Recent Developments in Radiocarbon and Stylistic Methods of Dating Rock-Art
Rosenfeld, Andree, Smith, Claire, Antiquity
Recent debate over the age of the rock-art of the Coa Valley, Portugal, has treated radiocarbon and stylistic approaches to rock-art dating as opposed. Several methods for 'directly' dating rock-art applied to the Coa rock engravings have yielded Holocene dates that are at odds with the Palaeolithic dates obtained by European researchers using established, stylistic methods (see Bednarik 1995a; 1995b; Watchman 1995; 1996), and European archaeologists are rejecting these direct dates (e.g. Clottes et al. 1995; Reevell 1995; Simoes de Abreu & Jaffe 1995; Zilhao 1995a; 1995b; Zuchner 1995). Here classic, stylistic methods have come into open - and extended conflict with the more recently developed radiocarbon methods. This apparent conflict demands an evaluation of the advantages and the limitations of both methods, as well as consideration of the rather different temporal information that each provides.
Accelerator Mass Spectrometer (AMS) radiocarbon dating enables the ages of extremely small samples of organic matter to be calculated. The possibility of dating of organic material in rock-art pigments by AMS was realized in the early 1980s (e.g. Van der Merwe 1982) and the first results came a few years later (e.g. Hedges et al. 1987; Van der Merwe et al. 1987; Loy et al. 1990; McDonald et al. 1990).
AMS radiocarbon dating was greeted with enthusiasm by rock-art researchers for it promised dates of greater precision and more importantly dates from very small samples from a great range of materials (Rosenfeld in press). Radiocarbon determination has since been applied to charcoal paintings (e.g. McDonald et al. 1990; David 1992; Chaffe et al. 1993; Clottes & Courtin 1993), to organic extracts from ochre paintings (Loy et al. 1990; Cole & Watchman 1992; David et al. 1993), to beeswax figures on rock (Nelson et al. 1993; 1995; Tacon et al. in press), to rock-surface accretions such as oxalate crusts (Watchman 1991; 1993; Cole et al. 1995; Cole in press) and to organic inclusions in siliceous skins (Watchman 1992; 1994) and in 'desert varnish' (Nobbs in press). For the first time, there has emerged the possibility of comparing radiocarbon-dated rock-art sequences with radiocarbon dates for other archaeological evidence (cf. Morwood & Smith 1994: 24).
Nevertheless, recent research has identified problems with radiocarbon dating of rock-art. Perhaps the most serious concerns the source of the carbon extracted for dating. While paints may have contained organic components, the problem of distinguishing organic material associated with rock-art production from other sources on the rock face remains. For example, McDonald et al. (1990) obtained inconsistent radiocarbon dates, ranging from 6085 b.p. to 29,795 b.p. from charcoal taken from the same motif at a rock-art site in the Sydney Basin, Australia. When this project was extended to include multiple sampled motifs from an additional three sites, all multiple sampled motifs returned age determinations that differ from each other by more than one standard deviation (McDonald in press: table 3).
A likely cause of this inconsistency is contamination by organic carbon from sources other than that associated with the painting event. The many potential sources of contamination include naturally occurring organisms on the rock face, such as algae, lichen, fungi, bacteria and others (see Watchman & Cole 1993; Ridges et al. in press). Remains of these can both predate and post-date the painting event and may be incorporated in unknown proportions in paint or in surface skins associated with paint. Contamination can also occur from deposits of inorganic carbonates, sometimes present as evaporites on rock surfaces, and from charcoal-bearing dust on the rock surface. Contamination can even be introduced during sample extraction and preparation (Armitage et al. in press; Campbell in press; Hotchkis et al. …