Abstract
The reanalysis of Mousterian assemblages by Binford and others has revealed major unresolved issues concerning how site differences are to be interpreted and what causes assemblages to differ. The argument proposed here is that the nature of the resource base under a given technology is the major determinant of assemblage variability. Scarce, unpredictable, and dispersed resources result in generalized and opportunistic exploitation patterns. This creates habitation sites with uniform assemblages characteristics, but does not create archaeologically visible procurement sites. In contrast, dense, predictable, and spatially clumped resources result in extended occupation of procurement sites and specialized tool assemblages. This creates a highly diverse set of archaeologically visible assemblages. These expectations conflict with expectations of other models.