Kozintsev_多变量统计分析方法

Copyright © 2010, Siberian Branch of Russian Academy of Sciences, Institute of Archaeology and Ethnography of the Siberian Branch of the Russian Academy of Sciences. Published by Elsevier B.V . All rights reserved.doi:10.1016/j.aeae.2010.02.014Archaeology Ethnology & Anthropology of Eurasia 37/4 (2009) 125–136E-mail: Eurasia@archaeology.nsc.ruARCHAEOLOGY,ETHNOLOGY& ANTHROPOLOGY OF EURASIAIntroductionRoutes of the early Caucasoid migrations to Siberia and Eastern Central Asia have become a focus of scholarly interest in recent years since this issue is closely related to that of the Indo-European homelands. Certain archaeologists believe that migrants from the Near East played a major role in the origin of Southern SiberianCRANIOMETRIC EVIDENCE OF THE EARLY CAUCASOID MIGRATIONSTO SIBERIA AND EASTERN CENTRAL ASIA, WITH REFERENCETO THE INDO-EUROPEAN PROBLEM *Measurements of 220 male Neolithic and Bronze Age cranial series from Eurasia were subjected to multivariate statistical analysis. The results support the idea that people associated with the Catacomb culture played a major role in the origin of the Afanasyev culture. Okunev people of the Minusinsk Basin, those associated with Karakol, Ust-Tartas, and Krotovo cultures, and those buried in the Andronov-type cemeteries at Cherno-ozerye and Yelovka were of predominantly local Siberian origin. The Samus series resembles that from Poltavka burials. The Okunev people of Tuva and probably Yelunino people were likely descendants of the Pit Grave (Yamnaya) and early Catacomb populations of the Ukraine. The same is true of the Alakul people of western Kazakhstan, who in addition, have numerous af ¿ nities amongst Neolithic and Early Bronze Age groups of Central and Western Europe. The probable ancestors of certain Fedorov populations were the Afanasyev tribes of the Altai, whereas other Fedorov groups apparently descended from late Pit Grave and Catacomb tribes of the Northern Caucasus and the northwestern Caspian. People of Gumugou are closest to Fedorov groups of northeastern Kazakhstan and Rudny Altai, suggesting that Caucasoids migrated to Xinjiang from the north rather than from the west. Describing the gracile Caucasoids of Siberia and Eastern Central Asia as “Mediterraneans” is misleading since they display virtually no craniometric ties with the Near Eastern, Southwestern Central Asian or Transcaucasian groups. The totality of evidence suggests that they were Nordics.Keywords: Indo-Europeans, Indo-Iranians, Tocharians, Southern Siberia, Western Siberia, Central Asia, Bronze Age, craniometry.cultures of the Bronze Age (Grigoryev, 1999; Bobrov, 1994; Kiryushin, 2004), and these theories are supported by those physical anthropologists who claim that all gracile Caucasoids are Mediterraneans, i.e. southerners by origin (see especially (Khudaverdyan, 2009)). Not long ago I expressed a similar view (Kozintsev, 2000).Recently, thanks to the work of a number of craniologists, S.I. Kruts in particular, the craniometric database related to the Bronze Age steppe populations of the Ukraine and Southern Russia has grown manifold. Its statistical analysis has led to the revision of earlier*Supported by the Russian Foundation for Basic Research (Project 09-06-00184a). A.G. KozintsevMuseum of Anthropology and Ethnography, Russian Academy of Sciences,Universitetskaya Nab.3, St. Petersburg, 199034, RussiaE-mail: agkozintsev@125ANTHROPOLOGY126 A.G. Kozintsev / Archaeology Ethnology & Anthropology of Eurasia 37/4 (2009) 125–136views. A more detailed craniometric comparison of each gracile Southern Siberian group with all others suggests that there is no reason to speak of migrations to Southern Siberia from the Near East, Southwestern Central Asia or the Transcaucasia, where Southern Caucasoids (Mediterraneans) were distributed (Kozintsev, 2007, 2008).Recently, an article by a group of French geneticists was published (Keyser et al., 2009), which reported on the analysis of DNA extracted from the bone samples taken from Andronov, Karasuk, Tagar, and Tashtyk human remains. Six genes controlling eye and hair color were studied. Most individuals buried in Bronze and Iron Age mounds in Southern Siberia (15 of 23, or 65 %) had light or mixed eye color, and 8 out of 12 (67 %) had fair or chestnut hair. Given that the Bronze Age people of the Tarim Basin (the likely proto-Tocharians), whose bodies are excellently preserved thanks to natural mummi¿ cation (Mallory, Mair, 2000), had the same hair color, and that a Russian admixture alone can by no means account for the depigmentation observed in modern natives of Southern Siberia and Kazakhstan, the conclusion is obvious. The principal source of early Caucasoid migrations to Siberia and Eastern Central Asia was located not in the Near East, but in Europe, moreover not in its southern part but in areas affected by the depigmentation process. Nearly eighty years ago this conclusion was reached by G.F. Debetz (1931), who compared cranial data on the Tagar people with the evidence of Chinese written sources. Apparently, the principal migration route of Caucasoid pastoralists from Europe to the east passed mainly along the steppe belt, and judging from archaeological data, the migration process continued throughout the 3rd millennium BC (Merpert, 1982; Semenov, 1993). But where was the source located? In the Eastern European steppes? In Central Europe?According to a view shared by most specialists, archaeologists and physical anthropologists alike, the Afanasyev culture was closely related to the Pit Grave (Yamnaya) culture, and its appearance in Gorny Altai and on the Middle Yenisei was caused by a migration from the Eastern European steppes. The possible role of Poltavka and Catacomb culture elements, too, has been discussed (Tsyb, 1981, 1984). The idea is supported by new radiocarbon dates indicating the coexistence of Catacomb culture with Pit Grave culture over most of the 3rd millennium BC (Chernykh, 2008). On the other hand, very early dates of the earliest Afanasyev sites in Gorny Altai (mid-4th millennium BC) suggest that the predecessors of the Pit Grave people, speci¿ cally those associated with the Khvalynsk and Sredni Stog cultures, as well as the proto-Pit Grave (Repino) tribes, might have taken part in Afanasyev origins. This suggestion was already made by physical anthropologists (Shevchenko, 1986; Solodovnikov, 2003).With regard to the post-Afanasyev Bronze Age cultures, the traditional idea that the Okunev culture is autochthonous has given place to theories stating that the Pit Grave and Catacomb traditions (Lazaretov, 1997), or those of Afanasyev culture, which were also introduced from Europe, were critical in Okunev origins (Sher, 2006). In terms of physical anthropology however, the presumed European ancestry of the Okunev people of the Minusinsk Basin, according to A.V. Gromov (1997b), pointing to af¿ nities with the Pit Grave and Catacomb people of Kalmykia, is rather indistinct and traceable mostly at the individual level if at all. The analysis of data concerning two independent trait batteries – craniometric and cranial nonmetric – suggests that the af¿ nities of the Okunev people of the Yenisei are mostly Siberian (Gromov, 1997a, b), and the integration of these data demonstrates that the unusual trait combination observed in Okunev crania is rather archaic (plesiomorphic) and may be more ancient than both the Caucasoid and Mongoloid trait combinations (Kozintsev, 2004). According to Gromov (1997b), the Okunev people resembled the Neolithic population of the Krasnoyarsk–Kansk region. The Karakol culture of Gorny Altai is similar to Okunev culture, and craniometric parallels between people associated with these cultures were also noted. However, Karakol crania are believed to exhibit a “Mediterranean” tendency (Chikisheva, 2000; Tur, Solodovnikov, 2005).The Okunev crania from Tuva and the Yelunino crania from the Upper Ob, especially the former, are much more Caucasoid (Gokhman, 1980; Solodovnikov, Tur, 2003; Kozintsev, 2008). This agrees with archaeological facts indicating the affinities of cultures such as Yelunino and Okunev of Tuva with Early Bronze Age cultures of Central and even Western Europe (Kovalev, 2007). The possible Caucasoid ties of other pre-Andronov tribes of Southern Siberia such as Krotovo (Dremov, 1997) and Samus (Solodovnikov, 2005, 2006) have been discussed by craniologists. K.N. Solodovnikov (Ibid.) believes that in all the above pre-Andronov groups, except the Okunev group of the Yenisei, these ties are Southern Caucasoid or Mediterranean which, in his view, is especially evident in the male series.The origin of the Andronov community is one of the pivotal points in Indo-European history. The predominantly Indo-Iranian or Iranian attribution of this community is beyond doubt (Kuzmina, 2007a, b; 2008). The relationship between its two constituents, speci¿ cally the Alakul (western) and Fedorov, which spread in an eastern direction up to the Yenisei, is less clear. The Alakul variety apparently originated earlier, in the 3rd millennium BC (Chernykh, 2008) and the cultures which contributed to its origin were Poltavka, Catacomb, and Abashevo. The origin of the Fedorov variety, which originated later and coexisted with Alakul over most of the 2nd millennium BC, remains obscure (Tkacheva, Tkachev, 2008).A.G. Kozintsev / Archaeology Ethnology & Anthropology of Eurasia 37/4 (2009) 125–136 127Craniologists have discovered that the Andronov community was markedly heterogeneous. People buried in graves with Alakul or mixed Alakul-Fedorov (Kozhumberdy) ceramics in western Kazakhstan displayed a trait combination which V.V. Ginzburg (1962) described as Mediterranean, and V.P. Alekseyev (1964) as leptomorphic. Ginzburg believed that this combination evidences the af¿ nities of western Alakul people with both the Timber Grave (Srubnaya) populations of the V olga steppes and those of Southwestern Central Asia (the Amu-Darya/Syr-Darya interÀ uve). The second idea was refuted by Alekseyev, who claimed that archaeological data point solely to western (Timber Grave) af¿ nities. Ginzburg ignored the critique and repeated his conclusion in the summarizing monograph (Ginzburg, Tro¿ mova, 1972). In this case, neither he nor Alekseyev used statistical methods and relied on typological assessments.As to the Fedorov populations, most of which display the characteristically “Andronov” trait combination believed to have derived from the Cromagnon variety, G.F. Debetz (1948) claimed that they had originated in the Kazakhstan steppes from whence they moved to the Yenisei. However, V.P. Alekseyev (1961) suggested that the Fedorov people of the Yenisei had descended from the Afanasyev populations of the Altai. Fedorov groups of the Upper Ob and the Altai deviate toward a gracile variety, traditionally described as Mediterranean. The presence of alleged “Mediterraneans” in these regions was explained differently: V.A. Dremov (1997) attributed it to links with the Alakul people, whereas K.N. Solodovnikov (2005, 2007) wrote about the pre-Andronov, speci¿ cally Yelunino substratum. People buried in Andronovo-type cemeteries in the Tomsk part of the Ob Basin at Yelovka II and in the Omsk stretch of the Irtysh basin at Cherno-ozerye I, according to Dremov (1997), differed from other Andronov groups and were autochthonous. Finally, the origin of the early Caucasoid population of Xinjiang, the members of which were buried at the Bronze Age cemetery of Gumugou (Mair, 2005), is a complete mystery (Han Kangxin, 1986; Hemphill, Mallory, 2004; Kozintsev, 2008).The objective of the present article is to explore the issue of early Caucasoid migrations to Siberia and Eastern Central Asia using a large craniometric database, much of which is unpublished.Materials and methodsOnly measurements of male crania were used. The number of the Afanasyev series is nine (six from the Altai and three from the Minusinsk Basin) (Alekseyev, 1961, 1989; Solodovnikov, 2003). The post-Afanasyev material consists of four Okunev series from the Minusinsk Basin (Gromov, 1997b), one Okunev series from Tuva (Alekseyev, Gokhman, Tumen, 1987), the Karakol series (Chikisheva, 2000; Tur, Solodovnikov, 2005), the Yelunino series (Solodovnikov, Tur, 2003), the Ust-Tartas and Krotovo series from Sopka-2 (Dremov, 1997; Chikisheva, unpublished), and the Samus series (Dremov, 1997; Solodovnikov, 2005). Seven Andronov samples were used. Two of them, from western Kazakhstan (Alekseyev, 1967) and from Yermak IV near Omsk (Dremov, 1997) represent mostly the Alakul variety. Fedorov samples come from Firsovo XIV near Barnaul (Solodovnikov, 2005), from other burial grounds on the Upper Ob (Ibid.) and from northeastern Kazakhstan (Ibid.)*, from Rudny Altai (Solodovnikov, 2007), and from the Minusinsk Basin (Alekseyev, 1961; Dremov, 1997). Also, measurements of two series from “Andronoid” burial grounds in Western Siberia at Yelovka II and Cherno-ozerye I (Dremov, 1997) were used, and also those of a Bronze Age series from Gumugou (Qäwrighul), Xinjiang (Han Kangxin, 1986).Unpublished measurements of Bronze Age crania from the Ukraine were kindly provided by S.I. Kruts; sources of data on most groups published by Russian scholars are cited in my previous publications (Kozintsev, 2000, 2007, 2008). Measurements of series from Central and Western Europe and the Near East were taken from a summary compiled by I. Schwidetzky and F. Rösing (1990).The total number of male cranial series representing the Neolithic and Bronze Age populations of Eurasia and used in this analysis is 220. One hundred and twenty-eight of them, mostly from the former Soviet Union, were studied according to the craniometric program employed by Soviet and modern Russian anthropologists. Fourteen traits were taken from it: cranial length, breadth, and height, frontal breadth, bizygomatic breadth, upper facial height which was measured to the “lower prosthion”, or alveolare, nasal and orbital height and breadth, naso-malar and zygo-maxillary angles, simotic index and nasal protrusion angle. Ninety-two series from Central and Western Europe and the Near East were studied by Western anthropologists. They were measured according to a smaller program from which nine linear dimensions were taken: cranial length, breadth and height, frontal breadth, bizygomatic breadth, and nasal and orbital height and breadth. Upper facial height was not used in this case to avoid confusion between the “anterior” and the “inferior” prosthion (the difference may be considerable).Measurements were subjected to the canonical variate analysis. Groups were compared pairwise using *Solodovnikov has reduced the size of this series by demonstrating on archaeological grounds that some crania previously believed to be Andronov actually represent other populations.128 A.G. Kozintsev / Archaeology Ethnology & Anthropology of Eurasia 37/4 (2009) 125–136the Mahalanobis D2 distance corrected for sample size (Rightmire, 1969). After correction, many distances become negative and should be regarded as sample estimates of zero or of small positive values.Creating a general classi¿ cation of all groups was not among the objectives of the present study. All methods aimed at such classification result in distortions. In cluster analysis, the distortions become progressively larger as the distances increase. By contrast, in two-dimensional projections determined by canonical variates or by nonmetric scaling axes, it is the closest ties that are distorted most in order to adequately render the most general pattern of group relationships. The broader the scope of the study, and accordingly, the wider the geographical range of the analysis, the more details are sacri¿ ced for the sake of the general classi¿ cation. These distortions may have contributed to the idea that all gracile Caucasoids are close relatives. While this may be true in a bird’s-eye view, a disregard for details in such a case may lead to a serious misinterpretation.Another advantage of using pairwise distances, rather than graphic methods of dimensionality reduction, stems from the fact that the latter are sensitive to the selection of groups. Thus, the sequence of groups in terms of the expression of Caucasoid versus Mongoloid traits depends on the way these extreme combinations are represented. Distances, by contrast are independent of this factor provided the standard within-group correlation matrix is used, as in this study.ResultsListed below are the smallest corrected D2 values (normally below 1.0), based on the fourteen-trait battery and ranked in an increasing order i.e. in the order of decreasing similarity. Minimal distances based on the nine-trait battery (D2 < 0.3) are given only for those groups which reveal at least one Central or Western European or Near Eastern parallel; these parallels are numbered as in the summary (Schwidetzky, Rösing, 1990). The only exception is the Afanasyev series from Saldyar I. It is quite small and reveals numerous af¿ nities, including early Central and Western European ones, which are not in ¿ rst place.Afanasyev, Ursul, Altai: Afanasyev, Nizhni Tyumechin (–1.12); Catacomb, Don (0.59); Timber Grave, Luzanovka, V olga (0.66); early Northern Caucasian culture, Kalmykia, group II according to V.A. Safronov (0.70).Afanasyev, Saldyar I, Altai: Pit Grave, V olga–Ural area (–2.49); Pit Grave, Orenburg region (–2.47); Pit Grave – Poltavka group,Volga–Ural area (–2.42); late Catacomb, Verkhne-Tarasovka, Lower Dnieper (–2.35); Afanasyev, Afanasyeva Gora (–2.35) and Karasuk III (–2.30); early Northern Caucasian culture, Kalmykia, group II (–2.10); Afanasyev, Minusinsk Basin, pooled (–1.87); Andronov, Firsovo XIV, Upper Ob (–1.82); Pit Grave, Yuzhny Bug (–1.79); Pit Grave, Ingulets (–1.51); early Catacomb, Molochnaya (–1.36); Timber Grave, Luzanovka, V olga (–1.34); Catacomb, V olga and Kalmykia (–1.23 in both cases). Numerous remaining parallels are mostly with Catacomb and Timber Grave groups.Afanasyev, Kurota II, Altai: Poltavka (–1.38); Afanasyev, Saldyar I (0.59); Catacomb, V olga (0.96).Afanasyev, Ust-Kuyum, Altai: late Catacomb, Samara–Orel watershed (–0.42); Pit Grave, Ingulets (–0.20); Pit Grave, Stavropol area (–0.07); Andronov, Minusinsk Basin (0.51); late Catacomb, Zaporozhye (0.80).Afanasyev, southeastern Altai: Afanasyev, Nizhni Tyumechin (–0.38); Catacomb, Don (–0.18); early Catacomb, Molochnaya (–0.09); Pit Grave, Kakhovka, Lower Dnieper (–0.05); Timber Grave, Yasyrev, Lower Don (0.16); Afanasyev, Saldyar I (0.22); Pit Grave, Yuzhny Bug (0.23); Timber Grave, Luzanovka, Volga (0.24); Pit Grave, Ukraine, pooled (0.47); early Catacomb, Verkhne-Tarasovka, Lower Dnieper (0.75); Timber Grave, Krivaya Luka, Lower V olga (0.81); Timber Grave, V olga, pooled (0.86); Timber Grave, V olgograd and Astrakhan regions (0.89); Pit Grave – Poltavka, Volga–Ural area (0.91); Catacomb, Kalmykia (0.95).Afanasyev, Nizhni Tyumechin, Altai: Afanasyev, Ursul (–1.12); Catacomb, Don (–0.86); early Catacomb, Verkhne-Tarasovka, Lower Dnieper (–0.76); Timber Grave, Yasyrev, Lower Don (–0.52); Afanasyev, southeastern Altai (–0.38); Timber Grave, Krivaya Luka, Lower V olga (–0.25); Timber Grave, Luzanovka, V olga (–0.07); Pit Grave – Poltavka, V olga–Ural area (0.27); Afanasyev, Saldyar I (0.56); Pit Grave, Ingulets (0.77); Pit Grave, Ukraine, pooled (0.83).Afanasyev, Altai, pooled: Timber Grave, Luzanovka, Volga (0.23); Catacomb, Don (0.29); Timber Grave, Bashkiria (0.79); Timber Grave, Krivaya Luka, Lower V olga (0.96).Afanasyev, Karasuk III, Minusinsk Basin: Afanasyev, Saldyar I (–2.30); Pit Grave, Volga–Ural area (–0.96); Afanasyev, Afanasyeva Gora (–0.54); Timber Grave, Volga–Ural area (–0.32); late Catacomb, Verkhne-Tarasovka, Lower Dnieper (–0.25); Catacomb, Volga (–0.22); Pit Grave, Orenburg region (–0.13); Catacomb, Crimea (0.01); Pit Grave – Poltavka, Volga–Ural area (0.24); Potapovka, V olga (0.73); Timber Grave, Yasyrev, Lower Don (0.90).Afanasyev, Afanasyeva Gora, Minusinsk Basin: Afanasyev, Saldyar I (–2.35); late Catacomb, Verkhne-Tarasovka, Lower Dnieper (–1.60); Timber Grave, Volgograd and Astrakhan regions (–0.66); Afanasyev, Karasuk III (–0.54); early Catacomb, Molochnaya (–0.46); Timber Grave, Yasyrev, Lower Don (–0.13); Pit Grave, Southern Bug (0.21); early Catacomb, Kakhovka,A.G. Kozintsev / Archaeology Ethnology & Anthropology of Eurasia 37/4 (2009) 125–136 129Lower Dnieper (0.44); Pit Grave, V olga–Ural area (0.85); Timber Grave, V olga (0.88); Pit Grave, Orenburg region (0.89); Timber Grave, Krivaya Luka, Lower Volga (0.96). The analysis based on the nine-trait set revealed one Western European parallel – with the Early Bronze Age (3rd millennium BC) group from Aveyron, France, (0.04), but ties with the steppe populations of the Eastern European Bronze Age are stronger.Afanasyev, Minusinsk Basin, pooled: Afanasyev, Saldyar I (–1.87); Timber Grave, Yasyrev, Lower Don (–0.71); Timber Grave, V olgograd and Astrakhan regions (–0.12); late Catacomb, Verkhne-Tarasovka, Lower Dnieper (–0.06); Pit Grave, Orenburg region (0.22); early Catacomb, the Molochnaya (0.24); Timber Grave, Luzanovka, Volga (0.64); Pit Grave, Volga–Ural area (0.65); Catacomb, Kalmykia (0.81); Pit Grave, Yuzhny Bug (0.82); Timber Grave, Volga (0.85); Pit Grave – Poltavka, V olga–Ural area (0.93); Timber Grave, Krivaya Luka, Lower V olga (0.95); Abashevo (0.99).Okunev, Uybat group, Minusinsk Basin: the only close parallel is with Okunev of the Tas-Khaza group (–0.95). Okunev of Chernovaya (1.44), and Karasuk (1.90) rank next.Okunev, Verkhni Askiz, Minusinsk Basin: Neolithic of Krasnoyarsk–Kansk area (–0.07).Okunev, Chernovaya, Minusinsk Basin: the only close parallel is with the Neolithic of Krasnoyarsk–Kansk area (0.36). The least removed among other populations is Okunev, Uybat group (1.44).Okunev, Tas-Khaza group, Minusinsk Basin: the only close parallel is with Okunev of Uybat (–0.95). The least distant among other groups is Karasuk (1.77).Okunev, Minusinsk Basin, pooled:the only close parallel is with the Neolithic of Krasnoyarsk–Kansk area (0.15). The least removed among other groups is Karasuk (3.37).Okunev, Aimyrlyg, Tuva: Pit Grave, Ingulets (–0.21); Timber Grave, Saratov region (–0.10); early Catacomb, Molochnaya (0.41); Timber Grave, Ukraine, pooled (0.45); Sapallitepe, southern Uzbukistan (0.67). Nine-trait set: early Catacomb, Molochnaya (–1.21); Tiefstichkeramik (related to Funnel Beaker), Ostorf, Germany, Late Neolithic, late 4th millennium BC (No.106) (–1.15); Afanasyev, Afanasyeva Gora (–0.76); Pit Grave, Ingulets (–0.53); late Catacomb, Verkhne-Tarasovka, Lower Dnieper (–0.39); Afanasyev, Minusinsk Basin (–0.37); Timber Grave, Volgograd and Astrakhan regions (–0.26); Abashevo (–0.26); Sapallitepe, Southern Uzbekistan (–0.02); Yelunino (0.01); Afanasyev, Saldyar I (0.24); Timber Grave, Khryaschevka, Volga (0.29).Karakol: the only parallel is with the Neolithic and Chalcolithic of Ust-Isha and Itkul, Upper Ob (0.98). The least removed among other groups is Yelovka II (3.87), whereas Okunev is much further (7.26).Yelunino:not a single close parallel. The least removed is Okunev of Tuva (1.56), next follow Djarat and Shengavit, Kura-Araxes culture, Armenia, 4th–3rd millennia BC (2.16); the pooled series of Kura-Araxes culture from Georgia ranks third (2.65), and Gumugou, Xinjiang, fourth (3.83). Nine-trait set: Poltavka (–0.13); Okunev of Tuva (0.01); early Catacomb, Molochnaya (0.01); Timber Grave, forest-steppe Volga area (0.22); Mierzanowice, Poland, Early Bronze Age, late 3rd – early 2nd millennia BC (No.173) (0.28).Ust-Tartas, Sopka-2: Krotovo, Sopka-2 (0.72).Krotovo, Sopka-2: Ust-Tartas, Sopka-2 (0.72).Samus:not a single close parallel. The least removed is Poltavka (1.18).Alakul, western Kazakhstan: early Catacomb, Molochnaya (–1.35); Pit Grave, Ingulets (–0.36); early Catacomb, Verkhne-Tarasovka, Lower Dnieper (0.44); late Timber Grave, V olga–Ural area (0.54); Kemi-Oba, Crimea (0.88). Nine-trait set: early Catacomb, Molochnaya (–1.39); Pit Grave, Ingulets (–0.88); Timber Grave, ground burials, Ukraine (–0.79), Pit Grave, Kakhovka, Lower Dnieper (–0.67); Parkhay II, Turkmenia, Middle and Late Bronze Age (–0.61); Tiszapolgar, Hungary, Chalcolithic, 5th–4th millennia BC (No.197) (–0.38); late Timber Grave, V olga–Ural area (–0.16); Rössen, eastern France, Neolithic, 5th millennium BC (No.43) (–0.09); Globular Amphorae, Germany and Poland, Early Bronze Age (early 3rd millennium BC) (No.192)(–0.07); Timber Grave, Ukraine, pooled (–0.03); Lengyel, Hungary, Neolithic, 5th millennium BC (No.40) (0.07); Meklenburg, Germany, Early Bronze Age, 4th–3rd millennia BC (No.107) (0.07); Aveyron, France, Early Bronze Age, 3rd century BC (No.99) (0.09); Unetice, Germany and Czechia, Bronze Age, 3rd–2nd millennia BC (No.208) (0.09); Linear Band Pottery, Neolithic, 6th millennium BC (No.14) (0.11); Pit Grave, Yuzhny Bug (0.20); Veterov, Austria, Bronze Age, III–II millennia BC (No.205) (0.21).Alakul, Yermak IV, the Irtysh: not a single close parallel. The least removed is the late Pit Grave group of Kalmykia* (1.32).Fedorov, Firsovo XIV, the Upper Ob: Afanasyev, Saldyar I (–1.82); Fedorov, Rudny Altai (–0.04); Catacomb, Kalmykia (0.06); Pit Grave – Poltavka, V olga–Ural area (0.43); Timber Grave, Luzanovka, V olga (0.87); Pit Grave, Orenburg region (0.90); early Catacomb, Molochnaya (0.94).Fedorov, the Upper Ob, pooled: Catacomb, Stavropol area (0.50); late Pit Grave, Kalmykia (0.80); Pit Grave, Stavropol area (0.90).Fedorov, Rudny Altai:Samus (–0.82); Afanasyev, Saldyar I (–0.71); Timber Grave, Luzanovka, Volga (–0.12); Fedorov, Firsovo XIV (–0.04); Pit Grave – *Group III according to V.A. Safronov (Shevchenko, 1986).130 A.G. Kozintsev / Archaeology Ethnology & Anthropology of Eurasia 37/4 (2009) 125–136Poltavka, Volga–Ural area (–0.02); early Northern Caucasian culture, Kalmykia, group II (0.61); Potapovka, V olga (0.63); Fedorov, northeastern Kazakhstan (0.79); late Catacomb, Kakhovka, Lower Dnieper (0.82), Catacomb, Don (0.83); Pit Grave, Orenburg region (0.88); Poltavka (0.99).Fedorov, northeastern Kazakhstan: late Pit Grave, Kalmykia (–1.50); late Catacomb, Yuzhny Bug (–1.44); Fedorov, Minusinsk Basin (–0.67); early Northern Caucasian culture, Kalmykia, group II (–0.59); late Catacomb, Kakhovka, Lower Dnieper (–0.47); Potapovka, V olga (–0.13); Catacomb, V olga (0.08); late Catacomb, Krivoy Rog, Upper Ingulets (0.16); late Catacomb, Ukraine, pooled (0.17); Pit Grave, Kalmykia (0.47); Pit Grave, Stavropol area (0.51); late Catacomb, Samara–Orel watershed (0.54); Timber Grave, Luzanovka, V olga (0.57); late Northern Caucasian culture, Kalmykia, group IV (0.72); late Catacomb, Zaporozhye, Lower Dnieper (0.74); Fedorov, Rudny Altai (0.79); Khvalynsk, V olga–Ural area (0.88).Fedorov, Minusinsk Basin: Fedorov, northeastern Kazakhstan (–0.67); late Northern Caucasian culture, Kalmykia, group IV (–0.09); Pit Grave, Stavropol area (–0.04); late Pit Grave, Kalmykia (–0.03); late Catacomb, Krivoy Rog, Upper Ingulets (0.31); late Catacomb, Samara/Orel interÀ uve (0.39); Afanasyev, Ust-Kuyum (0.51); late Catacomb, Kakhovka, Lower Dnieper (0.57); early Northern Caucasian culture, Kalmykia, group II (0.69); Timber Grave, Luzanovka, Volga (0.72); late Catacomb, Crimea (0.83); late Catacomb, Ukraine, pooled (0.84).“Andronoid”, Cherno-ozerye I, Omsk area: late Krotovo, Sopka-2 (0.91).“Andronoid”, Yelovka II, Tomsk area: not a single close parallel. The least distant is Irmen (1.01); those ranking next are late Krotovo, Sopka-2 (1.49), Cherno-ozerye I (3.42), and Karakol (3.87).Gumugou, Xinjiang: not a single close parallel. The least distant are two Fedorov series – from Rudny Altai (1.26) and from northeastern Kazakhstan (1.28).DiscussionAfanasyevThe results challenge the traditional idea that the sole and direct ancestors of the Afanasyev people were those of Pit Grave culture. Pit Grave af¿ nities rank ¿ rst only in the cases of Saldyar I and Karasuk III. Catacomb parallels are no fewer than those with Pit Grave, and in most instances they are the most pronounced. Every Afanasyev group has close ties with Catacomb groups. By contrast, not all Afanasyev series show close Pit Grave connections: these are absent in two groups of the Altai (Ursul and Kurota II) and in the pooled Altai sample. In half of the Altai series, ties with the Catacomb people of the Don are the most distinct, and the same is true of the pooled Altai group. Afanasyeva Gora and the pooled Minusinsk series are closest to the late Catacomb of the Lower Dnieper, whereas the series from Kurota II in the Altai, is closest to Poltavka. These results are matched by archaeological facts which, according to S.V. Tsyb (1981, 1984), evidence the importance of Poltavka and Catacomb cultures in Afanasyev origins.Strangely, similarities with Timber Grave people are no less numerous. In fact, for the pooled Minusinsk group they are more distinct than those with Catacomb and Pit Grave. The Timber Grave tribes could not have played any role in Afanasyev origins because they lived later; nor do any facts point to a reverse migration of the Afanasyev people or their descendants to Europe. The results can hardly be attributed to a slightly uneven representation of the three Eastern European cultures in the database, where the Pit Grave is represented by 15 series, the Catacomb by 18, and the Timber Grave by 16. More likely, these results testify to the considerable stability and relative homogeneity of the physical type of the Eastern European steppe populations over the Bronze Age despite the succession of cultures and apparently despite microevolutionary trends such as gracilization.Attempts at tracing the origins of this type have so far been unsuccessful. On the one hand, the Pit Grave people of the Lower Dnieper (Kakhovka and Kherson areas), the Catacomb people of the same region (Verkhne-Tarasovka, early group) and those of Kalmykia are similar to the Chalcolithic Khvalynsk population (5th–4th millennia BC). Accordingly, Khvalynsk might have been ancestral for some Eurasian steppe populations of the Bronze Age. Another Chalcolithic series which represents the Sredni Stog culture is more isolated, the least removed from it being various Afanasyev groups of the Altai and the Catacomb people of the Don. All these facts may point to the deep Eastern European roots of the Pit Grave, Catacomb, and partly Afanasyev communities.On the other hand, not all the Eastern European steppe populations of the Bronze Age appear to have been autochthonous. The analysis of a larger number of groups using the reduced trait battery reveals numerous early (4th millennium BC and earlier) Central and Western European parallels for groups such as the Pit Grave from the Ingulets and early Catacomb from the Molochnaya. These ties are especially evident in four gracile early Catacomb groups of the Ukraine, which show 14 close ties with Central and Western European populations and eight with those of Transcaucasia and Southwestern Central Asia. This apparently attests to migration, since the late Catacomb people are more robust, contrary to the normal diachronic trend (Kruts, 1990) and show no such ties. Nor are these af¿ nities shown by the Afanasyev people,。