The Denbigh Flint Complex in Northwest Alaska: A Spatial Analysis
The uniformity of Denbigh Flint assemblages across different ecological zones in northwestern Alaska suggests that the typologies we have constructed for the analysis of lithic remains are insufficiently sensitive to reveal behavioral differences between sites. By integrating typological, materials, and spatial data, however, we are able to detect some differences that relate to ecology, seasonality and external relationships. I focus here on a spatial analysis of multiple Denbigh components from two site areas: Cape Krusenstern and Onion Portage. I employ two different analytical strategies to identify possible toolkits: a factor analysis of artifact clusters within a fixed distance from hearths and within house floors, and a more flexible cluster analysis based on the nature of artifact clusters in a variety of contexts. The factor analysis reveals five types of meaningful clusters, each with assemblages related to different sets of activities. The second, more ad hoc clustering method is based on five spatial variables: house floors; hearths; near hearths; artifact concentrations unassociated with formal features; and areas of randomly dispersed artifacts, and is especially effective in revealing differences in seasonality. This approach reveals sets of activities during snow-free seasons, some specifically late spring/early summer or fall, versus activity sets that occurred in winter. From a more regional perspective, Denbigh implements are seen to have been brought to the sites in finished or near finished form, indicating that the earlier stages in tool manufacture occurred elsewhere. The apparent importation of Denbigh tools in finished form adds fuel to William Irving's argument that the finest of the Denbigh artifacts were produced by itinerant flintknapping specialists. The lithic analysis also informs us about the nature of external contacts, especially between the coast, and the Kobuk, Noatak and Koyukuk rivers. Finally, I demonstrate that these different analytic methods for studying prehistoric activities have their own strengths and weaknesses, and without good spatial data even the best of die methods has major limitations. This is a call to increased attention in our excavations to recording precise provenience of all lithic materials — flakes as well as formal artifacts, a formidable task, but one with rewarding results.