On the Cultural Geography of the Eastern Caucasus and Southern Caspian in the Mesolithic

At a time when Soviet Paleolithic experts were searching for specific features in the material remains of the Stone Age (1960s–1970s), in the eastern Caucasus the Chokh and Trialeti Mesolithic cultures were identified. These paleo-cultural studies aimed at the identification of groups of sites that would meet the notion of “archaeological culture” as a typological structure corresponding to the upper level of the triad: attribute–type–culture. Identification of a particular archaeological culture was generally recognized as proven if groups of products of specific types, or even of a single cultural form, were identified for a certain set of sites.

It should be admitted that the identification of the Stone Age cultures has in practice very often outstripped the methodological relevance of the research procedure. Even when the attribution of sites to one community seemed justified, questions were raised as to whether the entity in question was an archaeological culture (a narrow local unity), a cultural community (a group of related cultures) or a community made up of different sites that shared

a common developmental pattern. Addressing these issues was hampered by the difficulty of achieving the goal of typological analysis, namely the identification of ideal types (not replaceable by subtypes, supratypes, categories, etc.), which would be used for the comparative analysis of materials.

In the past, archaeological cultures were usually identified not on the basis of clearly defined typological characteristics, but often by intuition. In this way, most of the Upper Paleolithic and Mesolithic cultures of the Caucasus were identified (Bader, Tsereteli, 1989)—the Imereti, Chokh, Trialeti, Black Sea, and Gubs.

The Trialeti culture according to Soviet (Georgian and Russian) researchers

The Trialeti Mesolithic archaeological culture was identified in the 1970s by Georgian researcher M.K. Gabunia from the materials of two cave sites (more precisely, on the grounds under rock-shelters): Edzani and Zurtaketi, located in the southern spurs of the Trialeti Range (Gabunia, 1976; Gabunia, Tsereteli, 1977) (see Figure ). The area of the culture’s distribution was determined to be approximately within the territory of Eastern Georgia.

The lithic industry at the Edzani rock-shelter site consists of 21,628 items, of which 1910 show signs of secondary working. The Zurtaketi lithic assemblage is much more modest—386 items, 21 of them with traces of secondary working.

The materials of the culture under consideration were not dated until the last decade. Recently, the dates have been obtained for the Bavra Ablari rockshelter site (Georgia) and for the Mesolithic layer of Damjili Сave (Azerbaijan). Both sites are located in the area of the Trialeti culture. Four radiocarbon dates for the Mesolithic deposits of Bavra Ablari correspond to a range of 9500–8700 cal BP (Varoutsikos et al., 2017: 243). The Mesolithic layer of Damjili Cave is dated to 6400–6000 cal BP (Nishiaki et al., 2019). With these dates, the chronological range of the culture can be estimated as from 9500 to the end of 7000 cal BP.

As for the two sites from which the Trialeti culture was originally identified, their age was estimated purely on the basis of a comparison of typological features. The reliability of these relative dates raises significant doubt. As is known, M.K. Gabunia attributed the Zurtaketi site to the Early Mesolithic, and Edzani to the Late Mesolithic (Gabunia, 1976). This conclusion was based on the fact that Edzani, in contrast to Zurtaketi, yielded elongated asymmetrical triangles and Gravettoid points, as well as numerous backed bladelets. That is, morphological groups of items that, according to the modern approach, should belong to the Early Mesolithic were chosen

Map of the sites mentioned in the article.

1 – Chokh; 2 – Bavra Ablari; 3 – Edzani; 4 – Zurtaketi; 5 – Damjili; 6 – Kmlo-2; 7 – Komishan; 8 – Hotu; 9 – Kamarband; 10 – Ali Tepe; 11 – Dam-Dam-Cheshme; 12 – Djebel; 13 – Kaskyr Bulak.

as elements indicating the relatively late age of the site within the Mesolithic. This is the case with the cultures identified in the northeastern Caucasus (Amirkhanov, 1987), and the southern and northern slopes of the western Caucasus (Tsereteli, 1973; Leonova, 2015, 2019; Aleksandrova, Leonova, 2017; Bar-Oz et al., 2009).

The materials from the Edzani site do indeed show features indicating the Late Mesolithic. These probably include the small size of such items as segments and trapezoids. The tools fashioned with distributed retouch and isolated symmetrically stemmed points can be classified as distinctly late. The indisputable evidence of Neolithic material in the collection is a core for pressure-flaking using a lever (see (Gabunia, 1976: Pl. XV)).

Giving the typological description of the Edzani toolset, the researcher points to the “abundance of such tools as geometric microliths (among these, noteworthy are the large series of asymmetric triangular inserts, rather developed forms of segments and trapezoids, segments with blunted arcs, and trapezoids with blunted upper bases), tools resembling arrowheads, numerous various perfectly worked insert-bladelets; micro-endscrapers on bladelets; and rounded endscrapers on microflakes, etc.” (Gabunia, Tsereteli, 1977: 34).

When considering the composition of the Edzani collection from the point of view of modern science, it should be noted that typologically it is heterogeneous, and includes various diachronous components. Similar observations have also been made by other researchers (Kozlowski, 1999). The bulk of this collection relates to the Mesolithic; it reveals features that distinguish this industry from other Mesolithic industries of the western Caucasus. Specific to the Trialeti complex are small cores with flattened flakingsurfaces, sometimes showing disk-form (along with prismatic and conical nuclei) and, consequently, the great importance of flakes serving as blanks for tool manufacture. Peculiar are such distinctive tools as trapezoids (these are close to carinated pieces). The implements in the form of a blade, semicircular in plan view, with a solid, flat, thinning inverse retouch on one or both ends, from the Edzani assemblage, have long remained in the background (Gabunia, 1976: Pl. XI, 28 ). Most likely, it was a product of this type that was identified by modern researchers under the name of “Damjili-type tool” during the recent studies of the Mesolithic layer of the eponymous cave in Western Azerbaijan (Nichiaki et al., 2019).

Considering the Trialeti culture, Gabunia points to the absence of signs of the use of pressure technique for the production of blanks in it. However, this feature cannot be treated as culture-specific, because this technique is not typical of the Mesolithic of the whole eastern Caucasus.

In general, the above typological description of the Trialeti culture can hardly be regarded as complete enough to serve as a basis for unambiguous attribution of a certain site to this cultural formation. Judging by this characteristic, this formation can include the materials from a variety of Mesolithic sites on a huge territory. This is precisely what happened later with the transformation of Gabunia’s “Trialeti Mesolithic culture” into Kozlowski’s “Trialetien”.

Trialeti industry according to S.K. Kozlowski

S.K. Kozlowski repeatedly referred to the issue of the Trialeti culture (Kozlowski, 1994, 1996, 1999; Kozlowski, Aurenche, 2005) in connection with the study of the Early Holocene cultural geography in the area of the Fertile Crescent and adjacent regions. He drew conclusions about the cultural phenomenon in question on the basis of study (mostly according to the literature) of the materials of Edzani (Gabunia, 1976), Hallan Çemi (Turkey) (Rosenberg, 1994), Ali Tepe (Iran) (McBurney, 1968), Kamarband (Belt), layers 28–11 (Iran) (Coon, 1957), Chokh, layers E–C (Amirkhanov, 1987), Dam-Dam-Cheshme II, layers 7–3 (Korobkova, 1977), and Nevalı Çori (Turkey) (Schmidt, 1994).

To denote cultural formations spread in the Middle East to the north of the Zagros Mountains, in the Taurus Mountains and in the Caucasus, Kozlowski uses the concepts with different meanings in terms of scope: “Trialetien industry” (Trialetien in the broad sense), “typical Trialetien” (Trialetien in the narrow sense), and “Caucasian-Caspian cultural area”. These concepts do not imply clear attributes and are not organized into a strict scheme with hierarchical levels—although the idea of hierarchy is present here, even if not clearly.

According to Kozlowski, the Trialetien in the broad sense is a Late Pleistocene–Early Holocene industry common to populations who are not engaged in a production economy, and provide for their needs mainly through hunting. This industry is generally contemporaneous with the proto-Neolithic and pre- pottery (Early Neolithic) cultures of the Taurus, Zagros, and Mesopotamia, such as Mlefaatian, Nemrikien and partly PPNB (Pre-Pottery Neolithic B). All these cultures date back to between 11 and 7 ka BC. The north-south boundary of their area of distribution runs from the southern Greater Caucasus to the main ridge of the Zagros Mountains, while the east-west line runs from the Eastern Taurus to the Kopetdag and Nebitdag. In this vast territory, there is an area where the industry in question is represented in its typical form. This is a region that includes Southern Georgia (Edzani site), and the southern and western coast of the Caspian Sea (Ali Tepe, Kamarband, Dam-Dam-Cheshme II sites)* (Kozlowski, 1996).

In the area of distribution of the Trialetien, the Caucasian-Caspian cultural province is identified, which includes the territories of the Imereti, Black Sea, and Shan-Koba Mesolithic cultures** (Ibid.). The researcher writes that “with high probability, the Trialetien was an industry of hunter-gatherers in the forested territories of the Caucasus, Elbrus, Kopetdag, Nebitdag, Eastern Taurus, and the northern slopes of Zagros; possibly, it was spread on the Iranian Plateau” (Ibid.: 163).

In the Trialetien industry, primary flaking consisted of prismatic, sub-conical, and cube-like cores, which are designed to produce both blades and flakes. Discoidal cores on small chips or concretions, as well as on large flakes, regularly occur. Core-flaking was carried out using a punch technique.

According to Kozlowski, geometric microliths form a great part of the retouched tools in this industry. They are predominantly large in size (over 15 mm long), often made on blades or fragments thereof, but not on bladelets. The collections of sites of this type contain numerous “para-Gravettes” and long, narrow, and ordinary (small) segments; there are also elongated asymmetric triangles and isosceles triangles, as well as large asymmetric and symmetric trapezoids.

The beginning of the development of the Trialetien industry in its typical form—not later than

10.5 ka BC—is established by the materials of the southern Caspian region (Ali Tepe site). The second phase of the industry’s development—9 ka BC— is determined as the phase of the widespread use of trapezoids. The third phase, dated to 8–7 ka BC (Edzani, Hallan-Chemi stage), is distinguished by a significant decrease in the proportion of trapezoids in the toolset.

In the west and north of the common area of distribution of the Trialetien, the destiny of the culture is thought to have developed differently. Kozlowski writes: “In the Kura River valley [rough error in localization. – H.A. ], the Chokh variant of the Trialetien is represented in somewhat modified technological form with pottery, while in the southeast of Turkey the Trialetien is transformed to the local variant of PPNB, probably as early as the beginning of 7 ka BC (Çatalhöyük, early stage)” (Ibid.).

Chokh culture

The most representative site of this culture is the Chokh site, located in the central (mid-mountain) part of Dagestan (northeast of the Greater Caucasus), with cultural deposits from the Mesolithic, Neolithic, and the Bronze Age. The notion of the “Chokh archaeological culture” emerged in the mid-1960s with the replacement of the “stadiality” approach in explaining the Upper Paleolithic of the Caucasus by the concept of culturalism (Bader, 1965). Initially, the Chokh culture was perceived as mainly Late Paleolithic. Four of the lower six layers of the site were wrongly dated to the Upper Paleolithic (Kotovich, 1964). Almost 30 years after the first excavations, it has become clear that the upper layers of the site are Neolithic (layer C) and Bronze Age deposits (horizon C1), while the two lower lithological layers (layers D, E) contain Mesolithic archaeological materials (Amirkhanov, 1987). Then, the features of primary working techniques, flint tools of specific types, and groups of implements were revealed, that have been recorded in such a combination only at this site and (almost) nowhere else.

The most peculiar Chokh features were recognized to be the following implements: points (arrowheads) of the Chokh type (in four variants), knives with distal retouched backs, low elongated asymmetrical triangles, and cores of archaic shapes (discoidal, similar to Levallois), which occur in the materials of all stages of the Chokh culture development.

A distinctive component of this culture is trapezoidal pieces—always carinated, sometimes asymmetrical, with straight or slightly concave sides. It was noted that the flint tools from the Chokh site and their working technique were changing from the lower Mesolithic layer to the upper one very gradually, so that one cannot conclude about qualitative transformations. In the goods from the Neolithic layer, there appeared pottery, brand-new types of tools (harvesting-knives and grinders), and items indicating the start of house-building and the formation of new subsistence patterns.

Notably, the significant changes accompanying the introduction of Neolithic innovations, at a certain developmental stage of the Chokh culture, had almost no effect on the flint industry’s typology. The Neolithic layer’s assemblage shows continuity with the materials of the underlying (Mesolithic) layers. This is reflected in the basic types of culturespecific flint tools: Chokh-type points, knives with retouched distal slanted backs, elongated asymmetrical triangles, and certain varieties of carinated trapezoids. Cultural ties with the industries of the underlying Mesolithic layers are evidenced by the presence of sub-Levallois cores, which were also in use during the Neolithic. These forms differ from their archaic analogues, first, in their size (at the Chokh site, these are always small), and second, in their trend for production of small flakes to make arrowheads (mainly of the Chokh type). In other respects, the described cores correspond to the morphology and characteristics of almost all well-known (Lyubin, 1965) variants of pieces of this Middle Paleolithic category, including varieties with a faceted (in some cores from the Chokh site, faceting is replaced by a retouch) edge of the working part of the striking platform. Discoidal cores also occur sporadically in the Neolithic layer.

Among the typical elements of the lithic evolution of the Chokh culture, mostly important is the appearance, in the Neolithic layer of the site, of signs of the use of manual pressure technique for the production of blanks in the form of micro-bladelets.

The area of distribution of the Chokh culture, according to the modern data, includes the mountainous part of Dagestan, or the central part of the northeastern Caucasus.

One radiocarbon date was obtained for each Mesolithic layer of the site: layer D, excavations 1957, charcoal (IGAN_ams 6313), о 68.3 % 12,83012,959 cal BP; 2o 95.4 % 12,784-13,010 cal BP;

averaged date 10,341 BC; layer E, excavations 1980, bone (IGAN AMS 8112), о 68.3 % 12,83012,959 cal BP; 2o 95.4 % 12,784-13,010 cal BP; averaged date 10,872 BC. Judging by these dates, the Chokh culture dates back to the Late Dryas (ca 11,000 cal BC). Its Mesolithic phase probably lasted until the beginning of the Atlantic (late 7th millennium BC). This assumption requires confirmation by absolute dates. Research in this direction is ongoing, and will hopefully yield more definite results in the near future.

Discussion

The above-mentioned expert in the Late Stone Age archaeology, S.K. Kozlowski, studied and interpreted many Mesolithic materials of the eastern Caucasus. Especially noteworthy is his contribution to the study of the geography of the Near Eastern Caucasian cultures: in particular, relations between the Mesolithic cultures of the Caucasus and the cultures of Zagros, Taurus, Anatolia, southern Caspian region, Central Asia, and Iranian Plateau (Kozlowski, 1994, 1996, 1999; Kozlowski, Aurenche, 2005). Nevertheless, these works show certain factual errors and are unconnected with specific materials. For example, in the description of the Chokh variant of the Trialetien, he points to the Kura River valley in the southern Caucasus as the area of its distribution rather than the northeast of the Greater Caucasus (Kozlowski, Aurenche, 2005: 52). Another example: the Mesolithic culture of the western Caucasus is perceived by Kozlowski as identical to the Shan-Koba culture of the Crimea (Ibid.). The issue of the typological features bringing together the Mesolithic materials of the Crimea and Caucasus was also studied by other researchers (Bader, 1961); probably, these are the works that Kozlowski relies on. It is important to note that the researcher’s predecessors see the origin of this proximity in a single line of development of the cultures in the compared regions. It is not quite clear why Kozlowski, being a supporter of the concept of multilevel (four levels) manifestations of similarity between industries, has not adopted this point of view (Kozlowski, Aurenche, 2005).

The notion of Trialetien, proposed by Kozlowski, had attracted almost no attention from Russian researchers of the Caucasian Mesolithic, nor from their Southern Caucasian colleagues. Meanwhile, it deserves a thorough analysis. First, noteworthy is a lack of clarity in the methodological justification of the cultural community called the “Trialetien” and the extensiveness of its area of distribution. The typological justification for the Trialetien proposed by Kozlowski seems vague: the area of distribution of this industry included the whole Caucasus, part of Southeastern Europe (the Crimea), the Anatolian and Iranian highlands, Northern Mesopotamia, the southern Caspian region, and the western part of Central Asia (at least Kopetdag, Nebitdag). The excessive size of this area was evident to Kozlowski himself; in his 1996 work, the Crimea was excluded (Kozlowski, 1996).

Notably, the northern slopes of the eastern Greater Caucasus are not mentioned among the territories of the Trialeti industry (Trialetien). As noted above, this is due to the fact that Kozlowski erroneously believed the Chokh site (one of the principal sites of that industry) to be situated in the Kura Valley in the southern Caucasus. In fact, this site was situated in the northern part of the Greater Caucasus Range, and its typical landscape was not a river valley in the Caspian lowlands, but a mountain steppe on a plateau-like upland with absolute heights of 1700–1800 m above sea level.

If we analyze the technical-typological features of the Trialetien listed by Kozlowski in his publications of different years, we should note their almost complete coincidence with the features that we used to define the Chokh Mesolithic culture (Amirkhanov, 1987). According to Kozlowski, in the typological set of the Trialetien, only the Chokh-type points are absent. In terms of methodology, this fact is very indicative. The Chokh-type points are a culturedefining type; they form the basis of the specificity of the Chokh culture. To acknowledge this means to agree that the Chokh culture has a special place in the Trialetien. From this alone it follows that there is a need to structure the materials included in the broad concept of “Trialetien industry” and to justify this cultural community (if it really existed) as a multilayered and multi-component entity.

Earlier, the items similar in their typological status to Chokh-type points (in fact, culturally diagnostic forms) served as a basis for identifying archaeological cultures in the Mesolithic of the Caucasus. It is precisely because of this kind of diagnostic material that the Chokh culture was never considered analogous to the Trialeti. The common feature of the two cultures was that they had practically the same features that differentiated them from the Mesolithic materials of the western Caucasus. If this kind of cultural similarity between the Trialeti and the Chokh cultures has not been described in the literature, it is only because it was perceived as obvious and not challenged by anyone. Another reason was, as noted above, a certain ambiguity in the typological justifications for the identification of the Trialeti culture.

In view of the above, it is surprising to note the complete absence of any mention in Kozlowski’s works of the Chokh archaeological culture as a separate cultural entity in the region under study. This is despite the fact that the Caucasian culture has much more material and typological references than any other Mesolithic or Neolithic culture of that region. For example, there is a specific type of piece with the eponym “point (arrowhead) of the Chokh type”.

This lack of attention to the methodological aspect of the distinction between the notions of “typical Trialetien” and “Trialetien industry” could not but affect the identification of their place in the hierarchy of notions and the boundaries of the areas of cultural communities. The typological content of the first notion is more definite than that of the second. As for the typical Trialetien, according to Kozlowski, the list of the relevant sites, stricto sensu, curious as it may seem, lacks the Trialeti sites themselves, if we perceive the Trialeti culture as it was originally identified (Gabunia, 1976) and repeatedly described in the literature.

It is noteworthy that the notion of “typical Trialetien” (i.e. Trialetien stricto sensu ), according to Kozlowski, corresponds to the sites of the southeastern Caspian Sea coast (Kamarband, Hotu, Ali Tepe, etc.), the eastern Caspian region (Dam-Dam-Cheshme II, Djebel), and the northeastern Caucasus (Chokh). Notably, we have already identified exactly this group of sites as a certain broad community forming the “Southern Caspian Mesolithic area” (Amirkhanov, 1987: 202–203). We regarded it as a unity of related archaeological cultures, which have deep genetic roots and differ from those of the Northern Caspian Mesolithic area.

Now, almost 40 years after the publication of this point of view, representatives of the new generation of Iranian archaeologists have given the culture of the above region a slightly modified, but essentially similar name—“Caspian Mesolithic” (Jayez, Nasab, 2016). In this variant, the cultural entity in question is associated with the territory of the southeastern

Caspian coast within the north and northeast of modern Iran. In addition, the sites of this region were rather categorically excluded from the relatively narrow cultural entity classified by Kozlowski as the Caucasian-Caspian community within the Trialetien, but also from the Trialetien in its broadest sense. In the newer concept, the sites of the eastern Caspian region (Dam-Dam-Cheshme II, Djebel, Kaylu) are considered as cultural analogues of the Iranian sites of the southeastern Caspian region (Ibid.).

The Chokh site, located in the mountains that fringe the Caspian Sea from the west, was not included by the Iranian researchers in the Caspian Mesolithic area. This can be explained by ignorance of the Chokh materials. As noted above, Kozlowski handled the Chokh materials by relying on generalizations of the authors, most of whom had not seen the Chokh materials themselves. New researchers of the sites in northeastern Iran already use in their developments the third-level generalizations made by Kozlowski. This explains the uncertainty, in particular, about the boundaries of the “left wing” of the “Caspian Mesolithic”. In cases where these researchers have full knowledge of cultural formations, they are extremely precise in determining their localization. For example, they consider the central part of the Iranian Plateau (at least its eastern regions) as a region whose neolithicization was influenced by Zagros cultural impulses, but in no way by the Trialetien (Nasab, Solange, Shirvani, 2019).

Among the South Caspian Mesolithic sites, the closest to the Chokh site is Ali Tepe, the earliest in the specified group. Typologically, their similarity is expressed in the materials of both sites by the presence of elongated segment-like points, knives with distal retouched backs, asymmetrical triangles, and single carinated trapezoids. Individually, these implements are typical of both the Final Paleolithic and the Early Mesolithic of the Caucasus; but together in one toolkit they occur rarely, especially in combination with a flat core with a straight flaking surface. Each of the sites under consideration shows its own specific types of hunting-weapon: at the Chokh site, these are points (arrowheads) of the Chokh type, and at Ali Tepe the stemmed forms of arrowhead (with lateral and symmetrical marginal notches). Later on, in the southern Caspian region, the latter are replaced by large asymmetrical trapezoids, each with a notch on one lateral side, which might be called beveled points (arrowheads) with notches in their bases.

Notably, the materials of the Lower Mesolithic layer of the Chokh site and those of the Early Mesolithic of Ali Tepe are chronologically correlated.

Conclusions

The use of the notion of “Trialetien” proposed by Kozlowski in its broad meaning is feasible only to distinguish the “barbaric” Mesolithic huntergatherer cultures that spread north of the Taurus and Zagros mountains from the Early Neolithic cultures of the Fertile Crescent with production economies, which appeared there no later than the late 9th millennium BC.

Technologically, the “Trialetien” sites are united by the absence of any signs of the manufacture of stone blanks by pressure technique. This undoubtedly adds flavor to the industry, but is not enough to distinguish a specific cultural-chronological entity on this basis. After all, this feature is inherent in any cultural community that was not familiar with the lithic reduction-technique in question.

The present state of research on specific archaeological materials and issues of the Mesolithic in both the eastern Caucasus and the circum-Caspian area does not require any substantial revision of long-standing general assessments and descriptions of the cultural geography of the region in question at the turn of the Pleistocene-Holocene. Techno-typological analysis of specific materials allows the identification (in the eastern Caucasus, the southern Caspian Sea coast, and the eastern Caspian region) of a unity of related but at the same time independent archaeological cultures. These include such cultures as the Chokh in the northeastern Caucasus (Amirkhanov, 1987), possibly the Trialeti in the southern Caucasus (Gabunia, 1976), the Southern Caspian in the north and northeast of Iran (Jayez, Nasab, 2016), and the Balakhan in the eastern Caspian region (Korobkova, 1970). The community of these cultures has been previously substantiated, and the area of their distribution has been determined as the Southern Caspian cultural area (Amirkhanov, 1987).

Thus, the notion of “Trialetien”, introduced by Kozlowski at the end of the last century, but not very well established in the literature, seems redundant for the following reasons. Above all, a culture with this name (Trialeti) had previously been identified by another researcher on different grounds and in a different territory. Moreover, the introduction of a new concept adds nothing to the understanding of the cultural geography of the territories in question during the Mesolithic. The inclusion of the Trialeti archaeological culture in its traditional sense into the above-mentioned broad Southern Caspian cultural area can be regarded as a novelty. We can agree with this assumption, although Iranian researchers (Jayez, Nasab, 2016) deny the connection of the Mesolithic industry of the southern Caspian region (Komishan Сave) they study with the Trialeti culture. In our opinion, there are differences between the industries of these areas at the level of archaeological cultures, but this does not invalidate the similarities between the materials in question at the super-cultural level, i.e. at the level of a group of related cultures. This is what we had in mind when we proposed the notion of “Southern Caspian cultural area”. This view of the situation seems to correspond to the state of archaeological realities today.

Acknowledgements

This study was supported by the Russian Science Foundation, Project No. 22-28-00088.