(Translation : M. O'Farrell) Summary of Part 1: Introduction 1 - Research context The Kodžadermen-Gumelni?a-Karanovo VI (KGK) cultural complex developed in south-east Europe. It is principally located in the modern territory of Bulgaria, extending slightly into southern Romania (Danubian plain) and northern Greece (Eastern Macedonia and Greek Thrace). The highly mountainous relief is scattered with a few plains (Map 2) that delimit five principal regions: the North-East, the Thrace, the North-West, the South-West and the greek eastern Macedonia. The mountains encircling Bulgaria create a continental climate that differs little from that of Protohistoric periods, which was warm and humid at the end of the Neolithic and during the Chalcolithic. The sea-level, which had already risen significantly due to post-glacial climatic warming, continued to rise: at around –5000, the level was 15 meters below the current level, at –4000 it was 11 meters below, and at –3000 it was 7 meters below. H. Todorova, J. Lichardus and J.-P. Demoule propose three principal chronological systems, two with differing denominations for the periods (Tables 3, 4.1 and 4.2). These three authors recognize the Boian-Marica cultural complex, followed by the KGK complex in the east, and the Salcu?a-Krivodol-Bubanj Hum complex in the west. These ensembles constitute what J.-P. Demoule calls the “Graphite Ceramic Zone” (GCZ). The terminology adopted here concerns two levels: - at the synthetic level, the periods (Early, Middle and Late Eneolithic); - at the intra-site level, the cultures attested by the occupation levels (excavator's terminology). The dendrochronological calibration curves show strong undulations during the 5th millennium, resulting in a deviation of up to 840 years. Furthermore, many of the samples come from old excavations, and the great majority of these dates have very large deviations, ranging from ±100 to ±200. Whether calibrated or not, the dates obtained offer little information. The three periods overlap significantly and the late period appears to be contemporary with the entire duration of the early and middle periods. We thus retain the general estimation of an Eneolithic lasting approximately 600 years. The Early period, with the Boian-Marica complex in Bulgaria (thus including the “Middle Eneolithic” levels of Goljamo Del?evo and Ov?arovo), would have lasted from approximately 4900 to 4600 BC, and the Late Eneolithic from approximately 4600 to 4300 BC The Salcu?a-Krivodol-Bubanj Hum Ia complex is generally considered to be rather late in the Late Eneolithic period, contemporary with the end of the KGK. Generally, the Balkan Eneolithic covers the first half of the 5th millennium. 2 - Methodology The lithic material represents a coherent ensemble reflecting a balance between a group's need for a certain quantity and quality of siliceous stone objects and the technical choices available, and employed, from within the ensemble of the group's knowledge. While these choices are developed based on the technical knowledge of the group, they are also a function of its representations of this knowledge. Though influenced by physical constraints, these choices are largely the result of cultural rules, or to a lesser degree, “cultural motivations”. The lithic assemblage is thus rich in technical, economic and cultural information. This research focuses on the settlements and cemeteries of North-East Bulgaria during the Late Eneolithic. For comparative purposes, it also concerns the settlements of four other regions, as well as Early Eneolithic materials in the sites where they have been found. In this context, detailed reconstructions of the operational sequences of debitage, particularly through refitting, would have led to an overly fine level of analysis that is poorly adapted to the objectives. Therefore, based on technological readings of the lithic objects and reconstructions of the operational schemes (schéma opératoire), this study proceeded in four stages: - identification of the raw materials exploited and their origins; - recognition of the techniques of blank production, all laminar in this context, which represent the technical and conceptual skills and habits particular to a group; - characterization of the use of blanks through a preliminary classification of the tools made from them. This aspect reflects as much the collective mental representations of the form a tool “must” have as the more or less wide range of activities related to the use of these tools. These data were then organized according to three orientations: - the synchronic perspective at the supra-regional scale was based on comparison of the reference region (the North-East) with the other four regions, showing potential regional particularities; - the diachronic perspective, first regional then supra regional, was based on comparisons of Late Eneolithic settlements with Early Neolithic settlements in the five regions studied; - the integration of funerary data, emphasizing the role of knapped lithic objects within the ensemble of grave goods, which are all contemporary with the Late Eneolithic and known only in the North-East. For each of the sites studied, the entire assemblage of Eneolithic lithic materials was analyzed. The material is presented in the logical order permitting reconstruction of the large segments of the operational sequences (chaîne opératoire). The stigmata observed, when they are systematically associated, can reveal different technological groups. The interpretation of these groups in terms of knapping techniques is susceptible to modification. However, the technological groups thus distinguished remain significant of different knapping techniques. The propositions made concerning some of these techniques are based on comparisons between the archaeological materials and experimental objects produced by Jacques Pelegrin (CNRS UMR7055). The tool classification was based on the presence and location of macroscopic use traces on tool edges and on retouch characteristics. These types of traces are frequently present. Their extension and intensity often allow identification of non-retouched pieces that were used as tools, the distinction of intentional fractures from active parts of the tool, and the recognition of reutilizations associated with changing functions. The typology of these tools was developed based on a hierarchical association of several criteria: 1 - The presence of macroscopic use traces: dulling, splintering or use retouch, polish or shine, gloss. 2 – The location of traces when they develop on the retouch. 3 – The nature of the retouch when there are no traces, or when the traces extend over the entire piece. 4 – The sequencing of overlapping traces or retouch, or different shaping retouch events. 5 – The type of blank. 6 – Finally, in the case several overlapping retouch or macroscopic traces that are visible to the naked eye, the piece is attributed to the last tool type represented. This typology includes 16 tool classes, each regrouping one or several types. The material comes from tells. Mixings or intrusions between occupation levels (villages) in both vertical directions are always present and estimated to concern approximately 10% of the material. The study first proceeded level by level, and then by a synthesis of an ensemble of levels corresponding to the Early or Late Eneolithic period. Because they rely on the general coherence of the ensembles studied, it is impossible to identify recurrent traits in lithic productions based on only a few isolated pieces. Therefore, the small sample sizes from certain sites are considered only in terms of presence and never absence. Two of the settlements are reference sites (Smjadovo and Sitagroi). Each yielded more than 3,000 pieces (not including small debitage and retouch flakes), thus allowing a control for the representativity of the other sites studied. These two reference settlements are chronologically and geographically complementary. 3 - Raw material analysis An identical macroscopic description is made of the flints from all of the sites studied. An inventory of flint types is thus compiled and compared to currently available geological data concerning potential sources. Theses sources are then attributed to distances classes. Local, regional and exogenous distances are defined based on their spatial distance from the site and the time necessary to make a round trip to them and back, taking into consideration the landscape relief. These distance classes do not qualify the mode of access to the sources. It is often seen as evident that a local source implies direct procurement, while an exogenous source implies indirect procurement through exchange. However, the interpretative implications (economic, social or cultural) are more complex when we speak of “exchanges”, “importations”, or even “commerce”. Lacking clear identification of the precise origins of artifacts, and considering that several potential sites exist, the closest potential source is thus initially retained. In any case, with or without certitude concerning the location of procurement sources, technological study of all the varieties of flint in archaeological assemblages is necessary to provide decisive elements for these types of interpretations. Debitage analysis gives indications concerning the locations at which cores were prepared, shaped-out (mise en forme) and exploited, and consequently, the form in which the lithic objects were introduced into a site. It can also reveal different debitage productions. The simultaneous study of several Eneolithic sites within a vast zone complicated the task of identifying flint sources. On the contrary, it was a major advantage for interpreting the modes of access to the raw materials because it allowed comparisons of the distribution of these flints and the forms in which they appear. Summary of Part II : Settlements of the North-East region The five settlements analyzed, as well as the dozen other sites documented by surface collections, dredging or publication, are distributed over the entire North-East region (Map 107). Late Eneolithic debitage The raw materials exploited, consisting of brown-yellow flints, are attributed to potential local and regional sources. In all the sites, the debitage technology is characterized by indirect percussion for the production of wide, thick blade blanks, 10 to 15 centimeters long. The blades are curved in profile and have a large smooth butt with an incipient cone, a diffuse bulb and regular ridges. All stages of the operational sequence, except sometimes roughing-out (dégrossisage), are present, and were therefore practiced at each settlement, or in close proximity. These operations are carried out on a single, smooth striking platform. The roughing out, often performed by direct percussion (sometimes with a stone hammer) with no platform preparation, is not always performed at the settlement, but may also have occurred at the procurement site. The core is then prepared by the removal of large flakes by direct percussion, followed by crested blades by indirect percussion. The full debitage products are realized by indirect percussion with suppression or abrasion of the overhang. The full debitage phase begins very quickly on the still largely cortical debitage surface. The debitage platform is also sometimes cortical. Corrections of frequent hinges are made by the removal of elongated flakes in the same direction as the full debitage. The arch of the debitage surface (carène) is maintained by partial or unilateral crested blades, and its horizontal (cintre) and vertical (carène) arches by elongated flakes. This debitage constitutes the common technological base of all the sites, in spite of the diversity in its execution. These numerous elements (local flint procurement, local debitage, intra- and inter-site variability in execution) allow qualification of this domestic debitage. Alongside the domestic debitage, the sites yield a few blades made by pressure debitage. The observable elements indicate debitage on a single, smooth platform on a large, flat nucleus with a narrow debitage surface. The blades produced are long or very long (20 to 34 cm), wide and thin. Their profile is straight, with a slight curve at the distal end. Their smooth butt is very small, the bulb is distinct, and the ridges are straight. The length, (30 cm and greater) lightness, and regularity of these blades imply the application several hundred kilograms of pressure. Therefore, they could not have been produced by standing pressure debitage. Such pressure can be obtained only with a lever, a technique tested through experimentation. The sites never show evidence of the pressure debitage phases of the operational sequence and pressure blades are always rare, never exceeding 12% of the full debitage phase represented at a site. Therefore, these blades must have been brought to the site in the form of semi-finished products, even though the raw material used is identical to that used for indirect percussion. The production of very long blades by pressure debitage requires very large nodules of high quality, homogeneous, fine-grained flint. The sources of very high quality brown-yellow flint (such as those of the Razgrad region) are very rare. They are located within the local range of some of the sites studied, but at a much greater distance from others. The hypothesis of an extra-regional production of these blades is excluded since the raw material used is specific to the North-East. Two hypotheses must be considered for the identification of the producers of the pressure debitage blades (Table 108). In the first hypothesis (Table 108, 1), all of the settlements produce both domestic debitage and pressure debitage. This implies that: - each site would have free and direct access to a few sources of high quality raw material regardless of their distance from the site; - the nodules would have been knapped by pressure at a location different from that of the settlements ; - the knappers would have brought back to the settlement, either just a few blades sporadically collected from the series produced, or complete series of blades, but of which the majority was redistributed elsewhere. This implies that the redistribution of semi-finished products would be quantitatively greater further from the sources and the producers than in their proximity. According to the second hypothesis (Table 108, 2), all of the settlements produced domestic debitage, but only some produced pressure debitage. The habitants of the sites with only domestic production, and where we find semi-finished pressure debitage products, would have procured small quantities of blades from the sites where long blades were produced by pressure debitage. This hypothesis implies that: - only a small number of sites would have had free and direct access to the few sites with high quality raw material; - these production sites would have been the only ones at which pressure debitage was practiced, particularly lever pressure, and from which some of these products were distributed to other sites in the region. None of the sites studied can be identified as production sites of long blades by pressure, except perhaps Top?ii, or especially Kamenovo. These are the only sites that yielded pressure debitage waste products. We retain the second hypothesis of production by a few specialized sites in the North-East. This hypothesis will be supported or invalidated by the third part of this research consisting of the comparative study of settlements in other regions. Late Eneolithic tools Most of the tool blanks were produced by the domestic debitage by indirect percussion. The tool assemblages, consisting mostly of blade tools, are comprised of three principal types: scrapers, utilized blades and sickle elements. Scrapers are always the most abundant. Other types, such as burins and retouched blades, are always present but in smaller proportions. The remaining tools consist of a few retouched flakes, splintered pieces, pointed blades, truncated pieces, flint hammers (sometimes in the form of reused domestic cores) and arrow or spear points with variable dimensions and forms. Finally, knapped flint axes are present at most of the sites. Some were made from specialized debitage cores and, along with the full debitage blades, represent the other pressure debitage product of the lithic assemblages of these sites. There is no indication of shaping (façonnage) at these sites, despite the presence of broken axe edges and resharpening flakes. Therefore, the axes must have been introduced into the settlements as finished products. The tool assemblages are very homogeneous from one settlement to another, but with variable tool proportions, particularly of the three principal types. These variations suggest that similar activities were practiced at the settlements, but in different proportions. This indicates the practice of preferential activities. During all of the Late Eneolithic, neither the debitage, nor the tools show a chronological evolution and there are significant similarities among the sites. The debitage and tools are homogeneous in the KGK-VI region of the North-East. The preferential activities probably reflect a balance between the productions of different settlements, thus indicating a network of relations between the sites. The hypothesis of preferential activities implies that the KGK sites are closely related from an economic point of view. Diachronic perspectives With the exception of Smjadovo and Strašimirovo, all of the sites excavated contained Early Eneolithic occupation levels and/or so-called “Middle Eneolithic” levels. These levels yielded a domestic debitage by indirect percussion and blades made by pressure debitage, the latter arriving at the sites in the form of semi-finished products. In comparison to the Late Eneolithic, the Early Neolithic domestic debitage produces shorter, thicker and more curved blades with less regular ridges. We observe sporadic preparation of the striking platform by faceting during the debitage process. For the remaining phases of the operational sequence, the procedures are identical to those of the Late Eneolithic. The blades produced by pressure debitage during the early periods are as long as those of the Late Eneolithic. Within the ensemble of full debitage products, they represent the same percentage as those of the Late Eneolithic. The blades arrive at the settlements in the form of semi-finished products. The principal elements of the Early and “Middle” Eneolithic are tool assemblages already scrapers, sickle elements and utilized blades. Retouched blades, burins and truncated pieces are also well represented. The flakes are sometimes used or retouched. The settlements have variable tool proportions, again suggesting the practice of preferential activities. Propositions for interpretations of the Eneolithic industries of the North-East All the data allow identification of the principal characteristics of the Eneolithic lithic industries in the North-East (Table 109). Though all of these characteristics have multiple and complex meanings, we can emphasize the primary meaning (the most immediately perceptible), whether it be of a cultural, chronological or economic nature. First of all, we observe a significant common base, characterized by practices that are maintained throughout all of the Eneolithic. These are common technical traditions, but also economic networks comparable to relations between villages. The technical tradition, which participates in a common technical knowledge through space and time, constitutes one aspect of what we can more generally refer to as a cultural tradition. This homogeneity throughout the Eneolithic is coherent with the research of J. Lichardus, who considers the ensemble of the Early Eneolithic (Pan-European Late Neolithic): a Boian-Marica cultural complex composed of the Marica cultures in Bulgaria and the Boian cultures in Romania. A few changes seem to occur in the domestic debitage through time, including an improved mastery of debitage techniques and a relative diminution of striking platform preparation techniques (abandonment of occasional faceting). In the tool assemblages, the systematic, though slight, predominance of some tools such as scrapers, and the appearance of new tools such as knapped axes, also appear to be chronological characteristics. They perhaps reflect a greater development of certain activities. The near omnipresence of pressure debitage blades may reflect a constant need to have a small quantity of long blades (produced by specialists) on hand. Though these blades seem to have a non-specific function, they appear to be an economic and social necessity. This necessity seems to be associated with several levels of meaning. It indeed plays a role in both the cultural tradition (technical knowledge, practices, social relations) and within the network of economic relations. The economic networks constitute not only economic relations, but also social relations between the sites. Analysis shows that they are ingrained in cultural traditions. At least from the perspective of lithic tool production, the economic networks are organized into partially specialized and fully specialized productions, the former consisting of preferential activities at the settlements, and the latter of the specialized production of long blades. For these economic networks to function during the entire Eneolithic (at least half a millennium), a balance between the different productions would have been necessary. This balance suggests that there was a strong cultural, economic and social integration of the settlements in North-East Bulgaria during the fifth millennium. Summary of Part III: The settlements of the other regions The sites of Thrace yield a domestic debitage that is identical to that of the North-East, including: frequent exploitation of local raw materials and only occasional exploitation regional ones; light roughing-out, perhaps at the raw material source, and the remaining debitage conducted at the settlement or in close proximity. During the shaping-out phase, a smooth striking platform is prepared, perhaps accompanied by frontal or lateral crests. The debitage products are long blades (approximately 10 to 12 cm) with a curved profile, a smooth, generally thick butt with an incipient cone, an overhang that is often removed or abraded, and regular ridges. Maintenance of the blade debitage surface is always realized in the direction of the full debitage. The long blades produced by pressure debitage originate in the North-East and are brought to all of the settlements in the form of semi-finished products, in the same proportions as at the north-eastern sites (approximately 10%). The three principal tools are scrapers (more or less predominant), sickle elements and utilized blades. Retouched blades are well represented, as are burins to a lesser degree. The remaining tools include splintered pieces, pointed blades, truncated blades, utilized or retouched flakes, and diverse spear and arrow points. Finally, exogenous flint axes originating from the north-oriental region are present. The sites have clearly similar tool assemblages, but with different proportions of tool types, perhaps indicating preferential activities. The Thrace and the North-East regions have the same common tool base during the Early and Late Eneolithic. This is true for the domestic debitage and tool assemblages, for the procurement of raw materials for the domestic debitage, for the circulation networks and relations between long blade production sites in the North-East and non-blade production sites the North-East and Thrace, as well as for tool production. The circulation networks of long blades during the Early Eneolithic, and then long blades and knapped axes made from cores during the Late Eneolithic, are stable. The two regions are thus integrated within the same ensemble, which is coherent with the cultural entities defined through decorated ceramics. There is no significant difference that can be correlated with either of the local ceramic cultures, and this to such a degree that the identity of the lithic productions of the North-East and Thrace seems even stronger than the affinities shared between these two regions from the perspective of ceramics. This leads us to consider the North-East and the Thrace as a single region, which correlates with the Early (Boian-Marica) and Late (KGK) Eneolithic cultural complexes. The results of this study of the Thrace zone support the hypothesis that a few of the north-eastern sites were specialized in the production of very long blades. This specialized lithic production played a sufficiently important role for the long blades by pressure debitage to be distributed over nearly all of the sites in the eastern half of Bulgaria, thus circulating over distances exceeding 200 kilometers as the crow flies. Throughout the Eneolithic, the three other regions—the North-West, South-West and northern Greece—are clearly differentiated from the eastern half. For the domestic debitage, only the South-West seems to exploit a majority of regional flints, while the North-West and northern Greece procure local materials. Roughing-out, shaping-out and debitage are all realized at the site or in close proximity, except during the early period in northern Greece when roughingout and shaping-out are realized far from the settlement. In terms of debitage, all three regions produce rather short, thick blades, generally 6 to 10 cm long, that appear to be made by direct soft percussion. Another common point that distinguishes them from the eastern half is the common practice of maintaining the striking platform from the flanks or the base of the core. The domestic debitage of the North-West, South-West and northern Greece are thus similar in several ways, during the early as well as late periods. They differ, however, in two essential aspects: the practice of faceting and the occurrence of long blades made by pressure debitage. Faceting is increasingly practiced the further south we descend: it is frequent in the North-West, predominant in the South-West, and only attested in northern Greece. The long blades made by pressure debitage, originating in the North-East and imported in the form of semi-finished products, have variable proportions according to the sites, regions and periods. In contrast to the eastern half, no whole long blades were found. The longest fragments are only around 15 centimeters long. Moreover, while they seem to be present in every settlement of the Early Eneolithic, they are absent during the late Eneolithic in the South-West and North-West. Long blades represent a maximum of 7% of a settlement's full debitage in northern Greece during both periods and in the South-East in the Early Eneolithic, but attains only 3% in the North-West during the two periods. In terms of tools, the North-West shows greater differences, while the South-West and northern Greece are similar in many ways, both to each other and with the eastern half. In the south-western and southern zones, as in the eastern zone, scrapers are strongly represented in the assemblages, and utilized blades to a lesser degree. On the other hand, mesial fragments are preferred as blanks for scrapers (the predominant tool) in the western half, while proximal fragments, or whole blades, are more often chosen in the eastern half. Burins are nearly absent in the assemblages from the western half, while they are rather numerous in the eastern half of Bulgaria. In the Late Eneolithic, triangular arrow points with a deep basal notch are attested in the North-West and South- West, but are absent in the Eneolithic of the eastern half. No knapped axe on a pressure debitage core has been found at any site in the whole western half during the Late Eneolithic. Finally, sickle elements, made on blades or elongated flakes, from the South-West, northern Greece and the eastern half, show an oblique extension of gloss, while in the Early Eneolithic in the North-West they are on flake blanks with gloss parallel to the distal edge. During the Late Eneolithic, the domestic productions and tools of each of the three regions are clearly rooted in the local tradition of the preceding period. On the other hand, inter-regional comparisons, particularly with the eastern half, present a contrasting image in terms of the tools on one hand, and the circulation networks of goods on the other. The networks of relations in the North-West are not equivalent to those revealed in the eastern half. Though the circulation density of pressure debitage products in the North-East and Thrace was interpreted as a sign of a strong economic integration of the settlements of these two regions, the North-West sites are not strongly integrated (at least from the perspective of lithics) in the network of relations in the eastern half of Bulgaria. Therefore, the dissimilarities are associated with the ensemble of tool production in the North- West, and thus with a portion of the technical system itself, during the entire Eneolithic. The particularisms of the North-West reflect other traditions, conceptions and technical skills of debitage and shaping, as well as other cultural traditions and representations of tools. In the Early Neolithic, ceramic studies distinguish the Gradešnica culture in the North-West and the Pasarel group in the Sofia Basin, which present affinities and are partially contemporary with the Marica culture in Thrace and the Sava and Poljanica cultures in the North-East. No difference is perceptible between the lithic industries of the Sofia Basin and the North-West but there are clear differences with the Thrace and the North-East. During the Late Eneolithic, in terms of ceramics, the Krivodol culture of the North-West and the Sofia Basin are integrated in the Salcu?a-Krivodol-Bubanj Hum Ia complex. Though numerous ceramic elements are shared with the KGK, notably with the Thrace culture of Karanovo VI, the differences in ceramic decors and forms are sufficient to distinguish two cultural complexes. On the whole, the lithic tool production is in accordance with the different cultural denominations. The similarities between the South-West and northern Greece on one hand, and the North-West on the other, are most pronounced in the lithic debitage. Meanwhile, from the perspective of networks of relations, the integration of the South-West and northern Greece in the same network as that of the eastern half is more significant than that of the North-West. During the Late Eneolithic, these networks appear to remain stable in the Eastern half and concern all sites, though the contribution of long blades is possible but not proven. In any case, the networks of relations decrease in importance during the Late Eneolithic in the western half; either they continue to exist at only a few sites and in low proportions (North-West and perhaps South-West), or they possibly disappear in the South-West. The South-West shows significant regional particularities in both debitage and tools. From the perspective of ceramics, the Early Eneolithic presents strong similarities with the Thrace culture of Karanovo V, which belongs to the Boian-Marica cultural complex. It is sufficiently different, however, to merit the distinction of the Slatino “group” or “culture”. Still, while similarities exist between the Pasarel group in the Sofia Basin and the Gradešnica culture in the North-West, differences are also observable. During the Late Eneolithic, the “Pernik group” in the South-West is distinct even though it shows similarities with both the KGK complex in Thrace and the Salcu?a-Krivodol-Bubanj Hum Ia complex in the North-West. Here, the ceramic and lithic comparisons do not completely coincide. Throughout the Eneolithic, the South-West differs from the other regions in terms of debitage technology. During the Late Eneolithic, economic parallels are present in the tools, essentially with the eastern half. A unique point of stylistic comparison with the North-West is attested. Consequently, the domestic debitage, tools and long blades attest to a common cultural and economic tradition in the South-West and in northern Greece. In terms of lithic industries, this remarkable homogeneity incites us to integrate these two regions into a South-western ensemble that is distinct from the North-West and the Eastern half in terms of domestic debitage, but that is better integrated than the North-West in the network of relations of the Eastern half during the Early Eneolithic. It seems to be much less integrated in these networks during the Late Eneolithic. Summary of part IV: Funerary ensembles 1 – Funerary practices Funerary ensembles are documented only in the North-East (Map 259). Almost all of them belong to the Late Eneolithic. The two ensembles of Varna, along with that of Reka Devnja, are the only ones for which the settlements are unknown. 1.1 - Tombs The same skeletal positions are encountered in both periods: flexed, extended, and in a secondary deposit. Symbolic burials (containing objects but no skeleton), appear in the Late Eneolithic. They were simple or with masks (Table 260). The proportions of different tomb types are not known for the Early Eneolithic. During the late period, extended burials can be absent or predominant in small funerary ensembles but are always predominant in the large necropolises. Secondary burials are relatively rare. They perhaps represent the last traces of a practice held over from the Early Eneolithic. This practice was gradually replaced by symbolic tombs. The simple symbolic tombs are present in both the large and small ensembles and are much more numerous than those with masks, which exist only at Varna I. The orientation of the tombs varies within the same necropolis, as well as from one funerary ensemble to another (Table 261). 1.2 - Grave goods The materials that compose the grave goods were already diverse in the Early Eneolithic, but became even more varied the Late Eneolithic (Table 262). All of the materials found in the Early Eneolithic are also found in the Late Eneolithic. Most of them have potential local origins and almost all come from the North-East. But from the beginning of the Early Eneolithic, exogenous materials were also present: Thrace copper, and especially, spondylus artifacts and dentalis shells, whose nearest origins are the North-Aegean coast. These were accompanied by obsidian, whose sources are Tokai in Hungary, nearly one thousand kilometers away, or more likely Milos in the Aegean, approximately 800 kilometers away. The Late Eneolithic shows a general increase in the diversity of materials, but these materials had essentially local or regional origins. Tools are also more diverse during the Late Eneolithic: flint or copper arrow and lance points, and preforms of copper tools. But the most diverse class is that of “objects” (for definition, see p. 225), which includes figurines, diverse forms of ornaments, scepters, nails, diadems, tips, tubes, miniature axes, etc. (Table 215). The object categories (object type according to its material) show that the diversification observed in the Late Eneolithic particularly concerns metal objects (especially gold) and beads. The beads are made from gold, bone and spondylus in the Early Eneolithic, and joined by marble, black mineral, malachite, quartz, kaolin and snail shells in the Late Eneolithic. The lithic objects included indirect percussion debitage products, which are present in nearly all of the Early and Late Eneolithic tombs, along with long blades made by lever pressure debitage. It seems that the long blades, unretouched, were produced with the intention of immediate inclusion in human burials. This study of the funerary ensembles also revealed a third technological group: pressure debitage with a crutch. This production is evoked, though it has not been formally recognized, in the settlement material of the North-East. The majority of pieces deposited in both the Early and Late Eneolithic tombs are whole blades or fragments, which are most often unretouched. The only retouched tools represented in the Early Eneolithic are double truncated blades (tranverse arrowheads), while Late Eneolithic tools included retouched blades, burins, scrapers, retouched pieces, and arrow or lance projectile points. During the Late Eneolithic, the long blades were most often placed alongside the arm or near the thorax, or sometimes the hip. The blade fragments or shorter blades tended to be placed near the head. 2 - Social organization 2.1 - Varna I During the Early and Late Eneolithic periods, grave goods varied greatly from one tomb to another, both within and between ensembles. These differences, which are particularly evident at Varna (Varna I), have long been interpreted as representing of a hierarchical society. We thus considered it indispensable to define the degree of richness or poverty of each tomb. Counts of object categories per tomb show the qualitative and quantitative differences between grave goods. Our results show that the tombs contain from 0 to 25 object categories (Table 216 and Figure 217), from which we defined five classes of richness (Table 264). Very poor tombs are more numerous than very rich ones. However, while very rich tombs represent only 3.7% of all tombs, they contain 74.7% of the object categories. They also contain the largest number of objects per category. The majority of categories found in the poor and medium rich tombs were also included in the very rich tombs. Among the 32 categories exclusively present in the rich and very rich tombs, more than half (19) are objects that are unique within the entire necropolis. Others are present in at least two tombs and can be considered as specific to the rich and very rich tombs. As with other materials, gold is not exclusively reserved for the rich and very rich tombs. The material is therefore not a distinct symbol of power. The general multiplicity (objects, materials, number of individual object types within tombs, complexity of associations) is also evidence of longer or more complex fabrication sequences in the richer tombs. The maximum length of long blades increases with the richness of the tomb (Figures 234 and 266), even though shorter blades are present in all the tomb classes. The presence of longer blades, along with indirect indications such as fragments of different blades lined up to imitate a whole blade, show the value accorded to the notion of length in the funerary ritual of the Late Eneolithic and within the society.We found no link between the richness of a tomb, which is a symbol of status and power, and the producers of long blades by lever debitage. Very long blades had a completely symbolic value associated with high status or power. We found no correlations that could link objects with a particular sex, age or type of tomb. The only remarkable tendencies concern correlations between object categories and tomb classes. The richer the tomb, the more diverse are the materials contained within. The materials are thus both exogenous and local or regional. The very rich tombs “drained” all of the possible material sources, from those of Varna region to those from south of Aegean Sea. We thus propose the hypothesis that power was not based on the production of goods, but on the control of their circulation, which is not contradicted by the geographic situation of Varna. The social hierarchy expressed by the Varna necropolis seems to concern children as well. We must thus consider, at least in this region, the existence of inherited power as early as the Eneolithic, or at least the possibility that the high status of a parent could extend to their children. 2.2 – Comparisons with other Bulgarian funerary ensembles Social hierarchy did not originate at Varna I; it was already present in the tombs of Varna II, dated to the Early Eneolithic. With the exception of symbolic tombs, the same characteristics are already observable. One of the notable differences in the grave goods is the high representation of copper body ornaments at Varna II, compared with an under representation of copper tools. This relation is reversed in the Late Eneolithic. Another difference is the less frequent presence of gold in the rich goods and its lower diversity; during the Early Eneolithic, it is found only in the form of beads. The necropolises and funerary ensembles of the Late Neolithic (contemporary with Varna I) all show social differentiation in their grave goods (Table 267), with the same characteristics in equivalent richness classes. In all of these ensembles, richness is also characterized by the multiplication of objects, categories and materials. Similarly, the categories specific to rich and very rich tombs are absent from the poorer tombs. The grave goods of the Late Eneolithic of North-East Bulgaria are thus very homogeneous in their composition according to different classes of tombs. The heterogeneity results from inequalities from one site to another; Varna I is clearly the necropolis containing the richest tombs of this entire region. This notable differentiation between sites can be interpreted in terms of a hierarchical relationship between them. Summary of Part V: Synthesis and conclusions The organization of artifact production, here flint artifacts, is an indispensable element for characterizing of the economic aspects of a society. We use the term domestic unit to refer to the smallest spatial unit of production perceptible, here the village. Domestic production is defined as the production of a domestic unit destined for its own use. The term non-domestic production refers to all production occurring outside the domestic unit. Beyond the domestic or non-domestic character of a production, it is also necessary to know whether it is produced by everyone or by certain individuals only. The term specialist is used here in the general sense of an individual with skills and knowledge in a precise domain, which leads others to consider him or her as particularly competent. It does not imply a stratified social organization or economic structure. In this context, specialized production, along with other productions, can represent part of the domestic production. Specialists work for the benefit of the whole group. Their specialty does not give them a permanent status or an accumulation of riches. It is nonetheless evident that the degrees of “complex labor division” must be distinguished in order to identify economic and social differences between societies with specialized productions. At one extreme there is specialization, and at the other, professionalization, which can no longer be part of a domestic production. The varied degrees of complex labor division, conceived as a theoretical continuum, can coexist in the same society. Professionalization is defined as the existence of production activities practiced by only certain individuals. It is part of the stratified economic structure of a society in which the economy does not function at the scale of the village, but at regional or supra-regional scale. In the extreme form of this type of economy, certain individuals devote almost all of their time to one activity, requiring them to receive in exchange the subsistence means that they lack. The modern connotation of this term, particularly monetary, implies an accumulation of reserves, but it is necessary to distinguish the organization of production from the value of exchanges (the contract). Furthermore, the existence of money is not directly subject to the existence of professionals. Finally, the more common term craftwork is no less delicate than the term professionalization, though it is perhaps even more ambiguous. Regardless of its access to raw materials, the production of a given village can take two principal forms, domestic or nondomestic. It can be non-specialized or organized according to a complex labor division, ranging from specialization to professionalization. Tool blanks can be exchanged, used in their unmodified form, or transformed into tools (Figure 270). This technological study of debitage from the Bulgarian Eneolithic permitted distinction of a domestic production and a non-domestic production. The domestic production, recognized in all regions, is highly dominant (90% of the material from individual sites). It was practiced mainly with local flints, and to a lesser degree with regional flints. This domestic production differs according to different regions in terms of debitage techniques and the morphology of the blade products. Within each of the five regions, there is a debitage common to all the sites, representing technical skills and conceptions shared by all of the settlementss. No evidence has yet been found for specialized productions within the domestic productions studied. The roots of the regional domestic production during the Early Eneolithic, with no fundamental modification during the Late Eneolithic, allow us to consider it as traditional. Tool blanks made from domestic production activities are frequently found among the grave goods (Late Eneolithic, North-East), yet they represent only 30% to 40% of the worked lithic material. A roughly east-west division appears, between the North-East and Thrace on one side and the zone west of Bulgaria on the other. The North-East and Thrace regions constitute a clearly homogeneous ensemble from the beginning of the Early Eneolithic. The west Bulgaria is less homogeneous and seems to partially share a tradition, but with regional particularities. In general, preparation of the striking platform through faceting diminishes progressively from the south to the north and from the west to the east. Within the western half, we observe a progressive decrease of striking platform faceting from the south in eastern Greek Macedonia, until the north, in the North-West, while it remains very marginal in the ensemble of the eastern half. The great homogeneity of the eastern half of Bulgaria, which permits us to regroup the North-East and Thrace into a single large ensemble, is shown to correspond to the cultural division identified based on ceramics: the KGK complex, and the early period, the Boian-Marica complex. Alongside the domestic productions, many of the sites studied yield products from a non-domestic production. These products were made from the brown yellow flint of the Pre- Balkanic platform (North-East). They consist of long blades made by pressure debitage with a lever. These blades are at least 20 cm long. The longest attain 34 cm in the habitats and 44 cm in the necropolises. The “shorter” blades (12 to 20 cm long) were produced by pressure debitage with a crutch. There is no perceptible difference between the Early and Late Eneolithic, indicating that this non-domestic production is traditional. Pressure debitage with a lever imposes certain constraints: the near impossibility of an itinerant debitage, the availability of high quality raw material sources, access to those sources, and mastery of this technique. Except for one tell (Kamenovo) discovered after the completion of this study, and another one (Top?ii) that has been subject only to preliminary investigation, no habitat site from any region has yielded evidence of the first stages of the operational sequence of pressure debitage. The only pressure debitage cores found cannot be considered as production waste since they were all reworked to make axes. The whole or frag- mentary blades found at the sites were brought in the form semi-finished products, and the axes in the form of finished products. The five regions studied show more or less significant parallels (Table 272). It is possible to group the North- East and the Thrace regions since the proportions of long blades are greater there and remain stable during the two periods. Moreover, these pieces are attested in all the settlements. The three other regions differ as much from each other as they do from the eastern half of Bulgaria. They are generally characterized by the fact that the non-domestic debitage products are not found in all of the settlements. The non-domestic production of long blades attests to a complex division of labor and can be qualified as a specialized production. Realized only in the North-East, it is a highly sophisticated, probably non-itinerant production. It was practiced certain villages only in this region, and probably by only certain individuals within those villages. For the moment, we have found no element that would allow us to conclude that the sources exploited for the production of long blades were not also used for the domestic production. It is thus impossible to confirm that access to this raw material was indirect. The blades made by pressure debitage (lever and crutch) are more numerous in the funerary context of the Late Eneolithic, where they are also particularly long (Table 273). Though this specialized production is attested in all the regions studied, we exclude the term importation since it implies the import of a good from a location exterior to a given society. In all regions the majority of tools were made on blade blanks. The range of tools falls into 16 classes. It is rare to find all the classes in a single habitat. During the Early Eneolithic, the dominant tool varies from one site to another, while scrapers systematically dominate during the Late Eneolithic. Whatever the dominant tool, the proportions of each tool class vary from one site to another during both periods. This is attested in sites that were most often completely excavated and are comparable. This variation of tool proportions could reflect the existence of preferential activities in the sites. Knapped stone axes do not appear until the Late Eneolithic. They were made on pressure debitage cores or blocks. The axes made on cores, which are products of specialized, nondomestic productions, were brought to the villages in the form of finished products and sharpened in the villages. Apart from these axes and the systematic predominance of scrapers in the Late Eneolithic, the homogeneity between the Early and Late Eneolithic attests to the existence of strong traditions in the tool assemblages. No specific use of the long blades is observable. Consequently, the domestic tools of a village uniformly integrate the domestic production and the specialized non-domestic production. As already shown by study of the production, the grouping of the North-East and Thrace is confirmed by analysis of the tool assemblages. It is clear that these two regions form a coherent and homogeneous ensemble in eastern Bulgaria, which has the same technical and cultural tradition. The South-West and eastern Greek Macedonia can also be grouped into one south-western ensemble, whose rare dissimilarities are more closely related to variations in preferential activities than to real economic, technical or cultural differences. As in the domestic production, the North-West is distinct in terms of its tool assemblages. A parallel with the south-western ensemble is nonetheless observable. We found no elements suggesting that the domestic debitage products circulated from region to another. Only products of the specialized non-domestic productions (long blades made by pressure debitage and knapped axes made on pressure debitage cores) circulated. The term exchange is used here in the broad and “total” sense defined by M. Mauss and later used by C. Levi-Strauss and M. Sahlins. Exchange implies a relation between individuals, groups or societies. In an exchange relationship, the object exchanged is done so in function of a reciprocity (givereceive- give back), and correlatively represents a value. A considerable part of the value of an exchange concerns more the partners in the transaction than the objects themselves. An exchange is therefore a goal in and of itself, and not only a means of achieving an end. The exchange network observed between the five regions studied shows more or less strong relations. The availability of raw materials, the domestic production and tool assemblages in each of the regions do not indicate economic shortages or needs that would be fulfilled by the exchange of long blades between villages. Social interpretation of the lithic industry What then is the value of these blades, for which care was taken to make them as long as possible, which were exchanged all over, and which were eventually broken into pieces of the same smaller dimensions as the local tools? In all of the regions, the quantity of blades exchanged is relatively low and their quality is not the highest (not the longest blades). Is the simple presence of these blades sufficient to attribute them a value? A complete cohesion exists between the Graphite Ceramic Zone and the geographic distribution of long, pressure debitage blades. For the territories of the Graphite Ceramic Zone, which were not included in this study, the presence of these blades is noted in excavation publications. While long blades were exchanged within this zone, outside of it another lithic raw material, obsidian, was exchanged. The obsidian of Central Europe and the obsidian of the Mediterranean circulated over vast territories, encircling the Graphite Ceramic Zone but never penetrated it. The only known example is an obsidian blade (probably from Milos) found in a very rich tomb at Varna I. Even in the settlements located on the Aegean coasts, obsidian is absent from the levels attributed to the Graphite Ceramic Zone. A cultural value, in the sense of cultural identity, is thus revealed in the necessity of exchanging of long blades. The form in which long blades were exchanged (whole or fragmentary) varied regionally. One of the values carried by the long blades was their length, as is shown by analysis of the funerary ensembles. Moreover, long blades were proportionally much more abundant in funerary ensembles than they were in settlements. In the funerary ensembles, they were generally whole and sometimes particularly long. The grave goods of rich and very rich tombs are characterized by a greater quantity of objects, a higher diversity of object types and a greater diversity of materials. The composition of grave goods of rich, and especially very rich, tombs is principally distinguished by an overabundance of symbolic objects, ornaments and body ornaments. Moreover, the social hierarchy observed in the necropolises indicates that power could have been associated with children, even very young ones, suggesting that this power (or certain forms of it) was hereditary. We observe a direct link between the richness of deceased persons and the lengths of the blades (made through specialized production) deposited in their tombs. Meanwhile, we know from the settlements that blade length was entirely symbolic since all the blades were segmented at one time or another and no specific use of whole blades has been identified. Long blades were indeed produced by a specific, specialized technique that was not practiced in the villages. And it is this technique that was valued since the long blades had no indispensable utilitarian function. The pressure debitage technique may have been limited to a restricted production center because it is only there that artisans possessed the necessary skills. But it may also be because the occupants of the receiving villages refused to practice this technique, which could have motivated the need for exchange relations. The long blades also had a prestige value, symbolic of richness and power. This power belonged to those who could accumulate richness, which, as we see in their tombs, was represented by a great quantity of objects with diverse origins. All the elements of this power seem to have depended on the control of the circulation of goods more than on the control of their production. The power symbolized by long blades, which was also the power of the control of the circulation of goods, was above all possessed by the important persons of Varna. However, each village that possessed a few of these long blades (or only the individuals that possess them?) also benefited from a small part of this value, symbolic of prestige and power. This production was thus mostly professionalized and its value was more social and cultural than economic. The control of the circulation of goods, or of exchanges, therefore also played a role in the construction and reproduction of internal social relations. This was true for the KGK culture and, to a lesser degree, for the ensemble of the Graphite Ceramic Zone. Within this zone, long blades were one of the codes that permitted recognition. In this cultural (ethnic?) and social (tribal?) identity, long blades had a cultural and prestige value. They thus served as supports, not for tools, but for social relations. The Late Eneolithic of Bulgaria was integrated within a large pan-European horizon, both chronological and historical, which was defined by J. Lichardus and M. Lichardus-Itten as the “Early Chalcolithic”. Though the complex structural characteristics of the lithic industry appear to correspond to the Chalcolithization of the communities of the Balkans, this process was already well advanced in the Early Eneolithic (or Late Pan-European Neolithic). The passage from one period to the other was an in situ, continuous cultural process that was not associated with any massive population movement. It is thus not surprising that we find no punctual rupture. More generally, insofar as the Chalcolithization of the Balkans occurred in association with particularly spectacular manifestations, it is also logical that these manifestations had roots in the preceding period. This undoubtedly explains why Bulgarian researchers speak of the Eneolithic for the ensemble of these two periods. Finally, given that the lithic industry is by definition a very ancient tradition, we can suppose it was one of the domains in which the constitutive transformations of the Chalcolithization were initiated the earliest. It is, moreover, in part paradoxical that it was at the moment when metallurgy arose that the lithic industry in Bulgaria took its most spectacular forms and attained an unequalled technical apogee. Finally, in terms of inter-regional variability, the principal regional groups of the domestic lithic industry correspond to the divisions determined by ceramic typology, even though they are less fine. In particular, the lithic specificity of the North-West fits well with the ceramic specificity during the two periods (Gradešnica then Krivodol groups). This is also true for the affinities between the Bulgarian South-West and the Greek eastern Macedonia. The variations observed here in exchange networks between regions could also support possible reflections on the growing integration of different parts of the Graphite Ceramic Zone, or even on the beginning of a disintegration at the end of the period.