Determining Style in Palaeolithic Cave Art: A New Method Derived from Horse Images
Pigeaud, Romain, Antiquity
The dating of the art displayed in the cave discovered at Chauvet (Ardeche) back to 32 000 years ago, and therefore into the Aurignacian, has stirred up research into Palaeolithic cave art (Clottes 1998; Valladas et al. 2001). Painted caves were previously, for the most part, classified according to a stylistic chronology devised by Andre Leroi-Gourhan (1965), which would place the Chauvet paintings in the Solutrean period (between 22 000 and 18 000 years ago). Although some scholars still have reservations (Zuchner 1999; Pettit & Bahn, 2003), it appears reasonable to accept this new dating and to put aside an evolutionary stylistic chronological scheme. The implication is that Palaeolithic art is a dendritic phenomenon, which did not follow a linear path, evolving from a clumsy image to a naturalistic chef d'oeuvre. As Peter Ucko (1987) has suggested, it is possible to envisage the existence of several contemporary styles, be they due to differences in technical or artistic ability or to regional variations.
In this case, should we abandon stylistic analysis? This question has provoked a lively debate amongst researchers (Pettit & Bahn 2003; Lorblanchet & Bahn 1993; Otte 1997; Vialou 1999; Otte & Remacle 2000; Lorblanchet & Bahn, 2003). It soon became apparent that only a minority of painted caves could be dated, either directly or through stratigraphic relationships. Style therefore seems to be an inescapable tool, if one is to insert painted caves into a chronological framework of Palaeolithic cultures. One condition, however, is that an agreement can be reached on the meaning of 'style'. For James Sackett (1970: 370), style '(...) concerns a highly specific and characteristic manner of doing something, and that this manner is always peculiar to a specific time and place'. In other words, style is not what unites, but what differentiates. Choosing very broad stylistic criteria allows us to bring together a large number of sites, but this exercise leads nowhere. One such broad criterion, for example, is the one known as the 'duck's bill', a stylised representation of a horse's nose whose elongated oval shape recalls that of a duck's bill (Figure 1). Since the time of Henri Breuil (Capitan, Breuil & Peyrony 1910), the 'duck's bill' has been considered an archaic trait and it has been used to date certain caves and some portable artefacts to early phases of the Upper Palaeolithic. However, there are figures of horses displaying 'duck's bills' on sites dated well into the Magdalenian period, for example on plaques from Gonnersdorf in Germany (Bosinski & Fischer 1980). The 'duck's bill' is therefore a particular way of stylising the end of a horse's nose, but is not specific to a site, nor attributable to a particular chronological period. On its own, it is not a useful stylistic criterion and must be used with great caution.
[FIGURE 1 OMITTED]
Other supposedly 'stylistic' criteria, on the other hand, are nothing of the sort. The tail is placed low, for example, on many representations of Palaeolithic horses, such as those of the cave of the Trinidad de Ardales in Andalusia (Spain) (Cantalejo Duarte et al. 1997). This is surely an anatomical error, as a real horse's tail starts right above the anus. Of course, it is possible that the artist deliberately took some liberties with anatomical reality. I contend that the 'low tail attachment' criterion is not a reliable stylistic criterion, as the real intention of the artist--simple error or artistic licence--will always remain unresolved. To compare sites on this basis alone would be a risky undertaking.
The problem of these formal coincidences is particularly acute in the period that precedes the Magdalenian. My working hypothesis is that there is a fundamental break between the Magdalenian (c. 17000-9000) and the cultural phases that came before (Aurignacian, Gravettian, Epigravettian & Solutrean). …