CHAPTER 2 Seeking Answers

The end of the controversy over the antiquity and authenticity of Upper Palaeolithic art simply started another one that has endured to this day. Why did Upper Palaeolithic people make these images? The philosophical and religious turmoil that still rages around Darwin’s theory of evolution suggests that large, emotive issues are involved in all aspects of human origins.

Perhaps the most insidious of these issues, one that covertly skews our thinking, is our use of the word itself – ‘art’.¹ It is one of those terms (others crop up in subsequent chapters) that everyone believes he or she understands – until asked to define them. All too readily people assume that ‘art’, as they understand the term, is a universal phenomenon, and they tend to ascribe not only the word itself but also all its connotations to non-Western contexts. As a result, we have come to think of ‘the artist’ as a special kind of person who, because of some universal, almost mystical, principle of inspiration, is found in all societies. But notions of ‘art’ and ‘artists’ are formulations that are made at specific points in history and in specific cultures. For instance, ‘art’, as we think of it in present-day London, New York or Paris, did not exist in the Middle Ages, when people did not distinguish between ‘artisan’ and ‘artist’.The notion of inspired individuals who, by their almost spiritual status, are set apart from ordinary mortals is a concept that gained acceptance in the more recent West during the Romantic Movement (c. 1770–1848), a period when writers and philosophers asserted the ascendancy of individual experience and a sense of the transcendental.

Some writers on Upper Palaeolithic art, such as the Berkeley archaeologist Margaret Conkey, ² rightly highlight the dangers of importing Western connotations by the unthinking use of ‘art’. Certainly, the word has misled many researchers to understand Upper Palaeolithic imagery in terms of Western art. But, that said, I believe that we can become over-sensitive to this problem. ‘Art’ is a handy monosyllable, and, provided we are aware of the dangers of its Western connotations, we can use it with caution.

The various ways in which researchers have explained Upper Palaeolithic art constitute a long historical trajectory that moved from simplicity to more complex hypotheses and then on to the collapse of the whole interpretative enterprise, and thus to present-day agnosticism. Today many researchers believe that it is impossible to know what Upper Palaeolithic art was all about. I start with an apparently simple catch-phrase that brings into the foreground the problematic word.

Art for art’s sake

At first, notions of early ‘savage man’ tended to discredit any suggestion that Upper Palaeolithic images had a symbolic purpose. Art was simply decorative, and it would be foolish to see abstruse symbolic meanings in the images of savages. As early as 1864, Édouard Lartet and Henry Christy, a wealthy English banker who financed much of Lartet’s work, tried to get around the problem of such sophisticated portable art coming from people who were, by all accounts, startlingly primitive. They argued that the environmental conditions of the Upper Palaeolithic spawned an abundance of animals that made hunting easy and so, despite their primitiveness, people had plenty of leisure time during which they could decorate themselves, their tools and their otherwise drab living spaces. Art, then, was born of leisure: the images were made for simple enjoyment, fun and decoration once people had achieved a measure of control over their environment. Importantly, Upper Palaeolithic art had no symbolic content and was, in accordance with Romantic ideals, essentially an individual rather than a social activity.This rejection of symbolic interpretation of Upper Palaeolithic art and emphasis on individual ‘inspiration’ arose from the belief that such primitive people could not possibly have had a religion.

The idea of a link between leisure and the development of aesthetic practices was taken up by Édouard Piette in his study of mobile art. He deployed bouquets of romantic phrases, such as ‘exclusively artistic’, ‘seeking perfection in art’ and ‘eternally concerned with the cult of beauty’.³ Illustrations of the time show, principally, bearded Upper Palaeolithic men absorbed in solemn contemplation in the making of art, while women kneel nearby, assistants or acolytes rather than role-players.⁴ This idea of art as something exalted and something that could be achieved only after the natural environment had been tamed, is still kept alive by textbooks that argue that art is the fruit of an ‘aesthetic sensibility’, an innate desire to produce beautiful things, that can be indulged only in leisure time. This is the explanation for Upper Palaeolithic art that became known, not altogether correctly, as art pour l’art, art for the sake of art.

The art-for-the-sake-of-art explanation did not enjoy prolonged acceptance after it was first mooted. In 1987, long after it had been abandoned, John Halverson, an American art historian, tried to revive it, albeit in a much more sophisticated form, but he met with little success.⁵ From the beginning, there were two principal reasons why early twentieth-century researchers began to doubt art pour l’art as a general explanation for Upper Palaeolithic images. First was the discovery of more and more subterranean art. It was understandable that people would decorate their weapons, wear pendants, and make statuettes; after all, these were activities with which researchers themselves were familiar. But why would they make works of art in inaccessible, dark, underground places? The seemingly obvious fact that art is made to be looked at is contradicted by deep Upper Palaeolithic cave art.

Secondly, Sir Baldwin Spencer and F. J. Gillen’s 1899 ethnographic report from Australia told of people who lived under very harsh conditions but who nevertheless made a vast amount of complex art in rock shelters. Cartailhac himself also pointed to the ‘Bushmen’ of southern Africa as a people who made art in circumstances that he believed (incorrectly) left little time for leisure. Then, too, in the 1890s, Sir James Frazer’s compendious collection of ‘primitive’ folklore from around the world, entitled The Golden Bough (a phrase he derived from one of Turner’s paintings), began its influential career; it did nearly as well as On the Origin of Species. Frazer (1854–1941), a man of astounding scholarship and erudition, showed conclusively that ‘primitive’ people did have religion, though it has to be said that his monumental study is a mishmash of little understood exotica.⁶ All this evidence, first from deep cave art and then from Spencer and Gillen, and Frazer, seemed to point to the, for some people, unpalatable conclusion that Upper Palaeolithic people did, after all, have a religion of some kind. Art in such societies was therefore not necessarily art pour l’art.

Today we recognize other problems with the art pour l’art explanation. First, from the point of view of logic, it may be expressed thus:

1 How do we know that human beings have an innate aesthetic sense or drive?

2 We infer it from the existence of what we believe to be the beautiful creations of the earliest art and from objets d’art around the world.

3 We then use this inferred sensibility to explain the existence of art.

The argument is thus clearly circular: a state of mind is inferred from the art, and then used to explain it without any other supporting evidence.⁷ Some researchers may question this conclusion by pointing to the symmetry of very ancient – as much as one million years old – stone hand-axes and by suggesting that an aesthetic sense originated in an in-built predilection for symmetry. The idea is engaging but not convincing. If a desire for symmetry is indeed inbuilt, there is no reason to suppose that the symmetry of the hand-axes had anything to do with ‘aesthetics’; the symmetry more probably resulted from (not necessarily linguistically) learned motor skills than a desire for beauty. In any event, there is next to no evidence in Upper Palaeolithic art for an interest in symmetry.⁸

Moreover, art historians have developed much more penetrating explanations for the making of art. At the root of many of these explanations is the realization that the making of art is a social, not a purely personal, activity. Art serves social purposes, though it is manipulated by individual people in social contexts to achieve certain ends. Art cannot be understood outside its social context.

Another point follows directly from this one. It is not possible for someone to depict an animal without in some way giving resonance to part of its symbolic associations, and those associations are socially created and maintained. For instance, in Western thought the lion has associations of regality and strength, the lamb of innocence and gentleness. The lion does not stand for deceit, nor the lamb for treachery, no matter what an individual person may believe. If an Upper Palaeolithic artist were to depict an animal because it had exclusively personal and arcane associations for him or her, those associations would be lost on most viewers who would, inevitably, see the picture in terms of other, socially sanctioned, associations. True, as we shall see, individuals manipulate images and their meanings for their own ends, but they do so within limits established by society; otherwise their new ‘take’ on a symbol would be totally obscure to viewers.

Essentially, it is the context of an image that focuses its meaning. For instance, a depiction of a lamb on the wall of a nursery will resonate differently from a lamb in an ecclesiastical stained-glass window.Yet perhaps not entirely so, for the symbolic associations of animals are all interrelated; they constitute a richly affective spectrum rather than a conglomeration of separate meanings. William Blake wondered at the ‘fearful symetry’ of the tiger that burned brightly in the forests of the night, while T. S. Eliot startlingly took up another segment of the tiger’s semantic spectrum: ‘In the juvescence of the year Came Christ the tiger’.

In all societies, many animals have rich associations of this kind, and it seems highly improbable that Upper Palaeolithic people would, repeatedly and in remote contexts, have depicted horses and bison if those creatures did not have some non-trivial, shared meaning for them.Indeed, the narrow range of painted and engraved subjects suggests social norms rather than personal predilections. Upper Palaeolithic artists were, essentially though not entirely, bound by rules of custom.

These thoughts on a supposed innate aesthetic sense in Homo sapiens, a species-specific drive that led to the broad set of objects and behaviours that we term ‘art’, has been summed up in a striking way. At the beginning of his immensely popular The Story of Art, Ernst Gombrich wrote, ‘There is really no such thing as Art. There are only artists.’ That is the view that I take in this book. I seek a practical, down-to-earth goal: I enquire about the origin of image-makers (‘artists’, if you will), not the genesis of some impossible-todefine philosophical concept. I argue that the first image-makers were acting rationally in the specific social circumstances that I describe in later chapters; they were not driven by ‘aesthetics’. That amorphous, changing concept that we, in our particular place in history, call ‘art’ came after the first imagemakers.

The next attempt to understand the meaning of Upper Palaeolithic art was far more enduring than art pour l’art. It has indeed entered popular thought about ‘cavemen’ and has given rise to innumerable cartoons. It is also the explanation that is most commonly encountered not only in popular accounts of Upper Palaeolithic art but also in textbooks.

Totemism and sympathetic magic

At the beginning of the twentieth century, and in the years immediately following, anthropology and sociology were coming into their own as respectable fields of study, and popular interest in ‘primitive’ people around the world was growing as a result of European imperial expansion. Writers such as Franz Boas, Henry Schoolcraft and Lewis Henry Morgan in America, Sir Henry Maine, Sir Edward Tylor and Sir James Frazer in England, Sir Baldwin Spencer and F. J. Gillen in Australia, Arnold van Gennep, Marcel Mauss and Émile Durkheim in France, to mention but a few, were prominent in this tidal wave of interest. In 1865, Tylor (1832–1917) had sensed an affinity between magic and prehistoric art.⁹ But the writer who most effectively turned the insights of anthropology and ethnography to an understanding of Upper Palaeolithic art was Salomon Reinach.

In 1903, the year after Cartailhac’s mea culpa, Salomon Reinach (1858–1932), who was at that time curator of the Musée des Antiquités at Saint-Germain-en-Laye, published an article entitled L’art et magie: à propos de peintures et des gravures de l’Age du Renne in the journal L’Anthropologie. He proceeded from the fundamental assumption that the only way in which we can gain some understanding of Upper Palaeolithic art is by examining the lifeways of existing ‘primitive’ peoples. Here, right at the beginning of Upper Palaeolithic art research, Reinach created a problem that still dogs the footsteps of all workers: is it possible to understand Upper Palaeolithic art without recourse to analogy?¹⁰ By using analogies, do we not simply create a past in the image of the present?

Reinach tried to handle the problem by insisting that researchers should strictly limit their analogical sources to hunters and gatherers – those people who, in his view, are closest to the way in which Upper Palaeolithic people lived. He was a good deal more cautious than some of his successors. Nevertheless, it is today recognized that twenty-first-century hunters and gatherers are not living fossils of the Upper Palaeolithic but people with a long history. Still, it has to be admitted that there are commonalities among huntergatherer communities that probably bear some resemblance to the ways in which Upper Palaeolithic people lived. Being sure of exactly what those commonalities are and which features are culture-specific is the problem.

One of the ideas to emerge from a reading of the Australian ethnography was totemism. Today, researchers recognize that the word has been illdefined, ¹¹ but at the beginning of the twentieth century it had a fairly straightforward meaning. The word itself derives from a North American Indian Ojibwa word that denotes an animal, or sometimes a plant, that is the emblem of a clan. One could speak of ‘Eagle-people’ or ‘Bear-people’. Similar beliefs and social entities are found in Australia. Armed with this information, some researchers suggested that the images in the caves were totems. But the explanation did not acquire many supporters.¹² If the depicted animals were totems, so the counter-argument ran, one would expect to find clusters of images of single species, each cave being a kind of ‘totem-pole’. But, came the response, if a few totemic groups constituted a single residential group then a comparatively small number of species would be found together – which we do find. The argument seemed to go round and round.

Reinach, however, was taken with other information that Spencer and Gillen published in 1899 about the Australian Aboriginal Arunta people. According to those two writers, the Arunta painted pictures of certain creatures in the belief that their actions would cause the depicted species to breed and multiply. It seemed to Reinach that the animals depicted in the European caves were the very ones that Upper Palaeolithic people would wish to multiply. It was like playing a game of ‘snap’: two similar cards turned up, one in Australia, the other in western Europe. The match was convincing, and so Reinach inferred similar motives for the production of art in both contexts.

Later, the Abbé Henri Breuil and others extended this hypothesis to hunting magic. The images, they argued, were intended to give hunters power over their prey. This idea, of course, fitted in well with what people in Europe knew about witchcraft, wax dolls and pins.With, perhaps, witchcraft in mind, Breuil claimed that many images had spears or projectiles sticking into them. Perhaps, he thought, the act of painting these weapons effected the death of a real animal. Breuil may well have been especially influenced by Niaux, the cave that de la Vialle and many others visited in 1660, and even before the seventeenth century. Here there are indeed a number of paintings of animals with what could easily be taken to depict weapons stuck into them. But, overall, only 15 per cent of Upper Palaeolithic bison images seem to be wounded or dying.¹³ Most appear to be alive and well.

At first, the explanation seemed to explain why so many of the images were hidden in dark chambers: magic would be performed out of sight. Then, too, the explanation seemed appropriate to the species depicted in the caves.When images of felines and other dangerous animals were found, ones that it seemed unlikely people would consciously hunt or wish to increase, Breuil explained their presence by saying that the artists hoped to acquire the strength and hunting skill of the predators. The elasticity of the explanation did not end there. Quadrangular signs, such as those in Lascaux (Fig. 58), were said to depict traps into which animals would fall. Other signs were taken to be hunters’ hides or the dwellings of spirits.

Today, there is little point in arguing all the pros and cons of either art pour l’art or sympathetic magic. To be sure, the image-makers may well have taken pride in their handiwork, and hunting, a hugely important Upper Palaeolithic activity, may well have had something to do with image-making. But so many of the arguments for these two explanations can easily be turned on their heads. For some writers, the complex, entangled superimpositions of many engraved panels of images obliterated any possibility of aesthetic appreciation. But that is a purely Western supposition: the confusion of lines and difficulty of decipherment may, for all we know, have been considered aesthetically pleasing.Then, too, some writers argued that leisure time was necessary for the making of art, and that art would not have been made during seasons of privation and incessant searching for food. Others responded that it could equally be argued that rites of sympathetic magic would be performed precisely during those difficult times in order to ensure an adequate food supply. Similarly, depictions of weapons may be taken as an indication of hunting magic; but, then, where no weapons are shown, it may have been the simple act of making the pictures that was considered efficacious. Either way, the explanation emerges unscathed. All the debate that went on through the twentieth century certainly exposed the weakness of these explanations, but, ultimately, as in many research contexts, explanations are discarded only when better ones are devised.

Making judgments

How does one recognize a ‘better’ explanation? If we are to understand the rollercoaster history of Upper Palaeolithic art research in the twentieth century, we need some insight into how science advances – or fails to advance. Broadly speaking, there are two schools of thought on this matter; they are not necessarily exclusive, though they are sometimes presented as if they are. On the one hand, there are researchers totally committed to the objectivity of science; they believe that there are established rules and procedures of verification that lead to sure knowledge. At the other extreme, there are those who see all knowledge claims as relative and the product of social forces and cultural perceptions.

First, it needs to be repeatedly emphasized that no explanation of Upper Palaeolithic art can ever be ‘proved’. As we saw in Chapter 1, Darwin did not claim to have proved his explanation of the mechanism of evolution. Proof is a concept that should be restricted to closed systems, such as mathematics. At the end of an algebraic problem, a student can write ‘Q.E.D.’ – that which had to be demonstrated. But in archaeology, nothing of interest will ever be proved in that way. Philosophers of science have devoted much attention to this problem. We cannot now enter into a detailed account of scientific methodology, fascinating though that field is, but before we proceed to the far more complex explanations that succeeded art for art’s sake and sympathetic magic it will be useful to pause and consider how we shall be able to judge them. I therefore note a few of the common criteria by which scientists judge and compare hypotheses.

Fundamentally, an explanation for some phenomenon or other must accord with received, well-supported general work and with overall theory. One cannot, for instance, explain an aberration in a planet’s orbit by invoking laser beams directed at it from beings living in the vicinity of Alpha Centauri, while all other orbits are adequately explained by gravitational fields within the solar system.

Secondly, a hypothesis must be internally consistent. In other words, all parts of a hypothesis must depend on the same premises and must not contradict one another.

The third criterion is one on which, as we have seen, Darwin insisted. A hypothesis that covers diverse fields of evidence is more persuasive than one that pertains to only one, narrow type of evidence. For example, if the theory of gravity applied only to inanimate objects, such as tennis balls, and not to living creatures, such as people, its explanatory value would be so limited that scientists would reject it.

Fourthly, a hypothesis must be of such a nature that verifiable, empirical facts can be deduced from it: a hypothesis must relate explicitly to observable features of data. Art for art’s sake fails this test because there are no verifiable features of images that it explains and that cannot be explained by other hypotheses. Aesthetic judgment is so vague that any image can be declared to be the result of an innate aesthetic impulse, even if, by today’s Western standards, its draughtsmanship is not very ‘good’.

Fifthly, useful hypotheses have heuristic potential, that is, they lead on to further questions and research. Art for art’s sake fails on this criterion too because, once stated, it closes off further research. It (apparently) says all that there is to say.

These five points are but a brief summary of some of the ways in which scientists judge hypotheses. But it should not be thought that these criteria are a simple check-list that scientists can tick off and then reach agreement.Many of the criteria are themselves problematic when it comes to practical application.

A more serious reservation, one that we must accept if we are to understand the historical trajectory of Upper Palaeolithic art research, concerns the social embeddedness of scientific work.¹⁴ Social constructivists point out that science is a lot less objective than many of its practitioners would have us believe. Thomas Kuhn, ¹⁵ physicist turned historian of science, argued that scientific knowledge does not accumulate incrementally as scientists methodically tick off a list of agreed criteria. Rather, science advances by the comparatively sudden abandonment of one theoretical and methodological structure – or paradigm – and its replacement by another. Kuhn entitled his influential book The Structure of Scientific Revolutions. Such revolutions, he went on to argue, can be explained only if we see them in terms of the sociological characteristics of scientific communities. Even as the art of antiquity was produced in a specific social context, so too the production of knowledge about that art is today also socially embedded.¹⁶ Kuhn’s work has not gone unchallenged: the revolutions that he cites may be special cases. Moreover, the notion of ‘social construction’ has been applied too enthusiastically to a great many ‘constructs’ that are not entirely socially constructed – including the different kinds of environments in which people live.¹⁷ Still, it is fair to say that today researchers realize that social factors do influence the acceptance or rejection of new ideas.¹⁸

We need to have these preliminary points, both ‘objective’ and sociological, in mind as we move on to a much more thought-out explanation than art pour l’art or sympathetic magic, one that gripped the imagination of rock art researchers during the second half of the twentieth century and that contains elements that still permeate researchers’ thinking about Upper Palaeolithic art. We need to know why it was so exciting at the time and why it has been so comprehensively abandoned.

Structure and meaning

‘Structuralism’ was one of the great informing notions of the second half of the twentieth century. It was a philosophical movement that had roots deep in Western thought. It has also been a diverse movement that has opened up enquiries into the relationship between the human mind and the material world. To guide us through the maze of structural approaches it is useful to distinguish between, on the one hand, ‘structural analysis’, a general method of analysis that examines the ways in which a ‘structure’, framework or mental template, of which people may not be aware, orders the ways in which they think and act, ¹⁹ and, on the other, ‘Structuralism’ (with an upper-case S) which refers to a specific kind of structure comprising binary oppositions and mediations of those oppositions.²⁰ Both kinds of structural theory can be discerned in an important eighteenth-century book.

Giambattista Vico (1668–1744) was an Italian jurist and classical scholar. In 1734 he published Principii di una scienza nuova (Principles of a New Science). Impressed by the work of such natural scientists as Newton and Galileo, Vico proposed a science of human society, what today we would call ‘social science’, as opposed to the ‘natural science’ of physicists, zoologists, astronomers, chemists, and so forth.

Writing in a period of European expansion and intense interest in the ‘primitive’ people and ‘savages’ whom the explorers were encountering, Vico began by challenging the current notion that such people had a different kind of mind from ‘civilized’ people. Rather, he said, the myths and explanations that they gave for natural phenomena were not simply nonsense based on ignorance; they were ‘poetic’ or ‘metaphoric’ and not intended to be taken literally. Way ahead of his time, Vico was ignored and the notion of a ‘primitive mentality’ endured through to the beginning of the twentieth century – and beyond.

Vico was astonishingly modern, even post-modern, in another way. He argued that the human mind gives shape to the material world, and it is this shape, or coherence, that allows people to understand and relate to the world in effective ways. The world is shaped by, and in the shape of, the human mind, despite the fact that people see the world as ‘natural’ or ‘given’. In performing this task of shaping the world, humanity created itself. This being so, there must be a universal ‘language of the mind’, common to all communities. Structuring, making something coherent out of the chaos of the natural world, is the essence of being human. In succeeding chapters I argue that Vico was correct, though in a way that he may not have fully appreciated. The problem with his work is that ‘mind’ is taken as an intellectualizing organ that mediates between a material world and an inner, mental world. But, as we shall see, there is more to the mind than just intelligence: there is also consciousness.

The notion of structure was greatly developed by Ferdinand de Saussure (1857–1913), a Swiss linguist, and it is from linguistics that twentieth-century structuralism sprang. Saussure must surely be the only modern Western influential figure who did not write a book. The key text, Course in General Linguistics (1915), was compiled by (no doubt unusually attentive) students from notes taken during his lectures. For our present purposes, we need consider only a few of the important distinctions that he drew, ones that eventually influenced the study of Upper Palaeolithic art.

The first is between langue (language or grammar) and parole (speech or utterances). Langue is thus the structure and parole the individual items produced in terms of that structure. Saussure illustrated this distinction by pointing to chess: the rules and conventions of the game constitute the langue of chess, while embodiments of those rules in actual games are examples of parole. The similarity with Vico’s earlier ideas about ‘the language of the mind’ is obvious.

Saussure also distinguished between diachronic studies and synchronic studies. The diachronic approach, as in philology, studies the development of language through time; synchronic studies examine a language at a given time, that is, its structure, its langue. Structure is therefore synchronic.

Related to these notions is one that became all important in Upper Palaeolithic art research. Meaning consists not so much in things themselves as in the relationship between things. For example, ‘life’ can have no meaning without ‘death’, ‘light’ without ‘darkness’, and so forth.

There is, of course, much more to Saussure’s linguistics, but these few points enable us to turn now to a researcher whose most penetrating ideas have been consistently overlooked or undervalued, not because they are demonstrably wrong but because the social and political circumstances of Upper Palaeolithic art research in the second half of the twentieth century militated against their propagation.

Society, structure and Karl Marx

Today, Max Raphael, ²¹ a German Marxist art historian, is seldom remembered, and then only as the source of some of André Leroi-Gourhan’s and Annette Laming-Emperaire’s structuralist ideas. Before we consider the massive contribution that these two researchers made to the study of Upper Palaeolithic art, we need to resurrect Raphael, not to see him simply as a source from which others selectively borrowed.²² It is unfortunate that writers who review the history of Upper Palaeolithic art research remark only, if at all, on his inchoate structuralist explanation, that is, his indebtedness to the earlier writers to whom I have referred, and ignore his heavy emphasis on the social context and character of the art. It is impossible to understand Raphael’s structuralist contributions without an appreciation of his use of social theory. The nature of society was, after all, the foundation and fons et origo of his thought.²³

As a young student, Raphael moved to Paris where he visited the studios of Rodin, Picasso and Matisse.²⁴ After brief incarceration in French detention camps, he moved in 1941 to New York, where he died in 1952, his work on Upper Palaeolithic art incomplete.²⁵ Raphael wrote his book Prehistoric Cave Paintings in the aftermath of the Spanish Civil War and during World War II, when the threat of fascism was tangible. He claimed that art historians and archaeologists of his day were reluctant to allow that Upper Palaeolithic people practised an art that was fundamentally, though not in all respects, similar to that of the mid-twentieth century. They chose to belittle the achievements of Upper Palaeolithic people ‘in order to maintain the misleading doctrine of progress from nothing to something’, ²⁶ and, by implication, thereby to parade the achievements of capitalism. By the beginning of the twenty-first century the tide had turned. Now, many art historians and archaeologists share Raphael’s position (even though they may never have heard of him) and are at pains to emphasize the modernity of Upper Palaeolithic people, both behaviourally and conceptually, and even to push modern behaviour further back in time as they rehabilitate the Neanderthals of the Middle Palaeolithic.

Following in Saussure’s footsteps and distinguishing between diachronic and synchronic studies, Raphael began by arguing that the ‘transverse section of historical existence…unfolds the qualities contained in the longitudinal section’.²⁷ These qualities consist in ‘the economic reproductive process and the struggle of economy, society and politics against religion, morality, art and science’; both of these elements were in the hands of the ruling classes.²⁸ All the fundamental social categories, forces and tensions that Marxist theory identifies were thus present at the lowest stage of fully human social evolution; Raphael explicitly challenged the then-current view that hunting bands, both historic and prehistoric, were egalitarian.²⁹ Upper Palaeolithic peoples were, in Raphael’s telling phrase, ‘history-making peoples par excellence’.³⁰

Marxist social theory, a putative key to history-making, is a complex tradition, not a monolithic canon.³¹ Still, one may say, all too briefly, that purist Marxist theory argues that a social formation (a ‘society’) comprises an infrastructure and a superstructure. The economic base of a social formation, the infrastructure, includes the social relations of production that produce the material necessities of life. In doing so, it determines the superstructure, that is, ideology, belief, religion and juridical controls. The content, or ‘message’, of art is thus part of the superstructure, though its production may derive from the social relations of production, which are part of the infrastructure. The social relations of production govern who shall perform what work and how the product of labour shall be distributed. It is the task of a rigid Marxist art historian to uncover the ways in which the infrastructure determines the superstructure. This determination may be detailed or conceived in broader terms, depending on the individual writer’s view. At this infrastructural level, class struggle principally operates and may be masked by the ideology of the superstructure, an ideology fashioned by the ruling class to conceal, or ‘mystify’, the social relations of production.

Upper Palaeolithic art, says Raphael, tells us nothing about the people’s instruments of production, their hunting techniques, or their habitations – that is, about the material component of the infrastructure.What it does tell us about is social struggle, and it does so by means of a structured code. As in totemism, social groups were represented in the art by animals. Animals depicted in combat represent antagonistic ‘clans’ (a term he does not adequately define). Often, in such instances, the animals are depicted ‘crossed’, that is, partially superimposed with heads facing in opposite directions. These compositions depict conflict. Conversely, animals pictured one inside the other depict alliance in social struggle and may stand for ‘domination, mediation or a promise of support’.³²

Then, more perceptively, he noted that researchers have interpreted the apparent disorder of Upper Palaeolithic parietal art as an absence of unity.³³ Their error arises from their failure to allow that there are different kinds of order. If researchers seek only the kinds of order with which Westerners are today familiar, they will miss, or reject as disorder, all other orders. Time and again in the history of Upper Palaeolithic art research, workers have fallen into this trap and failed to discern an order different from the one which they were seeking.

Let us now focus on the kind of order, or structure, that Raphael discerned in Upper Palaeolithic art. Essentially, he challenged Breuil’s position that the images should be seen as isolated instances and emphasized instead that they should be studied as compositions, a notion that had been tentatively suggested for a few panels some decades earlier. Because the structures of these compositions recur in panel after panel and cave after cave, we can, Raphael believed, infer the compositional devices (langue) that underlay their construction. Further, he argued that researchers should interpret the elements in relation to the whole composition. In this he was echoing Saussure’s structural insistence on relationships: as ‘death’ is meaningless without ‘life’, so bison are (to a large degree) meaningless without horses. The notion of planned compositions was thus born: these compositions were instances of parole derived from a langue that could, in turn, be inferred from them.

Raphael went on to note that different species tended to predominate numerically in specific caves. For instance, at La Pasiega in Spain stags and does predominate. In Lascaux, it is horses and aurochs (wild cattle). Most significantly in the light of Leroi-Gourhan’s subsequent work, Raphael declared that, in Les Combarelles, in the French Dordogne, ‘horses are repeatedly represented as hostile to the bison and bulls’.³⁴ Then at nearby Font de Gaume, ‘the bison fight against the horses they have found there before them’.³⁵ Later in his book, he somewhat imprecisely developed the opposition between these two species: ‘In Palaeolithic art horses and bison represent man and society.’ ³⁶ But this was not all. He more specifically anticipated a key feature of Leroi-Gourhan’s work when he wrote that the magical battle of clans on the ceiling at Altamira was a vehicle for further meanings. The delicate hind confronting the large bison could, he claimed, be read as ‘…a conflict between feminine tenderness and masculine bulk’ (Pls 2 and 3). The opposition (female:male) thus acquired a ‘universal human significance’.

Raphael carried this binary opposition over to an interpretation of signs. He distinguished one set of signs as representing male sex: ‘arrow points, phallus, act of killing or death’. The other he saw as ‘feminine sex, woman, lifegiving, life’.³⁷ Faced with what Raphael calls ‘tragic dualism’, the Upper Palaeolithic artists ‘confronted the task of harmonizing the opposition’.³⁸ The ‘message’ of Upper Palaeolithic art was produced, reproduced and transmuted by means of oppositions and mediations. This is a fully Structuralist position, though it predates Lévi-Strauss.

Raphael’s work has been comprehensively ignored. As we shall now see, Leroi-Gourhan and Laming-Emperaire tacitly rather than explicitly developed his structuralist (and Structuralist) insights but completely passed over his social analyses. They were more interested in cracking the ‘code’ (langue) of Upper Palaeolithic art than in the functioning of the society that produced it. Similarly, writers on Leroi-Gourhan may mention Raphael fleetingly and merely as a source of Leroi-Gourhan’s and Laming-Emperaire’s ideas. Certainly, any murmur of Marx is suppressed.

Structuralism and mythograms

Our consideration of how Raphael’s social theory controlled his interpretation of Upper Palaeolithic art brings us to two researchers who, at least in my view, made the greatest twentieth-century contribution to the study of Upper Palaeolithic art.³⁹ They are Annette Laming-Emperaire and André Leroi-Gourhan (Fig. 13). Sadly, some recently published works, ones that in fact depend heavily on their ideas, hardly mention them; they are suffering the same fate as Raphael. Our historical rollercoaster ride continues.

Laming-Emperaire (1917–78), or Laming as she then was before her marriage, published a short book about Lascaux in 1959. The site had been sensationally discovered on 8 September 1940. Three schoolboys found the cave by chance when 17-year-old Marcel Ravidat’s dog, Robot, tumbled down a sink hole that had been known to local people since before World War I. Farmers had tried to block the hole to prevent cattle falling down it. The boys retrieved Robot and decided to return later with better equipment, including rope.

13 André Leroi-Gourhan, philosopher, ethnographer, archaeologist and innovator. He re-examined Upper Palaeolithic cave art and devised a new, all-embracing explanation.

On 13 September, Ravidat and three other friends, including two boys who had been evacuated from wartime Paris, returned. They widened the hole, and Ravidat was the first to slide down the steep slope beneath it. When they all gathered at the foot of the slope, they found themselves in a large subterranean chamber. They had a lamp, but could see little in the vast space. They advanced into what is now known as the Axial Gallery (Pls 21–23), where they first spotted images of animals. In great excitement they explored the rest of the cave and found more and more wonders (Pls 1, 20, 26–28).

Realizing the importance of the find, they resolved to tell their schoolmaster, Léon Laval. Because of his age, he was at first unable to get down the difficult incline, so he asked one of his former pupils to copy some of the images. Thus convinced of the importance of the discovery, he arranged for the Abbé Breuil to be contacted. Breuil had fled from German occupied Paris and was sheltering in Brive. He soon authenticated the discovery. It was a stunning find, unequalled until, in 1994, Chauvet and his two friends found the cave that features in our Time-Byte III and that now bears his name. No wonder that Lascaux provided material for a brand-new approach to and explanation for Upper Palaeolithic cave art.

In a general discussion of Upper Palaeolithic art that precedes her analysis of Lascaux, Laming-Emperaire seems to follow Raphael in a number of important ways. She

– questions the value of ethnographic parallels,

– argues that the difficulty of access to many subterranean images pointed to ‘sacred’ intentions,

– rejects any simple form of totemism,

– proposes that ‘the mentality of Palaeolithic man was far more complex than is generally supposed’, ⁴⁰ and

– argues that images should be studied as planned compositions, not as scatters of individual pictures painted ‘one at a time according to the needs of the hunt’.⁴¹

Her work thus shifted emphasis from magical concerns to symbolic meanings. More specifically she argued that horses and bovids are deliberately and repeatedly paired. She remarked that this sort of association raises ‘a problem which has hitherto never been considered at all’.⁴² She also wrote that deliberate associations of images ‘which have hitherto been interpreted as juxtapositions or superpositions should in fact be regarded as deliberately planned compositions’.⁴³

All these points seem to derive from Raphael, yet she does not mention him in the Lascaux book. Did she consider it imprudent to mention Marxist Raphael in the days of the Cold War? Yet, when she came to prepare her Ph.D. thesis, ⁴⁴ she acknowledged Raphael as the originator of the ideas that she later developed and said that Raphael let her have ‘typed notes’ as early as 1951.⁴⁵ Then, in an inexplicable volte-face, she concluded that Raphael’s hypothesis that some panels ‘represent the struggle of the sexes or of totemic clans…[is] unfortunately useless. It is difficult to accept Raphael’s conclusions.’ ⁴⁶

In La Signification de l’ Art Rupestre Paléolithique, ⁴⁷ her major work, Laming-Emperaire suggested that the horse may be associated with ‘femaleness’ and the bison with ‘maleness’. Later, she revised her views and, although she continued to allow that horse and bison were in some way fundamental to Upper Palaeolithic cave art, she rejected the notion that they ‘stood for’ sexual duality.⁴⁸ Instead, she developed Raphael’s suggestion that ‘clans’ and social organization, along with creation myths, were at the root of the art.

Laming-Emperaire died tragically in a motor accident in 1978 but the methodology and techniques that she proposed became the procedures that she, Leroi-Gourhan and numerous other researchers would follow. In her early Lascaux book, she advocated ‘compiling inventories and collating all the archaeological evidence in order to extract the utmost information from both sources’.⁴⁹ She argued that researchers should prepare ‘distribution maps according to the following criteria: position of the works in a cave, associated archaeological remains, signs of use, and content or form of a representation’.⁵⁰

Laming-Emperaire’s methodological suggestions were exactly what André Leroi-Gourhan (1911–86) adopted to prepare for his complex hypothesis. He was an anthropologist and ethnographer who taught at the Collège de France in Paris. Earlier in his career, he was an assistant director at the Musée de l’Homme along with Claude Lévi-Strauss (b. 1908), the enormously influential French savant and father of the Structuralist movement in anthropology and archaeology. To understand Leroi-Gourhan’s contribution to the study of Upper Palaeolithic art we must first consider Lévi-Strauss’s work.

The son of a musician and himself an accomplished musician, Lévi-Strauss studied law and philosophy; he acknowledges also the influences on his thought exercised by geology, Freudian psychology and Marx.⁵¹ Before World War II, Lévi-Strauss taught at the University of São Paulo and did ethnographic work in Brazil amongst the Caduveo and Bororo Indians. After leaving the university he led a year-long expedition to the South American Nambikwara and Tupi-Kawahib communities. But theoretical work interested him more than fieldwork, which he found slow and time-consuming.⁵² During the war he taught in the New School for Social Research in New York.⁵³ There he met and was influenced by numerous European linguists, including Roman Jakobson.⁵⁴ He was thus in New York at the same time as Raphael, but there does not seem to be any evidence that they met, though they may well have. After the war, he was appointed French Cultural Attaché in the United States of America. He returned to France in 1948.

Lévi-Strauss is a Marxist and thus has that in common with Raphael, though in exactly what sense he is Marxist is hard to say;⁵⁵ indeed, much of what Lévi-Strauss says is hard to pin down. The Cambridge anthropologist Edmund Leach⁵⁶ perceptively remarks that Lévi-Strauss seems to admire Verlaine’s dictum ‘pas de couleur, rien que la nuance’ (‘no colour, nothing but nuance’). Lévi-Strauss’s Structuralism played itself out in a number of arenas but most famously in his mythological studies. Of these, ‘The Story of Asdiwal’, ⁵⁷ a Northwest Coast of America Tsimshian myth, is the most persuasive, and the four volumes of Mythologiques (The Logics of Myth), an analysis of over 800 myths (plus more than 1,000 variations of them) from South America through to the Northwest Coast of North America, the most spectacular. Lévi-Strauss’s work is founded on the notion that binary oppositions and mediations of those oppositions constitute a unifying thread, a hidden logic, that runs through all human thought. Binary oppositions comprise two ‘opposite’ terms, for example, up:down, life:death and male:female. ‘Structure’ is concerned with logical categories like these and the kinds of relations between them. According to Lévi-Strauss, every society’s language, kinship system and mythology are variations on this binary theme.⁵⁸

Inevitably, in every such system of classification and relations there are anomalies: real history sometimes contradicts present conditions; and there are always people and animals that do not fit neatly into the system of categories. For Lévi-Strauss, ⁵⁹ ‘the purpose of myth is to provide a logical model capable of overcoming a contradiction (an impossible achievement if, as it happens, the contradiction is real)’. Because myth is ultimately fated to fail, large numbers of variants are generated, each slightly different from the others. As Lévi-Strauss himself says, ‘Thus, myth grows spiral-wise until the intellectual impulse which has produced it is exhausted.’ ⁶⁰

Leroi-Gourhan had little to say about the way in which his and Lévi-Strauss’s paths crossed – they sometimes took coffee in the same common-room – but there can be no doubt that he owed his philosophical and methodological position to Lévi-Strauss. Leroi-Gourhan’s references to his colleague at the Collège de France may be sparse, but it would be obtuse to suggest that he tried to claim originality for what were so clearly Lévi-Strauss’s developments of Structuralism: Leroi-Gourhan accepted his debt to Lévi-Strauss as patent.

Leroi-Gourhan says that he and his team of associates began work on his monumental The Art of Prehistoric Man in Western Europe ⁶¹ in 1957. He published a shorter book, Les Religions de la Préhistoire, in 1964. From the beginning, he says, he did not find the disorderly scatter of images that he expected to find; rather, he was impressed by ‘the unity each of the sets of figures embodies’.⁶² Early on, he and Laming-Emperaire realized that they were both ‘very much on the same track’. Leroi-Gourhan says that the grouping of images had been ‘entirely over looked’ until Laming-Emperaire noticed it, ⁶³ though we know that this was not true: Raphael had certainly noticed groupings. At this point, they decided to pursue their researches separately. When they later compared their finished work, they found that if they had both gone ‘off track’, they had ‘done so with the same sense of direction, which fact deserves at least a certain interest’.⁶⁴

As I outline Leroi-Gourhan’s work, it will be apparent that it seems to answer the five criteria that I mentioned for judging hypotheses. Certainly, it is in accordance with what was then a well-established philosophical position (Structuralism); it is internally consistent; it covers (to a certain extent) different bodies of data; it appears to connect with ‘real’ data; it seems to open up new avenues of research. Shortly, I shall show that some of these accords are illusory.

Leroi-Gourhan began by rejecting as too vague and imprecise all the old generalizations that had appeared in book after book for 50 years. For instance, the rather naive division of animals into food and dangerous creatures was, in his view, superficial and tendentious. Paralleling Raphael and Laming-Emperaire, he also rejected the use of ethnographic analogies; researchers should, he said, derive their explanations from the data of the art. Yet, like Raphael and Laming-Emperaire, he found that this admirable-sounding precept is impossible. He also decided to begin his study with parietal rather than portable art; the parietal images, he noted, remained where people had placed them, a key consideration if one is to study spatial structure. Importantly, he followed Raphael and Laming-Emperaire in acknowledging that the art was not a scatter of individual images, as Breuil had argued, but rather a carefully composed series of groupings that had to be understood in their relationships with one another.

A number of Leroi-Gourhan’s principles have their roots in Vico’s ideas. The first is that Upper Palaeolithic people had the same brain/mind as we and all people living today have: theirs was no ‘primitive mentality’. This was a point that Lévi-Strauss convincingly demonstrated in such works as The Savage Mind.⁶⁵ Another principle is that Upper Palaeolithic people imposed the ‘shape’ of their minds on their material world. A third principle is that close study of the products of the Upper Palaeolithic mind will reveal something of its functioning. This last point, it has to be admitted, has more than a whiff of circularity about it: if we decide, as a result of received notions from linguistics and Structuralism, that the mind works in a particular way and then analyse the data of the art in that very way, the outcome of our work will be to a certain extent predetermined.

More precisely, Leroi-Gourhan accepted Lévi-Strauss’s proposition that the mind works like a computer in binary oppositions and that meaning lies in the relationships between the elements of those oppositions. If, at this point, one recalls binary oppositions such as light:darkness, male:female, divine:human and so on, all seems well. But what binary oppositions are specifically discernible in Upper Palaeolithic cave art?

Setting out from these principles, Leroi-Gourhan analysed the numerical data of the images and their locations in 66 caves. He prepared detailed maps of the caves along the lines that Laming-Emperaire advocated and analysed the numbers of images and their locations by means of punch-cards. But, in all cases like this, the categories fed into the system and the possible relationships between images all derive from the mind of the researcher. In a sense, it was Leroi-Gourhan who was asking himself questions of his own devising.

Let us now consider the pivotal elements of Leroi-Gourhan’s rather complex hypothesis; he himself frankly refers to ‘a labyrinth of hypotheses’.⁶⁶

To complicate matters further, he changed his mind about aspects of his work in response to criticism and, especially, as a result of discrepancies between his explanation and the ‘real’ data of the art. We need not explore all the changes that he allowed, though the ‘dialectic’ between his thought and the criticisms directed at his work is fascinating;⁶⁷ we are more concerned with the foundations and principles of his work.

To begin with, he believed that he could divide the depictions of animals into four groups:⁶⁸

A Smaller herbivores: horse, ibex, stag, reindeer, hind.

B Large herbivores: bison, aurochs.

C Peripheral species: mammoth, deer, ibex.

D Dangerous animals: feline, bear, rhinoceros.

Then he went a stage further. He designated the Group A animals as ‘male’ and the Group B animals as ‘female’. Here he is not referring to sex of the individual images. Rather, he is saying that Group A species represent ‘maleness’ and Group B species ‘femaleness’, as metaphysical properties (Fig. 14). Group C comprised species that seemed to be painted peripherally to those of A and B. Group D was a separate group of animals that people would have feared.

Next he tackled the signs. Again seemingly influenced by more than just the germ of an idea in Raphael’s work he distinguished two categories: a and b. Group a comprised what he called ‘narrow’signs: hooked or ‘spear-thrower’ signs; barbed signs; single and double strokes; and rows of dots. Group b comprised ‘wide’signs: triangles; ovals; quadrangular signs; and claviforms. Both groups seem to be derived, he argued, from schematic representations of, respectively, male and female genitalia. This binary grouping (a:b) was linked to the A and B animal groups; ‘narrow’signs were considered male, and ‘wide’ signs female. The sets of groups, both animals and signs, could thus ‘stand for’ one another (Fig. 14). In short, Upper Palaeolithic art could be summed up by the principal binary opposition A + a (male) : B + b (female).

14 A tabulated summary of Leroi-Gourhan’s male : female binary opposition as manifested in images and signs. In the caves, these were ‘opposed’ or ‘paired’.

This was Leroi-Gourhan’s response to the suggestion that the images and the signs have nothing whatever to do with each other. On the contrary, he saw them as complementary signifying systems, perhaps like text and diagrams in a book – both say the same thing, though in different ways. In a subsequent chapter, I argue that some of the signs are actually of the same order and system as the images of animals; they are not different, though parallel, systems: some of the signs and the animals derive from the same source.

Leroi-Gourhan, then, saw A and B as constituting a binary opposition, with C animals as peripheral additions. He was thus able to construct a summary ‘mythogram’: A + B – C (the dash does not mean a minus sign). He argued that this mythogram was a mental template – the langue – of Upper Palaeolithic art for the entire period. Leroi-Gourhan thus took the Saussurian notion of langue and turned it into a Lévi-Straussian binary opposition.

We now come to the way in which Leroi-Gourhan related these sets of images to the caves (Fig. 15). First, he divided the caves into a number of areas: entrances, central areas, and deep areas (he also allowed for diverticules, small side chambers). Male animals and signs were, he claimed, distributed throughout the caves, while female animals and signs clustered in the central areas. In those central areas, the horse:bison/aurochs pairing was central; Group C animals were peripheral. The deep, end chambers were said to be associated with dangerous animals. The cave itself was thus an integral part of the ‘message’. He saw it as largely female. In his own words:

15 A simplified version of Leroi-Gourhan’s conception of an idealized Upper Palaeolithic cave. Entrance and deep areas flanked an all-important central area of ‘paired’ images and signs.

Palaeolithic people represented in the caves the two great categories of living creatures, the corresponding male and female symbols, and the symbols of death on which the hunters fed. In the central area of the cave, the system is expressed by groups of male symbols placed around the main female figures, whereas in the other parts of the sanctuary we find exclusively male representations, the complements, it seems, to the underground cavity itself.⁶⁹

In sum, Leroi-Gourhan⁷⁰ believed that the art was ‘the expression of ideas concerning the natural and supernatural organization of the living world’. Prudently, he added that, in that ancient thought system, ‘the two might have been one’. Here, we seem to have a clear version of Lévi-Strauss’s Structuralism: Leroi-Gourhan’s mythogram is apparently exhibiting and trying to overcome a fundamental contradiction, or parallel contradictions, as, says Lévi-Strauss, does myth. Can we echo Lévi-Strauss⁷¹ and say that Upper Palaeolithic art grew ‘spiral-wise until the intellectual impulse which…produced it [was] exhausted’? Is panel after panel, ranging in time from the beginning to the end of the Upper Palaeolithic, dealing with male:female, culture:nature, and other parallel oppositions and contradictions?

In a couple of absorbing conversations that I had with Leroi-Gourhan in 1972 in Valcamonica, Italy, I found that he was at that time convinced of the validity of the mythogram and the patterned distribution of motifs in the caves. It was the predictive potential of his work that he found persuasive: he felt that, in a cave he had never before entered, he could foretell the presence of, say, a horse to complement an image of a bison that he had just been shown – to the astonishment of his guide. But he registered an important reservation: he believed that the horse:bison :: male:female opposition was only one characterization of the mythogram. For him, the mythogram was a vehicle that could carry a wide range of meanings. Though he did not say so, he was perhaps influenced by Lévi-Strauss’s tireless reiteration of the great culture:nature and male:female binary oppositions, and he would have been better advised to choose some more neutral characterization of the mythogram than male:female.

When one compares Leroi-Gourhan’s work with the old art-for-art’s-sake and sympathetic magic explanations, one cannot but be astounded by his combination of industry and ingenuity. Why, then, did researchers come to abandon his explanation? In reciting his failings, some critics adopt an unbecoming triumphalism; they are more interested in his errors than his insights. Some illogically present him as paradigmatic of the failure that must necessarily await all who attempt to understand Upper Palaeolithic art.

Essentially, the principal reason for Leroi-Gourhan’s fall from academic grace has to do with the criterion that demands articulation with hard data. Although he himself stressed the empirical component of his work and believed that more empirical work in the caves would expand the explanatory potential of his insights, the opposite turned out to be the case.

For one thing, the diversity of the topography of the caves makes it impossible to compare them with one another in terms of entrance/central/deep/ diverticule areas (Figs 1, 2, 3, 7, 54, 56). Sometimes this problem led Leroi-Gourhan to identify a ‘central’ or a ‘deep’ area by means of the species depicted, not the topography – a circular argument. Sometimes his identification of a mark on a cave wall as a particular species seems to be a result of his preconception of that area of the cave. Then, too, his statistics have been criticized on various grounds.⁷² As the years went by and more and more caves were discovered, such empirical problems multiplied until researchers concluded that the whole Structuralist edifice was built on friable empirical foundations. The mythogram was too good, too neat, to be true. Leroi-Gourhan’s idea did not have the adaptability that Breuil’s sympathetic magic enjoyed. Ironically, Leroi-Gourhan’s was a ‘good’ hypothesis in that it had the capacity to be invalidated by empirical work; in Karl Popper’s term, it was refutable. Whatever other qualities the mythogram hypothesis may have, it foundered on empirical grounds.

Towards the end of the twentieth century the feature of the mythogram hypothesis that most troubled researchers was Leroi-Gourhan’s insistence that the formula for the embellishment of the caves persisted through the entire Upper Palaeolithic. His explanation was derided as ‘monolithic’. Such a long period, his critics reasoned, must have embraced far-reaching changes; there must have been more diversity through time and across space than he allows. This is, in fact, an equivocal objection. The passage of time does not cause change; it merely allows for it. Beyond researchers’ hunch that there must have been considerable change, there is no good reason why change must have taken place. Then, too, we need to be explicit about what sort of change we mean. If we say that the mythogram may have remained essentially the same, but its content, or meanings, may have changed, we have made a point with which Leroi-Gourhan agreed. Certainly, horse and bison are enduring Upper Palaeolithic motifs, but did they always carry exactly the same meanings? So we need to be more specific when we complain that so many changeless millennia seem an unlikely proposition. Both change and stasis must be demonstrated, not assumed.

Still, we must note exactly what was falsified. It was the grouping/pairing of species and their distribution through the caves that was shown to be incompatible with the more varied data – not the notion of structure. Raphael, Laming-Emperaire and Leroi-Gourhan may all have been wrong in the specifics of their hypotheses but correct in their beliefs that the caves are indeed patterned. The crux of the matter is to identify the criteria that inform, and will therefore reveal, the (or, rather, ‘a’) pattern. This is a point to which subsequent chapters return.

Strangely, although Leroi-Gourhan produced the most thorough-going use of Structuralism in archaeology, his influence was limited, even in his heyday and in his native France. This limitation may explain why it has now become easy to ignore him and yet still benefit from his work. It is also true that Leroi-Gourhan produced his Structuralist work at a time when archaeology, greatly influenced by the so-called New Archaeologists who strove for a strictly ‘scientific’ discipline, was turning its back on attempts to get at the meanings of ancient symbols and belief systems.⁷³ Scepticism about such matters became the order of the day.

A return to society…and more

Leroi-Gourhan’s work, in all its complexity and with all its problems, is the vertex of this chapter. It flourished (though it was contested from the beginning) for a couple of decades, but its abandonment was swift and decisive; the historical rollercoaster went into a steep plunge. Unfortunately, the entire structuralist position was discarded along with many of its valid principles. The baby went out with the bathwater before it could grow to maturity. As a result, the study of Upper Palaeolithic art is in danger of losing some of its greatest practitioners. Today, there seems to be what Lévi-Strauss⁷⁴ calls ‘a desperate eagerness which drives us to forget our own history’ – or at any rate to forget its positive elements as we concentrate on the failures and errors of our predecessors. Today few researchers want to be seen as structuralists. In 1997 Lévi-Strauss himself put the general point forcefully:

In the structuralist wake came an anti-theoretical age of agnostic empiricism.⁷⁶ Researchers did not abandon structuralism for another theoretical perspective that would make better sense of the data: on the contrary, they abandoned theory altogether. Or, at least, so they believed.

But it has not been entirely a Dark Age of obscurantism. Some researchers, such as Georges Sauvet, ⁷⁷ Denis Vialou⁷⁸ and Federico Bernaldo de Quirós, ⁷⁹ have usefully pursued structuralist explanations. For Sauvet, a pattern is discernible in formal analysis of signs; for Vialou, it is in the distribution of species in specific panels. Bernaldo de Quirós sees Altamira as a planned structure comprising three principal sections, or nuclei, linked by passages: ‘Indeed, the decoration of the cave seems to conform to a pattern that invariably takes advantage of the natural relief, as do the masks in the Horse’s Tail, the doe’s head in the Pit, and the figures of bison on the Great Panel’.⁸⁰ He believes that the sections of the cave were designed as a location for initiation rites.⁸¹ Structure is thus implicated in the reproduction of social order through initiation rites.⁸² Whether Sauvet, Vialou and Bernaldo de Quirós are correct in the specific patterns that they believe they have uncovered, the important point is that they continue to argue that there must be some pattern.

Other researchers have followed up social approaches to the problem.In her last statement, Laming-Emperaire⁸³ returned to Raphael’s sociological position. She remained convinced of the salience of horse and bison, but she reverted to the idea that the different species depicted in the caves represented social groups and that some of the panels told the myths of the origins of these groups. The art was essentially a social phenomenon. Mythogram and myth were thus coming together in her thought.

The mythogram alone did not, indeed did not attempt to, explain the greatest of all enigmas: why Upper Palaeolithic people made images in deep, dark caves. The number and types of data that it covered were strictly limited. The reason for this curious subterranean Upper Palaeolithic practice must have been social; there must have been enduring social pressures that kept driving people underground throughout the Upper Palaeolithic, and, if we restrict our enquiry to the ‘meaning of the message’ and ignore its social context, we shall unnecessarily hobble ourselves.

Shifting emphasis from the content of the message to its form and the reasons for it, some researchers have linked the startling topographical contexts of many images to ‘information transmission’ and have drawn on recent developments in ‘information theory’. Initiates and others, they argue, were psychologically prepared by the circumstances of the caves for the reception of crucial information about social customs and so forth.⁸⁴ Others have linked this supposed ‘information’ to relations between geographically scattered but interdependent communities that constituted alliance networks.⁸⁵ The ‘appearance of parietal art in Late Pleistocene Europe resulted from the closing of social networks under conditions of increasing population density’.⁸⁶ Antonio Gilman⁸⁷ developed a Marxist approach to this sort of alliance theory and argued that tensions within alliance networks led to the Middle to Upper Palaeolithic Transition and eventually art as a response to a need for symbols of identity. How classless societies could develop internal contradictions that would lead to major social change has long been a problem for Marxist theorists.⁸⁸ Unfortunately, Gilman has not followed up this work.

These information theory explanations in some ways state the obvious (pictures meant something) and, at the same time, accept the disputable proposition that pictures carry enormous loads of information, more than words. But as the art historian Ernst Gombrich⁸⁹ has pointed out, pictures have the power to move, but they in fact convey very little information: because people read pictures in different ways images always remain semantically equivocal. The best that can be said for pictures is that they trigger memories of information that has been absorbed in different ways, that is, by experience and verbally. So whilst the Upper Palaeolithic images may have sometimes functioned as mnemonics, their capacity to store or convey information was limited. They were not like the hard drives of modern computers.

One final explanation for the appearance of art requires comment because it is alluringly attached to evolutionary theory. Ellen Dissanayake of Edinburgh University argues that art is a characteristic of our species and defines it as a way of ‘making special’, a process that includes all the ‘arts’.⁹⁰ According to her, it is this ‘making special’ that is adaptive, not just image-making or one of the other ‘arts’. To my mind, this definition and explanation are too broad: we need to distinguish types of art and explore the specific problems that each poses. Then we need to explain art in social terms – how it functioned in society, not simply how it promoted survival.

These considerations bring us back to a central theme of this book – methodology. If we have learned anything from a century of research on the meanings and purposes of Upper Palaeolithic art it is that we have no agreed method for tackling the problems that they pose, and that we have to address different kinds of questions, such as the meanings of images and their strange placement in caves. To begin to confront such diverse issues we need to intertwine a number of strands of evidence;⁹¹ one ‘pure’ method, be it inductive, deductive or analogical, will not suffice, a point that I develop in Chapter 4.⁹² Proof, it is true, will remain elusive, but complementary types of evidence that unite to address the complex problems posed by Upper Palaeolithic art can, I believe, produce persuasive hypotheses. Alive to the influence of our own social context and the ways in which the ‘scientific’ community is moulded by social pressures, we need to develop a more complex method than those hitherto used to address specific, isolated aspects of Upper Palaeolithic art.

Throughout this chapter I have adopted what is essentially a modernist, or realist, approach. It would, I believe, be a serious error to suppose that, because science is a social activity, all explanations are simply products of their own times and therefore cannot be said to approximate more or less closely to a situation that was part of a past that did indeed happen as it did. I believe that it is possible to produce knowledge of a real world ‘out there’. We need not embark on an endless programme of deconstruction. We do not live in a world composed entirely of representations. Clearly, I cannot enter now into a full consideration of these contentious ideas. All I can do is state my position: I believe that some representations (such as hypotheses about the past) correspond more nearly to reality than others, and we can discern which they are. If this were not so, there would be no telephone, radio or interplanetary travel.

Allowing the effects of social contexts and at the same time emphasizing a real historical past and the possibility of constructing hypotheses that approximate to it, we can proceed to address the enigma of what happened to the human mind in the caves of Upper Palaeolithic western Europe. This is the epistemological principle on which the following chapters are posited. To identify a key evidential strand that is currently ignored, we need to go back to Vico and other early writers and recall that the human mind tends to impose itself, or rather aspects of its structure, on the stream of undifferentiated sense impressions that come to us in daily life. We must therefore look more closely at the brain, the mind, intelligence and the shifting, mercurial consciousness of human beings. But first we must see what happened in western Europe at about 45,000 to 35,000 years ago and understand that those events were but a slice of a much wider shift in human behaviour. The brain and changing social relations were shaping the mind.⁹³