Published online by Cambridge University Press: 02 January 2015
In order to obtain a good site chronology the radiocarbon measurements must be carried out on samples having a firm archaeological basis and high probability of association with the event or structure the date of which is required. Patently the radiocarbon date can be no better than the sample provided for measurements, and any missassociation at this stage cannot subsequently be rectified. Often discordant radiocarbon dates are explained away by supposing that the sample was ‘contaminated’ with earlier or later material. Calculations based on these surmises may often disclose that excessively large amounts of contaminants must be postulated to secure a given result.