In recent years, under the heading of the ontological turn, anthropology has mobilized the metaphysical category of ontology in what appears to be, firstly, an attempt to sublate the crisis of representation that problematized ethnography as an ultimately impossible task and, secondly, a refusal to reduce anthropology to the ontologization of human knowledge in the psycho-cognitivist style. The origin of this ‘turn’ is generally located in the work of French anthropologist Philippe Descola, who notably advocates for a structural inversion between the Naturalist (modern) ontology and the Animist (non-modern) cosmography, where one can be read through the lens of the other in the very project of anthropology: here, Naturalism supposes the continuity of physicalities (postulating a “Mononature”) and produces discontinuities at the level of Culture (postulating a “Multiculture”), while Animism supposes the continuity of Culture or “soul” (“Monoculture”) and sees discontinuities in Nature (“Multinature”). As Roy Wagner puts it, European praxis “makes souls” against the backdrop of a given material continuum; indigenous praxis “make bodies” against the backdrop of a given sociocultural continuum. In this model, we don’t have mind (or culture, or language) on one side, and being (or reality, or the world) on the other, but rather various ways of being, or various ways of treating nature. Anthropologists Martin Holbraad and Eduardo Viveiros de Castro discussed with us the implications of such a ‘turn’ in the articulation of universality and diversity.

Araweté shaman bringing down the dead and the gods to partake of a tortoise banquet (Ipixuna, Pará, Brasil, 1982). Photo: Eduardo Viveiros de Castro.

Araweté shaman bringing down the dead and the gods to partake of a tortoise banquet, Ipixuna, Pará, Brasil (photo by Eduardo Viveiros de Castro, 1982).

Glass Bead: Anthropology doesn’t accept anything as being generally true of all cultures unless the imaginaries, cosmologies and collectives it studies are equally seen as variants of each other: nothing could seem further from classical metaphysics. Still, for you the notion of being appears as the most powerful of the comparative operators under investigation: in your words, the aim of anthropology is to produce “equivocations” between ontologies, and its practice is cast as a “comparative ontography.” Not the comparison of ontologies, but comparison as ontology, in what appears to be less an ontology of comparison than a series of variations applied to the very domain of ontology. Can you describe the ways in which ontology—or rather, the ontological—ought to integrate the discourse of anthropology?

Martin Holbraad: I’m not sure these ideas necessarily all go together, or whether, at any rate, I can speak to them all. Given the nature of the anthropological debate on ‘ontology’ (a total buzzword by now, with all the problems that go with that), I think your point about comparison as ontology, as opposed to comparison of ontologies, is a good place to start. Indeed: the idea that if anthropologists are interested in ontology it must be because they want to chart and compare different ‘ontologies’ has proved incredibly hard to budge for those of us who propose the concern with ontology, not as a new way of configuring the object of anthropological inquiry, but rather as a different way of imagining the nature of that inquiry—its coordinates, its modus operandi, and ultimately its ‘point.’ The misreading is hardly surprising. In anthropology (and I guess the same may go for other fields) this is what ‘theory’ has tended to look like: a tussle about how to think of our object of study: is it culture, social structure, habitus, ideology, experience, praxis, values, cosmologies, collectives, deep structures, cognitive schemes? So when anthropologists bandy around ‘ontology,’ it is assumed that this must be put forward as a proposal for a new object of study. Indeed, some of the people who are associated with the so-called ontological turn in anthropology do just that, e.g., Philippe Descola’s project of identifying ethnologically “basic assumptions as to what the world contains and how the elements of this furniture are connected”1 and articulating comparatively four possible sets of them, or Michael Scott’s search for “root assumptions … concerning the essential nature of things and their relationships,”2 which, confirming the terms of your question, he calls “comparative ontology.” The lesson I took from Eduardo’s landmark essay “The Relative Native,”3 which in my view is the most explicit statement of an altogether different concern with ontology—which I share—is that ‘ontology’ for us anthropologists designates not an object of inquiry, but rather the kind of ‘problem’ such objects pose to us. In fact, the thought goes, the very reason for which we should give up on deciding between ‘culture,’ ‘structure,’ ‘praxis,’ or what-have-you, as the rightful object of anthropological inquiry, is that the most basic problem any given object of inquiry poses to the anthropologist is that of deciding what it might be in the first place. So if ontology integrates the discourse of anthropology, as you put it, it does so by rendering anthropological ‘questions’ as ontological ones. I’ll try to show some of the implications of this, hopefully with some examples too, as the discussion develops.

Eduardo Viveiros de Castro: Coming “deuterologically” after Martin, I have the privilege of not having much better to say. So let me try to say more or less the same as Martin did with a difference. Let me start with the comparison of ontologies vs. comparison as ontology contrast. Some of my writing on the polemical theme of ‘ontology’ may occasionally have given room for aligning my position with the comparison of ontologies pole. For example, the stark contrast—deliberately provocative and therefore outrageously simplistic—multiculturalism vs. multinaturalism could be read as positing a great divide between two major ontological orientations, namely, Modern Western and Amerindian, conceived as two opposed sets of fundamental presuppositions about the nature of ‘reality’ and the place (epistemological as well as ontological) of humans in ‘it.’ In fact, though, the contrast was meant to distinguish two opposed ‘anthropological’ outlooks, two anthropological ontologies, by which I mean at the same time, two ways of doing anthropology: (i) Western ortho-anthropology as the description of the sundry different human-cultural-spiritual ways of representing ONE reality-nature, versus another anthropology, the Amerindian hetero-anthropology as the nonrepresentational positing of a multiplicity of corporally-specific perspectives which ‘constitute’ reality (in the sense of being the stuff the world is made of); (ii) anthropology (“our” anthropology, the academic discipline engaged in the business of describing diverse human representations of both human and nonhuman Nature) as engaged in the comparison of itself—Western metaphysically-derived conceptual repertoire—with alien anthropologies that may turn out to have an entirely different set of conceptual presuppositions, including, crucially, different notions of anthropos and logos. This gives anthropology an ‘ontological’ quality or ‘twist’ (‘turn’), insofar as its task is now to compare conceptual ‘languages’ without presupposing a concept-independent, ‘exterior’ ontological realm of realia that could serve as a baseline for the comparison. Anthropology becomes a game with no superior arbiter or impartial umpire, a game whose rules change as the game goes on. “Comparison as ontology,” to my mind, means two things (again): (i) comparison collapses the epistemic and the ontological (word and world), as it sets as its task that of establishing the conditions under which alien concepts can be “translated” into our own conceptual language without having to leave the world they necessarily’ carry with them at the door; (ii) ‘ontologies’ only become understandable as continuous and contingent ‘variations’ (i.e., ‘comparisons’) of one another, both within and across those objects we used to call cultures, societies, cosmologies or intellectual traditions. Anthropology, through comparison, becomes a method, or perhaps a process, to discover difference and variation and to transform difference and variation into what there is: ‘being’ is being able to be other, being virtually other. Anthropology is not in the business of adjudicating on the correctness, consistency or even the proper grammatical ‘number’ of the noun ‘ontology’ (ontological monism vs. ontological pluralism, ontological monarchy vs. ontological democracy, etc.). In that sense, it is ontologically agnostic, or better yet, anarchist. ‘Ontology,’ as far as anthropology as an ontologically comparative endeavor is concerned, starts from, and with, the methodological principle according to which we do not know what being is without having first engaged in ethnographic (ontographic) fieldwork. ‘Ontology’ thus becomes an ‘outdoor science’ like field ecology or natural history.

GB: The ontological turn of anthropology seems to share some ground with various forms of contemporary thought characterized by the attempt to undo or overcome the transcendental turn of modern metaphysics and to resuscitate the precritical notion of reality in which humans are not subjects but one of many “actants.” If Kant took the category of thought to be universal, the critical project of structuralist and poststructuralist thought can be said to have sought to relativize and historicize this position, by maintaining that there are multiple and irreducible ways to apprehend the world, relative to cultures, subject positions, or “worldviews.” Where Kant built his theory in terms of the features intrinsic to any cognition, structuralism and poststructuralism often externalized this view, casting it in terms of discursive regimes, ideology, or linguistic practices. Despite its alleged departure from Kant, some forms of structuralist and poststructuralist thought in anthropology seem to have nonetheless reproduced the logic of its position, claiming that the real is accessible only as mediated or constituted by discourse, epistemic constructions, etc. How do you cast your projects with regard to the complex debt of anthropology towards Kant?

MH: Following Durkheim, the trend in anthropology is to think of ‘categories’ as objects of empirical investigation: the ‘x’ take time to be linear, while the ‘y’ take time to be circular, and our job as anthropologists is to collect data on the x and the y in order cogently to describe their respective ‘temporalities’ (understood ipso facto, as you say, as relative and historical). One could note that the term ‘category’ here tends to be used loosely, to include much more than Aristotle or Kant ever had in mind, including such things as concepts, assumptions, generalizations, schemes of classification, patterns of reasoning or even behavior—pretty much anything… Still, the point is that this idea that something even remotely equivalent to Kantian “categories of the understanding” could be rendered as objects of empirical description raises a majorly ‘critical,’ if you like, query: if these empirically describable categories are meant to vary from one place or time to another, then what kind of categories would WE need in order to describe them in any given case? There appear to be two options. Either we have (or could develop) a repertoire of “categories of the anthropological understanding,” let’s call them, which could encompass all possible empirical variation of indigenous ‘categories’; or, alternatively, we would have to accept that, in some instances at least, the variation we wish to describe might be inconsistent with the categories we can use to describe it, in which case our attempt to describe our varying ethnographic object would have to involve developing new anthropological categories for understanding it in any given case. The former option is interesting—it has something deliciously Aristotelian about it: the terms of anthropological description attain the status of ‘categories’ inasmuch as they become irreducible—i.e., the lowest common denominator of all the putatively varying indigenous ones. But note that this renders ethnographic variation superficial and derivative. The latter option, on its part, has something Kantian about it, since it adds to the task of ethnographic description a component reminiscent of transcendental deduction: what categories (though now I’d just call them concepts) do I as an anthropologist need in order sensibly (e.g., free from logical inconsistency) to describe my ethnographic material? E.g., what concept of time do I need in order to describe the putatively circular time of the x? What kind of thing must time be for me to be able to say, for instance, that the x take past events to return to the present, when, I take it, the whole point about past events (what makes them ‘past’) is that they’re gone forever? To the extent that there can be as many of these kinds of Kantian-sounding questions as there are ethnographic variations of such a concept as time, say, what we have here is a rather strange, but to me exciting, Kantian ‘relativism,’ with categories of the anthropological understanding multiplying in different directions through their exposure to ethnographic variation. And note, of course, that the manner of transcendental questioning that is required to set this multiplication in motion is ontological in the sense I indicated earlier: what must I take time to be (or not to be, for that matter) for my description of the ethnography of the x to make sense, is the question.

EVC: If we are (if we were) to stick to the Kantian project, I would observe that anthropology has always concentrated itself on Kant’s transcendental aesthetics and transcendental analytics (forms of sensibility plus categories of the understanding), even as the anthropological turn moved, starting with the French Sociological School, towards the empiricization and historicization of  the transcendental (social time and social space, mana as the ancestor of the category of causality, Foucaldian epistemes as modes of the historical a priori, etc.). So, just to carry on with the Kantian vocabulary, I would suggest it is about time anthropology starts to confront head-on the empirical (ethnographic) deconstruction of the transcendental dialectics. We should be studying the ideas of savage reason, not just the “categories of the savage understanding (mind).” This would imply, of course, rejecting the purpose of transcendental dialectics, namely, the famous critical ban on “going beyond sense-experience.” In other words, we must, in a sense, revert to a precritical ‘critical’ attitude towards the Kantian project, in order to reestablish “the transcendent” as a legitimate domain of speculation. After all, what are “institutions” like shamanism or divination, if not sui generis modes of speculative thought (shamanism as speculative oneirism, divination as speculative logics) that delve into what Kant dubbed the transcendent realm? One way to go beyond the Nature/Culture dualism that has been plaguing us since the dawn of Modernity is to thoroughly reconceptualize the notion of Supernature along non-Western, non-scholastic, non-Christian lines. Latour’s “mode of existence” that he calls Metamorphosis (see his An Inquiry into Modes of Existence) may offer an interesting starting point, the more so as—this is explicitly acknowledged by Latour himself—such “mode” is ontologically devalued or underdeveloped by the “Moderns” (it is reduced to the interiority of the subject’s unconscious), on the one hand, and is totally heterogeneous to the mode he calls Religion, on the other hand.

GB: If indeed the role of anthropology is not the study of representations anymore, but rather the study of ontologies, what is the role of the natural sciences (historically determined, as a matter of fact, by a Naturalist ontology) in your project?

MH: Much as I respect natural science in its own terms (and for all sorts of other reasons, albeit not uncritically), for me the most urgent task for anthropology in the twenty-first century is to stop thinking of itself by analogy to it. Very roughly speaking, I take the whole image of anthropologists as charters of the sociocultural world around us, in all its ‘complexity’ and ‘variation’ and so on, to be exhausted after an excellent innings, as the Brits say, of a century. There are myriad other ways to think of what we do as anthropologists, and exploring them in no way entails (necessarily at least) parting company with such cherished things as truth, method, discovery, rigor, argument, analysis, consistency, cogency, or whatever it is we think marks us out as serious people doing work that’s worth doing. Take for example the image of anthropology we’ve been discussing. Before I compared it with Kant because you invited us to do that, but I could just as well compare it with the rigors of, say, Grotowskian theater-making. Imagine that, roughly, as the task of training actors’ bodies to be so versatile as to be able to take on any theatrical character a play might throw at them: every bodily habit that gets in the way of taking on, say, Agamemnon must be painstakingly worked on in order to allow the space for that character to be embodied and, thus, expressed. It is just that order of rigor, and the kind of truth that emerges from it, that anthropology too can aspire to: let’s train our thinking—or whatever it is we use as anthropologists—to be so versatile as to be able to ‘take on’ (i.e., express, convey) any ethnographic materials life throws at us. And that, I suggest, involves above all working on our habits of thinking about what the things we encounter in our ethnographies might ‘be’—ontological habits that tend to get in the way.

EVC: Here I will just take the liberty to quote (not verbatim) something that I heard from Patrice Maniglier, one of the few contemporary philosophers who takes anthropology very seriously. Maniglier suggests that what physics has represented to all the other natural sciences and, more importantly, to philosophy as well (from metaphysics to political philosophy to ethics) since the Scientific Revolution of the seventeenth century—namely, the role of a Model, a sort of epistemic Ego-ideal—is bound to be played by anthropology in the present century. With its central intuition that “being is being a variant” (Maniglier is a Deleuzian structuralist, like yours truly), anthropology sets a conceptual agenda that may—just may, of course—become crucial for the Anthropocene epoch.

Training at Jerzy Grotowski's "Laboratorium" in Wrozlaw, Poland, 1972.

Training at Jerzy Grotowski’s “Laboratorium” in Wrozlaw, Poland (1972, all rights reserved).

GB: This issue of the journal mobilizes Castalia, a fictional province dedicated to the synthesis of knowledge in Hermann Hesse’s The Glass Bead Game, as a hyperbolic image of the ways in which humanism figured the project of reason of Western modernity. We’d like you to elaborate on the situation of what you call “comparative ontography” (and the distribution of agency therein) in relation to several threads of contemporary philosophy proposing to widen or reduce the circle of the human, namely antihumanism (according to which the human subject, classically understood, is not necessarily to be conceived as the privileged bearer of rationality), transhumanism (according to which humans are not the only rational agents), and post-humanism (according to which rationality extends itself beyond the human’s biological and symbolic terms). Could you situate your projects with regard to those schematic (albeit predominant) polarities of contemporary thought?

MH: It’s a good question because I think a lot of people tend to see a basic continuity between the kind of approach we have been discussing here and the broad trend in social theory at large (including many anthropologists, STS scholars as well as philosophers) to play around with, or even erase, the distinction between humans and nonhumans, to use Bruno Latour’s language. For me, the alliance is a fragile one. To the extent that recent revisions of the human/nonhuman distinction express a broader willingness to deny or otherwise experiment with ontological axioms in general, their resonance with the kind of anthropological program we have been discussing here is certainly there. But again, there’s also a basic distinction to be made. In my reading, a lot of the works associated with the various terms you mention—’posthumanism’ perhaps captures the trend as a whole—are revisionary: their aim is to replace the ontological framework of humanism with something else, presumably something better. The ecological politics and ethics that often drives this agenda bears this out—the idea being that the ‘modern’ anthropocentrism of humans vs. nonhumans constitutes the ontological premise of our current ecological malaises, and must therefore be urgently replaced by a less self-centered metaphysics that will create the ontological conditions for a better future. For me—and here I may be in disagreement with Eduardo—these questions of political, ethical or indeed ecological virtue are external to the categorical imperative of anthropology (I’m joking with Kant), which, as we have said already, is that of keeping constitutively open the horizon of what is conceptually possible. That is simply the decision of refusing to decide, much less legislate, as to which ontological regime is better or truer, e.g., is it humanism or anti/trans/post-humanism? Our job as anthropologists, according to this image, is to find ways not to decide on these things, in order the better to be able to allow the materials we study—our ethnographic exposures –to decide, as it were, for themselves. As human beings (or anti/trans/post-human beings for that matter) we may perhaps decide what we want. But as anthropologists we can’t. We are meant to be conduits, not legislators, and our job is to help to express things, not to recommend them.

EVC: Well, not wishing to start a long (uncontrolled) equivocal dialogue with Martin here, I’ll say that no effort to problematize the concepts of anthropos and logos (and the institutional apparatuses related to them) is foreign to anthropology as I (we?) understand it. I am in deep sympathy with currents of thought that seek to finish with the state of ontological exception (self-)granted to “Man” in the Modern era, the catastrophic political and ecological consequences of which is unnecessary to dwell upon. When the chips are down, it is not so easy to separate one’s role as an anthropologist from one’s predicament as an Earth being, a Terran. I would not wish anthropology to become a new transcendental arbiter that sets no legislation as the new rule of law. I am of the opinion that those new trends—not all of them, by any means—are actually allies of the anthropological position I (we?) defend. Political allies, I mean, i.e., people with whom one can find points of tactical, if not strategical, agreement. But epistemology and politics are just two names for the same thing. An anti-equivocal synonymy, if you will.

GB: Together with Morten Axel Pedersen, you wrote a text titled The Politics of Ontology: Anthropological Positions,4 in which you claim that “the politics of ontography resides not only in the ways in which it may help promote certain futures, but also in the way that it ‘figurates’ the future in its very enactment.” Can you explain the main ways in which the move away from representationalism and the mobilization of the register of “figuration” can have traction in the domain of politics?

MH: I guess my question for anthropology is how it can be ‘political’ without importing from beyond its own terms of inquiry criteria of good and bad politics, let alone ready-made definitions of what counts as being political in the first place. Can we in fact refuse to decide on anything beyond what our ethnographic engagements require, as I have suggested, without thereby being profoundly apolitical or even conservative? My own answer takes the form of something rather reminiscent of the Cartesian method of doubt. (We did Kant already, so why not invoke as approvingly that other supposed philosophical baddie?) If the no a priori ontological decisions imperative serves to denude the analytical terrain in order to allow space for ethnographically driven alterity to operate, then it works rather like Cartesian doubt: Don’t take anything for granted. But where Descartes ended up, of course, was in having to take for granted the very act of refusing to do so. Now, as we have seen, the anthropological ‘cogito’ too, if you like, has a structure of its own: it is ‘constituted’ by keeping open the possibility of difference. So the question then is what ‘politics,’ if any, such a manner of inquiry may itself instantiate. In other words, if we accept that anthropology is par excellence in the business of giving expression to difference, then whatever characterization we give that as a political act would count as having emerged directly from the terms of anthropological inquiry, rather than being imported from elsewhere. In the piece you mentioned in your question we pointed out that this way of thinking makes anthropology an inherently antiauthoritarian discipline. And we even used the word ‘revolutionary’ in that context, which I think annoyed some people who prefer to measure revolutionary credentials on the scale of activism and commitment to already established revolutionary causes. To be perfectly honest, having come of political age in Greece in the 1980s (Alexis Tsipras was in the year below me in school), this kind of revolutionary brow-beating holds no excitement for me. Certainly, as far as anthropology goes, I am much more interested in a different sense of political action, which is expressed most forcefully by Adorno in his account of the musical significance of Schoenberg’s serialism. At issue there was the instantiation of radical difference—and in Schoenberg’s case certainly wholesale musical revolution—in the form of one’s own labor. And that, in my own case, I guess would be anthropology: the form of its labor, which we have been discussing throughout this interview.

EVC:  Well, no quarreling with that. I just do not want to make anthropologists sound like those scientists who say that ‘science’ (and technology, the ultimately important practical consequence of science) is politically neuter, everything ‘political’ hinging on the uses others (non-scientists) make of it. The non-decouplable political aspect of anthropology lies not in its vouching for this or that ontological outlook, but in keeping all (ontological) options open as a matter of principle. If it belongs to the essence of anthropological ‘objects’ (very broadly speaking) the capability of being otherwise (sensu Povinelli), this means change—political, theological, economic change and what have you—is always on the table. But of course, just as “you don’t need a weatherman to know which way the wind is blowing,” as Bob Dylan famously sang, then we sure do not need anthropologists to tell us that European-born capitalist civilization is in its death throes, and is taking the planet with it to a very bad place, as far as many species (including our own) are concerned. Anthropologists just help us focus our gaze elsewhere (‘otherwise’) and show what is out there—show that there are a number of other possible worlds out there.

Interview conducted for Glass Bead by Vincent Normand.