Printer Friendly

Rock art and changing perceptions of southern Africa's past: Ezeljagdspoort reviewed.

Africa 48: 117-34.

LEWIS-WILLIAMS, J.D. & T.A. DOWSON. 1988. Signs of all times: entoptic phenomena in Upper Palaeolithic art, Current Anthropology 29: 201-45.

1989a. Images of power: understanding Bushman rock art. Johannesburg: Southern Book Publishers.

1989b. Theory and data: a brief critique of A. Marshack's research methods and position on Upper Palaeolithic shamanism, Rock Art Research 6: 38-53. Since copies of Bushman (San) rock paintings were first published just over a century and a half ago, approaches to the art have undergone significant changes: at different times, different aspects have been singled out for comment and interpretation. These changes did not happen randomly. Rock-art research, like any other academic pursuit, is historically situated: changes in research strategies and explanations are implicated in ideology and politics. This paper therefore investigates the kind of knowledge rock-art research produces and its relevance to contemporary thought.(1)

In South Africa, knowledge of indigenous peoples has been appropriated and used in political manoeuvres (Smith 1983; Parkington & Smith 1986; Wright & Mazel 1987; Hall 1988). In educational propaganda the Bushmen have long been presented as simple, helpless nomads, wandering aimlessly across a landscape they did not own. Because they are said to be inherently 'primitive', their artists are described as naive chroniclers of daily events. So conceived, their stature as human beings capable of creative thought and controlling their own destiny is obscured.

Today a new colonial ideology is building on this picture of primitiveness by holding them up as exemplars of natural ecologists. They are now said to live 'in harmony with nature', using but not destroying the bounty of their environment. Along with this emphasis on admirable people-nature relationships there is a concomitant portrayal of the Bushmen's supposed child-like, mystical wisdom, spiritual insight and almost supernatural powers. As presented by writers such as Laurens van der Post, they have been made to speak as if from an innocent Garden of Eden; Jungian archetypes come easily and platitudinously. Although this new ideology appears to exalt them, it is actually presenting them as an indissoluble part of primeval circumstances, a mystical people unable to look after themselves in the materialistic modern world. The spiritualizing of the Bushmen is simply a more palatable way of rendering the reality of their immediate day-to-day needs invisible and marginalizing them in political debates (Voss 1987; Barnard 1989; Biesele & Weinberg 1990; Marshall & Ritchie 1984; Mazel 1992).

Recent rock-art research has explicitly challenged both the primitive and the other-worldly stereotypes by revealing the complexity and subtlety of Bushman thought and art (cf. Lewis-Williams 1989; 1990). The rock paintings and engravings do not record a child-like interest in the events of nomadic life; nor are they reflections of an innate, romantic empathy with 'nature'. This research has been based on the copious 19th- and 20th-century Bushman ethnography. The complex and detailed fit between the ethnography from both centuries and specific features of the art confirms its relevance and suggests that the rock paintings and engravings can be explained, at least in part, in terms of widely held Bushman beliefs and widely practised rituals (Lewis-Williams & Biesele 1978; Lewis-Williams 1972; 1980; 1981a; Vinnicombe 1972a; 1976; Dowson & Deacon in press). The centrality of the ethnography in any study of the 'meaning' of Bushman rock art is now accepted by virtually all researchers.

As a contribution to rock-art research's challenge to demeaning stereotypes we consider a single rock painting in some detail. We begin by examining straightforward narrative views of the painting. We then show that ethnographic research can reveal otherwise unsuspected complexities. Finally, we argue that the results of neuropsychological research into altered states of consciousness, carefully situated in the specific historical circumstances of the Bushmen, can lead us still further towards a better grasp of what they achieved in their remarkable art and so challenge some current ideas about them.

Narrative views

In 1835 Sir James Alexander visited the Langkloof, a long, broad valley near present-day Oudtshoorn, to see 'drawings executed by the former occupiers of the country' (Alexander 1837: 314, 315). He was surprised to find that 'these rude attempts of uncivilised artists are not utterly devoid of merit'. When he came to comment on the paintings individually, he accepted each as a depiction of 'real life' and wrote about the use of the bow and arrow and, more elaborately, about one painting which he thought depicted 'an embassy of females suing for peace'. He went on to assert, 'No one can deny that their reception is a gracious one, to judge by the polite attitudes of the male figures, perhaps chiefs' (Alexander 1837: 316). Although the sex of none of the figures is clearly shown, Alexander saw one group as women, the other as men: in accordance with 19th-century colonial perceptions of gender, the women are subservient suppliants, the men gracious, polite chiefs.

But Alexander's approach ran into trouble when he saw the strange painting of apparently fish-tailed figures at Ezeljagdspoort (FIGURES 1 & 2; for colour photographs see PLATE II, page 222, this number; also Willcox 1963: plate 11 and Lewis-Williams & Dowson 1989a: figure 87e). He had to admit, 'We are unable to assist the reader, even by a conjecture, in elucidating the meaning of that which he here sees represented' (Alexander 1837: 317). Yet he went on to suggest that the painting may allude to 'the amphibious nature attributed to the whites by the natives in the olden day'.

Despite his bafflement, Alexander tended to see the painting in terms of race. Since his time, southern African rock-art researchers have often tried to identify different races supposedly represented in the art and to interpret specific paintings as depictions of racial conflict (e.g. Dart 1925; Willcox 1984; Woodhouse 1987; for further comment on this trend see Posnansky 1985; Lewis-Williams 1989; 1990). By the middle of the 20th century racial conflict had become the key tenet of a widely accepted 'stylistic' classification of the art (Lewis-Williams 1990). Researchers proposed that the art could be divided into two major periods, an earlier 'pre-Bantu' period comprising peaceful paintings and a later 'post-Bantu' period comprising restless paintings and scenes of conflict (van Riet Lowe 1952: 7; today 'Bantu' is used in a linguistic sense only.). Some rock-art research has thus presented the black farmers as intrusive disturbers of the Bushmen's Garden of Eden and, in doing so, supported the colonial segregationist ideology.

At the same time, the paintings themselves were seriously misrepresented. Because researchers were interested mostly in 'simple' paintings that seemed to fit their conception of the art as narrative, they overlooked puzzling features that did not appear to 'tell a story'. This can be seen in the copies Major C. Michell made for Alexander. In dealing with the Ezeljagdspoort group, he concentrated on getting the 'amphibious' aspects right, but he omitted other details such as the fingers at the ends of the elongated figure's arms. He also left out the two straight lines at this figure's shoulder, and he rather confused the undulating lines. Because these features were less comprehensible than the supposed fish-tails, which he could understand in terms of Western stories about mermaids and myths about people coming out of the sea, he did not take as much care with them. Far from being 'objective', his copy of the painting was skewed by his limited understanding of the Bushmen and their art.

When writers who see the art as principally narrative confront a patently 'non-real' painting like this they frequently identify it, as Alexander did, as a 'scene' from a myth rather than from daily life and so preserve their position, albeit in a broader sense. Indeed, a mythical interpretation of the Ezeljagdspoort painting that has become widely accepted seems to be strongly suggested by a 'fine old legend' that Wilhelm Bleek and Lucy Lloyd obtained from a Mr D. Ballot who had, in the 1870s, noted it down from Afrikaander, 'a very old Bushman still surviving in those parts' (Bleek 1875: 20). It tells of a young girl who was dragged into a dark pool by water-maidens (watermeide). When her mother found what had happened to her, she collected some herbs and ground them to a powder. She then scattered the powder on the surface of the water. As a result, the girl came out of the pool unharmed, though the water-maidens had licked her cheeks white because they loved her so much (Leeuwenburg 1970). (For more on the 12,000-page Bleek and Lloyd collection of Bushman ethnography see Lewis-Williams 1981a: 25-37; Deacon 1986; Dowson & Deacon in press.)

Leeuwenburg has argued that this legend, together with Afrikaander's claim to know many other stories about water-maidens, 'adds weight to the theory that the hunters had an extensive fund of folk-tales, and that pictorial representations of these tales can be found amongst the paintings' (Leeuwenburg 1970: 146). In his view, the legend shows that even 'non-real' paintings can be subsumed under the narrative rubric.

All too easily this kind of approach asserts rather than explicates associations between depictions and 'tales'. Ezeljagdspoort is no exception. A close reading of Bleek's (1875: 20) report shows that Afrikaander explained only the 'subject' of the painting. This is why he ignored the curious elongated figure and the 'fish-tails', the feature in which Alexander, Bleek, Lloyd, Ballot and others were most interested. Indeed, the manuscript of the tale does not mention the painting at all; on the contrary, Ballot states explicitly that the legend was obtained in response to the question, 'Do you believe in Watermeide?' Ballot then linked the legend to the Ezeljagdspoort depiction. Bleek (1874: 13), who, as a result of Orpen's (1874) work on the Bushman folk-tales and rock art of the southeastern mountains, believed that some paintings illustrated 'Bushman mythology', readily accepted Ballot's association. We, on the other hand, believe that Afrikaander may never have seen the painting and that there are no compelling reasons for linking his legend to it.

Ritual and art

Right or wrong about a connection between the legend and the Ezeljagdspoort painting, Bleek (1875: 13) realised, as long as 120 years ago, that Bushman rock art could be understood only from within, and not by taking an outsider's view and describing it in Western terms. Although this principle was lost sight of after his death, it later became central to southern African rock-art research. When, in the 1960s, rock-art researchers (e.g. Maggs 1967; Vinnicombe 1967; 1972a; 1972b; Lewis-Williams 1972; 1974b) tried to follow the quantitative techniques then being explored by 'mainstream' archaeology, they discovered that the numbers remained silent and that 'laws' were not forthcoming. While many archaeologists persevered with the admittedly exciting promise of the New Archaeology, some rock-art workers adopted a different position and sought what has since become known as an 'insider's view' (Lewis-Williams 1972; 1974a; 1974b; Vinnicombe 1972a; 1972b; 1975; 1976).

In southern Africa that view could be realised through 19th- and 20th-century Bushman ethnography. But it was soon recognized that the ethnography is 'fragmentary' (Lewis-Williams 1972: 64): extensive as it is, it is, of course, not a complete inventory of Bushman life and belief. Moreover, the ethnography provides no simple, straightforward explanations of the art. Bushman ethnography, like the art itself, is a text that requires close, critical reading and interpretation: the art does not 'illustrate' the ethnography, nor does the ethnography 'explain' the art in any direct sense. Both contain, in their own ways, metaphors that lay at the heart of Bushman belief (Lewis-Williams 1980). Meaning is created every time a researcher, in an effort to plumb these metaphors, interacts with the ethnography on the one hand and the art on the other, and then attempts to bring the two texts into mutually illuminating conjunction. Interaction between researcher and text, being always historically situated and specific, is as much an artefact of social circumstances as of data. These caveats, now becoming widely accepted in archaeology, are entered here to show that we are aware that our own interpretations of rock paintings are as complexly constituted as, say, Alexander's; the difference between our position and Alexander's is that we recognize at least some of the social and ideological factors bearing upon our work. Acknowledging these predispositions, we now explore changing ethnographic understandings of the Ezeljagdspoort 'mermaid' painting.

An initial ethnographic account was made possible by a Bushman's comment on a copy of the painting that Ballot made in the 1870s and that was then 'recopied from his sketches by Mr H.Ch. Schunke' (note on folder in South African Library). Like Michell's copy, this one was also made in the narrative, literalist tradition. Again, the 'fish-tails' are fairly accurate, but the copy omits entirely the puzzling undulating lines, the dots and the two straight lines at the elongated figure's shoulder; it does, however, show this figure's fingers. In 1878, three years after Bleek's death, Lloyd showed it to /Han=kass'o, a /Xam Bushman who came from the Kenhardt area (for further details of the informant's home territory, see Deacon 1986) and who was familiar with at least some aspects of rock art (Lewis-Williams 1981a: 30-31). Comments like his are so rare and valuable that each demands the closest possible analysis (Vinnicombe 1976: 314); a superficial reading can easily be misleading and fail to detect all-important metaphors (Lewis-Williams 1980; 1988).

/Han=kass'o's observations, as recorded by Lloyd (MS pp. L.VIII.1.6063-6068) and with her parentheses, run as follows:

I think that the rain's navel is that which goes (along here). I think that these people, they address the rain that the rain's navel may not kill them, that the rain's navel may be favourable towards them. That the rain's navel may not kill them. That the rain's navel might keep favourable towards them. This man he has hold of a thing which resembles a stick. I think that they are rain's people. I do not know them, for I behold that they are people. For, they have their arms: they resemble people. They feel that they are sorcerers, the rain's sorcerers they are: for this man is holding a thing which resembles /khoe (a curved stick used in making a Bushman house). I do not know whether it is a /khoe, for I see the thing resembles a /khoe. These people (i.e. those on the lower side of the line in the picture) I do not know whether the rain's navel divides them from the other people. People (they) are, sorcerers, rain's sorcerers. They make the rain to fall and the rain's clouds come out on account of them. Hence the rain falls, .... and the place becomes green on account of it. This thing (i.e. what we should have called the right arm of the rain figure), it is the one which resembles a caterpillar, the rain's caterpillar.

Unfortunately, Lloyd did not record the questions she put to /Han=kass'o, so we do not know how she may have directed his attention. We may, however, be fairly confident that she would have asked him about Afrikaander's watermeide legend. That he did not touch on this theme in any way suggests that his remarks can probably be seen as a denial of the watermeide interpretation. Instead he said, first, that the painting, in general and despite details he did not understand, had something to do with weather control and, secondly, that the figures were 'rain's sorcerers', that is shamans whose task it was to make rain and to protect people from thunderstorms (Bleek 1933a; 1933b; Lewis-Williams 1981a: 103-16). Parts of his explanation are enigmatic and require detailed exegesis. (For more on the essentially shamanistic nature of Bushman rock art throughout southern Africa see e.g. Huffman 1983; Lewis-Williams 1983a; 1983b; 1984; Maggs & Sealy 1983; Yates et al. 1985; Garlake 1987; Deacon 1988; Dowson 1992; Hall 1986; 1990; Manhire et al. 1986; Parkington 1989; Kinahan 1989; Mazel 1989; Lewis-Williams & Dowson 1989a.)

In the first remark Lloyd recorded, probably a response to a question from her, /Han=kass'o identified what we see as a greatly attenuated human figure as 'the rain's navel', saying 'that which goes (along here)'. His use of 'goes' and Lloyd's parenthesis suggest he was actually pointing to the long, curving line; he did not use the word 'man' as he did when he referred to other figures in the group. Furthermore, at the end of his statement he said, 'This thing (i.e. what we should have called the right arm of the rain figure), it is the one which resembles a caterpillar, the rain's caterpillar'. (A few pages later he explained that the 'rain's caterpillar' was large and poisonous and that it came out of the water.) If he thought that an arm of the elongated figure was a caterpillar, he clearly did not perceive the image in any way as human. When he used the phrase 'the rain's navel', he was thus referring only to the line without realizing it was an attenuated human figure. All his subsequent remarks must be read in the light of his perception of the curved figure.

The linear form of this depiction suggests that 'umbilical cord' would be a more apposite phrase, but Lloyd gives all the few recorded uses of the /Xam word (! hain) as 'navel'. 'The rain's navel' is in fact one of a set of anatomical metaphors that included 'the rain's legs', 'the rain's hair' and 'the rain's blood' (Lewis-Williams 1981a: 104). To clarify /Han=kass'o's metaphor we note an association between the human navel and the work of rain-shamans. In a statement given independently of the Ezeljagdspoort painting, Dia!kwain, another of Bleek and Lloyd's /Xam informants, said that people wishing to protect their camp from an 'angry' rain (a thunderstorm) do three things.

First, they stand in front of their dwellings, facing the direction from which storms come, and 'strike their (?navels) with their fists' and 'press their hand in their ?navel' (L.V.3.4993-4994; Lloyd's parenthesis). Lloyd placed a question-mark before 'navel' here, but the manuscript shows that it is the same /Xam word that /Han=kass'o used. Almost certainly Dia!kwain would have assisted with the translation by a gesture, and the word must, at the very least, have something to do with the general area of the stomach. On the other hand, Lloyd's question-mark suggests that 'navel' is not an exact equivalent of the /Xam word: if the word meant no more than 'navel', it would have been easy enough for him to indicate this by pointing. It is in fact possible that the /Xam word meant the stomach and diaphragm, not just the navel itself, because this part of the body, known to the modern Kalahari !Kung Bushmen as the g//abesi, is believed to be the seat of a shaman's potency and the place where it starts to 'boil' when he enters trance (Katz 1982). As the 'boiling' increases, the stomach muscles contract painfully, and the shaman bends forward (Marshall 1969: 363). It is therefore significant that the large, central 'fish-tailed' figure in the Ezeljagdspoort group is bending forward (for more paintings of forward-bending figures see Lewis-Williams & Dowson 1989a: figures 15, 20, 28, 29). The striking of the navel and the pressing of a hand into the stomach in the face of a thunderstorm may thus have been a ritual dramatization of some of the physical effects of trance: people performed these actions to evoke the power of trance experience as they tried to control the storm.

Secondly, Dia!kwain said that the shamans of the rain 'snap their fingers at the rain' while they are facing the threatening storm and pressing 'their hand in their ?navel' (L.V.3.4995). Snapping the fingers, or simply pointing, was a way of shooting supernatural potency into a person or an animal (Marshall 1969: 351-2; Katz 1982: 46, 168, 263; Lewis-Williams 1986b; for more on supernatural potency see Marshall 1969; Biesele 1978; Lewis-Williams 1981a; Katz 1982). Finger-snapping was therefore related to the striking of the navel in that both gestures were involved in ritual protection from storms.

Thirdly, Dia!kwain said that they apostrophized the rain: 'Thou shall falling turn back ... for thou dost not a little lighten, for it seems as if thou art very angry.' This is exactly what /Han=kass'o said when he commented on the figures in the Ezeljagdspoort painting: 'They address the rain that the rain's navel may not kill them, that the rain's navel may be favourable towards them' (L.VIII.1.6063-6064).

In sum, Dia!kwain's statements and !Kung beliefs about the g//abesi show that the 'navel' was probably associated, in ways that we do not fully understand, with supernatural power and shamanistic altered states of consciousness. Thus, when /Han=kass'o pointed to 'the rain's navel' in the picture, he was thinking, at least in part, of the 'angry' rain from which shamans of the rain protected people by shooting their power at it and apostrophizing it. He was saying that what he saw as a long, curving line (not as an attenuated human figure) represented the power of the rain: 'the rain's navel' was, in some respects, a metaphorical way of talking about that power. This reading of /Han=kass'o's phrase clearly accords well with the independent interpretation of the so-called 'thin red line' as a way of depicting supernatural power (Lewis-Williams 1981b), though the line can also be seen as a transformed shaman (Dowson 1988a).

Be that as it may, the most important point arising from /Han=kass'o's comment on the painting is that he recognized it as depicting the activities of shamans, despite his inability to provide an interpretation of all its various features. These activities go some way towards clarifying a number of features, including the two straight lines at the elongated figure's shoulder. Schunke's version of Ballot's copy omits these lines altogether. Wilman, who copied the painting in about 1917, seems to have mistaken the lines for the legs of another but partially faded elongated figure. Both Ballot and Wilman were puzzled by these lines, but ethnographic evidence and comparisons with other paintings suggest that they depict sticks. Two other figures in the group are holding similar pairs of sticks. When Bushman trancers bend forward acutely, they sometimes support their weight on two dancing-sticks (Wilmsen pers. comm.), a posture observed long ago by Burchell (1822: 63) and often depicted in rock paintings where the sticks are sometimes meticulously decorated with small red marks. Today !Kung healers do not seem to attach any special significance to dancing-sticks, though Marshall (pers. comm.) noticed that they sometimes wave them over their patients. An early-19th-century report, however, describes Bushman women placing 'two bits of reed across each other' on the back of a man who had collapsed during what appears to have been a curing, or trance, dance (Arbousset & Daumas 1846: 247). Using ostrich feathers, they then wiped the sweat off him, all the while 'leaping backwards and forwards across his back'. Their manipulation of two pieces of reed in cases of severe sickness and their own statement that their use 'exerts a salutary influence on the sick person' led Arbousset & Daumas to suggest that the use of reeds was part of a 'religious rite'. Noting their observations and the frequent depiction of two sticks in paintings of dances, Vinnicombe (1976: 314) concluded that they 'were connected with protective and curing medicine'. The ethnographic and painted evidence thus suggests that the sticks depicted in the Ezeljagdspoort painting, though meaningless to the early copyists, were highly significant for the artist and original viewers.

The positions of the pairs of sticks in the painting are also significant. Two pairs are held as if by dancers, but the pair Schunke omitted and Wilman misinterpreted are closely associated with the elongated figure's shoulder. Held sticks were in some general sense intelligible to the copyists, but their detached position next to the elongated figure meant nothing to them. The sticks, as well as one undulating line, are, however, close to the figure's n//ao, a 'hole' at the nape of the neck and between the shoulder-blades from which !Kung shamans say they expel the sickness they have removed from people (Biesele 1978: 929). This expelled sickness can be 'seen' by shamans but not by ordinary people (Biesele 1978: 929; England 1968: 427). Numerous trance figures, especially trance-buck (e.g. Lewis-Williams 1981a: figure 33e), have lines or dots emanating from this area, and these have been interpreted as part of the shaman-artists' perception of spiritual reality (Lewis-Williams 1981a: 93-101). In the light of /Han=kass'o's statement that the painting depicts rain-shamans, it is highly significant that the n//ao is closely associated with n!ao, a power that changes the weather and, in certain circumstances, brings rain (Biesele 1978; 1992; Marshall 1957). The placing of a pair of sticks near the elongated figure's n//ao may thus have held some special significance for the shaman-artist that linked trance dancing with n!ao and weather control. This interpretation of the sticks thus accords well with /Han=kass'o's statement that the painting depicts 'rain's sorcerers'.

All the ethnographic evidence we have adduced so far points towards the Ezeljagdspoort painting being in some way associated with the activities, beliefs and metaphors of Bushman rain-shamans, but it still leaves the matter of the curious 'fish-tails' unexplained. A decade and a half ago, Lewis-Williams (1977: 167) argued that /Han=kass'o's failure to comment on the ichthyoid feature, together with the similar appearance of some other paintings in the same region (Lewis-Williams 1977: figure 2), suggests that 'the mermaid effect has been created by poor draughtsmanship rather than an intention to depict human beings with ichthyoid tails'. Since then our increasing understanding of Bushman shamanism, metaphors and the rock art itself has led us to accept the draughtsmanship of these paintings and to advocate a more comprehensive interpretation. The route to this more persuasive explanation is via an until recently unexplored field of research.

Neuropsychology

Once the ethnographic and painted evidence for believing Bushman rock art to be essentially shamanistic had been accepted, research turned to neuropsychological studies of altered states of consciousness (e.g. Thackeray et al. 1981; Maggs & Sealy 1983; Lewis-Williams 1986a; Lewis-Williams & Dowson 1988; 1989b; Dowson 1989; 1992). Because the nervous system is a human universal and because the visual, somatic, aural, gustatory and olfactory hallucinations it produces are very similar in form whether they are induced by hallucinogens, hyperventilation, sensory deprivation and so forth, neuropsychology can give some idea of the structure and nature of hallucinations that is independent of cultural circumstances (e.g. Eichmeier & Hofer 1974; Siegel & Jarvik 1975; Reichel-Dolmatoff 1969; 1978; Lewis-Williams & Dowson 1988). On the other hand, much of the content of hallucinations is culturally controlled and therefore varies from society to society. Neuropsychological research may explain the origin of some forms, but it cannot explain the meanings and associations that these hallucinations had for those who experienced them (Dowson 1989). The meanings attached to hallucinations and the social function of altered states of consciousness are always culture-specific and historically situated. Our recourse to neuropsychology should not be seen as the reduction of all rock art to universal constants of any sort. Culture is always paramount. As we now show, neuropsychology and ethnography facilitate different but complementary approaches to interpretation.

Visual hallucinations

The earliest copy of the Ezeljagdspoort group, the one made by Michell, shows an undulating line curving round the top of the elongated figure. Schunke's version of Ballot's copy omits all the lines. Wilman combined one of the undulating lines with the two sticks associated with the elongated figure's n//ao so that they resemble the faded remains of a second elongated figure. Johnson's copy omits the undulating lines altogether and has the two sticks lower down the elongated figure's right (?) arm. In contrast to these copies, a recent tracing by Dowson shows two undulating lines and at least four dots. Acknowledging ever-present human error, we believe this copy to be more accurate than the earlier ones.

As we have argued, the reason why previous copyists omitted the undulating lines and dots is that they do not fit an essentially narrative view of the art. Today we have a better grasp of the 'non-real' elements of Bushman rock art, and it is these that explain the undulating lines and dots. In an early stage of trance, subjects 'see' luminous, pulsating geometric forms known as entoptic phenomena, form constants or phosphenes (e.g. Siegel & Jarvik 1975; Kluver 1942; Knoll & Kugler 1959; Horowitz 1975; Oster 1970; Eichmeier & Hofer 1974). The forms include zig-zags, grids, nested catenary curves, filigrees and, significantly, undulating lines and dots. Because these entoptic phenomena derive from the human nervous system, they may be experienced by all people who enter a trance state, no matter what their cultural background. In a deeper stage of trance entoptic elements combine with culturally determined hallucinations of animals, people, monsters and so forth.

Because we know from the ethnography that Bushman rock art is essentially shamanistic, it seems highly probable that the geometric motifs that are common among the rock engravings (Dowson 1992) but which are also found among the paintings derive from entoptic experiences. In some paintings deriving from deeper trance the entoptic elements are very prominent (e.g. Dowson & Holliday 1989); in other panels they have been more subtly blended with iconic elements (e.g. Dowson 1989). Indeed, numerous paintings throughout southern Africa combine zig-zags, dots, U-shapes and other entoptic forms with people and animals. The undulating lines and dots in the Ezeljagdspoort painting are thus consistent with deeper trance visions (Lewis-Williams & Dowson 1988). Just what shaman-artists thought about them is, of course, another matter. Perhaps they thought the lines and dots were supernatural potency, as they seem to have thought about dots and zig-zags in other paintings (see examples in Dowson 1989; Dowson & Holliday 1989).

Somatic hallucinations

Neuropsychology also informs on somatic hallucinations that involve alterations to a subject's perception of his or her body. Sometimes these alterations are radical, even to the extreme of dismemberment (Eliade 1972). In other instances they take the form of prickling sensations or the feeling that one has extra limbs and digits, a condition known as polymelia (e.g. Emboden 1979: 44).

One of the most common somatic hallucinations is associated with the sensation of rising up until one towers above and looks down on one's surroundings and one's body and limbs feel extremely attenuated. This feeling is one of the most frequently depicted hallucinations in Bushman rock art: shamans are often shown with long, thin bodies and limbs. One particularly clear example shows three such figures, each in a distinctive trancing posture: the middle one has its hands raised to shoulder height in a typical dancing position, and the one on the right has its arms in the backward posture that some !Kung Bushman shamans adopt when they are asking G//aoan (God) to put more potency into their bodies (Lewis-Williams 1986b). The figure on the left has its hands placed on its chest, a less well understood gesture that may be related to the prickling sensation Bushman shamans say they experience in their breastbones.

We suggest that this kind of somatic hallucination explains the curving, slightly undulating elongated human figure in the Ezeljagdspoort painting -- the one that /Han=kass'o did not realize was human and called 'the rain's navel'. Contrary to his interpretation, the figure is probably cognate with many others in southern Africa that depict trancing shamans experiencing the somatic hallucination of attenuation.

Metaphor and neuropsychology

One of the most useful results of the dual ethnographic and neuropsychological research programme has been the recognition that a number of Bushman metaphors originate in the strange and overwhelming sensations of trance experience, just as Westerners who have experienced certain altered states of consciousness resort to metaphors like 'high' (rising up and attenuation) and 'trip' (out-of-body experiences). Prominent among the Bushman metaphors for trance experience was 'under water' (Lewis-Williams 1980; 1988). Trance and submersion both involve feelings of weightlessness, affected vision, difficulty in breathing, the sense of being in another world, sounds in the ears and, finally, unconsciousness. Bushman shamans recounting their trance experiences therefore spoke of travelling under water (Biesele 1975; 1980) and diving into water-holes (Orpen 1874: 2, 10; Bleek 1935: 28, 32).

When individual shaman-artists came to paint their experiences, they explored the pictorial potential of trance metaphors, especially through animal behaviour (Dowson 1988b; Lewis-Williams 1981a; 1988). One particularly explicit painting shows fish associated with human figures, one of whose arms end in fish-tails. A combination of ethnographic, neuropsychological and painted evidence thus leads us towards accepting the Ezeljagdspoort 'fish-tails' as another depiction of the 'under water' experience.

Allowing that the Ezeljagdspoort figures may be ichthyanthropes, we argue that a more persuasive interpretation has been obscured because Alexander and all subsequent commentators set our minds running along amphibious lines. Our interpretation comes from neuropsychology, similar painted figures some of which are in the same region and Bushman beliefs about birds. We consider each of these three sources of evidence.

Neuropsychological and ethnographic research has shown that flight is another common (indeed, world-wide) metaphor for trance that arises from sensations of weightlessness, dissociation from one's body, journeys to distant places and changes in perspective that include looking down on one's surroundings (e.g. Barber 1971: 103; Siegel 1977). The Tukano shamans of the Amazon Basin, for instance, say that they fly on the winds and visit far-off places (Reichel-Dolmatoff 1978: 291).

Some paintings not far from Ezeljagdspoort take up this theme; they are very like the Ezeljagdspoort figures but they also have long arms curving back parallel to their sides. Another more recently recorded painting, in the adjacent Cango Kloof, shows 'diving' figures with aliform arms and forked tails. One of these has a conventionalized human head with the two antelope ears of a shaman's cap (Bleek 1935: 43-7; Lewis-Williams 1981a: 77). Such paintings, though not common, are widespread. One in the north-eastern Cape shows a fantastic bird standing on a man's head. In the Natal Drakensberg paintings show what appear to be human beings blended therianthropically with birds; these figures also have 'arms' in the backward trancing posture (Lewis-Williams 1981a: 88). Another brings together the 'flight' and 'underwater' metaphors by associating a bird and fish with trance dancers.

Do these paintings imply that Bushmen conceived of trance in terms of avian flight? To answer this question we return to the ethnography. One of Lloyd's informants said that Bushman shamans 'turn themselves into little birds' (Bleek 1956: 119). Dia!kwain also spoke of a shaman who turned himself into 'a little bird' that sometimes alighted on a person's head (Bleek 1935: 18-19). He did not say what species he meant by 'little bird', but in another, unpublished, statement he was more specific and said that the swallow was 'the rain's thing' (L.V.21.5698-5707). The word he used for swallow was !kwerri-/nan. The second part of this word is obscure, but the first part means to thunder or to strike with lightning. In /Xam idiom it was the rain (!khwa) that was said to strike (!kwerriten) (Bleek 1956: 465). Another /Xam word for swallow was /kabbi-ta-/khwa (Bleek 1956: 296). Here the first part of the word is obscure, but /khwa is another (perhaps erroneous) form of !khwa, rain (Bleek 1956: 296, 431); ta forms the possessive case. Thus whatever /kabbi and /nan may have meant, both words point to a close association between swallows and rain. Dia!kwain expressed this association when he said that swallows 'come when the rain clouds are in the sky' (L.V.21.5698-5699). In !Kung, swallows are g!ace'mhsi, 'children of the rain' (Biesele pers. comm.).

But the association goes further. Dia!kwain said that people should not throw stones (a sign of disrespect) at swallows because 'the swallow is with the things which the sorcerers take out |/ki, which also means 'to possess', as a shaman possesses potency~, which they send about. Those are the things which the swallow resembles' (L.V.21.5701-5702). A man who did not heed this injunction fell unconscious because 'the swallow had entered into him' (L.V.21.5706). Dia!kwain then clearly established the notion of shamanistic transformation into swallows by explaining that 'a sorcerer had come out of the swallow into him' (L.V.21.5706-5707). For the Bushmen, a swallow could be the embodiment of a shaman. The relevance of these beliefs to the Ezeljagdspoort painting becomes inescapable when we recall that /Han=kass'o identified the figures with forked tails as 'the rain's sorcerers' and added that they 'make the rain to fall and the rain's clouds come out on account of them' (L.VIII.1.6067-6068).

We therefore argue that the close links between shamans of the rain, swallows and clouds, should be seen together with statements about shamans turning into birds and paintings such as the ones in FIGURES 5 and 6. The ethnographic, neuropsychological and painted data thus combine to suggest very strongly that the Ezeljagdspoort group depicts not fish-people, as has generally been thought, but rather swallow-people, that is, shamans of the rain in the form of swallows, and that they are engaged in some sort of activity in the spiritual (hallucinatory) realm to protect people from the dangers of thunderstorms. The painting is thus a graphic realization of visual and somatic hallucinations that have been given content and meaning by the specific cultural and cognitive situation of the painter.

Beyond narration

So far we may have given the impression that our interpretations of the 'graphic realizations' of Bushman religion are in the narrative tradition. Whereas earlier writers saw the art as a narrative of daily or mythical events, we may seem to be arguing that the art is, in large measure, a narrative of hallucinatory experiences. In some ways this is true: the art does sometimes, though not always, record shaman-artists' visions. On the other hand, ethnographic and neuropsychological data together with features of the art itself show that both the paint and the rock face to which it was applied were significant (Lewis-Williams & Dowson 1990).

In the first place, the paint was more than prosaic colouring matter. The evidence for this comes from two ethnographic sources. In 1930 How (1962: 26-42) asked Mapote, an old man of Sotho and Pondomise descent who had learned to paint with Bushmen in their caves, to make some paintings for her. He said he would need qhang qhang, a glistening haematite, and that it would have to be prepared at full moon out of doors by a woman who would have to heat it over a fire until it was red-hot and then grind it to a powder. This pigment would then have to be mixed with the blood of a freshly killed eland. In the event, few of these conditions could be met, but Mapote none the less produced a few paintings for How. Some 50 years later an old woman of mixed Bushman descent then living in Transkei confirmed the use of eland blood as a medium (Jolly 1986; Lewis-Williams 1986c; Prins 1990; see also Ellenberger 1953). She explained that artists used eland blood to infuse their paint with supernatural power. Clearly, the preparation of paint was, at least in some circumstances, much more than a simple technological procedure.

Other ethnographic evidence suggests that the rock walls to which the paint was applied were also significant. A !Kung shaman said that his teacher, an older and more experienced shaman, had told him that he would enter the earth, that he 'would travel through the earth and then emerge at another place' (Biesele 1980: 56). This underground journey involved a passage through water, a belief recorded in the 19th century as well (Orpen 1874: 7; Bleek 1924: 17). These reports and indeed the world-wide occurrence of shamanistic subterranean and subaquatic journeys can probably be ascribed to the neuropsychologically determined experience of a vortex that engulfs subjects as they move into deep trance. Eventually, as they are drawn into the vortex, they feel dissociated from their bodies, and they become part of their own mental imagery (e.g. Siegel & Jarvik 1975: 128, 139; Siegel 1977: 134, 136).

The relevance of this experience to the rock art is suggested by the way in which paintings emerge from or enter into the rock face. Sometimes red lines fringed with white dots (Lewis-Williams 1981b; Dowson 1988a), depictions of therianthropes, fantastic eared and tusked serpents and also 'realistic' animals appear to enter and leave cracks or other inequalities in the rock face. In other instances depictions emerge from smears of paint. Many of these paintings incorporate elements that are clearly diagnostic of Bushman shamanism: a kneeling posture, arms held backwards, lines extending back from the nose that probably represent nasal haemorrhage, and fly-whisks (Lewis-Williams 1981a; Lewis-Williams & Dowson 1989a). For some southern Bushmen, the entrance to the tunnel, or vortex, seems to have been in the walls of rock shelters (Lewis-Williams & Dowson 1990).

The rock itself was like a veil suspended between this world and the spirit world: it was not a neutral tabula rasa on which artists could paint whatever they pleased. Rather, it was a significant and constraining context, in some ways like stained glass windows in a church: whatever appears in the windows, even apparently secular items such as animals or boats, are given, by their context, restricted symbolic meaning and heightened affective impact.

The active role which both rock and paint played in shamanistic rituals was suggested by the old Transkei woman of mixed Bushman descent to whom we have referred. She explained that, as shamans danced, they turned to face the paintings. It seems that potency went from an eland, via its blood, into the paintings and then, in other ritual circumstances, it flowed from the paintings into the shamans.

Something of what trancing Bushman shamans experienced as, seeking potency, they turned to the rupestral veil is suggested by neuropsychological research. A hallucinating Western subject, looking at a picture of a waterfall, said that he saw 'a remarkable brilliance' and a 'three-dimensionality'. 'The picture appeared to have a depth found in stereoscopic pictures' (Cohen 1964: 133). Other subjects have reported similar changes in depth relations and the way in which two-dimensional objects are seen in three dimensions (Kieffer & Moritz 1968). At the same time, hallucinations are projected on to walls and ceilings (Kluver 1926: 505, 506; Knoll et al. 1963: 208; Hofer & Osmond 1967: 13). Significantly, Szuman (cited by Siegel & Jarvik 1975: 109) described this experience as 'pictures painted before your imagination', and Siegel (1977: 134) likened it to 'a motion picture or a slide show'. These reports suggest that, as trancing Bushman shamans looked at the painted walls of rock shelters, their own hallucinations would have been projected on to and would have mixed with the paintings, themselves brilliant, animated and three-dimensional in their trance vision. Then, like some of the painted visions, the shamans would have penetrated the veil on their way to the spirit world.

The paintings were not merely a record or narrative of past religious experience, depictions of things that existed elsewhere or at another time, as a photograph represents a person or object that has its reality somewhere else. The paintings were things in themselves with their own power and life and which played an active role in shamanistic rituals. Probably, for trancing shamans, paintings were arrested visions with which they could articulate their own experiences. Each ritual performance gathered up past visions reified on the rock face and, combining them with visions of the moment, created a new experience: the paintings were, in this sense, living, powerful and active earnests of spiritual transcendence.

The Ezeljagdspoort painting was thus much more than a narrative record of a vision of shamans protecting people from the dangers of thunderstorms. It probably played an active role in the rituals of hallucination: trancing shamans were drawn into and through it as they journeyed to the spirit world to battle with the weather and the powers that threatened human existence. Through its immanent potency it both triggered and informed spiritual experience.

Conclusion

We have not explored all the possible connotations and implications of the Ezeljagdspoort painting. Shamanism is a highly complex phenomenon that permeates every part of a society and every part of a world view. Moreover, rain-making, a component of shamanism, is itself a complex activity with rich connotations and associations (Lewis-Williams 1981a: 103-16). All these connotations, together with some related to the ritual of painting itself rather than to specific depictions, probably resonated in the Ezeljagdspoort rock shelter.

Our preferred interpretation of the Ezeljagdspoort painting as an intrinsically powerful reification of a Bushman shaman-artist's hallucination of swallow-shamans engaged in a rain-protection ritual is, of course, a hypothesis, but one that differs from the essentially Western speculation of the narrative approach. By contrast, our ethnographic and neuropsychological hypothesis is more complex and is, moreover, supported by diverse evidence. We argue that the diversity of the evidence contributes to our hypothesis and, indeed, whole approach to southern African rock art in five ways. Most importantly, we know from the ethnography that Bushman rock art was, at least in very large measure, closely associated with trance experience. Secondly, we know, again from the ethnography, many of the metaphors Bushmen used to speak about this experience. Thirdly, we know, from the art itself, some of the diverse and often idiosyncratic ways in which Bushman shaman-artists depicted these metaphors. Fourthly, we know, from neuropsychology, the kinds, though not the 'meanings', of the somatic and visual hallucinations Bushman shamans experienced. Finally, we know from the ethnography, neuropsychology and the art that the paintings were not just pictures: they were powerful ritual objects, things in themselves. We are thus able to contain our interpretations within the milieu of Bushman experience, and this, of course, increases their acceptability. Our interpretation of the Ezeljagdspoort painting is not simply a new 'guess'; it has more supportive ethnographic neuropsychological evidence than competing explanations, and it is consistent with what we already know about the art. Most significantly, it draws our attention to and explains more features of the painted data than did the older narrative interpretations.

The methodology we have exemplified here enables us to uncover aspects of southern African rock art that eluded copyists and researchers for many years. We are beginning to understand some of the religious experiences of Later Stone Age hunter-gatherers that, not long ago, would have been considered forever lost. The task of rock-art research is, in the last analysis, to recover those 'lost' subjectivities; they are as central to our understanding of the past as technology, subsistence strategies and ancient climates. Today there is a welcome and growing emphasis in southern African Later Stone Age archaeology on social analysis (e.g. Wadley 1987; Kinahan 1989; Mazel 1989), but some accounts still give scant attention to the cognitive past.

Prehistoric people were no different from historical people; indeed, a distinction between prehistory and history subverts our attitude to the past, especially in South Africa. Without wishing in any way to diminish traditional archaeological concerns, we believe that, by excluding belief, ritual and 'spiritual' experience, as uncovered by rock-art research, from accounts of the southern African past or by restricting these components to a brief coda to a technological symphony, writers create, albeit inadvertently, an ideological view that separates the indigenous Khoisan inhabitants of the subcontinent from the stream of human history: thus dispossessed, they appear to be insentient pawns on an ecological chess board, moved hither and thither by factors beyond their control (Lewis-Williams 1985; Wadley 1987; Mazel 1989). With their ancestors bereft of quintessentially human feelings and beliefs and any truly humane history, modern people are easy prey to political positions that see them as expendable in an inexorable process of 'development'. On the other hand, by analysing individual rock-art depictions and panels of depictions in the detail that our method allows, we can recover some of the religious and intellectual experiences of the people who constituted at least the last phases of the southern African Later Stone Age. We can thus move towards restoring their full humanity and so dissolving the boundary between history and prehistory. Southern Africa's past need no longer be seen as sharply divided into two different kinds of discourse by the arrival of white settlers. That event was but one intervention in a long and varied human history.

Acknowledgements. We thank Thomas Huffman, Megan Biesele, Lyn Wadley, Anne Holliday, Geoff Blundell and Sven Ouzman who read drafts of this paper and offered useful comments. Lorna Marshall, Megan Biesele (who standardized our spelling of Bushman words), Mathias Guenther, Ed Wilmsen and Polly Wiessner have generously shared their work with us. The Librarian, Jagger Library (University of Cape Town), permitted quotation from the Bleek and Lloyd Collection, and the Director, South African Library, granted access to collections. We are grateful to a number of people who kindly typed a long series of drafts. The Rock Art Research Unit is funded by the Centre for Science Development and the University of the Witwatersrand.

1 Some writers prefer 'San' to 'Bushmen' on account of the latter's pejorative associations for some people in southern Africa. Unfortunately, 'San', a Nama word, also has negative connotations: it could be translated 'vagabond', and its use by archaeologists and anthropologists ascribes to the Nama antagonistic attitudes towards Bushmen. Because there are so many Bushman languages, there is no generic word to cover all groups. Along with writers such as Lorna Marshall, Megan Biesele, Rob Gordon and Alan Barnard, we retain 'Bushman' but explicitly reject all possible pejorative associations.

References

ALEXANDER, J.E. 1837. Narrative of a voyage of observation among the colonies of western Africa, in the flag-ship Thalia; and of a campaign in Kaffir-land, on the staff of the commander-in-chief, in 1835. London: Henry Colburn.

ARBOUSSET, T. & F. DAUMAS. 1846. Narrative of an exploratory tour to the north-east of the colony of the Cape of Good Hope. Translated from the French of the Rev. T. Arbousset by John Croumbie Brown. Cape Town: Robertson.

BARBER, T.X. 1971. Imagery and 'hallucinations': effects of LSD contrasted with the effects of 'hypnotic' suggestions, in S.J. Segal (ed.), Imagery: current cognitive approaches: 101-29. New York (NY): Academic Press.

BARNARD, A. 1989. The lost world of Laurens van der Post?, Current Anthropology 30: 104-14.

BIESELE, M. 1975. Folklore and ritual of !Kung hunter-gatherers. Ph.D. thesis: Harvard University.

1978. Sapience and scarce resources: communication systems of the !Kung and other foragers. Social Science Information 17: 921-47.

1980. 'Old K'xau', in J. Halifax (ed.), Shamanic voices: a survey of visionary narratives: 54-62. Harmondsworth: Penguin.

1992. 'Women like meat': Ju/'hoan Bushman folklore and foraging ideology. Johannesburg: Witwatersrand University Press.

BIESELE, M. & P. WEINBERG. 1990. Shaken roots. Johannesburg: EDA.

BLEEK, D.F. 1924. The Mantis and his friends. Cape Town: Maskew Miller.

1933a. Beliefs and customs of the /Xam Bushmen. Part V: The rain, Bantu Studies 7: 297-312.

1933b. Beliefs and customs of the /Xam Bushmen. Part VI: Rain-making, Bantu Studies 7: 375-92.

1935. Beliefs and customs of the /Xam Bushmen. Part VII: Sorcerors, Bantu Studies 9: 1-47.

1956. A Bushman dictionary. New Haven (CT): American Oriental Society.

BLEEK, W.H.I. 1874. Remarks on Orpen's Mythology of the Maluti Bushmen, Cape Monthly Magazine (n.s.) 9: 10-13.

1875. Brief account of Bushman folklore and other texts. Cape Town: Government Printers.

BURCHELL, W.J. 1822. Travels in the interior of Southern Africa I. London: Longman, Hurst, Rees, Orme & Brown.

COHEN, S. 1964. The beyond within: the LSD story. New York (NY): Atheneum.

DART, R.A. 1925. The historical succession of cultural impacts upon South Africa, Nature 115: 425-9.

DEACON, J. 1986. 'My place is the Bitterpits': the home territory of Bleek and Lloyd's /Xam San informants, African Studies 45: 135-55.

1988. The power of a place in understanding southern San rock engravings, World Archaeology 20: 129-40.

DOWSON, T.A. 1988a. New light on 'the thin red line': neuropsychology and ethnography. Unpublished BSc. Honours thesis. University of the Witwatersrand.

1988b. Revelations of religious reality: the individual in San rock art, World Archaeology 20: 116-28.

1989. Dots and dashes: cracking the entoptic code in Bushman rock paintings, South African Archaeological Society Goodwin Series 6: 84-94.

1992. Rock engravings of southern Africa. Johannesburg: Witwatersrand University Press.

DOWSON, T.A. & J. DEACON. In press. Bleek and Lloyd 1870-1991. Johannesburg: Witwatersrand University Press.

DOWSON, T.A. & A.L. HOLLIDAY, 1989. Zigzags and eland: an interpretation of an idiosyncratic combination, South African Archaeological Bulletin 44: 46-8.

EICHMEIER, J. & O. HOFER. 1974. Endogene Bildmuster. Munich: Urban & Schwarzenberg.

ELLENBERGER, V. 1953. La Fin tragique des Bushmen. Paris: Amiot, Dumont.

ELIADE, M. 1972. Shamanism: archaic techniques of ecstasy. New York (NY): Routledge & Kegan Paul.

EMBODEN, W. 1979. Narcotic plants. New York (NY): Macmillan.

ENGLAND, N.M. 1968. Music among the Ju/wa-si of South West Africa and Botswana. Ph.D thesis. Harvard University.

GARLAKE, P.S. 1987. The painted caves: an introduction to the prehistoric rock art of Zimbabwe. Harare: Modus.

HALL, M. 1988. Archaeology under apartheid, Archaeology 41(6): 62-4.

HALL, S.L. 1986. Pastoral adaptations and forager reactions in the eastern Cape, South African Archaeological Society Goodwin Series 5: 42-9.

HOFER, A. & H. OSMOND. 1967. The hallucinogens. New York (NY): Academic Press.

HOROWITZ, M.J. 1975. Hallucinations: an information-processing approach, in R.K. Siegel & L.J. West (ed.), Hallucinations: behaviour, experience, and theory: 163-95. New York (NY): John Wiley.

HOW, M.W. 1962. The mountain Bushmen of Basutoland. Pretoria: Van Schaik.

HUFFMAN, T.N. 1983. The trance hypothesis and the rock art of Zimbabwe, South African Archaeological Society Goodwin Series 4: 49-53.

JOLLY, P. 1986. A first generation descendant of the Transkei San, South African Archaeological Bulletin 41: 6-9.

JOHNSON, R.T. 1979. Major rock paintings of southern Africa. Cape Town: David Philip.

KATZ, R. 1982. Boiling energy: community healing among the Kalahari Kung. Cambridge (MA): Harvard University Press.

KEIFFER, S.N. & T.B. MORITZ. 1968. Psychedelic drugs, Pennsylvania Medicine 71: 57-67.

KINAHAN, J. 1989. Pastoral nomads of the central Namib Desert. Ph.D thesis. University of the Witwatersrand.

KLUVER, H. 1926. Mescal visions and eidetic vision, American Journal of Psychology 37: 502-15.

1942. Mechanisms of hallucinations, in Q. McNemar & M.A. Merrill (ed.), Studies in personality: 175-207. New York (NY): McGraw Hill.

KNOLL, M. & J. KUGLER. 1959. Subjective light-pattern spectroscopy in the electroencephelographic frequency range, Nature 184: 1823-4.

KNOLL, M, J. KUGLER, O. HOFER & S.D. LAWDER. 1963. Effects of chemical stimulation of electrically-induced phosphenes on their bandwidth, shape, number and intensity, Confinia Neurologica 23: 201-26.

LEEUWENBURG, J. 1970. A Bushman legend from the George district, South African Archaeological Bulletin 25: 145-6.

LEWIS-WILLIAMS, J.D. 1972. The syntax and function of the Giant's Castle rock-paintings, South African Archaeological Bulletin 27: 49-65.

1974a. Rethinking the Southern African rock paintings, Origini 8: 229-57.

1974b. Superpositioning in a sample of rock-paintings in the Barkly East district, South African Archaeological Bulletin 29: 93-103.

1977. Ezeljagdspoort revisited: new light on an enigmatic rock painting, South African Archaeological Bulletin 32: 165-9.

1980. Ethnography and iconography: aspects of southern San thought and art, Man 15: 467-82.

1981a. Believing and seeing: symbolic meanings in southern San rock paintings. London: Academic Press.

1981b. The thin red line: southern San notions and rock paintings of supernatural potency, South African Archaeological Society 36: 5-13.

1983a. Introductory essay: science and rock art, South African Archaeological Society Goodwin Series 4: 3-13.

1983b. The rock art of southern Africa. Cambridge: Cambridge University Press.

1984. Ideological continuities in prehistoric southern Africa: the evidence of rock art, in C. Schrire (ed.), Past and present in hunter-gatherer studies: 225-52. New York (NY): Academic Press.

1985. Social theory in southern African archaeology. Paper presented at South African Association of Archaeologists conference, Grahamstown.

1986a. Cognitive and optical illusions in San rock art research, Current Anthropology 27: 171-8.

1986b. Paintings of power: ethnography and rock art in southern Africa, in M. Biesele, R. Gordon & R. Lee (ed.), The past and future of !Kung ethnography: critical reflections and symbolic perspectives: 231-73. Hamburg: Buske Verlag.

1986c. The last testament of the southern San, South African Archaeological Bulletin 41: 10-11.

1988. Preparation for transformation: some remarks on the role of metaphors in San shamanism, in E.R. Sienaert & A.N. Bell (ed.), Catching winged words: oral tradition and education: 134-45. Durban: Natal University Press.

1989. Southern Africa's place in the archaeology of human understanding, South African Journal of Science 85: 47-52.

1990. Discovering southern African rock art. Cape Town: David Philip.

LEWIS-WILLIAMS, J.D. & M. BIESELE. 1978. Eland hunting rituals among northern and southern San groups: striking similarities,
COPYRIGHT 1993 Antiquity Publications, Ltd.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Lewis-Williams, David; Dowson, Thomas A.; Deacon, Janette
Publication:Antiquity
Date:Jun 1, 1993
Words:9441
Previous Article:The Chumash and the swordfish.
Next Article:Active tectonics and land-use strategies: a Palaeolithic example from northwest Greece.
Topics:

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters