Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, AFRICAN HISTORY ( (c) Oxford University Press USA, 2018. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 19 January 2019

Rock Art Research Methods

Summary and Keywords

Rock art is an archaeological resource with the potential to reconstruct aspects of the ideologies of prehistoric societies. Research methods are distinguished here from theoretical, interpretive frameworks. The methods discussed here concern the documentation of rock art, methods of working with the temporal dimensions of rock art (such as developing relative chronologies and dating), and the characterization of pigments. Nonetheless, the choice of research methods depends on an explicitly formulated, theoretically informed research question. Research aims will also determine the scope and scale of the documentation and chronological methods employed.

Fieldwork is a major and initial component of documentation and may involve surveying for rock-art sites. Researchers should experience rock art first hand. Digital mapping and imaging techniques are used routinely, but field tracings continue to be an important means of recording and interpreting the art. Computational photography includes enhancement software such as DStretch and other techniques that enable researchers to see details that would otherwise be invisible.

Temporality is a fundamental attribute of rock art, and the biggest challenge in this regard is to relate the chronological sequences on the rock face to other archaeological and environmental data and thus contextualize the rock art. Relative chronologies provide information about the order of image-making episodes at a site or in a particular region. Age determinations may be arrived at using correlative methods in which the art is dated by means of independently available age ranges. Accelerator Mass Spectrometry (AMS) radiocarbon dating is commonly used to date organic paint samples. Engravings are difficult to date; age ranges obtained from cation-ratio (CR) and varnish microlamination (VML) are regarded as approximations. Pigment analysis is used to characterize the inorganic components of paint and to detect the presence of organic components. Research methods are multidisciplinary and thus require a coordinated, unified approach in order to achieve the research aims.

Keywords: African rock art, archaeology, dating, digital image enhancement, research methods, rock engravings, rock paintings

Researching Rock Art: Methods

Rock art may be broadly defined as anthropogenic markings that—with the exception of art mobilier such as plaquettes and other types of portable art—are painted and engraved on living or natural rock and are fixed in the landscape. The making of markings on rock is a widespread, social phenomenon that extends from the early 21st century to tens of thousands of years ago—rock art occurs everywhere on earth except Antarctica. This definition excludes geoglyphs, megalithic structures, inscriptions, and monumental structures. Painted images are known as rock paintings (used here), or pictographs. Rock engravings (used here) are also known as petroglyphs.1

The collective term “rock art” is widely used, but the concepts of “art” and “artist” need qualification. Some argue that “art,” in the sense of the production of “art objects” made by “artists” for exhibition or sale, is but one of many social and historical contexts for image making.2 The terms “image” and “image maker,” respectively, would thus be more appropriate. The words “art,” “art object,” and “artist” are only retained here because they are used in common parlance.

The value of rock art lies in its potential to recreate the ideologies of past societies for which other material culture remains are frequently absent, fragmentary, or mundane. A primary aim of rock-art research is to understand how rock art relates to the social-relational matrix of which it is part—that is, to investigate the “meaning” or significance of the images and the process of image production and consumption within a particular image-making society.

It is crucial to understand that images themselves do not speak, nor can they be “read” as a text (i.e., they are not themselves “signs”), although of course they may be spoken about using signs (i.e., language). Indeed, “realism [in rock art] . . . is often more apparent than real . . . what appear to be naturalistic portrayals of animals and humans are often likely to be well-disguised spirits, their supernatural aspect betrayed by clues that gain meaning in a larger context.”3 For example, hunter-gatherer paintings of fish on the banks of the Senqu River in Lesotho were almost certainly not painted as a record of daily activities for the entertainment of the hunter-gatherer community or the benefit of researchers. The fish paintings are juxtaposed with paintings of a mythical rain animal, a dying eland, and clapping and dancing human figures, all of which have shamanistic associations. It is thus more likely that these motifs invoke the purported ability of shamans “to harness and make rain, using that power to produce socially desirable benefits, including perhaps opportunities for group aggregation around seasonally restricted spawning runs of fish.”4

As this analysis and much other research suggests, rather than being a vehicle for symbolic communication, the rock art produced by a particular social group is, in Alfred Gell’s words, “a system of action, intended to change the world rather than encode symbolic propositions about it.”5 Rock art, according to this understanding, has a “practical, mediatory role . . . in the social process” of that particular image-making group.6 The role of images and image making is an inextricable part of the society in which it was produced.

It is also useful to distinguish between referential and gestural rock markings, although the mark makers themselves may not have recognized this dichotomy and some markings may straddle these boundaries.7 Many (perhaps most) rock markings are referential—that is, they have a referent in the form of an object or an organism and are more or less recognizable depending on the conventions and techniques used to depict them.8 Gestural markings, such as scatters of peck marks, battered edges of rocks, and grooves, on the other hand, do not necessarily depict any “thing”; rather, these markings are evidence of the performance of repeated actions, which the mark makers may have regarded as important and significant. For example, repeated and patterned scratches as well as much longer and deeper grooves made on spiritually significant outcrops of pyrophyllite in South Africa were possibly made to produce a mineral pigment used in body decoration and grooves, but, at the same time, the large grooves were perceived as scarification marks on the body of a supernatural serpent.9

“Method” as distinct from “theory” is defined as the systematic procedures followed in the creation of rock-art data: documenting, recording, mapping, sampling, and analyzing. Methods themselves are theoretically informed. Data are not theory free, and all rock-art research methods are practiced within a theoretical and methodological paradigm.10 “Analysis” of rock art therefore means the interpretation of data following explicit theoretical, methodological, and social assumptions; for the purposes of this article, however, theoretical frameworks of interpretation are not considered.11 A weakly formulated hypothesis is not made any stronger by the use of accurate and precise recording equipment. Nor should it be necessary to point out that the notion of absolute “proof” is inappropriate; it is rather the relative strength of arguments advanced and how these manage to explain or account for the phenomena for which explanations are sought.12

Rock art is overwhelmingly an archaeological rather than a historical artefact.13 Given the incomplete and partially preserved nature of most archaeological data, archaeology, in contrast to history, typically focuses on longer-term processes and structures in the absence of documented evidence about individuals and historically attested groups. For historians this has important implications because the study of archaeological artefacts is characterized by “its own set of sources, intellectual history, ways of doing, and ways of knowing—its own disciplinary imperative” that is distinct from those of historians, although both disciplines are mutually dependent.14

The archaeological nature of most rock art imbues it with not just the theoretical, methodological, and technical complexities intrinsic to archaeological artefacts and the practice of archaeology generally but also with its own specific problems.15 These are: aspects of the spatial qualities of rock art as an artefact that was intentionally embedded in its environment, understanding the human image-making intervention on the rock surface (i.e., techniques and materials used to make the art, subject matter, authorship and identity, meaning and motivation), and matters concerning the passing of time (chronology, style, superimpositions, dating estimates, and weathering of the imagery).

There is an array of methods to collect rock-art data that can be used to document and record rock art, and to investigate—from micro to macro levels—the relationship between the rock art, its environment, its archaeological context, and its temporality. Implicit in and indispensable to these methods is the notion of a style or art tradition. Without a concept of style, typologies, however loosely or rigidly defined, and formal classification are impossible. Slippery a concept as it is, style is a fundamental attribute of all material culture.16


“Style” as it is used here encompasses the repertoire of subject matter of a particular body of rock art (e.g., depictions of a limited range of animal species, human images in particular postures, etc.), the distinctive manner in which the subject matter is depicted (including the substrate, ways of marking the substrate, colors employed, materials), and the characteristic ways in which the images are arranged on the rock face. Key to the notion of style is the assumption that a particular style is unique and that it is the result of deliberate and intentional practices, beliefs, and decisions of the image makers themselves (not the researcher).17 A particular style is thus assumed to be the product of a specific time period and social grouping and may be understood as a “temporally or culturally diagnostic trait” of the image-making society.18

The efficacy of the stylistic method depends in part on the body of imagery that is the focus of research. For example, images of horses in the Maloti-Drakensberg Mountains of Lesotho and South Africa may confidently be situated within a specific and restricted period of time, between the 1830s and 1880s. This window is bounded by the introduction of the horse into the area in the 1830s and the defeat of cattle-raiding groups by colonial authorities in the 1880s. Horse imagery in the Maloti-Drakensberg correlates with other distinctive motifs that include paintings of baboons as well as depictions of people with standardized details that are repeated from site to site. There is a consistent and limited repertoire of motifs depicted at the sites and arranged in a limited number of contexts. The images are created according to stylistic conventions and paint that contrast markedly with an earlier, mostly pre-colonial image-making tradition. The availability of historical documents for the period was also a valuable adjunct to this research and made it possible to persuasively link the rock paintings to a historically known group, the AmaTola.19

On the basis of visual, spatial, and archaeological lines of evidence Sven Ouzman has identified a small and localized 18th–19th century Korana tradition of rock art in the interior of South Africa comprising finger painting and rough brushwork of traditional aprons, humans, animals (including horses), serpents, geometrics, and paint smears in red, orange, and white.20

Research by Adrian Flett and Penny Letley on stylistic change in paintings of eland (Taurotragus oryx) in the uKhahlamba-Drakensberg Escarpment of South Africa defines Early, Peak, and Later Period phases to the calibrated radiocarbon dates obtained from calcium oxalate layers in the uKhahlamba-Drakensberg by Aron Mazel and Alan Watchman.21

Documentation Methods

The fixed nature of most rock art sets it apart from much other archaeological material and has important implications for rock-art research and the methods deployed to understand it.22 Recording rock art begins with deciding on the definition of the “site” itself. In some cases it will be difficult to demarcate discrete rock-art sites because the rock art covers a very large contiguous area, such as an entire gorge (e.g., the Didima Gorge in the uKhahlamba-Drakensberg), a massif (e.g., Dâures in Namibia, Gilf Kebir in south west Egypt, and Lovo in Democratic Republic of Congo), or a single image in a ritual seclusion space containing a single female kudu.23

The notion of the “complete” documentation of a site is an unattainable, empiricist ideal, notwithstanding advances in digital technology that allow for the recording of aspects of rock-art sites with great precision and accuracy.24 Recording and documentation is always for a purpose and is therefore selective. This applies to rock-art research projects as well as to conservation and recording projects whose main aim is to “preserve,” “conserve,” or to “comprehensively document” rock-art sites. Total non-selectivity, in which data are purportedly accurately and comprehensively recorded without “subjectivity” for the benefit of future researchers with new research requirements, is a pipedream. Short of the “total reproduction of all site data at atomic level” it is impossible to future-proof documentation.25


Ideally researchers should experience first hand the spatial dimensions of the rock art in its physical environment to appreciate the setting of the art and, crucially, to see the original rock art. Research thus usually involves visiting the rock art. Copies of rock art of whatever kind, whether sketches and tracings, photographic, or 3D and 4D scans, are intrinsically selective and partial regardless of the sophistication or expense of the techniques employed. Most importantly, the researcher needs to decide what needs to be recorded at a site. What is the purpose of the proposed documentation? The research aims will determine how detailed and accurate the records need to be. Does the research require comprehensive documentation of the site or only the sampling of motifs?

The initial phase of documentation then is to visit the site to familiarize oneself with the natural features of the site and the images it contains. It is during this phase that researchers may evaluate how the research aims relate to the data and consider refining and revising research goals. This is also the first opportunity to look at the art in its context and to look for and note details about the individual figures and their interrelationships and decide how to classify the collective arrangement of the images at a site into clusters or “panels,” as they are often called. It may be useful to contextualize the location of the site with regard to its macro environment. This kind of recording would identify the landscape features that make up the site (e.g., the site may be incorporated into larger landscape features such as an overhang in a cliff in a valley) and to quantify these by providing information about the relevant dimensions. Site recording forms are commonly used to record these data. Many research institutions have standard site record forms so that data are collected in a systematic and consistent fashion. Collected data are then usually integrated into a database.

Surveys of likely areas in which to locate and document rock art are important ways of gathering data; they are an important preliminary phase of research because they can reveal patterns in the rock-art data and can modify the scope and aims of the research. Local knowledge is often crucial to planning and executing field surveys, especially in extreme and topographically challenging environments.

Surveys for rock art in the great massifs of Gilf Kebir, Jebel Ouenat, Arkenu, and the surrounding smaller hills in the central Libyan Desert (Eastern Sahara) have enabled researchers to construct chronologies of various styles of rock art and to conduct research on hitherto unknown large and well-preserved rock art.26 This exploration work has also collected data on other archaeological data recovered from surface scatters of ceramics and lithics and from excavations. Systematic survey and recording of the Daurês massif (Brandberg) on the eastern fringes of the Namib desert was carried out by Harald Pager over several years, and many parts of South Africa have been surveyed for rock art, such as the Makgabeng Plateau in Limpopo Province.27

Surveys range in scope from ongoing rock art reconnaissance using digital apps by enthusiasts to fixed-period, fully funded, multidisciplinary exercises, such as the Wadi Sura Project.28 Unmanned Aerial Vehicles (UAVs or drones) are being used to locate and record rock art and other archaeological features, especially in less easily accessible areas.29

Observations about the geology and geomorphology of the survey area are relevant to understanding the formation of the rock-art site, as are the dimensions of the landscape features in which the rock-art site is situated and the dimensions of the site itself, and, of course, its geographical coordinates. After several years of field work in the Dâures massif (Brandberg) in Namibia, Tilman Lenssen-Erz has proposed that the artists “categorized and organized” their surroundings at site level as well as the wider landscape, which he calls their Gestaltung.30 He proposes five basic “dichotomic relations” that structure how particular rock-art sites were used. The argument is complex, but the point is that Lenssen-Erz identifies various types of rock-art sites based upon their size, layout situation, and the amount of other archaeological artefacts present.

In the course of fieldwork on the eastern fringes of the Namib desert, John Kinahan identified two rock-art sites that may have been implicated in the work of solitary, itinerant shamans in the second millennium ad.31 By considering the spatial dimensions of the sites, specific paintings of female kudu, as well as radiocarbon dates, he has argued that these sites were used by shamans who were providing their services to various communities, including neighboring Ovaherero agriculturists.

There are also other less apparent landscape elements to consider. Some researchers have also investigated the intangible role that sound may play at certain rock-art sites and how these phenomena contributed to ritual performance. Rock gongs have been noted at several African rock-art sites, such as those in and around Lake Victoria Nyanza,32 as well as certain rock-art sites in the southern African interior.33 The question of whether the rock gongs were in contemporaneous use with the image-making activities at the same place is one that has to be investigated and argued in each case.

Others have investigated the role of resonances and echoes at rock-art sites. In order to investigate whether the reverberation and echoes of rhythmic clapping such as that performed at San trance dances may have contributed to their ambience, Riaan Rifkin measured the level of echoes at Klipbak 1 engraving site in Northern Cape Province, South Africa.34 Aron Mazel attributes the high number of sites and paintings in the Didima Gorge, KwaZulu-Natal, South Africa, (3,909 paintings in seventeen sites) in part to its acoustic properties.35 Kinahan identifies a measurable degree of resonance within an engraved area enclosed by several large boulders at /Ui-//aes (Twyfelfontein) in Namibia.36

Depending on the research aims it may be useful to make plan and section drawings of the sites that include the location of natural features, as well as the location of the rock-art imagery.37 Other possibly relevant information includes the orientation of the site as well as general observations of the effects of the elements (sun, wind, and rain or moisture) on the site itself as well as the images it contains.

Most researchers will want to “capture” the imagery in situ and then take the recordings away for further detailed study and analysis. Freehand sketching or photography in conjunction with verbal (written or dictated) notes are useful mnemonics to familiarize the researcher with their data. These sketches and photographs would include particularly interesting images and also diagrams showing the relationships between images, as well as inclusion of details about the natural rock surface, such as ridges, cracks, and hollows, especially instances in which the artists appear to have made use of these features to situate images and compositions of images on the rock face. In many cases the artists seem to have regarded the rock surface as a meaningful surface and not an inert or neutral background.38 It is therefore important that the recording method takes into consideration the intentional interrelationships created by the artists between the image and its situation on the rock surface. At the same time it is important to note details about the rock face itself, such as its color, texture, and porosity, as such variables affect the visibility and possibly the placement of the images.39 In addition, it would also be useful to look for and note other features of the rock surface, such as the presence of abiotic and biotic factors, for example, areas of salt deposition; rock exfoliation; water flows; the presence of plants, lichen, and moss on the rock surface; the presence of insect and bird nests; as well as the presence of animals (e.g., reptiles, hyraxes, porcupines, carnivores) that make use of features of the site for habitation. Anthropogenic factors such as graffiti and the use of the site could also be observed and noted at this stage.40

Measurements of the dimensions of the images and of the distances between images and clusters of images would also be recorded at this preliminary stage; more accurate and detailed data-gathering procedures may be carried out later. Depending on the scope and aims of the research and the funds available, subsequent fieldwork might involve numbering and mapping the images and using Geographical Information Systems (GIS) onto a local XYZ coordinate system using a total data station, which can be georeferenced into a universal coordinate system using a Global Positioning System.41 Site record forms have been used routinely since the 1990s in digital form in Android applications and computer software.42

In addition the occurrence of superpositioning of images (i.e., instances in which one painting partially or wholly covers another) and renovation of images can be noted. Superpositioning provides data about the relative chronology of the images and is therefore a fundamentally important method. In this way successive “styles” or “traditions” can be sequenced in time. Such relative chronologies are important as a line of evidence that can be compared with chronometric age ranges (see Relative Chronology). These details may provide data about the symbolic, narrative, and chronological content of the images. The Harris matrix, a technique originally developed to determine the sequence of layers of archaeological deposits, has been employed in complex cases of superpositioning.43

It is useful at this stage to use image-enhancement software such as DStretch, which can be used on portable iOS and Android devices, in order to clarify details that cannot be seen with the naked eye. These attributes can be incorporated into the field sketches. In the case of rock paintings, the colors used to depict the images would be noted, as the choice of colors by the artists and indeed the source of pigments from which the paint was made often have important symbolic connotations. Changes in the use of colors as well as patterns of weathering of particular “styles” may also be significant in the investigation of the chronology of the images. However, variation in the degree of desert varnish on engravings is not a reliable indication of relative age because the condition of the varnish may be due to a number of factors that are unrelated to the age of the art.44 For rock engravings researchers could observe and record data regarding the technique or techniques used to create the imagery (e.g., pecking, scraping, or incising).

Recording Techniques

Having familiarized oneself with the site and its imagery by observing, noting, and perhaps sketching the features described above, the next phase is to accurately record the imagery in context. Digital techniques such as high-resolution photography, photogrammetry, and laser scanning are now commonplace. Up until the early 21st century, however, and in some cases even now, researchers value redrawn copies of field tracings for purposes of study and publication. Research into the acoustics of rock-art sites is also being carried out.45 Rifkin describes the use of a digital recorder (Sony IC) to record the sound stimulus and processing and analyzing the results with the Adobe Audition 1.5 sound-editing program. Such quantitative sound-level measurements aim to document objectively the presence of significant sound reflection at rock-art sites.

Geographical Information Systems (GIS) have been used to document what Thembi Russell describes as the vertical and horizontal dimensions of rock-art data.46 On the large-scale horizontal plane, the distribution of rock-art sites can be plotted in a landscape, while on a smaller scale, the patterning of images can be determined and recorded in the vertical dimension. Russell suggests that the horizontal dimension can be used to integrate data regarding the distribution of archaeological sites with early sheep remains with the spatial patterning of rock-art sites that feature images believed to be made by Khoekhoe herders.47 These data may be used to argue for the routes by which sheep first entered what is now South Africa.

Tracing and Copying

The technique of on-site, field, direct, or contact tracing, as it is variously known, has for many years been a mainstay of recording rock art.48 As with any recording techniques, tracings are not a substitute for the original imagery; a tracing is an interpretation of the form of the imagery and the interrelationships of the images. The technique involves placing a transparent or semi-transparent medium (such as drafting film Ozatex) over the imagery and then attaching the tracing film to the rock surface using tape. Where necessary, sections of tracing medium can be joined so that all the imagery to be copied can be included. The outlines of the images are then carefully traced onto the film using either a pencil or fiber-tipped pen. Features of the rock such as cracks and depressions that the tracer considers to be in meaningful interaction with the imagery are also recorded. The choice of tracing instrument and film varies according to the intricacy of the images, their size, and the techniques used to create the images. Very detailed paintings require a fine tracing instrument such as a clutch pencil with a thin lead (e.g., 0.3 mm) to copy the details accurately.49 In these instances the tracing film needs to be able to take the pencil mark. A thin, single matt drafting film is suitable for the purpose. Larger or less-detailed painted images, and engravings in general, could be traced using fiber-tip pens of varying sizes and a clear plastic tracing film.

Perhaps the most important aspect of making accurate copies of rock art, regardless of the techniques employed, is that it forces the researcher to carefully examine and classify the images in detail, and it is these details that often provide key data in understanding the rock art.50 The three-dimensional nature of the rock surface, however, means that tracings (which are two-dimensional) introduce spatial distortions, though these are not necessarily critical to understanding the imagery.

Concerns have been raised about the impact that direct tracing may have on the conservation and integrity of the images. The technique is obviously not suitable for images that are on friable surfaces, because the contact between the tracing film and rock surface could dislodge portions of the painted rock, thereby destroying parts of the image. There is also the danger that the tracing process might destroy or damage imagery inadvertently as a result of misjudgment or ignorance on the part of the tracer. In some countries tracing can only be carried out under permit from heritage authorities, while in other countries or provinces or states within a country field tracing is not permitted at all. There are alternative tracing techniques available, however, including the use of computer software to create digital tracings.51

The tracing of engravings should also take into consideration the friability and overall condition of the rock surface before the technique is applied.52 The use of rubbing techniques to copy engravings, in which paper is placed over the engraved surface and then rubbed with graphite, is now regarded as too invasive because the rubbing action could damage the engravings. The making of casts of engravings, especially latex casts, is also invasive and no longer acceptable practice because it leaves traces and stains on the rock surface. Not all engravings can be easily discerned or recorded; fine incisions and scratches may only be visible when the light is at a certain angle and are extremely difficult to trace. Computational photography includes techniques such as Reflectance Transformation Imaging (RTI) that uses the mathematical enhancement of surface features to enable researchers to see details that would otherwise be invisible.53

Digital Image Enhancement

In the past, analogue photography was considered inadequate for documenting rock art at the level of detail and accuracy required for study. Tracing was considered more accurate.54 At that time photographic prints and color transparencies could not accommodate the level of detail that an on-site observer could detect with or without the benefit of magnification; light conditions and the angle from which the photograph was taken could also introduce illusory details. Technological advances in digital imaging have largely eliminated these distortions and inaccuracies. The resolution of professional-grade digital cameras used in conjunction with professional-grade lenses of various focal lengths, lighting equipment (e.g., flash), and basic photographic techniques make it possible to see more detail more clearly in the images themselves than in the original motif on the rock face. The use of macro lenses to photograph rock art images at life size further enhances our ability to see details that would escape the unaided eye.55

The use of a panoramic tripod allows the creation of montages of the rock surface as well as 180- and 360-degree panoramas.56 These panoramas capture the site and its surrounds and allow the viewer to experience it from varying angles and distances. Software (e.g., Photoshop, PTGui, Gigapan) facilitates the stitching of multiple images to create these 2D models.57

Video documentation of the site, its context, details of the imagery, and the activities of researchers, visitors, community members, and so on can be a valuable record.58

Image-processing software, such as Adobe Photoshop and Lightroom, together with enhancement software, particularly DStretch, are fast replacing field tracing.59 It is now possible to enhance images (especially those painted in shades of red) that would otherwise be very difficult if not impossible to discern with the naked eye. This combination of software packages and a variety of workflows also makes it possible for researchers to produce digitally enhanced renditions (“tracings”) of rock art, thus obviating the need for field tracing.60

Regardless of the methods used, errors in recording do occur.61 Copies may omit key details or misconstrue the imagery.62 Sometimes one recording technique can help to correct a mistake that is present in a copy made using another technique. For example, Francis Thackeray and Thembi Russell published a tracing of an antelope with what they argue to be two parallel wounds on its neck, a detail that they consider to be an important symbolic element of the painting.63 The use of DStretch on a digital image of the original painting showed that the “wounds” are an optical illusion.64

Tracing photography and video are 2D recording techniques. They capture the horizontal and vertical dimensions of the imagery, usually from a frontal or orthogonal perspective in order to reduce distortion of the images. By definition, 2D recordings lack depth; this limitation in the face of developing 3D techniques does not render two-dimensionality obsolete, however.65 The 2D perspective remains essential for studying details of imagery, performing enhancements, and creating diagrammatic tracings.

Researchers and heritage practitioners are now creating 3D and 4D models of rock art to create the feeling of virtual space, of “being there.” 4D models incorporate the fourth dimension, that of temporality, and can be used to show changes in the evolution of rock-art imagery over time. Digital photogrammetry and laser scanning are used to produce 3D models. Photogrammetry can now be carried out with a minimum of training and with relatively inexpensive camera equipment and software. Structure from Motion technology creates high-resolution digital elevation models (DEMS) from multiple overlapping images taken from many different angles.66 These are processed using software such as Agisoft PhotoScan. This technology has been used in the Western Sahara by Saharawians in concert with researchers from Córdoba University who recorded the Great Shelter of the Small Wadi west of Taffariti using Structure from Motion technology.67 In south-west Egypt an interdisciplinary team financed by the German Research Council has used a suite of digital techniques to document rock art at Wadi Sura II, the “Cave of Beasts,” on the edge of the Gilf Kebir plateau, including digital photography and 3D laser scanning.68

Developing a Common Analytical Strategy

When planning to analyze, date, and sample rock art it is advisable for researchers and specialist colleagues to develop a “common analytical strategy.”69 Developing a rationale for attempting an age determination is a first, vital step. What does the researcher want to date and why? How will the data be collected, and what dating methods will be used? What are the criteria according to which the results will be interpreted, and what will the results contribute to existing knowledge?

Materials scientists may have little understanding of the physical research environment and the archaeological nature of the rock art. Stephane Hoerlé and colleagues point out that “reciprocal misunderstanding may arise from practical constraints specific to each side that are not properly expressed or explained to the partner.”70 These misunderstandings need to be identified and resolved. The complexities of the materials analysis of rock art require a combination of analytical techniques that are used in parallel and often on the same sample. The “complementary use” of the various techniques and thus the involvement of several specialists requires close cooperation and clear communication.71

Analysis of Pigments

Paint used to make rock art may consist of inorganic pigments and organic binders, although inorganic pigments are more frequently detected in pigment analyses, possibly because the organic binders have been decomposed by ultraviolet (UV) radiation to undetectable levels.72 Inorganic pigments cannot be radiocarbon dated because they do not contain carbon, but even very small amounts of organic material present in paint can be detected and analyzed and, possibly, dated (see Cosmogenic Techniques). It is not currently possible to identify the origin of the organic material in paints, but fat, blood, egg whites, egg yolks, honey, milk, plant juice and resins, oils, and urine have all been proposed, some with ethnographic support.73 In Lesotho, Marion How reported an artist asking for the blood of a freshly killed eland with which to make paint.74

Shades of red, orange, brown, and yellow are very common in rock paintings and often derive from “oxide/hydroxide minerals of various oxidation states and degrees of hydration.”75 Pigments may be sourced from nearby outcrops (as at Wadi Sura II, the “Cave of Beasts”), while in other cases people obtained pigments from thirty to one hundred kilometers away.76

Black pigments could be manufactured from manganese oxide or hydroxide and, less frequently, from charcoal. Charcoal is eminently suitable for radiocarbon dating (with the exception of misleading age estimates obtained from “old wood” and “old charcoal”).77

Paint manufacture may have involved several processes, including the acquisition of pigments, processing of pigments by heating them, grinding, and possibly adding organic binders.78 There is ethnographic support in southern Africa that paint was itself considered to be a potent substance.79 Pigment studies use various techniques to characterize pigments in terms of the chemical composition. Recent analyses of naturally detached paint fragments were studied using optical microscopy, scanning electron microscope Energy Dispersive X-Ray Spectroscopy (EDS), Raman spectroscopy, and Fournier Transform Infrared Spectroscopy (FTIR) to investigate their paint stratigraphy, composition, and provenance.80 A study by Jeffrey Hughes and Anne Solomon characterized samples of ochres and other pigments in KwaZulu Natal, South Africa, in terms of their physical characteristics, and chemical and mineralogical properties using transmission and scanning electron microscopy with associated elemental analysis using X-ray diffraction.81

Non-invasive, portable Raman spectroscopic devices in South Africa and X-ray diffraction and energy dispersive X-ray (EDX) techniques in the Gilf Kebir in south-west Egypt were used to identify locations on the rock face with pigments and substrata that would provide suitable samples for laboratory analysis.82 Working at Wadi Sura II in the Eastern Sahara, Sabine Krause and colleagues used computer-integrated infrared spectrometry and Munsell soil color charts to define the color range of the art. Pigment color and chemical analysis was carried out in the laboratory using a variety of methods, including digital infrared spectroscopy (IRS), video microscopy, and X-ray powder diffraction (XRD).83

Working with Temporality

The dimension of time is immanent in all archaeological artefacts; without chronological data rock art cannot be directly related to other archaeological artefacts. Miles Burkitt, writing in the 1920s on southern African rock art, proposed that rock paintings be studied in the same way as stone tools—that is, in terms of stratigraphic layers (corresponding to the superpositioning of images on a rock surface) and typology (i.e., style, or manner of depiction).84 The objective was to compile an overall sequence of painting styles based on superimpositions in the art; this sequence could then be related to the temporally corresponding excavated layers of occupation at rock-art sites. The achievement of this goal has proved to be more elusive than anticipated and continues to challenge researchers.85

The problem is to develop a shared chronology between the art on the rock surface and the other archaeological data, especially the stratigraphic layers at the foot of the painted or engraved rock surface. The art may have been made in minutes, hours, days, or weeks; to further complicate matters, rock surfaces at some sites are marked by repeated episodes of image making. In this way many isolated episodes of image making of unknown duration are built up on the rock surface. Excavated data, however, typically accumulate relatively slowly. It may take millennia for a few centimeters of deposit to accumulate and it is difficult to distinguish time periods of less than a few hundred years within the deposit.86 Thus, while it is known, for example, that in the prehistory of the Western Cape, South Africa, “the people who painted also gathered shellfish and plant foods, hunted and fished, established and maintained kinship and other links . . . ,” it is often difficult to obtain chronological links between the art on the rock surface and the accumulated layers at its foot.87

Moreover, the painted and engraved rock surface is an active interface between the rock interior and the ambient atmosphere.88 Evapotranspiration, deposition, and weathering are active weathering processes. The rock surface is thus the site of (largely poorly understood) biological and biochemical activity. Paintings may become encapsulated by accretions that prevent their erosion but which also make it very difficult to separate the paint from the accretion.89 The probability of contamination of paint samples must be contended with by following protocols such as taking multiple samples of the material to be analyzed as well as samples of unpainted rock for comparative processes.

Such is the importance of time that in the past rock-art research was confined to the margins of archaeological research, because contemporary researchers were unable to determine the age of rock art directly, and also because it was generally believed that the meaning of the art was beyond reach.90 The “patchiness” of rock-art data means that these problems remain fundamental to the conduct of rock-art research, but important advances in chronological techniques and interpretative research have nonetheless been made. By far the oldest and still important and valuable method of investigating and establishing chronology in rock art is relative chronology.

Relative Chronology

Relative chronology gives an indication of the age of an artefact in relation to its position with another artefact (see Correlative Techniques). Superpositioning, the placement of an image on top of another image, provides the most fundamental measure of chronological sequence. The lowest image is by definition older than the overlying image, although, crucially, the interval of time between image-making episodes is usually not known. Regional chronologies are compiled from data from all recorded sites in a defined area.

A chronological sequence has been constructed for South African rock engravings based upon styles based on technical methods of execution used (either “hairline” [incised], or pecked) and subject matter (“geometrics” and “animal representations” of different styles).91 Karl Butzer and colleagues were unable to establish if the hairline engravings were an earlier phenomenon dating to around 7000 to early 6000 bp. They found that the various styles of animal engravings “relate” to the Smithfield stone tool industry between 4000 and 3000 bp. Most of the geometric engravings are more recent and coincide with the appearance of pottery in the last 1,300 years.92

In some situations, such as the Wadi Sura in the Gilf Kebir in south-western Egypt, Wadi Sura–type paintings appear much more weathered than the subsequently painted cattle pastoralist images, an observation that suggests that the interval between the two traditions was perhaps thousands of years.93 Relative chronology derived from the analysis of superimpositioning of rock-art “styles” or “traditions” should ideally be evaluated with regard to age estimates obtained using dating techniques such as Accelerator Mass Spectrometry (AMS) radiocarbon dating (see Cosmogenic Techniques).

Certain sites in a particular region may be critical in determining the succession of styles in a particular geographical area; these “key” sites are particular rock-art surfaces with superpositions of various styles of art that enable their chronological succession to be established. For example, András Zboray identifies site KTW 21/Q in the Jebel Ouenat as a key locality, because here cattle pastoralist paintings are found superimposed over a giraffe hunting scene attributed to the Miniature style, which in turn overlap larger “Ouenat roundhead” figures.94 The particular sequence is thus taken to hold for the sequences in general (until proof to the contrary is found).

Zboray has proposed a sequence for the rock art of the Eastern Sahara based on superimpositions.95 Chronologies based on relative dating, correlative methods, and corroborated with radiocarbon estimates can be incorporated with other archaeological data to establish a “history” for the rock art of an area.96

The Harris Matrix

Originally developed to determine superpositional relationships in archaeological stratigraphy, the Harris Matrix has been applied to rock-art research to keep track of rock art superpositions.97 The technique works by identifying “stratigraphically associated motifs” which “are grouped into layers which themselves form sequences.”98 Proprietary Harris Matrix Composer software is available to construct the sequences, and the development of Geographical Information Systems (GIS) and 3D laser scanning has alleviated somewhat the painstaking, tedious, and time-consuming processes of recording and representing the imagery.99 These relative chronologies may then be related to 14C age determinations.100 In this way two independent lines of evidence are used to arrive at a more secure age estimate.

The Harris Matrix has been used in South Africa to establish relative sequences in parts of the uKhahlamba-Drakensberg, but nothing seems to have been published on its use elsewhere in Africa.101 Joané Swart reports that the sequences recorded at two uKhahlamba-Drakensberg sites correlate with some of the age determinations obtained by Aron Mazel and Alan Watchman.102 Presently, the Harris Matrix has been used at single sites, but the method could be extended to develop regional rock art sequences.103

The vulnerability of this powerful method lies in the formulation and definition of categories used to classify the painting sequences.104 Researchers have to formulate these based on criteria that they have devised: if the criteria imposed fail to place an image in the correct part of the sequence of superpositions, the matrix will not accurately reflect the actual relative sequence. Thus the method is most effective when dealing with what appear to be very different-looking and thus relatively easily defined traditions (e.g., the fine-line naturalistic paintings of the Western Cape, South Africa, that are overlain by handprints and finger-painted art).105 It is far more difficult to identify changes in style within a single tradition in a restricted area.106 It should also be noted that the sequence of painting or engraving on a surface of itself provides little if any data about the time interval between periods of image making. It has also been demonstrated that in some cases superpositioning may have been carried out more or less contemporaneously.107

Dating techniques provide information about the age of an artefact sample in various ways. The notion of calendrical or solar time that is so familiar does not apply to the temporality of rock art, which is measured in terms of a numerical rate of change in the condition of the artefact.108 The notion of an “absolute date” is inappropriate.109 Dating researchers speak instead of “dating” and “age estimates.” They urge the use of multiple independent lines of evidence and the use of different analytical techniques and different laboratories to reach an age estimate.110

Echoing the necessity for the development of a common strategy between archaeologists and other specialists, Karen Steelman and Marvin Rowe remind researchers that they should be cautious, even skeptical, of age estimates provided by specialists and that these results are only as good as the samples from which the date ranges are obtained.111

Correlative Techniques

Age determinations obtained by correlating the images with events for which age ranges are independently available are especially useful. Such events could be geological in scale: based on estimates about the rate of exposure of the engraved substrate by geomorphological processes, David Morris suggests that the age range for the Driekopseiland engraving site on the Riet River, Northern Cape Province, South Africa, falls within the past 2,500 years.112

Alternating layers of rock varnish formed in wetter and drier periods can be calibrated with the help of independent records of climatic shifts.113

Minimum and maximum ages of spalls of rock art can be determined by correlating the presence of these remnants in excavated, stratified layers of deposit at rock-art sites and in which charcoal deposits have been radiocarbon dated. In such cases it is important to establish the stratigraphic layers containing the rock art and the charcoal have not been disturbed by subsequent activity (e.g., by burrowing animals, erosion, or human activity), as such disturbances will alter or destroy the relationship between the rock art and the charcoal.114 There are several examples of this kind of correlative dating in southern Africa.115 Erich Wendt obtained age estimates of between 27500 and 25000 years bp for “art mobilier” recovered from a “Middle Stone Age Layer” at Apollo 11 Cave in southern Namibia, the oldest date yet obtained for rock paintings in Africa.116 Work by Francis Thackeray and colleagues describes radiocarbon dates associated with engraved stones from Wonderwerk Cave, Northern Cape Province, that suggest a minimum age of at least ten thousand years old.117

The depiction of extinct animals (such as the engravings of an extinct bovid, Bubalus antiquus, at Messak Settafet in the Libyan Desert), or the appearance of domesticated animals in the rock art of the Sahara (e.g., horses and camels) and in southern Africa (horses, sheep, and cattle), or items of technology (e.g., chariots in north Africa, trains in South Africa) and material culture (e.g., axes, shields in southern Africa) in a body of rock art may also usefully be correlated with supporting ages and historical events.118

Cosmogenic Techniques

Techniques that measure the decay of radiocarbon 14C isotopes119 and other cosmogenic nuclides of 3He, 10Be, 14C, 26Al, and 36Cl provide numerical values for the rate of decay of the radioactive molecules,120 which are converted to calendrical (solar) age ranges by means of calibration curves, often based on tree ring sequences.121,122 The development of Accelerator Mass Spectrometry (AMS) has enabled researchers to use carbon-containing samples as small as one millionth of a gram for 14C analysis.123 Interestingly, the first direct radiocarbon dating of a rock painting was determined on African rock art.124

Paint samples of between one hundred millionths of a gram to 1 mg are taken, and it is preferable to take several smaller samples from various parts of the rock painting. Where possible, samples are taken from areas of painted surface that are already spalling. Samples (including part of the underlying rock) are collected using a sterile surgical blade and rubber gloves are worn. Samples are wrapped in foil and bagged. Unpainted rock surfaces immediately next to the paintings are taken to establish the background levels of organic contamination.125

Various techniques have been developed to make sure that only the organic material associated with the paint is targeted for analysis. Improved techniques for developing pretreatment protocols for dating black pigment have been developed in research in South Africa.126

Recent AMS radiocarbon dates have been obtained using carbon blacks from short-lived organic material on different samples from Later Stone Age rock-art sites in Botswana, Lesotho, and South Africa; it is reported that paintings from south-eastern Botswana were made as early as 5723–4420 cal. bp.127

In what is believed to be the first direct dating attempt of Saharan rock art, Mori and colleagues used acid hydrolysis to separate amino acid content from the rock substrate of a paint sample at Lancusi Shelter (south-west Libya) before using AMS to obtain an approximate age (6145 ± 70 bp).128 This age range is consistent with the archaeology and climatic reconstructions of the central Sahara.

AMS radiocarbon dating of the calcium oxalate layers that form underneath and above pigment layers has enabled researchers to establish maximum (2900 bp) and minimum (2300 bp) ages for rock paintings.129

Dating Engravings

In contrast to dating rock paintings, in which the aim is to estimate the time of application of paint to the rock, rock engravings are dated from the time at which the rock surface was exposed by pecking, incising, or rubbing.130 Exposed boulders have a natural rock coating or “rock (or ‘desert’) varnish . . . a dark, paper-thin rind, consisting of approximately two thirds clay minerals that are biogeochemically cemented to the underlying rock surface by about one-fifth manganese, iron oxyhydroxides and other minerals accreted to the rock surface” that accumulates in a sequence of layers.131 The chemical composition varies regionally. Varnish is internally layered through microdepositional processes that are correlated with climatic shifts. This signature layering is used as the basis of most chronometric dating techniques. Because the accumulation of rock varnish is influenced by the local geology, the calibration of rates of chemical changes in the rock varnish must be done regionally.

All dating methods for engravings are “experimental” according to Dorn because of the complex and dynamic state of rock varnish.132 Biological activities, environmental factors, and major climate shifts influence the microchemistry of the rock varnish. The microchemistry may vary in composition across a rock face, nor is varnish growth evenly distributed, making sampling difficult.

The two principal methods of dating engravings are Cation Ratio (CR) dating, and Varnish Microlamination (VML) dating; AMS radiocarbon dating may also be used.133

CR dating relies on the differential rate at which some mobile trace elements in the rock varnish (in particular, potassium and calcium) leach out more rapidly than others, such as titanium. A calendrical age estimate is calculated through a regional calibration curve. The lower the CR value obtained, the older the specimen. Sampling involves taking a few samples from the engraving and using the average value to calculate the CR value. This practice means that CR dates have wide standard deviations and are therefore “approximations rather than absolute dates.”134 CR dating at Klipfontein engraving site in South Africa’s Northern Cape Province yielded CR values ranging from 9,02 to 0,59—that is, from about one hundred years ago to 10000 bp.135 The results also contained age ranges for each of the three styles of engravings noted.136

VML dating estimates also yield very general age ranges. They are based on the identification of regional varnish microstratigraphic signatures that result from climate changes. Microstratigraphic layers of orange-colored and manganese-poor varnish may be inter-bedded with black, manganese-rich layers that correspond to major climatic shifts from drier to wetter periods, respectively. Work by Mauro Cremaschi in the Messak Settafet plateau (Fezzan, Eastern [Libyan] Sahara) describes three microlayers of rock varnish, each of which formed during different climatic regimes.137 Combining these VML data with radiocarbon dating of archaeological deposits and data regarding Holocene climatic changes derived from cave deposits, Cremaschi suggests that most of the varnish developed between 6000 bp and 4000 bp “under a semiarid climate characterised by steppe-type vegetation, up to the onset of desert conditions.”138

AMS radiocarbon dating has been tested on organic particles encapsulated in the rock varnish based on the assumption that this carbon-containing material dates to the time at which the rock varnish began to develop.139 There is, however, a danger that the organics might turn out to predate the time of varnish formation or that the varnish might be contaminated by more recent organic material.140

It is important to evaluate published dating results in terms of various criteria in order to assess their reliability. These include the provision of an archaeological rationale for sampling, thorough recording of the site prior to any sampling activities, descriptions and visual documentation of the sampled image (including the number of samples and their respective weights), description of the sampling techniques used, and description of the pretreatment regimes used. In the case of AMS radiocarbon dating, radiocarbon laboratory identification numbers and raw radiocarbon dates with ±1 uncertainty should be supplied. It should also be specified whether these dates were corrected for 13C, calibrated dates and the software used to obtain these, and whether the intercept or Bayesian statistics methods were used. Finally, unsuccessful results should also be reported, detailing how many samples were rejected and the reasons for their rejection.141

Final Comments

Rock-art research methods as defined here have been restricted to discussion and evaluation of methods of documentation and contextualizing rock art in chronological order. The theoretical frameworks of interpretation that should inform the use of these methods in the practice of research have not been discussed. Because the techniques used to document and estimate the age of rock art are often adapted from developments in other areas of applied technology, it is to be expected that innovation and development in digital technologies and their application to documentation, mapping, analysis, and chronometry will continue. However, the application of technologically advanced techniques in documenting and dating of rock art—critical as these are in themselves for the archiving and conservation of rock art—needs to be appropriate to the primary research questions posed about this inherently meaningful resource of our prehistoric past.

Discussion of the Literature

The origins of African rock-art research methods date from the mid-19th century.142 With the indispensable assistance of local populations, rock-art enthusiasts explored parts of the continent for rock art. The main concentrations of rock art in Africa are in the Saharan regions, the Horn of Africa, central and eastern Africa, and southern Africa.143 Rock art in Africa is only absent from forested areas, very high altitudes, and of course from areas without suitable exposed rock surfaces. Parts of north Africa and southern Africa have been far more intensively and extensively investigated, while other parts of the continent, especially the central portion of sub-Saharan Africa, north of the Zambezi River, have been under-recorded.

Copies, often watercolors, and of varying degrees of accuracy, were initially the principal method of recording rock art, and dedicated researchers travelled vast distances under difficult conditions in order visit sites.144 The adoption of photography in the early 20th century was an important innovation that researchers soon adopted,145 although accurate field copies (but not tracings) remained the mainstay of rock art recording until the development of translucent films made it possible to make field tracings (also known as contact tracings because the tracing film is placed over the images to be copied). Since the 1960s and up until the early 21st century researchers have used various tracing methods in the field to produce copies that are accurate, detailed, and to scale. Photography assumed much greater importance with the availability of color slide film in the late 1960s, but until the introduction of digital imaging techniques, field tracings were regarded as the best possible method of capturing details of rock art, and still are by some. Harald Pager’s landmark book Ndedema used a unique technique he developed in which he worked on site, making use of oil-paints to color life-size black and white photographs of the rock art of the Didima Gorge.146

Until the mid-1960s, and even in the early 21st century, research into rock art has been considered as a challenge, because it was believed that the art could not be dated, nor could its meaning be recovered. Rock-art data could therefore not contribute to an understanding of prehistoric people beyond analysis of the subject matter of the art for “illustrations” of their technology and economic and social activities. Surveying and recording of rock art was regarded as the province of the amateur enthusiast, while the very few archaeologists in employment at the time focused on more tractable research problems. Interpretations of the art were regarded as a free-for-all, with various assessments of the meaning and significance of the art. These included notions of sympathetic magic, depictions of scenes from daily life, and paintings of animals assumed to be menu items.147

From the 1960s, archaeologists interested in rock art attempted to explore rock art using methods based on archaeological excavation, developing classificatory categories according to which data collected during systematic surveys could be analyzed and used to develop a cultural-historical sequence that could be correlated with other archaeological data obtained by means of excavation.148

Surveys of areas for rock art and the meticulous recording, classification, and analysis of these data revealed that image makers worked according to conventions, such as preferences for particular animal species and repeated motifs of human activities. At this stage certain researchers realized that the only way to understand these data was to reconstruct the significance of the images themselves.149 Significantly, this development occurred in southern Africa, where researchers had the “luxury of ethnography” in the form of extensive 19th- and 20th-century testimony by Khoe-San people.150 James David Lewis-Williams in particular began to contextualize motifs in the rock paintings of the Maloti-Drakensberg, especially, but not only, of eland in terms of this ethnography.151 He identified characteristic motifs in the iconography of this area that relate to the performance of healing rituals and out-of-body travel. His ethnographic approach has been applied to the fine-line hunter-gatherer art tradition south of the Zambezi River to suggest that this rock-art tradition may be understood as metaphors and manifestations of supernatural potency produced within a shamanistic worldview.152

Lewis-Williams’s dissatisfaction with theoretically uninformed research methods and his consequent focus on investigating the overarching metaphors in southern African hunter-gatherer art meant that other archaeologically important aspects—style and chronology—received less attention in his work.153 Nonetheless researchers within what has been called the “shamanist paradigm” have incorporated agency, gender, and change in their research.154 In the meantime, dating technologies have developed greatly since the 1970s. Style and chronology are important in order to place the art in a historical context. Interestingly, in other parts of Africa, especially in the Sahara, researchers have been very reluctant to use ethnography and analogy to investigate how rock art relates to the social-relational matrix of which it is part. Here researchers have expended considerable effort and expertise on recording the rock art and compiling stylistic chronologies. Similarly, work by Pager in southern Africa, including his prodigious documentation of the Dâures massif (Brandberg) in Namibia, consciously eschews interpretation.155 It is as though two separate approaches have developed over the past forty years: one approach (inspired by Lewis-Williams) that prizes interpretation, and the other approach that focuses on the compilation of chronologies based on superimpositioning of images and attempts to relate the chronology of the rock with other archaeological and historical elements. These two approaches, however, are not mutually exclusive.

In many parts of Africa there is still a need for basic survey and documentation of rock-art sites with a view to attributing authorship of the art and constructing their significance. Obviously technological (and especially digital) applications to aid in documentation and dating of rock art will continue to develop apace. Any methods of researching rock art will of necessity involve the theoretical frameworks that inform researchers’ choices of methods.

Primary Sources

For an extensive overview of theoretical frameworks, methods, and regional analyses of world rock art, the two standard reference works are the Handbook of Rock Art Research, edited by David Whitley, and Jo McDonald and Peter Veth’s edited volume, A Companion to Rock Art.156

Several collections of essays on interpretation, documentation, and method concerning rock art, mostly from Africa, Australia, Scandinavia, and North America, have been published in the past twenty years. These include Christopher Chippindale and Paul S. C. Taçon’s The Archaeology of Rock Art,157 Christopher Chippindale and George Nash’s The Figured Landscapes of Rock Art,158 and Benjamin Smith, Knut Helskog, and David Morris’s Working with Rock Art.159

As an introduction to African rock-art traditions, Alec Willcox’s The Rock Art of Africa160 remains the most comprehensive and illustrated overview of the art traditions of the continent. Alec Campbell and David Coulson’s African Rock Art: Paintings and Engravings on Stone161 is a richly illustrated selection of rock paintings and engravings from north, east, and southern Africa.

The recent documentation of Wadi Sura II (“Cave of Beasts”)162 is an example of a comprehensive and technologically sophisticated documentation of an extraordinary rock-art site in the Gilf Kebir, in south-west Egypt. Several essays describe various aspects of the methods used to document and to date the rock art.

Developments in technological applications for surveying, documenting, dating, and analyzing rock art are constantly being published in a wide range of journals concerned with archaeological science, digital imaging techniques, Geographical Information Systems (GIS), analytical methods, and other specializations. A special issue, “Digital Imaging Techniques for the Study of Prehistoric Rock Art” has been published in Digital Applications in Archaeology and Cultural Heritage.163

Journals specializing in cultural heritage management and conservation are sources of information about new methods and techniques that may have application to rock-art research.

The African Rock Art Digital Archive.

A database of several collections of rock-art color transparencies (i.e., slides), photographs, drawings, and texts.

African Rock Art Image Project.

A catalogue of approximately twenty-three thousand digital photographs of African rock art obtained from the Trust for African rock art and hosted by the British Museum.

iDAI.objects Arachne.

The African Archaeology Archives (University of Cologne, Germany) include rock art documented in Namibia (since 1963) and north Africa (during the 1980s) in the form of color transparencies (i.e., slides), photographs, drawings, and texts.

Oxford Handbooks Online: Archaeology.

A collection of articles and books on a wide range of archaeological topics, including The Oxford Handbook of The Archaeology and Anthropology of Rock Art, edited by Bruno David and Ian J. McNiven.

Further Reading

Bruno David, John Brayer, Ian McNiven and Alan Watchman, “Why Digital Enhancement of Rock Art Works: Rescaling and Saturating Colours.” Antiquity 75, no. 290 (2001): 781–792.Find this resource:

    Harris, Edward, and Robert Gunn. “The Use of Harris Matrices in Rock Art Research.” In The Oxford Handbook of The Archaeology and Anthropology of Rock Art. Edited by Bruno David and Ian J. McNiven. New York: Oxford University Press, 2017.Find this resource:

      Lewis-Williams, J. David. “Introductory Essay: Science and Rock Art.” South African Archaeological Society Goodwin Series 4 (1983): 3–13.Find this resource:

        Lewis-Williams, J. David. “Documentation, Analysis and Interpretation: Dilemmas in Rock Art Research.” South African Archaeological Bulletin 45 (1990): 126–136.Find this resource:

          Mazel, Aron. “Images in Time: Advances in the Dating of Maloti-Drakensberg Rock Art since the 1970s.” In The Eland’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, 81–97. Johannesburg: Wits University Press, 2009.Find this resource:

            Pearce, David G. “Changing Men, Changing Eland: Sequences in the Rock Paintings of Maclear District, Eastern Cape, South Africa.” American Indian Rock Art 28 (2002): 129–138.Find this resource:

              Pearce, David G. “A Comment on Swart’s Rock Art Sequences and Use of the Harris Matrix in the Drakensberg.” Southern African Humanities 18, no. 2 (2006): 173–188.Find this resource:

                Rudner, Ione. “Khoisan Pigments and Paints and Their Relationship to Rock Paintings.” Annals of the South African Museum 87 (1982): 1–281.Find this resource:

                  Russell, Thembi. “The Application of the Harris Matrix to San Rock Art at Main Caves North, KwaZulu-Natal.” South African Archaeological Bulletin 55 (2002): 60–70.Find this resource:

                    Russell, Thembi. “No One Said It Would Be Easy. Ordering San Paintings Using the Harris Matrix: Dangerously Fallacious? A Reply to David Pearce.” South African Archaeological Bulletin 67, no. 196 (2012): 267–272.Find this resource:

                      Swart, Joané. “Rock Art Sequences in uKhahlamba-Drakensberg Park, South Africa.” Southern African Humanities 16 (2004): 13–35Find this resource:

                        Swart, Joané. “Harassing the Matrix: A Reply to Pearce.” Southern African Humanities 18, no. 2 (2006): 178–180.Find this resource:

                          Vinnicombe, Patricia. People of the Eland: Rock Paintings of the Drakensberg Bushmen as a Reflection of Their Life and Thought. Pietermaritzburg: University of Natal Press, 1976.Find this resource:


                            (1.) Though the images may have been pecked, incised, or scraped.

                            (2.) Tim Ingold, The Perception of the Environment: Essays on Livelihood, Dwelling and Skill (London and New York: Routledge, 2000), 131. See also Alfred Gell, Art and Agency: An Anthropological Theory (Oxford: Clarendon Press, 1998).

                            (3.) Solveig Turpin, “Archaic North America,” in Handbook of Rock Art Research, ed. David S. Whitley (Walnut Creek, CA: AltaMira Press, 2001), 404.

                            (4.) William Challis, Peter Mitchell, and Jason Orton, “Fishing in the Rain: Control of Rain-Making and Aquatic Resources at a Previously Undescribed Rock Art Site in Highland Lesotho,” Journal of African Archaeology 6, no. 2 (2008): 203–218.

                            (5.) Gell, Art and Agency, 6.

                            (6.) Gell, Art and Agency, 6.

                            (7.) André Rosenfeld, “A Review of the Evidence for the Emergence of Rock Art in Australia,” in Sahul in Review: Late Pleistocene Archaeology in Australia, ed. Mike Smith, Matthew Spriggs, and B. Fankhauser (Canberra: Australian National University, 1993), 71–80; André Rosenfeld, “Rock Art and Rock Markings,” Australian Archaeology 49 (1999): 28–33; Josephine Flood, Rock Art of the Dreamtime: Images of Ancient Australia (Sydney: Angus & Robinson, 1997); and Josephine Flood, “Linkage between Rock Art and Landscape in Aboriginal Australia,” in Pictures in Place: The Figured Landscapes of Rock Art, ed. Christopher Chippindale and George Nash (Cambridge, UK: Cambridge University Press, 2004), 182–200.

                            (8.) Flood, Rock Art of the Dreamtime, 148–149; Flood, “Linkage between,” 91; and Rosenfeld, “Rock Art.”

                            (9.) Jeremy C. Hollmann, “Exploring the Gestoptefontein-Driekuil Complex (GDC): An Ancient Women’s Ceremonial Centre in North West Province, South Africa,” South African Archaeological Bulletin 68, no. 198 (2013): 156.

                            (10.) See, for example, James David Lewis-Williams, “Debating Rock Art Research: A Reply to Butzer,” South African Archaeological Bulletin 46 (1991): 49.

                            (11.) See Mark McGranaghan, “Ethnographic Analogy in Archaeology: Methodological Insights from Southern Africa,” in Oxford Research Encyclopedia of Africa History (Oxford University Press, 2017) for a discussion of the use of ethnographic analogy in archaeology and rock art interpretation.

                            (12.) James David Lewis-Williams, “Wrestling with Analogy: A Methodological Dilemma in Upper Palaeolithic Art Research,” Proceedings of the Prehistoric Society 57, no. 1 (1991): 160.

                            (14.) Phil Bonner, Amanda B. Esterhuysen, and M. H. Alex Schoeman, “Introduction,” in Five Hundred Years Rediscovered: Southern African Precedents and Prospects, ed. Natalie Swanepoel, Amanda Esterhuysen, and Phil Bonner (Johannesburg: Wits University Press, 2008), 16.

                            (15.) See Peter Mitchell, “Introduction to Archaeological Methods and Sources,” in Oxford Research Encyclopedia of Africa History (Oxford University Press, 2017).

                            (16.) For example, Julie E. Francis, “Style and Classification,” in Handbook of Rock Art Research, ed. David S. Whitley (Walnut Creek, CA: Altamira Press, 2001), 221–244; and Inés D. Sanz, “A Theoretical Approach to Style in Levantine Rock Art,” in A Companion to Rock Art, ed. Jo McDonald and Peter Veth (Chichester: Blackwell, 2012), 306–321.

                            (17.) For example, Francis, “Style and Classification,” 221–222.

                            (18.) Francis, “Style and Classification,” 222.

                            (19.) William Challis, “Taking the Reins: The Introduction of the Horse in the Nineteenth-Century Maloti-Drakensberg and the Protective Medicine of Baboons,” in The Eland’s People: New Perspectives in the Rock Art of the Maloti-Drakensberg Bushmen; Essays in Memory of Patricia Vinnicombe, ed. Peter Mitchell and Benjamin Smith (Johannesburg: Wits University Press, 2009), 104–107; William Challis, “Creolisation on the Nineteenth-Century Frontiers of Southern Africa: A Case Study of the AmaTola ‘Bushmen’ in the Maloti-Drakensberg,” Journal of Southern African Studies 38, no. 2 (2012): 265–280; and William Challis, “Binding Beliefs: The Creolisation Process in a ‘Bushman’ Raider Group in Nineteenth-Century Southern Africa,” in The Courage of //Kabbo: Celebrating the 100th Anniversary of the Publication of Specimens of Bushman Folklore, ed. Janette Deacon and Pippa Skotnes (Cape Town: UCT Press, 2014), 247–264.

                            (20.) Sven Ouzman, “The Magical Arts of a Raider Nation: Central South Africa’s Korana Rock Art,” South African Archaeological Society Goodwin Series 9 (2005): 101–113.

                            (21.) Adrian Flett and Penny Letley, “Style and Stylistic Change in the Rock Art of the Southeastern Mountains of Southern Africa,” South African Archaeological Bulletin 68, no. 197 (2013): 3–14; and Aron Mazel and Alan Watchman, “Dating Rock Paintings in the uKhahlamba-Drakensberg and the Biggarsberg, KwaZulu-Natal, South Africa,” Southern African Humanities 15 (2003): 59–73.

                            (22.) Christopher Chippindale and George Nash, “Pictures in Place: Approaches to the Figured Landscapes of Rock-Art,” in Pictures in Place: The Figured Landscapes of Rock Art, ed. Christopher Chippindale and George Nash (Cambridge, UK: Cambridge University Press, 2004), 1–36.

                            (23.) Harald Pager, Ndedema: A Documentation of the Rock Paintings of the Ndedema Gorge (Graz: Akademische Druck, 1971); Aron Mazel, “Time, Color, and Sound: Revisiting the Rock Art of Didima Gorge, South Africa,” Time and Mind: The Journal of Archaeology, Consciousness and Culture 4, no. 3 (2011): 283–296; Harald Pager, The Rock Paintings of the Upper Brandberg, Part I: Amis Gorge (Cologne: Heinrich-Barth Institut, 1989); Harald Pager, The Rock Paintings of the Upper Brandberg, Part II: Hungorob Gorge (Cologne: Heinrich-Barth Institut, 1993); Harald Pager, The Rock Paintings of the Upper Brandberg, Part III: Southern Gorges (Cologne: Heinrich-Barth Institut, 1995); Harald Pager, The Rock Paintings of the Upper Brandberg, Part IV: Umuab and Karoab Gorges (Cologne: Heinrich-Barth Institut, 1998); Harald Pager, The Rock Paintings of the Upper Brandberg, Part V: Naib Gorge (A) and the Northwest (Cologne: Heinrich-Barth Institut, 2000); and Harald Pager, The Rock Paintings of the Upper Brandberg, Part VI: Naib (B), Circus and Dom Gorges (Cologne: Heinrich-Barth Institut, 2006). Rudolph Kuper, Wadi Sura: The Cave of Beasts (Cologne: Heinrich-Barth Institut, 2013); Geoffroi Heimlich, Le massif de Lovo, sur les traces du royaume de Kongo (Oxford: Archaeopress, 2017); and John Kinahan, “The Solitary Shaman: Itinerant Healers and Ritual Seclusion in the Namib Desert during the Second Millennium AD,” Cambridge Archaeological Journal 27, no. 3 (2017): 553–569.

                            (24.) For example, André Rosenfeld, “Recording Rock Art: A Conflict of Purpose?” in Conservation of Rock Art: Proceedings of the International Workshop on the Conservation of Rock Art, ed. Colin Pearson (Sydney: The Institute for the Conservation of Cultural Material, 1978), 9–14.

                            (25.) Rosenfeld, “Recording Rock Art,” 10.

                            (26.) For example, Frank Förster, Heiko Riemer, and Rudolph Kuper, “The ‘Cave of Beasts’ (Gilf Kebir, SW Egypt) and Its Chronological and Cultural Affiliation: Approaches and Preliminary Results of the Wadi Sura Project,” in The Signs of Which Times? Chronological and Palaeoenvironmental Issues in the Rock Art of Northern Africa (Brussels: Royal Academy for Overseas Sciences, 2010), 197–216; Heiko Riemer, Franziska Bartz, and Sabine Krause, “New Rock Art Sites in the Gilf Kebir, SW Egypt: A Review of Recent Results from the Wadi Sura Survey 2009–2011,” Sahara 24 (2013): 7–26; András Zboray, “A Revision of the Identified Prehistoric Rock Art Styles of the Central Libyan Desert (Eastern Sahara) and Their Relative Chronology,” in The Signs of Which Times? Chronological and Palaeoenvironmental Issues in the Rock Art of Northern Africa (Brussels: Royal Academy for Overseas Sciences, 2012), 217–255; and András Zboray, “Wadi Sura in the Context of Regional Rock Art,” in Wadi Sura: The Cave of Beasts, ed. Rudolp Kuper (Cologne: Heinrich-Barth Institut, 2013), 18–23.

                            (27.) Pager, The Rock Paintings, Part I; Pager, The Rock Paintings, Part II; Pager, The Rock Paintings, Part III; Pager, The Rock Paintings, Part IV; Pager, The Rock Paintings, Part V; Pager, The Rock Paintings, Part VI; and Edward B. Eastwood, Johan Van Schalkwyk, and Benjamin Smith, “Archaeological and Rock Art Survey of the Makgabeng Plateau, Central Limpopo Basin,” The Digging Stick 19, no. 1 (2002): 1–3.

                            (28.) Janette Deacon, Nicholas Wiltshire, and Rika du Plessis, “Designing Digital Recording for Volunteers in Rock Art Surveys, Management Plans and Public Outreach in the Cederberg, South Africa,” African Archaeological Review 35, no. 2 (2018): 225–239; and Kuper, Wadi Sura.

                            (29.) Stephen Berquist et al., “A New Aerial Photogrammetric Survey Method for Recording Inaccessible Rock Art,” Digital Applications in Archaeology and Cultural Heritage 8 (2018): 46–56.

                            (30.) Tilman Lenssen-Erz, “The Landscape Setting of Rock-Painting Sites in the Brandberg (Namibia): Infrastructure, Gestaltung, Use and Meaning,” in Pictures in Place: The Figured Landscapes of Rock-Art, ed. Christopher Chippindale and George Nash (Cambridge, UK: Cambridge University Press, 2004), 131–150.

                            (31.) Kinahan, “The Solitary Shaman,” 553–569

                            (32.) George Jackson, J. Stephen Gartlan, and Merrick Posnansky, “Rock Gongs and Associated Rock Paintings on Lolui Island, Lake Victoria, Uganda: A Preliminary Note,” Man 65 (1965): 38–40; and Merrick Posnansky, Andrew Reid, and Ceri Ashley, “Archaeology on Lolui Island, Uganda,” Azania 40 (2005): 71–100.

                            (33.) Sven Ouzman, “Seeing is Deceiving: Rock Art and the Non-Visual,” World Archaeology 33, no. 2 (2001): 237–256; and John Parkington, David Morris, and Neil Rusch, Karoo Rock Engravings (Clanwilliam, South Africa: Clanwilliam Living Landscape Project, 2008).

                            (34.) Riaan Rifkin, “Engraved Art and Acoustic Resonance: Exploring Ritual and Sound in North-Western South Africa,” Antiquity 83, no. 321 (2009): 585–601.

                            (35.) Mazel, “Time, Color, and Sound,” 283–296.

                            (36.) John Kinahan, “The Dancing Kudu: Women’s Initiation in the Namib Desert during the Second Millennium AD,” Antiquity 91, no. 358 (2017): fig. 6.

                            (37.) See, for example, Pager, The Rock Paintings, Part I; Pager, The Rock Paintings, Part II; Pager, The Rock Paintings, Part III; Pager, The Rock Paintings, Part IV; Pager, The Rock Paintings, Part V; and Pager, The Rock Paintings, Part VI.

                            (39.) Rosenfeld, “Recording Rock Art,” 10.

                            (40.) Rosenfeld, “Recording Rock Art,” 10.

                            (41.) Hans Leisen et al., “New and Integral Approaches to Rock Art Recording as Means of Analysis and Preservation,” in Wadi Sura: The Cave of Beasts, ed. Rudolph Kuper (Cologne: Heinrich-Barth Institut), 42–49; Frank Förster, “Figuring Out: Computer-Aided Rock Art Recording and Analysis,” in Wadi Sura: The Cave of Beasts, ed. Rudolph Kuper (Cologne: Heinrich-Barth Institut, 2013), 50–53; and Shumla Archaeological and Research Centre (3 July 2018).

                            (42.) For example, Förster, “Figuring Out.”

                            (43.) Edward C. Harris, “The Laws of Archaeological Stratigraphy,” World Archaeology 11, no. 1 (1979): 111–117.

                            (44.) David S. Whitley, “In Suspect Terrain: Dating Rock Engravings,” in A Companion to Rock Art, ed. Jo McDonald and Peter Veth (Chichester: Blackwell, 2012), 605–624.

                            (45.) For example, Rifkin, “Engraved Art”, 595; and Kinahan, “The Dancing Kudu,” 1049.

                            (47.) Russell, “The Position,” 37–40.

                            (48.) James David Lewis-Williams, “Image and Counter-Image: The Work of the Rock Art Research Unit, University of the Witwatersrand,” African Arts 29, no. 4 (1996): 34–41, 93.

                            (49.) James David Lewis-Williams, “Image and Counter-Image,” 38.

                            (50.) Jannie H. N. Loubser and Paul Den Hoed, “Recording Rock Paintings: Some Thoughts on Methodology and Technique,” Pictogram 4, no. 1 (1991): 1–5.

                            (51.) For example, Jeremy C. Hollmann, “Digital Technology and Documentation of Hunter-Gatherer Rock Art in South Africa,” African Archaeological Review 35 (2018): 157–168.

                            (52.) Loubser and Den Hoed, “Recording Rock Paintings,” 2.

                            (53.) Mark Mudge et al., “Robust and Scientifically Reliable Rock Art Documentation from Digital Photographs,” in A Companion to Rock Art, ed. Jo McDonald and Peter Veth (Chichester: Blackwell, 2012), 644–659; and Cultural Heritage Imaging, consulted July 3, 2018.

                            (54.) James David Lewis-Williams, “Image and Counter-Image,” 38.

                            (55.) Jeremy C. Hollmann, “Bees, Honey and Brood: Southern African Hunter-Gatherer Paintings of Bees and Bees’ Nests, uKhahlamba-Drakensberg, KwaZulu-Natal, South Africa,” Azania: Archaeological Research in Africa 50, no. 3 (2015): 343–371; and Jeremy C. Hollmann, “Allusions to Agriculturist Rituals in Hunter-Gatherer Rock Art? eMkhobeni Shelter, Northern Ukhahlamba-Drakensberg, KwaZulu-Natal, South Africa,” African Archaeological Review 32, no. 3 (2015): 505–535.

                            (56.) See Robert Mark and Evelyn Billo, “A Stitch in Time: Digital Panoramas and Mosaics,” American Indian Rock Art 25 (1999): 155–168; and many others.

                            (57.) Elia Quesada et al., “Empowering the Local Saharawi People: Training for Rock Art Documentation in Western Sahara (DARSSO Project),” African Archaeological Review 35, no. 2 (2018): 241–256.

                            (58.) Paul S. C. Taçon, “Presenting Rock Art through Digital Film: Recent Australian Examples,” in Working with Rock Art: Recording, Presenting and Understanding Rock Art Using Indigenous Knowledge, ed. Benjamin Smith, Knut Helskog, and David Morris (Johannesburg: Wits University Press, 2012), 207–215.

                            (59.) For example, Robert Mark and Evelyn Billo, “Application of Digital Image Enhancement in Rock Art Recording,” American Indian Rock Art 28 (2002): 121–128; Bruno David et al., “Why Digital Enhancement of Rock Art Works: Rescaling and Saturating Colours,” Antiquity 75, no. 290 (2001): 781–792; Jean-Loïc Le Quellec, Frédérique Duquesnoy, and Claudia Defrasne, “Digital Image Enhancement with DStretch: Is Complexity Always Necessary for Efficiency?” Digital Applications in Archaeology and Cultural Heritage 2 (2015): 55–67; and Liam M. Brady and Robert Gunn, “Digital Enhancement of Deteriorated and Superimposed Pigment Art: Methods and Case Studies,” in A Companion to Rock Art, ed. Jo McDonald and Peter Veth (Chichester: Blackwell, 2012), 627–643.

                            (60.) Jeremy C. Hollmann and Kevin Crause, “Digital Imaging and the Revelation of ‘Hidden’ Rock Art: Vaalekop Shelter, KwaZulu-Natal,” Southern African Humanities 23 (2011): 55–76; and Hollmann, “Digital Technology.”

                            (61.) Dipuo W. Mokokwe, “Representing Southern African San Rock Art: A Move towards Digitisation,” in Working with Rock Art: Recording, Presenting and Understanding Rock Art using Indigenous Knowledge, ed. Benjamin Smith, Knut Helskog, and David Morris (Johannesburg: Wits University Press, 2012), 61–81.

                            (62.) James David Lewis-Williams, Thomas Dowson, and Janette Deacon, “Rock Art and Changing Perceptions of Southern Africa’s Past: Ezeljagdspoort Reviewed,” Antiquity 67, no. 255 (1993): 273–291; and Geoffrey Blundell and Edward B. Eastwood, “Identifying Y-Shape Motifs in the San Rock Art of the Limpopo-Shashi Confluence Area, Southern Africa: New Painted and Ethnographic Evidence,” South African Journal of Science 97 (2001): 305–308.

                            (63.) Francis Thackeray and Thembi Russell, “Rock Painting of Roan Antelope at Giant’s Castle, KwaZulu-Natal,” Digging Stick 21, no. 3 (2004): 18.

                            (64.) Le Quellec et al., “Digital Image Enhancement,” 59–60.

                            (65.) Inés D. Sanz, “Rock Art Recording Methods: From Traditional to Digital” in Encyclopedia of Global Archaeology, ed. Claire Smith and Jo McDonald (New York: Springer, 2014), 6531–6537.

                            (66.) For example, Natan Micheletti, Jim H. Chandler, and Stuart N. Lane, “Structure from Motion (SfM) Photogrammetry,” in Geomorphological Techniques (Online Edition), ed. Simon Cook, Lucy Clarke, and Jo Nield (London: British Society for Geomorphology, 2015), chap. 2, sec. 2.2.

                            (67.) Quesada et al., “Empowering the Local,” 241–256.

                            (68.) Leisen et al., “New and Integral Approaches,” 42–49.

                            (69.) Stephane Hoerlé et al., “Microanalysis and Dating for Rock Art Studies: Towards a Common Analytical Strategy,” South African Archaeological Bulletin 65, no. 192 (2010): 221.

                            (70.) Hoerlé et al., Microanalysis and Dating,” 221.

                            (71.) Hoerlé et al., Microanalysis and Dating,” 221–222.

                            (72.) Sabine Krause, Heiko Riemer, and Hans Leisen, “Paints and Pigments in the Rock Art of Wadi Sura,” in Wadi Sura: The Cave of Beasts, ed. Rudolph Kuper (Cologne: Heinrich-Barth Institut, 2013), 58–61.

                            (74.) Marion How, The Mountain Bushmen of Basutoland (Pretoria: Van Schaik, 1962).

                            (75.) Karen L. Steelman and Marvin W. Rowe, “Radiocarbon Dating of Rock Paintings: Incorporating Pictographs into the Archaeological Record,” in A Companion to Rock Art, ed. Jo McDonald and Peter Veth (Chichester: Blackwell 2012), 574.

                            (76.) Krause et al., “Paints and Pigments”; and Sven Ouzman and Johannes Loubser, “Art of the Apocalypse: Southern Africa’s Bushmen Left the Agony of Their End Time on Rock Walls,” Discovering Archaeology 2, no. 5 (2000): 38–45.

                            (77.) Steelman and Rowe, “Radiocarbon Dating,” 574.

                            (78.) Flett and Letley, “Style and Stylistic Change,” 13.

                            (79.) Patricia Vinnicombe, People of the Eland: Rock Paintings of the Drakensberg Bushmen as a Reflection of Their Life and Thought (Pietermaritzburg: University of Natal Press, 1976); and James David Lewis-Williams, Believing and Seeing: Symbolic Meanings in Southern San Rock Paintings (London: Academic Press, 1981).

                            (80.) Adelphine Bonneau, David. G. Pearce, and Mark Pollard, “A Multi-Technique Characterization and Provenance Study of the Pigments Used in San Rock Art, South Africa,” Journal of Archaeological Science 39 (2012): 287–294.

                            (81.) Jeffrey Hughes and Anne Solomon, “A Preliminary Study of Ochres and Pigmentaceous Materials from KwaZulu-Natal, South Africa: Towards an Understanding of San Pigment and Paint Use,” Natal Museum Journal of Humanities 12 (2000): 15–31.

                            (82.) Aurélie Tournié et al., “The First In Situ Raman Spectroscopic Study of San Rock Art in South Africa: Procedures and Preliminary Results,” Journal of Raman Spectroscopy 42 (2011): 399–406; and Krause et al., “Paints and Pigments,” 59.

                            (83.) Krause et al., “Paints and Pigments,” 58.

                            (84.) Miles Burkitt, South Africa’s Past in Stone and Paint (Cambridge, UK: Cambridge University Press, 1928); and James David Lewis-Williams, “Introductory Essay: Science and Rock Art,” South African Archaeological Society Goodwin Series 4 (1983): 3–13.

                            (85.) Royden Yates, Anthony Manhire, and John Parkington, “Rock Painting and History in the South-Western Cape,” in Contested Images: Diversity in Southern African Rock Art Research, ed. Thomas A. Dowson and James David Lewis-Williams (Johannesburg: Wits University Press, 1994), 29–60; Mazel, “Images in Time,” 81–97; Aron D. Mazel “Paint and Earth,” Time and Mind: The Journal of Archaeology, Consciousness and Culture 6 (2013): 49–58; and Judy Sealy, “Rock Art and Archaeology in the Maloti-Drakensberg: Then and Now,” South African Journal of Science 107, no. 3–4 (2011): 15–17.

                            (86.) Yates, Manhire, and Parkington, “Rock Painting,” 30; and Sealy, “Rock Art,” 2–3.

                            (87.) Yates, Manhire, and Parkington, “Rock Painting,” 30.

                            (88.) Chippindale and Nash, “Pictures in Place,” 5.

                            (89.) Steelman and Rowe, “Radiocarbon Dating,” 566.

                            (90.) Lewis-Williams, “Introductory Essay,” 3.

                            (91.) Karl W. Butzer et al., “Dating and Context of Rock Engravings in Southern Africa,” Science 203, no. 4386 (1979): 1210.

                            (92.) Butzer et al., “Dating and Context,” 1210.

                            (93.) Zboray, “A Revision,” 22; and Zboray, “Wadi Sura,” 238–239.

                            (94.) Zboray, “Wadi Sura,” 22.

                            (95.) Zboray, “A Revision,” 22; and Zboray, “Wadi Sura.”

                            (96.) David Morris, “Engraved in Place and Time: A Review of Variability in the Rock Art of the Northern Cape and Karoo,” South African Archaeological Bulletin 43 (1988): 109–121; and Siyakha Mguni, Archival Theory, Chronology and Interpretation of Rock Art in the Western Cape, South Africa, Cambridge Monographs in Archaeology 93 (Oxford: Archaeopress, 2016).

                            (98.) Harris and Gunn, “The Use of Harris Matrices,” 12.

                            (99.) The Harris Matrix Composer.

                            (102.) Swart, “Rock Art Sequences,” 33; and Mazel and Watchman, “Dating Rock Paintings,” 64–71.

                            (103.) Harris and Gunn, “The Use of Harris Matrices,” 7.

                            (104.) Swart, “Rock Art Sequences,” 13–14; and Harris and Gunn, “The Use of Harris Matrices,” 10.

                            (105.) Yates, Manhire, and Parkington, “Rock Painting,” 39.

                            (107.) James David Lewis-Williams, “The Syntax and Function of the Giant’s Castle Pock-Paintings,” South African Archaeological Bulletin 27 (1972): 49–65; and James David Lewis-Williams, “A Dream of Eland: An Unexplored Component of San Shamanism and Rock-Art,” World Archaeology 19 (1987): 165–177.

                            (108.) Whitley, “In Suspect Terrain,” 607.

                            (109.) Ronald I. Dorn, “Chronometric Techniques: Engravings,” in Handbook of Rock Art Research, ed. David S. Whitley (Walnut Creek, CA: Altamira Press, 2001), 167–189.

                            (110.) Dorn, “Chronometric Techniques,” 182–183; and Steelman and Rowe, “Radiocarbon Dating,” 578.

                            (111.) Steelman and Rowe, “Radiocarbon Dating,” 578.

                            (112.) David Morris, “Snake and Veil: The Rock Engravings of Driekopseiland, Northern Cape, South Africa,” in Seeing and Knowing: Understanding Rock Art with and without Ethnography, ed. Geoffrey Blundell, Christopher Chippindale, and Benjamin Smith (Johannesburg: Wits University Press, 2011), 37–53.

                            (113.) Dorn, “Chronometric Techniques,” 175.

                            (114.) Steelman and Rowe, “Radiocarbon Dating,” 576.

                            (115.) Wolfgang Erich Wendt, “‘Art Mobilier’ from the Apollo 11 Cave, South West Africa: Africa’s Oldest Dated Works of Art,” South African Archaeological Bulletin 31, no. 121–122 (1976): 5–11; Pieter Breunig, “Archaeological Investigations into the Settlement History of the Brandberg,” in The Rock Paintings of the Upper Brandberg, Part 1: Amis Gorge, ed. Harald Pager (Cologne: Heinrich-Barth Institut, 1989), 17–48; Aron D. Mazel, “Collingham Shelter: The Excavation of Late Holocene Deposits, Natal, South Africa,” Natal Museum Journal of Humanities 4 (1992): 1–52; Aron D. Mazel, “Dating the Collingham Shelter Rock Paintings,” Pictogram 6, no. 2: 33–35; Nicholas J. Walker, Late Pleistocene and Holocene Hunter-Gatherers of the Matopos: An Archaeological Study of Change and Continuity in Zimbabwe (Uppsala: Societas Archaeologica Upsaliensis, 1995); and Antonieta Jerardino and Natalie Swanepoel, “Painted Slabs from Steenbokfontein Cave: The Oldest Known Parietal Art in Southern Africa,” Current Anthropology 40, no. 4 (1999): 542–547.

                            (116.) Wendt, “Art Mobilier.”

                            (117.) Anne I. Thackeray et al., “Dated Rock Engravings from Wonderwerk Cave, South Africa,” Science 214 (1981): 64–67.

                            (118.) Mauro Cremaschi, “The Rock Varnish in the Messak Settafet (Fezzan, Libyan Sahara), Age, Archaeological Context, and Palaeo-Environmental Implication,” Geoarchaeology 11, no. 5 (1996): 393–421; Alfred Muzzolini, “Saharan Africa,” in Handbook of Rock Art Research, ed. David S. Whitley (Walnut Creek, CA: Altamira Press, 2001), 605–636; Anthony Manhire et al., “Cattle, Sheep and Horses: A Review of Domestic Animals in the Rock Art of Southern Africa,” South African Archaeological Bulletin Goodwin Series 5 (1986): 22–30; Jannie H. N. Loubser and Gordon Laurens, “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, ed. Thomas A. Dowson and James David Lewis-Williams (Johannesburg: Wits University Press, 1994), 83–118; Muzzolini, “Saharan Africa,” 621–622; Edward B. Eastwood and Cathelijne Cnoops, Capturing the Spoor: An Exploration of Southern African Rock Art (Cape Town: David Philip, 2006); Pager, Ndedema 119: 126; Vinnicombe, People of the Eland 10, no. 69: 273-274; Tim M. O’C. Maggs, “The Zulu Battle-Axe,” Natal Museum Journal of Humanities 5 (1993): 175–188; Loubser and Laurens, “Depictions of Domestic”; Hollmann, “Allusions to Agriculturist Rituals,” 516–518; Sven Ouzman, “Indigenous Images of a Colonial Exotic: Imaginings from Bushman Southern Africa,” Before Farming 2003, no. 1 (2003): 4, 11–12; and Challis, “Taking the Reins”; Challis, “Creolisation”; Challis, “Binding Beliefs”; and Hollmann, “Allusions to Agriculturist Rituals.”

                            (119.) Mitchell, “Introduction to Archaeological.”

                            (120.) Dorn, “Chronometric Techniques,” 169.

                            (121.) Steelman and Rowe, “Radiocarbon Dating,” 572.

                            (122.) Many of the sources in this article do not provide this information and therefore the calibration status of many of the estimates cited here is not known.

                            (123.) Steelman and Rowe, “Radiocarbon Dating,” 569.

                            (124.) Nicholas Johannes Van Der Merwe, Judy Sealy, and Royden Yates, “First Accelerator Carbon-14 Date for Pigment from a Rock Painting,” South African Journal of Science 83 (1987): 56–57. The dating was performed on a charcoal-containing painted figure in the Cederberg, Western Cape Province, South Africa.

                            (125.) Steelman and Rowe, “Radiocarbon Dating,” 570.

                            (126.) Adelphine Bonneau et al., “An Improved Pretreatment Protocol for Radiocarbon Dating Black Pigments in San Rock Art,” Radiocarbon 53, no. 3 (2011): 419–428.

                            (127.) Adelphine Bonneau et al., “The Earliest Directly Dated Rock Paintings from Southern Africa: New AMS Radiocarbon Dates,” Antiquity 91, no. 356 (2017): 322–333.

                            (128.) F. Mori et al., “Chemical Characterization and AMS Radiocarbon Dating of the Binder of a Prehistoric Rock Pictograph at Tadrart Acacus, Southern West Libya,” Journal of Cultural Heritage 7 (2006): 344–349.

                            (129.) Mazel and Watchman, “Dating Rock Paintings,” 91.

                            (130.) Whitley, “In Suspect Terrain,” 607.

                            (131.) Whitley, “In Suspect Terrain,” 607.

                            (132.) Dorn, “Chronometric Techniques,” 182; and Whitley, “In Suspect Terrain,” 619.

                            (133.) Whitley, “In Suspect Terrain,” 613.

                            (134.) Whitley, “In Suspect Terrain,” 609.

                            (135.) David S. Whitley and Harold J. Annegarn, “Cation-Ratio Dating of Rock Engravings from Klipfontein, Northern Cape,” in Contested Images: Diversity in Southern African Rock Art Research, ed. Thomas A. Dowson and James David Lewis-Williams (Johannesburg: Wits University Press, 1994), 189–197.

                            (136.) Whitley and Annegarn, “Cation-Ratio Dating,” 195.

                            (137.) Cremaschi, “The Rock Varnish,” 391, 393

                            (138.) Cremaschi, “The Rock Varnish,” 393.

                            (139.) Whitley, “In Suspect Terrain,” 613.

                            (140.) Whitley, “In Suspect Terrain,” 613.

                            (141.) Steelman and Rowe, “Radiocarbon Dating,” 572.

                            (142.) See Alex Willcox, The Rock Art of Africa (London and Canberra: Croom Helm, 1984), 1–5.

                            (143.) Willcox, The Rock Art, 19, map 4.1.

                            (144.) There are a great many publications of copies of rock paintings from the Sahara and southern Africa with far fewer from other parts of the continent, such as Mary Leakey’s Africa’s Vanishing Art: The Rock Paintings of Tanzania (Garden City, NY: Doubleday, 1983). North African examples include Leo Frobenius and Hugo Obermaier, Hadschra Maktouba (Munich: Wolff, 1925); several publications by Henri Lhote; and Raymond Vaufrey, “L’art Rupestre Nord-Africaine,” Archives of the Institute for Human Palaeontology 20–21 (1939). Southern African examples include Helen Tongue, Bushman Paintings (Oxford: Clarendon Press, 1909); the work of George Stow, all reproduced in Pippa Skotnes, Unconquerable Spirit: George Stow’s History Paintings of the San (Athens, OH and Johannesburg: Ohio University Press, 2008); and that of Abbé Henri Breuil in Namibia, for example: Abbé Henri Breuil, Philipp Cave: The Rock Paintings of Southern Africa; Volume Two (Paris: Trianon Press, 1957); Abbé Henri Breuil, The Tsisab Ravine and Other Brandberg Sites: The Rock Paintings of Southern Africa; Volume 3 (Paris and London: Trianon Press, 1971); Abbé Henri Breuil, Anibib and Omandumba and Other Erongo Sites: The Rock Paintings of Southern Africa; Volume 4 (Paris and London: Trianon Press, 1960); and Abbé Henri Breuil, The Sphinx and the White Ghost Shelters and Other Spitzkopje Sites: The Rock Paintings of Southern Africa; Volume 6 (Paris and London: Trianon Press, 1975). Justine Wintjes has published about southern Africa’s early rock-art recordings: Justine Wintjes, “A Pictorial Genealogy: The Rainmaking Group from Sehonghong Shelter,” Southern African Humanities 23 (2011): 17–54; Justine Wintjes, “The Frobenius Expedition to Natal and the Cinyati Archive,” Southern African Humanities 25 (2013): 167–205; and Justine Wintjes, “Glimpses Prior to Orpen: Pre-1874 Texts and Pictures about Bushman Rock Painting,” in On the Trail of Qing and Orpen, ed. José M. de Prada-Samper et al. (Johannesburg: Standard Bank, 2016).

                            (145.) Wintjes, “The Frobenius Expedition”; and Maria Wilman, The Rock-Engravings of Griqualand West and Bechuanaland, South Africa (Cambridge, UK: Deighton Bell, 1933).

                            (146.) Pager, Ndedema.

                            (147.) Lewis-Williams, “Introductory Essay,” 3–13.

                            (148.) Foremost among these in southern African research are Timothy Michael O’Connell Maggs, “A Quantitative Analysis of the Rock Art from a Sample Area in the Western Cape,” South African Journal of Science 63 (1967): 100–104; “Microdistribution of Some Typologically Linked Rock paintings from the Western Cape,” in Proceedings of the Sixth Pan African Congress on Prehistory and Quaternary Studies (Dakar: Pan African Archaeological Association, 1967), 218–220; Timothy Michael O’Connell Maggs, “Some Observations on the Size of Human Groups during the Late Stone Age,” Supplement to the South African Journal of Science, special issue 2 (1971): 49–53; Pager, Ndedema; Lewis-Williams, “The syntax and Function,” 49–65; James David Lewis-Williams “Superpositioning in a Sample of Rock-Paintings from the Barkly East District,” South African Archaeological Bulletin 29 (1974): 93–103; Patricia Vinnicombe, “The Recording of Rock Paintings: An Interim Report,” South African Journal of Science (1967): 282–284; Patricia Vinnicombe, “Rock-Painting Analysis,” South African Archaeological Bulletin 22 (1967): 129–141; and Vinnicombe, People of the Eland.

                            (149.) Lewis-Williams, “The Syntax and Function”; Patricia Vinnicombe, “Motivation in African Rock Art,” Antiquity 46 (1972); Patricia Vinnicombe, “Myth, Motive, and Selection in Southern African Rock-Art,” Africa 42 (1972): 192–204; and Vinnicombe, People of the Eland.

                            (150.) James David Lewis-Williams, “Rock Art and Shamanism,” in A Companion to Rock Art, ed. Jo McDonald and Peter Veth (Chichester: Blackwell, 2012), 17–33.

                            (151.) Lewis-Williams, Believing and Seeing.

                            (152.) For example, James David Lewis-Williams, “Quanto? The Issue of ‘Many Meanings’ in Southern African San Rock-Art Research,” South African Archaeological Bulletin 53 (1998): 86–97.

                            (153.) See, for example, comments by Yates, Manhire, and Parkington, “Rock Painting,” 29–60; and Mazel, “Images in Time,” 45

                            (154.) Colin Campbell, “Images of War: A Problem in San Rock Art Research,” World Archaeology 18, no. 2 (1986): 255–267; Thomas A. Dowson, “Reading Art, Writing History: Rock Art and Social Change in Southern Africa,” World Archaeology 25, no. 3 (1994): 332–344; Judith Stevenson, “Shaman Images in Rock Art: A Question of Gender,” in Representations of Gender from Prehistory to Present, ed. Moira Donald and Linda Hurcombe (London: MacMillan, 2000), 45–66; Geoffrey Blundell, Nqabayo’s Nomansland: San Rock Art and the Somatic Past (Uppsala: Uppsala University, 2004); Challis, “Taking the Reins,” 104–107; Challis, “Creolisation,” 265–280; and Challis, “Binding Beliefs,” 247–264.

                            (155.) Pager, The Rock Paintings, Part I; Pager, The Rock Paintings, Part II; Pager, The Rock Paintings, Part III; Pager, The Rock Paintings, Part IV; Pager, The Rock Paintings, Part V; and Pager, The Rock Paintings, Part VI.

                            (156.) David S. Whitley, ed., Handbook of Rock Art Research (Walnut Creek, CA: Altamira Press, 2001); and Jo McDonald and Peter Veth, eds., A Companion to Rock Art (Chichester: Blackwell, 2012).

                            (157.) Christopher Chippindale and Paul S. C. Taçon, eds., The Archaeology of Rock Art (Cambridge: Cambridge University Press, 1998).

                            (158.) Chippindale and Nash, Pictures in Place.

                            (159.) Benjamin Smith, Knut Helskog, and David Morris, eds., Working with Rock Art: Recording, Presenting and Understanding Rock Art Using Indigenous Knowledge (Johannesburg: Wits University Press, 2012).

                            (160.) Willcox, The Rock Art.

                            (161.) Alec Campbell and David Coulson, African Rock Art: Paintings and Engravings on Stone (New York: Harry N. Abrams, 2001).

                            (162.) Rudolph Kuper in collaboration with Franziska Bartz, Erik Büttner, Frank Darius, Frank Förster, Lutz Hermsdorf-Knauth, Sabine Krause, Hans Leisen, Heiko Riemer, Jürgen Seidel, and András Zboray, in Kuper, Wadi Sura.

                            (163.) Guillaume Robin, ed., “Digital Imaging Techniques for the Study of Prehistoric Rock Art,” special issue, Digital Applications in Archaeology and Cultural Heritage 2, no. 2–3 (2015): 35–232.