| Abstract: | Accurate bone density data are essential for assessing the influence of destructive processes in archaeological faunal assemblages. Unfortunately, the diversity of methods employed by different researchers to derive density values has resulted in recent confusion. Two recent publications in this journal [J Archaeol Sci 29 (2002) 883; J Archaeol Sci 29 (2002) 979] exemplify this state of misunderstanding. Both studies argued that the role of bone density in shaping archaeological faunal assemblages has been largely overrated, but both based their conclusions on density values that were inaccurately derived. The former employed a method of calculating bone density that has been largely discredited over the past decade [J Archaeol Sci 29 (2002) 883]. Within a larger discussion of zooarchaeological methodology, the latter provided an assessment of the current state of bone density research that inappropriately characterized the discrepancies between available sets of density data as a reflection of the differences between two technologies—photon densitometry and computed tomography [J Archaeol Sci 29 (2002) 979]. The actual dichotomy exists—irrespective of the technology employed—between studies that account for variation in the shape of bone cross-sections and those that do not. The different sets of density data currently available to zooarchaeologists vary tremendously in their accuracy. We review and evaluate the different techniques employed in the research of bone density patterns of mammalian fauna. Computed tomography produces the most accurate density data. For elements without medullary cavities, photon densitometry may provide density values of similar accuracy but only if a method of cross-sectional shape-adjustment is applied. |
