The Contamination of Pleistocene Radiocarbon Determinations in Australia

Article excerpt

Previous articles in ANTIQUITY have taken different views of the dating pattern for the human settlement of Australia. Is the apparent limit in the region of 35-40,000 years ago visible in the radiocarbon determinations a real date for the human presence? Or is it an artificial result of the dating method? A comparative study of the dating pattern in archaeological as against non-archaeological contexts may inform.

The radiocarbon 'barrier'

Roberts et al. (1994), in responding to a paper by one of us (Allen 1994), misrepresent the position taken there and the actual data on a number of occasions. At the same time they add nothing substantive to the main issue raised, that the present gap of c. 15,000 years between the dates for earliest human presence in Australia using luminescence and 14C techniques requires better explanation. On this point Roberts et al. (1994: 611, 615) reach the same intellectually unsatisfactory explanation as in their previous papers: 'the phenomenon of the radiocarbon barrier' is that determinations older than c. 35,000 b.p. 'are close to the theoretical limits of the method and that contamination by even a tiny amount of modern carbon could change a sample of "infinite" age into one with an apparent age of 40,000 years or less' (our emphasis. See also Jones 1982: 30; 1989: 762; 1993: 113; Roberts et al. 1990a: 126-7; 1990b: 95-6).

If the age limits in archaeological radiocarbon determinations for Australia arise from a technical contamination 'barrier' then it should also be seen in radiocarbon determinations from natural, non-cultural contexts (hereafter 'geological' contexts). If it is a real phenomenon, corresponding generally to the human settlement of the continent, then we should see different patterns in the determinations from each of the different contexts. We thus assembled 14C determinations from many Australian and New Guinean localities (the two countries being parts of the same Pleistocene landmass). These were divided into two groups, according to whether they derived from undisputed cultural contexts ('archaeological' contexts) or geological contexts.

We began by constructing lists of geological and archaeological uncalibrated determinations which exceeded 14,999 radiocarbon years in their mean ages, listed in the Appendix below. Both sets were assembled from either published sources and/or our own research data, or data offered to us by colleagues. We initially included all such determinations without reference to their locations, the types of material used for dating, whether the determinations were finite or infinite (i.e. older than a nominated age), whether or not determinations were considered by their publishers to be anomalous, and without particular reference to the year or place of their publication. Subsequently, we chose to exclude determinations for which we had no laboratory number (to avoid duplication, and also infinite determinations). This left a set of 217 geological and 254 archaeological determinations. The archaeological determinations used here derive from a search of Archaeology in Oceania, Australian Archaeology, Queensland Archaeological Review, Antiquity, Nature, books by Flood (1989) and White with O'Connell (1982), and the Terra Australis series and the Occasional Papers in Prehistory series published by the Department of Prehistory, Australian National University. These were augmented by all other published determinations over 14,999 b.p. known to us, regardless of publisher. Our own unpublished determinations from Tasmania were included, as were unpublished dates given to us by Richard Cosgrove and Peter Veth, whom we thank. Geological determinations were derived from Quaternary Research, Search, Australian Journal of Earth Sciences, Nature, Australian Journal of Science, Australian Journal of Botany, Palaeogeography, Palaeoclimatology, Palaeoecology, Colhoun (1985) and determinations kindly provided to us by Peter Kershaw, Geography and Environmental Science, Monash University. …