When using radiocarbon dates relating to archaeological materials from the last 10 000 years or so it is necessary to standardize or calibrate the laboratory determination (conventional radiocarbon age) to convert them into calendar years. This is especially important if radiocarbon determinations are to be compared with historically derived calendar dates, or dates determined by other means.
Laboratory determinations are given in radiocarbon years before present (RCYBP or BP), but radiocarbon years are not of equal length because of variations in the level of 14C in the atmosphere in the past. Special corrections for the effects of certain factors that alter the background level of 14C may also be necessary. Calibration is based on measurements of the 14C levels in material of known age, principally samples of ancient wood taken from dendrochronologically dated sequences. The earliest calibration curves, published in the mid 1970s, used samples from the bristlecone pine (Pinus aristat), an especially long‐lived tree species that grows in the Sierra Nevada of California. During the 1980s and 1990s regional calibration curves were developed for different parts of the world based on local dendrochronological sequences. A number of excellent computer‐based calibration programmes are available, including OxCal from the Radiocarbon Accelerator Unit in Oxford University, England, and CALIB from the University of Washington in Seattle, USA. Both can be downloaded from the world wide web.
Calibrated radiocarbon ages are conventionally indicated by the abbreviations bc/ad or cal.bc/cal.ad. However, because the calibration process involves using the standard deviations provided by the laboratory for the original determination, the calibration of a specific determination will result in an age range, the width of which will depend on the size of the standard deviation, the level of certainty used (typically 66 per cent or 95 per cent based on one and two standard deviations respectively), and the nature of the calibration curve in the region covered by the initial determination. Thus, calibrated radiocarbon dates are usually given as a range, for example 2030–2125 bc, which if calibrated using a single place of standard deviation would mean that there was a 66 per cent probability that the actual date lies within that range. When citing dates it is standard practice to give the calibrated age, the original determination, and the laboratory code (e.g. 2472–2404 bc (3925 ± 35 RCYBP) OxA‐5328). Within a book or academic paper all dates should be calibrated using the same procedures and the same calibration curve, and the parameters used should be summarized in the preface or some other convenient place. See also conventional radiocarbon age; corrected age; radiocarbon dating.
http://calib.qub.ac.uk/calib/ CALIB radiocarbon calibration programme, downloadable.
http://c14.arch.ox.ac.uk OxCal radiocarbon calibration programme, downloadable.