Rohit Shah

Chapter III

The painted dark

Lions in the Ardèche, pigs in Sulawesi, kangaroos in the Kimberley, and the long argument over why anyone painted in places no one could see.

In December 1994, three cavers, Jean-Marie Chauvet, Éliette Brunel, and Christian Hillaire, felt a draught coming from a pile of rubble in a limestone cliff above the Ardèche gorge in southern France. They cleared a passage, lowered themselves into a series of vast chambers, and swept their lamps across the walls. What they saw has been described, by nearly everyone who has seen it since, in the language of revelation: horses with open mouths, as if whinnying; a pride of lions, their heads drawn in overlapping outlines so that they seem to move; woolly rhinoceroses locking horns; mammoths, bears, bison, a panther, an owl. The images were drawn with a confidence and naturalism that seemed impossible for the dates later established for them. The oldest phase of painting at the Chauvet cave has been dated to around thirty-six thousand years ago.

Chauvet overturned a cherished assumption. For much of the twentieth century, cave art was understood to have developed gradually, from crude beginnings to the sophistication of Lascaux, about seventeen thousand years ago. Chauvet showed that some of the most accomplished images were among the oldest. Art, it seemed, had not evolved from simplicity to mastery. It had arrived, at least in Europe, fully formed.

Art had not evolved from simplicity to mastery. At least in Europe, it arrived fully formed.

Chauvet also held clues that pointed beyond art. In a chamber deep in the cave, the skull of a cave bear had been placed on a fallen block of stone, in a position that is very difficult to explain as accident. Dozens of other bear skulls lay on the floor around it. On a hanging rock pendant near the back of the cave, the lower body of a woman, drawn with emphatic genitals, merges into the head and forequarters of a bison. The cave floor preserved footprints, including those of a child, walking some distance through the darkness, and near them the prints of a large canid. Whatever happened in Chauvet, it was not a gallery. It was a place where people went to do something.

A doubted discovery

It was not always accepted that people so ancient could produce such images. In 1879, the amateur archaeologist Marcelino Sanz de Sautuola was digging in the floor of the cave of Altamira in northern Spain, when his young daughter María, wandering further in, looked up and saw the painted bison on the ceiling. Sautuola published the find and was widely accused of fraud; the paintings were simply too good, in the view of the leading prehistorians of the day, to have been made by the "savages" of the Stone Age. Only in 1902 did one of his fiercest critics, Émile Cartailhac, publish a public recantation, after similar paintings were found in French caves sealed by deposits that could not be faked. Sautuola had died by then.

The episode is worth remembering because it shows how deeply prejudice can govern interpretation. The ancestors were not permitted to be artists until evidence forced the issue. The same pattern would recur with Neanderthals, with Africa, and with Asia. Our assumptions about what "primitive" minds can do have repeatedly been proven wrong, and nearly always in the same direction.

The world map of the painted dark

For a century, Europe was the heartland of Palaeolithic art. The discoveries of the past decade have dismantled that geography.

On the Indonesian island of Sulawesi, in the limestone karst of Maros-Pangkep, researchers led by Maxime Aubert and Adam Brumm have used uranium-series dating of mineral crusts over paintings to establish extraordinary ages. A painting of a Sulawesi warty pig at Leang Tedongnge was dated to at least 45,500 years ago. A panorama at Leang Bulu' Sipong 4, dated to at least 43,900 years ago, shows small figures hunting pigs and dwarf buffalo, some of the figures apparently combining human bodies with animal features such as beaks or tails. In 2024, the team reported a scene at Leang Karampuang in which three human-like figures interact with a pig, dated to at least 51,200 years ago, the oldest known figurative art in the world.

These dates matter for two reasons. They show that figurative painting was being done in Southeast Asia at least as early as in Europe, which means the capacity was almost certainly carried out of Africa by the modern humans who dispersed across the world, rather than invented in any one place. And the Sulawesi images do something the European ones rarely do: they tell stories. Hunters confront animals. Figures that are part human and part animal act. The earliest known picture made by our species is, in effect, a narrative, and a strange one.

In northern Australia, the Kimberley region holds one of the world's greatest concentrations of rock art, including the elegant, tasselled human figures known as Gwion Gwion. A painting of a kangaroo there, dated using mud-wasp nests found above and below the pigment, is at least seventeen thousand years old. In Arnhem Land, rock art traditions span tens of thousands of years and continue to the present, painted and repainted by people who know the stories that go with them. That continuity makes Australia uniquely important: it is the one place where the question of what rock art means can sometimes be put to the people whose ancestors made it.

In central India, the rock shelters of Bhimbetka, on the edge of the Vindhya hills in Madhya Pradesh, were identified by the archaeologist V. S. Wakankar in 1957 and are now a UNESCO World Heritage Site. Hundreds of shelters bear paintings, most of them dating from the Mesolithic onward, showing hunts, dances, animals, and figures with elaborate headdresses. At Bhimbetka and at Daraki-Chattan, small cup-shaped hollows hammered into rock surfaces, known as cupules, have been claimed by some researchers as among the oldest rock markings on earth, possibly going back to the Lower Palaeolithic; the dating remains debated. What is not debated is that the Indian subcontinent holds one of the longest continuous records of human engagement with painted stone, extending into the living traditions of Adivasi communities whose own wall art continues motifs of dance, animal, and spirit.

A red-painted figure of a rider on a horse on the rock at Bhimbetka, Madhya Pradesh
PlateA rider on horseback painted on a rock shelter wall at Bhimbetka, Madhya Pradesh, one of the densest concentrations of rock art in the world. Paintings here span thousands of years. Photo: Marjolein Katsma / Wikimedia Commons

And in Spain, a team led by Dirk Hoffmann reported in 2018 that some painted marks, including a hand stencil at Maltravieso cave and ladder-like red lines at La Pasiega, were more than sixty-four thousand years old, which would place them before the arrival of modern humans in Europe and make Neanderthals their authors. The dating technique and its interpretation have been vigorously disputed. If the claim stands, the painted dark was not the exclusive domain of our species at all.

What is on the walls, and what is not

A curious consistency runs through the Ice Age art of Europe. Animals dominate. Horses, bison, aurochs, mammoths, ibex, deer, and in some caves lions, bears, and rhinoceroses appear in large numbers and are drawn with anatomical accuracy and a vivid sense of movement. Human figures are rare and, when they appear, are usually schematic, distorted, or merged with animals. There are almost no landscapes: no ground lines, no trees, no rivers, no sun. There are many abstract signs: dots, lines, grids, rectangles, and forms that look like combs, tents, or keys.

The animals painted are also not simply the animals eaten. At Lascaux, the most common animal on the walls is the horse, while the most common animal among the food remains is the reindeer, which is scarcely painted. At Chauvet, dangerous animals, lions, rhinos, bears, are unusually prominent. Whatever the artists were doing, they were not making a menu. They were making a selection, governed by rules we cannot read.

A painted horse from the Lascaux cave, shown in a museum replica
PlateA horse from Lascaux, shown in a faithful replica at the Anthropos museum, Brno. At Lascaux the horse is the most common painted animal, though reindeer dominate the food remains. Photo: HTO / Wikimedia Commons

The images were also placed with care. Painters exploited the natural relief of the rock, using a bulge to form the shoulder of a bison or a crack to form the line of a back, as if the animal were already there in the stone, waiting to be revealed. Some images are in large chambers where many people could gather. Others are in narrow crevices, high on ceilings, or down shafts, where only one person could see them, and where the painter had to work in contorted positions. Some are so deep inside the caves that reaching them requires hours of crawling. This, above all, is the fact that any theory must explain. Why paint where no one can see?

Whatever the artists were doing, they were not making a menu.

First theory: art for art's sake

The earliest interpretation, current in the late nineteenth century, was that the paintings were simply decoration, the product of leisure in a time of abundant game. It collapses on contact with the evidence of placement. No one crawls a kilometre into the earth to decorate a crevice no one else will ever enter. Whatever pleasure the painters took in their work, and the quality of the work suggests they took a great deal, pleasure alone does not explain the location.

Second theory: hunting magic

At the beginning of the twentieth century, the French priest and prehistorian Henri Breuil, who spent decades copying cave art and became its most influential interpreter, advanced the theory of sympathetic hunting magic. Drawing on the ethnography of Aboriginal Australia, especially the work of Baldwin Spencer and Francis Gillen, and on James Frazer's The Golden Bough, Breuil argued that the images were intended to secure success in the hunt and increase the fertility of game. To paint an animal was to gain power over it; some animals appeared to be marked with spears or wounds.

The theory has intuitive force and is still widely repeated. But its difficulties are serious. As we have seen, the animals painted do not match the animals eaten. Very few are shown wounded. And the theory rests on a mechanical conception of magic, in which images act on the world like tools, that owes more to Victorian assumptions about "primitive" thought than to what later ethnographers learned about how hunter-gatherers actually understood their relationships with animals.

Third theory: the structure of the cave

In the 1960s, the French prehistorian André Leroi-Gourhan, influenced by structuralism, took a radically different approach. Instead of interpreting individual images, he mapped the distribution of thousands of images across dozens of caves and looked for statistical patterns. He argued that the caves were organized as ordered sanctuaries, with certain animals, notably horses and bison, paired in central panels, and others relegated to entrances and deep recesses. He proposed that the pairings represented a fundamental duality, which he read as male and female, and that the abstract signs could similarly be sorted into male and female types.

Leroi-Gourhan's specific sexual interpretation has not survived, and his statistics have been criticized. But his method changed the field. He showed that the caves were not random accumulations of pictures but deliberately structured spaces, and that the meaning of an image depended on where it was placed and what it was placed with. Cave art, in his hands, became something closer to a text, or to architecture: a built environment for experience.

Fourth theory: the mind in the cave

The most influential and most contested modern theory came from southern Africa. In the 1970s and 1980s, the South African archaeologist David Lewis-Williams set out to interpret the rock art of the San, using nineteenth-century records of San testimony, notably those collected by Wilhelm Bleek and Lucy Lloyd, and modern ethnography of Kalahari San communities. He concluded that much San rock art was connected to the trance dance, the central religious ritual of San life, in which healers, through intense rhythmic dancing, clapping, and hyperventilation, enter altered states of consciousness to cure sickness, control rain, and travel to the spirit world. The art depicted the experiences of the trance, including bleeding from the nose, figures bending forward with arms held back, the transformation of people into animals, especially the eland, and the sensation of flying or going under water.

In 1988, Lewis-Williams and his colleague Thomas Dowson extended the argument to the European Upper Palaeolithic. Their premise was neurological: the human nervous system, they argued, produces broadly similar sequences of visual experience in altered states, whether induced by drugs, rhythmic stimulation, sensory deprivation, fasting, or pain. In the first stage, people see geometric forms, called entoptic phenomena: grids, dots, zigzags, nested curves, and filigrees. In a second stage, the mind tries to make sense of these forms, turning them into familiar objects. In a third stage, reached through an experience of passing through a vortex or tunnel, people enter a fully formed hallucinatory world, in which they may encounter animals and spirits and feel themselves transformed into animals.

Lewis-Williams pointed out that the abstract signs of cave art resemble entoptic forms, that composite human-animal figures resemble the transformations of the third stage, and that the cave itself, with its narrow passages leading into chambers, physically resembles the vortex that leads to the spirit world. In his 2002 book The Mind in the Cave, he proposed that the painted caves were understood as the underworld itself, or as the entrance to it, and that the images were not pictures of animals seen outside but visions fixed to the rock. The painter was not copying the world but capturing spirit-animals that emerged from behind the membrane of the wall. The hand stencils, in this view, may record a gesture of touching the membrane, or of reaching into the other world.

Lewis-Williams's three stages of altered consciousnesszigzag → serpentStage onegeometricformsStage twothe mindconstrues themStage threevortex andbeings within
FigureThe neuropsychological model of David Lewis-Williams and Thomas Dowson: geometric forms first, then the mind reading them as familiar things, then a passage through a vortex into a world of animals and part-human beings. Critics think it is too general; it remains influential.
The painter was not copying the world, but fixing to the rock the visions that emerged from behind it.

The theory is elegant and has been enormously influential. It has also drawn sustained criticism. Critics such as the archaeologist Paul Bahn and the anthropologist Patricia Helvenston have argued that the neuropsychological model is too general, since geometric patterns are common in art of every kind; that the San analogy cannot be carried across tens of thousands of years and thousands of kilometres; that the term "shamanism," taken from Siberian ethnography, is stretched to cover anything mystical; and that most cave images are calm, naturalistic depictions of animals rather than visionary distortions. Supporters answer that no rival theory accounts so well for the cave's depth, the composite figures, and the use of rock relief.

A fair assessment might be this. There is strong evidence that altered states of consciousness were important in many hunter-gatherer religions, and it is highly plausible that they played some role in Palaeolithic religion. Whether they explain the painted caves as a whole is far less certain. The truth is likely to be plural: different caves, different periods, and different images may have served different purposes, including initiation, storytelling, the marking of territory, the teaching of animal knowledge, and communion with spirits.

The shaft at Lascaux

One image stands at the centre of every argument. At the bottom of a narrow, five-metre shaft at Lascaux, a cave discovered in 1940 by teenagers exploring the hills above Montignac, is a scene unlike any other. A bison, apparently disembowelled, its intestines spilling out, stands with its head lowered and horns directed at a man. The man is drawn in a few crude strokes, in stark contrast to the virtuosic animals elsewhere in the cave. He is falling backward, or lying down, with an erect penis. His head is that of a bird, with four fingers on each hand. Beside him is a staff topped with the figure of a bird. Near them, a rhinoceros walks away.

Breuil read the scene as a hunting accident, perhaps a memorial to a hunter killed by a bison. Others have read it as a star map, or a myth. Lewis-Williams and others read it as a shaman in trance, his bird-headed form marking a transformation, the bird on the staff echoing the bird-topped staffs used by Siberian shamans to aid the flight of the soul, and the erection a known physiological effect of certain trance states. Every reading is conjectural. But the placement, in the most inaccessible part of the cave, and the singularity of the scene suggest that it was of special importance. It is, in a sense, the first story whose meaning we have forgotten.

The sorcerer and the lion-man

Composite beings, part human and part animal, recur throughout the earliest record of human imagination, and they are among the strongest hints of a mystical dimension to that imagination. At the cave of Les Trois-Frères in the French Pyrenees, a figure engraved and painted high on a wall of a chamber has come to be called the Sorcerer. In Breuil's famous drawing, it has the antlers of a stag, the eyes of an owl, the ears of a wolf, the tail of a horse, and human legs and genitals, and appears to be dancing. Later researchers have noted that Breuil's drawing embellished what is actually on the wall, which is more ambiguous. But the composite nature of the figure is not in doubt.

Older still is the figure known as the Löwenmensch, the lion-man, found in pieces in 1939 in the Hohlenstein-Stadel cave in southern Germany and reassembled over subsequent decades. Carved from mammoth ivory, about thirty centimetres tall, and dated to about forty thousand years ago, it has the head and body of a cave lion and stands upright on human legs. Experimental reconstruction suggests it would have taken hundreds of hours to carve with stone tools. It is the oldest known depiction of a being that does not exist in nature. And it was found in a chamber at the back of the cave that seems to have been used for special purposes.

The ivory lion-man figurine from Hohlenstein-Stadel, standing upright with a lion's head
PlateThe lion-man of Hohlenstein-Stadel, carved from mammoth ivory some 35,000 to 40,000 years ago and now in the Museum Ulm: a being part human, part lion. Photo: Thilo Parg / Wikimedia Commons, CC BY-SA 4.0

The lion-man is, in its way, as important as any god. It proves that by forty thousand years ago, at the latest, human beings were thinking beyond what they could see and giving material form to beings of the imagination. It was not a lion and it was not a person. It was something in between, and the space between human and animal, as we will see in the next chapter, was the place where much of primordial religion lived.

Nearby caves of the Swabian Jura yielded other astonishing objects from about the same period: flutes made from the wing bones of griffon vultures and swans and from mammoth ivory, including one from Hohle Fels found in 2008 with five finger holes, and a small ivory female figure, also from Hohle Fels, with exaggerated breasts and vulva and, in place of a head, a carved loop through which it could be hung. Music, sculpture, and imaginary beings appear together in the same caves, at the same time. Whatever was happening in these places involved sound, movement, and the imagination working at full stretch.

The sound of the cave

One of the more intriguing lines of recent research has concerned acoustics. The French researcher Iegor Reznikoff, testing the resonance of caves such as Niaux and Arcy-sur-Cure with his own voice, reported that painted panels and red dots tended to occur at points of strong resonance, where a voice or a hum would echo and amplify. In some narrow passages, red dots seem to mark places where the resonance changes. Other researchers have made related claims, and some have suggested that the natural stalactites of certain caves were struck like instruments; many bear signs of having been hit. These claims are not universally accepted, but they point to something that is easy to forget when looking at photographs. The caves were not experienced as silent picture galleries. They were dark, echoing, often wet, often cold, lit by flickering lamps that made images appear and disappear, and filled, in all likelihood, with voices, drums, and flutes.

Imagine the experience from inside. Hours of crawling through a narrow passage, the sounds of your own breath and the scraping of your body against rock. Then a chamber, where a lamp is raised, and a horse appears on the wall, its eye catching the light, its form shifting as the flame moves. A chant rises, and the voice returns from the rock, as if the cave itself were answering. Without sunlight, without landmarks, with the senses deprived in one direction and overloaded in another, the ordinary boundaries of perception begin to dissolve. It would be surprising if the experience were not, in some sense, religious.

Who went into the dark?

The popular image of the cave painter is an adult man, a hunter-shaman. The evidence complicates this. The archaeologist Dean Snow, analysing the proportions of hand stencils in several French and Spanish caves in 2013, argued that many, perhaps most, were made by women, although the method used to sex the hands has been questioned. At Rouffignac in France, researchers studying finger flutings, lines drawn with the fingers in the soft clay of the walls, identified flutings made by children, some as young as two to five years old, who must have been lifted up by adults to reach the high surfaces. Footprints in caves such as Chauvet and Tuc d'Audoubert include those of children and adolescents.

This suggests that the caves were not only the preserve of isolated specialists. Children were brought into the dark, perhaps as part of initiation, perhaps simply as members of their community. At Tuc d'Audoubert, deep inside the cave, two clay bison were modelled leaning against a rock, and in the clay nearby are heel prints that some have interpreted as the traces of a dance, perhaps performed by young people in a rite of passage. Initiation into adult knowledge, taking place in a dark, frightening, transformative space, is one of the most widespread patterns of religious life among known hunter-gatherers. The cave, on this reading, was a place where children became adults and learned the secrets that made their world intelligible.

The membrane

Let us return, finally, to the hand on the wall.

All of the theories we have examined, from hunting magic to shamanic vision, share an intuition that the images mattered because of where they were, and that where they were was not neutral space. The cave was a threshold: between light and darkness, between the surface world of daily life and the underworld, between the living and the dead, and between the ordinary state of consciousness and the extraordinary. The wall of the cave, in several traditions we know about, was understood as a boundary behind which the spirit world lay. Painting on it, touching it, blowing pigment over a hand pressed against it, might have been a way of reaching through that boundary, or of drawing out what lay behind it.

If this is right, then the painted caves were among the world's first sacred spaces, places set apart, where the ordinary rules of experience did not apply, and where human beings encountered powers greater than themselves. They were not temples in the later sense; there is no evidence of gods, priests, or institutional religion. What they suggest instead is a religion of relationship, especially relationship with animals, who dominate the walls, merge with human beings in composite forms, and appear to have been the principal inhabitants of the other world.

To understand that relationship, we must leave the caves and go out into the forests, tundras, and savannas where the painters lived, and consider how people who hunted animals, depended on them, and knew them intimately might have understood them, not as resources, but as persons. That is the subject of the next chapter.

Curated by Claude from my favourite reads. About this material