Radiocarbon Dating

(article by Marvin W. Rowe)

One of the more frequently asked questions about rock art is: How old is it? There was a revolution in our ability to answer that important question starting three decades ago (1977). The major advancement that laid the groundwork for the later dating revolution was the development of accelerator mass spectrometry (AMS). That innovation led to a drastic reduction in the amount of carbon necessary for a radiocarbon date - from a few grams to less than 1 mg of carbon. This reduction in sample size opened the way for even the small amounts of organic matter in rock paintings to be dated starting a decade later in 1987. Assigning painted images to a particular time period and, thus, a prehistoric culture, allows archaeologists to gain information on the artistic, cultural, technical and religious aspects of a people.

Most radiocarbon dates on rock paintings have been attained through three major techniques: (1) dating the carbon from black paintings with charcoal pigments; (2) dating other pigments that are also organic, and more rarely pigments that include discrete organic materials that can be removed physically; and (3) dating the carbon extracted from mineral pigmented paintings in which organic matter was added in the preparation of the paints. All of these techniques have distinctly different advantages and challenges in their application.

The first radiocarbon dates on rock paintings were in 1987 on charcoal pigments, in some ways the most straight forward method. Dating charcoal is the best tested technique as it has been used extensively since the beginning of radiocarbon dating. But it has two inherent problems that have been, for the most part, ignored. The first of these is encountered in the dating of all archaeological charcoal: the so-called “old wood” problem. This situation occurs when wood that has been dead for a long time, but has simply not decayed yet, is burned forming charcoal. It also occurs when the central portions of a very old tree are burned, again yielding charcoal. A second potential problem, particularly important for dating of charcoal paintings, is the so-called “old charcoal” problem. This situation occurs when the rock artist uses a piece of charcoal that just happens to have been lying on the ground for an undetermined length of time. A piece of charcoal may be quite old before being picked up to construct a drawing. A documented example of that occurred at a site in Australia. A historical charcoal graffiti, “Mr. C. B. Ross” was dated that was near an engraving, “C. Ross 1894”. The radiocarbon date indicated that the charcoal graffiti was about 1300 years old! This is a problem that is present in all dating of charcoal paintings, and one that is generally undetectable.

In the second category of rock art dating, organic pigments or organic inclusions, the problems are perhaps less difficult than for other techniques. Both these situations are rarely encountered. Dating of organic “beeswax” paintings in Australia is the only situation to my knowledge where replicate samples of a particular painting were dated by two different research groups, albeit using very similar techniques. Unfortunately, there was a statistically significant difference between the two ages. However, the difference in the ages, 4040 ± 80 years ago compared to 4460 ± 80 years ago, was not great (10 %). The main disadvantage of this is that such “beeswax” rock art is very rare, as far as I know occurring only in one area in Australia.

Inorganic pigments were far more frequently used than charcoal or other organic pigments in making rock paintings. Reds, oranges, browns, and yellows are iron oxide/hydroxide mixtures and black is quite often manganese oxide/hydroxide mixtures; none of these pigments can be dated directly. The third major dating technique utilizes plasma-chemistry to extract any organic material that is present in the mineral-pigmented paints. This technique also encounters several problems. The first is when no organic matter was added initially to produce the paint. In that case no date is possible by any method using radiocarbon dating. That situation is very clear: no carbon is extracted, so no date is possible. Similarly, some organic material may be extracted, but in amounts too small to attempt a date. Another problem is that the rock upon which a painting occurs may itself contain significant amounts of organic matter, enough in some cases to invalidate any attempted date. That can be ascertained by extracting carbon from a nearby rock surface. When that background contamination is found to be too high, no further attempt is made to date that painting. And finally, the chemical identity of the material being dated is unknown. And some organic materials are unsuitable for radiocarbon dating. The general validity of this method has been verified by comparing dates obtained from paintings for which archaeologists have inferred a date range based on archaeological inference; agreement was satisfactory, although the inferred age ranges are generally too large to strictly test the technique in detail. The plasma-chemical extraction method has also been validated by measuring the radiocarbon dates on archaeological materials that were dated by conventional AMS dating. The veracity of the method has been studied further by running replicate samples on the same painting, sometimes using different chemical pretreatments. The conclusion of all the studies seems to indicate that the method yields rock painting dates in the expected ranges, but with a higher than expected uncertainty, ~ ± 250 years.

Finally, one additional technique has been used only rarely, but is useful for both paintings and carvings. In many areas calcium carbonate and calcium oxalate mineral coatings form on top of the rock art. If that layer can be removed, dating the oxalates that have formed yields a minimum age for the art. This technique has provided useful constraints on the ages of rock art in a few cases.

I will close with one last problem that must be overcome for the final validation of all techniques for dating rock art. It can be summarized by a quote here from the book, Faust in Copenhagen: a Struggle for the Soul of Physics, describing the revolution in physics that occurred during the 20^th century.

A second opinion was going to be necessary, no matter how reliable … results were. There was always some possibility of error in such research, and the standard operating procedure for an important experiment was and still is to have it repeated in another laboratory. If results agree, the community can proceed with confidence.

Until more determinations are made on replicate samples by different laboratories, using different techniques, and agreement is observed, we cannot be completely confident in the dates. As another rock art researcher put it, “One of the basic requirements of science is the replicability of experiments.”

But, in spite of the potential problems that beset each technique, several hundred dates on rock paintings around the world have now been produced by several laboratories. We can be optimistic that most of those dates will stand the test of time and be accepted as reliable. Getting accurate dates on rock art is important as it will allow incorporation of rock art information into the other archaeology of a site.

Bibliography

M. W. Rowe, “Dating studies of prehistoric pictographs in North America”, In Discovering North American Rock Art, eds. C. Chippendale, D. S. Whitley & L. L. Loendorf, University of Arizona Press, Tucson, AZ, Chapter 12, pp. 294-319, 2005.

M. W. Rowe, Reflections on dating rock art, In Exploring the Mind of Ancient Man, ed. P. C. Reddy, Research India Press, New Delhi, India, Chapter 19, pp. 218-231, 2007.

G. Segrè, Faust in Copenhagen: a Struggle for the Soul of Physics, Viking, Penguin Group, NY, NY, p.183, 2007.