Structure: method or subversion of the social sciences?

It seems there’s no longer any real doubt as to the answer to this question, and that it is doubly negative. ‘Structuralism’, or what was designated as such mainly in France in the 1960s and 1970s (setting aside the question of other uses), is no longer regarded as a truly fertile method in the domains of sociology and anthropology, nor in those of linguistics and psychology, even if many of the concepts and schemata of thought that it put into circulation are still recognizable. A good portion of those who had enthusiastically adopted its language and objectives have turned away from it, in some instances towards methodologies that are more positivist, statistical or explicative, and in others towards participative inquiries, seeking a more immediate contact with ‘experience’. But one couldn’t really say it represented a ‘subversion’ either, since – apparently at least – the social and human sciences are holding up very well and still enjoy the same institutional legitimacy.

The moment has thus arrived to start singing the ‘swansong’, and probably to take note of a ‘collective shipwreck’, to use the expressions advanced by the sole work on the question available today, which is, truth be told, extremely mediocre. [1]

Our point of view will be different, for we figure that the question of identifying the exact content of the enterprise or intellectual adventure called ‘structuralism’ is still largely open to discussion and full of enigmas. But this question is itself indissociable from the question of knowing under what other forms and what other names the questions that gave rise to this enterprise are still posed today. We might even say it’s a question of knowing under what forms and names they can resurface, as soon as it becomes clear that the good health of the ‘social sciences’, the unity of their field, and the compatibility of their ‘methods’ are in truth extremely fragile. The problem of subversion thus presents itself again, but in a more ominous atmosphere, for we are no longer in a conjuncture of the ‘triumph’ of this project of scientific knowledge:

in many respects we risk entering into the ‘liquidation’ phase, to invoke a more or less expedient term of inventory.

The first condition for seriously discussing the lessons of structuralism is to realize that no unitary position was ever constituted under this name, not even in the sense of the extension of a model. Structuralism does not designate a school, then; it designates a movement, within a given intellectual conjuncture.

And what characterizes it above all, to borrow a key expression from Foucault, are its ‘points of heresy’. [2]

These appear to turn mainly on three large questions:

that of the constitution of the subject, that of the theoretical break or cut [coupure] of knowledge, and that of the universality of human nature. But before going further, we must first say a few words concerning what constitutes, prior to these divergences, the epistemological background common to the major ‘structuralist’ endeavours.


In spite of what has been put forth (and has been particularly supported by the received idea according to which structuralism has its roots in the generalization and exportation throughout the human sciences of a linguistic model, with the Saussurean theory of language and its typical dichotomies – synchrony/diachrony, language/speech, signifier/signified, and so on, – constituting in this respect at once the prototype of a structuralist approach and its logical organon), I don’t believe it is necessary to emphasize here before all else the primacy of the question of the sign and structures of signification. Or, rather, the importance of this question comes second, on the basis of an actively sought original solution to the methodological dilemma that was constitutive of the human sciences in the nineteenth and twentieth centuries, and that continues to accompany their institutional development.

Structure Method or subversion of the social sciences?

Étienne balibar

This dilemma presents itself as a conflict between an ‘explicative’ and a ‘comprehensive’ method, a ‘nomothetic’ and an ‘idiographic’ method, or a tradition that is naturalist, deterministic, and so on, and one that is hermeneutic. Ultimately, the structuralist programme coalesces around the project of overcoming the alternative between ‘individualistic’ analytical reductionism and ‘holistic’ organicism (to use the terminology eventually imposed by Anglo-Saxon efforts).

Approaching things from this point of view allows us to resituate the structuralist project in a double context. First of all, the philosophical context (which we will return to in our conclusion): one might suggest that the unilateral choice between a reductionist and an organicist perspective is typical of ‘scientific’ method ologies, whereas major philosophical projects are always concerned instead to overcome or relativize this abstract opposition. The fact that structuralism also sought exit from this dilemma is evidence of its philosophical dimension, but this does not mean that it was purely speculative. On the contrary, structuralism always sought to implement this overcoming in an immanent way, in the development of an effective knowledge that refers to objects.

Then there is a context that belongs specifically to the human sciences: what accounts for structuralism’s specificity must be grasped here through comparison with a certain number of previous and later endeavours, all of which nevertheless belong to the same grand historical conjuncture. Let us mention the effort to synthesize sociology and history developed by the Annales School and brought to its maximal conceptual precision by Fernand Braudel. Let us also mention the more recent pragmatic conception of ‘sociological reasoning’ as ‘natural reasoning’ developed by JeanClaude Passeron. [3]


As for Marxism’s contribution to this debate: as a discourse precisely situated at the intersection of philosophy and the social sciences (with or without reference to the ‘dialectic’), it has been and remains decisive, but according to very different orientations.

The conception of historical materialism proposed by Althusser around a notion of a structural causality that was specifically based on a distancing from both ‘mechanical causality’ and ‘expressive causality’ was in fact one of the driving elements of the structuralist movement as such (much more than Lévi-Strauss’s fairly conventional reference in his early texts to the determinant function in the last instance of material infrastructures and to Marx’s theory). The American ‘heretical’ Marxist Immanuel Wallerstein has been engaged in a totally different project since the 1970s, seeking to overcome the conflicts of method within nineteenth-century social science. In his analysis of the complementary aspects of the world economy, considered as a historical concrete system whose general laws reflect regulation in a given period, he has instead sought models on the side of systems theory and the ‘New Alliance’ theorized by Prigogine. [4] ‘Structure’, ‘system’ (words whose specific sense, if it could even be determined, is hardly important) are terms that are essential here in order to designate the mode of thought (and knowledge) of the totality that structuralism sought beyond these dilemmas. They have in common the emphasis on the relations to the detriment of the terms, or rather the postulation that the function and identification of the terms are entirely determined by the nature of the relations. Whence the affinity of structuralism with the mode of axiomatic thought in mathematics and the deductive sciences; for it is clearly necessary for the relations to be organized in a system, or considered as a totality, in order for this ultimate determination to be really thinkable. This is what allows structuralism to go beyond the theses of the positivist and critical traditions, which are also directed against ‘substantialism’. These put the emphasis either on the relation (as in the famous Comtean definition of law) or the function (as in Cassirer’s definition of the concept), but they leave the problem of the relation between the set and the individual totally unresolved.

Inversely, the structuralism that provides the means to characterize individuals as well as sets by a ‘second degree totality’ – the system of relations that assign them their respective places – must inevitably open onto a new, properly ontological dilemma, for which the two possible interpretations of the Althusserian notion of ‘support’ (Träger) provide a clear enough idea.Either the support is a singular existence constituted by the action of the structure – which determines all of its features, in other words, generates it – or, on the contrary (in the manner of the Lacanian ‘Real’), the support is an undetermined limit, whose singularity by definition exceeds all logical determination. We will return in a moment to this dilemma with regard to the constitution of the subject and the ‘point of heresy’ that it establishes in the structuralist tradition. It would not be difficult to show that it is exactly parallel to the one that opposes the syntactic-semantic orientation to the pragmatic orientation in contemporary logic.


The emphasis structuralism put on a new conception of totality, overcoming the aporias of the alternative between individual and set, corresponds to a specific conception of the correspondence between objects of knowledge and the construction of theories. No one has better isolated this conception in years than JeanClaude Milner. Having participated in its development, he ultimately extricated himself and countered with a return to a universalist ‘Galilean’ epistemology, based on a generalized conception of scientificity as the production of literal algorithms in the tradition of Chomsky and Lacan. [5]

What Milner has clearly shown is that the structuralist project, which is essentially anti-reductionist, tends to give substance to an ideal of science as being immanent to the domain of relations forming its ‘object’. As such, it is just as irreducible to an application or importation of concepts from physics or biology as it is to the transcendence of the mind. For this immanence to be so, it is necessary to establish a ‘natural’ correspondence between the domain (or the field of objects: what Althusser had called the ‘continent’), the concepts whose specificity gives rise to a ‘problematic’ (or a mode of determined ‘constitution’), and ultimately the procedures of verification and demonstration. We are thus at the opposite side of the idea of a mathesis universalis, in what Milner calls an epistemological ‘Aristotelianism’ (and whose slogan could in fact be the prohibition on metabasis eis al o genos, reiterated by Husserl at the beginning of the Logical Investigations).

Here we find explained the structuralist movement’s propensity to pass from a formalist to a historical epistemology, centred on the question of the formation of concepts proper to each science and on the search for the inaugural ‘break’ for new domains of scientificity.

That its major preoccupation has always been to trace the ‘frontiers’ of the domain of objects that correspond to science, or to locate the ‘mode of constitution’ of the objects of knowledge that renders them accessible to the concept, is also made clear. Finally, it becomes clear why it exercised a particular appeal for researchers in different disciplines, from linguistics to anthropology to psychoanalysis, who were concerned above all to escape from both the technicism of formal ‘models’ and the humanist litanies of consciousness [litanies humanistes de la conscience].6

On this basis we can thus situate the main ‘points of heresy’ or dilemmas of structuralism. It ought to be clear that none among them has yielded (or could yield) a definitive decision. And thus we understand how the structuralist movement had for its ‘end’ the yielding, in the best of cases, to diverse ‘post-structuralisms’: an ‘end’ not in the sense of an exhaustion of questions, or an avowal of impotence, but of an inevitable displacement.


The first dilemma is the one that concerns the constitution of the subject. After the somewhat simplistic debates over the opposition between the point of view of the structure and that of the subject (or over the misrecognition of the subject by the supporters [tenants] of the structure), which have certainly been sustained by adversaries and epigones, it is time to realize that all the major representatives of structuralism, in their respective domains, have concentrated their efforts precisely on the question of the subject with the intention of removing this notion from its transcendental indetermination. In so doing, they make it shift from a constituent to a constituted function, or consider it as an effect.

All the structuralists in this sense consider the subject to be ‘produced’, or rather, that there exist modes of production for the subjectivity effect. It is this common inspiration or problematic that allows us to understand, in this instance, the anthropological dimension present in all structuralist enterprises, at least if one admits that the proper object of anthropology is precisely the study of differential modes of subjectivity and forms of individual or collective (in fact, more fundamentally, transindividual) experience that correspond to them in the history of humanity.

Such a programme can seek theoretical precursors from various directions, for example in the classical theories of the passions and the imagination (Malebranche, Spinoza, Hume), especially since they underscore the dimension of misrecognition indissociable from any structure of the constitution of the subject. Or in the Marxian analysis of ‘commodity fetishism’, as it makes evident the forms of subjectivity (perception of the world and of the other) implicated in the very objectivity of value and market exchange. [7]

But it so happens that French structuralism drew the formulation of its problems essentially from the set of questions bequeathed and suggested by the work of Mauss, Lévi-Strauss’s presentation of which (as Claude Imbert never ceases to remind us) truly constitutes the key moment of ‘crystallization’ for the programme as a whole.

For the attempt to find in the very form of social relations, or their specific ‘logic’, the explanation for the conduct, strategies and modes of representation of oneself and others that form the secret of subjectivity effectively comes from Mauss, and more specifically The Gift. In this regard, Mauss completely reworked the Durkheimian legacy, moving past the juxtaposition of a naturalism of the social organism (or of the division of labour) and of a moralism or normativism of the ‘constraint’ society exercises over individuals.

He discovered in the constitution of the symbolic body (at once symbolizing and symbolized) the very point of indistinction between the individual and the set, between individual initiative and the transindividual unconscious. [8] From here, the alternatives play out.

They basically bear on two points. One concerns the relation between modes of subjectivation [subjectivation] and structures of subjection [assujettisement]. This ‘play on words’ that runs across the whole Western tradition has a particularly strong resonance in French, and this is perhaps what explains the fact that French structuralism wound up with it at the heart of its internal conflicts (between Lacan and Foucault,

Lacan and Althusser, Althusser and Foucault: all so many ways of thinking this articulation, by privileging one term or the other).

The other point concerns the relation of subjectivation to individuation, and leads to the confrontation between a conception of the individual subject that makes of it the ‘synthesis of structural determinations’, interiorized in a corporeal habitus or in a determined ideological position, and a conception of a subject that makes of it the lack, the ‘void’ abstractly common to all structures and consequently on the underside [endeçà] of determined forms of individuality, testifying to the impossibility of there being any subject that ever coincides perfectly with itself. Bourdieu is on one side here, Lacan on the other, along with all those for whom structuralism was generally an attempt to think, according to the analysis of Gilles Deleuze, the ‘flaw or default [défaut] of structures’, rather than their completeness and their efficacy. [9] One could say that Althusser, for his part, never ceased to oscillate between the different possibilities according to essentially political criteria.


The second dilemma can be presented as an antithesis within the interpretation of the idea of the theoretical break, which is itself constitutive of the idea of the relation between the subject and the forms of knowledge from the moment that the subject is precisely no longer constituent, but must be thought rather as an ‘effect’. How will its reinscription in the field of knowledge (or ‘theoretical practice’), for which it no longer constitutes the presupposed a priori, shielded from all contradiction by its universality and its simplicity, be effected? And how will this reinscription be compatible, not with a pure and simple relativism, but with a reworking of the ideal of universality?

I’ll leave aside here the ‘solution’ Foucault proposed in 1966 in The Order of Things, regardless of its own philosophical interest (the ‘disappearance’ of the subject of knowledge in the period between successive ‘epistemes’, which resonates at least formally with a certain Heideggerian idea of the history of being, but then the resurgence of the practical subject, or of a subjective engagement precisely in the form of ‘heretical choices’ that internally divide each episteme). Rather than constituting a problematic for the human sciences, this solution is basically a meta-discourse.

On the other hand, we can compare the two strategies respectively signalled by the Althusserian expression of the ‘epistemological break’ and the LéviStraussian expression of the ‘view from afar’ (the title given to the third volume of Structural Anthropology).

They are opposed term for term, departing from the formally common necessity for a deconstruction of evidence or a breaking of the ‘hermeneutic circle’ in which the subject of knowledge never faces the object except by having at its disposal a ‘pre-comprehension’ of its signification in advance. With the object in this case being human behaviour, or the social relation, it is important for the structuralist project to institute the conditions of a radical alterity in the place of all preestablished complicity, and yet also to transform this alterity into the very condition of conceptualization.

From the perspective of the epistemological break, this alterity is provided by the very development of the concept. It is thus essentially intellectual, even if it is later revealed to be the tributary of historical conditions that are much more material (in particular, a certain ‘position’ taken in the class struggle, and more generally in the conflicts between the ‘dominant’ and the ‘dominated’). From the perspective of the view from afar, theorized and practised by Lévi-Strauss, this alterity is provided by the cultural decentring of the observer, redoubled in his ‘self-consciousness’, wherein it takes the form of a conflict between two orientations that are both necessary. The participation and the retreat confront one another par excellence in the interpretation of the limits between the sacred and the profane, the normal and the pathological, and the violence and the institutions that belong to each ‘culture’.


To conclude, this opposition allows us to open on to a third ‘point of heresy’. We can fully illustrate it with the help of contradictory tendencies found in the work of Lévi-Strauss and the uses to which they can be put.

It is obviously an error to think that structuralism did nothing other than rediscover essentialist ideological themes concerning human nature in modern terminology, as various adherents to existentialism and the dialectic have rashly accused it of doing. It does not follow, however, that structuralism has nothing in common with the questions that proceed from the ‘loss’ of this ‘paradigm’ (as Edgar Morin would say).

On the contrary, the most interesting effect of its anthropological engagement is precisely that it brings about multiple theoretical possibilities in this regard, all of which concern the search for an elaboration of the concrete status of the universal as the correlate of the idea of the human race [en tant qu’il constitue le corrélat de l’idée d’espèce humain].

To some extent, the origins of this interrogation are located elsewhere in one of the major ‘practical’, but not institutional or technocratic, questions to which we owe the development of the human sciences in the second half of the twentieth century: the question of racism and the arguments one can oppose to it after the abjection of Nazism and colonialism became clear.

Faced with this, one of the possibilities leads not so much towards the biological naturalization of universals as towards a cognitivist program; that is, an interpretation of different ‘perceptions’ of the human as differing learning processes inscribed in the very constitution of the brain. The other possibility, which is also at least latently present in Lévi-Strauss (in his writings on history and those close to Devereux’s, from the pathological of one civilization to another), [10] seeks to think the universal not so much as a power or plasticity realized in differential learning experiences, but as the system itself, constituted by the double diversity of ‘cultures’ and individual ‘characters’.

Here we are more in the tradition of Fourier than of Helvetius, and it is not an accident that such considerations can appear particularly fruitful in a moment when we are faced with the problem of the alternative between a recrudescence of the war of the races, transmuted in exclusion, and the constitution of a new integral differentialism that can be drawn from the structuralist enterprise. This would be a differentialism that is not content to oppose historically collective sets like so many separate incommunicable ‘universes’ (which is basically just a resurrection of the old Humboldtian thematic), but which shows each individuality traversed, at least virtually, and constituted by the relation of all the forms of subjectivity (and their nonrelation: difference or irreducible alterity). This in no way suffices to practically assemble the subjectivities or identities in global space. But this is probably one of the only ways to ground its possibility apart from pious humanist wishes and litanies of respect for the Other.


Three points of heresy, then, but, as is immediately clear, three prospects as well, for the opening and renewal of debates that have lost nothing of their actuality. The question of knowing in which ‘domain’ of the division of intellectual labour these debates are playing out remains to be asked. At a certain level this question is futile; it only concerns the classification of ‘disciplines’ in the architecture of teaching establishments and in the structures of finance and power in the CNRS. But, on the other hand, it requires the following response: all of these questions are fundamentally philosophical. And this is how I would gladly put an end to the dilemma that I proposed myself at the outset: the importance of structuralism does not come so much from its having provided a ‘method’ for the social sciences or what in it allows us to ‘subvert’ their epistemological status, as from the manner in which it has reinscribed its problems in philosophy, contributing once again to the latter’s regeneration out of its other.

Translated by knox peden

The images in this piece are installation shots of Tehchning Hsieh, One Year Performance, 1980–1981, at FACT (Creative Foundation for Art and Technology), Liverpool Biennial,

November 2010.


This article has its origins in a talk at the Journées d’étude of the URA 1394 for Political, Economic, and Social Philosophy, CNRS–Université de Paris X–Nanterre, 6–7 April 1995; published as ‘Normes de scientificité et objet des sciences sociales’, in Tony Andréani and Menahem Rosen, eds, Structure, système, champ et théorie du sujet, L’Harmattan, Paris, 1997.

It is presented here as a basis from which to think the transdisciplinary dynamics of the concept of structure.

1. ^ François Dosse, Histoire du structuralisme, 2 vols, Editions La Decouverte, Paris, 1991 and 1992; republished Librairie Garnier–Flammarion, Paris, 1995. François Dosse, The History of Structuralism, 2 vols, Volume 1: The Rising Sign, 1945–1966, and Volume 2: The Sign Sets, 1967–present, trans. Deborah Glassman, University of Minnesota Press, Minneapolis, 1997. [Balibar’s reference to the ‘swansong’ alludes to Dosse’s Baudelairean titles for the two volumes in French; the title of the second volume, le chant du cygne, forms a homophonic pun with the title of the first, le champ du signe, the field of the sign. Trans.] 2 Michel Foucault, Les mots et les choses. Une archéologie des sciences humaines, Gallimard, Paris, 1966; The Order of Things, Random House, New York, 1970.

3. ^ Jean-Claude Passeron, Le raisonnement sociologique. L’espace non-poppérien du raisonnement naturel, Nathan, Paris, 1991.

4. ^ Immanuel Wallerstein, Impenser la science sociale. Pour sortir du XIXe siècle, PUF, Paris, 1995, translation of Unthinking Social Science: The Limits of NineteenthCentury Paradigms, Polity Press, Cambridge, 1991.

5. ^ Cf. in particular Jean-Claude Milner, L’Amour de la langue, Seuil, Paris, 1978; Verdier, 2009; For the Love of Language, trans. Ann Banfield. Palgrave Macmillan,

Basingstoke, 1990; L’Œuvre claire: Lacan, la science, la philosophie, Seuil, Paris, 1995; and the text on the Saussurean conception of the sign, unfortunately difficult to find today, ‘Retour à Saussure’, Lettres sur tous les sujets no 12, April 1994, ed. Le Perroquet, now available in Le Périple structurale: figures et paradigmes, 2nd edn, Verdier, Paris, 2008.

6. ^ The question had been posed with an exceptional clarity in the major book of G.G. Granger, Pensée formel e et sciences de l’homme, Aubier, Paris, 1960.

7. ^ On the anthropological dimension of Marx’s analyses of ‘commodity fetishism’, cf. among others, Jean-Joseph Goux, Freud, Marx, Economie et symbolique, Seuil,

Paris, 1973; Les iconoclastes, Seuil, Paris, 1978; and Alfonso M. Iacono, Le fétichisme: Histoire d’un concept, PUF Philosophies, Paris, 1992.

8. ^ The major function of Mauss’s œuvre is henceforth perfectly elucidated due to the work of Bruno Karsenti:

first in his short book, Marcel Mauss: Le fait social total, PUF Philosophies, Paris, 1994; and then above all in his doctoral thesis (Université de Lille III, 1996): L’Homme total: Sociologie, anthropologie et philosophie dans l’œuvre de Marcel Mauss, PUF, Paris, 1997.

9. ^ Gilles Deleuze, ‘How Do We Recognize Structuralism?’ (1972), in Desert Islands and Other Texts, 1954–1974, trans. Mike Taormina. Semiotext(e), New York, 2003, pp. 170–92.

10. ^ Cf. Claude Lévi-Strauss, ‘Cosmopolitanism and Schizophrenia’ (1983), in The View from Afar, trans. Joachim Neugroschel and Phoebe Hoss, University of Chicago Press, Chicago, 1992, pp. 177–85.