Abstract
Abstract The Levantine sites of Skhul and Qafzeh have been interpreted as indicating an early, short and unsuccessful expansion of Homo sapiens out of Africa. Chronometric age estimates, however, indicate a history of prolonged occupation, and suggest that Skhul (∼130–100 thousand years ago [ka]) may have been occupied earlier than Qafzeh (beginning ∼110–90 ka). Morphologically, the Skhul individuals can be described as somewhat more primitive in comparison to the Qafzeh fossils. Though the lithic assemblages of sites such as Skhul and Qafzeh are often described as being technologically similar, as part of the ‘Tabun C’ phase/industry, limited detailed information on the Skhul lithic assemblage has been published, and little comparative work has been conducted. Here, we present an analysis of the Skhul stone tool assemblage to describe its characteristics, to evaluate the lithic results against the fossil and chronological data, and for inter-site regional comparison. Our findings indicate that the Skhul lithic assemblage differs from other Levantine Marine Isotope Stage (MIS) 5 sites, such as Qafzeh. For example, there was more of an emphasis on diverse methods of point production at Skhul, and the available samples indicate a greater emphasis on preferential rather than recurrent Levallois reduction at Skhul. The current findings suggest that neither the Levantine Middle Palaeolithic in general, nor \MIS\} 5 assemblages in particular, were technologically homogeneous. These data are consistent with either a long occupation of the Levant by Homo sapiens in \{MIS\} 5, or at least two phases of occupation (early \{MIS\} 5 and mid to late MIS). Whatever the fate of the Skhul and Qafzeh population(s), their occupation of the Levant was neither short nor culturally uniform. Our findings add to the growing pool of evidence that the dispersal of our species ‘Out of Africa’ was more complex than hitherto thought. Further work on \{MIS\ 5e contexts in the Levant and elsewhere in Southwest Asia should be a research priority.
