| Summary: | This study is designed to investigate patterns of lithic technological variability in relation to settlement strategies that were employed by late prehistoric inhabitants of central and southern regions of interior British Columbia. The research contributes to current archaeological method through an experimental program of stone tool manufacture, and also to current understanding of Interior Plateau prehistory, through a multiregiohal analysis of technological variability. The first stage of the study involves conducting a controlled experiment, to determine the degree to which lithic debitage can be used to predict stages of chipped stone tool manufacture, and to devise an efficient means of classifying debitage into general reduction stages. The experiment is unique in providing control over the precise sequential removal of flakes, and also in examining quantitative variability in debitage that have been produced as the by-products of the manufacture of several tools and cores. The result of the experimental program is the formulation of a debitage classification that classifies flakes into early, middle or late reduction stages, and also into bifacial and bipolar reduction types. The archaeological analyses in the second major stage of the research use the debitage reduction stage classification and the occurrence of various lithic tools to examine the nature of interassemblage variability across the 38 sites from four regions of the Interior Plateau. A total of 14,541 flakes, 164 cores and 861 tools from the Eagle Lake, Mouth of the Chilcotin, Lillooet and Hat Creek regions are analyzed, using multivariate and bivariate quantitative methods. Three hypotheses relevant to lithic technology and hunter- gatherer archaeology are evaluated in this stage of the study. The analyses first employ the experimental debitage classification to obtain interpretable patterns of inter-assemblage similarities and differences. Multivariate analysis shows that several kinds of sites defined on the basis of features can be grouped by their predominance of early/core reduction, middle/wide ranging reduction, and late/ maintenance reduction debitage. The first formal hypothesis tested is that obsidian and chert raw materials should evidence patterns of conservation and economizing behavior by virtue of their geological scarcity in relation to vitreous basalt raw material. A series of chi-square tests demonstrates that debitage frequencies by reduction stage are proportionately equal for these three raw materials in all but the Mouth of the Chilcotin region. In all regions, except Lillooet where tool sample sizes are too small for reliable testing, tool sizes and scar counts show no significant difference attributable to raw materials. A slight trend is noted for chert tools to be larger and simpler than vitreous basalt or obsidian tools. A set of bivariate graphs demonstrates that while lithic raw materials may be reduced in highly similar manners, one raw material may have served to replace another. The second hypothesis, that tool curation and maintenance strongly affects assemblage composition, is first tested by examining tool assemblage measures that have been suggested by recent lithic technological models. Assemblages are highly variable with respect to the numbers of tools left at sites in relation to the intensity of tool maintenance that occurred at sites. The third hypothesis tested is that a set of site occupation purposes can be reliably predicted on the basis of debitage reduction stages and a functional tool classification. Using multiple discriminant analysis, house-pit sites are accurately predicted at an 80% rate, and lithic scatters without features are accurately predicted at a rate of 60%. Lithic scatters with housepits achieve 86% correct classification; lithic scatters with cachepits are correctly classified at a rate of 75%; and lithic scatters with fire-cracked rock are accurately predicted 80% of the time. The results of this analysis are further strengthened by removing an ambiguous assemblage from consideration. The most significant findings of the multiregional analyses are those of definite tool cuiration patterns as evidenced in the raw material analysis, and the occupation span inferences of the tool maintenance analysis. Overall, it has been demonstrated that an experimentally obtained stage classification of debitage enables the derivation of behavioral inferences that could not be currently obtained by other means. In its multiregional perspective, this study has shown that processes of lithic assemblage formation are largely independent of regional provenience and more dependent on settlement purpose. Overall, the greatest determinant of assemblage variability is inferred to be site occupation span. Arts, Faculty of Anthropology, Department of Graduate
|
|---|