CHAPTER 4 : Dating the Past : Key Factors of Radiocarbon Dating

• radiocarbon dating is universal, because the radioactive isotope 14C is formed continuously throughout the Earth's atmosphere by the effects of cosmic radiation.
• 14C has a known half-life, and decays at a known rate, but the original half-life was too low by around 3%; it is now judged to be around 5730 years, rather than 5568.
• the rates of formation and decay are in balance; cosmic radiation in the past should have maintained 14C in the atmosphere at a constant level. However, the level of cosmic radiation has fluctuated over time, perhaps in relation to sunspot activity and the Earth's magnetic intensity. This means that the formation of 14C in the atmosphere has varied; thus, samples from organisms that absorbed abnormally larger or smaller amounts of 14C will give misleadingly younger or older dates. In addition, calibration reveals that dates from the southern hemisphere are around 30 years too old compared with those from the northern hemisphere; this is probably because the greater area of oceans in the southern hemisphere has affected the distribution of 14C in the atmosphere.
• all life-forms contain carbon, and living organisms absorb carbon from the atmosphere, mainly in the form of carbon dioxide - photosynthesis by plants is one common mechanism. Animals and plants therefore maintain the same proportion of newly formed 14C as the atmosphere until their death, when it begins to decay. However, different isotopes of carbon are taken into organisms at different rates (fractionation); proportions of 13C and 14C must be checked and an adjustment made to the estimated date. Furthermore, marine organisms absorb 'old' 14C from sea water; samples taken from shells or bones of marine mammals give dates which are misleadingly early by several hundred years. Some of this old carbon is absorbed by humans - in Scotland, for example, by people living in the Mesolithic period on Oronsay (Richards and Sheridan 2000) and in the Viking period on Orkney (Barrett et al. 2000) - as well as by other animal species which eat large quantities of seafood.
• a calibration curve must be used to convert radiocarbon years into calendar years (fig. 4.17). Tree rings have revealed not only short-term fluctuations in 14C levels but also a long-term divergence between 14C estimations and calendar years that grows increasingly wider before c. 1000 BC. Samples with a radiocarbon age of 5,000-7,000 years require upward adjustment of as much as 500-1,000 years, while uranium-thorium dating shows that coral dated to around 26,000 BC by radiocarbon is actually 30,000 years old. Dendrochronology provides independently dated samples of wood from annual tree rings stretching back more than 11,000 years. Some bristlecone pines still growing in the south-western USA are more than 4,000 years old, while earlier samples come from dead trunks preserved in this semi-arid habitat, and from oak trees found in bogs or river sediments in Europe. Samples from marine corals have extended the calibration curve back to around 30,000 BC by comparing 14C with dates derived from uranium/thorium isotopes.
• a statistical estimation of error, expressed as a standard deviation, is attached to laboratory counts of radioactivity. Since isotope decays occur at random, a reasonably long counting period is needed to reduce this inherent error. Several counting sessions are carried out along with measurements of laboratory standards to monitor the performance of the equipment. The standard deviation derived from the counting statistics is preceded by '±'; fig. 4.18 shows how the reliability of a date may be envisaged.