Ricoeur Confronts Levi Staruss On Structuralism
Ricoeur Confronts Levi Staruss On Structuralism
Ricoeur Confronts Levi Staruss On Structuralism
A Confrontation
Paul Ricoeur: The methodological questions I should like to ask you are of three kinds; all
three concern the possibility of co-ordinating your scientific method—structuralism as a
science—with other modes of comprehension which are not built on a generalized linguistic
model, but consist of a recovery of meaning in reflective or speculative thought, in short,
what I have myself called a hermeneutic. The first question concerns the intransigence of the
method—its compatibility or incompatibility with other modes of understanding. This
methodological question is directly inspired by a meditation on the particular examples you
use in The Savage Mind1. I wonder to what extent your method’s success has not been
facilitated by the geographical and cultural zone to which it has been applied, i.e. the zone
of what used to be called totemism, of the ‘totemic illusion’, which is precisely characterized
by the extraordinary exuberance of its syntactic arrangements, and perhaps in compensa-
tion by the great poverty of its content; is it not this contrast which explains why structural-
ism has such an easy victory, in the sense that its victory leaves almost nothing behind?
57
My second question, then, is to ask whether there is a unity of mythical thought—
if there might not be other formulations of mythical thought which would lend
themselves much less easily to structuralism?
This doubt leads me to my third question: what becomes of the structure-event and
synchrony-diachrony relations as functions of other models? In a system in which
synchrony is more intelligible, diachrony seems to be a perturbation, a fragility; I
am thinking of the Boasian formulation you like to quote on the demolition of
mythical universes which collapse as soon as they have been constituted, because
their solidity is momentary— in some way, it only exists in the synchrony. If we
consider thought organizations falling within the scope of a tradition-event
relation rather than a diachrony-synchrony relation, we get a quite different
result. This third question is related to the problem of historicity which is the
object of your dispute with Jean-Paul Sartre at the end of the book.
In our study circle we also embarked on a discussion of the philosophy implicit in
your method, but we did not linger over it, for we thought it unfair to your work
to devote ourselves primarily to this question; for my part, I think it essential
not to pass on too quickly to a discussion of structuralist philosophy, for to do
so would make it impossible to devote enough time to the structural method. I
therefore suggested that we should save till the end a discussion of the different
philosophical possibilities that you have yourself combined in what seems to me to
be a hesitant way: is it a renewal of dialectical philosophy, or on the contrary a
kind of generalized combinatory, or finally even, as you put it, a materialism pure
and simple in which all the structures are natural structures’?
That is the scope of the questions I should like to ask you, but I leave it up to you
to take them in the order you prefer.
Claude Lévi-Strauss: A book is always something of a prematurely born
child and mine strike me as fairly repugnant creatures compared with
what I should have liked to have brought into the world, and ones
which I do not feel too proud of when they are exposed to other people’s
gaze; so I have not come here with the belligerent intention of making
an absolute defence of positions which I should be the first to admit
have their precarious side, as Paul Ricoeur has correctly pointed out.2
Allow me one preliminary observation. There has been some misunder-
standing, for which I am myself solely responsible, about the place this
book occupies in my work as a whole. In fact, it is not—to express it in
Ricoeur’s own terms—‘the last stage of a gradual process of generaliza-
tion’, ‘a terminal systematization’, ‘a terminal stage’. This might seem
to be the case, but my intentions were quite different. Just as Totemism3
was a preface to The Savage Mind as I explained, in the same way The
Savage Mind is a preface to a more important book; but because when I
was writing the former I was not sure that I would ever even begin the
1
Claude Lévi-Strauss: La Pensée Sauvage, Paris, 1962; translated as The Savage Mind,
London, 1966.
2
Lévi-Strauss is referring to Paul Ricoeur’s ‘Structure et Hermeneutique’, which
appeared in the same issue of Esprit as this interview, November, 1963, pp. 596–627,
and which had previously been communicated to him.
3
Claude Lévi-Strauss: Le Totémisme Aujourd’hui, Paris, 1962; translated as Totemism,
London, 1964.
58
latter, I preferred not to mention it so as to avoid the risk of having to
go back on my word later. So it should rather be seen as a kind of
pause in my thought, a sort of resting-place, a moment when I stopped
for breath and allowed myself to contemplate the surrounding land-
scape, precisely the landscape through which I shall not, cannot and
have no wish to travel: the philosophical landscape which I can see
from a distance but which I shall leave vague since it is not on my
itinerary.
Now, a pause between what and what? Between two stages in the same
undertaking, which might be defined as a kind of inventory of mental
constraints, an attempt to reduce the arbitrary to an order, so as to
discover a necessity immanent in the illusion of liberty. Thus, in The
Elementary Structures of Kinship4, I had chosen a domain which might
appear at first sight to be characterized by incoherence and contingency,
and which I tried to show could be reduced to a very small number of
significant propositions. However, this first experiment was insufficient,
because the constraints in the domain of kinship are not of a purely
internal order. By this, I mean that it is not certain that their origin is
drawn exclusively from the structure of the mind (esprit); they might
arise from the exigencies of social life, and the way the latter imposes
its own constraints on the exercise of thought.
The second stage, to be completely devoted to mythology, will be an
attempt to avoid this obstacle, since it seems to me that it is precisely
in the domain of mythology, where the mind seems most free to aban-
don itself to its own creative spontaneity, that it is significant to establish
its conformity to law. As far as kinship and marriage rules are concern-
ed, it might still be asked whether the constraints come from within or
without; this doubt is no longer possible as far as mythology is con-
cerned : if in this domain the mind is bound and determined in all its
operations, than a fortiori this must be the case everywhere.
Thus I am particularly grateful to Paul Ricoeur for stressing the fact
that there might be a relation between my undertaking and that of
Kantianism. In short, it is a matter of a transposition of Kantian inquiry
into the ethnological domain, with this difference, that instead of using
introspection or reflecting on the state of science in the particular
society in which the philosopher happens to be, I move to the limit: by
investigating what there might be in common between the humanity
which seems most remote from us and the way our own minds work;
by trying to disengage the basic and constraining properties of all mind,
wherever it be.
I wanted to get this clear first of all; I shall now pass on to Paul
Ricoeur’s first question, which seems to me to dominate his study: i.e.
the question as to whether mythology has a unique explanation.
59
with Ricoeur’s position, but from an ‘ultra’ of The Savage Mind, if I may
say so, who could well attack me for not having included within its
jurisdiction the Bible, the Hellenic tradition and a number of other
traditions as well. For either these bodies of work derive from mythical
thought, and if it is accepted that the method is valid for the analysis of
the latter it follows that it must also be valid for the former; or the
method is held to be inappropriate to these cases, thereby excluding
them from mythical thought. Then I should have been correct in
leaving them out.
5
E. R. Leach: ‘Lévi-Strauss in the Garden of Eden’, Transactions of the New York
Academy of Sciences, Series 2, 1961, pp. 386–96. See also his ‘Genesis as Myth’,
Discovery, May, 1962, pp. 30–5, and ‘The Legitimacy of Solomon’, European Journal of
Sociology, 7, 1966, pp. 58–101. Both essays are included in the volume Genesis as Myth,
London, 1969.
60
Even in the mythology to which my next book6 is almost completely
devoted, that of tropical America, I can see that there are heterogeneous
levels. So I have preferred to leave aside certain texts, at least for the
time being, because their internal organization seems to depend on
other principles: mixed with myths in South America there is an almost
novelistic literature which is perhaps open to structural analysis, but to
a refined and transformed structural analysis which I cannot embark on
for the moment.
So in this perspective a cautious attitude is called for: to embark only
on what seems likely to give worthwhile results; to leave the rest aside
for better times when the method has proved itself. This reserve seems
to me to be characteristic of any undertaking which claims to be scien-
tific. If the study of matter had started with a theory of crystallization,
many physicists could have said: that is not the only state of matter,
there are others you cannot deal with; to which the ancient crystallo-
graphers would no doubt have replied: yes, but these are the best or the
simplest properties, those offering a kind of short-cut to the structure;
because of this, we have set aside for the moment the question as to
whether the study of crystals explains the whole of matter, or if there
are really other things to consider.
Now for the philosophical objections which I shall deal with rapidly,
since Paul Ricoeur wants to leave them aside for the moment: he
stressed their ‘outline’ character, their hesitant side. I agree with him
completely. I have no wish to produce a philosophy, I have simply set
out to examine the philosophical implications of some aspects of my
work for my own personal profit. What I should like to say before
moving on is simply that where Paul Ricoeur sees two possibly con-
tradictory philosophies, one close to dialectical materialism, accepting
the primacy of praxis, and the other leaning towards materialism tout
court, I see rather two stages in a single reflection; but I only attach
secondary importance to this and I am quite ready to be taken to task
on this point by philosophers.
I also find myself in complete agreement with Paul Ricoeur when, no
doubt as a criticism, he defines my position as ‘a Kantianism without a
transcendental subject’. This deficiency arouses his reservations, but I
am quite happy to accept his formulation.
I come now to what seems to me to be Paul Ricoeur’s fundamental
objection, which he repeated just now, and which I have copied from
his essay in the following significant sentence: ‘It seems’, he says, ‘as if
one part of civilization, precisely the part which did not produce our
own culture, lends itself to the structural method better than any other.’
This raises a considerable problem. Are we dealing with an intrinsic
difference between two kinds of civilization, or simply with the relative
position of the observer, who cannot adopt the same perspectives vis-
à-vis his own civilization as would seem normal to him vis-à-vis a differ-
ent civilization? In other words, Ricoeur’s disquiet, his conviction that
6
Claude Lévi-Strauss: Mythologiques* Le cru et le cuit, Paris, 1964, Mythologiques **Du
miel aux cendres, Paris, 1966, and Mythologiques*** L’origine des manières de table,
Paris, 1968. The first volume has been translated as The Raw and the Cooked, Introduction
to a Science of Mythology, Vol. I, London, 1970.
61
if I wished to apply my method to the mythical texts of our own tradi-
tion (which, be it noted, I have carefully refrained from doing) I
would find that there was something left over, an irreducible residue
which it was impossible for me to absorb . . . As a member of my own
civilization who has internalized this mythical tradition, who has been
brought up on it, I willingly agree; but I wonder if some native sage,
who happened to read The Savage Mind and noted the way I treated his
own myths, could not, with reason, raise exactly the same objection.
When Ricoeur’s text opposes totemism and kerygmatism (a word with
whose meaning in contemporary philosophy and theology I am un-
familiar, but which from its etymology I take to imply the idea of a
promise, an announcement), I should like to ask him what could be
more ‘kerygmatic’ than the Australian totemic myths which are also
based on events: the appearance of the totemic ancestor at a certain
point in the territory, his peregrinations which have sanctified each
named place and which define for each native the motives for the
personal attachment which gives profound significance to the country,
and which are at the same time, on condition that the native remain
faithful to him, a promise of happiness, an assurance of safety, convic-
tion of re-incarnation? These profound certainties are found among all
those who have internalized their own myths, but they cannot be per-
ceived by those who study them from the outside and must be left on
one side by them. So much so, that this kind of bargain I am offered, of
a domain where structural analysis reigns supreme in exchange for
another where its powers are limited, well, I wonder, suppose I accepted
it, would it not lead, if not to a re-introduction of the traditional dis-
tinction between a primitive mentality and a civilized mentality, at
least to a more reduced, let us say miniature distinction between two
kinds of pensée sauvage: one falling entirely within the scope of structural
analysis, and another containing something else as well? I hesitate to
accept this bargain because it would give me more than I bargained for.
Perhaps I laid too little stress on this in my book: what I have tried to
define as pensée sauvage cannot as such be attributed to anything, be this
one part or type of civilization. It has no predicative character. Or
rather, what I mean by pensée sauvage is the system of postulates and
axioms required to establish a code which allows the least unfaithful
translation possible of ‘the other’s’ into ‘ours’ and vice versa, the set of
conditions in which we can best understand ourselves; certainly, there
is always something left over. Ultimately, pensée sauvage as I intend it is
merely the meeting-point achieved by an attempt at understanding, by
me putting myself in their place and by them being put by me in my
place. The circumlocutions best suited to an examination of its nature
appeal to the notions of geometric space, common denominator,
highest common factor, etc, and exlude the idea of anything belonging
intrinsically to one part of humanity, defining it absolutely. So, to sum
up, I feel that I am ultimately in total agreement with Paul Ricoeur,
except that the differential principle he postulates does not seem to me
to be in the thoughts themselves, but in the various situations which the
observer happens to be in vis-à-vis these thoughts.
Paul Ricoeur: But if I refer to your work, this permutation of observers does
not completely satisfy me; there are differences in the very object of study which
62
are not effected by a permutation of the roles of observer and observed. These are
the objective characteristics which ensured an optimal relation between diachrony
and synchrony in a cultural ensemble during the era of classical totemism. So it
cannot be the observer’s viewpoint which distinguishes one mythical ensemble from
another; they differ even from the same viewpoint; with the result that they all
fall within the scope of the structuralist approach, but with different degrees of
success; at the end of my essay I demonstrated the non-existence of the naturally
symbolic, that a symbolism only functions within an economy of thought, within a
structure; that is why hermeneutics will never be possible without structuralism.
The question which emerges is whether there do not exist gradients, or, if you
prefer, stages, corresponding to the prevalence of the synchronic over the dia-
chronic which condition the very existence of your profession as a structuralist. I
cannot believe that this is a question of the observer: temporality does not have
the same significance everywhere; precisely where you can say that synchrony is
strong and diachrony weak, this does not seem to me to be the product of our
position as observers, but results from a certain constitution of the ensemble you
are studying.
Claude Lévi-Strauss: That is quite correct. It can be explained by the fact
that you take the adjective ‘totemic’ to have a much wider meaning
than the one I give it. As an ethnologist, I use it in a narrow, technical
sense. Indeed I realise that throughout your article you assume a kind
of equation of pensée totémique and pensée sauvage. I think that the rela-
tion is quite different: totemism arises from pensée sauvage—as I have
strongly emphasized—but pensée sauvage extends far beyond the limits of
the religious and legal system which was (falsely) categorized within the
term ‘totemism’. Consequently when I refer to the ‘totemic void’ in
the great civilizations of Europe and Asia, I do not mean that the distinc-
tive features of pensée sauvage are not to be found there as well, in other
guises. The two problems cannot be raised on the same plane.
If the basis of your argument amounts to the claim that an objective
difference exists between our civilization and those of non-literate
peoples in that the former accept a historical dimension while the
latter reject it, we are in agreement, since I have myself strongly in-
sisted on the same point. But it seems to me that we are not talking
about the same history: the temporality you put forward as an intrinsic
property of certain forms of mythical thought is not necessarily a func-
tion of the objective historicity of our Western civilizations and of the
way they have ‘historicized’ their development. ‘Historicized’ myths
are well-known throughout the world; a particularly striking example
is the mythology of the Zuni indians of the South Western USA, which
has been ‘historicized’ (on the basis of material which is not so histori-
cized) by native theologians in a way comparable to that of other
theologians on the basis of the ancestral myths of Israel. As it appears
in your study, the difference does not seem to me to relate so much to
the existence of history in the mythology (for even the most ‘totemic’
Australian myths tell a story, they are in a temporality)—but to the fact
that this history is either closed in on itself, locked up by the myth, or
left open as a door into the future.
Paul Ricoeur: Do you think it is an accident that it was precisely the pre-
Hellenic, Indo-European and Semitic bases which made possible all the re-
63
interpretations which philosophers and theologians have given us? Surely this
must be related precisely to a richness in content which calls for a reflection on the
semantics itself, and no longer on the syntax. If we accept an ultimate unity in
the domain of myth, it must follow that other methods than your own could be
applied to totemism, that it is possible to consider what they say and not merely
the way they say it, that what they say is full of meaning, charged with latent
philosophies; that we can therefore expect totemism’s Hegel and Schelling.
Marc Gaboriau: There has been some discussion of history, of the ‘diachrony’;
I should like to ask a few questions on this subject, concerning more particu-
larly the problems of the ‘diachrony’. Why is it that a given society transforms
itself over time? In certain parts of your work—in particular in Structural
Anthropology7 and in your preface to Mauss’s Sociologie et Anthropologie8
—you insist on the fact that the factors behind transformation should not be
looked for in the social systems in isolation (the kinship system, mythology, etc ...)
but in the way they are superimposed on one another and articulated together. You
claim that the latter constitutes a series of factors which must be studied before
considering external influences. I should like you to throw more light on this first
series of factors: at the end of Structural Anthropology you introduce the
concept of a ‘structure of subordination’; but it seems to me that you use the
same term to mean two different things: on the one hand, social inequalities
64
(polygamy, privilege, etc), on the other, you seem sometimes to denote by this
term the superimposition of the different systems which constitute a society. Can
you make this point a little clearer?
Claude Lévi-Strauss: This is really two questions, is it not? First of all, a
general question. I must admit that I am unable to answer it. I think
that ethnology, sociology and the human sciences in general are unable
to answer it, because societies develop very largely through the action
of external factors which fall within the scope of history, not that of
structural analysis. So the construction of a theory of social develop-
ment demands the observation of a large number of societies which
have been immune from any external influence (and by external I
mean not merely the action of other societies, but that of biological and
other phenomena) which is obviously impossible. I often tell my
students that there would have been no Darwin if there had not pre-
viously been a Linnaeus; the problem of the evolution of species
could not have been posed without initially defining what is meant by
species and making a typology of them. Now we are far from possessing,
and perhaps will never possess, even a taxonomy of societies compar-
able with pre-Linnaean taxonomies, such as Tournefort’s. So I see this
as a question which—yes, we can speculate about, such speculation
is not futile—but which we can never say anything very serious
about.
As far as the other question is concerned, if there is an ambiguity in
my text—I must admit that it was a long time ago—I am sorry; it is a
matter of a translation from English, since it was originally in that
language. But I seen to remember limiting the expression ‘structures
of subordination’ by opposing it to ‘structures of communication’,
meaning thereby that there are in society two major structure types:
structures of communication which are reciprocal and structures of
subordination on the other hand which are univocal and not reversible.
It may be that there is somewhere an ambiguity between this particular
meaning and the one you have indicated, but it was not an intentional
ambiguity.
Marc Gaboriau: There is an ambiguity, especially is this text as compared with
others; notably the Mauss preface, where you attempt to explain the transform-
ations of societies by studying the articulation of the different systems. In particular,
you say that, in their very nature, these systems are never integrally translatable
one into another, and that therefore a society can never remain identical to
itself.
Claude Lévi-Strauss: Yes, what we are looking for, let us say, in a society
reduced to a certain number of structural agencies stacked one on top
of another or imbricated into one another, is the ways of establishing
the kinds of disequilibria which explain why a society would still
‘budge’, even if it were immune to external influence.
Mikel Dufrenne: I should like to return to the problem raised earlier of the
relations between syntax and semantics. I wonder if what you have just said
about the fact that as far as you are concerned meaning is always secondary to a
pre-mythical and non-significant given, is not to some extent in conflict with your
65
own analyses. For example, in the analysis of the Asdiwal myth9 in which you
show that ultimately, considering the behaviour of the Tsimshian, particularly
Tsimshian women, in respect to fish, man is identified with the fish, this suddenly
illuminates the rest of the myth. It is as if the previous analysis which revolved
around the oppositions: high-low, East-West, sea-mountain, etc, was in some
sense a preparation for this sort of ultimate revelation of meaning, and that
therefore, the meaning here is given differently through a kind of immediate act of
consciousness, and is not the result of a syntactical analysis. It may be true that in
mathematics, truly formal thought, semantics is always in some sense on a level
with syntax and subordinate to it, but in an analysis like the one above, or your
analysis of the Oedipus myth which suddenly shows that the lame Oedipus means
something in itself, i.e. one form of birth opposed to another, surely there is a
kind of revenge of semantics on syntax, an immediacy of a meaning which has not
been logically produced or revealed?
Claude Lévi-Strauss: It is my impression that in the examples you refer
to the meaning is not directly perceived but deduced, a reconstruction
based on a syntactical analysis. If my memory is correct, the passage
from La Geste d’Asdiwal demonstrates that a certain syntactical relation-
ship is not reversible (as opposed to what happens in grammar, where
both ‘Peter kills the bull’ and ‘the bull kills Peter’, are permitted). The
fact that a proposition can only be formulated in one direction means
that certain hypotheses can be made about the hidden movement of the
native thought. But this is, after all, my own hypothetical statement, so
I feel that it is eminently a ‘reconstruction’. To satisfy Dufrenne and
Ricoeur, I must add that, naturally, I am not in the slightest excluding
—it would be impossible, anyway—the recovery of meaning to which
Paul Ricoeur has alluded; perhaps the difference between us lies in the
fact that I think it is a supplementary means which we can use in an
attempt at post factum control of the validity of our syntactical operations.
We work in the ‘human sciences’ and we are men studying men, so we
can allow ourselves the luxury of trying to put ourselves in their places.
But this is at the last moment, our last satisfaction: to ask the question:
does it work? if I try it out on myself, will it go? Consequently, I
feel that from a methodological point of view, the recovery of meaning
is secondary and derivative compared with the essential work which
consists of taking apart the mechanism of an objectified thought; here
I can do no better than to take up the terms of Ricoeur’s critique; for it
does not seem to me to be a critique; it is just exactly what I set out to do.
Paul Ricoeur: If the meaning which I have recovered in this way does not in-
crease the understanding I have of myself or of things, it is not worthy of the name of
meaning. But there can be nothing of this sort if the syntactical inquiry stand out
against a background of non-meaning; for what else do we understand by the very
words meaning and non-meaning if not the episodes of a consciousness of
history which is not simply one culture’s subjectivity looking at another culture,
but truly a stage in the reflection which is trying to understand everything? In
other words, it is the particular discourses which have a meaning, it is the things
said and not simply their syntactical arrangements by an outside observer. I
9 Claude Lévi-Strauss: ‘La geste d’Asdiwal’, École pratique des hautes études, Section des
sciences réligieuses, Extr. Annuaire 1958–1959, pp. 3–43; reprinted in Les temps modernes,
March, 1961; translated in The Structural Study of Myth and Totemism, ed. E. R. Leach,
ASA Monograph No. 5, London, 1967.
66
realize that it is essential for the scientist to restrict himself to the arrange-
ments of which he has set himself up as the observer; that is how he avoids what I
have called the ‘hermeneutic circle’ which makes me one of the historical sectors
of that very content which is interpreted through me; to be a human scientist, I must
be beyond my own reach; but can I still talk of meaning and non-meaning if this
meaning is not an episode in a fundamental reflection or a fundamental ontology
(without here choosing between the two great traditions, Kantian and Hegelian).
Claude Lévi-Strauss: It seems to me that you are linking the notion of
person with the notion of discourse. But what do the myths of a society
consist of? They make up the discourse of that society, and there is no
personal transmission of this discourse; so it is a discourse which can be
collected just as a linguist who goes off to study a little known language
can hope to construct its grammar without bothering to know who
said what was said.
Paul Ricoeur: But if I do not understand myself better by understanding them,
can I still talk of meaning? If meaning is not a sector of self-understanding, I do
not know what it is.
Claude Lévi-Strauss: I find it quite legitimate that a philosopher who
posed the problem in terms of the person should raise this objection,
but I am not obliged to follow suit. What do I understand by meaning?
A particular flavour perceived by a consciousness when it tastes a com-
bination of elements of which any one taken alone would not produce
a comparable flavour. So, just as a laboratory worker trying to make a
chemical compound has many means at his disposal to test his result—
he has his spectrograph, and his reagents, but he is not usually content
with these for he also knows that he has a tongue, so he tastes, recog-
nizes the characteristic flavour and says: yes, that’s it—the ethnologist
also tries to recover the meaning and complement his objective proofs
by intuition. He is a being endowed with sensitivity and intelligence, he
has these means as well. So he tries to reconstitute the meaning; he
reconstitutes it by mechanical means, he constructs it, unwraps it. And
then, after all, he is a man, so he tastes it.
Jean-Pierre Faye: I should like to ask a question about contemporary myths. It
concerns language zones in which it is history which has been mythologized. As
opposed to cases of historicized myths, these are mythologized ‘histories’ (histori-
cal interpretations). Take the case of the German nationalist ideologies of the
inter-war period: this seems to me to be a ‘privileged’ field to which your criteria
might be applied. We have here a kind of language halo with a strong biological
charge, very close to the forms of ancient mythology and in which history is
absolutely inverted in the myth. An attempt to map out these different languages
gives, on the one hand, a kind of topology in which these languages represent very
precise intersections. On the other hand, they can also be treated as transforma-
tions of meaning, and in this respect they present two remarkable features: each of
them will allow of an inverse transformation. On the other hand, the combination
or composition of two of them gives ‘something’ (a signification) which itself in
turn belongs to the ideological ensemble, no doubt because we are dealing with a
retrograde system of thought which is consequently closed in on itself. This
‘axiom of closure’ is not to be found in the case of other ideologies, eg. those of
liberals or of the Marxist left. In the case of the nationalist ideology which
67
decked out National Socialism in Weimar Germany and which referred to itself
as the ‘National Movement’, the kind of closure which would seem to lend itself to
structural analysis is really present, if the word structure is being used in its
algebraic sense: a particular set has a structure if it is endowed with a well-
defined ‘law of composition’.
Claude Lévi-Strauss: I completely agree with you in thinking that noth-
ing bears a closer resemblance—formally speaking—to the myths of
what we call exotic or non-literate societies than the political ideology
of our own societies. Any attempt at a transposition of the method
would no doubt have to start with political thought rather than relig-
ious tradition. But is it necessary to single out any particular kind of
political thought? I am very reluctant to do so; it would seem to me,
for example, that the ‘mythology’ of the French Revolution would
show similar ambiguities to the one which you have referred to. After
all, the term ‘sans-culotte’ has had a spectacular career while its primitive
meaning has probably been lost and the affinity with ‘culot’, ‘culotté’
(cheek, cheeky) may have played a larger part in its success. But when
all is said and done, we always come back to the same point. The ques-
tion is to know whether what we are trying to attain is what is true in
and of the consciousness we have of it or outside this consciousness. I
believe it is perfectly legitimate to look inside, by a recovery of mean-
ing, except that this recovery, this interpretation philosophers or his-
torians give of their own mythology, I treat simply as a variant of that
mythology itself. In my analysis it becomes matter, objectified thought
once again. In other words, I am not at all contemptuous of efforts such
as the one I only know from the résumé given by Ricoeur, but—from
what I know of it from his résumé—and if, God help me, I had to attack
this kind of problem, I should see in it a variant of Biblical mythology
and stack it on top of the latter rather than putting it after the latter.
Paul Ricoeur: I did not say that meaning was meaning in or of consciousness;
meaning is first of all what instructs consciousness: language is above all a
vehicle for the meaning to be recovered and this potentiality of meaning cannot be
reduced to my own consciousness. The choice is not between the subjectivism of an
immediate consciousness of meaning and the objectivity of a formalized meaning;
between the two there is what meaning proposes, what meaning says, and it is
this ‘to-be-said’ and ‘to-be-thought’ which seems to me to be the other side of
structuralism; and when I say the other side of structuralism, I am, perforce, not
indicating a subjectivism of meaning, but a dimension of meaning which is ob-
jective as well, but the objectivity only appears in the consciousness which recovers
it. This recovery expresses the extension of consciousness by meaning rather than
consciousness’s jurisdiction over meaning. That is why I am not opposing sub-
jectivity to structure, but precisely what I call the object of the hermeneutic, i.e.
the dimensions of meaning opened up by these successive recoveries; which then
raises the question: do all cultures offer as much to be recovered, re-said and re-
thought?
Claude Lévi-Straus: A moment ago I was on the point of talking of privi-
leged examples—I shall come back to Ricoeur’s point via this detour
10
On this term, and in particular on the use made of it by the philosopher Heidegger,
see J-P Faye’s ‘Heidegger et la Révolution’, Médiations, Autumn, 1961.
68
—but are they really privileged? They are too rich in material whose
abundance overwhelms us. We are in an eminently favourable situa-
tion as far as exotic societies are concerned precisely because we know
almost nothing about them and this poverty is in some sense our
strength: we cannot avoid the essential . . .
Claude Lévi-Strauss: So the sign has quite simply been transformed into a
symbol.
Jean-Pierre Faye: Yes, but it has lost the connections which linked it to the
initial symbol.
Jean-Pierre Faye: Yes, but the later symbol is somewhat factitious, there is
something fabricated about it, whereas in retrograde political mythologies there is
perhaps what might be called a recourse to the umbilical cord. The political
signs of the Left or of liberalism are more ‘semiological’ and less ‘symbolic’, are
in some sense on the road leading to a mode of thought of Kantian (or Durkheim-
ian) type, Kantian thought being itself, as an historical fact, a by-product of
ideological liberalism (of which it is by rights the philosophical under-pinning).
But if, on the contary, it is a question of types of political thought which are
themselves ‘savage’, if it is a matter of ideologies more directly at grips with
mythologies, perhaps it would then appear that their ‘pensée sauvage’ is more
savage than yours . . . That is to say, it retains more of the element of participa-
tion . . . By ‘participation’ I mean the kind of double action of the sign, which on
the one hand operates within a certain structural circle and, on the other, hangs
on to its connections with a ‘nature’ of the language. Obviously, this linguistic
nature raises problems. But Heidegger’s insistence on always going back to the
origin of the language is, it seems, a movement quite distinct from that of
structuralism, and one which does not seem baseless. For just at the moment when
he was the dupe of an ideological language, Heidegger discovered that he was in
some sense verifying his philosophy of language. . . .
Kostas Axelos: I should like to ask a question which troubles me a lot, especially
since reading The Savage Mind. There could be said to be two forms of
genealogical thought: one a naïve genealogical thought according to which things
follow one another from generation to generation in space-time, and the other a
speculative genealogical thought, for example Hegel’s, according to which there
69
is a genealogical development, a Phenomenology of Mind, this genealogical de-
velopment being none other than the development of an original and total structure,
that of the Greater Logic. In my humble opinion, Hegel is the father of structural-
ism, so to speak, being at the same time the first to have made such use of genetic
thought. In genealogy the logos aspect of genealogy must also be understood. You
have broken out of the limited framework of a primitive mentality on the one
hand and a civilized thought, which begins when one requires it to have begun, on
the other, and you talk of a global pensée sauvage. I must ask a naïve question
which worries me perhaps because of its naïvety: when did pensée sauvage
begin in space-time? From what point on is it possible to talk of thought?
Claude Lévi-Strauss: That is a big question, but I am not sure why I am
expected to be able to answer it, for it is the question of the problem of
the origins of humanity, of what the anthropologists call ‘hominization’.
Since when have there been thinking beings? I have no notion, and I
doubt whether my colleagues in physical anthropology have any clear
ideas on the subject. I would go further and doubt even whether
theoretically we will ever be able to pinpoint a moment in this develop-
ment when man began to think, and I rather tend to suggest that
thought began before man.
Jean Lautman: I should like to return once again to the question of meaning, for,
ultimately, if Lévi-Strauss’s work worries me at all, it is in some way because he
tells us that we express ourselves when we do not think we an expressing our-
selves. My question is in several parts.
First of all, in Structural Anthropology, where you show that the shaman’s
method compares structurally with a psycho-analytic cure, I sensed a kind of
ambiguity: on the one hand, an underlying criticism of psycho-analytic therapy as
nothing new since it is the shaman’s method, on the other hand, a valorization
which I find much more comprehensible now that you have given us The Savage
Mind, in so far as where you are concerned either of these liberating expressions
and human self-revelations are valuable. Do you accept the statement that you
have in some sense proposed an attempt at the constitution of a collective psycho-
analysis which does not approach the individual structures of Mr X or even the
set of psychological structures in a society, but, aiming higher, the organizational
schema of all society? If so, I can understand the interest in linguistics you share
with the modern French school of psycho-analysis, and for similar reasons:
Zipf’s Law, for example, shows that when we speak, and think we are speaking
freely, we are in fact governed by structures older than the emergence of meaning in
our own thought.
The second part of my question is about history: in the critical comments on
Jean-Paul Sartre’s work that you put forward at the end of The Savage Mind,
I shall pass over the points on which there is general agreement with you, to
discuss the aspect of history you criticize: the fact that it uses a very meagre code;
the essential of this coding system is chronology, and this is ultimately an im-
portant but limited area of knowledge. For you do say that history is important
nevertheless. But it seems to me that you see history as consisting usually of the
obscuring of meaning, and meaning, to the extent that it is important, as much
better expressed at the moment it springs forth from the structures of society in
their first crystallization than in the ongoing development which is imposed on
them.
70
For my last point, I was astonished by your affirmation in the last few pages of
The Savage Mind that the recent paths of science lead to a rapprochement
with the world of matter via communication. You show that this process is, in
fact, the very process of magic thought, which has always approached the paths of
nature via the modalities of interpretation; personally, however, I am reluctant to
think that the paths of contemporary science and magic practices can be re-
absorbed into the same set. You have certainly shown that there is a structured set
in both cases, but—and here I cannot agree with your quotation of Heyting in
the same chapter—the structured systems at work in the societies you study are
totally saturated structured systems, whereas the axiomatic systems of contem-
porary thought are basically unsaturated systems. It seems to me that this
opposition has a wider scope, but it would be rash of me to ask you to take it
further.
Claude Lévi-Strauss: What huge problems! The first, on psycho-analysis:
I am trying to make an analysis of meaning, but why call it a psycho-
analysis? You have shown, I think, that what is not conscious is more
important than what is conscious. I could be said to be trying, in my
own way—that is, as an ethnographer—to participate in a collective
undertaking in which the ethnographer has only a modest part to play.
This undertaking is to find out how the human mind functions. So it
probably parallels part—I repeat, part—of what psycho-analysts are
doing, for I distinguish between two aspects of psycho-analysis: the
theory of the mind worked out by Freud, which is based on a critique of
meaning (and here I have a feeling that the ethnologist does the same
thing for collective ensembles that the psycho-analyst does for indi-
viduals); and, on the other, what might be called a theory of the cure
which I leave completely alone. For I do not believe for one moment
that the self-analysis undertaken by the mind will improve it; so from
this angle it is not a psycho-analysis; and I am completely indifferent as
to whether it improves or no. What interests me is to find out how it
works and that is all. So much for the first point.
As for the second, I am afraid there must be some misunderstanding,
and this is not the first time I have come across it. Ultimately, what is in
the last chaper is no critique of history, in the sense that it was not I that
started it. I have nothing against history; I have the greatest respect for
it; I read the works of historians with infinite interest, even passion,
and I have always maintained that it is impossible to embark on any
structural analysis without having first obtained from history all it is able
to give us in illumination, which is unfortunately very little when we are
dealing with non-literate societies. I merely sought to redress, or at least
I rebelled against, what seems to me to be a very manifest tendency in
contemporary philosophical France to regard historical knowledge as a
kind of knowledge superior to all others. So I limited myself to the
statement that history was one kind of knowledge among others, that
there could be no knowledge of continuity, only of discontinuity, and
that history was no exception in this respect. So I do not claim that
history’s code is any more meagre than any other: this would obviously
be inaccurate; merely that it is a code and that therefore historical
knowledge suffers from the same weaknesses as all of knowledge, which
is not to say that it is not very important. I should like to suggest (with
out any ill-feeling) that you twist my position for your own purposes
71
when you attribute to me a certain tendency to think that men express
themselves better in their crystallized institutions than in their historical
development. This raises a major problem which we have touched on
several times and should have dealt with, and which we shall now deal
with thanks to you: that is, the problem of diachronic structures. After
all, the fact that events are situated in time is not sufficient to exclude
them from structural analysis; it merely makes the latter more difficult.
But the linguists’ position on this point is quite clear: they accept that
there is a diachronic linguistics as well as a synchronic linguistics, but
the former raises many problems, the principal one being the necessity
to begin by revealing recurrent sequences in a development which does
not always allow of an isolation of terms of comparison. Perhaps with
the assistance of sociology, ethnography, and who knows what else,
history will one day achieve this, but that day has not yet arrived.
Consequently, it is better to leave the problem of diachronic structures
aside for the moment, and devote ourselves to those aspects on which
we have a firm grip.
Now for the third point. I must admit that in the last few pages of The
Savage Mind there is a little rather false lyricism, and that I even allowed
myself to say a little more than I should have—I have already been
reproached for this by our colleagues in the exact and natural sciences.
Nevertheless, I do not think I ever suggested an equation of modern
scientific thought and magical thought. As you put it yourself, one is
saturated and the other non-saturated; I think that is what I wrote
myself, in almost those same words, in the first chapter of the book,
when I said that a sign is an operator for the reorganization of a set,
whereas a concept is an operator for the opening-up of a set. It is
obvious that if I had set out to establish an equation of modern science
and magic, I would have been laughed at to my face, and rightly so. All
I was trying to point out was that as it progresses, modern science is
rediscovering in and through itself a certain number of things which
allow it to pass a more tolerant judgment on magical thought than it
has ever done before.
Jean Cuisenier: It is certainly difficult to apply structural linguistics to the
diachrony. But there is one case where there has been for some time an attempt to
apply analogous methods to the diachrony, that is, in political economy. In this
domain an interest appeared and grew in the study of types of fluctuation, the
registrations of long periods, the delimitation of certain sequential forms, etc . . .
Indeed, we have a large quantity of high-quality statistical information on the
19th century and much work has been done to disengage the principal types of
fluctuation from this material by empirical means. So there is one case—probably
a privileged case—where structural analysis is typically applied to sequences and
in which it has indisputably achieved a certain success. It seems to me that this is
because economic events fall very largely outside the conscious and voluntary
control of the human subjects they affect. A comparison, for example, of the
phenomena of kinship with economic phenomena is a comparison of analogous
things, for they are both phenomena which are only comprehensible over long
periods of time, and also phenomena which men find particularly difficult to grasp
or interfere with voluntarily. Now these are precisely the cases where structural
analyses, diachronic as well as synchronic, have been most remarkably successful.
It is certainly not chance that economics has been able to develop structural
72
analysis to an extraordinary degree of refinement through techniques such as the
tableau économique, national accounting, input-output matrices, etc. The
success and refinement of the analysis when it is applied to kinship structures and
the structures of the economy is an epistemological fact which has something to
teach us.
Claude Lévi-Strauss: Yes, I agree that it has something to teach us, but
this lesson is not completely optimistic, to the extent that economic
phenomena are an exceptionally favourable example, in that, firstly, we
are examining a society in which they have played an essential part for a
long time; secondly, they also have a rapid rhythm, a low periodicity;
in a century or a century and a half many things have taken place and it
is possible to extract a large number of recurrences; finally, our capital-
ist societies are such that all these phenomena have been directly or
indirectly written down and collected in documents. Even in the case of
language (although diachronic linguistics can count some considerable
achievements to its name) it is more difficult, because there are so many
things in the evolution of language which have completely escaped us,
since they were not transcribed at the time they could be observed, and
there is now no trace of them left, or almost no trace. We are not
always so lucky as to find favourable phenomena.
Pierre Hadot: You have dedicated your book to Maurice Merleau-Ponty and it
has also been pointed out to me that the expression ‘esprit sauvage’ occurs in
Merleau-Ponty’s work. Is there some relationship between your thought and his?
We have already discussed this a little amongst ourselves this year.
Claude Lévi-Strauss: I should say that the relationship was certainly not
reciprocal, in that Merleau-Ponty, from his writings and what he said
to me in conversation, had a much stronger impression that what I was
doing derived from his philosophical work than I had of the possibility
of joining him; probably because of a certain, perhaps provisional,
incompatibility in the way ethnologists and philosophers pose their
problems. Paul Ricoeur has emphasized this several times, and with
reason. The philosopher makes a kind of all-or-nothing demand which
my emphasis is not intended to criticize. He is immediately attentive to
the extension of a position into other domains, insisting that its con-
sistency must be retained, and when he sees a weak point in this con-
sistency he raises a fundamental objection, whereas the ethnologist is
more careless as to the future. He tries to resolve one problem, then
another, then a third. If there is a contradiction between the philo-
sophical implications of the three attempts, he does not torment
himself about it, for as far as he is concerned, philosophical reflection is
a means rather than an end.
Jean Conilh: In your book you explain that Western thought has always been
drawn towards pensée sauvage. I wonder if the problem you have raised is not
the following: each time we attempt an interpretation of savages, is this not
always ultimately a way of finding a meaning for them so as to understand our-
selves? In the 18th century we find writers discussing the Good Savage in relation
to the questions they were asking about themselves. In the bourgeois colonialist
epoch we find a conception of the primitive which presents him as inferior (pre-
logkal). I find it very remarkable that today economists and even novelists also
73
talk of structuralism and recognize themselves in your work. In other words,
have you not constituted a philosophy, and a philosophy of our time ? But in that
case, I can reject this philosophy and go back to primitive mentality, reading it at
another level, the level of symbols for example, and find a new meaning for it. In
short, is our problem to classify or to find a meaning?
Claude Lévi-Strauss: To be sure, I think that one of the reasons for the
attraction ethnology exercises, even on non-professionals, is that its
inquiries have powerful motivations within the heart of our society,
integrating as they do a number of our society’s dramas. But a distinc-
tion must be made: after all, what motivated the constitution of astro-
nomy?—preoccupations of a theological kind, or even the desire to
draw up horoscopes ensuring the success in love and war of the powers
of this world. However, these are not the real reasons for its import-
ance, which depends on results whose interest lies on another plane. So
I do not think there is any contradiction in this double aspect. We ought
to recognize that, whether we are ethnologists or are merely interested
in ethnology, it is for scientifically impure reasons; nevertheless, if
ethnology is to deserve recognition someday for its role in the con-
stitution of the human sciences, it will be for other reasons.
Paul Ricoeur: Perhaps we can find common ground precisely in this field to
which your work has led you. Do you put your philosophy down to transient,
impure personal motivations or do you think that there is a structuralist philosophy
in solidarity with the structural method? In the first case your work is philosophic-
ally neutral; it leaves us the responsibility of choosing at our own cost and risk . . .
Paul Ricoeur: For myself, I think that this implicit philosophy affects your
work itself, in which I see an extreme form of modern agnosticism; as far as you
are concerned there is no ‘message’: not in the cybernetic, but in the kerygmatic
sense; you despair of meaning; but you console yourself with the thought that, if
men have nothing to say, at least they say it so well that their discourse in
amenable to structuralism. You retain meaning, but it is the meaning of non-
meaning, the admirable syntactical arrangement of a discourse which has nothing
to say. I see you as occupying this conjunction of agnosticism and a hyperintelli-
gence of syntax. Thereby you are at once fascinating and disquieting.
74