Dating the Aegean Late Bronze Age with Radiocarbon*

When Cadogan wrote ‘Dating the Aegean Bronze Age without radiocarbon’ (1978), this writer agreed completely with the view that the absolute dating of the Aegean Late Bronze Age was already established. Correlations with Egypt and other parts of the eastern Mediterranean seemed secure, and virtually all Aegean prehistorians felt that radiocarbon dates could be ‘tested’ by comparison with the Aegean chronology. By this reasoning, the radiocarbon dates for Late Minoan (LM) I/Late Cycladic (LC) I Thera were rejected because they suggested a date about a century earlier than the ‘traditional’ dates (Betancourt and Weinstein 1976). The assumptions must now be reconsidered. The series of radiocarbon dates from the LC IA settlement at Akrotiri on Thera is one of the longest series from any European site. A subset of short-lived samples, processed by the University of Pennsylvania laboratory, remains today one of the most internally consistent groups of dates (Michael 1976, Michael and Weinstein 1977). The evidence from this group needs only a summary here because it is well known, having been discussed many times in the literature (Betancourt and Weinstein 1976 pp. 332-333, illustration 1, Michael 1977, Weinstein and Betancourt 1977). The entire series processed by the Pennsylvania laboratory is particularly useful because the species that yielded the charcoal were determined, and the contexts were archaeologically secure. If one omits from consideration the dates derived from undersized samples, those from contexts before the LM IA/LC IA destruction, and those from long-lived trees which should date building materials erected well before the destruction, one is left with ten samples, all from twigs, grain and other short-lived material. After calibration, this groups yields dates with a cluster in the seventeenth century B.C., by whatever valid calibration or statistical method which may be employed. In short, the series is insistent on a date in the seventeenth century B.C. for the end of LM IA and LC IA. The difficulties with this conclusion, admirably summarized by Cadogan (1 978), stem from the fact that this date is about a century earlier than the ‘traditional’ chronology based on interrelations with the eastern Mediterranean. Several explanations have been advanced in efforts to explain the discrepancy, but none are satisfactory. As Michael noted when he first advanced the theory, emissions of volcanic carbon dioxide which would raise the resultant dates of plants near fumaroles do not explain the consistent results obtained from several different materials taken from different storage areas (1977 p. 794). Samples taken from short-lived materials might be expected to yield deviant results with one or two dates