Claude Levi-Strauss - French social anthropologist - interview

UNESCO Courier, Oct, 1990 by Manuel Osorio

Claude Levi-Strauss You are the author of a considerable body of work which has won respect all over the world. Those who know your name are more numerous than those who are familiar with your work. And so we should like to ask on behalf of our wide international public some basic questions about the key elements in your thinking. But first of all, could you tell us something about your career, about your early days?

-- As a writer I was actually a comparatively late starter. My career has had its ups and downs. After embarking on the study of law I turned to philosophy and then moved on first to sociology--it was to teach sociology that I went to the university of Sao Paulo in Brazil--and then to ethnography. It was only during the war, when I had taken refuge in the United States, that I began to write. On the one hand I used some of my field studies on the Nambikwara, and on the other I began my theoretical work by writing The Elementary Structures of Kinship.

This early research was doubtless accompanied by anxieties...

-- I would prefer to speak of uncertainties rather than anxieties. It would be more exact. My theoretical thinking took a certain direction which I could not see clearly. Then, while I was in the United States in 1942, I met the linguist Roman Jakobson and discovered that these vague ideas existed in the form of a body of doctrine in structural linguistics. I was immediately reassured.

At that time we were both teaching at the New School for Social Research in New York, a kind of university in exile which had been founded by French-speaking refugee scholars. We attended each other's lectures. I found in Jakobson's teaching the interpretative guidelines I needed. He encouraged me and even urged me to go further with my lectures on kinship.

Lectures which led to a book...

-- Yes, my first major book, The Elementary Structures of Kinship, which was also my doctoral thesis.

What coincidences did you find between the linguistics of Jakobson and the ethnographical research you were then carrying out? Linguistics and ethnography are two very different disciplines.

-- The differences are enormous. I was in no way attempting to transpose mechanically into ethnology what was being done in linguistics. It was rather a question of the same overall inspiration. With this very clear feeling, right from the start, that we were certainly not talking about the same things and that we could not talk in the same way about different things. But linguistics confirmed two points which until then I had only vaguely realized.

The first is that, to understand very complex phenomena, it is more important to consider the relationship between them than to study each of them in isolation. The second is the notion of the phoneme, as Jakobson developed it. Phonemes are distinctive sounds such as e, t, q, a, b, c, which signify nothing in themselves. It is the way in which they are combined to make words which makes it possible to differentiate the meanings--p and t mean nothing but in English they differentiate pot from tot.

This idea seemed to me to be extremely fertile. Once again, it was not a question of comparing ethnological facts with phonemes--they are realities of a different order. But when I tried, for example, to understand the extraordinary and apparently arbitrary diversity of the marriage rules which prevail in a given society, I did not ask myself the question people had asked until then: what does this rule, taken by itself, mean within this society? On the contrary, I considered that the rule meant nothing in itself, but that the way in which all the rules were combined, in opposition or in juxtaposition, was a means of expressing certain significations. For example, these rules served to formulate or stimulate cycles of exchange within the social group. Cycles of exchange which are themselves significant and which vary from society to society.

How would you sum up this new method?

-- Let's say that in all human societies there are first-order ensembles--language, kinship, religion, and law, to name only a few. In the early stages of research, it is perhaps wise to consider them as separate units. But at a later stage, one must ask what relationship there can be between these first-order units and the more complex second-order units which they constitute when combined. We thus gradually reach the idea of the "total social fact" as formulated by the French sociologist and anthropologist Marcel Mauss. It comprises a variety of representations, at once linguistic, legal, religious, and so on.

But this is within a given society taken as a totality...

-- Initially yes. But later one tries to understand the correlations and oppositions that may exist between neighbouring or distant societies.

You have written of "hot" and "cold" societies to differentiate between societies with a historical perspective and static societies...

-- These are theoretical poles. All societies are actually located somewhere along a spectrum. So-called primitive societies are not primitive at all. Their ideal is not to change, to remain in the state in which the gods created them at the beginning of time. Of course they do not succeed. They also exist in history. But they have a tendency to neutralize changes, to maintain an ideal state--that represented by myths. On the other hand, in societies which I have called "hot", we try to define ourselves in opposition to our ancestors. Change is thus much more rapid. We are not only aware of the existence of history but we wish with the knowledge we have of our past to reorientate the future, legitimize or criticize the evolution of our society. History for us is an element of our moral conscience.