the ELA ND’s PEOPLE

New Perspectives in the Rock Art of the Maloti-Drakensberg Bushmen

Essays in Memory of Patricia Vinnicombe
Edited by Peter Mitchell and Benjamin Smith
 Wits University Press
1 Jan Smuts Avenue
Johannesburg
South Africa
http://witspress.wits.ac.za

Entire publication © Wits University Press 2009

Introduction and chapters © Individual authors 2009

First published 2009

This is the second volume in the Rock Art
Research Institute monograph series

ISBN 978-1-86814-498-3

All rights reserved. No part of this publication may be

reproduced, stored in a retrieval system, or transmitted in any
form or by any means, electronic, mechanical, photocopying,
recording or otherwise, without the express permission, in
writing, of both the copyright holders and the publisher.

Edited by Lee Smith

Cover design by Quba Design and Motion
Layout and design by Quba Design and Motion
Printed and bound by Ultra Litho (Pty) Limited
60 Chapter 5
 CHAPTER FIVE

Meaning then, meaning now:

Changes in the interpretative process
in San rock art studies

David Pearce, Catherine Namono and Lara Mallen

M
eaning is what it is all about today in the study of have never been replicated, and his tracings in the Brandberg
San rock art. A few recent studies consider dating (now the Dâures massif), Namibia (Pager 1989), set the
(e.g. Mazel & Watchman 1997, 2003), sequence standard for professional recording of rock art in southern
(e.g. Russell 2000; Pearce 2002, 2006; Swart 2004), chemical Africa. In addition to his graphical recording of rock art, he is
analysis (e.g. Hughes & Solomon 2000), quantification (e.g. also well known for his numerical recording: if his programme
Lenssen-Erz 2001) and conservation (e.g. Hœrle & Salomon of quantification was not the most extensive, it was certainly
2004; Hœrle 2005, 2006; Deacon 2006; Hall et al. 2007; published in the greatest detail (Pager 1971).
Meiklejohn et al. 2009). But these studies are in the minority Lewis-Williams’s contribution is also readily apparent.
and outside the mainstream of San rock art research. It is His use of ethnography, not in a general way, but in the
difficult for researchers at the beginning of the twenty-first explanation of details of ritual and belief (Lewis-Williams
century to imagine a time when it could have been different, 1981), is largely credited with our current understanding of
when meaning was not king; it seems inconceivable that a time San rock art. His introduction of neuropsychological insights
existed when we did not know something of what the art was is also well known (Lewis-Williams & Dowson 1988; Lewis-
about. Williams 2001a).
That was exactly the case, though, in the 1960s and 1970s Vinnicombe’s contribution, on the other hand, is less
when Patricia Vinnicombe was working on her book People clearly defined: she pioneered the quantitative approach in
of the Eland (Vinnicombe 1976). Other studies at the time southern African rock art studies; she made extensive use of
consisted mainly of guesswork by amateurs and quantification San ethnographies; she tried to link historical events to rock
by a new generation of professional researchers, of whom art; and her book, People of the Eland, is certainly well illustrated
Vinnicombe (1967a) herself was a pioneer. Vinnicombe with many excellent copies. Broadly speaking, however,
was faced with the fundamental problem of having to say Vinnicombe chose three interrelated ways of approaching
something about the art, of finding some sort of order to the the art: quantification, a study of San ethnography and a
paintings. In retrospect, the way of doing this may be obvious; consideration of regional history. She therefore collected
 Figure 5.1
it was not at the time. quantitative data on certain categories of images and then
Patricia Vinnicombe, David Lewis-
She was not alone in her quest. Together with Harald Pager sought information on these themes from San ethnographies
Williams and Emmanuel Anati at the
and David Lewis-Williams, Patricia Vinnicombe is one of the and ethnohistorical documents. Her approach is clear in the
Valcamonica rock art symposium in
three names most often cited as having ushered in this ‘modern’ structure of her book: there is a chapter on eland, another
Italy in 1972. From RARI archive.
era of San rock art research in southern Africa. Pager (although on baboons and so forth. Whilst her approach was extremely
making some limited use of ethnography) is probably best innovative in the southern African context, it was a product
known for his recording of the art. His magnificent, full-colour, of methodological movements in the wider intellectual world
full-size copies of the Didima Gorge paintings (Pager 1971) (Figure 5.1).

The Eland's People 61

BOX 5. The presentation of bushman rock art in the uKhahlamba-Drakensberg

Ndukuyakhe Ndlovu

The uKhahlamba-Drakensberg Park1 was inscribed as the with a challenge to reposition their strategy, following states that no visitor is allowed to access any painted
fourth South African World Heritage Site in November not only national but also international regulations. shelter except in the company of a custodian. However,
2000 after meeting four selection criteria, two each Before its declaration, biodiversity conservation took as with any legislation, the effective implementation of
for natural and cultural significance. This semi-circular precedence over cultural resources management. As a this policy is a great challenge. In addition, the growing
border between the South African province of KwaZulu- result, the professional staff of the current management interest in rock art has encouraged entrepreneurs to
Natal and the inland mountain Kingdom of Lesotho also authority still does not include any culturally sensitive produce products themed on rock art, such as clothing,
boasts a high level of endemic and globally threatened and professionally qualified archaeologists or kitchen utensils, stationery and so on. This entrepreneurial
plant species. Besides nature, the uKhahlamba- anthropologists. activity has further broadened the popular awareness of
Drakensberg Park contains many rock shelters with This lack of professional capacity seriously hampers rock art and ensured that more and more people want to
the largest and most concentrated group of paintings the management and presentation of rock art in the see examples of the original art.
south of the Sahara made by hunter-gatherer people. World Heritage Site. Although rock art has attracted This has led to the development of a greater public
Discovered about 150 years ago by Europeans, rock art the interests of many people for aesthetic reasons and rock art infrastructure. To accommodate those with
paintings were initially considered primitive and crude. academic research over the years, its presentation to the either disabilities or time constraints, two rock art centres,
However, these paintings are today thought not only to public has been very limited. Prior to the declaration, the namely Kamberg (Figure B5.1) and Didima (Figure
be of outstanding quality but also to give voice to the only ones who knew the localities of a few of the rock B5.2) and an open-air museum at Main Caves, Giant’s
complex spiritual beliefs of the precolonial Bushman art paintings were those who hiked the steep valleys Castle, have been constructed. This has changed the
inhabitants of the region. of the uKhahlamba-Drakensberg Park, the few tourists presentation of rock art in the uKhahlamba-Drakensberg
The research of Patricia Vinnicombe, among with a rock art interest, as well as local communities considerably. Whereas in the past rock art sites could
others, was instrumental in bringing the rock art of the and researchers. The location of all but a handful of rock be visited by only a few, the existence of these rock art
uKhahlamba-Drakensberg to the fore within southern art sites was, and still is today, a closely guarded secret. centres has ensured that any person with an interest
African archaeology. The current presentation of the Other sites were intentionally omitted from maps, be in rock art can be guided to one or more of a group of
area as a treasure trove of rock art and as the ‘Rosetta they for hikers, road users or others. This restriction in carefully chosen rock art sites deemed appropriate for
Stone’ of rock art interpretation is directly attributable to access to site location details was presumed to be in the public visitation. At Kamberg Rock Art Centre, visitors
research by Vinnicombe and others. Had it not been for interests of the proactive management of rock art. This have an opportunity to watch an audio-visual show
this work, the uKhahlamba-Drakensberg Park might not long-standing low-key approach to presentation has and to take a guided tour to Game Pass Shelter. Didima
have been proclaimed a mixed World Heritage Site and changed considerably, and in a positive direction, since Rock Art Centre features a display hall, traditional fireside
therefore the presentation of its rock art to the public the award of World Heritage status. storytelling, and an auditorium where visitors can watch
would be very different from what it is today. Following The new status of the uKhahlamba-Drakensberg an inspiring audio-visual show. There is also an option for
the declaration of the uKhahlamba-Drakensberg Park Park has been significant in luring tourists to the region, visitors to access three rock art sites close to the centre.
as a World Heritage Site, South Africa’s president at the boosting the steady increase in tourist numbers since Besides tourism, Bushmen in KwaZulu-Natal are
time, Thabo Mbeki, approved the use of rock art2 in the the ending of apartheid in 1994. A number of tourism increasingly requesting access to rock art sites to use
country’s new national coat of arms (see Box 3, Figure destinations market the uKhahlamba-Drakensberg Park them for spiritual purposes. This is part of the living
B3.2), the highest visual symbol of the state. There could not only for its pristine environment, but also as a rock heritage still evident in the uKhahlamba-Drakensberg
be no better demonstration of the significance of rock art destination. This has increased the demand for visits Park. Kamberg hosts the annual Eland Ceremony,
art for understanding the history of this country (Smith to painted rock art sites. However, this growing interest bringing together Bushman descendants from across
et al. 2000). puts rock art sites under considerable human threat. the uKhahlamba-Drakensberg to celebrate their heritage
The declaration of the uKhahlamba-Drakensberg To manage rock art sites better, the relevant provincial and to share it with others.
Park as a World Heritage Site presented its managers authorities have established a custodian policy which

62 Chapter 5
 Notes

1. In Zulu, the word uKhahlamba means ‘Barrier of

Spears’. Dutch-speaking European settlers in the
1830s named the mountains Drakensberg, meaning
‘Mountain of Dragons’.
2. The human figure comes from the Linton Panel, a
famous panel of rock art now housed and displayed
in the South African Museum in Cape Town. In
1918, this panel was removed from the farm of
Linton in the Maclear District of the Eastern Cape
Province (Lewis-Williams 1988).

References

Lewis-Williams, J.D. 1988. The World of Man and the World

of Spirit: An Interpretation of the Linton Rock Paintings.
Cape Town: South African Museum.
Smith, B.W., Lewis-Williams, J.D., Blundell, G. &
Chippindale, C. 2000. Archaeology and symbolism
in the new South African coat of arms. Antiquity 74:
467–468.

 Figure B5.1
Kamberg Rock Art Centre in the
central Drakensberg. From RARI
archive.

 Figure B5.2
Ndukuyakhe Ndlovu (left) and
the Didima Rock Art Centre in
the northern Drakensberg. From
RARI archive.

The Eland's People 63

In this chapter we recount briefly the various areas of
research to which Vinnicombe contributed, before moving
on to a discussion of the ways in which each of these has
changed and grown in the ensuing years. We do this by
following three interpretative methods she employed:
quantification, ethnography and history. We do not
provide an exhaustive review of all work conducted in
these three fields, indeed, these are the subjects of other
chapters in this volume. Instead, we highlight studies that
were methodologically important in the development of
research. For this reason, we do not discuss every piece of
research conducted in the past 30 years, particularly not
those that have come and gone without making any major
impact on the general trajectory of study. We follow this
methodological discussion with a consideration of how we
see San rock art research changing in the future.

Quantification

In the 1960s, Vinnicombe embarked on an innovative

programme of numerical analysis of San rock art images
in the southern Maloti-Drakensberg region. The novelty
of her programme was twofold. First, she introduced
quantitative techniques to the study of San rock art. Using
these techniques, she believed, researchers would be able
to unravel the meaning of the art. Her second innovation
was the development of a detailed set of categories and
subcategories of art. She used these categories in her own
studies, but, significantly, they were also adopted, with
some modification, by other researchers (Pager 1971;
Lewis-Williams 1972, 1974). France. Coming as he did from an ethnographic background,
Vinnicombe’s introduction of a quantitative programme the species and sex of animals depicted in the caves seemed
was very much in line with broader movements in obvious candidates for themes to use in ordering the imagery.
archaeology at the time. The then New Archaeology sought In accepting the primacy of species and sex and their
to make archaeology more ‘scientific’ so as to make the supposed symbolic significances in constructing his categories,
discipline more ‘objective’ and increase perceptions of its Leroi-Gourhan was conforming to a prominent theme in
status in relation to the natural and biological sciences. The Western anthropological thought. Ethnographic work from
collection and statistical analysis of quantitative data was many parts of the world had shown that people use animals as
one of the ways this paradigm was manifested. Raising the social and other kinds of symbols. His narrow focus on species
status of rock art research was certainly an explicit goal of and sex, of course, masked many other potential ways of
quantitative studies; at the time, rock art studies in southern categorising the images. One could, for instance, define equally
Africa were largely the domain of interested amateurs. valid categories on the basis of size, colour or technique of
Similar trends were manifest in other parts of the animal depictions, amongst many other characteristics. It was
world. In Europe, André Leroi-Gourhan (1968) was nevertheless this influential study that led Vinnicombe towards
studying rock paintings in the Upper Palaeolithic caves of quantification and ethnographic study in southern Africa.

64 Chapter 5
 Figure 5.2 The quantitative study that Vinnicombe undertook in and handprints. Although his initial intention was to
Vinnicombe argued that eland the Maloti-Drakensberg Mountains analysed a total of obtain an objective assessment of the content of the art, he
served as a link between the 8 478 individual images (Vinnicombe 1967a, 1967b, 1976). conceded that in some instances it was impossible to avoid
material and spiritual worlds of the A number of characteristics were recorded for each figure. a certain element of subjectivity (Maggs 1967: 100–101).
San. Traced and redrawn by Justine For each human figure, for instance, 20 different categories Pager also began a large recording programme in the
Olofsson. of data were recorded. She suggested that her numerical KwaZulu-Natal Drakensberg. It consisted of copying
techniques, in addition to helping to understand the art paintings and later quantifying them. He recorded a total
that was analysed, would provide a means of comparing of 12 767 individual images from more than 17 sites in
art in different regions of southern Africa. This proposition the Didima Gorge (Pager 1971: 321, 323). Following this,
was soon to be explored as researchers began using similar he began an even more ambitious recording programme
techniques in other areas. in the Brandberg of Namibia. In 1963 the University of
At the same time as Vinnicombe, Timothy Maggs (1967), Cologne had begun a major research project that focused
working in the Western Cape Province, used numerical on empirical accuracy of documentation and analysis
analysis to assess the content of rock art in that area. He of the rock art of southern Africa (Lewis-Williams 1990:
quantified features of 1 471 images of human figures, animals 126). This project grew into what later became known as

The Eland's People 65

BOX 6. Visitor attractions in the uKhahlamba-Drakensberg: Main Caves, Kamberg and Didima
Aron D Mazel

Inspired largely by the insights generated by Patricia I visited these attractions in 2007 as part of a project paintings of half-human half-animal figures.” Nowhere
Vinnicombe and David Lewis-Williams, the rock investigating the presentation of San archaeology are the Didima visitors informed that these are different
paintings of the uKhahlamba-Drakensberg have played and history (including rock art) in the uKhahlamba- ways of interpreting the same figures. Leaving aside the
a significant role during the last three decades in Drakensberg, and how this has responded to the findings merits of these competing hypotheses, it is clear that the
stimulating new interpretations of rock art throughout of local archaeological research since the 1980s. This visit visitors to these three attractions receive an inconsistent
the world. Alongside this development we have seen built on previous trips to Main Caves (2001) and Kamberg message regarding the interpretation of the rock art.
the construction of a substantial archive of the region’s Rock Art Centre (2002). The research has concentrated A similar picture emerges regarding the approaches
abundant rock art, a vastly improved understanding on the fixed displays and the Kamberg audio-visual taken to the presentation of the excavated hunter-
of the hunter-gatherer past through excavations, and presentation. This short insert briefly investigates the gatherer past. At Main Caves this issue is dealt with
the direct dating of the paintings (see Wright & Mazel approaches taken in the interpretation of rock art and of in relation to the excavation of Main Caves South
2007). It is therefore unsurprising that 1) archaeology, the excavated past. It is worth noting that although the and San relations with Bantu-speaking farmers and
and particularly rock art, featured prominently in the Didima and Kamberg visitor attractions have been called European colonists during the last 2 000 years. In
uKhahlamba-Drakensberg becoming a World Heritage ‘Rock Art Centres’, both have a wider display remit and contrast, at Kamberg the excavated past is completely
Site in 2000, and 2) that the region has been at the also cover other aspects of San archaeology and history. ignored in both the audio-visual presentation and the
forefront in the public presentation of rock art and For the rock art, the three attractions approach accompanying booklet (Blundell 2002). Furthermore,
hunter-gatherer archaeology and history in southern its interpretation differently. Kamberg only presents on the back cover of the commercially available video
Africa. With regard to the latter issue, the last decade the shamanistic explanation, pioneered by David of the audio-visual presentation, purchased by the
has witnessed the upgrading of the Giant’s Castle Main Lewis-Williams (1981; Lewis-Williams & Pearce 2004), author at Kamberg in 2002, it is erroneously stated that,
Caves visitor attraction (1998) and the development of while Main Caves and Didima acknowledge that “With the Bushmen now gone from the area, all that
rock art centres at Kamberg (2002) and Cathedral Peak more than one interpretation is currently favoured by remains is their rock art and their stories as told by Zulu
(2003) (see Ndlovu, Box 5 in this volume, for additional archaeologists (i.e. the shamanistic explanation and the communities of the Drakensberg”, without reference
background information). At Main Caves, new on-site mythological approach developed by Ann Solomon to the fact that there is much excavated hunter-
text panels were developed along with a boardwalk to [1997]). At Main Caves the shamanistic explanation gatherer material in the region (Mazel 1989; Mitchell,
provide visitors with improved access to the paintings and the mythological approach are presented on the this volume). At Didima there is some, but still limited,
(Figure B6.1a, b); at the Kamberg Rock Art Centre visitors same panel as competing approaches, while at Didima reference to the excavated record. It includes, however,
are shown an audio-visual presentation and have the they are presented separately in adjacent cases with no a mock Middle Stone Age (MSA) excavation, which is
option to visit Game Pass Shelter 1 and to purchase reference evident of their being competing hypotheses. completely misplaced as no MSA excavations have
a booklet entitled The Unseen Landscape: A Journey to This lack of acknowledgement is further reflected in the ever taken place in the uKhahlamba-Drakensberg, as
Game Pass Shelter (Blundell 2002); and, at the Didima explanation of therianthropic figures. In the ‘Mythology’ opposed to the highlands of Lesotho on the other side
Rock Art Centre at Cathedral Peak, there are museum- showcase at Didima it is stated that, “It is believed by of the Escarpment (Carter 1978). Neither at Kamberg
type displays supported by an audio-visual presentation; some that paintings of therianthropes, made up of nor at Didima is there any mention of the rock shelters
visitors have the option to visit Lower Mushroom animal and human features, are representations of the in the vicinity of the centres that have yielded extensive
Shelter, Brotherton Rock and Procession Shelter (Jeremy First San,” while in the ‘Trance Hypothesis’ showcase the and varied archaeological assemblages (e.g. Collingham
Hollmann pers. comm. 20 December 2007). All site visits text reads as follows: “It has been suggested that, in some Shelter and Mhlwazini Cave) and that have contributed
are done in the company of a guide, who also provides instances, the spiritual healer could harness the power substantially to the understanding of the uKhahlamba-
interpretation. of an important animal, which could possibly explain Drakensberg hunter-gatherer past (Mazel 1990, 1992).

66 Chapter 5
Finally, there is a series of displays at Didima about rock
art and ‘every day life’ covering ‘Fish & Fish Spearing,’ References Mazel, A.D. 1990. Mhlwazini Cave: the excavation of late
‘Plant Food & Foraging’ and ‘Paints & Preparation’, but Holocene deposits in the northern Natal Drakensberg,
except for brief mention that bone fish hooks have been Blundell, G. 2002. The Unseen Landscape: A Journey to Natal, South Africa. Natal Museum Journal of
found, there is no reference to excavated remains that Game Pass Shelter. Johannesburg: Rock Art Research Humanities 2: 95–133.
shed light on these themes. With this, the opportunity Institute. Mazel, A.D. 1992. Collingham Shelter: the excavation of late
to integrate information from rock art and the excavated Carter, P.L. 1978. The prehistory of eastern Lesotho. Holocene deposits, Natal, South Africa. Natal Museum
record was lost. Display of excavated artefacts would Unpublished PhD thesis. Cambridge: University of Journal of Humanities 4: 1–52.
have significantly enriched the exhibits and provided Cambridge. Solomon, A.C. 1997. The myth of ritual origins?
visitors with a greater appreciation of the uKhahlamba- Lewis-Williams, J.D. 1981. Believing and Seeing: Symbolic Ethnography, mythology and interpretation of San
Drakensberg hunter-gatherer past. Meanings in Southern San Rock Paintings. London: rock art. South African Archaeological Bulletin 52: 3–13.
To conclude, it is clear that while Main Caves, Academic Press. Wright, J.B. & Mazel, A.D. 2007. Tracks in a Mountain Range:
Kamberg and Didima visitor attractions have greatly Lewis-Williams, J.D. & Pearce, D.G. 2004. San Spirituality: Exploring the History of the uKhahlamba-Drakensberg.
improved the presentation of uKhahlamba-Drakensberg Roots, Expression and Social Consequences. Walnut Johannesburg: Witwatersrand University Press.
San archaeology and history for the public, some issues Creek: AltaMira Press.
have not been adequately dealt with and others have Mazel, A.D. 1989. People making history: the last ten
been omitted altogether. These will require redressing in thousand years of hunter-gatherer communities in the Personal communication
future initiatives. Thukela Basin. Natal Museum Journal of Humanities 1:
1–168. Jeremy Hollmann, 20 December 2007

 Figures B6.1a & b

Boardwalk at Main Caves, Giant’s Castle,
central Drakensberg. From RARI archive.

The Eland's People 67

the ‘Cologne School’. The University of Cologne, with them lacking, he argued cogently against the use of
funding from the German Research Council, supported quantification. His criticism was largely logical in nature.
Pager’s (1989, 1993, 1995, 1998, 2000) and later Tilman He pointed out that the programme of quantification was
Lenssen-Erz’s (2001) quantitative studies in the Brandberg. essentially empiricist (as distinct from good empirical
These studies yielded vast volumes of numerical data and work) and based on flawed inductive logic (Lewis-
statistical analyses. Proponents of the Cologne School Williams 1983, 1984). We do not review the ensuing
argue that empirical accuracy of rock art documentation debates here. As Thomas Dowson (1990: 172) points
allows for explanations for the rock art to evolve directly out, “[c]ritiques of the practice of empiricism in rock art
from the data (but see Lewis-Williams 1990). research have become almost as tedious as empiricism’s
The difficulty with all the quantitative studies is continued practice”.
that although they produced a great deal of information Now that few researchers persist with such
– researchers knew for the first time, for instance, the quantitative studies (but see e.g. Lenssen-Erz 1989, 2001;
percentages of depictions of eland in a given area – the Eastwood 2003, 2005), many researchers seem to ignore
meaning of the paintings did not flow from this information. this important debate from the 1980s. It is a great pity that
Vinnicombe’s analysis did, however, show that the this is the case because several of the points made apply
rock paintings were not merely a reflection of the daily more widely than to quantification alone. Although few
pursuits of San (Vinnicombe 1976: 347). She observed researchers now make explicitly empiricist arguments,
that selectivity of subject matter marked adherence to
the unwitting use of arguments of this kind seems to
specific conventions. She concluded, however, “the
be creeping back into rock art studies and, indeed,
presentation of numerical data in support of the selective
archaeology more widely.
and repetitive characteristics of the art is nevertheless
Possibly the greatest benefit (and certainly not an
relatively meaningless unless the rationale underlying
intended one) of the quantitative programme is that it
the preferences is understood” (Vinnicombe 1976: 347).
forced researchers to look in great detail at a very large
The selectivity shown by the painters themselves led
number of paintings. It was the resulting familiarity with
Vinnicombe (1976: 349) to seek explanations in a general
the art and its fine details that allowed for more detailed,
anthropological approach and, in particular, in a theory
ethnographically based interpretations. Arguments ceased
of ritual. She argued, for instance, that eland (Figure 5.2)
being made about ‘the art’ and instead were made about
served as a link between the material and spiritual worlds
individual figures and details of those figures.
of the San (Vinnicombe 1976: 353).
One study that attempted to look at the relationships Unfortunately, and as with empiricist arguments, this
between categories, and produced significant results, was extremely important experience of earlier researchers is
that of Lewis-Williams (1972, 1974), who quantified being forgotten. Many subsequent students of San rock
paintings in the Giant’s Castle area of the KwaZulu- art have not paid as much attention to these fine – but all-
Natal Drakensberg and the Barkly East District of the important – details of the paintings. When interpretative
Eastern Cape Province. Although most of the data he statements are made about ‘the art’ without recourse to
collected remained as opaque to understanding as those details of particular images, we end up with largely abstract
of other researchers collecting similar material, his statements that, although they may appear reasonable in
superpositioning study did produce substantive results: theoretical terms, are unable to explain details of rock art
images. An example of this type of explanation is the so-  Figures 5.3a & b
he found that certain categories of image occurred in
called ‘mythical’ interpretation of San rock art (Solomon David Lewis-Williams found that
statistically significant relationships with other categories.
many paintings depicted actions
This point, however, was where the limitations of the 1997), which may sound reasonable in abstract terms, but
and equipment related to the trance
study became apparent: it was not possible to say what which is unable to provide interpretations for individual
dance; conversely, he found few
the meanings of the relationships were. images. The problem results from a combination of
images that related to other rituals.
Lewis-Williams’s greatest contribution as far as researchers’ being insufficiently familiar with enough rock
Photo by Lorna Marshall. Photo from
quantification was concerned was probably his argument art in the field and failing to design research that takes
RARI archive.
against it. After he tried to use the techniques, and found sufficient cognisance of the fine detail of rock art.

68 Chapter 5
 Ethnography

Although Vinnicombe was not the first researcher to use

ethnography to help interpret San rock art, she did so in a
much more detailed way than many of her predecessors.
Whereas most earlier researchers tended to use ethnography
in rather general terms (e.g. Lee & Woodhouse 1970),
Vinnicombe used it more extensively and in greater detail.
The way she used the ethnography, however, did not differ
significantly from that of her predecessors. Like them, she
drew direct links between the pictures and the ethnography.
This approach is what has, rather disparagingly, been referred
to as ‘ethnographic snap’: researchers search ethnographies
for all the information they can find about a particular subject
in the paintings (eland, for instance) and then take that
information to explain paintings of that subject. Whilst the
information about the subject does likely reflect indigenous
knowledge on the topic, it cannot be said to necessarily
explain the meaning behind the painting of that subject.
In Vinnicombe’s case, this approach translated into
searching the published ethnographies for references
to animals and then trying to link the details from the
ethnographies to the paintings. In this process she correlated
some paintings with mythical accounts and others with
the rituals described. Unfortunately, she was never able to
link the mythical events she recounted to the paintings in
detailed ways. Despite this obvious shortcoming, some later
researchers have attempted similar correlations with similar
lack of success (e.g. Solomon 1997).
Lewis-Williams overcame this problem by using
ethnography in more specific ways. He was able to narrow
down the context of meaning by linking rock paintings to
San ritual (Lewis-Williams 1981). He discussed in detail the
most prominent rituals in San society: girls’ initiation, boys’
first kill, marriage and the trance dance. He expressed his
discussion explicitly in terms of Peirsian semiotics. Using the
language of icons, indices and symbols, he argued for the
specific function of signs in San society. He suggested that
the images depicted in San rock art were symbolic, and that
although many were polysemic, their context (i.e. a rock face)
focused the associations of the symbol (Lewis-Williams 1981,
1998, 2001b). The use of ethnography in this non-literal
way was a major step forward that has allowed for detailed
insights into San rock art. Unfortunately, this same step has
not been taken in most other branches of archaeology.

The Eland's People 69

In linking his ethnographic data from San ritual to rock
paintings, Lewis-Williams found that many paintings
depicted actions and equipment related to the trance
dance (Figures 5.3a, b); conversely, he found few images
that related to other rituals (Lewis-Williams 1981). John
Parkington and colleagues (Parkington 1996; Parkington &
Manhire 1997; Parkington et al. 1996) have subsequently
argued that some of these rituals are depicted in rock
paintings in the Western Cape Province, but Judith
Stevenson (2000) has shown that certain details suggested
to be related to initiation rituals may in fact relate to other
ritual and gender contexts.
Most subsequent uses of ethnography in the study of
San rock art have followed a similar method. Cases have
been multiplied, details refined and geographical spread
increased, but the underlying method has remained
essentially the same. A number of other methods for
studying San rock art have, however, grown out of the use
of ethnography. Probably the best known of these is the use
of neuropsychology.
Ethnographically informed studies of San art showed
that it was essentially religious, relating to shamanistic
practices, beliefs, experiences and cosmology (Lewis-
Williams & Dowson 1999; Lewis-Williams & Pearce 2004).
Shamanistic ritual (the trance dance) led participants
into altered states of consciousness (Katz 1982). This
insight from the ethnography led researchers to examine
the neuropsychological literature to find out what people
in altered states of consciousness actually experienced in
that state (for details see Lewis-Williams & Dowson 1988;
Lewis-Williams 2001a). Information on the phenomena
experienced by those in altered states proved useful in
understanding certain enigmatic San paintings (e.g. Lewis-
Williams 1995; Blundell 1998). Although knowledge of San
ethnography led researchers to pursue neuropsychology
as an interpretative method, it is important to note that it
itself is not an ethnographic method. It is an independent,
non-ethnographic line of evidence that can be pursued in
addition to ethnography.
While interest in neuropsychology was an outgrowth
of the ethnographic approach that developed after the  Figure 5.4
publication of People of the Eland, another outgrowth of Northern Sotho paintings have been linked to boys' initiation. Photo from RARI archive.
the ethnographic method, and the associated awareness
of symbolism, used extensively in recent research was one  Figure 5.5
which Vinnicombe’s work in some ways prefigured: natural Vinnicombe saw contact imagery as a narrative record that she hoped to correlate with historically
modelling. It, too, is not an ethnographic method, but one known events. Tracing by Pat Vinnicombe. Redrawing by Justine Olofsson.

70 Chapter 5
 stimulated (at least in southern Africa) by ethnography. potential to place rock art research more firmly within the
Natural modelling refers to the modelling of symbols and broader field of archaeology through its contribution to an
non-real, supernatural aspects of belief on real, natural understanding of social change. Vinnicombe devoted the
phenomena, often animals and their behaviour. In this opening chapters of People of the Eland to finding a link
way, symbols are often derived from natural phenomena between written colonial documents relating to the Maloti-
in a ‘logical’ (rather than arbitrary) way (Whitley 2001). Drakensberg Mountains and the San rock paintings found
Southern African researchers drew on methodologically in shelters throughout that region. She drew a distinction
innovative studies in North America (Whitley 1994, 2001) between images with ‘traditional’ subject matter and those
to undertake studies explicitly using natural models to that displayed colonial-period subjects or ‘contact’ imagery
explain features of southern African San rock art (Hollmann (Figure 5.5). Whilst she used an ethnographic approach
2002; Mguni 2004, 2005; Mallen 2005). The use of to interpret the supposedly traditional subjects, she
explicit natural models in this way allows researchers to envisaged the contact imagery as a painted substitute for
link observable features of the natural world to features historical records written by the San. She thus saw contact
of rock art not directly explained by ethnography. This imagery as a narrative record that she hoped to correlate
method, of course, is a type of analogy and subject to the with historically known events.
methodological constraints of that type of argument (see By her own admission (Vinnicombe 1976: 9), this
Wylie [1982, 1985] for methodological considerations in approach proved to be substantially more difficult in
analogy arguments). practice than she had envisaged. Indeed, Geoffrey Blundell
The success of the ethnographic approach in (2004: 63) has suggested that certain major dichotomies
understanding and interpreting San rock art has led the arose from her publication of People of the Eland. First, the
approach to be more widely used, both in other branches distinction she made between ‘contact’ and ‘traditional’
of southern African archaeology (Mitchell 2005) and to imagery, and the different approaches to their interpretation,
interpret other traditions of rock art in southern Africa. The proved to be a hindrance. Second, her attempt to write
most prominent of the other art traditions that ethnography an ‘insider’s’ history of the San relied almost entirely on
has helped interpret are those made by Northern Sotho- colonial outsider records. Despite her best efforts, she did
speaking people in the northern parts of South Africa. not manage to use the ‘contact’ art as evidence for the San
These studies have linked the so-called Late White rock art point of view. These issues have remained pertinent in
(Figure 5.4) to boys’ and girls’ initiation rituals (Moodley the decades since her study, and subsequent research has
2004, 2008; Namono 2004; Namono & Eastwood 2005). turned to various social theories in attempts to overcome
With the past success of the ethnographic method these dichotomies and incorporate San rock art into the
clearly demonstrated, it seems likely that it will throw writing of southern African history.
light on other, so far enigmatic, traditions of rock art in Colin Campbell (1986, 1987) made a ground-
southern Africa. The as yet uninterpreted geometric breaking contribution to this effort when he put forward
tradition attributed to herder people (Smith & Ouzman a structural-Marxist approach to understanding change
2004) is, for example, also likely to be understood through in the San rock art in the Maloti-Drakensberg region. He
ethnographic studies. used the notion of symbolic labour (an idea introduced
into Marxist theory by Maurice Godelier [1975, 1977] and
first applied to San rock art by Lewis-Williams [1982])
History to suggest that San interaction with colonists and Nguni
speakers led to social change through the development
Linking San rock art to historical documents represents an of a new element in the relations of production, what he
ongoing quest, which had its inception in the publication termed the “shamanistic relation of production” (Campbell
of People of the Eland (Vinnicombe 1976). Vinnicombe’s 1987: 46, original italics). Focusing on ‘contact’ rock art, as
goal of writing a history of the south-eastern San opened categorised by Vinnicombe (1976), he suggested that this
the way for what has become an increasingly important imagery was used as a backdrop that reinforced the new
research avenue, made all the more attractive by its social order (Figure 5.6). Importantly, Campbell was able

The Eland's People 71

  Figure 5.6
Both Patricia Vinnicombe and Colin Campbell
made a division between 'contact' and
'traditional' imagery. Images such as this one
were classified as 'contact' based on their
inclusion of horses. However, this kind of
classification led both researchers to ignore a
substantial corpus of rock art that may have
played a major role in interactions. Tracing by
Patricia Vinnicombe. Redrawing by Justine
Olofsson.

to see the San as active agents in their relationships with He argued that by manipulating the rules of painting,
others. However, Vinnicombe’s division between contact shamans were attempting to manipulate the cosmos.
and traditional imagery, which he maintained, meant that Importantly, he saw the art not only as a reflection of
he ignored a substantial corpus of rock art that may have this process but as actively involved in the negotiation of
played a role in the interactions that he considered. social change. Perhaps the most important problem with
Thomas Dowson (1994, 1995, 1998, 2000) took Dowson’s work is that he suggests that the social change
the process of historicising San rock art a step further by amongst the San who produced the paintings he considers
employing structuration theory (as described by Bourdieu was initiated by economic changes brought about by
1977; Giddens 1984). This approach provided a more interaction amongst San, Bantu speakers and Europeans.
nuanced explanation of social change in San societies In the absence of dating of the specific images he
within the Maloti-Drakensberg region. Focusing on the last considers, this position is an assumption based on broader
2 000 years, Dowson argued that interaction between San revisionist concepts of San history. Without direct dating
and newly arrived Nguni speakers precipitated change in or associations with contact period imagery, the phases he
San society, from which San shamans benefited. Focusing identifies could equally well illustrate changes amongst
on ‘traditional’ imagery, he suggested that three phases pre-contact hunter-gatherers. Dowson’s innovation in
of social change are evident in rock art in the Maloti- considering rock paintings to be part of active agency
Drakensberg Mountains: communal groups, consortium relationships should not, however, be underestimated.
groups, and pre-eminent shamans (Dowson 1994, 2000). Much recent research is based on this concept.

72 Chapter 5
 Figure 5.7
One marker of interaction is that domestic
ungulates took on symbolic meaning within
the San shamanic worldview. Tracing by
Patricia Vinnicombe. Redrawing by Justine
Olofsson.

In addition to these structural-Marxist and structurationist to have meaning within the San shamanistic world view
approaches, other perspectives that have been termed (Figure 5.7). Similarly, Simon Hall (1994) focuses on the
interactionist (Blundell 2004) have also been brought to southern Eastern Cape Province and argues that evidence
bear on the question of how to historicise San rock art. from excavated material, rock art and ethnographic data
Less explicitly theoretical, these studies have sought to suggests that San–Bantu-speaker interaction in that area
combine oral traditions, written historical information, was not a simple one-way process, but rather led to mutual
and archaeological and rock art data. An early study social restructuring.
focusing on images of domestic stock in San rock art Importantly, each of these studies focuses on a specific
found a correlation between paintings of fat-tailed sheep geographical area and the particular interactions amongst
in the Cederberg Mountains of South Africa’s Western the people within it, as opposed to wide, overarching
Cape Province and the location of known pastoralist sites generalisations. They also combine numerous strands of
during the colonial era (Manhire et al. 1986). Building on evidence, creating a detailed picture of the fit between
this early research, an important study by Johannes Loubser the rock art and the broader historical and excavated
and Gordon Laurens (1994) in the Caledon River Valley, data. However, they tend to focus only on explicit contact
Free State Province, considers the effects of the spread of imagery and do not consider the role of the broader corpus
Sotho-Tswana-speaking agropastoralists on the San rock of rock art in these processes of interaction.
art of the area. They conclude that the depiction of images A slight variation on these interactionist approaches
such as domestic ungulates, derived from a Sotho-Tswana is that put forward initially by Francis Thackeray (1984,
socio-economic world, suggests that these images came 1988, 1990), and more recently by Pieter Jolly (e.g. 1995,

The Eland's People 73

1996a, 1996b, 2005) and Frans Prins (1990, 1994, 1999).
Given the evidence for long periods of contact with Bantu
speakers, both Prins and Jolly have suggested that, contrary
to the earlier work of Vinnicombe and Lewis-Williams, San
paintings should not be interpreted exclusively in terms of
San ethnography, but that interpretation should consider
also the ethnography of Bantu speakers.
Jolly (1996b) suggests that the establishment of
a symbiotic relationship between the San and Bantu
speakers in the Drakensberg may have meant that the
San adopted some of the ritual and religious elements of
their agropastoralist neighbours, and that aspects of their
art should thus be interpreted from the Bantu-speakers’
perspective. Prins (1994, 1999) emphasises evidence
for intermarriage amongst San and Bantu speakers,
concluding that all San rock art may be seen as ‘contact’
imagery. Although Prins and Jolly do not conceptualise the
relationships between San and Bantu speakers in the same
way, both positions overlook evidence for multidirectional
and changing relationships in which the San often played
an influential role (recently, Jolly (2007) has softened his
previous position somewhat, and has acknowledged this
multi-directional flow of influence, although he maintains mounted raiders who occupied the area in the nineteenth  Figure 5.8
an emphasis on the dominance of agro-pastoral religious century. Prolonged precolonial interaction amongst hunter- Nomansland. The southern
beliefs). Important examples of this include linguistic gatherers, pastoralists and agropastoralists in this region, Drakensberg was the home of the last
evidence concluding that the Nguni languages from this he argues, resulted in a coalescence of certain cultural San painters. From RARI archive.
south-eastern region of South Africa have incorporated beliefs. Drawing on the ethnography of San, Khoekhoen
clicks from San languages (e.g. Louw 1977, 1979; Herbert and Bantu-speaking groups, all of whom are represented
1990a, 1990b, 1995). In addition to this, the anthropologist among the AmaTola, Challis highlights overlapping beliefs.
David Hammond-Tooke (1997, 1998, 1999, 2002) has The connection between this multi-ethnic group and the
argued extensively that certain distinctive cultural elements rock art allows Challis to circumvent one of the problems
found amongst south-eastern Nguni speakers are not found of earlier interactionist approaches, which focused on the
in any other Bantu-speaking group, but were borrowed directionality of cultural borrowing. Instead, the concept
from the San. His argument for cultural borrowing of ‘cultural coherence’ (Challis 2008: 146), established
contradicts the idea of a unidirectional flow of information with recourse to ethnographic and painted data, is used
from Bantu speakers to San, and emphasises the need for a to provide new insights into the worldview of a creolised
more nuanced conception of the ways in which San rock art raiding group. Importantly, Challis notes that ‘every
and historical information may be combined. instance of interaction was different’ (2008: 44). In focusing
The challenge of combining San rock art and historical on the particular case of the AmaTola, he is able to build a
information has been taken up by Sam Challis (2008), detailed and nuanced understanding of this group and a
whose work in the Maloti-Drakensberg region integrates subcategory of rock art that is linked to them. Individual
ethnographic and historical information with a recurring cases of focused and nuanced research such as this will
set of images: dancing groups, baboons, horses, and collectively allow us to build up a broader picture that more
baboon and horse therianthropes. Based on coherence of accurately reflects the complexity of inter-ethnic interaction
style and regional location, he argues that this category in the historical period and its relationship to the production
of images was made by the AmaTola, a creolised group of and consumption of rock art (see also Mallen 2008).

74 Chapter 5
 In contrast, Blundell (2004) has suggested that the key to be fair to say that we broadly understand what San rock
overcoming some of the difficulties inherent in previous art is about, both at the level of the art corpus and at the
approaches to historicising San rock art is theoretical. Blundell level of many classes of imagery. It is no longer necessary
suggests that the adoption of social theories of a structural for researchers to argue over this broad understanding
type, typical of the approaches already discussed, creates the or to justify it in each publication: the broad-level,
impression that the art was something that was intellectualised. ethnographically derived understanding of San rock art as
Drawing on ethnographic and painted evidence, he argues religious in nature and relating to shamanistic practices is
that the art was experienced rather than intellectualised now generally accepted and understood. This is, of course,
and that the production and consumption of rock art not to say that one cannot criticise details of interpretations
needs to be seen as an embodied process. Using theories of – many such details do require revision – and new emphases
embodiment to explain social change in the San communities can usefully be employed. The pre-Vinnicombe days of
in Nomansland (northern Eastern Cape Province; Figure 5.8), “gazing and guessing” (Lewis-Williams 2002), where any
Blundell argues that a particular category of anthropomorphic guess seemed as good as any other, are, thankfully, now
images, described as significantly differentiated figures (SDFs; long gone. One can no longer promote contrary opinions
similar to Dowson’s pre-eminent shamans), were depictions of of the art simply for the sake of having a different opinion
“powerful individual potency-owners” (Blundell 2004: 176). or appearing critical (Whitley 2006). Unless these contrary
He suggests that a progression in the painted evidence from opinions are strongly grounded in San ethnography, San
SDFs to large-headed SDFs represents a shift in the role of San history, details of the rock art, and are able to explain
ritual specialists through time. He argues that as global and even more specific details within San art than the current
local processes of interaction came to have more of an effect on explanation, they should be rejected. If researchers wish to
Nomansland, the changing role of San ritual specialists from contest the broad-level interpretation they need to present
healers to rainmakers served to decrease the importance of an alternative, more persuasive explanation that accounts
the post-cranial body in ritual and some classes of rock art for more of the data and is internally consistent, rather than
(Blundell 2004: 177). Arguing that the body is intimately tied simply making claims that cannot be substantiated by the
up with identity construction, he suggests that we should see available evidence.
this change in the depiction of images as evidence of a shift in The next challenge for researchers is not the continued
the way individual identity was constructed in ritual in the San justification of the overall understanding of the art, but an
communities of Nomansland (Blundell 2004: 177). engagement with the art on a much finer scale. This is, in
a way, what Vinnicombe was attempting to do in People of
the Eland. Subsequent research has now made such studies
Directions possible. Instead of understanding ‘the art’, researchers
need to begin to work out detailed interpretations for
A preliminary examination of the trajectory of rock art research individual shelters. The accumulation of and relationship
in southern Africa suggests that methodological changes have between paintings in a particular shelter has often been
played a far greater role in advancing study of the art than the discussed in terms of the physical sequence of paintings, but
application of new theories. In particular, the ethnographic much still needs to be done when we consider these same
method has proved pivotal in our current understanding of accumulations and relationships in cognitive, symbolic and
the art. Indeed, it is telling that this method has remained social terms. Similarly, the relationships between particular
constant whilst one theory after another has come and gone. sites need to be examined.
On the other hand, one should consider that it was probably In addressing these challenges researchers need to
broad theoretical changes that led to the adoption of the beware: a major trend in world archaeology is towards
ethnographic method in the first place, or at least provided theorisation. Whilst not in any way wishing to deny the
the conditions necessary for its adoption. Research so far has important role of theory (our discussion has illustrated
led us to a position in which we now know a great deal about how changing theory has helped refine interpretations),
San rock art, certainly much more than was known when researchers need to be cautious of relying too heavily on it.
Vinnicombe was writing People of the Eland. Indeed, it would Here we do not refer to the use of theory in interpretation,

The Eland's People 75

but to the imposition of wholesale theoretical explanations References
– gender or landscape, for instance – without recourse to
details of the art and ethnography. It is, after all, only one Blundell, G. 1998. On neuropsychology in southern African rock art
part of a suite of tools researchers need to use in crafting research. Anthropology of Consciousness 9: 3–12.
an interpretation. The relationship between theory and Blundell, G. 2004. Nqabayo’s Nomansland: San Rock Art and the
data is one that has changed repeatedly in studies of San Somatic Past. Uppsala: Uppsala University Press.
rock art. Researchers have to guard against the pendulum Bourdieu, P. 1977. Outline of a Theory of Practice. Cambridge:
swinging too far in either direction (Dowson 2001: 313). Cambridge University Press.
The unintentional lesson of quantification has much to Campbell, C. 1986. Images of war: a problem in San rock art
teach students answering these new questions: researchers research. World Archaeology 18: 255–267.
must engage with details of the rock art images. Campbell, C. 1987. Art in crisis: contact period rock art in the
A second area, on a much larger scale of analysis, south-eastern mountains of southern Africa. Unpublished MSc
which is also beginning to open up, is the consideration thesis. Johannesburg: University of the Witwatersrand.
of regional diversity. This is an area into which frequent, Challis, W. 2008. The impact of the horse on the AmaTola
but not particularly successful, forays have been made. ‘Bushmen’: New identity in the Maloti-Drakensberg Mountains
Vinnicombe’s suggestion that quantitative studies could of southern Africa. Unpublished D.Phil thesis. Oxford:
show regional variations in San rock art was certainly University of Oxford.
correct. Differences of composition and frequency can (and Deacon, J. 2006. Rock art conservation and tourism. Journal of
have) certainly be demonstrated. The next step is to answer Archaeological Method and Theory 13(4): 379–399.
the questions of why the differences occur and what those Dowson, T.A. 1990. Rock art research’s umbilical cord: a review
differences mean. Answering those questions will once of Africa Praehistorica, Vol. 1, The rock paintings of the Upper
more return researchers to consideration of fine details of Brandberg: Part I Amis Gorge, Heinrich-Barth-Institut, Köln.
the art and locally specific ethnographies and historical Cimbebasia 12: 172–176.
sources. Dowson, T.A. 1994. Reading art, writing history: rock art and social
The work Vinnicombe pioneered has come a long way change in southern Africa. World Archaeology 25: 332–345.
since she introduced her innovations. We are, however, by Dowson, T.A. 1995. Hunter-gatherers, traders and slaves: the
no means approaching the end of the intellectual trajectory ‘Mfecane’ impact on Bushmen, their ritual and their art. In:
upon which she embarked almost 50 years ago. The quest Hamilton, C. (ed.) The Mfecane Aftermath: Reconstructive
for more detailed meaning in rock art research continues. Debates in Southern African History: 51–70. Johannesburg:
Although now equipped with an ever-growing collection of Witwatersrand University Press.
interpretative tools, the task ahead remains a challenging Dowson, T.A. 1998. Rain in Bushman belief, politics and history:
and exciting one. the rock-art of rain-making in the south-eastern mountains,
southern Africa. In: Chippindale, C. & Taçon, P.S.C. (eds)
The Archaeology of Rock-Art: 73–89. Cambridge: Cambridge
Acknowledgements University Press.
Dowson, T.A. 2000. Painting as politics: exposing historical
Our thanks go to the editors for inviting us to contribute to this processes in hunter-gatherer rock art. In: Schweitzer, P.P.,
volume. We thank colleagues and a referee who commented on Biesele, M. & Hitchcock, R. (eds) Hunters and Gatherers in the
drafts of this chapter. The Rock Art Research Institute is funded Modern World: Conflict, Resistance and Self-determination: 413–
by the South African National Research Foundation and the 426. New York: Berghahn Books.
University of the Witwatersrand. Neither institution is responsible Dowson, T.A. 2001. Queer theory and feminist theory: towards a
for the views expressed herein. sociology of sexual politics in rock art research. In: Helskog,
K. (ed.) Theoretical Perspectives in Rock Art Research: 312–329.
Oslo: Novus Forlag.
Eastwood, E.B. 2003. A cross-cultural motif in San, Khoekhoe and
Northern Sotho rock paintings of the Central Limpopo Basin,

76 Chapter 5
 southern Africa. South African Archaeological Bulletin 58: Hœrle, S. 2006. Rock temperatures as an indicator of weathering
14–26. processes affecting rock art. Earth Surface Processes and
Eastwood, E.B. 2005. From girls to women: female imagery in the Landforms 31: 383–389.
San rock paintings of the Central Limpopo Basin, southern Hœrle, S. & Salomon, A. 2004. Microclimatic data and rock art
Africa. Before Farming [online version] 2005/3, Article 2. conservation at Game Pass Shelter in the Kamberg Nature
Giddens, A. 1984. The Constitution of Society: Outline of the Theory Reserve, KwaZulu-Natal. South African Journal of Science 100:
of Structuration. Cambridge: Polity Press. 340–341.
Godelier, M. 1975. Modes of production, kinship, and Hollman, J.C. 2002. Natural models, ethology and San rock-
demographic structures. In: Bloch, M. (ed.) Marxist Analyses paintings: pilo-erection and depictions of bristles in
and Social Anthropology: 3–27. London: Malaby Press. southeastern South Africa. South African Journal of Science 98:
Godelier, M. 1977. Perspectives in Marxist Anthropology. 563–567.
Cambridge: Cambridge University Press. Hughes, J.C. & Solomon, A.C. 2000. A preliminary study of
Hall, K., Meiklejohn, I., Arocena, J., Prinsloo, L., Sumner, P. & ochres and pigmentaceous materials from KwaZulu-Natal,
Hall, L. 2007. Deterioration of San rock art: new findings, South Africa: towards an understanding of San pigment and
new challenges. South African Journal of Science 103: 361–362. paint use. Southern African Humanities 12: 15–31.
Hall, S. 1994. Images of interaction: rock art and sequence in Jolly, P. 1995. Melikane and Upper Mangolong revisited: the
the Eastern Cape. In: Dowson, T.A. & Lewis-Williams, J.D. possible effects on San art of symbiotic contact between
(eds) Contested Images: Diversity in Southern African Rock Art south-eastern San and Southern Nguni and Sotho
Research: 61–82. Johannesburg: Witwatersrand University communities. South African Archaeological Bulletin 50: 68–80.
Press. Jolly, P. 1996a. Interaction between south-eastern San and
Hammond-Tooke, W.D. 1997. Whatever happened to /Kaggen? southern Nguni and Sotho communities c. 1400 to c. 1800.
A note on Cape Nguni/Khoisan borrowing. South African South African Historical Journal 35: 30–61.
Archaeological Bulletin 52: 122–124. Jolly, P. 1996b. Symbiotic interaction between black farmers and
Hammond-Tooke, W.D. 1998. Selective borrowing? The south-eastern San. Current Anthropology 37: 277–305.
possibility of San shamanistic influence on Southern Bantu Jolly, P. 2005. Sharing symbols: a correspondence in the ritual
divination and healing practices. South African Archaeological dress of black farmers and the south-eastern San. South
Bulletin 53: 9–15. African Archaeological Society Goodwin Series 9: 86–100.
Hammond-Tooke, W.D. 1999. Divinatory animals: further Jolly, P. 2007. Before farming? Cattle kept and painted by the
evidence of San–Nguni borrowing. South African southeastern San. Before Farming [online version] 2007/4,
Archaeological Bulletin 54: 128–132. Article 2.
Hammond-Tooke, W.D. 2002. The uniqueness of Nguni Katz, R. 1982. Boiling Energy: Community Healing Among the
mediumistic divination in southern Africa. Africa 72: 277– Kalahari Kung. Cambridge (MA): Harvard University Press.
292. Lee, D.N. & Woodhouse, H.C. 1970. Art on the Rocks of Southern
Herbert, R.K. 1990a. The sociohistory of clicks in southern Bantu. Africa. London: Purnell.
Anthropological Linguistics 32: 120–138. Lenssen-Erz, T. 1989. The conceptual framework for the analysis
Herbert, R.K. 1990b. The relative markedness of click sounds: of the Brandberg rock paintings. In: Pager, H. The Rock
evidence from language change, acquisition and avoidance. Paintings of the Upper Brandberg: Part I Amis Gorge: 361–370.
Anthropological Linguistics 32: 295–315. Köln: Heinrich-Barth-Institut.
Herbert, R.K. 1995. The sociohistory of clicks in Southern Bantu. Lenssen-Erz, T. 2001. Gemeinschaft-Gleichheit-Mobilität: Felsbilder
In: Mesthrie, R. (ed.) Language and Social History: Studies im Brandberg, Namibia, und ihre Bedeutung. Grundlagen einer
in South African Sociolinguistics: 51–67. Cape Town: David textuellen Felsbildarchäologie. Köln: Heinrich-Barth-Institut.
Philip. Leroi-Gourhan, A. 1968. The Art of Prehistoric Man in Western
Hœrle, S. 2005. A preliminary study of the weathering activity at Europe. London: Thames and Hudson.
the rock art site of Game Pass Shelter (KwaZulu-Natal, South Lewis-Williams, J.D. 1972. The syntax and function of the Giant’s
Africa) in relation to its conservation. South African Journal of Castle rock paintings. South African Archaeological Bulletin 27:
Geology 108: 297–308. 49–65.

The Eland's People 77

Lewis-Williams, J.D. 1974. Superpositioning in a sample of Louw, J.A. 1977. The adaptation of non-click consonants in
rock paintings from the Barkly East District. South African Xhosa. Khoisan Linguistic Studies 3: 74–92.
Archaeological Bulletin 29: 93–103. Louw, J.A. 1979. A preliminary survey of Khoi and San influence
Lewis-Williams, J.D. 1981. Believing and Seeing: Symbolic Meanings in Zulu. Khoisan Linguistic Studies 8: 21–89.
in Southern San Rock Paintings. London: Academic Press. Maggs, T.M.O’C. 1967. A quantitative analysis of the rock art
Lewis-Williams, J.D. 1982. The economic and social context of from a sample area in the western Cape. South African Journal
southern San rock art. Current Anthropology 23: 429–449. of Science 63: 100–106.
Lewis-Williams, J.D. 1983. Introductory essay: science and rock Mallen, L. 2005. Linking sex, species and a supernatural snake
art. South African Archaeological Society Goodwin Series 4: 3–13. at LAB X rock art site. South African Archaeological Society
Lewis-Williams, J.D. 1984. The empiricist impasse in southern Goodwin Series 9: 3–10.
African rock art studies. South African Archaeological Bulletin Mallen, L. 2008. Rock Art and Identity in the North Eastern Cape
39: 58–66. Province. Unpublished MA thesis. Johannesburg: University
Lewis-Williams, J.D. 1990. Documentation, analysis and of the Witwatersrand.
interpretation: dilemmas in rock art research. South African Manhire, A.H., Parkington, J.E., Mazel, A.D. & Maggs, T.M.O’C.
Archaeological Bulletin 45: 126–136. 1986. Cattle, sheep and horses: a review of domestic animals
Lewis-Williams, J.D. 1995. Seeing and construing: the making in the rock art of southern Africa. South African Archaeological
and ‘meaning’ of a southern African rock art motif. Cambridge Society Goodwin Series 5: 22–30.
Archaeological Journal 5: 3–23. Mazel, A.D. & Watchman, A.L. 1997. Accelerator radiocarbon
Lewis-Williams, J.D. 1998. Quanto? The issue of ‘many’ meanings dating of Natal Drakensberg paintings: results and
in southern African San rock art research. South African implications. Antiquity 71: 445–449.
Archaeological Bulletin 53: 86–97. Mazel, A.D. & Watchman, A.L. 2003. Dating rock paintings
Lewis-Williams, J.D. 2001a. Brainstorming images: in the Ukhahlamba-Drakensberg and the Biggarsberg,
neuropsychology and rock art research. In: Whitley, D.S. KwaZulu-Natal, South Africa. Southern African Humanities
(ed.) Handbook of Rock Art Research: 332–357. Walnut Creek: 15: 59–73.
AltaMira Press. Meiklejohn, K.I., Hall, K. & Davis, J.K. 2009. Weathering of rock
Lewis-Williams, J.D. 2001b. Monolithism and polysemy: Scylla art at two sites in the KwaZulu-Natal Drakensberg, southern
and Charybdis in rock art research. In: Helskog, K. (ed.) Africa. Journal of Archaeological Science 36: 973–979.
Theoretical Perspectives in Rock Art Research: 23–39. Oslo: Mguni, S. 2004. Cultured representation: understanding
Novus Forlag. ‘formlings’, an enigmatic motif in the rock art of Zimbabwe.
Lewis-Williams, J.D. 2002. A Cosmos in Stone: Interpreting Religion Journal of Social Archaeology 4: 181–199.
and Society through Rock Art. Walnut Creek: AltaMira Press. Mguni, S. 2005. A new iconographic understanding of formlings,
Lewis-Williams, J.D. & Dowson, T.A. 1988. The signs of all a pervasive motif in Zimbabwean rock art. South African
times: entoptic phenomena in Upper Palaeolithic art. Current Archaeological Society Goodwin Series 9: 34–44.
Anthropology 29: 201–245. Mitchell, P.J. 2005. Why hunter-gatherer archaeology matters: a
Lewis-Williams, J.D. & Dowson, T.A. 1999. Images of Power: personal perspective on renaissance and renewal in southern
Understanding San Rock Art (second edition). Johannesburg: African Later Stone Age research. South African Archaeological
Southern Book Publishers. Bulletin 60: 64–71.
Lewis-Williams, J.D. & Pearce, D.G. 2004. San Spirituality: Roots, Moodley, S. 2004. Koma: the crocodile motif in the rock art of
Expressions, and Social Consequences. Cape Town: Double the Makgabeng Plateau, Limpopo Province, South Africa.
Storey. Unpublished MA thesis. Johannesburg: University of the
Loubser, J. & Laurens, G. 1994. Depictions of domestic Witwatersrand.
ungulates and shields: hunter-gatherers and agro-pastoralists Moodley, S. 2008. Koma: the crocodile motif in the rock art of
in the Caledon River Valley area. In: Dowson, T.A. & the northern Sotho. South African Archaeological Bulletin 63:
Lewis-Williams, J.D. (eds) Contested Images: Diversity in 116–124.
Southern African Rock Art Research: 83–118. Johannesburg: Namono, C. 2004. Dikgaatwane tsa Basadi: a study of the link
Witwatersrand University Press. between girls’ initiation and rock art in the Makgabeng

78 Chapter 5
 Plateau, Limpopo Province, South Africa. Unpublished MA formations, and the unreliable informant. Natal Museum
thesis. Johannesburg: University of the Witwatersrand. Journal of Humanities 11: 43–62.
Namono, C. & Eastwood, E.B. 2005. Art, authorship and female Russell, T. 2000. The application of the Harris Matrix to San
issues in a Northern Sotho rock painting site. South African rock art at Main Caves North, KwaZulu-Natal. South African
Archaeological Society Goodwin Series 9: 77–85. Archaeological Bulletin 55: 60–70.
Pager, H. 1971. Ndedema: A Documentation of the Rock Paintings of Smith, B.W. & Ouzman, S. 2004. Taking stock: identifying
the Ndedema Gorge. Graz: Akademische Druk. Khoekhoen herder rock art in southern Africa. Current
Pager, H. 1989. The Rock Paintings of the Upper Brandberg: Part I Anthropology 45: 499–525.
Amis Gorge. Köln: Heinrich-Barth-Institut. Solomon, A.C. 1997. The myth of ritual origins? Ethnography,
Pager, H. 1993. The Rock Paintings of the Upper Brandberg: Part II mythology and interpretation of San rock art. South African
Hongorob Gorge. Köln: Heinrich-Barth-Institut. Archaeological Bulletin 52: 3–13.
Pager, H. 1995. The Rock Paintings of the Upper Brandberg: Part III Stevenson, J. 2000. Shaman images in San rock art: a question of
Southern Gorges. Köln: Heinrich-Barth-Institut. gender. In: Donald, M. & Hurcombe, L. (eds) Representations
Pager, H. 1998. The Rock Paintings of the Upper Brandberg: Part IV of Gender from Prehistory to Present: 45–66. London:
Umuab and Karoub Gorges. Köln: Heinrich-Barth-Institut. Macmillan.
Pager, H. 2000. The Rock Paintings of the Upper Brandberg: Part Swart, J. 2004. Rock art sequences in the uKhahlamba-
V Naib Gorge (A) and the Northwest. Köln: Heinrich-Barth- Drakensberg Park, South Africa. Southern African Humanities
Institut. 16: 13–35.
Parkington, J.E. 1996. What is an eland? N!ao and the politics of Thackeray, J.F. 1984. Masquerade and therianthropes in art. The
age and sex in the paintings of the Western Cape. In: Skotnes, Digging Stick 2: 2.
P. (ed.) Miscast: Negotiating the Presence of the Bushmen: 281– Thackeray, J.F. 1988. Southern African rock art and Xhosa beliefs
289. Cape Town: University of Cape Town Press. associated with abantubomlambo. Pictogram 1: 2–3.
Parkington, J.E. & Manhire, A.H. 1997. Processions and groups: Thackeray, J.F. 1990. On concepts expressed in southern African
human figures, ritual occasions and social categories in the rock art. Antiquity 64: 139–144.
rock paintings of the Western Cape, South Africa. In: Conkey, Vinnicombe, P. 1967a. Rock painting analysis. South African
M.W., Soffer, O., Stratmann, D. & Jablonski, N.G. (eds) Archaeological Bulletin 22: 129–141.
Beyond Art: Pleistocene Image and Symbol: 301–320. Berkeley: Vinnicombe, P. 1967b. The recording of rock paintings: an
California Academy of Sciences. interim report. South African Journal of Science 63: 282–284.
Parkington, J.E., Manhire, A.H. & Yates, R. 1996. Reading San Vinnicombe, P. 1976. People of the Eland: Rock Paintings of the
images. In Deacon, J. & Dowson, T.A. (eds) Voices From the Drakensberg Bushmen as a Reflection of Their Life and Thought.
Past: /Xam Texts and the Bleek and Lloyd Collection: 212–233. Pietermaritzburg: Natal University Press.
Johannesburg: Witwatersrand University Press. Whitley, D.S. 1994. Shamanism, natural modelling, and the
Pearce, D.G. 2002. Changing men, changing eland: sequences in rock art of far western North America. In: Turpin, S. (ed.)
the rock paintings of Maclear District, Eastern Cape, South Shamanism and Rock Art in North America: 1–43. San Antonio:
Africa. American Indian Rock Art 28: 129–138. Rock Art Foundation.
Pearce, D.G. 2006. A comment on Swart’s rock art sequences and Whitley, D.S. 2001. Science and the sacred: interpretive theory
use of the Harris Matrix in the Drakensberg. Southern African in U.S. rock art research. In: Helskog, K. (ed.) Theoretical
Humanities 18(2): 173–177. Perspectives in Rock Art Research: 130–157. Oslo: Novus
Prins, F.E. 1990. Southern Bushman descendants in the Transkei: Forlag.
rock art and rainmaking. South African Journal of Ethnology 13: Whitley, D.S. 2006. Is there a shamanism and rock art debate?
110–116. Before Farming [online version] 2006/4, Article 7.
Prins, F.E. 1994. Living in two worlds: the manipulation of Wylie, A. 1982. An analogy by any other name is just as
power relations, identity and ideology by the last San rock analogical: a commentary on the Gould–Watson dialogue.
artist of the Transkei, South Africa. Natal Museum Journal of Journal of Anthropological Archaeology 1: 382–401.
Humanities 6: 179–193. Wylie, A. 1985. The reaction against analogy. Advances in
Prins, F.E. 1999. Dissecting diviners: on positivism, trance- Archaeological Method and Theory 8: 63–111.

The Eland's People 79