Radiocarbon 32

Outlines the cause of the recently-discovered errors in dating (centred on the counting system) and explains the methods adopted to revise the affected results. Where revision has been possible dates have been found older by 200 to 300 radiocarbon years (on average). Individual revised results are tabulated: in particular the Cranborne Chase, Peel Castle Viking cemetery, and Burghfield Quarry results illustrate the helpful effects of the revision.

Summarizes guidelines accepted by most radiocarbon laboratories in the interests of producing consistently reliable dates.

M S Baxter gave the opening address (253-5) and closing remarks (393-7) which included a brief summary of progress and future plans. Several papers by Glasgow staff are included (T C Aitchison, E M Scott, M S Baxter, D D Harkness, G T Cook), on technical aspects of the intercomparison exercise. Roy Switsur (341-6) offers 'A consideration of some basic ideas for quality assurance in radiocarbon dating', and M G L Baillie (361-6) writes on 'Checking back on an assemblage of published radiocarbon dates', in which he warns against placing too much reliance on published dates produced in the past. For instance, although the Sweet Track in the Somerset Levels is now tightly dendrodated, a high proportion of the early dates for it were too young yet remain in the literature. To assist users to understand the limitations associated with past dates, it is shown that over a sample of sixty-two determinations for the English Neolithic, 34% were found to lie outside 200 radiocarbon years of their true radiocarbon ages. Radiocarbon dating is a fairly blunt tool, and laboratories should not make unsupported claims for the accuracy of their determinations. Austin Long (393-7) describes 'A quality assurance protocol for radiocarbon dating laboratories'.