This theory dates back to the French archaeologist Salomon Reinach, but was particularly promoted by the pioneering Abbé Henri Breuil. Making analogies with surviving hunter-gatherer peoples, Breuil and others contended that Paleolithic art represented a kind of hunting magic. This theory was later taken up by semi-Marxists and Marxists such as Hauser, Fischer and Sánchez Vázquez. There are various problems with it, but let us begin with the theory.
The magic theory
The theory starts with the proposal, convincing in itself, that it is unlikely that Paleolithic cave paintings were created to decorate the places where people lived. Paleolithic people did not live in caves, or only occupied the entrances of them, whereas some of the paintings are hidden in dark corners and can only be seen by lying down. They are also painted at conflicting angles and even on top of each other, which would ruin any decorative effect. Hauser articulated it well:
A whole series of indications argues against such an interpretation, above all the fact that the paintings are often completely hidden in inaccessible, absolutely unilluminated corners of the caves where they would have been quite impossible as ‘decorations’. Their palimpsest-like superposition, destroying any decorative effect from the very outset, also argues against such explanations. After all, the painters were not forced to paint their pictures one over the other. They had space enough. This very superposition of one picture over another points to the fact that the pictures were not created with any intention of providing the eye with aesthetic enjoyment but were in fulfilment of a purpose in which the most important element was that the pictures should be accommodated in certain caves and in certain specific parts of the caves...
Speculation naturally followed that this art had a ceremonial or spiritual significance. Hauser sees the inaccessibility of the paintings as proof of a ‘magical’ purpose.
The magic theory contends that during the Paleolithic period there was none of the complex separation of art and reality that occurs later. The act of acquiring power by imitation was extended by Paleolithic hunters and artists beyond tool-making and language, until there was no separation between the image of the animal and the animal itself. Hunters hoped that by making an image of an animal, they would acquire power over it. In Hauser’s words:
The Paleolithic hunter and painter thought he was in possession of the thing itself in the picture, thought he had acquired power over the object in the portrayal of the object. He believed the real animal actually suffered the killing of the animal portrayed in the picture. The pictorial representation was to his mind nothing but the anticipation of the desired effect; the real event had inevitably to follow the magical sample-action, or rather to be already contained within it, as both were separated from each other merely by the supposedly unreal medium of space and time... It was not the thought that killed, not the faith that achieved the miracle, but the actual deed, the pictorial representation, the shooting at the picture, that effected the magic.
Many of the cave images show animals surrounded by arrows and spears, or by human hunters, and they were even stabbed or shot at.
Horse from Lascaux. It appears to be under attack from spears or arrows.
Art was expected to assist in the hunt, and owes its finely-observed naturalism to the need to recreate the animals as closely as possible. The animals are drawn from the side, where their form was most apparent and complete; where they had distinctive features like antlers, the head would be turned so that these too could be seen at their most representative angle. Dance would have originated in the same way, as a group imitation of the animals that we would shortly hunt.
This might explain why human beings are depicted far less often than animals in Magdalenian paintings, and then in a less naturalistic form. Given the supposed power of an image to correspond with a real being, they may have worried that creating images of people could expose them to harm. The power of this correspondence between image and living person has never quite left us: as famously explored by Oscar Wilde in The Picture of Dorian Gray.
Paleolithic magic would not have resembled religion as we understand it today, with its elaborate rituals, scriptures, esoteric myths and hierarchies. There were neither the resources nor the social organisation for that. The magic was as direct a part of the hunt as the sharpening of a spear. This art was part of our determination to exert control over our environment: art, spirituality and science were one and the same.
Problems of the magic theory
One reason we need to examine the magic theory is that it was taken up by several Marxist writers in the first half of the twentieth century. Marxists worth their salt look to contemporary scientific thought, and sympathetic magic was for a time the predominant theory among archaeologists in interpreting Paleolithic art. “Any other explanation of Paleolithic art,” wrote Hauser, “…is untenable.” Fischer wrote that “The magic of tool-making led inevitably to the attempt to extend magic to infinity… Art was a magic tool.” In 1965, Sánchez Vázquez confidently asserted: “Today no one has any doubt about the close relationship between art and magic in the Upper Paleolithic period.” Max Raphael, in his influential book Prehistoric Cave Paintings, proposed that Paleolithic art was totemic, adding: “that totemism and magic co-existed in the world-view of the paleolithics in a specific manner is indisputable.” All of these writers either drew or are likely to have drawn theoretical backing from the writings of the Marxist archaeologist Gordon Childe.
The attraction of the theory for Marxists at that time is obvious: there is no doubt that ‘making alike’ played an essential part in our aesthetic development, and the magic theory is a plausible extension of it. Above all it emphasises humans acting to take control of their environment. But our archaeological knowledge and technique is much better today, and it is clear that there are limitations to what the theory can explain.
The first problem is that it relies upon ethnological comparisons with modern hunter-gatherer societies. Such societies are by their nature exceptional rather than typical. Most societies progress socially and technologically, so any that remain in the most ancient mode of production are anachronistic by at least 5,000 years. Even when a society is in arrested economic development, meanings will shift and change across thousands of years: modern Aborigines for example are unable to explain to researchers the art of their Paleolithic ancestors. We simply cannot know what Paleolithic people believed, because they left no written explanations of the meanings they attributed to their images.
Another problem is that the inaccessibility of parietal art, often claimed as evidence of ritual or magical purpose, is in fact exaggerated — many images are not in especially obscure places, especially at African sites. R Dale Guthrie suggests that Paleolithic people quite possibly did paint the great majority of their paintings in accessible public places  — it is just that the works in protected, inaccessible locations are the only ones that have survived.
Even if sympathetic magic was important to one Paleolithic culture, it may not have been for others. It does not explain the presence of non-prey animals (the Chauvet cave features a hyena and a leopard), or of abstract symbols and handprints. Not all the images of prey animals are highly lifelike. The painters of Lascaux hunted reindeer and left their remains in the caves, but the caves have no images of reindeer at all. It does not explain what to make of the statuettes of women — an association with encouraging fertility can not be assumed, as this was an impoverished society that probably wished to limit population growth as much as to increase it. The wary avoidance of the human image seems contradicted by the 155 engraved human portraits found in on the floor of a cave at La Marche in France , and although rare in Magdalenian paintings, they are not uncommon outside Europe. And alongside the fine cave paintings, there is a mass of other images, engravings and ornaments that is the stuff of everyday life and refuses to fit into the scheme of the theory of hunting magic.
A figure at Le Gabillou in France, which appears to be half-animal and half-human and has been described as a ‘sorcerer’. A similar figure was found at Les Trois Frères in the French Pyrenees. Such images may well be magical, but hunting magic does not seem adequate to explain them.
Hunting is only one element in Paleolithic art. There is imagery that suggests an understandable interest in sex, and perhaps rituals of fertility. The half-human, half-animal beings may be totemistic, as Raphael suggested. Ethnographic parallels have led to the proposal — for example from French prehistorian Jean Clottes and South African archaeologist David Lewis-Williams — that cave art is not magical but shamanistic: i.e., that they were part of a ritual that induced a trance-like state to make contact with the spirit world. As shamans danced themselves into an altered physiological state, they experienced brightness, patterns of light and shapes, and saw animals within those shapes. Paleolithic parietal art was the attempt to reproduce hallucinatory imagery.
Before we dismiss superstition and the mass of false beliefs that accompany it, we must remember that Homo sapiens was still a new kind of being, with an extraordinary consciousness which we were just beginning to try and understand. Our attempt to find meanings were derived from our lived experience at that time. Although magic is a fantasy, the idea of correspondence is still in use today, and not just in religious practice — we see it every time we burn a figure in effigy. And the correspondence of a drawn symbol with a word, which in turn represents a thing, is a requirement for the invention of writing.
The problems with the theory of sympathetic magic do not necessarily mean that the Marxists who adopted it were entirely wrong and that magic has no relevance whatsoever to Paleolithic art. It is the promotion of it in ways contradicted by evidence that is untenable. The case in support of the magic theory is well argued — but it can only be part of the answer. We must remember Randall White’s observation, cited in an earlier article, that no interpretation of Paleolithic art can be applied to every work.
In one sense there is no need to struggle for an explanation for what this art was ‘for’. Nobody would suggest that modern art had only one motivating cause, and there is no need to try and find one for Paleolithic art either. A cave painting created to assist the hunt may happily exist alongside a face whittled from bone for the pure fun of it. Art is humans’ attempt to affirm their intellectual, emotional and physical powers in concrete, sensuous form. As such, it is as diverse as humanity itself.
 Arnold Hauser, vol. 1 of The Social History of Art (1951). Hauser’s over-simplified view of prehistoric art has become dated.
 Ernst Fischer, The Necessity of Art (1959).
 Adolfo Sánchez Vázquez, Art and Society (1965).
 Max Raphael, Prehistoric Cave Paintings (1945). Max Raphael was a German art theorist who moved to France and then the US to escape Nazism. Adopting Marxism in the 1930s, he offered us some of the most intelligent and sensitive Marxist writing on art. This essay was an important part of a new generation of work in the 1940s and 1950s that sought to go beyond early interpretations of Paleolithic art. He proposed that the paintings were much more structured than previously thought.
 See R. Dale Guthrie, The Nature of Paleolithic Art (2006).
 I should add that there are doubts about the authenticity of these images, which would be remarkable if genuine.