One of the most disconcerting aspects of introducing the work of Lévi-Strauss to students is that those who are just beginning an anthropology course often seem able to grasp quickly and easily the main points of his work, while those who have a good anthropological training seem almost invariably incapable of understanding what he is saying. The reason is that Lévi-Strauss’s work deals with different questions from those which are traditionally assumed to be the subject-matter of anthropology. Trained anthropologists often wrongly assume that he is asking similar questions to theirs and therefore find the answers baffling. He himself rarely, if ever, places his work in relation to that of others in an illuminating way. The work of his colleagues is, for him, not a body of alternative theories but rather a resource to be exploited as data in the building of his system. As a result, his references to other anthropologists simply further obscure what he is doing, especially for those who know well the works he refers to.
The reason for this astonishing independence is partly a matter of personality. Like Rousseau, whom he so much admires, he is very much a ‘promeneur solitaire’. It is also a product of the way he entered anthropology and of the way his work has developed. Unlike most British or American anthropologists, he has never been trained in an academic anthropological tradition, and he has never had to prepare undergraduates for a degree. Lévi-Strauss’s anthropological career began in Brazil, where he chose to go and teach before the war in order to try his hand at ethnography. He succeeded in this to a certain extent and has produced a number of studies on Brazilian Indians. However, this introduction to his subject-matter often seems more important for the personal and moral significance it had for him than as a source of data. Again, he is unlike most Anglo-Saxon anthropologists in this. They tend to amass their piles of field notes and work from them as though these were the only reliable facts on the people they study. Lévi-Strauss by contrast makes great use of all information available from previous writers, whether anthropologists or not. As a result, his data are often much richer than those of his colleagues, but also much less contextualised and less systematic.
It is not in Brazil that Lévi-Strauss’s anthropology took on its unique character but in New York, where he fled the Nazi victory in France. He took a post at the New School for Social Research and met, among others, the linguist Roman Jakobson. The significance of this meeting is central, because it is the combination of Lévi-Strauss’s ethnographic knowledge and the theories of a particular school of linguistics called ‘structural linguistics’ which produced ‘structural anthropology’, from which the whole range of ‘structuralisms’ claim to derive. Structural linguistics was an amalgam of ideas sometimes seen as going back to the work of the Swiss linguist Ferdinand de Saussure, who himself was strongly influenced by the sociological theories of Durkheim. De Saussure’s main contribution was to show that the study of language need not be a matter of building up language families by presumed historical affinities. This had been the concern of earlier, mainly Germanic linguists, and still forms the basis of the language maps of school atlases. Another type of study was possible: looking at language as an internally-regulated system whose nature makes it possible for people to say what they want to say. This study of language was the study of the ‘structure’ of language. The ideas of de Saussure were further developed by a group of Eastern European linguists, the Prague School, with which Jakobson was associated. They demonstrated, with apparently extreme scientific rigour, the nature of the system governing the organisation of significant sounds in language – what is called phonemics. This scientific rigour predictably appealed to American linguists such as Bloom field, who tried to expand the principles of phonemics to all aspects of the study of language and so produced structural linguistics, the future Aunt Sally of Chomsky.
It was not the arid aspects of structural linguistics which attracted Lévi-Strauss – indeed, structural linguistics in the hands of Jakobson always retained a dynamic and human character – but rather its unexpected link-up with an influential mathematical-cum-psychological theory: cybernetics. Cybernetics developed rapidly during the war in America because of its military applications. Basically, it was the theory which made it possible to build the first artificial intelligences, or computers, which were to guide missiles so that they would follow the target – that is to say, modify their behaviour in response to the information they had gathered. What linked this type of engineering to linguistics was the fact that these early computers organised their complex working by means of simple binary contrasts. The phonemic theory was also based on simple binary contrasts, and thus language, too, could be seen as a device which could produce complex information-processing on this same simple logical basis. It was assumed that this similarity could not be coincidence, and with a little neurological evidence for support, the cyberneticians deduced that the similarities were to be explained by the fact that this was how all information-processing devices, and therefore the human brain actually worked. All at once it seemed as if human sciences, represented, by linguistics had joined up with natural sciences represented by neurology, as well as with engineering and mathematics.
It was against the background of excitement created by these congruences that Lévi-Strauss developed structural anthropology. The idea was quite straightforward and went as follows: what anthropologists study is the particular knowledge which people in exotic cultures hold. This may be knowledge which takes the form of myth or ideas about plants and animals or about what characterises different types of men. This last type of knowledge, social knowledge often takes the form of kinship systems. For Lévi-Strauss social systems, and kinship systems in particular, are ways whereby we classify other people so as to organise our behaviour towards them; they are not, as they used to be understood by British anthropologists, that behaviour itself.
Now, if anthropologists study the knowledge of fellow human beings, it inevitably follows that such knowledge will be organised according to the way in which the human brain organises information. If, as Lévi-Strauss and the cyberneticists believed, the brain stores and organises information by means of binary oppositions, then this mechanism will be the underlying principle of organisation behind what all anthropologists study. The knowledge of individuals and cultures is of course very complex, but that was no obstacle to believing that it was all simply organised in this way, because Lévi-Strauss could point to computers which were similarly able to handle complex information by the same very simple means. It was a matter of building up binary oppositions on binary oppositions.
This Lévi-Strauss first demonstrated at length in his book The Elementary Structures of Kinship. There he argued that the first step in the organisation of social knowledge was the establishment of a fundamental binary opposition between givers and receivers. The idea that gift-giving creates enduring social categories defined by complementary opposition goes back to Marcel Mauss, who, like Lévi-Strauss, saw the fact that the exchange of gifts creates a relation which transcends the act of giving itself as the very foundation of human society, or, to put it another way, as the foundation of social classification. The particular gift which Lévi-Strauss saw as primary was the exchange of women between groups. This gift exchange was made necessary by the rule of incest which forbade sexual relations inside the group. Lévi-Strauss picked on that particular gift relationship because he felt it was possible to demonstrate that all systems terms of kinship classification, but especially those he called elementary structures, could be shown to have developed by means of subsequent binary opposition built upon the basis of the fundamental binary contrast created by the original gift exchange of women.
All the rest of Lévi-Strauss’s work is an attempt to show that the same principles of organisation of knowledge, of structure, underlie other areas of human knowledge, whether this be plant and animal classification, technological knowledge or mythology. In The savage Mind, he demonstrates the all pervasiveness of the structural organisation of classification, drawing special attention to the phenomenon by which structured information concerning one domain – for example, animals – is made to impose structures on another domain – for example, social classifications. This use of the same structure for different domains explained the centrality of metaphor as an organizing principle of culture. It also showed that apparently bizarre and exotic customs such as totemism – the practice of referring to people as though they were animals – were only one manifestation of a much more wide-ranging and fundamental aspect of all knowledge.
There is clear shift of focus between the later and the earlier work of Lévi-Strauss. The earlier work, as we saw, is concerned with demonstrating that the varied and complex systems which the anthropologist records could have all been built up from the same or similar elementary oppositions by means of the same process: the way the human brain stores and organizes information. The later work is more concerned with the way humans having learnt already complex and varied systems of culture, further transform them. This secondary process is different in that it is not any more concerned with primal starting-points, but it is still governed by the same general principles of human neurological processes. The move away from questions relating to the original building up of structures to a consideration of the transformation of already constructed structures into further transformations takes Lévi-Strauss into theoretical problems about the history of knowledge through time.
He is well aware of an apparent paradox in this shift of emphasis. Structuralism as he sees it goes back to the turning away from historical questions, a turning away which characterises the work of de Saussure. The great linguist had, following Durkheim stressed the heed for synchronic studies – that is, the establishment of relations between features existing at the same time – as against diachronic studies – that is, the establishment of relations between features in a temporal sequence. In focusing on the transformation of structures through time, Lévi-Strauss in fact seeks to abolish the distinction between the two different types of study. He wants, however, to do this without losing the advantages which had been gained in establishing the contrast in the first place.
The notion of transformation which has come to dominate his work is the idea that humans react to new situations and new information, not by simply adding to knowledge, but by reorganising what they already know according to the laws which govern the structures of human cognitive processes. Indeed, any communication of information is likely to lead to the structural transformation of this information. The simplest example of transformation is inversion. For example, a myth may be passed on from one person to another and apparently change completely, but this change will, on closer examination, reveal itself to be a systematic inversion of all the internal relations of its structure. The structure will have been changed only slightly, but the appearence of the myth will superficially seem totally different. This, according to Lévi-Strauss, is the way culture develops.
The notion of transformation is not just a theory about knowledge in history, it is also a method. Insisting that culture, like all human knowledge, is structured by principles which relate to general features of the organization of cognition is easy enough, but the problem is how to demonstrate this. The difficulty with simply identifying a pattern in, for example, a myth is, that there is no way of showing that this pattern is not simply an arbitrary product imposed by the anthropologists, but that it has, real significance for the way the information is ordered in the minds of the people from whom the myth was collected. The way to overcome this problem, according to Lévi-Strauss, is to show that number of myths are related by means of structural transformation. He puts the matter quite emphatically in his latest book on masks: ‘Any myth or sequence in a myth would remain incomprehensible if each myth were not opposable to other versions of the same myth or to apparently different myths, each sequence opposable to other sequences in the same or other myths, and especially those whose logical framework and concrete content, down to the smallest details, seem to contradict them.’ In other words, Lévi-Strauss argues that it is only by showing that myths do have this type of relation that one can convincingly show that the structures demonstrated do have psychological validity.
The demonstration of such relations between myths is the subject-matter of by far the largest of Lévi-Strauss’s books, the volume Introduction to the Science of mythology. This is a tracing of long chains of myths, meandering along numerous paths throughout aboriginal South and North America. What links these myths, and therefore makes the tracing of these chains possible, is the demonstration that all the myths are transformations one of another according to principles compatible with the psychological theories on which all of Lévi-Strauss’s structuralism is based. His latest book, which has just been translated under the title of The Way of the Masks, is simply a continuation of the voluminous study of Amerindian myth which he had previously undertaken: however, it is a continuation with a difference. The principal subject-matter is no longer myth, but material objects, and the famous ‘coppers’ of the North-West Coast.
First, Lévi-Strauss considers the myths associated with the masks of North American Indians and, not surprisingly, he tries to link them up in his usual way, by means of structural transformation. He then goes on to show that very similar relations exist between the objects themselves. In particular, the book demonstrates an almost total inversion in the features of the two related masks. Even more boldly, the book suggests that masks and ‘coppers’ are similarly structurally-related. ‘Coppers’ are large, characteristically-shaped copper plates, sometimes painted, which were the most valuable prestige objects of a set of North-West Coast Indian societies. They are famous in anthropological literature because of their connection with the large-scale title-validating feasts called ‘potlatch’ – a type of feast which has been used as evidence for a wide variety of theories and hypotheses.
This extension of Lévi-Strauss’s structuralism is extraordinarily stimulating: both the main subject-matter of the book and the subsidiary matters which the author considers in what he calls ‘three excursions’. There are, however, many problems with this new extension, as there are indeed with the whole edifice of ‘structuralism’. The specific problems with this latest book concern the question whether it is legitimate to consider objects which, once finished, have a life independent of their makers as similar to unwritten myths, which do not. Of course, Lévi-Strauss is aware of the difficulty, but I do not feel he has really dealt with it satisfactorily. Another worry that anthropologists will have with this contribution is that the ceremonies and rituals during which the masks are used are hardly discussed in any detail in the book. This is surely strange and unjustified. One would have thought that these are inseparable from the objects themselves and need to be treated with equal care.
The problems with structuralism itself are, of course, much more wide-ranging and cannot be considered properly here. I shall merely list areas of concern. First, the strength of the psychological foundation on which the whole theory rests is very, very much in doubt. What were taken to be the certainties of cybernetics in the Forties and Fifties do not appear as such today. Secondly, the ‘demonstration’ of structures and transformations, in spite of repeated efforts at precision, still remains a long way from meeting scientific criteria. Thirdly, the special relationship of transformation and event, which the whole theory supposes, is still insufficiently analysed. These are all very serious objections and although Lévi-Strauss repeatedly tries to answer critics on these points, they mean, I am afraid, that the whole of his work is an indication of what structuralism could reveal rather than a genuinely convincing demonstration that the job has really been done.
Lévi-Strauss has, defined paths for future work will probably be much more fruitful than the many blind-alleys of earlier anthropological theory. His work is a truly remarkable achievement. Its originality is a direct result of the fact that he came to anthropology from the outside. He has been able to build up this ambitious edifice by remaining a promeneur solitaire. Only in this way could he have been so original and creative. It has enabled him to explore with wonderful clarity and empathy innumerable problems which had not even been considered. It has produced a work which is both an intellectual and an aesthetic delight. This isolation, however, has also cost him dearly. It explains the extent to which he has been unable to modify his theory in the light of new developments in anthropology and especially in psychology. He has done enough, however, for it to be clear where one can fruitfully look for further insights.
Send Letters To:
London Review of Books,
28 Little Russell Street
London, WC1A 2HN
Please include name, address, and a telephone number.