Are Victoria West cores “proto-Levallois”? A phylogenetic assessment

Abstract

Cores from South Africa assigned to the “Victoria West” industry have long been purported as a “proto-Levallois” core form, and thus regarded as ancestral to the Levallois prepared core technologies of the Middle Paleolithic and African Middle Stone Age. Similarities in form between Victoria West cores, in terms of surface morphology and the removal of large flakes from a prepared surface, led to hypothesized schemes of technological evolution from Victoria West cores through to fully developed Levallois cores. However, the phylogenetic basis of this Victoria West “proto-Levallois” hypothesis, and the assumptions of phylogenetic homology upon which it rests, have never been tested formally. In recent years, archaeologists have begun to use phylogenetic methods drawn from biology to test hypotheses of technological and cultural evolution. Here, the phylogenetic assumptions of the Victoria West “proto-Levallois” hypothesis are tested directly using a cladistic (maximum parsimony) protocol. The cladistic analyses indicate that Victoria West cores are not the basal sister taxon of a Levallois clade, as predicted by the proto-Levallois hypothesis. Moreover, character analyses demonstrate that several characters relating to core surfaces and flake scar morphology are not phylogenetically homologous, but result from convergent technological evolution within the Acheulean techno-complex. Post hoc analyses further determine that these results are not confounded by choice of outgroup or raw material factors. The results were also shown to be robust on the basis of the ensemble retention index statistic, bootstrap analyses, and permutation tests. Hence, it is concluded that Victoria West cores do not represent a “proto-Levallois” core form, and that the term “para-Levallois” should more correctly be applied on phylogenetic grounds. It is further argued that even in cases where different technologies are found to share phylogenetically homologous features, use of the term “proto” is questionable on theoretical grounds.