For more than three hours, a colleague and I walked through the grassy foothills of the Drakensberg Mountains in KwaZulu-Natal, meeting not a soul on the way. Ultimately, we came to a wide cave half-screened by bushes and a splashing waterfall. Behind this watery veil are some of the finest specimens of ancient San, or "Bushman," rock painting in South Africa. The water has not damaged them, although vandals have. We gazed at walls covered with more than 1,600 images of humans and animals engaged in myriad activities. That night, we slept in the cave, continuing our trip the next day.

That expedition, 10 years ago, was to obtain paint samples that might be radiocarbon-dated. One sample, from a painting of an eland (the biggest of all antelopes), contained microscopic quantities of organic material that allowed the image to be dated to about 400 years ago. Such a direct measurement is rare. Most pieces of rock art, painted in red, brown or yellow ocher--a hydrous iron oxide--contain no organic carbon. So radiocarbon dating, which measures the steady decline of the isotope carbon 14 in organic materials, cannot be used. Our earliest date comes from a Namibian cave, where excavated floors contained painted slabs between 19,000 and 26,000 years old. The oldest date we have for painting on cave walls indicates that mural art was being made at least 3,600 years ago.

Rock paintings and engravings, testimony to a once ubiquitous hunter-gatherer presence, are found from coast to coast in thousands of diverse sites in southern Africa. Some sites are sheltering sandstone caves with hundreds of images; others contain only one or two figures. Some paintings look exquisite, their lines and colors still fresh. Others are faint and crumbling, damaged by time, water and the graffiti of unthinking visitors.

By far the most common subjects in rock art are humans--usually shown in profile, sometimes unclothed--and a wide variety of animals. The most revered of the animals are the larger herbivores. The eland is widely celebrated, although different areas have their own favorites: the elephant in South Africas Cape Province, for example, and a species of antelope called kudu in Zimbabwe. A variety of other creatures are also pictured. Snakes, lions and fish are not uncommon in the art of the Drakensberg Mountains. Hippopotamuses, rhinoceroses, rhebok, baboons, ostriches and domesticated animals appear in the art of many areas. Rarer themes include the aardwolf, aardvark and other creatures both real and imaginary. With the exception of snakes and bees, the San people rarely painted reptiles and insects.

Rock art research is among the most demanding of archaeology's subdisciplines. Without recourse to conventional archaeological methods--weighing, measuring, mapping and statistical comparison--rock art research relies on theoretically and culturally informed interpretations, supported by particularly rigorous argument. We do know that the artists were among the earliest inhabitants of southern Africa, the ancestors of the modern-day "San" peoples. The term San is a linguistic label: the San and Khoekhoe--formerly Hottentot--languages make up the Khoisan group of many related languages and dialects, characterized by click sounds. The plant-gathering and hunting economy of the San has been extensively studied as a model for how people lived until relatively recent times, when animals and plants were domesticated.

Although rock art occasionally provides historical information, paintings and petroglyphs are not historical documents. It is only after the 15th century, when Europeans "discovered" southern Africa, that we begin to have a clearer picture of historical conditions. In 1652 the Dutch established the first permanent settlement in Cape Town. As the newcomers expanded their domain over the next three centuries, they frequently displaced indigenous peoples, whose traditional ways of life changed or disappeared entirely. In some areas, theft of cattle and horses by the San led to retaliatory raids by European farmers. Episodes are recorded in which entire San groups were massacred. Survivors of these communities were eventually absorbed into indigenous herding and farming societies or became laborers around European settlements.

The ancient art traditions had ceased by the 20th century. Today relatively few San speakers live in the old ways, except in parts of Botswana and Namibia. Only the wide distribution of archaeological sites, place-names and rock art alerts us to the vast areas once occupied by these peoples.

In studying the art, the archaeologist is forced to seek all imaginable clues. There are two classes of work: the paintings, which usually occur in caves and shallow shelters, and incised boulders and other surfaces that are found in the dry interior. The petroglyphs, which tend to be less figurative, have until recently attracted less attention than the paintings.

The style and, to a lesser extent, the subject matter of the paintings vary between regions. Often a single site includes works in several styles, so that it is impossible to tell whether it is the work of different artists or art from different historical periods. Early researchers suggested that simpler or less delicate images, in one color only, are the oldest, with color range and stylistic intricacy evolving through time. Today we know there is no such straightforward correlation. Some of the less accomplished work is probably the most recent--some perhaps made by shepherds and children.

Devotees have been trying to interpret rock art for more than a century. Those interpretations change with new knowledge, discoveries and intellectual currents. San testimonies would be extremely helpful in guiding us, but unfortunately, only one exists. It came from a Lesotho San man named Qing, who acted as a guide to a British official, Joseph Orpen, in the Lesotho Mountains in 1873. Qing was familiar with the making of rock paintings and commented on the paintings that they saw. Qing confirmed what some already suspected: that rock painting, as one contemporary European scholar wrote, was not "the mere daubing of figures for idle pastime" but "a truly artistic conception of the ideas which most deeply moved the Bushman mind."

In addition to Qing's direct testimony, researchers also draw on indirect accounts from San speakers. By far the richest body of material was collected a century ago, from people speaking a San language known as /Xam (the initial character is a click sound). In 1870 a group of /Xam San men from northern Cape Province were imprisoned in Cape Town for offenses ranging from stock theft to murder. Wilhelm H. I. Bleek, a German philologist, acquired custody of the men, who built huts at the bottom of his garden and worked as domestic servants. But their main task was sharing accounts of their traditions. While Bleek focused on the language, his sister-in-law, Lucy C. Lloyd, recorded thousands of pages of /Xam lore. A selection was published in Specimens of Bushman Folklore, written by W.H.I. Bleek and L. C. Lloyd (George Allen, London, 1911).

This extraordinary colonial encounter revealed the /Xam world: personal histories, myths, religious beliefs, and magical and mundane practices. Although by the late 19th century these people no longer practiced rock art, their commentaries have proved extremely valuable for interpreting it. Together with Qings account, the /Xam testimonies have helped show that African rock art is much more than mere decoration or reflections of everyday concerns. Instead rock art can best be understood as a religious art, reflecting the /Xam peoples relations with the spirit world and to ritual practices. And almost certainly the act of painting itself had magical importance.

A comparison of Qing's account with the /Xam testimonies shows broad similarities between /Xam and Lesotho San myths. Both San peoples esteemed a creator figure named /Kaggen. Both also spoke of underwater beings and of the creation of the eland. Qing supplied the long-sought link between rock art and myth, whereas the /Xam accounts provided crucial cultural detail that Qing's commentary lacked. Researchers since have relied heavily on both sources.

Window on Culture
SEVERAL SCHOLARS have noted the extraordinary similarities between the mythology of San groups far distant from one another in time and space. All San peoples tell of a primeval time when animals were people; after an initial creation event, they were differentiated. But these first people were often stupid, lacking customs and manners, and only after a second creation did they become real people.

Many stories recount the doings of these animal people. Some explain the origins of fire, heavenly bodies and other physical phenomena. We hear why the baboon has a hairless rump, why people marry and why death is inevitable. Other narrative themes include encounters with warlike neighbors or dangerous carnivores. Food is a constant preoccupation, with a surprising number of stories featuring autophagy--the eating of one's own body. The stories dramatize the dilemmas of existence that faced San hunter-gatherers and emphasize themes involving death and regeneration.

The belief that animals were once people allows an interpretation of therianthropes--figures both human and animal. Some of these paintings, and others of fantastic creatures, may portray beings from the primordial world. Alternatively, some researchers contend that they depict the shamans experience of physical transformation during a trance--when shamans enter the realm of the spirits of the dead.

Some experts, notably David Lewis-Williams and his colleagues at the University of the Witwatersrand, Johannesburg, have correctly observed that the art does not illustrate the mythology. They propose instead that rock art is connected to ritual--and to one ritual in particular: a healing dance that is still practiced by communities in Botswana and Namibia (these peoples do not make rock art). During a ritual dance that may last all night, shamans enter an altered state of consciousness induced by rhythmic movement, singing and clapping. In this hallucinatory state, they believe that they travel to the spirit realm to battle supernatural forces that cause illness.

Lewis-Williams and his associates have proposed that shamanic hallucinations may have prompted the first making of art, in Africa and elsewhere. They hypothesize that because humans all share the same neurological circuitry, visual hallucinatory forms should be similar throughout time--and that geometric designs drawn in the European Paleolithic and Bronze ages, as well as North American Indian art, may also be understood in terms of the healing trance dance and shamanic hallucinatory experiences.

It is certainly true that many creatures in San mythology are not portrayed in the rock art. Yet mythology does provide a crucial context for understanding ritual. Myths tell of the origin of death and disease, the trials of life that ritual practices address. Art-making can probably best be seen as being linked to ritual practices--such as rainmaking and initiation--recorded from recent San peoples.

RECENT STUDIES have shown that hunting scenes in rock art are not as common as early researchers believed. Some paintings originally thought to depict hunts almost certainly portray rainmaking. Testimonies from the /Xam show that they viewed the rain cloud as an animal walking the countryside on "legs" of streaming rain. Rainmakers had to lead a large herbivore from its home in a water hole, take it to a high place and slaughter it; where its blood ran, rain would fall. The rain animals depicted in rock art resemble large herbivores, such as cattle, hippos or antelopes, but often with strange features and proportions. The rain bull in myths and stories embodied the lethal thunderstorm, whereas the female rain animals brought gentle rain.

Qing, in Lesotho, also described rainmaking. He described one rock painting as depicting underwater beings who tamed "eland and snakes." This painting shows six humanlike figures and two bristly animals of no known species, one led by a thong attached to its nose, the other being approached by two men with spears. Despite their geographic separation, both Qing and the /Xam described markedly similar beliefs regarding rain.

Some have argued that rainmakers depicted in such paintings were living shamans, but there is ample evidence that they were in fact considered to be benevolent spirits--dead family members who helped their living kin. Qing, for example, described antelope-headed men in rock paintings "as men who had died and now lived in rivers." These men, as well as the underwater beings leading the rain animals, can best be interpreted as spirits of the dead. /Xam commentators specifically stated that "sorcerers of rain" were dead people, as were game sorcerers to whom the living would appeal for help in the quest for food.

Understanding San beliefs about rain is crucial to understanding their art. Like Kalahari peoples today, the San told of two important beings, a creator figure and a master of death and disease. In the /Xam narratives, this death deity is the Rain Bull. He is the dangerous thunderstorm and the water in the waterhole. People became stars after they died, which then fell into the water where the Rain Bull lived. Qings accounts of dead people living underwater derive from this same complex of beliefs.

Clearly, rock art images of people catching a rain animal allude to more than just controlling the weather. As master of the spirit world, the Rain Bull controls not just rain but also life and death, sickness and health. The "rain paintings" common in rock art can therefore be linked to people's efforts to prevent disease and misfortune and not just storms or drought.

Importance of Initiation
HUNTER-GATHERER societies such as the San are egalitarian, with both sexes having equal access to resources. Nevertheless, social distinctions do exist, and the two most important in these societies--gender and age--meet in the institution of initiation. Ceremonies for girls at puberty, copiously documented, seem to eclipse male initiation, which seems at most to have been a lesser celebration of a boy's first kill. Contemporary Kalahari hunter-gatherers also have conspicuous female initiations. Numerous /Xam stories tell of female initiates who disobeyed the puberty seclusion rules, sneaked off, and were then abducted and drowned by the Rain Bull. (After entering the cultural repertoire, the Rain Bull and other visual images may have been used in varied contexts for different purposes. Initiation paintings, for example, do not seem to involve interactions with the spirit world.)

/Xam initiation stories tell of the dangers attractive female initiates posed to men. The purpose of the rites, however, was probably equally to protect young women from inappropriate male attention. In the close proximity of band societies, members are highly dependent on social cooperation and harmony. Tensions between kin arising from sexual jealousies and misdemeanors may be socially and economically destructive. A purpose of initiation was surely to regulate interpersonal behavior and avoid such conflicts.

Though not abundant, some rock art sites and compositions may have been part of gender-specific initiation rites. Hugely voluptuous female figures with genital details, brandishing crescent-shaped objects, may well relate to female initiation. Paintings of womens dances may be linked to initiation or birth. A remote shelter high in the southwestern Cape Province mountains--unusual for its abundance of female figures and total absence of male imagery--may also pertain to female initiation or birth. Another composition apparently depicting a female initiation ceremony is found in a KwaZulu-Natal Drakensberg site. It shows a prone figure and three clapping women in a circular enclosure. Other figures dance outside. On the periphery of the composition (not shown in the illustration) is a male figure with considerably overemphasized genitalia.

Many sites contain a profusion of diverse imagery, different in theme and style. Some may have been used over centuries for a variety of purposes, others only once or twice, for a particular end. Recent interpretations have emphasized healing of the sick, an action only rarely rendered explicitly in rock art. However, some images previously thought to depict rain animals may be of the Rain Bull himself. Because he is the death figure, these images may have been painted in an effort to cure physical illnesses.

Of course, paintings and engravings depicting European colonists, wagons, soldiers and domestic animals may well be records of real events rather than ritual occasions. In addition, some rock art appears to reflect interactions between the San and other groups. John E. Parkington and his collegues at the University of Cape Town have suggested that handprints found along the southwestern coast, usually overlying earlier art, may have been the work of Khoi herders. Depictions of cattle introduced by the migrating herders and farmers, as well as iron artifacts, maize cobs and glass beads found in excavations, all testify to San involvement in other African economies.

New Interpretations
THE FINAL IMAGES themselves may not be alone in creating significance to the prehistoric artists. The act of painting itself may also have been important. Recent research in anthropology and art history has drawn attention to the process of art making and of the materials used, as opposed to creating an end product for viewing (as is usual in Western arts). This may be especially relevant to understanding San arts.

Probably each step of the painting process in San art carried cultural significance. One indigenous account from 1910 mentions ritual preparations that involved pigments being ground by women at full moon. In my research, I have emphasized the symbolic and spiritual significance of technically nonessential substances, such as fat and eland blood, that were said to be added to the paint mixture, presumably as magical aids. The act of making art, it appears, exerted magical effects that could influence the spirits and help control fate.

Powerful analytical techniques are now available for identifying these paint ingredients and culture-specific "recipes" of ancient paintings. This information would not only help us understand the significance of rock art but is also of interest for reasons ranging from resource exploitation to ancient trade to contemporary conservation.

I have researched the use of pigments and paints, with a view to identifying "magical" additives. This work, with the assistance of soil scientists at the University of KwaZulu-Natal, has involved a suite of techniques, including x-ray fluorescence (to provide quantitative data on minor and trace element composition), x-ray diffraction (to reveal crystal structure and parent rock types of paint ingredients), and environmental scanning electron microscopy (to yield qualitative data on elements present). Another promising technique that we have used experimentally is synchrotron radiation analysis. This technique, suitable for tiny samples, allows for x-ray fluorescence and x-ray diffraction of the same spot. So far these sophisticated research tools have generated more questions and problems than answers. The variability in both pigments and paints may be too vast to produce results relevant to answering archaeological questions.

The function and many meanings of rock art in history and prehistory still generate debate, although a broadly spiritual role is now well established. As an extraordinary and evocative record of the past, San rock art is becoming part of the culture of postapartheid South Africa. Yet paintings face many threats. Through the combined efforts of a spectrum of specialists, we hope to ensure that the rock art will endure as a testament to an ancient African culture, tragically displaced.

ANNE SOLOMON is a graduate of the University of Cape Town, where she obtained her Ph.D. She is a former postdoctoral research fellow of the Getty Center for the History of Art and the Humanities in Los Angeles and was, until recently, senior curator of archaeology at the Natal Museum in Pietermaritzburg. Her book on San arts, The Imagination of the San, is to be published in early 2005 by the University of KwaZulu-Natal Press.