The elementary structures of being (human)

This work is licensed under the Creative Commons | © Michael Lambek. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau4.3.028


The elementary structures of being (human)

Michael LAMBEK, University of Toronto

Comment on Descola, Philippe. 2013. Beyond nature and culture. Translated by Janet Lloyd with a foreword by Marshall Sahlins. Chicago: University of Chicago Press.

Imagine if Claude Lévi-Strauss had attempted to apply the methods and goals of The elementary structures of kinship to the substance of The savage mind, that is, to offer a systematic account and comparison of the possibilities and limits of the forms for imagining the place of human beings in the world, in relation to other kinds of being and in internal relation to ourselves. This is what Philippe Descola sets out to do in Beyond nature and culture (2013a). The measure of the book is how far he accomplishes this ambitious goal.

In fact, Beyond nature and culture is a magnificent achievement. It offers a beautiful and complex model and is written with great precision of thought and elegance of language, a tour de force of both coherent argument and ethnographic comprehension. But as it would be boring simply to sing its praises and churlish to poke holes in little pieces of the argument, I turn to some of the big questions it raises. My comments are directed to the larger frame and are intended to make certain assumptions and limits more explicit than perhaps Descola would like them to be.

This book is the latest phase and new face of structuralism (a far better term than ontologism …), reappearing in the Anglophone world after a hiatus but almost as brilliantly as the original work of the master. Along with Eduardo Viveiros de Castro and a few others, Descola rescues structuralism from the endless permutations of the Mythologiques and the loyal application to myth of the pseudo-mathematical formula Lévi-Strauss once put forward (Maranda 2001) and offers an implicit challenge to both the universalizing abstractions of poststructuralism and the sort of aimless empiricism at which Marshall Sahlins pokes fun in his Foreword and which Descola merely ignores. It is not a “turn” but a resurgence and renewal.

The structuralist program is best described in Descola’s own words. Descola is “trying to gain a better understanding of the principles according to which humans schematize their experience of things in such very different ways, welcoming non-humans into their collectives with varying degrees of liberality, and either actualizing the relations that they discern between existing beings in concrete systems of interactions or not doing so… . [These] forms of relating to the world seem to me neither limitless nor incommensurable” (2013a: 403). Furthermore, the “aim is limited to establishing the bases for a way of conceiving the diversity of the principles of a schematization of experience that is free of the preconceptions that modernity has led us to maintain regarding the state of the world” (405).

Deeply inspired by his teacher, Descola, as his title suggests, does not simply follow Lévi-Strauss but critically reexamines some of his central arguments. In particular, Descola appears to reject the centrality of the nature/culture opposition to all forms of human thought, considering it precisely and exclusively a modern preconception. In the course of my response I question this claim. I would like to take the “beyond” of his title in the sense of expansion rather than abandonment.

* * *

Descola’s argument proceeds in two stages. The first, larger step sets out four modes of identity or ontology. The second lays out six modes of relation and then tries to show a kind of compatibility (elective affinity?) between certain modes of ontology and certain modes of relation. Identity and relation approximate in some respects to what used to be called culture and society. This second phase of the argument probably deserves a book of its own.

I have great admiration for the project and the content of the book. The elaboration of the four modes of ontology is original and the discussion of the modes of relation refreshing. Descola’s models are at the very least extremely good to think with, heuristic, as he says. It is important to recognize these are four modes of ontology, not distinct ontologies, with the entailments of closure and mutual incomprehensibility that objectification would bring.

Niels Bohr is reputed to have said, “It is wrong to think that the task of physics is to find out how nature is. Physics concerns what we can say about nature …” Anthropology in Descola’s hands is about what we can say about what humans do say about the world, the ways in which humans describe and understand their relations to others in the world and thereby live together in it. Descola argues that the identifications and distinctions that humans find or make with other beings hinges in part on what kinds of internal divisions they recognize within themselves, that is, the parts of the world they find within themselves, that may connect them with other beings. From these identifications and distinctions, distinct modes of relations between beings follow, and Descola classifies these as well.

Descola argues that humans do all this without resorting to the concepts of nature and culture or rather, that resorting to an opposition between nature and culture is only one of several possible ways of doing so. Hence the nature/culture opposition is not universal and it is limited and defective as an analytic lens for the anthropological project. By contrast, I want to argue that we cannot talk about the “what” it is that people (including Descola) are doing or the subject about which they are thinking and talking without resorting to terms like nature and culture, albeit giving them more abstract reference and less specific sense than that characterized by their popular usage (itself quite heterogeneous) or their unique historical development in Europe. I think the terms or concepts are implicit in Descola’s argument. To go “beyond” nature and culture is not thereby to reject them, only to recognize a more expansive scope, meaning, and differentiation.

* * *

Consider the following two propositions: (1) Human beings are a natural species, just one among other species. This is to understand “oneself as another” (to borrow a phrase from Paul Ricoeur), the human as “another other”; (2) The human species, together with its products and achievements (and perhaps its defects), is distinctive in important respects, exhibiting a qualitative difference in being from the kind of differences (say, between chimpanzees and gorillas) manifest between other closely related species. I take both these points to be incontrovertible and to be features of the ontology underlying anthropology. However, as Descola demonstrates, it is evidently not the case that all societies have considered these two propositions in this manner. Certainly, for most of human history there was not a discrete domain or entity conceptualized as “nature,” nor another sphere clearly indicated as cumulative “history” or “culture,” but this did not preclude people from considering the similarities and differences between themselves and other animals and, to one degree or another, or in one form or another, to index their distinctiveness and their particular relationships, if not to actively recognize them. Indeed, it is one of the distinctive features of human beings that we not only can but also must do so.

What Descola does is to survey the range of fundamentally different (distinctive) ways in which human distinctiveness from and connections to others have been conceptualized. While he challenges the universality of a nature/culture model, such a model, in some manner, must haunt the project itself. Nature/culture is one way to name the human condition at the most general level.

Descola certainly complicates the question of nature and culture understood as a simple binary opposition and in that sense he certainly does go “beyond” it. But in fact, he replaces it by two binary oppositions, a substantive one between physicality and interiority, and a logical one between similarity and difference. These oppositions intersect one another, thereby producing four alternatives. From a “naturalist” perspective the oppositions are respectively: (a) what distinguishes humans (in my language, “culture”) from and relates them to other beings (“nature”); and (b) what is it within human beings, i.e., in the composition of humans, that is “natural” (similar or identical to other beings) and what is “cultural” (distinctive to humans). These are questions of external and internal relations respectively and they are good naturalist (or culturalist) questions, but they do not actually go beyond the problematic of nature and culture. Moreover, it is the brilliance of Descola’s book to show that every society negotiates these questions and hence that the issues are found well beyond the naturalist (“our”) mode of ontology.

The issues are intrinsic to the human condition, from whatever perspective that condition is viewed by humans. Indeed, it is our very ability to view our condition, and perhaps our compelling need to do so, as well as the ability or necessity of doing it from particular, hence multiple, perspectives, that gives rise to these questions. And so far as we know, and whatever the abilities of macaques to learn to wash sweet potatoes, chimpanzees to use a few phrases on a keyboard, or animists to think of other species otherwise, we are the only species able to reflect on our condition (and the condition of other kinds of beings) and to come up with the various solutions that Descola labels naturalism, animism, analogism, and totem-ism. I take this to be a naturalist, metanaturalist, or hermeneutic truth and I take it that Descola does so as well.

However brilliant the argument, Descola does not evade the paradox of locating the anthropologist’s position. Descola resembles other structuralists here in ignoring the hermeneutical questions to which the possibility of structuralism both gives rise and demands an answer (and I hereby point to previous responses from hermeneutics to previous generations of structuralists).

I hope it is clear that I do not find nature/culture dualism and Descola’s quadripartite model mutually exclusive but rather see them as different levels of abstraction. Descola himself cannot get by without the terms. Thus he says in his preface “the anthropology of culture must be accompanied by an anthropology of nature that is open to that part of themselves and the world that human beings actualize and by means of which they objectivize themselves” (2013a: xx). Nature and culture are intrinsically troubling categories insofar as (like mind and body1) they are neither mutually exclusive nor identical nor commensurable; there is no single or final way to assign or settle their relations and the plurality of ontological modes is empirical proof of this. The analysis and survey of the elementary differences in how this can be approached is precisely the value of this book. Yet in the end this lived incommensurability is human nature, is human culture. There is no one right or simple solution, hence many lines of thought have been proposed and these tend to be incommensurable to one another rather than contradictory.2 Our thoughts about the world and our place in it remain essentially unresolved, a world then of intellectual puzzle and pleasure but also fundamentally incomplete.

Finally, what are nature and culture to be replaced by? A single word, ontology? And would this replace Lévi-Strauss’s restless dualism or intellectual tension with a kind of complacent monism or intellectual resting place? I am sure this is not Descola’s intention but he might clarify his own distinctive position. He is certainly not among those ontologists who advocate a radical form of relativism. He eventually slips in an admission of being a naturalist (2013a: 244) and more explicitly on page 303. Instead Descola elaborates what I might call a kind of relative relativism, or what he describes in an important passage (305) as relative universalism. He appears to think the alternatives are finite in number and discontinuous from one another, but is equally at pains to show that the schemas are not rigidly bounded off from one another and that life is lived with a plurality of schemas. Of course, some leaps of understanding might be wider than others (a point he does not develop here).

* * *

I raise a few other points concerning the ontological modes (“dispositions of being”). Descola discerns four alternatives, clarifies how they look and how they differ from one another. One wonders though, where the criteria for distinguishing them come from, from which ontology or metaontology, as it were—and what would be the effect of classifying along different axes or criteria, thereby creating categories overlapping the ones he distinguishes. Why are there precisely four types, why do they draw from precisely two sets of oppositions, and why are the axes conceived as binaries?

Does the stability of the categories emerge from reason, from material factors, or from tradition? The totemic mode is largely restricted to Australia; it is interesting the other alternatives did not emerge or take hold there as well, as one might expect if the rationalist hypothesis held. Similarly, although animism is very widespread, Asia and the Americas have an ancient cultural connection. I think the point is that while the principles derive from the mind their coalescence and elaboration into fully fleshed and locally dominant versions is manifest through tradition.

The case remains open whether the four alternatives as discerned in the model are each fully commensurable with the others. Commensurability here is a product of the intersection of two lines of opposition. These oppositions may not be equivalent to one another and they may not intersect at right angles as the use of graphs supposes. To put this in another way, how comparable are the relations between any two modes along a single axis with the relations along a different axis?

Finally, what is the relationship of pure (ideal) types to actual life? Does Descola assume the actual manifestations are each internally consistent and complete? Is anything else a mere “hybrid?” Or are pluralism and mixture the norm and the “pure” type a rarity? This question is reminiscent of the debate surrounding Lévi-Strauss’s depiction of prescriptive marriage systems. Lest such skepticism, whether in an empiricist or philosophical vein, be perceived as mere sniping, it should be recognized that the very question defers to the much harder and more generative intellectual work of the original abstraction of the model.

* * *

This is theory that takes animism seriously, not by naively proposing an animist view of the world but by starting as if animism mattered. Like all ethnographers, Descola begins with the contrast between the society in which fieldwork is conducted and the society from which the anthropologist comes. Hence the Achuar are critical here and by extension Amazonia more generally.3 Toward the end of the book Descola ventures that the significance of functionalism in British social anthropology stemmed from the circumstance of studying societies in Africa and Asia where analogism reigns. Descola himself is determined to transcend ethnographic positionality and particularism and his ability to do so is a mark of the value and importance of structuralism by contrast with the kinds of dyadic “conversations” characteristic of most hermeneutic anthropology. He thereby also escapes a “West versus Rest” model.

Animism is defined as “the attribution by humans to non-humans of an interiority identical to their own” (Descola 2013a: 129). I don't know why Descola says “identical” rather than “similar.” In the animist mode humans perceive others as humans on the inside and understand that nonhumans perceive themselves as humans. However, Descola narrows Viveiros de Castro’s more radical rendition of perspectivism; only in some instances is there the further elaboration that nonhumans perceive humans in nonhuman bodily form.

Of the four modes animism is inevitably the most generative here; hence its contrasts with naturalism along one axis and with totemism along another are the most compelling. Analogism, which is placed in the far quadrant from animism, is perhaps the mode handled least convincingly, in part because it is so large and comprehensive and serves something of a remainder to the three more clearly demarcated modes.

Descola’s depiction of totemism is the inverse of that of Lévi-Strauss. He draws on Australian ethnography to show that Australians do identify substantively with particular other species by means of distinctive qualities that link them with a particular place, individual, moiety, et cetera. Totemism in this sense is limited largely to Australia (where it was pervasive). It is evident that totemism and animism are characteristic of most societies that do not go beyond a certain level of political and technological complexity.4 This is an observation that Descola tries to avoid, presumably for the good reason of not wanting to risk a materialist reduction or determinism. But perhaps the difference between animism and totemism is of a different order, or has developed along different paths, than the differentiation of and from the other two modes.

Analogism—which is close to what Lévi-Strauss described as the pensée sauvage as a result of his original critique of “totemism”—is attributed by Descola to premodern Europe, China, India, West Africa, the Andean and central Mexican civilizations, and presumably other places as well. Its wide spread suggests that it may represent a different level of abstraction from Descola’s version of totemism. Moreover, its emergence in central Mexico and the Andes (and presumably the Pueblo societies) within a sea of animism deserves further exploration. Later chapters, introducing concepts like predation, begin to explore the relations between societies where animist and nonanimist modes are respectively dominant. I find the idea that analogism encompasses holism, destiny, transcendent gods, ancestors, sacrifice, spirit possession, and reincarnation quite provocative for thinking about what anthropology has considered “religion.”5

Naturalism, like analogism, appears to cover a lot of ground, incorporating both Christianity and natural science and within the latter positions as variant as anthropology and evolutionary psychology. As Descola recognizes, some recent scientific work naturalizes mind and homogenizes individual differences, and politically there is evidently the treatment of people as “bare life.” Conversely, I would like to know where Descola would place ecological thinkers like Bateson or Rappaport. There is also the question of art. Take the integration of “nature and sculpture” at the marvelous Louisiana Museum of Modern Art outside Copenhagen. “Nature” itself is cultivated and sculpture placed in relation to trees and horizons to form a total landscape, at places barely emergent from the rock from which it is made. What would an anthropologist from Amazonia or Australia make of this? Are not both art and gardening forms of European practice that transcend the divisions of naturalism? Or should we think of them, in Lévi-Strauss’s sense, as merely mediating them?

* * *

A central metaphor of this book is that of ordering. If Descola sets out the “schema of practices … the manner in which … humans … organize their experience, in particular in their relations with nonhumans,” he does so “to set in order and compare the discouraging multiplicity of circumstances in this world” (2013a: 364). We might extend the parallel to set out the modes of anthropological approach to diversity. Structuralist ordering is clearly one of the most powerful and illuminating alternatives and this book exemplifies structuralism at its best. There is insight on every page of this deeply informed, cogently argued, comprehensive, and elegantly written work.


Bernstein, Richard. 1983. Beyond objectivism and relativism: Science, hermeneutics, and praxis. Philadelphia: University of Pennsylvania Press.

Descola, Philippe. 2013a. Beyond nature and culture. Translated by Janet Lloyd with a foreword by Marshall Sahlins. Chicago: University of Chicago Press.

———. 2013b. “Presence, Attachment, Origin: Ontologies of ‘Incarnates.’” In A companion to the anthropology of religion, edited by Janice Boddy and Michael Lambek, 35–49. Boston: Wiley-Blackwell.

Lambek, Michael. 1998. “Body and mind in mind, body and mind in body: Some anthropological interventions in a long conversation.” In Bodies and persons: Comparative perspectives from Africa and Melanesia, edited by Michael Lambek and Andrew Strathern, 103–23. Cambridge: Cambridge University Press.

Maranda, Pierre, ed. 2001. The double twist: From ethnography to morphodynamics. Toronto: University of Toronto Press.



Michael Lambek
Department of Anthropology
University of Toronto
1265 Military Trail
Scarborough, ON M1C1A4 Canada


1. My own line of thinking about these matters has been to pose with reference to mind and body (but it could apply equally to nature and culture or to monism and dualism) the paradox of both/and and either/or relations they inevitably generate (Lambek 1998). It is no criticism to say that Descola himself appears to wrestle or wriggle between monist and dualist tendencies; to oppose dualism is itself a dualist move.

2. I take “incommensurability” as meaning having no common measure, not as contradictory, mutually exclusive, or mutually incomprehensible (Bernstein 1983).

3. Indeed, the so-called ontological turn is a manifestation of the emergence to salience in anthropology of Amazonian ethnography and, by extension, animist ontologies elsewhere, notably Siberia.

4. Moreover, he does allow some overlap. The animism of the Ojibwa is inflected by a bit of totemism (and of course is where the word “totem” originates).

5. See also Descola (2013b).