Nqabayo's Nomansland: San rock art and the somatic past

STUDIES IN GLOBAL ARCHAEOLOGY 2 GEOFFREY BLUNDELL NQABAYO’S NOMANSLAND San Rock Art and the Somatic Past African and Comparative Archaeology Department of Archaeology and Ancient History, Uppsala University, Uppsala 2004 Rock Art Research Institute School of Geography, Archaeology and Environmental Science, University of the Witwatersrand, Johannesburg 2004 For DL-W and in Memory of PV (1932-2003) ABSTRACT Blundell, G. 2004. Nqabayo’s Nomansland: San Rock Art and the Somatic Past. Studies in Global Archaeology 2. Uppsala. 204 pp., 75 figures, 3 appendices. ISSN 1651-1255, ISBN 91-973212-0-6. The most significant challenge facing modern southern African rock art research is the integration of rock paintings into the construction of San history. This challenge is made all the more difficult because of poor chronological control over the images. In the absence of reliable dating techniques, the challenge to interdigitate image and history becomes a profoundly theoretical one. Drawing on theoretical studies of body and embodiment, this work takes up the challenge of incorporating rock paintings into the production of the past. Primarily concerned with a small area, previously known as Nomansland, in the south-eastern mountains of South Africa, the work uses embodiment as a tool for extending present interpretations of the art before moving on to arguing that a focus on body allows us to detect change in certain images in Nomansland. Finally, embodiment is used to re-evaluate present understandings of the social consumption of the paintings. In using embodiment to investigate issues of meaning, change and the production and consumption of San rock art, it becomes clear that this theoretical concept offers a way of incorporating rock paintings into the writing of San history in Nomansland. This work, then, contributes to the broader field of southern African San historiography, where the question of San interaction with other peoples is sometimes treated too simply and in a manner that is not consistent with broader postcolonial writing. Keywords: San, Bushman, southern Africa, Nomansland, Eastern Cape, Storm Shelter, rock art, shamanism, Kalahari revisionist debate, body, embodiment, somatic past, postcolonial, colonial Geoffrey Blundell, Rock Art Research Institute, University of the Witwatersrand, P. Bag 3, PO Wits 2050, Johannesburg, South Africa and African and Comparative Archaeology, Department of Archaeology and Ancient History, Uppsala University, Box 626, SE-751 26, Uppsala. E-mail: [email protected] ISSN 1651-1255 ISBN 91-973212-0-6 © Geoffrey Blundell Studies in Global Archaeology 2 Series Editors: Paul J. J. Sinclair Editor: Christina Bendegard Cover Design: Geoffrey Blundell Published and distributed by the African and Comparative Archaeology, Department of Archaeology and Ancient History, Uppsala University, Box 626, SE-751 26, Uppsala, Sweden Printed in Sweden by Elanders Gotab, Stockholm, 2004 A NOTE ON CONVENTIONS USED IN THE TEXT It is standard practice in South Africa to omit the names of farms on which rock art is located in order to protect the unmanaged sites from damage by human agency. Consequently, I have followed the practice of using site numbers to refer to rock art locales in the text. The site numbers that I use are those designated by the Rock Art Research Institute (RARI) at the University of the Witwatersrand in Johannesburg. The typical format of the the site number is RARIRSA-MEL9. The second segment refers to the country (in the case of this text RSA stands for South Africa and LES refers to Lesotho) and the third segment is the site number. At times, the site numbers can be confusing to the reader and so, where it is safe to do so, I have used more commonly known site names or well-known but invented designations. The vast majority of images used to illustrate this text as well as additional visual material may be accessed by locating the URL—Http://ringingrocks.wits.ac.za (note: no “www”)—and typing in the site number into the search function (use Roman numerals). Unless otherwise specified, all images are the product and property of RARI. CONTENTS CONTENTS................................................................................................................................................ 8 LIST OF FIGURES ................................................................................................................................... 11 ACKNOWLEDGEMENTS ....................................................................................................................... 13 INTRODUCTION: ‘THE BUSHMEN AREN’T FOREVER’ .................................................................. 16 Three Centuries of Misrepresentation .....................................................................................................21 The Revisionist Challenge and Beyond ...................................................................................................23 The Nomansland Project .........................................................................................................................28 CHAPTER 1: NQABAYO’S NOMANSLAND: PEOPLE, PLACES AND PICTURES............................ 34 The San of Nomansland ..........................................................................................................................34 Henry Francis Fynn’s Investigations ................................................................................................................. 35 Walter Stanford’s Interviews ............................................................................................................................. 36 The Discovery of Manqindi Dyantyi ................................................................................................................ 40 A Nomansland Rock Art Site ...................................................................................................................45 Nomansland’s Past in Paint and People ..................................................................................................49 CHAPTER 2: SAN ROCK ART AND HISTORY ..................................................................................... 53 The Hermeneutic Approach ....................................................................................................................53 The Use of Shamanism .................................................................................................................................... 56 The Applicability of Neurological Research ..................................................................................................... 57 Three Limitations of the Hermeneutic Approach ............................................................................................ 61 The Social Approach ................................................................................................................................62 Marxist Approaches ......................................................................................................................................... 63 Structurationist Approaches ............................................................................................................................. 66 Interactionist Approaches ................................................................................................................................ 69 Towards a New Theoretical Approach .....................................................................................................74 CHAPTER 3: THE SOMATIC PAST ....................................................................................................... 76 From the Body to Embodiment ...............................................................................................................76 Archaeology and the Body .......................................................................................................................79 Embodiment and San Rock Art ...............................................................................................................80 A Non-Structural Social Approach ................................................................................................................... 81 Embodiment and Identity ................................................................................................................................ 86 The Somatic Past ......................................................................................................................................87 CHAPTER 4: THE JAWS THAT BITE, THE CLAWS THAT CATCH: THE ELDRITCH IMAGES .... 89 Disease, Death and Disorder in San Cosmology .....................................................................................90 The Eldritch Images ................................................................................................................................97 RARI-RSA-GEN1 ......................................................................................................................................... 101 Junction Shelter .............................................................................................................................................. 102 RARI-RSA-MEL6 .......................................................................................................................................... 107 Storm Shelter ................................................................................................................................................. 109 Body and Interpretation in San Rock Art .............................................................................................. 111 CHAPTER 5: OTHER BODIES: THE TYPE 2 ROCK ART IMAGES .................................................. 113 Type 2 Rock Art ...................................................................................................................................... 113 A Different San Tradition? ............................................................................................................................. 114 A Bantu-Speaker or Khoekhoen tradition? ..................................................................................................... 117 The Nomansland Neighbourhood .........................................................................................................119 The Bhaca ...................................................................................................................................................... 121 The Mpondo .................................................................................................................................................. 123 The Mpondomise ........................................................................................................................................... 124 The Thembu .................................................................................................................................................. 124 The Phuthi ..................................................................................................................................................... 125 The Khoekhoen ............................................................................................................................................. 127 Nomansland and its Art in a Local System ............................................................................................128 CHAPTER 6: THE DEATH OF THE POST-CRANIAL BODY ........................................................... 131 Significantly Differentiated Figures ....................................................................................................... 132 Nomansland in a Global Colonial System ............................................................................................. 146 Explorers ....................................................................................................................................................... 146 Shipwrecks ..................................................................................................................................................... 147 Escaped Slaves ............................................................................................................................................... 149 The Bushman Wars ........................................................................................................................................ 149 Mission Stations ............................................................................................................................................. 150 The Death of the Post-Cranial Body ..................................................................................................... 153 CHAPTER 7: THE CONSUMPTION OF SAN ROCK ART AS AN EMBODIED EXPERIENCE ..... 158 The Pattern of an SDF Site .................................................................................................................... 158 Understanding Patterns in San rock art ................................................................................................. 162 Rethinking the Consumption of San Rock Art ...................................................................................... 166 The Social Implications of Patterned Sites............................................................................................ 170 CONCLUSION: TOWARDS A SOMATIC PAST OF NOMANSLAND ............................................... 175 ENDNOTES ........................................................................................................................................... 181 REFERENCES ........................................................................................................................................ 182 APPENDIX 1.......................................................................................................................................... 197 APPENDIX 2.......................................................................................................................................... 198 APPENDIX 3.......................................................................................................................................... 203 LIST OF FIGURES Fig. 1. Map of southern Africa ...................................................................................................................................................................................17 Fig. 2. Flowering Hoodia gordonii ..................................................................................................................................................................................18 Fig. 3. The hoarding on De Beers’ planned London retail outlet before intervention. .....................................................................................20 Fig. 4. The hoarding as altered by Survival International.. .....................................................................................................................................20 Fig. 5. Advert taken out by Survival International in a Flemish newspaper. .......................................................................................................22 Fig. 6. Map of the Nomansland Core Study Area. ..................................................................................................................................................30 Fig. 7. A view of the Nomansland study area in summer ......................................................................................................................................31 Fig. 8. Winter in Nomansland. ....................................................................................................................................................................................31 Fig. 9. Henry Francis Fynn ..........................................................................................................................................................................................35 Fig. 10. Sir Walter Stanford. .........................................................................................................................................................................................37 Fig. 11. The group of San that met with Stanford in the 1880s ............................................................................................................................38 Fig. 12. San man and woman. .....................................................................................................................................................................................39 Fig. 13. Reverse of photograph in Fig. 12. ................................................................................................................................................................39 Fig. 14. Manqindi Dyantyi at Ncengane Shelter. ......................................................................................................................................................41 Fig. 15. Manqindi Dyantyi walking in the Bushman’s Cuttings area. ....................................................................................................................42 Fig. 16. Map of proximity of rock art sites to places of historical significance. ................................................................................................44 Fig. 17. View of Storm Shelter Panel.........................................................................................................................................................................46 Fig. 18. The density of painted images at Storm Shelter. .......................................................................................................................................46 Fig. 19. Translucent white figure from Storm Shelter. ............................................................................................................................................47 Fig. 20. The primary distinctive anthropomorphic figure at Storm Shelter. .......................................................................................................48 Fig. 21. The secondary distinctive anthropomorphic figure at Storm Shelter.. ..................................................................................................48 Fig. 22. Strange white anthropomorphic figure from RARI-RSA-BAE2 ............................................................................................................48 Fig. 23. Image of antelope that is not in the classic San tradition. .......................................................................................................................48 Fig. 24. Black-and-white drawing of the primary panel at Storm Shelter. ..........................................................................................................51 Fig. 25. Hartebeest therianthrope. ..............................................................................................................................................................................72 Fig. 26. Therianthrope from Storm Shelter. .............................................................................................................................................................72 Fig. 27. Human figures running on all fours.............................................................................................................................................................73 Fig. 28. Translucent-white figure with red details ....................................................................................................................................................98 Fig. 29. Translucent-white figure with emaciated hips and exaggerated penis ....................................................................................................98 Fig. 30. Translucent-white figure underneath ‘classic’ San rock art ......................................................................................................................99 Fig. 31. A typical cluster of translucent-white figures from Nomansland. ........................................................................................................100 Fig. 32. A cluster of Eldritch Images from the Uniondale area. .........................................................................................................................101 Fig. 33. Composite image of Eldritch Images walking on a thread that connects the real and spirit worlds ..............................................102 Fig. 34. Rock art panel from RARI-RSA-GEN1. ..................................................................................................................................................103 Fig. 35. Harald Pager’s copy of the Fat-Tailed Sheep Panel at Junction Shelter. ..............................................................................................105 Fig. 36. Detail of Cluster 5, Section 2. .....................................................................................................................................................................106 Fig. 37. The Eldritch Images of RARI-RSA-MEL6. ............................................................................................................................................108 Fig. 38. Detail of Eldritch Image at RARI-RSA-MEL6. ......................................................................................................................................109 Fig. 39. Eldritch Image from RARI-RSA-MEL10.................................................................................................................................................109 Fig. 40. The cluster of Eldritch Images from Storm Shelter. ..............................................................................................................................110 Fig. 41. Detail of Eldritch Image from Storm Shelter. .........................................................................................................................................111 Fig. 42. Possible bisexual Eldritch Image. ..............................................................................................................................................................111 Fig. 43. Typical monochrome Type 2 image. ..........................................................................................................................................................114 Fig. 44. Bichrome Type 2 image................................................................................................................................................................................114 Fig. 45. Poster or Block Style image. ........................................................................................................................................................................115 Fig. 46. Type 2 image of horse and rider.................................................................................................................................................................115 Fig. 47. Indeterminate antelope of the Late White tradition. ..............................................................................................................................116 Fig. 48. Typical cluster of Type 2 images. ...............................................................................................................................................................117 Fig. 49. Khoekhoen geometric rock art image. .....................................................................................................................................................118 Fig. 50. Type 2 antelope. ............................................................................................................................................................................................118 Fig. 51. Stanford’s Map (modified) of the south-eastern seaboard. ....................................................................................................................120 Fig. 52. Map showing the distribution of SDF and Type 2 sites in the Nomansland study region. ..............................................................134 Fig. 53. RARI-RSA-PEL4. .........................................................................................................................................................................................135 Fig. 54. The SDF from RARI-LES-MTM1 ............................................................................................................................................................135 Fig. 55. RARI-RSA-BUR1. .......................................................................................................................................................................................135 Fig. 56. RARI-RSA-MEL6 ........................................................................................................................................................................................137 Fig. 57. SDF from Ncengane Shelter .......................................................................................................................................................................136 Fig. 58. RARI-RSA-WID2 .........................................................................................................................................................................................138 Fig. 59. SDF at RARI-RSA-RON1 ..........................................................................................................................................................................139 Fig. 60. SDF at RARI-RSA-TYN2. ..........................................................................................................................................................................139 Fig. 61. Nomansland LH-SDF ..................................................................................................................................................................................140 Fig. 62. Secondary SDF from Storm Shelter ..........................................................................................................................................................140 Fig. 63. The primary LH-SDF at Storm Shelter ....................................................................................................................................................141 Fig. 64. Dowson’s representation of the SDF from RARI-RSA-MEL8 ............................................................................................................142 Fig. 65. Reinterpretation of RARI-RSA-MEL8 as a LH-SDF. ............................................................................................................................142 Fig. 66. SDF with unique headdress. ........................................................................................................................................................................143 Fig. 67. LH-SDF with facial detail ............................................................................................................................................................................143 Fig. 68. Disembodied heads with other San images ..............................................................................................................................................144 Fig. 69. RARI-RSA-WAR5.. ......................................................................................................................................................................................145 Fig. 70. Detail of one of the disembodied heads at RARI-RSA-WAR5............................................................................................................145 Fig. 71. Humorous cartoon. ......................................................................................................................................................................................160 Fig. 72. Manqindi Dyantyi touching the rock art ...................................................................................................................................................161 Fig. 73. Type 2 image before the removal of a pigment sample ........................................................................................................................162 Fig. 74. Fig. 73 after the sample has been removed. ..............................................................................................................................................163 Fig. 75. Row of 21 finger-painted dots from RARI-RSA-STK1.........................................................................................................................164 ACKNOWLEDGEMENTS Of the many people who assisted in the production of this work, Paul Sinclair deserves most of my gratitude. He took me on as a student at a point when the thesis was already developed and gave me my own intellectual space while still managing to steer my thought around various obstacles. Together with his wife, Amelie Berger, Paul graciously hosted me on several visits to Sweden. In addition, the staff and students of African and Comparative Archaeology at Uppsala University, notably Christina Bendegard, Elisabet Green, Markku Pyykönen and Anneli Ekblom were always ready to assist and offer advice. Neil Price, who introduced me to Paul, guided me through the complexities and marvel- holiday home on the edge of Nomansland has been a welcome shelter from the bitter winter cold of the area on several occasions. I was also fortunate enough to meet Lyn Meskell while writing this work. Not only did she nudge me in the right theoretical direction but she kindly allowed me to read some of her unpublished material and made useful suggestions concerning the text. Like so many others, I owe an enormous intellectual debt to David Lewis-Williams whose lectures as an undergraduate first inspired me to take up the pursuit of rock art. His mentorship has been invaluable both within and beyond academia. Pat Vinnicombe spent time at the University of the Witwatersrand in 2000 where she not only encouraged me but also generously allowed me access to the archival material that she so painstakingly collected for People of the Eland. Between Pat and David there is very little that has not been said about Drakensberg San rock art and it has been a great privilege for me to inhabit a space contiguous to both of their offices at the Rock Art Research Institute. When Pat unexpectedly passed away in March 2003, it came as a shock to many researchers around the world. Without the work that she and David undertook in the Drakensberg in the 1970s, South African rock art research would not enjoy its current celebrated status. A substantial part of this work was undertaken while on a Coleman Fellowship in Art History at the Metropolitan Museum of Art in New York during 2000 and 2001. I am particularly grateful to Alisa Lagamma and Julie Jones of the Department of the Arts of Africa, Oceania and the Americas for getting me to New York. I am also grateful to Marion Burleigh-Motley, Marcie Karp and Allison Hawkins of the Education Department for looking after me and the other fellows so well. Of the many fruitful discussions I had with my fellows, I benefited most from those with Thomas Dale and Henry Drewal. The MetMuseum’s resources are unparalleled, but their most valuable assets are the librarians in the Goldwater Library; Leslie Preston, Joy Garnett and Ross Day were simply phenomenal in obtaining—often obscure—publications. Kendall McWilliam and Amy Chen saw to it that I had all I needed to function and assisted in ways too numerable to mention. Teresa Lai and Lasley Poe were simply marvellous and they offer a culinary tour of New York that is unsurpassed. While staying at International House, I had occasion to have many Ben Smith, the director of the Rock Art Research Institute, braved the considerable bureaucracy at the University of the Witwatersrand to allow me long leave so that I could undertake the writing of the thesis. Always eager to expand my horizons beyond the Drakensberg lous subtleties of Swedish society during my various and beyond San rock art, I have enjoyed many fieldtrips, forays to Uppsala. He and his wife, Linda, also kindly al- discussions and arguments with him and have come to lowed me to stay at their home on several occasions and appreciate his infectious enthusiasm for all things relatmade my time in Sweden a very pleasant experience. ing to rock art. I have had the pleasure of Carolyn Hamilton’s astute guidance on this and all my postgraduate work. Her 13 stimulating and intellectually enriching discussions with Alica Sekolec, Gabi Bockaj, Lauren van Vuuren, Jill and Joel Claassen, Tom Rossetti, Kai Liiva, Dorothy Ndletyana, Seema Srinath and Julia Goesser amongst others. I am also grateful to Jim and Liz Watson, Graeme Hardie, Silas Mountsier and Claire Dean for hosting me at various events while in the USA. Brian Wallace kindly brought books to me in NYC from South Africa and saw to it that I ventured beyond the boundaries of Manhattan. Neil Rimmer deserves a special mention for working his technical wizardry across the Atlantic and maintaining and fixing my computer. Over the course of doing the research and writing of this work, I have come to meet a great number of people in rock art research. Many of them have influenced my thinking through discussion, argument or through facilitating my visitation of rock art sites outside of my own research area. These include Jean Clottes and the Chauvet Cave team, Chris Chippindale, David Whitley, Larry Loendorf, Janette Deacon, David Morris, Sven Ouzman, Ed Eastwood, Cathelijne Cnoops, Knut Helskog and Ericka Englestad. Other academics that offered useful advice, pointed me in the direction of relevant literature and answered nagging questions include: Tony Traill, Himla Soodyall, Trefor Jenkins, Pieter Jolly, Frans Prins, Phil Bonner, David Hammond-Tooke (who also kindly allowed me to cite from his unpublished field notes and who sadly also passed away during the course of this work), Lyn Wadley, Catherine Mathers and Isak Niehaus. Brad Keeney and Ringing Rocks generously arranged for me to travel with RARI members to Tsumkwe in Namibia in 2002 where we were allowed to observe dances and meet with members of the local !Kung community. The rock art fieldwork involved long days of survey and recording in arduous terrain and inclement weather. Without the assistance of the following people in the field, this thesis would not have been possible. In particular Glen and Sue Tonkin, Gordon McKenzie, Ghilraen Laue, Paul Den Hoed, Alan Marshall, Demi Bender, David Pearce, Siyakha Mguni, Ken Garrett, Sam 14 Challis, Jamie Hampson, Phil Ferrer (who also kindly explained triboluminescence to me), Silvia Hillebrand, Alain King, Neil Lee, Jeremy Hollmann, Olivia Campbell, Sally Coleman, Maria Cruz Berrocal, Carol Wallace, Azizo Fonsecka, Sandile Majola, Robyn Pickering, Casey Clements-Bass, Rory Maclean, Nathan Honey, Sonette du Plessis, Ruben Mowszowski, Bjørn Helberg, Winnie Mokokwe, Leslie Zubieta, Justine Olofsson, Lara Mallen, Tamaryn Petersen, Leanne Douglas, Justin du Piesanie, Kerry Fatherley, Leila Henry and Robert Kruger assisted me greatly. Without the permission and kind assistance of the landowners of the present area that was Nomansland, the fieldwork would not have been possible at all. I am particularly indebted to Phyl Sephton, Sheila BellCross and Jim Feely for their hospitality, generosity and persistent enthusiasm. Bun and Val Mae Roberts, Pete Moore, Ronnie Moore, Adele Moore, Paul Sephton all kindly allowed me access to their farms to search for and record rock paintings. At Mondi NECF, Dave Butt, Andre Marais, Hendrik Bouwer, Helen Lechmere-Oertel and Luke von Vogt were particularly helpful. During the final stages of production, a number of people contributed in various ways, ranging from checking references and finding literature to preparing certain images. These include Sióbahn McCusker, Lara Mallen, Tamaryn Peterson, Justin du Pisanie and Kerry Fatherley. In particular, Willem Steyn, Liza Xenophontos and Azizo Fonsecka of the Ringing Rocks Digital Laboratory scanned numerous slides and re-drawings for me. Bonnie Williamson undertook the testing for haemoglobin/myoglobin in the samples that were removed from Nomansland. This project was funded by a number of sources: The Department of Archaeology and Ancient History, in the Faculty of Arts, at Uppsala University kindly awarded me a Research Development Grant for the year 20012002. African and Comparative Archaeology at Uppsala kindly financed the production of this book. The fieldwork was sponsored by the National Research Founda- tion of South Africa as part of their grant to the Rock Art Research Institute. Finally, my greatest thanks go to my family, John, Lyn and Shiona Blundell, Daphne Smit and to Ken Price, Laura Wallace and especially to Carol Wallace, for all their support in so many innumerable ways. 15 INTRODUCTION ‘THE BUSHMEN AREN’T FOREVER’ We’re doing what we can to pay back, but it’s a really fraught problem... especially as the people who discovered the plant have disappeared. I honestly believed that these bushmen had died out and am sorry to hear they feel hard done by. I am delighted that they are still around and have a recognisable community. The Observer, Sunday June 17, 2001 There are some 110, 000 extant San (Bushmen) scattered across South Africa, Botswana, Namibia and Angola today1 (Fig. 1). During the nineteenth and twentieth centuries, researchers studied intensively the click-languages, hunter-gatherer economy, mythology, folklore, archaeology, history and rock art of these autochthonous southern African people. So prolific is the past research on the San that they are sometimes described as the most studied people in the world. In the light of this extensive research, the statements made by Richard Dixey, the CEO of Phytopharm, that open this chapter seem even more remarkable. Phytopharm is a small pharmaceutical company, located in Cambridgeshire, England, that recently patented P57, an appetite-suppressing substance found in the Hoodia gordonii, a species of cactus endemic to parts of southern Africa (Fig. 2) inhabited by San. Because P57 appears to have very few side effects, Phytopharm believes that it will revolutionise the dieting industry in the Western world. On the strength of the research carried out so far, Phytopharm has licensed the drug to the US pharmaceutical giant, Pfizer, for $21 million. When Pfizer eventually releases P57 in pill form, the company stands to earn substantial profits in the $9 billion per annum slimming market (Barnett 2001). San’s indignation at their treatment attracted media attention and journalists soon began investigating the story. When reporters confronted Phytopharm, Dixey was genuinely surprised and explained that it was the South African Council for Scientific and Industrial Research (CSIR) that initially approached Phytopharm about the Hoodia cactus. The CSIR appear to have cut a deal with Phytopharm as far back as 1998 (Finch 2001). Dixey claimed that the CSIR had told him that the San tribes who used the cactus no longer existed and that they assured him that agreements were in place to help local communities. When reporters confronted the CSIR, Dr Marthinus Horak, the man in charge of the project, defended the deal. He claimed there were only a few hundred San left in South Africa itself and as they lived in isolated areas, they were very difficult to contact. Horak also claimed that the CSIR intended to contact the San at some point when clinical trials had been approved; to do so prior to this time, he contended, would raise the community’s expectations with promises that it may not have been possible to meet (Barnett 2001). In the outcry following the initial media reports, various San communities, through their legal representatives, managed to strike a deal whereby they will be In the negotiations between the various pharmaceuti- paid a percentage of the CSIR royalties (Barnett 2002, cal companies over the licensing of intellectual prop- Finch 2001). The success of the San communities is erty rights and patents, the San, whose knowledge and encouraging but there are nagging issues that still suruse of the plant initially alerted researchers to its prop- round this affair. It is remarkable, for example, that a erties, played no role. It is of little surprise, then, to community apparently so isolated and so difficult to learn that news of the negotiations caused an outcry contact can manage to contact the press, the CSIR and amongst the San and their legal representatives. The international pharmaceutical companies but that those 16 Fig. 1. Map of southern Africa showing Nomansland, the CKGR, rock art locales mentioned in this work and the distribution of known San languages. Place names in italics are older, colonial designations that are now no longer used (except in the case of “Transkei”2). organizations, given their substantial resources, cannot do the reverse. It is equally remarkable that those organizations can manage to collect samples of the Hoodia plant from the apparently isolated areas where the San live but find it too difficult to contact the San themselves. These ironies aside, issues that are more serious arise from this debacle. For example, how is it possible that a people so numerous and so intensively studied and publicised be thought to be extinct? in the light of ongoing social and political marginalization of San communities. The recent removal of G/wi and //Gana people living in the Central Kalahari Game Reserve (CKGR) in Botswana (see Fig. 1) and the subsequent involvement of Survival International is a powerful example. Between 1997 and 2002, the Government of Botswana removed some 2200 members of these two San communities from this reserve as part of their Remote Area Dweller (RAD) resettlement scheme. While some of these communiAnswers to such questions become more urgent ties relocated freely, others were removed against their 17 Fig. 2. Flowering Hoodia gordonii, Namibia. Image courtesy of Lytton John Musselman. will. A handful of the community chose to remain be- Beers claim that the facts of the situation are as follows: hind and they have persevered even after the governThere are no “massive diamond deposits” in the ment cut off their water supply in order to force them CKGR. Exploratory drilling has taken place only from the land (see Suzman 2002 for a more detailed at Gope and mining has been shown to be not discussion on the background to the CKGR issue). commercially viable. As a result the prospecting plant has been taken away, the shaft headgear is being dismantled, the test shaft capped and the Survival International, an organisation dedicated to the site made safe and secure. Mr Corry was informed protection of human rights for small-scale societies of this in March 2002. When drilling started there such as the San, has taken up the cause of the G/wi and were no Bushmen in the vicinity of the Gope prospect. Eventually a few families were attracted //Gana. In a highly visible campaign, the organizato the area by the water produced by the shaft tion (currently led by Steven Corry) has alleged that dewatering process. This water was supplied to economic motives lie behind the removal of the San. them for livestock use. Moreover, De Beers supThey claim that the Botswana government, in conjuncplied a pump and tank for a borehole to the west tion with mining companies, particularly the South of the Gope prospect, which was handed to the local community. Although exploration has endAfrican diamond-mining giant, De Beers, wishes to ed, those Bushmen who remain in the area conremove the San to facilitate mining of the gemstone tinue to have access to this water. Mr Corry was in the reserve. De Beers has issued a number of meinformed of this in March and again in July 2002. dia statements refuting Survival International’s accusations. In one of these statements (12 July, 2002), De De Beers has never sought the removal of the 18 Bushmen from the CKGR. Diamond mining does not require the removal or resettlement of any community, in Botswana or elsewhere. Indeed, we welcome the presence of local populations to whom we can offer employment. Where mining operations may conflict with present land usage— for example a farm or cattle post—the individuals involved receive adequate and appropriate compensation following extensive consultation and negotiation. This cannot be construed as the forced re-settlement of local communities. Mr Corry has been informed of our policy in this regard. We have received assurances from the Botswana Government that its plans for the resettlement of the Bushmen have nothing to do with diamond mining, but have been motivated by a desire to provide for their health care, education and other development needs. The programme is not restricted to the Bushmen of the Kalahari but has been applied throughout Botswana. These assurances have been given to Mr. Corry by the Botswana High Commissioner in London. Following the recommendations of this mission, Government took a decision through the Ministry of Commerce and Industry Circular No. 1 of 1986, that: the boundaries and status of the CKGR should be maintained as at present; the social and economic development of Xade (old) and other settlements in the Reserve should be frozen because they have no prospect of becoming economically viable; viable sites for economic and social development should be identified outside the Reserve and the residents of the Reserve be encouraged, but not forced to relocate to those sites; and that the Ministry of Local Government and Lands should advise Government on the incentives required to encourage residents in the Reserve to relocate. Such statements by De Beers and the Government of Botswana have not deterred Survival International and their campaign has become increasingly strident. In October of 2002, the organization altered the hoarding The Government of Botswana has also denied the covering De Beers’ planned retail outlet on the corner of claims of Survival International. In a statement (No Piccadilly and Old Bond Street in London; the outlet was Date) on their official web site, they offer the followscheduled to open on 21 November in the same year (see ing reasons for the removal of San from the CKGR: Survival International 30th October, 2002). The original billboard featured the internationally-famous modThe CKGR was established as a game reserve under the Game Proclamation through el, Iman, but Survival International replaced her image the High Commissioner’s Notice on the 14th with one of a San woman and they also substituted the February, 1961. The Game Proclamation was company’s famous logo: “A diamond is forever” with the superseded by the current Wildlife Conservaslogan “The Bushmen aren’t forever” (Fig. 3 and Fig. 4). tion and National Parks Act No. 28 of 1992, Far from being merely witty, the pun coined by Survival which maintained the area as a game reserve. International encapsulates many of the issues facing Over time, it became clear that many residents contemporary San and those who would research them. of the CKGR were becoming settled agriculturIt is, as we shall soon see, also a deeply problematic sloists, raising crops and rearing livestock. These gan. Together with the Hoodia debacle, the CKGR issue land uses, especially livestock husbandry, are not compatible with preserving wildlife resources. demonstrates the inability of various Western institutions to deal with the complexities of San identity in a In 1985, in recognition of increasing land postcolonial world. This failure to grasp adequately the use conflicts in the reserve, Government appointed a Fact Finding Mission to investigate intricacies of San identity is the result of many centuthe situation in the CKGR. The Mission found ries of Western representations of the San. That these that indeed some locations in the CKGR were representations are still very much alive in the West is rapidly evolving into settled agricultural comclearly apparent from the events that followed the almunities especially at Xade (old), Molapo, Metsiamanong, Gope, Mothomelo and Gugama. teration of the hoarding of De Beers’ London branch. 19 Fig. 3. The hoarding on De Beers’ planned retail outlet on the corner of Piccadily and Old Bond Street in London before intervention by Survival International. Fig. 4. The hoarding as altered by Survival International. Both Figures 3 & 4 are courtesy of Survival International. 20 Three Centuries of Misrepresentation In the face of ongoing provocation, De Beers has threatened legal action against Survival International (18 November, 2002). It is understandable that De Beers is eager to disassociate itself from the removal of the San from the CKGR. Widely regarded as a monopoly, the company has, over the last decade or so, tried to shed its negative image. Increasingly, the company is involved in initiatives to prevent the distribution of so-called blood diamonds from conflict areas in Africa (O’Ferrall 10th April, 2003) and it is, apparently, relinquishing its hold over the distribution of diamonds in an effort to be allowed to operate in the United States of America, where it is forbidden to do so under antitrust legislation (Keaton August 25, 2000). In the light of these efforts, the company takes the allegations made by Survival International seriously. Although ongoing and unresolved, the importance of the CKGR issue is in the perceptions that both Survival International and the Government of Botswana have of the San’s place in a world dominated by global capitalism and nationalistic concerns. While Survival International’s accusation that the Government of Botswana and De Beers are in league to appropriate the land of the G/wi and //Gana do not seem sustainable, both organizations clearly share a similar vision of the future of Botswana. This is clear from media statements issued by De Beers in response to Survival International’s actions (12 July, 2002): They also cloud the issue of what best constitutes a sustainable development strategy for the Bushmen—a matter which must reside with the Government and people of Botswana. Survival International may quite legitimately endeavour to influence government policy, but should not attempt to draw the diamond industry, which has been the life-blood of the economy of Botswana, into what is essentially a political and humanitarian issue. De Beers reiterates that it is committed to the continued development and welfare of all the people of Botswana. We must emphasise most strongly, however, that the re-settlement programme has nothing to do with the use of the country’s diamond resources. Since independence, these resources have been managed wisely by the Government to the benefit of its citizens in pursuit of economic and social development that has transformed Botswana into one of the few economic success stories in Africa. Diamonds, above all else, have been responsible for this sustained development and De Beers is proud of its contribution to Botswana’s success. In this statement, De Beers reiterates that diamonds are not the reason behind the resettlement scheme in the CKGR and the company reaffirms its commitment to economic development of Botswana. Importantly, the company believes that the sustainable development of the San “must reside with the Government and people of Botswana”. Indeed, De Beers does not offer any criticism of the Government of Botswana’s resettlement policies in this statement or, apparently, anywhere else. The Government of Botswana has made similar statements in which they put forward their reasons for the RAD resettlement programme: • • • • • • Undertake intensive development of remote settlements so as to bring them to levels comparable with the rest of other communities/ villages in the country; Promote production-oriented income and employment generating activities; Enhance the Remote Area Dwellers’ access to land and other natural resources; Encourage community leadership and active participation by Remote Area Dwellers in the election of their representatives in political and developmental organisation; Provide training and education to enable Remote Area Dwellers to lead self-sustaining livelihoods; and Promote cultural and economic advancement of Remote Area Dwellers by facilitating their integration into the mainstream society without any detriment to their unique culture and tradition. Clearly, both De Beers and the Government of Botswana see the future of the country as lying in the development of mining and industry. Not surprisingly, both organizations subscribe to a view of the world based on global corporate capitalism, national identity and eco21 indeed, these same reasons lie behind the perceptions of the San as isolated, few in number, extinct or verging on extinction that emerged during the Hoodia affair. These perceptions of the San are, in fact, relatively widespread. Almost since the beginning of European settlement in southern Africa, writers portrayed the San as a disappearing people. The settlers perceived the San as savages in the Hobbesian sense of the word: they were pathetic, wretched creatures of the earth, doomed to a life of hardship (Barnard 1999). Later on, these perceptions took on tones of social evolution. Many Westerners saw the movement of first the Khoekhoen pastoralists, then Bantu-speaking agropastoralists then European settlers into southern Africa as a process that evoked the American doctrine of manifest destiny. More powerful and advanced peoples inexorably replaced the supposedly primitive San. By the second half of the nineteenth century, it was near-universally accepted that the San were on their way to extinction. Writers such as George William Stow (1905) proclaimed the demise of the San in page after page Fig. 5. Advert taken out by Survival International in a Flemish of his The Native Races of South Africa in phrases such newspaper. The image plays to common perceptions of the San as as: “…thus perished the last ruler of the Tooverberg hunter-gatherers who are removed from the modern world. Bushmen” (p.177) and “…this ill fated race..”(p.183). nomic development. Within this vision of Botswana’s future, the place of the San, however, is a difficult one. The Government of Botswana obviously view the San as marginal and, apparently, wish to assist the San to enter ‘mainstream society’. In a quote, probably taken out of context, but nevertheless revealing (Goering 15th October, 2002: 2), Botswana’s President, Festus Mogae, asks the question: “How can you have a Stone Age creature continue to exist in the age of computers?” Some people have construed this statement as a desire on the part of the president to eradicate the San but the question was probably meant to refer to the vast difference in economic practices between the San and the rest of Botswana. Indeed, while some of the methods used by the Government of Botswana in their resettlement schemes are questionable, there is certainly no evidence of genocide of the San. There are, however, fundamental reasons that lie behind Mogae’s perceptions of the San as Stone Age relics; 22 Of course, there is an element of truth to these observations that the San were disappearing as by the end of the nineteenth century, most South African San languages were no longer spoken or only existed in fragments. Nevertheless, such views were only partly true for, in certain parts of southern Africa, Namibia and Botswana particularly, San people persisted. Yet, so entrenched was the idea that they were on their way to extinction that many believed that the San were pushed into the semi-arid regions of these countries by more powerful peoples. They only managed to survive there and it would not be long before they too would perish. Historians of southern Africa inadvertently took up many of these widespread perceptions of San demise during the twentieth century. Thomas Dowson (2000), for example, argues that in many historical works, the San function as a rhetorical device; they appear in the introductory sections of history texts but almost never feature in later, more modern history . Their contribution to southern African history is thus limited to a distant antiquity and writers neglect the San’s impact on the more recent past; for many historians, San history effectively ends with the arrival of other peoples in southern Africa. Not all historians have treated the San this way and some, such as John Wright (1971), have produced substantial histories of the San. Mostly, however, southern African historiography has placed the San on the periphery of historical analyses. While historians marginalized the San, anthropologists and archaeologists placed them at the centre of their studies. Beginning in the 1950s, the Marshall family began working with the !Kung San living in the Kalahari Desert of Namibia and Botswana (for descriptions of the Marshall Family work see Lewis-Williams 1999: ixxvi). Later on, anthropologists such as Richard Lee (e.g. 1979, 1984) and others built upon their work. Sociobiology and the renewed interest in evolution evident in Western academia during the 1950s and 1960s influenced much of the work that immediately followed on from that of the Marshall family; it was widely thought that the San were ideal models of past hunter-gatherer societies. Although much of this later ethnographic work on the San is more sophisticated and complex than critics sometimes allow, an unintended consequence of the major thrust of their research was to create a perception of the San as pristine, isolated and unchanging—a throwback to a time in the distant past. While historians wrote the San out of southern African history, anthropologists placed them in a static ethnographic present (for a discussion on the pitfalls of ethnography see Clifford 1988, Clifford and Marcus 1984). of the San. Ironically, Survival International, in its attacks on De Beers and the Government of Botswana, draws on similar stereotypes. For example, an advert taken out by the organization in a Flemish newspaper includes a photograph of the top half of a San hunter (Fig. 5). Rather than show the abject poverty of the living conditions of the San in the Central Kalahari, the organisation uses an image that evokes the hunter-gatherer past of the San. Their pun—“The Bushmen aren’t forever”— betrays an anxiety that the culture or way of life of the San is disappearing. This apprehension is founded on an idea of San society as timeless, unchanging and isolated from the global political economy. Indeed, the organization’s efforts are evocative of earlier attempts by people such as Donald Bain to create a reserve for the San wherein they could continue to exist, apparently, as they had done from time immemorial, as hunter-gatherers, living in harmony with nature (Gordon 1995, 1999). While Survival International’s efforts to assist the San are commendable, the perpetuation of these stereotypes contributes to a socio-political environment that continues to marginalise them (for a perceptive critique of Survival International’s role in the whole CKGR debacle and their subsequent defence, see Suzman 2002, 2003, Corry 2003). The Revisionist Challenge and Beyond Recognition of these problematic perceptions led some researchers to express a theoretical dissatisfaction with academic and popular discourse on the San. Known as revisionists, these writers first fully articulated their critique a little over a decade ago. Although there were earlier warnings of the dangers in the dominant discourse on the San (e.g. Marks 1972), it was the publication of Edwin Wilmsen’s Land Filled With Flies: It is not surprising then that, even though the San are A Political Economy of the Kalahari (1989, cf. Wilmsen much studied and very much alive in southern Africa, and Denbow 1990) that impacted most powerfully on Phytopharm should believe them to be extinct and it southern African San archaeology, anthropology and is equally unsurprising that the President of Botswana historiography. This work raised many questions about believes them to be relics of the Stone Age; these per- discourse on the San but, in particular, it questioned the ceptions are, in part, the result of some three-and-half degree to which the San are—and were for the last 2000 centuries of conditioning by Western representations years—isolated, pristine, egalitarian and unchanging 23 hunter-gatherer communities. Revisionists claim that researchers have ignored the interaction between San and other southern African communities (cf. Gordon 1992). Conventionally, southern African peoples are divided into four groups: San, Khoekhoen, Bantu-speakers and European colonists. Revisionists argue that any discussion of the San must consider their relationship to these other groups. The primary challenge laid down by the revisionists, then, is the production of a past that integrates the San into broader regional and even global political economies. Without this integration, according to the revisionists, we deny the San a history and construct them as cultural and temporal isolates. whom they interacted and, in many cases, acculturated. Wilmsen’s treatment of the San as a mere economic underclass has led Susan Kent (1992: 56) to accuse him of giving the Kalahari San “history while denying them autonomy”. For Kent, the central problem with both traditional and revisionist positions lies with the extrapolation from small groups to the San as a whole. She points out that both schools of thought have focused largely on the !Kung-speakers of northern Namibia and Botswana. The !Kung are part of the larger family of Khoesan languages whose chief identifying characteristics are their click consonants. Stripping away the clicks reveals syntax and grammar so divergent, that the variThe revisionists, in turn, have been accused by more ous languages are not merely mutually exclusive but are traditional writers of having both feeble evidence and radically so (Traill 1978, 1980, 1986). Given this diversity, misinterpreting that material (Solway and Lee 1990). In it is better to question the nature of interaction between spite of these criticisms, the central issues raised by the different San groups scattered in space and through revisionists have resonated widely in such diverse as- time, rather than make all-embracing claims for all San. pects of research on the San as popular writing (Barnard 1989, Voss 1987), photography (Gordon 1997, 1998), Underpinning these debates about diversity and spatiofilm production (Ruby 1993, Tomaselli 1992, 1993, temporal variation are deeper concerns about identity. Tomaselli et al. 1992), museum display (Davison 1991, Identity is, of course, a pervasive motif in contemporary Dowson and Lewis-Williams 1993, Wright and Mazel studies in the social sciences and humanities. Although a 1987) and, of course, academic discourse (Hudleson recurring theme, however, identity often appears in the 1995, Schrire 1992). The impact of revisionist thinking literature as a taken-for-granted. Far too often, writers on southern African archaeology, in particular, has been do not specify clearly their conception of the process immense and today it would be very difficult to write of identity-formation. At times, this omission leads to a about San people in the past without considering their rhetoric that suggests that identity-formation is an end relationship to other communities. While it is widely ac- in itself and, indeed, the ultimate goal of all social proccepted that integrating the San into broader local and esses. Arguments of this type tend to slip into a form global developments is essential, researchers disagree on of functionalism in which identity is seen as relatively the theoretical and empirical means needed to achieve fixed, homogeneous, natural and thus unchangeable. such integrative pasts. The revisionists, for example, Michael Rowlands (1994: 132) labels this approach to have only been able to produce inclusive pasts at the ex- identity ‘primordialist’. The primordialist understanding pense of San independence. Wilmsen, for example, ar- of identity influences much of the traditional research gues that the San of the Kalahari Desert of Namibia and that has been carried out amongst the Kalahari San. Botswana are not a distinct cultural entity but, instead, are merely an underclass in a patron-client relationship Since the early 1980s, primordialist understandings of with Sotho-Tswana-speaking people and that they have identity-formation have increasingly come under presbeen so for the last 2000 years. For Wilmsen, San iden- sure from, what Rowlands (ibid.) terms, ‘interactionist’ tity is subordinate and incorporate to that of the more approaches. These approaches see identity-formation as hierarchically organized Bantu-speaking peoples with socially produced. Such approaches to identity-forma24 tion are especially evident in studies of the formation of nation-states. Most famously, Benedict Anderson labelled nation-states as ‘imagined communities’ because, as he points out, “the members of even the smallest nation will never know most of their fellow-members, meet them, or even hear of them, yet in the minds of each lives the image of their communion” (Anderson 1983: 15). Today, it is not just national identities that are understood as ‘imagined’ but almost all forms of collective identity are, in some ways, understood to be socially produced. Indeed, identity is “…always mobile and processual, partly self-construction, partly categorization by others, partly a condition, a status, a label, a weapon, a shield, a fund of memories et cetera. It is a creolized aggregate composed through bricolage” (Malkii 1992: 37). The area of research where these sentiments have been most extensively discussed is that of postcolonial studies. Scholars in this field do not subscribe to a single, unified theoretical position but, instead, they engage in “a series of discussions about the sorts of cultural forms and identities created through colonial encounters” (Gosden 2001: 241). Although there are more practitioners, three of the most prominent writers in postcolonial studies are Gayatri Chakravorty Spivak (1987), Edward Said (1993) and Homi Bhabha (1994). Central to the discussions of these three writers is a conception of culture as non-essentialized (Gosden 2001)—that is to say, for these writers, collective identities are not impermeable, bounded ‘facts’ but are instead fluid and constructed. Bhabha, for example, argues that “colonialism is not about the meeting of different cultural forms, colonizer and colonized, who maintain their own separate identities, but about the creation of hybrid and creole cultures resulting from sustained colonial contact” (Gosden 2001: 241). Whereas traditional approaches to studying the San tended to assume a primordialist position on identity, one might expect the revisionists to have adopted an interactionist approach that emphasizes postcolonial ideas on hybridity and creolization but this is not the case (but, see Wilmsen 1996). Although revisionist efforts concentrate on the interac- tion between the San and other southern African peoples, they do not seek to understand that interaction as a process of creolization but, instead, they reduce that interaction to one in which the San are merely subordinate to the dominant Bantu-speakers. The revisionist emphasis on political economy is largely responsible for producing this view of a one-sided incorporation. Political economy models, often with strong marxist underpinnings, have pervaded southern African historiography for the last two decades and while they have produced brilliant and often radical alternatives to previous histories, they have often done so to the detriment of other facets of historical research. Of this trend, Jean and John Comaroff (1988: 6) note, “there remains a tendency, in historical sociology, to explain processes of domination in terms of political and economic forces. In the study of state formation and imperialism, realpolitik is given precedence over ritual, material factors over the moral suasion of the sign”. Political economy does not consider how the San might have contributed to the production of a hybridized culture in non-economic and non-political ways. Political economy models, then, are not a very useful way of understanding the complex processes of creolization and a better analytical tool is thus required for the production of a postcolonial past. A better analytical tool should recognize the San’s codependence on their Bantu-speaking neighbours and acknowledge their marginal status in that relationship without reducing them to an underclass that lacks the ability to influence and manipulate interaction. Such a tool exists in the concept of the subaltern. Loosely derived from the work of Antonio Gramsci the word ‘subaltern’ refers to subordinated or hierarchically marginal social groups. The word is without theoretical rigor but useful for critically discussing concepts of the sovereign subject (Landry and Maclean 1996: 203). The Subaltern Studies Collective of historians has brought this word to prominence in their reinterpretations of Indian history. This Collective has been trying to write a history of India from the perspective of the colonized rather than from the perspective of the colonizer. The two fundamental questions that they ask are: 25 …how can we touch the consciousness of the people, even as we investigate their politics? With what voice-consciousness can the subaltern speak? (Spivak 1988: 285). These questions are aimed at rethinking “Indian colonial historiography from the perspective of the discontinuous chain of peasant insurgencies during the colonial occupation” (ibid.: 283). This project is largely paradoxical because there is an absence of subaltern “memoirs, diaries, or official histories” (Landry and Maclean 1996: 203). Nevertheless, as members of the Subaltern Studies Collective argue, since the prose of colonial archives takes its form from the will of colonial administrators, it is predicated on the will of the colonized. Consequently, it should be possible to read the consciousness of the subaltern as a factor in the construction of those archival documents (cf. Dirks 1992). For the Collective, then, what is absent in the archival documents is as important as what is there. privileging of subaltern consciousness as autonomous (ibid.). Ultimately, for Spivak, the subaltern cannot speak through the writing of the Subaltern Studies Collective because the members of that group are also privileged. This last point of Spivak’s critique, however, is at odds with much of postcolonial thinking because it assumes that ‘subaltern’ is an essentialized group rather than being a hybridized group themselves. Nevertheless, we must take cognisance that any past produced by Western writers may not necessarily coincide with that produced by a San or San descendant. Nevertheless, if we take care to avoid Spivak’s criticisms, the concept of a subaltern offers a better way of understanding change in San communities than does the political economy model employed by revisionists. The concept allows us to write about the San as a marginalized group but without reducing them to a powerless economic underclass. Postcolonialism, then, emphasizes the perspective of the marginalized and disempowered in the colonial process while acknowledging that both colonizer Spivak (1996: 214), an eminent member of the Collecand colonized contributed to a hybrid social milieu. tive, both admires the efforts of her colleagues but also reads their work “against the grain” so to speak. For Uncovering the viewpoint of the marginal is not an Spivak, the two major contributions of the group are: easy task. There is often a lack of empirical evidence for their perspectives. In the absence of written matefirst, that the moment(s) of change be pluralized rial by the colonized, as with the Indian situation, some and plotted as confrontations rather than transition (they would thus be seen in relation to histories researchers hold archaeology up as a means through of domination and exploitation rather than withwhich we can uncover the perspective of the colonized. in the great modes-of-production narrative) and Through material culture remains produced by the colsecond, that such changes are signalled or marked onized, some argue, it is possible to retrieve something by a functional change in sign-systems. (ibid. : 205) that is not there in the written colonial documents. According to Spivak, then, the Subaltern Studies Col- While, in principle, this is liberating, in practice, the relective offers a new understanding of change that is not covery of such a perspective depends more on the relanecessarily tied to marxist-inspired political economy tionship between the material culture and archaeologimodels. Nevertheless, Spivak reads the Subaltern Stud- cal thought than it does on the material culture itself. ies Collective material with a critical eye. For her, their Given that the discipline grew largely out of colonial efforts are problematic because they assume a subaltern interactions (Gosden 1999, Rowlands 1998), archaeolconsciousness that is homogenous and autonomous; ogy carries with it a legacy of colonial power-relations. they assume that “there is a pure form of consciousness” At an obvious level, these power relations are evident, (ibid.: 286). As Spivak points out, “the colonized subaltern for example, in the size and location of museum collecsubject is irretrievably heterogeneous” (ibid.: 284) while tions. Those power relations are also evident, at a tacit “practical historiographic exigencies” do not allow the level, in the very theoretical positions that archaeologists 26 adopt. This point may be illustrated by a consideration of stone-age archaeological research in southern Africa. As in other parts of the world, this field of study categorizes lithic remains into technological categories (for reviews of southern African stone age research see Deacon and Deacon 1999, Deacon 1984, Mitchell 2002). These categories, their chronological frameworks and the technological functions of the different lithic forms that comprise them are at the heart of debates in southern African archaeology. Of course, this is not to say that southern African stone-age archaeology is empiricist and lacking in theory; there have been a number of theoretical efforts in recent years (e.g. Wadley 1997). Nor is it the case that southern African stone-age archaeology ignores the symbolic elements of material culture and most archaeologists today would agree that lithics do carry meaning and play an important social role (for more sophisticated analyses of technology and symbolism see Dobres and Hoffman 1999). It is the case, however, that there are very few studies that attempt to show what those meanings may have been in the past; it is just too difficult, if not impossible, many archaeologists contend, to recover those lost meanings. Moreover, much stone-age archaeology has tended to fall under the influence of processual archaeology. With its ‘scientific rationalism’, processual archaeology has tended to treat religion and symbolism as epiphenomenal and essentially unknowable in the archaeological past (Whitley et al. 1999). In those studies that do attempt to consider the symbolism, moreover, it is difficult to escape the impression that the symbolic aspects of the lithics are tacked onto the more important and manageable work of technological classification. In following this technological emphasis, archaeologists obscure areas of intellectual investment that may be more profitable. Paul Taçon (1991), Dorothy Hösler (1995) and, more recently, David Whitley et al. (1999) have shown that a consideration of the natural properties of archaeological materials are useful in understanding symbolism. For example, quartz, a prevalent material in the archaeological record, is not an ideal tool-making material. Although the substance can produce sharpedged tools for cutting, it is difficult to control the manner in which the material flakes. Quartz, however, has another important natural property; the crystal structure is piezoelectric (Whitley et al. 1999: 236). When subjected to pressure, quartz crystals produce a potential difference (voltage). A small amount of mechanical shock, such as rubbing two pieces of quartz together, will eject orbital electrons from the crystal’s atoms. When the electrons return to their atomic orbits, they lose energy in the form of a photon, whose frequency lies within the visible range of the electromagnetic spectrum. The ability of quartz crystals to convert mechanical energy (rubbing) into radiation energy (light) is called triboluminescence and it is this property that points to the prevalence of quartz at archaeological sites throughout the world in spite of its poor tool-producing properties. In cases where ethnohistoric data are available, in North America for example, the reasons for the widespread occurrence of quartz appears to be related to the articulation between the natural property of triboluminescence and the religious and symbolic cosmology of hunter-gatherers (Whitley et al. 1999) and not necessarily to the tool-producing properties of the material. Of course, the writers that I have mentioned here who are working on the natural properties of lithics are neither the first nor the only ones to do so. Indeed, Mary Douglas’ (1970) anthropological work in the 1960s and 1970s on natural symbols such as the pangolin was seminal. Nevertheless, these writers remain a minority within a discipline that is dominated by a technological paradigm that obscures other aspects of material culture, such as symbolism and religious connotations, that are as, if not more, important to understanding the role of those objects in the communities that made them. As political economy models marginalize ritual and sign, so the archaeological emphasis on technology places religion and symbolism on the periphery. The emphasis on politics, economics and technology may and often do, hinder the retrieval of a perspective of the colonized because they are not necessarily the mechanisms through which they communicated 27 or offered resistance. In order to produce a past that adequately considers a San prespective, then, we need to concentrate our efforts on those mechanisms of expression that we known were important to the San themselves and, more importantly, we need a theoretical framework that allows, as far as is possible, those mechanisms to take prominence in the research. ton, and Manhire 1990). Of these traditions, it is San rock art that is best understood thanks to a conceptual breakthrough in the mid-1970s (Lewis-Williams 1981, Vinnicombe 1976) that saw a move away from earlier, colonial and pejorative approaches to one of ethnographic-based interpretation. After nearly three decades of interpretation (see, for example, Lewis-Williams and Dowson 2000), San rock paintings and engravings are From the discussion in this chapter so far, it should amongst the best-understood rock art traditions in the be clear that a past that conforms to postcolonial ide- world. The developed understanding of the meanings als should comprise four fundamental characteristics: of many images makes the rock art a powerful data first, it should concentrate on processes of creolization set from which to investigate possible San perspecand hybridization. Second, it should focus on the col- tives of the historical processes that affected them. onized. Third, a postcolonial past should concentrate on the mechanisms that the colonized use for expres- Of course, researchers do not understand all aspects sion rather than simply on political economy. Finally, of San rock art equally and many issues are still deany theoretical approach used should allow, as far as bated vigorously. In particular, the integration of the possible, those mechanisms to ‘speak for themselves’ art into southern African history has attracted conrather than force them into specific Western theo- siderable attention since the breakthrough period of retical frameworks. In this work, I establish a founda- the 1970s. Researchers attempting to interdigitate tion from which we may begin to build a past of the rock art into the production of history have worked San that conforms to these four postcolonial ide- largely independently of postcolonial discussions and als. In taking steps towards producing a postcolonial these efforts, with one exception, did not influence past for Nomansland, I show how we may begin to the Kalahari revisionist debate significantly. This is inconstruct a San past that does not trivialize them or deed a pity because both debates would benefit from treat them merely as rhetorical introductory devices. a consideration of the rock art material and, in turn, efforts at integrating rock art into southern African history would benefit from the theoretical insights The Nomansland Project of the revisionist debate and postcolonial theory. Central to this work is the evidence of rock art—a form of expression that is known to have been im- Following Kent’s suggestion, this study takes as its startportant to the San. Within southern Africa, south of ing point the emphasis on the particularity of history the Zambezi River, there is an astonishing wealth of for small groups of San people who occupy a limited rock art. The vast majority of this art was made by the geographical area. Primarily an effort at writing a past of San, although some rock art was made by Bantu-speak- the San of Nomansland, a small area on the south-easting agropastoralist peoples (Hall and Prins 1993, Prins ern seaboard of South Africa that is part of the broader and Hall 1994), both independently and, as we shall area known as the Transkei2 (Fig. 1), the work is sympasee, in co-operation with San. There is also evidence thetic with revisionist ideals but is a conscious effort to that Khoekhoen pastoralists also made rock art (Dow- overcome some of their shortcomings. My goal here is son, Blundell, and Hall 1992, Manhire 1998, Manhire, to take the first steps toward writing a past of NomansParkington, and Van Rijssen 1983, Rudner and Rudner land San during the colonial period that considers the 1959, Smith and Ouzman 2004, Van Rijssen 1984, 1994, complexity of their interaction over time rather than Wadley 2001, Willcox 1959, 1960, 1984, Yates, Parking- simply reducing them to isolated relics or, on the other 28 hand, that treats them simply as an underclass of their more powerful Bantu-speaking neighbours. In order to do this, I draw on ethnographic, historical and rock art sources. Of these sources of evidence, it is particularly the rock art that is important. Between 1992 and 2004, in a period totalling some thirteen weeks that spanned both summer and winter months, together with colleagues and students, I visited 134 sites in Nomansland (see Fig. 6, Fig. 7, Fig. 8 and Appendix 2). Of these, 57 were located during the course of direct survey, 50 were located through the guidance of the local population and 27 were previously known to academia. The majority of these sites are situated in a small area that lies below the Drakensberg escarpment that is demarcated in the north by the Pitseng Valley, in the south by the town of Elliot, in the east by Ncengane Shelter and in the west by the Drakensberg escarpment. This demarcated area is referred to as the Nomansland Core Study Area throughout this work and it is illustrated in Figure 6. In addition to the sites that I visited in the Nomansland Core Study Area, I visited a further 16 sites in areas immediately adjacent to Nomansland for comparative reasons. These sites are mostly located in the region of the town of Rhodes in the Barkly East District, which is situated above the Drakensberg escarpment The images from these sites and their topographic, archaeological, ethnographic and historical contexts are discussed in seven chapters. I begin in Chapter One by describing Nomansland, its rock art and, in particular, the San people who lived there during the colonial period. The historical record for the San of Nomansland is substantial by comparison to other areas in southern Africa and, through a remarkable set of fortuitous events, we can see Nomansland from both the European colonizer and the colonized San perspective. On the one hand, there exist colonial observations of events in Nomansland during the nineteenth and twentieth centuries and, on the other hand, there exist documented oral accounts by the San of Nomansland, the people who lived with them and their descendants. It is the corpus of rock art, however, that makes Nomansland so remarkable. More so than any other area of southern Africa, we can tie San individuals to specific places in the landscape and sometimes those places are painted. I describe one of these places, known as Storm Shelter, in detail. Storm Shelter is an important site because, in some ways, it is similar to an archaeological type-site. The various images that comprise the painted panels at the site appear repeatedly at other Nomansland rock art sites. The images at Storm Shelter are mostly what we may label as ‘classic’ San rock paintings in that they are finely detailed and shaded polychromatic images. Nevertheless,there is another type of painted image at the site that is not of this classic tradition. These other images raise intriguing questions about the authorship of some of the Storm Shelter paintings. The rock art and rich historical data of Nomansland thus allow for the production of a past that emphasizes the perspective of the colonized. Chapter Two revisits the revolutionary period in southern African San rock art research from the late 1960s to the late 1970s. I discuss the present understanding of the meanings behind San rock art and I critically evaluate efforts at integrating San rock art into the writing of southern African history that developed after this period. In particular, I pay attention to the various theoretical frameworks used to incorporate rock art with historical events, evaluating their relative strengths and weaknesses. I conclude that, in spite of the valuable insights of these various approaches, we require a new theoretical framework to integrate rock art data with history. This new framework must facilitate the production of a past that seeks to understand San historical interaction as a complex process in which their identity was constructed and contested. In Chapter Three I discuss this new framework in detail. Here, I draw on material that has emerged from the social sciences and humanities over the last two decades, which considers the human body and embodiment as tools for social enquiry. Emphasizing ‘the body’, I argue, offers a way around some of the limitations of revisionist approaches and efforts made by rock art researchers to use the art as a source of historical information. It is particularly in the concept of embodiment 29 Fig. 6. A map of the Nomansland Core Study Area. The majority of the 134 sites that form the principal subject matter of this work are represented by black dots on this map. The positions of Ncengane Shelter and Storm Shelter are marked by stars. To the north-west of the five finger-like spurs of the Prentjiesberg lies the Drakensberg escarpment. 30 Fig. 7. A view of the Nomansland study area in summer. The basalt ridge of the high Drakensberg can be seen in the background. The rock art shelters are located in the sandstone hills, known as the “Little ‘Berg”, that can be seen in the foreground. Storm Shelter is situated in a valley hidden from view but in the far left side of this photograph. Fig. 8. In the winter months parts of Nomansland are often covered with snow and temperatures can fall as low as -14ºC. 31 that we can find a way of overcoming some of these problems. I outline a set of points taken from literature on ‘the body’ and embodiment that offer a strategy for the writing of a somatic past of Nomansland. In Chapter Seven, I consider the relation of the various images to one another and to the space in which they were made and viewed. Present understandings of the social production and consumption of San rock art emphasize production but pay too little attention In Chapter Four, I show how the strategy of the so- to consumption. Here again, a consideration of emmatic past is a valuable tool that allows us to add to bodiment and how it is governed by the physical space the existing, well-developed interpretative model of San of rock art sites in Nomansland allows for a reconrock art in the south-eastern mountains. I discuss the sideration of the social consumption of San rock art. Eldritch Images, a category of paintings that are rarely found in contexts that inform us about what they might In discussing the various images in these chapters, I represent. Drawing on widespread San ethnography, I show how the interaction between San and other peoshow that an understanding of San concepts of disease, ples in southern Africa is a complex, dynamic enterdeath and disorder and a consideration of how these prise rather than a simple political and economic one. concepts are physically embodied within certain beings, It is only by appreciating this complexity and dynamism allows us to identify this particular category of images. that we can begin to produce a past that takes into account the four principles of postcolonialism that I have In Chapter Five, I consider the unusual images found described. The production of such a past has repercusat Storm Shelter (and other Nomansland rock art sites) sions beyond academia. As we saw earlier, with regard that are not of the classic San type. The authorship to the controversy over the Hoodia plant, a perception of these images needs to be established. In consider- of the San as isolated and unchanging relics of the past ing potential candidates—the ‘other bodies’ that lived can lead to their exclusion from important economic in the Nomansland region—I show how Nomans- and political events of the present. In ignoring the inland has been part of a melting pot of diverse cul- teraction between San and other peoples in southern tural elements. It appears that the communities who Africa, archaeologists, anthropologists and historians made the classic art attempted to maintain a San iden- contribute to their present exclusion. Indeed, one retity in the face of this cultural diversity. Integral to visionist has even accused scholars who make rigid disthis identity was the control of the space of painting. tinctions between San and Khoekhoen, for example, as reproducing Verwoerdian ideology from the Apartheid Chapter Six considers how being part of a diverse cul- era (Schrire 1992). While this is perhaps overly harsh, tural milieu might well have challenged and changed San it is clear that writing about the San as an essentialized identity in the area. These changes are most evident in a group of people has important implications for their particular kind of painting that appears to be unique to present situation. On the other hand, some revisionists Nomansland and its hinterland. These paintings depict deny the San a status beyond being an economic unanthropomorphic images and accentuate selected parts derclass of Bantu-speaking society. Such claims provide of the human body. The emphasis of certain anatomical support for the continued marginalization and even elements comes at the expense of others. By emphasizing persecution of the San presently living in southern Afselected corporeal parts, I argue, individual artists shifted rica. In certain parts of the Kalahari Desert, the San the manner in which their identities were constructed in are forcibly moved off land that they have occupied order to take cognisance of their changing social milieu. for many generations in the name of mining and tourOne consequence of this changing milieu was a shift ism developments. Indeed, anthropologists who would in the ritual role of San painters, which further facili- study the San are often denied permits to do so and tated the emphasis on particular body parts in the art. tourism aimed at these indigenous peoples is discour32 aged. Those southern African governments who implement these actions and policies deny that the San exist as a separate, identifiable ‘ethnic’ group, claiming, instead, that the San are nothing more than Remote Area Dwellers. Given the revisionist claims that traditional views of the San play into certain political ideologies, it can be argued, conversely, that the revisionist approach also plays into contemporary political ideologies that support oppression of the San. Clearly, an approach to writing San history that neither isolates them nor treats them as a mere economic underclass of Bantu-speaking society is needed if we are to avoid offering support to political ideologies that seek to disempower them. 33 CHAPTER 1 NQABAYO’S NOMANSLAND: PEOPLE, PLACES AND PICTURES ‘Nomansland’ is a term that comes to us from the past and it refers to an area that is no longer called by that name. In the last 150 years, it has been labelled, either in its totality or in its various parts, East Griqualand, Transkei, North Eastern Cape and today north Eastern Cape. The multiplicity of names reflects the long-standing liminal status of this area of southern Africa. Situated in the interstice of the Natal and Cape colonies in the 1800s, this small area on the south-eastern seaboard of South Africa was very much betwixt-and-between (Fig. 1). It was in this remote and unexplored landscape, beneath the towering basalt peaks of the Drakensberg—a part of the larger south-eastern mountain complex— that some of the last San rock paintings were made and it was from here that the San launched their final resistance campaigns against the encroaching colonists. Both fearing and loathing the San, most colonists regarded them simply as ‘savage’ bandits who stole domestic stock. In a colonial mindset, the ‘savage’ San had no rights to land and the area that they occupied between the two colonies was, in a very real sense, ‘No Man’s Land’. While the word ‘Nomansland’ was used in the first part of the nineteenth century to indicate an absence of ownership, it is retained here because the subsequent labels that I have mentioned make a claim to ownership of land that belonged, at least from the last 20, 000 years, to the San. Indeed, as the historical material that I discuss in this chapter shows, there is evidence to suggest that small groups of San maintained a close relationship to specific areas of the Nomansland landscape throughout the colonial period (unless otherwise specified, I use the phrase ‘colonial period’ to refer to the period 1500-1900 AD). In particular, the evidence points to a close link between one group of San under the leadership of Nqabayo and a specific part of Nomansland. The irony of the possessive apostrophe in the title of this chapter thus counteracts 34 the notion that Nomansland was a ‘No Man’s Land’. While Nomansland was remote and inaccessible and while the San who lived there were closely affiliated with identifiable small areas, it would be a mistake to think of the groups as isolated and insular. When exactly they became part of broader local and global trade systems is not clear but oral and historical data suggest that they have interacted with Bantu-speaking peoples for at least the last 500 years and probably longer. This interaction took on many forms and varied over time; the San did not instantly become subordinates in a patron-client relationship, as revisionist arguments suggest for the Kalahari. Nor should it be assumed that the interaction between San and Bantu-speakers instantly produced a hybridized community in Nomansland. Instead, we need to see creolization as a process of domination and resistance that took place over a long period of time. Moreover, as will become clear in later chapters, the process of hybridization was made even more complex in Nomansland by colonialism. Remarkably, the San of Nomansland only became subordinate to their Bantu-speaking neighbours in the twentieth century. While these processes are discussed extensively in ensuing chapters, here I introduce the archaeological, historical and rock art material that comprises the evidence from which to construct San history in Nomansland. The San of Nomansland It is seldom the case that the individual artists behind the rock art images found throughout the world are known. Even in southern Africa, with its extensive ethnographic record, most of the artists behind the images remain nameless. Frustratingly, where names of artists are recorded, they are almost never tied to specific images in identifiable places. Of course, this dilemma is not exclusive to rock art studies but plagues most southern African hunter-gatherer archaeology. In some ex- ceptional cases, such as that of certain /Xam San from the north western parts of South Africa (Deacon 1986, 1988, 1996), ethnographic material allows for the tying of people to place and, more loosely, to the archaeology and rock art found there. The rarity of such occurrences makes the importance of Nomansland even more marked for, through remarkable serendipity, a number of sources of evidence allow us to link people to places over a period of some 150 years, from 1837 to 1990. Importantly, those places are sometimes painted and the people who lived there for that 150-year period made many of those pictures. The linkages between people, places and pictures in Nomansland are, arguably, the most concrete for any rock art area in the world. Even before the 150-year period, archaeological material provides a substantial record of San hunter-gatherer presence in the area. The earliest archaeological evidence for the occupation of Nomansland by hunter-gatherers dates from at least 29, 000 years ago (Opperman 1996). Although there are very few dates from archaeological deposits in the area, what is available suggests that hunter-gatherers were present from this time all the way through to the colonial period (Opperman 1987, 1996, 1999, Opperman and Heydenrych 1990). While the paucity of archaeological excavation in the area allows for only a hazy view of the ancient hunter-gatherer history of the area, a relative wealth of historical material allows for a much clearer view of the last 500 years. In particular, three major historical events provide substantial insights into who the San hunter-gatherers of the area were and how they changed between the mid-nineteenth and the late twentieth-century. Henry Francis Fynn’s Investigations The first of these events took place in 1848 when Harry Smith, the Cape Governor, requested Henry Francis Fynn (Fig. 9) to stay as British resident with the Mpondo, a Bantu-speaking group under the leadership of Faku, who lived in the Transkei region (Stapleton 2001: 71). Previously, Fynn had been a trader in Port Natal during the 1820s and in 1835 he had acted as a colonial agent for the Cape government to the Mpondo people. His command of the Nguni languages spoken by the Mpondo and his knowledge of local political matters made him an indispensable liaison between the Bantuspeaking peoples of south-eastern South Africa and the British colonies at the Cape and Natal. One of his duties during the 1848 expedition was to ascertain the relationships between the various peoples under Faku’s rule and the San living in the area that he governed, which included Nomansland. At this time, the farmers in the Natal Colony were experiencing heavy losses from stock raids. The farmers and colonial authorities suspected that the San living in Nomansland, in partnership with various Bantu-speaking groups, were guilty of these raids. Shortly after his arrival amongst the Mpondo, Fynn began his investigation of the relationship between the San and other people under the charge of Faku. It did not take him long to identify at least three distinct groups of San living in or adjacent to Nomansland in the midnineteenth century that, he believed, were responsible for the stock depredations in Natal. Fynn moved to Fig. 9. Henry Francis Fynn. It is through Fynn’s investigations in Nomansland during the mid-nineteenth century that three distinctive San groups can be identified. Reproduced from The diary of Henry Francis Fynn. 1950. Pietermaritzburg, Shuter and Shooter. 35 punish the San groups and their allies but his efforts were thwarted by Wesleyan missionaries living along the south-eastern seaboard who were sympathetic towards to the San. At the missionaries’ request, the colonial authorities removed Fynn and sent Walter Harding to investigate Fynn’s findings as well has his punitive actions. In the course of compiling his report, Harding interviewed members of the three San groups as well as Khoekhoen and Bantu-speaking people who knew them (see Harding 1850, 1850, 1850, 1850). Harding’s investigations support Fynn’s findings that there were three groups of San operating in the Nomansland region. While all three groups could and probably did make some of the images described in this work, the historical evidence suggests that one group in particular made many, if not most of the images discussed here. The three groups are the Thola and two other groups under the leadership of Mdwebo and Nqabayo respectively. stated that Mdwebo’s band was small, numbering only some thirteen men. The group had lived on the Mzimvubu River for some two years before moving closer to the Mkhomazi River. Owing to a number of encounters with colonial settlers, we can further establish the movements of Mdwebo’s group. In 1846, Jacobus Uys, the well-known Afrikaans-speaking pioneer farmer, twice encountered Mdwebo in the Mzimkhulu-MthamvumaBisi area. Information obtained from other sources shows that at this time Mdwebo’s band moved as far away from the escarpment as the Ngeli Mountains (Wright 1971: 126). In June 1848, James Melville learned from Bantuspeakers living on the Mkhomazi River that Mdwebo’s band was living at the Bisi River. In March the following year, Melville, learned from a local chief in southern Natal that a party of San living near the Bisi had moved away. At about this time, Mdwebo’s band was observed farther south amongst the Bhaca (ibid.: 126), an Ngunispeaking people, who had themselves moved into the According to the Fynn and Harding material, the Thola area of Nomansland after the Mfecane upheavals (ca. inhabited both sides of the Drakensberg Mountains close 1816-1824). Mdwebo’s band was thus slowly moving to the sources of the Mzimvubu River and hunted on southwards, quite possibly because colonial expansion the plateau areas beneath the high mountains. Although in the northern parts of the Natal Colony brought the exact numbers of this group are not known, Qinti, settlers ever closer to the Drakensberg Mountains. a son of Mdwebo, described the Thola as being large in number (Harding 1850, cf. Wright 1971: 126). In com- Walter Stanford’s Interviews parison, the groups under Mdwebo and Nqabayo were The second important historical encounter that sheds much smaller. According to Qinti, both Mdwebo’s group light on the San and which allows us to identify more and Nqabayo’s band were hostile towards the Thola. An accurately who the artists of some of the Nomansland account of an attack mounted by members of Mdwebo’s rock paintings were, took place in the 1880s. Sir Walter and Nqabayo’s groups in 1850 on the Thola confirms Stanford (Fig. 10), a magistrate in the Transkei area, enQinti’s statements (Z. 1850). The attack was, ostensibly, countered small San populations in the territories under to retrieve cattle that the Thola had stolen from colonial his command as well as a Bantu-speaking man who hade farmers yet, as historical evidence shows, both Mdwe- lived with the San in the mid-nineteenth century, at about bo’s and Nqabayo’s people were also implicated in stock the same time that Fynn was conducting his investigathefts and their attack was an attempt to shift colonial tions (Fig. 11). His discussions with these San groups suspicion away from them and on to the Thola. The and the Bantu-speaking man provide the most detailed hostility between Mdwebo and Nqabayo on the one description known of the movements of a particular San side and the Thola on the other suggests that the Thola band in Nomansland. One group of San with whom he did not make many of the images that I am concerned spoke comprised three individuals—the leader, Luhayi with here and we shall see the reason for this shortly. and another man, Mkahlila and a woman, Mamxabela. This man and woman were both between forty and Although Qinti described the Thola as numerous, he fifty years of age when Stanford first met them on the 36 Umnga River and then later again on the Tsitsa River (Maquarrie 1962). They related some of their life experiences in Nomansland to Stanford at these two meetings. Mkahlila told Stanford that he had been a small boy living at Gubenxa when a group of Thembu, a Cape Ngunispeaking people, under the leadership of Mgudhlwa attacked his group. The Thembu left him for dead but some of the band, who had been away hunting and so escaped the massacre, found him still alive. Stanford (Maquarrie 1962: 29) describes the woman, Mamxabela, who could still speak a little San when she met him, as follows: Mamxabela was a lively talkative little being. She told me that her husband who had died some years previously was a painter and she pointed to rocks in the distance where some of his work was to be seen. She evidently had a high opinion of his skill as an artist (ibid. : 29). The statements made by Luhayi and Mamxabela are corroborated by the statement of Silayi, a Thembu man, who lived and raided with Nqabayo’s group (Maquarrie 1962, Stanford 1910). In his statement made to Stanford on the 7th May 1884, Silayi describes how he had gone to live with Nqabayo’s people in about 1850. He had initially lived on the Xonxe River (White Kei) where there were still San groups living in close interaction with the Thembu who were under the leadership of Jumba. Silayi befriended Hans, a Khoekhoen man and his nephew, Ngqika. Hans was Ngqika’s mother’s brother and Ngqika’s father was a San man named Qako, who lived under the famous San leader Madolo (on Madolo’s history see Saunders 1977). Nqabayo’s group, according to Silayi, lived in the Drakensberg at the sources of the Xuka (the largest tributary of the Bashee River) and Qanqaru Rivers (today known as the Mooi River; it is a tributary of the Tsitsa River) and he first met them in the Umngqazo Mountains (also known as Apryntje’s Berg). From here, Silayi, together with Nqabayo’s band moved along the Kraai River, raiding livestock from Boer farmers near the town of Dordrecht before moving back over the mountains of the Ntunjankala (also known as the Gatberg). There are other places that Fig. 10. Sir Walter Stanford who encountered several small groups of San in the area formerly known as Nomansland in the 1880s. Stanford Papers, Jagger Library. Silayi mentions by name—the Indenxa River, the Tsonio River, Waschbank, the Kowe River (Slang River), Hlankomo—where the band either stayed or passed as they raided livestock from the colonists. In other places, such as downriver on the Gubenxa or at Maxongo’s Hoek, they fought skirmishes with Thembu and Boers. The names of these places mentioned by Silayi either are the same as or are places contiguous with the names that Mamxabela and Mkahlila mentioned in their interviews with Stanford; the landscape that these two and Silayi lived and moved in was the same. Importantly, according to Silayi, Mgudhlwa’s Thembu attacked Nqabayo’s band in 1858 because they had stolen some of his cattle (Maquarrie 1962: 36). The Thembu destroyed the group in this attack but Nqabayo and a few of his men got away, 37 Fig. 11. The group of San that met with Stanford in the 1880s. The elderly woman and the man with the pipe occur in several of his photographs. Stanford Papers, Jagger Library. seeking refuge under Mditshwa, the leader of the western Mpondomise (another Cape Nguni-speaking group) (Stanford 1910). Almost all the women and children in Nqabayo’s band died in the attack and the subsequent massacre of the San prisoners by the Thembu. Only two boys and one woman, Nqabayo’s daughter, escaped. was one of the three survivors of the women and children (see below). Luhayi and Mamxabela we cannot be so certain of but it is easy to entertain the idea that they were the other two who survived the attack at Gubenxa. Indeed, Mamxabela stated that she had joined Mditshwa’s band at some point in her life, further suggesting that she was a member of Nqabayo’s It is curious that the group that Stanford met at the group at the time of attack. This raises the possibility Umnga and Tstitsa Rivers also comprised three indi- that Mamxabela may have been Nqabayo’s daughter. viduals, two men and one woman—Luhayi, Mkahlila and Mamxabela. Mkahlila confirmed the linkage be- It is unfortunate that Stanford does not provide the dates tween his small group and Nqabayo’s by his account for either of the two occasions on which he met Luhayi, of his survival after the attack by Mgudhlwa on the Mkahlila and Mamxabela. Nevertheless, some clues to San at Gubenxa. Silayi, in his statement, confirms that the dates are amongst Stanford’s archival material at the this is where the attack occurred. Mkahlila, then, was University of Cape Town. Within this collection there a member of Nqabayo’s group and, almost certainly, are several black-and-white photographs of San that 38 Fig. 12. The man and woman observable in Fig. 11, standing and without their fellows. Stanford Papers, Jagger Library. Fig. 13. Reverse of photograph in Fig. 12. The text clearly identifies the man and woman as Mkahlila and Mamxabela respectively. Stanford Papers, Jagger Library. have been published previously (Jolly 1992). There is little accompanying information with the photographs but written on the envelope that contains the photographs is: ca 1885. In faint pencil on the back of one of the photographs (Fig. 12 & Fig. 13), is the following: ingly mute, both on the occasion of meeting Silayi and then later when he met the small group of San, it is clear that he maintained his interest in them until late in life. Two items amongst his personal papers held at the Jagger Library at the University of Cape Town attest to his sustained interest. First, in 1908, he collected a statement from Mgudhlwa (Stanford n.d.), the Them- Man: Kahlila 3 Woman: Mamxabele The same two faces on the front of this picture are identifiable in most of the other photographs (See Fig. 11). Quite clearly, Stanford must have met this small group of San shortly after meeting Silayi—probably within a year following that event. The encounters with both Silayi and then later with this small group intrigued Stanford and although his diaries are surpris- bu chief responsible for scattering Nqabayo’s small group of San. Mgudhlwa had this to say about the San: I remember the Bushman chief Madolo. He was living at Bangindhlala (Mt. Arthur Mission Station formerly known as the Bushman School). Ndhlela the father of Stokwe,4 Chief of a branch of the Quati tribe attacked Madolo there and broke up the tribe. Madolo fled to Sidutyini (St. Marks). Later Jumba and his tribe returned to Tem39 buland. It was about the time of Umhlahaza (1857) that I attacked the Bushmen at Gubenxa. The name of the chief was Nqabayo. It was before this that Silayi one of our young men had left us and joined these Bushmen. I attacked the Bushmen because they stole horses from us. Magombe and two others of the Bushmen stole the horses. They were followed. They fought. One of them was killed and one of our men named Nkani. A poisoned arrow killed him. At Gubenxa we surrounded the Bushmen and their families. We killed all the men who were there. Some—amongst them Nqabayo—were away at the time hunting. They escaped. We captured the women and children. On our way home and without my knowledge, the young men fell upon the women and children and killed them too5. Mgudhlwa’s statement further corroborates both Silayi’s statement as well as those of Luhayi’s group. The second item in Stanford’s papers is a letter written to Stanford by Arthur Gladwin of Tsolo on the 10th August 1909; in the letter, we learn of the fate of Luhayi’s small group: After staying with Mditshwa’s Mpondomise for some time after the attack at Gubenxa, Nqabayo returned to the Drakensberg Mountains and the last Silayi heard of him, he was at the sources of the Mzimvubu River (Maquarrie 1962: 36-37). Silayi left Nqabayo’s group before the attack to return to the Thembu. At the same time Hans and Ngqika left for Qokolweni Mission Station and afterwards they went to Adam Kok’s country, the site of present-day Kokstad (ibid.). It seems that even after the dispersal of many of its members, many members of Nqabayo’s group stayed amongst the Mpondomise, acting as rainmakers for them. Importantly, Stanford’s interviews with Silayi and Luhayi’s group some two decades later confirm Fynn’s suspicions that it was the San of Nomansland that were responsible for the raids on the Natal Colony. The Discovery of Manqindi Dyantyi The third historical encounter important to understanding the changing situation of the San in Nomansland occurred about 100 years after Stanford encountered Silayi and Luhayi’s group of San. In the mid 1980s, an old woman was found living near the present-day town of I went up on Friday to see the Bushmen. Ha LuTsolo, a short distance south-east of the area that once hayi is dead recently. Kahlila who was wounded at Gubenxa died. The old woman Mamxabelo is getwas called Nomansland (Jolly 1986, Lewis-Williams ting mentally and physically feeble and I’m afraid 1986). This woman is known as ‘M’ in much of the lither statement would not be satisfactory. Her son erature that concerns her in order to protect her identity however is an intelligent boy, he went through the but after her death in 1988, her full name has been disKimberley siege and he seemed very well up in their history. Would you be satisfied with a stateclosed as Manqindi or Maqoqa Dyantyi (Jolly and Prins ment made by him coached and assisted by the old 1994). Dyantyi was an elderly woman when she came to woman and the others, alone he may be inclined the attention of archaeologists (Fig. 14). In a number to romance after his Kimberley achievements of interviews with different archaeologists, she offered but in the presence of the others he might “toe the line”. Luhayi’s successor seems rather a ruff. information about her family and her past. According I again spoke to Ma Mamxabelo about going to to Manqindi Dyantyi, her father was a San man named Cape Town to permit Péringuey to take a clay cast Lindiso who had married an Mfengu woman and then of her but she is still afraid. I however later when settled down amongst the Mpondomise near Tsolo apart put it to her son who is quite game so I asked him to reassure his mother and get her consent. (Prins 1990: 110). Lindiso had been a rainmaker and an Mamxabelo tells me she was not with her peoartist and had made paintings in a rock shelter, known ple when Mgudhlwa killed them at Gubenxa but as Ncengane Shelter, which is in the Bushman Cuttings with her husband at Tabankulu (Gladwin 1909)6. Area close to the town of Tsolo (Fig. 16); Dyantyi took Gladwin’s letter confirms that Mamxabela was not archaeologists to the site on several occasions(Prins Nqabayo’s daughter and that Luhayi was one of the 1994). Dyantyi’s uncle, Masela, was also a rainmaker and three survivors of the Thembu attack at Gubenxa. an artist and her elder sister, Chitiwe, was also a rain40 Fig. 14. Manqindi Dyantyi at Ncengane Shelter. Behind her, the painted images are clearly visible; some of these images were made by her father, Lindiso, in the early decades of the twentieth century. maker but was not a painter (Prins 1990: 113). As we shall see in later chapters, Dyantyi’s testimony concerning her family, along with other oral histories sheds important light on San ritual and painting in Nomansland. Dyantyi’s identification of her father, Lindiso, is important because it allows us to link Fynn’s observations and the Stanford material to the twentieth century. In 1913 Louis Péringuey, the director of the South African Museum, wrote to Stanford asking for his assistance in contacting a Mamxabele so that he could learn more about the San method of preparing paint (Jolly 1999: 61). Stanford referred the letter to M. Apthorp at the Department of Native Affairs. Apthorp’s response dated 14 July 1913 reads (first published in Rudner 1982: 54): I regret to say that both Mamxabele, the old woman and the native mentioned in the communication are dead. The magistrate of Tsolo, however, got into communication with her son and the former has furnished the following report. Her son, Lindiso, states that he was told by his mother during her lifetime that white, red and yellow clay and charcoal made from the wood of the Coral Tree7 were mixed with water and the fat of the bushbuck or other animal. The pigment was then applied to the rocks etc. with a piece of grass, which was sometimes split to make it resemble a brush. Poponi, a relative of Lindiso’s, corroborates him and says that he remembers, on one occasion, seeing Mamxabele’s husband mixing and applying some paint. It has been pointed out that this is almost certainly the same Mamxabela that Stanford encountered in the 1880s (Jolly 1999: 61). This means that Mamxabela, the mother of Lindiso, is the grandmother of Manqindi Dyantyi and as Mamxabela was connected to Nqabayo’s band through her association with Luhayi, this means that Manqindi Dyantyi, as the granddaughter of Mamxabela, is descended from Nqabayo’s band. 41 Fig. 15. Manqindi Dyantyi walking in the Bushman’s Cuttings area. Ncengane Shelter is situated in the Inxu River Valley, which can be seen in the background. In addition to these three important historical encounters, there are also many other colonial observations as well as oral accounts by San descendants or Bantu-speakers who interacted with them. These accounts allow us to clarify further the history of Nqabayo’s people. In 1891, a San family was known to be living in a rock shelter overlooking the Inxu or Inqu River near Gqaqala (Gibson 1891: 34). These San were rainmakers for Mpondomise people and a number of other accounts in the first two decades of the twentieth century also mention San rainmakers in the Tsolo area (Callaway 1919, Hook 1908, Scully 1913, see Jolly 1992 for a compilation of these accounts; cf. Butler 2001). Whether these accounts refer to one family or several is unknown but evidence suggests that there were certainly several San people living in the Tsolo District. Oral testimonies collected from Mpondomise peoples by the anthropologist, David Hammond-Tooke (n.d.), provide some idea of the number of San people living 42 in the area in the first few years of the twentieth century. According to some Mpondomise peoples, in 1905 the chief magistrate of Tsolo ordered all the San living on the Umnga River (where Stanford had encountered Luhayi’s group in the 1880s) to be collected together; some thirty men, women and children were present. There were apparently three family units present, one of whom was under the leadership of Luhayi. These San were last seen on farms in the area of the town of Ugie. While the San of Nomansland can be tied to specific areas on the landscape where they seem to have stayed for many decades, it is clear that they were not isolated from other San. Nqabayo’s group, for example, was related to the San under the famous leader Madolo— mentioned by Mgudhlwa in his statement—that lived to the south west of Nomansland. As we have seen, a member of Nqabayo’s group, Ngqika, had a father that lived with Madolo. There were also other groups in and around Nomansland, such as that under the leadership of Flux Lynx, but the relationship between these groups and those under Mdwebo, Nqabayo and Madolo is not yet clear. It would seem that, at times, the San congregated around leaders such as Madolo in large numbers while at other times they splintered off into smaller groups such as that under Luhayi. What little evidence is available suggests that the various San groups in and around Nomansland—with the exception of the Thola—largely co-operated with one another. I have already mentioned, in the Introduction, the mutual unintelligibility of San languages (Traill 1978, 1980, 1995) so this co-operative nature between the groups living along the base of the Drakensberg suggests that they may have spoken the same language, although they may have communicated through another language such as Xhosa. Nevertheless, in the early 1930s, Dr H. Anders encountered two San men, one of whom he described as an old friend, in the Bushman Cuttings area, where Ncengane Shelter is situated. These two men still spoke some San and Anders recorded some 140 words. The language they spoke has been called !Gã !ne (Anders 1934/5, Traill 1995). Remarkably, the two men that Anders interviewed were called Lindiso and Poponi and, given their close proximity to Ncengane Shelter when Anders met them, and the fact that Apthorp’s letter describes Poponi as a relative of his, this is almost certainly the same Lindiso that was the father of Manqindi Dyantyi and the son of Mamxabela (cf. Jolly 1999). Nqabayo’s group and the San of Nomansland thus spoke !Gã !ne. The other group of San in the Nomansland area—the Thola—were enemies of Mdwebo and Nqabayo and their habitation above the escarpment in what is today Lesotho suggests that they spoke a different language, possibly Seroa (Traill 1995). There seems, then, to have been a loose coalition of San groups, speaking the same language, living along the foothills of the Drakensberg Mountains, in and around Nomansland for much of the nineteenth century. encounter with Luhayi’s band and with Silayi in the 1880s and the discovery of Manqindi Dyantyi in the 1980s— together with the many lesser accounts provide an unparalleled historical account of a specific group of San peoples over a period of some 130 years. The historical material allows us to link known San individuals over time and it allows us to link those individuals to a landscape (Appendix 1). Most importantly, that landscape is one in which there are numerous painted rock shelters and through the historical material that I have discussed we know of at least three generations of San artists who were painting in and around Nomansland between the mid-nineteenth century and possibly as late as the 1920s. The importance of all this historical material lies in the attribution of real people to the landscape of Nomansland. When speaking of the San of Nomansland, the word should not convey the usual impression of an impenetrable, undifferentiated mass of people. Rather, the San of Nomansland were individuals with children, parents, ancestors; each of those individuals had their own personal history and, in many cases, their own personal tragedy. It is these people—Nqabayo’s people—who were the true owners of Nomansland. Today, what was Nomansland includes the present towns of Ugie and Maclear and their hinterland. It is a long drive from any major urban centre in South Africa; the most common way of driving there is through the Kraai River Pass, past the road west to Dordrecht, down the precipitous Barkly Pass, often iced over in winter. As one approaches the town of Ugie, familiar names appear such as Xuka Drift Siding and the Gatberg with its distinctive hole can be seen immediately to the east of the road and situated beneath the picturesque Prentjiesberg—the contemporary name for the Apryntje’s Berg— lies the town itself built only a few years after Nqabayo’s group was attacked at Gubenxa. It is in amongst the hills and valleys of Ugie and nearby Maclear, around, on top of and adjacent to all the places that Silayi and the other members of Nqabayo’s group mentioned that one finds rock art (Fig. 16). Mamxabela stated that her husband The three historical encounters that I have discussed— was a painter and in response to Stanford’s questions, Fynn’s observations between 1848 and 1850, Stanford’s Silayi described how Nqabayo’s group had made their 43 Fig. 16. This map shows the close proximity of the rock art sites from the core study area in Nomansland to places identifiable from the Fynn, Stanford, Manqindi Dyantyi and other historical material. 44 paint brushes and pigment and then, apparently as an afterthought, he uttered, “They could paint very well”. The art that forms the principal subject matter of this thesis comes from the very areas that are mentioned by Silayi, Mamxabela, Mkahlila and Luhayi and many of the images that will be discussed here, almost certainly, were made by Nqabayo’s band and their descendants. Whether the Thola, Mdwebo’s band or other groups made some of the images as well is unknown as their do not appear to be observations of these two groups painting. It is unlikely that the Thola would have made the images discussed throughout this work because they occur within the area occupied by Nqabayo’s band that were, as we have seen, hostile towards them. It also seems that Mdwebo’s band was situated slightly north of the area where Nqabayo’s people painted, living amongst the Bhaca. Almost certainly then the images discussed here come from Nqabayo’s Nomansland and while we cannot say with certainty exactly which images they made, Nqabayo’s people probably made many, if not most of them and, more importantly, they would have known the locations of most of the sites and they would have seen many of the images. A Nomansland Rock Art Site It was in one of the remote valleys of the Ugie-Maclear area where Nqabayo’s people walked that, in December 1992, members of the Rock Art Research Unit (now, Rock Art Research Institute), University of the Witwatersrand, discovered a remarkable painted site, now known as Storm Shelter (Blundell and Lewis-Williams 2001, Fig. 17 & Fig. 18). Three years later, the site was recorded through the direct tracing method and comprehensive photography. Tracing, as it always does, revealed far more than casual observation. There are some 231 recognisable images in the panel that stretches approximately six meters in length and is nearly two meters in height (Fig. 24). Indecipherable remains attest to the even greater richness that was once visible. There are as many as five layers of paint in places and a great variety of pigments were manipulated in complex shading patterns during its construction. The imagery is of the classic San rock art of the south-eastern mountains as well as certain unique variations never before seen. Both ubiquitous and unique images shed important light on the meaning and significance of San rock art in Nomansland and as the various images will feature repeatedly in forthcoming chapters, they warrant some description here. As with most sites in the Drakensberg Mountains, there are large numbers of antelope in the panel that are identifiable as principally eland (Taurotragus oryx). The some 43 eland images are depicted in a variety of postures, including some that are in an upside down position; often the hair on the back of their necks is erect. Near the centre of the panel, where the largest eland are to be found, is the greatest concentration of superimpositioning (Fig. 18). Lower down, towards the right hand extremity of the panel, is a cluster of some 32 rhebok (Pelea capreolus), depicted in similar complex shading as the eland but, curiously, without any horns. The only other identifiable species of antelope in the panel is hartebeest (Alcelaphus buselaphus) with only some five images present. The numerical order of depictions of these antelope is something that occurs time and again at certain Nomansland rock art sites; I thus term this the 1-2-3 arrangement. While human figures outnumber depictions of animals in the panel, many of the anthropomorphic images have antelope heads or hooves. These therianthropic (part-human, part-animal) images are for the most part human in form, with the exception of one image, close to the centre of the panel. This image is almost entirely antelope in form except in posture; its legs articulate incorrectly for antelope legs (Fig. 26). This antelope in human posture casts doubt on any clear distinction between human and animal subject matter in the panel. This suspicion is abetted by another image depicted in a translucent-white colour, with human legs, arms and body but with an antelope head. It is one of a cluster of six images in similar translucent-white pigment, some of which appear to be human and others that are therianthropic (Fig. 19). The similar state of preservation and 45 Fig. 17. Storm Shelter. The large boulder that has fallen from the roof in such a way as to prevent domestic stock from rubbing against the images is clearly visible while the person provides a sense of the scale of the panel. The panel is about 6m in length and 1m in height. Fig. 18. The density of painted images at Storm Shelter is impressive, particularly toward the centre of the panel where there are as many as five superimposed layers of paint. Area of view is the 4m mid-section of Storm Shelter. 46 Fig. 19. One of the six images at Storm Shelter that are painted in translucent white. While the figure has a human body, it bends forward and has an antelope head. The figure is 10cm in breadth. tight clustering of these images suggest that they were all it represents something more than just a human being. painted at about the same time and comprise a “scene” Intriguingly, to the right of this figure is another, smaller in the sense that they appear to interact with one another. one with a similar large head and whose facial features include a chin, lower and upper lips, an eye that narrows Above and below these white images are the only other at both ends and a tapered ear (Fig. 21). This figure does representations of identifiable animals other than an- not, however, have a beret-like headdress but, instead, telope at the site—two felines. One of the felines is has a headdress that resembles the strange head of one covered in dots and has the appearance of a leopard of the eland in the centre of the panel. These distincwhile the other very large one has a flowing mane and thus appears to be a depiction of a male lion. This lion is painted partly over an enigmatic figure that is partly human in appearance but has non-human features such as a tapered ear (Fig. 20). Importantly, the figure has no legs and its body ends in a rounded waist. This ‘HumptyDumpty’ figure also has a head that is greatly exaggerated in size in relation to the rest of its body. The head is in profile and is depicted in considerable detail with chin, upper and lower lips, nose, eye and ear all present. The figure also has a headdress that resembles a French beret in form. The tapered ear, the large, narrowed eye and the missing legs of this figure, strongly suggest that tive anthropomorphic figures immediately catch the eye when glancing at Storm Shelter’s main panel because they stand out from the other images in terms of their size, detail, position in the panel and unusual facial features. Of all the curious images in the panel, the most intriguing is that of the thin red line, fringed by small white dots on either side. The line, between two and five millimetres in width and about 89cm in length, weaves its way through the small cluster of rhebok at the lower right of the panel, seemingly connecting them together. On a small boulder opposite the main panel at Storm Shel47 Fig. 20. The primary distinctive anthropomorphic figure at Storm Shelter. The complete figure is 30cm in length. Fig. 21. The secondary distinctive anthropomorphic figure at Storm Shelter. Note the detailed facial features. 23cm in length. Fig. 22. Strange white anthropomorphic figures, such as this one from RARI-RSA-BAE2, can be seen at Storm Shelter and at sites throughout Nomansland. Figure is 11.5cm in length. Fig. 23. Animals that appear to be depictions of eland but that are not as fine as ‘classic’ San rock art images are found at several Nomansland rock art sites, including Storm Shelter. 20cm in breadth. 48 ter are a number of antelope that appear to be representations of eland. These images are painted in a thin, powdery white pigment and are very different to the eland in the main panel that are thickly painted in diverse pigments. These are just some of the more prominent images in Storm Shelter; an exhaustive description of all the images would take too long for the purposes of this work. Importantly, the images at Storm Shelter may be seen at many rock art sites in Nomansland (Fig. 22 & Fig. 23). The process of studying new as well as previously known sites, revealed that many, but not all, of them show a pattern similar to that at Storm Shelter. In particular, this pattern is evident at sites with depictions of large anthropomorphic figures that stand out from the rest of the paintings, as do the pointed-ear figures at Storm Shelter. It is at the sites with distinctive anthropomorphic figures that the pattern is most apparent. At almost every one of these sites, there are depictions of the three antelope species—eland, rhebok and hartebeest and in that numerical order. There are the depictions of therianthropes, the clusters of translucent-white figures, felines, rain-animals and thin red lines fringed by white dots. Of course, not all sites have this pattern; there are smaller sites at which only certain of the elements are present. Still other sites have none of the elements. Nevertheless, the pattern is apparent at sites, such as Storm Shelter, where there are large anthropomorphic figures that stand out from the rest of the paintings. that I have identified as well as to provide an explanation for their consistent arrangement in a pattern. Doing so for an area like Nomansland has particularly important repercussions for San rock art research both in the Drakensberg and the rest of southern Africa. Nomansland’s Past in Paint and People The number of layers of paint at Storm Shelter suggests that the images were not all painted at the same time. The freshness of the white pigment at Storm Shelter and at other sites in Nomansland also suggests that many images were made in recent times. Of course, not all images are recent and some must date to a time before Nqabayo. Certainly, the dates for rock art in other parts of southern Africa suggest a great antiquity for some of the images. In southern Namibia, at Apollo 11 Cave (Wendt 1976), the oldest painted representational images in Africa are dated to about 27, 000 years before the present (3 C14 dates: 26300±400; 26700±650; 28400±450). In other parts of southern Africa, such as in the Matopos, paintings are dated to at least 10, 000 years before the present (Walker 1996: 11-14). These dates certainly give great antiquity to rock painting in southern Africa. The exciting discovery of incised ochre pieces at Blombos Cave, on the southern coast of South Africa, that date to between 75, 000 and 100, 000 years before the present, suggest that artistic production in southern Africa is far more ancient than anyone has dared to imagine so far (Henshilwood et al. 2001, 2002). Since, as we have seen, hunter-gatherers were occupyThere are also idiosyncratic images in the rock art sites ing Nomansland at least 29, 000 years ago, it may legitiof Nomansland. At Storm Shelter, there are two images mately be asked how many of the paintings in Nomansof mushroom-like objects in the main panel. These re- land date from the nineteenth and twentieth centuries main unidentifiable and have, as yet, not been found and how many from earlier periods. There are, however, anywhere else. While these idiosyncratic elements are four arguments that mitigate against the great antiquity important in obtaining insights into the art, the dif- of most of the rock art in Nomansland—subject matter, ficulty of accurately identifying what they represent preservation, direct dating and historical information. sometimes prevents adequate understanding. It is not my intention here to decipher all these idiosyncratic First, the subject matter of many of the images of images but rather to concentrate on the role of idio- Nomansland is something that has caught the attensyncrasy as part of the pattern of Storm Shelter type tion of many visitors to the area. Indeed, even Stansites in Nomansland. It is my intention in later chap- ford commented on the images that he saw. It apters, then, to elucidate both the elements individually pears that he did not see any of the painted sites or 49 Fig. 24. Black-and-white drawing of the primary panel at Storm Shelter. Solid=Black; Dense Stipple=Red; Other Stipples=Lighter Shades of Red images that members of Nqabayo’s band made and it was only many years after his encounter with Silayi and the others of Nqabayo’s band, that Stanford visited rock art sites in Nomansland. He describes this visit: I afterwards found that many of the Bushman paintings were obviously of a modern time. These I saw in the Maclear district during the Boer War. Encounters between white men on horseback carrying long rifles and Bushmen with their bows and arrows were depicted. Success always appeared to be on the side of the Bushmen, the white horseman shown as falling with an arrow through him (quoted in Maquarrie 1962: 29). Stanford was incorrect in his final observation and tragically, as we shall see in later chapters, in the long 50 run, the San were neither successful in actual or painted conflicts. Indeed, it is difficult to accept Stanford’s description as accurate and certainly it would be very unusual to find a painting depicting a colonist shot with an arrow. Nevertheless, it is important that it was the historical subject matter of the art of Nomansland that impressed Stanford. Survey work for this project as well as work carried out by Patricia Vinnicombe (1976) reveal a number of sites in Nomansland and adjacent areas with images of horses, wagons, farmers and soldiers with guns and people in colonial period dress. Indeed, at Ncengane Shelter itself, where Lindiso painted, there are images of wagons and horses with riders. The subject matter then, reveals that many paintings were made during the colonial period. Secondly, many of the images in Nomansland, whether d; Clear=White. Scale in centimetres. they are of recent historical subject matter or not, are done in white pigment. In southern Africa and other parts of the world, only a few images show traces of white as the pigment is fleeting. In Nomansland, many of the images still show significant quantities of white pigment, suggesting that they are relatively recent. Given the fragile nature of the soft sandstone of the south-eastern mountains on which the images are found and given that is the most significant watershed in southern Africa, the conditions are not conducive to the longevity of white or other pigments; elsewhere, such as in the Matopos, the hard-weathering granite and the semi-arid climate lend themselves to the preservation of white pigment. The abiding white pigment of Nomansland, then, further suggests a relatively recent age to the images considered here. The third reason for the argument that most of the images in Nomansland are relatively recent comes from Accelerated Mass Spectrometry (AMS) radiocarbon dates taken off samples removed from San paintings in an area to the north of Nomansland. There are very few radiocarbon AMS dates for rock paintings in the south-eastern mountains. In 1993, samples were taken from paintings in the northern parts of the mountain complex and two usable dates were obtained. Both dates are relatively recent in time: one date is 507-297 BP or AD 1443-1653 and the second is 690-50 BP or AD 1260-1900 (Mazel and Watchman 1997:448). More recently, a further eight AMS radiocarbon dates were obtained (Mazel and Watchman 2003). If the techniques used for the extraction, preparation and dating of these eight samples survive critical scrutiny, then 51 the age of some images in the south-eastern mountains is between 3000 and 4000 years BP Clearly, the existing dates for art in the south-eastern mountains do not extend nearly as far back as the earliest dates of the occupation of Nomansland by hunter-gatherers. This is not to say that paintings were not being made in Nomansland at that time. The antiquity of art elsewhere in southern Africa suggests that the San would have been making art in Nomansland at that time as well. Those ancient images, however, have probably all disappeared or only minute traces remain given the unfavourable conditions. Most images that are clearly visible in Nomansland then are probably of recent origin. The fourth and last reason that leads me to suggest that many of the images in Nomansland are of a recent period comes from the historical material that I have discussed. This material undoubtedly shows that paintings were made in the Nomansland region by Lindiso (and probably other San) during the first two decades of the twentieth century; the evidence also shows that Lindiso’s father (the husband of Mamxabela) was making art in the Nomansland region during the last half of the nineteenth century and, finally, Silayi’s testimony shows that the San were painting images in the Nomansland region in 1850. We can thus identify three generations of San artists in the Nomansland area, who painted over a period of approximately seventy years during the nineteenth and twentieth centuries. It is clear from the discussion of the historical material and the description of the rock art that Nomansland offers substantial material from which to write a postcolonial past. Indeed, there are few other places in southern Africa and the rest of the world where people, places and pictures can be tied to closely together. Whereas the Subaltern Studies Collective had to recreate a subaltern perspective by reading between the lines of colonial texts, in Nomansland, the quantity and diversity of ethnographic, oral and rock art material allow us to consider the historical processes affecting the area from a San perspective. Intertwining these diverse sources of 52 evidence is not a straightforward task; the historical material is sometimes contradictory and the rock art is not a simple narrative. Indeed, a number of researchers have tried to combine the art and historical material inn order to write a history for the San. These efforts have varying degrees of success and, in the next chapter, I discuss the principal attempts to link San rock art to history. CHAPTER 2 SAN ROCK ART AND HISTORY The greatest challenge to southern African rock art research in the past 25 years has been the integration of the imagery into the construction of San history. Although written from a position independent of postcolonial discourse, the efforts over the last 25 years have considerable pertinence to that discussion because, more so than almost any other indigenous archive, the densely communicative nature of rock art holds the potential for interpreting perspectives of the colonized. In order to unlock the full potential of the rock art to provide these perspectives, it is necessary to see how and why researchers have integrated the art into the writing of history and to establish the contributions and shortcomings of previous efforts. uninformed colonial approach, marred by its arrogant guesswork and racist underpinnings, came in the form of efforts to give the study of rock art ‘scientific’ credibility through the introduction of quantitative analysis (Lewis-Williams 1972, 1974, Maggs 1967, Vinnicombe 1967, 1967). In turn, numerical studies were abandoned when it became clear that they could not lead to the determination of ‘meaning’; sophisticated ethnographic-centred interpretations then succeeded in the mid-1970s, marking, what is widely regarded as, the turning point in southern African rock art research. In particular, two texts, People of the Eland: Rock Paintings of the Drakensberg Bushmen as a Reflection of their Life and Thought by Patricia Vinnicombe (1976) and Believing and Seeing: Symbolic Meanings in Southern San Rock Paintings by David Lewis-Williams (1981), are held to be the seminal works in the field. Importantly, both works focused on images from fieldwork areas in the south-eastern mountains; Vinnicombe worked just north of the Ugie-Maclear area but partly within Nomansland while Lewis-Williams, in addition to working farther north, analysed images in an area to the east but contiguous with Nomansland. Often spoken of in the same breath, Vinnicombe’s and Lewis-Williams’ works are very different in important ways. Understanding their differences leads to a better appreciation of the dichotomies that still face southern African San rock art research. Although chronologically later than Vinnicombe’s book, I consider Lewis-Williams’ work first because it is the approach pioneered by him that has been most influential. The difficulties of integrating rock art into San history may be traced back to the period between 1967 and 1977. Although the pivotal concerns of southern African San rock art research existed in various forms within the discipline previously, it was during that ten year period that they were given new definition and by force of the personalities who established them, new distinction and impetus. By the end of that decadelong period, southern African rock art research had undergone, in the Kuhnian (1962) sense of the phrase, a paradigm shift and the foundations for further research had been firmly established. Although diverse, the research efforts following the 10-year period may be divided into three main approaches. In this chapter, I revisit the so-called revolutionary period. I show how San rock art research developed and how several dichotomies emerged from that period. It is with these dichotomies that subsequent research has wrestled. The Hermeneutic Approach Believing and Seeing was initially produced as a doctoral The general trajectory of that ten-year period is well- dissertation within the Social Anthropology Departknown (see Lewis-Williams 1995, see Lewis-Williams ment at Natal University in South Africa in 1977; in and Dowson 1994). The reaction to the dominant, 1981 it was published as a book by Academic Press. The 53 central arguments in the text drew substantially on the 12, 000 or so pages of /Xam San belief, myth and folklore, collected by Wilhelm Bleek and his sister-in-law, Lucy Lloyd, in the 1870s (the collection is published in Bleek 1931, 1932, 1933, 1933, 1935, 1936, Bleek and Lloyd 1911, Lewis-Williams 2000). Bleek had obtained permission from the authorities at the Cape to have several /Xam San prisoners live with him at his home in Mowbray—including Diä!kwain and /Han≠kasso (who are mentioned in several places throughout the text)—so that he could work on the language. This remarkable ethnographic collection on the /Xam, a southern San people of the northern Cape Colony, had remained virtually obscure until the publication of Believing and Seeing (for discussions on the context, limitations and problems of the Bleek and Lloyd Collection see Deacon and Dowson 1996, Thornton 1983). The interpretative arsenal that the Bleek and Lloyd collection offered was supplemented by ethnographic studies of northern San, particularly !Kung speakers (but also other groups, such as the G/wi), in the Kalahari Desert of Namibia and Botswana; these were undertaken from the 1950s onwards (for an overview see Barnard 1992). Although temporally and spatially removed from the rock art, the demonstrably close similarities between the San of the Kalahari and those of the nineteenth-century /Xam (e.g. Lewis-Williams and Biesele 1978) allowed for the productive deployment of this literature for the interpretation of San rock art. narrative, the images were now seen as a complex system of metaphors and symbols. Believing and Seeing was the first substantial effort to establish the meanings behind the San rock paintings. Lewis-Williams’ work, and that of his colleagues, has been labelled as the trance hypothesis or more commonly, the shamanistic approach or shamanistic school. These labels carry unfortunate and misleading connotations and they do not adequately make a distinction between the conclusions at which Lewis-Williams arrives and the methods and techniques that he uses in his analysis of the rock art to arrive at those conclusions. For this reason, I label the approach started by Lewis-Williams as the ‘hermeneutic approach’. Of course,‘hermeneutic’ refers to different intellectual endeavours in different disciplines. Here, I use the word in its more widely accepted sense, as it pertains particularly to scripture, as the art or science of interpretation. It is fundamentally with how we may interpret San rock art in a manner that is consistent, as far as possible, with the way in which the makers of the images and their communities interpreted them that underpins Believing and Seeing and the subsequent work of Lewis-Williams. Following the publication of Believing and Seeing, San rock art research increasingly focused on the intrepretation of the meanings of the images. Importantly, the theoretical approaches of semiotics and Turnerian symbolic analysis that were employed in Believing and Seeing, were not social theories in the sense that they were models of how collectives of people interacted in space and through time. As we shall see, the employment of All this ethnographic material was structured by two these theoretical approaches allowed for the data—the theoretical approaches in Believing and Seeing—early ap- rock art images—to take primacy in the argument. proaches to semiotics, in particular, the work of Charles Sanders Peirce (1931-35) and Charles Morris (1946, The hermeneutic approach made significant strides 1964) and, more importantly, the symbolic anthropol- throughout the 1980s and 1990s in establishing the symogy of Victor Turner. Turner’s classic works in central bolic associations of particular images. As these efforts Africa (1966, 1967) provided the ideal framework for spread farther afield, it became increasingly clear that the sophisticated interpretation and integration of sym- the original formulation of Believing and Seeing needed to bolism and ritual within San society and then within the be modified. Originally, Lewis-Williams envisaged San rock art. Using Turnerian symbolic analysis, Lewis-Wil- rock art as implicated in a number of rituals in San soliams teased out and elucidated the semantic spectrum ciety—puberty rites, marriage, rain-making and curing of key symbols in San ritual—such as the eland—that or trance dances. However, when it came to illustrating were evident in the art. Instead of being seen as simple the various chapters, he found only a few images that 54 could be argued to illustrate puberty rites or marriage ceremonies while he had too many for the curing dance and rain-making sections. As more and more images were analysed, it was clear that the curing dance and rain-making pervaded the rock art of the south-eastern mountains. Moreover, it became clear that the dance and rain-making were linked in that they both concerned the activities of San ritual-specialists (known as medicinemen, shamans or owners-of-potency in the literature). Images far beyond the original study areas in the southeastern mountains were now interpreted in terms of the activities of San ritual-specialists. Two important developments fuelled the continued productivity of this approach. First, the ongoing ethnographic research in the Kalahari Desert continued to bring to light useful information that, in turn, helped researchers to understand the symbolic associations of the images (e.g. Lewis-Williams et al. 2000). Although the various Kalahari groups had no tradition of rock art themselves and in spite of the linguistic diversity amongst groups, they all practised the great circular dance so important to their ritual and symbolic lives. New insights into the dance supplemented the Bleek and Lloyd material and together provided a fruitful ethnographic interpretative framework. Second, in the early 1980s, the ethnographic material was supplemented by anthropological and medical literature on the neuropsychology of altered states of consciousness, which I will discuss in detail shortly. Armed with the dual ethnographic-neuropsychological tools, many images were now interpreted. Anthropomorphic figures that bend forwards, sometimes with arms back, now came to be understood as people adopting postures that were important to the dance. People in handto-nose postures or that were depicted with nasal bleeding came to be interpreted as participants of a dance. Paintings of eland were now understood to be the most powerful symbols of supernatural potency, a substance that pervades the San universe and that is essential to the success of the dance. Therianthropic images were interpreted as San ritual-specialists who were in the process of transforming or had already transformed into animal shape in order to travel to the spirit world. In addition to these, a host of other images came to be seen as being fundamentally linked to the dance and the activities of the ritual-specialists at the ceremony. Indeed, so detailed were some of the images of beings and animals in the spirit world, that it became accepted by those following Lewis-Williams that the paintings were done by the ritual-specialists themselves or under their explicit guidance because only they could have seen such creatures. By the end of the 1980s, the efforts of Lewis-Williams as well as those of his colleagues had established—what was now called—the shamanistic context of the rock art of the Drakensberg (Lewis-Williams 1990, 2003, Lewis-Williams and Dowson 1989, 1992) and many other parts of southern Africa (Dowson 1992, Garlake 1987, 1987, 1990, Huffman 1983, Yates, Golson, and Hall 1985, Yates, Parkington, and Manhire 1990) and even today this approach continues to yield new insights into the images (Lewis-Williams et al. 2000). These interpretative efforts, however, have not been without criticism. Although there have been many debates over the last two and half decades, much of the criticism of the hermeneutic approach is unfortunately marred by blatant factual errors, denialism and personal vitriolic attack. Indeed, it is now the case that edited volumes appear, whose entire self-professed purpose is to demolish the work of Lewis-Williams. In this atmosphere of frenzied hysteria, it is very difficult to sort nonsense from valid and insightful criticism. A further complication is Lewis-Williams’ extension of his research to the Upper Palaeolithic rock art of Western Europe and the widespread adoption of his ideas by other researchers. At times, it is difficult to ascertain whether criticism is directed at his southern African research or at his European efforts or at the work of those who follow him. Rather than wade through the confused morass of criticism and counter-criticism that can be read elsewhere (for an overview of global debate on shamanism in rock art see Pearson 2002), I consider two aspects that dominate present critical discussion of the hermeneutic approach and that are important because they impact directly on discussion in ensuing chapters. The first of these revolves around the validity of the word ‘shaman’ 55 in San communities and the second concerns the applicability of neurological research in interpreting San rock art. Following this, I will outline and discuss three aspects of the hermeneutic approach that give it its powerful heuristic potential but that also limit the approach. The Use of Shamanism A persistent criticism of rock art research conducted under the influence of Lewis-Williams concerns the words ‘shaman’ and ‘shamanism’. In extreme form, some critics go so far as to claim that there is no evidence in the extensive ethnographic record of San communities for shamanism (Hromnik 1991). These claims, however, are usually made from a position of little knowledge of San ethnography. Indeed, it was anthropologists and not rock art researchers who first used the word ‘shaman’ to describe San ritual-specialists. As far back as 1975, Mathias Guenther (1975/1976), an anthropologist with extensive experience in the Kalahari Desert, used the word ‘shaman’ to describe the activities of ritual-specialists at curing dances at Ghanzi. Importantly, Guenther uses the word when referring to the ritual-specialists of a number of San groups. Other writers working with San (Hewitt 1986) as well as researchers undertaking crosscultural studies (Halifax 1980, 1982, Noll 1983, 1985, Winkelman 1989) have used the word ‘shaman’ to refer to San ritual practitioners (cf. Lewis-Williams 1992). It was only in 1987, over a decade after Guenther used the word, that San rock art research first adopted ‘shaman’. Previously the word ‘medicine-man’ had been favoured but was abandoned because it was felt that it carried pejorative connotations (Lewis-Williams Pers. Comm.). More serious criticism than denialist claims that the San have no shamans comes from anthropologically-influenced literature. Some critics have argued that the words ‘shaman’ and ‘shamanism’ are used too freely (e.g. Kehoe 2002). These critics contend that the word ‘shaman’ refers to the specific cultural context of certain Siberian peoples, particularly the Tungus. Although the word has come to be used for communities in North America, South America, Oceania and Africa, critics argue that this widespread use of the word has more to do with the 56 influence of writers such as Mircea Eliade (1972), who created a view of indigenous people as primitive. By continuing to use ‘shaman’, so the argument goes, one continues to support this primitivist position. Debates concerning the definition of shamanism are not peculiar to rock art research and as far back as 1853, about the time when Silayi’s sojourn with Nqabayo’s group was coming to an end, scholars in Europe were contesting accepted ways of understanding Siberian shamanism (Price 2001: 4). Most recently, Roberte Hamayon (1998) has argued that too much emphasis has been placed on defining shamanism in terms of ‘trance’ and ‘ecstasy’ and that what is needed is a definition that emphasizes the social situation of shamans. While this intention is admirable, Hamayon does not quite succeed in providing such a definition herself. While at one level the discussions around the use of the terms ‘shaman’ and ‘shamanism’ in rock art research and elsewhere are semantic issues, at another level they are also about how we understand the widely variant practices to which the terms refer. Ideally, it is accepted anthropological practice to use the terminology that various communities use themselves. The various phrases used by San throughout southern Africa all refer in some way to ‘owner-of-potency/energy/power’. Amongst the !Kung of Namibia and Botswana, the word is n/omkxaosi while amongst the /Xam the word is !gi:xa (Lewis-Williams 1992). Although in the remaining 140 words of the !Gã !ne language there is no mention of an appropriate term, it is probable that a phrase similar to !gi:xa was used as words that are cognates appear in the Cape Nguni languages and Khoekhoen languages that are found around Nomansland; the Cape Nguni languages are known to have been heavily influenced by Khoesan languages. While using the indigenous terms may be preferable, the clicks do not roll off Western tongues easily and by avoiding the use of ‘shaman’, one may obscure what are real similarities between different peoples. For these reasons, I retain the words ‘shaman’ and ‘shamanism’ throughout the thesis, but I use the words interchangeably with ritual-specialist, owner-of-potency/potencyowner and, less frequently, I use n/omkxaosi or !gi:xa. The Applicability of Neurological Research While some writers deny that the San have shamans, others claim that there is no evidence for trance, ecstasy or altered states amongst the San (Hromnik 1991). While these claims are easy enough to refute through reference to ethnographic investigations that identify such states amongst the San (e.g. Howell 1979: 51, Katz 1982, Katz and Biesele 1986: 221, Lee 1967, 1968, 1984: 103, Marshall 1969: 349, Shostak 1981: 10), neurological research remains one of the most controversial issues in rock art research and there are those who would gladly see its demise. Such a radical view, however, ignores physiological reality and is debilitating. In order to understand why it occupies this controversial position, we need to consider briefly the historical trajectory of neurological studies and how they came to be incorporated into rock art research (for more comprehensive overviews see Blundell 1998, Lewis-Williams 2002). Within rock art research the word neuropsychology has come to be used to refer to wide ranging studies that concern various states of human consciousness in such diverse fields of study as medicine, psychology and anthropology. Although there were earlier studies on various states of consciousness, it is only in the twentieth century that significant advances were made. The first major insights came from the work of Heinrich Klüver (1966), who recorded the verbal accounts of people under the influence of the hallucinogenic substance mescal, extracted from the cactus Echinocactus Williamsii. Although most of his research was conducted during the 1920s, his major study was only published much later, in 1942. Later on, Max Knoll experimented with electro-shock treatment on volunteers (Knoll and Kugler 1959, Knoll et al. 1963). As Klüver had done, Knoll meticulously documented the experiences of his patients. Joseph Eichmeier and Oskar Höfer (1974), two of Knoll’s students, continued his pioneering work in electro-shock treatment research. In much of this early work, the lack of sophisticated equipment for imaging neurological activity made researchers reliant on the testimonies of patients or volunteers. As such, these works are largely phenom- enological in that they document the bodily experiences of people while in altered states of consciousness. Early studies of altered states of consciousness, such as those conducted on electro-shock therapy patients, were sporadic and largely undertaken in isolation from one another. In the late 1960s and early 1970s, this situation changed as research into altered states gained rapid momentum with the invention and subsequent widespread use, of LSD in Western societies. Researchers such as Gerald Oster (1970) and Ronald Siegal (Siegel 1977, Siegel and Jarvik 1975, Siegel and West 1975), for example, obtained extensive accounts, including drawings, from people taking the psychoactive substance. At about the same time that that this work was being undertaken, Geraldo Reichel-Dolmatoff (1969, 1972, 1978, 1978) was conducting ethnographic work on the Tukano and other Native American groups in the Amazon Basin, who ingested powerful hallucinogenic plants such as Banisteriopsis caapi. While the work by medical researchers on Western subjects and anthropologists on small-scale communities still emphasized the experience of altered states, other researchers were studying the effects of altered states, particularly those induced by the ingestion of LSD, on the construction of individual identity (Barr et al. 1972, Grof 1979). It is perhaps this area of investigation into identity-formation that, most accurately, can be called neuropsychological research. More recently, with advances in medical technology, research has concentrated on mapping activity areas of the brain during various stages of consciousness. Richard Cytowic (2000), a medical researcher, for example, has mapped activity areas of the brain during synesthetic episodes. Synesthesia, a complex process by which the brain confuses one sense with another, may be induced in altered states but the condition also occurs naturally (ibid.: 6). Natural synesthetes regularly confuse one of their senses with another, for example, taste with touch. Typically the experience is unidirectional; taste sensations may be construed as touch by tactile sensations are not confused with taste. Working with a synesthetic volunteer, Cytowic injected the pa57 tient’s head with Xenon gas, which is radioactive but biologically inert. Xenon gas clusters around electrical current in the brain and neurological activity can thus be mapped. During periods of synesthesia, Cytowic found that the gas clustered in the limbic system of the brain. The limbic system lies beneath the cortex and is thus an evolutionary older part of the brain. Moreover, the limbic system is regarded as the seat of emotion while the cortex is regarded as the base of analytical reason. Other researchers, such as Eugene d’ Aquili and Andrew Newberg (1978, 1986, 1993, 1993, 1999, 1999), have used similar neurological mapping techniques during investigations of the experiences of religious mystics. Drawing on laboratory work, these two have surveyed religious mystics through time and across cultural differences. They conclude that many of the famous mystics through time have suffered from pathological conditions that have produced altered states. As such they argue that ‘God’ is wired into the human brain. Neurological mapping such as that carried out by Cytowic and the work of d’ Aquili and Newberg have greatly extended the earlier phenomenologically orientated efforts. What makes all this work on altered states—both phenomenological and biological documentation—important is that many researchers working in these fields have pointed to the similarities between drawings made by people of their experiences during altered states and images found in rock art. Long before rock art researchers were aware of this research, neurological researchers were aware of the potential of their work to account for certain rock art images. In particular, the work of Knoll and his successors, Eichmeier and Höfer, is important in this regard. Knoll pointed to the possibility that images in southern African rock art were the product of altered states of consciousness. Eichmeier and Höfer went much further. They called attention to the similarity between the imagery of electro-shock induced altered states and the imagery of many different art traditions throughout the world. In page after page of their mammoth Endogene Bildmuster (1974), they document these similarities in various rock art traditions. The sheer scale of their work, however, serves to undermine their argu58 ment. In many instances, the similarities they point to are superficial and their arguments are often unconvincing. The superficial correlation between the imagery of altered states and that found in various artistic traditions is a problem that plagues rock art research to this day. The reason for this is, in part, the manner in which neurological research was incorporated into rock art studies. It was North American researchers who first laid the foundations for later research. Ken Hedges (1976, 1979, 1982, 1983, 1985, 1992, 1994) and Thomas Blackburn (1977) described certain Californian rock art images as the product of altered states of consciousness. Their pioneering efforts were built upon by David Whitley (e.g. 1987, 1992, 1994, 1998, 2000, 1999) and others (e.g. Boyd 1996, Boyd and Dering 1996, Francis and Loendorf 2001, Loendorf 1994). In other parts of the world, researchers such as Emmanuel Anati (1981) and Robert Bednarik (1984) dabbled briefly with the relation between the imagery of altered states and rock art. In southern African archaeology, the similarities between certain rock paintings and engravings and the imagery of altered states was first noticed by Francis Thackeray and colleagues (Thackeray et al. 1981) in the early 1980s. It was this paper that alerted David LewisWilliams and his colleague, Thomas Dowson, to the potential of neurological research for rock art studies. Consuming the vast and diverse literature, Lewis-Williams and Dowson (1988) developed what is known as the neuropsychological model. It was only after the publication of this model in 1988 that neuropsychology became such a controversial issue within rock art research (for debates on the controversy see Bahn 1996, 1997, Bednarik 1990, Lewis-Williams and Clottes 1998, Lewis-Williams and Dowson 1990). That it should arouse such controversy at this late stage, after decades of medical and anthropological research suggesting its usefulness, is remarkable and requires discussion. The so-called neuropsychological model put into a coherent form the numerous, diverse accounts of people undergoing altered states of consciousness. As the model was largely constructed from people’s descrip- tions of their experiences in altered states, it is more of a phenomenological than a psychological model. Lewis-Williams and Dowson observed that people’s descriptions showed remarkable similarity cross-culturally and seemingly independently of the method of induction used. This similarity lay not only in the types of experiences but also in their progression. Lewis-Williams and Dowson described this progression as having three stages: initially people undergoing altered states see a limited range of luminescent geometric images that move about in the visual field in complex ways. These images are often described by medical researchers as phosphenes or endogenous percepts, but LewisWilliams and Dowson labelled them ‘entoptic’, meaning within vision, because they appeared to be experienced even if subjects closed their eyes. The ability to experience these entopic images independently of an external light source meant that their origin was within the optic system itself. As they are built into the optic system, entoptics are experienced cross-culturally. With time, subjects begin to interpret these images from cultural and individual backgrounds. A grid pattern, for example, seen in the first or entoptic stage may be interpreted by a San person as a honeycomb while someone from a Western perspective may interpret the same grid pattern as a bank of television screens. Similarly, someone experiencing a particular entoptic form will construe the image differently if they are angry or sexually aroused (Clottes and Lewis-Williams 1998). After experiencing this construal stage for some time, people undergoing altered states, report a sensation of passing or falling through a tunnel or vortex that is sometimes characterized by a light at one end. Following this sensation, people have complex experiences such as aural hallucinations and somatic sensations such as synesthesia (confusion of senses), polymelia (extra digits) and attenuation (a feeling of being stretched out). Lewis-Williams and Dowson labelled this third phase as the iconic stage. While the model was based on descriptions of people’s experiences, the commonality of those experiences seemed to point to universal physiological structures underlying these experiences. It should be stressed that this three-stage description is a model and it is thus a useful but imperfect way of discussing very diverse and complex neurological processes. Well aware of the limitations of models, LewisWilliams and Dowson attached some caveats. First, the three stages are not ineluctable; although it appears that people do pass through an altered state in this order, it is not clear that they always experience entoptics before moving into the more complex stages or whether some people can pass directly to the second and third stage. Second, the three stages or not discreet packages but are best understood as a building up of different kinds of experiences: entoptics thus occur through all three stages and construal occurs in the iconic phase as well. The three-stage neuropsychological model has found widespread acceptance in rock art research in different parts of the world but it has also attracted criticism. The three most substantial criticisms that have been levelled at the model concern the physiology of altered states, the predictive value of the model and its apparent deterministic quality. I consider each in turn. Researchers sometimes argue that one of the limitations of neuropsychological research is that too little is known at present about where and how altered states occur in the brain. This limited knowledge of the physiology is a problem when applying the model to the distant past. Jeremy Dronfield (1996), for example, points out that although the people of the distant past were also Homo sapiens sapiens, it is likely that there were slight differences in the morphology of the human brain. Research on feline brains, for example, shows that there are differences between the structures of domestic cats and those of the large wild carnivores even though they are only separated by a few thousand years. If similar differences existed in the human brain in the distant past, then the experiences of altered states would have been ‘qualitatively’ different to the sorts of experiences collected by researchers in the twentieth century. This criticism is not entirely convincing for two reasons. First, altered states appear to be a product of the mammalian and not just of the human brain. As far as can be told, all mammals—except for the spiny anteater 59 (Echidna tachyglossus/ Echidna zyglossus) (Cytowic 2000: 60)—experience altered states, ranging from dreams through to animals, such as deer and bear, that actively seek out psychoactive substances, consume them and then exhibit behaviour similar to that of humans experiencing altered states. If such similar states can be observed in mammals as a whole with such enormous time depths underlying their different evolutionary trajectories then it seems reasonable that people in the distant past would have exhibited similar reactions in altered states. Second, what is known about the physiology of such states suggests that they have a long history in the human brain. The fact that Cytowic’s work shows that synesthesia and thus altered states of consciousness, have primacy in the limbic system suggests that they are a very ancient element of the human brain. Together these two points suggest that the degree to which altered states were ‘qualitatively’ different—a vague phrase in any event—in the distant past is probably negligible. The model was originally intended as a tool to argue for the origin in altered states of certain geometric images found in the European Upper Palaefolithic. Lewis-Williams and Dowson thus demonstrated that two other rock art traditions—southern African San rock art and Coso Range Native American rock art—were known ethnographically to be shamanic and implicated in altered states. It could be argued, convincingly, that certain images—particularly geometrics—in both those traditions were thus derived from altered states. Formally similar images occurred in the Upper Palaeolithic but they lacked ethnographic context. By extension, then, the geometric images in the Upper Palaeolithic suggested that the rock art of that period was also implicated in shamanism. Today, the argument for the shamanistic qualities of the Upper Palaeolithic has advanced substantially and depends not only on these geometric forms but on other images, such as therianthropic figures, in addition to other contextual evidence (see Clottes and Lewis-Williams 1998, Lewis-Williams 2002). widespread adoption of the ideas by researchers working in other parts of the world. Many of these researchers used the model to argue that geometric images in rock art traditions found throughout the world were entoptic and that those arts thus had to be shamanic. A great deal of this work—but by no means all—exhibits a flippant quality in that many writers play a kind of ‘neuropsychological snap’: whenever a geometric image is found, it has to be entoptic and entoptics are the uncritically treated as evidence for shamanism. In making superficial correlations, researchers revert to the weak argumentation of Eichmeier and Höfer and, in so doing, undermine more serious efforts to demonstrate the origin in altered states of specific images found in certain rock art traditions. The obsession with entoptic images has done much to harm the use of neurological research in rock art studies. This is a pity because entoptics are the least interesting and least informative aspect of the model. Far more revealing are the complex experiences of the second and third stages that often go beyond the visual (Lewis-Williams and Dowson 1990, Ouzman 2001). Whether a particular cluster of dots, parallel lines or zigzags is entoptic or not is, at the end of the day, not particularly interesting. Those particular dots, parallel lines and zigzags had meanings for the people who made them and it is those meanings that are important. While neuropsychology can provide insight into the origin of those images it cannot normally provide those meanings; meaning must come from the ethnographic or other contextual evidence. The third and final criticism that I consider here also stems from the overemphasis on the first or entoptic stage of the model by researchers. Some critics argue that the model removes agency and self-determination from those who made the art. In effect, they say, it reduces people to automatons who are slaves to the physical processes of their brains. As the model was originally formulated, culture and agency only come into play in the second and third stages and the emphasis on the first stage could lead some critical observers Unfortunately, an unexpected consequence of the de- to the conclusion that the model is thus deterministic. velopment of the neuropsychological model was the The tight linking of social experiences to physiology in 60 research on shamanism and altered states reminds some researchers of socio-biology, with all its attendant social evolutionist arguments and the potential of those arguments to justify the entrenchment of social inequalities. Owing to this criticism, in part, researchers who use the model constantly point out that it does not deny agency or culture. Jeremy Dronfield (1996), for example, argues that even though people in altered states experience a range of entoptic phenomena, they choose which images to depict and those choices are culturally and individually informed. Lewis-Williams (1997) himself has argued increasingly for a view that treats altered states of consciousness as a manipulable resource that is both culturally and individually specific. Not only does cultural and individual background affect the choice of images to depict but they also affect the range of images seen during altered states. Indeed, culture mediates the entire neuropsychological model, from the culturally informed technique of induction to the kind of visual and physiological phenomena experienced by those in altered states to the choice of which of those experiences is regarded as important enough to play a role in the post-ecstatic socio-cultural milieu. In addition to the cultural situation of the model, research, such as that of Cytowic’s, points strongly to the primacy of individual emotion over rational thought in altered states. Far from being automatons, then, people in altered states have a significant stake in what kinds of experiences they seek and how they want those experiences to play out. neutic approach has been so successful is because it ignored the dating of southern African rock art. There are two reasons for this: first, the art is very difficult to date and both the small number of existing dates and their poor resolution mean that they are not good enough for purposes of understanding detailed change in the art in a specific location. Second, coming as it did, from a social anthropological background and not an archaeological one, meant that the kind of culture change that archaeologists working in other areas of southern Africa were interested in was not a major concern for those following the hermeneutic approach. The suspension of chronology in hermeneutic efforts led to a broad-based interpretative framework that in its most fundamental formulation held to the view that, given the sources available to us, we can say with varying confidence—depending on the particular image and available ethnographic and neurological evidence—that at some point in time and for some San people, at least, the image carried this or/and that meaning. This is not to say that an image had no other meanings or that those meanings remained static through time; what those meanings are and how they changed, however, need to be demonstrated and not merely assumed; otherwise efforts at understanding the art revert back to a position of mere assertion. It is important to keep in mind that while the hermeneutic approach considers chronology in only the most cursory terms, it does not dismiss the idea of change in the art. Three Limitations of the Hermeneutic Approach While the criticism levelled against the use of shamanism and neurological models in rock art research are pertinent to any study that wishes to determine the meanings of San rock art images, I now turn my attention to three aspects of the hermeneutic approach that have, paradoxically, allowed for the success of the approach but have also placed limitations on the ability of the approach to adequately consider the relationship of the images to history. These three characteristics concern chronology, regionality and a focus on the image. Closely related to the suspension of chronology in the hermeneutic approach is the treatment of space as uniform. Moving out from the original study area, proponents of the hermeneutic approach suggested a similar interpretative framework for other rock art regions in southern Africa. Selected images from diverse and widely separated areas were explained as shamanistic, using the well-established tools of ethnography and neuropsychology. Of course, such efforts, while for the first time allowing researchers to see meaningful similarities in the art across space, tended to mask variation and It is sometimes said that one of the reasons the herme- divergence. Regional differences were played down in 61 favour of a broad-based general meaning. The suspension of chronology and the uniformitarian treatment of space made the hermeneutic approach somewhat antiarchaeological. After all, southern African archaeology was—and still is—concerned with change over time and variation across space. Indeed, prior to the ten-year revolutionary period, these concerns were fundamental to southern African rock art research, which was very much an enterprise undertaken by archaeologists. Both the suspension of chronology and the treatment of space as uniformitarian stemmed from the hermeneutic focus on imagery and its emphasis on problem-orientated research. Growing as it did out of the reaction to the naïve empiricism of quantitative studies, the hermeneutic approach was initially influenced by Karl Popper’s philosophy of conjecture and refutation (e.g. Lewis-Williams 1984). As such, the hermeneutic approach was problem orientated: specific images required interpretation through the collation, interpretation and application of ethnographic data. The social production and consumption of the art was initially of less importance to this approach. While in one sense these three characteristics are weaknesses, they are also strengths. Suspending time and space allowed the hermeneutic approach to establish a broadbased interpretative framework for the art; it provided the requisite tools for the translation of a previously opaque ‘language’. In Ray Inskeep’s (1971: 101) wellknown formulation, it was this approach that allowed researchers to move from learning about the art to learning from the art; the stage had been set for research that was concerned with the changing social role of the art. The Social Approach The difficulties of establishing the social role of the art, however, are apparent in the second seminal text to emerge from the mid-1970s—People of the Eland. Some eight years in the making, the scope of People of the Eland was epic. The first part of the text draws extensively on colonial accounts, particularly those in 62 the KwaZulu-Natal archives, in order to write a history of the Eastern San. The San rock art imagery that is Vinnicombe’s data comes from Nomansland, but slightly north of the area occupied by Nqabayo. The area is well-known, thanks to the efforts of Vinnicombe, for its representations of colonial settlers, soldiers, horses, guns and military encounters—what we may call ‘contact’ images. Importantly, although Vinnicombe is cautious in her statements, she often implies that the imagery possibly relates to actual events. In later chapters of her book, Vinnicombe discusses images that are seemingly non-contact and more ‘traditional’ such as depictions of antelope and other animals. To interpret the imagery discussed in these chapters, she draws on the ethnography of both the Kalahari Desert as well as the published Bleek material on the /Xam. As Lewis-Williams does, Vinnicombe also makes extensive use of the well-known testimony of Qing, a San from the south-eastern mountains, who commented on paintings to Joseph Millerd Orpen in the early 1870s (Orpen 1874). In contrast to Lewis-Williams, however, Vinnicombe made very little use of the extensive unpublished Bleek material. In her final chapter, Vinnicombe brings together all the elements that she has discussed in the second part of her work under a theoretical umbrella. This umbrella comprises elements of—as in Lewis-Williams’ case—Victor Turner’s ideas on symbolism and ritual. In addition, presumably through her exposure to Edmund Leach (p. xvii) she drew on Lévi-Straussian concepts of metaphor, but most importantly she was influenced by the Functionlaist theory of Alfred Reginald Radcliffe-Brown. Drawing on this theoretical cocktail, Vinnicombe emphasized that the art must be understood as part of the broader social and cognitive system of the San: eland, for example, were symbols of the San collective. It is in this last chapter that Vinnicombe introduces the notion that some of the ubiquitous eland imagery was produced for psychological and social harmony—an argument in accordance with a Functionlaist position. She argued: Through the act of painting and re-painting, eland for instance, the mental conflict involved in destroying a creature that was prized and loved by their deity—of killing in order to live yet at the same time incurring the displeasure of their Creator— was, I believe, ritually symbolised and resolved. The emotional catharsis experienced by Bushmen after participation in the ritual Medicine Dance has been stressed by all authors on the subject and the fact that paintings were emotionally significant to the Bushmen was poignantly demonstrated by Stow a century ago. (Vinnicombe 1976: 350) Crucially, in tying matters together, Vinnicombe does not refer again to the images showing ‘contact’ between San and colonists, pointing to the difficulty of incorporating those images into an ethnographic-based interpretation. Three important dichotomies emerged from the two seminal texts that I have discussed and the related works of Lewis-Williams and Vinnicombe. First, a split emerged between understanding the ‘meaning’ of the images—the hermeneutic approach—and understand- that she was not certain how to achieve this for the ‘contact’ images. The problem is even more acute in the historian John Wright’s (1971) Bushman Raiders of the Natal Drakensberg, written some years prior to People of the Eland and utilizing the same archival sources that Vinnicombe drew upon. Wright illustrates his text with ‘contact’ images but makes little effort to explain these images using a San perspective beyond suggesting that they were possibly depictions of real events. The emergence of these three dichotomies, far from debilitating rock art research, offered new opportunities for ground-breaking research. It soon became clear that the general progression of the history of the San of the south-eastern mountains and Nomansland during the nineteenth century was well understood, albeit from a colonial perspective. Although the hermeneutic approach persevered, the task facing researchers now was the integration of the imagery itself into this historical material; producing a history of the San that not only relied on colonial records but included their art became the Holy Grail of southern African rock art research. The allure of success in achieving this was obvious; rock art research, having relocated from Social Anthropology at the University of the Witwatersrand to the Department of Archaeology would contribute to the understanding of change. In so doing, it would possibly shed its ambiguous position within the broader field of archaeology where it was sometimes seen as less serious than main-stream, excavation archaeology. Moreover, this could be realized in a manner that was cognitive as opposed to the technological approaches that characterised most archaeology until the early-1980s and which I discussed in the Introduction. Throughout the course of the 1980s and 1990s, then, efforts were made to integrate the art into what was known about the development of San history in the south-eastern mountains. These efforts included marxist, structurationist and interactionist approaches. ing the social production and consumption of the art—the social approach. Vinnicombe (1972, 1972) herself raised this distinction some years before People of the Eland in her perceptive critique of the art for art sake approach to San rock art. Yet, this distinction, as Vinnicombe realized, was only an analytical one. In the world of the San artists, meaning and social production and consumption would not have been easily distinguishable. Second, a distinction was made between ‘contact’ imagery and seemingly ‘traditional’ subject matter in the art. The ‘traditional’ subject matter was to be understood by reference to ethnography and neuropsychology. The ‘contact’ images proved more difficult to understand than traditional ones; at best, they possibly represented actual events. This difficulty gave rise to a third dichotomy: efforts to write a history for the San were largely written from the perspective of colonial records and did not really succeed in treating the art as evidence for the point of view of the colonized. Although Vinnicombe was very much aware of Marxist Approaches the need to see the images from a San perspective, the As research emerged out of the 1970s, it became apuncomfortable division of People of the Eland showed parent that the integration of rock art into the produc63 tion of San history required that the ‘contact’ images be understood in terms of hermeneutic advances already made and still being made. Quite clearly, a simple correlation between the paintings and the actual events they were based upon could not be drawn and the images had to be understood as part of the broader shamanistic understanding of the art. It was thus clear that a strong social theoretical framework was also a requisite. The structural-functionalism employed by Vinnicombe was no longer adequate because of the teleological nature of the argument in addition to the other well-known flaws of such approaches (see Cohen 1968 for criticisms of structural-functionalism). With Vinnicombe’s departure to Australia, it was Lewis-Williams who made the next important contribution to the social approach. Reacting against the earlier innatist-aesthetic (also known as the art for art’s sake approach) approach, which placed the production and consumption of the art at an individual’s psychological level, thereby denying the possibility of understanding the social context of the art, as well as the simplistic Functionlaist argument of the sympathetic magic hypothesis, he introduced a structural-marxist model into southern African San rock art studies. Drawing on the work of Maurice Godelier (1975, 1977, 1978) and Jonathan Friedman (1978), in particular, his focus was very much on “the articulation between the art and the economic base of the San social formation” (Lewis-Williams 1982: 431). applied to San ethnography and only then to the art. The argument is successful when the structural-marxist model is applied to San hunter-gatherer social organisation but it becomes awkward when the art has to be included in the model. The imagery is treated largely as a passive mnemonic or didactic aid. For example, Lewis-Williams argues that the art “provided a permanent backdrop to daily social relationships pointing to the social and economic order which the medicine men worked to maintain” (ibid.: 438). At best, the art actively reduced the fears of novice shamans who had not seen the fearsome spirit world before by illustrating what that world looked like (ibid.: 431). The treatment of material culture as passive and reflective, rather than constitutive and active is a particular problem with some marxist approaches (Davison 1991). While this is not a novel criticism of the structural-marxist approach in San rock art research, the contradictory employment of the model is a more substantive problem. Structural-marxism is a model, like most marxist forms of thinking, which is particularly pertinent to understanding radical social change. Yet, Lewis-Williams used the model to show how the symbolic labour of San shamans “operated on the conditions of social cooperation and on the renewal of nature” (LewisWilliams 1982: 433). As such, the model was used to show how the social structure was maintained and reproduced in the face of contradictions in San society. The nexus of this articulation was the shaman’s symbolic labour. In particular, Lewis-Williams was concerned with the social role of such activities as curing the sick and the parallel task of warding off sickness-carrying spirits-of-the-dead, controlling the movements of the game, controlling rain and going on out-of-body travel to visit camps with access to resources. Lewis-Williams showed how these four acts of symbolic labour operated on the social relations of the production process (ibid.). By extension, the art which depicted these acts of symbolic labour also operated on the social relations of production. This is very close to Victor Turner’s particular brand of Functionlaist and marxist analysis. Influenced by Max Gluckman’s (under whom he studied) school of situational analysis, Turner accepted that the normal state of affairs in society was one of contradiction and conflict. His interest was in how ritual operated to reproduce society (as well as reproduce the contradictions) in the face of these contradictions. It is clear that Turner’s influence extended to this first marxist-inspired intervention into San rock art studies. The construction of his argument here is important. Crucially, Lewis-Williams did not address the ‘contact’ It begins with a theoretical framework, which is then images in this paper. The explanation for this was: 64 It is true that some paintings apparently portray historical events such as cattle raids (Vinnicombe 1976:44-48) and so seem to count against my argument that many of the artists were medicine men who painted depictions associated with trance performance, but these “narrative” scenes constitute only a small proportion of the art and it is not with them that I am now concerned. The emphasis of the art lies rather on antelope and ritual activities (Lewis-Williams 1982: 434). Despite the reaction against Functionlaist approaches, the first structural-marxist intervention into southern African San rock art was not that different from Functionlaist approaches (Lewis-Williams, Pers. Comm.). Nevertheless, it was the first serious attempt at explicitly deploying a social theoretical framework in the context of southern African San rock art and also the first such effort in southern African archaeology (see Barham 1992 for a discussion on the successes and failures of marxist approaches in southern African archaeology). It was up to one of Lewis-Williams’ students, Colin Campbell, to use the structural-marxist approach to understand change in southern African San rock art. In a superb but little known Masters Dissertation, Campbell applied the model to the ‘contact’ images of the Drakensberg (Campbell 1987). Whereas the tacit distinction made by Vinnicombe between images of colonial period subject matter and traditional images had given rise to the distinction between ‘contact’ and pre-contact images, Campbell now extended the category of ‘contact’ images to include depictions showing interaction not only between the colonists and San but also between the San and Cape Nguni-speaking peoples. Instead of the four areas of symbolic labour that LewisWilliams looked at, Campbell focused only on three areas—what he called the San symbolic labour triad of rain-making, control of animals and healing (1987: 37). Visitation to other, distant camps on out-of-body journeys is not as significant for Campbell as it is for LewisWilliams. Nevertheless, like Lewis-Williams, Campbell considers how shamanic symbolic labour contributed to the maintenance and reproduction of San society. Yet he went further and showed how, through interaction with Cape Nguni-speakers and colonists, the symbolic labour provided the basis for change in San communities in the south-eastern mountains. Drawing partially on the archival material that Vinnicombe and Wright had brought to light coupled with an analogous situation amongst the various San communities of Ghanzi during the 1970s, Campbell argued for the advent of: a new element in the relations of production: access to certain resources and distribution of the product came to be controlled by the shamans. I refer to this new element in the relations of production as the shamanistic relation of production (original emphasis) (ibid.: 46). What both the Ghanzi and archival material showed was that San interaction with Cape Nguni-speakers centred on symbolic labour: San ritual-specialists cured the sick, controlled the movement of game and, most importantly, controlled rain. In return, they received payment in the form of material goods such as cattle and other livestock. The San also raided livestock from colonial farmers and this too led to the influx of material items into their communities, particularly horses, which they did not seem to trade with Cape Ngunispeakers. This influx of items into a traditional egalitarian society whose social organization did not really allow for the accumulation of material goods produced a surplus that accumulated largely to the shamans. In turn, according to Campbell, this led to class-distinction amongst the San of the south-eastern mountains. Another salient aspect is Campbell’s argument that symbolic labour was a site of struggle between the San and the Cape Nguni-speaking communities. On the onehand, the San were—and still are in southern Africa— regarded as rainmakers without equal. Their ability to produce rain was an obvious advantage over the agrarian Cape Nguni-speaking groups and gave them a ritual niche that also had political and economic advantages. On the other hand, the Cape Nguni-speaking groups were more numerous and better organized than the San. Initially, San used their ritual position to their advantage 65 but over time they were forced into a more subordinate relationship either through violent means or through economic dependency on Cape Nguni-speakers. As the nineteenth-century progressed and colonial efforts became more successful in stemming San raiding, they were forced into ever-closer interaction with Cape Ngunispeakers. The balance of power shifted to give the Cape Nguni-speakers the upper hand (Campbell 1987: 60). practice theory, particularly Anthony Giddens’ (1979, 1984, 1991) structuration theory. Rising out of dissatisfaction with early social theories but in particular structuralism and marxism, structuration theory attempted to bridge a number of dichotomies between phenomenological approaches and their ‘tyranny of the individual’ and objectivist approaches with their ‘tyranny of society’. Giddens was not alone in these endeavours and his ideas are part of a broader theoCampbell’s argument was a major step forward in the retical interest in practice. Pierre Bourdieu (1977), for ongoing efforts to articulate rock art with other histori- example, as Giddens did, sought ways in which to uncal material. In many ways, his work was a precursor to derstand better the manner in which individuals and Kalahari revisionist efforts to challenge the dominant society interacted with one another. It was, however, discourse of the San as isolated relics from the Stone Giddens’ ideas on agency and the duality of structure Age. Yet, his arguments were also more nuanced than that had the greatest impact in San rock art research. many of the revisionist arguments. Whereas Wilmsen, for example, denies San autonomy, Campbell offers a In a series of papers, Thomas Dowson (1994, 1995, more sophisticated view of the interaction between San 1998, 2000) used structuration theory to understand and Bantu-speakers—one in which the San were not the social change of San in the south-eastern mounsimply passive victims but active agents who resisted tains during the last 2000 years (but see Kinahan 1989: domination. Nevertheless, there are some limitations to 20ff, 1991 for work on the rise of individuals in Nahis arguments. In particular, his definition of ‘contact’ mibian rock art). Structuration theory had a number of images is limiting. His emphasis on images depicting advantages over structural marxism in that it offered horses, cattle, guns and other elements clearly associated a way of understanding change in society as ongoing; with interaction gives the impression that it is only these structural marxism’s heredity was one of social theory images that should be understood as a product of in- designed to understand radical, revolutionary change. teraction. There were other images in the art that lacked The definition of radical, revolutionary change was the obvious elements of ‘contact’ but that were never- not something that structural marxists always defined, theless related to interaction. Indeed, as the revisionist yet, the impression is that revolutionary social change challenge showed, the last 2000 years of San history in takes place in months or, perhaps, a few years at best. southern Africa had to be understood as the product of The ongoing change that the San of the south-eastern interaction between peoples. As discussed in the Intro- mountains experienced in the last 2000 years could duction, much rock art of the south-eastern mountains thus not be considered revolutionary; indeed, even the seems to fall predominantly into this time period and changes they experienced in the nineteenth century clearly most of the images need to be understood as a seemed to take place over a number of decades and product of this interaction. By focusing on images that structural marxism was inadequate as a model to undemonstrated interaction to a Western eye, the broader derstand this more gradual change. It is perhaps this San perspective of colonialism was still lacking; it would inadequacy that lies behind the Functionlaist tendentake another theoretical innovation to appreciate how cies of both Lewis-Williams’ and, to a lesser degree, the colonial process impacted more widely on their art. Campbell’s arguments. Given that there was no clear evidence of radical change in the art itself, there was Structurationist Approaches a need to explain continuity to so some degree. StrucThis innovation appeared in the form of agency and turation and agency thus apparently offered a way of 66 explaining gradual but persistent change in San society. Drawing on the same analogy with the Ghanzi material as Campbell and, influenced by John Kinahan’s (1991) work in Namibia, Dowson argued for a more gradual progression in the changes of the rock art of the south-eastern mountains. Importantly, this progression was not merely established from the ethno-historical sources but the art itself now became evidence in its own right and formed a vital strand in the argument. The imagery itself was no longer treated as merely the passive mnemonic or backdrop of structural-marxist approaches. It was now seen as an active constituent in the negotiation of change in San society. Structuration theory thus seemingly offered a theoretical framework that allowed for more interpretation of a San as opposed to colonial perspective of the historical processes in the south-eastern mountains. Like Campbell, Dowson accepted the historical trajectory of the San in the south-eastern Mountains as laid out by Wright and Vinnicombe. Yet, where Campbell argued for the emergence of a shamanic mode of production, Dowson argued for a more nuanced progression. Importantly, he contended that this progression could be seen in the art itself. The progression was one through three phases: egalitarian, consortium and pre-eminent. Dowson argued that the earliest paintings visible—judging from their position underneath other images—were largely monochromatic, showing very little differentiation in size, colour and detail and referred to a time when San social structure was egalitarian. Later paintings, while still of the same size, showed variation in colour and detail and were made at a time when the hunter-gatherers’ social environment was changing. These paintings show the gradual rise of individuality amongst the San. The latest images were of individuals who were painted different to all other human images in a panel. These figures were often larger and painted in significantly more detail and colour than surrounding images. These final paintings, he labelled as pre-eminent and he argued that they depicted the rise of powerful, individual shamans. Given their features, such as nasalhaemorrhage, therianthropic features, postures such as arms-back or bending forward, these images were clearly identifiable as shamans. As San society entered into ever-increasing interactive relationships with initially Bantu-speakers and then colonists, the egalitarian structure was gradually replaced by one of hierarchy in which the shamans, as Campbell had argued, would have enjoyed the most benefits. Here, the art not only reflected this transition, it was actively used to negotiate it. Shaman-artists manipulated the various rules and resources of painting in order to depict themselves as individuals, initially as small groups or consortia of individuals but later as individual, pre-eminent shamans. An important aspect of Dowson’s work is that he stresses the importance of theory in the writing of San history. Dowson (1993) criticized researchers—most notably Aron Mazel (1992)—who wanted to write San history by considering the archaeological depositional remains but not the rock art. Dowson (1993: 642) claimed that such efforts were chronocentric: Indeed, ‘history’ is a Western construct that privileges events that can, more or less, be pinpointed in time. Mazel takes ‘history’ as a given and therefore writes a Western history (that pinpoints certain kinds of events and ignores or marginalizes others) of colonial times in Southern Africa. What is needed is a new concept of ‘history’ that breaks from emphasising a chronology of only certain kinds of events, one which accepts other evidence and other kinds of statements and constructions. In response, Mazel (1993) argues that until southern African San rock art could be placed into a more secure chronological framework, efforts to use the images to write history are misplaced. Admirably, Mazel (1996) has embarked on a project to obtain direct dates from the rock paintings and so far a handful of images have been dated. While this work is important and provides a very broad chronological framework, the number of dates, their poor resolution and the inescapable doubts about the integrity of the material actually dated do not inspire confidence in many researchers that direct dating is the answer to the problems of integrating rock art into the writing of San history. By contrast, Dowson begins 67 from the assumption that most of the rock art of the south-eastern mountains dates to the last 2000 years. As such, most of it stems from a period of interaction between San and other people. This assumption, together with his use of structuration theory, allows Dowson to avoid writing a history based on a series of absolute dates. His argument for a progression from communal through consortium to pre-eminent paintings is thus careful to avoid implying that the three phases replace one another. Instead, he argues that some San shamanartists would have resisted efforts by those who wanted to establish a hierarchy of ritual-specialists; communal paintings would thus still be made while consortium and pre-eminent images were being produced. It is best therefore to see the progression as the building up over time of different types of paintings rather than three discrete phases. The importance of the phrasing of his argument in this way lies in its avoidance of an absolute chronology and its reliance on a relative one. Dowson’s claims are, in fact, testable because analyses of layers of superimposed rock paintings through Harris matrices, for example, could reveal whether this progressive sequence was present at rock art sites. By phrasing historical arguments in theoretical discourse such as structuration theory, some of the debilitating aspects of a lack of good chronology can be overcome. Nevertheless, historical arguments that draw on rock art would obviously benefit from better absolute dates. At present, however, dating techniques are still not sufficient to be able to assist with working out the chronology of paintings within a single panel. In any single rock painting panel in the Drakensberg, there may be a few hundred images and simply to work out what was painted before or after what (where there is no superimpositioning) would require an accuracy from radiocarbon dates that present standard deviations do not allow. The failure of scientists to come up with a reliable and accurate technique to date southern African rock art should not, however, hold back efforts to use the imagery in the writing of history. As Timothy Yates (1993: 35) points out in a discussion of Scandinavian rock art: 68 …the way forward for rock art analysis is not to address issues of chronology but to theorize the art— a theorization which must extend way beyond the stale discussions of terminology—and study its appearance and meaning in local and regional terms. Despite the advances made by Dowson, there were considerable theoretical and empirical shortcomings in his argument. His use of structuration theory, for example, may be questioned in light of criticisms levelled against Anthony Giddens in recent years. Following the initial enthusiasm with which Giddens’ structuration theory was greeted, criticism of his approach has grown in recent years. Lynn Meskell (Meskell 1999: 25), for example, points out that: The problem with theories like structuration (Giddens 1984), which are widely cited in the archaeological literature, is that they treat society as an ontology which is somehow independent of its members (Meskell 1998: 157). The agency Giddens allows individuals gives them the power of reflexivity, but not of motivation: they are doomed to be perpetrators rather than architects of action. While he recognizes that selves are reflexively made and individuals do contribute to social influences (Giddens 1991: 2-3), this is rarely demonstrated in his over-arching theories. His individuals are overrationalized and over-socialized (Craib 1992). Within Dowson’s work, this problem translates as an overemphasis on constraint. This emphasis is exacerbated by Dowson’s use of images from diverse areas in the south-eastern mountains. As an important contributor to the hermeneutic advances made during the late 1980s and early 1990s, Dowson’s method of fieldwork was to select images from widespread areas in order to show the applicability of the shamanistic interpretation. Consequently, Dowson’s argument creates the impression that San rock art throughout the south-eastern mountains moves slowly and inexorably through a three stage progression. More importantly, the progression that he argues for can rarely be seen within a single rock art panel and, as I have pointed out, he does not provide studies of superimpositioning to demonstrate the sequence. The examples he uses to illustrate his progression also come from different areas with in the southeastern mountains. The communal and consortium images come from the western side of the south-eastern mountain complex, from that area generally referred to as the Malotis and north of the Orange River while his pre-eminent category images are all taken from south of that river and from the south-east and south-central south-eastern mountains. The pre-eminent images seem thus to be restricted to a smaller area within the south-eastern mountain complex and his progression ultimately does not hold up to critical empirical scrutiny. In the folded mountains that run along the southern part of South Africa, farther south and to the west of Nomansland, archaeological evidence suggests that the San were a changing society long before Bantu-speaking people arrived in southern Africa. Some of the richest San burials discovered in all of southern Africa are found in these folded mountains and the areas immediately surrounding them (Wadley 1996). The burials often, but not always, have grave goods; in some cases the bodies appear to have been covered in ochre and ostrich eggshell beads while in other burials warthog tusks are found (Hall 2000, Hall and Binneman 1987). Of particular importance are burials in which painted stone pieces are interred with the body (Binneman and Hall 1993, Lewis-Williams 1984, Rudner 1971). In a recent discovery at Tierkloof on the southern coast, the body of a San person was found buried in the rear of a painted shelter. The arms of the body were folded over a painted slab of stone, as if it was clutching the slab (Binneman 1999, Pearce 2002). This example, together with the many other rich burials from the southern part of South Africa, suggests that, if not hierarchical, the San communities of these areas certainly singled out individuals for special treatment. The dates for most of these burials and painted stones are between 6500 and 2000 years ago (see Pearce 2002: 29), long before Bantu-speaking or Khoekhoen people arrived in South Africa. This is clear evidence that the San identified ‘special’ people within their communities independently of any interaction with other peoples; not only were they capable of change but they were changing long before the arrival of other people. Therefore, the possibility exists that the pre-eminent figures were part of San rock painting in the southern south-eastern mountains long before the arrival of Khoekhoen and Bantuspeakers. Dowson’s argument that the interaction between San and Bantu-speakers led to the rise of San individuality is therefore somewhat misleading, because individuality was certainly present in the San communities along the southern coast before 2000 years ago. A more pressing concern with Dowson’s argument is that it is constructed in such a way as to imply that all change in San social life and art comes about only through interaction with other peoples. The progression from communal through consortium to pre-eminent, according to Dowson, needs to be understood as the result of the shift in economic practices brought about by the arrival of Bantu-speakers and Europeans in South Africa. This problem is not limited to Dowson’s argument but occurs in many efforts to discuss changes in San hunter-gatherer social organization. Indeed, Campbell’s argument is similarly constructed and so are arguments that have been made more recently. Ørnulf Gulbrandsen (1991), for example, in a paper discussing the implications of the revisionist debate for understanding egalitarianism identifies three conditions under which leaders can emerge in San society. These are, first, when circumstances create communities much larger than a band, second, when nomadism is restricted and people are confined to a site and, third, when resource and social problems are created by external agricultural groups (Gulbrandsen 1991: 99). Although he does not explicitly link his first two points to the appearance of agricultural groups, they are common consequences where Bantu-speaking agriculturists have impinged on San areas in southern Africa. In arguments such as these and even more so in certain revisionist writing, the impression is created that the San Interactionist Approaches only changed through interaction with Bantu-speakers. In addition to structural-marxist and structurationist ap69 proaches, a number of efforts to link rock art to history have emerged since the mid-1980s. In contrast to the theoretically-informed approaches that I have discussed, these studies share an emphasis on the empirical and do not draw explicitly on social theory. Most notably, these studies have been conducted in the mountainous regions along the western seaboard and the south-eastern mountains. Indeed, one of the earliest interactionist studies considered images of domestic stock—cattle, sheep and horses—throughout southern Africa but focused on the rock art of the Cederberg mountain area and the Drakensberg (Manhire et al. 1986). This study shows that paintings of fat-tailed sheep in the Cederberg, tend to coincide “approximately to the area occupied by pastoralists during the early colonial era” (ibid.: 27) while in the Drakensberg, paintings of horses and cattle tend to be far more frequent in the southern parts than in the northern area of the range (ibid.). es of conflict between San and Sotho-Tswana speakers. Another important study is that made by Simon Hall (1994). Drawing on excavated material, rock art imagery and ethnographic-based models of hunter-gatherer and farmer interaction, Hall traces the changing nature of San-Bantu-speaker interaction in the southern Eastern Cape Province. As with Loubser and Laurens, Hall emphasises the particularity of interactive processes between these people for small areas. In contrast to early revisionist positions, Hall uses the available strands of data to point out that interaction was not simply a one-way process of acculturation by the Bantu-speakers of the San. Instead, there were “constructive economic interactions and exchanges” between the San and their Bantu-speaking neighbours (Hall 1994: 81) that mutually restructured both sides. Hall’s study of San-Bantu-speaker interaction is thus the closest yet to achieving some of the four postcolonial principles This early work has since been built upon, particularly in that were outlined in the Introduction. Unfortunately, the south-eastern mountains. Jannie Loubser and Gor- the archaeological material that Hall draws on is, by his don Laurens (1994), for example, have undertaken a own admission, biased to hunter-gatherer archaeology large-scale study of some 200 rock art sites in the Cale- and so we do not yet have much insight into how indon River Valley, an area situated on the western side of teraction affected Bantu-speaking groups in this area. the south-eastern mountains. Loubser and Laurens draw on a wide range of sources including Sotho-Tswana oral While these studies are remarkable because of the extraditions, historical documentation, archaeological and ceptionally rich data and the insightful linkages that rock art data. They consider the impact that the spread researchers make between the rock art (both in terms of Sotho-Tswana-speaking agropastoralists in the Cale- of distribution and meanings) and the broader archaedon River Valley had on San rock art. In a similar fashion ological and historical context, the lack of an explicit to Campbell, they conclude that aspects of the Sotho- theoretical framework sometimes leads to interpretaTswana socio-economic world, such as domestic ungu- tions that lean towards functionalism. So, for example, lates, were incorporated into the production of rock art interaction between the San and Khoekhoen and San in ways that transmuted their realistic aspects into non- and Bantu-speakers is understood in terms of stress, real images that carried meaning within a San shamanic which leads to increased ritual and a concomitant surge worldview. Loubser and Laurens’ work is important be- in painting activity. Yet, it is not explained why stress cause they do not offer sweeping claims for the entire should lead to an increase in the frequency of paintings south-eastern mountains but, instead, they emphasize and there are undertones here of Vinnicombe’s Functhe localized trajectories of San interaction with other tionlaist explanation that the act of painting was catharpeople. Nevertheless, as Campbell did with contact im- tic. Moreover, these approaches share with Campbell the ages between San and colonists, they only consider im- limitation of a consideration of imagery that is explicages that are explicitly of interaction between San and itly related to contact; they do not attempt to consider Sotho-Tswana speakers, such as cattle, sheep and imag- the broader corpus of art as the product of interaction. 70 While the interactionist studies that I have discussed consider the effects of the spread of Khoekhoen pastoralists and Bantu-speaking agropastoralists on San rock art, they all hold to the idea that the tradition of painting remained a San one. A number of researchers, however, have attempted to understand San rock art from the perspective of the Bantu-speakers. Francis Thackeray made one of the earliest efforts in this regard (Thackeray 1984, 1988, 1990, 1993, Botha and Thackeray 1987). More recently, Pieter Jolly has developed Thackeray’s earlier suggestions. Jolly has studied interaction between San and the Bantu-speaking peoples in the Drakensberg, with special emphasis on San-Basotho and SanNguni interaction (Jolly 1994, 1995, 1996, 1996, 1996, 1997, 1998, 1999). His work includes archival analysis, documentation of oral histories and rock art interpretation. He draws on this diverse material to argue for the establishment of symbiotic relationships between San and Bantu-speaking peoples. This symbiosis, according to Jolly, “resulted in the San’s adopting some of the religious concepts and ritual practices of black farmers. We may therefore need to interpret the symbolism and religious ideology underlying some of the art in terms of the dynamics of ideological change resulting from interaction between the San and others…” (1996: 280). Jolly, for example, argues that many therianthropic images in the art probably wear antelope masks and are not, as Lewis-Williams holds, necessarily shamans transforming into animal form while in trance (see for example Jolly 1995). He notes amongst Bantu-speaking communities masks are sometimes worn at certain rituals and that when we see images in the art with masks, they may be interpreted as depictions of those rituals. Jolly is cautious here and he does not suggest that all therianthropic images be understood in this manner. His caution aside, the argument suffers from several flaws. First, there are many therianthropic figures in the art that are clearly not just human figures wearing animal masks. It has long been recognized that many San images blur the boundaries between animal and human characteristics. For example, in Nomansland there are images whose features are almost entirely antelope (Fig. 25). In one example from Storm Shelter (Fig. 26), a figure has an antelope head and hooves. The posture, however, is not that of an antelope since the image is depicted as sitting on its haunches. Most strikingly, the front legs articulate incorrectly for it to be antelope—they bend as human arms do. This image has many antelope characteristics and cannot be interpreted merely as someone wearing a mask. On the other end of the spectrum, in another shelter in Nomansland, RARI-RSA-MOZ1, there are paintings of three figures whose attributes are all human but they are depicted as if running on four legs (Fig. 27). Even when there are no antelope hooves or heads present, their postures suggest that they are partly animal. Both these painted shelters come from the very heart of the area where San interacted with Bantu-speakers intensively during the nineteenth century but they cannot be explained simply as San adopting Bantu-speaking customs of wearing masks. Recently, Jolly (2002) has acknowledged that there is a greater diversity of therianthropic images in San rock art and that some of them are clearly the product of shamanic transformation but he maintains that some may be better understood as San adopting Bantu-speaking rituals and practices. The very diversity of therianthropic images makes his task of demonstrating this even more difficult. One has to argue that specific images must be understood as something very different from all the other therianthropic images. The task is further complicated because there is a distinct lack of evidence to support the directionality of the movements of cultural traits. The number of observations of Bantu-speakers wearing masks in South Africa are minimal and the Nguni and Basotho peoples are certainly not known for masked traditions. Such traditions are far more common amongst the Bantu-speaking peoples of central and western Africa. In the absence of any direct evidence to support his arguments for the flow of directionality, Jolly, a priori, asserts that the direction of cultural flow was from Bantu-speakers to San. The only evidence he offers to support this is a string of analogous examples from other areas where farmers and hunter-gatherers have interacted, both within and 71 Fig. 25. Many images in Nomansland are a combination of human and animal features. This hartebeest has human arms instead of front legs and holds a digging stick with a bored stone. Approximately 20cm in breadth. Fig. 26. Although the anatomical features of this image are antelope, the front legs articulate incorrectly and probably represent human arms. Solid is black, clear is white and stipple represents various shades of red. Scale in centimetres. without Africa. In these cases the flow of information has been from the farmers to the hunter-gatherers. This inductive reasoning, however, can easily be overturned in the case of the San because where evidence does exist for cultural borrowing it is overwhelmingly in favour of movement from the San to the Bantu-speakers. For example, linguistic studies show that the Nguni groups with which the San had such a long and close interaction have a number of clicks in their language. No other Bantu-speaking languages throughout all of sub-Saharan Africa have these clicks and, with the exception of a handful of click languages spoken by hunter-gatherers in east Africa, they do not occur in Africa outside of the southern portion of the continent. Linguists have thus long concluded that the Nguni-speaking languages of the south-eastern part of South Africa adopted the clicks from the San (Herbert 1990a, 1990b, Lanham 1964, Louw 1974, 1977, 1979, Traill 1995). It is not just merely clicks but words that have been taken over by Nguni-speakers. For example, the word amagqirha in Xhosa, referring to diviners, is a cognate with the word !gi:xa in the /Xam San language and means owner-of-potency (Lewis-Williams 2003: 112). In addition to these linguistic elements, David Hammond-Tooke (1997, 1998, 1999, 2002), the doyenne of Bantu-speaker anthropology, has argued that there are a number of cultural elements in the Bantu-speaking communities of south-eastern South Africa that are found amongst the San but are not found amongst other Bantu-speaking groups. Taken together, this suggests that the flow of cultural borrowing was largely from San to Bantuspeakers and not, as Jolly contends, vice-versa. Underlying these specific problems with Jolly’s approach are two larger critical issues—the first concerns the atheoretical nature of his argument and the second his ahistorical understanding of interaction. I discuss each in turn. 72 Jolly does not employ the marxist-inspired politicaleconomy approach of the Kalahari revisionists. As I pointed out in the introduction, that model, with its Fig. 27. While many images that are predominantly antelope in their characterisitics have human features, some images that are mostly human in form exhibit animal charactersitics. In this example, several human figures run on all fours, which suggests that they have adopted animal characteristics. Each figure is approximately 20cm in breadth. strong economic emphasis, results in a view in which the San are merely an underclass of Bantu-speaking society. By abandoning this model, Jolly adopts a position whereby the San of the south-eastern mountains exist in their own right, but are nevertheless subordinate to Bantu-speaking communities. Although he does not point this out himself, this is a radically different position from the revisionists. The problem, however, is that Jolly does not replace the political-economy model with another explicit social theoretical framework. Instead, he opts for an approach to interaction that relies heavily on ideas of cultural borrowing. In this view, southern African societies are divided into discreet and clearly bounded ‘packages’—in Jolly’s case, the San and Bantu-speakers. These discrete groups then exchange items of ‘culture’, be they simple material things such as clubbed sticks or more complex things such as initiation rituals and religious beliefs. In this formulation, the main task of historical analyses becomes an investigation of the direction in which ‘things’ moved between different ‘cultures’. Did the San or the Bantu- speakers have it first and who borrowed it from whom? Jolly’s view of culture is thus very close to the essentialist perspective that postcolonial theory challenges. A second important problem with Jolly’s version of interactionism is the lack of temporal diversity in his argument. He claims that the San of the Drakensberg have been in a “sustained symbiotic interaction” with their Bantu-speaking neighbours (Jolly 1995: 69). The first problem here is with the word ‘sustained’. While, in the article from which this quote is taken Jolly is not explicit about the time period that this word covers, elsewhere he makes it clear that he is interested in San interaction with Bantu-speakers from about 1400 until the last quarter of the nineteenth century (Jolly 1996: 30)—a period of some 450 years. It is clear from Campbell’s work, however, that, even in space of a few decades in the nineteenth century, the nature of interaction between the San and Bantu-speakers varied greatly. If San-Bantu-speaker interaction varied so much within so short a space of time, then it would almost certainly 73 have varied considerably over the 450-year period with which Jolly is concerned, yet he does not account for this change. Although in another article Jolly (1996) makes a distinction between pre- and post-Mfecane periods of San-Bantu-speaker interaction and he also points to the differences between Nguni-San and Sotho-San interaction, he only describes the historical data and never analyzes the material in terms of changing power dynamics. This problem is exacerbated by his exclusive emphasis on San-Bantu-speaker interaction. He (Jolly 1995: 76), rightly observes that Campbell places more emphasis on San-European interaction. Significantly, Jolly never considers that interaction nor does he consider San-Khoekhoen interaction in the southeastern mountains. This is not a standing criticism if one’s aim is explicitly to study San-Bantu-speaker interaction but it is a problem if one fails to consider how other events and processes affected that interaction. Wright (1971), Vinnicombe (1976) and particularly Dowson (1994, 2000) all show how European colonialism profoundly affected San-Bantu-speaker interaction. By downplaying these other aspects, Jolly removes interaction from its local, regional and global context and he creates the impression that San-Bantu-speaker interaction took place in a spatio-temporal vacuum. Although he aims to introduce a historical perspective back into San rock art studies, he ends up reproducing the very ahistoricity that he would challenge. into arguing for cultural borrowing and then demonstrating the directionality of that borrowing—as we have seen, a very difficult task. Ultimately, Jolly suffers from the same problems that the revisionists experience. By treating interaction as a predominantly unidirectional affair, with the San being passive consumers of Bantu-speaking ‘culture’, Jolly denies the San agency and fails to consider adequately how the San resisted domination by Bantu-speakers. In doing so, he reduces interaction to an unchanging process that has remained the same for four-and-half centuries. Towards a New Theoretical Approach The three post-revolutionary period approaches to San rock art research that I have discussed—structural marxism, structuration theory and interactionist approaches—all have positive and negative attributes. Nevertheless, none of them is suited to the writing of past for Nomansland that conforms to the four postcolonial principles that were identified in the Introduction. The interactionist approach of Jolly lacks an adequate theoretical formulation of social interaction and thus tends to reduce San-Bantu-speaker relations to an unchanging formula. Structural-marxism and structuration theory, on the other hand, are social theories that are fundamentally concerned with change. These two approaches, however, deal with abstract concepts such as society, structure, class and function. These concepts From this discussion, it should be clear that, as with are not always helpful when it comes to understandstructural-marxist and structuration approaches, inter- ing San society. As Tim Ingold (1999: 399) points out: actionism, as espoused by Jolly, has both advantages …the forms of hunter-gatherer life cannot be and disadvantages (cf. Jolly 2000, Prins 1999). His work understood as instances of any essential kind of has been beneficial in that it has forced researchers to society. The distinctiveness of hunter-gatherer soconsider more broadly, the influence of Bantu-speakers ciality lies in its subversion of the very foundations upon which the concept of society, taken in any of on San rock art. He has also brought the revisionist its modern senses, has been built. Hunter-gatherdebate into rock art studies, something that Campbell ers show us how it is possible to live socially, (that and Dowson did not accomplish. Nevertheless, Jolly’s is, to conduct one’s life within an unfolding maposition differs fundamentally from that of the foretrix of relationships with others, human and nonhuman) without having to “live in societies” at all. most revisionists in that he does not adopt the political-economy model but falls back upon an essentialist understanding of society. While this allows him to The difficulty for researchers becomes the interdigitamaintain the cultural integrity of the San, it forces him tion of the social theories, which have explicit struc74 tural models about the workings of a society, with the actual rock art data. Often in efforts with explicit structural social theoretical frameworks, like structural marxism and structuration theory, the data—in small and sometimes almost imperceptible ways—are ‘massaged’ in order to fit the abstract theoretical framework. Most attempts at producing a history of the San using their rock art have drawn of these abstract concepts that I have mentioned and consequently struggle uncomfortably to marry theory and data. In a similar manner, Meskell (1999: 25) critiques the structuration approach of Giddens: …my real point on contention is Giddens’ temporal chauvinism, which accords modern society certain freedoms and self-construction, set against the narrow confines of traditional or non-modern societies. For Giddens, tradition structured action and ontological frameworks in non-modern societies (1991: 48), which is simplistic and reductive in the main. In terms of the individual, he quotes Baumeister’s notion (1986) that the ‘idea that each person has a unique character and social potentialities that may or may not be fulfilled is alien to pre-modern culture’ (Giddens 1991: 74). While Giddens wants to temper this to some degree, I think the general tenet continues to colour his work. His position is formulated from ignorance of rather than any specific reference to other cultures or ancient ones (original emphasis). ment that the conceptual structures we employ be made shifting rather than rigidly fixed. The difficulty for any research, then, that would consider the San’s changing relationship with other southern African peoples is the degree to which social theories are applicable and relevant to these peoples. Moreover, the degree to which data are manipulated to ‘fit’ the theories is also an issue. On the other hand, efforts, such as those of the interactionist approach, that lack an explicit theoretical framework, do not escape these problems and they tend to reduce complex processes of interaction to simplistic ideas of cultural borrowing. In terms of the four criteria listed in the introduction— emphasis on hybrid identity, stressing the perspective of the colonized, study of the mechanisms of expression used by the colonized and the deployment of a theoretical framework that does not force those mechanisms into preconceived ideas of social structure—the three approaches discussed in this chapter are only partially successful. While the marxist, structurationist and interactionist approaches all stress the colonized perspective and while all three efforts emphasize rock art as a mechanism of expression of the colonized, none of the approaches consider the art as a process of hybrid identity and all three force the images into a preconceived structural framework, largely modelled on Western societies. Clearly, any effort to interdigitate San rock art with Critiques of structural theories such as those of history requires a theoretical framework but one structural marxism and structuration theory offered that avoids the shortcomings of the structural aphere by Ingold and Meskell respectively are produc- proaches that I have discussed. In the next chaping a growing realization within archaeology and an- ter, I turn my attention to a theoretical approach thropology that social theory is not a universal giv- that offers, to some degree at least, a way around en. Daniel Miller and colleagues (Miller, Rowlands, some of shortcomings that I have described here. and Tilley 1989: 3), note of this growing realization: It is increasingly being recognized that the classic sociological traditions of the 19th century, stemming from Marx, Freud, Comte, Weber and Nietzsche, have specific preoccupations that cannot be universalized. There is a need for a more radical recontextualization of social theory than has hitherto been the case. Part of this recontextualization will involve a require75 CHAPTER 3 THE SOMATIC PAST In this chapter, I turn my attention to a theoretical framework that potentially offers a way around some of the limitations of the theoretical approaches discussed in the previous chapter. This ‘new’ theoretical approach allows for the construction of San history in ways that are more in line with contemporary postcolonial thinking. Instead of treating the San as simply an underclass or alternatively as an isolated, discreet cultural entity, this new approach allows us to consider how the San contributed to hybridized southern African cultures in ways that changed over time and across space. Ironically, this ‘new’ approach is not new but, in fact, has been under development within the social sci- potential of the body lies in the fact that it is not a model—in the sense of a functionalist, structural-marxist or even structurationist model—of how societies or cultures operate. Instead, the body offers a much looser, but theoretically informed, framework from which to approach the various strands of data that can be drawn upon to write San history. While data does not speak for itself, the benefit of writing about the body is that it does not impose abstract concepts of how society functions on data, but instead, allows one to select an aspect of the data and to talk about it theoretically; data are thus given primacy but in a way that is not simple empiricism. ences and humanities since the early 1980s. This frameThe enormous possibilities of the body as a theoretiwork centres on the materiality of the human body. cal tool have enticed many researchers across the social From the Body to Embodiment sciences and humanities and the plethora of new ideas Although apparent in earlier social and anthropological and case studies can sometimes seem overwhelming research (Shilling 2001), the body has gained increasing in their diversity. This is not to say that studies of the attention as a focus of study in recent years in sociology body are a chaotic, unstructured hodgepodge of ide(e.g. Shilling 2001, Turner 1996, Turner 2003), anthro- as. Instead, the analytical issues on which researchers pology (e.g. Blacking 1977, Csordas 1994, e.g. Lambek focus give the diversity of studies their cohesiveness. and Strathern 1998), history, feminism and gender stud- The sociologist Bryan Turner (1996: 24ff) identifies ies (e.g. Butler 1993, Jacobus, Keller, and Shuttleworth three major analytical issues that have emerged in the 1990). This growing interest, according to Emily Martin study of the body. First, the body is analyzed as a set of (1990), is partly due to a change in Western perceptions social practices; the human body has to be constantly of the body. This change is one from a Fordist perspec- and systematically produced, sustained and presented tive where the body is socially organized for efficient in everyday life and therefore the body is best regarded mass production through a disciplined order in time as a potentiality that is realized and actualized through a and space to a late capitalist perspective of the body as variety of socially regulated activities or practices. Secsocially organized for flexible response in a world where ond, the body is studied as a system of signs, that is time and space are rapidly collapsing. Importantly, while as the carrier or bearer of social meaning and symbolthe growing interest in the body possibly stems from ism (1996: 26). Third, there is an interest in the human changing perceptions and experiences of the body in body as a system of signs which stand for and express the West, it does not follow that theoretical approaches relations of power (1996: 27). More recently, a fourth to the body are necessarily modelled on Western socie- analytical issue has emerged that considers the body ties, as are some of the structural theoretical approach- as lived experience. Often spoken of as embodiment, es previously used by rock art researchers. The great this topic requires more discussion than the first three. 76 The current interest in embodiment is partly the result of a growing awareness of the divergent philosophical underpinnings of contemporary social theories. Underlying different corporeal studies are dissimilar strands of phenomenological philosophy. As the study of what is perceived, phenomenology is not a school of philosophy but “is rather a movement whose proponents, for various reasons, have propelled it in many distinct directions” (Kockelmans 1995: 578). Although phenomenological philosophies can trace influences to Georg Hegel (1770-1831) and Friedrich Nietzsche (1844-1900) and, indeed, to Medieval times, most modern phenomenological thinkers take as their starting point the philosophy of Edmund Husserl (1859-1938) (Johnstone 1993: 555). Following Husserl, all modern phenomenological writers stress the sovereignty of experience (Ferguson 2001). Phenomenologists thus share an interest in analyzing and describing consciousness. For Husserl, all consciousness is tied to objects (ibid.). Indeed, phenomenologists hold to the idea that there is a world external to human existence but that world cannot be known. All that can be known are our perceptions of that world. Phenomenologists, then, argue that philosophy must undergo a radical change of attitude whereby philosophers turn “from things to their meanings, from the ontic to the ontological, from the realm of the objectified meaning as found in the sciences to the realm of meaning as immediately experienced in the “lifeworld” ” (Kockelmans 1995: 579). Nevertheless, as they are very much associated with modernity, phenomenologists tend to see experience as falling into the mutually exclusive realms of subject and object (Ferguson 2001: 233). This emphasis on the subjective or objective nature of experience leads to the major differences amongst phenomenological philosophers. These differences are particularly apparent in the works of Martin Heidegger and Maurice Merleau-Ponty, two of the most influential phenomenologists in current archaeological theory. concrete situations of action” (Guignon 1995: 317). For Heidegger, there is no pregiven human essence but, instead, humans are defined through their actions. Human actions, moreover, are embedded in particular contexts so that one is not free to do simply as one sees fit but one’s cultural context, for example, places limits on what one can do. His later work, however, tends to stress the constraint of environment of human action. Indeed, the later Heidegger was concerned with escaping the subjectivism that he felt was dominating Western thought (1995: 319). For him, things such as a Greek temple shape the world in which people live and so constitute the kinds of people that can live in that world (1995: 319). The emphasis on the constraint of environment on human action has earned Heidegger’s later works the unflattering description of “anti-humanist”. Merleau-Ponty’s (1962) work has not had as much influence as that of Heidegger. Nevertheless, his work, while not without its faults (for a discussion of the strengths and weaknesses of Merleau-Ponty’s philosophy see MaCann 1993), has more relevance for studies of the body. This is because Merleau-Ponty “…sought to emphasize not only the existential (worldly) nature of the human subject but, above all, its bodily nature. Thus his philosophy could be characterized as a philosophy of the lived body or the body subject…” (Madison 1995: 484). Merleau-Ponty, then, was the first to make the body central to philosophical analysis. Key to his philosophy is the notion of embodiment, which draws a distinction “between the “the objective body”, which is the body regarded as a physiological entity and the “phenomenal body”, which is not just some body, some particular physiological entity, but my (or your) body as I (or you) experience it” (Leland 1995: 221). The relationship between phenomenological philosophy and social theory over the course of the twentieth century has, at times, been one of inMartin Heidegger’s work is often divided into early difference and at other times one of unease. Inand late periods. His early work, influenced by Hus- deed, as Harvie Ferguson (2001: 242) points out: serl who was his teacher, was an attempt “to develop Sociologists and social theorists have found it a highly original account of humans as embedded in 77 easier to ignore than to criticize or assimilate phenomenology. It has been (wrongly) identified as both ‘psychological’ and ‘idealist’; positions from which sociology regards itself as having a special responsibility to win conviction. The incorrect association of phenomenology with psychology is apparent in the works of theorists such as Giddens and Bourdieu. Both Giddens’ The Constitution of Society (1984) and Bourdieu’s Outline of a Theory of Practice (1977) begin by drawing a distinction between phenomenological and objectivist approaches to social theory. On the one hand, they claim that phenomenological approaches to social theory are those that stress the paramount role of the individual in social interaction at the expense of concepts of constraint of social structure. Objectivist approaches, on the other hand, are those that overemphasize the role of social structure at the expense of individual agency. Giddens’ development of the duality of structure and Bourdieu’s idea of habitus are both conscious efforts to overcome this dichotomy between phenomenological and objectivist approaches. The importance of Ferguson’s point is that it shows that the association of phenomenology with social theories that stress the individual is misleading because phenomenological approaches themselves fall into studies that emphasize either subjectivity or objectivity. being”) to argue that there is no pre-given human essence. Foucault, employing Heidegger against French existentialism, created an archaeological study of systems of knowledge (“discursive formations”) that exist and develop independently of human intentions and beliefs and as a result he rejected the centrality of the subject in the study of changes in knowledge. (Turner 2003: xvi). There is thus a tendency in Foucault’s work on bodies towards objectivism—the abstract concepts of society, structure, formation and so forth have primacy over the individual agent. Those following Foucault have sometimes exacerbated this tendency to objectivism because they attempt to “slavishly imitate” him (Turner 2003: xvi). Merleau-Ponty’s thinking, on the other hand, has impacted most powerfully on those researchers who deny or downplay the idea of social structure. For proponents of this kind of phenomenology, the individual agent is paramount and concepts of social structure are thought to be debilitating to the study of human action. These approaches are most obvious in certain strands of gender studies and feminist critique. Although diverse and complex, feminist studies of body tend to fall into one of three groups (Grosz 1994, cf. Meskell 1999: 39-42). One group stresses the female body as the critical metaphor and motif in the subordination of women. A second group emphasizes the social conThe dichotomy between subjective and objective emstruction of the body. A third group focuses on sexuphases within phenomenological thought is apparent al difference. While some of the writers that fall into from the influence that Heidegger and Merleau-Ponty these three groups adopt a Foucauldian (and through have had on social theorists who consider the body. On him Heidegger’s objectivist version of phenomenolothe one hand, Heidegger’s work has profoundly influgy) approach that sees the body as inscribed (e.g. Butler enced Michel Foucault’s studies of the history of ideas, 1993), other writers, such as Moira Gatens (1996) and in particular his three-volume history of sexuality and Susan Hekman (1990) concentrate on the diversity of his history of madness. In these works and, indeed, in his experience of the human body (for a review of changother writings, Foucault sees the human body as a proding feminist approaches to the body see Hughes and uct of its social landscape. For him, body is something Witz 1997). It is the work of writers such as Gatens that various social institutions work on and inscribe. and Hekman that a shift from the study of the body Turner describes Heidegger’s influence on Foucault: to a study of embodiment is most apparent. Embodiment thus refers to this emphasis on the diversity of Heidegger’s philosophy of time had displaced the bodies as lived experience as opposed to Foucauldian conscious subject of traditional Kantian philosophy by developing the concept of Dasein (“thereanalyses that stress ‘the body’ as socially inscribed. 78 At a semantic level, then, the words 'body’ and ‘embodiment’ refer to different things. The first: …suggests a reified object of analysis, whereas ‘embodiment’ more adequately captures the notion of making and doing the work of bodies—of becoming a body in social space” (Turner 1996: xii). At a deeper, theoretical level, the two concepts refer to two very different philosophical views of social interaction. ‘Body’ has come to be associated with the philosophy of Heidegger—as largely expressed through Foucault—and an objectivist view of social interaction in which abstract notions of society and structure are paramount and individual experience subordinate. ‘Embodiment’, on the other hand, is associated with the philosophy of Merleau-Ponty and emphasizes individual corporeal experience. Both emphases are evident in contemporary archaeological studies of body. Archaeology and the Body Archaeology is often regarded as suffering from a theoretical time lag in comparison to the other disciplines within the humanities and social sciences. The last two decades, however, have done much to change the image of archaeology as a theoretically impoverished discipline. Inspired by broader post-modern movements, post-processual archaeology has adopted a wide range of theoretical viewpoints, the complexity of which seems to increase with each new theoretical development (for overviews see Hodder 2001, Johnson 1999, Shanks and Tilley 1987, Shanks and Tilley 1992). As part of this growing theoretical awareness, many archaeologists took up the ideas of Bourdieu and then Giddens concerning practice and agency in the late 1980s and early 1990s. Importantly, as we saw in Chapter Two, both Bourdieu and Giddens set out respectively to bridge the divide between theories that stress individuals and theories that stress social structure. While some archaeologists still use agency, there has been a marked move away from such attempts to bridge the division to explicitly phenomenological approaches in the discipline. This is perhaps . most evident in archaeological studies of landscape Landscape studies have proliferated in the discipline in recent years and phenomenology is often touted as a radical new way of approaching the past (Bender 2001). Indeed, it is claimed that phenomenological approaches offer “guidance towards a new perspective and experience of archaeology, a new exploration of the past” (David and Wilson 1999: 295). While the potential of phenomenological approaches for archaeology is certainly great, in practice this potential is rarely realized. Indeed, archaeology often uses phenomenology as a tacit circumvention of issues of cultural relativity. Archaeologists apparently ‘immerse’ themselves into the landscape and through observation of the relationships between topographical and archaeological features are then able to discern the symbolic importance of the landscape for past peoples. Gro Mandt (1995: 278), for example, comments on the Vingen rock-engraving sites in Norway: The tiny fjord, surrounded by steep and seemingly impassable mountains, with its large number of rock pictures gives the impression of a very special—even a sacred—place. The view from Vingen towards a characteristic mountain formation called Hornelen adds to this impression. The association of concepts such as sacred and special with this landscape are, of course, imposed and draw on Western perceptions of landscape that, in all likelihood, have very little in common with those of the people who made the engravings. While Mandt’s work is blatant, other phenomenological approaches in archaeology often put forward similar perceptions of landscape in less obvious fashion. In such instances, phenomenological theory functions to mask perceptions of landscape that have been inherited from the enlightenment. There are at least three such inherited perceptions that act as conceptual filters in supposedly phenomenological studies of landscape (Smith and Blundell 2004). First, the archaeologist’s choice of study area is likely to be determined in part by their perceptions of landscape. Second, phenomenological approaches to archaeological landscapes are likely to overemphasize the importance of macro-topographical features at the expense of microfeatures, which ethnographic studies of small-scale peo79 ples show are often more important. Thirdly, landscape archaeology concentrates on topography as boundary when territorial matters are often of less concern to Non-Western peoples. The widespread occurrence of these three filters in archaeological studies of landscape stems from the incorrect conflation of the concept of individual with the pronoun ‘I’—what we may call the empathetic trap of phenomenological studies. rials are to be found, studies of embodiment are opening new avenues of research. This is particularly true of ancient Egyptian and Mayan archaeology (Meskell and Joyce 2003). In Egypt, Lynn Meskell (1999) discusses the lives of tomb builders at Deir el Medina, a New Kingdom village (c. 1500-1100 BC), through the theoretical filter of embodiment. Meskell describes her programme for the study of embodiment as such: It is remarkable, considering archaeology’s current and growing interest in phenomenological approaches, that the discipline has not embraced studies of embodiment, with their phenomenological underpinnings, more fervently. Indeed, the discipline has been slow to take up somatic studies in any form (Knapp and Meskell 1997, Meskell 1996, Meskell 1998, Meskell and Joyce 2003). Those archaeological efforts that have considered bodies, perhaps unsurprisingly, have concentrated on human burials (Shanks and Tilley 1982). These efforts have largely looked at the body as a social object. As Meskell (1999: 42) points out archaeology has tended to First, there is the materiality of the body: the way we eat, sleep, bleed, menstruate, feel pain and so on. Although we might experience these incidents in variant ways, we always start within a single kind of body, to be human is to be embodied. Secondly, there are the elements of construction, the social setting and constitution of the body, depending largely on cultural context, whether it be Sambia or classical Greece. Each society has a corporeal style. For some the body is not skin-bound, but may connect to other bodies, ancestors, spirits and so on. Bodies cannot be considered as ahistorical or non-cultural. Thirdly, there are the operations of sex and/or gender upon the body plus all the other identity markers of sexuality, age, race, ethnicity, disability et cetera …And, lastly, there is the individual dimension: what is uniquely our experience of living in and through our own specific bodies (1999: 37). …be too easily seduced by the body’s overt aesthetic possibility, the promise of new avenues to ancient sexualities and the straightforward power dynamics of the Foucauldian body politic. Archaeology tends to align itself with an inscriptive model”. Within this inscriptive perspective, the emphasis is on how bodies are constructed, controlled and manipulated by institutions of power. Archaeology has been less concerned with approaches that are interested in the body “…as it is experienced and rendered meaningful…” (ibid.). Consequently, Meskell argues that many “…archaeologists still bypass the embodied individual in favour of a body which is a passive reflector of large scale social processes—or what I term the society-in-microcosm model” (Meskell 1999: 43). The analysis of the embodied individual is, however, not an easy undertaking in many archaeological contexts. Often, there is a paucity of data that could be used to produce such studies. It is therefore understandable that many archaeologists would prefer an inscriptive model. In places where a rich diversity of archaeological mate80 This programme is taken further and used as a comparative tool in Embodied lives: Figuring Ancient Maya and Egyptian Experience (2003). In this work, Meskell and Rosemary Joyce compare ancient Mayan and Egyptian constructions of the body showing how at a surface level they appear similar but at a deeper level, they are constructed on very different understandings of corporeal experience. These two works are important because they take embodiment out of a purely theoretical discussion and show how the concept can be used as a tool for archaeological analysis. Embodiment and San Rock Art Although Meskell and Joyce show the way for archaeologies of embodiment, there are still too few such studies within the discipline. Even efforts that use more of an objectivist approach to ‘the body’ are infre- quent. The paucity of body studies is particularly evident in rock art research. This is somewhat surprising given the abundance of rock art images globally that blur animal-human boundaries or exaggerate—sometimes grotesquely so—selective human physiological features; these aspects of rock art offer rich intellectual pickings for studies of embodiment. The relative absence of corporeal studies in rock art is lamentable because, in southern African San rock art research at least, embodiment offers a way around some of the shortcomings of structural approaches to the incorporation of the imagery into the writing of history. Moreover, it provides a tool for understanding the art as a form of identity-construction, which is as we have seen in the Introduction, at the heart of the Kalahari revisionist debate. I discuss each of these in turn. A Non-Structural Social Approach As outlined in Chapter 2, attempts to incorporate rock art into the production of southern African history have largely adopted social theories of a structural type. When researchers use theories of this type, they are forced to follow a particular analytical route. First, the particular social theory has to be mapped onto diverse ethnographies in order to create a general model of San society. Because the social theories are largely modelled on Western societies, the relevance of these theories to hunter-gatherer communities has been called into question and consequently the value of such ‘mapping’ is questionable. Second, the rock art has to be placed within a particular box on the diagrammatic representation of how society operates. In marxist approaches this is often at the ideological level while in structuration approaches, this is at the level of, what Giddens calls, rules and resources. The placement of rock art in these pre-defined boxes tends to create the impression that the art was something that was intellectualized. In later chapters, I will point to evidence that suggests the art was something that was bodily experienced rather than simply intellectualised. Finally, structural theories all assume particular models of social change—marxist schools of thought, drawing on Hegelian dialectic, posit revolutionary shifts in the modes of production while structuration theory uses a recursive approach that emphasizes gradual but continual change. The degree to which different San communities underwent radical or persistent daily social change is itself a chronological issue that needs to be argued for specific cases rather than theoretically imposed. Studies that concentrate on body allow researchers to partially overcome these three problems typical of structural theories. In particular, the concept of somatic society, proposed by Turner, offers ways to by-pass these problems. The notion of a somatic society is one in which the major moral, personal and political concerns are problematized in and expressed through the conduit of the body (Turner 1996: 6). It is particularly in perceptions of disease and bodily excess (Williams 1998) as well as in attitudes towards social fluids (Turner 2003) that one can see the expression of these concerns. The ways in which these moral, personal and political issues are expressed through the body differ through time and across space. This spatio-temporal contextual aspect of somatic society makes it a particularly useful concept for archaeology. Underlying these differences, there are four fundamental tasks concerning the body that confront every society (Turner 1996: 38). These are: • • • • The reproduction of populations in time The regulation of bodies in space The restraint of the ‘interior’ body through disciplines The representation of the ‘exterior’ body in social space Consequently, it should be possible to analyse any social group, including the San, in terms of how they play out these issues. For Turner, the abstractions of class, structure and function have value only in their relation to human physicality and it is the human body and not these abstractions that should form the axis of all sociological enquiry. As Chris Schilling (2001: 446) points out, Turner’s efforts to establish a sociology of the body in terms of these four points has many influences, not least of which is the theory of Talcott Parsons. This gives a 81 structural quality to Turner’s theory of bodily order as expressed in the four points listed. Turner, himself, has recognized the limitations of his earlier work and has shifted toward a position of phenomenological embodiment (see Preface to Turner 1996). Nevertheless, while Turner’s ideas on bodily order still have a structural quality to them, they are still a radical alternative to social theories that previously dominated Western discourse. Although the framework still draws on insights from sociology, it diverges from the abstract, ‘structured’ models of Western society that rock art researchers have traditionally used. Unlike structural social theories, then, somatic society is not an attempt to define a wideranging, abstract concept of society. Rather, it offers a framework that theorizes something that is a universal, the human body, in a sustained and questioning manner. published the first social analysis of the Great Dance. While her lucid explanation of the symbolic qualities of the dance are a model of ethnographic insight, Marshall tends to stress the harmonious aspects of the dance. She notes, for example, that amongst !Kung-speakers, the community would often conspire to have two men who were in dispute dance one behind the other. When one collapsed in ecstasy, the other would support him and they would often heal each other. In so doing, physical closeness was transmuted into social harmony and the tension between the two antagonists was thus resolved. For Marshall, then, the dance alleviated social tension and functioned to produce social co-operation. This Functionlaist interpretation is de rigueur for its time and it does not detract from Marshall’s work which remains a superb piece of social and symbolic analysis. It is possible to describe a San somatic society in terms of the four themes that Turner identifies. However, as the central concern here is with how to use San rock art to write a history, a description of San somatic society should limit itself to those aspects of body that are directly pertinent to the art. As we have seen in Chapter 2, the hermeneutic approach established that the majority of San rock art is implicated in what is known variously as a curing dance, medicine dance, healing dance, or trance dance. Here, I prefer the term ‘Great Dance’ as it evades some of the negative connotations of the other terms and it avoids defining the dance too narrowly. It also stresses the primacy of this ritual and the embodied performance it entails over other San dances. Remarkably, her original work remains the single most penetrating social and symbolic analysis of the Great Dance to date. The ethnographers who followed her to the Kalahari Desert were more interested in subsistence and other facets of San life than in the dance. Consequently, the anthropological analyses carried out after the 1950s tend to be largely derivative of Marshall’s original work and the most substantial analyses of the dance, since Marshall’s original article, have been dominated by psychiatric investigations (Katz 1982, 1997, Keeney 1999, 2003). While these investigations are excellent in their emphasis on the individual experiences of the dance, they too tend to emphasize the positive healing aspects of the dance; rarely do they try to understand the dance as a vehicle for social negotiation and contestation. We are thus faced with an unusual situation in which the dominant understanding of the Great Dance is still heavily influenced by Functionlaist thinking. There are very few analyses of the dance that examine how the ritual might produce social conflict, or at least in Victor Turner’s terms, function to reproduce the already existing contradictions in San society. How the ritual might contribute to the production of differentiation and inequality amongst the San is an issue seldom considered in the ethnography of the Kalahari San groups (a notable exception to this pattern is the work of Almost all San groups throughout southern Africa practise the dance. While there are, obviously, differences in the way the Great Dance is practiced amongst different San groups, it is the astonishing similarities across impermeable linguistic barriers that are so striking. For a ritual that is so important to San people through time and across space, it is remarkable how little social analysis of the dance has taken place. Although dances fitting this description were observed amongst the San in the colonial period (e.g. Arbousset and Daumas 1846), it was only in the late 1960s that Lorna Marshall (1969) 82 Mathias Guenther, which I will discuss in detail shortly). Women or young girls do most of the singing and clapping at the dance. They have the crucial role of reguA closer look at the mechanics of the dance, however, lating the availability of supernatural potency. They do reveals that it can be interpreted as a vehicle for creating this by varying the rhythm and pitch of the singing and social inequality amongst the San. In the Kalahari Desert clapping. When one of the dancers or shamans becomes of the last few decades, the dance is usually performed ecstatic and loses control, perhaps falling into the fire or at night and the ritual can occur several times a week. running out into the darkness to ward off malevolent Typically, the dance is circular in form, taking place spirits, the singing and clapping slow down in intensity around a fire. The women clap and sing songs while and volume, sometimes coming to a complete stop (on most of the men and some women dance. This con- the role of San music in the dance see England 1968). In tinues until some of the dancers and even some of the slowing down the physical performance of the dance, choral performers become ecstatic, often collapsing or the San believe that they control the ebb and flow of falling into the fire. When ecstatic dancers behave in this supernatural potency and control is a central concept manner, other participants at the dance support them. in working with the energy. In limited and controlled When one or more of the performers enters trance, the quantities, the potency is beneficial, but in excessive intensity of the singing, clapping and dancing wavers quantities it is highly dangerous. The !Kung San, for exwhile others attend to the collapsed person. The mo- ample, believe that dancers filled with potency can kill mentum of the singing and dancing soon resumes after small children by pointing and snapping their fingers at some minutes have passed. The dance proceeds in this them (Katz 1982: 46, 168, 263, Marshall 1969: 351-2). manner of undulating intensity until late at night, sometimes only ending in the early hours just before dawn. Control of potency is also an important criterion in making a further distinction between the various performers Underlying the various actions of the performers at this at a dance. It has often been noted that most of the men ritual is the San concept of supernatural potency. Called in certain !Kung-speaking communities of the Kalahari n/om by !Kung-speakers and !gi by /Xam-speakers, this Desert partake in the dance. From this observation, it supernatural potency permeates the universe and re- is sometimes claimed that up to two-thirds of the men sides in certain animate and inanimate things. The San in San society are shamans. Observations of !Kung San regard animals, especially the eland, as reservoirs of this communities in present-day Namibia, however, suggest potency. The potency is also found in the songs that the that these claims do not adequately take into account women sing and they are often named after a particular San perceptions of differences amongst the various animal, carrying labels such as ‘giraffe song’ or ‘eland performers at the Great Dance. In July 2002, together song’. The potency is important to the dancers because with colleagues from the Rock Art Research Institute, I they require it to enter the spirit realm where they per- was fortunate enough to visit a !Kung San community form various tasks such as the fighting of malevolent in Tsumkwe, on the border of Namibia and Botswana. spirits, making rain and visiting relatives in distant plac- Bradford Keeney, a psychiatrist who has been working es on out-of-body journeys. Importantly, dancers need with San n/omkxaosi in the Kalahari Desert for several to activate this potency through their physical actions years, kindly arranged the expedition. Working with and once ignited they must control the energy. It is best Keeney and his translator over several days, we were into think of the Great Dance as a performance with set vited to observe two dances and we were allowed to inroles and specific actions to be carried out by the indi- terview participants before and after the performances. viduals who fulfil those roles. There are three fundamental performance roles in the dance: singing/clapping, During these few days, it became apparent that this pardancing and shamanic activities. I discuss each I turn. ticular community made a distinction between ‘danc83 ers’ and n/omkxaosi, meaning owner-of-potency. It is only the n/omkxaosi that may properly be called shamans and at any particular dance there are usually only a few of them. N/omkxaosi tend to be older men in the community while dancers are usually younger. Indeed, this particular community only recognized two elderly men at the two dances that I observed as n/omkxaosi as well as several women. All the other, younger, men were referred to as dancers. The distinction, then, between n/omkxaosi and dancer is partly age-based. This is perhaps not surprising given the lengthy process by which one becomes a n/omkxaosi. Amongst !Kungspeakers, there is no formal initiation into the category of owner-of-potency. Instead, becoming a shaman is a gradual transition from dancer to n/omkxaosi. Activating n/om is a painful experience and carrying out shamanic labour in the spirit world is described by many n/omkxaosi as a profoundly frightening experience. Accordingly, many young dancers decide that the transition from dancer to n/omkxaosi is too taxing and they do not advance. Sometimes a young dancer will seek out an experienced n/omkxaosi to guide him through the painful and frightening experiences of the dance. In the absence of any obvious initiation ritual to mark the transition, it is not exactly clear when a dancer becomes an owner-of-potency. Nevertheless, it seems that any change in status is closely tied to corporeality. That body figures prominently in the shifting status from dancer to owner-of-potency is evident from the Tsumkwe community that we visited. The two elderly n/omkxaosi of the community both appeared physically feeble during the course of interviews. Indeed, one of the two had recently been treated for tuberculosis. At the night time dances, however, with their shirts removed and in dancing attire, they transformed into physically imposing participants. Both were vigorous and enduring dancers and their healing prowess was visible to all. At times, for example, one could see the abdominal contraction and expansion that comes with the hyperventilation experienced from prolonged periods of dancing. These contractions are important because the !Kung San believe that the abdomen is where 84 the n/om ‘boils’ before it travels up along the spine and reaches the head where it explodes and catapults the dancer into the spirit world. Since the energy is invisible to all except the n/omkxaosi, the physical manifestation of the boiling energy in abdominal contraction and expansion is important because it provides a visible sign—along with the profuse sweating—that the dancer is activating and working with the energy. In comparison to the n/omkxaosi, the young dancers seemed awkward and reluctant participants at the dance. Some young male dancers would leave the dance circle and converse with members of the community or any other onlookers, only to return to the dance a little later. As the night progressed and the dance intensified some returned to dance and others even assisted n/omkxaosi with their healing activities. Yet, the participation of the younger dancers never reached the level of involvement of the n/omkxaosi and their participation at the dance was half-hearted in comparison to the two elderly n/omkxaosi. Of course, their lacklustre performance could be attributable to a number of reasons. The presence of outsiders at the dance could have made the younger members of the community reluctant to participate. This reluctance may be the result of increased exposure amongst !Kung San youth to Christian-influenced education. We need to recognize, therefore, that the level of participation by dancer in the dances may have varied over the last few decades of ethnographic research in the Kalahari Desert and, by extension, over time and across space between different San communities. Nevertheless, the distinction made by the Tsumkwe community under discussion between mere dancers and owners-of-potency demonstrates that the dance does not necessarily promote egalitarianism, as is usually suggested, but that it is also a vehicle for the production of social difference. This difference, moreover, is based on gender but also on age. The most important factor in this differentiation, however, is bodily prowess in the various techniques of the dance and its associated shamanic activities. The importance of physical prowess at the dance to the status of a shaman is evident beyond the ritual it- self. After a dance has taken place, San communities in the Kalahari often critically discuss the performance of various participants. This critical discussion often centres on the physical performance of the dancer during the ritual. Moreover, dexterity at the dance leads to benefits in daily social life. Supernatural potency is associated with sexual potency and those who become known for their abilities in the various techniques of dancing and shamanic activities receive substantial sexual interest from the women. At times, amongst !Kung communities, this leads to jealousy between older men who are no longer capable of performing vigorously at the dance and younger, more virile, participants. mentioned, the concept of somatic society holds that the major moral, personal and political concerns of any community are problematized in and expressed through the conduit of the body (Turner 1996: 6). Considering the rock art in Nomansland from the perspective of somatic society allows us to speak about broad-based meanings that are related to shamanistic belief and practices but it also allows us to see the depiction of those bodies as an expression of the major moral, personal and political concerns of San society. In this view a shaman’s body is not simply a religious symbol but also a political one. By allowing us to see the imagery at the same time as both meaningful in a religious and political sense, a somatic lens allows us to bridge the dichotomy By analysing the Great Dance through a somatic lens, between meaning and motivation that has hampered it soon becomes clear that the ritual is far more than southern African San rock art research since the 1970s. simply a medium for the promotion of harmony and egalitarian social structure. Indeed, the dance produces In so doing, we place the art at the centre of any soor reproduces a number of differences in San commu- cial discussion concerning the San. Whereas strucnities between men and women, young and old and sha- tural-marxist and structuration had to map their remans and ordinary people. Underlying all these conflict- spective social theories onto San society and then fit ing relationships is the physical skill of participants at the art into a ‘slot’ on a diagram of supposed social the dance. The importance of this analysis of the Great structure, embodiment allows us to treat the art as a Dance from a perspective of ‘the body’ is in the implica- direct metaphorical comment on San social processes; tions it holds for interpreting San rock art. In Chapter 2, in this sense, embodiment allows us to upfront the we saw how the hermeneutic approach established that data rather than force it into preconceived schemas San rock art, principally in the Drakensberg but also of social interaction modelled on Western societies. elsewhere, was implicated in their shamanistic religious practice and belief. The art depicts the Great Dance, An emphasis on body and embodiment then allows us fragments of the dance such as postures and corporeal to escape some of the limitations of structural-type aspects (e.g. nasal haemorrhage) particular to the ritual, theories to understanding the social production and visions and somatic experiences generated in the altered consumption of the art. In Chapter Two, I pointed states of the dance (e.g. therianthropic figures), symbols out that one of the most significant problems facing of supernatural potency (eland) and metaphors (such post-revolutionary-period southern African rock art as death and underwater) of the trance experience. In a research was the bringing together of meaning and broad sense, the art refers to meanings associated with motivation in the analysis of imagery. Whereas rethe dance. However, as I have argued, real dances pro- searchers have approached meaning from an extemmote social differentiation and that discrimination is poral and extra-spatial perspective, studies of motivabased on bodily performance. We may thus reasonably tion have always referred to the social forces driving expect that rock art, which depicts certain aspects of the production and consumption of the imagery, which the dance, is also implicated in the production of social are largely accepted as varying through time and over inequality and that such discrimination is tied to the rep- space. Unfortunately, approaches to motivation have resentation of human corporeality. Moreover, as I have largely drawn on structural-type social theories and 85 have struggled, with varying degrees of success, to fit Nevertheless, within the broader archaeological field meaning into these social frameworks. Embodiment of- embodiment studies strongly emphasize identity. fers a way of bridging meaning and motivation because Meskell (1999: 35-36), for example, states that: it allows us to treat the images as carrying shamanic meanings that are at the same time social interventions. A significant dimension of the individual, personhood and identity is surely the locus of the body itself: at times a somatic grounding for the self and Embodiment and Identity at other times merely a linked entity. Experiences While body as a tool for social investigation allows us to and constructions of the body provide an imporovercome some of the shortcomings of the structural tant existential grounding for individuality and for theories previously used by rock art researchers, it also the processes of individuation in any given culture. allows for a more central consideration of processes of identity-formation and contestation. A persistent arguWe have already seen how, by adopting a corporeal apment in corporeal studies is that bodies are inextricably proach, the Great Dance can be re-examined as a ritulinked to identity. Timothy Yates (1993), for example, in al that produces social hierarchy, rather than one that what was a pioneering effort at analysing rock art from leads to social cohesion. The implication of bodies in the theoretical perspective of body, pointed to the link identity, however, allows us also to consider the Great between body and identity in Scandinavian rock art. In Dance as a process of identity construction. Guenhis analysis, Yates draws heavily on critical discussions in ther, for example, describes the role of identity-formapsychoanalysis that consider the relationship of body to tion amongst San groups in Botswana. Working in the identity. In line with some feminist writers, he points out Ghanzi district of western Botswana in the late 1960s that the privileging of sexual physiology as an indicator and early 1970s, he (1975, 1975/1976) considered the of identity is a cultural construct (ibid.: 49). This manchanging relationship of San people to their European ner of defining identity is one that is largely followed in and Bantu-speaking neighbours (cf. Guenther 1999). the West but it is not a human universal; in some socieThis relationship was one in which San were being deties, for example, having a penis is not necessarily comprived of their traditional hunting and gathering way mensurate with maleness. Yates concludes that identity of life and were becoming labourers for their agropas“is achieved through signification, not through natural toralist co-habitants of the region. Hunger, malnutrior biological processes contained within the body ittion and disease were constant concerns for the San self ” (ibid.: 69). Yates uses such insights to criticize cerof Ghanzi who were from diverse linguistic groups, tain Scandinavian rock art studies that assume that the including Nharo, !Kung, G/wi, Tssau, G//ana and !Xõ absence of women in the rock carvings of the region (Guenther 1975/1976: 47-48). Guenther (1975/1976: is always an empirical issue. Furthermore, he contends 49) argues that in Ghanzi the dance became a mechthat such arguments make the mistake of defining idenanism for the San to cope with their predicament: tity in non-discursive terms—complete bodies already distinguished as male and female (ibid.: 46). While his For the farm Bushmen the trance dance is an emotionally envigorating (sic.) and satisfying socritique is poignant, ultimately, Yates does not drive cial event because of its therapeutic, integrative home the message; he does not himself offer any arguand cathartic qualities. These are what attract a ments for how identity was constructed in Scandinapeople whose existence is full of oppression and vian rock art in ways that transcended the depiction of deprivation. The ritual offers relief from these pressures and is thereby intrinsically ‘revitalizing.’ sexual organs. Nevertheless, Yates’ pointed the way for future corporeal analysis of rock art but, unfortunately, as mentioned earlier, there have been very few efforts While this statement echoes the harmonious, functionto build on his formative ideas within rock art studies. list-inspired interpretations offered by earlier research86 ers, Guenther goes on to point out three important revitalizing qualities of the Great Dance as performed in Ghanzi during that time. First, the dance contributes to a positive sense of self amongst those who participate. At the dance the San speak of themselves as downtrodden and oppressed whereas in daily life they are more self-effacing. Moreover, at the dance items of material culture that are European or belong to Bantu-speakers are shunned and the participants talk about an earlier time before these outsiders came and when life was better. Second, the dance generates a sense of a San identity that transcends the linguistic divisions amongst the various groups. The structural form of the dance, for example, “is ‘hybrid’ in form and content, consisting primarily of ≠’au//ei and Nharo attributes and the Great Dancers are members of all the various linguistic groupings” (Guenther 1975/1976: 51). Third, the dance appears to assist the various San communities to articulate more clearly a strategy for resisting exploitation and oppression. According to Guenther (1975), this strategy revolves around owners-of-potency who have acquired wealth and prestige. It is around these prestigious individuals that San identity at Ghanzi took shape in the 1970s: As the main actor in the dance, the San ritual were an individual feels relief from the pressures of contact and asserts his personal and collective identity, the dancer becomes a powerful rallying symbol that represents the San people and their culture (Guenther 1975/1976: 52). therefore makes a useful analytical tool for approaching San history. As I pointed out at the start of this text, issues of identity underlie much of the Kalahari revisionist debate. On the one hand, some writers who adhere to a more traditional view subscribe, either implicitly or overtly so, to a primordialist or essentialist view of societies that sees them as bounded and clearly recognizable entities. Even some revisionist-inspired studies, such as Jolly’s work on San-Bantu-speaker interaction, cling to this way of conceptualising collective identities. On the other hand, revisionist approaches collapse all differences to class distinction. Importantly, the difference between these two views is not merely an empirical one; it is also a profoundly theoretical one. Traditional views often follow a Functionlaistinspired view of societies as cohesive and integrated wholes while revisionist approaches follow marxistinspired political economy models that privilege economics as the primary mechanism for constructing social identity. In their most basic formulation, both these ways of seeing San identity have a profoundly negative impact on the present-day socio-political situation of living San, as we have seen in relation to the Hoodia cactus and CKGR issues. Body and particularly the notion of embodiment offer a way around this dilemma by emphasizing identity as something that is individually experienced, not merely socially inscribed. The Somatic Past These three observations go beyond the traditional ethnographic observations that the dance operates to produce social harmony to suggest that San identity— both individual and collective—and the construction of that identity viz. a viz. ‘other’ people is inextricably tied to the dance. It is particularly through the efforts of the shamans that this identity is expressed and, as we have already seen, powerful dancer/shamans are identifiable as such largely through their physical abil- Body and embodiment offer theoretical tools from which to approach the integration of San rock art into history that is more in line with postcolonial aspirations than structural-marxist, structuration or interactionist approaches. Nevertheless, as we have seen in this chapter, theoretical discussions of body are diverse and sometimes opaque. If we are to use body and embodiment successfully to link rock art to history in Nomansland, then we need to take the often complex and diity at the Great Dance. At one level then, the dance verse ideas and develop a strategy with which to inteconstructs identity, both collective and individual, and it grate the images into history. While this strategy may does so through the embodied actions of the dancers. have applicability for rock art in other parts of southern Embodiment is vital to identity-construction and it Africa, it is intended here as a guide from which to ap87 proach the rock art of Nomansland. This strategy— which I term the somatic past—comprises four points: 1. Painted or engraved anthropomorphic bodies need to be treated as the basic unit of analysis for a consideration of change in the rock art of Nomansland 2. Where the evidence allows, a somatic analysis of San rock art should approach the imagery from a perspective of embodiment—that is the images need to be seen as an expression of how the world is perceived by individuals and how their identities are tied to their physical being. 3. Where an embodied approach is impractical or unachievable, analyses should nevertheless consider ‘the body’. Such an approach treats painted bodies as microcosms of the social conditions of the group that produced the images. The ways in which bodies are represented in the art, for example in various postures, are thus not only considered as depicting meanings associated with shamanism but are at the same time metaphors for the state of the social group that produced them. 4. The somatic past should not simply consider the meanings of images from the perspective of body or embodiment but should approach the entire production and consumption of the imagery from a corporeal perspective. This fourfold strategy, I believe, offers a productive way of incorporating the rock art of Nomansland into the production of San history for the area. The proof of the pudding is in the taste and the applicability of this strategy needs to be demonstrated. In order to show its applicability, I begin, in the next chapter, by considering how corporeality allows us to augment and extend the current interpretative understanding of San rock art. 88 CHAPTER 4 THE JAWS THAT BITE, THE CLAWS THAT CATCH: THE ELDRITCH IMAGES Images of flying antelope have long been debated in San rock art research. These images, characteristically, have antelope heads and hooves but they have human arms or wing-like protrusions, often in an arms-back position. They occur throughout large sections of the south-eastern mountains. Originally, researchers suggested that these flying-buck (also sometimes called ‘ales’) depicted spirits-of-the-dead. Neil Lee and Bert Woodhouse (1964, 1968) were among the first to put forward such a suggestion. Shortly thereafter, Harald Pager (1971) extended the argument to include both flying antelope and therianthropic images and a few years later Vinnicombe (1976: 239) made a similar argument. She suggested that spirits-of-the-dead were very much associated with wind in San cosmology. As the flying-buck images suggested flight, by extension they depicted these supernatural beings. With the development of the hermeneutic approach, these interpretations of the flying buck images were challenged in the 1970s and 1980s. Lewis-Williams (1975: 422, 1981: 8788) argued that the ethnographic evidence to suggest that these images were spirits-of-the-dead was flimsy. There was, however, significant evidence that pointed to an interpretation that suggested that they depicted transformed shamans on out-of-body journeys. For close on two decades, the discussion concerning whether or not flying buck depict spirits-of-the-dead or transformed shamans lay dormant until Anne Solomon, one of the most vehement critics of Lewis-Williams, resurrected the earlier suggestions of Lee, Woodhouse, Pager and Vinnicombe that some images in San rock art depict spirits-of-the-dead. Solomon believes that San rock art research has become too one-dimensional in its concentration on shamanism. She argues that research should focus on variability within the art and so she suggests that some images of therianthropes in the art represent spirits-of-the-dead and are not neces- sarily images of transformed shamans (Solomon 1997: 9). The primary challenge that Solomon raises for contemporary researchers is that of differentiating between images of shamans and spirits-of-the-dead. This is not as straightforward a task as it might appear. The ethnographic material from diverse San groups shows that both shamans and spirits-of-the-dead share many important features. Both shamans and spirits-of-the-dead, for example, can turn into animal form. The difficulty then is how to distinguish images of shamans from depictions of spirits-of-the-dead in the rock art. Unfortunately, Solomon does not provide any criteria upon which to base such discrimination and thus she does not answer her own challenge. Solomon also appears to be unaware that Dowson (1994), who, as we saw in Chapter Two, was one of the arch-proponents of the shamanistic interpretation of San rock art, argued that certain rock painting images depict spirits-of-the-dead. Dowson based his argument on the context of a single painted image in Nomansland but, as he was primarily interested in other aspects of the art, he did not go on to suggest criteria for discerning images of shamans from those of spirits-of-the-dead in San rock art more generally. Criticism of the shamanistic approach as only one-dimensional, then, is misleading. Nevertheless, the important question raised but not answered by Solomon is this: how do we identify variability in San rock art, when so much of it is overtly shamanic and concerned with the experiences of the Great Dance? In this chapter, I argue that a closer consideration of San concepts of embodiment allows researchers to identify more variability in San rock art than has hitherto been the case. In particular, embodiment allows for the discrimination between images of shamans and images of spirits-of-the-dead. In assisting researchers to make this distinction, I argue, embodiment is a useful tool for interpretative analysis. As such, it provides a 89 theoretical tool with which to bridge the divide between meaning and social motivation in San rock art research that has been apparent since the so-called revolutionary period that I discussed in Chapter Two. I begin by describing the cross-cultural similarities of San concepts of the spirits-of-the-dead. These cross-cultural similarities show that both shamans and spirits-of-thedead share a number of aspects that make it difficult to distinguish one from the other and, consequently, it is difficult to identify differences in the art. Nevertheless, a consideration of how San ideas of disease, death and disorder are corporeally manifested offers a way around these difficulties. With a clearer idea of the bodily distinctions between shamans and spirits-ofthe-dead, it is easier to look for differences in the art and so, in the last part of this chapter, I turn my attention to a distinctive category of painted image found at Storm Shelter and at other Nomansland rock art sites. While the images are clearly shamanic, they have bizarre corporeal qualities that suggest that they are something more than just depictions of potency-owners. Disease, Death and Disorder in San Cosmology In spite of the linguistic differences and spatial distances between the various San groups, they share some remarkable similarities (Lewis-Williams and Biesele 1978). The most important of these is, as we have seen in Chapter Three, the performance of the Great Dance. The similarities of the dance between the different groups are perhaps not so surprising when one considers the similar existential and ontological reasons behind the ritual. We have already seen how the dance acted as a vehicle for the creation of a PanSan identity amongst the diverse groups living in the Ghanzi region in the 1970s. According to Guenther, there are also other reasons for performing the dance at Ghanzi—these concern hunger and disease. Of the first of these he states (Guenther 1975/1976: 46): Hunger and malnutrition are the constant existential concerns of the farm Busmen and this is one of the reasons for the high incidence of stock theft committed by the San. Some white farmers suffer heavy losses through stock theft 90 and this is the major reason for the strongly unsympathetic regard in which the whites hold the San. Stock theft, along with squatting, petty crime and poaching, are also the main offences for which farm Bushmen are arrested and imprisoned in the local jail. The local farmers and administrators refer to these problems as the “Bushman Problem”. The San add hunger and disease to the list and refer to these problems as shda8 (Kgalagari-derived term meaning “suffering”). The term has wide currency among farm Bushmen who use it epigrammatically to sum up the conditions of life of a farm Bushman. Hunger and malnutrition are, of course, the opposite of a preferred, healthy life and they are the anthithesis of the normal social order and so they can be considered to be disorder. Of the second concern, Guenther (1975/1976: 46-47) notes: Disease is the other existential concern for the San of Ghanzi. The incidence of disease has drastically increased in the area as a result of contact which has introduced new diseases to these people. These are of two types: organic diseases such as tuberculosis, smallpox, chickenpox, venereal disease and “social” diseases, mainly witchcraft and sorcery. The latter were introduced by the black peoples; the San’s own religious system has always been devoid of these “black arts”. Brian Turner (2003: 1), speaking about social organization more broadly, points out that all “disease is disorder—metaphorically, literally, socially and politically”. Like hunger and malnutrition, it is the opposite of a normal, preferred and ordered existence. Ultimately, if not corrected, hunger, malnutrition and disease can lead to death; they are thus afflictions that must be overcome if the normal order of things is to be restored. Hunger, malnutrition, disease, death and their treatment and prevention are aspects of San thought that are negotiated at the Great Dance. To understand how San deal with these concerns at the dance, we need to appreciate the religious cosmology that drives the dance. The most extensive studies of San religious beliefs have been undertaken amongst the !Kung and other groups living in Namibia and Botswana. While not as extensive as the Botswana and Namibia material, the available ethnographic evidence for the southern San groups suggests that they shared similar religious cosmologies. For this reason, I concentrate on the !Kung in the following discussion on San religious cosmology and I note where there are interesting similarities or differences in cosmology between the !Kung and other groups. Most San and, indeed, Khoekhoen (see Schapera 1930 for a discussion on the religious similarities and differences amongst Khoekhoen and San groups), groups believe in a supernatural world, occupied by various gods and spirit-people. This other world is said to exist either in the sky or underground. In many ways, this spirit world parallels the real one. The supernatural and real worlds are interconnected and humans can journey to the spirit world while the inhabitants of that alternative reality can travel to this one. Different San groups appear to believe that there are diverse access points to the supernatural world; one that appears to be common to most groups is the waterhole. Once access to the supernatural realm is gained, the principal conduit for moving in that world and returning to this one are invisible threads or webs that hang from the sky. It is along these threads that shamans in trance and creatures from the spirit world move (Katz 1997, LewisWilliams et al. 2000). Elizabeth Marshall Thomas (1959: 147) describes the !Kung version of this other world: that inhabit the other world. In particular, most San groups in Namibia and Botswana appear to believe in dualistic deities—a paramount and a lesser god. The function and morality of these two gods, however, appears to vary from group to group. Some groups see the paramount god as more benevolent while the lesser god is malicious; in other groups, the paramount god is just as malevolent as the lesser one. In addition to the two gods, all San groups believe in various kinds of spirit-beings that live with either or both of the gods. The !Kung, for example, believe in the knee-knee-none people who are spirit-beings that have human-like form but that are remarkable for having no knees (Marshall 1999: 247-248). Not all San groups, however, believe in this specific category of supernatural beings. Nevertheless, one kind of spirit-being that all San groups believe in is the spirits-of-the-dead. Remarkably, the word for spirits-of-the-dead is similar across different linguistic groups in the Kalahari. The !Kung word is //gauwasi and other groups often have a similar-sounding word. It is not just in name, however, that the spirits-of-the-dead are similar. They have comparable functions and motivations that span linguistic divides. The similarities in the function and motivation of spirits-of-the-dead stems ultimately from widespread Kalahari San beliefs about life after death. The G/wi, for example, are typical in that they believe in life after death, but not in the idea that one’s behaviour in life influences one’s condition after death. There is no Heaven or Hell and every individual’s lot is the same after death (Silberbauer 1965: 102). The G/wi Under the world is another world just like this one, and, indeed, all Kalahari San groups appear to believe with trees and omarambas, hills and pans, a sun that upon death, human beings are transformed into and a moon, Bushmen, Bantus and Europeans spirits-of-the-dead and then they live for eternity in the just like the people here and the bottom of our spirit world. Typically, one of the two gods creates spirworld is that world’s sky. There is no other connection between our world and the lower world its-of-the-dead from deceased human beings. Amongst except possibly through certain deep waterholes, the !Kung, it is ≠Gao N!a, the paramount god, who for it was through a waterhole that the great god manufactures spirits-of-the-dead. In the process of climbed from that world when he first came here. manufacturing a spirit-of-the-dead, an important disIn addition to perceiving waterholes as an interface tinction is drawn between a person’s life, /xwa, and a between the real and spirit worlds, various San groups person’s soul, /n (Marshall 1999: 27). According to the share great similarities in their beliefs about the beings !Kung, both substances exist in the body but have im91 portant differences. Lorna Marshall (1999: 27) describes and blood south and then east to where the great the !Kung perception of /xwa (life) and /n (soul): god, ≠Gao N!a resides. While the !Kung believe that it is specifically the spirits-of-the-dead who carry the It exists in the torso, in all the vital organs, in soul and other body parts to the gods in the sky, a the abdomen, in the blood, in the heart, lungs, number of Kalahari San groups believe generally that throat and mouth and everywhere in the head. once the soul escapes from the body it travels to god’s It does not exist in the arms or the legs, the !Kung believe, because they know that human village in the sky after death (Barnard 1992: 252). beings and animals can be wounded in a limb or even lose one and still not die, whereas a Once the soul has reached its destination, it is transwound in one of the vital parts is likely to kill. formed into a spirit-of-the-dead by one of the gods. The people with whom I spoke thought that life must be especially concentrated in the heart. A The !Kung state that it is ≠Gao N!a who turns souls into wound there is sure to kill. When life dies in the spirits-of-the-dead. He does this by making a fire under body, it stays there, dead, as the body itself is the tree that is said to live close to his house. On the fire, there but dead. It is the spirit that does not die. he places a pot with ingredients that have powerful supernatural potency. These ingredients cannot be found Although Marshall uses the word ‘spirit’ here, I prefer on earth. He then hangs the soul, heart and blood from the word ‘soul’ because it allows us to make a distinction the tree (which the !Kung say also has vast quantities of between the supernatural substance and the being that is potency). The smoke from the pot then rises to meet made from that substance, for amongst the !Kung, spirthe soul and the two tangible body parts. Within the its-of-the-dead are said to be made from a person’s ‘soul’. smell of the smoke are carried the supernatural energies of the unearthly ingredients that boil. As the potency Whereas /xwa (life) is tied to the body, even after death, of the smoke flows around the soul, heart and blood /n (soul) is transcendent. The soul is closely associated suspended in the tree, it transmutes them into a new with breath and more generally with air. This is also the body and a spirit-of-the-dead is formed. Once the new case amongst other San groups. Carlos Valiente-Noailles spirits-of-the-dead are fashioned, ≠Gao N!a rubs their (1993: 197) notes that amongst the G/wi, breath and spirbodies with a special fat called !thu. As with the ingreit are closely connected. As one G/wi San commented: dients of the pot, !thu is comprised of elements that “When somebody dies, the air inside him escapes out. are otherworldly and extremely potent. After the spirThe body is buried, but the air that has escaped makes its-of-the-dead have been created, they live with ≠Gao (becomes) this spirit.” Amongst the /Xam, similar beliefs N!a in a largely subservient role (Marshall 1999: 28-29). appear to have been held. Diä!kwain (Bleek and Lloyd 1911: 397), for example, linked death to wind: “When Life with ≠Gao N!a is not that different from living on we die, our [own] wind blows; for we, who are human earth and the spirits-of-the-dead have similar implebeings, we possess wind; we make clouds when we die.” ments, tools, weapons and clothing to those of mortals. They eat the same plant and animal foods. They are, For the !Kung, the process of becoming a spirit after however, particularly fond of honey, a substance redodeath involves the soul as well as physical body parts lent with supernatural potency. The daily activities of the (Marshall 1999: 27-28). Upon death, existing spirits-of- spirits-of-the-dead are also similar to those of people the-dead visit the corpse and remove the soul through on earth. So, for example, spirits-of-the-dead perform the top of the head. They also take the heart and the dances, as do living San. Megan Biesele (1993: 71-71) reblood of the person and they carry these three ele- counts a remarkable story from a !Kung shaman, Kxao ments towards the western sky first, where the lesser Giraffe, who made a journey to the spirit world where god, //Gauwa, lives. They then take the soul, heart he saw God’s House. On his journey, Kxao Giraffe en92 countered spirits-of-the-dead. He described them thus: And the spirits were singing. The spirits were having a dance. I began to dance it, too, hopping around like this. I joined the dance and I danced with them, but Kaoxa9 said to me, ‘Don’t come here and start to dance like that; now you just lie down and watch. This is how you should dance,’ he said, as he showed me how to dance. So the two of us danced that way. We danced and danced…Yes, my friend. Now up there in the sky, the people up there, the spirits, the dead people up there, they sang for me so I can dance. !Kung who have seen spirits-of-the-dead describe them as resembling their earthly, physical forms but with one difference; instead of curly hair they have straight hair, similar to that of Europeans. While straight, the hair retains its original dark colour. As humans do, spiritsof-the-dead grow older but, before they die, ≠Gao N!a rejuvenates them. If children die, however, they remain children within the other world (Marshall 1999: 28). Indeed, while in form they resemble their human selves, in substance, spirits-of-the-dead are similar to air. One G/wi San described the spirits-of-the-dead to ValienteNoailles (1993: 197): ‘Spirits are a kind of air and if the air is too dry, as right now, it may be because somebody has died, a bad person who will bring on drought’ (cf. Marshall 1999: 27 on the !Kung association of spirits with air). While the similar resemblance between the spiritual form and the previously lived body makes the spirits-ofthe-dead recognizable, they are invisible to most people and can only be seen by powerful potency-owners, usually during the Great Dance. Richard Lee (1979: 107) describes the interplay between form and substance in the way spirits-of-the-dead are perceived by the !Kung: The healers in trance see the //gangwasi10 in a variety of forms. To some they look like real people. You can touch them and feel their flesh. To others, they appear like smoke, transparent and ephemeral. One healer described them as having only one leg, standing in midair. In many ways then, life after death is a continuation of the activities of a mortal existence but with a certain sense of loss. This sense of loss seemingly becomes more pronounced with time and their eternal existence is sometimes frustrating to the spirits-of-the-dead who become bored. This is particularly the case regarding sexual partners. As on earth, spirits-of-the-dead have spouses. Importantly, however, if they tire of their wives or husbands, the //gauwasi may take an attractive young mortal as an additional lover or spouse by killing them and taking their soul back to the spirit world. Indeed, the spirits-of-the-dead seem to be preoccupied with stealing the souls of people, not only to be their lovers but also for companionship and seemingly out of anger. Amongst, the G/wi, for example: The spirit of the deceased is believed to be malign and resents being deprived of the company of his family and band and will revenge himself on anybody whom he can catch. He will also try to catch somebody from his family or band to keep him company. The ghosts of people who have died in old age are less dangerous as they have regarded the approach of death calmly and are ready to die when they must. But the ghost of a younger person will, if you come near his recent grave, “catch you, blind your eyes and rob you of your spirit so that you wonder lost in your own country, not knowing your wife or your children”, as one informant stated (Silberbauer 1965: 102). Typically, spirits-of-the-dead carry out their mischief at dances where they attempt to abduct the souls of those attending. The !Kung say that it is the firelight at the Great Dance and the singing and clapping that attracts the spirits-of-the-dead (Marshall 1969) and that they lurk in the shadows just beyond the edge of the fire, watching and waiting for an opportunity to strike. In addition to their desire to ‘catch’ members of their family and band and to abduct lovers from dances to take with them to the spirit world, the spirits-of-the-dead also carry out the orders of one of the gods. Amongst the !Kung, they obey ≠Gao N!a but sometimes the //gauwasi also take orders from the lesser god, //Gauwa. Their principal task under these orders is to afflict people with illness and disease. They do so by shooting invisible arrows of sickness into the people who attend the Great Dance. 93 As they bring illness and disease and because they steal the souls of the living, the //gauwasi are generally feared and there is a strong avoidance behaviour associated with recent graves amongst the Kalahari San groups because the spirits-of-the-dead are believed to lurk there. As Silberbauer (1965: 102) points out for the G/wi, the corpse is buried in a squatting position in a grave and the death is mourned with loud lamentation for three days. All the personal possessions of the deceased are broken and placed on the grave to mark it and warn others to avoid the area. Then the G/wi abandon the place and they do not return while the living members of the group still remember the deceased. Although, as I have mentioned, the ethnographic material on southern San beliefs concerning spirits-of-thedead is not as extensive as the material collected from the Kalahari groups, available evidence suggests that the southern groups held similar beliefs. The //Xegwi of north-eastern South Africa, for example, appear to have believed in dualistic deities. When a person died their soul was thought to have travelled to the greater of these two deities, /a’an (Potgieter 1955: 29). Moreover, it appears that the //Xegwi greatly feared spirits-of-the-dead. Eward Potgieter (1955: 18) comments that it “is evident that the remaining relatives fear the spirit of the deceased, except in cases where he or she was a close beloved relative. Other spirits, however, may return to worry the living and may do so when funeral rules are not observed.” In addition to the // Xegwi, other southern San groups also appear to have held similar beliefs about the dead. A remarkable narrative given by Diä!kwain (Bleek and Lloyd 1911: 364371) concerning two encounters with spirits-of-thedead shows that the /Xam believed in these beings11: We buried my wife in the afternoon. When we had finished burying her, we returned to the home of my sister, Whai-ttu1 and the other people, whence they had come forth. They had come to bury my wife with me; and we went away, crossing over the salt pan. And we perceived a thing which looked like a little child, as it sat upon the salt pan, seeming as if it 94 sat with its legs crossed over each other. And my sister, Whai-ttu, spoke, she questioned us: “Look ye! What thing sits yonder upon the salt pan? It is like a little child.” And !Kweiten ta//ken [another sister] spoke, she asked us: Why is it that this thing is truly like a person? It seems as if it had on the cap which Diä!kwain’s wife used to wear.” And my sister, Whai-ttu, spoke, she answered: “Yes, O my younger sister! The thing truly resembles that which brother’s wife was like.” It did thus as we went along, it seemed as if it sat looking (towards) the place from which we came out. And //Ku-an spoke, she said: “The old people used to tell me, that the angry people will wont to act thus, at the time when they took a person away, they used to allow the person to be in front of us, (so that) we might see it. Ye know that she really had a very little child, therefore, ye should allow us to look at the thing which sits upon the salt pan; it strongly resembles a person, its head is there, like a person.” And I spoke, I said: “Wait! I will do thus, as I return to my home, I will see, whether I shall again perceive it, as it sits.” And we went to their home. And we talked there, for a little while. And I spoke, I said to them that they appeared to think that I did not wish to return (home); for the sun was setting. And I returned on account of it. I thought that I would go in the same manner as we had come; that I might, going along, look whether I should again perceive it, as it sat. Going along, I looked at the place, where it had sat; because I thought that it might have been a bush. I saw that I did not perceive it, at the place were it had sat. And I agreed that it must have been a different kind of thing. For my mothers used to tell me that, when the sorcerers are those who take us away, at the time when they intend to take us quite away, that is the time when our friend is in front of us, while he desires that we may perceive him, because he feels that he still thinks of us. Therefore, his outer skin2 still looks at us, because he feels that he does not want to go away (and) leave us; for he insists upon coming to us. Therefore, we still perceive him on account of it. My sister’s husband, Mansse3, told us about it, that it had happened to him, when he was hunting about, as he was going along, he espied a little child, peeping at him by the side of a bush. And he thought: ‘Can it be my child who seems to have run after me? It seems to have lost its way, while it seems to have followed me.’ And Mansse thought: “Allow me to walk nearer, that I may look at this child (to see) what the child (it) be.” And Mansse saw that the child acted in this manner, when the child saw that he was going up to it, that he might see what child it was, he saw that the child appeared as if it feared him. The child sat behind the bush; the child looked from side to side; it seemed as if it wanted to run away. And he walked, going near to it; and the child arose, on account of it. It walked away, looking from side to side; it seemed as if it wanted to run away. the skin. In both encounters described in this narrative, it appears that the apparitions are not true spirits-ofthe-dead. Instead, they are encounters with the person’s ghost before it has been transformed into a spirit-ofthe-dead. These ghosts resemble the external physical form of the living persons because they have retained the external skin. The ghosts are child-like, seemingly confused, fearful of humans and apparently harmless. These ghosts appear to have been placed before their relatives by the true spirits-of-the-dead, who are variously identified in the narrative as the ‘angry people’ and the ‘sorcerors’ (!gi: xa, shamans). It is the spiritsAnd Mansse looked (to see) why it was that the of-the-dead, particularly those who are shamans in the child did not wish him to come to it; and the spirit world that, according to Diä!kwain, take the ghosts child seemed to be afraid of him. And he exaway. In Lloyd’s footnote, Diä!kwain states that when a amined the child; as the child stood looking at person dies, they do as the /nū people. The word /nū him. He saw that it was a little girl; he saw that 4 the child was like a person. In other parts (of in /Xam signifies dead, departed or spirit (Bleek 1956: it) it was not like a person; he thought that he 350). Intriguingly, in Diä!kwain’s narrative, he quotes // would let the child alone. For a child who was Ku-an as saying that it was the ‘angry people’ who took afraid of him was here. And he walked on, away the dead. The /Xam word for angry people is !nu while the child stood looking from side to side. And (as) the child saw that he went away from ke (Bleek 1956: 482). While the click is different from it, it came forward (near the bush), it sat down. that of the word /nū, signifying spirit, !nu ke is often used in the Bleek and Lloyd texts to describe ‘dead people’ or In this moving passage, Diä!kwain recounts the death ‘spirit people’ (Hollmann 2002: 563). It is clear then that, and burial of his first wife, Mietche, who died of illness in this passage, Diä!kwain explains that upon dying, a in about 1863 (Bleek 1932: 249, cf. Deacon 1996: 32ff). person’s ghost or soul is taken by the spirits-of-the-dead. After burying her, Diä!kwain, together with his two sisThe spirits-of-the-dead, moreover, are described as anters, encounters an apparition that resembles his wife. gry and, according to Diä!kwain, they are also shamans. Indeed, the ghost’s identity is partly determined by the cap that it wears, which is the same as the one worn by Diä!kwain’s observation that it is shamanic spirits-ofDiä!kwain’s wife when she was alive. Yet, at the same time, the-dead who send the apparitions and who come to the apparition is child-like in form. Later on, Diä!kwain claim the person after death is important because it is the returns to the place of sighting but the apparition has principal task of living shamans to fend off the spiritsleft. Diä!kwain also describes an encounter between a of-the-dead. Indeed, as I noted earlier, it is only the n/omspirit and his brother-in-law, Mansse, who was married kxaosi amongst the !Kung who can even see the spiritsto a third sister of his, /A-kkumm. Mansse encounters of-the-dead and so only they can ward off their attacks. the spirit that, as with the spirit of Diä!kwain’s wife, in The range of techniques that shamans deploy against some aspects is child-like in physical form but in other the //gauwasi appears to be similar amongst the various ways it is not like a human. Indeed, in one footnote Kalahari groups. Shamans may talk, argue or plead with recorded by Lloyd, Diä!kwain explains that it is the part the //gauwasi (Lee 1979: 103, Valiente-Noailles 1993: of the apparition that remembers its relatives that still 198). Failing this, they shout insults at the spirits-ofresembles its human form. This part is identified with the-dead and may even resort to physical means. Often, 95 for example, at a dance a shaman will run out into the The ambivalent attitude towards the //gauwasi is not as darkness beyond the firelight, throwing sticks or hurling contradictory as it might at first appear. Given that the insults at the spirits-of-the-dead. These insults are di- spirits-of-the-dead are often recently deceased relatives, rected at the anatomy of the //gauwasi, especially their it is understandable that the San have positive feelings sexual organs and bodily secretions. Marshall (1999: 87) for them. Moreover, the !Kung appear to have a deep records some of these insults as “uncovered penis”, understanding of the motivations behind the actions of “hyena penis”, “lion penis”, “thrown-away penis” and the //gauwasi and, at times, this understanding produc“filthy face”. Potency-owners may also use their divin- es sympathy for the plight of the spirits-of-the-dead. ing switches to flick away the arrows of sickness shot by According to Lee (1984: 109), one of the best explathe //gauwasi. In addition, the shamans remove any in- nations he was given for the actions of the spirits-ofvisible arrows of sickness that the //gauwasi shoot into the-dead, came from a !Kung woman called Chu!ko: people by ‘sucking’ or ‘snoring’ them out of the patient. “Longing,” she said, “longing for the living is The shamans may then expel the sickness out the back what drives the dead to make people sick. When of the neck or sneeze it out, transmitting it back to one they go on the road that leads to the village of the of the spirits-of-the-dead, who then takes it away (Bar//gangwasi they are very, very sad. Even though nard 1992: 252). Amongst the G/wi, “the men absorb they will have food and company and everything 12 they need there, they are not content. They miss a small amount of evil and then go out into the night their people on earth. And so they come back to and cough it out of themselves” (Silberbauer 1965: 99). us. They hover near the villages and put sickness into people, saying, ‘Come, come here to me.’” Ethnographic observations such as these have led reChu!ko’s view was corroborated by others, she searchers to characterize the relationship between the “made the process of death a struggle between two spirits-of-the-dead and living shamans as an antagonistic loving sets of relatives, one living and the other and martial one. In spite of such observations, San attidead, each wanting the individual for themselves.” tudes to the //gauwasi seem to be ambivalent. Recently, Bradford Keeney (2003) has suggested that the aggresWhile the !Kung understand the motivations of the sive relationship between shamans and the spirits-of//gauwasi and even have sympathy for them, they the-dead has been overemphasized amongst the !Kung. nevertheless are dangerous because they bring death, He refers to interviews that he has conducted with indisease and disorder and it is these three things that dividual n/omkxaosi where they describe how they apthe !Kung fear more than the //gauwasi themselves proached and were assisted by a spirit-of-the-dead, usu(Marshall 1999: 30). This distinction between the beally a close relative. Instead of fearing the spirits-of-theings themselves and the things that they bring from dead, Keeney suggests that the relationship is far more the spirit-world helps to explain the ambivalent atco-operative. Other ethnographers also mention the amtitudes that San have towards the spirits-of-the-dead. bivalent attitudes towards the //gauwasi. Marshall (1999: 30), for example, noted that amongst the !Kung, the respect word for spirit-of-the-dead, /airisi, was, in fact, In spite of such ambivalent attitudes, it is abundantly rarely used and, in spite of what people said about the clear from the widespread ethnographic material on dangerous and fearsome nature of the //gauwasi, their spirits-of-the-dead that they are a chaotic influence; behaviour towards the spirits-of-the-dead was usually they are the bringers of death, they are the purveyors of casual, suggesting that they were not greatly feared. This disease and they are the principal agents of disorder in was not, however, the case with attitudes and behaviour San cosmology. It is at the Great Dance that they contowards the greater and lesser gods, who were always re- duct the work of death, disease and disorder; they cause spected and feared and who were never treated casually. death by stealing the souls of young people; they bring 96 disease on the orders of the gods in the form of tiny invisible arrows of sickness; and in so doing they create disorder in San social life. It is the task of the owners-of-potency to negotiate with and combat the spiritsof-the-dead and, in so doing, to restore life, health and order. Potency-owners restore life by travelling to the spirit world, where they retrieve the souls that have been abducted; they restore health by removing the arrows of sickness that the spirits-of-the-dead shoot into people from the edge of the firelight and they restore order by repairing conflicts between real people. The functions of real shamans and shamanic spirits-of-the-dead are thus inverted and it is in the embodiment of these opposite functions, I argue, that we may find criteria with which to distinguish between images of real shamans and depictions of spirits-of-the-dead in rock art. The Eldritch Images So far we have seen that it is difficult to establish recognizable differences between the world of the living and the world of the dead. Both worlds are similar in appearance and the inhabitants—people in one and spirits-of-the-dead in the other—of both worlds do the same things such as dancing and singing. A further complication is that spirits-of-the-dead can be shamans and real shamans talk about their experiences in altered states of consciousness as dying (Katz 1982); the blurring of death and shamanic experience thus makes it difficult to distinguish between living shamans and spirits-of-the-dead. Both the spirit and real worlds, moreover, are inextricably linked through the invisible threads that act as conduits for shamans and spirits-of-the-dead to move along. As the two worlds are so similar and interlinked, it is difficult to extract criteria from the ethnographic material on the real and spirit worlds that would allow researchers to distinguish between images of shamans and depictions of the spirits-of-the-dead. Possibly, a consideration of what spirits-of-the-dead look like may assist but, as we have seen, San encounters with these beings are often individually specific and such idiosyncratic experiences are not very useful for establishing criteria that could be used for making distinctions in widespread areas of rock art. Recalling one of the comments collected by Lee and reproduced earlier in this chapter, there are, for example, few anthropomorphic images that can be argued to have only one leg in San rock paintings. More widely held observations about their appearance may be more useful but existing ethnographic information is not rich in these descriptions. The interweaving of the two worlds and the similarities between the occupants of those worlds, together with the relative absence in the ethnographic record of widely held ideas about their appearance, make it very difficult to discern between images of spirits-of-the-dead and living shamans in San rock art. Jeremy Hollmann (2003) has recently taken up this point and has suggested that it is very difficult to distinguish between the characteristics of spirits-of-the-dead, living shamans and other spirit-world beings such as the so-called primal time people in San ethnographic. Following Solomon’s suggestion, he argues that images of therianthropic figures, for example, probably refer to all three types of spirit being and are not simply depictions of transformed shamans. Nevertheless, I would suggest that it is, indeed, possible to make a distinction between some of these spirit beings and real shamans. By considering first the distinction between spirits-of-the-dead and the three things—death, disease and disorder—that they bring to the world of the living and second, the embodied nature of these three things, I argue that we may indeed draw a distinction between images of spirits-of-the-dead and images of living shamans in the rock art of Nomansland and, indeed, in other southern African rock art areas. Following Turner’s notion that all disease is disorder and that such disorder is typically expressed in and through the body in any given society, we may hypothesize that disease, death and disorder will find corporeal expression in San rock art. If the spirits-of-the-dead are the carriers of hunger, malnutrition and disease, then we might reasonably expect any potential depiction of such beings to show corporeal abnormalities that mark them as something very different from the 97 Fig. 28. Translucent-white figures sometimes have small features, such as the nasal bleeding and bow in this figure, that are depicted in red. The figure plays a musical bow and this activity, together with the nasal bleeding, indicates that the figure is associated with the shamanic activity of the Great Dance; the unusually grotesque penis and peculiar foot, however, suggest that the figure possibly portrays a more sinister being. Solid=Red; Clear=Translucent White. Scale in centimetres. Fig. 29. Translucent-white figures in Nomansland are instantly recognizable by their distinctive corporeal features. This figure is typical in that it is depicted with emaciated hips and a penis that is exaggerated in size. The figure also has a feline-like head. Approximately 15cm in length. norm. Can one, for example, identify diseased and malnourished bodies in the art? If so, is it possible to determine whether these are images of spirits-of-thedead or images of human beings who have been afflicted by the disease that they send? In answering these questions, however, caution needs to be exercised. It is difficult to ascertain what a corporeal abnormality is in San rock art because, as we have seen in Chapter Two, altered states of consciousness lead to complex somatic experiences that are frequently depicted. Corporeal abnormalities in the art, such as extra digits and unnaturally elongated bodies, may depict various sensations of trance such as polymelia and attenuation. Theoretical prediction thus needs to be supplemented with both ethnographic and painted-contextual information. nounced bodily abnormalities. These images also occur at Storm Shelter. Typically, the images that comprise this category have four important characteristics. First, the most notable aspect of this class of images is the pigment. Typically, they are white in colour but the quality of the pigment differs from the white that is used in other San paintings (See Fig. 22). The white that is most commonly used in Nomansland rock art is a thick white to off-white pigment that is usually opaque. The images that make up the category under discussion, however, are painted in a thin white pigment that is often trans- lucent. While they are almost completely white, other pigment is sometimes used to indicate small features, such as dark red for nasal haemorrhage (Fig. 28). In some cases, the white of the images has a slight reddish tinge. Second, the translucent-white figures are often There is nevertheless a distinctly recognizable catego- clustered together, as at Storm Shelter, in small groups ry of painted image in Nomansland that shows pro- of seemingly related images. These small groups usu- 98 Fig. 30. Translucent-white figures are most commonly depicted over other San rock art images. Nevertheless, in some instances, such as in this example, ‘classic’ San rock art images are painted over the translucent-white figures. This suggests that the translucent-white figures cannot be interpreted simply as a recent chronological development in San rock art; instead their widespread occurrence throughout southern Africa suggests that they need to be seen as a more enduring element of San rock art. Solid represents black, clear indicates white and stipples stands for various shades of red. Scale in centimetres. ally number from three to ten figures but, at some sites, there may be many more within a cluster. Although rare, isolated single figures of this category do occur. When in interactive groups, the translucent-white figures often appear to be more dynamic than many other images within a particular painted panel. They may run with legs outstretched in a non-realistic manner or they may adopt various peculiar postures with arms raised in the air. Indeed, they are very similar in dynamism to images found in the rock art of Arnhem Land, Australia, that have been called Dynamic Images (Smith, Chippindale, and Taçon 2000). Third, the translucent-white images are most commonly—but not always—painted over other images in a panel in Nomansland. When over translucent-white figures, the other San rock paintings are classic in appearance, showing that the two types of images were made concurrently (Fig. 30). The translucent-white images thus cannot be interpreted as a simple chronological development within the San rock painting tradition. Fourth, the translucent-white images are most commonly anthropomorphic and frequently therianthropic. There are only a few examples of animals painted in this pigment in Nomansland. In addition to the therianthropic features, many of the translucent-white images have features such as nasal blood and are in bending-forwards or arms-back-postures, clearly associating them with the Great Dance and San shamanism. The fifth characteristic is the grotesque corporeal aspects of these images. They are often depicted with grossly exaggerated penises, claws in the place of hands and feet, ferocious teeth in the mouth, oversized fingers, emaciated hips and skeletal bodies (Fig. 29 & Fig. 31). Indeed, they appear to be diseased and malnourished. Their bizarre qualities and their translucentwhite colour make these figures readily visible at any Nomansland rock art site and unfortunately they were labelled the ‘weird whites’ when they were first discovered (Vergnani 1999). Subsequently, they have been relabelled ‘Eldritch Images’ (Blundell and Lewis-Williams 2001). Whatever label one attaches to them, they are 99 Fig. 31. A typical cluster of translucent-white figures from Nomansland. Note the claw-like feet, grotesque heads and exaggerated penises. Scale in centimetres. certainly so bizarre that Alice may well have expected to encounter them on her travels in Wonderland. It is particularly the ‘biting jaws and catching claws’ of the Eldritch Images that impress and call to mind the poem of the Jabberwocky in Lewis Caroll’s (1872) Through the Looking Glass and What Alice Found There. in San rock art. Their widespread distribution, both in space and time, suggests that whatever the Eldritch Images represent, it must have been widely shared amongst different San groups. Of course, while they are recognizable across vast areas in southern Africa, there are, nevertheless, slight regional variations amongst the images. The Eldritch Images found in the Maloti MounThe Eldritch Images occur throughout the south-east- tains (the western portion of the south-eastern mounern mountains, outside of Nomansland and are found tains), north of the Orange River, for example, tend to in other rock art areas of southern Africa. A cluster of be smaller and generally more animated than those of at least nine Eldritch Images (Fig. 32) has been found Nomansland. They are often depicted in running posin the Cape folded mountains as far west as the town tures and they are often portrayed as chasing other Elof Uniondale (Laue 1999). In the north-eastern part of dritch Images. They also sometimes have pronounced South Africa, in the Bongani Game Reserve Area, a sin- ears that give them a Mickey Mouse-like appearance. gle eldritch image has been found amongst the many paintings that are found on the granite boulders of the In spite of this widespread distribution, Eldritch Imarea (Challis and Hampson 2000: Pers. Comm., for a de- ages are found in frustratingly few painted contexts scription of the rock art of the area see Hampson et al. that provide explicit clues to their interpretation beyond 2002). Farther north, in the domed-shaped shelters of their obvious shamanic associations (Fig. 33). They are the Matopos mountains of Zimbabwe, Eldritch Images rarely depicted interacting with anthropomorphic or have also been observed (Mguni 2003: Pers. Comm.). animal images except for other eldritch figures. NevAs I mentioned in Chapter One, the rock art of the ertheless, at a few sites in the south-eastern mountains, Matopos appears to be at least 10, 000 years old and this Eldritch Images are found in informative contexts that suggests that the Eldritch Images are an enduring image provide further clues as to what they represent. Here 100 Fig. 32. A cluster of Eldritch Images from the Uniondale area, several hundred kilometres west of Nomansland. Scale in centimetres. I discuss four such sites. Two are from outside of the Nomansland Core Study Area, while two are from within the area. The first site, a shelter visited by Stow in the second part of the nineteenth century, is from the Maloti Mountains and has paintings of Eldritch Images that are typical of the western section of the south-eastern mountain complex above the Orange River and which are thus slightly different from those of Nomansland. The second is located in the Drakensberg and thus the eastern section of the south-eastern mountains but is far to the north of Nomansland. The third site is from Nomansland in a valley very close to the ravine in which Storm Shelter is found. The fourth and final site that I consider here is Storm Shelter itself. RARI-RSA-GEN1 The first site, RARI-RSA-GEN1, is located on the western side of the south-eastern mountains, in the Maloti range and is situated just north of the Orange River, not far from Nomansland. The site was visited by George William Stow at some point between the years 1867-1882. He copied some of the images at the site, which he later published. Subsequently, it was visited by Dorothea Bleek and in more recent times by a number of southern African rock art researchers. The site is best-known for a bizarre creature that is widely accepted to be a depiction of a rain-animal (Fig. 34). Both northern and southern San groups perceive rain as a dangerous creature from the supernatural world (Lewis-Williams 1981: 103ff). Moreover, there is a widespread distinction made between male and female rainanimals—the male being far more violent and dangerous. In order to make it rain, /Xam San ritual-specialists entered the supernatural world through trance. In the spirit world, they sought the rain-animal and lured it back to the world of the living. In order to soothe the ferocious creature, the potency-owner would sometimes wave an aromatic herb, buchu ((Barosma betulina, B. crenulata and B. serratifolia), under its nose. Once the creature was calmed, they led it from the supernatural world into this one where they either slaughtered it or cut it so that its blood and milk would fall as precipitation. Images of large creatures that are composite animals and are not identifiable as a particular existing species are often depicted in contexts that point to their identity as rain-animals and such creatures are a readily identifiable category of images in San rock art throughout the south-eastern mountains (see Wood101 eral anthropomorphic figures in thin translucent-white pigment. The technical drawing used in Figure 34 only shows four such figures but a black-and-white photograph reveals that there are at least three more such figures which are faded. These translucent-white figures are typical of the Eldritch Images found in the Maloti Mountains. As I have already mentioned, they are usually smaller than those found in Nomansland, tend to be even more animated and have a Mickey Mouselike appearance. The Eldritch Images in this panel are physically very different to the other four anthropomorphic images, one of which is clearly a shaman. Fig. 33. Composite image of Eldritch Images walking on a thread that connects the real and spirit worlds. Eldritch Images are rarely found in informative contexts. When they are found in such contexts, as in this image, they engage in activities that are also shamanic practices. The occurrence of these Eldritch Images in such close proximity to a rain-animal and the difference between them and the other four anthropomorphic images provides useful insights into what they represent. According to /Xam San, the spirits-of-the-dead are said to live with the rain-animal and sometimes they even ride the animal. /Han≠kasso told Bleek and Lloyd (1935: 304-305) that a person wishing to make rain must first speak to the dead “who are with the rain” and then to the rain itself. He said that the spirits-ofthe-dead had bound the rain as with a horse’s reins and that it was the task of the potency-owners to loosen the thong, thus allowing the rain to fall. The potencyowner then broke the “rain’s ribs” and scattered them over dry land. Their close association with the rainanimal at RARI-RSA-GEN1 suggests that the Eldritch Images here may in fact depict spirits-of-the-dead but such arguments by association are usually the weakest forms of argument in rock art research. At the next site that I discuss, however, the evidence that the Eldritch Images depict spirits-of-the dead is more extensive. house 1992 for a survey of images of rain-animals). At site RARI-RSA-GEN1, the rain-animal is slightly flaked, but much of the painting is still visible. Clearly evident is the animal’s large penis (the creature only has a single front and a single back leg), signifying that it is a male rain-animal. Its large teeth are formidable and the human figure lying prone beneath its path is testa- Junction Shelter ment to its ferocity. Leading the animal is a figure that The second site that I discuss here was recorded by is clearly identifiable as a shaman from the nasal blood Harald Pager and published in 1971. In a remarkable and dried animal bladders in the hair. This shaman fig- effort, Pager documented eighteen rock art shelters in ure is painted in a grey-black pigment. Two other fig- the Didima Gorge of the northern Drakensberg. His ures behind the rain-animal are painted in similar pig- recording technique was to make life-size black-andment. In addition to these three grey-black figures and white photographs of the shelters. Spending some two the one lying prone beneath the rain-animal, are sev- years in the gorge, working in the shelters from the ac102 Fig. 34. Rock art panel from RARI-RSA-GEN1. A rain-animal is clearly associated with figures depicting shamans as well as Eldritch Images. Clear=Translucent White; Solid=Red; Stipple=gray-Black. Scale is in centimetres. tual paintings, he then painted in oil-based pigments the rock art images in colour onto the black-and-white photographs. Although there are occasional errors in his recordings, both in the Didima Gorge and elsewhere (see, Blundell and Eastwood 2001, King 1998), Pager’s copies remain some of the most accurate representations of San rock paintings. One particularly remarkable rock art shelter where Pager worked is situated at the confluence of two rivers in the Didima Gorge and is thus called Junction Shelter. In this shelter, there are several painted rock panels. On one of these panels are many images of fat-tailed sheep and it is in this Fat-Tailed Sheep Panel that another informative context exists that helps to shed further light on the Eldritch Images. appears to have two distinct parts (Fig. 35). The eastern area of the panel, Part A, is comprised principally of images in various shades of red ochre while the western area of the panel, Part B, is made up of images that are painted in a thin, translucent, white pigment. Part A may be further analytically subdivided into five clusters of images: The Fat-Tailed Sheep Panel is 3.95m x 1.2m in size and Cluster 2 is located just beneath Cluster 1. In this clus- Cluster 1 is situated in the top right-hand section of Part A. In this cluster are depicted 11 anthropomorphic figures— holding bows—in running postures. They are painted in a red pigment. Some remains of pigment suggest that previously there were several more anthropomorphic figures in the cluster. In addition to these anthropomorphic images, there is also a single painting of an eland. 103 Fig. 35. Harald Pager’s copy of the Fat-Tailed Sheep Panel at Junction Shelter. A spatial separation is evident between the red figures of Part A an ter are six anthropomorphic figures, which are smaller and have thinner bodies than those of the first cluster. These images are also red in colour. The remains of an indeterminate antelope are also present. In addition, there are 11 anthropomorphic images that are painted in a thin and translucent-white pigment. These figures are thin and small in comparison to the red images of Cluster 1. There is also a single running anthropomorphic figure that is very similar to those of Cluster 1. ing from their hunting bags that they carry over their shoulders. Flywhisks are items used only at the Great Dance (Blundell 1993, Marshall 1969: 358, 1999: 69), while some also have infibulated penises (a common—but poorly understood—feature where a bar is depicted across the penis), suggesting that this is not merely a simple depiction of hunters returning from a successful hunt, but that the image makes reference to the release of supernatural potency from dead antelope, so necessary for the performance of the dance. Cluster 3, immediately below Cluster 2 comprises 11 anthropomorphic figures in running postures, Cluster 5 may be further divided into Section 1 and Secwhich are red in colour. A single eland is also painted. tion 2. Section 1 comprises at least 22 figures and the remains of several indeterminate images. Most of the imCluster 4, to the left of Cluster 2, comprises a fur- ages are red but three are in a yellow colour. One of the ther 15 anthropomorphic figures, this time in walk- red figures holds a bow with an arrow in position, while ing and not running postures. Over their shoulders, some figures hold sticks. At least two red figures are they appear to carry the carcasses of killed antelope. depicted in the arms-back position, a posture associated Some of the figures have divining switches protrud- with the Great Dance. Section 2 (Fig. 36) is a composi104 nd the translucent-white images of Part B. The panel is 3.95 metres in length and 1.2 metres in height. tion that is roughly arranged in three horizontal rows, situated one above the other. The top row comprises 17 anthropomorphic figures painted in a light red colour, all with an erect or semi-erect penis. In six of these figures the penis is infibulated. Most of the anthropomorphic figures also hold a staff in one hand and they appear to depict a line of dancers. A little to the right of this top row are three anthropomorphic figures, painted in a darker red colour, that carry hunting bags. Below the row of 17 dancing figures are a further 18 anthropomorphic images. Some of these are painted in a darker red than that of the figures of the top row while others are of a similar colour. One anthropomorphic image in this middle row is in the hand-to-nose posture, another reference signifying that this scene depicts a dance. To the right of the middle row, is an image of a woman with a baby on her back. She holds a digging staff with bored stone. There are also three other human figures in red pigment, one of which holds a long staff and has a divining switch protruding from his shoulder. Below the middle group, the lowest row comprises four running figures that hold bows. These are smaller than the figures in the top and middle row. To the left of these four images, is an anthropomorphic figure in red and the remains of what was probably a similar figure. To the right of the four running figures are at least two more red human figures and the remains of a third. From the postures, such as hand-to-nose and the line of men with dancing sticks, the images in Section 2 clearly portray a dance. Interspersed amongst the figures of the dance are at least ten anthropomorphic figures, painted in a thin white translucent pigment. Where superimposed, these figures are over the red images. Most of these white images appear to be normally proportioned. One image (A), however, is elongated, which as I have mentioned, is a typical experience of the altered states of consciousness of the dance. At the right edge of the 105 Fig. 36. Detail of Cluster 5, Section 2. The Eldritch Images are interspersed between the red dancing figures. ticipants. This figure then possibly represents a ghost or a spirit-of-the-dead. Another figure in translucent white that is possibly a representation of a spirit-of-the-dead is opposite the woman with a child on her back (C). The digging stick with bored stone that she holds is typically used to dig up roots and bulbs but also has a ritual function as it is used amongst the /Xam to bang on the ground in order to call up the spirits-of-the-dead (Bleek 1935: 35-36, 41-43). This image could thus be read as a woman, very much associated with a dance, calling up ple by the spirits-of-the-dead. The small size of the im- a spirit-of-the-dead, represented by the figure in white. age is also intriguing because it is clear from the /Xam testimonies that ghosts are child-like in stature and ap- If these two images are representations of either spir- topmost row, just behind the dancers, is a short figure that is clapping but only has three fingers (B). Moreover, the figure’s head is oddly shaped and the neck is not indicated. Instead, the neck and head of the figure merge into one with the abdomen. The face of this figure also lacks the rudimentary features so common in most San rock paintings. Parts of the anatomy are thus recognisable as human but others are not. Indeed, the figure brings to mind Diä!kwain’s comments about the indistinct qualities of ghosts that are placed before peo- pearance. The position of this figure at the edge of the dance is also important. As we saw in the ethnographic discussion about death, disease and disorder, spirits-ofthe-dead are attracted to the dance by the sounds of singing and the flickering firelight. It is from the edge of the dance, that they shoot arrows of sickness into the par106 its-of-the-dead or ghosts about to become spirits-ofthe-dead, then the possibility arises that all the thin, translucent-white anthropomorphic figures in the panel depict such beings. Indeed, in Part B of the Fat-Tailed Sheep Panel, which is to the right of Part A, there are significantly fewer red anthropomorphic figures, but far more translucent-white images. The majority of these white images are of fat-tailed sheep, which were introduced to South Africa from the north about 2000 years ago (Henshilwood 1996, Mitchell 2002: 227ff., Sealy and Yates 1994). In addition to the fat-tailed sheep, there are several peculiar animals in a similar translucent-white pigment in Part B. There are, for example, a fat-tailed sheep with two tails, an indeterminate creature that appears to have wildebeest qualities, a small zigzag snake with an antelope head and several indeterminate creatures that appear to be conflations of antelope and fat-tailed sheep characteristics. There are also two paintings of baboons. In addition to the animalimages, there are also many running anthropomorphic figures in translucent white. In contrast to the red images, both in Part A and Part B, the translucent-white images are almost stick-like in appearance. Moreover, while there are three red anthropomorphic figures in running postures at the bottom of Part B, most red figures in this area are less animated than the white figures. It is interesting to note here that there is a general, but not exclusive, organisational separation between the red ochre images and the translucent-white images in the Fat-tailed Sheep Panel. Most of the red ochre images are in Part A (the eastern side) of the panel while most of the translucent-white images are in Part B, (the western side) of the panel. The red and white images overlap slightly in the central area of the panel. Another interesting feature of Part B is that it is situated on a rock surface that is clearly demarcated by an upper and lower ledge. Towards the west, the bottom ledge begins to rise to meet the upper step. In so doing, the surface on which the images of Part B are painted becomes tapered. It is into this narrowing area, towards the west, that most of the translucent-white animals appear to be moving. Given that features of the rock surface are often incorporated into San rock paintings and that clear ethnographic and painted evidence shows that the world of the spirits was thought to lie behind the rock surface (Lewis-Williams and Dowson 1990), it is possible that the tapered area was thought of as a conduit into the world behind the rock surface. The broad division between the two areas of painting may imply a distinction here between the real and spirit worlds, with the world of living in the east section of the panel and the world of the dead in the west section and it is at the dance where the two worlds overlap. Between these two worlds, a number of figures, both shamans and spirits-of-the-dead move backwards and forwards. The Eldritch Images at Junction Shelter and their painted context, more so than the images at RARI-RSA-GEN1, suggest that Eldritch Images depict spirits-of-the-dead. In Nomansland itself, however, there are two shelters that provide further convincing evidence for this suggestion. RARI-RSA-MEL6 A third site that offers contextual information that assists in identifying what the depictions of Eldritch Images represent comes from the Nomansland area. Site RARIRSA-MEL6 is located in a valley close to the ravine in which Storm Shelter occurs. The site is well known for a remarkable and prominent figure of a shaman (Fig. 37). Indeed, Dowson includes this figure as one of his preeminent figures. The image has an articulated face with nose, upper and lower lips and chin. Four red stripes are painted on the side of the face; two are painted on the forehead area while a single line is painted from the nose and another one from the mouth. These lines have been previously interpreted as the nasal blood and mucus smeared back across the face, so characteristic of the behaviour of San potency-owners (Lewis-Williams and Dowson 1989). The figure has a unique fezlike headdress. In addition, there are two streamer-like appendages attached to the top of the headdress. On the figure’s back is a bag with three protruding objects that are probably divining switches. Prominent figures such as this image at RARI-RSA-MEL6 are a characteristic feature of some rock art sites in Nomansland and in Chapter Six, I discuss these images in more detail. Surrounding the prominent figure are many images that appear to represent various items of San material culture. Some of these appear to be bags and others represent digging sticks with bored stones affixed to them. 107 Fig. 37. The Eldritch Images of RARI-RSA-MEL6 appear to lurk at the edge of the dance. The figures at this site, however, are not painted in translucent-white pigment. Scale in centimetres. Still other items are unidentified. In addition to the material culture items there are twenty anthropomorphic figures in seated postures, surrounding the prominent figure. Eight of these anthropomorphic images, to the right of the prominent figure, are in clapping postures. These clapping postures and the blood smeared over the prominent figure’s face all show that this panel depicts a Great Dance. In addition to the eight clapping figures, there is a single peculiar image to the right of the prominent figure. This peculiar image is anthropomorphic but has arms and legs that end in claws rather than fingers and toes. To the left of the prominent figure are a further three anthropomorphic images with feet and hands that end in ferocious claws. One of these clawed figures also has two fearsome teeth projecting from the lower jaw (Fig. 38). These figures with jaws that bite and claws that catch are mostly positioned between the prominent figure and the other seated human figures that are depicted with limbs that terminate normally; they are thus placed at the edge of the dance circle as it were. The 108 placement of the figures and their ferocious claws and, in one example, teeth, led Thomas Dowson (1994) to argue that these images depict spirits-of-the-dead. As I mentioned at the beginning of this chapter, however, Dowson did not go further and suggest criteria for identifying similar representations of spirits-of-the-dead. Importantly, none of the images that Dowson argues are depictions of spirits-of-the-dead at RARI-RSAMEL6 are painted in a translucent-white pigment. Instead, the bodies of these figures are painted in various shades of red pigment, while the figure with the large, pointed teeth has an abdominal area painted in a thick white/off-white pigment. The claws on the ends of the hands and feet, as well as the teeth of the one image, are also painted in white. The use of two separate colours—one for the majority of the body and the other for the jaws and claws—at this site, is a technique that is also occasionally found in Eldritch Images painted in translucent-white pigment. Indeed, at RARI-RSA- Fig. 38. Detail of one of the Eldritch Images at RARI-RSAMEL6. Notice the ferocious jaws and sharp claws. Solid=Red; Clear=White/Off-White. Scale in centimetres. Fig. 39. Compare this figure from RARI-RSA-MEL10 with Fig. 38. This figure is painted in a translucent-white pigment with red claws. Scale in centimetres. MEL10, a site situated in a valley adjacent to the one in which RARI-RSA-MEL6 is found, is a painting of a peculiar running figure in the thin, transparent-white pigment typical of the majority of Eldritch Images (Fig. 39). The figure, however, has claws on one hand and on one foot that are painted in a red pigment. San artists, then, used similar techniques at RARI-RSA-MEL10 and RARI-RSA-MEL6 and the only difference between the images with claws at the two respective sites is in the use of colour—at one, red was used for the body and white for the claws; at the other white was used for the body and red for the claws. San use of pigment in the rock paintings of the south-eastern mountains is generally quite variable and attempts to work out a pattern of symbolic association of colours by a careful consideration of images and ethnographic evidence have proved futile (Lewis-Williams 2000: Pers. Comm.). It is surprising then but not inexplicable that the Eldritch Images at RARI-RSA-MEL6 should be painted predominantly in red. In spite of the differences in colour, the images at RARI-RSA-MEL6 have similar grotesque and bizarre physical characteristics to the image at RARI-RSA-MEL10 and, indeed, to the other translucent-white images in Nomansland that they must be considered as Eldritch Images. As Dowson argued, these Eldritch Images at RARI-RSA-MEL6 appear to portray spirits-of-the-dead and, by extension, we may infer that the other Eldritch Images of Nomansland, painted in translucent white, represent similar beings. Storm Shelter The fourth and final painting that I discuss here that sheds light on the Eldritch Images is Storm Shelter itself. In Chapter One, I pointed out that the site is very much the equivalent of an archaeological type site and this is certainly the case for the Eldritch Images. It is in the main panel at Storm Shelter that the most informative context for the Eldritch Images—so far discovered—is found. As described in Chapter One, close to the large feline at the top left of the panel is a cluster 109 Fig. 40. The cluster of Eldritch Images from Storm Shelter. The oral emanation from the shaman flows into the face of one of the Eldritch Images. Scale in centimetres. of seven images in translucent-white pigment (Fig. 40). Small parts of some of the figures are painted in light red pigment. They are typical of the Eldritch Images of Nomansland in that they are therianthropic. The images cluster together and appear to be associated with one another. Four of the figures are arranged in a linear fashion, moving from right to left, in what appears to represent a progression through the various phases of a dance. The rightmost figure stands upright and holds a dancing stick. The figure to the left of this one also stands upright but is in the arms-back position, which is adopted by Kalahari potency-owners when they ask God to place the supernatural potency into their bod- tency advances up the spine and explodes in the head. Above and to the left of this line of four dancing figures are a further three Eldritch Images. One of these figures bends forward while another stands upright. The third figure has a strange, feline-like head and a skeletonized pelvic area as well as an unusual emanation from the foot area (Fig. 41). Above the figure with the skeletal frame is a tall anthropomorphic figure that is clearly not an eldritch image (Fig. 40). This figure has white lines smeared back from the nose across the face. These lines almost certainly represent the mucus and possibly blood that emanates from the noses of par- ticipants at the dance that is smeared back across the ies (Lewis-Williams 1981: 88). The figure to the left of face. The figure also holds a divining switch, indicatthe arms-back image bends over as the potency begins ing that it is a potency-owner. From this figure’s mouth to ‘boil’ in the stomach area, causing painful cramps. a red line originates and apparently moves downwards The next figure in the progression has an antelope toward the cluster of Eldritch Images but, noticeably, head and is transforming into animal-form as the po- to the figure with the secretion from the foot. The line 110 Fig. 41. Detail of Eldritch Image from Storm Shelter. The foot emanation can clearly be seen in this photograph. See Fig. 40 for scale Fig. 42. A possible bisexual Eldritch Image. About 15cm in height. reaches the eldritch figure and zigzags across its face, the case that what is being depicted in this section of ending in the region of the figure’s mouth and nose. Storm Shelter is a potency-owner coughing or sneezing While nasal blood and mucus are commonly represent- out sickness and sending it back to a spirit-of-the-dead, ed in the San rock art of Nomansland and the south- represented by the eldritch image, with the foot secreeastern mountains more generally, this is the only im- tion. By extension, then, we may suggest that all the Elage—so far discovered in Nomansland—with a clearly dritch Images in this panel represent spirits-of-the-dead. oral emanation. The red colour might at first suggest blood but such a condition would obviously be danger- Body and Interpretation in San Rock Art ous and would require treatment and in such a scenario, The contextual information from the four shelters that one would expect the image to be depicted as a patient I have discussed strongly suggests that the Eldritch Imand not as a shamanic healer. Instead, the answer to ages represent spirits-of-the-dead. On their own, howwhat the red discharge depicts is to be found in the ever, the evidence from the four sites may not be enough testimonies of Kalahari and /Xam San healers, who, to convince the most sceptical reader. Yet, a consideraas we saw earlier, cure patients by snoring out sickness tion of how various San groups perceive disease, death through their noses and then expelling the sickness and disorder as embodied in the spirits-of-the-dead through the back of their necks or by coughing it out. greatly advances the limited informative contexts. Time Once they expel the sickness, it goes back to the spirits- and again, even when they do not occur in informative of-the-dead, from whence it came. It is thus likely to be contexts, the Eldritch Images have bodily characteris111 tics that clearly mark them out as something different from the other images within a panel. Their skeletal and grotesque appearance is particularly important because they display the physical characteristics of hunger and malnutrition that, as we have seen, are the existential preoccupations of Kalahari San. Indeed, their appearance is of diseased and ill beings yet they are not found in contexts which suggest that they are patients being healed by shamans. If they are not afflicted patients but nevertheless are diseased, ill and malnourished, then they are almost certainly representations of spirits-of-the-dead, who carry these afflictions with them. ethnographic accounts, widely held amongst different San groups, of the spirits-of-the-dead (but see Ouzman and Loubser 2000 for a slightly different interpretation.). In this chapter we have seen how by using concepts of corporeality and embodiment, we can answer Solomon’s challenge to San rock art research and how we can make a distinction between images of shamans and images of spirits-of-the-dead. As such, embodiment offers a theoretical tool with which to extend hermeneutic efforts to establish the meanings of certain images when there is either a lacuna of direct ethnographic evidence or a paucity of informative painted contexts in San rock art. Furthermore, the grotesque shape and size of the Having established its credibility as a tool for hermepenis on some of the Eldritch Images remind us of neutic analysis, we are now in a position to ascertain Marshall’s observations of the insulting comments the value of embodiment as a toll for the analysis of that !Kung owners-of-potency make about the sexual the social production and consumption of San rock art. organs of the //gauwasi. Indeed, one of the Eldritch Images in Nomansland is possibly bisexual (Fig. 42), having both breasts and a penis. Figures with ambiguous sexual characteristics are extremely rare in San rock art and they are clearly not the norm. By painting an eldritch image with bisexual features, it is possible that the artist or artists were making a statement about the figure’s non-conformity to the norm and as we have seen, spirits-of-the-dead deviate from the corporeal norm of human beings. In this possible bisexual example and in the many other images with markedly bizarre sexual organs, we have further evidence that Eldritch Images represent spirits-of-the-dead. In addition to the corporeal qualities of the Eldritch Images, the colour and quality of the pigment used for their depiction adds further support for the spirit-ofthe-dead interpretation. Although very little is know of San colour symbolism, some evidence comes from the nearby Bantu-speaking groups who believe that white is a colour particularly associated with the world of spirits (Ngubane 1977). The translucent quality of the white pigment also calls to mind Lee’s observations that the !Kung often describe the spirits-of-the dead as having the transparent and ephemeral qualities of smoke. In many ways, then, the Eldritch Images apparently conform to 112 CHAPTER 5 OTHER BODIES: THE TYPE 2 ROCK ART IMAGES Southern African rock art studies, for most of the Type 2 Rock Art nineteenth century, have concentrated on San paint- The images at the main panel at Storm Shelter, that I ings and engravings. Occasionally, research has pointed described in Chapter One, may be characterised as clasto the possibility of a different authorship for images sic San rock paintings because they are typical of the that do not comply with the techniques, subject mat- fine-lined, shaded polychromatic representations found ter and manner of depiction of San rock art. Recently, throughout the south-eastern mountains. The classic this research has gained momentum and it is becoming San rock painting tradition also includes mono- and accepted that most of southern Africa’s major linguis- bi-chromatic images and like the shaded polychrome tic groupings participated in the making of rock art. images these are executed in a fine-line and in a paint Specifically, three different traditions are now identi- medium that appears to be a well-integrated mixture fied with San-, Khoekhoen- and Bantu-speaking peo- of pigment and binder. In addition, there are curious ples. Research, however, has yet to move from efforts images at Storm Shelter that do not fit with the conto provide cultural provenance to rock art to the more ventions of the classic San tradition. These images are complex issue of the role of the art in the construc- found on the side of a small boulder that faces towards tion of identity. Nevertheless, it is unavoidable that re- the shelter and thus the main panel. The images are in search must begin by attaching some cultural label to a faded white pigment and they lack the sophisticated traditions that are clearly different from San rock art; shading of the antelope of the main panel. From the only once this is done can we begin to question the manner in which the images are depicted, it is possible complexity that the art has played in identity-formation. to identify them only as quadruped animals. Nevertheless, eight sites, other than Storm Shelter, have been In this chapter, I begin the process of using body and found in Nomansland with images that are similar to embodiment as tools to understand the social produc- those on the boulder and these images exhibit antelope tion and consumption of the rock art by considering the features. As they mimic features—such as the projecauthorship of images in Nomansland that are visibly dif- tion formed in the antelope’s neck by the anterior exferent to classic San rock paintings. After describing the tremity of the thyroid cartilage of the larynx, antelope differences between these images and San rock paint- horns and dewlaps—found in classic San paintings ings, I discuss the various possibilities for authorship of eland, it is likely that they represent this antelope. and I argue that these images cannot be understood by making simplistic correlations between material culture While the subject matter is possibly similar to that of and cultural identity. Instead, they need to be understood classic San rock paintings, the technique and manner as a hybrid product stemming from the interaction be- of execution are not. Typically, these images are monotween San, including Nqabayo’s group, and the other chrome (Fig. 43) or bichrome. When they are bichrome, peoples occupying the Nomansland area. By under- the interior body of the animal has blocks of pigment standing Nomansland as part of a local system instead contained within the body of the animal that differ in of an isolated locale, we shall see in the ensuing chap- colour from the rest of the figure (Fig. 44). The colters how other people influenced the production of San ours in which both the blocks of pigment and the rest rock painting and San perceptions of their own identity. of the animal’s body are painted include red, white and 113 Fig. 43. A typical monochrome Type 2 image. Black represents red. Scale in centimeters. Fig. 44. Bichrome Type 2 image. Stipple represents yellow and black indicates red. Figure is approximately the same size as that of the monochrome image in Fig. 43. Approximately 21cm in breadth. yellow. Importantly, the pigment in which these three colours are painted is thin and powdery in comparison to the classic San images at Storm Shelter and elsewhere in Nomansland (See Fig. 23). There is usually very little variation in posture and they show far less diversity than classic San paintings of antelope. In all cases observed so far the powder-pigment quadrupeds are painted from the side. The differences in pigment quality, manner of depiction and variability between the powder-pigment quadruped images and classic San rock paintings are sufficient to suggest that they are of a separate tradition. I therefore label them Type 2. end of the shelter, or on separate boulders, as at Storm Shelter. The pronounced dissimilarities between Type 2 and classic San images as well as the occurrence of Type 2 images on surfaces close to but (mostly) separate from San rock paintings raises important questions about the authorship and significance of these images. Do the differences, for example, represent a breakdown or change in the classic San rock painting tradition or were Type 2 images painted by people other than the San? I consider each of these possibilities in turn. A Different San Tradition? Although there are regional variations throughout Type 2 images are often found at sites with classic San southern Africa, San rock painting is easily identifiable rock paintings. At only one site of the present sample as a corpus of art in places ranging from the Matopos (RARI-RSA-ESP3) of nine are Type 2 images found in Zimbabwe (see, for example, Garlake 1987, Garlake at shelters by themselves with no other San rock paint- 1995, Walker 1996), the Brandberg in Namibia (see, for ings being present in the same shelter. Most intriguingly, example, Lenssen-Erz and Erz 2000) and the Drakenfive of the known Type 2 sites are also shelters where sberg in South Africa (see, for example, Lewis-Williams distinctive anthropomorphic figures with facial features 2003, Lewis-Williams and Dowson 1992). The paintings are painted. There is thus a very close association be- in these areas (Fig. 1) are characterised by fine detail, tween these images and sites with prominent (Dowson’s complex shading and detailed depictions of antelope pre-eminent category) figures. With one exception, and humans. The art of the three areas differs, however, when found at San rock art sites, the Type 2 images oc- in the subject matter and the emphases of particular cur on rock surfaces separate to that on which classic aspects of antelope and humans. While differences can San paintings are found. These surfaces may be at the be ascertained between the San rock paintings of geo114 Fig. 45. Loubser’s Poster or Block Style images from the eastern Free State. Black represents red, stipple is light red and the uncoloured areas are white. Scale in centimeters. Fig. 46. Type 2 image of horse and rider. Such figures strongly suggest that Type 2 images are a late feature of Nomansland rock art and probably date from the colonial period. Approximately 20cm in breadth. graphically distant areas, it has proven far more difficult to determine if there are variations in San rock paintings over time in a specific region. Nevertheless, as we saw in Chapter Two, Loubser and Laurens (1994) have demonstrated that there is a shift over time in the San painting traditions in the Caledon River Valley on the western side of the south-eastern mountains. They observe paintings of eland that appear less refined than classic San rock paintings in the area and label these ‘poster’ or ‘block’ style images. Poster style eland are usually painted over classic San rock paintings, indicating their more recent age. These eland are done in a powdery pigment similar to Type 2 images in Nomansland (Fig. 45). They are also monochrome or bichrome but they do not have the same spatial arrangement of pigment within an image as in the Type 2 paintings. Whereas Type 2 images sometimes have small blocks of colour enclosed within the central part of the animal body, poster style images usually have a red body and white neck/head area. Moreover, poster style eland are easily identifiable as eland whereas the Type 2 quadruped images present more difficulties for identification. A further difference is that poster style eland are usually depicted with front and back legs articulated slightly inwards while Type 2 images are portrayed with straight legs (Compare Fig. 43, Fig. 44 & Fig. 45). These differences discount the possibility that Type 2 images are equivalent to the poster style and the presence of a single poster style image amongst the sites surveyed for this work (RARI-RSABOU1) would seem to count against any argument that Type 2 is simply a regional variation of poster style. Nevertheless, the appearance of a different style or tradition of San rock art within the Caledon River Valley raises enticing possibilities that Type 2 images could possibly be explained as the development of a new San painting tradition. At one shelter (RARI-RSA-LAB11), there are also images of riders on horseback juxtaposed with Type 2 eland and executed in the same technique (Fig. 46), demonstrating that at least some images of the tradition stem from the colonial period and thus suggesting that Type 2 might be a late development in San rock painting, made at a time when the classic way of making images changed. However, there are images of wagons, colonists discharging muskets and horses with riders all painted in the classic San way in Nomansland. 115 Fig. 47. Images of indeterminate antelope of the Late White tradition. Images in this tradition are finger-painted while Type 2 images are not finger-painted. This suggests that classic San rock art continued to be made while Type 2 images were being produced. Type 2 cannot thus be understood as being similar to poster style in that it replaces classic San rock art over time. The position of the images on completely different panels from San rock art is important in this regard. Superimposition is a widely documented and important aspect of San rock painting. It occurs throughout southern Africa (see, for example, King 1998, Mguni 1997, Russell 2000) and in rock art traditions known to be of great antiquity, such as in the Matopos of Zimbabwe. In the Drakensberg, including Nomansland, the superimposition of images is also common. At Storm Shelter, for example, there are as many as five layers of painting, one on top of the other. There appear to be ‘rules’ to the way in which one image is superimposed over another throughout the south-eastern mountains. Images, for example, are never superimposed over others in a way that completely obscures 116 them. Instead, they are placed over others so that earlier paintings always remain visible. ‘Rules’ such as this one suggest that superimpositioning is best understood as the building up of layers of meaning, each artist adding their contribution and thus enhancing the meanings of earlier images (on the implications of superimpositioning for meaning see Lewis-Williams 1972, 1974, 1992). If San made Type 2 images, it seems likely that they would have superimposed them over existing San paintings, as with the poster style images in the Caledon River Valley. The placement of Type 2 images on rock panels separate from San fine-line paintings and the lack of any superimpositioning in this tradition require a break so radical with convention that they suggest that someone other than San made these images. This suggestion finds support in evidence from the historical record of Nomansland. The material collected by Fynn, Stanford and the interviews conducted with Dyantyi as well as the other observations of the San Fig. 48. Compare this typical cluster of Type 2 images to the Late White figures in Fig. 47. Solid represents Red and clear is White. Scale in centimetres. of Nomansland all suggest that Nqabayo’s people and their descendants remained within a relatively small area in Nomansland from at least the mid-nineteenth century (and almost certainly earlier) all the way through to the 1980s. Silayi, as we have seen, claimed that they could paint very well and his description of the techniques of paint mixing and application conform well to what is known from other sources about the making of classic rock painting (compare, for example, Silayi’s description to other accounts in Rudner 1982). This suggests that Nqabayo’s San produced rock art in the classic San tradition. Moreover, Mamxabela’s husband was a painter and her son, Lindiso, was the last known San rock artist of Nomansland. Significantly, there are no Type 2 paintings at Ncengane Shelter, where Lindiso painted as recently as the 1920s. While some images at the shelter are cruder than others, they all ‘fit’ with the classic San tradition. Most importantly, none of them is executed in the powder-pigment associated with Type 2. The presence of Nqabayo’s people in the area for so long and their continued making of classic San rock art until about the 1920s further supports the idea that someone other than the San made Type 2 images. Type 2 images, then, cannot be understood to be the work of San artists simply breaking with the classic tradition. On the surface it would thus appear that they must be attributable to a different people, possibly Khoekhoen or Bantu-speakers. In the next section, I consider whether either scenario is plausible. A Bantu-Speaker or Khoekhoen tradition? As I mentioned at the start of this chapter, it is now widely acknowledged that there are at least three major rock art traditions in southern Africa—San, Khoekhoen and Bantu-speaker rock art. The most widely known of these three is the ‘Late White’ art. As the name suggests these images are painted almost exclusively in a white pigment (Fig. 47). In contrast to the fine brushwork of classic San rock paintings, the Late White images are finger painted and, consequently, the pigment is often 117 Fig. 49. This geometric form is probably a Khoekhoen rock art image. Fig. 50. Compare this Type 2 image with the Khoekhoen geometric form in Fig. 49. Clearly, Type 2 images are not of the Khoekhoen tradition. thickly applied to the rock surface (Prins and Hall 1994). It is often but not always the case that Late Whites are painted over San images, which is why they are described as ‘late’. Late White images include depictions of trains, colonial forces on horseback and even a camel (Smith and van Schalkwyk 2002), which lends further support to the belief that they are relatively late in the southern African rock art sequence. Late White paintings are found throughout southern Africa but also occur in parts of central and east Africa (Smith 1997). As one moves south, however, the density of Late White sites appears to decrease. In South Africa, for example, there are only a few hundred known Late White sites in comparison to the tens of thousand San rock art sites. The vast ma- is not a shamanistic rock art. Indeed, all available eth- jority of South African Late White sites are found in the northern part of the country, particularly in the Makgabeng Plateau. South of this plateau, Late White sites are not common but they do occur. In Nomansland, amongst the present sample of sites, there is a single site with Late White images. At one, RARI-RSA-CRA6 there are at least five images of humans, all in handson-hip postures and with penises exaggerated in size. nographic and oral history evidence suggests that the images were made to act as didactic and mnemonic devices in Bantu-speaker’s initiation rituals. In certain cases, the art also acted as an expression of political dissent towards colonial rule (Smith 1998). In addition to the Late White paintings, Bantu-speakers also appear to have made engravings. Along the south-eastern seaboard, engravings of linked circular motifs appear to be associated with Zulu-speaking peoples. The current hypothesis on these images is that they represent a map of the central cattle pattern settlement so closely associated with Nguni peoples (Maggs and Ward 1995). In addition to the identification of Bantu-speaker rock art, claims have been made that Khoekhoen pastoral- ists also made paintings in parts of South Africa (Dowson, Blundell, and Hall 1992, Manhire 1998, Manhire, Parkington, and Van Rijssen 1983, Rudner and Rudner 1959, Van Rijssen 1984, 1994, Wadley 2001, Willcox 1959, 1960, 1984, Yates, Parkington, and Manhire 1990). Typically, these images are finger-painted, usuA number of sources of evidence link Late White rock ally in red pigment and there is no shading as amongst art to Bantu-speakers and in parts of central Africa to- the San fine-line tradition. Moreover, while represenday, it is still being made by Chewa people (Smith 1995). tational images do occur in this Khoekhoen tradition, Although the art is made in rock shelters with San paint- the imagery is predominantly geometric in form (Fig. ings and possibly draws on the perceived supernatural 49). It is also becoming apparent that there is an associpotency in San rock paintings, the Late White tradition ated geometric tradition in the engravings (Smith and 118 Ouzman 2004). Arguments that link the Khoekhoen to these painted and engraved geometric traditions rely strongly on regional correspondences between the rock art and areas known to be occupied through archaeological and historical evidence by Khoekhoen. The symbolism of these images and the purpose for making them remain, at present, speculative. Nevertheless, the linking of a rock art tradition to Khoekhoen people creates an important distinction between three different painted rock painting traditions in southern Africa—San Fine-line (classic) rock paintings, characterised by naturalistic depictions of humans and animals, Bantu-speaking Late White finger paintings characterised by stylized images of animals and humans and Khoekhoen finger-painted tradition characterised by geometric motifs rather than representational imagery. Importantly, we have already seen that it is unlikely that Type 2 images in Nomansland are a development of the San fine-line tradition. They cannot, however, be described as fitting with a Bantu-speaking Late White tradition as Type 2 images are not finger painted and they are never painted in thick white pigment. Type 2 images also do not fit with a Khoekhoen geometric tradition for similar reasons (Fig. 50). We cannot thus arrive at a simple correlation between Type 2 images and San, Bantu-speaking or Khoekhoen cultural identity. Instead, following postcolonial arguments, we need to understand Type 2 images as produced out of the interaction between various people in Nomansland. The Nomansland Neighbourhood in the ethnographic literature of southern Africa in the following discussion. While I use these labels, it must be kept in mind that such terms have come under scrutiny and that current thinking regards these terms as complex social constructs that have their own history and concomitant shifts in designations. Both in southern Africa and globally, questions have been raised about the extent to which such labels create the impression of a static, essentialist society. Following this line of questioning, researchers have begun to consider how ‘ethnic’ identities have been constructed. In southern Africa, for example, Carolyn Hamilton (1998) has shown that Zulu identity and what we understand to be the essential traits that make up ‘Zuluness’ today are the result of representations, both Western and Zulu-speaking, that have been made historically for political and other reasons. While Hamilton’s work is exceptionally nuanced, sometimes writers following this line of thought tend to create the impression that because labels of identity are social constructs, the identities to which they refer don’t exist. Yet, anyone who has undertaken fieldwork amongst indigenous communities in southern Africa can testify that the labels used here are ones that are used by people themselves. While such labels are constructed, then, the identities to which they refer are very real. Moreover, some writers tend to see the construction of these labels as a purely Western creation. Such a view marginalizes the role played by indigenous peoples in their own identity-construction. This view also plays into the hands of those who would dispossess the extant San of southern Africa on the grounds that they do not exist as a separate group but are simply an economic underclass. It would be erroneous to think that the marginal ge- While I accept that the labels that I use here are conographical position of Nomansland meant that Nqa- structed and that they have a long semantic history, it is bayo’s group or, indeed, the other San living there were not my intention to discuss that history (indeed, each ‘pristine’ and isolated. Indeed, the San of Nomansland term would require a work of substantial length) but, inand along the whole eastern seaboard were caught up stead, to use the terms as historically received categories in complex and diverse interactive processes with their that are largely accepted in southern Africa. As the disKhoekhoen and Bantu-speaking neighbours. In order cussion proceeds, it will be clear that these labels do not to make sense of the complex and fluid relationship of refer to groups with defined and impermeable boundthe San to the various peoples in an around Nomans- aries but that they describe groups usually comprised land, I use cultural labels for people that are common of people from diverse cultural backgrounds. Those 119 120 Fig. 51. Stanford’s Map of the south-eastern seaboard, modified to include the known San groups in the area as well as Nomansland’s neighbours. groups, moreover, are not static and unchanging but often have a complex history. Using these labels, however, assists in understanding the very complex processes of interaction that have taken place in Nomansland over the last 500 years or so. I consider, in turn, the Bhaca, Mpondo, Mpondomise and the Thembu and their interactions with the San of Nomansland (Fig. 51). The Bhaca The Bhaca, according to their oral history, are descended from a small clan of Nguni-speakers occupying the upper Pongola River, beneath the Lebombo Mountains between Swaziland and South Africa (Hammond-Tooke 1962: 2). This tribe, known as the Zelemu, moved southwards sometime between 1734 and 1800. By 1800 the Zelemu were neighbours to the Wushe, a related clan that recognized a common ancestor. Early in the 1800s the Zelemu-Wushe became entangled in the Mfecane. Shaka sent warriors against the ZelemuWushe in about 1821 and they fled south, across the Mkomanzi River. Here the Zelemu-Wushe formed an alliance with the Cunu, a group that had also fled from Shaka some years earlier. Together, these groups conducted various raiding expeditions against other tribes in the region, pushing farther south still where they had successes against the Mpondomise in what is now the Mount Ayliff district. Their success, in turn, forced the Mpondomise farther south until they defeated the Zelemu-Wushe near the present-day town of Tsolo. The Zelemu-Wushe returned to the Mkomanzi River but were attacked by the Zulu, losing many cattle in the fight. The group again moved south and settled between Rode and the Mgano Mountains in what is now the Mount Frere area, close to the Core Nomansland Study Area. From here, they raided the Xhosa, Thembu, Mpondomise and Sotho. These raids were, in fact, movements of the whole tribe; entire families and stock moved, stopping only to plant sorghum and then moving on again after harvest (Hammond-Tooke 1962: 5). In 1830, Shaka’s military made a final advance on the Zelemu-Wushe. While crossing the Nunge Mountains, a severe snowstorm overtook his troops at night and many died of exposure; the next day the Zulu retreated. The Zelemu-Wushe attributed the victory to the magic of their leader, Madzikane, who was a reputed herbalist and keeper of the tribe’s sacred medicines (ibid.: 6). It is reputed that the thick smoke from Madzikane’s ritual fire turned into the clouds the brought snow and sleet. It appears that it is from this event that the ZelemuWushe received the name Bhaca from the Zulu, which means ‘Those who hide away and conceal themselves’. Madzikane died within four years of this event while on a raid against the Thembu and other groups. Ngcaphayi (not to be confused with Ngqabayi/Nqabayo) now became leader of the Bhaca. Under Ngcaphayi, the Bhaca became notorious raiders throughout the eastern seaboard area. The politics between the various groups adjacent to Nomansland at this time appears to have been complex and shifting and various Nguni groups allied themselves to others at certain times only to be enemies later. In 1835, the Bhaca became allies to the Mpondo under Faku and they settled in Mpondoland. Together with the Mpondo, the Bhaca invaded Thembuland on three occasions, removing virtually all the Thembu’s cattle. The Bhaca also conducted raids on the Mpondomise (Hammond-Tooke 1962: 6) without their Mpondo allies. Eventually, however, they fell out with the Mpondo and battles were fought at Lusikisiki and Mkatha forcing the Bhaca to withdraw from Mpondoland. Some years later, in 1844, the Bhaca invaded Mpondoland and attacked Faku, their former ally, at Mkhata. Ngcaphayi was killed at this battle and the Bhaca split into two groups. One section, under Mdutyana moved back to the Mzimkhulu area while the other under Makhaulu, Ngcaphayi’s son, established a settlement at Lutateni. As Makhaulu was still too young to rule, this section was ruled for many years by Diko, assisted by Ngcaphayi’s remarkable widow, Mamjucu. The Bhaca made peace with the Mpondo after this battle but this was not the end of their involvement with raiding. It was only four years after the battle in which the famous Bhaca leader, Ngcaphayi, was killed that Fynn was sent to stay with Faku so as to facilitate his investigations into the identity of the San raiders in Nomans121 land who were stealing colonial stock in the Natal Colony. In order to accomplish his mission, Fynn sent spies to the various groups of San living in Faku’s territory in December 1848 and it is through his efforts that, as we saw in Chapter One, we can identify the various groups of San in Nomansland. After a month of observation, one of these spies returned to tell Fynn that he had observed San with cattle and horses living near the Bhaca leader Mchithwa. According to the spy, Mchithwa’s brother was also looking after some of the San’s cattle. Fynn requested a meeting with the Bhaca and in response, Sawodi, another of Mchithwa’s brothers, visited Fynn. Fynn accused the Bhaca of being involved in the raids on Natal together with the San and he demanded that they surrender the San to his authority or face armed intervention. According to Fynn, Sawodi did not deny any of these accusations and promised to deliver the San within a week. Sawodi, however, did not surrender the San; instead, Mamjucu, who, as I have mentioned, was Ngcaphayi’s widow and ruler with Diko of one section of the Bhaca sent three messengers to Fynn. They reported that the San were away hunting and so nothing could be done about them. They requested Fynn to send someone to oversee the capture and detainment of the San and he obliged by despatching an observer who returned after a short while to describe a split amongst the Bhaca. One faction of the Bhaca, under Mamjucu and Sawodi, wanted to comply with Fynn’s request while another, under Mchithwa, protested their innocence and refused to give up the San. Fynn’s observer also reported that the San living with the Bhaca were related to the Mpondomise leader, Mandela, who was also implicated in the raids on Natal. both Mamjucu and Mchithwa were supportive of the San. Mchithwa, in particular, was in collusion with the San and a mixed group of people, mostly Mfengu and Mpondomise, under Hans Lochenberg (a coloured) on the upper Tina River. Fynn, frustrated with the Bhaca and the lack of action on the part of the Natal Colony took matters into his own hands. He pressurized Faku into obtaining the stolen colonial cattle back from the Mpondomise and intimated that a force from the Natal Colony would be sent against the San. He wrote to the Wesleyan Missionaries closest to the Bhaca and Mpondomise and warned them that a force would be sent against the Mpondomise chief, Mandela. With Faku’s help, Fynn arrested Mchithwa and an Mpondo force was sent against the Bhaca. Some 140 cattle were seized from the Bhaca and some were also taken from Lochenberg. Fynn also demanded from Lochenberg and Mandela the stolen cattle in their possession with a threat of military action if they were not forthcoming. Fynn’s threats, however, were thwarted by the Wesleyan missionaries who disapproved of his forceful efforts in the face of, what they believed was, a lack of evidence concerning the complicity of Nomansland San in the raiding of colonial livestock. Emboldened by their support, the Bhaca placed themselves under the protection of the missionaries to avoid further demands by Fynn. Lochenberg then proclaimed his innocence as well as that of Mdwebo’s San and demanded the return of the cattle that had been taken from him. Lochenberg claimed that the guilty San were the Thola. Mamjucu also visited Fynn and, for the first time, declared the Bhaca and Mdwebo’s San to be innocent. The missionaries completely undermined Fynn’s authority and influence and, as his role and status were in any case vague, he could The evidence against the Bhaca mounted in April 1849 not enforce his demands. On the 4th July 1849, seven when Fynn took a statement from a fugitive Bhaca who months after beginning his investigation, Fynn wrote had been at a dance at Mchithwa’s settlement when the to the Natal Colonial authorities requesting an invesSan arrived with cattle, some bearing colonial brands. tigation. After some deliberation, the Natal authorities The San sold the Bhaca ox-tails and then held a feast decided not to accede to this request but in early 1850, at their own settlement at which some of the Bhaca however, with another agenda in mind, Walter Hardwere present. Moreover, Fynn received evidence that ing was despatched to inquire about the San raids and the split amongst the Bhaca was contrived and that Fynn’s actions, but at the same time to obtain a cession 122 of territory from Faku. The historical evidence strongly suggests that Nomansland San, especially Mdwebo’s group, had close ties to the Bhaca. Together they raided and the spoils of their forays into the Natal Colony were often shared even if they did not raid together. The Mpondo Mpondo oral history suggests that they moved into the Transkei region from farther north and closer to the Drakensberg. They were observed in the area that they presently occupy by survivors from the Stavenisse in 1686 (Soga 1930: 302). As with the other groups living in the south-eastern region of South Africa during the nineteenth century, the Mfecane shaped the Mpondo. When Shaka’s warriors invaded Mpondoland in the early 1820s, the Mpondo under Faku, fled south across the Mzimvubu River. On the withdrawal of the Zulu, Faku’s followers returned to Mpondoland, only to flee again when Shaka’s warriors invaded a second time in 1828. This time he fled south and west and for the next twenty years settled in the Mngazi River Valley, returning to Mpondoland only in 1841 (Kuckertz 1990: 26). One of the most powerful groups along the eastern seaboard, the Mpondo managed to maintain economic and political independence until late in the colonial period (Beinart 1982), even though the governments of both the Natal and Cape colonies desired Faku’s land. Colonial authorities sought to exploit the power of the Mpondo in the mid-nineteenth century as a restraint against the San raiders of Nomansland. During a raid on Natal in February 1846, for example, the San stole a large herd of cattle and the Natal government demanded from Faku the return of the stock because, according to a treaty signed with him in 1844, the San were living in the area under his control. Initially, Faku ignored the demand because he did not view the scattered San as part of the Mpondo Kingdom (Stapleton 2001: 68). When the Natal Colony pressed their claim, Faku denied responsibility for the San and requested colonial assistance to drive the San out of his territory. The reasons for Faku distancing himself from the San are intriguing for the Mpondo leader appears to have had close ties with them throughout the early and middle nineteenth century. In 1830, W.B. Boyce, the first Wesleyan missionary to Mpondoland, observed San living along the Mzimvubu River trading elephant ivory with “Faku’s people”, as he put it, in exchange for corn and tobacco (Steedman 1835: 280). Jolly points out that this indicates that Faku acted as middleman, bartering ivory from the San hunter-gatherers and then trading it with European settlers (Jolly 1996: 48). More than just this trade relationship, the Mpondo were allies to the Bhaca and on several occasions co-operated in attacking or raiding other groups along the south-eastern seaboard. The Bhaca, as we have seen, were very close to the San of Nomansland and it is likely that on co-operative military ventures between the Mpondo and the Bhaca, San were also present. Faku almost certainly did struggle to control the various San groups in his territory. The San of Nomansland appear to have been semi-autonomous, co-operating with the neighbouring Nguni-speaking groups when it was useful to them, at other times doing as they chose without regard for the implications of their actions for their neighbours. Yet, his desire to distance himself from his former allies and trade partners requires explanation. Faku, largely through the efforts of the Wesleyan missionaries in his area, was aware that the Natal Colony coveted land under his control for further settlement. The San raids on the farmers of Natal offered the colonial authorities an opportunity to press Faku to give up land and, on occasion, Boer commandos requested permission from Faku to pursue a group of San raiders, probably with the added intention of surveying the land under his control. Faku, drawing on missionary support and advice, played a sophisticated diplomatic game, requesting assistance from the Cape government while denying any control over the San. The fact that shortly afterwards, Faku’s troops attacked San groups suggests that he had more knowledge of who the raiders were than he let on to the colonial authorities and that he had, to some degree, influence over them. The historical material, then, on the relationship of the San to the Mpondo is not as extensive as thet material on San-Bhaca relationships. What evi123 dence is available suggests that, at the very least, they were trading partners and through this relationship, the San communities along the south-eastern seaboard were brought into the broader economic environment. The Mpondomise The Mpondomise and the Mpondo are, genealogically, closely related and, according to oral histories, come from a common ancestor (Soga 1930: 334ff). These two tribes, supposedly, once occupied the area in, what is today, known as Giant’s Castle, at the foot of the Drakensberg Mountains. From here they moved southwards, the Mpondo first, followed later by the Mpondomise and settled in the Transkei areas where they are found today. The Mpondomise settled between the Tina and the Tsitsa Rivers. According to oral history, one Mpondomise chief, Ncwini, had a San wife by whom he had a son called Cira. It is from Cira, according to Soga (1930: 338), that the main line of the Mpondomise is descended. The Mpondomise’s recognition that they descend from a San woman probably softened their attitude to the San of Nomansland. The Bhaca, for example, attacked the Mpondomise on a number of occasions during the nineteenth century, raiding them for livestock. We have already seen that the Bhaca were closely allied to Mdwebo’s San in Nomansland, protecting them from demands of Fynn. If the Bhaca were raiding Mpondomise livestock, then some of Mdwebo’s San must have also been complicit in these clandestine and martial activities. The Mpondomise then probably had cause to be antagonistic to the San. Yet, it is amongst the Mpondomise that the remnants of Nqabayo’s group found a home after the Thembu under Mgudhlwa attacked them in 1858. Indeed, after Nqabayo’s band was attacked, Nqabayo himself went to live amongst Mditshwa’s people. Mditshwa was leader of the Mpondomise living west of the Tsitsa River in what is today the Tsolo district. Jolly (1992: 89) lists a number of accounts of observations of San rainmakers living amongst the Mpondomise in the last decade of the twentieth and the first two decades of the twenty-first century. Some of these accounts almost certainly refer to the rem124 nants of Nqabayo’s group and some of them report San people living in what is probably Ncengane Shelter on the Inxu River. As with the Bhaca, the historical evidence for San-Mpondomise relationships is substantial and this material shows that they largely co-operated. The Thembu The Thembu’s ancestral relationship to the other Nguni groups living in south-eastern South Africa is difficult to ascertain. The are possibly more closely related to the Sotho (Soga 1930: 466) than to the Mpondo, Mpondomise and Bhaca. Whatever their relationship to these groups, they appear to have settled along the south-eastern seaboard long before the Mfecane forced the other Nguni groups into the area. The date for this occupation, from oral accounts, seems to have been some time after 1620. It appears that movement of peoples from the north into what is today KwaZuluNatal Province split the Thembu into two groups. One group stayed behind in KwaZulu-Natal Province while the other moved south. The survivors of the wreck Stavenisse, who called them “Temboes”, encountered this second group in 1686 around the Mtamvuna River. Thembu oral tradition, however, mentions that a battle occurred in about 1650 on the Msana River, a tributary of the Bashee (Soga 1930: 468). According to Soga (ibid.: 470), the graves of identifiable Thembu leaders that date back to seventeenth century can be found in the area that the group occupies today, suggesting that they have been there for some centuries. This area is to the south of the Core Nomansland Study Area. Soga (1930: 479) states that the Thembu, historically, were less aggressive than the other groups along the south-eastern seaboard and that they certainly seem to have been on the receiving end of the many military actions that took place in that area over the last 500 years. As already mentioned, the Bhaca and Mpondo invaded Thembuland on three occasions, removing most of the Thembu’s cattle. At other times, the Thembu seem to have been more closely allied to groups such as the Mpondo. One Thembu leader, Ngubencuka, married a daughter of Faku, the great Mpondo chief. The Them- bu’s relationship to the San, however, appears to have been less amicable than that between the Bhaca and the San. Soga mentions that during Ngubencuka’s leadership, a section of the Thembu took the area of presentday Queenstown from the San. Of course, it was the Thembu leader Umgudhlwa who attacked Nqabayo’s San and whose men, without his permission, murdered the captured women and children. The Thembu attack on Nqabayo’s group should come as no surprise given that Nqabayo’s people had close relations with Mdwebo’s San who, in turn, allied themselves to the Bhaca— enemies of the Thembu for much of the nineteenth century. The available historical evidence, then, suggests that, of all the groups in the Nomansland region, the Thembu had the most hostile relationship to the San. Yet, on certain occasions, individual Thembu, such as Silayi, went to live with the San for extended periods. The Phuthi While the Bhaca, Mpondo, Mpondomise and Thembu are all Nguni-speaking peoples, living below the Drakensberg escarpment, above the escarpment a number of Sotho-speaking groups are known to have had close ties to San groups. The Bafokeng, for example, have oral traditions that refer to intermarriage between themselves and San (Ellenberger 1912: 18-19). Another group that was observed to have close relations with the San during the nineteenth century are the Phuthi. Phuthi oral traditions suggest that they came from the Giant’s Castle area, where the Mpondo and Mpondomise also claim they originated, and that they moved into northern Lesotho where they settled with a group known as the Maphuthing (ibid.: 24ff). For a period of some fifty years they stayed with the Maphuthing, adopting the name ‘Phuthi’ (meaning duiker). A schism appears to have taken place and the Phuthi then left the Maphuthing and slowly migrated southwards down the Caledon River Valley, eventually settling in the area of Mohales Hoek toward the end of the eighteenth century. In about 1795, as the Phuthi were moving towards Mohales Hoek, a child called Moorosi was born who, during the first half of the nineteenth century, would become their leader of great renown. During those first few decades of the nineteenth century, the Phuthi appear to have forged a very close alliance with the San groups in what is today southern Lesotho. As with Nqabayo and Mdwebo below the escarpment, this relationship seemed to be one of co-operation on raiding forays of livestock from European colonists or Bantu-speaking groups. The Mfecane disruptions that affected the southern Lesotho area from the early 1820s onwards appears to have given impetus to the economy of raiding that was affecting the entire eastern seaboard. At this time, the Phuthi split into two groups— one group, including Moorosi, went to live with the Mpondomise under Myeki below the escarpment (Jolly 1994: 56). Other members of the Phuthi chose to remain with the San along the Blekana and Tele Rivers on the escarpment just above Nomansland (Ellenberger 1912: 159). Moorosi returned after a short stay with the Mpondomise and the Phuthi now took up residence in the rock shelters around the present-day town of Lady Grey. From here, together with San groups, they raided the cattle of Nguni-speaking peoples below the escarpment, especially the Thembu. In order to facilitate further their raiding activities on the Nguni-speaking groups below the escarpment, the Phuthi moved into the eastern Witteberg, about the area of Lundean’s Nek; from here it was easy to launch expeditions into Nomansland and its surrounding areas (ibid.: 161). It was here, in 1824, that a group of people from various Basotho clans, known as the Motleyoa attacked the Phuthi forcing them to flee towards the Kraai River. In 1825, Moorosi led the Phuthi and defeated the Motleyoa and they settled the Lundean’s Nek area once again. Shortly thereafter the Thembu raided the area and the Phuthi lost most of the stock that they had previously taken from these Cape Nguni-speakers. At about the same time, Moshesh, the emerging paramount leader of the Basotho, sent his warriors to raid livestock from Moorosi. When they found no livestock, his troops took Phuthi hostages. Some of these were released but the young men were all taken to Moshesh at Thaba Bosiu and they were only released after the Phuthi paid tribute to him and subjugated themselves to his authority. In order to pay their debt to Moshesh, the Phuthi now 125 raided below the escarpment regularly. On two occasions, in September 1828 and then again at the beginning of 1829, Moshesh himself joined with Moorosi and raided livestock from the Nguni-speaking groups on the south-eastern seaboard. It appears that it was the Thembu who suffered mostly from these raids (Ellenberger 1912: 190-195). It is difficult to determine the extent of San involvement in these raids but their can be little doubt that they joined Moorosi on many occasions. Indeed, the relationship between the two groups was so close that Moorosi took two San women as wives and had a number of children by them. Moreover, during the Mfecane period, the Phuthi attracted individuals who were dispossessed from many different groups. Under Moorosi, the Phuthi, the San and people from other diverse cultural backgrounds were unified into a political entity that carried powerful influence in southern Lesotho and the Cape Colony for much of the nineteenth century. Then, in 1878, Moorosi’s son, Doda, was arrested for stealing colonial stock and conflict between the Phuthi and the Cape colonial authorities followed (Gill 1992: 15-16). The colonial authorities, together with members of various Sotho-speaking clans, laid siege to Moorosi’s stronghold at Thaba Moorosi. For most of 1879 the siege continued but in November that year, a way was found into the stronghold and Moorosi and many of his followers were killed. Many of those that escaped sought refuge in the remote mountains of southern and eastern Lesotho. Qacha’s Nek and Quthing. Mapote accepted the invitation and he travelled to How’s home. Here, he told her of his time spent with the San some 60 years previously and how, as a young man, he used to paint with the San in rock shelters. How asked him to make two rock paintings. It was, however, a long time since he had last painted and so he thought of asking an old friend to join him. In what has become a famous passage, How (1962: 32) describes Mapote’s pondering: “I will ask…”, he said. “I will ask…” He put his hand over his eyes and said again, “I will ask…” Then he took his hand from his eyes, looked at me and said, “They are all dead that I could ask”. Nevertheless, without assistance, Mapote still man- aged to produce two painted stones. The stone that he painted for How is approximately 20 x 15 centimetres in size. In this small area, Mapote produced four images—an eland, a hartebeest, a warrior holding an Nguni shield and two San with bows and arrows. Yet, the importance of Mapote’s encounter with How to rock art research lies not so much in the images that he produced but in his description of the process of the production of painted imagery. Mapote described that two important ingredients in the paintings were ochre pigment, which he called qhang qhang and blood. The best qhang qhang, according to Mapote, came from the high basalt mountains and sparkled while eland blood was the most sought after binder for the pigments. In the absence of eland, Mapote used blood from a cow In 1930, 61 years after Moorosi’s defeat, a remarkable in his paintings (for more on the ritual production and meeting took place in the remote mountains on the consumption of San rock art see Lewis-Williams 1995). edge of Nomansland between one of the survivors Even more important, Mapote told How (1962: 33) of the siege and battle at Thaba Moorosi and the wife that, together with his half-San stepbrothers, who were of a colonial official, Marion Walsham How. How was the sons of Moorosi’s two San wives, he used to paint staying at Qacha’s Nek, north of the Core Nomansland at one end of a shelter while the San proper painted at Study Area, at the time when she came to hear of a 74- the other end. Together with the other historical mateyear old man called Mapote who knew about San paint- rial discussed in this chapter, Mapote’s testimony proing. She sent an invitation to Mapote who lived in the vides important insight into the changing social conremote mountains to visit her. Mapote was the son of text of the production and consumption of rock art Moorosi and one of his lesser wives, an Mpondomise in Nqabayo’s Nomansland during the last 500 years woman. After the defeat at Thaba Moorosi, he moved and, it provides clues to the authorship of the Type 2 to the remote mountains of eastern Lesotho between rock art. The hisotrical material on Phuthi-San interac126 tion, while substantial, does not, at present, shed light on the relationship between the San living above the escarpment with those under the leadership of Mdwebo, Madolo and Nqabayo below the Drakensberg. Griqua dispatched a scouting party of over a hundred about mid-year in 1859 (Ross 1976: 97-98). This party travelled south, crossed the south-eastern mountain escarpment close to the modern town of Dordrecht (ibid.: 98). From here, they headed north-eastwards, The Khoekhoen passing close to the present-day town of Maclear and The various neighbours of the San of Nomansland that thus just skirting Nqabayo’s territory, before ending I have discussed so far are all Bantu-speaking groups. up in the region of Mount Currie. Finding the area Importantly, there were also Khoekhoen groups with attractive, the scouting party returned to Philippolis. whom the San of Nomansland had to interact. The relationship between the Khoekhoen and the San is the In 1860, the Griqua began to depart for Nomansland. subject of considerable and ongoing archaeological After relocating closer to the south-eastern mountains and historical debate. One side of the debate holds to from Philippolis, during the summer of 1862-3 they the view that the Khoekhoen and the San are linguisti- crossed this formidable terrain (ibid.: 102-103). This cally, culturally and economically distinct peoples. The time, however, they headed through the mountains of other side argues that it is impossible to separate the Lesotho. The terrain was more arduous that that of the two groups from one another and that San are prob- original scouting route. In addition, the weather in the ably dispossessed Khoekhoen (Schrire 1980, 1992). mountains was always unpredictable, in both summer Whatever the eventual outcome of this debate may and winter, and the various Basotho groups, as well as be, it is clear that in Nomansland, a distinction was the San living with them, raided and harassed the Griqua made between the two. Silayi, for example, identified on their journey. Eventually after just over two years of his friend, Hans (who introduced him to Nqabayo), as travelling, the Griqua reached Mount Currie below the a Khoekhoen. Ngqika, Hans’s nephew, was descended escarpment and established a settlement on the slopes of from a Khoekhoen woman and a San man (Maquarrie the mountain. Yet, the trek had left them impoverished 1962: 31). So, while Khoekhoen were living with the and demoralised (ibid.: 103).The Griqua were not only San groups of Nomansland, at least by the middle of comprised of Khoekhoen but some of their members the nineteenth century, a distinction was made between were San or, at least, descended from San. In addition, the two groups by the indigenous occupants of the area. there were white traders that trekked with the group when it moved to Nomansland. In their diverse makeOf the various Khoekhoen groups that affected up, the Griqua paralleled the San groups of NomansNomansland, the most powerful were the Griqua. In land. Once settled, the Griqua naturally attracted the the late 1850s, the Griqua were settled at Philippolis, in dispossessed and disaffected and the Mount Currie the central plains of South Africa. Although they lived habitation became home to many individuals from dilargely autonomously from colonial control, their inde- verse backgrounds. Two of these individuals were Hans pendence was threatened at this time because the Boers and Ngqika, who after the defeat at Gubenxa, eventuin what is today the Free State Province desired their ally migrated to Griqua country (Maquarrie 1962: 36). land. When it seemed that there was no way to maintain their sovereignty, they decided to move from the area. It is difficult to ascertain the degree to which the GriA number of possible areas for translocation presented qua migration and settlement affected the production themselves but the one that seemed the most attractive of rock art in Nomansland. It is likely that some Griwas Nomansland. A notorious raider, Smith Pommer, qua individuals joined the San groups of the area and who had spent time in the Nomansland region, had in- it is also probable that some San left the Nomansland formed the Griqua about the area (Shepard 1976). The area and went to join the Griqua at Mount Currie and 127 the later at Kokstad, where they established their new base in 1872. Yet, these sorts of movements appear to be quite rare. Both Hans and Ngqika had links to the Khoekhoen and so it was a natural step for them to join them once the power and influence of Nqabayo were curtailed after the defeat at Gubenxa. Yet, the various encounters by Stanford and others with the San in the Ncengane Shelter area, that I described in Chapter One, suggest that most of Nqabayo’s group and their descendants stayed in the area where Silayi encountered them in 1850 or so through to the second part of the nineteenth century and even into the twentieth century. This suggests that many San did not join the Griqua but preferred to remain autonomous. Nomansland and its Art in a Local System In spite of its geographical isolation, then, Nomansland was not a pristine Eden in which the San lived idyllic lives as hunter-gatherers, independent of the wider southern African social milieu. Yet, at the same time, the San were not simply an underclass of their Bantu-speaking neighbours. They employed a range of economic strategies, including trade in the form of ivory, bribery/tribute in the form of stolen stock paid to their neighbours and at other times, they sold their ritual services, particularly as rainmakers, for payment. When it suited them, they entered into alliances with various neighbouring peoples; at other times, they acted as independent agents, even stealing from their own allies. Adopting these various economic strategies allowed Nqabayo’s followers and their descendants to retain some control over their destiny all the way into the first two decades of the twentieth century. Yet, the larger local forces operating on the Nomansland region slowly forced them to amalgamate with their Bantu-speaking neighbours. This amalgamation was rarely one of a simple patronclient relationship but, instead, was one based on the San’s liminality in which they manipulated the symbolic capital of their ritual status for socio-economic benefit. Moreover, the relationship between the San of Nomansland and their neighbours was not homogenous, neither 128 in time nor in space, but differed greatly depending on the particular time and the particular group of Bantu-speakers with whom they interacted. The San, for example, were closer to the Bhaca, the Mpondomise and the Phuthi than they were to the Thembu. Here again, it seems that Nqabayo’s group were close to the Mpondomise while those under Mdwebo were closer to the Bhaca and above the escarpment the San groups were closer to the Phuthi. These different alliances almost certainly created difficulties for the San groups. On the one hand the San groups were fairly close to each other and there appears to have been significant fluidity between them, so that members of Madolo’s group were related to people in Nqabayo’s group. Nqabayo was also closely allied to Mdwebo’s group. Yet, Nqabayo’s relationship with the Mpondomise was at odds with Mdwebo’s relationship with the Bhaca for the Mpondomise and the Bhaca, as we have seen, were, for much of the nineteenth century enemies. The Phuthi, on the other hand appear to have been close to the Mpondomise. The oral history of both groups claims that they originally occupied the area in the Giant’s Castle area, in the northern Drakensberg before migrating southwards along different routes. Moorosi also had an Mpondomise wife and, possibly, this is why the Phuthi were so close to the Mpondomise. A more intriguing possibility is that the San living with the Phuthi were related in some way to the San living with the Mpondomise (i.e. Nqabayo’s people) and it was this relationship that facilitated close contact between the Mpondomise and Phuthi. A close relationship between the San living amongst the Phuthi and those under Nqabayo and Mdwebo is implied by the 1850 expedition carried out by these two leaders that I described in Chapter 1. In that venture, members of both Mdwebo’s and Nqabayo’s groups went up the escarpment and crossed the Witteberg into southern Lesotho before following the Orange River upstream to where they attacked the Thola. Their route would have taken them through the heart of Moorosi’s Phuthi territory and given the general state of unrest in the area during most of the nineteenth century, it seems that it would be impossible for them to do so undetected. This suggests that the San groups below were on cordial terms with those above the escarpment. ing Nomansland and its surroundings were creolized. Whatever the relationship between the San of Nqabayo, Mdwebo and those living with the Phuthi, it is clear that by the nineteenth century, the San of Nomansland were caught up in a complex and constantly changing set of alliances with their Bantu-speaking neighbours. Their relationship with their neighbours, as the Mpondomise’s recognition of their descent from a San woman suggests, stretched much further back in time. This interaction was not merely one of superficial interchange of cultural traits and ideas between two discrete cultural groups but was one founded on genetic and linguistic interchange. Indeed, the linguistic creolization of Nguni-speaking and San peoples is evident in the presence of click consonants in the Ngunispeaking languages of south-eastern South Africa. The Nguni languages, as we have already seen, are the only members of the Bantu-speaking family to include click consonants. Moreover, certain words in the Nguni languages are cognates of words in San languages (Herbert 1990a, 1990b, Lanham 1964, Louw 1974, 1977, 1979, Traill 1995). The interaction between the San and the Nguni peoples went further than cultural and linguistic creolization; the preponderance of Khoesan genes amongst the various Cape Nguni groups demonstrates that there was substantial physiological interchange (Jenkins, Zoutendyk, and Steinberg 1970). The production and consumption of some of the rock art of Nomansland, at least, must be understood within this complex and changing context of San society within a local system and it is through these local interactions that we may understand Type 2 rock art images. Mapote’s description of the division between San and non-San areas of painting in shelters is important in this regard. At the start of this chapter, I noted that Type 2 images are mostly found in association with sites where large, distinctive anthropmorphic figures can be seen and that they only occur rarely by themselves. With one exception, all known Type 2 images at sites with classic rock paintings, occur on rock surfaces separate from San images. At Storm Shelter they are found on a boulder but, most commonly, they are found on rock surfaces that are situated towards one end of the shelter and that are spatially separated from classic San rock paintings. The separate placement of Type 2 rock art suggests that these images were made by people living together with San groups who produced rock art in conjunction with them but within spatially discrete areas in the same shelters. Type 2 images, then, were probably produced by non-San people painting in together with the San of Nomansland. As we have seen, Mapote lived with the San who were associated with Moorosi and who made paintings in Lesotho. The area is spatially so close to Nomansland and the time when Mapote was living with Nqabayo’s band comprised individuals that were of San, the San is so close in time to Nqabayo’s active occupaKhoekhoen, Thembu and mixed descent. The different tion of that area that its seems the only reasonable conBantu-speaking groups in and around nineteenth-cen- clusion to determining the authorship of Type 2 images. tury Nomansland and, indeed, throughout the Transkei region can also not be thought of as essentialized, dis- We have seen in Chapter One that Nqabayo’s group crete cultural packages. They were largely shaped by the included people from diverse cultural backgrounds. Sievents of the Mfecane, during the first two decades of layi was a Thembu and he was introduced to Nqabayo the nineteenth century but also from historical process through his friend Hans, who was of Khoekhoen destretching further back in time. Groups such as the Phuthi scent. Clearly then, Nqabayo’s group, as were the San and Bhaca were comprised of different clan lineages and groups living with Moorosi during the nineteenth cenincorporated refugees from other groups. Although we tury, was an amalgam of different cultural groupings may refer to the groups by the labels of San, Thembu, that orbited around a core of San people. What gave Mpondo, Mpondomise, Bhaca and so forth, clearly, by these groups their cohesiveness was the shared ecothe mid-nineteenth century the various groups occupy- nomic practice of raiding but, as Mapote’s testimony 129 shows, it would be fallacious to regard these groups as entities sharing a common identity. They appear to have been highly fractious and people who joined them could just as easily leave in a short time. Silayi only spent three years with Nqabayo’s group before leaving and the 1858 massacre of the women and children seemed to encourage those of part-San and non-San descent to scatter. Hans and Ngqika, as we have seen, went to live with the Griqua. Within these volatile groupings, a new and hybrid social life was fast evolving and new art forms, such as the Type 2 images were produced. Yet, it would be a mistake to see these new expressions as simply co-operative ventures with equal contributions by people of different cultural backgrounds. Clearly, classic San rock painting remained a San practise and while other people were allowed to paint, they were restricted as to where they were allowed to do so. This suggests that San rock painting in Nomansland during the mid-nineteenth century, like the Great Dance at Ghanzi during the mid-1970s, became a vehicle for the expression of San identity. The presence of other bodies from different cultural backgrounds and of different physiological types in the very shelters in which San were painting, however, would soon lead to a shift in the way that certain San created their identity. In the next chapter, I discuss how changing constructions of San identity can be perceived in the portrayal of the human body in the rock art of Nomansland. 130 CHAPTER 6 THE DEATH OF THE POST-CRANIAL BODY The term ‘Nomansland’ implies a peripheral area, iso- ments. One such category of images has been recoglated from broader socio-political developments. As nized by Thomas Dowson (1994) and labelled as ‘prewe have seen in the previous chapter, however, the eminent’. Dowson argues that these images were the San groups in Nomansland had complex and chang- culmination of a three-stage progression over time ing relationships with their neighbours. These relation- that saw an increase in San concepts of individualships were not simple exchanges of ideas and artefacts ity. The pre-eminent images, as the last expression of but, instead, were complex processes that involved in- this process of individualization, represented powertermarriage and co-habitation at various times. These ful, individual pre-eminent shamans—a kind of ‘big long-term interactions produced hybridized socio- man’. This three stage progression was the result of cultural groups in Nomansland and its environs. Im- the social stratification of San groups through their portantly, as I have argued, some of the Nomansland ritual interaction with their Bantu-speaking neighrock art needs to be understood as a product of local bours. This interaction, although originating earlier in processes in the area. These local processes of inter- time, was intensified with the advent of colonialism. action in Nomansland, however, were themselves tied to broader, global developments. The most important While I have already briefly discussed the theoretiof these worldwide developments to affect Nomans- cal and empirical problems of Dowson’s argument in land is, of course, colonialism. Although the official Chapter Two, I take that discussion further here and date for the colonization of South Africa is 1652, it I consider his arguments in light of the discovery is clear that from the earliest voyages of discovery by of Storm Shelter and several other painted sites in Portuguese mariners in the Fifteenth Century, south- Nomansland that depict images that fall into his preern Africa’s indigenous populations were brought into eminent category. Drawing on this new material and the global system. How far back before colonialism the principles of the somatic past, I argue that a more we may trace the impact of global systems on south- nuanced interpretation of this category of pre-eminent ern Africa’s people is, however, still a debated issue. In images is possible. I begin by describing the characterthe Kalahari Desert, for example, Wilmsen and Den- istics of Dowson’s pre-eminent category. I have menbow (1990: 499ff.) have argued that the San were part tioned at several places in the text the occurrence of of global trade networks as far back as 1000 years be- large, distinctive anthropomorphic figures at rock art fore the present and probably even earlier. Whatever sites in the Nomansland Core Study Area. These figthe antiquity may eventually turn out to be, it is clear ures are present at Storm Shelter and it is at sites with that, for at least the last 500 years, southern Africa has these distinctive anthropomorphs that Type 2 images been subject to the effects of global colonial processes. are mostly found. These distinctive anthropomorphic figures correspond to Dowson’s pre-eminent category. As with the Type 2 images that need to be understood I re-label these pre-eminent category anthropomorphs in terms of local processes, some of the rock art in as Significantly Differentiated Figures (SDFs) in order Nomansland needs to be explained in terms of the to separate the social and political context of these iminteraction between global forces and local develop- ages from their technical and artistic attributes. I then 131 show that the newly discovered images allow us to see a change in the depiction of the bodies of these figures over time. As such, these rock art images show a changing experience of embodiment within the San communities of Nomansland over time and this shift is partially explicable by a consideration of historical material that shows that the communities of the south-eastern seaboard in general and Nomansland in particular were creolized amalgams of different population groups. tain complex show that there are also no SDFs known north of the Orange River15. Moreover, in her survey of the Drakensberg, the eastern side of the south-eastern mountain complex, Vinnicombe (2000: Pers. Comm.) also did not observe any SDFs. Her survey area began just north of Nomansland and extended to just south of the central Drakensberg region. As we have seen in Chapter Two, David Lewis-Williams and Harald Pager worked to the north of Vinnicombe’s survey area in the 1970s and 1980s and, significantly, they did not record any images in their publications that could be argued Significantly Differentiated Figures to be SDFs. SDFs thus appear to be limited to the The most noticeable characteristic of the SDFs is their southern parts of the south-eastern mountains, south apparent limited distribution within the south-eastern of the Orange River and their highest concentration mountains. All known SDF sites so far found occur is in the Nomansland region in the Storm Shelter and south of the Orange River and most are located within adjacent valleys. While the possibility exists that some the Nomansland area (Fig. 52). There are, at present, sites with SDFs may still be identified in parts of the at least four known SDF sites outside of but close to south-eastern mountains, it would seem to be the case the Nomansland Core Study Area. The first of these that, at the very least, SDFs increase in frequency the sites—RARI-RSA-PEL4—is situated above the escarp- farther south one moves in the mountain complex. ment in the Lundean’s Nek area, where the Phuthi and their San allies also resided during parts of the nine- Their limited distribution in the south-eastern mounteenth century (Fig. 51). RARI-RSA-PEL4 was first tains suggests that they may have been made by single published by Walter Battiss in 1948 and subsequently group of San or San groups living next to one another visited by Harald Pager and then also by Bert Wood- that were in contact. It is noteworthy that most SDF sites house and Neil Lee13. A second site—RARI-LES- occur in or immediately next to the part of NomansMTM1—is in southern Lesotho, close to the famous land occupied by Nqabayo’s people. Indeed, Ncengane mountain fortress of Thaba Moorosi, discussed in Shelter, where Lindiso made what are probably the last Chapter Five (Fig. 54). The remaining two sites—RARI- San rock paintings ever to be produced, has an SDF RSA-DIN1 and RARI-RSA-BUR1 (Fig. 55)—are situ- (Fig. 57), further supporting a link between Nqabayo’s ated above the escarpment and come from compara- San or their immediate ancestors and the SDF shelters. tive material in the Barkly East area that I mentioned If Nqabayo’s band made the SDFs, then they would in the Introduction. Other than these four sites, all obviously be recent paintings. Although undated by sciknown SDF sites are located in the Nomansland Core entific techniques, at one SDF site, the subject matter Study Area and most are in a cluster of valleys imme- of the paintings suggests a relatively recent date for the diately next the one in which Storm Shelter is located. SDF. At RARI-RSA-MEL6, the SDF is painted in sevAlthough large sections of Lesotho, including the southern parts close to Nomansland, were surveyed by researchers working out of Roma University in Lesotho in the 1980s, no SDF images were noted (Smits 1998: Pers. Comm.)14. Intensive surveys of the Maloti Mountains on the western side of the south-eastern moun132 eral different colours (Fig. 56). One of these colours is a very unusual orange that does not appear to occur at any other Nomansland site. To the left of the SDF is a separate but related scene that includes several bizarre images. The same orange pigment on the SDF is also present on some of these extraordinary figures, suggesting that they were painted from the same paint pot and thus at the same time as the SDF. One of these peculiar figures has a human leg and appears to bend forward. Most notably, however, this figure has the head and horns of a cow. The presence of domestic animal features on this figure allows us to set a chronological limit on the date of its creation. Cattle first entered South Africa about 2000 years ago and then slowly moved south. It is thus highly unlikely that the strange cattle therianthrope would have been depicted before then. If, as I have suggested, the SDF was painted at roughly the same time as this odd figure, then it was probably created in the last 2000 years and probably even more recently. At RARI-LES-MTM1 (Fig. 54), the SDF wears what appear to be cattle horns; if they are, they too add a similar chronological control to that of RARI-RSA-MEL6. While the available chronological evidence is at present limited, it nevertheless suggests a relatively recent date for the SDFs. We cannot, at this time, however, be certain that Nqabayo’s group painted the SDFs. What is certain, however, is that they knew of the SDFs. Indeed, it would be difficult for any visitor to a rock art site to miss an SDF as they have a number of characteristics that make them readily identifiable. They are typically anthropomorphic and are commonly therianthropic, often having hooves instead of feet and other antelope features such as eyes or pointed ears. Their therianthropic characteristics tie the images to experiences and beliefs surrounding the Great Dance. The frequent depiction of nasal haemorrhage in the SDFs, a feature that we have already seen was common to the San dances performed in the south-eastern mountains, supports this connection. These features do not, of course, mark the SDFs out as exceptional because they are common to many painted figures in the southeastern mountains. What makes them noticeably different is their size, which varies from 28cm (see Fig. 58) to 70cm (see Fig. 59) in height (outside the Core Nomansland Study Area, one SDF is 90cm in height). In comparison to other anthropomorphic images in rock art areas elsewhere in southern Africa, such as in the Matopos of Zimbabwe, this is not large. Nomansland SDFs, however, are large in comparison to other paintings within the south-eastern mountains and, most importantly, they are large in relation to other images within the panel in which they are painted. Their relative size compared to other images makes them easily recognizable in any shelter in which they are found. The large relative size of the SDFs extends to certain anatomical elements. Some of the images, for example, have unusually large penises (but not grotesque as in the Eldritch Images) in comparison to other San rock paintings within the south-eastern mountains as a whole and within the panel in which they occur specifically (Fig. 56 & Fig. 58). Their imposing size is underscored in some examples by an embellishment in the number of accoutrements, such as bows or divining switches, which the SDFs carry. One SDF, for example, is depicted with sixteen bows protruding from a hunting bag (Fig. 64 & Fig. 65). Clearly, this is an unrealistic number. As with the large overall size of the figures and the various exaggerated anatomical parts, the inflated number of material items associated with the SDFs serves to draw attention to these images in densely painted panels. However, it is not possible, amongst the 134 Nomansland Core Study Area rock art sites, to identify an SDF at more than one shelter; each SDF is unique to a particular shelter. In addition to their large size and exaggerated number of material items, some SDFs are painted in astonishing detail. At RARI-RSA-WID2, for example, the SDF has a profiled face with the upper and lower lip as well as the nose being present (Fig. 58). The figure appears to have either a headdress or braided hair. Two thin red lines are painted from the nose and across the face and down the neck of this SDF; the lines represent the emanation of nasal blood experienced during observed dances in the south-eastern mountains. The remarkable detail of the face and the head extend to the rest of the body. The infibulation on the figure’s penis is remarkably elaborate. Below the waist, two delicate antelope legs with hoofs are depicted. The legs are painted in white but both are fringed by a thin red line on the back side of the limb. By contrast, the hoofs and fetlocks (that part of the leg where a tuft of hair grows from the 133 Fig. 52. Map showing the distribution of SDF and type 2 sites in the Nomansland study region. Stars=SDF/LH-SDF Sites; Diamonds=Type 2; Dots=Rock art sites. Note how the SDF/LH-SDF and Type 2 Images cluster together around the Storm Shelter area. 134 Fig. 53. RARI-RSA-PEL4. The arrow points to the SDF. The site was first published by Walter Battiss in 1948 but he apparently did not observe the large-headed central figure. The head of this figure is painted in white but the nose and a section of the cheek are painted in red. The figure wears a distinctive striped headdress and, although only faintly visible in this photograph, is associated with a staff-like object that is astonishingly similar to the staff-like objects associated with the SDFs at RARI-RSA-MEL8 and with the primary SDF at Storm Shelter. Fig. 54. The SDF from RARI-LES-MTM1. This site, as is RARIRSA-PEL4 (Fig. 53.), is located in the area adjacent to Nomansland and is known to have been occupied by Moorosi and the San under his leadership. This figure’s headdress possibly includes cattle-horns. Orientation is as the original. Solid=Red; Clear=White; Stipple= Lighter Red. Figure is about 36cm in length. Fig. 55. There are several sites adjacent to the Core Nomansland Study Area as marked out in Fig. 52 where possible SDFs are found. I include RARI-RSA-BUR1 (seen here) in the list of known SDF sites. Solid=Red; Clear= White; Stipple=Lighter Shades of Red. The anthropomorphic figure on the right is about 90cm in height; the one on the left about 85cm in length. 135 Fig. 56. The SDF at RARI-RSA-MEL6 is painted in similar pigment to an associated figure with cow-like features. This suggests that at least som Nomansland landscape. This is a typical SDF site with a large, anthropomorphic figure that is painted in considerable detail. Solid=Red/Orange; pastern-joint) of this therianthropic figure are painted in black. A thin red line fringed by white dots appears to weave around the upper body of this SDF. In addition, the figure holds a long staff and a divining switch. Figures such as the one that I have just described from RARI-RSA-WID2 led Dowson to argue that the SDFs were far more detailed than the other images with which they are juxtaposed. Yet, this is only true for a small number of the SDFs; most are, in fact, depicted in less detail than the other images found in the same rock art panel. For example, at RARI-RSA-RON1, a shelter with numerous fine-line and shaded images, parts of the SDF are painted in less detail than the other images in the panel (Fig. 59). This particular example is the largest of the known SDFs in the Core Nomansland Study Area, Fig. 57. The SDF from Ncengane Shelter, on the right in this photograph, has a large head painted in white pigment. The presence of such a figure at the shelter where Lindiso painted strongly suggests that, if Nqabayo’s group and their descendants were not the makers of these images, they were at least familiar with them. Photograph courtesy of Knut Helskog. Figure is approximately 50cm. 136 me of the SDFs date from the last 2000 years and they must be considered as a product of interaction between the San and other people on the Clear=White; Stipple=Off-White. Scale in centimetres. being some 70cm in size. The figure has a profiled face with a nose and upper and lower lip depicted. Painted around the head area are numerous tiny white dots. While the facial detail is similar to that of surrounding images, the body below the head is painted with few features. There are no accoutrements, infibulation of the penis or any finely detailed legs and hooves, as we have just seen with RARI-RSA-WID2. Even more remarkable than the lack of detail on the body below the head, the figure has no hands. It does not appear that the hands were once painted and have now faded as no trace of any pigment is detectable whatsoever. If this was a case of pigment fading over time then the hands would almost certainly have been painted in fugitive white. The face of the image and the dots surrounding the head area, however, are all in white pigment and have preserved well, further counting against any suggestion that the hands of this particular SDF were ever depicted. By contrast to the SDF, almost all anthropomorphic and therianthropic images in the rest of the panel (and at other Nomansland sites) have substantial detail in the body area below the head. Indeed, it is only in the area of the head that the SDF exhibits more detail as no other image has so many dots painted in the cranial area. A lack of detail can also be seen in the SDF from RARI-RSA-TYN2 (Fig. 60). This SDF is therianthropic and has an anthropomorphic body but a long hartebeest-like head. Parts of the head are painted in different colours and the attention to detail in the facial area is extensive. The figure’s hands are also depicted in some detail and it is clear that this SDF has claw-like digits instead and hands and fingers. The level of detail, however, decreases further down the body. There is very little detail on the abdomen below the arms and the figure is represented as only having one leg; that leg, moreover, terminates at the knee area so that the SDF is portrayed with only the upper part of one leg. The emphasis on the head of the SDFs at RARI-RSARON1 and RARI-RSA-TYN2 is a feature that extends to certain unusual SDFs at other Nomansland rock art 137 Fig. 58. RARI-RSA-WID2 is a classic example of an SDF site. The scene is of a Great Dance and there are several dancing and clapping figures. To the left, an SDF is painted in different colour and far more detail than any of the other figures. The figure holds a divining switch, has antelope hooves and nasal blood smeared across the face. The large creature to the far left is probably a depiction of a rainanimal while the antelope between the rain-animal and the dance is a classic Type 2 image. Thisis the smallest of the known SDFs. Red is solid and lighter shades are stippled. White is clear. Scale in centimeters. sites that have unnaturally large heads. The size of the head in these SDFs varies but they can take up as much as a third of the entire image (Fig. 61). In some rare cases, the heads are as large as the rest of the body. Because of the large head size, I term these particular images, Large Headed-Significantly Differentiated Figures (LH-SDFs). Although, as we have already seen, Battiss published the SDF site RARI-RSA-PEL4 (Fig. 53) as long ago as 1948 he appears to have been unaware of the SDF at the site, which is an LH-SDF, and these figures have remained an unrecognized category of San rock paintings in the south-eastern mountains. It was only with the discovery of Storm Shelter, which has two clear examples of LH-SDFs (Fig. 62 & Fig. 63), that the possibility that such images may be more frequently depicted in the rock art arose. In the course of survey for the Nomansland Project, a handful of further examples 138 were found. Of the 134 sites that form the core sample of this study, only twelve depict SDFs (this number excludes the four sites outside the Nomansland Core Study Area) and at seven of these there are LH-SDFs. The majority of these are found in the valleys immediately adjacent to Storm Shelter. One of these seven sites, RARI-RSA-MEL8, was identified by Dowson as depicting a normal SDF but has been re-evaluated and identified as illustrating an LH-SDF (Fig. 64 & Fig. 65). At the time that Dowson was conducting his research, Storm Shelter had not yet been discovered and the possibility of LH-SDFs was still not widely contemplated. The SDF at RARI-RSA-MEL8 is, in any event, not clear and is flaking in certain sections. In the light of the discovery of other LH-SDFs, however, RARIRSA-MEL8 can be re-interpreted as an LH-SDF site. Fig. 59. The largest of the known SDFs in the Core Nomansland Study Area is approximately 70cm in height and is located at RARI-RSA-RON1. Note the lack of detail in the abdominal areas below the head. Solid represents red. White is indicated as clear. Scale in centimetres. Fig. 60. As with the SDF in Fig. 59, the SDF at RARI-RSA-TYN2 shows a marked lack of detail in the abdominal area below the head. Only one leg and only the upper part of that leg are portrayed in this image. Figure is approximately 35cm in length. As do normal SDFs, LH-SDFs have faces that are painted in profile with clearly articulated features. These commonly include the eye, nose, upper and lower lips, chin, forehead and, more rarely, an ear. Sometimes, these facial features are taken from the animal world and at Storm Shelter, for example, the two LH-SDFs both appear to have antelope eyes as well as pointed ears that are clearly not human. Moreover, those aspects of the faces of LH-SDFs that are human often appear to be drawn from physiological features that are distinctly non-San. The various combinations of animal and human physiological features make each LH-SDF unique and the distinctive nature of each of these figures is further enhanced by the exclusive headdress of each image. Typically, the headdresses have the appearance of items of apparel and some appear beret-like in form (e.g. Fig. 63) while others are more fez-like in appearance (Fig. 66) and still others look like a form of woollen cap (e.g. Fig. 53). In some cases, however, the headdresses are clearly non-realistic and they may, for example, take the form of lines of dots trailing from the head of the figures (Fig. 61). The distinctive facial features and headdresses make each LH-SDF unique and, as with the normal SDFs, the same LHSDF cannot be identified in more than one shelter. At Storm Shelter, however, there are two images that may be classified as LH-SDFs and at certain other sites more than one normal SDF may be present. At Storm Shelter, one of the LH-SDFs is positioned at the top left-hand side of the panel and is seemingly superimposed by a feline. The other is in the centre of the panel and is not superimposed by any other images. Typically, however, when there is more than one 139 Fig. 61. Certain SDFs in Nomansland are remarkable for their enormous heads. Some of the large-headed SDFs have facial features that appear to be derived from a wide-range of population groups including European, Asian, Bantu-speaking and San populations. Colour is entirely white. Scale in centimeters. Fig. 62. One of the secondary SDFs from Storm Shelter. The figure has a typical profiled face, including upper and lower lips, nose, ear and eye. Notice the large size of the head in relation to the rest of the body. Black represents red and white is clear. Scale in centimeters. SDF or LH-SDF present at a site, one of the images will be larger than the others. This is the case at Storm Shelter where the LH-SDF associated with the feline is larger than the one in the centre of the panel. In such cases, we may refer to the larger of the two images as the primary SDF or primary LH-SDF and the others as secondary SDFs or secondary LH-SDFs. At Storm Shelter, both the primary and secondary LH-SDFs have similar facial features—both have a narrowed eye and a pointed ear. Nevertheless, they have different headdresses; one has a beret-like headdress while the other appears to have a kind of skin draped across its face. While similarities amongst the figures thus occur between LH-SDFs within shelters, these similarities do not extend to images in different shelters and it appears to be the case that each LH-SDF is unique to the particular rock shelter in which it is found. Nevertheless, three of the LH-SDFs, at three different sites have a peculiar characteristic that suggests whomever made each image was aware of one or both of the other shelters. This feature is an extended staff, painted in white and red. Towards the top end of the staff in all three examples is a small protrusion that extends either side of the main shaft. The left and right extensions of these protrusions are painted in different colours. In all three cases the LH-SDF associated with the staff has one arm extended towards the staff in such a way so as not to touch the object; it is as if the figure reaches for the staff. Two of these three examples come from adjacent valleys; the Storm Shelter primary LH-SDF (Fig. 63) and the re-interpreted LH-SDF at RARI-RSA-MEL8 (Fig. 65) are in adjacent valleys, while the third image from RARI-RSA-PEL4 (Fig. 53), is as I have mentioned previously, from the Lundean’s Nek area, above the escarpment and a relatively short 140 Fig. 63. The primary LH-SDF at Storm Shelter. The figure has been intentionally depicted as if behind the feline. Note the ‘HumptyDumpty’ body and the absence of legs. Note the staff-like object, which is nearly identical to that at RARI-RSA-PEL4 (Fig. 53) and RARI-RSA-MEL8 (Fig. 65). White is clear and stipple represents various shades of red. Scale in centimeters. distance away. The occurrence of this staff in such similar contexts suggests that either the three images were made by the same artist/artists or by people who were in contact with one another. In Chapter Five, I discussed the possibility of relationships between the San living with the Phuthi who occupied the Lundean’s Nek area and those under Nqabayo and Mdwebo, who lived below the escarpment. We also saw how Nqabayo’s and Mdwebo’s groups passed through Phuthi territory when they attacked the Thola. These historical accounts clearly show that the LH-SDFs at Storm Shelter, RARIRSA-MEL8 and RARI-RSA-PEL4 could easily have been made by the same group or by groups in contact with one another. Nevertheless, in spite of their similar staffs, the significant differences between the three LH-SDFs at these three sites suggest that the images are meant to be seen as as different from one another. While they all show a concern with the head and facial area, most LH-SDFs show a similar lack of detail in the abdomen as the SDFs at RARI-RSA-RON1 (Fig. 59) and RARI-RSA-TYN2 (Fig. 60) that I described earlier. At RARI-RSA-MEL9, for example, the SDF has a large head with facial details that include a chin, upper and lower lips, nose and an eye (Fig. 67). The face is painted in an off-white colour while the rest of the figure’s body is painted entirely in a black colour with the exception of three small patches of red pigment on the top of the head that probably represent a headdress. The figure appears to hold a number of arrows or arrow-like objects that have differently coloured sections, possibly indicating the linkages common in real and painted link-shaft arrows. If these are meant to represent arrows, they are greatly exaggerated in size. In addition to the arrow-like objects, there is a second figure on the back of the SDF. Sometimes when owners-of-po141 Fig. 64. Dowson’s representation of the SDF from RARI-RSAMEL8. Note the exaggerated size of the penis and the unrealistic number of bows (16) in the hunting bag. Red is represented by solid and white depicts a lighter shade of red. Figure is 52cm in height. tency dance and they wish to instruct a novice in the workings of supernatural potency, they may carry the trainee on their backs (Katz 1982: 47, Keeney 1999: 25). Such ‘piggy-backed’ figures may also represent spiritsof-the-dead, which shamans also carry on their backs (Hollmann 2003: 95-96, Keeney 2003: 98). This further emphasizes that SDFs and LH-SDFs are depictions of owners-of-potency. Whatever the figure on the back represents, it is clear that the lower body of this LHSDF—even allowing for fading of the paint—was never depicted in as much detail as the head and facial regions. Fig. 65. Reinterpretation of RARI-RSA-MEL8 as a LH-SDF. Note that the figure is associated with a staff-like object that is similar to objects associated with SDFs at Storm Shelter and RARI-RSAPEL4. Red is solid, stipple is light red. Figure is 52cm in height. described it in Chapter One, the image has a ‘HumptyDumpty’-like appearance (Fig. 63). As with the SDF at RARI-RSA-RON1, the primary LH-SDF at Storm Shelter has no hands and the figure’s arms end simply as stumps. Again, as with the SDF at RARI-RSA-RON1, this appears to be a deliberate omission by the painter/ painters as there is no evidence of pigment fading over time and substantial portions of the image are painted in white, which, as I have mentioned, is the most fugitive of the pigments. While the LH-SDF at Storm Shelter has no legs and no hands, there are some images at rock art sites immediately adjacent to the Nomansland Core In more extreme examples than at RARI-RSA-MEL9, Study Area that completely lack a post-cranial body. certain anatomical features below the head may be entirely absent in LH-SDFs. The primary LH-SDF at One such site is RARI-RSA-CLO1 where San painted Storm Shelter, for example, has no legs; instead, as I a single human head in outline (Fig. 68). At another site, 142 Fig. 66. Each SDF and LH-SDF has a unique headdress. This suggests that the figures depict specific individuals. Solid indicates Red/Orange while stipple represents an off-white colour. Scale in centimeters. Fig. 67. The post-cranial bodies of some LH-SDFs are treated in far less detail than the head. This figure has a detailed face but the abdominal areas have very little detail. Solid represents black. Stipple indicates red. White is portrayed as clear. Scale in centimeters. RARI-RSA-WAR5 (Fig. 69 & Fig. 70), there is a two metre-long panel on which are painted several intriguing images. On the far left of this panel are two large anthropomorphic figures while on the far right there are two eland and a cluster of four Eldritch Images, one of which has two dancing sticks. In the centre of the panel, in the lower section are two anthropomorphic figures that are over 35cm in height; both are highly detailed and both have elaborate heads. In addition to these two figures with highly detailed heads and three other less-detailed anthropomorphic figures, there are five bodiless heads painted in profile in the central area of the panel. While the neck areas of these five figures are depicted, it is clear that the rest of the body below the neck area was never painted. The largest of the heads is just over 23cm in size and is depicted with a nose, upper and lower lips and a chin. From the nose, blood streaks across the face and the image as an eye tapered at both ends in a manner reminiscent of the two LH-SDFs at Storm Shelter. On the head is a crown-like head-dress. To the left of this figure, are a further four heads, painted slightly smaller and with less distinctive facial features than those of the large head. Nomansland and its environs are the only southern African rock art areas so far known where images of LHSDFs and images of anthropomorphic heads without bodies are found in suhc numbers. In other San rock art areas, such as the Matopos and the Brandberg, anthropomorphic images are typically conventionally proportioned and fully formed. Both these rock art areas are of significant antiquity, 10, 000 years BP for the Matopos 143 Fig. 68. The increasing importance of the head in the rock art of Nomansland and the concomitant demise of the post-cranial body find their ultimate expression in images such as this one, where paintings of disembodied heads exist together with other San images. Solid represents red. Stipple represents lighter shades of red. White is portrayed as clear. Scale in centimetres. (Walker 1996: 11-14) and at least 2700 years BP for the Brandberg (Breunig 1985). In both areas, the art occurs on hard-weathering and long-lasting granite and the art thus probably predates much of rock art in the Drakensberg, which we have seen, is on more fragile sandstone. Moreover, the LH-SDFs and the bodiless-head images are a very small percentage of the total number of Nomansland anthropomorphic rock art images. These three points—the limited geographic spread of SDFs, LH-SDFs and bodiless heads; the lack of these types of images in other southern African rock art areas known to be of greater antiquity than the visible images of the south-eastern mountains; and the relatively small number of these images in comparison to normally proportioned and full-bodied anthropomorphic images in Nomansland—all suggest that, over time, there was a progression in the rock art images of Nomansland. This progression was one in which the SDFs first emerged and then, over time, there was a shift in the prominence of certain body parts; the cranial area became more important while the post-cranial body de144 clined in significance. It is unlikely that the progression went in the other direction, from a portrayal of heads only to the representation of the entire human body. Such an argument would presuppose that Nomansland rock art was a tradition that developed completely independently from and in a direction completely opposite to that of any other San rock art area in southern Africa. Rather, it makes more sense to see the process amongst the SDFs in Nomansland as one that moved towards the declining importance of the post-cranial body. This process of the diminishing post-cranial body is different to the progression that Dowson argues exists in the rock art of the south-eastern mountains. In Chapter Two, we saw that his suggestion that individuality increased in San society and rock art over time through their interaction with others has difficulties. Importantly, I do not disagree with Dowson that the SDFs depict powerful individual potency-owners; what I dispute is that those dominant individuals only appeared after the San began to interact with their Bantu-speaking neigh- Fig. 69. RARI-RSA-WAR5. In this panel, several figures have been depicted without bodies. Some of these images are shown with the nasal blood characteristic of the Great Dance. Clear represents white. Solid depicts red. Scale in centimetres. bours. The burial evidence from the southern Cape is important in this regard, as it suggests that the San had well developed concepts of individuality and, indeed, notions of powerful individual shamans long before they interacted with Bantu-speakers. What does appear to be a progression related to San interaction with their Bantu-speaking neighbours and others, however, is the diminishing importance of the post-cranial body in the SDFs and LH-SDFs over time. As the evidence from RARI-RSA-MEL6 suggests that some of the SDFs, at least, were made during the last 500 years, the waning significance of the post-cranial body in these figures needs to be explained in terms of the changing social conditions in the Nomansland region at this time. I have already discussed the local socio-political conditions of the Nomansland area in Chapter Five. From about 500 years ago, however, the local interactions between the San and their neighbours were increasingly brought into the global colonial system. I now turn my attention to the impact of the colonial process on the local interactions between the San and their neighbours in Nomansland. Once we understand the interplay be- tween local and global forces, we are in a better position to comprehend why the post-cranial body diminished in significance in the rock paintings of Nomansland. I describe five aspects of the colonial process and its impact on the Nomansland region. These are explo- Fig. 70. Close-up of one of the disembodied heads at RARI-RSAWAR5. Note the nasal blood. The head is 23cm in height. 145 ration, shipwrecks, slavery, war and mission stations. Nomansland in a Global Colonial System Colonialism in southern Africa, and in other parts of Africa, is often understood in terms of a frontier (Alexander 1984, Gilliomee 1979). This is particularly the case with historical studies of the south-eastern seaboard. On the one hand, Bantu-speaking agropastoralist communities are seen as migrating southwards along the eastern shores of Africa. On the other hand, European colonists are understood as moving outwards, north and east, from the Cape Colony. In their move eastwards, they encountered the Bantu-speakers and a series of frontier wars took place between the two groups. There have been a number of sophisticated discussions on the usefulness of the concept of frontier to historical analysis and some theorists have laboured on defining more precisely different types of frontiers that account more adequately for different types of interaction (Kopytoff 1987). Nevertheless, the word ‘frontier’, even in all its theoretically defined permutations, still conjures up the image of map with a line drawn across the surface, dividing colonizer and colonized (Legassick 1980). The image of this line across a map also creates an image of two groups of people, with two distinct and incompatible cultural packages, who meet at this line. Typically, conflict follows. While the notion of frontier may be useful for understanding historical processes in some parts of the world, I argue, in this chapter, that it is not a useful way of understanding the social processes in south-eastern South Africa, including Nomansland, over the last 500 years. In particular, colonialism cannot simply be described, either as an ideology or in its physical manifestation, as a process that slowly moves across a map, continually pushing the line that demarcates the frontier ever farther into the interior. Indeed, the idea of a line on the map masks more complex processes of social and cultural hybridization that were taking place long before the official boundaries of the Cape Colony extended to the edges of Nomansland. I discuss these processes by considering the role of explorers, shipwrecks, escaped 146 slaves, war and, finally, mission stations in the production of hybridized cultures along the south-eastern seaboard. Explorers By the mid-eighteenth century, official knowledge of the territories to the east of the Cape Colony extended only to the territory north and south of the Outeniqua Mountains. In order to assess the potential of the country east of these mountains, the first official colonial expedition to these territories was organized in 1752. At this time, hunters and trekboers (pioneer farmers on the edge of the Cape Colony) had already passed these mountains to the grasslands beyond (Forbes 1965: 7ff.). Indeed, Hermanus Hubner had undertaken an extensive hunting expedition into the territories along the south-eastern seaboard in 1736 in order to obtain ivory through trade. The Hubner expedition probably ventured as far as Mpondoland, between the sea and the area that would become known as Nomansland. Hubner’s expedition was attacked on their return journey to the Cape. Hubner’s expedition was not officially sanctioned and it is only known because of the attack. It is thus very likely that Hubner’s was not the first expedition to seek ivory beyond the boundaries of the Cape Colony. Indeed, the returning survivors of Hubner’s party encountered other trading and hunting expeditions on their way out of the Cape Colony. It was partly as a result of unsanctioned expeditions such as Hubner’s that the Dutch authorities at the Cape decided to investigate the areas beyond the Colony’s official eastern borders. In order to assess the potential of the land beyond their boundaries, officials at the Cape dispatched Ensign August Frederik Beutler with an assistant, Carl Albregt Haupt, to the land beyond the eastern boundary of the colony in 1752. The expedition of 71 people and 11 wagons, including a small boat, left Cape Town on 29 February of that year. A few days after the expedition left, on 31 March and then again, two days later, on 2 April, the expedition met solitary French sailors. These two sailors had been aboard the French sloop Le Necessaire from which they had been sent ashore to fetch water on 27 February. Their boat overturned in the surf, leaving the two sailors stranded. These two sailors informed Beutler that the French were exploring the south-eastern shores of South Africa from Mauritius with the view of establishing a settlement along the south-eastern seaboard (Forbes 1965: 9). Already, by the mid Eighteenth Century, then, explorations were taking place, by either foot or ship along much of the south-eastern coastline by different European colonial powers. municating in a sophisticated manner. The Khoekhoen were thus on good terms with the Cape Nguni-speaking peoples. That the guides would mislead the expedition at the request of Cape-Nguni speakers suggests that, already at this early stage, there was a shared culture of suspicion and even resistance to colonialism amongst the different peoples along the south-eastern seaboard. Shipwrecks Although the early travellers passed through the Eastern Cape and came close to Nomansland, their journeys were usually transitory and the effects of their stay seemingly brief. There were, however, more enduring interactions between European settlers and the indigenous peoples along the south-eastern seaboard. Almost from the beginning of European exploration of the southern African coast, numerous vessels wrecked along those shores, leaving stranded survivors. In the case of the already-mentioned Le Necessaire, the sailors were fortunate that they were close to the Cape Colony and thus encountered settlers who could assist them. Other survivors were not so fortunate and before the south-eastern shore was settled, they were faced with the daunting prospect of walking back to the Cape or, alternatively, heading north for the Portuguese settlements in what is today Mozambique. The arduous and hazardous journey in either direction forced some of those who were stranded to seek an alternative life amongst the indigenous populations living along the south-eastern coast. On their way through and then out of the Cape Colony, a number of Khoekhoen guided Beutler’s party. On the 27th June, four months into the journey, at a point close to the Kei River, an altercation occurred between Beutler and his guides. The guides had led the party to the Kwenurha River where they could apparently find no crossing place. Beutler suspected that the Khoekhoen guides were deliberately misleading him at the insistence of the local Cape-Nguni groups, who, it appears, wished to discourage the advance of the expedition. Indeed, Vernon Forbes (1965) comments that the route taken by Beutler’s expedition is often peculiar and at odds with the topographically obvious route and it thus appears that either the guides were incompetent or they deliberately misled Beutler. Suspecting the worse, Beutler’s party tied up their guides and threatened them with death. Under duress, the guides did find an easier way past the Kwenurha River, suggesting that Beutler’s suspicions were well founded. Having crossed the river, Beutler eventually got as far as the Qora River in early July before return- One of the most famous of these examples is that of ing, eventually reaching Cape Town on 6 November, the Grosvenor, a British East Indiaman that wrecked of after some eight months in the field (Forbes 1965: 19). the Transkei coast in 1782. Of the 110 survivors, only a handful ever returned to the Cape Colony after four Beutler was not the last explorer to venture along the months of walking. There were nine women who sursouth-eastern seaboard and, indeed, as the Hubner ma- vived the wreck and who, together with the captain and terial shows, he was not the first. The importance of his some sickly and injured people, could not keep up with expedition, however, lies in the rebellion of his guides. the other survivors and so headed inland up the Umtata The events surrounding their restraint at the Kwenurha River. Rumours that the indigenous people along the River clearly illustrate that the indigenous peoples along coastline had abducted these women prompted two resthe south-eastern seaboard were not isolated from one cue expeditions. The first left for the site of the wreck another but, in spite of different languages and the as soon as the first survivors reached the Cape Colony spatial distances between them, were capable of com- but was unsuccessful in locating the women. For years 147 after the wreck, however, stories of European women living amongst the Cape Nguni-speaking peoples trickled back to the Cape Colony. These rumours led to a second expedition some eight years later. Known as the van Reenen expedition, this second venture also failed to locate any living survivors of the Grosvenor. This second expedition did, however, discover a woman of European descent living amongst the Mpondo people (Kirby 1954: 4-5). Various historical sources show that she survived a shipwreck at the age of seven that occurred along the south-eastern coast sometime between 1740 and 1750, some 30-40 years before the Grosvenor. It appears that she was English because she was named ‘Bess’, a corruption of the abbreviated ‘Beth’ for ‘Elizabeth’, and because the girl had an Indian servant, who also survived the wreck. The Mpondo named Bess ‘Gquma’ and she married a man called Sango, who was leader of the AmaTshomane clan. Together they had five children. She died sometime in the first two decades of the Nineteenth Century and was buried where she had lived on the Umgazana River. The descendants of Gquma and Sango were eventually so numerous that they were identified as a distinct group and known as the Abalungu, a name meaning ‘white people’(Kirby 1954). A number of historical sources also mention the AmaMholo (referring to Moors), a group of people living in the Transkei areas that were seemingly descended from Lascars cast ashore in the early eighteenth century (ibid.). Gquma was certainly not the only foreigner to marry into one of the Bantu-speaking groups along the southeastern seaboard of South Africa. Indeed, one of the Grosvenor survivors, John Bryant, was too badly injured to attempt the walk back to Cape Town. Together with another survivor of the wreck who was mentally inhibited, Bryant remained on the beach where the Grosvenor wrecked. Bryant recovered from his injuries and eventually married into one of the Cape Nguni-speaking groups along the coast. In 1825, Fynn (1950: 12), a much younger man than he was when he went to the Nomansland area in 1848, met one of Bryant’s children in the Transkei territories. The survivors of the Gros148 venor were not the only people to stay behind and settle amongst the local communities along the south-eastern seaboard. Indeed, one calculation of the colonial period shipwrecked survivors known from historical sources to have remained behind amongst the indigenous communities of the south-eastern seaboard, places the number at well over 1000 people (Bell-Cross 200: Pers. Comm.). It is clear from the material pertaining to shipwrecks along the south-eastern shores of South Africa that Europeans and Asians were marrying into Cape-Nguni communities at least from the early 1700s and probably much earlier. These unions were not simply ones in which cultural ideas were exchanged; they were relationships that produced children whose physical characteristics were a combination of features from European, Asian and African peoples. While there is no substantial evidence concerning the intermarriage of shipwrecked people with San, I have already noted the intermarriage between Nomansland San and their neighbours, specifically the Mpondo and Mpondomise, throughout the south-eastern parts of South Africa. If European and San genes were extensive amongst the Mpondo and Mpondomise, then it is not infeasible that many individuals of these two groups carried genes from both groups. From a genetic perspective, then, the communities of the south-eastern seaboard were far more mixed than traditional anthropological labels might lead one to expect. It is thus not useful to understand colonialism as a process entailing the movement of a frontier along the south-eastern seaboard. Instead, it is better to see the area over the last 500 years as one undergoing a process of creolization in which San, Khoekhoen, Bantu-speaker, European and Asian genes moved freely across supposed cultural divisions. Together with genetic changes, new ideas and practices emerged out of these interactions. The appearance of these new bodies and new ideas posed significant challenges to San potency-owners and, as we shall see shortly, it is in the rock art of Nomansland that we can see the taking up of these challenges. Escaped Slaves A day after the survivors of the Grosvenor began their long walk to the Cape, they encountered a man called Trout. Trout was of Malay descent and he had been a slave in the Cape Colony. He managed to escape and made his way along the south-eastern coastline to the Transkei area, where he settled amongst the Cape-Nguni groups. Trout spoke Dutch, which fortunately one of the Grosvenor survivors understood and so they were able to obtain useful information for their journey. Trout warned the survivors that the Cape was far off and that there were many large and dangerous rivers that they would have to cross on their walk. He added that they would encounter immense difficulties from the many peoples and wild animals they would meet. Before he left them, Trout gave them one last piece of advice—he urged them to stay along the coast, because inland they would encounter the ‘Boshemen Hottentots’, who would almost certainly kill them all (Kirby 1953: 73). Later on, the van Reenen expedition also met Trout and he told them that the people who had survived the wreck but did not reach the Cape were all dead (Kirby 1958). Trout was certainly not the only escaped slave to end up in south-eastern South Africa. The presence of escaped slaves at a place so far away from the Cape Colony might seem odd at first. As a number of historians have pointed out, however, the policies of the Cape Colony towards escaped slaves and those that assisted them created a situation where escapees fled towards the east and thus towards the Transkei area. In order to cope with the problem of escaped slaves, the Cape colonial authorities threatened to punish by death anyone found harbouring or assisting escapees. The indigenous groups living in and immediately adjacent to the Cape thus chased escaped slaves away or attacked them, forcing them to travel well beyond the boundaries of the Colony in order to reach safety (Penn 1999: 73ff.). Moving northwards into the semi-arid areas of the Karoo was a particularly difficult route to take, and so many escaped slaves opted to take the coastal route, going eastwards as this area was more lush and food and water were available to those who knew how to obtain them. As with the shipwrecked survivors, this eastward movement of escaped slaves brought people from diverse linguistic and cultural backgrounds into contact with the indigenous communities living along the southeastern shoreline. Again, this contact was not simply superficial but—as with Trout—involved long-term co-habitation and intermarriage that produced children and new ideas and practices. It is clear from the material on shipwrecks and escaped slaves that Nomansland and its environs were places of genetic and socio-cultural hybridization long before Europeans settled in the area. The Bushman Wars While the settlement of shipwrecked survivors and escaped slaves from the Cape Colony did much to bring the indigenous people along the south-eastern seaboard into the global colonial system, these interactions were initially limited. Then, in 1774, an event took place that was to lead to a significant increase in the interaction between the San of the south-eastern areas and other peoples. This event would also mark the beginning of the process that would ultimately see the south-eastern San reduced to a servile status both within European colonial and Bantu-speaking communities. In that year war erupted along the eastern and northern boundaries of the Cape Colony between Trekboers and San. The Trekboers were continually encroaching on San land and they were destroying the wild animal population, upon which the San relied for meat. While this led to tensions, a more immediate cause for anger was the Trekboer habit of abducting San children for slavery. It is not surprising then that the San launched a highly mobile hit-and-run campaign against the Trekboers. They began what was one of the first wars of resistance against colonialism in South Africa. The raids that the San conducted were, to the largest possible degree, for the purposes of obtaining domestic stock from the Trekboers to eat. Repeatedly, where the San could have massacred the settlers, they did not do so. Sadly, the reverse was far from the case and even before the war of 1774 the destruction inflicted by the colonists was horrific. Two accounts by observers illustrate the extent of the carnage: 149 In this expedition they had killed about a hundred and made prisoners of twenty, chiefly small children. Some of whom they at this present juncture had with them. It was said that in a similar expedition in 1765, 186 had been killed. None of the Christians that went on this expedition were either killed or wounded (Thunberg n.d.). I have heard of one man, who is represented as an estimable character in other respects, declare, that, within a period of six years, the parties under his orders had either killed or taken 3,200 of these unfortunate creatures (Colonel Collins in 1809. See Moodie 1960: Part V: 7). In the light of these quotes, Trout’s warning about the ferocious ‘Boshemen Hottentots’ to the survivors of the Grosvenor should come as no surprise—years of harsh war and the brutal treatment dealt to them by the Trekboers had hardened them. In spite of being described as ferocious, the San were not victorious against the Trekboers. The hit-and-run tactics that they employed during the 1774 war produced only limited successes. The Trekboer commandos, mounted on horseback and armed with guns, easily caught the San who were on foot and had only bows and arrows. While the poisoned-tip arrows caused great fear amongst the colonists, the weak bows did not allow the arrows to be fired over great distances and while some Trekboers died from poisonedarrow wounds, the numbers of wounded and dead were startlingly disproportionate to the numbers of San who were killed. While resistance did not end in 1774, many of the San living along the eastern boundary of the Cape Colony were forced into indenture on colonial farms from this time onwards. As in the Ghanzi District of Botswana during the 1970s, a distinction emerged between San living on farms and those that managed to cling to their hunting and gathering lifestyle. In order to maintain their independence from the Trekboers, some San would have had to move eastwards, away from the Cape Colony’s boundaries. Shortly after the 1774 war, the First Frontier war erupted between the Cape Colony and the Cape Nguni-speaking Xhosa. This was followed by several more engagements that continued well into the nineteenth century. The general disruption 150 of these wars caused an influx of Khoekhoen, Ngunispeakers, San and escaped slaves into the areas along the south-eastern seaboard adjacent to Nomansland. The trickle of people into the area, that had begun with shipwrecked survivors and escaped slaves, was turning into a flood by the start of the Nineteenth Century. Many of the refugees, wanderers, escapees and others who flowed into the Nomansland area landed amongst the various San groups living there. With the possibility of sustaining a hunting and gathering life diminishing with each passing year, the San turned increasingly towards raiding. As they acquired horses and guns from the encroaching colonists, they became more mobile and more formidable and as livestock raiding became the way of living during the nineteenth century, the San and others adapted quickly. By the mid-nineteenth century, when Silayi went to live with Nqabayo’s people, virtually none of the !Gã !ne groups living in the Nomansland area were composed solely of San. As we have seen, Silayi was introduced to Nqabayo through Hans, a Khoekhoen man. Nomansland is thus best thought of as home to a number of groups whose core membership was San, but with people from different linguistic and cultural backgrounds joining the core for varying durations of time. Sometimes, those who joined the San would stay for a short period but others, such as Silayi, stayed with them for several years. Whatever the duration, ultimately those who joined the San either left voluntarily, to return to their own people, as Silayi returned to the Thembu, or they were forced to leave when the San groups were destroyed or disbanded through military action. As the non-San membership of these groups was rarely if ever permanent, I term the loosely affiliated groups of San and non-San that was so common in Nomansland and its environs during the nineteenth century, Core San-Transient Follower Groups. Mission Stations The political and economic unity of these Core SanTransient Follower Groups was based very much on livestock raiding, from either European colonists or from the powerful Bantu-speaking groups living along the eastern seaboard. It is within this pattern of livestock raiding that mission stations first became important to the Core San-Transient Follower Groups. When mission stations were established in the Transkei region from the 1820s and onwards, they soon provided the nuclear hubs around which the dispossessed and disaffected Core San-Transient Follower Groups orbited. The stations provided food and shelter and places around which to congregate but, most important of all, the missionaries at these stations proved to be useful interlocutors between the colonial authorities and the peoples of the Nomansland area. Some of the mission stations were specifically established for the San, while others were built amongst the various Cape Nguni-groups. The growing importance of missionaries in the interaction between colonial authority and the people of Nomansland and adjacent areas led some groups to settle near mission stations. Perhaps, the best known of these groups was that under the San leader, Madolo. Madolo’s career is well documented and thus allows us a glimpse into the shifting relationship between Core San-Transient Follower Groups, missionaries, Cape Nguni-groups and colonial authorities. The colonists knew Madolo as Madoor or Madur and he lived in the area immediately south of the Mt. Arthur Mountains. Stow (1905: 201) claims that Madolo originally occupied land just outside of the town of Whittlesea. Christopher Saunders (1977), has tracked the life and movements of Madolo more closely than Stow. According to him, sometime around the year 1835, Madolo moved away from Whittlesea to the tributaries of the Kei. Xhosa and Thembu encroachment on Madolo’s land seems to have prompted this move. Their intrusion, in turn, was the result of colonial advancement to the west as well as the product of displacement of various groups during the Mfecane to the east (Saunders 1977: 146). In 1838, James Read Senior, of the London Missionary Society, who was stationed at Philipton on the Kat River was told by one of his congregation, who was himself of San descent, of relatives who were living a ‘wild existence’ at the tributaries of the Kei River. An exploratory party sent from Philip- ton discovered Madolo north of the Black Kei River (ibid.: 147). In September 1839, Read moved to the area that Madolo inhabited and established a new mission that became known as the ‘Bushman School’ (later it was called New Bethelsdorp and still later Freemanton). This is the same ‘Bushman School’ that Umgudhlwa mentions in his statement to Stanford, given in Chapter One. Shortly after being established, the mission station became embroiled in a dispute between Mfengu and Mpondomise and the Mpondomise laid siege to the mission; Madolo’s San and the Mfengu broke the siege by attacking the Mpondomise but lives were lost on both sides (ibid.:148). Following this engagement, the station flourished in the early to mid 1840s. In 1842, there were 15 San families at the mission station and by the mid 1840s some 300 heads of families. Many of these were Khoekhoen- and Bantu-speakers and Madolo appears to have been in charge of all these groups (ibid.:148). In 1846, another ‘frontier’ war broke out between the Cape Colony and the Xhosa. The Colony requested Madolo to send a force to protect the Kat River Settlement during this so-called War of the Axe. Madolo led a force of some 200 men, including San, Khoekhoen, Mfengu and some Thembu to assist the Colony. In their 15 months of service—for which they received no compensation—Madolo’s group distinguished themselves in engagements with the Xhosa and their Thembu allies (ibid.:149). The war, however, saw the decline of the mission station and Madolo and his followers became poverty-stricken; Madolo himself complained to visitors that he did not get out hunting after the war (ibid.:150). In 1849, land was allocated to Madolo’s group as well as to other San under the leadership of a man called Flux (Fluks) Lynx. Soon, however, others were settling on their land. In September 1849, a Thembu group moved onto the land allocated to Flux Lynx and he appealed to the colonial authorities. He never received a reply and after a year of no action on the part of the Cape Colony, Lynx took matters into his own hands in September 1850 and he attacked the Thembu. The Thembu, however, successfully repelled Lynx and he lost ten of his men in the engagement. The colonial authorities sent 151 troops to quell the fighting and arrested 23 of Lynx’s followers, confiscating 19 rifles in the process (ibid.:151). Lynx, together with some of his followers, decided to flee colonial jurisdiction. Madolo did the same and he crossed the Kei River out of the Cape Colony. Madolo and Lynx then settled in rock shelters along the White Kei River where they numbered some 62 armed people. San appear to have fought to retain their identity as a separate people. It appears that not only were they successful for much of the Nineteenth Century in maintaining their separate identity but that they exercised control and leadership over people from many different cultural backgrounds. The ‘Bushman School’, for example, was home to numerous San, Khoekhoen and Bantu-speaking people, at times numbering several hundred When yet another war erupted in 1850 between the individuals and Madolo was their acknowledged politiCape Colony and the Xhosa, Madolo, once held to be a cal leader. Importantly, as Saunders points out, Madolo close ally, became regarded as an enemy by the colonial retained his ‘San’ identity throughout his life. The politipowers. Shortly thereafter, their status in the colonial cal ascendancy of Madolo and to a lesser degree that of government’s eyes shifted again during the Mlanjeni Nqabayo and Mdwebo, is even more remarkable given War. During this conflict, Madolo’s people again fought the supposed political marginalization and subservience for the Cape Colony against the Thembu. In a remarka- of San when interacting with Bantu-speakers. In the ble irony, the Cape government did not reward Madolo case of Nomansland, revisionist arguments that conwith land at the end of the war but, instead, offered tinually treat San as an underclass clearly do not hold. the Thembu some of Madolo’s land. In 1854, the Cape Colony appointed a committee to investigate the affairs The mission stations on the edge of Nomansland were of the ‘Bushman School’; the committee found that natural gathering places for these Core San-Transient Madolo had been treated with great injustice in spite of Follower groups and it was through the missionaries his significant contributions in the Cape Colony’s vari- that the San negotiated with colonial authorities. Inous ‘frontier’ wars. The committee’s findings were never deed, it appears that the missionaries were often mataken up and little was done to assist Madolo (ibid.:152). nipulated by the San in Nomansland. Nqabayo and Nevertheless, he seems to have stayed in the rock shelters Mdwebo, as we have already seen in Chapter Five, along the White Kei until 1855, when St. Mark’s Mission drew on Wesleyan missionary support—albeit through Station was established close to the shelters. According their Bhaca allies—in order to thwart Fynn and it was to Stow (1905: 204), who visited these shelters in the ultimately through their efforts that the missionarlate 1860s and met some San painters in the shelters, ies pressurized the colonial authorities to recall Fynn Madolo was last heard of in 1856 when he left the area from his position as agent with Faku. Key to the misand moved still closer to the Drakensberg Mountains. sionary defence of the San and their Bhaca allies was their belief in the innocence of the groups in raiding Madolo was not the only San with political influence activities. Yet, as Silayi testified some thirty years after in the Nomansland area during the nineteenth century. Fynn’s allegations, Nqabayo’s group (and almost cerNqabayo attracted Khoekhoen and Thembu followers tainly Mdwebo’s as well) were guilty of stock theft. This and, Mdwebo’s group, who numbered only fifteen fami- clearly demonstrates that the San and the Bhaca (who lies of San, was observed at times to have several hun- were implicated in the raids) misled the missionaries. dred followers. By the mid-nineteenth century the San groups in the Nomansland area were no longer homog- The historical evidence concerning exploration, shipenous entities. Instead, these groups were composed of wrecks, escaped slaves, warfare and mission stations a core of San and a transient membership of other cul- that I have discussed show that although Nomansland tural identities. As we have already seen in Chapter Five, was only settled relatively late in the colonial period, it in spite of these heterogeneous cultural groupings, the has been part of the global colonial system for at least 152 500 years. From early on in colonial expansion and exploration, the south-eastern seaboard has been involved in processes of cultural interaction, that cannot adequately be described in terms of a frontier. More than a simple swapping of ideas and practices across impermeable cultural boundaries, these processes entailed prolonged periods of creolization and hybridization and the emergence of new identities. Indeed, the intermarriage of people from European, Asian and Bantu-speaking backgrounds, taken together with the widespread observation noted in Chapter Five of San intermarrying with Bantu-speakers, produced not only new cultural forms but also new physical traits. Rather than see Nomansland and the south-eastern seaboard as made up of different and discrete cultural identities, the area should be understood as one of shifting identities As Clifton Crais (1992: 14) points out about the social conditions of the south-eastern seaboard during and before the colonial period, the area was “…a site of ethnic ambiguity and intensive social construction. New identities were assembled, older ones reshaped”. It was often around the various San groups in and around Nomansland that many of these identities were fashioned and individuals from very diverse cultural backgrounds coalesced around influential San leaders. This was particularly the case from the late eighteenth and early nineteenth century, following the devastation of the various so-called ‘Frontier Wars’. These mixed groups of Core San-Transient Followers became militarily powerful, particularly during the mid-nineteenth century and they conducted raids on the livestock of the colonists as well as that of the powerful Cape Nguni-speaking groups in the area. Indeed, at times they even engaged in open warfare against the more powerful Bantu-speaking groups in the area. Within these mixed groups, as we have seen in this and the previous chapter, the San largely maintained their own identity. The military influence and raiding capability of these San groups, however, began to wane from the 1850s onwards. Through colonial intervention, as in the case of Madolo and Lynx and through direct attacks, such as that of Umgudhlwa on Nqabayo’s band, the San military presence was largely destroyed and the large groups of Core San-Transient Followers were scattered. From this time on, the San of the south-eastern seaboard seem to have existed as small groups only. Yet, some of these small groups managed to retain their autonomy for many decades and, as we saw in Chapter One, descendants of Nqabayo’s band still lived largely independently at the turn of the twentieth century. It is against the historical background of the re-fashioning of identity over time that I have outlined here that the diminishing importance of the post-cranial body in the SDFs of Nomansland and its environs needs to be understood. In the final section of this chapter, I discuss how these historical circumstances would have led to the bodily changes in the rock art of Nomansland that I have outlined. The Death of the Post-Cranial Body Earlier in this chapter, I mentioned that I do not contest Dowson’s argument that the SDFs represent powerful shamans. Indeed, the features of some of the newly discovered images support his claims. In particular, the association of many of the SDFs with staffs is informative. Staffs appear to be very much associated with authority in the historical San and Khoekhoen communities in South Africa. Colonists on the border of the Cape Colony, for example, from the later 1700s nominated a local Khoesan person as the ‘chief ’ of a particular group and then issued them with a staff of office so that they could be identified as leaders of their respective groups (Newton-King 1981: 68). There is also a suggestion that the /Xam identified staffs with authority. Diä!kwain (Bleek 1935: 11-12), for example, commented on a copy of a rock painting of one man, five women and small antelope: They seem to be dancing, for they stand stamping with their legs. This man who stands in front seems to be showing the people how to dance; that is why he holds a stick, for he feels that he is a great man. So he holds the dancing stick, because he is the one who dances before the people, that they may dance after him, for the people know, that he is the one who always dances first, because he is a great sorcerer. 153 Direct comments on rock art images by San, such as this one made by Diä!kwain, thus support an interpretation of the SDFs/LH-SDFs associated with staffs as depictions of authoritative shaman figures. In addition to such features as staffs, there are often specific aspects of a particular SDF or LH-SDF that further suggest that they represent prominent and powerful potency-owners. A good example of this is the primary LH-SDF at Storm Shelter. This figure, with a pointed ear, an unnaturally large eye and beret-like headdress reaches toward a staff. The figure appears to be painted underneath the large feline that is also present in this upper left side of the main panel. Normally, traces of the underlying picture are visible through the semitransparent paint of the overlying image in superimpositioning. Careful scrutiny of the Storm Shelter primary LH-SDF, however, does not reveal any trace of pigment or line of the large anthropomorphic figure underneath the feline (Fig. 63). Indeed, it appears that the feline was painted first and that the LH-SDF was painted afterwards and arranged in such a manner as to suggest the relationship of superimposition. Alternatively, if the feline were painted first, then the painter intended the LH-SDF to appear as if emerging from behind the feline. Whatever the painter’s intention, it is clear that the LH-SDF is intimately linked to the feline. As we saw in Chapter Five, superimposition in San rock art is an important technique by which overlying and underlying images comment, build upon and reinforce the meaning and significance of one other. It is thus no coincidence that the feline and the primary LH-SDF at Storm Shelter are closely associated and the significance of this relationship needs to be sought in San ethnographic beliefs about felines as well as depictions of felines in the rock art of south-eastern mountains. Felines play a similar symbolic role in many San groups. This role is one that is often opposed to animals that are not meat-eaters. The !Kung, for example, draw a distinction between carnivores and herbivores (Biesele 1978: 927-928) and the relationship between the two serves to characterize a number of oppositions in their thinking. Men, for example, are related to women in the same 154 way that carnivores are related to herbivores. Men thus stalk women, who are their prey and who are referred to as ‘meat’ (Biesele 1993). Other San groups also make the distinction between carnivore and herbivore. The !Ko talk of biting and non-biting animals (Vinnicombe 1976: 216) and amongst the /Xam, women spoke of their husbands as beasts of prey (Bleek 1933: 303, 388). Another important opposition associated with the carnivore-herbivore relationship is that between hostile and amicable; carnivores, particularly felines, are thought to be ‘angry’ and are associated with darkness. They share these characteristics with the spirits-of-thedead who, as we saw in Chapter Four, are angry and are attracted by the firelight of the dances at night. As with the spirits-of-the-dead, various San groups use respect words for felines. At night, the !Kung, for example, prefer to use the avoidance word, n!e, instead of the usual, //ka, for lion (Marshall 1969: 352). The /Xam also used respect terms for lions (Bleek 1932, Bleek and Lloyd 1911: 183, Currlé 1913: 117, Mackenzie 1871: 151). While felines share some similarities with spirits-of-thedead, they are also associated with potency-owners. Felines, particularly lions, can change into human form (Bleek 1932: 61), while owners of potency can change into lions (Bleek and Lloyd 1911: 187, Heinz 1975: 29, Katz 1982: 115, 227); felines, as well as potency-owners, can transform into hartebeest (Bleek 1932: 62); felines, as do shamans, can make it rain (Bleek and Lloyd 1911: 261); and both felines and potency-owners ‘know’ things through dreaming (Bleek 1932: 55). This close relationship between felines and San shamans is frequently depicted in the rock paintings in the south-eastern mountains (Lewis-Williams 1985). Depictions of felines will sometimes show nasal bleeding or they will be depicted as if pursuing other images, such as flying antelope, which, as I mentioned in Chapter Four, represent transformed shamans on out-of-body travel. The feline at Storm Shelter, with which the primary LH-SDF is associated, is not painted with nasal haemorrhage and it does not pursue transformed shamans but, through its superimposed relationship with the LH-SDF, is clearly associated with shamanic iconography. Importantly, the presence of a mane allows us to identify this image as a lion. It is unusual that species of felines at rock art sites are identifiable because, often, diagnostic features such as the mane of a lion and the spots of a leopard are omitted. It is a widely held belief, amongst the various San groups, that only the most powerful potency-owners can turn into feline, particularly lion, form (Lee 1968: 46). The false superimposition of the primary LH-SDF at Storm Shelter, then, appears to be an intentional effort to associate the figure with that of the feline. Linking the figure to this powerful animal thus makes a statement about the abilities and authority of the LH-SDF. As only the most powerful shamans can turn into feline form, this LH-SDF probably depicts an influential and important owner of potency. The SDFs and the LH-SDFs thus appear to portray powerful individual San ritual-specialists. The unique face and head of each SDF, especially the LH-SDFs, supports this argument. While many anthropomorphic figures in the rock art of Nomansland have faces and possibly represent individual San, the LH-SDFs, with their large heads and detailed faces appear to represent specific individuals more so than any other image in the Nomansland region. If they represent specific individuals then they are a form of portraiture. Of course, they are not portraits in the Western sense of the word because they often have features that are not realistic, such as the pointed ears of the two LH-SDFs at Storm Shelter. Instead, they are portrayals of individuals as they appear in the spirit world, behind the rock surface. Their non-real and animal features raise the question of how they were recognized as depictions of specific individuals by viewers of the art. The answer to this question lies in the distinctive headdress of each SDF and LH-SDF. We saw in Chapter Four that headdresses are often associated with individual identity amongst the /Xam. Diä!kwain, for example, identified the apparition of his wife by the cap that it wore when her other features were already partly transformed so as to make her only vaguely recognizable. In addition to the animal and non-real features of the images of the LH-SDFs, aspects of the facial features of these images are distinctly non-San. Indeed, they often appear to include physical features that are more closely associated with Bantu-speaking, Asian and even European groups. Indeed, the most intriguing of the LH-SDFs so far discovered is at RARI-RSA-MIE1 (Fig. 61). This figure is some 53.5 cm in size and is entirely of white pigment, but a thicker pigment than that of the translucent Eldritch Images. The figure has an unnaturally large head, with a twin trail of small white dots coming from the back of the head. As do many SDFs and LH-SDFs, the figure holds a long staff in its hand, thereby showing that it represents a powerful individual potency-owner. In the other hand, the figure appears to hold a handkerchief. Remarkably, the RARI-RSA-MIE1 LH-SDF has a straight nose and a large, rounded eye which are facial features characteristic of Europeans. Of course, while the figure has what are possibly European physical features, it also has antelope hooves, a clear marker that it is a transformed shaman. The figure is thus more than simply a depiction of a European. This LH-SDF thus probably depicts a transforming San potency-owner in the spirit world, who has incorporated aspects of European facial physiology. As the San were increasingly exposed to other peoples with different physical profiles during the last 500 years, they included these ‘other bodies’ in their painted representations of themselves. Images such as this one suggest that not only were the San painters of Nomansland cognisant of the processes of hybridization and exchange but that they actively sought to manipulate them in the painted depictions of their experiences of the spirit world. By including physical features from ‘other people’ in their depictions, the San artists of Nomansland were negotiating their identities as part of the hybridized communities along the south-eastern seaboard. In negotiating these new identities, however, the San did not simply paint themselves as ‘other’ but, instead, sought to refashion their identities by incorporating elements of ‘other’ physiologies into their San shamanic religious practise. The most obvious way of including such features is in the facial area as the post-cranial bodies of different groups are not that different. Consequently, 155 the decline of the post-cranial body and the emergence that as San hunter-gatherers were increasingly forced to of large heads with detailed facial features grew out of become sedentary, rainmaking became more important the refashioning of identity in the Nomansland area. for the San in their dealings with their Bantu-speaking neighbours. The influx of goods led to an economic In addition to the re-shaping of identity, a shift in the stratification that largely benefited the ritual-specialists. ritual role of potency-owners in Nomansland over time Dowson extended this argument and he suggested that is probably also responsible for the diminishing impor- the paintings became a vehicle for entrenching inequalitance of the post-cranial body. Lewis-Williams (1981: ties between potency-owners and ordinary members of 77) has pointed out that there are several overlapping the San communities. While working from different thecategories of San ritual-specialists, including shamans oretical perspectives, both authors saw the professionwho heal, shamans concerned with rain-making and alization of San ritual skills as leading to the influx of rain-control and those that are concerned with game material goods, which, in turn, led to social stratification. control. Amongst the Kalahari groups, healing appears to be the most widespread and significant of these tasks. As rain-making and rain-control became more imporThe vast numbers of images that refer to the curing ac- tant for the San in the Nomansland area, it is likely that tivities at the Great Dance that are found throughout the curing played less of a role. San healing, as we have south-eastern mountains suggests that a similar situation seen, is an activity very much directed at the post-crapertained for the San groups in this area. Nevertheless, nial body, particularly from the stomach to the neck. At as I mentioned in Chapter One, visitors to the Nomans- the dance, the owners-of-potency will lay their hands land region in the late nineteenth century noted that on anyone present who is thought to be ‘sick’. The sickthe San who were living in the area were famous rain- ness is removed from the patient by moving it along the makers; there is almost no comment about their healing spine and out of the back of the neck. The owners-ofskills. In particular, the San living around the Ncengane potency do this in two ways: they snort the sickness in Shelter area in the late nineteenth and early twentieth through the nose as they move their heads along the centuries were rainmakers to the Mpondomise (Calla- spinal pathway or they use a divining switch that they way 1919, Gibson 1891, Hook 1908, Scully 1913). We rub along the patient’s back, moving the sickness uphave also seen that Manqindi Dyantyi disclosed that her wards along the body until it is expelled out of the back father, Lindiso, her uncle Masela and her sister Chitiwe of the neck. Importantly, healing practices aimed at the were all rainmakers. Indeed, the independence of the removal of ‘sickness’ from the abdomen is a specifically San in and around Nomansland in the latter decades of San practise in southern Africa. Most Bantu-speaking the nineteenth century, in large part, depended more healing practices centre on the divination of the cause on their ritual rainmaking prowess whereas, in earlier of the sickness; remedies normally involve the ingesdecades, it had depended on their military proficiency. tion of various plant medicines (see Hammond-Tooke 1989). In addition, the cause of illness amongst BantuA number of researchers who have used the rock art speaking communities is usually witchcraft or behaviour to assist in the writing of San history have commented that is displeasing to the ancestral spirits. Unlike San on this evidence concerning the importance of rain- communities, however, the cure to affliction amongst making in the relationship between Nomansland San Bantu-speaking communities is the destruction of the and their neighbours. Both Campbell and Dowson, for agent sending the illness or proper veneration of the example, as we have seen in Chapter Two, argued that ancestors through various forms of sacrifice. In San San rainmaking activities for their neighbours led to the communities, as we saw in Chapter Four, the illness is influx of wealth. As San ritual-specialists made rain, so removed from the patient’s body and sent back to the they were paid in various forms. Campbell suggested spirits-of-the-dead. The San and Bantu-speaking sys156 tems of healing are thus very different and it is likely that, as the San of Nomansland became more and more dependent on their ritual role amongst the Cape Ngunispeaking groups over time, they slowly abandoned their traditional concepts of illness and bodily healing, opting for the beliefs and practices of their Bantu-speaking neighbours. With the decline in curing activities in Nomansland and the increase in rain-making and raincontrol, the abdominal areas below the head would have become less significant for San shamanic practice. In a context where the face and head were already becoming important in the reshaping of individual San identity, the changing ritual role of potency-owners would have hastened the decline of the post-cranial body in the rock art. The progression, then, from SDFs to LH-SDFs and ultimately to images of heads only, allows us to construct a history of the San body in the rock art of Nomansland. As we have seen from the theoretical discussion in Chapter Three, the body is intimately tied to identity construction. We should thus see the decline of the importance of the post-cranial body as evidence for a shift in the way individual identity is constructed within the San communities of Nomansland. Such a view, coming from a theoretical concern with the body, is more nuanced than a structurationist perspective that argues for the gradual rise of individuality, as seen in the SDFs, amongst the San of the south-eastern mountains over time. In addition to these shifts in identity, the changing ritual role of San potency-owners in Nomansland, from healers to rainmakers, also served to make the body below the neck less important to San shaman-artists. By drawing on the principles of the somatic past, we have seen in this chapter and Chapter Five, how rock art can be incorporated into the production of San history in the Nomansland region. In the next chapter, I turn my attention to how concepts of embodiment can shed further light on the way that the San socially consumed the painted images. 157 CHAPTER 7 THE CONSUMPTION OF SAN ROCK ART AS AN EMBODIED EXPERIENCE A humorous cartoon, illustrated in Figure 71, provides the central theme for this chapter. The cartoon shows a group of well-dressed and sophisticated socialites from the contemporary art world. They appear to attend the opening of an exhibition, except, in this case, the gallery is an Upper Palaeolithic cave, modelled on Grotte de Chauvet near Avignon in south-eastern France. The caption captures their jargon-filled and platitudinous conversation and the humour of the cartoon comes from the disjunction between the Palaeolithic rock art and the modern language that the socialites use to describe those ancient images: clearly, most readers do not think that Upper Palaeolithic painters appropriated “quotidian domestic icons”. This cartoon neatly captures a dilemma in rock art research. So often researchers write about gender, agency, power relations and other theoretical concerns without giving much thought to how those social issues were brought into operation through the art. Indeed, in much theoretical writing about rock art, it is as if the socialites of the cartoon were merely substituted with the ancient communities who made the images. We get a picture of Upper Palaeolithic image-makers or San rock painters standing casually before the images and debating their social impact; in such a view, the makers of the images are passive consumers of their work. In this chapter, I consider whether the notion of embodiment allows us to overcome this tendency in rock art research to treat the original image-makers as reactive viewers of their own handiwork. I contend, must begin by considering how the physical properties of both the images and the space of viewing those images impacted on the human body. By considering these properties and the way that they affected the body, together with ethnographic material and the concept of transference, I argue, we may better model San consumption of the imagery. Finally, I consider the broader social implications of the pattern that I have described. Importantly, as it has received much attention previously, I do not consider the production of rock art in any detail here, but only in so far as it pertains to consumption (see Lewis-Williams 1994, 1995). The Pattern of an SDF Site In Chapter One, I pointed out that Storm Shelter and other sites with Significantly Differentiated Figures have a pattern. This pattern includes a number of categories of images, some of which have been discussed in the literature previously and others which I have considered in previous chapters. At the heart of this pattern are the images of SDFs and LH-SDFs. Once these are present at a painted shelter, it is a reasonable expectation that the other images that comprise the pattern will also be there. These other categories of images, or elements, include the Eldritch Images; anthropomorphic dancing or clapping figures; anthropomorphic figures in exaggerated running postures (known as the runrun posture); a thin line that is often red and fringed with white dots but may also take on other forms; I begin by describing the ‘pattern’ of images at Nomans- three species of antelope that are usually depicted in land SDF/LH-SDF sites. I then evaluate recent efforts a numerical hierarchy that I termed the 1-2-3 arrangethat seek to understand the composition of rock art ment in Chapter One, in which eland (these are often sites. Any analysis of the pattern of Nomansland rock arranged as if they represent a herd, with one or two art that I describe needs to take care to avoid some of large bulls and many smaller cows and younger males) the pitfalls of these recent efforts. Any understanding are the most numerically prominent, then rhebok and of the patterned nature of Nomansland rock art sites, then hartebeest; at least one feline; and at least one rain158 animal. In addition, closely associated with the SDF/ LH-SDF sites but usually depicted on a different, adjacent or nearby rock surface are the Type 2 images. Analysis shows that these elements are frequently all or mostly present at the SDF/LH-SDF sites (Appendix 2). It is important to keep in mind that constructing categories of rock art images and then quantifying them is an approach now widely criticized in rock art studies (Lewis-Williams 1990, 1995, Lewis-Williams and Dowson 1994). In the case of the Nomansland SDF/LHSDF sites, for example, it is difficult to know at every shelter how many images have faded away or have flaked off as some sites are poorly preserved. Some of the elements of the pattern may thus have been present previously but, now, they are no longer visible. Moreover, some of the elements that I have just identified are present at some non-SDF/LH-SDF sites. It is thus difficult to know what elements are exclusive to SDF/ LH-SDF sites and which ones are not. In spite of this and other problems with such analyses, it is remarkable how frequent the elements that I have identified are present at SDF/LH-SDF sites. It is the regular occurrence of the elements that I have identified as present at SDF/LH-SDF sites, however, which is remarkable. With these problems in mind, I suggest that Nomansland SDF/LH-SDF sites are patterned and, as I mentioned in Chapter One, that this pattern does not reside so much in the arrangement of elements as it does in the presence of those elements. Indeed, no apparent patterned arrangement of these images is immediately evident at SDF/LH-SDF shelters and at some sites, for example, the run-run figures are painted at the top of a panel, while at others they are painted at the bottom. Nevertheless, future research may point out that the elements that I have identified are also spatially arranged. For example, it is common to find clusters of rhebok at the lower end of a panel and to one side of the majority of images. Spatial layout studies may thus show a statistical propensity for the placement of images, such as these, in a particular area on a painted rock surface in relation to the other elements. Moreover, these elements are al- most certainly not the only ones of the pattern—future research might well bring to light other repeated motifs in the sites that appear to us now only as random images. In addition, there may be further aspects to the pattern of SDF/LH-SDF sites that are not yet observable. Preliminary tests of pigment samples with a colourimetric test kit made by Bayer Diagnostics, known as ‘Hemastix’, has shown the presence of haemoglobin/ myoglobin in certain images, for example, that suggests an intriguing pattern as well. The tetramethylbenzidine and buffer chemicals in the Hemastix strips react with both haemoglobin and myoglobin and the more molecules of these that present, the darker the colour on the strip. The test strips have a sensitivity of 0.15-0.62 ng/μl, which is comparable to 5-20 intact red blood cells per microlitre (Williamson 2000: 28). Unfortunately, the strips also react with positively charged metal ions that may be found in manganese, copper and magnesium. Importantly, chlorophyll, commonly found in the cyanobacteria and lichens that may occur on rock surfaces and on painted images contains magnesium. Testing of pigment samples may therefore produce a positive result that is false. Residue specialists claim to have devised a technique for eliminating such false positives (Loy and Matheson 2000). By adding an equal volume of 0.5M disodium-ethylenediaminetetra aceticacid-disodium salt (Na-EDTA) of pH 8.0 to the extracted sample and then allowing it to mix for one minute, specialists point out that positive results based on the positive ions present in the magnesium in chlorophyll are eliminated from the test while the haemoglobin and myoglobin are not. By testing a sample using Hemastix strips, adding Na-EDTA to positive samples and then re-testing with Hemastix, false positive results may be eliminated. Even with these precautions, researchers point out that the test is still a presumptive one (Williamson 2000: 29). Ultimately, at issue is the range of substances that might yield false positives. Nevertheless, if Hemastix tests yielded consistent results on archaeological materials of specified and repeated criteria such as in repeated pigment types in San rock 159 Fig. 71. A humorous cartoon raises important questions about the way we think about the social production and consumption of rock art. art, then confidence in the technique would increase. In 1998, a permit was sought from and granted by the South African Heritage Resources Agency for the removal of samples of pigment from selected sites in Core Nomansland Study Area. The motivation for this application came from the evidence of Mapote. He explained to How that the binder in the images was blood and preferably eland blood. The aims of removing the samples were to establish the extent to which blood was present in the San rock art of the area and to determine whether blood was present in certain images but not in others. In total, 81 samples were removed from 46 images at 10 sites. A stainless steel scalpel was used to remove the pigment that was to be sampled (Fig. 73 & Fig. 74). This scalpel was dipped in alcohol before and after each sample removal and then allowed to dry naturally. At each site, a new scalpel was used while the old one was discarded. Each sample was removed as carefully as possible and all efforts were made to cause as little damage to the images as possible. In most cases, the samples were removed from areas that showed micro-flaking and from where it was apparent that the 160 pigment would flake away in the foreseeable future. Operating in this manner, only very tiny grains were taken from each image. The majority of the 46 images that were tested were sampled at least twice, with a few images being sampled three or four times, always from different areas. Each sample was placed in a sterile plastic tube and then they were removed from the field to the University of the Witwatersrand Archaeology Department where they were tested for haemoglobin using the Hemastix technique by Bonnie Williamson. Of the 81 samples taken, 12 were not subjected to the Hemastix test; this was done to allow for the possibility of the development of new tests in the future that would either show the present testing to be correct or false—in other words, these 12 samples acted as a control. Each of the 69 samples that were tested was placed into distilled water. After a few minutes, a Hemastix was inserted into the water. On achieving a positive result, Na-EDTA was added to eliminate any positive ions from known potential contaminants. Williamson then retested the samples using new Hemastix strips. As the second test involved a dilution of the sample, a positive result was always weaker on the Hemastix scale than with the first Fig. 72. By contrast with the cartoon in Fig. 71, Manqindi Dyantyi’s oral account and practical demonstration suggest that the production and consumption of rock art, at least in a southern African context, was a far more embodied and immediate process than some theoretical archaeological approaches would have us believe. test. A negative test, reflected as a zero score on the Hemastix strips, was interpreted as the absence of haemoglobin/myoglobin. A positive result, usually about half the score of the first test, was interpreted as indicating the presence of haemoglobin/myoglobin (Appendix 3). tive and two were negative. Both Eldritch Images that were sampled turned out to be negative. A single feline image was tested and returned a negative result while a single rain-animal also tested negative. The Hemastix testing, then, suggests that SDFs and LH-SDFs regular- The results of the testing revealed an intriguing pattern in that certain images consistently tested negative for haemoglobin/myoglobin while others frequently tested positive. The image types that were sampled and tested included SDFs/LH-SDFs, non-SDF anthropomorphic figures, fine-line eland, Type 2 eland, thin red lines and Eldritch Images. Eight SDFs/LH-SDFs were sampled and five tested positive for the presence of haemoglobin/myoglobin, while three, notably all at Storm Shelter, returned a negative result. Three non-SDF anthropomorphic images were sampled and they all tested negative. Six fine-line eland were sampled and of these four tested positive, one was negative and one was not tested. Of the twenty-one Type 2 images (all eland) that were sampled, fifteen were positive, five returned a negative result and one was not tested. Four sections of thin red line were sampled—one was not tested, one was posi- ly show positive results for the presence of haemoglobin/myoglobin while the other non-SDF anthropomorphic images are more frequently negative. Both fine-line eland and Type 2 eland showed regular positive results. The Eldritch Images as did the thin red lines tended to be negative while the small sample size for rain-animals and feline images makes the result negligible until more of these image types are tested. The preliminary tests for haemoglobin/myoglobin thus suggest that certain of the elements that are identified as belonging to the pattern of SDF/LH-SDF sites typically incorporate blood as a binding agent while other images do not. Although more work is needed, these preliminary results suggest that elements that I identified earlier are not the only aspects of the pattern but that the constituent ingredients of the paint of those elements are also patterned. It is the repeated presence of the identified elements at SDF/LH-SDF sites as well as the recurring 161 Fig. 73. The leg of a Type 2 Image before the removal of a pigment sample for testing of the presence of haemoglobin/myoglobin. The area immediately under the step, where clear micro-flaking is present, is where the pigment was sampled. use of blood as a binder for some but not other ele- man in San art (Dowson 1994, 1995, 1998, 2000). While ments that need to be explained. Explaining such pat- these more recent efforts have considered idiosyncratic terns in rock art is, however, not an straightforward task. imagery in relation to other images within a particular site, they have seldom considered the interplay between Understanding Patterns in San rock art eccentric images and more conventionalized motifs There has been a growing interest over the last decade within a site. Nevertheless, in recent years some efforts or so in issues of idiosyncrasy, individuality and agen- have been made to understand the form and composicy in rock art research. Early studies of individuality tion of rock art panels in their entirety. These efforts tended to concentrate on selected images from diverse are largely influenced by art historical perspectives. sites and they emphasized how those images were idiosyncratic developments or plays on more widely depicted themes. An apparently unique painting of two crabs, for example, has been argued to represent an idiosyncratic portrayal of the underwater metaphor associated with altered states of consciousness, that is more commonly symbolized by paintings of fish in the south-eastern mountains (Dowson 1988, Lewis-Williams et al. 1986). More recent studies of individuality have stressed the rise of the powerful individual sha162 One of the most significant art historical efforts is that of Anne Solomon, who is, as I mentioned in Chapter Four, critical of Lewis-Williams’s work. Her critical position stemmed, initially, from her work on gender in rock art (e.g. Solomon 1992, 1994, for a critique of Solomon’s gender work see Stevenson 1995) but in recent years, she has shifted to a focus on myth and its relationship to rock art imagery. From this perspective, she has questioned some of central tenets of the sha- Fig. 74. The same image as in Fig. 73 but after the sample has been removed. manistic approach. As we saw in Chapter Four, for example, she has resurrected the debate concerning the distinction between images of spirits-of-the-dead and images of shamans. Recently, the differences between Solomon and Lewis-Williams have been published in a debate spanning several issues of the South African Archaeological Bulletin (Lewis-Williams 1998, 1999, Solo- Pierce-Morris semiotics in Believing and Seeing. Although, as Solomon points out, semiotics has always had a close relationship with structuralism, Lewis-Williams’s use of the Pierce-Morris material only formed a small part of that work and, as I have pointed out in Chapter Two, the symbolic interactionism of Victor Turner was far more influential. After this initial effort, semiotics played a negligible role in Lewis-Williams’s work and neither he nor Vinnicombe have ever undertaken an explicit structuralist analysis of San rock art, such as that made famous by André Leroi-Gourhan (1968, 1982) for the Upper Palaeolithic images of western Europe. In reply- mon 1997, 1999, 2000). While this published debate shows that there are fundamental theoretical differences between them, to a younger generation of scholars, Solomon’s move toward placing myth at the centre of her investigation places her thinking much closer to that of Lewis-Williams than was previously the case. ing to her critique, Lewis-Williams has criticised Solomon’s earlier gender work, in turn, as being structuralist Nevertheless, Solomon is critical of both Lewis-Wil- because it was based on constructions of lists of biliams and Vinnicombe. The principal accusation she nary oppositions. In response, Solomon points out that: makes is that their theoretical orientation is structuralThis minimal definition inevitably oversimpliist (Solomon 1997: 10). Solomon bases her claim that fies matters. Whilst Lévi-Strauss did maintain Lewis-Williams is a structuralist because of his use of 163 Fig. 75. Row of 21 finger-painted dots from RARI-RSA-STK1. The dots become progressively more human to the left. Transformations of dots such as these are probably the result of synesthetic experiences, where two senses, in this case touch and vision, are confused. Solid=Red; Clear=White. Scale in centimetres. that humans characteristically order and classify experience using binary oppositions, this does not mean that every time one encounters contrasting dualistic terms that one is dealing with a structuralist analysis! The founding metaphor of structuralism is rather the depth: surface dualism, where depth refers to ‘mental structures’ or a ‘cognitive template’ and ‘surface’ to cultural ‘expressions’ as the realisation of those ‘deep structures’ (Solomon 1999: 52). for example) are related to each other. It is, of course, one of the central tenets of Lévi-Straussian structuralism that societies can be described by producing lists of oppositions in the manner that Solomon does for the San and that these oppositions can then be arranged to show how they are structurally related to each other. Indeed, early anthropological work on gender followed a similar approach to that of Solomon in setting up lists of dualisms and describing societies in this manner. This kind of early gender work has been criticized—in many cases by gender theorists themselves—as being overly influenced by structuralist thinking (e.g. Ortner 1974). While Solomon too is self-critical of her early work on gender, her more recent material shows even stronger strains of structuralist thinking and, in light of her accusations against Vinnicombe and Lewis-Williams, this newer work needs to be carefully considered. While Solomon correctly points out that the use of dualistic terms does not necessarily incriminate one as a structuralist, her earlier gender work did more than just casually mention dualisms. She describes San cosmology in lists of binary oppositions and discusses how certain oppositions are related to one another. Importantly, in structuralist thought, the dualistic relationship between A and B, day and night for example, is not as important as the observation that A and B (day and night) are related to one another as C and D (man and woman, As I have mentioned, Solomon’s recent efforts have fo164 cused on San myth and its relationship to rock art. It is ironic that she should concentrate on myth because it was in analyses of myth that the structuralism of LéviStrauss was most successful (Leach 1970). Indeed, even today, long after the shortcomings of structuralism have been exposed and widely accepted, anthropologists continue to use structuralist analysis for myths. Lewis-Williams (1996, 1997), for example, has undertaken analyses of San myths that begin from a structuralist perspective but try to go beyond and so counter the limitations of the approach. Importantly, he has never applied these structuralist approaches to the rock art itself, using the method only for analysis of myth. Solomon, however, shows remarkable structuralist tendencies in trying to link myth to image. In attempting to make this link, Solomon begins from a position where she maintains that Lewis-Williams’ has overemphasized the role of ritual in the art and downplayed the role of myth. Although this is not strictly true, the relationship of myth to the art is something that is imperfectly understood. Janette Deacon, too, has considered this affiliation and she argues that art, ritual and narrative draw on “the same reservoir of experience” (a useful phenomenological approach). For Solomon, however, this is not enough. She claims that “…there may be structural similarities which go beyond this” (Solomon 2000: 280). She then attempts to show similarities between elements of the formal organisation of San narratives and the composition of rock art panels. She considers some San rock painting images in the south western Cape that have been described as ‘circular’, ‘spiralling’ and ‘centrifugal’ by Pippa Skotnes (1994). Solomon points out that the circularity and non-linear form of these images is paralleled in the mythology. Rather vaguely, she suggests that such “non-linear compositions were matched to their subject-matter” (ibid: 280). Concerning other images—the so-called mythic woman forms—she states that: “Although these figures can by no means be said to illustrate the myths, there are parallels in their construction” (ibid: 281). Although she is not explicit, the implication is that since myth and art share “structural similarities” (her phrase), they derive from a deeper, cognitive template. By her own definition, given earlier, Solomon is thus a structuralist. One of the main criticisms of structuralism, of course, is that it is synchronic and therefore ahistorical (see, for example, Clarke 1981). Most structuralist analyses do not consider how the sets of binary oppositions that they construct are historically created and negated. Nor, in their analyses of myth, do structuralists normally consider how myths are historically situated narratives; instead, they treat them as cognitive phenomena existing out of time. These criticisms apply equally to the work of Solomon. Although she accuses Lewis-Williams of ignoring or downplaying change in the art, she never attempts to demonstrate change in the art herself. Her earlier work on gender did not show how sets of oppositions are temporally situated and she did not try to show how gender relations amongst the San changed over time. Her more recent work on myth is also ahistorical and she does not attempt to show how San myths are historically constructed narratives. Of course, these are very difficult tasks and given the lack of dates for San rock art and a paucity of material in which to situate myths historically, perhaps impossible. One might be tempted to say that these criticisms are therefore unduly harsh were it not for the fact that Solomon levels these very same accusations at LewisWilliams. It is a remarkable quality of Solomon’s writing that she is often guilty of the very theoretical misdemeanours that she accuses Lewis-Williams of doing. The structuralist pitfalls into which Solomon stumbles are to be avoided in any effort to understand the pattern of the SDF/LH-SDF sites at Nomansland. On the one hand, we should thus not expect to find a rigid pattern at every site; indeed, such an expectation would presuppose a cognitive template to which San painters were enslaved. Nor should we see the pattern as something out of time. On the other hand, traditional archaeological approaches, which emphasize time, are also not entirely useful ways of understanding the pattern at Nomansland sites. Traditional archaeological approaches to rock art tend to treat the images as analo165 gous to the stratigraphic layers of deposit; one layer of motifs successively replaces another as one stratum is laid over another in time. It is this kind of analogy that makes the application of Harris Matrices so attractive to rock art researchers. David Pearce (2001), however, has criticized the application of Harris Matrices in rock art studies by applying the technique to four SDF/LHSDF shelters in Nomansland. The technique, he argues, cannot be used to show a simple progression of one layer of images over another. Instead, in Nomansland, he was able to show that there is no clear-cut overlay of the various elements of the pattern. Eldritch Images, for example, are at times over other paintings and are at other times under other motifs. While far more sites need to be studied this way, Pearce’s work suggests that we cannot just understand the pattern at Nomansland rock art sites simply as a function of time from a rather naïve archaeological viewpoint. We are thus faced with a dilemma: we cannot consider the pattern as a planned construction as this runs the risk of being structuralist and we cannot simply treat the elements of the pattern as motifs that spread through the southeastern mountains over time. A different way of explaining this pattern is needed and it is to embodiment that, I argue, we need to turn, to explain this pattern. Rethinking the Consumption of San Rock Art in a squatting position; it is therefore unlikely that they would have been uncomfortable as was Skotnes, when viewing the images that she describes. Nevertheless, if we can avoid falling into such traps, Skotnes’ suggestion that more attention needs to be paid to considering how the space of viewing structured the viewers’ experience of the images is important. I now consider the space of viewing at Nomansland rock art sites. The SDF-LH-SDF sites are typically positioned in rock shelters on steep slopes. The distance between the painted surface and the edge of the shelter, beyond which the floor falls steeply away is usually no more than one or two metres. It would be physically impossible to hold a Great Dance in a circular formation in these shelters. Such activities could possibly have been structured as linear events and the length of the shelter floor used but it appears that it would have been very difficult in some sites and almost impossible in others to hold a dance in front of the images. At Storm Shelter, for example, the large boulder before the images prevents all but the most restricted physical movement and only allows a few people to view the images concurrently (Fig. 16). The limited space of the SDF/LH-SDF shelters is important because Manqindi Dyantyi told researchers that the San of the area used to dance in Ncengane Shelter before the images. Ncengane Shelter is different to other SDF/LH-SDF sites in Nomansland in that it has a larger floor area and dances would have been possible. When Dyantyi took archaeologists to Ncengane Shelter, she explained that the San used to perform dances in the shelter and that the participants would turn with open hands towards the paintings to harness the supernatural potency that resided in the images. Moreover, as I noted earlier, Mapote told Marion How that paintings were made from blood, preferably eland blood, which means that the images were made from the very substance most redolent with supernatural potency. It is thus not surprising that the Great Dance took place in painted rock shelters and that participants turned towards the images to harness the potency. If we are to circumvent this dilemma and if we are to avoid treating the consumption of the art as in the cartoon, then we need to consider rock art as an embodied process. By this, I mean that more consideration should be given to how the properties of images and the sites in which they occur structure the corporeal viewing or experience of the images. Pippa Skotnes (1994) has done this for a single site in the Western Cape, where she observes that the images are on the ceiling and that to see them one has to squat uncomfortably and look upwards, thereby placing stress on the neck. The danger here is that we fall into the empathetic trap that we saw, in Chapter Three, is typical of some phenomenological approaches to landscape archaeology. San are of course well-known for being able to sit very comfortably for hours on end Another important property of Nomansland rock art 166 is the size of the images themselves. San rock art in the south-eastern mountains, generally, is noted for its small, highly detailed images. The images, even when they would have been newly painted, cannot be seen from very far and viewers of the art need to get their eyes as close as possible to the art to see the pin-head detail of the images. This would have been true for the San who looked at the art as well. The viewing distance could not have been more than two metres away at the most for the majority of SDF/LH-SDF sites. Almost certainly, this distance would have been smaller because of the small size of the images. At Storm Shelter, for example, the viewing distance is restricted to approximately one metre by the fallen boulder. This short viewing distance is noteworthy because, as one gets closer to the images, a person’s range of view becomes restricted so that eventually the images take up all of one’s vision. Indeed, one becomes immersed in the images very quickly simply by viewing them. It is the limited space of viewing and the spatial constriction of vision that is important in understanding how the art was consumed at Nomansland SDF/LH-SDF sites and points to a difference in the way a Western viewer, such as Skotnes, and a San person would have looked at and thought about the images. the term is difficult to explain, it is best understood as a way in which Native Americans “look at reality” so as to make “it possible for them to know something by temporarily turning into it.”(ibid.: 61). Transformation, then, may be understood as knowing by becoming. Importantly, central to this concept of transference are various states of consciousness. According to Highwater, people ...like American Indians, who do not normally make a distinction between dreaming and waking, are capable of a type or projection or transference which they experience as “transformation” (ibid.: 61). Clearly then, transformation is to be understood as something that involves moving between various states of consciousness without necessarily making a distinction as to which state is more real than the other. It is also clear from this statement that Highwater does not see transformation as something exclusive to Native Americans but something that is found amongst other small scale shamanistic communities—“people like American Indians”. In so far as it entails making little or no distinction between the truth value of normal and altered states of consciousness such as dreams, transformation would appear to be widespread. Certainly amongst the San, ideas of shamanic transformaJamake Highwater, a Native American descendant, has tion into various animals are widespread (Lewis-Wilwritten extensively about the differences between Western liams 1985, 1992, 1996, 1997) and little distinction is and, what he calls, primal ways of thinking. In particular, made between various states of consciousness. Highhe draws a distinction between art in the Western world water goes further than this and argues that transforand art amongst Native Americans. Drawing explicitly mation is not merely specific to Native Americans and on the phenomenological ideas of Merleau-Ponty, High- people “like them” but can also be found amongst cerwater (1982: 55) observes that in the Western world, the: tain sections of Western society, most notably amongst certain post-Renaissance Western artists (ibid.: 62-63). “conceptualizing” of art into something special called “Art” produced a wide separation between commonplace experience and specialized forms of expression” whereas for “primal peoples, on the other hand, the relationship between experience and expression has remained so direct and spontaneous that they usually do not possess a word for art” (original emphasis). One such artist identified by Highwater is the Russian painter, Wassily Kandinsky (1866-1944). Kandinsky is often called the father of ‘spiritual’ art because of his short but seminal essay entitled Concerning the Spiritual in Art and Painting in Particular (Kandinsky 1977). In December 1911, at an artist’s conference in St. Petersburg, Kandinsky read this essay to the audience. In 1912 the Highwater uses the term ‘transformation’ to describe the work was published in German and it was reprinted twice ‘primal’ approach to art. Although he acknowledges that in the same year. Two years later the work was translat167 ed into English (Lipsey 1997: see Endnote 1 for the full publication history). In this now famous and influential essay, Kandinsky became one of the first artists to chart a course away from the Renaissance tradition of objective representation. In his works after 1911, he became increasingly interested in expressing the essence of objects rather than their surface reality. His work and ideas influenced some of the most famous twentieth century Western artists, including Paul Cézanne, Piet Mondrian and members of the Cubist movement (Lipsey 1997). It is well-known that some of these artists, in their search for the spiritual, turned to the arts of indigenous peoples around the world. Indeed, Kandinsky himself was influenced by Siberian shamanic peoples in his quest for the spiritual (Weiss 1995). It is thus not surprising that Western artists such as Kandinsky and others should grasp Highwater’s concept of transformation. axiom that creativity is an experience rather than an abstract idea…(Cytowic 2000: 56). What makes Kandinsky’s synesthesia so important in this context is that—as we have seen in Chapter Two— people in altered states sometimes have synesthetic episodes that are normally so rare amongst people in normal states of consciousness (Cytowic 2000: 127ff.). The very shamans that Kandinsky was interested in Si- beria and, indeed, elsewhere in the world, were at times able to have the experiences of confused senses that he endured as part of his daily life. Kandinsky’s appreciation of the concept of transformation, then, stems not only from his interest in shamanism but also from his natural empathy for altered states. It is from this perspective of the neurological experiences of synesthesia and altered states that we can best understand Highwater’s concept of transformation and, I argue, it There is, however, a further reason that would have is in the altered states experienced by San potency-ownled Kandinsky to an appreciation of transformation. ers that we need to seek a new understanding of the Kandinsky had the rare pathological condition of syn- social consumption of southern African San rock art. esthesia, that affects about 10 people in a million naturally (Cytowic 2000: 10). As we saw in Chapter Two, The complexity of synesthetic experiences is somesynesthesia is a complex neural process that appears to times expressed by people who have consumed LSD. affect the Limbic System of the brain, and it is under- A male in his mid-twenties, for example, ingested LSD. stood as the confusion of senses. Synesthetes typically While under the influence of the drug, he looked at “hear colour”, “feel sounds” and “taste shapes”. Kan- a painting, similar to Salvador Dali’s The Persistence of dinsky’s synesthesia can be seen in some of his musi- Time and he saw a projection of himself moving cal efforts. In addition to being a painter, Kandinsky across the painting. In another case, a physician who was a pianist and in the same year that Concerning the had ingested the powerful hallucinogenic drug, mesSpiritual in Art and Painting in Particular was published, cal, experienced a wide range of complex and contogether with his friend from the famous Blue Rider fused sensations. His experiences were originally degroup, Thomas De Hartmann, he produced a one-act scribed in German by K. Beringer (1923, 1927) and opera entitled The Yellow Sound. It is not only in his mu- then translated by Heinrich Klüver (1966: 71-72) sic that he drew on his synesthetic experiences but also in his painting. For example, he used musical terms to describe his images and he called them “compositions” and “improvisations” (ibid.: 55). In using his synesthetic experience of confusing sight with sound, Kandinsky wished to push aside analytic explanations and move himself and his audience closer to the quality of direct experience that synesthesia represented. He grasped the 168 The subject stated that he saw fretwork before his eyes, that his arms, hands and fingers turned into fretwork. There was no difference between the fretwork and himself, between inside and outside. All objects in the room and the walls changed into fretwork and thus became identical with him. While writing, the words turned into fretwork and there was, therefore, an identity of fretwork and handwriting. ‘The fretwork is I.’ The subject’s identification with the fretwork image eventually extended to engulf all aspects of his experience in the altered state: I am fretwork; I hear what I am seeing; I think what I am smelling; everything is fretwork . . . I am music, I am climbing in music; I am a touching fretwork; everything is the same (Klüver 1966: 22). Certain rock art images in the south-eastern mountains appear to be related to synesthetic experiences, similar to these and those experienced by Kandinsky. In particular, certain images of dots appear to be related to such states (Dowson 1989). At one site, for example, RARI-RSA-STK1, there exists a row of 21 painted dots (Fig. 75). These dots, painted in red were made by the finger-painting technique. Some of the dots are connected by a thin red line. The dots on the right side of the painting have few features but as one moves left the dots begin to show human-like limbs extending from the abdominal finger dots until they are clearly depictions of human beings. Lewis-Williams and Blundell (1997) have argued that this row of finger painted dots represents a synesthetic experience in which vision and touch have been combined. Such combinations of visionary and tactile sense are more widespread in San rock art than was previously thought (Ouzman 2001). It is clear from Manqindi Dyantyi’s statements and actions that, at Ncengane Shelter, tactile and visual aspects of San rock art combined. When she described how the San had turned to the images to harness their supernatural potency, she herself turned towards the images and touched them (Fig.72). led Lewis-Williams and Dowson to suggest that the rock surface was like a veil between the real and spirit worlds. Since their original observations, it has become clear that the rock surface is an integral part of rock art in North America, the Upper Palaeolithic caves of Europe and Scandinavian hunter-gatherer rock art. Together with Dyantyi’s testimony and Mapote’s evidence, the use of the rock surface in the creation of imagery has led to the suggestion that San rock paintings were more than just images—they were things in their own right and, probably, some of the images not only represented them but were the manifestations of the animals, creatures and beings that inhabited the spirit world behind the rock surface (Lewis-Williams 1995). By touching the images, then, San were coming into physical contact with the spirit world and its inhabitants. The topographical properties of the sites themselves, where there is usually only a small distance in which to view the images, as well as the small and detailed imagery, the neurological evidence of synesthetic states, together with Dyantyi’s evidence all suggest that under certain circumstances, the consumption of the rock art images at SDF/LH-SDF sites in Nomansland involved altered states of consciousness in which there was a blurring of image, body and identity. In some cases, the San viewers of the art might have seen a projection of themselves moving across the painted panel while in others they may have felt themselves merge with the images. The distinction between image and viewer would have been eroded and, in some cases, transference would have taken place and the viewers of the images would have become the images themselves. Owners-of-potency would The touching of rock art images is important because, thus have become their SDF/LH-SDF portraits and, as it is increasingly becoming apparent, both in South with the physician and the fretwork, ‘The image is I’ Africa and in other parts of the world, that the rock would have been an appropriate phrase. It is from this surface is an integral part of the composition of some perspective of the embodied process of viewing the imrock art images. This suggestion was first put forward ages that any effort to consider the social nature of the by Lewis-Williams and Dowson (1990) with regard to art must begin. In the next section, I consider the social the San rock paintings of the south-eastern mountains, implications of the embodied consumption of rock art. where there are numerous paintings that show images either emerging from or entering into natural features, such as cracks, in the rock surface. This observation 169 The Social Implications of Patterned Sites One of the reasons usually advanced to support the idea that the San are egalitarian is their lack of ownership of material items. Gift exchange systems are then held up as examples of the lack of ownership of material goods amongst San people. Many San groups in the Kalahari Desert practice a form of gift exchange, known amongst the !Kung as hxaro. Typically, hxaro involves the establishment of long-term gift exchange partnerships between individuals in widely scattered groups. Items of material culture are then exchanged at regular intervals. Individuals may have many giftexchange partners during their lifetime and a gift that they receive from someone, in turn, may be given to someone else (Cashdan 1985, Marshall 1961). In this way, material items have a wide circulation within the Kalahari Desert. The gift exchanges function, in part, as a mechanism of survival. In times of drought, exchange partners can call on one another for food and water. The exchange of material items thus lends itself to an economic system that does not place much emphasis on the accrual of worldly possessions. The relatively mobile life of San hunter-gatherers in former times also contributed to the low number of owned possessions. Of course, today, things have changed in the Kalahari Desert and many San are not as mobile as previously and they have begun to collect material goods. While hxaro makes it seem as if the San, traditionally, had no concept of ownership, there are certain aspects that show that they did have a concept of ownership of certain things under certain circumstances. 1986) and the !Kung often have rights in many overlapping n!oresi. Indeed, to use the resources of a particular n!ore one simply joins a camp or one asks permission (Lee 1979: 336). While !Kung n!oresi are not clearly defined, some other San groups, the G/wi for example, consider the boundaries more important. Natural landmarks mark these boundaries. Yet, as George Silberbauer (1981: 193) states, ‘The boundaries of a territory are roughly defined by landmarks or, more correctly, in terms of areas surrounding these landmarks.’ For the G/wi, it is the areas of natural resources between the major landmarks that are more important. The ownership of waterholes is important because such places are intimately tied to San concepts of rainmaking and rain-control. The /Xam San, as we have seen in Chapter Four, have a very elaborate system of beliefs concerning rain and its association with waterholes. They conceptualise rain as creature that is clearly a conflation of real and non-real elements. Called !khwa: -xa: xoro in the /Xam language, the name translates as ‘water ox’. Rain animals are sexed with males being ferocious, aggressive and associated with violent thunderstorms and lightning, so common in Nomansland, while female rain-animals are calmer and are associated with soft, gentle, soaking rain. In order to make it rain, an owner-of-potency must ‘enter’ a waterhole during altered states of consciousness in order to locate and entice rain-animals from the spirit world into this one. This is a dangerous task particularly if it is a rain bull and those who enter the waterholes sometimes take with them the aromatic herb, buchu, in order to One area where such a concept of ownership can be hold it under the nose of the rain-animal so that the seen is in San ideas of territory. For the !Kung, a ter- scent will have a soothing effect. If the rain-animal is ritory is defined in terms of the natural resources that calm then it is easier to capture by slipping a noose over occur there. Collections of natural resources are called its neck or tying a thong to it is nose. Once captured, n!oresi (Marshall and Ritchie 1984: 82). Each territory the creature must be led into this world where it is maor n!ore has an owner, known as kxai k’xausi (Marshall nipulated across the landscape and either killed or cut; 1976: 184) and is usually formed around a permanent the resultant blood and milk becomes precipitation. or semi-permanent waterhole. It is only the waterhole that is really ‘owned’ in any sense of the word (Smith Belief in the capturing and controlling of rain animals 1994: 373). Rights to use the n!ore are inherited through appears to have been limited to the southern San as kinship or established through residence (Barnard the Kalahari groups do not appear to have such de170 veloped concepts of rain-animals, although they make rain in other ways and sometimes speak metaphorically of rain as if it were a creature (Lewis-Williams 1981: 104, Marshall 1999: 165-167). From the substantial ethnographic material on rain-making and rain-control, it is clear that the San believe rain-animals to be large quadruped creatures that are ferocious (especially the male). Rain-animals can have large teeth and hooves with which they bite or trample on those who would capture them. While there is some evidence to suggest that rain–animals are partly modelled on real beasts such as hippopotamus and rhinoceros (Ouzman 1996), it is clear that the creatures derive many characteristics from the hallucinatory experiences of San potencyowners. They are thus a conflation of real and imaginary imagery. The detailed ethnographic description of rain-making and rain-control has allowed researchers working within the hermeneutic approach to identify images in the rock art that can be argued persuasively to represent rain-animals. In the south-eastern mountains, there are rock art images of quadruped creatures that are clearly conflations of real and imaginary attributes. As we saw in Chapter Four, they are often painted with large teeth and sometimes they are depicted as biting or stomping on surrounding human figures. In some cases, human figures stand before the creatures holding up sticks with attachments that recall the use of buchu to soothe the violent tempers of the beasts. By using the attributes listed in the ethnography then it has become relatively easy to identify rain-animals at many sites throughout the south-eastern mountains (see Woodhouse 1992 for an overview of rain-animals). are painted in a white pigment and on one of them are painted small red dots in the abdominal area. The characteristics of these images suggest that they depict rain-animals and no other images at other Nomansland sites look identical to these. The unique quality of the rain-animal images at Storm Shelter, elsewhere in Nomansland and the in the south-eastern mountains as a whole, has significant implications for understanding San concepts of ownership of the rock art sites. Although rain-animals are readily identifiable as a class of images in San rock paintings, they are remarkable because no two are exactly alike. In contrast to other images such as antelope or thin red lines, each rain-animal image is unique. Indeed, there are seldom more than one or two rain-animals at any single rock art site in the south-eastern mountains. Rock art sites in Nomansland follow this trend as well. At Storm Shelter, there are two quadruped images in the centre of the main panel that are superimposed by other images. They script) have taken up Dowson’s earlier suggestions and they have considered notions of ownership of rain-animals in the ethnography and the art. Following on from David Whitley’s (1998) important work in the far west of North America, they consider the importance of place in rain-making and rain-control. They argue that the Bleek and Lloyd ethnographic material suggests that rain was made in specific locations. //Kabbo, for example, spoke of riding “the rain up the mountain on top of which I always cut the rain” (Bleek Thomas Dowson (1998) has suggested that the idiosyncratic nature of rain-animals in the south-eastern mountains may be understood partly as the result of their imaginary nature and partly as a result of San concepts of ownership of rain. He notes that amongst the /Xam San, wind and rain were associated with individual people and that when talking about weather, the San often use possessives such as: “You know that when father used to shoot game, his wind blew like that” (Bleek 1932: 329). The pervasive notions of ownership of elements of weather in the /Xam ethnography leads Dowson to suggest that paintings of rain-animals were ‘owned’ by the San who painted them. He states (1998: 82): The important point for the role of rock-art in historical processes is that it seems highly probable that a shaman-artist could point to a specific depiction and say, ‘That is my !khwaka xoro (rain-creature)’.Other people could look at an imposing potency-filled painting of a rain-creature and say, perhaps with awe and respect, ‘That is so-and-so’s rain-creature’. More recently, Lewis-Williams and Pearce (Manu- 171 1935: 309, 310). Lewis-Williams and Pearce point out that the word ‘always’ suggests a specific mountain on which //Kabbo made rain. Another of the San who offered information to Bleek and Lloyd, Diä!kwain, described how rainmakers or rain-controllers killed and cut the rain-animal away from their camp, apparently out of sight of the rest of their community (Bleek 1935: 377). The comments by //Kabbo and Diä!kwain suggest that rain-making was done in a special place to which the rainmakers returned time and again. they occupy in the cosmological beliefs suggests that San concepts of ownership extended, to some degree at least, to rock art sites. This suggestion finds support from the imagery at Nomansland rock art sites. I have already argued in Chapter Six that the SDFs/LHSDFs are a form of portraiture and that they represent individual San owners-of-potency. At sites that have a number of these images, one always stands out as the primary SDF/LH-SDF, although there may nevertheless be multiple differentiated figures at the site. Moreover, the SDF/LH-SDF images are site specific and they do not appear to occur outside of a single rock shelter. The limitation of these portraiture-like images to a specific shelter suggests that the individual potency-owners that they represent are people who were associated with that particular rock art site. The individual qualities of rain-animals at these sites also support a link between individuals and painted sites. As we have seen, rainmakers returned to the same place to make rain and they ‘owned’ rain-animals and by extension, they owned paintings of rain-animals, which are the actual inhabitants of the spirit world, not merely representations thereof. The idiosyncratic nature of painted rain-animals suggests that they were associated with the work of a single shaman-artist. The close association of rock shelters and images with individuals suggests that there was, to some degree, a concept of ownership of rock art sites amongst the San of Nomansland and possibly throughout much of the south-eastern mountains. The SDF/LH-SDF images would, in part, be a way of indicating such ownership. Their suggestions find support in the testimony of Manqindi Dyantyi. As we saw in Chapter One, her father, Lindiso and her uncle, Masela, were painters and rainmakers. They used to visit Ncengcane Cave in order to make rain and to paint. It is important that Dyantyi, in a number of interviews with archaeologists, made no mention of her father and uncle painting or making rain in another cave. Dyantyi’s elder sister, Chitiwe, was also a rainmaker but was not a painter. According to Manqindi, when Chitiwe was approached to make rain, she would leave home in the morning in secret and go to a rock shelter where she would stay for most of the day, returning only in the afternoon (Prins 1990: 113). It seems likely that this shelter was also Ncengcane Cave. The observations of San in the Nomansland area around the turn of the twentieth century that are mentioned in Chapter One, also point to the association of San rainmakers with specific rock shelters. It would thus appear that rain-making and rain-control took place at specific sites in Nomansland to which San ritual-specialists returned time and again to perform ceremonies associated with weather. If sites were owned in Nomansland and if the SDF/ LH-SDF images are a form of portraiture of the ownFrom the ethnographic evidence concerning n!oresi, ers, then we may model the social implications of the waterholes and rain-animals, it is clear that, at the very patterning of the sites. It is likely that access to proleast, San have a weak concept of ownership of place ducing and consuming the images was not entirely reand certain spirit-world beings. Importantly, in San cos- stricted amongst the San communities in Nomansland mology, both waterholes and rock art sites are portals during the nineteenth century, as access to most activito the supernatural world. As rainmakers ‘entered’ wa- ties amongst extant San groups appears to be open to terholes during altered states of consciousness to cap- all who have the desire to learn. Nevertheless, anyone ture rain-animals, so other potency-owners ‘entered’ wishing to paint at an owned site would, as with the case the spirit realm at the rock face. The similar place that of waterholes, have to seek permission to do so from 172 the owner. While this was probably merely a formality in the past, the situation changed when San groups in the area began to include Khoekhoen and Bantu-speaking members. As we saw in Chapter Five with the Type 2 images, when non-San people joined San communities, there was a spatial separation of painting; San painted in one side of a shelter, while non-San painted on the other side. By enforcing a spatial separation in shelters and by keeping the production of different types of image separate, the communities of Nomansland and its environs thus tried to maintain a distinct San identity. It is likely that the control of the space of production also extended to control over the production of the images and the subject matter of the images became less a matter of choice of individuals and, over time, more a matter of the will of the owners-of-potency. Moreover, this control probably also extended to the circumstances under which the images were viewed or socially consumed; anyone wishing to undergo transference—becoming an image—and thereby experiencing the spirit world would have probably done so under the supervision or guidance of the owner of the site. It is in the extension of control owners-of-potency over the space of rock art sites and the production and consumption of the images themselves that we may understand the pattern of SDF/LH-SDF sites in Nomansland. If the images are not only representations of the spirit world but the manifestations of that spirit world, then they are patterned because their existed amongst the San communities who made them consensus over what the spirit world should look like. This consensus can be explained in several ways. For one, the panels can be seen as constructed. In other contexts, notably at Çatalhöyük, the famous Neolithic settlement in Turkey, researchers have suggested that the design of houses and the choice of painted images on certain murals were a conscious effort to construct a cosmos; the space of living was thus also the space of the other world (Lewis-Williams 2004). The word ‘construction’, however, implies a plan of some sort from which people were working and while it is probably appropriate to the Neolithic architecture at Çatalhöyük, it is not appropriate for southern African San rock art where, as we have seen earlier in this chapter, the dangers of falling into a structuralist position are ever present. If the pattern at SDF/LH-SDF sites in Nomansland was a conscious construction then the repetition of that pattern in site after site would mean that the San who made the art were simply following a cognitive template. Instead, through a consideration of the historical material in this and previous chapters, which strongly suggests growing control over time of the production and consumption of the rock art by owners-of-potency, it appears that there was a growing restriction on the elements that formed the subject matter of that cosmos. At the outset of this chapter, we saw that traditional ways of understanding the consumption of rock art tended to treat the original viewers of the art as detached observers, who were somewhat analogous to the contemporary Western art viewers in the cartoon of Figure 71. Rather than take this approach, I have suggested that the consumption or viewing of the rock art should be considered from the perspective of embodiment. From this position, we can draw on the multiple ethnographic and historical strands of evidence for the Nomansland region to put forward a model of how the images at SDF/LH-SDF sites would have been consumed. At the heart of this model lies the notion of transference—a process by which viewers of the art became the very images that they observed. The evidence of Dyantyi and Mapote suggest that the images were the actual inhabitants of the spirit world and that, through transference, San observers were projected into that spirit world. This new way of understanding the viewing of the art as an embodied process, together with historical material that suggests increasing control over the spirit world by owners-of-potency, strongly suggests that the pattern at SDF/LH-SDF sites may be understood as an increasing restriction on the elements that constituted the spirit world. This limitation was itself a result of time, which allows us to escape the problems of other approaches to seeing patterns in San rock art that tend to treat panels from a structuralist perspective where the patterns are constructions that exist out of time. 173 It is from the perspective of the art as a space for the experience of transference that I suggest future efforts to understand the social issues surrounding the production and consumption of the images should begin. One of these issues that might be profitably considered concerns images that are absent. As we saw in the Introduction, postcolonial theorists consider what is absent as important as what is represented. An analysis of ‘missing’ images in rock art is not a straightforward task. By its selective nature, there are innumerable ‘absences’ in the art. At Storm Shelter, the SDF ‘type site’ for Nomansland, for example, only three of the many species of antelope are depicted. The presence of these three species can be explained by recourse to San ethnography (in fact the presence of only one, the eland, has been explained satisfactorily). If one were to ask the question differently and enquire why the other antelope species are absent, the answer would inevitably be trivial. Nevertheless, if one accepts the view that the art is ideological (and therefore masks social inequality) and that it helps to constitute social reality rather than merely reflect it, then what is absent is as important as what is present. Although theoretical expectations might not be met in the case of the antelope, the principle of absent images takes on greater significance when considered in conjunction with what is known about the meanings of the art. If the art is, as Lewis-Williams has demonstrated, principally concerned with the ritual of the Great Dance, the symbols of supernatural potency associated with the dance and the experiences of the dancers, both real and hallucinatory, then we are in a better position to discern what has been ‘lost’ in the transmutation from ritual and experience to representation of that ritual and experience. 174 CONCLUSION TOWARDS A SOMATIC PAST OF NOMANSLAND Archaeology, as a discipline, is often concerned with the study of origins. Some of the most celebrated issues in the subject, for example, concern the emergence of humans and lithic technology (e.g. Isaac 1989), the development of modern human behaviour (e.g. Mcbrearty and Brooks 2000), the origins of domestication and agriculture (e.g. Cowan, Watson, and Benco 1992, Harris 1996, Reed 1977, Rindos 1984, Thorpe 1996) and the beginnings of image-making (e.g. Lewis-Williams 2002). An obvious related interest concerns the transition between periods of significant difference, such as that from the Upper Palaeolithic to the Neolithic in Europe (e.g. Childe 1936, 1942). The emphasis on origins and transitions has meant that archaeology has, arguably, paid less attention to understanding how and why certain practices and ideas come to an end without any discernable transition. Those efforts that do consider the archaeology of termination tend to do so from the perspective of culture history, where ‘cultures’, such as the Maya or Anasazi for example, are seen as meeting an abrupt end, either through environmental catastrophe or through the agency of other cultures. Such explanations for the termination of cultures give rise to ideas of the kind that Stow popularised, where every last San perished dramatically in some forsaken spot of southern Africa. Stow’s comments, as we saw in the Introduction, helped produce perceptions—such as those held by the Chief Executive Officers of powerful drug companies—that the San are extinct. It is clear from the detailed documentary evidence from Nomansland, which I discussed in Chapter One, that such ideas concerning the termination of San rock painting as a practice are too simplistic. Indeed, a great deal of this work has been an effort to counteract these simplistic perceptions and to show how varied and complex the socio-historical processes were that led ultimately to the end of San rock painting in the south-eastern mountains. In order to understand these varied and complex processes, I argued, in Chapter Two, that southern African rock art research requires a sophisticated theoretical apparatus that overcomes some of the difficulties of previous approaches, such as marxist and structurationist theories. In Chapter Three, it was suggested that the theoretical concepts of ‘the body’ and embodiment offered such tools. The utility of these conceptual devices for extending the present hermeneutic model of southern African San rock art was then demonstrated in Chapter Four. Through a consideration of widespread San concepts of death, disease and disorder and how they are embodied, I argued that it is possible to identify the Eldritch Images, with their grotesque features and skeletal appearance, as representations of the spirits-of-the dead. Embodiment, however, is not simply a tool for extending and adding nuances to the present understanding of rock art; it also has powerful heuristic capabilities for the study of the changing social production and consumption of the images. In Chapter Five and Chapter Six, I considered how ‘the body’ and embodiment could be used as tools for an analysis of change in the rock art of Nomansland. First, in Chapter Five, I described images, that I labelled as Type 2, that are clearly not classic San fine-line paintings. These images, furthermore, do not conform to traditions of rock art that have been attributed to either Khoekhoen or Bantu-speakers. Indeed, the images, as do the peoples of Nomansland and its environs, need to be understood as the product of local historical processes of creolization. In Chapter Six, the argument was extended to a consideration of the global colonial forces affecting Nomansland. In particular, the LH-SDFs, it was argued, need to be seen as a product of global colonial forces. Finally, in Chapter Seven, I argued that 175 the production and consumption of rock art needed to be seen as an embodied process in itself, rather than as a series of detached social and political statements in which the body simply reflected social issues. as an undifferentiated part of the greater south-eastern mountains. At some point in the past, at a time still unknown, the San groups of Nomansland and its environs began to paint SDFs. While Dowson argues that they are the product of the recent colonial The principal aim of this work, as I stated at the out- time period, we have seen that natural inequalities in set, was to take the first steps toward writing a past of the Great Dance, could have led to the production of the San in Nomansland that considered the complexity such images at a much earlier time without necessarof their interaction over time. Rather than see them as ily requiring interaction. These SDFs were depictions isolated groups in the traditionalist sense or simply as of powerful individual potency-owners, who probeconomic dependents of other southern African peo- ably ‘owned’ the rock art sites and controlled access ples, as some revisionist positions advocate, we needed to the production and consumption of the images. to work towards writing a past that is more in line with postcolonial ideals. I pointed out four such postcolonial With the arrival of other communities in southern Afideals—a concentration on creolization and hybridiza- rica, beginning at about 2000 years ago, the San entered tion, a focus on the subaltern or colonized, an analysis into varied and complex relationships with both Bantuof the mechanisms that the colonized use for expres- speaking and Khoekhoen peoples. In the south-eastern sion (such as rock art) and a theoretical approach that mountains and particularly in Nomansland and its enviallowed, as far as possible, those mechanisms to ‘speak rons, this relationship was not one of immediate domifor themselves’ rather than force them into specific nation by the new groups, who were often more numerEurocentric theoretical frameworks. I have argued that ous, but was one in which shifting power relations often by drawing on the principles of the somatic past—an favoured the San. Moreover, the relationship between analysis of anthropomorphic bodies in rock art, the use different Bantu-speaking groups and the San groups in of embodiment and where this is not possible a con- Nomansland was variable, both in time and space and sideration of ‘the body’ as a tool for interpreting these we cannot speak simply of a broad San-Bantu-speaker anthropomorphic images as well as the process of their relationship. We saw in Chapter Five, for example, that social production and consumption—we have a strat- during the nineteenth century, the San of Nomansland egy by which to attain the four ideals of a postcolonial were closely allied to the Bhaca and the Mpondomise past. After using this strategy to discuss Eldritch Im- but were hostile toward the Thembu. With the arrival ages, Type 2 rock art, SDF and LH-SDF figures as well of explorers and the subsequent European settlement as the conditions surrounding their social production of South Africa, the social complexity of the Nomansand consumption in separate chapters, we are now, at land environment increased even further. In Chapter the end of this work, able to bring the various strands Six, we saw how survivors of shipwrecks (both Eurotogether and to put into place the rudimentary struc- pean and Asian), escaped slaves from the Cape Colony ture of the somatic past of Nqabayo’s Nomansland. and dislocated Khoekhoen, Bantu-speakers and San peoples all moved into the Nomansland region and Nomansland rock art shares many similarities with im- mixed with the San groups already there. In such a ages from other areas in southern Africa. There are, complex mix, Nomansland and its peoples can only be nevertheless, a number of distinctive images and the seen as creolized peoples. Importantly, the processes of most prominent of these are the SDFs and the LH- hybridization and creolization brought about by these SDFs. These images suggest that the historical trajec- new peoples arriving in Nomansland was not instantatory of Nomansland and its rock art was different from neous but, rather, was one in which identities were ofother areas; Nomansland cannot simply be treated ten contested. An important aspect of this contestation 176 was the rock art and the space in which it was produced. This is quite clear at SDF sites where Type 2 images are mostly peripheral to classic San fine-line paintings. From Mapote’s evidence, we interpret these images as paintings made by people living with San groups. The influx of all the different peoples into Nomansland brought about a shift in San perceptions of the body. Over time, there was a decline in the importance of the post-cranial body of the SDFs. This decline seems to begin at the extremities; feet and hands disappear first and then legs begin to diminish in significance. At the same time that the post-cranial body diminishes in importance, the significance of the cranial area grows. The size of the SDF heads increases and their faces become more detailed so that LH-SDFs emerge. Ultimately, this process ends in images that simply portray heads with detailed faces but no post-cranial abdomens. This course of the death of the post-cranial body, I argue, is partly the result of the changing ritual role of potencyowners from traditional healing to one of paid rainmaking. In this shift, the post-cranial body, as the locus of healing activity, would have become less important. Indeed, the painters of the images, who were the owners-of-potency themselves, were probably trying to encourage the acceptance of such a shift, which benefited them economically, to the rest of their communities. features that are specifically non-San was possibly an attempt by individual potency-owners to manipulate their personal identities to be more appropriate for the diverse and complex creolized milieu during the colonial period. Importantly, the identities created by the images and the social negotiation which they permitted, were put into operation through the embodied mechanisms of consuming the art that involved transference. Through synesthetic or other neurological processes of altered states of consciousness, the distinction between image and viewer became blurred. Potency-owners thus became one with their SDFs/LH-SDF representations in the spirit world; people, place and painting became one. Each image of a potency-owner was, moreover, also a marker of that person’s ownership of the site. Any image that was made by another person was only allowed through the control of the potency-owner. Consuming the images, as producing them, was thus almost certainly a controlled and ritualized process in which potency-owners manipulated the embodied experiences of viewing the images for their benefit. It is the increasing control of the production and consumption of the images that leads to standardization of subject matter at SDF/LH-SDF sites in Nomansland. Over time, especially with the increasing inclusion of non-San into the San groups of Nomansland, the potency-owners’ control of the space of painting became tighter. Nevertheless, the emphasis of certain body parts at the Ultimately, the control of sites and the production of expense of others in the SDF and LH-SDF images can- images became a secret as the San sought to mainnot be seen simply as a painted statement about an indi- tain their ritual control and to retain their San identity. vidual’s political status; they also need to be understood as an expression of shifting concepts of identity where This is clearly the case at Ncengane Shelter. Manqindi the head and facial regions were becoming important as Dyantyi described how her father, Lindiso, painted in markers of identity. The LH-SDFs are thus probably a secret at the site. Lindiso would take along her elder form of portraiture but, as the faces contain non-real sister, Chitiwe, to Ncengane Shelter but while she parand animal features, they are representations of power- ticipated in rain-making rituals, she did not ever observe ful owners-of-potency as they appear in the spirit world. her father paint. Dyantyi’s observations on the secrecy Moreover, the faces appear to include physiological fea- surrounding her father’s painting activities are suptures taken from European, Bantu-speaking and Asian ported by Lindiso’s apparent reluctance to declare his communities. At one level, this reflects the actual histor- artistic habits to Europeans. When Dr Anders spoke ical processes of creolization, which were both cultural to Lindiso and Poponi about the !Gã !ne language in and physical. At a deeper level, the inclusion of facial the 1930s in the Bushman Cuttings area, Lindiso did 177 not tell him that he was a painter. Perhaps, in the context of the conversation, this did not seem important. Nevertheless, when Lindiso was requested to ask his mother, Mamxabela, about the ingredients of the paint, he also failed to mention his status as a painter. This is clear from Apthorp’s letter (see Chapter One): the same shelters but on different surfaces, these limited co-operative ventures came to an end and the San began making art in secret. This was almost certainly to protect their status as ritual-specialists amongst the Mpondomise and other Bantu-speaking communities along the south-eastern seaboard and the income that came with that status. In such a scenario, San identity Her son, Lindiso, states that he was told by became very much tied to paintings and the rock shelhis mother during her lifetime that white, red ters in which they were painted. Indeed, when Lindiso and yellow clay and charcoal made from the was still a child, the San of Nomansland were still livwood of the Coral Tree were mixed with water and the fat of the bushbuck or other animal. ing in rock shelters even though they were very much incorporated into Mpondomise society and could easily Lindiso only relayed to Apthorp what Mamxabela told have settled in built structures. By continuing to live in him and he did not offer any additional information. He rock shelters, they retained the one thing that still linked could easily have told Apthorp further, helpful informa- them to their past and their identity. Even after Lindiso tion or even shown him how paint was produced but, in- married and settled in a homestead, he, together with his stead, he remained silent and offered nothing else. Lind- brother, continued to visit rock shelters to make images. iso, as we saw, in Chapter One was described as intelligent and he had survived the Kimberley siege during the An- The broad outline of the process that led to the end glo-Boer War. It is therefore unlikely that he misunder- of San rock painting in Nomansland that I have outstood the importance of Apthorp’s question. As it was a lined here requires further work in the future to flesh directive from Stanford, acting on behalf of Péringuey, out more of the detail. Specifically, better chronologito him, requesting information from Mamxabela, it is cal control will enable more detailed analysis of some equally unlikely that Apthorp neglected to mention any- of the shifts in and between the rock shelters of the thing as important as this in his report. It would seem, Core Study Area. Nevertheless, the writing of a past for therefore, that Lindiso purposively hid the fact that he Nomansland should not be constrained by the absence was a painter from Apthorp and the other Europeans he of sufficiently reliable and numerous direct dates. The encountered; the act of making images was his secret. writing of those pasts should be constructed in such a way as to circumvent the need for chronological specifThe secrecy surrounding Lindiso’s painting activities is ics or, alternatively, they should be set up as testable intriguing because it is clear that San rock painting in hypotheses for the future. In addition to tighter chronoNomansland and in the south-eastern mountains as a logical control, there is a need for work on establishing whole, was not always a secret activity. Qing, for exam- tighter links between excavated material and rock paintple, spoke openly to Orpen about the meanings of the ings in Nomansland. Excavations of the SDF/LH-SDF paintings. When the San working with Bleek and Lloyd shelters, for example, could possibly reveal whether or were shown copies of rock art images, they also spoke not the material culture is site specific in a manner remiabout them openly. Mamxabela spoke proudly of her niscent of the images themselves. Of course, both dathusband’s prowess as an artist to Stanford and pointed ing work and excavation are undertakings for the future out shelters in which he had painted. It thus appears and they will almost certainly add nuance and insight that at some time during Lindiso’s life, the practice of into the general understanding of the Nomansland rock painting in Nomansland became a clandestine past that I have described here. Nevertheless, I believe San activity. Whereas previously, the San had allowed that the broad structure of the progression that I have Bantu-speakers to make images alongside their own in outlined will hold and that it is in the move towards 178 secrecy that the seeds were sown for the end of San rock painting in Nomansland. The more controlled the production and consumption of images became, the fewer people had access to the skills and knowledge required to produce and consume them. This situation was possibly exacerbated by the gradual amalgamation of San with their Mpondomise neighbours, who were, traditionally, a much more patriarchal society. It is remarkable that Lindiso did not pass on to either of his daughters the knowledge required to continue the rock painting tradition. Perhaps, if he had a son, in his new patriarchal environment, he may have passed on the skills. Whatever the reasons may be, it is clear that San rock painting, as a long-lived and widespread practice in Nomansland came to an end with Lindiso and his generation. This is not the same as Stowein dramatics that would see the end of the San as a people, nor is it the case that the production of San identity ended with the demise of the painting tradition; the criteria that were used to constitute ‘San’ and ‘Sanness’ now simply shifted elsewhere, away from the rock art. Some decades after Lindiso stopped painting, in the first few decades of the twentieth century, his last daughter, Manqindi Dyantyi passed away. Dyantyi’s children appear to have all left the Tsolo area and after her death, her last daughter moved to Soweto (Prins 2004 Pers. Comm.), thereby finally breaking the link between people, places and pictures that makes Nomansland such an important place for the study of San history. Nowhere is that break more powerfully felt than at the rock art sites themselves. These sites are scattered over private farm land that is parcelled into clearly defined units, marked and divided by fences. In the more remote parts of the Nomansland Core Study Area, the farms are usually only for grazing and they are not occupied. Often, the silence at the sites seems incongruent with the richly detailed paintings that testify to the vibrant practices of image-making that once took place there. the town of Ugie, the names of the places mentioned by Silayi appear: Xuka Drift Siding, Gatberg and Prentjiesberg—places where Nqabayo and his people lived, fought or passed by on their raiding expeditions. Yet, if he were somehow brought to life, Nqabayo would hardly recognize these places where he once lived. Today, much of the natural grassland has been planted by forests, which will ultimately be turned into paper that, in turn, will be used to produce books and academic theses. While the forestry has brought economic and environmental problems to the area, they have also, ironically, led to better protection of the paintings. The single greatest threat facing the rock art of Nomansland is domestic stock, particularly cattle, sheep and goats that enter the rock shelters and rub against the images, thereby abrading the paintings. Moreover, they stir up the dust on the shelter floors and the dust settles on the images, bonding with them over time. Indeed, one of the reasons for Storm Shelter’s good preservation is the enormous boulder before the images that restricts access by animals to the paintings. By removing the domestic animals, then, the forestry companies have, unintentionally, ensured that the art of Nomansland will endure for longer than it would have. Passing Ugie, one approaches the most prominent town in the area, Maclear. It is from this town that one approaches Storm Shelter and other sites that are described in this work. The town itself is named after Sir Thomas Maclear who was the Royal Astronomer at Cape Town during the later part of the nineteenth century. As Martin Hall (1996) has pointed out, Maclear was also the neighbour of Wilhelm Bleek in Mowbray and he was displeased that the /Xam San prisoners, with whom Bleek was working, were allowed to live on the adjoining property. On 7 April 1875, Maclear (n.d.) encountered Diä!kwain and described the event in his diary: “Disgusting…in the Paddock…The person, a Bushman of Dr Bleeks. Puks pulled him off. I always dreaded the proximity of Dr Bleek’s Bushman”. To see these painted sites, one must drive through the In the light of such a statement, it is intriguing to wonKraai River Pass, past the road west to Dordrecht and der what Maclear would have thought had he known down the precipitous Barkly Pass. As one approaches that the town named in his honour was so close to the 179 heart of the Nqabayo’s country. Perhaps, the remoteness of the settlement and Nomansland from Cape Town would have mitigated his fears of San proximity. Ironies such as this litter the Nomansland landscape and, today, elements of the colonial period live, sometimes uncomfortably, side by side with those of modern South Africa. Indeed, it is a short distance from Maclear to the Transkei territory, where, formerly, there was an Apartheid Era homeland. Today, many people from the former Transkei have moved to the Maclear area in search of employment. What was formerly a town controlled by whites only has become a colourful multi-cultural community, more true to Nomansland’s hybridized and creolized past than the isolationist policies of Apartheid. The town is still remote from South Africa’s major centres, but, in spite of this, Maclear is a bustling urban settlement with several supply stores, butcheries, drinking places and even a hospital. More informal trading structures line the pavements on busy week mornings and one can purchase a range of plants taken from the countryside by traditional healers, whose knowledge about the medicinal properties almost certainly stems from the older, indigenous knowledge of the San. The inhabitants of the town represent the full gamut of South Africa’s diverse political spectrum and include conservative, liberal and radical views. At times these diverse views flare up in heated arguments while at other times people are more focused on the business of living. There are still many social, political and economic obstacles to overcome in the area. Issues of land restitution and the wider repairing of the damage done by Apartheid still have to be faced. Moreover, the effects of widespread environmental and economic destruction brought about by forestation in the area are still to be fully felt. Most concerning of all is the widespread devastation of HIV-AIDS and outside the small towns of the area, fresh graves grow at a disturbing rate. In such a scenario, the future of the rock art shelters where Nqabayo’s group and their descendants painted seems uncertain and the future of Nomansland as a whole is unclear; it remains, very much, No Man’s Land. 180 ENDNOTES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. This number is taken from the Working Group for Indigenous Minorities of Southern Africa (WIMSA) Uniform Resource Location (www.san.org.za). Although ‘Transkei’ was formerly the name of an Apartheid Era ‘Homeland’, which now no longer exists, the name—which means across the Kei River—is still used to refer to the general area. While the spelling is slightly different on this photograph, it is clear that the name refers to Mamxabela. Comma added. I have substituted Stanford’s spelling of "Ngqabayi" with the modern spelling "Nqabayo". I have also added a full stop after the end of the second sentence before Ndhlela. All else is as it appears in Stanford’s typed version of Mgudhlwa’s statement. I have corrected the spelling of Péringuey’s name in this letter. I have substituted the name that Apthorp used for this tree with the modern version, which is less pejorative. My italics. ≠Gao N!a in the orthography followed in this work. //gauwasi in the orthography followed in this work. All superscript notes here refer to Bleek and Lloyd’s original notes: 1. Whai-ttu means “Springbok Skin”. 2. That part of him (with) which he still thinks of us, is that with which he comes before us, at the time when the sorcerers are taking him away; that is the time when he acts in this manner. For, my mother and the others used to tell me, that (when we die) we do as the /nū people do; they change (?) themselves into a different thing. 3. My sister, /A-kkumm’s husband it was who told us, that he had perceived a child who was afraid of him. It wanted to run away. 4. At one time, when he looked at it, it was not like a person; for, it was different looking, a different thing. The other part of it resembled a person. I take this to mean sickness. See Marshall 1999: 29 for the !Kung equation of evil with sickness. Present efforts to re-locate this site have been unsuccessful. I have also viewed the vast majority of the several thousand slides in this collection, now in the possession of Lucas Smits in The Netherlands, and other than Thaba Moorosi, I could not identify any image that qualified as an SDF. Surveys by Ouzman and Blundell of the territory adjacent to the north bank of the Orange River in the early 1990s yielded no images that were convincing examples of SDFs. Extensive work by Jannie Loubser and colleagues throughout the Caledon River Valley on the western side of south-eastern mountains has also not produced any clear SDFs. 181 REFERENCES Alexander, J. 1984. “Early Frontiers in Southern Africa,” in Frontiers: Southern African Archaeology Today, 207 edition. Edited by M. Hall, G. Avery, D. M. Avery, M. L. Wilson, and A. J. B. Humphreys, pp. 12-23. Oxford: B.A.R: International Series. Anati, E. 1981. The Origins of Art. Museum 33:200210. Anders, H. 1934/5. A Note on the South Eastern Bushman Dialect. Zeitschrift für Eingeborenen Sprachen 25:81-89. Anderson, B. 1983. Imagined Communities: Reflections on the Origin and Spread of Nationalism. London: Vaso. Arbousset, T., and F. Daumas. 1846. Narrative of an Exploratory Tour to the North-East of the Colony of the Cape of Good Hope. Cape Town: Robertson. Bahn, P. G. 1996. Comment on Dronfield’s “the Vision Thing: Diagnosis of Endogenous Derivation in Abstract Arts”. Cambridge Archaeological Journal 6:55-57. —. 1997. Membrane and Numb Brain: A Close Look at a Recent Claim for Shamanism in Palaeolithic Art. Rock Art Research 14:62-67. Barham, L. S. 1992. Let’s Walk before We Run: An Appraisal of Historical Materialist Approaches to the Later Stone Age. South African Archaeological Bulletin 47:44-51. Barnard, A. 1986. Rethinking Bushman Settlement Patterns and Territoriality. Sprache und Geschichte in Afrika 7:41-60. —. 1989. The Lost World of Laurens Van Der Post? Current Anthropology 30:104-114. —. 1992. Hunters and Herders of Southern Africa. Cambridge: Cambridge University Press. —. 1999. “Images of Hunters and Gatherers in European Social Thought,” in The Cambridge Encyclopedia of Hunters and Gatherers. Edited by R. B. Lee and R. Daly, pp. 375-383. Cambridge: Cambridge University Press. Barnett, A. 2001. “In Africa the Hoodia Cactus Keeps Men Alive. Now Its Secret Is ‘Stolen’ to Make Us Thin.,” in The Observer. London. —. 2002. “Fat Windfall for the San,” in The Observer. London. Barr, H. L., R. J. Langs, R. R. Holt, L. Goldberger, and G. S. Klein. 1972. Lsd: Personality and Experience. Wiley Series on Personality Processes. New York: Wiley-Interscience. Battiss, W. 1948. The Artists of the Rocks. Pretoria. Red Fawn Press. 182 Bednarik, R. G. 1984. On the Nature of Psychograms. The Artefact 8:27-32. —. 1990. On Neuropsychology and Shamanism in Rock Art. Current Anthropology 31:77-80. Beinart, W. 1982. The Political Economy of Pondoland 18601930. Johannesburg: Ravan Press. Bell-Cross, S. 2004. Number of Shipwrecked People who Married into Indigenous Communities along the South-Eastern Mountains. Maclear: Personal Communication. Bender, B. 2001. Landscapes on-the-Move. Journal of Social Archaeology 1:75-89. Beringer, K. 1923. Experimentelle Psychosen Durch Mescalin. Zeitschrift ges. Neurol. Psychiat. 84:426433. —. 1927. Der Meskalinrausch. Monogr. Gesamtgeb. Neurol. & Psychiat. 49:1-315. Bhabha, H. 1994. The Location of Culture. London: Routledge. Biesele, M. 1978. Sapience and Scarce Resources: Communication Systems of The!Kung and Other Foragers. Social Science Information 17:921-947. —. 1993. Women Like Meat: The Folklore and Foraging Ideology of the Kalahari Ju/’Hoan. Johannesburg: Witwatersrand University Press. Binneman, J. 1999. Mummified Human Remains from the Kouga Mountains, Eastern Cape. The Digging Stick 16:1-3. Binneman, J., and S. Hall. 1993. The Context of Four Painted Stones from the Eastern Cape. South African Journal of Field Archaeology 2:89-95. Blackburn, T. 1977. Biopsychological Aspects of Chumash Rock Art. Journal of California Anthropology 4:88-94. Blacking, J. 1977. The Anthropology of the Body. Association of Social Anthropologists of the Commonwealth. London: Academic Press. Bleek, D. F. 1931. Customs and Beliefs of the /Xam Bushmen. Part I: Baboons. Bantu Studies 5:167179. —. 1932. Customs and Beliefs of the /Xam Bushmen. Part Ii: The Lion. Bantu Studies 6:47-63. —. 1932. Customs and Beliefs of the /Xam Bushmen. Part Iii: Game Animals. Bantu Studies 6:233249. —. 1932. Customs and Beliefs of the /Xam Bushmen. Part Iv: Omens, Wind-Making, Clouds. Bantu Studies 6:324-342. —. 1933. Beliefs and Customs of the /Xam Bushmen. Part V: The Rain. Bantu Studies 7:297-312. —. 1933. Beliefs and Customs of the /Xam Bushmen. Part Vi: Rain-Making. Bantu Studies 7:375-392. —. 1935. Beliefs and Customs of the /Xam Bushmen. Part Vii: Sorcerors. Bantu Studies 9:1-47. —. 1936. Beliefs and Customs of the /Xam Bushmen. Part Viii: More About Sorcerors and Charms. Bantu Studies 10:131-162. —. 1956. A Bushman Dictionary. Vol. 41. American Oriental Series. New Haven: American Oriental Society. Bleek, W. H. I., and L. C. Lloyd. 1911. Specimens of Bushman Folklore. London: Allen. Blundell, G. 1993. The Divining Switch as an Item of Material Culture in Space and Time. BA (Hons), University of the Witwatersrand. —. 1998. On Neuropsychology in Southern African Rock Art Research. Anthropology of Consciousness 9:3-12. Blundell, G., and E. B. Eastwood. 2001. Identifying YShape Motifs in the Rock Art of the LimpopoShashi Confluence Area. South African Journal of Science 97:305-308. Blundell, G., and J. D. Lewis-Williams. 2001. Storm Shelter: A Rock Art Discovery in South Africa. South African Journal of Science 97:43-46. Botha, L. J., and J. F. Thackeray. 1987. A Note on Southern African Rock Art, Medicine-Men and Nguni Diviners. South African Archaeological Bulletin 41:71-73. Bourdieu, P. 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press. Boyd, C. E. 1996. Shamanic Journeys into the Otherworld of the Archaic Chichimec. Latin American Antiquity 7:152-164. Boyd, C. E., and J. P. Dering. 1996. Medicinal and Hallucinogenic Plants Identified in the Sediments and Pictographs of the Lower Pecos, Texas Archaic. Antiquity 70:256-275. Breunig, P. 1985. New Excavations in the Brandberg Provide Clues to the Age of the Rock Paintings. The Digging Stick 2:3. Butler, G. 2001. Sister Mariya’s Account of San Painting in the Tsolo District in the Late 1920’s and 1930’s. Southern African Field Archaeology 10:1618. Butler, J. 1993. Bodies That Matter: On the Discursive Limits of “Sex”. New York: Routledge. Callaway, G. 1919. Mxalmi, the Feaster. New York: Macmillan. Campbell, C. 1987. Art in Crisis: Contact Period Rock Art in the South-Eastern Mountains of South- ern Africa. M.Sc. Thesis, University of the Witwatersrand. Cashdan, E. 1985. Coping with Risk: Reciprocity among the Basarwa of Northern Botswana. Man 20:454-474. Challis, W., and J. Hampson. 2000. “Verbal Report (with Tracing) to G. Blundell of Eldritch Image at Bongani Game Reserve, Eastern South Africa.” Johannesburg. Childe, G. V. 1936. Man Makes Himself. London: Watts. —. 1942. What Happened in History. Harmondsworth: Penguin. Clarke, S. 1981. The Foundations of structuralism: A Critique of Lévi-Strauss and the structuralist Movement. Brighton: Harvester Press. Clifford, J. 1988. The Predicament of Culture: Twentieth-Century Ethnography, Literature, and Art. Cambridge (MA): Harvard University Press. Clifford, J., and G. E. Marcus. Editors. 1984. Writing Culture: The Poetics and Politics of Ethnography. Berkeley: University of California Press. Clottes, J., and J. D. Lewis-Williams. 1998. The Shamans of Prehistory: Trance Magic and the Painted Caves. New York: Abrams. Cohen, P. S. 1968. Modern Social Theory. London: Heinemann. Comaroff, J., and J. Comaroff. 1988. Through the Looking-Glass: Colonial Encounters of the First Kind. Journal of Historical Sociology 1:6-32. Corry, S. 2003. Bushmen-the Final Solution and Blaming the Messenger. Before Farming:1-4. Cowan, C. W., P. J. Watson, and N. L. Benco. 1992. The Origins of Agriculture: An International Perspective. Washington: Smithsonian Institution Press. Craib, I. 1992. Anthony Giddens. London: Routledge. Crais, C. 1992. The Making of the Colonial Order: White Supremacy and Black Resistance in the Eastern Cape, 1770-1865. Cambridge: Cambridge University Press. Csordas, T. J. 1994. Embodiment and Experience: The Existential Ground of Culture and Self. Vol. 2. Cambridge Studies in Medical Anthropology. Cambridge: Cambridge University Press. Currlé, L. C. 1913. Notes on Namaqualand Bushmen. Transactions of the Royal Society of South Africa 3:113-120. Cytowic, R. E. 2000. The Man Who Tasted Shapes. Cambridge (MA): The MIT Press. d’ Aquili, E. 1978. The Neurobiological Bases of Myth and Concepts of Deity. Zygon 13:257-275. —. 1986. Myth, Ritual and the Archetypal Hypothesis. Zygon 21:141-160. 183 d’ Aquili, E., and A. B. Newberg. 1993. Liminality, Trance and Unitary States in Ritual and Meditation. Studia Liturgica 23:2-34. —. 1993. Religious and Mystical States: A Neuropsychological Model. Zygon 28:177-200. —. 1999. The Neuropsychological Basis of Religions, or Why God Won’t Go Away. Zygon 33:187-201. d’Aquili, E., and A. B. Newberg. 1999. The Mystical Mind: Probing the Biology of Religious Experience. Minneapolis: Fortress Press. David, B., and M. Wilson. 1999. Re-Reading the Landscape: Place and Identity in Ne Australia During the Late Holocene. Cambridge Archaeological Journal 9:163-188. Davison, P. 1991. Material Culture, Context and Meaning: A Critical Investigation of Museum Practice, with Particular Reference to the South African Museum. Ph.D, University of Cape Town. De Beers. 12 July, 2002. “Media Release: Diamond Mining and the Bushmen of the Central Kalahari.” —. 18 November, 2002. “Media Release: The Bushmen of the Central Kalahari.” Deacon, H. J., and J. Deacon. 1999. Human Beginnings in South Africa: Uncovering the Secrets of the Stone Age. Cape Town: David Philip. Deacon, J. 1984. The Later Stone Age of Southernmost Africa. Cambridge Monographs in African Archaeology; 12. Oxford, England: B.A.R. —. 1986. “My Place Is the Bitterpits”: The Home Territory of Bleek and Lloyd’s Xam San Informants. African Studies 45:135-155. —. 1988. The Power of a Place in Understanding Southern San Rock Engravings. World Archaeology 20:129- 140. —. 1996. “Archaeology of the Flat and Grass Bushmen,” in Voices from the Past: /Xam Bushmen and the Bleek and Lloyd Collection. Edited by J. Deacon and T. A. Dowson, pp. 245-270. Johannesburg: Witwatersrand University Press. —. 1996. “The /Xam Informants,” in Voices from the Past: /Xam Bushmen and the Bleek and Lloyd Collection. Edited by J. Deacon and T. A. Dowson, pp. 11-39. Johannesburg: Witwatersrand University Press. Deacon, J., and T. A. Dowson. Editors. 1996. Voices from the Past: /Xam Bushmen and the Bleek and Lloyd Collection. The Khoisan Heritage Series. Johannesburg: Witwatersrand University Press. Dirks, N. B. 1992. Colonialism and Culture. Ann Arbor: University of Michigan Press. Dobres, M.-A., and C. R. Hoffman. 1999. The Social Dynamics of Technology: Practice, Politics, and World 184 Views. Washington: Smithsonian Institution Press. Douglas, M. 1970. Natural Symbols: Explorations in Cosmology. London: Barrie and Rockliff the Cresset P. Dowson, T. A. 1988. Revelations of Religious Reality: The Individual in San Rock Art. World Archaeology 21:116-127. —. 1989. Dots and Dashes: Cracking the Entoptic Code in Bushman Rock Paintings. South African Archaeological Society Goodwin Series 6:84-94. —. 1992. Rock Engravings of Southern Africa. Johannesburg: Witwatersrand University Press. —. 1993. Changing Fortunes of Southern African Archaeology: Comment on A.D. Mazel’s ‘History’. 67:641-644. —. 1994. Reading Art, Writing History: Rock Art and Social Change in Southern Africa. World Archaeology 25:332-345. —. 1995. “Hunter-Gatherers, Traders and Slaves: The ‘Mfecane’ Impact on Bushmen, Their Ritual and Their Art,” in The Mfecane Aftermath: Reconstructive Debates in Southern African History. Edited by C. Hamilton, pp. 51-70. Johannesburg: Witwatersrand University Press. —. 1998. “Rain in Bushman Belief, Politics and History: The Rock-Art of Rain-Making in the SouthEastern Mountains, Southern Africa,” in The Archaeology of Rock-Art. Edited by C. Chippindale and P. S. C. Taçon, pp. 73-89. Cambridge: Cambridge University Press. —. 2000. “Painting as Politics: Exposing Historical Processes in Hunter-Gatherer Rock Art,” in Hunters and Gatherers in the Modern World: Conflict, Resistance and Self-Determination. Edited by P. P. Schweitzer, M. Biesele, and R. Hitchcock, pp. 413-426. New York: Berghahn Books. Dowson, T. A., G. Blundell, and S. Hall. 1992. Finger Paintings in the Harts River Valley, Northern Cape Province, South Africa. South African Journal of Field Archaeology 1:27-32. Dowson, T. A., and J. D. Lewis-Williams. 1993. Myths, Museums and Southern African Rock Art. South African Historical Journal 29:44-66. Dronfield, J. 1996. The Vision Thing: Diagnosis of Endogenous Derivation in Abstract Arts. Cultural Anthropology 37:373-391. Eichmeier, J., and O. Höfer. 1974. Endogene Bildmuster. Munich: Urban and Schwarzenberg. Eliade, M. 1972. Shamanism: Archaic Techniques of Ecstasy. New York: Routledge and Kegan Paul. Ellenberger, D. F. 1912. History of the Basuto: Ancient and Modern. London: Caxton Publishing Company. England, N. M. 1968. Music in the Societies of Certain Bushman Groups in Southern Africa. Ph.D. thesis, Harvard University. Ferguson, H. 2001. “Phenomenology and Social Theory,” in Handbook of Social Theory. Edited by G. Ritzer and B. Smart, pp. 232-248. London: Sage. Finch, J. 2001. “Bushmen Aim for Cactus Rich Pickings,” in The Guardian. London. Forbes, V. S. 1965. Pioneer Travellers of South Africa: A Geographical Commentary Upon Routes, Records, Observations and Opinions of Travellers at the Cape: 1750-1800. Cape Town: A.A. Balkema. Francis, J. E., and L. L. Loendorf. 2001. Ancient Visions: Petroglyphs and Pictographs from Wind River and Bighorn Country, Wyoming and Montana. Salt Lake City: The University of Utah Press. Friedman, J. 1978. The Evolution of Social Systems. London: Duckworth. Fynn, H. F. 1950. The Diary of Henry Francis Fynn. Pietermaritzburg: Shuter and Shooter. Garlake, P. S. 1987. The Painted Caves: An Introduction to the Prehistoric Rock Art of Zimbabwe. Harare: Modus. —. 1987. Themes in the Prehistoric Art of Zimbabwe. World Archaeology 19:178-193. —. 1990. Symbols of Potency in the Paintings of Zimbabwe. South African Archaeological Bulletin 45:1727. —. 1995. The Hunter’s Vision: The Prehistoric Art of Zimbabwe. London: British Museum Press. Gatens, M. 1996. The Imaginary Body. London: Routledge. Gibson, A. G. S. 1891. Eight Years in Kaffraria. London: Wells Gardner. Giddens, A. 1979. Central Problems in Social Theory: Action, Structure and Contradiction in Social Analysis. London: Macmillan Press. —. 1984. The Constitution of Society: Outline of the Theory of Structuration. Cambridge: Polity Press. —. 1991. Modernity and Self-Identity: Self and Society in the Late Modern Age. Cambridge: Polity Press. Gill, S. J. 1992. “D.F. Ellenberger: A Brief Sketch of His Life & Work,” in History of the Basuto: Ancient and Modern. Edited by S. J. Gill, pp. 5-55. Morija: Morija Museums and Archives. Gilliomee, H. 1979. “The Eastern Frontier,” in The Sahping of South African Society, 1652-1820. Edited by R. Elphick and H. Gilliomee. Cape Town: Longman. Gladwin, A. 1909. “Letter to Colonel Stanford,” in The Sir WEM Stanford Papers, pp. 1-2. Cape Town, Jagger Library. Godelier, M. 1975. “Modes of Production, Kinship, and Demographic Structures,” in Analyses and Social Anthropology. Edited by M. Bloch. London: Malaby Press. —. 1977. Perspectives in Marxist Anthropology. Cambridge: Cambridge University Press. —. 1978. Infrastructures, Societies and History. Current Anthropology 19:763-771. Goering, L. 15th October, 2002. “Angering the Ancestors in Botswana,” in Chicago Tribune, pp. 2. Gordon, R. J. 1992. The Bushman Myth: The Making of a Namibian Underclass. Boulder: Westview Press. —. 1995. Saving the Last South African Bushmen. Critical Arts 9:28-48. —. 1997. Picturing Bushmen: The Denver African Expedition of 1925. Cape Town: David Philip. —. 1998. “Backdrops and Bushmen,” in The Colonizing Camera. Edited by W. Hartmann, J. Silvester, and P. Hayes, pp. 111-117. Cape Town: University of Cape Town Press. —. 1999. “Scene(S) at the Exhibition: Bain’s “Bushmen” at the Empire Exhibition, 1936,” in Africans in Show Business. Edited by B. Lindfors, pp. 266289. Bloomington: Indiana University Press. Gosden, C. 1999. Anthropology and Archaeology: A Changing Relationship. London: Routledge. —. 2001. “Postcolonial Archaeology: Issues of Culture, Identity, and Knowledge,” in Archaeological Theory Today. Edited by I. Hodder, pp. 241-261. Cambridge: Polity Press. Government of Botswana. No Date. “Relocation of Basarwa from the Central Kalahari Game Reserve.” Grof, S. 1979. Realms of the Human Unconscious: Observations from Lsd Research. London: Souvenir Press. Grosz, E. 1994. Volatile Bodies: Toward a Corporeal Feminism. Bloomington: Indiana University Press. Guenther, M. 1999. Tricksters and Trancers: Bushman Religion and Society, First edition. Bloomington: Indiana University Press. Guenther, M. G. 1975. The Trance Dancer as an Agent of Social Change among the Farm Bushmen of the Ghanzi District. Botswana notes and records 7:161-166. —. 1975/1976. The San Trance Dance: Ritual and Revitalization among the Farm Bushmen of the Ghanzi District, Republic of Botswana. S.W.A Wissenschaftliche Gesellschaft 30:45-53. Guignon, C. B. 1995. “Heidegger, Martin,” in The Cambridge Dictionary of Philosophy. Edited by R. Audi. Cambridge: Cambridge University Press. Gulbrandsen, Ø. 1991. “On the Problem of Egalitarianism: The Kalahari San in Transition,” in The 185 Ecology of Choice and Symbol: Essays in Honour of Fredrik Barth. Edited by R. Grønhaug, G. Haaland, and G. Henriksen, pp. 81-110. Bergen: Alma Mater Forlag. Halifax, J. 1980. Shamanic Voices. Harmondsworth: Penguin Books. —. 1982. Shaman: The Wounded Healer. New York: Crossroad. Hall, M. 1996. “The Proximity of Dr Bleek’s Bushman,” in Miscast: Negotiating the Presence of the Bushmen. Edited by P. Skotnes, pp. 143-159. Cape Town: University of Cape Town Press. Hall, S. 1994. “Images of Interaction: Rock Art and Sequence in the Eastern Cape,” in Contested Images: Diversity in Southern African Rock Art Research. Edited by T. A. Dowson and J. D. Lewis-Williams, pp. 61-82. Johannesburg: Witwatersrand University Press. —. 2000. Burial Sequence in the Later Stone Age of the Eastern Cape Province, South Africa. South African Archaeological Bulletin 55:137-146. Hall, S., and J. Binneman. 1987. Later Stone Age Variability in the Cape: A Social Interpretation. South African Archaeological Bulletin 42:140-152. Hall, S., and F. E. Prins. 1993. Late White Rock Paintings of Central and Southern Africa. Lantern 42:2. Hamayon, R. N. 1998. “’Ecstasy’ or the West-Dreamt Siberian Shaman,” in Tribal Epistemologies: Essays in the Philosophy of Anthropology. Edited by H. Wautischer, pp. 175-187. Aldershot: Ashgate. Hamilton, C. 1998. Terrific Majesty: The Powers of Shaka Zulu and the Limits of Historical Invention. Cape Town: David Philip. Hammond-Tooke, W. D. 1962. Bhaca Society, a People of the Transkeian Uplands, South Africa. Cape Town: Oxford University Press. —. 1989. Rituals and Medicines: Indigenous Healing in South Africa. Johannesburg: AD Donker. —. 1997. Whatever Happened to /Kaggen? A Note on Cape Nguni/Khoisan Borrowing. South African Archaeological Bulletin 55:22-24. —. 1998. Selective Borrowing? The Possibility of San Shamanistic Influence on Southern Bantu Divination and Healing Practices. South African Archaeological Bulletin 53:9-15. —. 1999. Divinatory Animals: Further Evidence of San-Nguni Borrowing. South African Archaeological Bulletin 54:128-132. —. 2002. The Uniqueness of Nguni Mediumistic Divination in Southern Africa. Africa 72:277-292. —. n.d. “Unpublished Field Notes.” 186 Hampson, J., W. Challis, G. Blundell, and C. de Rosner. 2002. The Rock Art of Bongani Mountain Lodge and Its Environs, Mpumalanga Province, South Africa: An Introduction to the Problems of Southern African Rock-Art Regions. South African Archaeological Bulletin 57:15-30. Harding, W. 1850. “Report of Walter Harding,” in Cape Government House 8/23, pp. 325-337. Cape Town. —. 1850. “Statement of Mangana,” in Cape Government House 8/23, pp. 406-408. Cape Town. —. 1850. “Statement of Qinti,” in Cape Government House 8/23, pp. 414-417. Cape Town. —. 1850. “Statement of Willem Lochenberg,” in Cape Government House 8/23, pp. 408-411. Cape Town. Harris, D. R. 1996. The Origins and Spread of Agriculture and Pastoralism in Eurasia. London: UCL Press. Hedges, K. 1976. Southern California Rock Art as Shamanic Art. American Indian Rock Art 2:126-38. —. 1979. The Rancho Bernardo Style in Southern California. American Indian Rock Art 5:51-8. —. 1982. Phosphenes in the Context of Native American Rock Art. American Indian Rock Art 7/8:110. —. 1983. “The Shamanic Origins of Rock Art,” in Ancient Images on Stone: Rock Art of the Californias. Edited by J. Van Tilburg. Los Angeles: UCLA Rock Art Archive. —. 1985. Rock Art Portrayals of Shamanic Transformation and Magical Flight. San Diego Museum Papers 18:83-92. —. 1992. “Shamanistic Aspects of California Rock Art,” in Calfornia Indian Shamanism. Edited by L. J. Bean, pp. 67-88. Menlo Park: Ballena Press. —. 1994. “Pipette Dreams and the Primordial SnakeCanoe: Analysis of a Hallucinatory Form Constant,” in Shamanism and Rock Art in North America. Edited by S. A. Turpin, pp. 103-124. San Antonio: Rock Art Foundation. Heinz, H. J. 1975. Elements Of!Ko Bushman Religious Beliefs. Anthropos 710:17-41. Hekman, S. J. 1990. Gender and Knowledge: Elements of a Postmodern Feminism. Cambridge: Polity Press. Henshilwood, C. S. 1996. A Revised Chronology for Pastoralism in Southernmost Africa: New Evidence of Sheep at C. 2000 B.P. From Blombos Cave, South Africa. Antiquity 70:945-949. Henshilwood, C. S., F. d’ Errico, R. Yates, Z. Jacobs, C. Tribolo, G. A. T. Duller, M. N., J. C. Sealy, H. Valladas, I. Watts, and A. G. Wintle. 2002. Emergence of Modern Human Behaviour: Middle Stone Age Engravings from South Africa. Science 295:1278-1280. Henshilwood, C. S., J. C. Sealy, R. J. Yates, K. CruzUribe, P. Goldberg, F. E. Grine, R. G. Klein, C. Poggenpoel, K. L. van Niekerk, and I. Watts. 2001. Blombos Cave, Southern Cape, South Africa: Preliminary Report on the 1992 - 1999 Excavations of the Middle Stone Age Levels. Journal of Archaeological Science 28:421-448. Herbert, R. K. 1990a. The Relative Markedness of Click Sounds: Evidence from Language Change, Acquisition and Avoidance. Anthropological Linguistics 32:295-315. —. 1990b. The Sociohistory of Clicks in Southern Bantu. Anthropological Linguistics 32:120-138. Hewitt, R. L. 1986. Structure, Meaning and Ritual in the Narratives of the Southern San. Hamburg: Helmut Buske Verlag. Highwater, J. 1982. The Primal Mind: Vision and Reality in Indian America. New York: Meridian. Hodder, I. Editor. 2001. Archaeological Theory Today. Cambridge: Polity Press. Hollmann, J. 2002. Natural Models, Ethology and San Rock-Paintings: Pilo-Erection and Depictions of Bristles in South-Eastern South Africa. South African Journal of Science 98:563-567. —. 2003. Indigenous Knowledge and Paintings of Human-Animal Combinations: Ostriches, Swifts and Religion in Bushman Rock-Art, Western Cape Province, South Africa. M.A. Dissertation, University of the Witwatersrand. Hook, D. B. 1908. With Sword and Statute. London: Greaves, Pass and Co. Hösler, D. 1995. Sound, Color and Meaning in the Metallurgy of Ancient West Mexico. World Archaeology 27:100-113. How, M. W. 1962. The Mountain Bushmen of Basutoland. Pretoria: Van Schaik. Howell, N. 1979. Demography of the Dobe!Kung. New York: Academic Press. Hromnik, C. A. 1991. Bushmen Shamans: An Hallucinatory Dream. The Digging Stick 8:6. —. 1991. A Testament to the Shamanistic Hallucinatory Trance Theory of the Southern African Rock Art. Rock Art Research 8:99-102. Hudleson, J. E. 1995. “One Hundred Years among the San: A Social History of San Research,” in Speaking for the Bushmen. Edited by A. J. G. M. Sanders. Gaborone: Botswana Society. Huffman, T. N. 1983. “The Trance Hypothesis and the Rock Art of Zimbabwe,” in New Approaches to Southern African Rock Art, vol. 4. Edited by J. D. Lewis-Williams, pp. 49-53. Cape Town: The South African Archaeological Society Goodwin Series. Hughes, A., and A. Witz. 1997. Feminism and the Matter of Bodies from De Beauvoir to Butler. Body & Society 3:47-60. Ingold, T. 1999. “On the Social Relations of the Hunter-Gatherer Band,” in The Cambridge Encyclopedia of Hunters and Gatherers. Edited by R. B. Lee and R. Daly, pp. 399-410. Cambridge: Cambridge University Press. Inskeep, R. R. 1971. “The Future of Rock Art Studies in Southern Africa,” in Rock Paintings of Southern Africa. Edited by M. Schoonraad, pp. 101-104. Johannesburg: South African Association for the Advancement of Science. Isaac, G. L. 1989. The Archaeology of Human Origins. Cambridge: Cambridge University Press. Jacobus, M., E. F. Keller, and S. Shuttleworth. 1990. Body/Politics: Women and the Discourses of Science. New York: Routledge. Jenkins, T., A. Zoutendyk, and A. G. Steinberg. 1970. Gammaglobulin Groups (Gm and Inv) of Various Southern African Populations. American Journal of Physical Anthropology 32:198. Johnson, M. 1999. Archaeological Theory: An Introduction. Oxford: Blackwell. Johnstone, M. 1993. “Phenomenology,” in Key Ideas in Human Thought. Edited by K. McLeish. London: Bloomsbury. Jolly, P. 1986. A First Generation Descendant of the Transkei San. South African Archaeological Bulletin 41:6-9. —. 1992. Some Photographs of Late Nineteenth Century San Rainmakers. South African Archaeological Bulletin 47:89-93. —. 1994. Strangers to Brothers: Interaction between South-Eastern San and Southern Nguni/Sotho Communities. MA thesis, University of Cape Town. —. 1995. Melikane and Upper Mangolong Revisited: The Possible Effects on San Art of Symbiotic Contact between South-Eastern San and Southern Nguni and Sotho Communities. South African Archaeological Bulletin 50:68-80. —. 1996. “Between the Lines: Some Remarks on Bushman Ethnicity,” in Miscast: Negotiating the Presence of the Bushmen. Edited by P. Skotnes, pp. 197-209. Cape Town: University of Cape Town Press. —. 1996. Interaction between South-Eastern San and Southern Nguni and Sotho Communities C. 1400 to C. 1800. South African Historical Journal 35:30-61. 187 —. 1996. Symbiotic Interaction between Black Farmers and South-Eastern San. Current Anthropology 37:277-305. —. 1997. An Evaluation of Recent Oral Evidence Relating to Nguni Diviners and the South-Eastern San. Kronos 24:119-128. —. 1998. “Modelling Change in the Contact Art of the South-Eastern San, Southern Africa,” in The Archaeology of Rock Art. Edited by C. Chippindale and P. S. C. Taçon. Cambridge: Cambridge University Press. —. 1999. The Identity of San Informants from the Tsolo District, Eastern Cape. South African Archaeological Bulletin 54:61-63. —. 2000. Nguni Diviners and the South-Eastern San: Some Issues Relating to Their Mutual Cultural Influence. Natal Museum Journal of Humanities 12:79-95. —. 2002. Therianthropes in San Rock Art. South African Archaeological Bulletin 57:85-103. Jolly, P., and F. Prins. 1994. M.-a Further Assessment. South African Archaeological Bulletin 49:16-23. Kandinsky, W. 1977. Concerning the Spiritual in Art, and Painting in Particular. New York: Dover. Katz, R. 1982. Boiling Energy: Community Healing among the Kalahari Kung, First edition. Cambridge (MA): Harvard University Press. Katz, R., and M. Biesele. 1986. “Kung Healing: The Symbolism of Sex Roles and Culture Change.,” in The Past and Present Of!Kung Ethnography: Critical Reflections and Symbolic Perspectives. Essays in Honour of Lorna Marshall. Edited by M. Biesele, R. J. Gordon, and R. B. Lee, pp. 195-230. Hamburg: Helmut Buske Verlag. Katz, R., Biesele, M. and St. Denis, V. 1997. Healing Makes Our Hearts Happy: Spirituality and Cultural Transformation among the Kalahari Ju/’Hoansi. Rochester: Inner Traditions. Keaton, J. August 25, 2000. “De Beers Buffs Its Image: Cartel out, Competition in at Diamond Leader That’s Seen to Eye U.S. Entrée,” CNNMoney. Keeney, B. Editor. 1999. Kalahari Bushman Healers, 1st edition. Vol. 3. Profiles of Healing. Philadelphia: Ringing Rocks Press. —. Editor. 2003. Ropes to God: Experiencing the Bushman Spiritual Universe. Profiles of Healing. Philadelphia: Ringing Rocks Press. Kehoe, A. 2002. Emerging Trends Versus the Popular Paradigm in Rock Art Research. Antiquity 76:384-5. Kent, S. 1992. The Current Forager Controversy: Real Versus Ideal Views of Hunter-Gatherers. Man 27:45-70. 188 Kinahan, J. 1989. Pastoral Nomads of the Central Namib Desert. Ph.D.Thesis, University of the Witwatersrand. —. 1991. Pastoral Nomads of the Central Namib Desert: The People History Forgot. Windhoek: New Namibia Books. King, A. G. 1998. Superpositioning and the Art of Fallen Rock, Junction Shelter, Didima Gorge, KwaZulu Natal. BSc (Hons) dissertation, University of the Witwatersrand. Kirby, P. R. Editor. 1953. Source Book on the Wreck of the Grosvenor. Cape Town: Van Riebeeck Society. —. 1954. Gquma, Mdepa and the Amatshomane Clan: A by-Way of Miscegenation in South Africa. African Studies 13:1-24. —. 1958. Jacob Van Reenen and the Grosvenor Expedition of 1790-1791. Johannesburg: Witwatersrand University Press. Klüver, H. 1966. Mescal, and Mechanisms of Hallucinations. Chicago: University of Chicago Press. Knapp, A. B., and L. M. Meskell. 1997. Bodies of Evidence in Prehistoric Cyprus. Cambridge Archaeological Journal 7:183-204. Knoll, M., and J. Kugler. 1959. Subjective Light-Pattern Spectroscopy in the Electroencephelographic Frequency Range. Nature 184:1823-1824. Knoll, M., J. Kugler, O. Höfer, and S. D. Lawder. 1963. Effects of Chemical Stimulation of Electrically-Induced Phosphenes on Their Bandwidth, Shape, Number and Intensity. Confinia Neurologica 23:201-226. Kockelmans, J. J. 1995. “Phenomenology,” in The Cambridge Dictionary of Philosophy. Edited by R. Audi. Cambridge: Cambridge University Press. Kopytoff, I. 1987. The African Frontier. Bloomington: Indiana University Press. Kuckertz, H. 1990. Creating Order: The Image of the Homestead in Mpondo Social Life. Johannesburg: Witwatersrand University Press. Kuhn, T. S. 1962. The Structure of Scientific Revolutions. Chicago: University of Chicago Press. Lambek, M., and A. Strathern. 1998. Bodies and Persons: Comparative Perspectives from Africa and Melanesia. Cambridge: Cambridge University Press. Landry, D., and G. Maclean. Editors. 1996. The Spivak Reader: Selected Works of Gayatri Chakravorty Spivak. New York: Routledge. Lanham, L. W. 1964. “The Proliferation and Extension of Bantu Phonemic Systems Influenced by Bushman and Hottentot,” in Proceedings of the Ninth International Congress of Linguists. Edited by H. G. Lunt. The Hague: Mouton. Laue, G. 1999. The Rock Art of the Uniondale District: The Significance of Regionality in Southern African Rock Art. Honours dissertation, University of the Witwatersrand. Leach, E. R. 1970. Claude Lévi-Strauss. New York: Viking. Lee, D. N., and H. C. Woodhouse. 1964. Rock Paintings of Flying Buck. South African Archaeological Bulletin 19:71-74. —. 1968. More Rock Paintings of Flying Buck. South African Archaeological Bulletin 23:13-16. Lee, R. B. 1967. Trance Cure of The!Kung Bushman. Natural History 76:31-37. —. 1968. “The Sociology Of!Kung Bushman Trance Performance,” in Trance and Possession States. Edited by R. Prince, pp. 35-54. Montreal: R.M. Bucke Memorial Society. —. 1979. The!Kung San: Men, Women and Work in a Foraging Society. Cambridge: Cambridge University Press. —. 1984. The Dobe!Kung. New York: Holt, Rinehart and Winston. Legassick, M. 1980. “The Frontier Tradition in South African Historiogaphy,” in Economy and Soceity in Pre-Industrial South Africa. Edited by S. Marks and A. Atmore, pp. 44-79. London: Longman. Leland, D. 1995. “Embodiment,” in The Cambridge Dictionary of Philosophy. Edited by R. Audi. Cambridge: Cambridge University Press. Lenssen-Erz, T., and M.-T. Erz. 2000. Brandberg: Der Bilderberg Namibias: Kunst Und Geschichte Einer Urlandschaft. Stuttgart: Jan Thorbecke Verlag. Leroi-Gourhan, A. 1968. The Art of Prehistoric Man in Western Europe. London: Thames and Hudson. —. 1982. The Dawn of European Art: An Introduction to Palaeolithic Cave Painting. Cambridge: Cambridge University Press. Lewis-Williams, J. D. 1972. The Syntax and Function of the Giant’s Castle Rock Paintings. South African Archaeological Bulletin 27:49-65. —. 1974. Superpositioning in a Sample of Rock-Paintings in the Barkly East District. South African Archaeological Bulletin 29:93-103. —. 1975. “The Drakensberg Rock Paintings as an Expression of Religious Thought,” in Valcamonica Symposium ‘72-Actes Du Symposium International Sup. Les Religions De La Préhistoire. Edited by E. Anati, pp. 413-426. Capo di Ponte: Centro Camuno di Studi Preistorici. —. 1981. Believing and Seeing: Symbolic Meanings in Southern San Rock Paintings. London: Academic Press. —. 1982. The Economic and Social Context of Southern San Rock Art. Current Anthropology 23:429449. —. 1984. The Empiricist Impasse in Southern African Rock Art Studies. South African Archaeological Bulletin 39:58-66. —. 1984. “Ideological Continuities in Prehistoric Southern Africa: The Evidence of Rock Art,” in Interfaces: The Relationship of Past and Present in Hunter-Gatherer Studies. Edited by C. Schrire, pp. 225-252. New York: Academic Press. —. 1985. Testing the Trance Explanation of Southern African Rock Art: Depictions of Felines. Bolletino del Centro Camuno di Studi Preistorici 22:4762. —. 1986. The Last Testament of the Southern San. South African Archaeological Bulletin 41:10-11. —. 1990. Discovering Southern African Rock Art. Cape Town: David Philip. —. 1990. Documentation, Analysis and Interpretation: Dilemmas in Rock Art Research. Review of “the Rock Paintings of the Upper Brandberg. Part I: Amis Gorge”. South African Archaeological Bulletin 45:126-136. —. 1992. Ethnographic Evidence Relating to ‘Trance’ and ‘Shamans’ among Northern and Southern Bushmen. South African Archaeological Bulletin 47:56-60. —. 1992. Vision, Power and Dance: The Genesis of a Southern African Rock Art Panel. Fourteenth Kroon Lecture. Amsterdam: Stichting Nederlands Museum voor Anthropologie en Praehistorie. —. 1994. Rock Art and Ritual: Southern Africa and Beyond. Complutum 5:277-289. —. 1995. Modelling the Production and Consumption of Rock Art. South African Archaeological Bulletin 50:143-154. —. 1995. “Perspectives and Traditions in Southern African Rock Art Research,” in Perceiving Rock Art: Social and Political Perspectives. Edited by K. Helskog and B. Olsen. Oslo: Novus Forlag. —. 1996. “A Visit to the Lion’s House: The Structure and Metaphors of a Nineteenth Century Bushman Myth,” in Voices from the Past: /Xam Bushmen and the Bleek and Lloyd Collection. Edited by J. Deacon and T. A. Dowson, pp. 122-141. Johannesburg: Witwatersrand University Press. —. 1997. Art, Agency and Altered Consciousness: A Motif in French (Quercy) Upper Palaeolithic Parietal Art. Antiquity 71:810-830. —. 1997. “The Mantis, the Eland and the Meerkats: Conflict and Mediation in a Nineteenth Century San Myth,” in Culture and the Commonplace: 189 Anthropological Essays in Honour of David Hammond-Tooke, vol. 56(2). Edited by P. McAllister, pp. 195-216. Johannesburg: Special Edition of African Studies. —. 1998. Quanto? The Issue of ‘Many’ Meanings in Southern African San Rock Art Research. South African Archaeological Bulletin 53:86-97. —. 1999. “Foreword,” in Nyae Nyae!Kung: Beliefs and Rites (by Lorna J. Marshall), pp. ix-xvi. Cambridge (MA): Peabody Museum of Archaeology and Ethnology. —. 1999. ‘Meaning’ in Southern African San Rock Art: Another Impasse? South African Archaeological Bulletin 54:141-145. —. 2000. “Pers. Comm. On the Omission of Colour Symbolism Chapter from “Believing and Seeing: Symbolic Meanings in Southern San Rock Paintings”.” Johannesburg. —. 2000. Stories That Float from Afar: Ancestral Folklore of the San of Southern Africa. Cape Town: David Philip. —. 2002. The Mind in the Cave: Consciousness and the Origins of Art. London: Thames & Hudson. —. 2003. Images of Mystery: Rock Art of the Drakensberg. Cape Town: Double Storey. —. 2003. “Use of the Words ‘Medicine-Man’ and ‘Shaman’.” Johannesburg. —. 2004. Constructing a Cosmos: Architecture, Power and Domestication at Çatalhöyük. Journal of Social Archaeology 4:28-59. Lewis-Williams, J. D., and M. Biesele. 1978. Eland Hunting Rituals among Northern and Southern San Groups: Striking Similarities. Africa 48:117-134. Lewis-Williams, J. D., and G. Blundell. 1997. New Light on Finger-Dots in Southern African Rock Art: Synesthesia, Transformation and Technique. South African Journal of Science 93:51-54. Lewis-Williams, J. D., G. Blundell, W. Challis, and J. Hampson. 2000. Threads of Light: Re-Examining a Motif in Southern African San Rock Art. South African Archaeological Bulletin 55:123-136. Lewis-Williams, J. D., and J. Clottes. 1998. Shamanism and Upper Palaeolithic Art: A Response to Bahn. Rock Art Research 15:46-50. Lewis-Williams, J. D., P. den Hoed, T. A. Dowson, and G. A. Whitelaw. 1986. Two Crabs in a Box. The Digging Stick 3:1. Lewis-Williams, J. D., and T. A. Dowson. 1988. Signs of All Times: Entoptic Phenomena in Upper Palaeolithic Art. Current Anthropology 29:201-245. —. 1989. Images of Power: Understanding Bushman Rock Art. Johannesburg: Southern Book Publishers. 190 —. 1990. Reply to Bednarik’s “on Neuropsychology and Shamanism in Rock Art”. Current Anthropology 31:80-84. —. 1990. Through the Veil: San Rock Paintings and the Rock Face. South African Archaeological Bulletin 45:5-16. —. 1992. Rock Paintings of the Natal Drakensberg. Pietermaritzburg: Natal University Press. —. 1994. “Aspects of Rock at Research: A Critical Retrospective,” in Contested Images: Diversity in Southern African Rock Art. Edited by T. A. Dowson and J. D. Lewis-Williams, pp. 201-221. Johannesburg: Witwatersrand University Press. —. 2000. Images of Power: Understanding Bushman Rock Art, Second edition. Johannesburg: Southern Book Publishers. Lewis-Williams, J. D., and D. G. Pearce. Manuscript. “Southern African San Rock Painting as Social Intervention: An Epistemological and Methodological Study of Rain-Control Images.” Lipsey, R. 1997. An Art of Our Own: The Spiritual in Twentieth Century Art. Boston: Shambhala. Loendorf, L. L. 1994. “Finnegan Cave: A Rock Art Vision Quest Site in Montana,” in Shamanism and Rock Art in North America. Edited by S. A. Turpin, pp. 125-137. San Antonio: Rock Art Foundation. Loubser, J., and G. Laurens. 1994. “Depictions of Domestic Ungulates and Shields: Hunter/Gatherers and Agro-Pastoralists in the Caledon River Valley Area,” in Contested Images: Diversity in Southern African Rock Art Research. Edited by T. A. Dowson and J. D. Lewis-Williams, pp. 83118. Johannesburg: Witwatersrand University Press. Louw, J. A. 1974. The Influence of Khoi on the Xhosa Language. Limi 2:43-93. —. 1977. The Adaptation of Non-Click Consonants in Xhosa. Khoisan Linguistic Studies 3:74-92. —. 1979. A Preliminary Survey of Khoi and San Influence in Zulu. Khoisan Linguistic Studies 8:89-21. Loy, T., and C. D. Matheson. 2000. Screening for Prehistoric Blood Residues: Using the Hemastix Colourimetric Test. Journal of Archaeological Science. MaCann, C. 1993. Four Phenomenological Philosophers: Husserl, Heidegger, Sartre, Merleau-Ponty. London: Routledge. Mackenzie, J. 1871. Ten Years North of the Orange River: A Story of Everyday Life and Work among the South African African Tribes from 1859-1869. London: Frank Cass & Co. Maclear, T. n.d. “Diary of Sir Thomas Maclear,” in Cape Archives A515/72. Cape Town. Madison, G. B. 1995. “Merleau-Ponty, Maurice (190861),” in The Cambridge Dictionary of Philosophy. Edited by R. Audi. Cambridge: Cambridge University Press. Maggs, T. M. O. C. 1967. A Quantative Analysis of the Rock Art from a Sample Area in the Western Cape. South African Journal of Science 63:100-104. Maggs, T. M. O. C., and V. Ward. 1995. Rock Engravings by Agriculturist Communities in Savanna Areas of the Thukela Basin. Natal Museum Journal of Humanities 7:17-40. Malkii, L. 1992. National Geographic: The Rooting of Peoples and the Territorialization of National Identity among Scholars and Refugees. Cultural Anthropology 7:24-44. Mandt, G. 1995. “Alternative Analogies in Rock Art Interpretation: The West Norwegian Case,” in Perceiving Rock Art: Social and Political Perspectives. Edited by K. Helskog and B. Olsen, pp. 263291. Oslo: Novus Forlag. Manhire, A. 1998. The Role of Hand Prints in the Rock Art of the South-Western Cape. South African Archaeological Bulletin 53:98-108. Manhire, A. H., J. E. Parkington, A. D. Mazel, and T. M. O. C. Maggs. 1986. “Cattle, Sheep and Horses: A Review of Domestic Animals in the Rock Art of Southern Africa,” in Prehistoric Pastoralism in Southern Africa, vol. 5, The South African Archaeological Society Goodwin Series. Edited by M. Hall and A. B. Smith, pp. 22-30. Cape Town: The South African Archaeological Society Goodwin Series. Manhire, A. H., J. E. Parkington, and W. J. Van Rijssen. 1983. “A Distributional Approach to the Interpretation of Rock Art in the South-Western Cape,” in New Approaches to Southern African Rock Art, vol. 4. Edited by J. D. Lewis-Williams, pp. 29-33. Cape Town: The South African Archaeological Society Goodwin Series. Maquarrie, J. W. 1962. The Reminiscences of Sir Walter Stanford. Vol. 2. Cape Town: Van Riebeeck Society. Marks, S. 1972. Khoisan Resistance to the Dutch in the Seventeenth and Eighteenth Centuries. Journal of African History 13:55-80. Marshall-Thomas, E. 1959. The Harmless People. London: Secker and Warburg. Marshall, J., and C. Ritchie. 1984. Where Are the Ju/Wasi of Nyae Nyae? Changes in a Bushman Society, 19581981. Cape Town: Centre for African Studies University of Cape Town. Marshall, L. 1961. Sharing, Talking and Giving: Relief of Social Tensions Among!Kung Bushmen. Africa 31:231-249. —. 1969. The Medicine Dance of The!Kung Bushmen. Africa 39:347-381. —. 1976. The!Kung of Nyae Nyae. Cambridge: Harvard University Press. Marshall, L. J. 1999. Nyae Nyae!Kung: Beliefs and Rites. Vol. 8. Peabody Museum Monographs. Cambridge (MA): Peabody Museum. Martin, E. 1990. The End of the Body? American Ethnologist:121-. Mazel, A. D. 1992. Changing Fortunes: 150 Years of San Hunter-Gatherer History in the Natal Drakensberg, South Africa. Antiquity 66:758-767. —. 1993. Rock Art and Natal Drakensberg HunterGatherer History: A Reply to Dowson. Antiquity 67:889-892. —. 1996. “In Pursuit of San Pre-Colonial History in the Natal Drakensberg: An Historical Overview,” in Miscast: Negotiating the Presence of the Bushmen. Edited by P. Skotnes, pp. 191-195. Cape Town: University of Cape Town Press. Mazel, A. D., and A. L. Watchman. 1997. Accelerator Radiocarbon Dating of Natal Drakensberg Paintings: Results and Implications. Antiquity 71:445-449. —. 2003. Dating Rock Paintings in the UkhahlambaDrakensberg and the Biggarsberg, KwaZuluNatal, South Africa. Natal Museum Journal of Humanities 15:59-73. Mcbrearty, S., and A. S. Brooks. 2000. The Revolution That Wasn’t: A New Interpretation of the Origin of Modern Human Behavior. Journal of Human Evolution 39:453-563. McLeish, K. Editor. 1993. Key Ideas in Human Thought. London: Bloomsbury. Merleau-Ponty, M. 1962. The Phenomenology of Perception. London: Routledge and Kegan Paul. Meskell, L. M. 1996. The Somatisation of Archaeology: Discourses, Institutions, Corporeality. Norwegian Archaeological Review 29:1-16. —. 1998. “The Irresistible Body and the Seduction of Archaeology,” in Changing Bodies, Changing Meanings: Studies of the Human Body in Antiquity. Edited by D. Montserrat, pp. 139-161. London: Routledge. —. 1999. Archaeologies of Social Life: Age, Sex, Class Et Cetera in Ancient Egypt. Oxford: Blackwell Publishers. Meskell, L. M., and R. A. Joyce. 2003. Embodied Lives: Figuring Ancient Maya and Egyptian Experience. New York: Routledge. 191 Mguni, S. 1997. The Evaluation of the Superpositioning Sequence of Painted Images to Infer Relative Chronology: Diepkloof Kraal Shelter as a Case Study. BA (Hons) Dissertation, University of Cape Town. —. 2003. “Verbal Report to G. Blundell on the Presence of Eldritch-Like Images in the Rock Art of the Matopos, Zimbabwe.” Johannesburg. Miller, D., M. Rowlands, and C. Tilley. 1989. “Introduction,” in Damination and Resistance. Edited by D. Miller, M. Rowlands, and C. Tilley, pp. 1-26. London: Unwin Hyman. Mitchell, P. 2002. The Archaeology of Southern Africa. Cambridge: Cambridge University Press. Moodie, D. 1960. The Record: Or a Series of Official Papers Relative to the Condition and Treatment of the Natives. Amsterdam: A.A. Balkema. Morris, C. 1946. Signs, Language and Behaviour. New York: George Braziller. —. 1964. Signification and Significance. Cambridge, Massachusetts: The M.I.T. Press. Newton-King, S. 1981. “Background to the Khoikhoi Rebellion of 1799-1803,” in Collected Seminar Papers on the Societies of Southern Africa. London: Institute of Commonwealth Studies, University of London. Ngubane, H. 1977. Body and Mind in Zulu Medicine. London: Academic Press. Noll, R. 1983. Shamanism and Schizophrenia: A StateSpecific Approach to ‘Schizophrenia Metaphor’ of Shamanic States. American Ethnologist 10:443459. —. 1985. Mental Imagery Cultivation as a Cultural Phenomenon: The Role of Visions in Shamanism. Current Anthropology 26:443-461. O’Ferrall, R. M. 10th April, 2003. “Transcript of Remarks Made at the Center for Strategic and International Studies, Washington Dc, Entitled “the Kimberley Process to Curb Conflict Diamonds: Early Achievements, Congressional Action, Next Steps-Critical Challenges in Phase Two of the Kimberley Process”.” De Beers Group. Opperman, H. 1987. The Later Stone Age of the Drakensberg Range and Its Foothills. Bar International Series 339, Cambridge Monographs in African Archaeology 19. Oxford: B.A.R. —. 1996. “Excavation of a Later Stone Age Deposit in Strathalan Cave a, Maclear District, North Eastern Cape, South Africa,” in Aspects of African Archeaology. Papers from the 10th Congress of the Pan African Association for Prehistory and Related 192 Studies. Edited by G. S. Pwiti, R, pp. 335-342. Harare: University of Zimbabwe. —. 1996. Strathalan Cave B, North-Eastern Cape Province, South Africa: Evidence for Human Behaviour 29 000-26 000 Years Ago. Quaternary International 33:45-53. —. 1999. A 300 Year-Old Living Floor in Strathalan Cave a, Maclear District, Eastern Cape. Southern African Field Archaeology 8:76-80. Opperman, H., and B. Heydenrych. 1990. A 22 000 Year-Old Middle Stone Age Camp Site with Plant Food Remains from the North-Eastern Cape. South African Archaeological Bulletin 45:9399. Orpen, J. M. 1874. A Glimpse into the Mythology of the Maluti Bushmen. Cape Monthly Magazine 9:113. Ortner, S. 1974. “Is Male to Female as Nature Is to Culture?” in Woman, Culture and Society. Edited by M. S. Rosaldo and L. Lamphere, pp. 67-88. Stanford: Stanford University Press. Oster, G. 1970. Phosphenes. Scientific American 222:8387. Ouzman, S. 1996. Thaba Sione: Place of Rhinoceroses and Rain-Making. African Studies 55:31-59. —. 2001. Seeing Is Deceiving: Rock-Art and the NonVisual. World Archaeology 33:237-256. Ouzman, S., and J. H. N. Loubser. 2000. Art of the Apocalypse. Discovering Archaeology 2:38-44. Pager, H. 1971. Ndedema. Graz: Akademische Druck. Pearce, D. G. 2001. Harris Matrices in Southern African Rock Art: Towards a Chronology of the San Rock Art of Maclear District, Eastern Cape Province. B.Sc. (Honours), University of the Witwatersrand. —. 2002. The Tierkloof Burial Stones: The Art and Context of Painted Stones in the Eastern Cape, South Africa. M.Sc., University of the Witwatersrand. Pearson, J. L. 2002. Shamanism and the Ancient Mind: A Cognitive Approach to Archaeology. Walnut Creek: Altamira Press. Peirce, C. 1931-35. Collected Papers of Charles Sanders Peirce. Cambridge, Massachusetts: Harvard University Press. Penn, N. 1999. Rogues, Rebels and Runaways: Eighteenth-Century Cape Characters. Cape Town: David Philip. Potgieter, E. F. 1955. The Disappearing Bushmen of Lake Chrissie: A Preliminary Survey. Pretoria: J.L. Van Schaik. Price, N. S. 2001. “An Archaeology of Altered States: Cognition, Cosmology and World-View,” in The Archaeology of Shamanism. Edited by N. S. Price. London: Routledge. Prins, F. 2004. “Departure of Manqindi Dyantyi’s Daughter from Nomansland.” Kimberley. Prins, F. E. 1990. Southern Bushman Descendants in the Transkei: Rock Art and Rainmaking. South African Journal of Ethnology 13:110-116. —. 1994. Living in Two Worlds: The Manipulation of Power Relations, Identity and Ideology by the Last San Rock Artist of the Transkei, South Africa. Natal Museum Journal of Humanities 6:179193. —. 1999. Dissecting Diviners: On Positivism, TranceFormations, and the Unreliable Informant. Natal Museum Journal of Humanities 11:43-62. Prins, F. E., and S. Hall. 1994. Expressions of Fertility in the Rock Art of Bantu-Speaking Agriculturists. African Archaeological Review 12:169-201. Reed, C. A. Editor. 1977. Origins of Agriculture. The Hague: Mouton. Reichel-Dolmatoff, G. 1969. El Contexto Cultural De Un Alucinogeno Aborigen, Banisteriopsis Caapi.Academia Colombiana De Ciencias Exactas, Fisicas Y Naturales. —. 1972. “The Cultural Context of an Aboriginal Hallucinogen: Banisteriopsis Caapi,” in Flesh of the Gods: The Ritual Use of Hallucinogens. Edited by P. Furst, pp. 84-113. London: Allen and Unwin. —. 1978. Beyond the Milky Way: Hallucinatory Imagery of the Tukano Indians. Los Angeles: UCLA Latin America Centre. —. 1978. “Drug Induced Optical Sensations and Their Relationship to Applied Art among Some Colombian Indians,” in Art in Society. Edited by M. G. a. V.Megaw, pp. 289-304. London: Duckworth. Rindos, D. 1984. The Origins of Agriculture: An Evolutionary Perspective. Orlando: Academic Press. Ross, R. 1976. Adam Kok’s Griquas: A Study in the Development of Stratification in South Africa. Cambridge: Cambridge University Press. Rowlands, M. 1994. “The Politics of Identity in Archaeology,” in Social Construction of the Past: Representation as Power. Edited by G. C. Bond and A. Gilliam, pp. 129-143. London: Routledge. —. 1998. “The Archaeology of Colonialism,” in Social Transformations in Archaeology: Global and Political Perspectives. Edited by K. Kristiansen and M. Rowlands, pp. 327-333. London: Routledge. Ruby, J. Editor. 1993. The Cinema of John Marshall. Reading: Harwood Academic Publishers. Rudner, I. 1982. Khoisan Pigments and Paints and Their Relationship to Rock Paintings. Annals of the South African Museum 87:1-281. Rudner, I., and J. Rudner. 1959. Who Were the Artists? South African Archaeological Bulletin 14:106-108. Rudner, J. 1971. Painted Burial Stones from the Cape. Supplement to the South African Journal of Science Special Issue 2:54-61. Russell, T. 2000. The Application of the Harris Matrix to San Rock Art at Main Caves North, KwaZulu-Natal. South African Archaeological Bulletin 55:60-70. Said, E. W. 1993. Culture and Imperialism. New York: Knopf. Saunders, C. 1977. Madolo: A Bushman Life. African Studies 36:145-154. Schapera, I. 1930. The Khoisan Peoples of South Africa: Bushmen and Hottentots. London: Routledge & Kegan Paul. Schrire, C. 1980. An Enquiry into the Evolutionary Status and Apparent Identity of San Hunter-Gatherers. Human Ecology 8:9-32. —. 1992. The Archaeological Identity of Hunters and Herders at the Cape over the Last 2000 Years: A Critique. South African Archaeological Bulletin 47:62-64. —. 1992. The Archaeological Identity of Hunters and Herders at the Cape over the Last 2000 Years: A Critique. South African Archaeological Bulletin 47:62-64. Scully, W. C. 1913. Further Reminiscences of a South African Pioneer. London: Fisher Unwin. Sealy, J. C., and R. Yates. 1994. The Chronology of the Introduction of Pastoralism to the Cape, South Africa. Antiquity 68:58-67. Shanks, M., and C. Tilley. 1982. “Ideology, Symbolic Power and Ritual Communication: A Reinterpretation of Neolithic Mortuary Practices,” in The Archaeology of Contextual Meanings. Edited by I. Hodder, pp. 129-154. Cambridge: Cambridge University Press. —. 1987. Social Theory and Archaeology. Cambridge: Polity Press. —. 1992. Re-Constructing Archaeology: Theory and Practice, Second edition. London: Routledge. Shepard, J. 1976. In the Shadow of the Drakensberg: The Story of East Griqualand and Its People. Durban: T.W. Griggs & Co. Shilling, C. 2001. “The Embodied Foundations of Social Theory,” in The Handbook of Social Theory. Edited by G. Ritzer and B. Smart, pp. 439-457. London: Sage. 193 Shostak, M. 1981. Nisa: The Life and Words of A!Kung Woman. London: Allen Lane. Siegel, R. K. 1977. Hallucinations. Scientific American 237:132-140. Siegel, R. K., and M. E. Jarvik. 1975. “Drug-Induced Hallucinations in Animals and Man,” in Hallucinations: Behaviour, Experience, and Theory. Edited by R. K. S. a. L.J.West, pp. 81-161. New York: Wiley. Siegel, R. K., and L. J. West. 1975. Hallucinations: Behavior, Experience, and Theory. A Wiley Biomedical Publication. New York: Wiley. Silberbauer, G. B. 1965. Bushman Survey. Gaberones: Bechuanaland Government. —. 1981. Hunter and Habitat in the Central Kalahari Desert. Cambridge: Cambridge University Press. Skotnes, P. 1994. “The Visual as a Site of Meaning: San Parietal Painting and the Experience of Modern Art,” in Contested Images: Diversity in Southern African Rock Art Research. Edited by T. A. Dowson and J. D. Lewis-Williams, pp. 315-330. Johannesburg: Witwatersrand University Press. Smith, A. B. 1994. “Metaphors of Space: Rock Art and Territoriality in Southern Africa,” in Contested Images: Diversity in Southern African Rock Art Research. Edited by T. A. Dowson and J. D. LewisWilliams, pp. 373-384. Johannesburg: Witwatersrand University Press. Smith, B. W. 1995. Rock Art in South-Central Africa. A Study Based on the Pictographs of Dedza District, Malawi and Kasama District, Zambia, University of Cambridge. —. 1997. Zambia’s Ancient Rock Art: The Paintings of Kasama. Livingstone: National Heritage Conservation Commission. —. 1998. “Rock Painting: South Africa’s First Protest Art?” in Business Day, pp. Cover Page. Johannesburg. Smith, B. W., and G. Blundell. 2004. “Dangerous Ground: A Critique of Landscape in Rock-Art Studies,” in The Figured Landscapes of Rock-Art: Looking at Pictures in Place. Edited by C. Chippindale and G. Nash, pp. 239-262. Cambridge: Cambridge University Press. Smith, B. W., C. Chippindale, and P. S. C. Taçon. 2000. Visions of Dynamic Power: Archaic RockPaintings, Altered States of Consciousness and Clever Men in Western Arnhem Land (Nt), Australia. Cambridge Archaeological Journal 10:63101. Smith, B. W., and S. Ouzman. 2004. Taking Stock: Identifying Khoekhoen Herder Rock Art in South- 194 ern Africa. Current Anthropology 45:Forthcoming. Smith, B. W., and J. A. van Schalkwyk. 2002. The White Camel of the Makgabeng. Journal of African History 43:235-254. Smits, L. 1998. “Pers.Comm. On Aral Survey and Sdfs in Lesotho.” Ellecom. Soga, J. H. 1930. The South-Eastern Bantu. Vol. Special number of Bantu Studies. Supplement No 4. Johannesburg: Witwatersrand University Press. Solomon, A. 1992. Gender, Representation and Power in San Ethnography and Rock Art. Journal of Anthropological Archaeology 11:291-329. —. 1994. “’Mythic Women’: A Study in Variability in San Art,” in Contested Images: Diversity in Southern African Rock Art Research. Edited by T. A. Dowson and J. D. Lewis-Williams, pp. 331-371. Johannesburg: Witwatersrand University Press. —. 1997. The Myth of Ritual Origins? Ethnography, Mythology and Interpretation of San Rock Art. South African Archaeological Bulletin 52:3-13. —. 1999. Meaning, Models and Minds: A Reply to Lewis-Williams. South African Archaeological Bulletin 54:51-60. —. 2000. “Ethnography and Method in Southern African Rock Art Research,” in The Archaeology of Rock Art. Edited by C. Chippindale and P. S. C. Taçon, pp. 268-284. Cambridge: Cambridge University Press. —. 2000. On Different Approaches to San Rock Art. South African Archaeological Bulletin 55:77-78. Solway, J. S., and R. B. Lee. 1990. Foragers, Genuine or Spurious? Situating the Kalahari San in History. Current Anthropology 31:109-146. Spivak, G. 1987. In Other Worlds: Essays in Cultural Politics. London: Routledge. Spivak, G. C. 1988. “Can the Subaltern Speak?” in Marxism and the Interpretation of Culture. Edited by C. Nelson and L. Gossberg, pp. 271-313. London: Macmillan. —. 1996. “Subaltern Studies: Deconstructing Historiography,” in The Spivak Reader:Selected Works of Gayatri Chakravorty Spivak. Edited by D. Landry and G. Maclean. New York: Routledge. Stanford, W. E. 1910. Statement of Silayi, with Reference to His Life among the Bushmen. Transactions of the Royal Society of South Africa 1:435-440. —. n.d. “Statement of Mgudhlwa,” in The Sir WEM Stanford Papers. Cape Town. Stapleton, T. J. 2001. Faku: Rulership and Colonialism in the Mpondo Kingdom (C. 1780-1867). Waterloo: Wilfred Laurier University Press. Steedman, A. 1835. Wanderings and Adventures in the Interior of Southern Africa. Vol. 2. London: Longman. Stevenson, J. 1995. Title Man-the-Shaman: Is It the Whole Story? A Feminist Perspective on the San Rock Art of Southern Africa. M.A., University of the Witwatersrand. Stow, G. W. 1905. The Native Races of South Africa. London: Swan Sonnenschein. Survival International. 30th October, 2002. “Iman Turns into a ‘Bushman’: De Beers’ New London Store Targeted.” Suzman, J. 2002. Kalahari Conundrums: Relocation, Resistance and International Support in the Central Kalahari. Before Farming:1-10. —. 2003. Response to Stephen Corry. Before Farming:410. Taçon, P. S. C. 1991. The Power of Stone: Symbolic Aspects of Stone Tool Use and Tool Development in Western Arnhem Land, Australia. Antiquity 65:192-207. Thackeray, A. I., J. F. Thackeray, P. J. Beaumont, and J. C. Vogel. 1981. Dated Rock Engravings from Wonderwerk Cave, South Africa. Science 214:6467. Thackeray, J. F. 1984. Masquerade and Therianthropes in Art. The Digging Stick 2:2. —. 1988. Southern African Rock Art and Xhosa Beliefs Associated with Abantubomlambo. Pictogram 1:23. —. 1990. On Concepts Expressed in Southern African Rock Art. Antiquity 64:139-144. —. 1993. New Directions in the Study of Southern African Rock Art. African Arts 26:74-75. Thornton, R. J. 1983. “This Dying out Race”: W.H.I Bleek’s Approach to the Languages of Southern Africa. Social Dynamics 9:11-28. Thorpe, I. J. 1996. The Origins of Agriculture in Europe. London: Routledge. Thunberg, C. P. n.d. Travels in Europe, Africa, and Asia: Made between the Years 1770 and 1779. London: W. Richardson. Tomaselli, K. G. 1992. “Myth, Racism and Opportunism: Film and Tv Representations of the San,” in Film as Ethnography. Edited by P. I. Crawford and D. Turton. Manchester: Manchester University Press. —. 1993. “The Post-Apartheid Era: The San as Bridge between Past and Future,” in Eyes across the Water Ii. Edited by R. M. Boonzajer Flaes and D. Harper, pp. 81-89. Amsterdam: Het Spinhuis. Tomaselli, K. G., T. Gabriel, N. Masilela, and A. Williams. 1992. People of the Great Sandface. Visual Anthropology 5:153-166. Traill, A. 1978. “The Languages of the Bushmen,” in The Bushmen: San Hunters and Herders of Southern Africa. Edited by P. V. Tobias, pp. 137-147. Cape Town and Pretoria: Human and Rousseau. —. 1980. “Phonetic Diversity in the Khoisan Languages,” in Bushmen and Hottentot Linguistic Studies, 1979. Edited by J. W. Snyman, pp. 167-189. Pretoria: University of South Africa. —. 1986. Do the Khoi Have a Place in the San? New Data on Khoisan Linguistic Relationships. Sprache und geschichte in Afrika 7:407-430. —. 1995. “The Khoesan Languages of South Africa,” in Language and Social History: Studies in South African Sociolinguisitics. Edited by R. Mesthrie. Cape Town: David Philip. Turner, B. S. 1996. The Body and Society: Explorations in Social Theory, 2nd edition. London: Sage. —. 2003. “Foreword: The Phenomenology of Lived Experience,” in Embodied Lives: Figuring Ancient Maya and Egyptian Experience. Edited by L. M. Meskell and R. A. Joyce. New York: Routledge. —. 2003. Social Fluids: Metaphors and Meanings of Society. Body & Society 9:1-10. Turner, V. W. 1966. The Syntax of Symbolism in an African Religion. Philosophical Transactions of the Royal Society 251:295-303. —. 1967. The Forest of Symbols: Aspects of Ndembu Ritual. New York: Cornell. Valiente-Noailles, C. 1993. The Kua: Life and Soul of the Central Kalahari Bushmen. Rotterdam: A.A. Balkema. Van Rijssen, W. J. 1984. Southwestern Cape Rock ArtWho Painted What? South African Archaeological Bulletin 39:125-129. —. 1994. “The Question of Authorship,” in Contested Images: Diversity in Southern African Rock Art Research. Edited by T. A. Dowson and J. D. LewisWilliams. Johannesburg: University of the Witwatersrand Press. Vergnani, L. 1999. “In South Africa, Researchers Encounter Spirit Paintings by a Lost People,” in Chronicle of Higher Education. Vinnicombe, P. 1967. The Recording of Rock Paintings-an Interim Report. South African Journal of Science 63:282-284. —. 1967. Rock Painting Analysis. South African Archaeological Bulletin 22:129-141. —. 1972. Motivation in African Rock Art. Antiquity 46:124-133. —. 1972. Myth, Motive and Selection in Southern African Rock Art. Africa 42:192-204. —. 1976. People of the Eland: Rock Paintings of the Drakensberg Bushmen as a Reflection of Their Life and 195 Thought. Pietermaritzburg: Natal University Press. —. 2000. “Pers. Comm. On Observation of Sdfs in Southern Drakensberg Mountains.” Johannesburg. Voss, A. E. 1987. The Image of the Bushman in South African English Writing of the Nineteenth and Twentieth Centuries. English in Africa 14:21-40. Wadley, L. 1996. “The Bleek and Lloyd Records of Death and Burial, and the Problems That These Present for Archaeologists,” in Voices from the Past: /Xam Bushmen and the Bleek and Lloyd Collection. Edited by J. Deacon and T. A. Dowson, pp. 271-286. Johannesburg: Witwatersrand University Press. —. 1997. Our Gendered Past: Archaeological Studies of Gender in Southern Africa. Johannesburg: Witwatersrand University Press. —. 2001. Who Lived in Mauermanshoek Shelter, Korannaberg, South Africa. African Archaeological Review 18:153-179. Walker, N. 1996. The Painted Hills: Rock Art of the Matopos, Zimbabwe - a Guidebook. Exploring Zimbabwe. Gweru: Mambo. Weiss, P. 1995. Kandinsky and Old Russia: The Artist as Ethnographer and Shaman. New Haven: Yale University Press. Wendt, W. E. 1976. ‘Art Mobilier’ from the Apollo 11 Cave, South West Africa: Africa’s Oldest Dated Works of Art. South African Archaeological Bulletin 31:5-11. Whitley, D. S. 1987. Socio-Religious Context and Rock Art in East-Central California. Journal of Anthropological Archaeology 6:159-188. —. 1992. Shamanism and Rock Art in Far Western North America. Cambridge Archaeological Journal 2:89-113. —. 1994. “Shamanism, Natural Modeling and the Rock Art of Far Western North American HunterGatherers,” in Shamanism and Rock Art in North America. Edited by S. A. Turpin, pp. 1-44. San Antonio: Rock Art Foundation. —. 1998. “Finding Rain in the Desert: Landscape, Gender and Far Western North American RockArt,” in The Archaeology of Rock-Art. Edited by C. Chippindale and P. S. C. Taçon, pp. 11-29. Cambridge: Cambridge University Press. —. 2000. The Art of the Shaman: Rock Art of California. Salt Lake City: The University of Utah Press. Whitley, D. S., R. I. Dorn, J. M. Simon, R. Rechtman, and T. K. Whitley. 1999. Sally’s Rockshelter and the Archaeology of the Vision Quest. Cambridge Archaeological Journal 9:221-247. 196 Willcox, A. R. 1959. Hand Imprints in Rock Paintings. South African Journal of Science 55:292-298. —. 1960. Who Were the Artists? Another Opinion. South African Archaeological Bulletin 15:23-25. —. 1984. The Rock Art of Africa. Johannesburg: Macmillan. Williams, S. J. 1998. Bodily Dys-Order: Desire, Excess and the Transgression of Corporeal Boundaries. Body & Society 4:59-82. Williamson, B. S. 2000. Prehistoric Stone Tool Residue Analysis from Rose Cottage Cave and Other Southern African Sites. Ph.D, University of the Witwatersrand. Wilmsen, E. N. 1989. Land Filled with Flies: A Political Economy of the Kalahari. Chicago and London: The Chicago University Press. Wilmsen, E. N. (1996). Decolonising the Mind: Steps Towards Cleansing the Bushman Stain from Southern African History. Miscast: Negotiating the Presence of the Bushmen. P. Skotnes (ed.). Cape Town, University of Cape Town Press: 185189. Wilmsen, E. N., and J. R. Denbow. 1990. Paradigmatic History of San-Speaking Peoples and Current Attempts at Revision. Current Anthropology 31:489-524. Winkelman, M. 1989. A Cross-Cultural Study of Shamanistic Healers. Journal of Psychoactive Drugs 21:17-24. Woodhouse, H. C. 1992. The Rain and Its Creatures: As the Bushmen Painted Them. Rivonia: William Waterman. Wright, J. B. 1971. Bushman Raiders of the Drakensberg 1840-1870. Pietermaritzburg: University of Natal Press. Wright, J. B., and A. D. Mazel. 1987. Bastions of Ideology: The Depiction of Precolonial History in the Museums of Natal and KwaZulu. Southern African Museums Association Bulletin 17:301-310. Yates, R., J. Golson, and M. Hall. 1985. Trance Performance: The Rock Art of Boontjieskloof and Sevilla. South African Archaeological Bulletin 40:7080. Yates, R., J. E. Parkington, and A. H. Manhire. 1990. Pictures from the Past: A History of the Interpretation of Rock Paintings and Engravings of Southern Africa. Pietermaritzburg: Centaur. Yates, T. 1993. “Frameworks for an Archaeology of the Body,” in Interpretative Archaeology. Edited by C. Tilley. Oxford: Berg. Z. 1850. “African Discovery-Treatment of the Natives,” in The Natal Witness. Pietermaritzburg. 197 Ngqika Unknown Father Mkahlila Nqabayo Unknown Daughter of Nqabayo Silayi Ciyo Membership of Nqabayo’s Group c. 1850 Nkwinti Tyazo Jan Hans Ndaralu (Married to Silayi) Mamxabela Unknown Father Appendix 1. Diagram of known members of Nqabayo’s Group about 1850. Links between the second and third tier indicate known parentchild relationships. Site Eldritch Danc./Cl. Run-Run Line 1-2-3 Arr. Feline Rain/A Type RSA ANG 1 None No Yes No No No No No No RSA ANG 2 Ind. Ind. Yes Yes Ind. Ind. Ind. Ind. No RSA ANG 3 None No Yes No Yes No No No No RSA BAE 1 Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA BAE 2 1 SDF Yes Yes Ind. Ind. Ind. Yes Ind. No RSA BAE 3 None Yes Ind. Yes Ind. Ind. Ind. Ind. No RSA BAE 4 Ind. Ind. Possible Ind. Ind. No Ind. Ind. No RSA BAE 5 None No Ind. Ind. Ind. No Ind. Ind. No RSA BAE 6 None No No No No No No No No RSA BAE 7 None No No No No No No No No RSA BAE 8 None No Yes No No No No No No RSA BAE 9 None No No No No No No No No RSA BAE 10 None No Yes No No No No No No RSA LAB 1 2 LH-SDFs Yes Yes Yes Yes Yes Yes Yes Yes RSA LAB 2 None No No No No No No No Yes RSA LAB 3 1 LH-SDF Yes Yes Ind. Yes Yes Ind. Yes Yes RSA LAB 4 None No Ind. Ind. Ind. No Ind. Ind. No RSA LAB 5 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA LAB 6 None Yes Yes Yes No No No No No RSA LAB 7 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA LAB 8 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA LAB 9 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA LAB 10 None Yes Yes No Yes No No No No RSA LAB 11 None Ind. Yes Ind. Ind. No Ind. Ind. Yes RSA LAB 12 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. No RSA CHA 1 3 SDFs Yes Yes Yes Yes Yes Yes Yes No RSA CHA 2 None No No No No No No No No RSA CHA 3 None No No No No No No No No RSA CHA 4 None No No No No No No No No RSA BOU 1 None Ind. Ind. No No No No No Poster RSA BRC 1 Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. 198 SDF/LH-SDF 2 A P P E N D I X 2 P.1 RSA CRA 1 None No No No No No No No No RSA CRA 2 None No Yes Ind. No Ind. Yes Ind. No RSA CRA 3 None No No No No No No No No RSA CRA 4 None No No No No No No No No RSA CRA 5 None No No No No No No No No RSA CRA 6 None No No No No No No No Lt. White RSA DEE 1 Ind. Yes Yes Ind. Ind. Ind. Ind. Ind. No RSA ESP 1 Ind. Ind. Yes Ind. Yes Ind. Ind. Ind. No RSA ESP 2 Ind. Ind. Yes Ind. Ind. Ind. Ind. Ind. No RSA ESP 3 None No No No No No No No Yes RSA FIR 1 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA FEN 1 None No Yes No No No No No No RSA FEN 2 Ind. No Ind. Ind. No No No Ind. No RSA FEN 3 Ind. Yes Yes Yes Ind. Ind. Ind. Ind. No RSA FST 1 None No Yes No No No No No No RSA FST 2 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. No RSA GAB 1 None No Yes No No No No No No RSA GLN 1 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA GTT 1 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA GTT 2 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA GTT 3 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA GTT 4 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA GRV 1 None No Yes No No No No No No RSA GRV 2 None Ind. Yes Ind. Ind. Ind. Ind. Ind. Possible/excluded RSA HIL 1 Ind. Ind. Ind. Yes Yes Ind. Ind. Ind. Ind. RSA HIL 2 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA HIL 3 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA HIL 4 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA MCO 1 None No Yes No No No No No No RSA MCO 2 None No No No No No No No No RSA MPN 1 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. No A P P E N D I X 2 P.2 199 RSA MEL 1 Ind. Yes Yes Ind. Ind. Ind. Ind. Ind. No RSA MEL 2 Ind. Ind. Yes Yes Ind. Ind. Ind. Ind. No RSA MEL 3 None No Yes No No No No No No RSA MEL 4 None No No No No No No No No RSA MEL 5 None No No No No No No No No RSA MEL 6 1 LH-SDF Yes Yes Ind. Ind. No Ind. Ind. No RSA MEL 7 Ind. Ind. Yes Ind. Ind. Ind. Ind. Ind. No RSA MEL 10 None No Yes No No No No No No RSA MEL 8 1 LH-SDF Yes Yes Yes Ind. No No No No RSA MEL 9 1 LH-SDF Yes Yes Yes Yes Yes Yes Yes Yes RSA MIE 1 1 LH-SDF No Yes Ind. Ind. Ind. Ind. Ind. Yes RSA MIE 2 None No No No No No No No Yes RSA MNT 1 None No No No No No No No No RSA MOZ 1 None Yes Yes No No No No No No RSA MOI 1 None No Ind. Yes No No No No No RSA MOI 2 None No No No No No No No No RSA RIR 1 Ind. Yes Yes Ind. Ind. Ind. Ind. Ind. No RSA ROY 1 Ind. No Yes No No No No No No RSA ROY 2 None No Yes No No No No No No RSA RON 1 1 SDF Ind. Yes Ind. Ind. Ind. Ind. Ind. No RSA SHA 1 None No Yes No No No No No No RSA SHA 2 None No Yes No No No No No No RSA SHA 3 None No No No No No No No No RSA SHA 4 None No No No No No No No No RSA TAL 1 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. No RSA TER 1 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. No RSA TER 2 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. No RSA TER 3 None No No No No No No No No RSA WEL 1 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. No RSA WEL 2 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. No RSA WEL 3 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. No RSA WEL 4 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. No 200 A P P E N D I X 2 P.3 RSA WEL 5 None No Yes Yes No No No No No RSA WEL 6 None No No No No No No No No RSA WEL 7 None No Yes No No No No No No RSA WLL 1 None No Yes No No No No No No RSA WLL 2 Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA WLL 3 None No No Yes No No No No No RSA WLL 4 None No No No No No No No No RSA WID 1 Ind. Ind. Ind. Ind. Yes Ind. Ind. Ind. No RSA WID 2 1 SDF Ind. Yes Ind. Yes Ind. Ind. Yes Yes RSA WID 3 None No Yes No No No No No No RSA WID 4 None Yes No No No No No No No RSA WOO 1 Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. No RSA NCE 1 1 SDF/1 LH-SDF Yes Yes Ind. Ind. Ind. Ind. Ind. No RSA POE 1 None No No No No No No No No RSA POE 2 None No No No No No No No No RSA PRH 1 None Ind. Yes No No No No No No RSA PRH 2 None No No No No No No No No RSA ZAM 1 Ind. Yes Ind. Ind. Ind. Ind. Ind. Ind. No RSA ZAM 1 None No No No No No No No No RSA BEN 1 None Yes Yes Ind. Ind. Ind. Ind. Ind. Ind. RSA BOK 1 None Yes Yes Yes No No No No No RSA LAC 1 None No Yes No Yes No No No No RSA PRS 1 None No Yes No No No No No No RSA MOX 1 None Ind. Yes No No No No No No RSA ROG 1 Ind. Yes Yes No No No No No No RSA OXC 1 Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA FRH 1 Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA SPR 1 None No Ind. No No No No No No RSA TYN 1 None No No No No No No No No RSA TYN 2 1 SDF Yes Yes Yes Ind. Ind. Ind. Ind. Ind. RSA TYN 3 Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. A P P E N D I X 2 P.4 201 RSA TYN 4 None No No No No No No No No RSA TYN 5 None No No No No No No No No RSA ULV 1 None Yes Yes Yes No No No No No RSA ULV 2 None No No No No No Yes No No RSA ULV 3 None Ind. Ind. Ind. Ind. Ind. Ind. Ind. Ind. RSA ULV 4 None No No No No No No No No RSA PIT 1 Ind. Yes Yes Yes Ind. Ind. Ind. Ind. Ind. RSA CHL 1 None Yes No No No No No No No A P P E N D I X 2 P.5 Appendix 2 (Pages 198-202). List of 134 sites in the Core Nomansland Study Area and table of presence and absence of elements of the pattern at the sites. Ind. refers to indeterminate. Elements may be indeterminate because of poor preservation or because the density of images at the sites does not allow for adequate identification until the site has been completely traced. Appendix 3 (Pages 203-204). Table of pigment samples. 0= No haemoglobin/myoglobin present. Positive results are indicated by +: the greater the number, the greater the presence of haemoglobin/myoglobin. DNT=Did Not Test. Adapted from Williamson, B. 2000: 119-120. 202 No. Site 1 RARI-RSAMEL6 Check 0 After EDTA — Off White DNT Brown-red +1 DNT +4 trace DNT +1 — DNT — Kit on back Lower thigh Lower thigh Shin Upper thigh Upper thigh Behind hump Brown-Red Brown-Red Brown-Red Brown-Red Brown-Red Brown-Red Dark Red +1 0 trace 0 0 trace trace — — — — — — — — — — 0 — trace Eland Eland Eland Hindquarter Mid-section On back, before hindquarter Dark Red Dark Red Dark Red 0 trace +2 — — +1 — — — 1a Type 2: Eland Dewlap Powder/Red trace — — 1b Type 2: Eland Lower body, before hindquarter Powder/Red 0/trace — — 17 2a Type 2: Eland Dewlap Powder/Red trace — — 18 19 20 21 2b 3a 3b 4a Type 2: Eland Type 2: Eland Type 2: Eland Type 2: Eland DNT Powder/Red Powder/Red Powder/Red DNT trace 0 trace DNT — — — DNT — — — 22 23 24 25 26 27 28 29 4b 5a 5b 6a 6b 6c 6d 7a Type 2: Eland Type 2: Eland Type 2: Eland SDF (major) SDF (major) SDF (major) SDF (major) Eldritch Image DNT Behind neck Midsection On back in midsection DNT Dewlap Hind leg Face-chin Back of head Bow string Head-dress Back of head DNT +1 DNT +3 0 DNT trace DNT +3 — DNT — DNT — — 0 0 — — — — 30 7b Eldritch Image Chin 0 — — 31 7c Eldritch Image Nasal blood DNT Powder/Red Powder/Red Off White Black Off White Red Translucent/White Translucent/White Red 0 — — 7d 1a Eldritch Image SDF (minor 1) DNT Face DNT Red DNT 0 DNT — DNT — 1b 1c SDF (minor 1) SDF (minor 1) Red Red 0 DNT — DNT — DNT 2a 2b 2c 3a 3b 3c 4a 5a 6a 7a 8a 1a SDF (major) SDF (major) SDF (major) SDF (minor 2) SDF (minor 2) SDF (minor 2) Feline Eland Eland Eland Eland Thin Red Line Headdress Lower abdomen, seated Headdress Face-chin Lower abdomen Oral emanation Face Upper thigh Forequarter Front leg Head Front leg DNT Section Red White Dark red Red White Dark red Light red White Red White DNT Red 0 DNT 0 0 0 0 0 +5 0 +4 DNT DNT — DNT — — — — — +5 — +4 DNT DNT — DNT — — T/0 — 0 — — 2a Thin Red Line Section Red 0 — — 2 3 4 5 6 7 8 9 10 11 RARI-RSAMEL8 12 13 14 15 16 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 RARI-RSAMEL9 Storm Shelter RARI-RSALAB3 Samp/ No. 1a Image Position of sample Colour Hemastix SDF (major) Head Orange-Red 1b 1c 1a SDF (major) SDF (major) SDF (major) Hand DNT Shoulder 1b 1c 2a 2b 3a 3b 4a SDF (major) SDF (major) Anthropomorph Anthropomorph Anthropomorph Anthropomorph Eland 4b 5a 5b — DNT DNT 203 49 2b Thin Red Line Dot White 0 — — 3a 4a 5a 6a 6b Eldritch Image Type 2: Eland Type 2: Eland Type 2: Eland Type 2: Eland Head Hind leg DNT Hindquarter Dewlap 0 trace/0 DNT trace +3 — — DNT — +2 — — DNT — — 55 56 7a 8a Type 2: Eland Type 2: Eland Hindquarter Front/shoulder 0 trace/0 — — — — 57 9a Type 2: Eland Back +2 +1 — 58 10a Type 2: Eland Hindquarter White Powder/Red DNT Powder/Red Powder/ White Powder/Red Powder/ White Powder white Powder/Red +2 trace 1a Type 2: Eland Upper Hindquarter +4 +4 — 1b 2a 1a Type 2: Eland Type 2: Eland Thin Red Line Stomach Patch Upper foreleg Dot Powder/ Yellow Powder/Red Powder/Red White 0 trace/0 0 — — — 0 trace/0 — 1b 2a 1a Thin Red Line Thin Red Line Rain-Animal Section Section Dot (1) Red Red Red +3 0 0 +1 — +3 — 1b 2a Rain-Animal Anthropomorph Dot (2) Stomach DNT +1 DNT +1 DNT — 2b 2c Anthropomorph Anthropomorph DNT Head DNT 0 DNT 0 DNT — 3a 1a Anthropomorph SDF (major) Thigh Thigh Red White& Red Tinge DNT White & Red Tinge Red White 0 trace — — trace 1b 1a SDF (major) Type 2: Eland Hand area Hindquarter White White 0 +4 — +3 — — 2a 3a 3b 4a 4b 5a 6a 7a Type 2: Eland Type 2: Eland Type 2: Eland Type 2: Eland Type 2: Eland Type 2: Eland Type 2: Eland Type 2: Eland Back Fore Section Rump Hump Hindquarter Hump Hump Hump White White Red White Red Red Red Red +4 +4 0 +4 0 0 0 0 +4 +4 — +4 — — — — — — — — — — — — 50 51 52 53 54 59 60 61 62 63 64 65 66 67 RARI-RSALAB3 RARI-RSAESP3 RARI-RSAWID1 RARI-RSAWID2 68 69 70 71 72 73 74 75 76 77 78 79 80 81 204 RARI-RSAMIE1 RARI-RSAMIE2