INTRODUCTION RAW MATERIALS

Transcription

1 es Abstract A series of more than 3000 chipped stone arti Thessaly. At the same time, characteristic changes can be seen such as: a greater role of local and mesolocal raw materials, manifested in the gradual atrophy of highly specialized lithic production that was earlier carried out in workshops and the transfer of production to household clusters. The continuation of these processes can be seen in the Early Neolithic of the western and northern Balkans which is of particular importance in view of the presence at Mavropigi of both painted as well as ceramics. Keywords: chipped stone, raw materials, lithic production, obsidian, silex blond INTRODUCTION The chipped stone industry at the Early Neolithic site of Mavropigi, comprising more than 3000 artefacts, is one of the largest inventories analyzed so far from the territory of northern Greece of this period. The raw material economy, technology, tool morphology, and organization of production within social structures have been analyzed according to the three phases of occupation on site dated at between 6,600 and 5,900 cal BC. This enabled the recognition of the phase of the Neolithic, dynamics which constitute an important part of the entire cultural evolution of the early farmers in the territory between the Aegean and the Adriatic basins. RAW MATERIALS Quartz, possibly of local origin, is best represented. This raw material, relatively of poor site in a full cycle of production. This is evidenced by the presence of cores and a numerous group of debitage quartz pebbles were worked using also splintering technique. The low quality of this raw material causes the rarity of quartz blades and retouched tools. Radiolarites, deriving most probably from a fairly small distance were, at least partially, worked This is documented by a greater quantity, more than blades, the presence of cores and chips, and the relatively low tool index (for

2 72. Most radiolarites could originate from about 50 km north of the site from the Voras mountain range comprising in part Mesozoic limestones. However, the nature of the were obtained from secondary deposits, namely from river gravels. Several types, of differing colour, have been distinguished, but at the present state of investigations we cannot ascertain whether they derive from the same deposit areas. The largest group are reddish radiolarites that occur as relatively large concretions in secondary deposits. The numbers simultaneously, the absence of cores indicate that complete blades or tools were brought to the site from specialized workshops situated in the vicinity of the raw material deposits, in all likelihood a fairly great distance away from the settlement at Mavropigi. Obsidian from the island of Melos is the most remote and important extralocal raw material found on site. Two types of obsidian a more frequent grayish, weakly transparent type, and a black, transparent type. As a rule the black variant, that occurred as larger concretions, was used for blade production. A possibility that the two types come from two different outcrops on Melos cannot be excluded. The island of Melos is situated at a distance of about km away in a straight line from the site. Possibly, obsidian artefacts (mainly as complete blades) found their way to Macedonia through Early Neolithic settlements in Thessaly where frequent use of this raw material was registered (Perlès, 2001). likelihood, obsidian blades were also endowed, prestige goods. silex blond played an important role (Perlès, until now not. Some scholars indicated possible sources in the or on the western coast is likely that silex blond originates from the Silex blond was also registered in layer X in the Franchthi Cave in the south (Perlès, 1989), Cyclope Cave on the island of Gioura (northern and, 2008). Tellenbach, 1983), Beotia (Sarakenos Cave Sampson., 2009), and in the Peloponese., 1996). The distance between the southernmost sites with silex blond and its most northern spread is about 300 km. Some researchers believe that itinerant specialized knappers produced long blades using their own stores of this raw material (Perlès, 2001). However, it seems more plausible that artefacts from silex blond circulated on the basis of exchange of various goods between the inhabitants of a chain of neighbouring settlements. The remaining raw materials (siliceous rocks of various colours and transparency, chalcedony, opals and quartzite), too, arrived in the settlement at Mavropigi most probably as part of exchange, but they are represented by only several, up to a dozen or so artefacts. Phase I Stratum 3 Merely 13 artefacts were associated with Phase two small bladelet paucity of the inventory in the oldest settlement phase could be the result of its destruction by younger settlement features. However, this could that the oldest settlement phase was a relatively brief episode, when the site was being settled by the group who had arrived in this area with its own with items obtained by exchange but also through attempts at exploitation of the raw materials near Among the exploited raw materials all the major groups that subsequently occur in the younger levels are represented. As the series is so

3 Early Neolithic settlement of Mavropigi in western Greek Macedonia 73 nonetheless, the presence of obsidian is noticeable (three three specimens including one four specimens, including three silex blond (one specimen) were registered. Local and mesolocal raw materials are: reddish radiolarite (4 specimens including three from One piece is burnt. Inventory structure sieving, almost a third are blades, followed by blade tools. Other artefacts occur as single items. The structure of lithics in this strata indicates workshops, typical for the Early Neolithic in mostly during limited tool rejuvenation. x 19 mm (Pl. 1.1), an initial, multiplatform core on a fairly large quartz pebble, measuring 74 x 69 x 58 mm (Pl. 1.2). ) coring angle of more than 90 degrees, from quartz, Blades are from obsidian (one specimen), greyish one), silex blond (one) and reddish radiolarite (one). One specimen is burnt. Blade width is between 6 to 17 mm, the majority are in the mode of mm just as the blades from the younger strata or punctiform, the coring angle is straight, with unidirectional dorsal pattern (except for one blade). The blades are regular, with parallel edges, detached from cores in advanced stages of reduction. Blade. 4) There were one retouched blade, an unretouched blade used as a sickle insert, and a burin. One burin spall has been also attributed to this level. a lateral truncation burin on a mediolithic cm long) blade, with transversal retouch, from obsidian (Pl. 1.3). a primary burin spall split off from a retouched edge of an obsidian blade (Pl. 1.4). a mesial fragment of a macroblade with denticulated inverse retouch, chocolate coloured a subcrested blade with parallel sickle gloss and inverse microscars; transparent, greyish Phase II Stratum 2 Stratum 2 provided 389 artifacts. The most numerous series are artefacts from quartz ( %). Other raw materials both mesolocal (radiolarities) as well as extralocal ( obsidian) play a minor role. Most quartz artefacts (210) come from the western where, in all likelihood, quartz production concentrated. The next raw material as to the number of artifacts produced is reddish radiolarite. Possibly, it was sometimes worked at the settlement, but the main body of artefacts were produced small number of chips found at the site (Fig. 48). Moreover, no concentrations of artifacts from zone of the settlement.

4 74 xtralocal raw material is greyish ifacts 5.2%), supplied as blades (11 specimens), and tools (5 specimens). Extralocal obsidian is also represented by blades only (11 specimens 2.9%). also artefacts from silex blond (8). Other raw materials (11 types of siliceous rocks macroscopically Inventory structur The structure of major technological categories Flakes and splinters together are more numerous than blades, and the tool index is relatively low. especially of artefacts from quartz. is 1.2% of the inventory. Four cores are from quartz and one from reddish radiolarite. Among the quartz specimens one is a small core fragment, and the others are initial cores, made on quartz pebbles without preliminary preparation. They are between 41 to 53 mm long, 28 and 41 mm broad and from 19 to 48 mm thick. The radiolarite core is an initial specimen for production, on a pebble, probably from river gravels (Pl. 1.7). All the cores were discarded in the early stage of reduction either because the raw material was of poor quality or the knappers skill was ) Stratum 2 provided 15 splintered pieces: from quartz 10, from reddish radiolarite three, one from dark radiolarite and one from silex blond (Pl. 1.9, 10). Flake or blade fragments were reduced as splintered pieces; thus, they are as a rule fairly small: length is between 19 and 38 mm, width shows greater variability between 8 and 45 mm. Only one quartz specimen is fairly large, measuring 64 x 66 x 20 mm (Pl. 2.1). Splintered pieces include: bipolar, tripolar as well as quadripolar specimens (Pl. 1.10). Scars short (Pl. 1.9). ) Stratum 2 yielded a large series of chips accounting for as much as 28.7% of the total. Most chips were discovered in the western (96 specimens 82.7%). Quartz chips are as much as 92.2% of the total number of chips. Such distinct Fig. 48. Structure of major technological categories in Phase 6 splinters, 7 chips Fig. 49.

5 Early Neolithic settlement of Mavropigi in western Greek Macedonia 75 processing, in all likelihood in the area around the western. Stratum 2 ly from quartz (74.4%). Besides quartz, a variety of different raw materials were represented, including a specimen blond. Flakes are relatively small in size: between 12 and 50 mm long, and between mm broad. They were detached from cores in advanced stages of reduction only one cortical and two with lateral cortex were found. Flake unprepared (25%). The technique of detachment of as much as 25% of the punctiform or linear butts. the upper strata the proportion of specimens with Fig. 50). at the settlement as there are no core tablets or The inventory contained, besides, 75 splinters, most from quartz (89.6%). They are smaller than mm broad. There were also or blades (6 from quartz, one from radiolarite, one from indeterminate rock), and four radiolarite and quartz chunks. ) The group of blades shows a great variety of raw specimens), specimens), obsidian (6 specimens), and quartz (8 specimens). Other raw materials, including silex blond, are represented by one to three specimens. The frequency of the various raw materials in the blade group is distinctly different from that in the Fig. 50. Table 1. Raw materials Flakes Blades 0,6 1,2 19,2 3,5 1,7 Obsidian 1,2 14,3 Reddish radiolarite 6,7 22,8 1,7 0,6 Yellow radiolarite 1,8 3,5 Dark radiolarite 0,6 5,3 Silex blond 0,6 5,3 Chalcedony 0,6 1,7 Quartz 81,5 14,3 Greenish radiolarite 1,2 1,7 Burnt and undetermined 2,4 5,2 To detach blades the pressure or the punch techniques were employed. The specimens produced using these methods had parallel edges (70%) and perpendicular butts, less often slightly

6 76 Most blades are broken. Providing blade fragmentation was an intentional operation then we can ascertain that the most desirable part of a blade was its mesial fragment without the thickest, proximal part and the slightly hinged distal part. Fragments like this account for 41.8% of all the broken blades. The proportion of proximal and of distal fragments is the same (23.2% each). However the index of blades with only the distal part broken off is slightly higher. Complete blades are between 22 and 72 mm long, and between 8 and 22 mm broad. The width of fragments indicates that they originated from blades of similar dimensions. The frequencies of butt types indicate that blow platforms. Butt retrimming or preparation was sporadic, but nevertheless more frequent production. Moreover, blades detached from cores with unprepared that resulted in linear or punctiform butts. The remaining blades are with a 90 or, less often, with an acute core angle (3 specimens). Only three blades had traces of cortex on the dorsal surface: two specimens had lateral and one distal cortex. The dorsal pattern indicates blade detachment from cores in advanced stages of reduction. Only one blade showed perpendicular dorsal pattern probably traces of preparation. This blade was made from A blade from reddish radiolarite had an opposite dorsal pattern. ) Stratum 2 provided 20 retouched tools and 7 sickle inserts without retouch. A variety of raw materials were used: transparent (9): mainly reddish variety (6) and single blackish (2), and violet (1) specimens. Other raw materials are less frequently represented: silex blond (4), 2 1), chalcedony (1) and obsidian (2). The raw materials that occur sporadically were used for the manufacture of burins (obsidian, chalcedony) and truncations The most numerous group of tools were blades with lateral retouch used for sickle inserts (5), as well as unretouched sickle inserts (7). a fragment of an on a convex blade. There are traces in the form of polishing and microstriations in the central section of the front (white ish transparent concave lateral retouch on the opposite edge (reddish radiolarite Pl. 2.6), distal retouch (chalcedony Pl. 2.7), mediolithic obsidian blade (Pl. 2.8). ) a transversal truncation on a macroblade sinister edge (Pl. 2.9), a double oblique truncation on a macroblade, silica gloss (Pl. 3.1). truncations, from black radiolarite (Pl. 2.10). an alternate perforator shaped in the proximal part of a bilaterally retouched macroblade, from transparent,. ) a macroblade with partial microretouch ; reddish radiolarite (Pl. 3.2), a distal fragment of a blade with partial obverse microretouch on one lateral edge; violet radiolarite (Pl. 3.3),

7 Early Neolithic settlement of Mavropigi in western Greek Macedonia 77 a fragment of an obsidian blade with mesial lateral retouch, denticulated retouch: one from dark radiolarite, the other from silex blond. 6) two fragments distal and proximal of mediolithic blades with obverse retouch on one edge and on the break (reddish radiolarite Pl. 3.4, 5), two mediolithic blades with parallel, bilateral silica gloss, and irregular microscars (one from reddish radiolarite, the other from grey,, a fragment of a macroblade with retouch on the proximal break; parallel lateral gloss (whit greyish, proximal oblique gloss (reddish radiolarite Pl. 3.9). a macroblade with splintered retouch in the distal part which resembles the technique by which inverse microretouch, and individual obverse scars ( a macroblade with two Clactonian notches on one edge. (Pl. 3.11). two blades fractured in the distal part, with oblique gloss on the fracture (one from smooth ish, three blade fragments with parallel gloss on one edge ( silex blond, transparent, two fragments of macroblades with bilateral, parallel and oblique gloss (one from silex blond, the other from dark radiolarite Pl. 3.16, 17). Phase IIIa and IIIb Stratum 1 The assemblages of the youngest phase of the settlement have been analyzed phases as one unit. Because the examination of pottery stylistics has shown that the lower portion of Stratum 1 somewhat different, this phase ( ) have been studied apart. No the artefacts and activities performed in the various zones of the settlement, or possibly even different occupational episodes. Raw material structure The youngest strata provided 2272 chipped stone artefacts. The assemblage is varied in terms of exploited raw materials: macroscopically 25 different raw materials were distinguished, represented by several, up to more than a thousand artefacts. Quartz, possibly of local origin, is best represented (1080 artefacts, i.e. 47.5% of the inventory).the low quality of this raw material causes rarity of quartz blades (3.7%) and retouched tools (1.4%) (Fig. 51.1). The second group of raw materials are radiolarites mesolocal origin. They site. This is documented by a greater quantity chips, and the relatively low tool index (Fig. 51.2). Radiolarite was used to make all the artefacts in the depot (area 801, feature 37) where about core reduction were deposited. This depot will be analyzed separately. cores (37) from Stratum 1 (P were made exclusively from quartz or radiolarite. The assemblage contained 160 obsidian artefacts i.e. 7.0% of the total inventory (Fig. 51.4). These comprise a high proportion of blades retouched tools (7.5%). Next is the proportion of weakly transparent with good cleavage, accounting

8 78 for 6.4% of the assemblage. The artefacts consist mainly of blades (45 specimens 30.6%), and tools (38 specimens 25.7%) (Fig. 51.3). Stratum 1 yielded 46 artefacts from silex blond which is 2.0% of the assemblage. Blades and tools account for as much as 86% of this (Fig. 51.5). Such an inventory structure is also found at other Early Neolithic sites in Greece and worked in specialized workshops. Major technological categories The structure of major technological categories is presented in Fig. 52. The most frequent group are together with the splinters they account for 32.6%. Blades (18.4%) and tools (9.1%) are less frequent. Also the ratio of splintered pieces (7.4%) and cores (1.6%) is low, but the chips (mostly The above list does not take into account chips the technological groups. This group of artefacts has never been employed at any other site Fig. 51. c blades, d tools, e splintered pieces, f splinters, g chips) Fig. 52.

9 Early Neolithic settlement of Mavropigi in western Greek Macedonia 79 territories (for example at the Early Neolithic site of Ecsegfalva in the Hungarian Plain Mateiciucová, s done the artefacts obtained by means of this procedure were, too, analysed separately. fragments: types of radiolarite one each; quartz 24 and a burnt specimen. Most cores (17 specimens) were found in sector 172. Most cores are on quartz chunks (11), also from reddish (1) and yellow radiolarite (1) and a burnt specimen from indeterminate rock. Reduction was restricted to several removals, and then the core was discarded. Cores are from about 30 to the case of 5 specimens the knapper intended to from quartz, the burnt specimen, a specimen from reddish radiolarite). The remaining specimens were discarded in a very early stage of reduction so that the intention of the knapper and the desired four specimens from reddish radiolarite, a specimen from grey radiolarite, and four specimens from quartz. (Pl. 4.1) are located on the narrower face of the situated on the narrower face of the concretion allowed to obtain fairly regular blades from quartz a raw material with poor cleavage. Cores from radiolarite are robust, with a fairly broad heavily exhausted; one specimen has lateral preparation (Pl. 4.2). Another core from radiolarite exhibits characteristic features of the Middle Palaeolithic reduction sequence; in all likelihood, it had been found by the inhabitants These cores had opposite platforms: 5 specimens are from quartz and 2 from reddish radiolarite (Pl. 4.4, 5). Two of these cores are in initial and two in the early stages of reduction (Pl. 5.1). Two cores are is a discoidal There were also core fragments, all from quartz. The majority of splintered pieces were made from quartz ( % of all splintered pieces), followed by specimens from reddish radiolarite ( % of all splintered pieces Pl. 5.4). Other raw materials comprise from one to 7 specimens. fragments from extralocal raw materials such as obsidian (Pl. 5.2,3) and silex blond (Pl. 5.5 from grave 7) were worked using the splinting technique. The high skill of knappers using this technique is evidenced by the presence of blades with bladelet scars detached by thin bladelets was recorded in the materials from Argissa (Perlès, s 13, 14). between 15 to 52 mm long, between 4 to 52 mm wide, and between 2 to 28 mm thick. Quadripolar specimens are rare in the examined series only 2 such specimens were found. Some researchers interpret splintered pieces as a special type of cores used for the production of small blanks, especially at sites situated at a greater distance from raw material deposits. Others treat splintered pieces as a kind of tools examined series both interpretations are plausible. from quartz ( % of all chips), which material was worked locally. Chips made from reddish radiolarite account for 10.3% of all chips. They come from tool repair, as well as of shaping tools. Obsidian chips are, too, relatively numerous

10 80 extralocal raw materials may derive from tool retouching (or repair). Chips were concentrated in sector 172 (450 specimens 64.5%). The chips described above come from the number of chips that were counted separately. sieving a total of 4300 chips were recovered from the site. Most chips are from quartz and other raw materials are few (e.g., obsidian 32 specimens, silex blond 7 specimens ). The proportions of chips from local raw materials and site, and tools from these raw materials were rejuvenated also at the site. The horizontal distribution of chips shows their concentration in the zone with the rectangular, overground dwellings, as well as in the areas between the dwellings north of the central were found, possibly detached in the course of exploitation of obsidian blanks. Flakes are the most numerous group in the inventory. Together with splinters and waste from core preparation they comprise 735 specimen, 32.6% of all artefact from this phase. The main splinters, i.e. 56.1% of the debitage). Production of quartz debitage concentrated in sector 172 near a splinters), in the same area where most of the chips were concentrated. Flakes from reddish radiolarite are the next most numerous group (124 specimens). The biggest concentration of dish radiolarite was, too, in sector there were no specimens from silex blond which was represented only by splinters. Flake size is between 15 to 64 mm, splinters are slightly smaller from 12 to 54 mm (Fig. 53 a, b). Cores from which the almost without preparation. Flake butts, especially Flakes from reddish radiolarite are more of core tablets and only one platform trimming cores. The differences in the frequencies of butt types between radiolarite and quartz specimens both are raw materials worked locally is the effect of raw material properties rather than that of different technological traditions (Fig. 54). A hard hammer was used t ; this is documented by a high proportion of specimens with an obtuse core angle slightly higher in the case of radiolarite specimens than in the case of small: they account for merely 3.4% of all 5.7%. Analysis of dorsal patterns shows the (42.6%). These are followed by the dorsal pattern were probably obtained from change of core orientation rather than from core preparation. Among number of quartz and radiolarite specimens is almost equal, but taking into account the orientation cores were mostly radiolarite. Only 18 on the dorsal face. blade) which is 18.4% of the inventory. The largest series were blades made from reddish radiolarite, numbering 111 items (Fig. 55). Blades from this raw material account for 26.5% of all blades, whereas the overall proportion of reddish radiolarite in the inventory is about 18%. The second largest group are obsidian blades (85 blades 20.3% of all blades, whereas obsidian artefacts account for 7.0% of the entire assemblage). There were 41 blades from quartz (i.e. 9.8% of all blades, whereas quartz artefacts account for 47.5% of the assemblage in Stratum 1 Pl. 5.9). A less numerous group are blades from silex blond (21 specimens, i.e. 5.0% of all blades). Other raw materials are represented by single blades.

11 Early Neolithic settlement of Mavropigi in western Greek Macedonia 81 Fig. 53. (Startum 1): a length, b width Fig. 55. Raw material frequency (a) in the whole inventory of Fig. 54. A major part of blades was found in sector 172 between the rectangular houses north from the central (146 specimens) and in sector 801, mostly in central (71 specimens). Sector 117 yielded a smaller number of blades (46 specimens), but in this sector the features of sparsely inhabited, to the west and south from central, yielded 47 blades and 30 blades respectively. Most blades have been preserved as blade fragmentation was intentional or whether unretouched blades were accidentally broken during use. A number of blades show nibbling signs. Mesial fragments predominate (47.9%), while at most Neolithic sites proximal fragments are most frequent. There were only 44 complete blade, between mm long; the most common length

12 82 mode is mm. The largest blades (more than 50 mm) are only a fraction (9%) of all blades. by the comparison of the width of complete and fragmented blades (Table 2). Blades preserved as fragments independently of the raw material were to a greater degree standardized, with similar proportions and probably similar dimensions. They were detached technique was used on extralocal raw materials ( silex blond ) and on raw materials worked could have been commonly known in the Early Neolithic and used by both, local and itinerant, specialized knappers. As a rule blades were detached from cores the frequency of blade butt types (Table 3). No major differences have been registered in the method of platform preparation depending on proportion of punctiform and linear butt together the series of examined artefacts was very small merely 18 specimens). Narrow blade butts are usually perpendicular to the axis, and the core angle is straight in 85% of the blades. Almost all the blades come from advanced stages of core reduction. Three blades were cortical, and 8 with lateral cortex. As far as dorsal pattern is concerned: 95% of specimens show unidirectional scars, only 3 show opposite scars, and 3 are with perpendicular scars. Bulbs are fairly distinct, although nearly 20% of the As we have already mentioned blades and fragments were often used without retouching. Analysis of the width of sickle inserts indicates that fairly standardized blades were selected. The most favoured raw material for sickle inserts was reddish radiolarite; quartz was not used. Retouched and unretouched tools Raw materials structure of tools The most frequent raw material used for tool production was (6 out of 13); it is less frequent among blades with lateral retouch used for sickle inserts (7 out of 26), denticulated tools (2 out of 15). Single specimens The next, most common group of raw materials used for tools are radiolarites of various colours (reddish, yellow, nearly black). They are best represented among truncations (10 out of 16), retouched blades used as sickle inserts (10 out of 26), and unretouched sickle inserts (18 out of 45). Sporadically radiolarites are represented Specimens from extralocal raw materials i.e. silex blond and obsidian are few but they are distributed in a number of tool groups. Silex blond occurs among burins (5 out of 16) and Table 2. Width of complete blades and blade fragments from 1) Width Complete blades <4 0,3 40,9 26,3 36,3 52,3 17,3 20,4 2,2 2,5 0,9 Blade fragments Table 3. Butt type N % Unprepared butts 9 6,3 Single blow butts 77 53,8 Dihedral butts 6 4,2 Facetted butts 16 11,2 Linear butts 15 10,5 Punctiform butts Total 143

13 Early Neolithic settlement of Mavropigi in western Greek Macedonia 83 burins, truncations, blades with marginal retouch used as sickle inserts and unretouched blades occur as single items. Obsidian specimens occur among truncations (2 out of 13), retouched blades (3 out of 17), and unretouched sickle inserts (3 out of 45 this proportion could be higher, but identify this function macroscopically). There are individual obsidian burins and specimens with Quartz tools are relatively few (14 specimens) four one), and a one). Other raw materials among tools are represented rocks (brown, chocolate coloured, Cretaceous, Tool morphology ) a specimen on a microlithic blade, with an asymmetrical, rounded front (Pl. 6.1). The proximal sections of lateral sides show crushing the front (rounding of the edge and perpendicular striations). Made from, weakly, lateral retouch (Pl. 6.2). Made from quartz, high, front. Made from radiolarite (Pl. 6.3). polishing, made from blackish radiolarite. 6.5), made from pink quartzite. poles of the splintered piece (Pl. 6.6); made from silex blond. a fragment of an atypical, dihedral burin made in the proximal part of a blade from Cretaceous, part there are two lateral burin scars from the butt, whereas in the distal part there are two form of microscars and microstriations on the (microscars) also on the central section of the lateral edge. Made from, weakly, (Pl. 6.9), a microlithic obsidian bladelet (Pl. 6.10). mesial fragment of a mediolithic blade from, mediolithic blade from yellow radiolarite (Pl. 6.12), a lateral angle burin on a splintered piece from green radiolarite (Pl. 6.13), obsidian (Pl. 6.14), a single blow burin made on the distal part of a bladalet with lateral denticulated retouch (Pl. 6.15). White greyish on the proximal part of a microlithic blade with bilateral retouch; silex blond. Next to the burin scar there are microtraces from crushing which suggests this burin is accidental (Pl. 7.1), a single blow, transversal burin made on the distal part of a mediolithic regular blade from shaped as a result of the pressure on the lateral edge (Pl. 7.2), a blade from silex blond with the proximal blow burin scar detached from the base (Pl. 7.3), three fragments of burins from: transparent radiolarite. blade, used as a sickle insert on its left edge, opposite to the burin scar (Pl. 7.4), a macrolithic blade worked by splintered

14 84 the left lateral edge. The tip of the burin shows sickle gloss (Pl. 7.5). Made from beige, weakly mediolithic blade used as a sickle insert (Pl. 7.6)., a mesial truncation burin shaped in the distal part of a macroblade from reddish radiolarite Pl. 7.7), mediolithic blade used as a sickle insert (Pl. 7.8); from silex blond,, subsequently used as a splintered piece (Pl. 7.9). Made from beige snap, on a mediolithic blade from,, two primary burin spalls; one from an unretouched edge of an obsidian burin (Pl. 8.2), the other from a slightly retouched edge of a burin, reddish radiolarite (Pl. 8.3), a very slender primary burin spall detached from a blade with a lateral notch, from silex blond (Pl. 8.4), a secondary burin spall from silex blond (Pl. 8.5), 7 fragments of burin spalls from: silex blond, obsidian, and reddish radiolarite (4 specimens). two truncations: one on a macrolithic, the other on a mediolithic obsidian blade; shaped by distal break (Pl. 8.6, 7), a narrow truncation on a regular macroblade, White greyish trasparent, ish, transparent, a distal fragment of an oblique truncation with lateral denticulated retouch (Pl. 8.10). Reddish radiolarite, a distal fragment of a straight truncation on a macroblade (Pl. 8.11). The distal part is thinned, a distal fragment of a concave truncation on a regular blade (Pl. 8.12). Blackish radiolarite, a distal fragment of a convex truncation on a mediolithic blade (Pl. 8.13). On the proximal break there is also occur on the edge of the truncation. Yellow radiolarite, a fragment of a convex truncation on a mediolithic blade with steep lateral retouch. Reddish radiolarite (Pl. 8.14), (Pl. 8.15), from reddish radiolarite. a proximal fragment of an atypical truncation with inverse, slightly concave retouch; (Pl. 8.16); yellow radiolarite with a dark band under the cortex, two microlithic truncations, possibly fragments of trapezes (Pl. 8.17, 18), from reddish radiolarite. a transversal truncation on a macrolithic blade used as a sickle on its sinister edge. Along its entire length the edge shows inverse microscars (Pl. 9.1), a convex truncation on a narrow macroblade from silex blond ; the dexter edge shows silica gloss and random microscars (Pl. 9.2), the proximal part of a mediolithic blade with an oblique truncation and inversely retouched section of lateral edge; at the truncation oblique, an oblique truncation with retouch extending onto the lateral edge. Oblique silica gloss can be seen at the retouched truncation (prior to the retouch) and on the proximal break (subsequent to the break) (Pl. 9.4). Reddish radiolarite, an oblique truncation on a microlithic bladelet with parallel gloss of the dexter edge (Pl. 9.5). Reddish radiolarite,

15 Early Neolithic settlement of Mavropigi in western Greek Macedonia 85 a fragment of a bladelet with a distal notch, used as a sickle insert (Pl. 9.6). Cretaceous, an oblique distal truncation on a mediolithic blade, with the tip broken off. Oblique silica gloss at the truncation (Pl. 9.7)., lateral retouch on the dexter edge. The proximal truncation was shaped subsequently to the oblique silica gloss (Pl. 9.8). transparent a sickle insert shaped on an oblique, weakly convex truncation; reddish radiolarite (Pl. 9.9). an arched backed piece on a mediolithic blade (Pl. 9.10). Black radiolarite, irregular, steep, bidirectional retouch. The lateral from the use as a sickle insert (Pl. 9.11). Reddish radiolarite. a symmetrical trapeze with the tips polished after the truncations had been retouched (Pl ). three asymmetrical trapezes from the same raw material: one with a distal oblique truncation, the other with a straight proximal truncation (Pl. 9.13, 14). The third trapeze on a thin blade had ventral impact fractures in the proximal part (Pl. 9.15), two trapezes approximating rectangles: one on a thin blade with truncations shaped by obverse retouch, made from chalcedony (Pl. 9.16), the other on a thicker blade with bidirectional retouch (made from reddish radiolarite Pl. 10.2), (Pl. 10.1), two trapezes with convex truncations extending to the back (chalcedony Pl and of the specimen shows traces of breakage and secondary retouch on the proximal truncation;, an asymmetrical trapeze with one oblique obverse and the other convex inverse truncations. Made from silex blond (Pl. 10.5), a trapeze with an oblique distal truncation and end (Pl. 10.6). 2 fragments of trapezes: one from reddish radiolarite (Pl. 10.7), the other from white opal (Pl. 10.8). a macrolithic from silex blond, with bilateral obverse retouch in the distal, a macrolithic blade with partial bilateral, a macrolithic blade with partial bilateral retouch, from silex blond (Pl ). The three fragments of macrolithic blades with partial, lateral microretouch. Two fragments are from obsidian (Pl , 13), and one from silex blond (Pl ), two blades with unilateral continuous inverse retouch. One specimen is from reddish radiolarite (Pl. 11.1), a mediolithic blade with lateral convex retouch, from reddish radiolarite (Pl. 11.2), a fragment of a macrolithic blade with discontinuous alternate retouch, from silex blond (Pl. 11.3), two fragments of mediolithic blades with alternate microretouch; one from obsidian, the other from reddish radiolarite (Pl. 11.4, 8), three fragments of mediolithic blades with partial inverse microretouch; from, a kind of with lateral alternate notches, from obsidian (Pl. 11.9), a blade with lateral inverse retouch, from, 5 fragments of mediolithic blades with irregular retouch, from transparent

16 86 a macroblade with parallel gloss and inverse micro a mesial and a proximal fragment of mediolithic blades, with parallel sickle gloss on the dexter edge and nibbling (Pl , 12). radiolarite, a proximal fragment of a macroblade with partial sickle gloss and partial bilateral retouch of the sinister edge (Pl ). Beige, weakly, a mediolithic blade with sickle gloss and microretouch of the dexter edge (Pl ), reddish radiolarite, three mesial fragments of macroblades with parallel gloss and obverse microscars on the dexter edge (Pl. 12.1); rough chocolate coloured inverse microscars on the sinister edge (Pl. 12.2);,, with parallel sickle gloss on both edges and irregular microretouch on one or both edges ish represented by two specimens, the remaining raw materials, single specimens, retouch and microscars on the dexter edge. Reddish radiolarite (Pl ), a fragment of a mediolithic specimen with oblique gloss in the distal part. Reddish radiolarite (Pl ), a blade with bilateral alternate retouch; parallel gloss on the lateral edge and oblique gloss prior to the distal break (Pl ), smooth,, an irregular mediolithic blade with oblique gloss in the proximal part, and obverse microscars ish, a mediolithic blade with a break in the distal part which is prior to the oblique gloss with microscars; reddish radiolarite (Pl ), a blade broken in the distal part. The break is earlier than as a sickle insert (oblique silica gloss on both sides in the proximal part). On both edges there are microscars in the central section. Yellowish radiolarite (Pl ), four fragments of blades with irregular retouch. Two are with sickle gloss on both edges (reddish radiolarite and chalcedony Pl , 17), and two with gloss on one edge (transparent whit two macrolithic blades with the tips thinned by, the ventral side; reddish radiolarite (Pl. 13.3), a proximal part of an obsidian macroblade executed from the break on the dorsal side (Pl. 13.4), a distal fragment of a microlithic blade with inverse retouch of one lateral side and a thinned tip. Silica gloss on both edges (yellow radiolarite Pl. 13.5), three retouch, from of the specimens was burnt., quartzite (Pl. 13.7),, from black radiolarite (Pl. 13.9), with bilateral retouch (Pl ), (Pl ). from reddish radiolarite (Pl. 14.1),

17 Early Neolithic settlement of Mavropigi in western Greek Macedonia 87 a retouched quartz splinter, radiolarite. on the proximal break (Pl. 14.2), from transparent, proximal retouch, from reddish radiolarite (Pl. 14.4), two almost the entire circumference, from quartz (Pl. 14.5, 6), a denticulated macrotool on a quartz core (Pl. 14.7), a fragment of a denticulated tool shaped on, two fragments of notched tools: one from chalcedony, the other from quartz (Pl. 14.9, 10), a fragment of a tool with denticulated retouch, from reddish radiolarite (Pl ), four quartz. 14 specimens with parallel sickle gloss on one edge. A complete specimen was made on an exceptionally large macroblade ( specimens are fragments (proximal, medial and These specimens were made from a variety of chalcedony, silex blond are represented by single specimens. Two specimens were made from reddish radiolarite and two from white greyish two specimens have oblique gloss: one in the proximal part, 15.11); the other at the distal break, from reddish radiolarite (Pl ), 6 specimens have parallel gloss on both edges. Three of them are on macroblades: two proximal fragments, one from reddish radiolarite (Pl ) 15.14), and a mesial fragment from transparent ish (Pl. 16.1, 2) and one from silex blond, a mesial fragment of a mediolithic blade from reddish radiolarite, with oblique gloss in the distal part, three core preparation hafted obliquely as sickle inserts, made from reddish radiolarite and, distal retouch on the edge used as a sickle insert (Pl. 16.3), four splintered pieces with oblique or parallel sickle gloss used as sickle inserts; whit blond. One of the specimens is burnt (Pl. 16.4, 5), a chip detached from the edge of a sickle insert from Moreover, there were 16 fragments of sickle inserts: 13 specimens with parallel gloss, including 3 specimens with gloss on both edges; from reddish radiolarite 3, from other radiolarites 2, from whit each. One specimen is burnt, 3 fragments with oblique gloss (2 from ish There were also 3 small fragments of undetermined retouched tools. A depot (cache) of lithic artefacts from Stratum 1 The pit 37, without ceramics, was a depot consisting of 95 radiolarite artefacts from four cores: a tablet from platform rejuvenation 1 (Pl. 17.5). 4. reduction 7 (Pl ). 5. retrimming 1 (Pl.17.6).

18 (Pl ) (Pl ). 8. blade most, however, unidirectional chips a fragment of a broken pebble. There were t as well as two artefacts detached from the same core but which did not The features of bulbs and butts of the artefacts in the depot indicate the use of a hard hammer. The comparison of the debitage products stored features in Stratum 1 shows that the technique of detachment of debitage products was slightly different, evidenced by a smaller proportion than in S the depot and from other features in Stratum 1 Stratum 1 (Fig. 57). This results from the fact that of core reduction. The pit with the store of debitage products did not exhibit typical features of a waste pit, as it contained lithic artefacts exclusively, nor did it have the nature of a workshop. Thus, the deposition of the artefacts was undoubtedly intentional and suggests a special role of the depot. Thrifty raw materials economy could be one explanation but which seems more likely it is quite possible that their deposition was related to rituals and symbolic activities. Differences between the inventories from Phases IIIa and IIIb Stratum 1 pottery recovered from the central and from few chipped stone artefacts that were analyzed together with the total inventory of Stratum 1 (see above). ies of some raw materials such as quartz, radiolarite or obsidian are different (Fig. 58). The quantitative structure The smaller proportion of quartz in in comparison with the later and the earlier phases is more likely to have been functionally determined. This interpretation is approximates the structure of the whole Stratum 1 rather than Stratum 2.The slightly higher index of blades and tools in Phase different settlement zones (Fig. 59). forms with lateral, rather irregular, often produced by a technique similar to the splintered technique, which caused a resemblance to Truncations also occurred (three specimens Pl. 8.8, 12; 9.9). Moreover, an asymmetrical distal proximal part (Pl. 7.10) and a burin spall were recovered. THE EVOLUTION OF THE CHIPPED STONE INDUSTRY IN THE LOCAL SEQUENCE Changes in the raw materials structure (Stratum 3) is very small. The group who settled own store of raw materials, mainly obsidian in the form of blades which account for a third of the inventory. Such a structure points to a southern direction of contacts, in all likelihood with the territory of Thessaly. Less important are greyish transparent which source is unknown, and reddish radiolarite, most probably mesolocal.

19 Early Neolithic settlement of Mavropigi in western Greek Macedonia 89 Fig. 56. by single blow, 3 dihedral, 4 facetted, 5 punctiform, 6 linear Fig radiolarite, 4 silex blond, 5 quartz Fig. 57. the whole Stratum 1 Fig blades, 5 tools, 6 splintered pieces, 7 chips Trace amounts of extralocal silex blond in Stratum 3 which of Mavropigi (probably in Albania) separated from it by mountain ranges could be the evidence blond is abundant at Early Neolithic sites. local quartz and also though to less extent radiolarites increased. Quartz was worked in a special zone of the settlement, near the western of all extralocal rocks such as obsidian, played a minor role. On the other hand, their presence on site indicates that systematic contacts were (Stratum 3).

20 90 less important, whereas considerable increase of the role of radiolarite processing is recorded (Fig. in the higher proportion of radiolarite tools (from 20% in Stratum 2 to 26% in Stratum 1). Among extralocal raw materials the proportion of obsidian again grows, whereas that of other raw materials remains more or less unchanged. An exception is silex blond which presence in the tool group drops (from 13.3% in Stratum 2 to 8.4% in Stratum 1 Fig. 60). The raw materials composition stability of contacts throughout the entire time span of the Early Neolithic settlement at Mavropigi. Extralocal raw materials point to a southern direction of these contacts, notably with Thessaly. The structure of major technological groups The structure of major technological groups is it manifests the relation between producers and Fig. 60. radiolarite, 4 other radiolarites, 5 silex blond, 6 quartz the inventory, similarity to Early Neolithic sites with painted pottery in the eastern and the central Balkans (Perlès, 2001;, 2007) can be seen. Thus, completed blades and blade products are absent, the series of chipped stone artefacts is small. a third are blades, Early Neolithic workshops are not known, undoubtedly lithic production was carried out in specialized workshops, in the vicinity of raw material deposits. This required highly specialized knappers who most probably held a high rank in the social hierarchy. At the same time, some lithic artefacts, for example obsidian blades, had not only a functional role but also prestige and tools could have been related to their use as sickle inserts since the process of harvesting as a ritual could have played an important role in the sphere of beliefs of agricultural groups. took place in the organization of chipped stone production. Some implements notably those continued to be imported as completed blades materials (quartz) were produced at the site. The question remains whether these changes were demand for chipped stone artefacts, or whether they were the effect of the process of adaptation to local conditions, including the location of raw material deposits, both in the sphere of production organization as well as in social relations that transfer of production to settlements or even to spheres of the expansion of the Early Neolithic. in the lithic production organization appear, as quartz which

21 Early Neolithic settlement of Mavropigi in western Greek Macedonia 91 has poor cleavage (48% in Stratum 1 as compared to 67% in Stratum 2) is partially replaced by radiolarite, mainly the reddish type (18.6% in Stratum 1 as compared to 9.9% in Stratum 2). An interesting example of local radiolarite processing is a depot of debitage products in pit 37, from preliminary and advanced stages of reduction of 4 cores. The depot did not contain lying Endröd 39 in the Hungarian Plain (, 1981), where, too, the Early Neolithic raw materials dominated, the debitage from the stored in a clay pot. Technology Throughout the levels the blade technique is the most common. Generally, this technique is similar to the macroblade technique of the Early Neolithic in the eastern and central Balkans (Gatsov, 2000, 2009). The reconstruction of series from (mainly in Stratum 2) and radiolarite (mainly in Stratum 1) were found. As a rule, these are either that, even in the case of some quartz cores, were used for the detachment of fairly regular blades. At all stages of the site occupation blanks were detached by means of the punch or pressure local raw materials (quartz and radiolarite) do not basically differ from blades from extralocal raw materials (obsidian ) as regards the technical attributes. No cores show preliminary preparation or retrimming. The only exception is a larger number of prepared butts in the debitage products in Stratum 2 compared with Stratum 1. This difference could be the result of a greater importance of local production in Stratum 1 and inferior skill of knappers in this phase of the settlement. A similar tendency is indicated by the changes in the morphology of blades between Strata 2 and 1, namely: smaller blade standardization and the manufacture of smaller and narrower blades in Stratum 1 than in Stratum 2 (Fig. 61). Tool evolution of burins (including one burin spall that could not retouched blades and unretouched sickle inserts. Although the assemblage is small, the major tool categories are typical of the Early Neolithic of Greece. trapezes, perforators, truncations and blades with also typical of the Early Neolithic of the Eastern Balkans where just as at Mavropigi blades with marginal retouch predominate. Blades with marginal retouch, often with sickle gloss, are predominant in 1), where they account for nearly a fourth of all the tools.the blade tools, known already in lower strata to sickle inserts, both retouched (as retouched blades, truncations and burins) as well as unretouched ones. The proportion of harvesting tools is very high: almost a half of all the tools of sickle inserts. Fig. 61.

22 92 The location of sickle gloss on the inserts points to the use of two types of hafts: with a continuous straight cutting edge (parallel gloss), obliquely hafted inserts (oblique gloss). The relation between retouch and sickle gloss indicates that sickle inserts were either retouched prior to use, or from other tools in a secondary stage of use. HORIZONTAL DISTRIBUTION OF FINDS AND THE FUNCTIONAL DIFFERENTIATION OF THE EXCAVATED AREA Lithic artefacts found on the site surface have not been included in this section as they could have come from various phases of the occupation of the site and, moreover, their horizontal distribution is random. basically no pits have been preserved, a large number of lithic artefacts was found on the were collected from the surface was square 801 (i.e. above the central ). But the main part only few pits were registered. is probably the result of intensive erosional had destroyed the remains of hypothetical features. the central on the other hand, is associated with m and weaker erosion in this area. The small number of lithics that could be makes it impossible to determine horizontal distribution in this phase. concentrate mainly in squares 172 and 132, i.e. number of artefacts from quartz was recovered, especially from square 132. This means that in this zone of the settlement quartz was worked in situ. activities that required a variety of implements, as well as tool repair were concentrated in square 172. This was a zone north of the central, in between the and the group of east part of the site. Almost a half of all the lithic artefacts from P zone. The artefacts were found both in the areas pits. Most obsidian artefacts, various harvesting were, too, recovered from square 172. Lithic production, domestic tasks, and tool repair were carried out between the dwellings as well as in their interior (about 80 artefacts were found inside dwellings 7, 6, 2). Large quantities of chips and splinters in the dwellings indicate that splintered technique was employed to repair and modify tools. However, because tools in the dwellings are not numerous we can conjecture that they were taken away as valuable components of the equipment of a household when the inhabitants abandoned their houses. The lithic production itself took place, as a rule, in the areas exterior to dwellings and pits. Lithic artefacts were also recovered from graves of grave 7 there were two such artefacts from silex blond : a retouched blade (Fig. 62 upper, Pl. 10.9) found below the skeleton, and a splintered piece on blade (Pl. 5.5). There was also a regular seven splinters that could have got into the grave accidentally as part of the. THE PLACE OF THE INDUSTRY FROM MAVROPIGI IN THE EARLY NEOLITHIC OF THE EASTERN BALKANS of the Early Neolithic settlement agglomeration in Thessaly (Gallis, 1992, 1994) and to the west of Early Neolithic sites in Macedonia

23 Early Neolithic settlement of Mavropigi in western Greek Macedonia 93 Fig. 62. Artefacts from grave 7 (upper retouched blade from silex blond, lower radiolarite blade) such as Nea Nikomedea (Rodden, 1965) and Giannitsa B (Perlès, 2001). Up till the 1990 s the settlement network in western Macedonia seemed poor, but subsequent rescue excavations enabled to uncover more than 40 new sites of the, 2005;, 2013). Mavropigi is one of the oldest Early Neolithic sites in northern Greece, dated to the time interval between 7,758 ±60 (DEM 1685) and 7,135 ±10 (DEM 1751), 2013). These dates are comparable with the dates of the sites of Nea Nikomedea (7,780 ±270 7,557 ±90 BP), Achilleion (7,611 ±83 BP). Mavropigi is an exceptional Early Neolithic site as the number of chipped stone artefacts is much larger than at the Thessalian sites. The site was excavated using exploration techniques Thus, the inventory obtained from Mavropigi is exceptionally complete of which we cannot be so sure in the case of sites in Thessaly investigated in the 50s. The investigations conducted so far at the Early Neolithic sites with painted pottery in the eastern Balkans have enabled us to formulate a lithic production: a) blade in its nature. Blades were produced using the punch or pressure technique, b) a preference can be seen for some raw ( silex blond acc. to C. Perlès) and Melian obsidian, c) extralocal raw materials were processed in workshops, mainly in the vicinity of deposits, and (Renfrew, 1984). Consequently, only small series of artefacts, mainly blades and tools, occur at the site d) tool production was based on blade blanks retouch predominating in assemblages also burins, truncations, less frequently trapezes The majority of blades were used without secondary preparation, usually as sickle inserts.

24 94 This model was registered at numerous sites of the Early Neolithic with painted ceramics in the territory of Thessaly: Achilleion (Elster,, 1996),, 1962; Zumbusch and, 1971; Tellenbach, 1983). A number of sites of the Neolithic with similar assemblages although without burins (Gatsov, 1993). Among Early Neolithic sites in Thessaly Argissa is particularly noteworthy, believed to be associated with monochrome ceramics (Perlès, 2001). The lithic assemblage from Argissa is relatively numerous (359 artefacts), with most artefacts made from obsidian (47.4%). This assemblage indicates that obsidian reached Thessaly as partially worked lumps. From these nodules blades were detached with measurable attributes similar to those from Mavropigi (the the splinter technique is also present at Argissa, used on broken blades. Tools are represented by blades with lateral retouch, burins, retouched, 1983). Just also occur (Perlès, hand, tools that are absent in Mavropigi, occur (Tellenbach, pieces (Tellenbach, The site of Achilleion in Thessaly, too, belongs to the classical Early Neolithic model of chipped stone industry (Elster, consecutive levels at this site the number of chipped stone artefacts increase gradually from important raw materials, which E. Elster thought to be local, followed by imported obsidian. The proportion of obsidian oscillates from 31.4% Achilleion show fairly regular dorsal pattern and occur as fragments. Tools are almost exclusively laterally retouched blades, usually with denticulated Elster does not mention any other tools. Both unretouched as well as retouched blades used as sickle inserts show parallel sickle gloss. As the Early Neolithic spread to the north this either by changes in social organization (the shift of lithic production to the level of household that provided local raw materials. The changes consisted in: a) the use of local lithic raw materials, b) the shifting of at least part of production to c) as a result, at some sites the quantity of increased, d) in addition to blade tools, that continue to, notched have not been fully published, but chronologically the site spans a horizon contemporaneous with Mavropigi, and its lithic industry is similar (Elster, 1977). According to E. Elster assemblages from this horizon are dominated by macroblade technique and the most frequent tools are laterally (or bilaterally) retouched blades, frequently used as inserts for harvesting tools. At Nea Nikomedea just as at Mavropigi local quartz was worked blades found in the structure interpreted as a temple (Rodden, 1964). Blades in the depot are from local raw material, while technology and size indicate a higher degree of standardization of production (Perlès, 2001:202). Unlike at Mavropigi, at Nea Nikomedea obsidian was not recorded although the site is situated close to the coastline. This suggests that obsidian was not supplied to Greek Macedonia directly from Melos via a sea route, but via Thessaly. Mavropigi is of key importance for the early stage of this transformation in the organization of production, combining the features of the classical Balkan model (especially as regards extralocal raw materials) with new tendencies such as exploitation of local raw materials and the shifting at least part The presence at Mavropigi of decorated pottery compels us to look into the relation of the industry at Mavropigi to the industries that accompany such ceramics at

25 Early Neolithic settlement of Mavropigi in western Greek Macedonia 95 western Balkan sites. Unquestionably, the dates from Mavropigi in the interval of 6, 5,900 cal BC are earlier than most dates from the sites with ceramics from the territory of Montenegro or Dalmatia that are not, 2009). An exception is provided at C with incised pottery was dated by Sordinas older than at Mavropigi. Above level C, within th th millenium cal BC transition a level with, 2003). New dates of the Neolithic levels at Sidari between 7,500 ±60 BP and 7,370 ±80 BP (Berger, 2014) are only slightly younger. The chipped stone industry from Sidari in the in the level with sherds individual blade tools appear (among others The industry at Sidari differs not only from the Mavropigi assemblages but also from the industries associated with,, 1999). The industries from these two sites represent a local Mesolithic tradition using microblade technique for the production of trapezes and microlithic segments, truncations Castelnovian Mesolithic is continued both in level, 1985;, 1994). The sites with ceramics from the territory of Dalmatia dated at to th millennium cal BC, in turn, are characterized by highly advanced macroblade technology based blond. This technology is known at the sites dated to th millennium such,, 2008) and Tinj (Müller, 1994). The macroblade technology at these sites contained macroblade elements, which, however,, 1966). The variability of lithic traditions in the polycentric origins of this complex with Greek Macedonia as one of its centres, probably independent of other centres, in the western Balkans. The hypothesis that claims the existence of many centers of the origin of to, 2011), but this of this complex. Acknowledgments Thanks are due to the NCN (Polish National REFERENCES (to parts 1 and 2) The Neolithic and Bronze Age of Northern Greece.. Museum Archaeologicum, Zadar. néolithique sur le site de Sidari (Corfou, Grèce). Nouvelles données géoarchéologiques et BONGA L. (in press). Early Neolithic pottery in Greece: interactions and reconsiderations, Conference, Rethymnon, Crete.. Šibenik. Muzej Grada Šibenika.

26 On the trail of Neolithic mice and men towards Transcaucasia: zooarchaeological clues from Nakhchivan (Azerbaijan) Philadelphia. ELSTER E.S University of California, Los Angeles, PhD Thesis. ELSTER E.S The chipped stone industry.. Monumenta. Ephoria of Antiquities, Larisa. and topographical; work in Neolithic Thessaly.. Ministère de la Culture. Edition. Jagiellonian University, L. Nikolova (ed.). BAR. PWN, Warszawa. J Culture). origin and spread of the Western Linear Pottery culture: between foragers and food producing lifeways in Central Europe., 19, prehistoric sites in western Macedonia: prefectures New evidence on the beginning of farming in Greece: the Early Neolithic settlement of Mavropigi in western Macedonia (Greece Aegean Basin and the Balkans in the Neolithization Societa industries from Neolithic levels at Lerna.. Wydawnictwo Uniwersytetu Warszawskiego, Warszawa.

27 Early Neolithic settlement of Mavropigi in western Greek Macedonia 97 cultural changes in prehistoric Macedonia: recent,. Prosinac, Zadar.. Beograd Universita, Beograd. Varia Archaeologica Hungarica, G.N. Bailey, E. Adam. E. Panagopolou, C. Perlès, British School of J. Beoessneck, M. Hopf (eds) Beitrage Habelt, Bonn des Adriaraumes.. Berlin. Papathanassopoulous (ed.) M Neolithic sites in the Marmara. PAPAGEORGOPOULOU Ch Ancient DNA: Anthropological remains from the early sites at Mavropigi, Xirolimni and Pontokomi of the 1, Aiani Aurenche, J.Cauvin (eds) Oxford,. Cambridge University Press, Cambridge., Archaeology (Argolid, Greece), British School of Athens, Suppl. 25, Athens. on 14C data: chronological frameworks in the Neolithic and Chalcolithic of southeastern Europe, C. Renfrew (ed.) RODDEN G.J Recent Discoveries from Prehistoric Macedonia. 5, RODDEN R.J An early Neolithic village in Greece. from Palaeolithic to the Early Bronze Age. British School at Athens 10,

28 98 TELLENBACH M Materialien zum Präkeramischen Neolithikum in Süd Ost Europa. University Studio Press, Thessaloniki (in Greek). Archaeobotanical remains from Mavropigi near and Creem on Red Designs on Early Neolithic Ceramics from Nea Nikomedeia. in Macedonia.

29 Early Neolithic settlement of Mavropigi in western Greek Macedonia 99 Pl. 1.

30 100 Pl. 2.

31 Early Neolithic settlement of Mavropigi in western Greek Macedonia 101 Pl. 3.

32 102 Pl. 4.

33 Early Neolithic settlement of Mavropigi in western Greek Macedonia 103 Pl. 5.

34 104 Pl. 6.

35 Early Neolithic settlement of Mavropigi in western Greek Macedonia 105 Pl. 7.

36 106 Pl. 8.

37 Early Neolithic settlement of Mavropigi in western Greek Macedonia 107 Pl. 9.

38 108 Pl. 10.

39 Early Neolithic settlement of Mavropigi in western Greek Macedonia 109 Pl. 11.

40 110 Pl. 12.

41 Early Neolithic settlement of Mavropigi in western Greek Macedonia 111 Pl. 13.

42 112 Pl. 14.

43 Early Neolithic settlement of Mavropigi in western Greek Macedonia 113 Pl. 15.

44 114 Pl. 16.

45 Early Neolithic settlement of Mavropigi in western Greek Macedonia 115 Pl. 17.