The shamanistic theory
by Mikey Brass (1997/1998)

THE DEMISE OF EMPIRICISM AND THE RISE OF TRANCE

By the early 1970s the realisation had dawned that, despite all the information gathered by means of data collections, a hypothesis explaining the meaning of rock art had still to be put forward. As the last artists had passed away a century before, first Vinnicombe and later Lewis-Williams turned to the ethnographic record - from the analysis of the modern Kalahari San by anthropologists, to the Bleek and Lloyd records and to the accounts left by Orpen. But in fact this promising line of investigation had begun in the late 19th century, until brought to a temporary halt by Bleek's death in 1875 and Orpen's developing interest in other matters. Lloyd, continuing Bleek's work, did not hold the same fascination for rock art (Lewis-Williams 1983: 4).

The empiricist approach that reigned supreme for the majority of the twentieth century is exemplified by the following statement of Willcox in criticism of Lewis-Williams' approach: "In my opinion, the minimum hypotheses as the raisons d'être of the representational rock art of southern Africa are the following: (1) to record important or pleasant events in the life of the community or in the experience of the artist; (2) to instruct the young or illustrate folktales; (3) to give pleasure to the artist through his work and his recreation on the rock, to be seen again, of what pleased him at first view, coupled with the aesthetic experience and receiving admiration for his skill. This motive I take to account for the great bulk of the art, and it does not conflict with, but would accompany, Reasons 1 and 2... The importance of the pleasure principle can hardly be overstated." (Willcox 1983: 540) In essence, the most simplistic explanation is the right one.

However, 'simple' explanations assume a range of interests and values on the part of the artist that are, in most cases, clearly a Western cultural construction. In reality it ranks the lowest in terms of the scientific evaluation of data by virtue of the fact it is ill-defined by nature and consequently its application is problematic (Lewis-Williams 1984: 58). In summary, Willcox's interpretations of the rock art subtly implies a constructional understanding of the artists' system of values and the way that symbols expressed those values. Therefore an understanding of San values and beliefs is essential to a scientifically-applicable approach to interpreting the San rock paintings (Lewis-Williams 1983: 541).

CONCEPTS AND METAPHORS FROM BUSHMAN ETHNOGRAPHY

Lewis-Williams' trance hypothesis is based upon the many similarities between that which has been observed among modern 20th century Kalahari San, that expressed in the Bleek and Lloyd records and that which is evident in the rock art paintings.

Modern San shamans, when asking for supernatural potency, have their arms behind their back. This occurs when the intensity of the trance dance increases and the shaman feels the energy or potency beginning to 'boil' inside him (Lewis-Williams 1990: 8; Van der Merwe 1990: 14). Sticks were an integral part of the trance dance. Women rarely participated in the dance, instead clapping and singing medicine songs which helped activate the potency. This potency is believed to originate in the stomach, travelling up the spine and exploding in the head, the result of which is the dancer entering trance (Van der Merwe 1990: 14). This dance can also be observed in the painting at Orange Springs, eastern Orange Free State (Dowson 1994: 336).

Nasal haemorrhaging also occurs in the curing ritual whereby the blood was wiped on the body of the patient in the belief that its smell would keep evil spirits at bay (Lewis-Williams 1982: 433).

Yet caution has to be exercised in proposing that all depictions of rock art figures supporting themselves with sticks are representations of shamans in trance dancing. The possibility exists that some of these depictions are the expression of outside influences on San society, a form of symbiotic relationship (Campbell 1986: 255). One such opinion has been voiced by Jolly in his examination of the paintings from the cave at Melikane, Lesotho (Jolly 1995: 68). In this region the San were in contact with Nguni and Sotho groupings. Jolly believes they were influenced by their ritual practices, and points out that the bending posture of the Melikane figures, who support themselves with sticks, are reminiscent of the identical posture adopted by Nguni diviners who also support themselves with two sticks (Jolly 1995: 72).

ANIMAL METAPHORS

It is not only straight forward trance dance icons that are represented in the rock art, but also its symbols. The trance hypothesis proposes that a relationship was seen between a trance dancer symbolically 'dying' and a dying antelope by the southern San. Such an analogy was drawn upon because both staggered, sweated and suffered nasal bleeds (Lewis-Williams 1982: 434). Trance also resulted in the sensation of the shaman's hair growing, paralleled by a dying eland's hair standing on end. Moreover, a dying eland was seen to be charged with the power a shaman drew upon as he symbolically 'died' (Lewis-Williams 1982: 434). This led to the depiction of a relationship between a trancer and a dying eland, which is recognised by its front legs collapsing, blood pouring from its nose, and almost always its head is down. The eland is thus a primary symbol of trance potency (Van der Merwe 1990: 15).

However the San concept of animal potency is not limited to the eland. The !Kung have ritual dances at the time of the year when bees are swarming, believing that bees possess the potency which is harnessed by the shamans (Lewis-Williams 1983:6). This symbolism of potency can be seen in the painting from Cullen's Wood, Barkly East where bees are depicted over the heads of dancing men.

THE SYMBOLIC WATER

The sensation experienced by the medicine on entering into trance was often expressed in terms of water (Lewis-Williams 1990: 10). This is the consequence of the roaring sound heard during particular altered states of consciousness, and symbolised in San mythology as a great body of water through which the shaman had to pass to enter the spirit world of the dead (Lewis-Williams 1990: 10-11).

However, an hypothesis that emphasises the primacy of mythology results in an interesting variant. The /Xam myths, as recorded by Bleek and Lloyd, describe the waterhole as the realm of the spirits of the dead, as the realm of sickness. Solomon believes that these testimonies refer not primarily to trance but rather to the set of beliefs held regarding mortality, and that these are notions of 'factual' death rather than any hallucinations (Solomon 1997: 7).

Thus 'falling' into and through water is visualised in terms of 'death'. In spatial terms a journey to the netherworld of the spirits of the dead and their mythical ancestors of 'The First Creation'. These mythical ancestors were regarded as being primitive, and to descend to their world was thought of as a "temporal reversion and a cultural regression" (Solomon 1997: 8). Thus the San concept of time was not uni-linear, allowing spirits or trancers to travel regularly between the past and the present.

Both these explanations draw on Orpen's statement that the figures at Melikane represented those who had "died and now lived in rivers" (Orpen 1874: 2 cited in Solomon 1997: 3). On the other hand, as Qing kept in close contact with the Sotho, Jolly regards his remarks as probably expressing Phuthi notions associated with visits by diviners in trance to underwater realms. Jolly concludes that the reference to the 'dead' is likely to be a metaphor for trance experience, as the trance hypothesis proposed, but with the moderating influence of Phuthi religious ideology (Jolly 1995: 74-5).

THERIANTHROPES

Lewis-Williams interprets depictions of so-called "flying buck" as "trance buck" - shamans in trance. The painting from Glenavon, Barkly East depicts some of the "trance buck" in a kneeling posture (the position a shaman often falls to when in trance). Two of these dancers have antelope heads and all have their arms in the characteristic backward posture adopted when receiving potency (Lewis-Williams 1982: 435). Thus these figures are "dying" in trance and the transformation of the dancer into the buck is taking place (Lewis-Williams 1982: 435; Lewis-Williams 1987: 172).

The linking of the shamans and animals with which they fused, however, is sometimes almost virtually or fully complete, thus rendering detection very difficult. However, nasal blood is one of the evidences for painted shamans in trance (Lewis-Williams 1987: 172). In fact the bleeding from the nose is one of the most important arguments against the interpretation of therianthropes as hunters dressed in animal masks, others being that often therianthropes bear flywhisks - an important trance dance symbol - and, instead of feet, have hoofs (Lewis-Williams 1987: 172).

However Jolly puts forward a case for the Melikane therianthropes being instead representations of ritual functionaries - his likely candidate being either Sotho or Nguni, and not San, adopting ceremonial dress embodying an antelope's head and skin. He backs up this hypothesis by citing Walton (1957: 279 in Jolly 1995: 72) who "remarked that Sotho ritual functionaries wear animal-head masks". There is also the striking resemblance between the Melikane paintings and those of a Sotho-Tswana "buck-jumper" photographed in 1934 who clasps onto two sticks and is covered by an antelope's skin and head (Jolly 1995: 72). This use of two sticks bending forward is also used on occasions by Xhosa diviners (Botha & Thackeray 1987: 72). The argument that these ceremonial dresses were probably too impractical is countered by the photograph and records of similar skin dresses in other parts of the world (Jolly 1995: 72 citing Saunders 1989, 1994: fig. 3). Jolly also believe that the Melikane painting techniques differ from those depicting a fusion of man and animal, as the animals' heads seem to be attached to, and the arms and legs distinctly separate from, what appears to be a kaross (Jolly 1995: 72). Jolly believes that further research could likely prove this to be the case with certain other therianthropic depictions.

This issue of ritual raises the question of whether there is any relationship between therianthropes and the San mythology, a line of inquiry which Solomon (1997) addresses. According to this hypothesis, therianthropes are most likely the spirits either of the ancestral San (in their 'First Creation' or therianthropic form) or of dead people considered powerful magicians while alive. When this perspective is taken into account then the references to water and death, together with the spoiling of the eland, become clear. The therianthropes can indeed be men who have died and now live in rivers, and were spoilt at the same time as the eland, rather than products of hallucinatory experience." (Solomon 1997: 9)

THE AMAZING BODY

Lewis-Williams has hypothesised that the elongated figures in the rock art are the representations of shamans' trance sensations of becoming taller (Solomon 1992: 308; Van der Merwe 1990: 15). However, in developing the hypothesis of the shamans' trance experiences being culturally determined, Solomon notes that gender conventions in recorded San records express femininity in terms of being "round/low/broad/short" and masculinity as "tall/narrow/slender" (Solomon 1992: 302). Solomon consequently hypothesises that these assertions of masculine form determined the form feelings of bodily distortion in trance (Solomon 1992: 308).

However, there are features of people in the rock art that can only be satisfactorily explained at present by the trance hypothesis - such as lines rising out of a shaman's head, which are said to be the shaman's spirit leaving his body and could only be seen by other shamans in trance (Lewis-Williams 1983: 9). For the /Xam informants of Bleek, out-of-body travel was important and some referred to these journeys as taking place in the form of an animal, the favourite being a lion (Lewis-Williams 1982: 436; 1992: 57). The aim of this transformation was to frighten away any lurking evil spirits which could also have assumed the shape of an animal.

THE RAIN-ANIMALS
The practice of rain-making is reflected in the San art in differing ways. According to Lewis-Williams, some of the rain-animals are probably narrations of the capture of a real rain-animal while others represent symbols of trance hallucinations (Van der Merwe 1990: 15). A favourite rain-animal was the eland; and unnaturally-tusked figures have also been suggested to relate with rainmaking (Prins 1990: 120). The Bleek and Lloyd records speak of "ritual practitioners...riding a 'rain-animal' to the top of a mountain and killing it so that its blood would fall as rain" (Lewis-Williams 1992:57). It was believed that rain would fall where the blood of the rain animal had run.

However Solomon believes that rain can be construed in gender terms as well, by pointing out that the !Kung believe fat rain-bringing clouds are feminine and wispy rainless clouds are male. There is a similar belief about the 'footprints' of the rain - marks left by the rain on the ground that are broad and shallow are those of the female rain, while smaller and deeper marks are those caused by the male rain (Solomon 1992: 302).

Solomon also points out that the importance of the eland as a rain animal, together with being a sexual and generalised fertility symbol, has a lot to do with its feminine characteristics of fat, docility, herbivorous and its link to water (Solomon 1992: 303).

THE SYMBOLISM OF LINES

There are many instances in the rock art paintings of human figures being connected to each other and to animals by means of a red line fringed with white dots (Lewis-Williams 1982: 437). The trance hypothesis links the white dots to the shaman's footprints and the whole thing to the fundamental issue of potency with which the hypothesis deals with.

Solomon approaches this question from her gender hypothesis, in which she notes "blood (as feminine) and water (as masculine) are gendered symbols of life and death, respectively" (Solomon 1992: 298). One of Bleek's informants described the red lines with white dots as representing the "rain's navel", and therefore Solomon follows the thread that these lines might be a way of representing the contrast between blood and water. Where the blood of the rainmaking animal soaked the ground, rain was believed to fall. Thus the white dots probably represent rain drops, or the rain's 'footprints', rather than the footsteps of a shaman (Solomon 1992: 317). There are paintings of shamans in trance in which dots portray the sweat from a shaman's armpits. The red line is interpreted as depicting blood.

SIGNIFICANCE OF THE ROCK FACE
Some of these lines will stretch for some metres through a rock shelter and link "widely separated paintings and apparently weaving in and out of the rock face.... It enters a tiny inequality on the rock or simply ends, only to reappear a few centimetres away" (Dowson & Lewis-Williams 1990: 5). Frequently figures seem to appear from or disappear into cracks and steps in the rock face. "Others are 'folded' into concave right angles; still others come off the edges of convex right angles. Some are fitted neatly into facets or hollows in the rock and a few incorporate nodules of rock." (Dowson & Lewis-Williams 1990: 5).

To explain these occurrences, Dowson and Lewis-Williams turn to the inner workings of the body's nervous system as there is one feature of trance that is neurologically determined (rather than the mind culturally valuing certain hallucinations above others) - the image of a tunnel, the sensation of travelling through one often being compared - on account of the rushing roaring heard - to that of the sound of rushing water (Dowson & Lewis-Williams 1990: 9, 10). The tunnel was equated by the San with a hole in the ground through which the spirit world could be accessed (Dowson & Lewis-Williams 1990: 10, 11).

Dowson and Lewis-Williams believe that often shamans entered trance in the presence of the rock shelter. Depth perception changes in trance and so their hypothesis proposes that "the walls of shelters and especially holes and inequalities in the rock surface may, under such conditions, have appeared to be entrances to tunnels leading to the spirit world. Shamans were drawn into these tunnels as they headed for the other realm" (Dowson & Lewis-Williams 1990: 12).

CONCLUSION

The shamanistic model developed by David Lewis-Williams is currently the most widely prevailing explanation of the San rock art paintings, and has as its central role the San trance dance. In the Kalahari today, women sit around a fire while singing medicine songs containing supernatural potency and clapping to its rhythm. The men, of whom normally roughly half are shamans, dance (Lewis-Williams 1987: 166). They tremble and stagger before entering trance, after which they lay their hands on various people to draw real or imagined sickness out of their bodies. Shamans enter a deeper level of trance, collapse and endure various hallucinations - for example out-of-body experiences. In this hypothesis trance is primary and mythology secondary - the trance experience resulted in the San mythology.

What Lewis-Williams' hypothesis does not incorporate is the issue of the symbolism of gender - how gender is portrayed in the rock art. With the social and economic gender divisions in San society, this aspect was sadly neglected until Solomon's' article (1992) and even now its relationship to the art requires further investigation. Solomon extended her investigations to include the possibility of the primacy of San mythology to trance (1997). In essence, Solomon puts forward the possibility that it was San mythology which culturally influenced which visions the shamans experienced in their altered state of trance and, consequently, the paintings are best viewed in association with the religious beliefs of the San.

Although the possibility of cross-cultural religious influences can never be ruled out (Jolly 1995) Lewis-Williams, Solomon and Jolly all emphasise, albeit in different ways the role of the trance dance. And one of the major underlying reasons for performing the trance dance among the modern Kalahari San is to emphasise the group's social and spiritual cohesion.