From the Yenisei to the Yukon: Interpreting Lithic Assemblage Variability in Late Pleistocene/Early Holocene Beringia

2. On Late Upper Paleolithic Variability in South-Central Siberia: Rethinking the Afontova

33 2. On Late Upper Paleolithic Variability in South-Central Siberia: Rethinking the Afontova and Kokorevo Cultures kelly e. graf O ne of the major themes in the study of Beringian lithic variability is the relationship between the Nenana and Denali complexes in central Alaska. For years many researchers have argued that these were two distinct technological complexes produced by two groups of people, while others contended that the assemblages represent different behavioral facies of the same cultural tradition (Bever 2001; Goebel et al. 1991; Graf and Goebel 2009; Hoffecker and Elias 2007; Hoffecker et al. 1993; Powers and Hoffecker 1989; West 1996). Likewise, researchers working in south-central Siberia have been divided on a similar problem in their region. Throughout the twentieth century, Russian Paleolithic archaeologists explained local archaeological variability as the result of different cultures, and for the most part this type of interpretation has remained (e.g., Abramova 1967, 1979a, 1979b, 1989; Akimova et al. 2005; Drozdov et al. 1990; Lisitsyn 2000; Mochanov 1977; but see Astakhov 1986; Vasil’ev 1992, 1996, 2001). In the Yenisei River region, Abramova (1979a, 1979b) argued that lithic variability in several late Upper Paleolithic (LUP) sites, dating between about 16,000 and 12,000  C BP (19,000–14,000 cal BP), resulted from the presence of two nearly coeval archaeological cultures: the Afontova culture, including the sites of Afontova Gora 2, Afontova Gora 3, Kokorevo 2, Kokorevo 3, Tashtyk 1, and Tashtyk 2; and the Kokorevo culture , represented by the sites of Kokorevo 1, Kokorevo 4, Novoselovo 6, and Novoselovo 7. According to Abramova (1979a, 1979b, 1989), both cultural traditions were essentially contemporaries that originated from two groups of people using two technologies deeply rooted in different traditions. The Afontova culture emerged first at about 16,000  C BP (19,200 cal BP) in the Krasnoiarsk area at the Afontova Gora sites and was represented by a flake-based, Mousterian-derived technology; the Kokorevo culture emerged farther south in the middle Yenisei River area from a middle Upper Paleolithic blade and bladelet manufacturing culture. This notion has continued, and in fact Akimova et al. (2005) recently argued that the multicomponent Listvenka site was initially settled by the Afontova culture before about 15,000  C BP (18,500 cal BP), but soon afterward the more southern Kokorevo culture penetrated north and occupied the site for about 3,000 years, with a reoccupation by Afontova culture peoples during the terminal Pleistocene . Recent reevaluation of the radiocarbon chronology for the region, however, shows that cultural occupations of Afontova Gora and Kokorevo sites are completely contemporaneous , with no clear spatial or chronological separation (figure 2.1) (Graf 2008a, 2008b, 2009). Abramova (1979a, 1979b, 1989) defined these two LUP cultures primarily by noting differences in relative frequencies of lithic artifact types. Afontova culture (AC) sites became distinguished by high frequencies of flakes produced typically from amorphous cobble cores, bipolar cores, gravers, side scrapers, end scrapers, and pièces esquill és manufactured mainly on flakes, as well as a general lack of burins. In contrast, Kokorevo culture (KC) sites are characterized by blade- and bladelet-based technologies with burins and end scrapers on blades being common but gravers and bipolar cores being rare in the assemblages . Abramova (1989) held that microblades are present in both cultural traditions; however, they were used as microtools at AC sites and served as insets for composite tools at KC sites. Further, she argued that osseous projectile points from AC sites are typically round in cross section , whereas those found at KC sites are flat and slotted. Another diagnostic feature of the KC is the presence of slab-lined hearths (Abramova 1979b). By explaining the lithic variability in this way, 34 Kelly E. Graf the late glacial. The second is the proposal that differences in site types played a role in producing the patterns observed (Vasil’ev 1992, 1996, 2001). Vasil’ev (2001) fervently argued that we should examine LUP lithic variability in the Yenisei River region from a behavioral perspective . I concur and suggest that to better understand the behaviors that produced lithic differences and similarities between these sites we need to go beyond simply investigating relative frequencies of artifact types and undertake research on toolstone procurement and selection, technological organization, provisioning strategies, and foraging /land use behavior. In this chapter I address the Afontova-Kokorevo problem by analyzing lithic data from several LUP sites to tackle three objectives. First, I reexamine frequency data of several descriptive lithic variables used...