebbles overlain by white clay and interpreted as dwelling floors by their excavators are at the depth of 4.50 - 5m below the top of the tell. The uppermost 2.50 - 2.70m were most probably occupied during the Middle and Late Neolithic and Early Copper age, as unstratified sherds are known which date to these periods (AFigs 5.7.1b –5.7.5).

The main field technique used during the new investigation was a control profile along the exposure cut by the river. The profile was 7m long and 2m deep. It is difficult to evaluate the amount of soil that was removed to clear the profile but the layer of humus removed from the tell was 10-15 cm thick. During my museum study in 2000, I was able to count 758 body sherds, 87 fragments of rims and 106 fragments of bases from the total excavated area of 10-14m2. They confirmed the above stated chronology and derived from both fine and coarse ware. In 10 out of over 100 storage units (plastic bags), it was possible to identify fragments from one and the same vessel, which however did not make a complete vessel. Five pieces of daub were also found during the new investigations. Burnt daub was found during the field survey on the opposite bank of the river, which was supposed to belong to the tell area prior to the moving of the riverbed. Sherds and bones are the other finds across the surveyed area, which now suffers from long-lasting and intensive cultivation. The museum storage bags contained several pottery objects that had traces of a large quantity of organic material (most probably straw) in the clay fabric. As they were not whole, it is difficult to assume their function but the shape suggests some kind of weights (loom and/or net). Another 6 kg of pottery sherds was also excavated during the renewed investigations. Over 500 animal bones or fragments of animal bones derive from this relatively small excavated area. Together with my observations from other tell excavations where animal bones were not found with such frequency, the latter suggests that the excavated area was some specific area for depositing food remains, including animal bones. However, given the present state of investigations, conclusive claims cannot possibly be made.

The last finds class to be mentioned in this short section on the archaeological evidence from Klisselika tell are the stone tools. Special investigations have not been undertaken and claims for the kind of raw material, as well as the tool types, are made according to the general knowledge of the team members. Fifteen flint artefacts altogether were discovered during the recent excavations. Nine of them were called tools, four were plunging blades, one scraper and a core. The latter, in fact, was a black opal core. During my museum study, I found a further 57 pieces of opal deriving from the tell. They were of different colours, mainly black, and of varying shape, size and stage of erosion. Another five quartz tools are present in the museum storage units, as well as three tools of an unknown type of stone. Eight flint flakes, one tool, 5 plunging blades and a fragment of a translucent flint tool complete the assemblage. It was claimed that opal was used for tool production instead of flint, as the former was abundant within the study area, in contrast to the availability of the latter (Leshtakov et al. 2001). Indeed, such a concentration of opal tools and raw materials is an important indicator for the potential use of this mineral as a flint substitute.

5.7.2 Plant remains

Archaeo-botanical study of the charred macrofossils and plant impressions has revealed that the main cultivated species at Klisselika tell were T. dicoccocum and peas (Popova 1985). A detailed list of species and the context within which the plant remains were found have not been provided.

5.7.3 The site and its surroundings according to the GIS analyses

Tell Klisselika is located on a terrace at 140-164 masl (CDFig.135), most probably close to a palaeo-channel of the river Sokolitsa. It is on a 1-2° slope (CDFig.136) with a Southern aspect (CDFig.137). The actual size of the site is difficult to establish due to various past and present destructions, that is why some approximate estimations were done pointing that the tell was not higher then 10m and its area varies between 1 - 1.56ha.

Visibility from the site is very good over the immediate surroundings within a 1 - 1.5 km radius (CDFig.138). Only to the North West of the tell was the visibility patchy and generally not very good. The same patchy view is valid for the area 8-9km to the West along the valley. The neighbouring Gudgova tell is visible from the site but since the former is later than Klisselika tell, the intervisibility was most probably important in the foundation of the first settlement of Gudgova tell. Both Iskritsa sites are visible from Klisselika but such visibility would have started to be an issue in the final stages of the tell occupation (early Chalcolithic), when the pits and the buildings at Iskritsa were built.

Ten meters were added to the terrain model surface of the site that should correspond to the height of a mature tell. Viewshed analysis run with this additional height shows the same general visibility over the valley but is less patchy in comparison to the first viewshed analysis (CDFig.139). The panorama over the surrounds of the site is much better, reaching almost 3km to the South, South East and East but still does not exceed 1.4km to the North. MIBC2 became visible and, as in the Gudgova tell case it may have some importance in terms of intervisibility, when the barrow was founded, since the latter did not existed during the time of Klisselika habitation.

Cost distance analysis (CDFig.140) shows that equal efforts were needed to reach the two Neolithic sites along the Sokolitsa valley - Mednikarovo tell and Obrutchishte flat site (first cost strip), while the other Neolithic site – Ovcharitsa II, located in the Northern part of the study region (7th cost strip) could be reached only after a day’s journey. The long sequence of the Klisselika tell bears traces of habitation during the Middle and Late Neolithic, to which periods the other three sites were dated. Whether or not the sites were occupied contemporaneously is difficult to say at the present state of the investigation.

The distribution of sites intensifies during the Copper Age, as some sites are situated in area with quick access to Klisselika tell (both Iskritsa sites), despite the continuous pattern of dispersed location – Galabovo tell in the second cost strip and Polski Gradets tell in the 5th cost strip. The chronology of the Polski Gradets tell will be discussed in section 5.9.1 but it is noteworthy that its Neolithic date and hence a possible connection with Klisselika tell should not be excluded.

The route network repeats the already discussed main routes along the two river valleys (CDFig.141) (CDFig.142). The route that connects Klisselika tell with Ovcharitsa II is one of the examples demonstrated and discussed earlier for the GIS ability to identify least-cost rather than least-distance routes. The path follows the main South route to the East, crosses the study area to join the North route 4 km South East of Galabovo tell and then follows the route until it reaches Ovcharitsa II. Common sense logistic analysis that does not use GIS tool should outline a route that goes due North regardless of the landscape particularities (CDFig.143).

There is one path whose use during the Neolithic is feasible but not sure – the path to Polski Gradets tell (CDFig.144). It generally the same path that connects the tell with both Iskritsa sites but instead of starting to the East from the Iskritsa site, it starts from Klisselika to the West, following the main South route for 1.3 km before reaching the point at which the path ascends to the North. The landscape and site visibility is the same as from the Iskritsa path, which confirms the claim for the possible choice of places for site locations with regards to their visibility from earlier paths (CDFig.145).

The visibility along the main South route between Klisselika and the sites in the valley is almost identical to the visibility from Gudgova tell (a few more visible areas around Gudgova tell from the paths from the latter) and should be discussed in section 5.8.3.

The panorama from the path Klisselika - Ovcharitsa II shares first the visibility from the main South route and then the visibility from the main North route (CDFig.146). Tell Mednikarovo is visible from the path, which, together with the data form section 5.4.2, draws to the conclusion that, within the Neolithic route network, there was almost complete site intervisibility (except for Obrutchishte) from the routes connecting the Southern and Northern parts of the study area.

Resources and land use

The SCA for the Klisselika tell follows the pattern of the previous sites. Distribution of soil types around the site given in Table 5.7.1. and CDFig.147 shows that meadow and especially smolnitsa soils are spots within a consistent spread of cinnomonic forest soil. Smolnitsa soil is excluded from the following analysis of the exploitation area, since its distribution in the first three circles is insignificant and whether or not the areas were exploited does not influence the final figures for site exploitation area.

Table 5.7.1 Soil distribution around tell Klisselika
Distance from site Meadow Smolnitsa Cinnomonic
0-500m 35ha 5ha -
500-1000m 77ha 39ha 62ha
1000-1500m 113ha 34ha 237ha
1500-2000m 85ha 101ha 362ha
2000-2500m 92ha 161ha 451ha
2500-3000m 100ha 154ha 606ha
3000-3500m 81ha 116ha 812ha
3500-4000m 55ha 155ha 711ha
4000-4500m 99ha 120ha 770ha
4500-5000m 117ha 131ha 760ha

Exploitation area

The site population was difficult to estimate given the imprecise data on site area. For that reason, a range of values was used to generate a reasonable suite of estimates. If the site area was 1ha, the population should vary between 125 and 168 (following Russell 1956 and Todorova et al. 1983). If the site area was 1.56ha, the number of people should vary between 195 and 264. In the first case, between 26,250 kg and 35,280kg of annual grain crop was needed to meet dietary requirements, requiring the cultivation of between 131 and 176ha of arable land. For the second case, the figures are 40,950 to 55,440 kg of annual crop, requiring 205 to 277ha of arable land.

Following the pattern of the previous studies, the first 0.5-km circle around the site is excluded from the arable land estimations. As Table 5.7.1 shows that, within an area of 500 to 1500m around the site, there was enough potential arable land to meet dietary requirements for the full range of population estimates. In the case of the lower values (131-176 ha), even just the meadow soil was enough to produce the necessary amount of grain. However, the patchy distribution of meadow soil, as well as the need for fallow land, especially in the case of long-term exploitation as required by tell populations, suggest the joint use of meadow and cinnomonic forest soils in a segmented pattern of cultivation with shifting fallow/arable land. In the case of the higher values (205-277ha), there is still enough arable land within a 1.5km radius but only 49ha are left free of any cultivation. This figure suggests very intensive deforestation in order to free the closer areas of cinnomonic forest soil. As long as there is no direct evidence to support such site-oriented forest clearance, I would suggest that a larger area was incorporated within the Klisselika site exploitation area. Another 500m-radius circle around the site provides a sufficiently large area of additional cinnomonic forest soil, facilitating a more flexible pattern of forest clearance, not necessarily concentrated in rings around the site. Indirect evidence for the minor impact of forest clearance may come from the absence of severe erosion, which not only let the Klisselika inhabitants remain on the tell for centuries but also facilitated the foundation of a new settlement 1km North East of the first one. To summarise, the area between 500-2000m from the site contains enough arable and browse land to sustain a long-term agro-pastoral subsistence strategy of fallow/arable land rotation, as well as some natural vegetation comprising forest, bushes and shrubs.

Catchment area

The evidence for the wider catchment area of tell Klisselika is sparse. A Spondylus bracelet found during the later excavations points to contacts with the Black Sea coast more than 100 km to the East. At a closer distance of 1 to 10 km are opal and quartz deposits, which may have, been exploited by the site inhabitants. Antler tools betoken hunting activity and prey accessibility should not have exceeded 10 km, as there was no severe deforestation around tell Klisselika in the Neolithic and Copper Age.

5.7.4 Summary and discussion

Although the evidence from Klisselika tell is somewhat inconsistent and sparse, the full range of social practices, subsistence strategies, local production and exchange directions are recognisable that were more fully developed in the later periods.

The scatters of burnt daub suggest secondary use of daub and the possibility of the controlled use of fire. So far, evidence for massive fires has not been reported. Structured deposition can be suggested on the basis of the unusually dense deposition of bones and unconfirmed presence of one pit7. The claim for fragmentation practices is supported by the evidence that there were vessels with matching sherds, which did not make a complete vessel.

The subsistence of Klisselika tell occupants was most probably mixed farming, with cultivation, stock-breeding, hunting and gathering. Some crop rotation is presumable on the basis of the patchy soil distribution. The antler tools point to some hunting activity.

The relatively large number of chipped stone artefacts and local raw materials suggest the exploitation of local source(s) and on-site production of flint tools. Exotic artefacts in the excavated data are limited to Spondylus, indicating participation in an extended exchange network.

5.8. Gudgova tell

5.8.1 General information and earlier studies

The site of Gudgova (also known as Mudrets I) was excavated for the first time in 1973, when two 40 x 10m trenches were laid out in the central and the Southern parts of the tell and excavated to bedrock (AFig.5.8.1b). The results of the excavations carried by M. Dimitrov have not been published and the materials from the tell have restricted access in the museum storerooms of the Stara Zagora Historical Museum. The only available stratigraphic information for these early excavations is Parzinger’s (1993:114) mention of a 3m-thick Early Chalcolithic layer and a 1.60m-thick Late Chalcolithic layer, with no mention of any BA deposits at all. Investigations were renewed in 1992-1994 and, for a very short time, in 1998 by the team of the Maritsa Iztok Expedition. The aim of the new excavations was to clarify the stratigraphy and chronology of the tell, as well as to put the site in a broad palaeo-environmental and settlement context (Leshtakov n.d.b, 1995). During the new investigations, the old profiles were cleaned but sterile ground was not reached (AFig.5.8.1a). The BA layer was established to be 2 - 2.20m in thickness, the Late Chalcolithic more than 2.50m in thickness, comprising 17 building horizons, and the initial occupational sequence was found to be 2.50 - 3m in thickness, dating to the Early Chalcolithic (Leshtakov et al. 2001). Area excavations were made only on the top of the tell, where six successive building horizons were distinguished in the BA occupation.

Archaeological evidence

During the renewed excavations, sterile ground was not reached, so no evidence for the initial occupation of the tell is known. Archaeological material from the Early Copper Age has not been found either. Eleven building horizons were found after the cleaning of the North profile of the central trench, with six more in the West profile. Very few comments were made on the Chalcolithic stratigraphy. It was claimed that the 17 building horizons are rebuildings of the settlement area, not just reconstructions of existing dwellings (Leshtakov et al. 2001). Nine of the 17 Late Copper Age horizons are defined by house floors, while the remaining eight occupational layers were identified on the basis of beaten clay levels. At a depth of 2.38m from the surface in the West profile, a dwelling with 4 rebuilding phases was found to contain an oven, which was destroyed by Bronze Age postholes. Above this dwelling, a second one was built using a lighter construction but which was destroyed by fire. The burning of the dwelling was claimed on the basis of the 5-15cm-thick layer of burnt rubble above the oven. The burnt rubble was followed by a hiatus 15-50 cm in thickness (Todorova, N. n.d.).

The Bronze Age occupational levels appeared in the North profile as two different layers. The first one, immediately overlying the hiatus, is black-grey and consists of at least two horizons. Initially the layers were dated to the earliest stage of the EBA - Ezero A (Leshtakov n.d.b). In the final publication of the tell chronology, this stage was not mentioned and the date of the next BA layer - EBA3 - was accepted as valid for the whole Bronze Age occupation (Leshtakov et al. 2001). The two BA layers were not divided by a hiatus and the upper one is brown –ochre in colour, consisting of at least four horizons.

The investigations in 1992 included the cleaning of the profiles of the old trenches and two new 5 x 5m sondages. Parts of four dwellings were discovered, related to the II - IVth building horizons of the upper BA layer (AFig. 5.8.2). In 1993, a 5 x 5m grid was established on the same orientation as the sondages of the 1970s excavations (AFig. 5.8.1b) and an area of 250m2 was excavated. The following season, the investigated area was enlarged to 650m2. A consistent vertical and horizontal stratigraphy of the whole excavated area has not been provided. During my study of the field documentation, I encountered great difficulties in interrelating features in coherent vertical and horizontal units, that made me suggest possible reasons for the lack of general plans related to the stratigraphic sequence of the tell. First, the early investigations have devastated the site in two ways – not only cutting into the thick cultural layer but also heaping the excavated soil around the mound. Secondly, the financial restrictions of the investigations determined a specific field technique (cleaning profiles and limited excavated surfaces) that failed to clarify the stratigraphy of the upper layers, now additionally damaged by contemporary cultivation. Nonetheless, the investigators themselves confessed that it was not relevant to relate absolute depths from the top of the tell to any consistent building horizon (Leshtakov n.d.c), although such an attempt was made in the various site reports. However, in the absence of a formal horizontal and vertical sequence, the available evidence proved insufficient to reconstruct such a sequence. This is the reason why the description of the site is organised not according to its stratigraphical progression from earlier settlements and features to later occupational stages but in terms of the evidence for building features, artefacts and social practices.

Building features

A complex of three houses was excavated in squares M19/Q19. Only their Southern parts were preserved, the Northern parts being destroyed by one of the 1970s trenches (AFig. 5.8.3A). Postholes and beaten clay floors were found but no trace of any oven or hearth. It was suggested that the houses were built on a North - South orientation, with two rooms and an area of over 50m2 (Leshtakov n.d.c). Another complex of three houses was found in squares M19/P19, below the former dwelling complex. Two of the dwellings had their East and West walls preserved, while the third had just with its East wall (AFig. 5.8.1B). The Northern parts of the houses were destroyed by the 1970s trench, while their Southern parts fall into an un-excavated area. In the Easternmost house, an oven was found, which was claimed to be built over an earlier oven in the previous building horizon. However, the deviation in plan to the West means that there was no strict continuity of re-building. Only the house with the oven in square O19/P19 was recorded as being burnt. However, another burnt house was excavated in squares N14/15-O14/15. The dwelling was rectangular with a possible apsidal North wall, a North - South orientation, two rooms and two entrances on the Eastern side (AFig. 5.8.3C). The preserved dimensions are 4.60m in width and 8.50m in length. The postholes on the short sides were in two rows, interpreted as evidence of house reconstruction. The trench found in the South part of the dwelling was interpreted as either a shed or an additional wall support. The same support function was assigned to the three posts around the entrance. Since no hearths or ovens were found inside the house, their location was presumed to be in the unexcavated area to the West; an argument supporting such a claim was the location of the oven from the upper horizon in the Western part of the dwelling. However, this interpretation assumes an unproven continuity of building feature and cannot be accepted. Part of a burnt dwelling was excavated in P18; only its Eastern wall was preserved – with a construction claimed to be of pisé (AFig. 5.8.3D). The investigated part of the dwelling had dimensions of 3.80m (North – South) by 1.80m (East – West). The house had a North - South orientation but lacked ovens and hearths. On the floor, there were at least 25 loom weights, some of them only sun-dried and not fired. Next to the weights, two very small gold rings were found that were interpreted as belonging to a chain necklace. Sieving the soil from the house yielded numerous small flint flakes.

Artefacts

Chipped stone assemblage

The chipped stone assemblage from Gudgova tell has not been consistently investigated. The Copper Age tools (AFig. 5.8.11) were not studied and the information from the inventory books for 111 whole and 91 fragments of flint artefacts was not related to the study of the BA assemblage (AFig. 5.8.17). The latter consists of 186 artefacts, identified as two cores, 16 flakes, 78 retouched tools, 54 blades, 27 small chips, 1 natural piece, 1 repairing flake, 3 flakes from preparation (re-working debitage) and 4 amorphous fragments. Thirteen types of raw material were recognised, with sources similar to those of the Galabovo lithic assemblage. In addition to the exposures mentioned there, the possible sources of two types of raw materials were located in the area around the Chirpan hills, some 100km to the North West. Primary and secondary production of the flint tools was presumed to have taken place outside the tell (Zlateva –Uzunova, 2003). Such conclusions do not correspond to the excavator’s claim for on-site flint production on the basis of the presence of one flake in the house in square P18 (Leshtakov n.d.c). Unclear aspects include the presence of the two cores (AFig. 5.8.17A) and their relation to the opal pieces mentioned in the petrographic study and the opal debitage recorded in the field diary. Last but not least, comments have not been made on the possible link between the opal source 1km to the North West of Gudgova tell (pers. comm. P. Karacholov) and the opal pieces from the tell. Given the present condition of the data, conclusive claims cannot be made but, on the basis of presence of small flakes/debitage, 2 cores, one repairing flake, a natural piece and several amorphous pieces, as well as the proximity of the opal source, I would suggest that some form of chipped stone production was practiced on or near the tell.

Pottery

The renewed investigations have produced a huge amount of archaeological material that is very selectively and sparsely published. A representative selection of artefacts is presented here combining published and unpublished material (AFig. 5.8.5-19), aiming to illustrate the typical range of objects, shapes, decoration and use of raw material in the Late Copper and Early Bronze Ages in Thrace. During the early excavations, only the whole and restorable vessels were collected, while fragments from non- restorable vessels were secondarily re-deposited on the tell. In 1994, one such depot was re-excavated in squares N13-P13, where a pile of Late Chalcolithic sherds yielded fragments from at least 200 vessels (Stoyanov n.d.). They were from both fine and coarse ware and with different shapes and patterns of decoration (AFig. 5.8.8). Another 20 fragmented (AFig.5.8.5E, H-N; AFig.5.8.7A-F, H, J) and two whole vessels were published that derive from the cleaning of the profile (Leshtakov et al. 2001). Their actual quantity is much bigger but not known because of the specific Bulgarian recording standards8. The same uncertainty is valid for the BA pottery as well.

Other artefacts

Apart from the whole and fragmented vessels, a large quantity of artefacts was collected during the renewed investigations. The type and number of artefacts are summarised in Table 5.8.1.

Type Whole Fragmented In preparation Total
Chipped stone tools(flint and 4 from opal) 111 91 - 202
Polished stone tools 34 36 1 71
Bone/horn/ antler tools 35 23 N /a 58
Gold pendants 2 - - 2
Clay whorls 53 21 3 77
Clay weights* 64 26 7 97
Clay altars - 4 - 4
Clay lids 4** 1 - 5
Clay figurines - 5 - 5
Clay strainers 1 1 - 2
Models of wheel 2 2 N /a 4
Sling bullets 1 - - 1
Clay reels 2 - - 2
Clay funnels 1 1 - 2

Table 5.8.1 Type and number of artefacts from tell Gudgova. * At least 12 of them are surely net weights. ** One is claimed to represent an oven.

Social practices

Burning houses

The main stratigraphic profile of the tell in the central 1973 trench shows evidence for burning in both the Chalcolithic and BA layers but not along the whole profile (AFig. 5.8.1a). Apart from the three dwellings explicitly mentioned to be burned, there was more evidence for fire found on the tell. The information derives from the field documentation and was not properly incorporated in the final interpretation of the site. Chalcolithic fires appear to be present in two case – in N. Todorova’s (n.d.) study of the Copper Age stratigraphy but without a section; and in square R18, where the soil excavated in the 1970s contained a huge amount of burnt clay. BA burning of houses could be traced in squares M18, M19, M20, L20, where compact areas of burnt rubble were found. In L20 under the rubble, there was a layer containing spots of ash and charcoal. In squares N20, M20, M19, N19, O19, the soil was full of fragments of burnt daub. Therefore, it is likely that some of the fire products not explicitly connected to the above-discussed built features - but found within the same or neighbouring squares - were in fact related to the burning of the houses. It is also likely that there were subsequent fire events as the two house complexes are found generally one after another in one and the same squares M19-P19. The presence of areas of compact rubble suggests a massive in situ fire but the data from the central profile (AFig. 5.8.1a) show no evidence of totally devastating fires covering the whole site. Hence, I should assume that burning of individual BA houses at the Gudgova tell was a deliberate and controlled process. The secondary use of daub is difficult to investigate, given the present condition of the data. In addition, the long-term modern cultivation of the tell has destroyed to a great extent the in situ surface situation. The only formal comment on the tell’s burnt houses is for a continuous and peaceful re-occupation despite the burning of the buildings ! The arguments were that the dwellings were empty of any house inventory and the house plans of later phases were superimposed upon earlier building plans. Evidence for the repeating house layouts were not provided, nor were any causes for the fires (Leshtakov et.al. 2001). However, the data from Gudgova tell confirm the observations from previously discussed sites for the controlled firing of individual structures.

Structured deposition

Structured deposition in pits was explicitly commented on in two cases. The first case is a pit in M19, probably belonging to the last occupational BA level, that has destroyed parts of two previous horizons. The pit was 60 cm deep and 120cm in diameter, filled with gray-black soil and fragments of pithos. No interpretation or relation to some of the other excavated features was presented.

The second case was in P18, in which a pit from the IInd BA horizon was cut into a pit from the IIId BA horizon. The earlier pit was interpeteted as rubbish dump because it contained charcoal, layers of ashes, fragments of animal bones and a few sherds, as well as having a location 40 -50 cm from a house. The later pit consisted of domestic and wild animal bones deposited in a 20-cm-thick layer of crumbly gray soil, among which cattle and red deer bones were recognised (AFig. 5.8.4b). Apart from the few sherds found among the bones, a funnel9 was deposited very close to the pit mouth (AFig.5.8.4B). The funnel was accepted as a symbol of dairy production and its final deposition made investigators infer ritual deposition (Leshtakov n.d.d)

One more feature type reveals a certain deposition pattern, which is not common and puzzled the investigators. These are clay-made features with white clay plastered floors and walls from 5 to 7cm high (AFig. 5.8.4a). The features have an entrance and were interpreted as grain-driers. In one of the features, animal bones were found, while, in general, they were filled with gray-black crumbly soil.

Fragmentation

Probably the most striking example of fragmentation practice comprises the LCA sherds (AFig. 5.8.8) from more than 200 vessels found in the secondary depot during 1994 (Stoyanov n.d.). Whether or not there were matching sherds distributed on the tell or on the surrounding sites is not possible to conclude in the present state of the investigations. The data from table 5.8.1, however, shows a high percentage of broken objects on the tell, which, combined with the evidence for the LCA sherds, suggests that deliberate fragmentation was practised on Gudgova tell. The pieces of broken objects may have been kept on the tell as a resource for personal enchainment through objects or may have been brought onto the tell as a result of such practices (Chapman 2000).

5.8.2 Plant remains

Chalcolithic

Eleven samples were processed by flotation from the Late Copper Age occupation levels. In general, they contained only single grains and only two samples provided a more consistent pattern. The first one, from a house context, contained barley, vetch and lentils. The other sample, from a pottery scatter context, contained T. dicoccum, T. compactum and T. spelta. The last is a rare species in prehistoric times in the Balkan Peninsula (Popova 2001). The distribution of botanical remains from both Copper and Bronze Age contexts is summarised in Table 5.8.2.

Twenty-five samples of carbonised wood have also been studied (n = 206 fragments). Ten tree taxa have been identified – oak, elm, maple, hornbeam, alder, birch, hazel and some unidentified fruit species.

Table 5.8.2 Plant remains from Gudgova tell
Species Chalcolithic Bronze Age
T. monococcum + +
T. dicoccum + +
T. compactum + -
T. spelta + -
Hordeum vulgare + +
Hordeum vulgare var. nudum - +
Panicum miliaceum - +
Lens culinaris + +
Vicia ervilia + +
Lathyrus sp. - +
Bronze Age

Ten samples were processed for flotation from the BA occupation layers. The data is summarised in Table 5.8.2. Two of them are of particular interest. One sample, from a house floor, contained einkorn, barley, millet, lentils, vetch and vetchling. The most frequent plant is vetch (53%). Cornel, orach and fat hen were also present. The other sample contained a large quantity of acorns, cornelian cherry stones and 23 whole (?) carbonised plums. Other gathered species included elder and grape pips. Six weed species were identified, of which four were more widespread – Chenopodium album, Polygonum aviculare, Galium aparine and Brassica compestra Samples of carbonised wood from a dwelling floor in O19/P19 have been studied. They showed the use of oak, elm, maple, hornbeam and mountain ash in house construction.

The past vegetation around the Gudgova tell was interpreted as a deciduous oak forest with some hazel, alder and birch growing alongside the rivers. Forest clearance was also suggested to have taken place, as elm and maple appeared in both Chalcolithic and BA samples, on which basis they were accepted as perennial species. These species like sunlight and clay soils and are characteristics for forest clearings (Popova 2001).

5.8.3 The site and its surroundings according to the GIS analysis

The site was located 800m North of the left bank of the river Sokolitsa. It is on a 1-2° slope (CDFig.148) with a South West aspect (CDFig.149), at 152 masl (CDFig.150). The visibility from the site is very limited. It is mainly over the area 1.5km South of the tell and some spots to the West along the North part of the valley (CDFig.151). There is a more consistent visible area at 9.3-10.7 km to the West, roughly before the Obrutchishte site. All the sites in the valley are visible, while only Atanasivanova mogila is on the edge of a visible/invisible area. The panorama over the immediate area around the tell improves when 8m were added onto the surface that correspond to the height of the mature tell (CDFig.152). The visibility from the mature tell is consistent around the site and in particular better in comparison to the previous viewshed to the areas North and North East of the tell. The general visibility over the valley to the West is less patchy and has the same long-distance visible spot near Obrutchishte. All the sites in the valley are visible and Atanasivanova mogila is in the visible area. This means that Atanasivanova mogila became visible with the growing of the Gudgova tell.

The cost surface analyses (CDFig.153) and the site distribution are summarised in Table 5.8.3:

Table 5.8.3 Site distribution around Gudgova tell
N of cost strip Sites located in the cost strip
0 Atanasivanova mogila, both Iskritsa sites, Klisselika tell
1 MIBC, KMBC, Mednikarovo tell
2 Obrutchishte
3 Galabovo tell, Kurdova mogila, Taniokoleva mogila – all locations, Tcherniova mogila – all locations
4 Manchova, Goliamata, Malkata and Ovchartsi barrows, Barrow 4, Goliama Detelina flat site
6 Aldinova barrow, Polski Gradets tell
7 Polski Gradets pit site, Ovcharitsa I and II, Gonova barrow

The interrelation between Gudgova tell and its possible contemporary sites shows a dynamic pattern in terms of cost. In both the Chalcolithic and the BA, there were sites with relatively easy access and sites that were at a substantial cost distance. Therefore, the location of Gudgova tell could be interpreted as a deliberate choice in consideration of the previous (e.g. tell Klisselika) and contemporary (e.g. tells Galabovo and Polski Gradets) sites, in which the accessibility of sites may has been an important factor. In addition, when the site was founded, the visual link with the adjacent earlier Klisselika tell may have also played a crucial role.

The logistics network derived from the cost surface analyses closely resembles the network of the Iskritsa dwelling site but since Gudgova tell is in the Eastern part of the valley, there are some differences that should be discussed here (CDFig.154). The presence of the main valley routes is confirmed and there are six paths that connect the tell with sites located North of the Sokolitsa valley.

Starting from East to West, the first of these paths leads to the Polski Gradets pit site (CDFig.155). The track of the path matches the Neolithic/Chalcolithic route from the Sokolitsa valley to Polski Gradets tell, with an additional segment of 2 km to the North East. This is important evidence for the possibility of the re-use of earlier routes during the Bronze Age, which, in its turn, justifies the choice of the location of the seven sites (Aldinova mogila is re-placed by Ovcharitsa II) visible from the route (CDFig.156). The contrast is that, during the Neolithic and Chalcolithic, they were visible places, while, during the Bronze Age, they were visible sites.

Roughly 10 km before reaching the Polski Gradets site, the path bifurcates and the new branch ascends to the North West towards Goliamata (CDFig.157) and Malkata mogila (CDFig.158) and Barrow 4 (CDFig.159). After 2 km, the path joins the route from/to the Iskritsa dwelling zone to/from these three barrows. The visibility from the paths is very similar (CDFig.160), (CDFig.161) – good over the Eastern part of Sokolitsa valley, the hills in the East part of the study area and the Northern parts of Ovcharitsa valley. There are visible spots or strip- like visibility over the central and Western parts of Sokolitsa valley and over the hills in the central part of the study area. The path to Barrow 4 has lower visibility towards the Ovcharitsa valley (CDFig.162).

The other two almost fully matching paths from the Iskritsa dwelling site network are the routes to Manchova (CDFig.163) and Kurdova barrows (CDFig.164). The only difference is in the initial segment of the route, which forks from the main South route 1 km East of Iskritsa and ascends for 3 km to the North North West. The visibility from the path is very good over the Sokolitsa valley and the hills at the Eastern and the Southern parts of the study area and strip-like over the hills in the central part of the region (CDFig.165). The path to Manchova mogila has a better view towards the Ovcharitsa valley (CDFig.166).

The point where the path to Kurdova and Manchova mogila branches from the main South route is the point of bifurcation for the path to Taniokoleva mogila (CDFig.167). It ascends to the North North West for 3 km and then descends for 4.2 km to the North West, passing by the all four possible locations of the barrow. The visibility from the path (CDFig.168) is very similar to the visibility from the path to Kurdova mogila but the more Eastern location of Taniokoleva mogila assures better visibility towards the Western hills of the central study area, while, from the path to Kurdova mogila, the Eastern areas are visible.

The last two paths from East to West are the tracks to MIBC. There is one path that starts from the above-discussed point (to the three barrows) and leads to MIBC2-4 (CDFig.169), (CDFig.170), (CDFig.171). It ascends for 1.1 km to the North West, then 800m to the North when it divides into 3 segments that lead to each of the barrows. The visibility from the path is very good over the Eastern part of Sokolitsa valley and the hills to the South of it (CDFig.172). The segment to MIBC2 adds some visible spots towards the Central and Eastern hills of the study area (CDFig.173).

The path to MIBC1 splits from the main route North of Iskritsa and ascend for 2.3 km to the North West until it reaches the barrow (CDFig.174). The visibility from the path is similar to the visibility from the paths to MIBC3and 4 but less consistent over the Southern hills (CDFig.175).

The last track to be discussed is the path to Polski Gradets tell (CDFig.176). It starts due North of the tell, ascending for 1.5 km, then continuing to ascend to the North North East for 3.5 km and finally descending Northwards for 2.5 km. The visibility from the path is mainly toward the Sokolitsa valley and the area to the South, as well as over the Eastern part of the study area (CDFig.177). There are some visible strips over the central hills and gullies and the Northern parts of the Ovcharitsa valley.

It is interesting to point out that the adjacent Klisselika and Gudgova tells do not share a common track to Polski Gradets tell. This is important evidence for availability of alternative routes, which means that, if the aim of the journey between Polski Gradets and the area North of the present village of Mudrets was not based on least-cost access, there was an opportunity to use the other route.

The last issue to be discussed in the Gudgova tell GIS analyses is the visibility from the tell to Galabovo tell. This is the main South route discussed in several GIS sections and in many details of its track and visibility on the Galabovo case study. The following comments are on the visibility of the main South route if the direction of movement was from East to West.

The first segment of the route between Gudgova and Klisselika tells (CDFig.178) assures good visibility over the area around the two sites. There is also a patchy view over the Northern parts of the Sokolitsa valley and consistent visible spot 8.5 km to the West near the Obrutchishte site (CDFig.179). Both Iskritsa sites and MIBC2 are visible from this little segment.

The path to the Iskritsa sites extends the view to the East and assures a very good panorama over the Sokolitsa valley and the foothills of the Sakar mountain (CDFig.180). The segment to the Iskritsa pit site provides better visibility to the East of the valley in comparison to the path to the Iskritsa dwelling site (CDFig.181). All the sites in the valley plus MIBC2 are visible from the paths.

The visibility to the next site – Atanasivanova mogila - adds some newly visible areas that make the view denser but the panorama over the valley and the sites is generally the same as in the case of Iskritsa (CDFig.182).

In addition, the path to Mednikarovo provides a good view over the foothills of Sakar to the South of Klisselika tell, as well as over the areas South of the Sokolitsa valley around Mednikarovo (CDFig.183). Again, all the sites in the Sokolitsa valley are visible.

The last segment of the main South route before the destination of the Galabovo tell - to Obrutchishte site - assures good visibility further East along the valley, as well as to the South of the valley over the foothills of Sakar (CDFig.184). Galabovo tell is already visible from this path.

In summary, the panorama along the Sokolitsa valley from East to West provides a high level of site visibility, as well as landscape visibility, as visibility broadens while moving to the West.

Resources and land use

The size of the site is 1.7ha, which, according to the population estimation accepted in the study, should accommodate from 212 to 240 persons. The necessary minimum annual crop of 44,520 - 50,400kg to sustain such a population requires 222 - 252ha of arable land. The soil distribution around Gudgova tell is given in Table 5.8.4.

Table 5.8.4 Soil distribution around the tell Gudgova
Distance from site Meadow soil Smolnitsa Cinnomonic
0-500m 40ha 40ha 1ha
500-1000m 85ha 80ha 49ha
1000-1500m 56ha 69ha 230ha
1500-2000m 91ha 21ha 421ha
2000-2500m 109ha 73ha 530ha
2500-3000m 54ha 131ha 660ha
3000-3500m 84ha 149ha 756ha
3500-4000m 99ha 86ha 798ha
4000-4500m 74ha 85ha 749ha
4500-5000m 40ha 187ha 603ha

Exploitation area

Table 5.8.4 and CDFig.185 show that the soils around the tell show a patchy distribution, consisting of three main types – meadow, smolnitsa and cinnomonic forest soil. Such a dispersed distribution implies a certain type of cultivation, in which the particular knowledge of soil characteristics is crucial – e.g. smolnitsa is difficult to process under many circumstances but, in favourable times, can be very fertile. Hence, its cultivation involves a high effort/high yield strategy, in contrast to cinnomonic soil cultivation, which was lower risk/lower yield. The SCA was performed for all four different combination of soil types – meadow/smolnitsa, meadow/cinnomonic, smolnitsa/cinnomonic, meadow/smolnitsa/cinnomonic – in order to explore the extent to which the patchy soil distribution may have affected the exploitation area of the Gudgova tell.

In the case of joint meadow/smolnitsa exploitation, the area 500-2000m from the site contained sufficient arable land to sustain the estimated population. The distribution of these soil types would have allowed a fallow/arable land rotation in a segmental cultivation.

The similar distribution of meadow and smolintsa soil up to 500-1500m distance from the site defines the area as sufficient for subsistence exploitation in both combinations - meadow/cinnomonic and smolnitsa/cinnomonic. In both models, such an exploitation area assumes total deforestation. Since there is no evidence to support intensive forest clearance around the tell, the exploitation area should probably be enlarged up to 2000m from the site.

In the last case, in which all three soil types were cultivated, the area up to 500 - 1500m from the site contains enough arable land for a successful agrarian regime for the inhabitants of the tell. At the same time, there was no need for full deforestation of the area, which facilitated a segmental system of fallow/arable based on the patchy soil distribution. The area was previously cultivated by the occupants of tell Klisselika, which means that the region was already deforested and some soil exhaustion could be anticipated.

Therefore, the exploitation area of Gudgova tell is to be enlarged up to 2000m from the site, within which there is enough arable land for each of the four combinations of soil use, no total deforestation, availability of fallow land and the opportunity for segmental cultivation practices.

Catchment area

The objects and finds excavated on Gudgova tell define a broad catchment area of the site that is to be interpreted in terms of both the mobility of the tell inhabitants and then existence of short- and long-distance trade and/or exchange.

The chipped stone assemblage indicates a small-scale catchment area from 1 to 30 km, as well as a medium-distance network of up to 100km (the Rhodopes and the Chirpan hills) and a long-distance catchment area from Northeast Bulgaria across the Stara Planina mountain range.

The minerals from which the polished stone tools at Gudgova tell have been made are summarised in Table 5.8.5.

Table 5.8.5 Rocks used for polished stone tools at Gudgova tell. Source: Matchev, n.d.
Material Balls Polishers Whetstones Axes Maces Adzes Scrapers Pieces Moulds Figurines Weights
Quartz + +
Aplite 1 1 +
Sandstone + + 1
Diorite 1 +
Dacite 1
Amphibolite
Basalt + +
Andesite/Basalt + +
Amphibolite +
Marl 1 +
Argilite 1
Andesite +
Rhyo-dacite +
Opal 2
Marble 1

With the exception of the opal exposure 1km North West of the tell, the distribution of these rocks varies between 10 and 50km from the tell . An important exception to this middle-distance catchment area is the stone axe made of glaucophane schist (AFig. 5.8.19A). So far, this type of metamorphic rock has not been identified in Bulgaria but, as long as there are other types of metamorphic rocks in Bulgaria, it is plausible that glaucophane schists once existed but have been heavily eroded (Machev, n.d.). Until more evidence to support such a claim is available, I would suggest that the stone axe was a long-distance import to the Gudgova tell. Such a type of rock is distributed in the Southern Aegean islands (Machev, n.d.). The presence of such an exotic object suggests a long-distance specialist exchange network. The axe is fragmented, which is a strong evidence for the social practice of personal enchainment. If the axe was brought whole on the tell, there is a possibility for another important social activity – the practice of gift exchange, whether of complete axes or fragments of axes (Chapman 2000). Trade contacts have been assumed between the island of Microvouni and Galabovo tell during the MBA on the base of stylistic similarities in pottery (Leshtakov 1996). The axe fragment from Gudgova tell suggests earlier contacts between communities in the study area and the Aegean, that, together with the evidence for figs from Galabovo tell, define the Aegean area as a recurrent partner in small-scale, infrequent and therefore significant prehistoric interactions.

Gold sources in Maritsa Iztok occur mainly in river sand sediments. Whether the gold of the two rings was of local origin or the ornament was imported is not possible to establish without scientific analysis.

The plant remains from Gudgova tell outline a different direction of human contacts in the later prehistory of the study region. The plum tree (Prunus sp.) is not a potential species in Bulgaria as so far the wild taxa was not identified. Its initial distribution area is thought to be the Caucasus (Popova 1994). However, the long distance – over 1500km - and the presence of whole fruits excludes the possibility of direct import of fruits. It is more likely that seeds were brought and planted in the region. A few years after the initial publication of the collective find of plums on the Gudgova tell, some new discoveries were made, which may change the current hypothesis towards a possible origin in the Balkans. First, Prunus domesticus has been found in Pre-Cucuteni III layers at Ruşeşti Noi and in the Cucuteni A2 layer at tell Poduri, 480km North East of the study area, as well as at other Cucuteni – Tripolye settlements (Monah at al. 1997). However, recent investigations in Bulgaria have shown that plums were gathered even in the Early Neolithic (Marinova 2002a), suggesting a much earlier migration of the wild species. Since the evidence from the two countries is not correlated, it is difficult to assess the origin of the plums in Gudgova tell – whether as local domestication in the Neolithic or as domesticated imports from the Cucuteni area. However, it is sure that, by the mid-5th Millennium Cal. BC, plums had already been introduced into the Danube basin. The presence of 23 whole fruits rather than just seeds presumes the existence of plum trees in the study region during the EBA.

The presence of at least 12 net weights supports the hypothesis of net fishing that may have taken place close to the tell, as well as at a distance. The same broad catchment range is valid for hunting. Although there is very little evidence for hunting – several antler tools and a deer skull, perhaps a hunting trophy - it is likely that the Gudgova occupants would have culled animals in the surrounding woods. The proximity of Sakar mountain foothills defines an area only 5-10 km from the site, with the probability of very rich game reserves in later prehistory.

5.8.4 Summary and discussion

Like all sites discussed so far, the Gudgova tell presents evidence for social practices of fragmentation, structured deposition and the burning of houses. Its particular location only 1km from the earlier and contemporary Klisselika tell, raises questions about preferences for site location. Since it is not known whether there was contemporary habitation on both sites, conclusive claims are not feasible. However, there are at least two reasons for the shift in settlement location. First, if the bounded space on Klisselika restricted further expansion, some families or the whole community have moved away but still very close to their old settlement. Or secondly, unresolved social issues forced the community to re-negotiate the existing social order, for which a new dwelling place was needed. At the same time, the link with the ancestors was equally important and this new place should be related to the ancestral tell. Such a link between the two places was made by the visual connection between the two tells. Thus, simultaneously, there was a physical separation but a symbolic link with the ancestors’ tell, which constitutes one of the forms of the ancestor cult.

Natural resources were not a constraint in the shift of the site location, since there was intensive occupation from the Early Neolithic up to the end of the EBA facilitated by the abundance of suitable resources. Such long-lasting human occupation also suggests that the subsistence strategies during that time were well balanced and did not lead to drastic environmental changes.

Some production processes may have taken place on the tell but the only more or less secure evidence is for flint production.

The presence of exotic objects (glaucophane axe) and non- local objects (flint) places the Gudgova tell in a wider network of trade with exotic objects and in a smaller network of commodity exchange.

5.9 Polski Gradets tell

5.9.1 General information and present condition of the data

The tell near the modern village of Polski Gradets was investigated in 1987. During the autumn field survey of the MI expedition, pottery was found dating to the Late Chalcolithic, EBA, MBA, Late Roman, Medieval and pre-modern periods. The sherds derived from soil removed from the tell during the excavation of grave pits for the AD 19th century cemetery. Two 5 x 5m squares were excavated on the flat, upper, Western part of the tell in order to establish the stratigraphy and chronology of the site. During the excavations, four pre-modern graves were excavated and an additional four grave pits were identified. The 19th century graves have destroyed the cultural layer up to 2m in depth and no undisturbed contexts were found during the only investigation of this tell.

Archaeological evidence

Chalcolithic archaeological features were not reached in the 50 m2 excavated area. However, on the basis of numerous Late Copper Age sherds found on the tell surface, it was concluded that the tell was occupied at the time of the Late Chalcolithic. Considering the height –more than 8m - it has been suggested that the site was founded during the Neolithic (Leshtakov et al.2001), although no other evidence was given to support such a claim. Two arguments oppose this hypothesis: (a) no Neolithic sherds were found on or near the tell; and (b) tell Gudgova is of similar height and lacks Neolithic occupation! Bearing in mind that, until secure data is provided, any conclusive comment is precluded, I would conclude on the basis of the present evidence that the site was occupied from the start of the Chalcolithic up into the BA. The BA layer is about 2m thick. The pre-modern grave pits at Polski Gradets tell were up to 2m deep and seemed to destroy several BA horizons and at least one Chalcolithic horizon.

Three successive building horizons were identified during the excavations. They consisted of three overlying dwellings, whose plans and size were not possible to establish due to the limited excavated area and subsequent destruction.

The lowest dwelling was burnt and its wall rubble was immediately overlain by the floor of the next house. The latter was of beaten clay, with brown –red spots, interpreted as a result of fire. An oven and a hearth were also found in the dwelling. The floor was overlain by a layer of ash and charcoal, interpreted as the debris of the burnt roof of the house. The soil above the dwelling was full of burnt wall rubble; it also contained sherds and animal bones. The last dwelling had a beaten clay floor, two ovens, a hearth and a built-in storage vessel. Above the floor, there was a layer of ash and charcoal, overlain by small and medium-sized pieces of burnt house rubble.

The soil in the whole excavated area, in general, contained sherds, burnt house rubble, small and medium-sized stones and ash and charcoal. The sherds from the last two dwellings dated the building horizons to the EBA2 stage. The few sherds published from the site confirm this chronology (AFig. 5.10.3). It was mentioned that there were unstratified MBA sherds on the tell as well, while the chronology of the lowest building layer was not discussed.

5.9.2 The site and its surroundings according to the GIS analysis

Tell Polski Gradets is located on a hill, at 189-213 masl (CDFig.186) It is on a 2-3° slope (CDFig.187) with a South West aspect (CDFig.188). Although on a hill, the visibility from the site is very limited – to no more than 2.6km to the South of the tell (CDFig.189). There is a strip-like view over the hills West of the tell and a more consistent view over the Northwesternmost edge of the study region. Three barrows were probably visible from the tell (surely two and one – Tcherniova barrow - with one out of its four possible locations). Better but still limited is the visibility when 10m are added to the site location surface (CDFig.190). More patchy strips are visible over the Western hills towards the central part of the study area and the panorama around the tell itself is more consistent. There are a few more visible areas to the North, North West and North East of the site. In addition to previous sites, two more possible locations of Tcherniova barrow are visible.

The cost surface results (CDFig.191) are summarised in Table 5.9.1.

Table 5.9.1 Site distribution around Polski Gradets tell
N of cost strip Sites located in the cost strip
1 Polski Gradets pit site
2 Aldinova barrow, Ovcharitsa I and II
3 Gonova barrow, Goliamata, Manchova, and Ovchartsi barrows, Goliama Detelina flat site Tcherniova mogila – all locations
4 Taniokoleva mogila – all locations, Malkata mogila, Barrow 4
5 Kurdova mogila, Galabovo tell, MIBC
6 Obrutchishte, Atanasivanova mogila, both Iskritsa sites, Mednikarovo, Klisselika and Gudgova tells
7 KMBC

There is a clear pattern of BA site location in areas of easier access. Pre–BA sites only began to appear in the 6th cost strip (with one exception – Ovcharitsa II is in the second cost strip). Therefore, it could be concluded that the reduction of the cost of site accessibility had happened during the BA, when denser settlement networks developed in the valley of the Ovcharitsa and the interfluve between the Sokolitsa and Ovcharitsa valleys.

The logistics network derived from the cost surface has a different overall pattern from those discussed so far, since it is made from a site that is located in the North part of the study region (CDFig.192). However, there are major similarities that confirm the presence of recurrent tracks in the later prehistory of the study region. The main South and North routes are the same in general and only the differences are going to be discussed here. There are two main differences in the North route and the reason for them is the change of the direction of movement - from East to West. The first difference is in the path to Aldinova mogila (CDFig.193) - the segment from Polski Gradets tell to the main North route differs from the segment that leads to the tell if the movement was from the West (e.g. Galabovo tell). Once the main North route is reached, the paths follow a common segment to Aldinova barrow. The second difference is in the segment towards the Goliamata and Malkata barrows. The branch that diverges from the main North route 8 km from Polski Gradets bifurcates after 300m, with the East segment leading to Barrow 4, and the West segment to Goliamata and Malkata barrows (CDFig.194). The latter climbs the hills with the least slope to the South West until the barrows are reached.

There is one more difference in the path to Ovcharitsa II, which starts from Polski Gradets tell due North, descending and ascending for a total of 4 km (CDFig.195). It is noteworthy that the paths to the neighbouring Ovcharitsa I and Gonova barrow (CDFig.196) do not climb the steeper hill to the North East but follow the main North route in the valley of the Ovcharitsa.

The main difference in the South route consists in the approach to each individual site. So far the access has been chain-like, from East to West or from West to East. The pattern from Polski Gradets tell is different, where the main traffic is from North to South. The Sokolitsa valley (respectively the main South route) is approachable by three main paths. The route across the study region from Polski Gradets to Klisselika tell bifurcates when the Sokolitsa valley is reached, with the left wing leading to Klisselika tell, while the right wing leads to the Iskritsa sites. Both paths were discussed in previous sections – the only difference here is that the movement was in the reverse direction.

The other three sites in the Southern part of the valley – Mednikarovo, Atanasivanova mogila and KMBC - are to be reached while following the main North route, crossing the study area over the contemporary mining area (a path already discussed in previous case studies when the region was crossed from South to North) and, once in the valley, every segment to the site is the same as described in each of the three individual logistics network. The last route that crosses the study region is the one from Polski Gradets tell to Gudgova tell (see section 5.8.3).

The track to the MIBC starts from the path Polski Gradets tell – Klisselika tell. Since the latter and Polski Gradets tell appeared to be contemporary, comments on this particular path are noteworthy. The route follows the least steep hills throughout the whole journey. It starts to the North of the tell, then descends left to the West South West and after 4.3 km climbs up again to the South for 4.8 km; then turns left to the West, following the ridges of the hills for 3.6km, when it bifurcates, with the Northern segment leading to barrows 3 and 4 (CDFig.197), (CDFig.198), the Southern segment to barrows 1and 2 (CDFig.199), (CDFig.200).

Viewshed analyses were performed only for paths that are new in the discussed logistics networks or that have not been discussed earlier.

The visibility from the paths to MIBC3 and 4 (CDFig.201), (CDFig.202) is almost identical – a good view over the hilly areas in the Southern and Eastern parts of the study area, as well as to the Northern parts of the Ovcharitsa valley. There are also visible strip-like views over the hills and the gullies in the central part of the region. The path to MIBC1 (CDFig.203) adds to this general panorama a few more visible areas along the Sokolitsa valley and in the central part of the study area. The path to MIBC2 (CDFig.204) provides some additional visible areas along the Sokolitsa valley, which assure the visibility over the Iskritsa pit site and Klisselika tell. The path to MIBC2 has a view towards Ovchartsi barrow, as well. All the paths share a common visibility over the seven barrows located in the Northern part of the study area.

The specific pattern of barrow intervisibility when approached from the South is valid here as well. From the paths to MIBC3 and 4 – barrow 1 is not visible; the path to barrow 2 lacks a panorama over MIBC4 and 1, and finally, from the path to MIBC1, barrow 4 cannot be seen.

The visibility from the route to Ovcharitsa II is mainly along the path – 1.3 km to West and 2.6 km to the East (CDFig.205). The areas South of the tell and some Northern parts of the Ovcharitsa valley are also visible. Five barrows (surely four and one with one out of four possible locations) and one flat site could be seen from the path.

The route to Aldinova mogila has a very good view over the North Eastern parts of the study area and the Northern parts of the Ovcharitsa valley (CDFig.206). There are also some strip-like views towards the central part of the region. Despite the patchy visibility to the East of the path, two sites contemporary with the path could be seen. Another five barrows (surely four and one with two out of four possible locations) and one enclosure are visible from the path, as well.

Resources and land use

There are two basal sizes for the Polski Gradets tell mentioned in the available literature:– 150 to 170m (MI report 1987) and 120m in diameter (Leshtakov et al. 2001). These figures resulted in two very different estimations of exploitation area presented in Table 5.9.2

Table 5.9.2 Estimation of exploitation area according to different site size estimates
Site area Population number Annual crop Arable land
1.1ha 141-192 29 610 - 40 320kg 148 – 201ha
2.5ha 318-432 66 780 – 90 720kg 334 – 453ha

Exploitation area

Calculations were performed for both ranges and it became apparent that, in the case of Polski Gradets, the population number affects the size of the possible exploitation area. Variables of the exploitation area are also due to the specific soil distribution around the site given in Table 5.9.3:

Table 5.9.3 and CDFigs.207 show a different pattern of soil distribution from those discussed so far for the MI sites. There is no meadow soil within a radius of 1000m around the tell and, instead, there is a new type of soil cover - – the rendzina type. The two main soil types were cinnomonic forest soil and smolnitsa, with a prominently zonal distribution. Estimations of the possible exploitation area followed a different pattern from the mechanism applied in the previous cases. First, the circle of 0 - 500m was included in the calculations, as it does not contain any meadow soil for pasture. Meadow soil was not taken into consideration for arable land calculations, since it appears at distances of 1000 - 1500m from the site in quantities more relevant for pasture than for cultivation. Three combinations of soil use were used to estimate the Polski Gradets exploitation area- only cinnomonic forest soil, cinnomonic forest soil and smolnitsa and a combination of cinnomonic forest soils, smolnitsas and rendzinas.

Table 5.9.3 Soil distribution around the Polski Gradets tell
Distance from site Meadow soil Rendzina Smolnitsa Cinnomonic forest soil Without soil Initial pedogenesis Artificial soil
0-500m - 31ha 18ha 28ha 2ha - -
500-1000m - 31ha 73ha 72ha 9ha - -
1000-1500m 15ha 13ha 134ha 177ha 1ha - -
1500-2000m 13ha 7ha 224ha 300ha 15ha - -
2000-2500m 24ha 35ha 276ha 342ha 33ha - -
2500-3000m 32ha 30ha 261ha 444ha 112ha - -
3000-3500m 18ha 16ha 319ha 495ha 173ha 4ha -
3500-4000m 60ha - 349ha 384ha 180ha 15ha -
4000-4500m 67ha - 436ha 343ha 114ha 60ha 21ha
4500-5000m 147ha - 261ha 408ha 54ha 39ha 56ha

In the case of the lower population estimate of 141-192, the area from 0 to 1500m around the tell contained enough arable land if all three soil types were cultivated or if the combination was restricted to cinnomonic forest soil and smolnitsa. If only cinnomonic forest soil was used, then the exploitation area should be enlarged to 2000m around the site.

For higher populations in the range 318-432, the exploitation area increases to a radius of 2000m for the use of all three soil types use, or for joint cultivation of cinnomonic forest soil and smolnitsa use; and up to 2500m if cinnomonic forest soil alone was cultivated.

In both cases, the defined exploitation area contains enough arable/fallow land, natural forest vegetation and pasture and browse land. The pattern of soil distribution suggests some form of zonal arable/ fallow rotation as well as some crop rotation. Cultivation of the rendzina soil would introduce some patchy cultivation practice, as this soil was located in two patches around the tell. However, the area of rendzina soil is only 75ha within a radius of 1.5km and the calculations have shown that its cultivation does not change the exploitation area size. This suggests that the rendzina soil was not relied upon as an important arable resource, possibly because it was an unknown quantity for interfluvial agriculture

5.9.3 Summary and discussion

The investigation of the Polski Gradets tell exploitation area has two important implications. First, meadow soil was not a crucial prerequisite for site location. Secondly, in cases where the site population exceeded 200 people, in order to keep the exploitation area closer to the site, the tell inhabitants may have started to cultivate the smolnitsa, that is difficult to till but very fertile. Indirect evidence for possible smolnitsa exploitation may be the fact that the site was founded on a place without meadow soil – a type relatively easy to cultivate. The Late Copper Age pottery found on the tell is typical for the KGK VI complex, that is known to comprise experienced agricultural communities. As discussed earlier (p. 237.), the initial occupation of this site is not known but, on the base of the resource distribution, it is likely to suggest that the first settlement was not before the beginning of the Copper Age. The argument for such a hypothesis is that any agricultural group needs social time to adjust its subsistence strategy and technologies to new or variant ecological conditions. The transition from alluvial cultivation to smolnitsa processing is not impossible during the Neolithic but a certain time was needed to explore the area and the available resources, as well as to develop the necessary knowledge and skills to cope with smolnitsa cultivation. So far, no Neolithic site is known in the study region that completely lacks meadow soil within a 1-km radius of the site. During the Neolithic, the study area was not densely settled and there were huge alluvium areas, which were not inhabited. It was also possible that the initial occupation was not connected to questions of soil availability at all. Its subsequent development into a tell, however, suggests that the site location was not accidental, since the successful sustaining of a long lasting tell occupation requires the ready availability of critical soil resources. Therefore, it could be concluded that the development of the Polski Gradets site into a tell validated the initial settlers’ choice of location and assumes a successful, long-term agro-pastoral subsistence strategy.

Polski Gradets tell was an important landmark in the landscape since it forms part of a recurrent pattern of long-distance visibility from most of the sites located along the Ovcharitsa valley (for details see the sections on visibility in Chapter 6). It is possible that the social landscape discourse prompted the spatial distancing but, at the same time, visual reference was made to an ancestral place as a deliberate act of relating-at-a-distance.

5.10 Polski Gradets pit site

5.10.1 General information and earlier studies**

The Polski Gradets pit site was excavated during several archaeological seasons in 1995-98 and 2002. Since the site has not been fully excavated, its actual size is not known. So far, more than 75 pits, 13 graves, a building and two basilicas were found (AFig.5.10.1). The features and materials date from the EBA, LBA, EIA, Roman/Late Roman and Medieval periods. The site is still under investigation and it is not published yet. The following description summarizes the data from the field diaries and site documentation for the features dated to the EBA and LBA (Georgieva and Nikov n.d. a-d).

Archaeological evidence

The 13 graves were all dated to the LBA on the basis of the grave goods. These data are summarized in Table 5.10.1.

Table 5.10.1 Evidence for Late Bronze Age graves, Polski Gradets pit site
Grave N Number of individuals Grave goods Special features Traces of burning
1 1 small clay cup, numerous small snails, large bronze needle, worn-out flint blade, pieces of a bronze ornament Missing mandible, replaced by the small clay cup The soil under the skeleton is mixed with charcoal
2 1 ? ? The soil under the skeleton is mixed with charcoal
3 1 Bronze chisel (AFig. 5.10.2N) - -
4 1 - The skeleton lies on a patch of medium-sized and small broken stones -
5 1child 3 clay beads, sea shells, two perforated snails (AFig. 5.10.2B, C, F, H, J, L, O, P) - -
6 1 Perforated sea shells, bone whorl, bronze needle (AFig. 5.10.2D, E, G, I) Destroyed by later pit Small pieces of charcoal in the fill.
7 1 - - -
8 1 child - - -
9 1 - - -
10 1 A fragment of a vessel Soil mixed with daub fragments -
11 1 - - Large amount of charcoal around, above, under the body
12 At least 5 - Disorderly spread of bones and fragments of bones; broken stones among the bones Charcoal, burnt daub fragments
13 1 Bronze arrowhead, 2 bone buttons with incised decoration, bone arrowhead, whetstone    

Until the discovery of grave 12, the site was considered to be a LBA flat cemetery. The disorderly spread of human bones in grave 12 puzzled the excavators, who informally interpreted it as evidence for human sacrifice (AFig.5.10.4). The cemetery was flat and maybe there were no grave indications on the surface, so a later pit may have destroyed grave 6. There is no single unifying pattern followed in the burials. Some of them contained grave goods, others had traces of burning, while still others had a combination of both. There were also some particularities that do not appear to be a recurrent pattern. This pattern suggests the material embodiment of a wide variety of different identities within individuals and the community at large.

The other prehistoric features comprise three pits containing EBA material.

Pit N12 is 81cm deep and its base followed the slope of the terrain. The initial deposits in the pit were large pieces of daub. Among them were found a loom weight and fragments of several spindle-whorls (AFig. 5.10.5A, K, E). Up to 50 cm above the daub pieces, there were other layers of daub pieces, small pieces of burnt daub and isolated pottery sherds. In this layer, a dog skeleton was excavated. Some of the sherds among the daub pieces were covered with small and medium-sized stones. Charcoal and animal bones were also found in the layer. A spot of ash and pottery sherds were discovered in the next layer, 15cm higher up. The soil in the pit is black, oily and mixed with stones. The only mentioned datable materials derive from the uppermost layer, which contains BA sherds.

Pit N 17 is up to 1m deep and its base is covered by layer of black soil, mixed with a large amount of charcoal (AFig. 5.10.5b). In the Southern part of the pit, this layer is covered by a black-gray soil mixed with sherds and pieces of daub. The next layer is up to 20cm thick - a brown-gray soil mixed with single sherds and animal bones in the Northern part of the pit (AFig. 5.10.5D), with more sherds as well as pieces of daub in the Southern part. That layer was sealed by a 20cm-thick layer of yellow-brown soil, with single sherds in the Northern part and a large quantity of sherds and pieces of daub in the Southern part. In the Northern part of the pit, two horns (AFig. 5.10.5H) with polished points were found that were interpreted as being worn out from intensive use. A spot of ash, animal bones and sherds overlaid this sealing layer in the Southern part of the pit. In the upper black oily layer, there were no sherds or any other archaeological material in the Northern part of the pit; and single sherds, a small amount of daub and charcoal and a flint tool (AFig. 5.10.5I) in the Southern part. At around 90 cm in depth, in the layer, a restorable/whole EBA vessel was found that has dated the pit (AFg. 5.10.5G).

Pit N58 is 60 - 70cm in depth and contains an EBA vessel that was claimed to be broken on the spot. The fragments were large and were found at a depth of 50cm. The layer at 40 - 50cm in depth consisted of light-brown clay soil, mixed with lots of charcoal, pieces of daub, sherds, animal bones and burnt daub. The layer at 50 - 70cm has the same characteristics. The base of the pit follows the uneven slope of the terrain. Different combinations of the mentioned components, with stones in addition, form the successive layers of the pit. There are visible changes in the soil texture but as the pit was excavated using a) mechanical layers and b) in four segments, the pattern of deposition is difficult to establish. The datable sherds from the layers above the layer with the BA vessel derived from EIA vessels. One sherd from the Medieval period was also found, close to the surface.

The pit has a complex horizontal and vertical stratigraphy, with many sherds that were post-BA; it is made more complex by the fact that the soil matrix is similar to the fill. It was suggested that there was more than one feature. It is possible that the initial pit was left open and subsequent sedimentation and/or erosion led to further in-filling of the feature. It is also likely that the digging of pits during the EIA led people to discover BA pottery scatters or pits, with subsequent careful re-arrangement of the content of their pit, incorporating the earlier material. The Medieval sherd was found close to the surface and may indicate some later activity, not necessary connected with the earlier pit/s.

5.10.2 Plant remains

The plant impressions on the daubs from pit 58 have been investigated. They contained negative traces of einkorn, bread wheat and barley.

5.10.3 The site and its surroundings according to GIS analysis

It was not possible to establish the exact size and location of the Polski Gradets pit site, since the site is still under investigation and has not yet been published. It is known, however, that it is a flat site of probably more than 1 ha in area. Four possible corners of the site were chosen in accordance with the site documentation. In order to simplify the analysis, a single dot was chosen to represent the site. This dot, respectively the cell in the grid, is roughly in the middle of the four possible corners of the site. The other possible locations are not displayed, since they present points, while this is a flat site with an extensive horizontal stratigraphy. However, each GIS analysis has been checked against all the possible corners and the results are summarised in the text.

The Polski Gradets pit site is located on a hill, at 189-213 masl (CDFig.208). It is on a 2-3° slope (CDFig.209) with a North West aspect (CDFig.210) and has very restricted visibility. It is patchy around the site – 2km to the West and North West and 1 km to the South (CDFig.211). There are some visible spots in the Northernmost parts of the study area. Only Ovcharitsa II is visible from the site.

The cost surface analysis (CDFig.212) and site distribution are given in Table 5.10.2:

Table 5.10.2 Site distribution around Polski Gradets pit site
N of cost strip Sites located in the cost strip
1 Polski Gradets tell, Ovcharitsa I and II
2 Aldinova and Gonova barrows
3 Ovchartsi barrow, Goliama Detelina flat site, Tcherniova mogila – all locations
4 Taniokoleva mogila – all locations, Goliamata, Malkata and Manchova, barrows, Barrow 4
5 Kurdova mogila, Galabovo tell, MIBC 2-4
6 MIBC1, Atanasivanova mogila, both Iskritsa sites, Klisselika and Gudgova tells
7 Mednikarovo tell, Obrutchishte, KMBC

The Polski Gradets pit site emerged in an area with easy and quick access to one earlier and possibly contemporary site. The latter was even visible from Polski Gradets pit site. Relatively less accessible in terms of cost but still in close vicinity were another two contemporary barrows. During the LBA, the pattern of high accessibility of contemporary sites is still valid but the barrows with their possibly contemporary burials are further away than the EBA sites.

The logistical network derived from the cost surface is similar to the previously discussed network for the Polski Gradets tell, with two important differences (CDFig.213). First, there is no separate route to Gudgova tell. In order to reach the tell, one should follow the route to the Sokolitsa valley and then via Klisselika tell to approach Gudgova tell from the East (CDFig.214). Secondly, there are two routes to the Western and Eastern edges of the Iskritsa flat site. The dwelling site is to be reached from the East via the route to the Sokolitsa valley (CDFig.215) (described in the previous case study as the route to Klisselika tell and the Iskritsa sites), while the Western part of the site is approached from the West via Atanasivanova mogila (CDFig.216) (the detailed track description is given in the previous case study). Thus, there is a possibility for a round trip between two sites, in which the onward and return trips cross the region following different tracks. There is no evidence that the Iskritsa sites and Polski Gradets pit site were contemporary. The reason for the detailed discuss on this particular pattern is because it confirms the possibility for alternative routes observed in the previous case study (p.282). Indirect confirmation for the alternative pattern is the lack of a special route between the contemporary Polski Gradets pit site and Gudgova tell. The onward journey may have been through the path shown in CDFig.214 and derived from the Polski Gradets pit site cost surface. The return journey, however, may have been via Polski Gradets tell, following the path not shown on CDFig.214 but which exited between the Polski Gradets and Gudgova tell as the previous case study cost surface analysis has confirmed.

There are minor differences from the logistics network of the Polski Gradets tell in the paths towards the sites in the North part of the study region. The main North route is used to connect the Polski Gradets pit site with the sites in the Western part of the study area, while there are four different paths to the sites North of the Polski Gradets pit site.

The route from the Polski Gradets pit site to Ovcharitsa II starts in a North North Easterly direction and then descends to the North West, following the least steep slopes of the hills (CDFig.217). From the path, there is very good visibility over the Western and North Western parts of the study region and the Northern part of the valley of the Ovcharitsa (CDFig.218). There is an invisible spot North of Ovcharitsa II and consistent invisible areas to the East, West and North of Ovcharitsa I and Gonova barrow. Four barrows (surely three and one with one out of four possible locations) and one flat site are visible from the path.

The Polski Gradets pit site – Ovcharitsa I site path follows the same track for 2km and then turns right to the North East, descending into a small gully for 800m and finally climbs up for 200m to the site (CDFig.219). The path follows the least steep slopes in this generally steep area. There is a good general visibility over the Northeasternmost part of the study area but the consistency of the view is interrupted by a totally invisible central part of this Northeasternmost area (CDFig.220). The view over the Northern part of the valley of the Ovcharitsa is patchy. Ovcharitsa II, Gonova barrow and one of the possible locations of Tcherniova mogila are visible from this path.

The same visibility occurs from the route to Gonova barrow, as only a 500m-long segment is needed to reach the site from Ovcharitsa I (CDFig.221). Therefore, it shares the same general visibility but it is better over the previously invisible middle part (CDFig.222). This is due to the higher location of the barrow. All of the remaining three sites in the Northern part of the study region – Ovcharitsa I and II and Aldinova barrow - are visible from the path, together with one of the possible locations of Tcherniova mogila.

The route from Polski Gradets pit site to Aldinova barrow takes a different direction, which initially descends into the valley to the North West and then climbs up again for the last few hundred meters following the least steep slopes (CDFig.223). Only the Northeasternmost part of the study region is visible (CDFig.224). The panorama from the path over that area is generally good but there is almost totally invisible strip around the sites of Ovcharitsa I and Gonova mogila. The Gonova mogila is visible, however, as it is on a hill. The other visible sites are the Ovchartsi barrow and Ovcharitsa II.

The last path to be discussed is Polski Gradets pit site - Polski Gradets tell (CDFig.225). It winds round the highest hill in the study area, initially descending and ascending to the South West for 1.1km and then following the ridge of the hill to the South for another 1.1 km. The visibility from this path is mainly over the hilly areas in the Eastern part of the study area (CDFig.226). There are scattered visible spots toward the central, North West and Northern parts of the study area. Three barrows (surely two and one with one out of four possible location) and one flat site are visible from the path.

In summary, the movement between Polski Gradets pit site and the contemporary EBA or LBA sites, which in the same time were in an area with least cost accessibility, assured an almost complete site intervisibility. The number of sites during the EBA around Polski Gradets pit site followed a relatively dense site distribution pattern, which changed to a more dispersed pattern during the LBA. The movement between Polski Gradets pit site and these distant sites followed the main valley routes and shared their high site visibility. If a LBA flat site existed North of Gudgova tell, there was an opportunity for round trips across the study area. The link between some LBA burials in EBA barrows and the flat cemetery near Polski Gradets will be discussed in Chapter 8.

5.10.4 Summary and discussion

The Polski Gradets pit site has not been fully published and the rare references to it consider the site as ritual. The long duration of similar activity (it was specially underlined that the fill of the pits was astonishingly similar despite the huge chronological differences) on one and the same place was not discussed in detail. The choice of the site has been suggested to be related to either gold sources in the Goliamata reka, flowing next to the modern village, or the presence of two types of rocks that may have been exposed as surface outcrops at the time of the active use of the site (pers. comm., K. Nikov). It may be noted, however, that there is no trace of on-site processing of any of these resources.

The Polski Gradets pit site contains evidence for social practices already discussed in previous case studies and generally barely discussed in Bulgarian prehistory and usually connected with some ritual activities. The pits provide evidence for certain types of activity that were either practiced on the site or somewhere off-site. The presence of fire products betokens fire concentrated in small areas (ash and charcoal), as well as fire over wider areas, outside the pits (burnt daub). It is likely that fire products (ash, charcoal, burnt daub) derive from burnt houses from (? nearby) settlements but it is also possible that the burning took place at the site. The presence of buildings on the site was not discussed on in the field documentation. Indirect evidence for fire at the site derives from the sondage at the South East edge of the site, deliberately located at the lowest area between two slopes. The sondage contained large pieces of burnt house rubble and numerous EIA, LIA, Roman and Late Medieval sherds. The location of the sondage and the condition of the sherds explain the apparent association of pottery of very different ages as a result of long-lasting erosion. Burnt house rubble, however, appeared to be in situ. At the time of my research, the site was still under investigation and the finds under the burnt debris were not yet apparent.

All the pits contain fire products (ash, charcoal or burnt daub), with evidence for secondary use of daub (pieces of daub), feasting (animal bones) and deliberate fragmentation (sherds, fragments of whorls). The BA vessel that was broken on the spot could be interpreted as trizna – a ritual pottery scatter. Special patterns of structured deposition could be observed only in pit 17, in which the very striking North/South division of finds expressed contrasts in both quantity and diversity, that ultimately results in different use of the two parts of the pits. Such a pattern of the deliberate deposition of contrasting objects may symbolize some ideological contrast (e.g. culture/nature) or some specific tension in social discussion.

Given the present paucity of contextual data, such patterns cannot yet be adequately interpreted. But such data, however, reinforces the interpretation of all of the pits and graves as the result of deliberate and controlled acts, i.e., structured deposition. The graves contain traces of similar depositional activities to those in the pits – fire products, the secondary use of daub and fragmented objects, even though they post-date the pit deposits by more than a millennium.

Graves N1 and 12 betoken a certain type of post-burial activity, involving body fragmentation. The only body part claimed to be missing from the inhumation in grave 1 was the lower jaw (Georgieva and Nikov n.d.c). The drawing of the skeleton, however, shows other bones to be missing as well, most obviously the pelvis. There were no obvious traces of severe later destruction (viz., looting). In such a case, two interpretations are possible. First, the mandible and the pelvis decayed because of soil acidity and the cup was placed close to the mouth to imitate drinking. Secondly, if the missing bones are the result of poor documentation (only the mandible is mentioned as missing in the diary), then the body may have been defleshed, the mandible taken and the skeleton was buried with the cup replacing the lower jaw. It is also possible that the disarticulation of the body occurred at a later stage, when the grave was re-opened (if it was closed on the first place at all). The hypothesis for post-burial activity is most likely, as there are some peculiarities in the grave. The flint tool and the ornament fragments were found in a small pit under the skull. The position of the bronze needle is not clear enough to conclude whether it was put near the skull or it was stuck into the head. Therefore, it could be inferred that the deceased was treated in a specific way that included post-mortem activities in which keeping part of the dead body among the living was important. Indirect evidence for memorialisation is the excavators’ claim for post-burial trizna, with the pottery scatter found some cms above the pelvic area.

Since anthropological investigations of the osteological remains in grave 12 have not yet been made, it is not known whether they represent complete skeletons of a certain number of individuals or there are body parts of many individuals. Therefore, conclusive claims are not yet possible but a brief evaluation of possible explanations is necessary. The presence of burnt daub pieces and charcoal links the grave to a further four graves in the cemetery that contained the same components. So it is possible that it was a collective burial, of whole bodies and body parts. The deposit may also have been subsequently destroyed by looting or other activities. It is also likely that the bodies were accidentally discovered during pit-digging activities and then were re-buried without following the anatomical order of the body parts.


  1. During the first three working seasons 13 building horizons have been identified. In 1995 the coal dust layer was removed and one or two more horizons have been observed. Since the presence of the last 15th horizon has not been confirmed, the total number of the Bronze Age building horizons is accepted to be 14.↩︎

  2. The animal bones were studied by G. Ribarov. The data presented in current study is a summary of his unpublished report.↩︎

  3. Current summary of plant remains evidence is made after few articles of Popova (1991, 1995, 1995a, 1998 together with Bozilova, 2001)↩︎

  4. For the purposes of simplicity the introduction of the method for population estimations are given for just one figure. In the following case studies the estimations are made for a range of figures.↩︎

  5. The position of the original riverbeds of both rivers is difficult to ascertain, as huge hydro-engineering work was done to drain the area for opencast mining. Additionally, this information is considered as secret, so it is not available for the public. Today, the mouth of the Sokolitsa lies South of the town of Galabovo.↩︎

  6. The paths Atanasivanova mogila – Northern sites join the main North route. Therefore they share the site visibility already discussed in Galabovo case study (see above, p. 177, 179- 181) and it not discussed in Atanasivanova case study.↩︎

  7. During my museum study, I came upon a single mention of pits at Klisselika but no further comments were made.↩︎

  8. In Bulgarian excavation technique, pottery is not usually weighed. During the museum study, it was not possible to evaluate the amount of the pottery excavated during the late investigations. However, it is sure that it is more than 25 storage units, with sizes 54/22/25cm (0.7 cu. m.).↩︎

  9. In the site diary, the only funnel is mentioned as a find in the pit but the inventory book contains two more artefacts claimed to be found in the same pit, that are shown here on AFig. 5.8.4 A, C↩︎