The middle paleolithic prepared-core stone tools associated with neandertals are called:

1.7.8 Garchi I

The Mousterian artefacts from the Garchi I might be related to the Mezin paleosol complex. Though stone tools are found in abundance at the base of the exposure, their original position may be only inferred from the rare occurrence of flint or chert pieces in the paleosol (Pavlov and Makarov, 1998). Fig. 1.7.16 gives a comprehensive idea of the Kama third terrace, with the Late Pleistocene loess-soil formation overlying a glaciofluvial facies complex dated to the Dnieper glaciation. The Mousterian artefacts in Garchi I most probably correspond to the sandy loam horizon separating the Mikulino interglacial soils from those of the Early Valday interstadials (Brörup-Amersfoort). The soils overlying the Dnieper series form a part of a paleosol complex correlatable with the Mezin paleosol complex known from the central and southern regions of the East European Plain.

In the overlying series of the Valday loess there is a horizon of the Middle Valday (Bryansk) soil, the upper part of which contains the cultural layer of the Late Paleolithic site of Garchi I. The site age is determined by the radiocarbon date on charcoal at 28,750 ± 795, TUa-941/32,800 ± 830 cal. yr BP/and by a series of OSL dates on samples taken from soil and loess horizons (Svendsen et al., 2008). The Late Pleistocene loess-soil formations are noted for a complicated structure related to the multiple phases and cyclicity in the climate and landscape evolution during the early and middle Valday. Particularly distinct variations in the lithology of the uppermost horizons have been identified at the sites in the levels immediately underlying the habitable area.

The sequence exposed at Garchi I displays cryoturbations not only traceable in the Middle Pleistocene glaciofluvial silts and sands, but also disturbing the above-lying soil profiles (Fig. 1.7.17). A similar picture is observed in the scarp of the lowered third terrace near Zaozerye: faults of noticeable amplitude are distinctly seen in layers lowering towards a wide balka opening into the Kama valley. In common with the area of the site itself, the time of the deformations coincides with the permafrost degradation at the end of the Late Valday, when massive ground ice melted in local depressions.

Strange as it may seem, no signs of polygonal ice wedges has been recorded in close vicinities of Late Paleolithic sites and localities in spite of continuous permafrost present there in the Late Pleistocene (Velichko, 1973; Velichko et al., 1982, 1997Velichko et al., 1982Velichko et al., 1997; Butakov, 1986).

It seems possible that under conditions of increasing aridity towards the end of the late Valday glaciation the formation of polygonal ice wedge systems in the middle reaches of the Kama depended mostly on the local hydrological characteristics, so that large cryogenic wedge-like structures could develop both on lower geomorphic levels (Zaozerye) and on elevated surfaces (Garchi I) (Fig. 1.7.18). At the Zaozerye site there are cryoturbations distinctly seen in the system of faults on the side of the Chusovaya valley.

Noteworthy is the presence of several superimposed paleosols in the geological context of the Zaozerye cultural layer. That may indicate a prolonged period of low sedimentation rate (including the Mikulino Interglacial, Early and Middle Valday). As a result of that superposition, the cultural layers of the considered sites occur on the top of humified loams that are actually fragments of partly redeposited paleosols dated to the Mikulino Interglacial and to warmer phases within Early and Middle Valday.

The sedimentation rate increased considerably after the early man had left the region in the Late Valday. That acceleration resulted in accumulation of loess loam 5 m thick including two soil horizons, rather weakly pronounced. The horizons presumably correspond to intervals of a relative stability of the paleosurface.

Middle Palaeolithic

The study of the Levantine Mousterian benefited from the long stratigraphy of Tabun cave and from a host of other excavated sites. The Middle Palaeolithic in the Taurus–Zagros ranges differs in part from that of the Levant, but resembles the industries from the Caucasus region. The basis for the general subdivision of the Levantine industries is the sequence of Tabun Cave. The increase in the number of well-published lithic assemblages including the data pertaining to the variability within each of the Mousterian industries was the result of the advent of technological studies centered on recognizing the reduction strategies, expressed in both, the basic modes within the Levallois concept, as well as other lithic production concepts. Although the morphological variability of the lithic artifacts from various sites is not fully portrayed through the simplified model based on the Tabun Cave sequence, it seems that the overall chronological trend is clear, and that the general entities are valid archaeological ‘cultures’ in time and space, as follows:

‘Tabun D type’ assemblages are stratigraphically located at the base of the Mousterian sequence in multilayered sites such as Tabun Cave, unit IX, Yabrud I Rockshelter, Hayonim Cave layers F and lower E, and in Douara Cave layer IV, Abu Sif, and the open-air site of Rosh Ein Mor in the Negev highlands.

This entity is characterized by the production of elongated blanks sometimes defined as blades, pointed, and when retouched are known as Abu Sif points. The blanks were removed from cores with basal preparation and correction of the core convexity conveys the impression that the reduction had been bidirectional. In Hayonim Cave and Rosh Ein Mor, both the laminar and Levallois methods core reduction strategies were practiced. Other tool types are the burins that occur in this industry.

Most Early Mousterian assemblages in the southern Levant were produced by this particular Levallois method and could be called Abu Sifian. In the northern Levant, the Early Mousterian Hummalian industry from sites in the el-Kowm Basin (northeast Syria) is essentially a bladey industry, mostly non-Levallois. Further north, in the Caucasus Mountains, a similar bladey Mousterian is known from Djrujula Cave, and a few other sites such as Koudaro Cave, and is sometimes named Djrujulan. Currently, the dating for all these industries ranges from about 250 000 to c. 150 000 years, and are therefore of similar ages to the ‘Tabun D type’.

The ‘Tabun C-type’ assemblages are found in multilayered sites above the ‘Tabun D-type’ ones (Tabun I 18-26 or layer C in Garrod's excavations), Hayonim Cave upper layer E, and below the ‘Tabun B type’ in Tabun Cave. Most of the assemblages are characterized by the dominance of oval–rectangular blanks, best described from Qafzeh cave. The common products of this method are suboval and subquadrangular flakes, sometimes of large dimensions, struck from Levallois cores through centripetal and/or bidirectional exploitation. Triangular points appear in small numbers in definite horizons, such as layer XV in Qafzeh. Other generally similar assemblages were reported from Skhul layer B, Naamé, Ras el Kelb, Ksar ’Akil XXVI, and Quneitra where the lithic assemblage is also considered as having particular traits.

‘Tabun B type’ comprises assemblages dominated by the production of mainly flakes and triangular Levallois points, frequently with a broad-base – the typical ‘chapeau de gendarme’ – striking platform. They are removed from unidirectional convergent Levallois cores. There is a striking similarity between Kebara units IX and X, Tabun layer B both in Mount Carmel and Tor Faraj in southern Jordan. In Amud and Tor Sabiha Caves, unidirectional convergent cores yielded narrower and more elongated triangular flakes, sometimes called ‘leaf-shaped’ flakes. It is worth noting that blades do occur in all of these assemblages sometimes comprising up to 25% of the blanks such as in Kebara units XI and XII, and Amud B1. Similar ‘Tabun B-type’ assemblages occur in Bezez Cave layer B, Ksar Akil layer XXVIII (Lebanon), Dederiyeh (northern Syria), and Um el Tlel in el-Kowm Basin.

For historical reasons, it should be noted that the production of long and narrow flakes by unidirectional convergent mode of flaking, and the variable frequencies of blades in ‘Tabun B-type’ assemblages, were interpreted as suggesting that this industry could have been the technological progenitor of the bladey Initial Upper Palaeolithic (IUP). While this suggestion means that the technical change, or the first step of the Upper Palaeolithic Revolution, took place in the Levant, it also has biological implications. For scholars who classify the human fossils associated with the ‘Tabun B-type’ industries (Kebara, the woman from Tabun, Amud, Dederiyeh) as Western Asian Neanderthals, it means that the descendants of this population were those who initiated the new technological revolution of the Upper Palaeolithic. This stands in contradiction to the genetic evidence that finds correlation between modern humans and Upper Palaeolithic industries. Hence, the emergence of the new lithic reduction sequences and tool types should be sought in a different region such as the Nile Valley or East Africa.

The Middle Palaeolithic assemblages uncovered in the caves of the Zagros and Taurus resemble those of the southern Caucasus. The common knapping methods are discoiadal and Levallois and the frequencies of retouched pieces are very high. The dates from Karain Cave (Turkey), Shanidar (Iraq), and Ortvale Klde (Georgia) indicate that these assemblages, rich in well-retouched and resharpened points and sidescrapers, characterize mostly the Last Glacial Mousterian. Surface collections in Anatolia, north of the Taurus ranges, demonstrate the presence of Levallois-dominated industries.

One of the cultural characteristics of the Middle Palaeolithic of the region are the burials. The best known are from Skhul and Qafzeh cave, embedded in ‘Tabun C-type’ assemblages, and considered as archaic modern humans, possibly related to the finds in Omo-Kibish and Herto in Ethiopia. Late burials are those of the Neanderthals in Kebara, Tabun, Amud, Dederiyeh, and Shanidar caves (see Excavation and Recording Techniques).

Faunal analysis indicated that humans hunt what was available in the vicinity of their seasonal camps. In the Levant, the main game animals were the gazelle and the fallow deer, while in the Lebanese mountains the fallow deer and the wild goat were very common. In Shanidar the wild goat was the common hunt and in the Caucasus the thur (Capra Caucasica). Both the identified hunting seasons and differences in the number of retouched pieces versus cores and debitage products indicate that Middle Palaeolithic humans were mobile, employing various habitats, either along altitudinal transects (Zagros and Caucasus) or more along topographic lows (the Levantine landscapes). Meager evidence concerning the vegetal diet was provided by carbonized plants from Kebara Cave. In spite of the poor preservation, it seems that in all the Mediterranean areas plants were more important as a stable source of food than animal tissues.

Archaic Humans and Neanderthals in Western Asia

The study of the Levantine Mousterian benefited from the earlier discovery of the long stratigraphy of Tabun Cave (Mount Carmel). Dorothy Garrod, the original excavator, identified three major layers that more recently were adopted as terms for classifying the industries, from late to early, as ‘Tabun B-type’, ‘C-type’, and ‘D-type’. Assemblages, generally similar to those of Tabun Cave, were discovered in other sites in the Levant. In addition, the radiometric chronology (basically through thermoluminescence dating (TL), electron spin resonance (ESR), and 14C; (see CARBON-14 DATING; ELECTRON SPIN RESONANCE DATING; LUMINESCENCE DATING) facilitated the ordering of the three different industries, briefly described here.

The earliest, Tabun D-type, is characterized by the production of elongated blanks defined as blades and when retouched, known as Abu Zif points. The blanks were removed from cores with basal preparation that occasionally produced the impression that the reduction sequence was bidirectional. This type of industry was found all over the Levant, in the Caucasus, and beyond the Caspian Sea. Their Levantine and Caucasus dates are from 250 000 to 130 000 years ago, and probably 90 000 years ago further east. It is quite possible that their presence and dates mark the dispersal pattern of the archaic modern humans, the population that made them. Unfortunately, no fossils were found in the excavated sites.

The more recent Mousterian is the Tabun C-type, best described from Qafzeh Cave (near Nazareth, Israel). The common products of this method are suboval and subquadrangular flakes, sometimes of large dimensions, struck from Levallois cores through centripetal and/or bidirectional exploitation. Triangular points appear in small numbers. The industry dates to 140 000/130 000–90 000/85 000 years ago and contained human burials in Skhul and Qafzeh caves, as well as the use of seashells for body decorations, and plenty of red ocher.

The latest Mousterian is Tabun B-type, dominated by the production of mainly flakes and triangular Levallois points, frequently with a broad-base striking platform removed from unidirectional convergent cores. In several sites such as Tabun, Kebara, Amud, Dederiyeh, and Shanidar Caves, Neanderthal burials were exposed. It seems that the arrival of this new population in the Levant, and the Zagros Mountains, was triggered by the cold conditions across Europe during OIS 4. Additional evidence for the presence of Neanderthals was discovered in Sakajia cave, and is evidenced by their stone industries in many localities in the Caucasus mountains and foothills.

Hence, the Levant was occupied by a northern population, the Neanderthals. However, after about a span of 40 000/30 000 years, there is a wealth of evidence for a major change. New groups of people took over the region at about 50 000/45 000 years. This was the wave often referred to as the recent dispersal of modern humans, the bearers and initiators of the Upper Palaeolithic, a change that is also seen as a technological and cultural revolution.

Early Valdai = Zyrian (=Early Weichselian) Glaciation

Mammalian faunas of this age have been distinguished mostly on the basis of fossils from Mousterian Paleolithic sites. About 80 sites of this age from Northern Asia and Eastern Europe have yielded mammal assemblages. Because of the difficulties in dating most sites of this interval, we include here a discussion of faunas correlated with MIS 5 a–d, MIS 4, and the first part of MIS 3 (117–35 kyr BP). Seven mammal assemblages with several subcomplexes have been distinguished from this interval. The wide territories of Northern Asia and Eastern Europe have been occupied by the mammoth mammal assemblage, including four sub-assemblages: Arctic (I), European-Siberian boreal (IIa), North Caucasian boreal (IIb), and Altayan (III) (Fig. 4 and Tables 1 and 2) (Baryshnikov and Markova, 2002). The index mammal for all of these sub-assemblages was woolly mammoth (Fig. 5). The mammoth Transbaikalian assemblage (IV) also includes woolly rhinoceros, wild yak, and saiga antelope. Steppe animals (wild horse, saiga, primitive bison, steppe pika, marmot, ground squirrels, steppe and yellow lemmings, and others) dominated the Eurasian steppe assemblage (V). Reindeer and Pleistocene bison were also typical in this area. The desert assemblage included two variants: Caspian desert (VIa) and Central Asian (VIb) desert assemblages (Fig. 5 and Table 2). The mountain assemblages contained characteristic faunas, including a number of endemic species. The Crimean mammal assemblage (VII) (Fig. 5 and Table 2) included woolly mammoth, woolly rhinoceros, wild horse, Pleistocene ass, saiga, red, roe and giant deer, mountain sheep and goat, cave bear, cave hyena, steppe and yellow lemmings, and ‘obscurus’ vole. The Caucasian mammal assemblage (VIII) included wild boar, red deer, roe deer, elk, Caucasian goat, leopard, cave lion, and Prometheus' mouse. Species indicative of warm climate include Kudaro's cave bear, small porcupine, forest dormouse, pine vole, and others. These animals inhabited the southern slopes of the main Caucasus mountain system. Species typical of steppe-like landscapes lived in the Northern Caucasus region, including northern pika, wild sheep, and mountain ground squirrel. Several now-extinct species survived in the Caucasian Mountains, including a species of bear that is similar to Deninger's bear, Merck's rhinoceros, and a small porcupine. This indicates that the Caucasian Mountains were a biotic refuge during the Late Pleistocene (Fig. 5 and Table 2) (Baryshnikov and Markova, 2002). The Central Asian mammal assemblage (IX) includes Asiatic wild ass (Equus hemionus), red deer, Siberian roe deer, primitive bison, Siberian goat, snow leopard, and large porcupine (Fig. 5 and Table 2).

Table 1. Mammoth mammal assemblage. Early Valdai=Zyrian (=Weichselian) glaciation

Reproduced from Baryshnikov GF and Markova AK (2002) Animal world (mammal assemblages of the Late Valdai). Chapter 7. In: Velichko AA (ed.) Dynamics of Terrestrial Landscape Components and Inner Marine Basins of Northern Eurasia during the Last 130 000 years, pp. 123–137. Moscow: GEOS, with permission from GEOS.

Table 2. Early Valdai (early Weichselian) montane assemblages

Reproduced from Baryshnikov GF and Markova AK (2002) Animal world (mammal assemblages of the Late Valdai). Chapter 7. In: Velichko AA (ed.) Dynamics of Terrestrial Landscape Components and Inner Marine Basins of Northern Eurasia during the Last 130 000 years, pp. 123–137. Moscow: GEOS, with permission from GEOS.

The Early Valdai mammal fossils indicate the unique structure of mammalian assemblages of this time. Most of the faunal assemblages include species associated with a variety of habitats, but they are dominated by tundra and steppe species, now found in different ecosystems. The overlap of their ranges during the Pleistocene produced assemblages of mammal species for which there is no modern analog. Forest animals lived mostly in mountain regions, including endemic species in different mountain systems. The rarity of localities containing forest mammal remains indicates the absence of a continuous forest zone during this interval. This allowed Arctic animals to range farther south, and steppe mammals to range farther north and west. These nonanalog, ‘mixed’ faunas reflect the strong influence of the Scandinavian and Ural ice sheets and some Siberian smaller glacial caps over Eastern Europe and northwest Asia, in combination with decreased annual temperatures and precipitation, and a wide distribution periglacial landscapes. These unique mammal assemblages from the last glaciation have been described in many papers (Vereshchagin and Baryshnikov, 1980; Vangengeim, 1977; Smirnov, 1996; Sher, 1971; Markova et al., 1995, and others).

Transitions in Lithic Production after 300 ka

In western Asia (including the Caucasus), changes in lithic production resulted in the type of Mousterian industries seen in Europe, whereas changes were more muted in southern Asia, and resulted in a non-Mousterian Middle Paleolithic. In China and the Far East, changes were even more subtle, and stone tool assemblages continued without major industrial innovations. Outside the Levant, details are still sketchy as to when and how these late middle Pleistocene transitions occurred.

The Lower to Middle Paleolithic transition in the Levant occurred ca. 250 ka (Shea, 2003) (Table 1). The immediate precursor of the Levantine Mousterian is the Yabrudian industry, which dates from 350 ka, and is the earliest regional lithic variant in Asia. It is found in Syria, Lebanon, and northern Israel, and features steeply retouched scrapers and relatively thick flakes (Figure 1); the Levallois technique is absent, and bifaces are rare or absent. The principal Middle Paleolithic industry of the Levant is Mousterian. Its most conspicuous attribute is the use of recurrent Levallois core reduction strategies to produce triangular and subtriangular flakes. The earliest Mousterian lithic assemblages are distinctive, with high values for Levallois and laminar indices. The transition from hand-held handaxes in the Lower Paleolithic to the use of hafted tools (i.e., stone points or blades mounted on wooden shafts) in the Middle Paleolithic may be considered a profound technological shift, as later hominids employed stone projectile weaponry to obtain game (e.g., Shea (2003)). Early Mousterian industries in the Levant range from ca. 250 to 128 ka, thus including a transition from warm and humid conditions during oxygen isotope stage (OIS) 7 to colder and more arid conditions in OIS 6. It is not clear which hominin(s) made the earliest Levantine Middle Paleolithic tools because both Neanderthals and early modern humans in the Levant are associated with the same types of Middle Paleolithic assemblages.

Table 1. Periodization of the Levantine Middle Paleolithic (after Shea 2003: Table V)

The Lower to the Middle Paleolithic transition in South Asia is poorly dated, though there is good stratigraphic evidence to indicate that the transition from large bifacial industries of the Acheulean to prepared core and flake industries of the Middle Paleolithic occurred by ca. 150 ka (Misra, 2001; Petraglia et al., 2003) (Figure 2). Acheulean industries in India may have been made by Homo heidelbergensis, based upon the identification of a cranial fragment from the Narmada river valley (Cameron et al., 2004). The hominin(s) that made the stone tools associated with the Indian Middle Paleolithic remain unknown but likely to have been H. sapiens by at least 60 ka (Mellars, 2006), if not earlier (Petraglia et al., 2007; Clarkson et al., 2012).

The Early Middle Paleolithic in Central Asia is poorly dated, but those from the last glaciation are Mousterian, or have Mousterian affinities. Several caves in Uzbekistan, notably Obi Rakhmat (ca. 60–90 ka), Anghilak (ca. 36–45 ka), Khudji (ca. 40 ka), and Teschik-Tasch, show that Neanderthals were present in OIS 4–3 (Glantz et al., 2008). The best-dated Early Middle Paleolithic assemblages are from Tajikistan and Uzbekistan; these date from OISs 5 and 7, and show an increase in Levallois flaking. Their makers are unknown, but likely Neanderthal.

Although a Middle Paleolithic has been recognized in China on the basis of stone tools classifiable as scrapers, points, and choppers, these assemblages are so similar to earlier ones that it is doubtful whether it is useful to regard them as Middle Paleolithic rather than a later variant of the Early Paleolithic (Gao and Norton, 2002). Additionally, very few Chinese sites are sufficiently well-dated and -documented to allow detailed analyses of the changes that occurred between 300 and 50 ka. The best-documented late middle Pleistocene site in China is currently the cave of Panxian Dadong (ca. 100–250 ka) in southern China (Miller-Antonio et al., 2004); although lithic reduction sequences were complex, they were not dependent upon the type of prepared-core technologies so common in western Asia.

Middle Paleolithic or Middle Stone Age

As in southern Africa, the middle Paleolithic (or Middle Stone Age) of northern Africa probably began by 300–250 ka. It includes two major units, the Mousterian and the Aterian, which overlapped considerably in time. The Mousterian predominates in northeastern Africa and the Aterian in northwestern, but the Mousterian is known as far west as Morocco and the Aterian as far east as the Nile Valley. Within the Nile Valley, other variants have been defined, such as the Nubian complex and the Taramsan, but these are undated. In fact, the only long Middle Paleolithic sequence is from the Haua Fteah, Libya, which was excavated in the early 1950s and is therefore essentially undated.

In artifacts, the Mousterian is very similar to Charentian and Denticulate Mousterian as known in Europe and southwestern Asia, and has a much higher frequency of retouching than is found in southern Africa. Bifacial foliates occur in the Nubian Complex in the Nile Valley and the Red Sea Hills and also in the Aterian to the west. The Aterian is best known for the presence of tanged tools (not merely tanged points), which are generally taken to indicate hafting. Wurz et al. (2005) concluded that there is the same kind of variability in the MSA technology from Klasies as there is in the Middle Paleolithic of the Nile Valley. However, the latter remains undated and, at this level of technology, convergence is not surprising.

During a warm period about 100–90 ka, the northern African moderns moved briefly into the Levant, as did several other African animals. This colonization failed and the Africans returned south when the climate became colder. The successful and permanent occupation beyond Africa was closer to 50 ka, or later (Hublin, 2000).

There is very little information about when the middle Paleolithic ended. In the Nile Valley, Vermeersch (2000, table 14.1) estimates that the middle Paleolithic ended about 80–70 ka. The Sahara could be inhabited only during periods of increased rainfall, and the last wet period before the final Pleistocene lies far beyond the range of radiocarbon dating. Thus, the end of the middle Paleolithic could be dated only when techniques were developed that extend beyond 40 ka (luminescence, ESR, uranium-series and others). Age estimates for the last wet phase are available for at least seven regions of the Sahara. They range for a low estimate of ≥60 ka in the Tadrart Acacus in central Libya to high estimates of ≥80 ka in the southcentral Sahara and the region of the Great Eastern Erg. The middle Paleolithic in the Sahara must therefore all be older than 65–70 ka (Figure 6).

There are few reliable absolute dates for the end of the middle Paleolithic. The date of 45 ka at the Haua Fteah is minimal. At Jebel el Gharbi, western Libya, two sites have radiocarbon dates of about 44–43 ka (Garcea, 2004, pp. 33–34). Dating the end of the Middle Paleolithic is particularly contentious in the Maghreb, where the Mousterian sometimes underlies the Aterian. Recent critical compilations of dates for the Aterian (Cremaschi et al., 1998, table I; Bouzouggar et al., 2002, table 7; Close, 2009, table I) indicate that it may not have lasted much (if at all) beyond about 40 ka.

Personal gear in the Middle Palaeolithic

Recent studies on Middle Palaeolithic transfers highlight the potentially flexible nature and long life span of some of the items conveyed over distances >20 km. For example, in the Quina Mousterian, large side-scrapers and limaces are shown to be both tools and potential supplies of raw material in the form of long thin retouch or resharpening flakes, which can in turn be retouched into tools, sometimes testifying to several cycles of exploitation. The same seems to hold for bifaces in the Mousterian of Acheulean Tradition (MTA). In addition, it is demonstrated that in some assemblages cores were frequently worked from flakes. This suggests that large flake blanks were transported both as potential tools (to be retouched) and as raw material supplies (to produce flakes). Considering that predetermined products such as Levallois preferential and preparation flakes and reduced cores are also part of the mobile tool-kit, the overall technological pattern of transport is somewhat analogous to the one associated with low quantities in the Upper Palaeolithic (Figure 9), and points to a higher degree of technological planning than previously thought. Interestingly, in some sites, this pattern is documented for both exotic and local raw materials, giving new insights into site function and duration of occupations.

11.2 Middle Paleolithic

The sites and localities belonging to Stage I are dated to the interval from c.115 to 55 (up to 35) ka BP (Fig. 11.2). The epoch formerly known as Mousterian is now denoted by a broader term “Middle Paleolithic.” The term is applied to the industries with diversified flake tools and widely (though not universally) spread Levallois techniques. It spans a broader time interval than the Mousterian and includes some sites formerly attributed to Late and Final Acheulian. The industries considered to be Middle Paleolithic are diversified and include the Mousterian sensu stricto, Micoquian, Seclinian, Epi-Acheulian, etc.

Traces of human penetration onto the plains of temperate zones in extratropical Eurasia are known from Western Europe and the southern margins of Central and Eastern Europe, as well as from Siberia (Velichko, 1997; Derevyanko and Markin, 1992; Ivanova, 1965; Markov and Velichko, 1967; Praslov, 1969). There is no evidence of long-term settlements, and individual sites are separated in time by considerable intervals. That may be interpreted as a result of short-term occupation or individual tentative episodes of the human invasion (raids “there and back again”).

The data available on most of open sites of that time (including fossil soils, pollen and spore assemblages, and faunal assemblages) suggest the site functioning fell mostly on the interstadials of the first half of the Early Valday (Ivanova, 1977, 1982Ivanova, 1977Ivanova, 1982; Velichko, 1988). The majority of the sites occurs in middle latitudes on plains and forelands of Europe (Le Moustier, La Ferrassie, Neanderthal, Steinheim, Molodova 1, Korman’ 4, Stinka, Rozhok, etc.) and Northern Asia (Okladnikova, Denisova, Strashnaya, and other caves sites at Altay; localities on the Yenisey and Angara rivers; and those in the Transbaikal region). The landscape reconstructions based on pollen data suggest a combination of forest and steppe formations (Grichuk, 1982, 1989, 2002Grichuk, 1982Grichuk, 1989Grichuk, 2002).

Most of the sites of that stage are represented by disturbed fragments of cultural layers, so it is hardly possible to reconstruct their original structure. The most promising sites for the purposes of reconstructing environments of the Mousterian communities in the central Eastern Europe are Khotylevo I (the Bryansk region) and Sukhaya Mechetka (the Volgograd region), the latter being the only one with stone inventories in abundance, faunal remains, and hearth structures (Moskvitin, 1962; Grishchenko, 1965). Pollen assemblages suggest the prevalence of steppes with patches of forests (pine) in river valleys. The faunal remains are also dominated by steppe species. At Khotylevo I located farther north than Sukhaya Mechetka, the cultural layer contains stone tools and animal bones. As follows from the fauna composition, there were mostly open, and not forested, spaces in the region at the time of the settlement functioning. The site is tentatively dated at the Early Valday (Zavernyaev, 1972; Velichko, 1988).

There are also distinct signs of the Middle Paleolithic human penetration farther north in Eastern Europe at the beginning of the Early Valday. That is suggested by localities (not very numerous) in the middle reaches of the Kama River (Peshcherny Log, Garchi I). At that time the region presented an alternation of open treeless landscapes and open forests of spruce and birch. At Garchi I, stone implements of the Mousterian appearance have been collected from the beach at the base of the exposure. The original geological context of the findings and their initial position in the loess-soil sequence is still not clear (Pavlov and Makarov, 1998; Svendsen et al., 2008). Pavlov et al. (2004) identified the position of a few implements within a soil complex. The latter bears traces of cryogenic deformations probably related to the beginning of the Early Valday cooling.

Judging from the results of recent investigations, the range of the Middle Paleolithic localities in Siberia was close to that of the Early Paleolithic. It included the Altay, Kuznetsk Depression, valleys belonging to the drainage basins of the upper and partly middle Yenisey, Angara, the upper reaches of the Lena, and the Transbaikal region. Most of the Paleolithic localities are confined to rugged areas in mountains of South Siberia, in the upper reaches of large rivers. The region is noted for mosaic structure of landscapes, which apparently was favorable for a wide dispersal of the primeval human groups.

Important Terms

The Middle Palaeolithic

The Middle Palaeolithic archaeological record of Inner Asia is more detailed. It is part of a Eurasia-wide technological and cultural sequence. Three major divisions of stone tool technology have been identified at Middle Palaeolithic sites: Mousterian forms, Levallois type technologies, and incipient large blade industries with connections to Upper Palaeolithic technologies. These technologies form a chronological separation among sites, and also demonstrate the relationship of Inner Asia Middle Palaeolithic technologies with those of Western Eurasia where the same broad sequence has long been known. Further links to Western Eurasia were the discovery in 1939 of a Neanderthal sub-adult skeleton at the cave site of Teshik-Tash and the 2003 excavation of sub-adult teeth and cranial fragments from the Obi-Rakhmat rockshelter, which may also be related to Neanderthal populations. Both sites are located on the western slopes of the Tian Shan Mountains of Uzbekistan.

Some of the most important Middle Palaeolithic sites of Inner Asia are those of the Anui River valley in the Altai Mountains. The principal Middle Palaeolithic sequence is found at the sites of Denisova Cave and Okladnikov Cave with the former providing the bulk of the record. The Denisova cave habitation dates back possibly as early as 130 000 years ago and covers the extent of the Middle Palaeolithic. Also recovered at Denisova Cave were teeth from a hominid identified as Neanderthal-like. This evidence and the skeletal finds from Teshik-Tash and Obi-Rakhmat, lend credence to the argument that much of this region was inhabited by hominids akin to the Neanderthals of Europe or other archaic human species, though this question is still controversial. Other cave sites such as Ust' Kan, Kara-Bura, and Obi-Rakhmat, as well as open air scatters and flint quarries such as Kapchigai, located in the Altai and Tian Shan Mountains, testify to the effective adaptations of Middle Palaeolithic groups of the region.

Middle Palaeolithic artifacts are found on the eastern flanks of the Altai and Tian Shan Mountains as well, though no sites of the depth and scale of the western Altai sites are known. The recently excavated cave at Tsagaan Agui in central Mongolia and the Tolbaga site near Lake Baikal represent the Middle Palaeolithic technologies of central Inner Asia. The transition to new stone tool technologies of the Upper Palaeolithic occurred by 40 000 years ago. Many major sites, such as Tsagaan Agui, Obi-Rakhmat, and Okladnikov Cave, contain continuous sequences from Middle Palaeolithic to Upper Palaeolithic materials. There are also new transitional sites such as Kara-Bom, the earliest Upper Palaeolithic site yet discovered at 43 000 BP.