The Antiquity of Man

   Home   |   About   |   What is Evolution   |   World Archaeology & Palaeoanthropology   |   North African Archaeology   |   Pseudoscience   |   Recommended Readings   |   Reviews   |   Links   |   Search   |   Contact

Anne Solomon. 1999. Magic, diversity and the ethnographic contexts of San rock paintings in southern Africa. IRAC conference

Diversity in San art has long been recognised and described, but the significance of this diversity remains poorly studied and understood. The problem of diversity and how it is to be interpreted has, however, come further to the fore in various ways in recent research.

Archaeologists studying San rock art have tended to analyse it primarily in terms of iconography (subject matter), whereas the stylistic and formal dimensions of the art have received less archaeological attention (cf. Skotnes 1991,1994). Art historians, on the other hand, emphasise the visuality of the rock art, focussing rather on graphic devices, style, spatiality, etc.(cf. Davis 1985, 1990,1996). San art, from the Cape to Tanzania, does display a number of similarities in subject matter (such as the predominance of humans and large herbivores, but in terms of style and form it is far more diverse. Art historians in particular have drawn attention to this formal diversity, with some researchers suggesting that "San art" in fact comprises "different arts" (Battiss 1948, Skotnes, pers.comm.). The extent to which one "sees" diversity, and the extent to which it is seen as "meaningful" are thus dependent on analytical methods and presuppositions (cf. Solomon 1998,1999).

I have considered the problem of diversity in various ways in my recent research. For example, are different treatments of a given motif "meaningful", or are they to be seen merely as "variation", idiosyncratic (cf. Lewis-Williams and Dowson 1989) and individual (cf. Dowson 1988)? Art historical approaches lean towards seeing different visual treatments as more significant (or potentially more significant) than do archaeological approaches, and integrating these insights has been one focus of my research. I have also been considering differences in the art of different regions as a product of historical differences between San painting communities. Recent work on contact and interactions between San hunter-gatherers and pastoralist or farming neighbours suggests that this is a necessity (e.g. Yates et al. 1994, Jolly 1996, Prins and Woodhouse 1996). Another question which informs my research is: In view of the well-documented diversity in the parietal art, which spans at least 3 500 years (Jerardino and Yates *), are we justified in interpreting it all within a single explanatory frame (which in the case of San art, has been shamanism and a so-called "shamanistic cosmology")?

Diversity, ethnography and analogy

The long-standing reliance on analogies derived from ethnographies of recent and contemporary Kalahari San (as a way of illuminating the art(s) of southern San art from earlier times) has implications for the ways in which diversity is understood. The analogical method, as developed particularly by Lewis-Williams (e.g. 1980, 1981; see also Stow 1905, Vinnicombe 1976, Lee and Woodhouse 1970 and various other works which utilise ethnographies), seeks and, by definition, depends on drawing out, similarities, whilst differences tend to be under-examined. The analogical method is therefore implicated in creating a picture of San societies, and San practices, as more similar than they probably were. Consideration of the ethnographies of documented San societies suggests that important differences exist(ed) between different groups in different times and places, and that this has implications for the use of ethnographic analogies, and for explanation of a varied body of art in near-identical terms.

Analogies at some level are undoubtedly valid. As Barnard (1992:3) has written, "In spite of differences associated with their subsistence pursuits, many otherwise diverse Khoisan peoples share a great number of common features…these features are not randomly distributed; nor have they simply diffused from one group to another as single culture traits. They represent structures held in common across economic, cultural, linguistic and 'racial' boundaries". In rock art research, analogies are drawn between groups such as the 20th century !Kung or (Ju/'hoasi) and Nharo of Botswana and Namibia, the 19th century /Xam and Maluti San in South Africa, and the prehistoric painters. It is argued that these groups are united by the same, or essentially similar "shamanistic cosmology" (e.g. Lewis-Williams 1998). This then permits analogies to be made. "Cosmological similarities", however, are not an unproblematic basis for analogy.

Similarities and differences in belief/cosmology and magical practices

Ethnographically recorded San societies clearly share a mythological tradition, as well as belief in a greater and lesser god, whilst magical practices focus on interactions with the spirits, who are seen as the source of illness and misfortune. At this most general level, diverse groups apparently do indeed share a 'cosmological' frame. However, not only are there many differences within this frame, but the ethnographic evidence suggests that there were different ways in which this generalised cosmology was actualised in practice. For example, although the !Kung and /Xam ethnographies both describe the existence of a greater and lesser god, there are nevertheless considerable differences in the ways in which they are conceptualised (see Barnard 1992 for an overview). For example, the Kung 'lesser god' does not appear in animal form, as the Rain Bull, as does the /Xam lesser deity.

Similarly, the mythological traditions of the various groups all centre on the initial creation of the world by the 'great god', and a phase when humans and animals had not been separated; animals could talk, and the first San had animal-like characteristics, insofar as they were said to lack customs, manners and social sense (Bleek and Lloyd 1911, Hewitt 1986, Guenther 1989, Woodhouse 1984). Guenther (1989) has dubbed this phase "primal time". After an event which Guenther refers to as "the second creation", animals and humans were separated; animals became animals and prey and lost their capacity for speech, whereas humans became fully cultured and socialised, but lost their prior immortality. Though the broad outlines of the myth tradition are very similar, the details are highly variable from group to group, even on a smaller scale than that of southern Africa. For example, stories describing the second creation are completely different amongst the 19th C /Xam and Maluti San (both in South Africa), even though both narratives describe the same mythological transition. In other words, though the forms of belief appear similar, the "content" may be very different.

In another dimension, although the broad contours of !Kung and /Xam models of illness are, in many ways, very similar, there appear to be some crucial differences in magical practices. As Bahn (1997) has also noted, the 19th C /Xam texts nowhere describe a communal trance dance comparable to the !Kung's. Accounts of healing do not mention group involvement, and the /Xam account of a dance known as the //k:en dance (DF Bleek 1935:11-15) does not mention curing, though it mentions bleeding from the nose and learning of 'sorcery'. All that can be deduced from the description of the //k:en dance is that it probably refers to a supernatural potency, and is affiliated in some way to magic.

Another problem that arises in comparisons of !Kung and /Xam curing is the argument that both are shamanistic (despite the statement by Katz (1982), author of the definitive study of !Kung curing, that they have "no shamanistic tradition"). One of the /Xam accounts that is said to describe the experiences of a "shaman" in trance apparently refers to spirit possession. In Lewis-Williams' reading, it describes a man, in the throes of violent trance, 'turning into' a lion, sprouting lion's hair and then being rubbed with fat by other people to normalise him (Lewis-Williams 19*). However, it is better interpreted as describing how a curer took a spirit from an afflicted person into his own body and battled with it. It is simply not clear from the account whether the lion's hair belongs to the spirit; whether the possessed man takes on aspects of what has possessed him (in my view, the most plausible interpretation); or, for that matter, whether the account is metaphorical. It is only by inference, and by analogy with Northern San trance performance that it can be interpreted in terms of trance and shamanic experiences. This does not deny that spirit possession may involve altered states; but it does reject the primacy accorded to ASCs in interpretations of San art.

A number of other references in the /Xam texts to spirits and their doings are interpreted within the shamanistic school of explanation as referring to shamanic experiences, but, as I have argued elsewhere (Solomon 1997), the majority of these describe the activities of the spirits, and not shamanic experiences in altered states. In this respect, I support the earlier interpretations of Pager (e.g. 1971,1975a,b,1994 [1984]), Woodhouse (e.g. 1984) and others who have suggested that therianthropic figures in the art represent both mythological beings - the beings from primal time, when humans and animals had not been separated - and spirits of the dead. Commentary on rock paintings of therianthropes by Qing, a 19th century Maluti San man, to the British official, Joseph Orpen (Orpen 1874), supports this interpretation (though Lewis-Williams (1998) disputes this). The details of these arguments are being further developed in ongoing research.

The gist of my perspective on San art is that it whilst a great deal of it probably does relate to relations with the spirit world, this does not necessarily indicate shamanism, and the emphasis accorded to ASCs is problematic (though altered states may have been entered, they are not necessarily central, nor the source of the imagery in art or myth). Rather, I contend that religious beliefs and practices focus on spirit possession (an anthropologically distinct entity), and a variety of magical practices arising from belief in the influence of these spirits who are influential after death; the spirits include both the beings from primal time, who were vanquished at the time of the second creation, and the spirits of dead people. The Maluti San (Orpen 1874) and /Xam testimonies (Bleek and Lloyd 1911, unpublished materials in the Bleek and Lloyd Collection) - especially the former - apparently emphasise the role of the mythological beings (although the /Xam texts also refer to deceased relatives (see below). It is important to note that even in the Kalahari today, different groups emphasise different aspects of the cosmology which appears to be held in common more widely. Elaboration of trance performance amongst the Nharo in Botswana, for example, has been analysed in terms of their status as de-culturated "farm Bushmen", and conscious efforts to assert a traditional identity (Guenther *). Similarly, the /Xam texts emphasise female far more than other recorded groups initiation (there are no less than 16 menarcheal tales in the Bleek and Lloyd Collection [Hewitt 1986]). Such differences of emphasis must surely be seen as products of historical differentiation in the beliefs of the various groups, irrespective of general cosmological principles held in common.

Magical practices recounted by /Xam narrators to the Bleek and Lloyd family have been interpreted by Lewis-Williams and his colleagues as "shaman[ist]ic", on the basis of various traits as outlined by Eliade (1964). However, as I have mentioned, except for accounts of curing (which apparently describe spirit possession, not shamanism as such), the majority of these accounts may be interpreted as describing the activities of the spirits, not living curers. Accounts of rain 'sorcerors', game 'sorcerors' and even some of the personages described as 'sorcerors of illness' may all be interpreted as spirits of the dead of various kinds. Rain and game sorcerors were apparently benevolent deceased relatives who might be called upon, formally or informally, for assistance, whereas sorcerors of illness were the disease-causing spirits. Accounts of the dead game sorcerors' control over animals have been interpreted in the shamanistic model as evidence for the shamanic "master of animals", whereas references to 'underwater' in accounts of rain-making have been interpreted in terms of the shaman's experiences of disorientation in a trance state. However, 'underwater' refers rather to the death realm, the subterranean home of the spirits, and is not amenable to a interpretation primarily in terms of trance: the /Xam texts explicitly state that rain-makers were dead people (DF Bleek 1933:305, Solomon 1997). Similarly, some /Xam texts which have been seen as evidence for trance and shamanism appear to describe sorcery and "black magic" (Solomon 1997). The /Xam may be more analagous to people of Khoisan extraction, in a neighbouring area in the 1950s than they are to the Kalahari !Kung at the same time. These people's magical repertoire was described by Carstens (1966), and centred on blikdraers (practitioners of black magic) and bossiesdokters, healers who counteracted the blikdraers' sorcery (cf. Barnard 1992:197).

The interpretation of /Xam accounts as describing 'shamanic' experiences has meant that all these activities have been seen as different aspects of the same phenomenon, or variations on the same theme. However, it seems that /Xam magical practices comprised a range of activities, centring on interactions with the spirits, but nevertheless varied and distinct from curing as such. As far as can be ascertained from the available ethnographies, the /Xam accounts do not describe a communal trance dance as recorded from the Kalahari, but individual curing, as well as informal appeals to the game and rain sorcerors, and black magic. Magical activities may be differentiated according to whether they are part of group communal ritual, or are private, informal, non-ritual or opportunistic. Recognition of such differences in the magical repertoire of a single group leads to the proposition that diversity within the art of a single region (or even site) may relate to site specialisation, i.e. the use of different types of sites, in different locations, for different purposes. Some of these may be archaeologically identifiable today.

For example, I have previously suggested that the site of Keurbos, which - very unusually - contains predominantly female figures, including some in a sexually explicit hand-to-groin posture, and no male figures whatsoever - may relate specifically to female initiation. Similarly, researchers have long regarded some sites (e.g. Bamboo Hollow; Gxalingenwa Shelter 1) as specialised rain-making sites. (Of course, many sites are so complex that painting episodes cannot be distinguished in this way, but that does not destroy the utility of the exercise). I have been especially interested in the figures which Pager (e.g. 1971) called alites (also known as flying buck, or in Lewis-Williams' work, 'trance buck'). I accept Pager's interpretation of them as spirits of the dead, rather than that of Lewis-Williams and his colleagues, who interpret them as representing the shaman's experiences in deep trance. A characteristic of these figures in the KwaZulu-Natal Drakensberg is that they are often found as isolated or quasi-isolated figures (i.e. in isolated small panels), in non-habitation sites. Not only do most of these sites provide no shelter, but they are also unsuitable for groups of people (some barely provide standing space). As such, it seems problematic to relate them to communal ritual in which the whole group (or a substantial portion thereof, e.g. one or other gender) participates.

In the light of the argument that I have summarised, viz. that southern San art and some of the diversity within it may be linked to a variety of different magical practices, I suggest that such sites represent individual and/or informal and/or opportunistic magical activities - specifically engagement with the spirits. It may be that an unsuccessful hunter or gatherer might have painted when failing in the food quest, as a component of asking benevolent deceased relatives (the game sorceror spirits) for assistance. Some might mark a place where a spirit was seen, or was believed to dwell. Other sites might be sorcery sites, as in the Australian art (cf. Layton 1992). Though there are no ethnographic accounts concerning the motivations for painting, it is possible - and desirable - to model a range of possible scenarios within which painting may have occurred, and a range of reasons for producing the visual imagery.

It may also be possible - in some instances - to correlate 'stylistic' factors, iconography and site type/location in a multi-factorial analysis, and to address the question of 'virtuosity', which is seldom considered in archaeological and iconographical studies (Skotnes, pers. comm.) If painting was potentially carried out by any individual (not just medicine people) as part of informal magical activity, then it might also be possible to factor in graphic skill. In other words, one might hypothesise that informal activity may correlate with monochrome paintings of lesser virtuosity, in particular types of sites, whereas some of the elaborate, polychrome panels, in large habitation sites - paintings which show extraordinary virtuosity - may be correlated with formal ritual, and the participation of more specialised artists. Undoubtedly there will be problems with such an approach, but as a way of addressing the theoretical problems which arising from separating form and content (cf. Davis 1990), it seems a useful avenue to explore.


Diversity in San rock art may indeed by understood in terms of individual idiosyncrasy, but it is also possible and desirable to consider diversity in relation to both historical differences amongst the prehistoric artists in different times and places, and, on a sub-regional basis, in relation to different magical practices in the repertoire of a single group. Explanation of most, or all "San art" in a single frame, without accommodation of "diversity ' in its several dimensions, is problematic. The root of the problem, in my view, lies partly in the ethnographic method itself, by means of which parallels in 'belief' etc. seem to collapse into near-identicality. I would like to conclude by referring again to Barnard's analysis of similarities, and the argument that "especially in the case of the Bushmen…religious notions have a fluid character which has led historically to cross-cultural uniformity, and at the same time, to intra-cultural diversity" (Barnard 1992:3,263-4). The essence of his argument, as I understand it, is that similarities between the various Khoisan groups are the product of historical sedimentation. Religious belief is notoriously conservative, and less subject to local adaptation than aspects of subsistence activities, and similarities in San 'cosmology' may be little more than a reflection of this. As such, similarities do not necessarily indicate that San practices were really similar or the same (this may be so in some instances, but further demonstration of the reality of similarities is required in each case). It is necessary to demonstrate that similarities at one level (e.g. 'cosmology') translate into similarities at another (e.g. ritual). Though the forms of Khoisan religion are similar, this does not guarantee that their 'content' is identical; the issue of form and content is linked to problems of qualitative analysis which deserve further attention. We need to be more aware of differentiations in various dimensions, in order to arrive at a less general, more nuanced and more historical understanding of San art.


Battiss, W. 1948. The artists of the rocks. Pretoria: Red Fawn Press.

Bahn, P. 1997. Membrane or numb brain: a close look at a recent claim for shamanism in Palaeolithic art. Rock Art Research 14(1):62-68.

Barnard, A. 1992. Hunters and herders of southern Africa. Cambridge: Cambridge University Press.

Bleek, W.H.I. 1874. Remarks by Dr Bleek. in Orpen (1874). Cape Monthly Magazine 9:10-13

Bleek, W.H.I. and Lloyd, L. 1911. Specimens of Bushman folklore. London: George Allen.

Bleek, D.F. 1933b. Beliefs and customs of the /Xam Bushmen. Part VI: Rain-making. Bantu Studies 7:375-392.

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

Bleek, D.F. 1936. Beliefs and customs of the /Xam Bushmen. Part VIII: More about sorcerors and charms. Bantu Studies 10:163-199.

Carstens, W.P. 1966. The social structure of a Cape Coloured reserve. Cape Town: Oxford University Press.

Davis, W. 1985. Present and future directions in the study of rock art. South African Archaeological Bulletin 141:5-10.

Davis, W. 1990. The study of rock art in Africa. In P. Robertshaw (ed). A history of African archaeology, pgs. 271-95. London: James Currey; Portsmouth (N.H.): Heinemann.

Davis, W. 1996. The deconstruction of intentionality in archaeology. In: W. Davis, 1996. Replications: archaeology, art history and psychoanalysis. Pennsylvania: Pennsylvania State University Press.

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

Guenther, M.G. * Guenther, M.G. 1989. Bushman folktales: oral tradition of the Nharo of Botswana and the /Xam of the Cape. Stuttgart: Franz Steiner Verlag Wiesbaden.

Hewitt, R.L. 1986. Structure, meaning and ritual in the narratives of the southern San. Hamburg: Helmut Buske Verlag (Quellen zur Khoisan-Forschung).

Jerardino, A. and Yates, R. 199*. A fortuitous fall:

Jolly, P. 1996. Symbiotic interaction between black farmers and south-eastern San: implications for southern African rock art studies, ethnographic analogy and hunter-gatherer cultural identity. Current Anthropology 37(2):277-306.

Katz, R. 1982. Boiling energy: community healing among the Kalahari !Kung. Cambridge, Mass.: Harvard University Press.

Layton,R. 1992. Australian rock art: a new synthesis. Cambridge: Cambridge University Press.

Lee, D.N. and Woodhouse, H.C. 1970. Art on the rocks of southern Africa. New York: Charles Scribner and Sons.

Lewis-Williams, J.D. 1980. Ethnography and iconography: aspects of southern San thought and art. Man (n.s.) 15:467-482.

Lewis-Williams, J. D. 1981. Believing and seeing: symbolic meanings in southern San rock paintings. London: Academic Press.

Lewis Williams* Lewis-Williams, J.D. and Dowson, T. 1989. Images of power: understanding Bushman rock art. Johannesburg: Southern Book Publishers.

Lewis-Williams, J.D. and Dowson, T. 1994. (eds.), Contested images: diversity in southern African rock art research: 331-371. Johannesburg: Witwatersrand University Press.

Orpen, J.M. 1874. A glimpse into the mythology of the Maluti Bushmen. Cape Monthly Magazine 9:1-13.

Pager, H. 1971. Ndedema. Graz: Akademische Druk.

Pager, H. 1975a. The antelope cult of the prehistoric hunters of South Africa. In: E. Anati (ed). Actes du Symposium International sur les religions de la prehistoire, pgs. 401-421. Capo di Ponte: Centro Camuno di Studi Preistorici.

Pager, H. 1975b. Stone Age myth and magic. Graz: Akademische Druk.

Pager, H. 1994 [1984]. San trance performances documented in the ethnological reports and rock paintings of southern Africa. Rock Art Research 11(2):88-100.

Prins, F. and Woodhouse, H. 1996. Subsaharan Africa: the state of the art: rock art in southern and tropical Africa- the last five years. In: P. Bahn and Fossati, A.(eds). Rock art: news of the world 1. Oxford: Oxbow Books.

Robertshaw, P. 1990. A history of African archaeology. London: James Currey; Portsmouth (N.H.): Heinemann. Skotnes, P. 1991. Is there life after trance? De Arte 44:16-24

Skotnes, P. 1994. The visual as a site of meaning: San parietal painting and the experience of modern art. In: T.A. Dowson and D. Lewis-Williams (eds). Contested images: diversity in southern African rock art research, pgs. 315-330. Johannesburg: Witwatersrand University Press

Solomon, A. 1997. The myth of ritual origins? Ethnography, mythology and interpretation of San rock art. South African Archaeological Bulletin 52:3-13.

Solomon, A. 1998. Myth, shamanism and San rock art. Paper delivered at the Valcamonica Symposium, Sulzano, Italy.

Stow, G. W. 1905. The native races of South Africa. Lon-don: Swan Sonnenschein

Wilman, M. 1968 [1933]. The rock engravings of Griqualand West and Bechuanaland, South Africa. Cape Town: A.A. Balkema.

Woodhouse, H. 1984. When animals were people: A-Z of animals of southern Africa as the Bushmen saw them. Johannesburg: Chris Van Rensburg Publications.

Vinnicombe, P. 1976. People of the eland: rock paintings of the Drakensberg Bushmen as a reflection of their life and thought. Pietermaritzburg: University of Natal Press

Yates, R., Manhire, A. and Parkington, J. 1994. Rock painting and history in the south western Cape. In: J.D. Lewis-Williams and Dowson, T. (eds.), Contested images: diversity in southern African rock art research, pgs.29-60. Johannesburg: Witwatersrand University Press.

back to top of page