Skip to Content.

Kent Archaeological Review extract

Radiocarbon dating.
by Peter Couldrey.

Since the publication of the first radiocarbon dates in 1949 by W F Libby, this method of obtaining 'absolute' dates for archaeological deposits has become widespread. No longer do prehistorians have to rely solely on the typological relationship of artefacts as a means of establishing a chronology. Instead dates can now be obtained as far back as about 60,000 BC without reference to trade, invasion or general diffusion. As a result, not only have the actual dates of events changed, but our interpretation of prehistory has been radically altered. For example, passage graves like those at Newgrange in Ireland and Maeshowe in Orkney were considered to have derived from the east Mediterranean (via Spain, Portugal and Brittany) and the so-called 'Wessex' culture of the British early Bronze Age was thought to have been dependent on Mycenaean Greece. Now, as a result of radiocarbon dates, the western cultures in both these cases are generally accepted as beginning earlier than those in the east, so this method of dating has proved to be of great importance to archaeologists. Unfortunately though, individual dates cannot be accepted as undisputed 'scientific' facts. It is intended here to outline the main assumptions behind the method and to note some problems involved.

The theory may be summarised as follows: Radiocarbon (or Carbon-14) is a radioactive isotope of carbon formed in the earth's outer atmosphere as a result of cosmic ray bombardment. These carbon-14 atoms combine with oxygen in the same way as the common form (carbon-12) to produce carbon dioxide. This then enters plants during the process of photosynthesis, and oceans as dissolved carbonate. In this way small quantities of radiocarbon are absorbed by all living organisms.

Because they are radioactive, the carbon-14 atoms immediately begin to decay at a known rate -- the half-life. (For any quantity of radioactive material, the half-life is the time taken for half of the original amount to decay). This time remains a constant. The rates of production and decay are such that the ratio of carbon-14 to carbon-12 atoms in the atmosphere also remains constant.

While an organism lives it will continue to absorb carbon dioxide and its supply of radiocarbon will be constantly replenished. But when it dies, this will cease and the ratio of carbon-14 to carbon-12 will decrease. Because the decay rate is known, the time that has elapsed since the death of the organism can be calculated after measuring the remaining ratio of carbon-14 to carbon-12 atoms. Naturally the age obtained in this way is in the form of years before the present (bp). By convention, to avoid the inconvenience of having to know the year in which the laboratory calculations were made, dates published in this form represent the number of radiocarbon years before 1950 AD.

Since its early development, several of the original assumptions behind the theory have been questioned. For example, it was thought that the carbon-14 atoms were evenly distributed over the earth's surface. This is now known to be false, though the divergence is very slight. Similarly, whereas it was previously assumed that all living organisms contained a standard ratio of carbon-14 to carbon-12 atoms, some plants are now known to contain more radiocarbon than others -- often depending on the nature of the soil. However this can now be accounted for and in both these cases the effect on the date is neglible.

Nevertheless there are two recent discoveries which have shown that dates calculated earlier need revision. The original half-life, used by Libby, was 5568 years. This figure is now considered to be too short and 5730 years is regarded as being more accurate. But because many dates have already been calculated using the old half-life this short one is still used until a new one is accepted by international agreement. The necessary correction can be made by multiplying the radiocarbon dates (bp) by 1.03. This alteration of the value of the half-life in no way effects the relative order of the dates.

More serious errors have been found as a result of work done using dendrochronology (tree-ring dating). By submitting a sample of wood to dendrochronological and radiocarbon methods, comparisons may be made between the results. It has been found that the radiocarbon dates gradually diverge from the 'true' dates and, after about 1500 BC, are constantly too young. Similar results were found by testing the radiocarbon method on historically datable material from Egypt. This work has shown that the ratio of carbon-14 to carbon-12 atoms in the atmosphere has not been constant throughout time. The advantage of using independently dated material to test radiocarbon method is not just that dates can be seen to be different, but that this difference can be measured. With this data a relationship can be established between radiocarbon dates and true calendar dates. Then, if such a relationship can be shown to be universally applicable, it will be possible to convert all radiocarbon dates to calendar dates. But this is not as simple as it first appears. The first curve published See Footnote [1] covering the time back to 5300 BC was drawn by 'cosmic schwung' (or eye) and though generally accepted, is disputed in detail. Even so, perhaps the main difficulty is in the use of such a curve, because it has "wriggles" which often result in one radiocarbon date being corrected to produce two or more possible 'true' dates. This is not to say that the 'wriggles' are wrong. Some consideration has been given to smoothing them out, but they are a fundamental part of the curve. Indeed, it is now certain that single radiocarbon dates are open to a wider degree of error than was previously supposed. Much work is being done on this subject: new calibration curves have been proposed See Footnote [2] and statistical procedures are being developed for studying the calibration of floating tree-ring chronologies. See Footnote [3] We still await a universally accepted calibration curve.

In spite of these difficulties, the overall relative accuracy of radiocarbon dates is generally accepted. For instance a series of twenty-six dates for samples from Sitagroi in north Greece showed good general agreement with their straigraphic order. See Footnote [4] Even without precise calendar dates radiocarbon dating is a valuable aid to our interpretation of the past, though such interpretation is only valid when the limitations of the method are understood.

When using the dates, the following convention should be observed: radiocarbon dates are denoted by bp, bc and ad; and corrected dates are denoted by BP, BC and AD. Furthermore, it should be remembered that no radiocarbon date is of any value unless the associated fieldwork is thorough. The date represents the time of the death of the organism from which the sample is taken. It is important that the relationship between the death of the organism and the archaeological context is understood An unstratified or contaminated sample is useless.


Footnote 1.

Suess, H E 'Bristlecone pine calibration of the radiocarbon time-scale 5200 BC to present' in Olsson, Ingrid U. (Editor) Radiocarbon variations and absolute chronology: Proceedings 12th Nobel Symposium. 1970. Return to the paragraph.

Footnote 2.

Switsur, V R 'The Radiocarbon Calendar Recalibrated' Antiquity, volume XLVII, Number 186, 1973. Return to the paragraph.

Footnote 3.

Clark, R M and Renfrew, C. 'A Statistical approach to the calibration of floating tree-ring chronologies using radiocarbon dates' Archaeometry, volume 14(I), 1972.
Clark, R M and Sowray, A. 'Further statistical methods for the calibration of floating tree-ring chronologies' Archaeometry, volume 15(2), 1973. Return to the paragraph.

Footnote 4.

Renfrew, Colin 'Sitagroi and the Prehistory of South-East Europe', Antiquity, volume XIV, Number 180, 1971. Return to the paragraph.
Close Window
Accessed this page via search engine or bookmark? Full K A R Index here.