of assemblage composition, allowing the definition of six distinctive facies. Bordes could find no association between facies and environment or between facies and time, leading him to suggest that each distinctive suite of artifacts represented a different cultural tradition. The interdigitating of facies over a long period of time in a relatively small area led Bordes to describe these cultural traditions as tenacious.

In the 1950s Bordes’s approach to assemblage analysis and interpretation was pioneering. Earlier studies, such as that of denis peyrony, had characterized assemblages in terms of the presence or absence of a few distinctive artifact types (the type fossil approach). In contrast, Bordes’s approach involved the analysis of entire artifact assemblages and separated information about the way that tools were made (technology) from the types of tools that were produced (typology).

Since Bordes’s studies, archaeologists have come to appreciate that a whole array of variables has contributed to the similarities and differences in the composition of the assemblages they study. However, four variables are considered fundamental to any attempt to understand assemblage composition: the flaking qualities and edge properties of the raw materials from which the artifacts were made, the techniques used to make the artifacts, the uses to which the artifacts were put, and culturally ordained rules about what an artifact of a particular function should look like (this variable is usually referred to as style). In the latter part of the twentieth century archaeologists strove to disentangle the impact that each of these variables had on the composition of artifact assemblages from different time periods.

Even as Bordes was discussing the impact of cultural traditions on artifact assemblages, other archaeologists, such as charles mcburney and j. d. clark, were investigating the possibility that recurring assemblage types might represent seasonal or functional differences in activities undertaken at different sites. These types of inquiries are often linked to the work of l. r. binford, who, in the 1960s and early 1970s, published a series of papers exploring the possibility that variations in the composition of Lower and Middle Paleolithic artifact assemblages might result from differences in function rather than from different traditions for making tools designed for particular tasks—that is, culture (see, e.g., Binford and Binford 1969). One stimulus for this idea came from studies of Lower Paleolithic assemblages and the observation that the geographic and temporal scale of variation in artifact assemblages was far greater than the spatial and temporal boundaries of cultural entities. However, the key stimulus came from studies of the material culture of contemporary hunting-and-gathering communities. The Binfords observed that these communities used different tool kits at different times and in different places, in response to seasonal and temporal variation in the availability of plant and animal resources. It was argued that the same organizing principles—the use of different items of material culture for different purposes in different contexts—would have influenced the behavior of prehistoric hunters and gatherers and, therefore, the composition of the archaeological assemblages they left behind.

Binford and Binford used multivariate statistics to explore the co-occurrence of artifact types. This technique revealed the existence of five recurring sets of artifact types, each of which was interpreted as a functionally distinct tool kit. The three sites investigated were found to contain different groups of tool kits, and each was consequently interpreted as a different type of site within a wider settlement system. Almost two decades later microscopic analyses of the wear traces preserved on the edges of Middle Paleolithic artifacts from the Middle East and France provided a direct test of the functions of the tool clusters identified by the Binfords. Perhaps the most important result of these use-wear studies was the observation that, during the Middle Paleolithic, there was no direct relationship between the form of a tool and the use to which it was put (Shea 1992). This undermined the specific functional interpretations that the Binfords had offered for each of the tool clusters they had described but not necessarily the underlying logic of their approach.

Subsequent studies (such as those of H.L. Dibble) of the impact of edge resharpening on