Variability and chronostratigraphy of levallois technology in middle and initial Upper Paleolithic of Mongolia

Purpose. According to a traditional point of view, the flake, blade and point are regarded as resulted blanks for the Levallois technology. However, there are no well-established criteria for determining Levallois blades in Siberia and the northern part of Central Asia. This category is likely to be a core trimming element in flake and point production or may result from using non-Levallois technology. The article presents research of Levallois morphology, variability, specificity of flaking technology as well as geography of its distribution and chronology in Mongolia. It is possible to study flakes and tortoise-like cores on the basis of Levallois points and their cores if they are considered determining categories of products and substrate in Levallois flaking. The article describes the distribution of different variants of Levallois on the territory of Mongolia. The stratified and surface sites with this technology, which are known to the researchers, are clustered in Gobi Altai (Tsagaan Agui Cave, Chikhen Cave, Chikhen 2 site), Lake Vallley in Central Mongolia (Orok-Nor 1 and 2, Nariin-Gol 17a), Orkhon River Valley in Central Mongolia (Orkhon1 and 7 and Moyltyn am sites) and Northern part of Central Mongolia (Kharganyn-Gol 5 site). Levallois technology is presented by different variants in these sites and is presented both in the Terminal Middle Paleolithic and Initial Upper Paleolithic assemblages. Known industries demonstrate the final products of Levallois flaking or cores. All the reduction stages are not present, and the deficiency of refittings challenges reconstruction of Levallois operational sequences. Results. Our analysis of published and recently excavated new materials shows that three main variants are defined in the chronostratigraphic classification of Terminal Middle - Initial Upper Paleolithic complexes of Mongolia. They are associated with presence or absence of Levallois technology and its percentage in the industries. At first, it is centripetal Levallois methods and unidirectional convergent Levallois method presented in the Terminal Middle Paleolithic assemblages with Upper Paleolithic components, with Orkhon-1 (H 3 of area 1, 2), Moyltyn am (H 5 and 4) and Kharganyn-Gol-5 (H 6-7 ) considered as conveniently homogenic complexes. Secondly, it is Initial and Early Upper Paleolithic blade industries without evidence of Levallois technology present in Tolbor-4 (H 6-5), Tolbor-16 (H 7a, 7b, 5/6), Tolbor-15 (H 7-5). Thirdly, it is centripetal and convergent unidirectional methods associated with Upper Paleolithic sub-prismatic technology and primary Upper Paleolithic tool kit. These assemblages are presented in chronologically late layers of Early Upper Paleolithic in Chikhen Agui (layer 3), Chikhen-2 (layers 2.5-3) and Moyltyn am (H 2). Conclusion. The latest Levallois industries are synchronous to the earliest Initial Upper Paleolithic complexes of Mongolia. The blade of the Initial Upper Paleolithic is likely to have intrusive character and is not associated with the Final Middle Paleolithic genetically. The period of their coexistence was short.