Interview with Lévi-Strauss (1972)
The following text is a LLM (Claude Sonate 3.5) translation of an interview with Claude Lévi-Strauss from 1972. The interview, which follows on from a prior interview in 1967, was conducted by Raymond Bellour. Both interviews have been described as ‘a sort of work-in-progress report on the different volumes of the Mythologiques cycle, which appeared over a seven-year period, between 1964 and 1971’*. Nonetheless, this later interview, provides opportunity for more overall reflection. Taken together, the texts offer an informed, specialised dialogue on Lévi-Strauss’ project.
* C. Johnson (2003) ‘Lévi-Strauss in his Interviews’, Nottingham French Studies, Vol. 42, No.1, pp.33-47.
Interview with Claude Lévi-Strauss
Original French published in Claude Lévi-Strauss: Textes de et sur Clause Lévi-Strauss, edited by Raymond Bellour and Catherine Clément (Paris: Idées-Gallimard, 1979, pp.157-209).
RAYMOND BELLOUR: “The Raw and the Cooked,” the first volume of your “Mythologiques,” emphasized, by its very title, the transition from nature to culture through an opposition based on the absence and presence of cooking. The second volume, “From Honey to Ashes,” similarly opposed what you called “the surroundings of cooking.” The third volume, “The Origin of Table Manners,” suggested, again through the mediation of the culinary metaphor, the birth of morality. How does “The Naked Man,” the fourth and final volume of “Mythologiques,” fit into this development, to which its title, at first glance, appears unrelated?
CLAUDE LÉVI-STRAUSS: And yet, it was clear in my mind from the beginning that if the word “raw” should appear first in the series of titles, the word “naked” should be the last. Indeed, carried by myths from tropical America to the northern regions of North America, I found myself faced with this seemingly paradoxical situation: these are manifestly the same myths or, better yet, it is the same myth that emerges in such distant places of the New World. But at the same time, from one region to another, from one linguistic family, from one cultural group to another, a transformation occurred within the myths. For if, for the Indians of tropical America, the passage from nature to culture is symbolized by that from raw to cooked, it is symbolized for these North American Indians by the invention of adornments, ornaments, and beyond that, by the invention of commercial exchanges.
RB: So you find between naked and clothed the same opposition as between raw and cooked.
CLS: Yes, that is to say, a hero who, in tropical America, finds himself in a state of rawness before accessing culture, finds himself in North America reduced to a state of nakedness. This raw/naked transformation is not, moreover, unique to the Indians of America since it exists in many languages, notably in French and English. When we say “ride bareback” (“monter à cru”), we mean without a saddle, so in a sense naked, when the English say “to sleep in the raw,” they mean to sleep without pajamas.
RB: You said in your third volume that you had developed, in “The Raw and the Cooked,” a logic of sensible qualities, in “From Honey to Ashes” a logic of forms, in “The Origin of Table Manners” a logic of propositions. Could you, on the one hand, situate these terms, and on the other hand, specify how the approach of the fourth volume is ordered at this level?
CLS: Let’s first say that the three logical levels remain present throughout all four volumes. It’s more a shift in emphasis than exclusive categories. I wanted to show, in “The Raw and the Cooked,” that very elementary and neglected sensible qualities (the opposition between raw and cooked, fresh and rotten, dry and wet, high and low) were used by mythical thought as symbolic tokens that can be differently distributed and allow the formulation of certain logical propositions. As things become more complicated, that is, as I incorporate new myths into the analysis to be able to pursue it, these simple oppositions give way to oppositions, no longer between terms but between relationships of terms: this is what I wanted to indicate in the following two volumes by indicating that, progressively, this logic became more flexible and complex.
In the fourth volume, these same three logical forms obviously develop, but it insists more than the previous ones on the relationship between logical constructions that have their own determinism and the techno-economic infrastructure of the populations considered. Thus, the transformation of rawness into nakedness, which puts the problem of the origin of cooking in the background, but instead insists on the establishment of commercial exchanges (consumer goods, adornments, ornaments, clothes, but also matrimonial transactions) as a symbol of the passage from nature to culture, would be inconceivable if we were not dealing with populations that do not practice agriculture, unlike those of South America that I had dealt with, but live from fishing, hunting, gathering, and collecting, and compensate for this apparent rusticity with a fantastic development of trade. For these populations, who organize large intertribal fairs, main fairs and secondary fairs, where neighboring, friendly, and sometimes even hostile peoples met, the great privilege of civilization is no longer, as in South America, a relationship of opposition between man and animal (man eats cooked, animal eats raw), but a relationship of opposition between peoples whose commercial aptitudes allow for a diversified menu and those who are reduced to living off their own production alone.
RB: You say: “The relationship between logical constructions that have their own determinism and the techno-economic infrastructure of the populations considered.” How exactly do you conceive this relationship that you expressed in “The Savage Mind” in terms that already constituted an inflection of the classical model of the Marxist tradition?
CLS: You’re actually asking two questions there. First, if I use infrastructure and superstructure in the classic way. Yes, I think so, with nuances of course, since what is called in Marxist terminology relations of production does not present itself exactly in the same way in societies that have a very rudimentary technology and in ours. As for the question of determinism, I don’t think Marx himself ever conceived of a rigorous determinism that goes in only one direction: it is clear that when I refer to infrastructures, it is mostly to emphasize that there is nothing in human thought that is not thought of the world, understanding by world the physical world and the social world taken together. Of course, mythological constructions constitute kinds of relays, if I may say so, which themselves subsequently intervene in the determinism of thought. It’s not something simple, it’s not a cause-and-effect relationship: man thinks about the world according to certain mental constraints, and the way he thinks about the world largely determines how he acts on it.
In other words, infrastructures never act as such, that is, as a reality that would be given objectively and externally to man. They act only insofar as they are thought, and from the moment they are thought, they are already put into a certain form that has something obligatory about it. The mind does not find itself face to face with a world that is completely external to it: it apprehends it in the form of a text whose elaboration begins at the most elementary levels of sensibility. I was very struck, I say in the conclusion of “The Naked Man,” by various researches on sensory psychology and more particularly on vision. The fact, for example, that we do not see objects as they are or as we suppose they are, but based on a real coding that takes place at the cerebral level and which even, in some animals, begins at the level of the retina. This coding operates by pairs of oppositions: between immobility and movement, color and absence of color, horizontal direction or vertical direction, or oblique direction, and so on; which shows that what we call infrastructure is the indirect product of a reconstruction that is the work of the nervous system operating according to principles and methods that do not differ substantially from the way the understanding itself works.
RB: How were you led to operate in such a logical way this passage, already largely initiated in the third volume, between South America and North America? Indeed, you study, even if it seems difficult to say in front of the mass of volumes, only a relatively small number of myths (about a thousand) involving, both in the South and in the North, an equally small number of populations, even if the survey, in this last volume, knows a greater geographical extension. Do you feel that this relatively closed space that you managed to determine across the two Americas was strictly necessary to support the type of demonstration you were looking for? Could you have done it, for example, by deepening a survey limited to one or the other of the two hemispheres?
CLS: I started, as you know, with a Bororo myth that I had chosen because I knew this population, and because we have very rich ethnographic material about them. But I only gradually understood the profound reasons for this choice: this myth in fact occupies a pivotal position between two considerable mythical sets, present in both South America and North America. I thus progressively realized that several paths opened up from this myth, depending on which aspect one emphasized, and that one could almost, by barely forcing things, deduce a priori all the great themes of American mythology. I only understood this while writing the fourth volume.
That said, there were certainly other possible itineraries. But I don’t think I could have achieved a comparable result by limiting myself to one of the two hemispheres and developing the paradigm more completely in each case. The fact is that at a certain point in my analysis, at the beginning of “The Origin of Table Manners,” I hit a wall, for the simple reason that the mythological corpus of South America is much less rich than that of North America (not intrinsically, of course, but because much more work has been done in the North, and for longer). Certain elements were lacking, which were instead offered by North American myths. I was thus forced into this shift, which itself led to many others. I would never have reached this end if I had not put side by side mythological data from very diverse regions: it was only on the condition of involving peoples different in language, culture, and way of life that I could manage to bring out what they have in common, that is to say, as I try to show at the end of this last volume, the preferential use of a certain method of cooking food – the use of the earth oven – which seems to me to provide the profound reason for this identity, at first sight incomprehensible, between the myths of the northwest of North America and central South America. Let’s say, in a word, that I could never have brought out the invariant if I had not agreed to broaden the field of comparison, because only a broad comparison could allow it to stand out.
RB: I was struck by the fact that in this fourth volume, where all the threads of your investigation come together, you emphasize more than ever the demonstrative possibilities of your approach, always seeking to find in the empirical corpus the relationships previously deduced by a purely logical approach. I’m thinking, for instance, of these myths from the West Coast of North America, particularly the Coos myths, which suddenly gather the essential elements that were until then articulated in a deductive manner.
CLS: When I began my investigation, I hadn’t yet examined these myths. As I gradually understood that I had to turn to the North to find a solution to my problem, I discovered that everything I had obtained by deductive means was empirically realized there. Like a laboratory experiment that allows verification through synthesis of the results of successive and partial analyses. It’s the only type of demonstration we can aspire to in the human sciences. There’s often a tendency to contest the validity of our assertions, on the grounds that it’s impossible to check whether they are true or false. In the human sciences, we don’t work on objects of the physical world, but on representations, and we can never be sure that beyond or below the level of consciousness or unconsciousness where we place ourselves, there aren’t always other levels behind – and so on, as if in an abyss. Physicists and biologists agree, at a given time, on the reference plane where it’s appropriate to situate oneself to work usefully. In the human sciences, on the contrary, this reference plane remains permanently debated. The only demonstrations we can claim are those that allow us to explain more things than we could before. This doesn’t imply that they are true, but only that they prepare the way for other demonstrations that will come later to explain even more, and this indefinitely, without ever accessing a definitively acquired truth.
RB: Before continuing, I’d like to ask you to clarify a point that concerns the text of the myth as it appears in “Mythologiques.”
Does the summary you make from the existing material, that is, from the text collected by informants, have for you a pre-analytical value, determined by the requirements of systematization?
CLS: You’re raising the problem of the model there. We must distinguish several phases in its elaboration. First, the analysis of the myth itself, which involves readings, re-readings, incubations, until gradually essential contours appear that will provide a guiding thread. This isn’t a summary, but rather the extraction of a secret architecture. Then there’s a practical problem: how to explain this to the reader? We must first give them a summary, then, from this summary, try to reconstruct the myth as it appeared in extenso in the source used. The summary thus fulfills a dual function: it’s a didactic means for the benefit of the reader, and at the same time, by anticipation, a kind of approximation of the model we’re trying to construct.
RB: I assume you always work on the complete text, and never from the summary you make of it.
CLS: Absolutely. I would even say that it’s impossible to work otherwise than on the full text. Often it’s seemingly insignificant nuances, repetitions, what might seem the least important, the least interesting that suddenly, through an inversion effect like we have in optical illusions, makes what appeared to be contingent content become form and key to meaning.
The ideal would obviously be for the reader to have the sources in hand and also use only the sources. As this isn’t possible, we must give them a substitute. But all the argumentation, which is built from the summary, actually tries to reconstruct the text in its entirety, or at least to return to it.
RB: What is fundamentally the reason for the excessive length of the full text of the myth?
CLS: It’s very variable. One of the great difficulties in analyzing myths, or at least in writing books about myths, is that the author inevitably imprints his mark on the summary he gives of them and makes them a much more homogeneous reality than they are. Whereas from one population to another and from one informant to another within the same population, there are considerable stylistic differences. In some cases, the myth is too long because of what appears to us as repetitions: which may be true on the level of the story that the myth tells, but is not at all on the level of its sociological function, its poetic value, or its stylistic expression for the population concerned. It also happens that at the moment when we use it, we don’t need the entirety of the myth, but only certain elements, whereas if we had taken things from another end, in another sense, other aspects of the myth would be brought to light.
RB: This touches on a fundamental problem: that of the expressivity of the myth, which you affect through the summary, but also by the fact that, as you have often emphasized, the myth is not studied in its language, therefore never as text, but as another language, that which properly constitutes its story.
On the other hand, “Mythologiques” tells us nothing, except by default, about the concrete experience of the speech of the myth in the life of a tribe, about this lived moment of its recitation-transformation which constitutes, in a sense, its highest moment of truth.
CLS: Which would indeed be very important to describe and analyze.
RB: Do you think this is possible?
CLS: Most certainly. First, because in the ethnographic literature, there are many indications about this: about the fact, for example, that in certain populations there’s a season when myths can be told and a season when it’s forbidden; that, in certain groups, they can be told during a period of the day but not just any, only at night or during the day, that they must be listened to in a certain position, either sitting or lying down. On the other hand, it’s a study that can be done in the field, in populations that continue to tell their myths. There’s an admirable document in this regard, it’s the film that Marchand made with Lizot among the Yanomami in Venezuela. One witnesses the narration of a myth: it’s worth all possible descriptions, all possible analyses. One observes that the diction differs from that of everyday language, and that truly the narrator lives the myth while telling it, that he makes the events come alive through his mimicry, his diction, his gesticulation.
RB: Such an analysis would then imply a polyvalence of levels comparable to that which you implement in your analysis of Baudelaire’s “Les Chats.”
CLS: Absolutely. And myths should also be analyzed in this way. Subject, of course, to not wanting to paint a large fresco as I tried to do in “Mythologiques,” but by conducting an intensive study on the myths of a single population.
RB: How would you envisage solving the problems that this kind of analysis seems to raise? Because the myth is, if one can say so, both less and more than a text. On one hand, it doesn’t have the well-defined and relatively closed character of the text (even if the text, we know today, includes its perpetual part of incompleteness): it seems in this respect difficult to analyze a myth as one can do, even if it’s not without some artifice, with a poem or a short story. This is why you speak, I assume, of the myths of a population, as one can consider the work of an author, to reach a higher degree of constraint. On the other hand, as we have just seen, the myth is more than a text, and even more than a recited text, since it invents itself in being told and transforms itself in believing to repeat itself, and that its conditions of enunciation are thus perfectly consubstantial with it (which has in the text, as an equivalent, only the always fugitive exteriority of the time of the creative experience). Isn’t there a set of difficulties intrinsic to the extensive analysis of myth?
CLS: I admit that it can be difficult to analyze a myth if one doesn’t have several variants at one’s disposal. But a sufficiently long myth often contains all of its variants, in the form of segments given in succession, but in fact superimposable in the manner of the musical form called: theme and variations. Moreover, sufficient knowledge of the language would allow studying the myth not only, as I was obliged to do, from a comparative perspective, from a semantic and ethnographic angle, but also from other angles: stylistic, prosodic, syntactic, grammatical, phonetic and phonological, each of which offers formal differences in relation to the others, but also coincidences or resurgences of invariant formal characteristics. Thus, the analysis of a myth would approach that of a poem or a short story. Is it really certain, moreover, that in these last two cases, the poem or the short story don’t constitute variants in relation to other poems, other short stories by the same author? That “Les Chats,” for example, is not, as a poem, a variant of other poems by Baudelaire where these felines, lovers, scholars are mentioned; and, on a formal level, as a sonnet, a variant of all the other sonnets in Les Fleurs du mal? To take another example, true both on the level of poetry and music, how can one doubt that the third act of Parsifal is a variant of the same act of Die Meistersinger? Here and there, an older and experienced man (Gurnemanz or Sachs) steps aside for a younger and exceptionally gifted one whom he installs. Here and there, their long tête-à-tête precedes the march to consecration with, between these two phases, an intermediate moment of appeasement and rediscovered unanimity (“Good Friday Spell” or “Quintet”). One wouldn’t make a satisfactory analysis of one without knowing and bringing in the other. Therefore, the analysis of Die Meistersinger should have waited until Parsifal was written, and that of Parsifal will remain incomplete because we will never know the opera that Wagner would have written next if he hadn’t died (just as we will never know the variants of Pelléas that The Fall of the House of Usher would have revealed). The differences and difficulties seem to me less great than your remarks suggest. In fact, I remember dedicating one of my first courses at the Collège de France, that is, thirteen one-hour lessons, to the analysis of a single Iroquois myth; and my work on The Story of Asdiwal, published in 1958, also focused on a single myth, admittedly studied from three variants. It’s only recently that I realized I had neglected a fourth one which, like a laboratory experiment, could have invalidated the hypotheses elaborated from the other three – or confirmed them, which I hope to show soon.
RB: Doesn’t it seem to you, in a comparable order of ideas, that ritual poses even more radically the problem of the level of analysis? At the end of The Naked Man, you devote a few very enlightening pages to the comparison of ritual and myth, which end up opposing them as the continuous and the discontinuous, living and thinking. How can one conceive, under these conditions, the idea of an approach comparable to that of Mythologiques, a “ritology,” and is it even possible?
CLS: Perfectly possible since we frequently observe that what one society tries to “say” in the form of myth, a neighboring society chooses to “say” in the form of ritual.
I have given examples of this in Structural Anthropology (Chapter XII), in a recent volume in honor of E. E. Evans-Pritchard (The Translation of Culture, 1971) and throughout Mythologiques; for instance in The Raw and the Cooked, when I use a Sherente ritual to interpret the myth of this tribe and neighboring variants on the origin of fire.
The reason is that ritual, as we can grasp it in its empirical manifestations, is always imbued with implicit mythology, and we are therefore not practically dealing with different essences. It’s only when, through an ideal analysis, we purify the notion of ritual to rid it of all the implicit mythology it always contains in fact, that we realize that myth and ritual have opposite purposes.
RB: Your research on myths was born, you said, from a kind of feeling of irritation with the arbitrary and illogical character that these myths that fascinated you seemed to present down to their smallest details. Do you have the impression today of having really managed to reduce them? Or are there still details in the myths you’ve worked on that have remained inexplicable to you? In this case, do you think this is due to a defective state of the corpus (either lack of ethnographic information or lack of necessary variants) or to an intrinsic difficulty in entering the frameworks of your formalization?
CLS: One never manages to reduce everything, or else research work would stop. Others will come – at least I hope so – who will take up the myths I’ve worked on and, perhaps aided by my groundwork, will perceive new relationships, integrate additional details, and advance the analysis beyond the point where I left it myself. To answer your question more specifically, yes, there are here and there details that remain inexplicable; but not because of a deficiency in the method: simply because they refer, or could refer, to ethnographic or other facts that the available literature doesn’t provide. For example, was a certain type of mask that has now disappeared, whose founding myth resembles the founding myth of another type attested in a neighboring tribe, the same mask or a different mask? Does the prohibition, mentioned by a myth, of revealing that one has consumed a certain species of fish with the bones, stem from the fact that this consumption was really forbidden, or is it a mythical elaboration whose explanation must be sought in the context? When a myth attributes to an animal a certain behavior in specific circumstances, does it take up an empirical observation, or an independent but unfounded belief, or does this attribution result from an internal constraint of the myth? One constantly encounters such problems, some of which can be solved while others resist, temporarily or permanently… A myth from the Pacific coast tells that if a fish of a certain species risks surfacing, its stomach comes out of its mouth and it dies. As this fish has rich and precise mythical functions, it would be crucial to know if this infirmity has a real basis, or if it results from the symbolic properties attributed to the animal because of these functions, which can then be illuminated, specified or developed.
RB: You have considered in the whole of Mythologiques, as we have seen, nearly a thousand myths. Do you think that using a larger number of myths – starting with all those that serve you to open as many virtual paths during the third and fourth volumes – would be likely to modify the conclusions you draw at the level of the entire American continent?
CLS: What have I tried to do? To describe a mythological discourse of the American Indians. But it’s not their only mythological discourse, far from it, since I have used only 10%, perhaps even 5% of the available corpus that currently represents North and South American mythology. The explanation of the rest would first give us other mythical discourses. I simply think that the one I have retained has exemplary value and that the others, if we had them, when we will have them, constitute either parallel discourses, or discourses that intersect the one I have retained, at certain points of its itinerary. If we were able (as we will be one day provided that others tackle myths in the same spirit as me), if we had the total discourse (and this, note well, is a purely theoretical notion, because there is no total discourse: I have used only a small part of the myths, but we possess only a tiny part of the myths that have existed), I hope that this total discourse would have nothing contradictory with the one I have uncovered. It would obviously be much richer, more varied, but it would be substantially of the same type.
RB: Do you, on the other hand, have the feeling that in the eventuality of an extension, of course still quite hypothetical, of your attempt in the perspective of a general mythology, one would find oneself among the multiple mythologies in a relationship of convergence of the same order, or that on the contrary one might find oneself faced with more or less radical divergences?
CLS: You’re raising there the problem of the unity of mythical thought throughout the world: it’s not a problem whose solution can be prejudged. It’s only to the extent that studies of the same order are made on other corpora that we will be able to know if this unity exists, in whatever form, or if on the contrary it’s appropriate to distinguish mythological species as we distinguish living species; and if these mythological species differ from each other like species of mammals, that is to say by preserving a certain kinship, or like species belonging to quite different zoological families, or even like plant or mineral species in relation to animal species.
CLS: Of course. All I can say for now is that we can be fairly certain that there exist in other regions of the world levels of mythical thought that are of the same order, as extremely rapid surveys already allow us to verify (but without prejudging that all levels of mythical thought are of the same order: even for America, I have carefully left aside certain levels of mythical thought, for example those that manifestly proceed from an elaboration by brotherhoods of philosophers, of sages; they have, of course, relationships with those I have considered, but would still need to be analyzed in themselves, for themselves, independently).
RB: Which would then make the differences not only in mythical content, but at the very level of forms, systems of relationship between terms.
CLS: Of course. All I can say for now is that we can be fairly certain that there exist in other regions of the world levels of mythical thought that are of the same order, as extremely rapid surveys already allow us to verify (but without prejudging that all levels of mythical thought are of the same order: even for America, I have carefully left aside certain levels of mythical thought, for example those that manifestly proceed from an elaboration by brotherhoods of philosophers, of sages; they have, of course, relationships with those I have considered, but would still need to be analyzed in themselves, for themselves, independently). I’m thinking of Luc de Heusch’s research on the mythology of a certain region of Africa, as well as various research on Australian mythology. I’m especially thinking of the Center for Comparative Research on Ancient Societies: the work of Jean-Pierre Vernant, Pierre Vidal-Naquet, Marcel Detienne seems to show that there are in Greek mythology certain levels where one finds oneself almost on an equal footing with American thought.
RB: This is a fascinating problem because, unlike American myths and so many others that are strictly oral in character, Greek myths are already textualized myths.
CLS: One could answer you that all collected myths are, by necessity, textualized. This is a condition that poses all sorts of problems. It’s curious that in Greek myths, whether one has one or several versions, a particular solution is given to a mythological problem. This is a point to which my attention was drawn when reading Detienne’s book, The Gardens of Adonis. He gathers there an Adonis mythology that reminded me of American myths, undeniably of the Adonis type. I’m not saying by this that they have the same origin, because we know absolutely nothing about that, but simply that from a typological point of view, they are classified in the same family of myths. Now, to the Adonis problematic, that is to say that of a mediating individual who finds himself placed between extreme terms, in this case a celestial goddess and a chthonian goddess, Aphrodite and Persephone, the Greek versions we possess give a unique solution, namely that Adonis dies, and that he consequently finds himself disjoined from his two mistresses. On the contrary, the equivalent American myths consider all possibilities: there are indeed several ways for a mediation to fail, either the mediator joins one of the two poles and completely disjoins from the other (and in this case, not always the same one), or he disjoins from both. Now the American myths present all three solutions.
RB: As if written mythology had made a choice among the possibilities.
CLS: Yes, unless Greek popular mythology, from which the written tradition is derived, was already working in this way. Vernant, Detienne, and Vidal-Naquet have very often noted, and this seems quite right to me, that the Greeks were already doing structural analysis of their mythology themselves: the work that analysis accomplishes on savage myths appears on the surface, if I may say so, in Greek myths. One could go so far as to say that the myths themselves do their own structural analysis: the names of the characters, of the divinities, for example, already express the structural content of the myths. So, in one sense, it’s almost the same thing, in another it’s very different.
RB: This reflexive gap of the myth on itself is perhaps not indifferent to the phenomenon of philosophical emergence which you have many times noted that, for reasons that are difficult to determine, it occurred in Greece, contrary to the continents of savage thought where it could have happened. In any case, if I understand correctly, you seem to suggest, without being able to prejudge it yet, that there is between the various mythological species a certain convergence or at least relationships close enough that we can rather distribute them in a vast table of variations than in the form of several heterogeneous tables.
CLS: Yes, but with one reservation, and on which I insist a lot. Mythical thought is a bit like a layered cake: levels of mythical thought are superimposed. I have tried to follow one layer, and if I think that this layer, by its structural properties, is found in other mythical spaces, that doesn’t mean that in these other mythical spaces, there don’t exist, just as in America, other layers.
RB: Could you specify what you mean exactly by “layer” or “level” of mythical thought? For example, how is the layer you have chosen to study defined, in the American continent, in relation to others that you have, for one reason or another, preferred to exclude (without perhaps being able, moreover, to assign them exact limits yet). On the other hand, is this delimitation relative to criteria of an empirical, geographical, racial, socio-historical order, or is it purely conceptual, that is to say, operating a division according to its own conditions of logical possibility?
CLS: Certain layers are easily distinguished by objective characteristics. Thus, even among peoples said to be savage, there can be popular mythologies and learned mythologies; there can also be mythologies relatively free from ritual and mythologies that are, on the contrary, closely intertwined with it. But let’s limit ourselves to the case of a more or less homogeneous mythology and start from a myth chosen (at least we believe so) at random. The interpretation of this myth will bring in others, those others in turn, and so on, but it is doubtful that our journey will have allowed us to exhaust the corpus.
I will then call a layer the set of myths that we will have effectively traversed or followed. Another starting point would allow us to penetrate and follow another layer, it being understood that often the layers will join to merge, or cross to go in different directions. After several operations of this type, we would better understand the architectonics of the system under consideration. In the last volume of Mythologiques, I sketched (regarding the group of myths called the “battle of fawns against bear cubs”) what could be a journey along a layer different from the one I was evolving in. We are too early in such studies for me to answer you about the nature of these layers and their degree of reality. At the point where I am, they seem to me only to translate the fact that the network of interconnections between myths is so complex that it is practically impossible to follow all the paths at once.
RB: Let’s now come to a set of methodological questions posed by Mythologiques. First, the codes, whose constitution and traversal regulate the course of the book. Do you think you have identified in the myths you have considered all the codes that are at play to constitute this “intercode” by which you define myth? On the other hand, is the distinction between codes, which diversifies them for example into astronomical, sociological, culinary, acoustic, alimentary codes, etc., in your opinion, quite rigorous?
CLS: (1) No, certainly not; it is perfectly conceivable that other codes can be identified. (2) These codes are defined by a certain number of categories which are our own categories of thought. For the myths themselves, the distinction does not exist, since their object is precisely to unify or transcend all these codes and to use not one or the other or this or that, but a super-code formed of all of them. These are distinctions that are made for the need of analysis, but which in themselves correspond to nothing real.
RB: Next, binarism, which you return to more insistently than ever in the fourth volume: first in the chapter that announces the closure of the investigation, the “binary operators,” then in a more theoretical way in the “Finale” where it allows you to make a fragile but precise junction between the human sciences and the exact sciences in the form of natural sciences. What do you think is the consequence of the fact that binarism thus functions sometimes at the most microscopic level when it organizes contrasted elements in pairs and at the level of a general principle insofar as it transcends the categories of different scientific cuts to find a kind of objective foundation in the order of life sciences?
CLS: Several aspects of binarism must be distinguished. First, binarism exists. Since linguists use it and since we very often see it objectively attested in the ways of thinking of the peoples we study. I’ve often been told that the opposition between nature and culture, for example, which I make such great use of, was a creation of ethnologists and that it couldn’t be superimposed on the systems of thought they study: I believe nothing is more false. Ethnologists could only conceive of this opposition because they borrowed it from their object of study. Of course, it’s not always expressed so directly; it can just as well be in the form of an opposition between the inhabited village and the bush, the forest and the cleared land, cooking and rawness, etc., but it’s always given to us by the material of our studies.
On the other hand, binarism constitutes a kind of common denominator or least common multiple, I don’t know, as soon as it’s a matter of translating the productions of one mental experience into the language of another mental experience. And this doesn’t imply its objective existence but simply its convenience. It’s a code that allows translating a message from one language to another, or rather finding the common elements of two messages, at the level where they make sense to each other.
If we take into account these two observations, on the one hand the objective existence of binary oppositions, on the other hand their methodological and practical effectiveness, it’s probably because they correspond to something at a more fundamental level. This is where we can turn to the natural sciences and ask them why, on the one hand, binarism appears in surface structure, on the other hand if it isn’t a convenient tool because it corresponds to certain modes of functioning of the brain itself and the sense organs. Physiologists and anatomists tell us passionately interesting things in this regard. They show us that the brain functions, perhaps not exclusively, but in any case very largely, as a binary machine, they show us, as I was saying earlier, that the coding that takes place at the very level of sensibility is a binary coding and that thus, when we come to use binarism, we are perhaps at the point where contact can best be made between the very foundations of thought and sensibility and their exercise.
RB: Don’t you ever have the impression of forcing the original text of the myth somewhat in the name of the binary opposition rule, perhaps even from the summary but more essentially during the analyses? Indeed, one sometimes believes to be faced with operations that border on prestidigitation, both obvious and half-justified. I’m thinking, for example, of a formulation that seems almost to be a pun: “The grandmother copulates backwards there instead of doing the reverse of copulating.”
CLS: Puns, like proverbs, riddles or – more particular to our civilization – advertising slogans, make constant use of elementary processes that we observe at work in mythical thought. Your remark therefore seems very pertinent to me. In the formula you quote, what is there other than a condensed way (but which the context makes explicit) of specifying two things? First, that if a society considers a way of copulating normal, one can copulate backwards there: from behind instead of from the front, the woman in a superior position instead of an inferior position, etc. Then, that if, as I have tried to show, indigenous thought establishes a homology between copulation and food consumption, it will perceive the reverse of food consumption, that is to say excretion, as homologous to the reverse of copulation. Now it’s not me, but the variants of myths that anyone can read, that apply themselves to exhausting all the resources of such a combinatorics. That said, I by no means limit myself to an analysis by binary oppositions. The index of The Naked Man contains about twenty references to “triads,” that of The Origin of Table Manners about thirty, and, in a note in From Honey to Ashes (p. 74), I give numerous examples of the use of a model that is not binary, but analogical.
RB: Binary oppositions manifest themselves in your analyses in the form of a set of operations, mechanisms that ensure this “impoverishment of the real” consubstantial with structural analysis; you evoke at the end of The Naked Man these “clear-cut relationships such as contrariety, contradiction, inversion or symmetry.” These are the same operations whose “very loose meanings” you deplored from The Raw and the Cooked onward. Do you have the feeling that these terms have become more precise through the progression of the volumes? And what do you think are the problems posed by the relative indefiniteness of your operative terms?
CLS: In my language, these terms are only approximations and will remain, I’m afraid, as loose and as floating until the end. The reason is that I’m not a good enough logician to try to deepen the meaning of the operations I perform on myths and to classify them. I work on myths as a primitive, compared to what could be done or what should be done. Young mathematicians now talk to me about new developments in logic where it seems that one wouldn’t even need to worry about the nature of the relationships that exist between terms. This is what’s called category theory: it’s based on the notion of morphism, that is to say elements or relationships that can be manipulated without having to question the logical nature of one or the other. Perhaps there will be a way to overcome the difficulty there one day.
RB: Can you specify how you envisage this relationship between the analysis of myths as you practice it and this “science of myths,” this “true science” in relation to which you position yourself so modestly from The Raw and the Cooked onward?
CLS: I don’t think we should adopt a mystical attitude towards this problem. Several aspects must be distinguished. Let’s first get rid of the use of calculators and computers. There are mythological corpora that are enormous and prodigiously complicated, a bit similar to the Brontë Juvenilia: the plots are obscure, the characters constantly change names during the narrative. The confusion is such that we would need to have computer resources at the outset to unravel the text. I think that computers would render us the greatest services there, to constitute vast indexes and give us very quickly the correlations that appear in the corpus between chosen elements. We could then start working, under conditions similar to those in which we found ourselves without computers, but with a much more reliable basis: this doesn’t actually pose any problem.
Let’s come to the point of a true logico-mathematical treatment which I don’t claim at all. To establish transformation relationships between myths, I perform a certain number of operations of which I don’t really know what they are. I would very much like someone more competent than me in logic, and who has time to waste, to lean over my work to try to determine the nature of these operations by trying to classify them. Indeed, I don’t know if they are very numerous because they are all different from each other; or if on the contrary they are always the same and thus reduce to a small set of recurrent operations that are used each time according to needs. If these operations are extremely numerous and if, at the limit, there are as many operations as interpretations, they amount to a simple description, we have done nothing other than describe the material in a language different from that in which it was formulated at the start, and we escape any possibility of logical treatment. If, on the other hand, these operations are relatively fewer than the cases in which they are used, we can say that we are faced with a logic defined by a set of rules and pose all sorts of problems: are these operations hierarchical? does the use of one operation at a certain moment entail the use of another operation at another moment? can we define the recurrent rules of such a grammar? If this were possible, we could then attempt to generate myths by implementing these operations in accordance with the rules that have been defined. All this would be very interesting but would go far beyond what I claimed to do.
RB: In any case, the very stake of Mythologiques as such seems to me to lie essentially in this impossibility of any strictly logifiable categorization; it constitutes the very engine of the book, its spiral development. There is something here that seems perfectly consubstantial with the human sciences as soon as they are exercised in the perspective of an analysis of signs, and particularly of large signifying units.
Insofar as they are obliged to laterally reflect their operations, we witness a permanent doubling of the object that prohibits the very notion of properly scientific result, in favor, inversely, if one can say so, of an equivalent spoken in another language, more logified, more conceptual obviously, but which engenders a speech, here a text without true beginning or end, rather than, in the strict sense, a science of language.
CLS: I quite agree with you and that’s why if the human sciences must look towards the natural sciences and the exact sciences, they cannot, congenitally, join them, due to the very nature of their object. I believe that we are fatally condemned to always wander between two poles, that with one hand we cling to the sensible and with the other to the intelligible. All we can hope for is that there doesn’t appear too visible a hiatus, that there is a moment when we, who try to do the work of the scientist, take over from the poet, the artist or the savage, and another moment when scientists in the full sense of the term can themselves take over from what we do.
That’s why it’s so annoying to sometimes be confronted, by specialists in logic or natural sciences, with the argument that our hypotheses are not, as they say, falsifiable, in the sense that, in biology as in physics, one can show that a hypothesis is true or false. I believe this is an impossibility consubstantial with the nature of the things we operate on. Biologists or physicists can reach a consensus on the level of reference where, at a certain stage of development of their science, it is appropriate to place oneself, whereas our level of reference is always in question. There is no agreement between us on this. We oscillate, if one can say so, like Cartesian divers, between the various possible levels of reference.
RB: This is to respond to the unrefined criticisms of scientific impatience. But it seems to me that your attacks, in the “Finale” of The Naked Man where you look back on your entire enterprise and respond to the various objections it has raised, are addressed especially and in a much more virulent way to those you qualify perhaps a bit too globally as “philosophers”.
CLS: Let’s say a riposte rather than an attack. It’s philosophers who attacked me and to whom I resigned myself to respond after keeping silent for a very long time. But when I speak of “philosophers”, I don’t mean “all philosophers”.
RB: What are these criticisms that these philosophers addressed to you?
CLS: Essentially two kinds of criticisms. The first is that the type of analysis I engage in is desiccating, situated on a purely intellectual level, ignoring all the data of sensibility and affectivity. The second, that my analysis of myths would impoverish them to the point of removing all meaning from them.
It seems to me, on the contrary, that structuralist thought is a thought that constantly seeks – and I believe this is a reason for the attraction it exerted, at least for a certain time – to reconcile the sensible and the intelligible and to refuse to compartmentalize man, as precisely some philosophers do who establish a separation between the domain of science and another that would belong to them properly. I believe that it is only possible to understand man from the moment when the type of explanation one seeks aims to reconcile art and logic, thought and life, the sensible and the intelligible. If these terms appear to us separated, as disjointed, we must blame only the congenital infirmity of our knowledge.
On the other hand, I have tried to highlight the richness of the type of myth analysis that I propose or that I bring back to honor after Plutarch, who is its initiator. It makes myths gush forth, contrary to what one might believe, infinitely more meaning than the banalities that one was generally content to find in them. If something emerges from the whole enterprise, it’s what I would call the pregnancy of detail: there is nothing in the myth that one can ignore, nothing that one can say is bizarre or absurd to preserve the purity of some great eternal truths. There is nothing that should not be taken into account, consequently that doesn’t have a meaning. I have precisely tried to show what this meaning is, or what these meanings are.
RB: When you say: there is nothing in the myth that doesn’t signify, one can’t help but think of the “everything signifies” which Barthes made the indispensable prerequisite of S/Z. It’s to this, no doubt, that we owe the feeling throughout Mythologiques and especially in this last volume which would like to grasp them all, of a sort of vertigo: for if the perpetual ascent from detail to overall law alone allows conferring on it a validity of dynamic reality, it ceaselessly in return unrealizes itself to the point of vertigo, under the ever-increasing flowering of detail which thus constitutes the particular tragic of this kind of books.
CLS: The great difficulty comes from the fact that the reader, whoever they may be and even if they have ethnological training, fatally approaches these myths from the outside. I took three years to write this last volume while no one, naturally, will devote three years to reading it. And yet, all this time was necessary to impregnate myself to such an extent with the substance of the myths that I knew them practically by heart.
On the other hand, this last volume, already much larger than the previous ones, would like to replace the three or four other volumes that should really have been devoted to the same material. But I was, throughout my work, obsessed by what happened to Saussure with the Nibelungen, on which he worked for years, to which he devoted countless notebooks. One realizes very well by reading them that as his fascinating study progressed, he became so crushed, drowned by his own materials that he could no longer hold the thread.
This was the major peril that awaited me throughout the composition of Mythologiques. I gave myself the inviolable principle that I should not succumb to it, and that I had to at all costs, even at the cost of trying the reader, conduct the enterprise to its term. Finally, it must be considered that a book like The Naked Man and more broadly the whole of Mythologiques, is also a repertoire of several dozen research themes that I was content to sketch out, to put in place, so that others, if they wish, can consider them successively and deepen them.
RB: You speak there, and this struck me during this interview, as if these researches were really accumulable. How can one envisage that multiple volumes comparable to Mythologiques come to be added to it without fearing that the mass will find itself obscuring what it should illuminate? One can only think of it, it seems to me, on the sole condition of this logico-mathematical treatment that you evoked, capable on the one hand of reducing the books seized in the reduplicative fatality that precisely makes them books, on the other hand of gradually substituting for their impossible stacking.
CLS: Not exclusively or necessarily. The great scientific hypotheses rest on an overwhelming mass of detailed work, of experimental protocols and it’s only from time to time – every twenty, thirty years, more – that a mind sometimes gifted for this kind of enterprise synthesizes them. By describing a considerable number of mythological universes, we will not put ourselves in a situation different from that in which are normally found, in the domain of the human sciences themselves, ethnographers, linguists, historians who accumulate descriptions of societies, languages, epochs: and then, one fine day, along comes a Durkheim, a Saussure, a Febvre who gives meaning to all that.
It’s banal to say it, but scientific progress takes shape in the manner of an iceberg: the cultured public only perceives the emerged tip, without suspecting the enormity of the mass of analyses and documents that supports it. The Mythologiques enterprise rests on the hundreds of works cited in the bibliography of each volume, and a provisionally satisfactory theory of myth, if it ever sees the light of day, will itself rest on dozens of enterprises comparable to mine. One must rid oneself of the philosophical illusion that there exist royal roads or shortcuts on the route to truth.
RB: In a final section, I’d like to try to situate some problems raised by your conception of structural analysis. We can thus question the relationships between your work and that of Georges Dumézil. In a recent interview (1) where you don’t fail to acknowledge, as always, what you owe them, you highlight two fundamental distinctions: one, practical, which stems from the fact that Dumézil works largely on dead languages and must thus give a preponderant part to philological work; the other, theoretical, which opposes two axes of concerns: that of Dumézil, who “uses Indo-European myths to brilliantly demonstrate a fundamental characteristic of Indo-European societies and thought, namely the predominance of a tripartite structure”, and yours, which studies myths in themselves to uncover certain fundamental properties of mythical thought.
Is it the difference in objects that seems to you to motivate the difference between two types of structural approach, one based on the similarity between terms, the other on their opposition? Or do you think, on the contrary, that both methods are applicable to all the objects considered? This would then imply that the myth, considered in both cases only at the level of its story, could be the subject of two different readings, whose complementarity and modes of articulation it would be essential to be able to measure. Would it be possible, in your opinion, to discern in savage thought, and particularly in that of the American Indians as it transpires through their myths, something comparable to the Indo-European tripartition, that is to say, a set of structural terms that would underpin less logical operations determined by a set of material conditions than a global representation of the social, economic, political, in a word historical, reality of the societies considered? This is in fact the question that Pierre Smith and Dan Sperber ask, in other terms, when, at the end of their study “Mythologiques by Georges Dumézil” (2), they invoke the need to “specify in what way it is permissible to abstract the theoretical device that is symbolism from this more concrete ensemble that constitutes ideology”.
CLS: Dumézil makes constant use of a method that he might not like to hear called structural but which seems authentically so to me, for particular purposes; and Saussure, whom we were talking about earlier, applies a method of already structuralist spirit for a much more narrowly historical demonstration, namely that the Nibelungen are a chronicle of the first kingdom of Burgundy. It is therefore not, as you seem to think, two methods, but the same method applied for different purposes: either, as you say, to bring to light certain properties of mythical thought, or to uncover a conception of the world specific to one or several societies. In fact, I have the impression of doing both things at once. Thus, when the analysis of myths allows me to uncover the meaning of the opposition between the continuous and the discontinuous in American thought and societies, the particular place occupied by what I have called chromatism, these are ideological factors that are situated at a level formally comparable to the principle of tripartition in Indo-European thought. Reciprocally, Vernant, Detienne and Vidal-Naquet work on Greek materials, therefore Indo-European, to reach certain differential characteristics of Greek society within the Indo-European ensemble (a narrower task than that proposed by Dumézil), which does not prevent that, on the level of myth analysis, some of their results corroborate those which, on a much broader level, I believe I have reached. To take another example, the Zuni Indians “historicize” their myths in a way quite similar to that which Dumézil was able to highlight among the Romans. Let’s not try to decide a priori in matters that we know so poorly, let’s not introduce from the outset distinctions of principle such as that between symbolism and ideology. Let’s rather apply ourselves, each of us, to the tasks for which we feel a taste and believe we have a vocation; and we will gradually realize that they converge.
RB: For several years now, you have given several examples of “outings” that show the possibility of applying your method of myth analysis to cultural objects of another order. You say, on the other hand, in the broadcast that Michel Tréguer dedicated to you four years ago: “Politics, the sense of history, our attitude towards our future or our past, this seems to me obviously charged with mythology and with a very comparable mythology, and probably studiable according to the same methods as those we are going to apply to distant examples (…) … we can, by studying these means of mass communication and the way they develop, find in a nascent state phenomena that are not so different from those we are going to look for in extremely exotic populations.”
How do you understand the possibility that you seem to advocate here of constituting and studying a contemporary mythology?
CLS: It’s not at all that I advocate it. I simply mean that it would be conceivable, although certainly much more difficult than in the case of savage thought. If you take history, I agree with Evans-Pritchard when he emphasizes that we cover two completely different things by “history”. There are on the one hand the events as they occurred, on the other hand the way we represent these events, which is the only way we can access them. There is an infinity of events since at every minute, every second something has always happened: consequently we choose the events, but we don’t choose them in the same way, we don’t establish the same relationships between them, we don’t give them the same meaning. This history which is a collective representation or a representation for a certain fraction of the collectivity (because there are several collective representations of the same history in a society), this collective representation of events seems to me close, by its nature or by its function, to what we call myth in societies without writing. I therefore believe that there would be a possibility of applying a similar method to history understood in this way. Of course, it would not be a matter of treating history itself as a myth, that is to say, of constituting a mythical history from and in addition to history but, considering the representation of the history of France that people on the right or people on the left might have, and among people on the left, socialists or radical-socialists, or communists or leftists, to investigate how certain distortions are constituted at the level of the narrative: and I don’t mean by this: distortions in relation to reality, but a system of transformations ordered as a set of autonomous variants.
RB: But don’t you have the impression that on the one hand we would then be faced with extremely heterogeneous discourses in relation to each other, that is to say presenting a much less strong degree of autonomy than that which myth maintains in relation to itself, and that on the other hand we would risk ending up discovering only things that we already know? I thought about this in relation to American cinema which seemed to me to particularly justify this mode of approach: it is indeed, among the means of mass communication, one of the most homogeneous that one can imagine. Now I’m afraid that a study of films that would aim to bring to light the structural invariants of American cinema would end up ultimately highlighting, with more or less descriptive success, more or less rigorous information, only a certain number of commonplaces, or rather obviousnesses, that is, things that we already know or at least largely suspect.
CLS: What you say there is quite right, and has often struck me when reading various attempts at structural analysis of literary works or others from our own society. One often has the impression of finding the “in the manner of” Racine, by Muller and Reboux, which is not only a pastiche of Racine, but at the same time that of a school textbook containing an extract from a tragedy by this author, embellished with explanatory notes attributed to a certain Mr. Dragonfly, I believe, a secondary school teacher, and which are reduced to a rather grotesque paraphrase of what was already in the text. These failures prove how difficult this kind of study is, and you gave a reason for it earlier by emphasizing that we are faced with discourses that are often heterogeneous in relation to each other.
But that’s not the only reason. What has generally been done in this field remains terribly superficial, and therefore uninteresting, because we don’t measure the scope of the task that a serious structural analysis of literary works that are both complex and extensive would represent. I don’t think I’m mistaken in saying that the analysis of a sonnet, that is to say fourteen half-lines, could require up to a hundred pages. Imagine what it would take to treat a novel according to the same principles! But – and this is another difficulty – would the principles be exactly the same? This question has hardly been asked; yet, a geological formation extending over tens of kilometers, and whose structure can be analyzed, is not structured in the same way or to the same degree as a block of crystal. The latter is rigorously structured at all levels, and the geometric shape given to the eye directly translates the arrangement of molecules which translates that of atoms. The genesis of the crystal merges with its form, diachronic analysis merges with synchronic analysis. On the contrary, the two remain distinct in the case of the geological formation I mentioned a moment ago, where levels that can be said to be structured are mixed with others where disorder and probability reign. Perhaps there are differences of the same order between poetry and novelistic prose, and one cannot even exclude that certain literary productions are as subject to contingency and the play of probabilities as, for example, a pile of rubble or a scree… Not everything is structured, and there is not necessarily structure everywhere. The question of principle that one must always ask oneself is: assuming that structure is present, where does it lie? To arrive at something a little new, it is only at this level that one should place oneself, and it rarely coincides with the patent level with which one too often contents oneself. That’s why, when one analyzes a poem, for example, one doesn’t limit oneself to commenting on the meaning, but one begins by decomposing it according to its multiple levels, phonological, syntactic, prosodic, etc.
RB: This would explain why structural analyses conceived according to your perspectives can only be successfully practiced on cultural objects of our societies on the sole condition of restricting the choice of corpus and pluralizing the levels of analysis, as opposed to what you did in Mythologiques. Whether it’s an object of artistic production or a layer of more deliberately sociological signs, what we need to find there, for us who are more or less its contemporaries, is this value of difference, of “exoticism”, Segalen would say, that myth possesses in itself obviously as a natural production of societies that are radically other to us. Hence the obligation of very circumscribed and polyvalent fields, so that the uncovering of a latent functioning can escape the banality of the general idea. Which Barthes understood well, for example, either by pinning down in his Mythologies a trait deepened in a spiral on itself (and enhanced by the work of writing), or by limiting his Fashion System to the written fashion garment, or to a rather narrow fringe of social discourse. But on both sides, nothing really equivalent to this signifying universality which gives itself, and for us without any banality, through the analysis of myth, to the very extent that these discourses, these productions were born, ultimately, as you show for the novel in already classic pages of The Origin of Table Manners, only from the bursting and decomposition of myth.
That’s why analyses similar to yours are either very closed in on themselves, or on the contrary can sometimes be practiced as relays, so to speak, in overall studies that cannot be reduced to them. I’m thinking in this sense of Norman Wachtel’s book, The Vision of the Vanquished. In this global study of the representation that the Indians of Peru have of the Spanish conquest, a study conducted according to the norms of historical and ethnographic description, the second chapter, “The Dance of the Conquest”, presents an analysis of a manifestation of Peruvian folklore, still alive today, which celebrates the death of Atahualpa. This analysis, conducted strictly according to your method of binary oppositions and by comparison of narrative variants, allows to put into perspective, by what it reveals of the contents of indigenous thought, the elements of the historical investigation.
CLS: You are quite right to cite Wachtel’s beautiful book which, from a methodological point of view – and I should have referred to it earlier – is halfway between Dumézil’s enterprise and mine. To return to your initial considerations, I quite agree that the pluralization of levels of analysis can compensate, to a certain extent, for the excessive (and often illusory) familiarity we feel with the most superficial level or levels. Why is that? Because the others normally escape conscious apprehension and it is therefore with respect to them that we find again this distance which constrains us to perceive only essential properties. Ethnology always consists of studying man from “the other end”, and it is by studying objects from the other end – the one about which we have no preconceived ideas because, most often, we didn’t even know it existed – that we can best preserve the benefit of the method. Now, don’t forget that even in the observer’s society, there are “other languages”, such as music, with respect to which we are not, at the outset, in very different conditions from those in which we find ourselves for the study of exotic mythologies.
RB: You pose, as a “working hypothesis”, in the “Finale” of The Naked Man, that “the field of structural studies includes four families of major occupants, which are mathematical beings, natural languages, musical works and myths”. This choice clearly manifests an exclusion: that of all the arts of figuration, from painting to cinema. Could it be that image, figure lends itself more difficultly to structural reduction, for example at the level of tonalities, of colors which bring into play intensities difficult to order in discrete elements? Wouldn’t this be recognizing on the other hand a limitation to the very principle of structural analysis, of the same order as that assigned to it by Jean-François Lyotard when, bringing into play the discourse/figure opposition, he detects at the principle of any artistic production an irreducibility to systematization of a linguistic order?
CLS: I said “major occupants”, but they are certainly not the only ones. And it seems to me that I have just partially answered your question: these are the domains about which it is most difficult to “make sentences”. Not impossible, certainly. Many sentences have been made about music, but their inadequacy to the object leaps to the eye, it seems to me, in a more immediate and more obvious way than in the case of the sentences – often inadequate, too – that have been made, let’s say, about painting. I see nothing comparable in musicology to the work of a Panofsky or even his predecessors. Which doesn’t prevent Panofsky from being a great thinker of structuralism, and therefore that plastic and graphic arts are also within the purview of structural studies. If, however, from this point of view, I don’t place them exactly at the same rank, it’s for reasons that you will excuse me from repeating since they are developed in the “Overture” of The Raw and the Cooked. You certainly put your finger on an additional difficulty by emphasizing that the different modes of aesthetic – or other – expression are unequally “codable” by means of discrete units. But I’ve already said it – structural analysis is by no means prisoner of binary oppositions. It can also resort to analogical models, that is to say where the relations between the elements vary more or less, in direct or inverse ratio to each other.
RB: You have often insisted, and you do so again here, on this pregnancy of detail whose systematic highlighting constitutes the major benefit of structural analysis, insofar as it allows explaining what had hitherto remained inexplicable by ordering a multiplicity of apparently disparate elements within the framework of a logic. This is why your method of myth analysis has thus found itself, for you as for many others, to be generalizable well beyond its object: hence these “outings” that you practice, prudent no doubt, limited, but which possess a considerable value of provocation and model. This is a situation to which only psychoanalysis is comparable: Freud’s work is entirely carried by the all-powerful determination of detail, and by this capacity to extend, from its points of origin, to the multiplicity of imaginary productions. There is a community of movement here that pushes all the more to question the increasingly total silence that you observe towards psychoanalysis. For if, originally, you receive from it a theoretical inspiration in the general form of the idea of the unconscious, if it then serves as a comparative and often contrastive reference for some of your studies in Structural Anthropology and Totemism Today, it is, if we except a few passing remarks at the end of The Naked Man, totally absent from Mythologiques, that is, in the very space where this pregnancy of detail triumphs to the point of exhaustion. (At most, if I’m not mistaken, it makes an ironic appearance in From Honey to Ashes, where one of the analyzed myths is found to perform the “first psychoanalytic cure in history”.)
I was very struck in this sense by the few remarks you shared with Barthes after your reading of S/Z (cf. here, p. 484). You articulate several codes (sex, name, kinship), operating a partial overlap between myth and text that you situate elsewhere in these terms in the final pages of The Naked Man:
“The difference between individual works and myths recognized as such by a community is not of nature, but of degree. In this regard, structural analysis can legitimately apply to myths from the collective tradition and to works by a single author.”
We find ourselves here, in fact, faced with a set of paradoxical relationships.
Barthes’ analysis, indeed, which has as its central object to exhaust the detail of the text, is largely situated in a psychoanalytic perspective, on the one hand classifying the determining effects according to which the code is structured within what he calls the “symbolic code”, on the other hand insistently supporting his theory of writing on Freudian and Lacanian terminology. On your side, your beginning of analysis of Balzac’s short story which finds itself, not without irony, bringing to light the incest taboo (that is, the usual fodder of psychoanalysts) shows well how a text can be illuminated without your approach putting into play at all what governs the Freudian approach, namely the production of desire and repression. How do you conceive, if you think it conceivable, a possible articulation between structural analysis and the Freudian perspective? Can’t we imagine that by a very tempting inversion effect, psychoanalytic reading can also contribute to illuminating myths? I’m not thinking so much there of the semantics of the statement itself, which is the level where you have chosen to situate your analysis: your criticisms like those of Vernant have rightly insisted on the often hasty reduction to which psychoanalysts subject cultural objects of societies too different from ours. But insofar as psychoanalysis essentially has as its object the relationship of enunciation and statement, doesn’t it seem to you that the Freudian perspective enters fully into any theory and any analysis of myth as a global object, that is to say, of individualized, historicized myth, if one can say so, by its conditions of production and its expressive thickness, in the sense we were evoking earlier?
Otherwise, what status should be given to the Freudian discovery, that is, to the emergence of desire within any discourse?
CLS: I retained from Freud much more than the idea of the unconscious: first the confirmation, as I had already learned from Marx, that the essential practical function of consciousness is to lie to itself; then and above all that, behind the apparent arbitrariness and irrationality of certain constructions of the mind, it is possible to discover a meaning. These are intellectualist and rationalist lessons where desire, which you speak of at the end, doesn’t have much to do with it. Besides, one can speak of an individual’s desire, not of the desire of a society, of institutions or customs, and, for the ethnologist who studies collective beings, the notion seems devoid of operational value.
See, for example, Róheim’s work, for which I have, moreover, great respect (because being a psychoanalyst allowed Róheim, in the field, to ask properly ethnographic questions but which, because of their different training, ethnographers probably wouldn’t have thought of). It happens that Róheim, reflecting on the Yurok (a small coastal population of California), encountered on his way their version of the myth that served me as a reference throughout Mythologiques. Now this myth, he believes he can interpret it entirely in terms of what he considers to be the specific characteristics of the Yurok psyche, and not only – which could have been legitimate – the differences that this version presents compared to others. But, at no point, does Róheim care about being in the presence of a pan-American myth whose particular version to the Yurok can only acquire meaning through differential contrast with other versions. He seeks to reduce a mythical ensemble whose area of diffusion is enormous to a narrowly localized psychic ensemble – a typically reductionist attitude – whereas one can only hope to reduce to each other the differences that manifest themselves between the versions and which, each on its own account, brings into play the entire techno-economic infrastructure of each society. But no: one possesses or believes to possess a key, and it is understood in advance that this key must open all locks. Yet in fact it opens none, at least not by itself.
There is, at the outset, something in common between the ethnologist and the psychoanalyst – which allows Devereux to speak of their complementarity: both seek, in the presence of a society or an individual, to decipher a message that only reaches them coded; and they can only succeed by relating this message and its code to a context which is, in one case, the ethnography of the society considered and, in the other case, the history of the individual studied. But, precisely, this context is not the same. Moreover, the psychoanalyst would be wrong if he thought he could ignore the ethnographer’s context, because each individual is, first, a member of a social environment that belongs to a given society; while the reverse is not true of the ethnographer unless he specializes in the study of the particular problem posed by the relationship of the individual to the group in a determined society. But, as far as the analysis of myths is concerned, these only exist as such insofar as they are collectively assumed by the group, and transmitted by it as it received them, as messages coming from elsewhere and in which the particularities introduced by successive individual narrators have gradually eroded.
I am therefore willing to admit that psychoanalysts and we do, in the purely intellectual part of our work and when it’s only a matter of understanding, the same thing: but we don’t study the same beings or rather (because I’m not unaware that societies are composed of individuals) in a very complex being, we don’t focus on the same network of articulations. Nothing would be lazier, at least for the ethnologist, than to let oneself go to supplying the gaps in one’s network by stuffing it with elements borrowed from another.
This doesn’t mean that considerations of a psychoanalytic type never intervene in our studies. One of my courses at the Collège, whose material didn’t find its place in Mythologiques but which I might publish one day, tended to show that certain South American myths make great use of categories, such as those of orality and anality, that psychoanalysis has only rediscovered. But in this case, we find that there is already psychoanalysis in myths, which is the very opposite of putting myths into psychoanalysis.
RB: Perhaps all these preceding questions can be reduced to a single question. Isn’t there, in general, a kind of contradiction, insofar as the human sciences cannot access the status of the natural sciences, in wanting to maintain them as you do in the perspective and model of a strict structuralism which seems to be able to constitute itself only at the price of more or less costly disjunctions practiced on the one hand between objects, in the broadest sense of the term, on the other hand within the objects themselves that you consider?
CLS: It would indeed appear according to some that we sacrifice the essential; that entirely occupied with inventorying and combining the bones of our skeletons, we forget that to make them live, the flesh of the muscles must move them and blood must circulate in them. I could say again: let each one do what suits their nature; there are physiologists and there are anatomists, one type of study does not condemn the other, they complement each other. But this would be neglecting a serious question; for this kind of criticism, which generally originates from philosophers, amounts to the reproach of limiting ourselves to those aspects of reality that we believe we can make a little more intelligible, instead of – as we are advised – piling up pell-mell what, in the current state of knowledge, we hope to understand, and what we perhaps feel, but don’t understand. It seems wiser to me to pose at each moment the problems that we can solve and, except for philosophers whose role it is, to leave the others aside for the time being. I don’t mean by this that philosophers don’t do us a service by shouting: attention, you don’t explain everything. But we are in our role by answering them that we explain what is knowable, that is to say what we can form other than confused ideas about.
RB: I wasn’t exactly targeting this kind of too general objections, you would say too “philosophical”. I wanted to return to and insist, by formulating them more globally, on the questions that preceded, in particular the one about psychoanalysis to which you seem to me to have only half answered. This, insofar as the cleavages that you practice find themselves specifically eluding the fundamental dimension of the unconscious as production of desire.
CLS: But is this the fundamental dimension of the unconscious? I’m not at all convinced of that. To limit myself to one example, as revolutionary as Freud’s treatment of dreams was, it seems to me that the general interpretation he gives of them – through the symbolic realization of a desire – remains singularly narrow; it probably applies to certain dreams, which are neither the richest nor the most interesting, and even there, I doubt it’s sufficient. What is, in my opinion, fundamental in dreams, is the highlighting of this essential tendency of the mind, even when left to its automatisms, to integrate heterogeneous data which, in the case considered, consist of bits of lived events, images, current organic sensations… This need, this requirement for integration are of an intellectual order much more than affective, and it would be muddying the waters to try to pass them under the rubric of desire, a notion which, I repeat, from the point of view of my research, is not operational. This exists, of course, but covers obscure forces of which we don’t even know if they are of a psychic or organic order and which, in both cases, come down to one and the same reality: the surge of life. Now, until further notice, we don’t know what life is, and if we are one day to know a little more about it, we will owe it to biologists, chemists, perhaps also physicists. I see no interest in disguising biological realities that elude us under psychological formulas that are sometimes beautiful, but always hollow: simple labels attached to the multiple ways in which consciousness reacts to forces that it undergoes without knowing or understanding them.
Our adversaries always return to romantic positions of which existentialism remains a faithful interpreter. They need to grasp the lived experience as it manifests itself, raw and bleeding, to the consciousness of Peter, Paul, Jack, Denis or Henriette… What happens in people’s consciousness is very interesting, but only insofar as by criticizing it, we can access the way things happen outside of it: not the subjective mush, but competing orders, of an organic, intellectual or social nature, of which affectivity only reverberates, on the level of individual consciousness, the shocks, conflicts or difficulties of adjustment.
Let’s take the example of marriage. The conscious ideology of certain societies can pretend that women are given without counterpart, bought or stolen. But, if the objective study of alliances reveals preferred degrees, networks tending to take on a certain form with, as a result, a more or less homogeneous distribution of women between lineages or families, we can conclude that what really functions in society is an exchange mechanism, and that because of a certain state of tension between the exchanging units or for any other cause, they conceal it from themselves by disguising it under fallacious covers.
RB: It’s this kind of determination that you’re aiming at, I suppose, with this assertion that sounds a bit like the ultimate term of the “Finale” of Mythologiques: “The fact of structure is primary.”
CLS: Not the ultimate term, because the “Finale” ends on a rather moralistic note, and you broaden the scope of an assertion that is placed at the beginning, in a discussion on the genesis of structures. But let’s take it in the sense you understand it. I’m going to perhaps appear scientistic to you: insofar as science exists and has a certain importance – not through its applications from which we are beginning to have less to expect than to fear, but because it makes us understand – it teaches us that the universe is composed of molecules and atoms, approximations no doubt, but of structural arrangements. What we call desire, production, drive, affect, what do I know more, is only the confused and obscure way in which we feel the effects of complex imbalances belatedly appeared with life itself, between these structural arrangements.
To say then that “the fact of structure is primary” will mean that life, and thought which is only a manifestation of it, relay, complicate and develop these structural arrangements that underlie them. From elementary particles to the genetic code, from the genetic code to language and operations of the understanding, we are confronted with structures. For, between thought and life, I believe that there can be no radical discontinuity. We are so constituted that we apprehend them as distinct orders, whereas it is rather the two ends of a chain whose intermediate links remain invisible to us because they weld behind our back. We contemplate the ends while ignoring where and how they meet.
All we can do is try to go back on each side a little further back and groping, to decrease the interval where the unknowable is established; what we will gain thus will only be in terms of structures, but without being able to conceal from ourselves that the point of junction – supposing it has a concrete reality – will always elude us. Limit of structuralism certainly, but which merges with the limit of all knowledge. For, in the much more restricted domains where we confine ourselves, we too can only claim to make a very small bit, each time, push back the frontier; that is to say, bring into the framework of a structural interpretation what remained below and so on, without it being possible or even conceivable – will you allow me to add: fortunately? – that the enterprise ever reaches its term.
Paris, summer 1972.
(1) By Georges Kukudjian, Le Magazine littéraire, Lévi-Strauss dossier, no. 58, November 1971.
(2) Annales, nos. 3-4, May-August 1971.
Notes on Structuralism & AI 