Transformation transformed

This work is licensed under the Creative Commons | © Philippe Descola. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau6.3.005


Transformation transformed

Philippe DESCOLA, Collège de France

The key methodological tool of Lévi-Straussian structural anthropology is the group of transformation. A structure only acquires an analytical dynamism thanks to its capacity to organize the transformations between the models of a same group of phenomena. The paper explores various approaches to this concept and builds on these results to examine the consequences of apprehending ontological pluralism as a group of transformation.

Keywords: structuralism, Lévi-Strauss, transformation, ontology, the nonhuman

I welcome the opportunity that the invitation to deliver this keynote lecture has offered me to reflect on my own relationship with structuralism, on the part it has played in the development of my work, and on the influence that one of its key figures, Claude Lévi-Strauss, has exerted upon me. Lévi-Strauss was my doctoral supervisor, an ideal one in that he left me entirely free to organize my work as I wished, to the point that he politely declined to read the manuscript of my dissertation—saying he had entire confidence in my abilities—before I officially registered it at the École des Hautes Études en Sciences Sociales (EHESS), so that I had no idea of the type of critical comments he would make during the formal defense. They mainly concerned my style, which he found too ornate. I had asked Lévi-Strauss to supervise my thesis for a variety of reasons. First, because reading Tristes tropiques when I was sixteen or seventeen was probably the initial shock that finally led me to become an anthropologist. However, I decided to read philosophy first, but I was among the few students at the École Normale Supérieure who did not limit himself to reading Lévi-Strauss’ more philosophical works, La pensée sauvage or Anthropologie structurale, but also studied the more technical books, such as Les structures élémentaires de la parenté and the first volumes of Mythologiques. I greatly admired him, and when I finally took it upon myself to ask his secretary for an appointment to expose my doctoral project I was as intimidated as if I had gone to visit Kant in Königsberg. Here at last was a huge figure of the social sciences, someone who was able to deal with arcane social facts by showing how they followed a complex logical pattern, and doing this with the clear and incisive mind of a great philosopher and the elegance and sensibility of a great writer. Lévi-Strauss was also at the time [34]the only person in Paris qualified to supervise doctorates on Amazonian Indians, and although he was very close to retirement, this was among the reasons why I asked him to be my supervisor.

However, both during the time I was struggling to write my dissertation on the relationship between an Amazonian people and its environment, and for a few years after, when I began teaching at the EHESS myself, I did not have the impression that I was a committed structuralist or that I was following a structuralist method. As was quite common at the time, my theoretical outlook was an improbable mixture of Gallic Marxism, semiotics, Husserlian phenomenology, and Lévi-Straussian structuralism, with small pinches of Sahlins, Douglas, Leach, and more exotic authors such as Georges-André Haudricourt, Gilbert Simondon, or Nicholas Georgescu-Roegen. Within this eclectic combination, I had the impression that structuralism was just a component and certainly not a driving force. And it was only when I began to give papers abroad, and when the audience would react to what I had just said by qualifying it as a structuralist outlook, that I realized how deeply I had been influenced by Lévi-Strauss and structuralism. I was not unlike Monsieur Jourdain in Molière’s Le bourgeois gentilhomme, who, upon learning the meaning of the word “prose,” exclaimed: “For more than forty years I have been speaking prose without knowing it!” Likewise, for several years I had been a structuralist without being fully aware of it. This lecture will thus have a rather personal tone, for which I hope to be forgiven, as I will endeavor to make amends for my initial blindness by trying to state some of the merits of structuralism and explaining how I came to use some of its basic methods and intuitions.

It might be apposite to begin by stating briefly what, in my view, constitutes the originality of structural anthropology. There is a dual aspect to it: one is a method for knowing and analyzing certain types of social facts, inspired by structural linguistics, that Lévi-Strauss and a few others following him have used with great success; but it is also a particular point of view on the very nature of social facts and on the epistemological conditions of getting to know the phenomenal world, which Lévi-Strauss developed alongside his method, and which came to represent for most observers outside anthropology, primarily the philosophers, the heart of his doctrine. Now, one can take advantage of the method without sharing the whole gamut of moral and philosophical convictions which condition the Lévi-Straussian vision of human experience, a vision that Lévi-Strauss himself had never any desire to impose upon anyone else.

What does the method consist in? According to Lévi-Strauss, it requires, first, the isolation of sets of phenomena that are suited for this kind of analysis, essentially those that belong to what he calls the superstructures, that is, elements of the world that the unconscious activity of the mind reputedly organizes in significant and systematic sets: the rules of kinship and marriage, classifications, myths, food norms, or artistic forms. In those systems, one looks for the relations, realized or potential, which connect elements characterized not by intrinsic properties but by their position vis-à-vis one another. The aim is to bring to light in a table of permutations the whole set of possible combinations between these elements. The structural model which results from this operation does not aim at a faithful description of any social situation; it is a heuristic device which provides the syntax of [35]the transformations which allows one to move from one variant to another within a class of phenomena. This basic principle of structural analysis was aptly defined by Jean Pouillon when he wrote that “structuralism properly speaking begins when one admits that different sets can be brought together, not in spite, but in virtue of their differences, that one then tries to order” (1975: 122). Strictly speaking, structural analysis in anthropology amounts to that: it reveals and orders contrastive features so as to discover the necessary relations organizing certain sectors of social life, such as the set of culinary techniques or of the ways to exchange potential spouses between individuals and groups. In sum, it is a very efficient method to reach the objective which any anthropological analysis should aim at: the detection and ordering of regularities in statements and practices.

To this general method one must add a point of view which is properly Lévi-Straussian on the nature of sociality and the purposes of the anthropological inquiry, with which one may partially agree or wholly disagree. I may say a few words about some aspects of it with which I am particularly concerned. The first one is the marked semiological inflexion of Levi-Strauss’ approach: for him, social life is a network of exchange of different kinds of objects which circulate like linguistic signs. Language provided him initially with a model for the whole of social life and offered a hope that the study of the latter might one day become as scientific as the study of the former. Language, and the language-like properties of symbols, is thus fundamental for him, both in its Saussurean dimension (meaning as arising out of contrastive features) and in its praxeological dimension (social interaction envisioned as an exchange of linguistic signs). And what Lévi-Strauss sometimes calls the symbolic function is a language-like ability of humans to give sense to the world by detecting in it salient features that can be organized in contrastive sets. Now, the emphasis put on symbolism and symbolic signs presents a problem for anthropologists like myself whose ambition is to widen the scope of the social sciences, to accommodate in them more nonhuman beings, and thus to shift the focus away from the sociocentric analysis of conventions and institutions toward the interactions of humans with (and between) animals, plants, physical processes, artifacts, images, and other forms of beings. For Lévi-Strauss both anticipates this project and conflicts with it. He anticipates it when he assumes, in the Mythologiques, for instance, that the world is to be seen not as the exclusive playground of humans, but as a vast array of meaningful differences between qualities and beings that can be systematically organized according to these differences. What I see implied there are the premises upon which I have been working that the sources of the plurality of beings and of regimes of existence lie at a deeper level than the sociocultural one traditionally studied by anthropology. This level, which, to borrow from Husserl, may be called antepredicative, is the one where humans and nonhumans become aware of each other and develop modes of relating prior to the usual processes of categorization and communication embedded in historically and linguistically contingent frameworks. However, these nonanthropocentric assumptions are at the same time contradicted by Lévi-Strauss’ dependency upon the language-like properties of social life. If one yields to this dependency, one fails to purge anthropology of its anthropocentrism, because nonhumans, devoid of linguistic abilities and capacities for symbolism, will always remain the passive objects of human cognition and inventiveness, mere bundles of qualities that humans detect and organize in symbolic [36]patterns. If nonhumans are to become agents in their own right, then they have to be able to escape this symbol-induced passivity.

There are several strategies to circumvent this problem, but I must admit that none is entirely satisfying. For example, one may focus on nonsymbolic signs—images, for instance— and study how the latter behave as iconic agents in the mesh of social life; this is what I have attempted to do in recent years (see, e.g., Descola 2010). Or one may adopt an approach in terms of Gibsonian affordances, giving some active leeway to the object represented for it to draw the perceiver into its sphere of influence. Or one may try to avoid altogether the use of the words “symbolic” and “symbolism,” even eschew the very idea of representation, thus hoping that the processes they qualify will just disappear as problems by not being referred to; this is the most common strategy. Another solution, which was propounded recently by Eduardo Kohn in his book How forests think (2013), is quite attractive but poses problems of its own. Following Peirce’s triadic semiology, Kohn argues that iconic signs (which share likenesses with what they stand for) and indexical signs (which are in a relation of spatial or temporal contiguity with what they represent) have to be brought into the anthropological agenda, not only as supplements to symbolic signs and to enrich humans semiosis, but also because icons and indexes are the signs that nonhuman organisms use to represent the world and communicate between life-forms. Their study would thus offer a way to bring together humans and other living beings within a more embracing semiosis and would provide a foundation for an anthropology which Kohn calls “beyond-the-human.” In other words, there is no need to shun representation as if it were a malevolent and Eurocentric process of dissociation between things and words that impedes a true access to, and a wholesome experience of, the richness of the world. For there are other forms of representation than those which use symbols, so that making present something absent converts all beings who possess this disposition—all organisms, according to Kohn—into selves. The forest that thinks by itself is an interesting way out of the forest of symbols which has thus far restricted the scope of anthropology—or perhaps we had better call this new science an ecology of relations; but it is far from solving all the problems of anthropocentrism. In particular, as I have argued elsewhere (Descola 2014), it restricts agency to semiosis and semiosis to organisms, thus leaving unaccounted for a great number of nonhuman and nonsemiotic agents nevertheless playing an important role in the social field, from artifacts to the climate.

But let us return to Lévi-Strauss. A second matter of concern with his idiosyncratic view of social life is the enigmatic role played by what he calls “the unconscious” (l’inconscient). According to him, the unconscious activity of the mind accounts for the structure and the functioning of symbolic systems, their variable contents being a contingent effect of the physical and historical environments within which they are deployed. Hence his well-known statement that anthropology is first of all a psychology. Although I am myself convinced that a considerable part of our cultural dispositions are acquired, used, and transmitted in a nonconscious way, I am rather reluctant vis-à-vis the Lévi-Straussian project of extracting from the black box of the unconscious little pieces of objectified thought embodied in institutions so as to deduce from them laws of the operation of the mind. Rather than deducing these kinds of dispositions from the social expression that they render [37]possible, as Lévi-Strauss does,1 I find it more reasonable to give cautious credit to what cognitive psychology has begun to teach us about the mechanisms allowing the acquisition and stabilization of nonpropositional knowledge. Finally, I share with Lévi-Strauss a general monist gnoseology which is marked both by a rejection of the Cartesian type of cognitive realism and by the assumption that there exists a physical continuity between, as he says, “the states of subjectivity and the properties of the cosmos,” although I find such an epistemological outlook among other authors too: Maurice Merleau-Ponty, Francisco Varela, James Gibson, or even, as Lévi-Strauss himself acknowledges, Spinoza.

If we leave aside Lévi-Strauss’ personal views on the nature of social life and on the philosophy of knowledge, and return to the method of structural anthropology, what the latter brings, compared to other great interpretive models of social phenomena such as historical determinism or functionalism, is the idea that no human phenomenon has a meaning in itself, that it only becomes relevant when it is contrasted with other phenomena of a similar kind—a system of marriage with other systems of marriage, a variant of a myth with other variants of a myth; so that the object of the inquiry is less the description of the phenomenon than the logic of its contrasts. This method often proved to be highly productive, in particular because it demoted the question of causality—whether environmental, cognitive, technical, functional, economic, political, or ideological—which has long vitiated all attempts at explaining social facts. Instead it gives preeminence to the conditions of composition of common worlds, that is, the principles which govern the compatibility and incompatibility between institutions, practices, ideological systems, sets of values, techniques, forms of exchange. It is thus more faithful to the mission of anthropology as I understand it. This mission is to bring to light these principles of composition, to understand why, in given circumstances, some elements have coalesced in a certain way to create polities of a certain kind; it is also to explain why these relations of compatibility and incompatibility recur under similar forms in very different parts of the world, why certain features do or do not mix harmoniously with others. Structural anthropology is a very efficient tool for revealing the laws of coalescence of these aggregates, precisely because it is attentive to the systematics of differences and because it eventually allows the integration within a unified model of local structures elicited from different fields of practice.

Technically, structural anthropology achieves the ends I have just sketched thanks to a very original tool which I have also used myself with profit: transformation. According to Lévi-Strauss himself, transformation is the keystone of the type of analysis which he practices and it is also what I find most fertile in his approach. He borrows from linguistics the notion of structure, understood as a system of contrastive oppositions, but he endows it with an analytical dynamism which results from the capacity of the structure to account for the ordered transformations between models of the same group, that is, models which are applied to the same set of phenomena. A structure is not a system. For a structure to exist there must be between the elements and the relations of various sets invariant relations which [38]allow moving from one set to another by the means of a transformation. Now Lévi-Strauss uses the operative tool of transformation in two quite distinct ways, which relate to two different morphogenetic traditions, that of the biologist D’Arcy Wentworth Thompson, whom he claims to follow, and that of Goethe, to whom he pays lip service;2 it is the latter which directly inspired me.

The Goethean type of transformation is the one that Lévi-Strauss implements in Les Structures élémentaires de la parenté, although he does not say so himself, but states instead that he was inspired in that book by D’Arcy Thompson.3 In the analysis of marriage systems the invariant relation is the exchange of women as an expression of the principle of reciprocity, itself the positive form of incest prohibition: the men abstain from the women of their own group and give them to the men of another group so as to obtain spouses in exchange. All the forms of matrimonial exchange that Lévi-Strauss analyses are as many transformations of this original principle, forms that he studies as a function of the growing complexity they exhibit in relation to the simplest sociological form that the principle of reciprocity can take on; this is the dualist organization where a society is divided into two classes who exchange women according to the rule of exogamy. This elementary form, known as restricted exchange, is the absolute minimal organization of reciprocity below which no social life is possible. From there a series of transformations follow which develop the possibilities of the initial invariant: the generalized exchange of the Australian marriage systems with four and eight sections, which are developments of the logical possibilities of the initial reciprocity invariant when the groups of wife givers and wife receivers are subdivided according to their place of residence; then the system of the Kachin of Highland Burma, in which generalized exchange is combined with bride purchase so that women circulate between groups in one direction while bridewealth circulates in the opposite direction; and finally a series of combinations of restricted and generalized exchanges in China and India, which are complex transformations of the original structure of reciprocity, with China combining restricted exchange in the peasantry and generalized exchange in the aristocracy, while India undergoes a dual evolution from generalized exchange toward hypergamy in the higher castes, and toward the closure of the lower castes within a system of restricted exchange.

This arborescence of the forms of marriage bears a strong resemblance to the methodical variations of an Urform in Goethe’s sense, the original form being in that case the exchange of women ruled by the principle of reciprocity, of which Lévi-Strauss unfolds all the logical consequences in as many morphological types of marriage alliances. Just as Goethe (1790) entertained the hope of discovering one day the Urpflanz, the original plant, the prototype from which could be derived by transformation the whole set of characteristics of all vegetal species, both [39]actual and logically possible, Lévi-Strauss sees in the principle of reciprocity the original form of all possible types of marriage alliance, of which he propounded the law of development. And just as Goethe opposed Linnaeus in botany because he rejected the idea of a static table of attributes, exhaustive though it may be, in favor of the deduction of a principle of transformation of biological forms from a complex initial combination, in the same way Lévi-Strauss opposes Radcliffe-Brown in the domain of social morphology when he considers as initial the structure which develops on the widest scale the logical possibilities of the principle of reciprocity, rather than retain as a point of departure the simplest forms of marriage from which the more complex would be derived.

By contrast with this conception of the variation as development of a complex prototype, in the analysis of myths Lévi-Strauss adopts an entirely different approach to variation which he claims to have borrowed also from D’Arcy Thompson.4 In his magnum opus On growth and form (1917), the latter propounds to use a geometrical grid of transformation to move from the form of an organism to the form of another organism by a process of continuous deformation and without resorting to some complex initial form out of which all the other forms of organisms would be derivable. Independently of any evolutionary lineage, one can thus move from the skull of an extinct species of rhinoceros to the skull of a contemporary species of tapir, and from this to the skull of a horse. Lévi-Strauss adapts this method to the analysis of myths. A mythical transformation is the group constituted, on the one hand, by the set of variants of a myth that keep the same structure, including by inverting it, and, on the other hand, by the set of myths, often originating from neighboring societies, that can be shown to transform each other by mutually borrowing episodes—which Lévi-Strauss calls mythemes—the motives of which they will invert or the function of which they will swap. The transformation of organic forms and the transformation of mythemes thus proceed in the same way, by a continuous series of small variations within a group.

However, as Lévi-Strauss himself acknowledges, there exists a major difference between myths and organisms: while the specific or generic differences between organisms depend in the end on discontinuities introduced by the genetic code, the group of transformation within which myths are contrasted constitutes a virtual continuum upon which the analyst performs arbitrary cuts (Lévi-Strauss 1971: 603–4). It is the analyst who chooses, according to his or her knowledge and his imagination, from the huge mass of myths and variants of myths, the mythemes which will yield the highest contrastive and semantic output. From this point of view, the variants of a myth do not carry by themselves a principle of discontinuity. In other words, the transformation by continuous variation that D’Arcy Thompson operates on the morphology of organisms requires that stable forms be given as templates both at the starting point—the rhinoceros—and at the end point—the horse; this is necessary for bringing to light the physico-mathematical principles which allow movement from one form to the other by a deformation without loss [40]of continuity. There is no such thing in the structural analysis of myths: each analyst will trace the path of transformation which his or her fancy will suggest to follow. While the mythical transformation introduces discontinuity by the means of contrastive operations upon a set of phenomenal properties relating to aspects of the world that can be shown to have common semantic properties, the anatomical transformation introduces continuity by the means of deformations exercised upon a set of phenomenologically dissimilar forms between which affinities are hypothesized.

Whether it concerns organisms, images, social types, or the semantic units of certain kinds of statements, the transformation of one form into another can thus be undertaken either in a “Goethean regime” or in a “Thompsonian regime.” In the former case, the transformation is the development in different forms of a kind of initial blueprint itself constructed by comparing empirical objects pertaining to the same set. In the latter case, transformation is a deformation by continuous variation in a space of coordinates that applies to forms that are already given. The first method, which Lévi-Strauss follows in Les Structures élémentaires de la parenté, is also the one which inspired my own approach when I undertook to organize the diverse forms of continuity and discontinuity between humans and nonhumans out of an initial relation between interiority and physicality (Descola 2013). I will take the liberty of recalling the guiding lines of my own reasoning as an illustration of the very theme of the symposium: keeping structuralism alive (and keeping thought alive with structuralism).

Simply stated, the ontological group of transformation that I developed in Beyond nature and culture purports to contrast modes of identification, that is, framing devices that regulate habitus, guide inferences, filter perceptions, and are largely the products of the affordances which the world offers to the specifically human dispositions. A fundamental function of these framing devices is to ascribe identities by lumping together, or dissociating, elements of the lived world that appear to have similar or dissimilar qualities. My argument is that one of the universal features of the cognitive process in which such dispositions are rooted is the awareness of a duality of planes between material processes (which I call “physicality”) and mental states (which I call “interiority”). By using this universal grid, humans are in a position to emphasize or minimize continuity and difference between humans and nonhumans. This results in a fourfold schema of ontologies, that is, of contrastive qualities and beings detected in human surroundings and organized into systems, which I have labeled “animism,” “totemism,” “naturalism,” and “analogism,” thus giving new meanings to well-worn anthropological concepts.

The range of identifications based on the interplay of interiority and physicality is in fact limited: when confronted with an as yet unspecified aliud,5 whether human or nonhuman, a human cognizant subject can surmise either that this object has a similar interiority and a different physicality, and this I call “animism”; or that the object is devoid of interiority but possesses a kind of physicality similar to that of the subject, and this is what I call “naturalism”; or that this object shares with other humans and nonhumans elements of physicality and interiority that are [41]similar but which altogether differ from those that other humans share with other nonhumans, and this I call “totemism”; or, finally, that this object’s interiority and physicality are entirely distinct from the subject’s own, even though they display small enough differences to allow for relations of correspondence, and this I call “analogism.” Each of these combinations affords a glimpse on a more general principle governing the distribution of the continuities and discontinuities between any human subject and the objects of its environment. Each of these modes of identification serves, moreover, as a touchstone for singular configurations of cosmological systems, of conceptions of the social link and theories of otherness that are as many instituted expressions of more entrenched mechanisms of recognition of the other.

Now this ontological group of transformation bears a resemblance to what Lévi-Strauss sometimes calls the “order of orders” (1958: 347), namely the upper level of structural articulation of the various systems composing social life. With the slight difference that the articulation here is not a function of an integration of levels already analytically defined, but results from a hypothesis as to what comes first in the experience of the world: namely discerning qualities in the objects that surround us and inferring the kind of relations that they afford. However, the matrix of identification does not work as a philosophical prime mover; rather it functions as a sort of experimental device that allows me to capture—thus to bring to existence—and to classify—thus to combine—certain phenomena to highlight the syntax of their differences. But there is more. By adopting this device, I wanted above all to remain faithful to this basic principle of structural analysis which holds that each variant is a variant of the other variants and not of any of them in particular which would be privileged. For if I gave the ontological matrix a fundamental position, on the other hand, none of the variants that it allows (animism, naturalism, totemism, analogism) and none of the variants detectable in other systems, which are as many transformations of the matrix—in the sociological, praxeological, epistemic, cosmological, spatiotemporal, or representational orders—can claim to predominate over any other variants. This was a requirement which I had set myself from the start so as to produce a model of intelligibility of social and cultural facts that would remain as neutral as possible in relation to our own ontology, naturalism. And this is why naturalism is only one of the four ontological variants in the matrix.

Structuralism in general thus provides the fairest form of symmetrization that anthropology can afford. By symmetrization I mean an attempt, proper to anthropology and one of its main claims to distinction, to render compatible on an equal footing the cultural features of the observer and those of the observed. For it became clear to all lucid anthropologists early on in the development of the discipline that ethnological comparativism, whether implicit or explicit, always presupposed in one way or another that the point of view of the analyst doing the comparison encompassed the point of view of the members of the societies that were being compared, or at least set a convenient point of reference for its evaluation. To redress this imbalance, a number of forms of symmetrization have been devised, the most common apart from structural analysis being to generalize the range either of a local concept—totem, mana, taboo, shaman, or hau—or of a local epistemic stance—Dumont’s idea of the hierarchical encompassment or Marilyn [42]Strathern’s notion of the person as an objectification of relations—and thus confer to the knowledge of the observed an equal, even competing, status to that of the observer. Another form of symmetrization is to transform an account of a native way of thinking into a more or less systematized corpus similar to a philosophical doctrine, at least in its mode of presentation. This is also an ancient tendency in the West, and one that even predates the former type of symmetrization, since it has been a characteristic feature over several centuries of a certain type of missionary anthropology.6 More recent cases are the famous Philosophie bantoue of Father Placide Tempels (1945), which triggered a heated debate among African philosophers, or even Eduardo Viveiros de Castro’s latest version of perspectivism (2014). All these forms of symmetrization are wanting for reasons that it would take too long to develop here, but the cardinal one is that they remain an idiosyncratic exegesis which upsets, and bypasses, the pragmatic conditions of utterance and of reception of the propositions which reputedly provide the operationality of the practice or norm transformed into an analytical concept or a philosophical doctrine.

The kind of symmetrization which structural analysis practices does not purport to generalize the range of a local concept or epistemic stance, or to offer a philosophical countermodel inspired from a local way of thinking, but to compose a combinatorial matrix that would account for all the states of a set of phenomena by bringing to light the systematic differences which oppose its elements. Why is it a symmetrization? Because, in accordance with standard structuralist procedure, the totalization is never given ab initio, as the starting point from which the Sirius of anthropology might structure the world under his imperial gaze, but results from the always uncompleted operation by the means of which cultural features, norms, institutions, qualities, propositions, are constituted as variants of one another within a set. And this set may not only be reconfigured differently if new elements are added; it has no other raison d’être than to subsume the variations for which it provides the encompassing framework. Far from being the “intellectual ideology, and the immanent logic, of a new, technocratic totalitarianism,”7 this type of symmetrization is in no way claiming a universalist position of detachment; for it is entirely dependent upon the multiple properties that people detect here and there in phenomena, and it thus requires nothing more in terms of an overhanging epistemic point of view than acquiring some knowledge on the diversity of the objects one deals with, a modest claim for what is after all a scholarly undertaking. Nevertheless, I have no qualms in admitting that the combinatorial option is as wanting as the two other forms of symmetrization, and for reasons that are altogether different: it requires a general knowledge of the institutions and ways of life of other peoples that only the West has been until now in a position to produce.[43]

As such, the structural-ontological approach depends upon a project of knowledge which is quite particular, not so much because of its universalist goal—for there were numerous systems of knowledge elsewhere which purported to account for everything—but because of the requirement of worldwide exhaustiveness of the empirical data upon which it rests.


Descola, Philippe, ed. 2010. La Fabrique des images: Visions du monde et formes de la représentation. Paris: Somogy/musée du quai Branly.

———. 2013. Beyond nature and culture. Translated by Janet Lloyd. Chicago: University of Chicago Press.

———.2014. “All too human (still): A comment on Eduardo Kohn’s How forests think.” Hau: Journal of Ethnographic Theory 4 (2): 267–73.

Descombes, Vincent. 1989. Philosophie par gros temps. Paris: Éditions de Minuit.

Diamond, Stanley. 1974. In search of the primitive: A critique of civilization. New Brunswick, NJ: Transaction Publishers.

Goethe, Johann Wolfgang von. 1790. Versuch die Metamorphose der Pflanzen zu erklären. Gotha: C.W. Ettinger.

Kohn, Eduardo. 2013. How forests think: Toward an anthropology beyond the human. Berkeley: University of California Press.

Lévi-Strauss, Claude. (1949) 1967. Les structures élémentaires de la parenté. Paris: Mouton & Co.

———.1958. Anthropologie structurale. Paris: Plon.

———. Du miel aux cendres. Paris: Plon.

———. 1971. Mythologiques, IV. L’homme nu. Paris: Plon.

Lévi-Strauss, Claude, and Didier Eribon. 1988. De près et de loin. Paris: Éditions Odile Jacob.

Petitot, Jean. 1999. “La généalogie morphologique du structuralisme.” Critique 620–1: 97–122.

Pouillon, Jean. 1975. Fétiches sans fétichisme. Paris: François Maspero.

Tempels, Placide. 1945. La Philosophie bantoue. Translated by A. Rubbens. Élisabethville (Lubumbashi): Lovania Éditeur.

Thompson, D’Arcy Wentworth. 1917. On growth and form. Cambridge: Cambridge University Press.

Viveiros de Castro, Eduardo. 2014. Cannibal metaphysics. Translated by Peter Skafish. Minneapolis: Univocal.[44]

La transformation transformée

Résumé : La clé de voûte méthodologique de l’anthropologie structuraliste lévistraussienne est la notion de groupe de transformation. Une structure acquiert un dynamisme analytique seulement si elle possède la capacité d’organiser ces transformations grâce aux modèles d’un même groupe de phénomène. Cet essai explore différentes approches de ce concept et s’appuie sur leurs résultats pour examiner les conséquences de l’appréhension du pluralisme ontologique comme un groupe de transformation.

Philippe DESCOLA is chair of the “anthropology of nature” at the Collège de France and professor at the École des Hautes en Sciences Sociales, Paris. Following extensive research in Amazonia, Descola developed a comparative approach to relations between humans and nonhumans, and now studies the anthropology of images, non-Western and Western.

Philippe Descola
Collège de France
52, rue du Cardinal-Lemoine
F-75005, Paris


Keynote lecture for the symposium “Living Structuralism/Le structuralisme vif,” Toronto, October 1–3, 2015.

1. For instance, the three cognitive imperatives that Lévi-Strauss elicits out of the working of marriage systems: the necessity of rule, the notion of reciprocity, and the synthetic character of the gift (Lévi-Strauss [1949] 1967: 98).

2. Lévi-Strauss acknowledges in passing that the morphogenetic tradition of which structuralism can be seen as one of the offshoots can be traced back to Goethe (Lévi-Strauss 1958: 354). In his article on the morphogenetic origins of structuralism, Jean Petitot (1999) reviews some of the sources that were relevant for Lévi-Strauss in this domain, but without contrasting the ways he used them, as I am attempting to do here.

3. This is the claim he made in a book of interviews with Didier Eribon (Lévi-Strauss and Eribon 1988: 159).

4. The most explicit reference to D’Arcy Thompson is at the end of L’homme nu (Lévi-Strauss 1971: 604–6) and explains how transformation is central to the structural analysis of myths; there are also two passing mentions of D’Arcy Thompson, one in Anthropologie structurale (1958: 358), the other in Du miel aux cendres (1966: 74n).

5. Aliud, a concept denoting an undetermined alter, that is, one which has not been specified yet by a relation, is borrowed from Descombes (1989: 89).

6. The Historia general de las cosas de la Nueva España, compiled in Nahuatl by Bernardino de Sahagún in the sixteenth century, is probably the earliest example of this trend, while the Lettres édifiantes et curieuses written by the Jesuits from China are their most celebrated expressions, if only because of the role they played in the formation of Leibniz’s ideas.

7. “For structuralism, epitomized in Lévi-Strauss, is the intellectual ideology, and the immanent logic, of a new, technocratic totalitarianism” (Diamond 1974: 297).