Introduction

The Middle Paleolithic in Greece is known from finds from a number of cave and open sites (Darlas, 1994; Darlas and Lumley, 1999; Kyparissi-Apostolika, 1999; Panagopoulou et al., 2002-2004; Papaconstantinou, 1988; Papagianni, 2000), but until recently no representative multilayered layer was recorded. a monument of this era.

Excavations of the Middle and Upper Paleolithic stratified site of Klisura, cave No. 1 (hereinafter referred to as Klisura - 1), in the east of the Peloponnese Peninsula (Figs. 1, 2) provided important materials on the stratigraphy of the monument's deposits, technical and typological features of stone industries, and various aspects of human activity. So far, only the results of studying the uppermost Mousterian layers VII, VIIa, VIII, and X exposed by a small pit in 1997 have been published (Koumouzelis et al., 2001b). In these deposits, microlithoid industry is represented; the tool kit is dominated by scrapers, convergent products are found, and forms such as Keene and products of Levallois splitting are very scarce. Excavations in 2001-2006 revealed many new Middle Paleolithic layers with open hearths, abundant faunal remains, and rich stone industries, including numerous retouched tools (various scrapers and pinnacles), as well as nuclei and cleavages. Oddly enough, the upper part of the Middle Paleolithic column was represented by Levallois forms, elongated points, and convergent scrapers.

This publication describes the Mousterian industries and presents the results of their comparative technological and typological analyses. The study is based on a sample of 37,922 artifacts from 14 layers. In addition, materials from layer XX roofing were also used in some comparisons.

The Middle Paleolithic member with a thickness of up to 6.5 m is composed of numerous layers with evidence of intensive habitation (layers VI-XX a-g, counting from top to bottom); dates for it have not yet been determined. A radiocarbon date of 40,010 ± 740 BP (GifA-99168) was obtained for layer V, which overlies the Middle Paleolithic sequence and represents an Early Upper Paleolithic industry with segmented plates with a blunted edge (uluzzo) [Ibid]. In the 1997 pit, layer VI was found to be mixed with the overlapping Uluzzo layer [Ibid], but it included a pure Mousterian complex in the rest of the excavation area.

page 2

Klisura deposits-1. Middle Paleolithic sequence

The Middle Paleolithic deposits of Klisura-1 contrast markedly with the Upper Paleolithic ones. The latter, containing more components of anthropogenic origin, are more friable and stony. Their color varies from gray to whitish, which is explained by the saturation of calcite ash and other combustion products. In addition, these layers are characterized by well-defined clay hearth structures overlain by ash interbeds (Karkanas et al., 2004; Koumouzelis et al., 2001a, b). In contrast, the upper part of the Middle Paleolithic strata (layers VII-XV) is gradually but distinctly transformed into more compact brownish or reddish fine-grained deposits, which include a series of stony interlayers in the back of the cave. These features are a consequence of the participation of more clay and other clastic materials in the sedimentation process, which indicates the leading role of natural processes in the formation of the upper part of the Middle Paleolithic layers. At the same time, they also contain an anthropogenic component in the form of separate layers of multicolored combustion products. This upper part of the Middle Paleolithic column shows several episodes of natural and anthropogenic sedimentation, separated by clear horizons of erosion. Layers XI-XIV are represented by reddish horizons with a high clay content, alternating with thin multicolored combustion horizons. Forming the first ones,

1. View of the Klisura Gorge from the Argotic Gulf (a) and from the Larissa fortress in Argos (b).

2. Klisura Cave No. 1.

page 3

3. Western excavation profile of Klisura-1. A part of the Middle Paleolithic sediment sequence (layers XVI-XXc) is visible with gray interlayers at the top (layer XVI saturated with ash), a stony layer XVIII in the middle, and burnt materials (layers XHa and XXc).

most likely, it was the result of rainwater flows moving clastic materials from the hills above the cave. Layers XI-XIV also contain well-developed carbonate crusts, which are preserved due to the cementation of the underlying surfaces and mark small breaks in sedimentation. The multicolored burning horizon in layer XI marks the end of formation of a pack of layers XVI-XI. Part of this pack was eroded, resulting in a depression in the entrance part of the cave. Layers X and IX are strongly lightened in some places and have a whitish color, although they also include darker (from dark gray to black) interlayers formed, apparently, due to the cementation of burnt remains by surface water. Another erosion episode led to the formation of a trough that cut off all previously deposited layers of the Middle Paleolithic sequence. A chute runs from the northern profile at the back of the cave to the western profile. It is filled with materials of layers XIa and XIb, which are a mixture of burnt residues and sand deposits with weakly expressed stratification. It is not yet clear whether this filling is the result of the development of natural processes or human activity. In the back part of the cave, it and the layers following layer XI lie directly under layers VIII-VI, which is a consequence of the combined impact of anthropogenic and natural processes. They are clayey, have a brownish tint, but also contain inclusions of burnt residues, which gives them a darker color in some places. Thus, the upper part of the Middle Paleolithic sequence was affected by erosion and a number of other erosion processes.

The upper and lower parts of the Middle Paleolithic sequence are sharply different from each other: clay, reddish, and stony inclusions of the upper layers clearly contrast with the underlying fine-grained, silty, and ashy layers. These differences may be related to changes in the karst configuration of the cave. Around the time of the formation of layer XV, a landslide probably occurred in the back of the cave, creating a crevasse that has survived to this day. Through it, rainwater, red soil, and rock fragments began to enter the cave, which affected the nature of deposits in the upper part of the Middle Paleolithic sequence. The lower part of the stratum (layers XVI-XX) is represented by homogeneous fine-grained deposits with signs of post-depositional chemical modification, especially noticeable in the lower layers of layer XX (Fig. 3). This modification was manifested in the formation of phosphate aggregates around limestone blocks. As we move from the upper layers to the lower ones, there is also a gradual increase in their slope towards the entrance.

Layers XVI and XVII consist of alternating gray and whitish horizons of considerable extent. Layer XVIII is brownish-gray, silty, and contains many bones and stone artifacts. The high content of the latter gives the layer a rocky appearance. Layer XIX is dark gray, silty, and homogeneous. Layers XXa, XXc and XXf are saturated with burnt material in the form of whitish and black lenses. Layers XXb and XXd are light brown and silty, with fewer signs of anthropogenic impact compared to layers XVIII and XIX. A local variation of the XHa layer is the XXa1 layer, which is reddish in color and is distinguished by its high content of red soil. Layers XXe and XXf are moderately cemented due to a large number of phosphate veins and couplings. In addition, they contain a significant sandy component. The XXg layer is loose, gray, and sandy. Layer XXI is found only near the eastern wall of the cave, where it is represented by dark red deposits lying directly on the rock bottom. It consists of red soil with a large number of limestone blocks and crushed stone. There are traces of intensive phosphate transformation. Contact between layer XXI and

page 4

Stone inventory of the Middle Paleolithic layers of Klisura-1, copies.

Stone tools

VI

VII

VIla

VIII

X

XIb

XII

XIII

XIV

XV

XVI

XVII

XVIII

XIX

Total

Scales

1 711

1189

1 334

1 270

820

101

29

82

649

698

1 542

1 543

4 141

493

15 602

Pieces and debris

607

481

244

425

297

15

2

13

40

34

64

75

259

19

2 575

Preforms

0

0

0

0

2

1

1

0

1

0

3

0

2

0

10

Nuclei

36

23

29

43

49

12

17

10

35

19

40

56

148

70

587

Flakes

571

705

742

945

1 044

287

263

241

638

547

862

1 288

4 791

1 489

14413

Plates

41

22

22

20

33

9

17

14

18

14

20

41

439

265

975

Tools

103

95

108

240

340

106

89

61

212

150

254

202

1 028

677

3 665

Bumpers

1

0

0

5

2

0

0

0

2

3

2

1

12

0

28

Tool renewal chips

1

0

0

0

7

2

6

2

4

6

6

3

8

4

49

Choppers

0

0

0

0

2

0

0

0

1

0

0

0

1

0

4

Pieces with test chips

2

0

0

0

0

0

0

0

0

0

0

5

5

2

14

Total

3 073

2515

2 479

2 948

2 596

533

424

423

1 600

1 471

2 793

3214

10 834

3 019

37 922

the above-lying sediments are dissonant, which may be a reflection of sharp natural changes in the corresponding period.

It is still too early to draw definitive conclusions, but it can already be assumed that the main part of the Middle Paleolithic sediment pack was formed in wetter conditions than in the Upper Paleolithic period. The sandy lenses in its lower part were apparently left by karst waters in underground tunnels that formed in the sediments already formed at that time. The latter probably also resulted from an increase in the ground water level, which contributed to a change in the chemical composition of sediments. The phosphate in them appeared, most likely, as a result of the putrefaction of guano, a large amount of which had accumulated by that time in the cave.

Composition of artifacts

The general structure of Middle Paleolithic complexes is characterized by the predominance of small cleavage and retouching products (flakes) up to 1.5 cm in size (according to preliminary estimates, they usually make up 30-55% of artifacts), small fragments (up to 20 %; they are especially numerous in the uppermost Middle Paleolithic layers X-VII), and small flakes (see table).. However, in the middle part of the column (layers XIb, XII, XIII), flakes are less common (less than 20 %). The frequency of occurrence of tools varies from medium to quite high (from 3.7 to 20 %) and contrasts with the small number of nuclei. If we exclude flakes, chunks and fragments from the calculations, then the predominant category of stone tools will be flakes (70-83%). Tools will occupy the second position (11-25 %; the lowest rates fall on the upper layers VII and VIIa) (Figure 4). The lowest percentage of nuclei (2.3 - 4.4%) and plates (1.5 - 4.4%). At the same time, if the proportion of nuclei remains low in all

Figure 4. Distribution of the main categories of artifacts in the Klisura-1 layers.

page 5

Fig. 5. Limas (1, 5), scrapers: trapezoid (2), angular (b, 7), segmental (8), double (9), lateral with distal and obushkovy thinning (10), lateral on the plate (11); spiky (3); spiky on the plate (4); Levallois flake (12). 1-chalcedony; 2, 3, 5, 7 - 12 - radiolarite; 4-flint; 6-volcanic rock. 1,3,8-10,12 - layer XIV; 2, 6, 11 - XI; 4-XIX; 5-X; 7-XII.

6. Nuclei in the Klisura-1 layers.

The frequency of occurrence of plates increases slightly in the lowest part of the Middle Paleolithic thickness, reaching 6,8-10,5 % in layers XVIII and XIX. Pieces with test chips, rocks (raw materials), as well as chocks, are either absent or presented in very small quantities in several layers, mainly in the lower part deposits. The ratio of the number of tools to the number of nuclei remains high in all Middle Paleolithic complexes: from 3.6 : 1 to 9.6 : 1. Calculations of the quantitative ratio of blanks and nuclei also gave high values: usually from 20: 1 to 35: 1, with the largest ones for the upper and lower layers. Thus, the composition of the stone inventory of the Middle Paleolithic complexes indicates that the primary splitting, manufacture and use of tools took place directly on the monument. As for the few differences in the composition of artifacts between layers, the most significant among them is the increase in the number of plates and plates in the most ancient Middle Paleolithic complexes.

Raw material

Radiolarite was the dominant raw material for artefacts in all Paleolithic layers of Klisura-1 (Fig. 5, 2, 3, 5, 7 - 12); such products in the Middle Paleolithic complexes account for 60-80 %. In second place - flint (15-32 %). Other rocks (including limestone, quartz, and chalcedony) played an insignificant role; products made from them do not exceed 4 %, usually accounting for 1-2 % (Figs. Such characteristics of raw materials as the small size of the initial pieces, tiles and fragments, as well as their average quality, which in many cases caused unintentional kink (longitudinal or loop-like) of chips, remain unchanged. The peculiarities of raw materials (especially some types of radiolarite) also explain the high percentage of flakes with a flat impact bump, which can lead to a false conclusion about the use of a soft bump.

Nuclei

Three types of nuclei predominate: those with traces of unidirectional (longitudinal), centripetal chips, and disc-shaped ones (Fig. 6). Unidirectional cleavage nuclei for flakes and plates/plas-

page 6

7. Levallois nuclei (1, 4) and flakes (9, 10, 12, 15) , nuclei with unidirectional negatives (2, 6) , nuclei with traces of a centripetal removal system (3, 5), a nucleus with signs of bipedal cleavage (7), an orthogonal nucleus (8), a toothed tool (77), edge (debordant) chip (73), plate (14). 1-layer XII; 2, 4, 6, 7,10, 11 - XV; 3-X; 5, 13, 14-XI; 8, 9-XIII; 12, 15-XIV.

8. Plates (7, 3), rib plate (2), plate core (4), Levallois flakes (5-7, 15), Levallois plate (8), Levallois nuclei (9, 11-13), Levallois points (10, 14). 1 - layer XVIII; 2,3 - XXc; 4, 12, 15 - XVII; 5, 7 - 9, 11 - VIII; 6-X; 10-VII; 13-XXd; 14-XIV

tinok (fig. 7, 2, 6, 7; 8, 4) they are most numerous (up to 50 %) in the two upper (VII and VIIa) and in the lowest (XV, XVIII and XIX) layers. Often they carry a broad, far-reaching negative of the final cleavage. Individual nuclei have traces of rib preparation and a partially rotatable working surface. The impact pads are usually smooth. Bipedal nuclei are rare, but in the upper and lower layers (VII and XIX) they are more common than in others.

Plate/plate nuclei mostly have a narrow front formed at the end of the chip or tile. Orthogonal nuclei (see Figures 7, 8) are found in relatively small numbers; among them there are samples of polyhedral shape. 7, 3, and 5) and disk-shaped nuclei (there are either equally or more of them) predominate in the layers of the middle part of the column (VIII or XIV). Levallois nuclei and cleavages are represented in a limited number, and there is no tendency to increase their role in the Middle Paleolithic layers. The frequency of occurrence of Levallois nuclei in the upper and lower layers is the same, and only in the middle part of the column they are somewhat more numerous. The Levallois nuclei for flakes usually exhibit a linear structure (see Fig.. 7, 7; 8, 11 - 13) or a recurrent withdrawal system (see Figures 7, 4; 8, 9). Nuclei for points and plates are rare. In the lower layers, many nuclei are represented by fragments and indeterminate, strongly worked samples. Raw rock samples are usually small in size. Larger pieces with test chips were found only in layers XVII-XIX. Prenuclei are rare; they are found in small numbers in layers X, XV, XVII, and XIX.

page 7

Flakes and plates

In most Middle Paleolithic layers, there are few plates. In layers VII-XVII, their index does not exceed 5 (excluding layer XII). A significant increase in the plate/plate index (from 9.7 to 16) was recorded in the lower three layers of the Middle Paleolithic column (Figure 9). Rib plates also appear here for the first time (see Fig. 8, 2) and flakes. An increase in the plate index in these layers corresponds to an increase in the number of nuclei with signs of unidirectional cleavage (mostly they are strongly worked). At the same time, the predominance of the same nuclei in the upper layers of VII and VIIa (where there are even nuclei for plates: 1 specimen - in layer VII and 5 specimens - in layer VIIa) did not lead to a similar change in the nature of blanks.

Flakes are mostly "non-Vallois" with unidirectional or centripetal negatives on the dorsal surface. The Levallois index is low (from 1 to 7) in all Middle Paleolithic complexes (see Fig. 9). Levallois billets are found in all layers (see Fig.. 7, 9, 10, 15; 8, 5 - 8, 10, 14, 15), and, although their number varies, there is no trend this variability is not observed.

The plates bear mostly negatives of unidirectional cuts (see Figs. 8, 1), and centripetal cutting is rare, including on Levallois chips (see Figs. 8, 8). The cut pattern of the dorsal surface of chips usually corresponds to the pattern observed on the nuclei (an exception is shown by finds from layer VIII). Parallel cleavage nuclei and their corresponding flakes predominate over nuclei and flakes with centripetal negatives in the upper layers, while the situation is reversed in the middle part of the column (layers X-IX). Then there is a relative equilibrium between different types of faceting (although nuclei with traces of a centripetal removal system, including disc-shaped ones, remain more numerous), and then parallel cleavage again becomes predominant. In layer XVIII, the dominance of the unidirectional lift system (48.7 %) over the centripetal one (10.8 %) is even more pronounced than in the upper layers. The rarity of primary and semi-primary plates and flakes in all complexes indicates that the initial cleavage was carried out outside the monument. As for the preparation of impact pads, all Mousterian layers are dominated by chips with smooth pads (40-60%); there is no obvious tendency to reduce their number. The proportion of roughly prepared sites (dihedral and polyhedral) remains stable; only in the middle part of the column is there a slight increase in the number of polyhedral sites. The frequency of occurrence of fine faceting features increases sharply in layer VIII, and especially in layer X, but a significant increase in IL is observed only in the second of these two complexes. As you move from top to bottom, a gradual decrease in the percentage of faceted sites is recorded. The complex of layer X (and probably also the industry of the underlying layer XI) it seems to be the most "Levallois" in the entire Clisura-1 column in terms of the nature of site preparation, Levallois index, types of nuclei and blanks. In general, in-

9. Plate indices (IIam) and Levallois (IL) for the Klisura-1 layers.

page 8

Fig. 10. Site faceting indexes for Klisura layers-1.

The faceting dex is low or medium (Figure 10); its maximum values correspond to layer X (IF1 = 43.1; IFs = 28.4). The overall faceting index fluctuates much less than the fine faceting index, and shows greater stability when compared across layers. By the nature of the sites, plates and flakes are usually similar. Significant differences between them on this basis were recorded only in the earliest Middle Paleolithic complexes. In layer XVIII, for example, faceted areas are much more typical for flakes (IFss = 8.4; IFs = 16.5; IF1 = 32.1) than for plates (IFss = 3; IFs = 7.5; IF1 = 20.4), whereas smooth (58.3 vs. 51.8%), linear (10.6 vs. 3 %) and point sites (17.5% vs. 1.9%) are more common on plates than on flakes. Plates with parallel - cut features usually have smooth areas (67.9 %, and 83.4% with linear and dotted ones), while faceted ones are very rare (4.1 %). Levallois blanks, on the contrary, are characterized by a high faceting index (IFss = 45.6; IF1 = 86.8), although they sometimes have smooth areas (10.8%). The technique of obtaining blanks is quite common: well-developed impact bumps (> 70 %) and obtuse angles (70-90 %) indicate the use of a solid bump. The production of plates identified in the lower layers was also based on the use of a solid bump: the corolla (lip) on the inner side of the site is very rare here (2.9 %); the percentage of flat impact bumps reaches a minimum value compared to other layers (5 vs. 11 - 24.4 %).

The "microlithoid" appearance of the industries and the presence of small plates are probably explained both by the nature of most varieties of local raw materials and by the peculiarities of the splitting strategies used by the cave inhabitants. Artifacts larger than 70 mm are rare in all layers (often such items are made of quartz, limestone, and sandstone). However, for example, in layer XVIII items with a length of >50 - 70 mm are more numerous (67 pieces) than in layer XVII (4 pieces). The nuclei are usually small in size (maximum length 68 mm, average-less than 35 mm, mostly 25 - 32 mm); they are either strongly worked or decorated on small initial rock parts. Some non-retouched blanks reach a length of 60 - 70 mm (which exceeds the size of the nuclei), but on average "microlitoid" -20 - 27 mm long, inferior in size to retouched tools (the average length of the latter is 30 - 35 mm). The few rocks with negative test chips found during the excavations are also small in size. It seems that larger non-retouched plates and flakes, like tools, tend to the lower layers (XVIII, XIX).

Among the actions performed on the monument, an important place was occupied by the primary splitting of the stone. However, the testing and decortication of the nodules was clearly carried out outside the parking lot; this is evidenced by the low number of primary chips and prenuclei. The proportion of chips without crusts is especially high in the layers of the middle part of the thickness (70-80% vs. 60-64% in other layers). Thus, the reduction sequence reflected in the materials of the monument is not complete. Its initial results

page 9

the stages are not presented, because the corresponding operations were performed outside the cave. All subsequent stages, on the contrary, are identified quite clearly, especially the final stages of activation and depletion of nuclei, as well as the re-registration of some of them into weapons. The same applies to different stages of tool life, including receiving blanks, retouching, re-registration, scrapping, and discarding.

Splitting methods

Upper layers VII, VIIa, VIII, and X reflect the coexistence of several cleavage systems (Koumouzelis et al., 2001b). Their use is recorded in the rest of the Middle Paleolithic complexes of the monument for obtaining flakes by a recurrent method: 1) unidirectional recurrent chipping without preliminary preparation of the working surface ("flat") and impact platform (naturally smooth or created by a single chipping), supplemented by bipedal and orthogonal chipping; 2) centripetal chipping from disc-shaped or conical/biconic nuclei [Sitlivy, 1996], one - or two-sided; 3) centripetal non-Levallois planar cleavage; 4) centripetal recurrent and linear Levallois cleavage. As a result of applying different methods, various chips were obtained. The unidirectional cleavage method produced broad and sometimes somewhat elongated massive flakes; centripetal non-Valloise cleavage resulted in short, rather thick, often marginal flakes (see Fig. 7, 13) with a wide prepared impact pad; preferential flakes with traces of centripetal cutting were removed from Levallois nuclei with signs of linear and centripetal recurrent cleavage (see Fig. 5, 12; 7, 15; 8, 5 - 7), as well as secondary flakes (see Figs. 7, 9, 10; 8, 15). Impact pads are faceted, sometimes prepared with a single chip.

These are the main methods of cleaving used throughout the entire period of formation of the Middle Paleolithic layers of Klisura-1. It should be noted that the Levallois strategy was used less frequently than the others. In addition, the use of several additional technologies was recorded: 5) Levallois convergent chipping to produce wide points (see Figures 8, 10) was used sporadically, for example, in layers VII, XIV and XIX; 6) Levallois recurrent unidirectional chipping to produce elongated blanks (see Figures 8, 8), for example, in layer VIII, and even more often in the lower layers 7) non-Valloise plate methods: a) operation of flat or partially rotated cleavage surfaces without their preliminary preparation (see Figs. 8, 4); non - facetted sites (mainly strongly worked, one-or less often two-site end or polyhedral nuclei); b) operation of pre-prepared ones by creating a side or central layer (Fig. 8, 2); the platforms are smooth (rare bulk nuclei with rib residues, in addition, there are a series of rib plates, gable and side). These two methods of triggering nuclei produced long, narrow, thin or thick plates with a straight profile and triangular or trapezoidal cross-sections, as well as small plates, including plates. Plates are found in small numbers in all Middle Paleolithic complexes, but evidence of the use of bulk Upper Paleolithic splitting is recorded only in the lowest layers (XVIII, XIX, XX).

The materials of the Middle Paleolithic layers do not allow us to speak about any directed changes in technological priorities. Plates are better represented in the early complexes than in the late ones, where their production becomes quite rare.

Tools

The role of the main raw material for the production of tools in all layers belongs, as noted, to radiolarite: 50 - 82% of all things with traces of secondary processing are made of it (see Figure 5). Flint is used in 11 - 30% of cases. Other rocks (quartzite, chalcedony, limestone, and volcanic rock; see Figs. 5, 6) were subjected to secondary processing only sporadically, although quartz tools are found in all layers. The use of the same rocks and approximately the same proportions is noted in the Upper Paleolithic layers (Koumouzelis et al., 2001a). In cases where the original shape of the chips is not significantly changed by retouching, you can determine the type of workpieces selected for the production of tools. Often guns of all classes were made on small and rather thick flakes. For scrapers and knives, edge chips (for example, 10-17% in layers XIII - XVI) and asymmetric short flakes, as well as chips with a natural edge, were often chosen. Primary chips were used as blanks for tools sporadically and in limited quantities (for example, 2.4 % in layer XI and 7 % in layers XIV and XV). Radiolarite tiles were selected even less frequently. Sometimes worked nuclei were re-formed into tools (scrapers). Retouched Levallois blanks are found in non --

page 10

11 Levallois and lamellar cleavages in the Klisura-1 layers.
a large number, excluding layer X. Plates were selected much more frequently (see Figures 5, 4, and 11); the highest proportions of tools on plates were recorded in layers XVIII, XIX, and XX (18-23%), and as we move up, this indicator gradually decreases (Figure 11). Most often, tools were made on non-Valloise flakes (>77 %).

Middle Paleolithic complexes are characterized by small tools (average length 30.5-35.5 mm), although in the lowest layers there is a slight increase in the size of blanks, especially plates and tools on plates. Tools were usually designed by means of semi-circular and steep scale retouching, while other types of it, including sub-parallel, flat, kina, bifacial, plano-convex and haphazard, were used less often. Kin and polukin retouching has been used sporadically for the design of massive scrapers, limes, and other convergent tools. It is especially common in the lowest layers (for example, on 5.3 % of all whole tools and on 7.6 % of scrapers and pegs in layer XIX). As a rule, retouching is marginal and not intensive. Many tools are only marginally modified blanks with marginal retouching. However, there are examples of exciting, intense one-sided retouching, as well as bifacial processing (scrapers, points). Traces of intensive shaping retouching are especially common in the lower layers. In all Klisura-1 complexes, the dominance of facial retouching is reflected (for example, 89.9% in layer XVI), signs of ventral retouching are rare (for example, 5.7 % in layer XVI and 8.2 % in layer XV; these are the maximum values). The frequency of bifacial retouching is approximately 3 % (for example, in layers XV and XVI); traces of alternating retouching are rare. Various types of truncation and thinning (basal, distal, and obuskovy) of the scrapers, knives, and especially the spikelets are observed, although not very often. It appears that there was a relationship between the nature of the blanks and how they were used. Ordinary flakes and some plates served as tools without any processing or were partially modified (retouching, thinning, truncation). Most of these blanks were transformed by flake and marginal retouching into single scrapers, toothed products (see Figs. 7, 11), etc.D. The form of such things then underwent only minor changes. On the contrary, large and thick flakes and plates were subjected to intensive processing by means of sheer, steep, gripping and bifacial retouching, which turned them into various scrapers (with one or more working edges), limas, and pegs. Thanks to this modification, the original mass of the artifact was significantly reduced, and sometimes the original type of weapon was changed. Tools that have undergone intensive transformation look asymmetrical relative to the central axis and somewhat disproportionate (high cross-sections, limes), there is too much width and shortening (some angular and transverse scrapers and even sharp points) or narrowness and elongation, giving them a pseudo-plate character; some tools - with two parallel or convergent working edges.

Among the items with secondary processing elements, scrapers dominate - from 63 to 84.7 % of all tools (excluding debris). The highest percentage of scrapers is in the middle part of the column (layer XIII), and the lowest values of this indicator are noted at the top-

page 11

12. Types of gun kit in the Klisura-1 layers.

Fig. 13. Points: elongated (1, 4, 11, 12, 17, 18), with distal thinning (2), short triangular (3), bifacial (5, 6, 8), Tejac (7), with signs of proximal thinning (9, 10, 14), asymmetric (13, 15), leaf-shaped (16), convergent scraping (19). 1, 4, 11, 14 - layer XIV; 2, 3, 6, 10 - XI; 5 - XV; 7, 17-XIII; 8-XIX; 9-XXc; 12, 15 - VIII; 13-VII; 16-XX; 18, 19-X.

in the upper (VII, VIIa) and lower (XVIII, XIX) layers (Fig. This fact reflects a significant increase in the proportion of Upper Paleolithic tools in the upper layers (16.8 - 18%), and in the lower layers of spikelets, flakes, and plates with signs of retouching. The gradual disappearance of Upper Paleolithic tools (scrapers, incisors, punctures, and especially numerous chisel-shaped items) in Layers VII, VIIa, and VIII, until their complete replacement by Mousterian forms (scrapers, spikelets), starting from layer XI, is well documented. In the lower layers (XIV-XIX), Upper Paleolithic tools are absent or isolated, even though the proportion of plates among them significantly increases (layers XVIII, XIX). Mousterian pinnules are found in all layers (excluding layer VIIa). Their specific weight among tools is small (1.4 - 3.7 %); it slightly increases only in layers XVIII (4.7 %) and XIX (7.5%). The frequency of occurrence of such types of tools as toothed products and retouched flakes is low or moderate (2.6 - 11.9 and 7 - 17%, respectively). Notched forms and knives with a natural edge are even rarer. The share of the latter in some cases reaches 4.4-6.5 % (layers XV-XIX; there are also individual knives with a retouched edge). Other types of tools are represented by individual items scattered in different layers (scrapers / raclette, flakes with truncation and thinning, beak, combined tools, "chopper"). Quite a lot (approx. 10 %) of unidentifiable gun wreckage; in some cases

page 12

they account for up to 25% of products with traces of secondary processing (layer XVI).

Spiky points

There are two main classes of spikelets: Mousterian elongated (see Fig. 5, 4; 13, 1, 11, 12, 17, 18) and short, often massive ones (see fig. 13, 3, 5, 6, 14). Other types, such as Tejac spikelets (see Figures 13, 7) and Kinson spikelets, are very rare. Tejac pinnacles are found only sporadically, for example, in layers XIII and XIV. Mousterian pinnacles of both categories are divided into symmetrical and asymmetrical; they can be triangular in plan (see Fig. 5, 3; 13, 3, 9, 10), triangular-versatile, angular (see Figs. 13, 75), punctured, beaked, segmented, and leaf-shaped (see Figs. .13, 16, 17), including double-ended ones. Depending on the nature of the workpieces (plates, flakes, and even tiles), they vary in the degree of elongation and massiveness. There are also different types of retouching used for their production: marginal, scaly and cool exciting, kina, and occasionally bifacial (see Figs. 13, b, 8). Some pinnacles have undergone distal or basal, ventral or dorsal thinning. Their end parts are widely represented, although the exact typological identification of the latter, as well as a number of other convergent objects, is not always easy. Spikelets of various types are found throughout the Middle Paleolithic layers, but the full range of diversity of these tools is represented only in the lowest XVIII, XIX and XX.

Scraped it

Single scrapers predominate over those with multiple working edges and make up, as a rule, 60-78% of tools in this category, only in layer VII their share decreases to 52.2 %. The frequency of side scrapers is high (see Figs.5, 10, 11). They are usually about 40-50%, in layers XIV and XVIII - about 60 %, and in layers X and XIX - up to 63 % (Fig. Cross scrapers (Fig. 15, 9, 20; 16, 2, 3) they are not so numerous, although they are also quite numerous (most often about 10 %, and in layers VIIa and XVII-up to 15-16 %). 15, 19) make up from 5 to 13-16 %. Double scrapers

14. Main types of scrapers in the Klisura-1 layers.

15. Lateral (1-4), double (5-7), angular (8), transverse ventral (9), angular (10-12, 16), convergent (13, 14), trapezoid (75), oval bilateral (17, 18), beveled (19), transverse kina type(20). 1, 3, 4, 11, 14, 19, 20 - layer XIV; 2, 8, 9, 13 - XV; 5-7, 10, 18-XI; 12-X; 15-17-XII.

page 13

16. Scrapers: double-truncated and cast (7), transverse (2, 3), angular (4), segmental two-sided, prondnik-type knife-like (5), angular (b), convergent (7, 8), trapezoidal (9, 10), segmental two-sided (77), Limasi (12, 13). 1 - layer XII; 2,3,5 - XV; 4, 11, 12 - X; 6, 9 - XVI; 7 - VII; 8 - XIII; 10-XI; 13-XVIII.

(see figure. 5, 9; 15, 5 - 7) most complexes contain about 10-14 %; the lowest concentrations are found in layers XX (4 %) and XVI (6 %), and the highest in layer VII (23 %). Frequency of convergent scrapers (see Fig. 15, 75, 14; 16, 7, 8) it varies from 2-5 to 10 %, reaching a maximum value of 22 % in layer XX. Angular (dejete) (see Fig. 5, b, 7; 15, 10 - 12, 16; 16, 6) - they show a similar trend and make up from 5 to 9-14% of the guns in this category. Limasi (see fig. 5, 7, 5; 16, 72, 13) They are not present in all complexes, but they are very characteristic of the middle and especially lower layers (1-2%). The highest frequency of convergent and angular scrapers is recorded in layers VIII, XI-XIII (20 %) and XX (36 %, according to incomplete calculations). Convergent tools in general (including pinnacles) are generally very characteristic of the Middle Paleolithic layers of Klisura-1 and account for an average of 10-25% of tools. Other scrapers, including corner ones (see fig. 15, 8; 16, 4), with traces of alternating retouching, with signs of core thinning (undergrowth) (see 16, 7), as well as with a continuous working edge covering the entire perimeter and sometimes with elements of bifisial retouching, trapezoidal (see Fig. 5, 2; 15, 15; 16, 9, 10), rectangular, oval (see Figures 15, 17, 18), and segmented (see figure. 5, 8; 16, 5, 11) forms are rare and not represented in all layers. Segmental scrapers are more often found in the lower layers - from XIV to XX. In many layers, the share of scraper fragments (12 - 27 %) and chips of their re-registration is high. The first ones are absent only in layer XIII, and in layers VIII, XI, XV and XIX they are few (1 - 5 %), which is compensated by the abundance of fragments of other tools. A high percentage of fragmented weapons can be a consequence of the significant intensity of their use, activation and re-registration, and in some cases reflect the impact of post-deployment processes. To make a choice between these hypotheses, you need to test them specifically.

Conclusion

The analysis of statistical data, technology and morphology of tools gives the impression of homogeneity of the Middle Paleolithic complexes of Klisura-1. The similarity is evident in the nature of raw material selection, the" microlitoid " nature of the industry, the incompleteness of the cleavage cycles presented in the materials (testing and decortication of the initial cleavage items was carried out outside the parking lot), the degree of harmony of the nuclei and re-registration of some tools, the use of planar parallel/discoidal/centripetal cleavage strategies in combination with Levallois technologies (preferential and recurrent methods for obtaining flakes and sometimes also points and plates), features of site preparation (faceting indices are low and medium; sites created by a single chip predominate), the dominance of the mousterian gun set (single side scrapers, many scrapers with several working edges, as well as convergent tools, including pointers), the predominance of the front half-round and steep scaly retouching, supplemented by the use of a large number of kin's retouching, exciting and bifacial, on the one hand, and subtle marginal -

page 14

with another. As for the differences and changes, it is necessary to note the presence of a significant number of tools of Upper Paleolithic types, primarily chisel-shaped, in the uppermost layers (VI-VIIa) and their disappearance after layer X; reflection in the lowest layers (XVIII-XX) of the use of Upper Paleolithic methods of bulk splitting, along with Levallois methods of manufacturing flakes, aimed at getting plates and plates. Plate splitting, however, did not lead in this case to the appearance of Upper Paleolithic tools, except for retouched plates and Middle Paleolithic tools on plates. Other technological and typological features vary only slightly from layer to layer. Thus, the upper Middle Paleolithic complexes differ from the rest mainly in typological terms (the presence of a significant Upper Paleolithic component), and the lower ones - in technological terms (plate splitting). Levallois splitting methods are recorded in all complexes, and they are best represented in some layers of the middle part of the Mousterian sequence. This picture, despite the great similarity of the monuments, generally does not agree well with the scheme established for Asprohalico (Papagianni, 2000), where the "micro - moustache" lies above Levallois-mousthier. The absence of bifacial products in the upper layers of Klisura-1 and their presence (albeit in small quantities) in the lower layers also runs counter to the general direction of local evolution. In general, the new comparative data obtained for Klisura-1 indicate a significant technological diversity with the homogeneity of many complexes. The materials of the monument demonstrate the presence of both common features with local Middle Paleolithic industries, as well as unique typological and technological features and specific trends.

Acknowledgements

The research was made possible thanks to the financial support of the National Foundation for Scientific Research and the University of Liege (2000-2001) and the Belgian Ministry of Science and Culture (MO/38/003 and MO/38/010), as well as a scholarship received by one of the authors (K. A. Kropotkin). Sobchik) from the National Endowment for Scientific Research and the University of Liege.

The authors are grateful to Dr. Rebecca Miller for her assistance in preparing the text.

List of literature

Darlas A. Le Paleolithique inferieur et moyen de Grece // L'Anthropologic. - 1994. - Vol. 98, N 2/3. - P. 305 - 328.

Darlas A., Lumley de H. Palaeolithic research in Kalamakia Cave, Areopolis, Peloponnese // The Palaeolithic Archaeology of Greece and Adjacent Areas. ICOPAG Conference, Ioannina, 1944 / Eds. G.N. Bailey, E. Adam, E. Panagopoulou, С Perles, K. Zachos. - L.: British School at Athens Studies, 1999. -Vol. 3. - P. 293 - 302.

Karkanas P., Bar-Yosef O., Goldberg P., Weiner S. Diagenesis in prehistoric caves: the use of minerals that form in situ to assess the completeness of the archaeological record // J. of Archaeological Science. - 2000. - Vol. 27. - P. 915 - 929.

Karkanas P., Koumouzelis M., Kozlowski J.K., Sitlivy V., Sobczyk K., Berna E., Weiner S. The earliest evidence for clay hearths: Aurignacian features in Klissoura Cave 1, southern Greece // Antiquity. - 2004. - Vol. 78, N 301. - P. 513 - 525.

Koumouzelis M., Ginter В., Kozlowski J. K., Pawlikowski M., Bar-Yosef O., Albort R. M., Litynska-Zajac M., Stronewicz E., Wojtal P., Lipecki G., Tomek E., Bochenski Z. M., Pazdur A. The Early Upper Palaeolithic in Greece: The Excavations in Klissoura Cave // J. of Archaeological Science. - 2001a. - Vol. 28. - P. 515 - 539.

Koumouzelis M., Kozlowski J., Escutenaire C., Sitlivy V., Sobczyk K., Valladas H., Eisnerat-Laborde N., Wojtal P., Ginter B., Kaczanowska M., Kazior B., Zieba A. La fin du Paleolithique moyen et le debut du Paleolithique superieur en Grece: la sequence de la grotte 1 de Klissoura // L'Anthropologic. - 2001б. - Vol. 105. -P. 469 - 504.

Kyparissi-Apostolika N. The Palaeolithic deposits of Theopetra Cave in Thessaly (Greece) // The Palaeolithic Archaeology of Greece and Adjacent Areas/Eds. G.N. Bailey, E. Adam, E. Panagopoulou, C. Perles, K. Zachos. - L.: British School at Athens Studies, 1999. - Vol. 3. - P. 232 - 239.

Panagopoulou E., Karkanas P., Esartsidou G., Kotjabopoulou E., Harvati K., Ntinou M. Late Pleistocene Archaeological and Fossil Human Evidence from Lakonis Cave, Southern Greece // J. of Field Archaeology. - 2002 - 2004. - Vol. 29, N 3/4. - P. 323 - 349.

Papaconstantinou V.S. Mcromousterien: les idees et les pierres. Le Mcromousterien d'Asprochaliko (Grece) et le probleme des industries microlithiques du Mousterien: These de Doctorat - Univ. Paris X, 1988. - 546 p.

Papagianni D. Mddle Palaeolithic Occupation and Technology in Northwestern Greece. The Evidence from Open-Air Sites. - Oxford: British Archaeological Reports. -2000. - 218 p. - (BAR International Series; N 882).

Sitlivy V. Le Paleolithique moyen ancien: variabilite technologique, typologique et fonctionnelle en Europe // Prehistoire europeenne. - 1996. - Vol. 9. - P. 117 - 155.

The article was submitted to the Editorial Board on 06.02.07.

page 15

Permanent link to this publication: