rock art and ritual: southern africa and beyondmsu.ac.zw/elearning/material/1299832184lewis-williams...

Complutum, 5, 1994: 277-289

ROCK ART AND RITUAL:SOUTHERN AFRICA AND BEYOND

J. David Lewis- Williams*

ABSTRACT.- Southern African Rock Art is taken as a starting point in order to argue the shamaniccharacter of European Upper Palaeolithic Parietal Art. The author identifies different stages ofproduction and consumption of rock-art depictions. All of them are embedded in rituals that consti-tuted, reproduced and sometimes subverted power relations. Mlultiple ethnographic analogies areuseful in order to build up an account of the changing association between art and ritual in the Up-per Palaeolithic of Western Europe.

IRESuwEv.- El arte rupestre sudafricano sirve de punto de partida para argumentar 'el cardcterchaminico del arte parietal paleolitico. El autor identifica los diferentes estadios de produccibn yconsume de las representaciones rupestres. Todos ellosforman parte de rituales que conforman, re-producen y a veces subvierten las relaciones de poder. El empleo de la analogia etnografica multi-ple es extremadamente itil si se quieren detectar las transformaciones en la asociacion entre arte yritual durante el Paleolitico Superior en Europa Occidental.

KEY'ORDS: Rock Art. Bushmen. Shamanism. Social Relations. Upper Palaeolithic.

PALABRAS CL4 E: Arte Rupestre. Bosquimanos. Chamanismo. Relaciones Sociales. PaleoliticoSuperior.

1. INTRODUCTION Setting aside Romantic ideas about art, I ar-gue that the production and consumption of rock art,

One of the most debilitating concepts in like all image-making, was embedded in the social,rock art research has been the autonomy of art. No- economic and intellectual circumstances of the cornm-tions of 'the artist' as isolated from the ebb and flow munity in which it was made (e.g. Wolff 1981). Likeof daily life, as an Olympian commentator on the foi- language and, indeed, other genres of material cultu-bles and tragedies of humanity, and as an ascetic, re, rock art did not merely reflect the society ininspired individual working not for gain but for the which it was made, its economy, power structures,sake of art itself have silently -for they are seldom myths and so forth. It also constituted. reproducedexplicitly stated- undermined many studies of hun- and sometimes subverted power relations. In thatter-gatherer rock art. Yet these ideas about artists sense, the making of each rock art image was aand their work are not universals: the notion of the socio-political intervention that either undenrwroteartist's separation from ordinary people and the so- power relations based, often, on the possession of ar-cial processes of production developed at a particular cane knowledge by a select group, or challengedtime in Western history. Essentially Romantic in ori- existing structures by attempting to shift the locus ofgin, the genesis of these ideas was historically situa- divine sanction. Exactly what .those social relationsted in a specific social, economic and intellectual were and exactly how any given rock art images re-milieu; in that context, the Romantic notions about produced or subverted them are specific historicalart and artists played an ideological role, often mas- questions.king social relations and asymmetrical power struc- If rock art is seen as active in the constitu-tures. tion of social relations, rather than as a passive re-

* Rock Art Research Unit. University of the Witwaterstrand. 2050 Johanesburg. Africa del Sur.

278 J. D. LEWIS-WILLIAMS

flection of society, it cannot be studied and under- southern Africa: they constitute a temporal and spa-stood in isolation from power relations and, moreo- tial mosaic that gives indications of what happenedver, from other expressive forms, such as myth and at specific times and in specific places. (For a fullerritual, that similarly reproduce or subvert social rela- account of nineteenth-century Bushman shamanismtions. I therefore begin with a very brief review of see Lewis-Williams 1981; Lewis-Williams and Dow-nineteenth-century Bushman (San1) shamanism and son 1989).the politico-religious setting of the production and Like all forms of shamanismn nineteenth-consumption of nineteenth-century Bushman rock century Bushman religion was constituted by institu-paintings (I do not consider southern African rock tionalised altered states of consciousness. As in allengravings; see Dowson 1992). I then discuss in mo- shamanistic religions, certain kinds of trance expe-re detail the relationship between southern Bushman riences and hallucinations were accorded the statusrock paintings and ritual. In doing so, I show that the of visions; people in general accepted the shamans'rituals associated with the making and viewing of the accounts of their visions as insights into what waspaintings, as well as the ritual context from which happening in the spirit world. It seems that approxi-much of the art's subject matter derived, were deeply mately half of the men and a third of the women in aembedded in social relations and that these rituals nineteenth-century Bushman camp may have beenconstituted and in some instances subverted those re- shamans, as was the case in some mid-twentieth-lations. Finally, I move on to some broad and preli- century Kalahari Bushman communities (Marshallminary implications for the study of West European 1969; Katz 1982). Being a shaman did not conferUpper Palaeolithic art and ritual. any general authority; nor did it bring with it any

special privileges. Shamans were, however, respectedand some enjoyed considerable prestige.

2. BUSHMAN SHAMANISM Bushman shamans entered an altered stateof consciousness either at a large communal trance

Research conducted over the last twenty dance or in more private circumstances. Trance wasyears has eroded the older notions of art-for-art's- induced not through the ingestion of psychotropicsake and hunting magic as explanations for southern drugs but by intense concentration. prolongedAfrican rock art. Today most researchers accept that rhythmic dancing, audio-driving and hyperventila-much of this art was implicated in the beliefs and ri- tion (but see Winkleman and Dobkin de Rios 1989).tuals of Bushman shamanism (for a history of sou- In an altered state of consciousness, shamans curedthem African rock art research see Lewis-Williams the sick, controlled the movements of antelope herds,in press a). The last paintings were made towards the journeyed to god. went on out-of-body travel to dis-end of the nineteenth century, and our information tant Bushman camps and made rain by capturing aabout Bushman religion in that century comes prin- (hallucinatory) 'rain-animal'. To achieve these endscipally from three sources: first and foremost, the vo- they harnessed a supernatural potency, !gi: or //ken,luminous W.H.I. Bleek and L.C. Lloyd Collection of that was associated with large game animals, espe-verbatim texts that were obtained from Bushmen cially the eland. Shamans were said to 'possess', forwho came from what is now the Northern Cape Pro- example, eland. giraffe or gemsbok potency.vince: secondly, J.M. Orpen's comparatively brief ac- The alternative reality to which animal-count of the mythology of the Maluti San of what is potency gave access, associated as it was with thenow Lesotho; thirdly, some less satisfactory but none Bushmen's comparatively egalitarian social structure.the less valuable material that the missionaries T. was not as complexly constructed as that of someArbousset and F. Daumas collected in the 1830s in other, more complex shamanistic societies. Yet, likewhat is now the eastern Orange Free State and Leso- many other shamanistic views of the spirit world. thetho. I supplement these nineteenth-century sources Bushmen's alternative reality was essentially imma-with material collected in the twentieth century from nent; like altered states of consciousness themselves.Bushman communities still living to the north in the it was 'next to', or 'within', people, not situated at aKalahari Desert. Finally, although there are now no remote distance; it interdigitated with the world ofrecognizable southern Bushman communities, I also daily life. At the same time, the cosmos, of whichdraw on information obtained from a descendant of this alternative reality was a part, was, like the cos-one of these communities and from a Basotho man moses of many shamanistic societies, conceived of aswho knew them well. We must, of course, bear in tiered. The Bushman cosmos comprised three levels:

·'/~" nind that these ethnographic and historical sources (1) the surface of the earth, the level of daily life; (2) ( .cannot be projected to cover all rock art throughout a spiritual underworld that was associated particu-

ROCK ART AND RITUAL: SOUTHERN AFRICA AND BEYOND 279

larly but not exclusively with the dead; and (3) a spi- political importance and that some shaman-artistsritual realm above the earth that was associated with used rock art to negotiate their positions in the chan-god, the spirits and also with shamans. The spirits ging and complex southern African society of that ti-were, however, not located exclusively in the realm me (Dowson 1994).above; they also walked the earth with people. Most Broadly speaking, Bushman rock art com-importantly, the shamans had the ability to transcend prises a range of depictions most of which are refera-the three-tiered cosmos and thus to move between ble to one or other aspect of Bushman shamanism.realities. Bushman shamans were essentially media- These depictions include: trance dances; shamans

' tors of the cosmos. identifiable by a number of features, postures andThe principal nineteenth-century southern gestures; animals, like the eland, that were conside-

Bushman spiritual figure was /Kaggen, a name that red sources of supernatural potency; 'scenes' that ap-is frequently translated as 'the Mantis', although the pear to record historical events, but that sometimesinsect Mantis religiosa was only one of his many ma- incorporate shamanistic elements (Campbell 1986);nifestations. He created all things, yet was, at the sa- various activities that shamans conducted in the spi-me time, mischievous and often stupid (Schmidt rit world, such as out-of-body travel and the capture1973; Lewis-Williams 1981: 117-126). He was, mo- of a rain-animal: shamans partially transformed intoreover, the original shaman. and he created and gave animals (therianthropes); other experiences, inclu-to human shamans the supernatural animal-potency ding shamanistic hallucinations; and 'abstract', geo-that they harnessed to achieve their ends. His favou- metric motifs that probably depict the geometricrite creature was the highly potent eland; when he entoptic mental images experienced in an early stagecreated the eland, an event recounted in a cycle of of trance (for reviews of these various classes of ima-myths, he in fact created the basis for the whole sha- gery see Lewis-Williams and Dowson 1989; Lewis-manistic enterprise. Although he was especially asso- Williams 1981, 1990; Dowson 1992).ciated with the eland, he created and protected allanimals, releasing them to hunters only under certaincircumstances. He protected his animals from hun- 3. BUSHMAN RITUAL AND ARTters and used various ruses to outwit them. In thissense, he was a Lord of the Animals, a figure found The range of subject matter in Bushmanin many shamanic religions, rock art throughout southern Africa suggests that it

On the other hand, nineteenth-century was concerned not with a single, monolithic 'mea-Bushman shamanism departed in a number of ways ning' but rather a closely interrelated set of mea-from the 'classic' shamanism of central Asia. -For nings, a broad diversity situated within an essentialexample, although Bushman shamans used dancing unity. This tension between diversity and unity deri-rattles, flywhisks and tortoise-shell censers, they did yes from the socio-political role of the art (Dowsonnot have the highly elaborate, symbolic costumes and 1994) and its association with rituals that, like theparaphernalia that are associated with some other Bushmen's cosmos, were posited on altered states ofshamans. Nor was shamanic power concentrated in consciousness. Indeed, rituals involving altered statesthe hands of one or only a few people; the egalitarian of consciousness were implicated in a series of fourideals of Bushman society (perhaps not always reali- stages in the production and consumption of Bush-sed; Gulbrandsen 1991: Wilmsen 1989) militated man rock art: (1) the acquisition of some of the art'sagainst the centralization of power. Anyone could try subject matter; (2) the making of paint; (3) the pain-to become a shaman. Those who did not manage to ting of the images in the rock shelters: (4) the use ofmaster altered states of consciousness were not in the paintings once they had been made. I considerany way despised. In 'traditional' Bushman society each of these stages in turn.shamanic power was thus generally separate frompolitical power, which was itself distributed throug- 3.1 The acquisition of imageryhout the social unit. In recent times, however, whensome Kalahari Bushman communities lost their land There were two contexts in which Bushmanand were reduced to wage-earning serfs, shamans be- shamans acquired their visionary insights into thegan to emerge as political leaders (Guenther 1975). spiritual world. Each was associated with two relatedToday these political shamans do not produce rock oppositions: first, society and the individual; se-art, but rock paintings in the southern Drakensberg condly, socially informed and sanctioned visions andsuggest that eighteenth-century and nineteenth- the novel, unexpected visions that altered states ofcentury shamans in that region did begin to assume consciousness inevitably produce.


The principal Bushman shamanic ritual. the informing social influences may be. the human brainone that brought together all people no matter what in an altered state of consciousness always producestheir age or sex, was the curing. or trance, dance. novel, or aberrant hallucinations. Most people igno-One of the earliest accounts we have of such a dance re these sports of the human nervous system becausewas recorded by Arbousset and Daumas in the 1830s. they are seeking specific kinds of visions that theyIt is. of course, an 'outsider's' view and is shot can understand and that will make them feel part ofthrough with these missionaries' manifest distaste for a social group. But some people seize upon halluci-indigenous beliefs and rituals. The dance, they said. natory novelties and then present them to others aswas performed by moonlight and consisted of "irre- specially privileged insights that set them abovegular jumps... They gambol together till all be fati- others or, more forcefully, that challenge the wholegued and covered with perspiration" (Arbousset and structure of power relations. In some circumstances,Daumas 1846: 246-7). The dance was so "violent" the individual visionary thus opposes social cons-that some dancers fell to the ground "exhausted and traints. Both these attitudes to the mental imagery ofcovered with blood, which pours from the nostrils". altered states of consciousness are found among theThese "exhausted" dancers were cared for by some of Bushmen.the women. About forty years later. Orpen (1874: 10) The role of individual Bushman shamans isrecorded a similar dance which. he said. was circu- also seen in the way that they understand dreams, thelar. He too noted that dancers fell down "as if mad second and more personal context in which they ob-and sick" and that blood ran from the noses of some. tain visions. The Bleek and Lloyd collection records

The wider importance of blood in Bushman the ways in which nineteenth-century shamans maderitual and art will become apparent as I proceed. He- rain and went on out-of-body journeys whilst in are I note only that Bleek and Lloyd found that it was dream (Lewis-Williams 1987). In a particularly stri-shamans who suffered nasal haemorrhages and that king and well documented twentieth-century instan-they rubbed their blood on those whom they wished ce, Beh, a Kalahari !Kung woman, dreamed of ga-to heal; they believed that its smell would keep evil Iloping giraffes (Biesele 1993). When she awoke, shespirits at bay. Twentieth-century researchers in the was able to discern in the rhythm of their poundingKalahari did not encounter much nasal bleeding hoofs the metre of a song. She was not herself a sha-among shamans, but they were told that a shaman man, but when she sung the song to her husband.may bleed in especially challenging or dangerous cir- who was a shaman, he instantly recognised it as acumstances (Marshall 1969: 374; Lewis-Williams new source of animal-potency. In a comparatively1981: 81). The depiction of nasal haemorrhage is short time, the giraffe 'medicine song' had spreadone of the distinguishing features of paintings of across the Kalahari and was being sung along side oftrancing shamans. Neither Arbousset and Daumas, the older songs. such as eland and gemsbok. Beh andnor Orpen, nor Bleek and Lloyd seem to have had her husband became well-known, though not politi-any understanding of trance, but today it is clear that cally powerful, people. As I have said. all people ex-these early writers were in fact describing trance perience unusual visions and dreams, but only a fewdances similar to those still performed in parts of the seize upon them and recognize their potential.Kalahari Desert. This tension between personal revelations

During the course of a present-day trance and socially sanctioned visions is evident in the rockdance in the Kalahari, when a number of shamans art (Dowson 1988). Many rock art motifs are wides-are in trance, one may draw the others' attention to pread. The eland. for instance, is the most frequentlywhat he or she believes he or she can see. perhaps a depicted animal in most regions of southern Africa.number of spirit-eland standing in the semi-darkness Yet, idiosyncratic motifs do occur. For these motifsbeyond the light of the fire. The others look in the di- to have been intelligible to other people they mustrection indicated, and then they too see the same vi- have fallen within the broad, general framework ofsions. There is thus a sharing of insights that makes Bushman symbolism and experience. A unique pain-for commonality of visions. Moreover, the describing ting of crabs, for instance, develops the Bushmanof visions after evenryone has returned to a normal metaphor of being underwater as a way of expressingstate of consciousness is a further powerful influence the sensations of trance experience (Dowson 1988:on what people 'see' in future trance experiences. fig. 3). Crabs as a motif did not become accepted byPeople tend to hallucinate what they expect to other shaman-artists. so we can conclude that the ar-hallucinate. tist who painted these unique images retained, unlike

At the same time. there are forces pulling in Beh and her husband. the special insight: the presti-the opposite direction. No matter how powerful the ge of having acquired a unique insight into spiritual


things was not shared and thus dissipated amongst mic clapping that helps the dancing shamans to entershamans in general. Whether this prestige developed trance. Within the contexts of both the dance and theinto political power is a question that we cannot now preparation of paint men and women co-operated. Itanswer. does, however, seem that women were nevertheless

generally accorded a supportive rather than a central3.2 The manufacture of paint or equal role. So constituted, the dance and the pre-

paration of paint were both potential ritual arenas forVery little was recorded about the wavs in the negotiation of gender roles: although women

which Bushman artists prepared their paint. Perhaps seem to have been generally subordinate to men,the early writers considered this too prosaic a matter their contribution to shamanic rituals was neverthe-to warrant their attention. There is, however, an im- less crucial, and individual women could no doubtportant account that strongly suggests that the ma- have manipulated these male: female relations toking of paint was far from prosaic. their own advantage.

In the early 1930s Marion Walsham How After some commercial ochre, specially pur-was able to converse with a seventy-four year old Ba- chased for the occasion (How deemed her piece ofsotho man, Mapote. who. as a young man, had lear- qhang qhang too precious to be used). had beenned to paint with Bushmen in their caves. He was a ground to a powder, Mapote asked for another highlyson of the Basotho chief Moorosi, and he had half- significant ingredient for his paint: "the blood of aBushman stepbrothers, the sons of Moorosi's Bush- freshly killed eland" (How 1962: 37). Qhang qhang,woman wives (How 1962: 33). According to Mapote. he said, was the only pigment that the Bushmen mi-the "true" Bushmen painted at one end of the cave, xed with eland blood. If the blood were not fresh, itwhile he and his half-Bushman stepbrothers painted would coagulate and not soak into the rock. As Howat the other end. A distinction between paintings at observes, the need for fresh blood implies that pain-opposite ends of rock shelters has not been observed, ting took place after a successful eland hunt Mapoteso we do not know how general this separation may then set about painting an eland because, as he put it,have been or, indeed, how many Basotho people we- "the Bushmen of that part of the country were of there taught how to paint. It may have been something eland" (ibid.: 38).that happened only once. The importance of eland blood as an ingre-

When How produced some red Bushman dient in the manufacture of red paint was confirmedpigment that a friend had given her some time before and enlarged upon in the early 1980s by an old wo-she met Mapote, he declared it to be authentic man of Bushman descent, known as 'M', who was li-"qhang qhang"; it "glistened and sparkled" in con- ving in Transkei, to the south of the Drakensbergtrast to commercially available ochre which was dull (Jolly 1986; Lewis-Williams 1986; Prins 1990). Herby comparison. Qhang qhang was dug out of the father had been a shaman-artist, and she pointed outhigh basalt mountains, and many Basotho people re- paintings that he had made. Her elder sister, whogarded it as a "powerful medicine" that would ward had died a few years before she was interviewed, hadoff lightning and hail (ibid: 34). Not only the Bush- been taught her father's shamanic skills and had beenmen themselves but also neighbouring people thus well known locally as a rain-maker. According to M,believed one of the pigments to have supernatural the whole sequence of events started with a ritualisedpowers. eland hunt. She explained that a young girl accom-

The transformation of this highly prized panied a group of hunters who went out after anpigment into paint was. according to Mapote, accom- eland. This girl 'hypnotised' the eland by pointing anpanied by ritual procedures. He said that a woman arrow at it; on the arrow was 'medicine' that hadhad to heat the qhang qhang at full moon out of been prepared by shamans. Dazed, the eland wasdoors until it was red hot. It was then ground bet- then led back, again by supernatural means (thoughween two stones until it was a fine powder. the movements of exhausted or wounded eland can

The role of a woman in the preparation of in fact be fairly easily controlled), to a place near thepigment recalls the part played by women at a trance rock shelter where the people were living and wheredance. Although up to a third of the women in a Ka- the paintings were to be made (Jolly 1986).lahari Bushman camp may be shamans, they gene- The people then prepared a mixture of elandrally sit in a circle with all the other women around a blood and fat. M explained that eland blood contai-central fire; some may dance with the men for a whi- ned supernatural potency (Jolly 1986: 6). Used inle. It is all the women together who supply the vital scarification rituals, this mixture of blood and fat im-singing of 'medicine' songs and the complex rhyth- bued the recipient with eland potency. She went on


to say that eland blood was also used in the prepara- perienced the kaleidoscopic world of trance- to con-tion of paint and that a painting made with eland tradict someone else's account. Rock paintings de-blood was a kind of storehouse of potency. picting therianthropes, greatly elongated figures, bi-

It should be noted that a dying eland bleeds zarre animals and so forth may have. in some sense,from the nose, even as a shaman who, in the Bush- paralleled the modem Bushman shamans' verbal re-men's own phrase, is 'dying' in trance bleeds from ports of spiritual things. By looking at these pain-the nose. Qing implied this parallel when he told Or- tings, people could obtain a vivid idea of whatpen (1874: 2) that the therianthropic figures in the shamans saw in the spirit world. This parallel bet-art (shamans partially transformed into animals: ween paintings and verbal accounts of trance expe-Lewis-Williams 1981) had entered trance "at the sa- rience should, however, not be taken too far. Theme time as the elands and by the dances of which potent ingredients qhang qhang and eland blood cau-you have seen paintings" (for a fuller interpretation sed paintings to be powerful 'things-in-themselves',of Qing's highly complex statement see Lewis- not just pictures or representations of other, muchWilliams 1980). more important, things.

Two kinds of blood were thus involved in The status of rock art images as somethingthe production of rock paintings. First, shamans bled more than mere pictures is further seen in the way infrom the nose when they entered trance to obtain vi- which some of them enter or leave cracks, steps orsions of the spirit world. At this time, their 'dying' in other inequalities in the rock face. Sometimes an an-trance and bleeding paralleled the nasal bleeding of a telope, a snake or a rain-animal is painted in such asdying eland. Secondly, the potent blood of an eland way that it seems to be emerging from behind thewas, at least sometimes, used to make rock art ima- rock surface. This feature of the art is probably rela-ges of these visions. Moreover, the accounts provided ted to the Bushman belief that the spirit world is rea-by Mapote and M corroborate one another in sugges- ched by means of an underground journey. For someting that different interest groups, men and women, nineteenth-century shamans this journey started bywere involved in various ways in the manufacture of diving (in their trance experience) into waterholespaint and, further, that the manufacture of some (Bleek 1935; cf. Biesele 1978). It seems probable thatpaint was highly ritualized. rock shelters were also sometimes seen as potential

entrances to the spirit world and that the rock face3.3 The making of rock paintings was a kind of 'veil' suspended between this world

and the spirit world (Lewis-Williams and DowsonThe ways in which Bushman artists applied 1990). Shaman-artists used their skills to coax the

paint to rock surfaces has been much debated. They animal and spirit inhabitants of the spirit world fromachieved remarkably fine lines and delicate detail; behind the rock and then, using ritually prepared,the lines are often as fine as those made by a thin potent paint, to fix these visions on the rock for all tolead pencil. How (1962: 33) noted that Mapote made see.small brushes from feathers and tiny reeds, but the Unfortunately, we do not know if this fixingvery finest lines must have been made with some- of visions was accompanied by rituals or how athing even finer, perhaps a quill or a sharp bone shaman-artist prepared him- or herself for the task.point. For instance, was the fixing of visions considered as

The delicate workmanship that is evident dangerous as spiritual journeys to the other world?everywhere in southern Africa suggests that it is Perhaps, like the dance in which visions were acqui-most unlikely that shamans would have painted whi- red, this fixing of visions was also considered an ap-le in trance; if not actually unconscious, they tremble propriate occasion for the singing of 'medicine songs'violently. More probably, they painted while in a to strengthen shamans in their work. Whatever thenormal state of consciousness, recalling their vivid case, it seems unlikely that Bushman artists wereglimpses of the spirit world and making powerful anything like the detached ascetics of the Westernimages of those visions and of eland, their principal Romantic fiction.source of animal-potency. Like Wordsworth's obser-vation on poetry, Bushman rock art should probably 3.4. The ritual use of southern Africanbe seen as powerful emotion recollected in tranqui- rock paintingsllity. In the Kalahari today, people listen intently as,the day after a dance. shamans recount their spiritual Once made, many of the images seem to ha-experiences; each account is accepted as a revelation ve continued to perform a significant ritual function.even if it seems -to ordinary people who have not ex- The rock shelters were not simply 'galleries', as wri-


ters on rock art often call them, where people could believe, her way of saying that the sight of the pain-view 'works of art'. As I have already argued, the tings deepened the dancers' trance experience.paintings were 'things-in-themselves', not just pictu- More than that, the fixed visions already onres of things that existed elsewhere. Many of them the rock face probably contributed to the dancers' ha-were, moreover, made with special, ritually prepared Ilucinations, informing and constraining the streampaint and thus became reservoirs of supernatural of mental images that the human nervous systempotency. produces in altered states of consciousness. The pain-

Again, there is, unfortunately, little ethno- ted images thus became part of a complex ritual ofgraphic information on precisely what happened to dancing, singing and clapping that controlled thethese potent images after shaman-artists had made spiritual, or hallucinatory, experiences of shamansthem. On the face of it, it seems unlikely that the and, possibly, other people as well.paintings would have simply dropped out of the am- As time went by, certain rock shelters acqui-bit of Bushman ritual and belief. Indeed, such evi- red more and more of these potent spiritual images.dence as we do have suggests that they continued to In some shelters paintings were done one on top ofplay an important ritual function, another, thus building up multiple lavers of images,

For instance, M said that, if a 'good' person the oldest ones fading into a blurred red background.placed his or her hand on a depiction of an eland, the The potency of paint, the ways in which some pain-potency locked up in the painting would flow into the tings enter or leave the rock surface, and the way inperson, thus giving him or her special powers. To de- which the potency stored in the images could be tap-monstrate how this was done, she arranged my fin- ped, all suggest that some of the most densely pain-gers so that my entire hand was on a depiction of an ted rock shelters must have been regarded as placeseland. As she did so, she cautioned that, if a 'bad' of exceptional potency. Moreover, there can be littleperson did this, his or her hand would adhere to the doubt that the shamans who lived and painted in tho-rock and the person would eventually waste away se shelters enjoyed enhanced prestige and, in moreand die. recent times, political power as well (Dowson 1994).

The importance of touching, and not merely The art was therefore not something separate fromlooking at, rock paintings is supported by evidence daily life but was deeply embedded in the negotiationfrom the Western Cape Province where there are pat- of socio-political relations. The making of paintingsches of paint that have been rubbed smooth (Yates et did not merely reflect those relations. On the con-al. 1991). It is not entirely clear what the patches we- trary, the making of paintings should be seen as are rubbed with, but the smoothness of the rock, parti- highly ritualized way of reproducing, entrenchingcularly in the centre of the patches, is easily dis- and, in some instances, challenging power structures.cerned. Similarly, the making of the positive hand- Seen in the light of the evidence I have pre-prints that are common in some parts of the Western sented, the southern African painted images wereCape was probably closely associated with ritual tou- clearly not just pictures of events or exercises in aes-ching of the rock rather than with the making of thetics. Rather, they were an integral part of a series'pictures' of hands. of complex ritual procedures in which, first, social

There is thus evidence that some of the tensions and, secondly, tensions between socially ac-paintings were not made merely to be looked at, like ceptable visions and those novel visions thrown upworks of art in a Western gallery. After they had by the nervous system were mediated. Ritual, in thisbeen made, they continued to be involved in rituals sense, was 'theatre': some of the 'props' were rock artin ways that we do not fullh understand, but it does images that proclaimed and, indeed, were palpable,seem that physical contact with some of the images tangible evidence for the shamans' access to spiritualfacilitated the acquisition of supernatural potency. realities.

In addition to physical contact, the imageswere important visually as well, but, again, notsimply as Western notions of art suggest. This con- 4. BEYOND SOUTHERN AFRICAtention was borne out by M. She demonstrated how,long ago, Bushmen had danced in the painted rock If the intimate relationship that existed bet-shelter to which she took Jolly and me and how they ween southern African rock art and ritual is to be ta-had raised their arms and turned to the paintings ken as the source of an analogical argument that willwhen they wished to intensife their potency. As they extend this sort of relationship to other rock arts, thedanced and looked at the paintings, potency flowed structure of the argument must -in each case- befrom the images and entered into them. This was, I clearly stated and subjected to scrutiny. I therefore


restrict the 'beyond' of my title to the Upper Palaeoli- ful images was played out, power groups must havethic parietal art of Western Europe, leaving aside the employed ritual to restrain and socialise the expe-rock arts of the Americas. Australia and so forth. riences of individuals. At the same time. rituals mustThis restriction is useful because the structure of the have been developed to present the 'spiritual' expe-analogical argument linking Upper Palaeolithic and riences of those whose visions were sanctioned to thesouthern African rock art has been set out elsewhere rest of the community, that is. to those who did notand need not be repeated in detail here (Lewis- have personal access to the revelations of altered sta-Williams and Dowson 1988. 1992; Lewis-Williams tes of consciousness and who were therefore politi-1991). As in my discussion of Bushman ritual and cally and socially to some extent disadvantaged. Itrock art I concentrate on Upper Palaeolithic parietal seems reasonable to conclude that the whole processpaintings rather than engravings, of the production and consumption of Upper Palaeo-

Briefly, the argument is founded on what I lithic parietal art was ritualised and that these ritualsargue are secure 'relations of relevance' (see, for were, broadly, shamanic.example, Wylie 1988) between altered states of cons- Attempts to discover some of the elementsciousness and a defined progression of mental ima- and forms of these rituals must take full cognizancegery that develops from the luminous. geometric of the temporal and geographical diversity of Upperentoptic phenomena seen in light trance to the Palaeolithic art. The 20000- year period must not beoverwhelming hallucinations of monsters, people homogenised into a replica of any single ethnogra-and emotionally charged objects of deep trance. If phically observed society; multiple analogies and in-depictions formally referable to all stages of this pro- terpretations will be required to build up a multi-gression of mental imagery are found in a rock art, component mosaic that fits the highly diverse empi-there is a strong implication that those images deri- rical evidence of Upper Palaeolithic an. Becauseved, at least in part, from the experiences of altered such a task would require too much detail for a paperstates of consciousness. If it is known that the rock such as this, the suggestions that I make should beart was made by a hunting and gathering society, seen as highly general and in need of temporal andthere is a further implication that the people practi- geographical refinement. I simply suggest somesed shamanism in one form or another. Confidence broad features of the temporal trajectory of Upper Pa-is lent to this form of argument by the southern Afri- laeolithic ritual and art.can case. We know independently of the art (that is.from the ethnography) that its makers practised aform of shamanism. The formal parallels that are 5. UPPER PALAEOLITHIC RITUALdiscernible between southern African rock art ima- AND ARTgery and the mental imagery of altered states of cons-ciousness can therefore be explained by the universa- Taking the four stages that I have identifiedlity of the human nervous system and the ways in in the production and consumption of southern Afri-which it behaves in altered states. It is thus the uni- can rock art as a model, I argue that (1) Upper Pa-versalitv of the human nervous 'system that provides laeolithic peoples' acquisition of imagery, (2) theira 'bridge' between nineteenth-century Bushman rock manufacture of paint, (3) their making of rock artart and Upper Palaeolithic parietal art (see also images. and (4) the ways in which thev subsequentlyBrown 1991). Both arts may be legitimately descri- used these images were all ritualised.bed as 'shamanic'. I now take these previously publis-hed conclusions as read and move on to elaborate 5.1. The acquisition of imagerythem.

The altered states of consciousness that were As Dowson and I have argued (Lewis-implicated in the production of Upper Palaeolithic Williams and Dowson 1988), much, but not all, ofparietal art must have been to some, no doubt tempo- the imagery of Upper Palaeolithic art was acquiredrally and geographically varying, extent institutiona- by shaman-artists in altered states of consciousness.lised. One of the reasons for institutionalising altered During the course of the Upper Palaeolithic the ri-states of consciousness would have been to mediate tuals of image-acquisition almost certainly variedthe tension between the free-flow of mental imager' considerably in. first. the ways in which these ritualsand the specific, limited range of socially sanctioned were socially situated and, secondly, the ways inimages that makes sense in a given society. As the which altered states were induced and experienced.tension between the novel, or aberrant, mental ima- At some times, especially but not exclusi-ges of individuals and the canon of socially meaning- vely at the beginning of the Upper Palaeolithic, and


in some places these rituals of image-acquisition we- to make large images nearer the entrances, was pro-re probably communal. Like the Bushmen's trance bably considered to have special, supernatural pro-dance, these rituals probably brought together large perties, as were at least some of the Bushmen'snumbers of people, and shamans obtained their vi- paints. At present we know that Upper Palaeolithicsions in the midst of society,. witnessed not only by people took great care with the manufacture of paintother shamans but also by ordinary people. Under and that they used different recipes (e.g., Clottes etsuch circumstances, there was a direct, visible asso- al. 1990; Clottes 1993; Ballet et al. 1979; Vandiverciation between the acquisition of visions and the 1983; Leroi-Gourhan and Allain 1979; Lorblanchetbody social. Ordinary people supported the shamans et al. 1990). These recipes should be considered toand encouraged them in the face of spiritual hazards. see if they contain any evidence for rituals. WhateverSuch power relations as were under-written by the such a study may reveal, it seems likely that Upperpossession of visions were thus reproduced and en- Palaeolithic people would have considered the mate-trenched by direct, virtually unmediated relations rials necessary for the fixing of visions to have pro-between shamans and ordinary people. perties commensurate with the potency of the visions

In such communal circumstances, the va- themselves and that the preparation of paint wouldrious participants would have experienced a range of have become hedged around with prohibitions andaltered states of consciousness. Those who were most rituals that would have defined social relations.intensely seeking visions may have used psychotropicdrugs to induce deep trance. Others. caught up in the 5.3. The making of rock paintingsritual dancing and music, believed that they couldshare some but not all of the insights that the leading The sheer quantity of paint needed for ma-shamans were experiencing. Still other people, on king large images seems to imply rituals of somethe fringe of the activity, were probably less intensely sort. The images in the Hall of the Bulls and theswept along by the ritual:; they experienced euphoria Axial Gallery at Lascaux, for instance, demandedand ecstasy but did not themselves see visions. On large quantities of paint and the construction of scaf-the other hand, some people may have felt themsel- folds (Leroi-Gourhan and Allain 1979). The makingyes to be opposed to the whole ritual procedure. of these-images therefore involved the active partici-

More complex social and ritual relations- pation of a large number of people who went out tohips are suggested by those Magdalenian images that find the pigment and, possibly, the medium, whoare deep underground and can be viewed by only a brought them back to the cave and mixed them, whofew people, perhaps only one person, at a time (Ben- built the scaffolds and, finally, who applied paint toder 1989; Lewis-Williams and Dowson 1993). These large areas of rock surface. The rituals that attendedremote images imply something like a vision quest, the making of these paintings were, I argue, differentas practised by, for instance, North American shama- from those performed in the depths of the caves, as,nic groups (cf. Whitley 1992), but not by the Bush- for example, in the Chamber of the Felines at Las-men. During the Magdalenian, shamans in search of caux, where images were delineated by only a fewrepeated visions and novices seeking them for the strokes. Communally produced art needs to be distin-first time seem to have separated themselves from so- guished from individually produced art; both kindsciety and, in the remote, dark, silent recesses of the of art imply rituals that define and reproduce socialcaverns, sought the altered states of consciousness relations, but they' do so in different ways.that would provide their visions. In some instancesthe often hastily executed rock art images of the re- 5.4. The ritual use of Upper Palaeolithicmote areas may have been made in a light state of al- rock arttered consciousness as the questers' mental imagerywas projected onto rock surfaces rather like a slide or A further implication concerns the ways infilm show (Lewis-Williams and Dowson 1988); in which painted and engraved images were used. Theother cases, the questers probably made images after large, impressive images in such places as the Hall ofthey had reverted to a normal state of consciousness, the Bulls at Lascaux or the Salon Noir at Niaux pro-as did the Bushmen. bably performed a function similar to many Bush-

man rock art images. Placed at or within compara-5.2. The manufacture of paint tively easy reach of the entrances to Upper Palaeoli-

thic caves, these images probably prepared questersThe paint that vision questers took with for the visions that they would see in the depths of

them into the depths of the caves, as well as that used the caves. As I have said, people hallucinate what


they expect to hallucinate (or what they have been 5.6. Ritual cosmology and powerdeliberately led to believe they will hallucinate), andwe may postulate the performance of preparatory ri- Although there were almost certainly majortuals that dramatically revealed communally made differences between the rituals associated with pain-entrance art to vision-seekers (and to others as well) tings and engravings near or at the entrances to Up-in an attempt to inform the hallucinations they would per Palaeolithic caves and those situated in theexperience in remote solitude. Such rituals would ha- depths, art in all parts of the caves displays a charac-ve reduced the personal element by controlling, to teristic that suggests an important, fundamental corn-some extent, the range of hallucinations that the ner- monality. In all parts of the caves, artists exploitedvous system generates and by alerting questers to natural features of rock surfaces in ways that implyonly certain kinds of mental imagery. Indeed, social that they believed the animals they. were depictingcontrol of altered states of consciousness lies at the existed behind the rock face; their task was to enticeheart of shamanism. these spirit-animals through the rock so that they

could establish a spiritual relationship with them that5.5. Ritual, sound and art would empower them to perform their shamanic

tasks (Lewis-Williams in press b). The Upper Pa-Another common component of ritual deser- laeolithic cosmos was, as in many other shamanic so-

ves mention. Sound plays an important role in sha- cieties, probably conceived of as tiered. Within suchmanic rituals, not only amongst Bushman groups but a cosmology, the spirit world was probably believedworldwide (for a review see Dobkin de Rios and Katz to be underground, and journeys into caves were pro-1975). Rhythmic and audio driving induce altered bably believed to be journeys into that realm. Ritualsstates of consciousness and provide a framework for formalising the manufacture of paint, the making ofa visionary's concentration. Musical instruments are painted images and, indeed, the uses to which thealso used to imitate sounds made by animals and images may have been subsequently put were pro-birds. There is evidence that music, or at any rate bably all posited on the existence of a subterranean,sound, was a component of Upper Palaeolithic ri- animal-filled realm that held the supernatural po-tuals. A number of 'flutes' have been found in West tency that sustained the universe and, more espe-and East European Upper Palaeolithic sites. Further, cially, that shamans sought and harnessed for theHuyge (1991) argues that the 'horn' held by the so- good of society.called Venus of Laussel is a scraped idiophone. The- Control of this potency was reserved andse and other instruments, such as bull-roarers and protected for limited numbers of people by means ofdrums, could have been used to suggest the presence rituals. It seems that, during the course of the Upperof animals as well as to provide a hypnotic rhythm Palaeolithic, as the depths of the caverns were in-(see also Waller 1993). Given the Upper Palaeolithic creasingly explored, society became more and moreunderstandings of the underworld and the rock face complex and hierarchical, with political power beingfor which I have argued, the suggestion that sounds increasingly concentrated in the hands of a few sha-were produced by striking stalactites is particularly mans (Lewis-Williams and Dowson 1993). The who-interesting (e.g. Dams 1984, 1985; cf Needham le complex ritual process of acquiring visions,1967, and Tuzin 1984): some of the struck stalactites making rock art and using the images was implicatedwere marked with geometric motifs that probably de- in power struggles, as was Bushman art during therived from entoptic phenomena seen in an early stage nineteenth century (Dowson 1994).of trance. I argue that the striking of stalactiteswould, in itself, have been a way of arousing andcommunicating with the spirit world that would have 6. FUTURE WORKworked together with paintings, singing and dancingto constitute complex ritual sequences. An associa- Taking southern African rock art as a star-tion between sound and art is also suggested by Rez- ting point, I have shown that the acquisition of vi-nikoff and Dauvois (1988) who argue that there is a sions, the making of paint, the placing of images oncorrelation in some Ariege caves between areas of re- the walls of rock shelters and the subsequent use ofsonance and the presence of rock art. Maximum re- the paintings were stages in the production and con-sonance, they claim, is achieved by the human voice, sumption of Bushman rock art and that these stagesand this leads them to suggest that chanting may ha- were all embedded in rituals that reproduced socialve been part of rituals that were associated with rock relations. Within this framework there was, never-art (Scarre 1989). theless, room for individual Bushman shamans to ac-


quire novel visions and to use these visions and the forth to build up an account of the changing associa-representations that they made of them to subvert so- tion between art and ritual in the Upper Palaeolithiccial relations for their own political ends. of Western Europe.

Further, I argue that Upper Palaeolithic pa-rietal art was, like southern African rock art, sha- Acknowledgementsmanic and that many of the images in the caves ofWestern Europe depict visions. The same series offour stages in the production and consumption of I am grateful to colleagues who read drafts

of this paper and offered useful comments: Geoffthese depictions as I identified in the production and of ts p and offerd useful comments: Geoff

Blundell, Thomas Dowson and Anne Holliday. Deni-consumption of southern African rock art probably unde Thomas Dowson and Anne Hoda De-obtained in Western Europe. As in southern-Africa. se Voorvelt and Anne Holliday kindly typed successi-each stage was probably set in a ritual context; in- ve drafts. The Rock Art Research Unit is funded by

the University of the Witwatersrand and the Centredeed, it would be hard to imagine that the 'fixing' the the Witwatersrand and the Centrefor Science Devielopment; neither institution is res-of potent, supernatural images was not ritualised. If or Scence lopment; neither institution is res-

as Dowson and I have argued (Lewis-Williams and ponsible for the views epressed here. I am gratefulto numerous people who made it possible for me toDowson 1988), altered states of consciousness lay at to numerous p whomade t ossble fo me tovisit a large number of Upper Palaeolithic caves inthe heart of Upper Palaeolithic religion, those states vst a large number of Upper Palaeoltc s

must have been institutionalised and ritualised. Fi- France: they include Norbert Aujoulat Paul Bahn,nally, the series of rituals associated with the produc- Robert B 0goun, Jean Clottes, Brigine Delluc, J.

Gaussen, Andrd Leroi-Gourhan, Michel Lorblanchet,tion and consumption of rock art must have changed Pe , n -e Ra, D achiduring the course of the Upper Palaeolithic as social ler Jean-P igaud, D. Sacchi and

Georges Simonnet.relations changed. Simonnet.These very general observations and hypo-

theses need to be refined and tested. First, we mustreject Romantic notions of 'the artist' as an isolatedgenius driven by some powerful aesthetic imperati-ve.Instead. we should construct hypotheses that si-tuate the production and consumption of Upper ....Palaeolithic art in the changing social and economic 'Bushman communities speak many mutually unintelligible langua-

processes of the time in which it was made (cf. ges. There is therefore no single Bushman word to cover all groups."Bushman" is, in the view of some writers, a pejorative and sexist

Lewis-Williams 1982). Secondly, if we accept this word, although some of the people themselves do choose to use it.theoretical position, we can start using multiple eth- "San", a Nama (Khoekhoen) word, is preferred by many, but by no

nographic analogies with strong relations of relevan- means all, academic writers. Unfortunately, it means something like"vagabond" and is therefore also pejorative. There is no unanimity on

ce, the distribution of art within the caves, data which word should be used In using "Bushman", I explicitly rejectretrieved from excavations. paint analyses and so any pejorative or sexist connotations.

REFERENCES

ARBOUSSET, T. & DALUMS, F. (1846): Narrative of an BIESELE, M. (1978): Sapience and scarce resources:exploratory tour to the north-east of the colony of communication systems of the !Kung and otherthe Cape of Good Hope. Translated from the foragers. Social Science Information 17: 921-French of the Rev. T. Arbousset, by John Croum- 947.bie Brown. Robertson. Cape Town. BIESELE, M. (1993): "Ibmen like meat": the folklore

BALLET, O.; BOCQUJET, A.; BOL-CHEZ, R.; COEY, J.M.D. & and foraging ideology of the Kalahari Ju/'hoan.CORTU, A. (1979): Etude technique de poudres co- Witwatersrand University Press. Johannesburg.lorres de Lascaux. In A. Leroi-Gourhan & J. BLEEK, D.F. (1935): Beliefs and customs of the/XamAllain (eds.): Lascaux inconnu. CNRS. Paris: Bushmen. Part VII: Sorcerors. Bantu Studies 9:171-174. 1-47.

BENDER, B. (1989): The roots of inequality. In D. Mi- BROWN, D.E. (1991): Human universals. McGraw-l1er, M. Rowlands & C. Tilley (eds.): Domination Hill. New York.and resistance. Unwin and Hyman. London: CAMPBELL, C. (1986): Images of war: a problem in83-93. San rock art research. World Archaeology 18:


255-268. / cial context of southern San rock art. Current An-CLorEs, J. (1993): Paint analyses from several Mag- thropology 23: 429-449.

dalenian caves in the Ariege region of France. LEWIS-WILLA.Ms, J.D. (1986): The last testament ofJournal of Archaeological Science 20: 223-235. the southern San. South African Archaeological

CLOTyES, J.; MENU, J. & WALTER, P. (1990): New light Bulletin 41: 10-11.on the Niaux paintings. Rock Art Research 7: 21- LEWIS-WILLIAMS, J.D. (1987): A dream of eland: an26. unexplored component of San shamanism and

DAMS, L. (1984): Preliminary findings at the 'organ' rock art. WobrldArchaeology 19: 165-177.sanctuary in the cave of Nerja, Malage, Spain. LEWIS-WILLIAMS, J.D. (1990): Discovering southernOxford Journal ofArchaeology 3: 1-14. African rock art. Cape Town: David Philip.

DAMS, L. (1985): Palaeolithic lithophones: descrip- LEWIS-WILLIAMs, J.D. (1991): Wrestling with ana-tions and comparisons. Oxford Journal of Ar- logy: a problem in Upper Palaeolithic art re-chaeology 4: 31-46. search. Proceedings of the Prehistoric Society 57

DoBnIcr DE Rios, M. & KATZ F. (1975): Some relations- (1): 149-162.hips between music and hallucinogenic ritual: the LEwIs-WrILLIAMS, J.D. (In press, a): Perspectives and'jungle gym' in consciousness. Ethos 3: 64-76. traditions in southern African rock art research.

DowsON, T.A. (1988): Revelations of religious rea- In K. Helskog (ed.): Proceedings of the Alta Con-lity: the individual in San rock art. World Ar- ference on RockArt.chaeology 20: 116-128. LEWIS-WILLIAMS, J.D. (In press, b): The incorporation

DowsoN, T.A. (1992): Rock engravings of southern of natural features of rock surfaces into SouthernAfrica. University of the Witwatersrand Press. African and West European Upper PalaeolithicJohannesburg. parietal art: a cross-cultural, cross-temporal study

DOwsON, T.A. (1994): Reading art, writing history: of religious belief. L'Anthropologie.rock art and social change in southern Africa. LEWIS-WILLAMs, J.D. & DOWSON, T.A. (1988): SignsWorldArchaeology 25(3): in press. of all times: entoptic phenomena in Upper Pa-

GUENTHER, M.G. (1975): The trance dancer as an laeolithic art. CurrentAnthropology 29: 201-245.agent of social change among the farm Bushmen LEWIS-WILLIAMS, J.D. & DOWSON, T.A. (1989): Imagesof the Ghanzi District. Botswana notes and re- ofpower: understanding Bushman rock art. Sou-cords 7: 161-166. them Book Publishers. Johannesburg.

GULBRANDSEN, O. (1991): On the problem of ega- LEWIS-WILLIAMS, J.D. & DOWSON, T.A. (1990):litarianism: the Kalahari San. In R. Gronhaug, G. Through the veil: San rock paintings and the rockHaaland, & G. Henriksen (eds): The ecology of face. South African Archaeological Bulletin 45:choice and symbol: essays in honour of Fredrik 5-16.Barth. Alma Mater Forlag. Bergen: 82-110. LEWIS-WILLIAMS, J.D. & DowsON, T.A. (1992): Art ru-

How, M. W. (1962): The mountain Bushmen ofBasu- pestre San et Paleolithique superieur: le lien ana-toland. Van Schaik. Pretoria. logique. L :Anthropologie 96 (4): 769-790.

HUYGE, D. (1991): The 'Venus' of Laussel in the light LEWIS-WILLIAMS. J.D. & DOWSON, T.A. (1993): On vi-of ethnomusicology. Archeologie in Vlaanderen sion and power in the Neolithic: evidence from1: 11-18. the decorated monuments. Current Anthropology

JOLLY, P. (1986): A first generation descendant of the 34: 55-65.Transkei San. South African Archaeological Bu- LORBLANCHET, MVI.: LABEAU. M.; VER,_ET. J.L.: FITTE. P.:lletin 41: 6-9. VALLADAS, H.; CACHIER, H. & ARNOLD. M. (1990):

KATZ, R. (1982): Boiling energy: community-healing Palaeolithic pigments in the Quercv. France.among the Kalahari 'Kung. Harvard University Rock Art Research 7: 4-20.Press. Cambridge. MARSHALL, L. (1969): The medicine dance of the

LEROI-GOURHAN, A.; ALLAIN, J. (1979): Lascaux incon- !Kung Bushmen. Africa 39: 347-381.nu. CNRS. Paris. NEEDHAM, R. (1967): Percussion and transition. Man

LEWIS-WILLIAMS, J.D. (1980): Ethnography and ico- 2:606-614.nography: aspects of southern San thought and ORPEN, J. M. (1874): A glimpse into the mytholog, ofart. Man 15: 467-482. the Maluti Bushmen. Cape Monthly Magazine (n.

LEW.s-WILLIAMS, J.D. (1981): Believing and seeing: s.) 9(49): 1-13.symbolic meanings in southern San rock pain- PRINS, F.E. (1990): Southern-Bushman descendantstings. Academic Press. London. in the Transkei - rock art and rain-making. South

LEWIS-WILLIAMS, J.D. (1982): The economic and so- African Journal of Ethnology 13: 110-116.


REZN.zOFF, I. & DAUVOIS, M. (1988): Bulletin de la So- cal Journal 2: 89-113.ciete Prehistoriques Francaise 85: 238-246. WILMSEN, E.N. (1989): Land filled with flies. The

SCARRE, C. (1989): Painting by resonance. Nature Chicago University Press. Chicago & London.338: 382. WINKELMAN, M. & DOBKIN DE Rios, M. (1989):

SCHMIaDT, S. (1973): Die Mantis religiosa in den Psychoactive properties of !Kung Bushman medi-Glaubensvorstellungen der Khoesan-Volker. Zeit- cine plants. Journal of Psychoactive Drugs 21:schrift fur Ethnologie 98 (1): 102-127. 51-59.

TuzN, D. (1984): Miraculous voices: the auditory ex- WOLFF, J. (1981): The social production of art. Mac-perience of numinous objects. Current Anthropo- Millan. London.logv 25:579-596. WYLE, A. (1988): 'Simple' analogy and the role of

VANDIVER, P. (1983): Palaeolithic pigments and pro- relevance assumptions: implications ofarchaeolo-cessing. Unpublished M.Sc. thesis. M.I.T. Bos- gical practice. International studies in the philo-ton. sophy of science 2: 134-150.

WALLER. S.J. (1993): Sound and rock art. NVature 363: YATES, R.; A. MANHIRE. (1991): Shamanism and rock501. paintings: aspects of the use of rock art in the

WHITEY, D.S. (1992): Shamanism and rock art in far south-west Cape, South Africa. South African Ar-western North America. Cambridge Archaeologi- chaeological Bulletin 46: 3-11.

rock art and ritual: southern africa and beyondmsu.ac.zw/elearning/material/1299832184lewis-williams...

Documents