political alliances with neighbouring farming communities. Silayi, a Thembu man who went to live with the Bushmen in the 1850s, told W.E. Stanford about the raids he participated in; he makes it plain that they were not simply raiding for food. Although Silayi's account is of a time towards the end of the period under discussion, his cannot have been the only such experience; other black people must have done similar things and earlier. Silayi does not say what happened to all the livestock taken, but on one occasion two horses were given to their Bushman chief.48 Black chiefs like Mandela, Mchithwa and Moorosi offered the Bushmen protection in return for livestock and ivory.49 Providing ivory and livestock, then, gave the Bushmen enormous political advantages during a time when their survival was being threatened more and more.
Impact of the 'Mfecane' on the Art
In this brief account of a fairly complex matter I refer to only two themes in the art, themes used by shaman-artists to establish and reinforce their power at this particular time: depictions of elephants, and conflict scenes.
In the south-eastern mountains paintings of elephants and scenes of elephant hunts are not infrequent. Given the shamans' control over other spheres of life, such as hunting,50 it is likely that they would also have tried to appropriate the ivory trade to bolster their own status. This is not to say that all elephant hunters were necessarily shamans. As hunters engaged shamans for success in antelope and ostrich hunting, they probably also appealed to them to ensure success in the acquisition of ivory. Depictions of elephant were thus making statements about the power and the position of the shaman: they were statements about and evidence for the shaman's control over resources. They made paintings of elephant hunts (Figures 5 and 6) and elephants (Figure 7) to reinforce their control over Bushman involvement in the ivory trade. Depictions of elephants, included in panels with clear trance associations, seem to have become symbols of the shamans' political power. The paintings were thus actively involved in the reproduction and entrenchment of the shamans' power.
This leads me to depictions of conflict. Although reportedly numerous, in fact conflict scenes are not found in large numbers. Early rock art scholars concentrated on themes such as these because they seemed to match their stereotype of the Bushmen as hostile raiders. The biggest concentration of paintings of conflict scenes occurs in the lower Caledon River Valley. As Cobbing has shown, this is an important region for discussing the struggles of the 'mfecane' period.51 Despite the
Figure 7. Rock painting of an elephant from the Giant's Castle area in the Natal Drakensberg. This depiction is part of a large and complex panel which is replete with trance imagery. This depiction is thus unequivocally linked to shaman ideology.
Figure 8. Portion of a large panel of paintings depicting Bushmen/farmer conflict, Wepener District, Free State (after Stow and Bleek, Rock Paintings, plates 61, 62).
Figure 9. Portion of a large panel depicting farmer/farmer conflict, Rouxville District, Free State (after Stow and Bleek, Rock Paintings, plate 37).
problem of exact dates for each of the paintings, the concentration of fight scenes in this area can hardly be fortuitous.
The Caledon Valley paintings depict Bushman/Bushman, Bushman/farmer (Figure 8) as well as some farmer/farmer conflicts (Figure 9). An examination of these scenes in terms of Bushman beliefs shows they are not simple records of actual events. Many have elements that unequivocally relate to trance experience and thus place the paintings in the realm of shaman ideology.52
Traditionally, Bushman shamans in trance fought off marauding spirits and malevolent but nameless shamans. With the development of relations with the Bantu-speaking farmers and the colonists, this shamanistic activity was extended to include these new sources of conflict.53 The shamans now used their powers to engage the intruders on their land. This is not, of course, a fully adequate explanation of the pictures. We need to enquire about the social relations issuing in the fights and also about the place of the shaman-artists themselves in these relations.
In the upheavals of the late eighteenth and early nineteenth centuries Bushman groups and individual shamans were drawn into conflicts between Bantu-speaking chieftainships. It may be that even as shamans had long made rain for specific leaders of farming communities, so those leaders called upon the Bushman shamans in times of strife with other farming groups. For instance, Moorosi, chief of the Phuti, was supposed to have been related to Bushmen.54 He acted as a protector to some Bushman groups, but D.F. Ellenberger reports that during the 'mfecane' Moorosi's people were subservient to the Bushmen.55 So too, I argue, were shamans' supernatural powers harnessed by participants in the 'mfecane' struggles. Shaman-artists then used their art to negotiate this new opportunity for developing their own political power. As they had painted elephant hunts so they came to paint conflict between Bantu-speaking groups and thereby enhanced their status. The power struggles of the 'mfecane' period were thus multidimensional.
Conclusion
Ironically, Moffat was right when he said, 'Hard is the Bushman's lot', a sentiment that has been picked up again and again. But Moffat could not have been aware of how his and other early commentators' attitudes were to underwrite the Bushmen's historical and political marginalisation; history said little for William Macmillan's 'unlucky Bushman' in 1927 and does not add much more today. In 1972 Shula Marks observed, 'History tends to be the history of the successful, and the Khoikhoi herders and San hunter-gatherers . . . have all but disappeared from twentieth-century South Africa, at least in their earlier guise. Most historians writing of South Africa dismiss them in passing.'56 Nothing has changed much. The role of Bushmen has merely changed from passing reference to a very generalised introductory chapter.57 Historical constructions at a more general level have also ignored the Bushmen. For the eastern Cape, Gerrit Harinck accords greater significance to Khoi–Xhosa interaction; the Bushmen were simply 'eradicated and dispersed'.58 Using both excavated material remains and rock art, Simon Hall has shown that 'this is a gross over-simplification'.59 By taking seriously both parts of the twofold archaeological record, excavated material remains and rock art, we shall be able to start producing a history of southern Africa that includes all the people who were involved, in no matter what capacity. More specifically, and for the purposes of this volume, Bushman art, until recently a neglected part of the archaeological record, can play a major role in developing a new understanding of the 'mfecane' period, thus drawing the hunter-gatherers back into history. The 'invisible' Bushmen were active participants who should not be ignored. There is much that the Bushmen in their roles as hunter-gatherers, traders and slaves in a time of colonial oppression can contribute to the history of southern Africa through their art.
1.J.B. Wright, 'Political Mythology and the Making of Natal's Mfecane', Canadian Journal of African Studies, 23, 2 (1989), 273.
2.Many writers prefer 'San' to 'Bushman' because of the latter's pejorative associations for some, but by no means all, people in southern Africa. Unfortunately, 'San', a Nama word, also has negative connotations: it could be translated 'vagabond', and its use by historians, archaeologists and anthropologists ascribes to the Nama antagonistic attitudes towards Bushmen. Because there are so many Bushman languages, there is no generic