Archaeologists are aware that many factors change archaeological sites after they have been initially deposited. One kind of post-depositional phenomena that could change the material record is the scavenging and reuse of manos and metates from older sites by the later inhabitants of an area. If this has occurred, even on a limited basis, grinding tools may be disproportionately represented on older sites. In this paper I will argue (1) that the scavenging and reuse of grinding stones by hunter-gatherers should be expected on theoretical grounds under many circumstances and that this behavior has occurred in the Great Basin and elsewhere, and (2) that there is a statistically significant bias in the occurrence of grinding stones toward Late Prehistoric sites in the eastern Great Basin. I will then discuss the consequences of such a pattern for archaeological interpretation of site function.

Excavations in the Sevier Desert of western Utah (Fig. 1) have produced architectural evidence for temporary habitation in small brush, wickiup-type structures between A.D. 1000 and 1100 by people with Fremont material culture. The inventory includes Fremont ceramics, a single corn cob, and projectile points common to, but not exclusive to, Fremont sites (Cottonwood Triangular, Rose Spring Cornernotched, and Desert Side-notched). The site broadens our understanding of Fremont architectural and settlement diversity. Several alternatives for eastern Great Basin Fremont economy are presented in fitting this site into the regional subsistence/settlement pattern(s).

This paper describes an experiment in which the Bustos site is treated as if it were an ethnoarchaeological situation in addition to being an interesting example of a late prehistoric site in the Great Basin. Ethnoarchaeological studies of site formation processes shaping the structure of sites (e.g., Kent 1984, 1987; O'Connell 1987) have been important to our understanding of site function, task organization, duration of occupation, seasonality, and the role of storage. The archaeological visibility of lightly constructed, perishable structures is another issue. A high percentage of past hunter-gatherer residential behavior has likely left archaeologists with a disproportionate number of lithic scatters, while decayed structures, requiring far more attention to locate, go unrecognized. Inferences from ethnoarchaeology conducted in other regions of the world have been applied to a case in the northeastern Great Basin to help identify the location of small structures whose superstructures have vanished (Simms and Heath n.d.). These kinds of studies hold implications for the use of negative evidence (i.e., the absence of residences) to interpret site function, as well as the assessment of significance, and policies for survey and test excavations in cultural resource management.

As part of an earlier study (Simms 1984), data on the costs and benefits of obtaining native food resources in the Great Basin were generated for use in foraging models developed from evolutionary ecology. Portions of these data are presented here for the benefit of researchers interested in the acquisition costs and nutrition of wild foods.

Wetland Adaptations in the Great Basin. Joel C. Janetski and David B. Madsen, eds. Provo: Brigham Young University Museum of Peoples and Cultures Occasional Papers No. 1, 1990, v + 285 pp., $17.00 (paper).

An apparent correlation between the expansion of pinyon pine into the northern Great Basin and prehistoric settlement shifts in the Reese River Valley and Grouse Creek region (Fig. 1) is explored using data on the costs of pine-nut procurement relative to alternative resources. A model of diet breadth is used to develop predictions about the timing of pine-nut use in the two cases. The diet-breadth model does not accurately predict the timing of pine-nut use in a third case, the Owens Valley (Fig. 1), leading to a discussion of other possible adaptive constraints. This study is about more than prehistoric pine-nut use. The issues confronted in these cases can probably not be resolved without interaction between theory and data. Such interaction can work to evaluate a typically incomplete and frequently misleading archaeological data set, as well as the underlying theories. There is a basic premise here that data, direct or otherwise, rarely speak for themselves and that a Baconian stance, requiring us to refrain from theorizing until "all" the data are in, shortchanges the scientific endeavor.