What Cave Art Means

View into Lascaux, Dordogne, France. Photo J.M. Geneste. Courtesy Centre National de Préhistoire, Périgueux, France.


THERE ARE IMAGES on the walls of caves, whether we put them there or not. Or, more precisely, we create images on the walls of caves, whether with charcoal and manganese or simply with our imaginations. Michelangelo’s well-known claim that he simply released from stone what was already there is straightforwardly true of Paleolithic artists. They placed their lines where the contours already suggested animal motion.

When I had occasion to remark early in my cave-art education that the pair of clay bison sculptures (ca. 15,500 years before the present) located in France’s Tuc d’Audoubert cave are relative rarities, since most cave art is painted on the walls, a veteran of the field corrected me. “It’s all sculpture,” she said. It is all “sculpture,” though most of it was done for us by the same natural forces that brought forth the underground spaces hosting the works. The caverns’ many undulations, outcroppings, fissures, and declivities were then enhanced by human hands, and sometimes saliva: as in the common crachis technique of spitting on a surface and then rubbing on the pigments. Other techniques include using water or plant oils as mediums and applying colors by means of pads, brushes, hands, or blowing—either through a tube or directly from the mouth.

The Grotte des Merveilles cave is found in Rocamadour, France, a historic town with a medieval chapel (likely the most impressive in the region) built into the limestone cliffside as if its builders, too, were only drawing out—more elaborately, with greater refinement—what was already there. Outside the Grotte, in early May, I watched a snail inscribing its trail of slime across a road. Inside, I saw what the scholarship calls “punctuations”: clusters of dots left there at some point in the Magdalenian (17,000–12,000 bp)—perhaps as clan identifiers, or as trail markers for those on long journeys, or ritual or sacred symbols, or units of reckoning, or any number of other purposes besides. They are the bare minimum traces of intentionality, and to see them as existing across an ontological divide from the snail’s trail may well be an act of faith or species-based pride.1

I confess I am wary of such pride, and my exposure to Paleolithic art has only reinforced this wariness. Over the past several decades anthropologists such as Philippe Descola and Tim Ingold have done considerable work to show that in most human societies there has been no presumption of a sharp boundary between the natural world and the built cultural milieu, between wilderness and settlement, and in consequence there has been no default presumption of a fundamental difference between the ways in which human art transforms the environment and those in which animals and plants do the same.2 The authors base their arguments on evidence from the ethnography of Amazonia and Sápmi, focusing on groups of people who, respectively, perceive peccaries and reindeer as in some deep and real sense equal actors, as persons if not as human beings, in a shared sociocosmic reality. It helps to make sense of cave art, I have found, to entertain the idea that the inhabitants of Paleolithic Europe made sense of the world around them through a similar sort of ontology.

Human beings arrived in the caves long after the bears did; perhaps our ancestors got the idea from them. We know that the European bear cult survived intact into the Middle Ages, parallel to the construction of the cathedrals, and that it still survives vestigially in parts of the Pyrenees and the Carpathians. At festival times in those parts one can still become a bear. Official royal genealogies of Scandinavia still began, into the sixteenth century, with hybrids of human-bear pairings. The historian Michel Pastoureau has conjectured that these historically attested practices and beliefs are likely the remnants of a singularly important place for the bear in the symbolic construction of social reality in European prehistory. The bear was not just one species among many that was “good to think with,” as Claude Lévi-Strauss famously said of animals, but was conceived literally as human kin.

Bears are not given particular prominence, at least numerically speaking, in Paleolithic figurative art, but then again we have no evidence that the animals represented on the cave walls were the most important ones, at least in a practical or material sense, for Paleolithic people. We know, for example, that there is a reverse correlation between the frequency of representations of a given species and its importance as a staple of human diet.3 What we can affirm with some conviction is that the caves were the space of human and ursine kin. Although they inhabited the space at different times, they were always aware of one another, in some alternative taxonomy that is not too hard to imagine, as fellow spelaeus species.

Historians often cite Hagia Sophia as an example of the continuity of symbols over the course of transformations in their interpretation. Archaeologists of the future will do well to understand that Muslims valued architectural features and even symbols that had first been installed by Christians, even as their meaning changed. This principle might also apply not only to the gradual, empirically verifiable, transitions from the Châtelperronian culture (45,000–36,000 bp, considered by some the oldest material culture of anatomically modern humans in Europe, associated with the Grotte des Fées cave in Châtelperron) to the Solutrean (22,000–17,000 bp, named for the style of carved stone found at the Roche de Solutré) to the Magdalenian (named for the type of specimens found at La Madeleine in Dordogne), but also to the mythological transition from bear to human.

The markings of bear claws on the walls of caves inhabited in other periods by human beings are held to be incidental, the by-products of some more fundamental bear activity. By contrast, it has been common to suppose that the creation of human art is both the execution of an intention and an end in itself. Even if there is almost always some ulterior purpose for the creation of art (such as service to the community or state, or, in more recent times, the accumulation of profit), the painting of human hands on the walls of the caves is—unlike the raking of bear nails on stone—generally supposed to involve, at a minimum, a conscious desire for the figures of hands to appear in rock.

It is a coincidence, though it feels like something more, that the Dordogne region of France features not only limestone caves decorated with mammoths, aurochs, indecipherable tectiform signs and punctuations, but also, outside, splendiferous medieval cathedrals with stained glass, votive candles, statues, paintings, and myriad crosses. The cathedrals are often held to be the more formidable achievement, in part because they were built up by intention and will, while the caves were arranged by nature, and then only faintly modified with pigments.

You can discern the cathedral builders’ intention in the rule-boundedness of their constructions, in the angles and proportions and the relative uniformity of the individual sites over the course of several centuries. Every cave by contrast is unique, and there appears to be no rule governing an individual cave’s particular configuration. There are laws of geomorphology, but the shape or size of any given cavern is dictated entirely by the unrepeatable conditions of its singular formation.

The art inside the caves is, by contrast, remarkably uniform; transformations in style and technique are measured not in centuries but in tens of millennia. It is only when the artist happened to use charcoal, an organic residue, that absolute dating by means of carbon 14 is possible; when the artist chose ferrous oxide instead, for example, only a very rough date can be estimated, based on stylistic clues that now appear frustratingly stable. One wants change, if only in order to help establish chronology. If a key feature of art is innovation, then we might do better to see the Paleolithic figures of horses and lions not as art but as a natural manifestation of humanness, generated by an encoded species-specific behavior like avian nest construction.

One seriously regarded, if highly contested, theory of the origins of stone hand-axes or bifaces among our hominid ancestors (long before the cave art of our fellow Homo sapiens) has it that these were not tools at all, but rather an evolutionally determined universal behavior among male hominids, a product of sexual selection that signaled to females of the species,the makers’ selectively advantageous mastery of symmetrical forms.4 If they could produce a perfect biface in stone, then potential female mates might believe that they could also produce bilaterally symmetrical offspring. We see similar activity in the males of a number of bird species, whose members build special mating nests, the principal purpose of which is to impress females with their talent for symmetry, their ability to reproduce in artifice what is already written into the body of every vertebrate.


SOME SPECIALISTS in parietal art, notably Michel Lorblanchet, insist that the most basic method of research in this area consists of personally copying, by hand and as accurately as possible, what is observed on the walls.5 Notoriously, the Lascaux cave—having long been closed to visitors, whose presence, whose very breath, would quickly degrade the site’s masterpieces—features a near-perfect simulacrum for tourists to visit. Lascaux II is often derided as the culmination of a dispiriting trend toward artifice in the contemporary world already analyzed by social critics like Jean Baudrillard, yet in fact the site has great value for researchers seeking to understand the original Lascaux. The production of the simulation enables specialists to approximate what the early modern philosopher Francis Bacon thought of as “maker’s knowledge,” which, in this case, is extended from the understanding of artifacts to the most adequate knowledge one can have of the natural world.

Ten thousand years or so divide the last European cave art from the rise of textual traditions telling us what ancient (though far from prehistoric) people felt and thought when they visited caves. Is it prudent to make use of these sources to help us imagine our way back in time, across the gap of millennia? No, but we must use whatever resources we have at our disposal. Historians have access to texts that supplement and give a first-person account of the world of the past people who interest them, not least of the material vestiges these people left behind. But prehistorians, having no such supplement, must struggle to work their way back to a plausible first-person perspective. In so doing, they must also rely on the faculty of the imagination—a dangerous stimulant in the best of cases, and often a gateway to recklessness. Caution, therefore, is advised.

Among the Greeks it was common for seers of various sorts to descend into caves in order to induce ecstasies and visions there in the dark silence.6 Later, the Roman polymath Pliny the Elder mentions “the priestess of the Earth,” who, “when about to prophesy, drinks bull’s blood before she goes down into the caves.”7 The ancients also noticed the transformative effect that a cave has on the conscious mind when the walls are partially illuminated by flickering light. We may have been too quick to take Plato’s allegory of the cave, perhaps the best-known thought experiment in the history of philosophy, as a mere allegory, whose events transpire in a cave but illustrate a more general point about knowledge and cognition. Could it be that Plato was in fact speaking from a familiar perception, shared with his readers, regarding the particular way the mind works precisely in the condition of encavement? 

Pliny, who believed mining to be a transgression against nature, a rape of the earth driven by greed, mocked those who discerned, in the depths of the earth, figures and shapes that are really only figments of their deranged minds. In the Middle Ages, there was a propensity to take these figures as quasi-real, as a subclass within the broad category of “games of nature,” which also included fossils: forms of beings generated in a material substratum of rock, which for its part is incapable of carrying them to completion as living and moving animals. They are animals forever frozen in stone as a result of the over-fecundity of Earth. Not only the earth but its caves, too, are gendered female, and associated with the female reproductive system. In Classical antiquity, caves and mine shafts were often called the matrix. The term (literally “pregnant female animal”) is sometimes translated as the “bowels” of the earth, but more correctly understood as its “womb.”

It was only in the seventeenth century that systematic arguments finally did away with such games of nature. In a remarkable text of 1636, discussed at length in Gaston Bachelard’s La Terre et les rêveries du repos, Pierre-Jean Fabre confronts us with a description of the features of a cave wall that leaves us long uncertain as to whether he is describing bare geomorphology, or rather the first certain identification of Paleolithic art in the modern period.

I saw in the caves and caverns of the earth in the country of Languedoc close to Sorèze, in a cave popularly known as the Tranc del Caleil, traits of sculpture and of imagery more perfect than one could have hoped. The most curious can go and see them; they will observe, inserted in and imposed on the rocks, a thousand sorts of figures that ravish the gaze of spectators.8

But our hope that Fabre might have been in the presence of true artifice is ultimately vanquished: “Never did a sculptor enter there to carve and to chisel the image . . . which should lead us to think that Nature is endowed with gifts and with knowledge that its Creator gave it in order to know how to work in diverse ways, in all sorts of materials.”9

By the time of philosopher G.W. Leibniz’s Protogaea of 1691–93, the separation between natural forms and mental fantasies was complete, and ensured that the perception of figurations in cave walls could only be the stuff of fiction. “As to the supposed appearance of the pope’s tiara . . . ,” he writes, “and all sorts of other shapes etched in the stone of Eisleben, I consider these to be, not games of nature, but of the human imagination, which sees battles in the clouds and hears its favorite melodies in the sound of bells or the beating of drums.”10

This purge of imagination from our reading of cave walls may have considerably delayed our proper identification of Paleolithic parietal art until the late nineteenth century. The prehistoric artists, we may responsibly conjecture, given that we know they had the same cognitive-perceptual apparatus as we do, were only using their imaginations, reading figures into the rock forms that were not, strictly speaking, there. Given the lingering memories of animals seen outside the caves, they were predisposed to make those forms reappear, just the way the human cognitive-visual apparatus always works—in part by receiving what is there and in part by complementing it with what we have already seen, what we expect, what we want.

Why were our progenitors seemingly blind to so many other features of the natural world? We do not know. Cognitive scientists today speak of “plant blindness”:11 we see an image of a lush green landscape with two animals frolicking in the middle of it. What does our cognitive-perceptual apparatus pick out as most salient? Not the plants. Chimpanzees have been observed awe-struck before waterfalls. Were not Paleolithic humans also impressed? They may have been, but they left no record of this impression on the walls of the caves. Their imaginative life, if the figurations on the walls are any accurate measure of its variety, was singularly centered on their fellow animate beings, the kind that move about, that breathe and perceive, just as they experienced themselves doing. Why such a limited range? Again, we do not know.


IF I HAVE DWELLED at some length on the cognitive history and phenomenology of human beings in caves, it is because I agree with Bachelard that much can be learned from meditation on our experience of elements and milieus. It is the poetics of the space that explains the particular poiesis within it, whose material traces we seek to understand.

You cannot understand Paleolithic cave art if you have not seen it in situ. This is not, or not only, because the real thing reveals details undetectable even in high-resolution images; it is rather because the works gain their sense from their milieu. Cave art is not art that happens to be found in caves. Even if we do not know what its function was—shamanism? a visual aid for storytelling? a record of hunting successes?—we can nonetheless be confident in conjecturing that at least many of the animals were painted into the walls of the caves because they were first seen in the walls of the caves, and that the Paleolithic artists saw these walls in much the same way that Fabre describes them.

The guides at the Grotte des Merveilles are aware that the speleothems of their cave, the Escherian columns and organlike flowstone, are more commanding than the faded animals and scattered dots, and they fill their narration with invitations to meditate on the grotesque forms that seem present in the place. An old man’s head emerges out of the floor, the stub of a stalagmite. The floor appears littered with calcitic phalluses.

The art in the Font-de-Gaume cave was discovered in 1901, and after inspection by the great prehistorian Henri Breuil this marked the beginning of French scholarly study of parietal art.12 The Altamira cave in Spanish Cantabria had been discovered in 1868 and written about in 1880, but the following two decades witnessed bitter controversy, often underlain by national prejudice, about the true origins in time of Altamira’s remarkable representations. For years prior to Font-de-Gaume’s “discovery” local farmers had used the cave to shelter their draft animals. Did they notice the images on the walls? It is possible that they did, but that this did not seem particularly noteworthy to them; a nineteenth-century French farmer likely would have presumed the paintings to be old, but not old in any way worthy of solemnly respectful preservation, of state intervention for the protection of a national landmark and of cultural heritage, even of primordial proto-Frenchness.

In the deep recesses of Font-de-Gaume a horse appears. Under the installed electric light the horse is easy to spot, and it appears as a two-dimensional painting. When the light is turned off, and replaced by the faint beam of a flashlight, the lines of the painting itself disappear, and what instead emerges is the vivid three-dimensional impression of equine haunches. The painting fades, and the sculpture appears, and we see again that the former was made only to draw out the latter.


IT HAS BECOME a truism in the study and in the public presentation of Paleolithic cave art that we will likely never know what it means. This seems an exaggeration of healthy caution. We know, already, that a representation of a bison “means,” among other things, the bison represented. We know that Paleolithic artists were singularly interested in the nonhuman mammals with which they shared an environment. Wide-focused quantitative surveys of forms and patterns have moreover revealed much that European cave art is not. Unlike Australian and southern African art from the same broad period, cave art in Europe offers no depiction of landscape, no horizon, no vegetation, almost no depiction of human-animal interaction, almost no hunting scenes, no obvious interaction between different species of any sort (though there are overlapping species painted at different times). If we exclude the nonrepresentative symbols, then, the focus of European cave art is remarkably narrow: it is the depiction, more or less naturalistic though removed from the environment, of various species of megafauna. Alain Testart has gone so far as to attribute a taxonomic function to the pictures, isolating and differentiating the various natural kinds.13

Nor do the regularities stop there. The relative absence of hybrid creatures, particularly of human-animal combinations, speaks against the interpretation of cave art as serving a shamanistic function. In the absence of horizons, there appears to be no interest in European caves in separating the world above from the world below (though the caves themselves may have been seen as portals to the underworld, thus obviating the need for representation of such thresholds), and so no separate realms for the shaman to mediate. There is the rare depiction of what may be a biped with antlers at Trois Frères, and at Pech-Merle there are human female figures that seem frozen halfway in metamorphosis into four-footed bovids. But to imagine a woman on all fours is not necessarily to entertain fantasies of the sort that cannot be realized in this world.14

Conspecifics are moreover depicted together in limited and stereotyped ways. They are typically either facing one another, or facing the same direction, or, finally, arranged en éventail, in the form of a fan, with several individuals superposed, each facing the same way, with the lower ones extending only somewhat further forward than the higher ones. It is likely that the animals depicted in the latter arrangement are not several individuals but one, and that the repetition is intended to convey motion. This effect is most evident when a flickering light is moved over the wall of the cave, and it is difficult not to imagine that this effect was often conjured in connection with storytelling. The prehistorian Marc Azéma has been the most active over the past several years in arguing for this kinetic or cinematic function of much cave art,15 and by now the idea of the figures en éventail as short cartoons, as “Paleolithic GIFs,” enjoys considerable currency.

Perhaps. But we should not become overconfident and abandon critical skepticism altogether. In any case, there are many figurations besides those en éventail, and no single theory of European cave art’s function can possibly account for all its various conventions and figurations. It is the supposedly nonrepresentational forms that, for obvious reasons, are the most resistant to interpretation. Some of them seem traceable to representations: just as our letter “A” began with the Egyptians as a bull’s head, the aviform signs seem to have begun as birds, even though in their developed form they look nothing like birds; and the triangles, with or without a line down the middle, seem to be abstractions of vulvas, and thus, no matter how minimal and stylized, to have to do with fecundity, life, or something else within that broad semantic cluster. But signs are never so stable as we imagine, even when they are anchored in something as basic and slow-changing as biological reproduction. It is at least possible that we are correctly identifying the vulvas of Paleolithic cave art as vulvas but missing entirely what these may have meant for their creators.

A plausible case of such misinterpretation arises with the figure of the “Wounded Man” of Pech-Merle. A rare depiction of a human form, probably a man, perhaps sporting a tail, and perhaps a bird’s head, though one far too big for his body, the Wounded Man also seems to be transpierced by arrows or spears. Four shafts penetrate his torso and come out the other side. Or perhaps there are eight shafts. There is simply not enough Paleolithic art representing spears in bodies for us to know what the conventions for representing transpierced figures might be. There is very little art representing animals wounded by human hunters: this would amount to interaction between species, which, again, is largely unknown in the tradition—an absence that plainly indicates we are observing the material remains of a lost sign system, in which rules we cannot now fully reconstruct dictated that some things could be represented but not others, and some together with others, and some not. And so the Wounded Man is called by this name only in deference to the earliest prehistorians, who often took for granted that conventions are durable across the ages.

The first French scholars of cave art were mostly abbots and priests, and they saw their work as contributing to the vast project of Catholic apologetics, though fueled also, no doubt, by a considerable dose of worldly curiosity. Their explicit concern was to find, in the symbolic world of early inhabitants of the continent that would much later become Europe, evidence for an awareness of the existence of a higher realm beyond the senses, and thus of a mental and spiritual capacity that could eventually accommodate the truth of Christianity. Over the twentieth century the discipline would develop into a proper science, yet an unusual one. To master it a student must know something of radiocarbon isotopes, of psychoanalytic theories of genital symbolism, and of everything in between. But in France the project has generally remained a variety of spiritual apologetics, however superficially secularized: one of interpreting prehistoric material traces as signs of a break with nature, rather than as evidence of our continuity with nature, of human inhabitation of a particular environment, and of symbiotic relations with other beings within that environment.

Whatever reasons we have to doubt the durability of signs across the ages, we may be certain that the cave-art mammoths are mammoths and the bison, bison. This is not convention, but visually persuasive representation, and the testament of this tradition’s undeniable naturalism and theriocentrism. Scholars have shied away in recent decades from the view that this preoccupation with animals, and with their artistic representation, amounted to a magical practice, or to ritual invocation of the spirits of the beasts for shamanistic purposes. It is just as likely that the representations were supplements to a cultural practice of storytelling, aided by images that appeared to move along the walls under the flickering flame, for no other reason than that cave artists were, as we are, members of the species Homo narrans: people who tell stories. There is a sort of magic when we go to the movies, too, after all. Exterior-world beasts, reproduced in caverns, brought there by human beings intent on conjuring in their imaginative realm—their interior life in a double sense—what mattered most to them in the external domain, in reality: this is what is certain in Paleolithic cave art. This is in part, but only in part, what cave art means.