J Archaeol Sci 32 (12)

The author argues that many studies of the human behavioural significance of frequencies of cut-marked mammal bones focus on determining the so-called `butchering pattern', and that few analysts comment on the tremendous range of variation in frequencies and anatomical distributions of cut marks across multiple assemblages of remains of a taxon. Such variation is evident even when faunal remains are associated with technologically, temporally, and environmentally similar cultures. This kind of variation is illustrated with frequencies of cut marked bone specimens comprising major limb joints of two artiodactyl genera from two sites in the northwestern United States. Three hypotheses are tested. The first and second hypotheses (one per genus) predict that the frequencies of cut-marked remains of a taxon from one site will match those frequencies evident on the remains of the same taxon at the other site. Both hypotheses are falsified. The third hypothesis is that remains of the larger taxon at each site will display more cut marks than the remains of the smaller taxon at each site. This hypothesis is statistically falsified at one site but not the other. Refutation of the hypotheses suggests that well-founded interpretations of frequencies of cut-marked remains may require unique kinds of contextual data.

The Loch of Clickimin, Shetland, is well known for its broch and associated monuments supposedly of the Bronze and Iron Ages, although the former date is disputed. Pollen, diatom and sedimentological investigations permit fresh insights into the landscape and economic impacts of those who constructed the monuments, reveal that the environs of the site experienced environmental change from Neolithic times onward and provide the first Holocene radiocarbon dates from the immediate locality. The broch-building period witnessed a continuation of pre-existing pastoral husbandry for which heather burning may have exacerbated the natural spread of blanket peat. Initially, no evidence was adduced for local arable activity in the present investigation, and this was thought to be consistent possibly with the low numbers of excavated querns and the absence of cereal macrofossil finds when compared with other Shetland broch sites. The application of 'rapid scanning' techniques, however, led to the discovery of a consistent cereal-type pollen representation from Bronze Age times onwards. The discrepancies between the palaeoenvironmental evidence and the environmental and palaeoeconomic inferences made by the excavator of the site are explored. Evidence is presented that both supports and contradicts previous assumptions surrounding an important archaeological site.