Title: Trance Performance: The Rock Art of Boontjieskloof and Sevilla
Abstract: A sample of paintings from the south-western Cape was analysed for evidence of trance symbolism. These symbols are found to exist, despite differences in appearance between the paintings in this sample and those from the Drakensberg. Such differences are deemed to be superficial. The implications of these interpretations are considered briefly with respect to two image types: 'group scenes' and 'conflicts'. * Manuscript received July 1985, revised September 1985 Introduction The rock paintings of the south-westem Cape have been partially described many times, and in recent years several research projects have recorded site numbers, distributions and contents from particular regions of this wider area (Maggs 1967a; Manhire 1981; Janssen van Rijssen 1981; Halkett 1982). Recent explanatory models for southern African rock art have, however, been derived from the paintings of the south-eastern escarpment. Here, Patricia Vinnicombe stressed the central, and mystical, role of the eland in the paintings and ethnography, basing her interpretations on extensive field recording (Vinnicombe 1972a, 1972b, 1975, 1976). David Lewis-Williams developed this anthropological paradigm further, arguing that a number of metaphors for trance experience and related symbolic concepts were central to understanding the meaning of the art (Lewis-Williams 1980, 1981a, 1981b, 1982, 1983a, 1983b). Apart from Maggs and Sealy's (1983) exploration of a specific trance theme there has been no attempt to take Lewis-Williams's explanatory hypotheses from the context of the Drakensberg and systematically test their value in the south-western Cape. Such an exercise has more than parochial value. An important part of LewisWilliams's exposition is the re-statement of McCall's (1970) suggestion that a 'Pan-San cognitive system' exists over much of southern Africa (Lewis-Williams & Biesele 1978; Lewis-Williams 1980, 1981a). Pan San beliefs are said to include a set of concepts and perceptions of the world that unify all rock art, mythology and the ritual and trance experiences of prehistoric and contemporary San. For example, identification of the same sets of key concepts suggests that rock painting in Tanzania (Leakey 1983) is related to trance performance known from far to the south (Lewis-Williams 1984a:229, and in press). Similarly, it is inferred that such conceptual systems are of some antiquity, as evidenced by art mobilier dated to at least 2 000 and perhaps 20 000 years before the present (LewisWilliams 1984a). In effect, the trance hypothesis frees rock painting interpretation from the constraints of both time and space. In reply to criticism from Willcox (1983) on this point, Lewis-Williams (1983:542-543) writes: the interpretation I have offered is not tied to any specific antelope . . . elsewhere, as one may expect, other animals are emphasized but the same key metaphors are depicted and these show that we are dealing with the same fundamental beliefs. In order to explore the generality of the trance hypothesis, we have analysed in detail a suite of rock paintings from the south-western Cape to establish whether or not trance-related images are present, thus following the example of Huffman's (1983) Zimbabwe study. Research Methodology Recording rock painting is part of a wider research programme, conducted by the Spatial Archaeology Research Unit (SARU) at the University of Cape Town, which is concerned with the prehistory of settlement pattems in the south-western Cape. Rather than selectively analysing sites chosen from this general area, we decided to concentrate intensively on two adjacent kloof systems, Boontjieskloof and Sevilla, located north-east of Clanwilliam, beyond the Pakhuis Pass (Fig. 1). These kloofs were chosen because of their known concentration of well preserved images (Johnson et al. 1959; Maggs 1967a, 1967b; Johnson & Maggs 1979; unpublished records). As Lewis-Williams has shown, the identification of trance metaphors often demands that detail is well preserved, thus particularly qualifying Boontjieskloof and Sevilla for the kind of study we have undertaken. This area was systematically searched and all archaeological occurrences were noted and plotted on 1:50 000 maps, while paintings were photographed and recorded by the method developed by Alfers (1982) for Project ARAL in Lesotho. This method entails tracing from photographic slides rather than the in situ image. The full results have been presented by Golson (1983) and the site records are housed in the Spatial Archaeology Research Unit. Although Lewis-Williams (1984b) has pointed out that the trance hypothesis does not stand or fall according to the frequency with which definite trance related scenes occur in any given area, total photographic recording is valuable for several reasons. First, it allows re-evaluation of the complete set of paintings at any time, so avoiding the selection of a sample on the basis of cursory field analysis. Second, and following from the first, the ARAL technique provides a relatively unbiased, complete and detailed permanent record; this is particularly important in this area because of the partial inundation of a few sites by dam construction in Boontjieskloof since our study was completed. Thirdly, photography reduces the amount of costly field time against a field tracing programme of equal intensity. The selection of paintings presented in this paper was made from the 83 sites discovered, which together contain a total of 3 398 images. The process of tracing and analysis has led to the realization that many paintings in the south-westem Cape are remnants. This implies more than the obvious fact that a proportion of all the paintings has survived; of greater importance are indications that individual images may be differentially preserved. Common examples are human figures with the head form usually termed 'hookhead', and paintings of eland. In a few well preserved examples it can be seen that the 'face' of the human figures, and the neck and legs of the eland, are painted in whites or yellows. Because of the kinds of minerals that comprise the paints, these colours are fugitive, or relatively impermanent, in contrast to the darker, lasting iron oxides which form the remainder of the image. These more durable portions are usually all that is left of eland and hookheads. From extensive field observation, we know that generally less than 15% of eland and hookheads survive in clear bichrome in the southwestern Cape, and this provides some indication of the general rate of survival of these pigment types, irrespective of context. In the case of eland and hookheads, the missing components are easy to infer, but this is not always the case for other images. In some examples, subtle detail was provided by the addition of white pigment and, if missing, this could not be inferred. Pigment degradation is an equally important factor where whole images may have been lost, as is clearly the case at SEV 14 (Fig. 5) and SEV 1, or Veg 'n Vlug (Fig. 10 and Johnson & Maggs 1979: plate 67). This has profound implications for rock painting analysis, for a casual glance through Lewis-Williams's illustrations will show that a considerable amount of detail contributory to his interpretations is in whites and yellows. However, this paper attempts to show that, while the presence of such details may facilitate interpretation, their absence and other representational differences do not mask the presence of certain key metaphors and the basic similarities between rock paintings in the south-western Cape and those described in the Drakensberg and elswhere. S. Afr. archaeol. Bull. 40: 70-80. 1985 This content downloaded from 207.46.13.129 on Sun, 26 Jun 2016 07:38:26 UTC All use subject to http://about.jstor.org/terms The South African Archaeological Bulletin 71
Publication Year: 1985
Publication Date: 1985-12-01
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
Cited By Count: 61
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot