Cannibal metaphysics: Amerindian perspectivism

With an introduction by Peter Skafish

RP 182 () / Article

From anthropology to philosophy: Introduction to Eduardo Viveiros de Castro

Peter Skafish

Can anthropology be philosophy, and if so, how? For philosophers, the matter has been and often remains quite simple: anthropology’s concern with socio-cultural and historical differences might yield analyses that philosophy can put to use (provided that it condescends to examine them), but only rarely does anthropology conceive its material at a level of generality or in relation to metaphysical issues in their positivity that would allow it to really do philosophy, especially of an ontological kind. Anthropologists, on the other hand, tend not to disagree, whether out of a preference for local problems or from the more canny recognition that even the best philosophers prove quite adept at mistaking modern ideological values for transcendental concepts. Such perspectives, however, are proving outmoded in the face of a now sizable group of thinkers, ranging from Bruno Latour and Isabelle Stengers to Marilyn Strathern to François Jullien, whose questions, concepts, objects and methods belong in different ways to both anthropology and philosophy, and who moreover propose that certain aspects of anthropology – analyses of scientific practices, knowledge of cultural variation, and an old thing called structuralism – are key to a new metaphysics as empirical, pluralistic and comparative as transcendental, unifying and general.

One of the chief instigators of this new approach is the Brazilian anthropologist and now philosopher Eduardo Viveiros de Castro, whose 2009 Métaphysiques Cannibales perhaps marks the first case of an ‘actual’ anthropologist, in the disciplinary sense, explicitly undertaking such a philosophy. (And ‘real’ philosophers agree: the book was published alongside those of Étienne Souriau, Tristan Garcia and Graham Harman in the Presses Universitaires de France series, edited by Quentin Meillassoux, Patrice Maniglier and others, entitled ‘MétaphysiqueS’.) A native of Rio (and carioca irony) who did fieldwork with a Northeastern Amazonian Indian group known as the Arawaté, Viveiros de Castro is widely known in social anthropology for showing that what falls under the domain of ‘social’ and ‘human’ relations for such Amazonian peoples is so broad – animals, plants, spirits are all conceived as persons – that modern distinctions between nature and culture, animals and humans, and even descent and marriage ties are effectively inverted. A generalized ‘potential’ or ‘virtual affinity’ obtains (‘affinity’ is the kinship term for relations established through marriage) wherein beings, because they are all initially related and thus ‘social’, must be established as ‘natural’ and substantial in the same way that conventional, cultural ones elsewhere have to be.

The means of doing that, from hunting to ritual to shamanism, involve contending with the additional fact that every relatable entity is conceived as having, whatever its bodily form, a soul – intentionality and apperception – of a ‘human’ character, and that all beings thus perceive themselves as humans, and other beings as either animals or cultural artefacts. Jaguars, for example, are thought to see themselves as humans, to see humans as human prey like peccarys and monkeys, and their own food as that of humans (blood as manioc beer). Successfully negotiating one’s relations with other beings therefore requires adopting their perspectives, as shamans do when they become animals, in order to know what they see things as being, and thereby in turn anticipating and knowing them as definite beings. What emerges from this ‘perspectivist’ universe, Viveiros de Castro continually emphasizes, is an ontology that reverses the terms of one of our most fundamental metaphysical dualisms. Because perspectivism confers on all beings the same ontological status, and distinguishing between them requires knowing the differences between their bodies, ‘culture’ becomes the underlying domain uniting beings in Amazonia and nature the differential, separating one. A ‘multinaturalism’ effectively prevails that is the converse of our naturalist multiculturalism.

Yet how Amazonian cosmology might function as a metaphysics for us goes well beyond its upsetting of our certainties about nature and culture and right to the core of contemporary philosophical debate. Although Métaphysiques Cannibales might appear to confirm, as nothing else has, that the forms of thought of indigenous peoples accord a central role to relations, virtualities and becomings in the way Deleuze claimed, Amerindian thought flips more than just the metaphysical poles by which modernity orients itself but also the very terms by which even critical, minor philosophies have conceptualized them. Viveiros de Castro’s frequent recourse to structuralist notions of inversion, duality, analogy and reciprocity primarily derives not from the influence Lévi-Strauss holds on his thinking but from the fact that the latter elaborated structuralism on the basis of North and South American materials rooted in conceptual schemes that are anything but a metaphysics of logocentrism or transcendence. When set next to Deleuze (especially the Deleuze of the remaining Deleuzeans), Amerindian cosmology can arguably be seen as regarding as actual the kinds of differential, relational realities Deleuze saw as only virtual, and phenomenal entities as being much more insubstantial than present. Amerindian metaphysics can effect for us a (quasi-dialectical) transformation of Deleuze’s thought that thereby connects both of them to the still poorly recognized challenge to the philosophies of difference of another ‘transformist’ thinker, Catherine Malabou, who has argued that the history of philosophy should be treated as a transformation group in Lévi-Strauss’s sense, and thus as subject to constant reversals.

An examination of these consequences and their stakes, however, can only come after a reading of Métaphysiques Cannibales, whose first two chapters are presented here as a single text, with the hope that the inventiveness, passion and vision of their author might help further decentre ‘philosophy’ so that it may finally become the sort of decolonial, polytraditional endeavour that it will have to be to retain its relevance in the future ‘multiversal’ world that has already begun to arrive.

Cannibal metaphysics

Amerindian perspectivism

Eduardo Viveiros de Castro

I once had the intention of writing a book that would have been something of a homage to Deleuze and Guattari from the point of view of my discipline; it would have been called Anti-Narcissus: Anthropology as Minor Science. The project was to characterize the conceptual tensions animating contemporary anthropology. From the moment I had the title, however, the problems began. I quickly realized that the project verged on complete contradiction, and the least misstep on my part could have resulted in a mess of not-so anti-narcissistic provocations about the excellence of the positions to be professed.

It was then that I decided to raise the book to the rank of those fictional works (or, rather, invisible works) that Borges was the best at commenting on and that are often far more interesting than the visible works themselves (as one can be convinced of from reading the accounts of them furnished by that great blind reader). Rather than write the book itself, I found it more opportune to write about it as if others had written it. Cannibal Metaphysics is therefore a beginner’s guide to another book, entitled Anti-Narcissus, that because it was endlessly imagined, ended up not existing – unless in the pages that follow.

The principal objective of Anti-Narcissus is, to mark the ‘ethnographic’ present in my fashion, to address the following question: what do anthropologists owe, conceptually, to the people they study? The implications of this question would doubtlessly seem clearer were the problem approached from the other end. Are the differences and mutations internal to anthropological theory principally due to the structures and conjunctures of the social formations, ideological debates, intellectual fields and academic contexts from which anthropologists themselves emerge? Is that really the only relevant hypothesis? Couldn’t one shift to a perspective showing that the source of the most interesting concepts, problems, entities and agents introduced into thought by anthropological theory is in the imaginative powers of the societies – or, better, peoples and collectives – that they propose to explain? Doesn’t the originality of anthropology instead reside there – in this always-equivocal but often fecund alliance between the conceptions and practices coming from the worlds of the so-called ‘subject’ and ‘object’ of anthropology?

The question of Anti-Narcissus is thus epistemological, meaning political. If we are all more or less agreed that anthropology, even if colonialism was one of its historical a prioris, is today in the process of completing its karmic cycle, then we should also accept that the time has come to radicalize the reconstitution of the discipline by finishing the job. Anthropology is ready to fully assume its new mission of being the theory/practice of the permanent decolonization of thought.

But perhaps not everyone is in agreement. There are those who still believe that anthropology is the mirror of society. Not, certainly, of the societies it claims to study – of course no one is as ingenuous as that anymore (whatever…) – but of those in whose bowels its intellectual project was engendered. We all know the popularity in some circles of the thesis that anthropology, because it was supposedly exoticist and primitivist from birth, could only be a perverse theatre where the Other is always represented or invented according to the sordid interests of the West. No history or sociology can camouflage the complacent paternalism of this thesis, which simply transfigures the so-called others into fictions of the Western imagination having no say in the matter. Doubling such a subjective phantasmagoria with an appeal to the dialectic of the objective production of the Other by the colonial system quite simply piles insult upon injury, proceeding as if every ‘European’ discourse on European traditional peoples serves only to illumine our ‘representations of the other’ and thereby makes a certain theoretical postcolonialism the ultimate stage of ethnocentrism. By always seeing the Same in the Other, by thinking that under the mask of the other it is always just ‘us’ contemplating ourselves, we end up contenting ourselves with a mere shortcut to the goal.

On the contrary, a veritable anthropology, as Patrice Maniglier has put it, ‘returns to us an image in which we are unrecognizable to ourselves’ since what every experience of another culture offers us is an occasion to experiment with our own – and far more than an imaginary variation, such a thing is a putting into variation of our imagination.1 We have to grasp the consequences of the idea that those societies and cultures that are the object of anthropological research influence, or, to put it more accurately, co-produce the theories of society and culture that it formulates. To deny that would be to accept a unique kind of constructivism that, at the risk of imploding in on itself, inevitably circles back to the same easy story: anthropology always constructed its objects badly, but when the authors of the critical denunciations put pen to paper, the lights came on, and it begin to construct them correctly. In effect, an examination of the readings of Fabian’s Time and Other (1983) and its numerous successors makes it impossible to know if we are once more faced with a spasm of cognitive despair before the inaccessibility of the thing in itself or before the old illuminist thaumaturgy where an author purports to incarnate a universal reason come to scatter the darkness of superstition – no longer that of indigenous peoples but of the authors who proceeded him. The de-exoticization of the indigenous, which is not so distant from all this, has the countereffect of a rather strong exoticization of the anthropologist, which is likewise nearby. Proust, who knew a thing or two about time and the other, would have said that nothing appears older than the recent past.

Blocking this type of epistemo-political reflex is one of the principal objectives of Anti-Narcissus. In order to realize it, nevertheless, the last thing we should do is commit anthropology to a servile relationship with economics or sociology whereby it would be made, in a spirit of obsequious emulation, to adopt the metanarratives promulgated by these two sciences and whose principal function would seem to be the repressive recontextualization of the existential practice(s) of all the collectives of the world in terms of ‘the thought collective’ of the analyst.2 The position argued here, on the contrary, affirms that anthropology should remain in open air, continue to be an art of distances keeping away from the ironic recesses of the Occidental soul (while the Occident may be an abstraction, its soul definitely is not), and remain faithful to the project of externalizing reason that has always so insistently pushed it, much too often against its will, outside the stifling bedroom of the Same. The viability of an authentic endoanthropology, an aspiration that has for numerous reasons come to have first priority on the disciplinary agenda, thus depends in a crucial way on the theoretical ventilation that has always been favoured by exoanthropology – a science of fieldwork in a truly important sense.

The aim of Anti-Narcissus, then, is to illustrate the thesis that every non-trivial anthropological theory is a version of an indigenous practice of knowledge, all such theories being situatable in strict structural continuity with the intellectual pragmatics of the collectives that historically occupied the position of object in the discipline’s gaze.3 This entails outlining a performative description of the discursive transformations of anthropology at the origin of the internalization of the transformational condition of the discipline as such, which is to say the (of course theoretical) fact that it is the discursive anamorphosis of the ethnoanthropologies of the collectives studied. Using the example, to speak of something close at hand, of Amazonian notions of ‘perspectivism’ and ‘multinaturalism’ – the author is an Americanist ethnologist – the intention of Anti-Narcissus is to show that the styles of thought proper to the collectives that we study are the motor force of anthropology.

A more profound examination of these styles and their implications, particularly from the perspective of the elaboration of an anthropological concept of the concept, should be capable of showing their importance to the genesis, now under way, of a completely different conception of anthropological practice. In sum, a new anthropology of the concept that can counter-effectuate a new concept of anthropology, after which the descriptions of the conditions of the ontological self-determination of the collectives studied will absolutely prevail over the reduction of human (as well as nonhuman) thought to a dispositif of recognition: classification, predication, judgement, and representation… Anthropology as comparative ontography – that is the true point of view of immanence.4 Accepting the importance and opportunity of this task of thinking thought otherwise is to implicate oneself in the project of an elaboration of an anthropological theory of the conceptual imagination attuned to the creativity and reflexivity of every collective, human or nonhuman.


Thus the intention behind the title of the book being described is to suggest that our discipline is already in the course of writing the first chapters of a great book that would be like its Anti-Oedipus. Because if Oedipus is the protagonist of the founding myth of psychoanalysis, our book proposes Narcissus as the candidate for patron saint or tutelary spirit of anthropology, which (above all in its so-called ‘philosophical’ version) has always been a little too obsessed with determining the attributes or criteria that fundamentally distinguish the subject of anthropological discourse from everything it is not: him (which is to say us), the non-Occidental, the non-modern, the nonhuman. In other words, what is it that the others ‘lack’ that constitutes them as non-Western and nonmodern? Capitalism? Rationality? Individualism and Christianity? (Or, perhaps more modestly, pace Goody: alphabetic writing and the marriage dowry?) And what about the even more gaping absences that would make certain others nonhumans (or, rather, make the nonhumans the true others)? An immortal soul? Language? Labour? The Lichtung? Prohibition? Neoteny? Metaintentionality?

All these absences resemble each other. For, in truth, taking them for the problem is exactly the problem, which thus contains the form of the response: the form of a Great Divide, the same gesture of exclusion that made the human species the biological analogue of the anthropological West, confusing all the other species and peoples in a common, privative alterity. Indeed, asking what distinguishes us from the others – and it makes little difference who ‘they’ are, since what really in that case matters would just be ‘us’ – is already a response.

The point of challenging the question, ‘what is proper to Man?’, then, is absolutely not to say that ‘Man’ has no essence, that his existence precedes his essence, that the being of Man is freedom and indetermination, but to say that the question has become, for all-too-obvious historical reasons, one that it is impossible to respond to without dissimulation, without, in other words, continuing to repeat that the chief property of Man is to have no final properties, which apparently gives it unlimited rights to the properties of the other. This response from ‘our’ intellectual tradition, which justifies anthropocentrism on the basis of this human ‘impropriety’, is that absence, finitude and lack of being [manque-à-être] are the distinctions that the species is doomed to bear, to the benefit (as some would have us believe) of the rest of the living. The burden of man is to be the universal animal, he for whom there exists a universe, while nonhumans, as we know (but how in the devil do we know them?), are just ‘poor in world’ (not even a lark…). As for non-Occidental humans, something quietly leads us to suspect that where the world is concerned, they end up all the same reduced to the smallest portion of it. We and we alone, the Europeans, would be the realized humans, or, if you prefer, the grandiosely unrealized, the millionaires, accumulators, and configurers of worlds. Western metaphysics is truly the fons et origio of every colonialism.

In the event that the problem changes, so too will the response. Against the great dividers, a minor anthropology would make small multiplicities proliferate – not the narcissism of small differences but the anti-narcissism of continuous variations; against all the finished-and-done humanisms, an ‘interminable humanism’ challenging the constitution of humanity into a separate order.5 I will re-emphasize it: said anthropology would make multiplicities proliferate. Because it is not at all a question, as Derrida opportunely recalled,6 of preaching the abolition of the borders that unite/separate sign and world, persons and things, ‘us’ and ‘them,’ ‘humans’ and ‘nonhumans’ – easy reductionisms and mobile monisms are as out of the question as fusional fantasies – but rather of ‘unreducing’ [irréduire] (Latour) and undefining them, by bending every line of division into an infinitely complex curve. It is not a question of erasing the contours but of folding and making them dense, of diffracting and making them iridescent. ‘This is what we are getting at: a generalized chromaticism.’7 Chromaticism as the structuralist vocabulary with which the agenda for its posterity will be written.


The draft of Anti-Narcissus is beginning to be completed by certain anthropologists who are responsible for a profound renewal of the discipline. Although they are all known figures, their work has not at all received the recognition and diffusion it deserves – even, and especially in one case, in their countries of origin. I am referring in the latter instance to the American Roy Wagner, who should be credited with the extremely rich notion of ‘reverse anthropology’, a dizzying semiotics of ‘invention’ and ‘convention’, and his visionary outline of an anthropological concept of the concept; but I am referring also to the English anthropologist Marilyn Strathern, to whom we owe the deconstruction/potentiation of feminism and anthropology, just as we do the central tenets of an indigeneous aesthetic and analysis forming the two flanks of a Melanesian anti-critique of Occidental reason, and even the invention of a properly post-Malinowskian mode of ethnographic description; and to that Bourguignon Bruno Latour and his transontological concepts of the collective and the actor-network, the paradoxical movement of our never-having-been modern, and the anthropological re-enchantment of scientific practice. And to these can be added many more recently arrived others, who will go unnamed since it would be largely impossible to do otherwise without some injustice, whether by omission or commission.

But well before all of them (cited or not) there was Claude Lévi-Strauss, whose work has a face turned towards anthropology’s past, which it crowns, and another looking into and anticipating its future. If Rousseau ought, by the former’s account, to be regarded as the founder of the human sciences, then Lévi-Strauss deserves to be credited not only with having refounded them with structuralism but also with virtually ‘un-founding’ them by pointing the way towards an anthropology of immanence, a path he only took ‘like Moses conducting his people all the way to a promised land whose splendor he would never behold’ and perhaps never truly entered.8 In conceiving anthropological knowledge as a transformation of indigenous practice – ‘anthropology seeks to elaborate the social science of the observed’ – and the Mythologiques as ‘the myth of mythology,’ Lévi-Strauss laid down the milestones of a philosophy to come positively marked by a skin of interminability and virtuality.9

Claude Lévi-Strauss as the founder, yes, of poststructuralism… Just a little more than ten years ago, in the postface to a volume of L’Homme devoted to an appraisal of the structuralist heritage in kinship studies, the doyen of our craft made this equally penetrating and decisive statement:

One should note that, on the basis of a critical analysis of the notion of affinity, conceived by South American Indians as the point of articulation between opposed terms – human and divine, friend and foe, relative and foreign – our Brazilian colleagues have come to extract what could be called a metaphysics of predation.… Without a doubt, this approach is not free from the dangers that threaten any hermeneutics: that we insidiously begin to think on behalf of those we believe to understand, and that we make them say more than what they think, or something else entirely. Nobody can deny, nonetheless, that it has changed the terms in which certain big problems were posed, such as cannibalism and headhunting. From this current of ideas, a general impression results: whether we rejoice in or recoil from it, philosophy is once again centre stage. No longer our philosophy, the one that my generation wished to cast aside with the help of exotic peoples; but, in a remarkable reversal [un frappant retour des choses], theirs.10

The observations also marvellously sum up, as we will see, the content of the present essay, which was in fact written by one of these Brazilian colleagues.11 Indeed, not only do we take as one of our ethnographic axes this properly metaphysical use South American Indians make of the notion of affinity, but we sketch, moreover, a reprise of the problem of the relation between, on the other hand, the two philosophies evoked by Lévi-Strauss in a mode of non-relation – ‘ours’ and ‘theirs’ – and, on the other, a philosophy to come that structuralism contains.

For whether we rejoice in or recoil from it, philosophy really is what is at stake… Or, rather, the re-establishment of a certain connection between anthropology and philosophy via a new review of the transdisciplinary problematic that was constituted at the imprecise frontier between structuralism and poststructuralism during that brief moment of effervescence and generosity of thought that immediately preceded the conservative revolution that has, in the last decades, showed itself particularly efficacious at transforming the world, both ecologically and politically, into something perfectly suffocating.

A double trajectory, then: an at once anthropological and philosophical reading informed, on the one hand, by Amazonian thought – it is absolutely essential that what Taylor has stressed are ‘the Amerindian bases of structuralism’ be recalled – and, on the other, by the dissident structuralism of Gilles Deleuze.12 The destination, moreover, is also double, comprising the ideal of anthropology as a permanent exercise in the decolonization of thought and a proposal for another means besides philosophy for the creation of concepts.

But in the end, anthropology is what is at stake. The intention behind this tour through our recent past is in effect far more prospective than nostalgic, the aspiration being to awaken certain possibilities and glimpse a break in the clouds through which our discipline could imagine, at least for itself qua intellectual project, another denouement (to dramatize things a bit) than mere death by asphyxia.


Such a requalification of anthropology was what Tânia Stolze Lima and I wanted to contribute to when we proposed the concept of Amerindian perspectivism as the reconfiguration of a complex of ideas and practices whose power of intellectual disturbance has never been sufficiently appreciated (even if they found the word relevant) by Americanists, despite its vast diffusion in the New World. To this we added the synoptic concept of multinaturalism, which presented Amerindian thought as an unsuspected partner, a dark precursor [un précursor sombre], if you will, of certain contemporary philosophical programmes, such as those developing around theories of possible worlds, those from the start outside the vicious dichotomies of modernity, or still those that, having registered the end of the hegemony of the sort of critique that demands an epistemological response to every ontological question, are slowly defining new lines of flight for thought under the rallying cries of transcendental empiricism and speculative realism.

The two concepts emerged following an analysis of the cosmological presuppositions of ‘the metaphysics of predation’ evoked above. We found that this metaphysics, as can be deduced from Lévi-Strauss’s summary, reaches its highest expression in the strong speculative yield of those indigenous categories denoting matrimonial alliance, phenomena that I translated with yet another concept: virtual affinity. 13 Virtual affinity is the schematism characteristic of what Deleuze would have called the ‘Other-structure’ of Amerindian worlds and is indelibly marked by cannibalism, which is an omnipresent motif in their inhabitants’ relational imagination. Interspecies perspectivism, ontological multinaturalism and cannibal alterity thus form the three aspects of an indigeneous alter-anthropology that is the symmetrical and inverse transformation of Occidental anthropology – as symmetrical in Latour’s sense as inverse in that of Wagner’s ‘reverse anthropology’. By drawing this triangle, we can enter the orbit of one of the philosophies of ‘the exotic peoples’ that Lévi-Strauss opposed to ours and attempt, in other words, to realize something of the imposing programme outlined in the fourth chapter, ‘Geophilosophy’, of What Is Philosophy?… even if it will in each case be at the price – but one we should always be ready to pay! – of a certain methodological imprecision and intentional ambiguity.


Our work’s perfectly contingent point of departure was the sudden perception of a resonance between the results of our research on Amazonian cosmopolitics (on the perspectivist multiplicity intrinsic to the real) and a well-known parable on the subject of the conquest of the Americans recounted by Lévi-Strauss in Race and History:

In the Greater Antilles, a few years after the discovery of America, while the Spaniards were sending out Commissions of investigation to discover whether or not the natives had a soul, the latter spent their time drowning white prisoners in order to ascertain, by long observation, whether or not their bodies would decompose.14

The author perceived in this conflict between the two anthropologies a baroque allegory of the fact that one of the typical manifestations of human nature is the negation of its own generality. A kind of congenital avarice preventing the extension of the predicates of humanity to the species as a whole appears to be one of its predicates. In sum, ethnocentrism could be said to be like good sense, of which perhaps it is just the apperceptive moment: the best distributed thing in the world. The format of the lesson is familiar, but that does not lessen its sting. Favouring one’s own humanity to the detriment of the other’s causes one to resemble the contemptible other in an essential way. Since the other of the Same (of the European) shows itself to be the same as the other of the Other (of the indigenous), the Same ends up unwittingly showing that it is the same as the Other.

The anecdote fascinated Lévi-Strauss enough for him to repeat it in Tristes Tropiques. But there he added a supplementary, ironic twist, this time noting a difference (rather than a resemblance) between the parties. While the Europeans relied on the social sciences in their investigations of the humanity of the other, the Indians placed their faith in the natural sciences; and where the former proclaimed the Indians animals, the latter were content to see the other as gods. ‘Both attitudes show equal ignorance,’ Lévi-Strauss concluded, ‘but the Indian’s behaviour certainly had greater dignity.’15 If this was really what happened,16 it would require us to conclude that, despite being just as ignorant on the subject of the other, the other of the Other was not exactly the same as the other of the Same. We could say that it was its exact opposite except for the fact that the relation between these two others of humanity – animality and divinity – is conceived in indigenous worlds in completely different terms than those we have inherited from Christianity. The rhetorical contrast Lévi-Strauss draws works by appealing to our cosmological hierarchies rather than to those of the Taino.17

In any case, it was consideration of this disequilibrium that led us to the hypothesis that Amerindian ontological regimes diverge from the widespread Western ones precisely with regard to the inverse semiotic functions they respectively attribute to soul and body. The marked dimension for the Spanish of the Antilles incident was the soul, whereas the Indians emphasized the body. The Europeans never doubted that the Indians had bodies – animals have them as well – and the Indians in turn never doubted that the Europeans had souls, since animals and the ghosts of the dead do as well. Thus the Europeans’ ethnocentrism consisted in doubting that the body of the other contained a soul formally similar to the one inhabiting their bodies, and the Indians’, on the contrary, entailed doubting that the others’ souls or spirits could possess a body materially similar to theirs.18 In the semiotic terms of Roy Wagner, a Melanesianist who will quickly reveal himself a crucial intercessor in the theory of Amerindian perspectivism, the body belongs to the innate or spontaneous dimension of European ontology (‘nature’), which is the counter-invented result of an operation of conventionalist symbolization, while the soul would be the constructed dimension, the fruit of a ‘differentiating’ symbolization that ‘specifies and renders concrete the conventional world by tracing radical distinctions and concretizing the singular individuals of this world’.19 In indigenous worlds, on the contrary, the soul ‘is experienced as … a manifestation of the conventional order implicit in everything’ and ‘sums up the ways in which its possessor is similar to others, over and above the ways in which he differs from them’;20 the body, on the contrary, belongs to the sphere of what comes from the responsibility of agents and is one of the fundamental figures of what has to be constructed against a universal and innate ground of an ‘immanent humanity’.21 In short, European praxis consists in ‘making soul’ (and differentiating cultures) on the basis of a given corporeal-material ground – nature – while indigenous praxis consists in ‘making bodies’ (and differentiating species) on the basis of a sociospiritual continuum, itself also given… in myth, as we will see.

Wagner’s conceptually dense and quite original theoretical system resists didactic summary; thus we request that the reader directly encounter its most elegant and realized presentation in The Invention of Culture. Grosso modo, the Wagnerian semiotic can be said to be a theory of human and nonhuman practice conceived as exhaustively consisting in the reciprocal, recursive operation of two modes of symbolization: (1) a collectivizing, conventional (or literal) symbolism where signs are organized in standardized contexts (semantic domains, formal languages, etc.) to the extent that they are opposed to a heterogeneous plane of ‘referents’ – that is, they are seen as symbolizing something other than themselves; and (2) a differentiating, inventive symbolic (or figurative) mode in which the world of phenomena represented by conventional symbolization is understood to be constituted by ‘symbols representing themselves’ – that is, events that simultaneously manifest as symbols and referents, thereby dissolving the conventional contrast. It should be observed, first of all, that the world of referents or the ‘real’ is defined here as a semiotic effect: what is other to a sign is another sign having the singular capacity of ‘representing itself’. The mode of existence of actual entities qua events or occasions is a tautology. It should be stressed that the contrast between the two modes is itself the result of a conventionalist operation (and perception): the distinction between invention and convention is itself conventional, but at the same time every convention is produced through a counterinvention. The contrast is thus intrinsically recursive, especially if we understand that human cultures are fundamentally in conflict over the mode of symbolization they (conventionally) privilege as an element appropriated for action or invention, in reserving to the other the function of the ‘given’. Cultures, human macrosystems of conventions, are distinguished by what they define as belonging to the sphere of the responsibilities of agents – the mode of the constructed – and by what belongs (because it is counterconstructed as belonging) to the world of the given or non-constructed. The core of any and every set of cultural conventions is a simple distinction as to what kind of contexts – the nonconventionalized ones or those of convention itself – are to be deliberately articulated in the course of human action, and what kind of contexts are to be counterinvented as ‘motivation’ under the conventional mask of ‘the given’ or ‘the innate’. Of course […] there are only two possibilities: a people who deliberately differentiate as the form of their action will invariably counterinvent a motivating collectivity as ‘innate’, and a people who deliberately collectivize will counterinvent a motivating differentiation in this way.22


The anthropological chiasm Lévi-Strauss opened on the Antilles incident agrees quite well with two characteristics recently distinguished in Amazonian ethnography. It unexpectedly confirmed, first, the importance of an economy of corporeality at the very heart of those ontologies recently redefined (in what will be seen to be a somewhat unilateral fashion) as animist.23 I say ‘confirmed’ because this was something that had already been abundantly demonstrated in the Mythologiques, as long as they are taken to the letter and thus understood as a mythic transformation of the mythic transformations that were their object. They describe, in other words, in prose wedding Cartesian rigour to Rabelaisian verve, an indigenous anthropology formulated in terms of organic fluxes and material codings, of sensible multiplicities and becomings-animal instead of in the spectral terms of our own anthropology that the juridical-theological grisaille (the rights, duties, rules, principles, categories and ‘moral persons’ conceptually formative of the discipline) overwhelms by comparison.24 Second, Amazonianists have also perceived certain theoretical implications of this non-marked or generic status of the virtual dimension or ‘soul’ of existents, a chief premiss of a powerful indigenous intellectual structure that is inter alia capable of providing a counterdescription of the image drawn of it by Western anthropology and thereby, again, of ‘returning to us an image in which we are unrecognizable to ourselves’. This double materialist-speculative torsion applied to the usual psychological and positivist representation of animism is what we called ‘perspectivism’, by virtue of the analogies, as much constructed as observed, with the philosophical thesis associated with this term to be found in Leibniz, Nietzsche, Whitehead and Deleuze.


Numerous, even virtually all, peoples of the New World, as diverse ethnographers have (unfortunately too often in passing) noted, share a conception of the world as being composed of a multiplicity of points of view. Every existent is a centre of intentionality apprehending other existents according to their respective characteristics and powers. The presuppositions and consequences of this idea are nevertheless irreducible to the current concept of relativism that it would at first glance seem to evoke. They are instead arranged on a plane orthogonal to the opposition between relativism and universalism. Such resistance on the part of Amerindian perspectivism to the terms of our epistemological debates throws doubt on the transposability of the ontological partitions nourishing them. This is the conclusion a number of anthropologists have arrived at (although for very different reasons) when asserting that the nature/ culture distinction – that first article of the Constitution of anthropology, whereby it pledges allegiance to the old matrix of Western metaphysics – cannot be used to describe certain dimensions or domains internal to non-Western cosmologies without first making them the object of rigorous ethnographic critique.

In the present case, such a critique demanded that the predicates arranged in the paradigmatic series of ‘nature’ and ‘culture’ be redistributed: universal and particular, objective and subjective, physical and moral, the given and the instituted, necessity and spontaneity, immanence and transcendence, body and spirit, animality and humanity, and so on. The new ordering of this other conceptual map led us to suggest that the term ‘multinaturalism’ could be used to designate one of the most distinctive traits of Amerindian thought that emerges upon its juxtaposition with modern ‘multicultural’ cosmologies: where the latter rest on the mutual implication between the unicity of nature and the multiplicity of cultures – the first being guaranteed by the objective universality of bodies and substance and the second engendered by the subjective particularity of minds and signifiers25 – the Amerindian conception presupposes, on the contrary, a unity of mind and a diversity of bodies. ‘Culture’ or subject as the form of the universal, and ‘nature’ or object the particular.

The ethnography of indigenous America is thick with references to a cosmopolitical theory describing a universe inhabited by diverse types of actants or subjective agents, human or otherwise – gods, animals, the dead, plants, meteorological phenomena, and often objects or artefacts as well – equipped with the same general ensemble of perceptive, appetitive, and cognitive dispositions: a soul. This interspecific resemblance includes, to put it a bit performatively, the same mode of apperception: animals and other nonhumans having a soul ‘see themselves as persons’ and therefore ‘are persons’: intentional, double-faced (visible and invisible) objects constituted by social relations and existing under a double, at once reflexive and reciprocal, which is to say collective, pronomial mode. What these persons see and thus are as persons, however, constitutes the precise philosophical problem posed by and for indigenous thought.

The resemblance between souls does not entail that what they express or perceive is likewise shared. The way humans see animals, spirits and other actants in the cosmos is profoundly different from how these beings see them and see themselves. Typically, and this tautology is something like the degree zero of perspectivism, humans will under normal conditions see humans as humans and animals as animals (in the case of spirits, seeing these normally invisible beings is a sure indication that the conditions are not normal, i.e. sickness, trance and so on). Predatory animals and spirits, for their part, see humans as prey, while prey see humans as spirits or predators. ‘The human being sees himself as what he is. The loon, the snake, the jaguar, and The Mother of Smallpox, however, see him as a tapir or a pecari to be killed’, remarks Baer of the Matsiguenga of Amazonian Peru.26 In seeing us as nonhumans, animals and spirits regard themselves (their own species) as human: they perceive themselves as or become anthropomorphic beings when they are in their houses or villages, and apprehend their behaviour and characteristics through a cultural form: they perceive their food as human food – jaguars see blood as manioc beer, vultures see the worms in rotten meat as grilled fish – and their corporeal attributes (coats, feathers, claws, beaks) as finery or cultural instruments, and they even organize their social systems as human institutions, are with chiefs, shamans, exogamous moieties and rituals.

Some precisions prove necessary. Perspectivism is only rarely applied to all animals (even as it encompasses nearly all other beings – at least the dead), as the species it seems most frequently to involve are the big predators and scavengers, like jaguars, anacondas, vultures and harpy eagles, and the typical prey of humans – wild boar, monkeys, fish, deer and tapirs. In fact, one of the fundamental aspects of perspectivist inversions concerns the relative, relational status of predator and prey. The Amazonian metaphysics of predation is a pragmatic and theoretical context highly favourable to perspectivism. That said, there is scarcely an existent that could not be defined in terms of its relative position on a scale of predatory power.

For if all existents are de facto not necessarily persons, the fundamental point is that there is de jure nothing to prevent any species or mode of being from being one. The problem, in sum, is not one of taxonomy, classification or ethnoscience.27 All animals and cosmic constituents are intensively and virtually persons, because all of them, it does not matter which, can reveal themselves to be (transform into) a person. This is not a simple logical possibility but an ontological potentiality. ‘Personhood’ and ‘perspectiveness’ – the capacity to occupy a point of view – is a question of degree, context and position rather than a property distinct to specific species. Certain nonhumans actualize this potential more fully than others, and some, moreover, manifest it with a superior intensity than does our species and are, in this sense, ‘more human than humans’.28 The question furthermore possesses an essentially a posteriori quality. The possibility of a previously insignificant being revealing itself (to a dreamer, sick person or shaman) as a prosopmorphic agent capable of affecting human affairs always remains open; where the personhood of being is concerned, ‘personal’ experience is, rightly, more decisive than whatever cosmological dogma.

If nothing prevents an existent from being conceived of as a person – as an aspect, that is, of a biosocial multiplicity – nothing at the same time prevents another human collective from not being considered one. This is, moreover, the rule. The strange generosity that makes peoples like Amazonians see humans concealed under the most improbable forms or, rather, affirm that even the most unlikely beings are capable of seeing themselves as humans is the double of the well-known ethnocentrism that leads these same groups to deny humanity to their fellow men [congénères] and even (or above all) to their closest geographically or historical cousins. It is as if, faced with the courageously disenchanted maturity of the old Europeans, who have long been resigned to the cosmic solipsism of the human condition (sweetened, it is true, by the consolation of intraspecific intersubjectivity), our exotic people oscillate perpetually between two infantile narcissisms: one of small differences between fellow peoples [congénères] that sometimes resemble each other too much, and another of big resemblances between entirely different species. We see how the other(s) can never win: at once ethnocentric and animist, they are inevitably immoderate, whether by lack or excess.

The fact that the condition of the person, whose universal apperceptive form is human, could be ‘extended’ to other species while ‘denied’ to other collectives of our own suggests from the very outset that the concept of the person – a centre of intentionality constituted by a difference of internal potential – is anterior and logically superior to the concept of the human. Humanity is the position of the common denominator [congénère], the reflex mode of the collective, and is as such derived in relation to the primary positions of predator and prey, which implicate other collectives and personal multiplicities in a situation of perspectival multiplicity.29 The resemblance or kinship [congénèrité] arises from the deliberate, socially produced suspension of given predatory difference and does not precede it. This is precisely what Amerindian kinship consists of: ‘reproduction’ as the intensive stabilization and/or deliberate non-achievement of predation, in the fashion of the celebrated Batesonian (or Balinese) intensive plateau that so inspired Deleuze and Guattari. In another text of Lévi-Strauss’s that (not by chance) deals with cannibalism, this idea of identity-by-subtraction receives a formula perfectly befitting Amerindian perspectivism:

The problem of cannibalism … does not consist in seeking the why of custom, but, on the contrary, how it uncovers this interior limit of predation to which social life perhaps boils down.30

This is nothing more than an application of the classic structuralist precept that ‘resemblance does not exist in itself; it is only a particular case of difference, one in which difference tends toward zero.’31 Everything hinges on the verb ‘tend,’ since, as Lévi-Strauss observes, difference ‘is never completely annulled’. We could even say that it only blooms to its full conceptual power when it becomes as small as desired – the difference between twins, for example, as an Amerindian philosopher would say.


The notion that actual nonhumans possess an invisible prosomorphic side is a fundamental presupposition of several dimensions of indigenous practice, but it is only foregrounded in the particular context of shamanism. Amerindian shamanism could be defined as the authorization of certain individuals to cross the corporeal barriers between species and adopt the perspective of their allospecific subjectivities so as to administer the relations between them and humans. By seeing nonhuman beings as they see themselves (again as humans), shamans become capable of playing the role of active interlocutors in the trans-specific dialogue and, even more importantly, able to return from their travels to recount them; something the ‘laity’ can only do with difficulty. This encounter or exchange of perspectives is not only a dangerous process but a specifically political art: diplomacy. If Western relativism has multiculturalism as its public politics, Amerindian shamanic perspectivism has multinaturalism as its cosmic politics.

Shamanism is a mode of action entailing a mode of knowledge, or, rather, a certain ideal of knowledge. In certain respects, this ideal is diametrically opposed to the objectivist epistemology encouraged by Western modernity. The latter’s telos is provided by the category of the object: to know is to ‘objectivize’ by distinguishing between what is instrinsic to the object and what instead belongs to the knowing subject, which has been inevitably and illegitimately projected onto the object. To know is thus to desubjectivize, to render explicit the part of the subject present in the object in order thereby to reduce it to an ideal minimum (and/or to amplify it, so as to obtain spectacular critical effects). Subjects, just like objects, are regarded as results of a process of objectivation: the subject constitutes or recognizes itself in the object it produces and knows itself objectively when it succeeds in seeing itself from the outside and as a thing. Our epistemological game, then, is objectivation; what has not been objectified simply remains abstract or unreal. The form of the Other is the thing.

Amerindian shamanism is guided by the inverse ideal: to know is to personify, to take the point of view of what should be known or, rather, the one who should be known. The key is to know, in Guimaraes Rosa’s phrase, ‘the who of things’, without which there would be no way to respond intelligently to the question of ‘why’. The form of the Other is the person. We could also say, to utilize a vocabulary currently in vogue, that shamanic personification or subjectivation reflects a propensity to universalize the ‘intentional attitude’ accorded so much value by certain modern philosophers of mind (or, more accurately, philosophers of modern mind). To be more precise, since the Indians are perfectly capable of adopting ‘physical’ and ‘functional’ attitudes (in Dennett’s sense) in their everyday life, we will say that we are faced here with an epistemological ideal that, far from seeking to reduce ‘ambient intentionality’ to its zero degree in order to attain an absolutely objective representation of the world, instead makes the opposite wager: true knowledge aims to reveal a maximum of intentionality through a systematic and deliberate ‘abduction of agency’. To what we said above about shamanism being a political art we can now add, by way of clarification, that it is a political art. For the good shamanic interpretation succeeds in seeing each event as being, in truth, an action or expression of intentional states or predicates of such and such an agent. Interpretive success, that is, is directly proportional to the successful attribution of intentional order to an object or noeme.32 An entity or state of things not prone to subjectivation, which is to say the actualization of its social relation with the one who knows it, is shamanically insignificant – in that case, it is just an epistemic residue or impersonal factor resistant to precise knowledge. Our objectivist epistemology, there is no need to recall, proceeds in the opposite direction, conceiving the intentional attitude as a convenient fiction adopted when the aimed-for object is too complicated to be decomposed into elementary physical processes. An exhaustive scientific explanation of the world should be capable of reducing every object to a chain of causal events, and these, in turn, to materially dense interactions (through, primarily, action at a distance).

Thus if a subject is an insufficiently analysed object in the modern naturalist world, the Amerindian epistemological convention follows the inverse principle that an object is an insufficiently interpreted subject. One must know how to personify, because one must personify in order to know. The object of the interpretation is the counter-interpretation of the object.33 The latter idea should perhaps be developed into its full intentional form – the form of a mind, an animal under a human face – having at least a demonstrable relation with a subject, conceived as something that exists ‘in the neighbourhood’ of an agent.34

Where this second option is concerned, the idea that nonhuman agents perceive themselves and their behaviour under a human form plays a crucial role. The translation of ‘culture’ in the worlds of extrahuman subjectivities has for its corollary the redefinition of several natural objects and events as indexes from which social agency can be inferred. The most common case is the transformation of something that humans regard as a naked fact into another species’ artefact or civilized behaviour: what we call blood is beer for a jaguar; what we take for a pool of mud, tapirs experience as a grand ceremonial house; and so on. The artefacts possess an ambiguous ontology: they are objects, but they necessarily indicate a subject since they are like frozen actions or material incarnations of a nonmaterial intentionality. What one side calls ‘nature,’ then, very often turns out to be ‘culture’ for the other.

We have here an indigenous lesson anthropology could benefit from heeding. The differential distribution of the given and the constructed must not be taken for an anodyne exchange, a simple change of signs that leaves intact the terms of the problem. There is ‘all the difference of/in the world’35 between a world that experiences the primordial as bare transcendence and pure anti-anthropic alterity – as the nonconstructed and noninstituted opposed to all custom and discourse36 – and a world of immanent humanity, where the primordial assumes a human form. This anthropomorphic presupposition of the indigenous world is radically opposed to the persistent anthropocentric effort in Western philosophies (the most radical included) to ‘construct’ the human as the nongiven, as the very being of the nongiven.37 We should nevertheless stress, against fantasies of the narcissistic paradises of exotic peoples (or Disney anthropology), that this presupposition renders the indigenous world neither more familiar nor more comforting. When everything is human, the human becomes a wholly other thing.

So there really are more things in heaven and earth than in our anthropological dream. To describe this multiverse, where every difference is political (because every relation is social), as though it were an illusory version of our universe – to unify them by reducing the inventions of the first to the conventions of the second – would be to decide for a simplistic, politically puerile conception of their relation. Such facile explanations end up engendering every sort of complication, since the cost of this ersatz ontological monism is its inflationist proliferation of epistemological dualisms – emic and etic, metaphoric and literal, conscious and unconscious, representation and reality, illusion and truth (I could go on…). Such dualisms are dubious not because all conceptual dichotomies are in principle pernicious but because these ones in particular require, as the condition of the unification of worlds, a discrimination between their respective inhabitants. Every Great Divider is a mononaturalist.

Translated by Peter Skafish


1 Patrice Maniglier, ‘La Parenté des autres’, Critique 701, Ocober 2005, pp. 773–4.

2. Harri Englund and James Leach, ‘Ethnography and the Metanarratives of Modernity’, Current Anthropology, vol. 41, no. 2, 2000, pp. 225–48. See also Lévi-Strauss’s distinction between anthropology qua a ‘centrifugal science’ adopting ‘the point of view of immanence’, and economics and sociology, the ‘centripetal sciences’ that attribute a ‘transcendental value’ to the societies of the observer. ‘Critères scientifiques dans les disciplines sociales et humaines’, Revue internationale des sciences sociales, vol. 16, no. 4, 1964, pp. 579–97.

3. Which does not at all mean that the former and the latter are epistemologically homogeneous from the point of view of the techniques in play and the problems implied. See Marilyn Strathern, ‘The Limits of AutoAnthropology’, in A. Jackson, Anthropology at Home, Tavistock, London, 1987, pp. 59–67.

4. See, on comparative ontography, Martin Holbraad, ‘Estimando a necessidade: os orácules de ifá e a verdade em Havana’, Mana, vol. 9, no. 2, 2003, pp. 39–77. The point of view of immanence is not exactly the same perspective as Lévi-Strauss’s in the passage cited above.

5. See Patrice Maniglier, ‘L’Humanisme interminable de Lévi-Strauss’, Les Temps Modernes 609, June–August 27 2000, pp. 216–41.

6. Jacques Derrida, L’Animal que donc je suis, Galilée, Paris, 2006.

7. Gilles Deleuze and Félix Guattari, A Thousand Plateaus, University of Minnesota Press, Minneapolis, 1987.

8. This allusion to Moses can be found in ‘Introduction to the Work of Marcel Mauss’, in Marcel Mauss, Sociologie et Anthropologie, PUF, Paris, 1950, pp. ix–lii.

9. On the philosophy to come of Lévi-Strauss, see Klaus Hamberger, ‘La Pensée objectivée’, in Michel Izard, Lévi-Strauss, Éditions de l’Herne, Paris, 2004, p. 345.

10. Claude Lévi-Strauss, ‘Postface’, L’Homme 154–155, April–September 2000, pp. 719–20.

11. See my ‘A propriedade do conceito: sobre o plano de imanência amerindio’, XXV Encontro Annual da ANPOCS, Caxambu, 2001, for another commentary on this passage, which has also been brilliantly discussed by Maniglier in ‘Des us et des signes. Lévi-Strauss: philosophie pratique’, Revue de Métaphysique et de Morale 1, 2005, pp. 89–108.

12. See Anne-Christine Taylor, ‘Don Quichotte en Amé- rique’, in Michel Izard, ed., Lévi-Strauss, Éditions de l’Herne, Paris, 2004, pp. 92–8. On this view of Deleuze, see David Lapoujade, ‘Le Structuralisme dissident de Deleuze’, in A. Akay, Gilles Deleuze için = Pour Gilles Deleuze, Akbank Sanat, Istanbul, 2006, pp. 27–36.

13. For the chief formulations of the idea, see Tânia Stolze Lima, ‘The Two and Its Many: Reflections on Perspectivism in a Tuna Cosmology’, Ethnos, vol. 64, no. 1, 1999, pp. 107–31, and Um Peixe Olhou para Mim: O Povo Yudjá e a Perspectiva, Edunesp/NuTI/ISA, São Paolo, 2005. See also Viveiros de Castro, ‘Cosmological Deixis and Amerindian Perspectivism’, Journal of the Royal Anthropological Institute, vol. 4, no. 3, 1998, pp. 469–88; ‘Perspectivisimo e multinaturalismo na Amé- rica indígena’, A Inconstância da Alma Selvagem, Cosac and Naify, São Paolo, 2002, pp. 345–99; ‘Perspectival Anthropology and the Method of Controlled Equivocation’, Tipiti, vol. 2, no. 2, 2004, pp. 3–22; ‘Exchanging Perspectives: The Transformation of Objects into Subjects in Amerindian Cosmologies’, Common Knowledge, vol. 10, no. 3, 2004, pp. 463–84. In what follows, I repeat themes and passages already known to the anthropological public, but which other readers will benefit from having reprised.

14. Claude Lévi-Strauss, ‘Race and History’, Unesco, Paris, 1952, p. 12.

15. Claude Lévi-Strauss, Tristes Tropiques, Penguin Books, New York, 1973, p. 76.

16. As Marshall Sahlins observed in How ‘Natives’ Think: About Captain Cook, for Example, University of Chicago Press, Chicago, 1995, the association of colonial invaders with local divinities, a phenomenon observed in diverse encounters between the Moderns and indigenous peoples, says much more about what the Indians thought about divinity than about what they thought of Europeanness or modernity.

17. The anecdote was taken from Oviedo’s History of the Indians; it should have taken place in Hispanolia, in the inquiry undertaken in 1517 by priests of the order of St Jerome in the colonies, and Puerto Rico, with the heuristic submergence of a young Spaniard, who was caught and then drowned by Indians. It is an argument that, moreover, demonstrates the necessity of pushing ‘the archaeology of the human sciences’ back until at least the controversy of Valladolid (1550–51), the celebrated debate between Las Casas and Sepúlveda on the subject of the nature of American Indians. See Anthony Pagden, The Fall of Natural Man: The American Indian and the Origins of Comparative Ethnology, Cambridge University Press, Cambridge, 1982.

18. The old soul has been going incognito since being rechristened as culture, the symbolic, mind… The theological problem of the soul of others became the philosophical puzzle of ‘the problem of other minds’, which currently extends so far as to include neurotechnological researches on human consciousness, the mind of animals, the intelligence of machines (the gods having transferred themselves into Intel microprocessors). In the last two cases, the problem is whether certain animals would not, after all, have something like a soul or a consciousness – perhaps even a culture – and, reciprocally, if certain material non-autopoietic systems lacking, in other words, a true body could show themselves capable of intentionality.

19. Roy Wagner, The Invention of Culture, University of Chicago Press, Chicago, 1981, p. 42.

20. Ibid., p. 94.

21. Ibid., pp. 86–9. Here I am myself ‘innovating’ on Wagner, who does not in The Invention of Culture raise the question of the status of the body in the differentiating cultures.

22. Ibid., p. 51.

23. The theme of animism was recently redeployed by Philippe Descola, who of course pays unstinting attention to Amazonian materials. See Descola, ‘Societies of Nature and the Nature of Society’, in A. Kupter, Conceptualizing Society, Routledge, London, 1992, pp. 107–26, and ‘Constructing Natures: Symbolic Ecology and Social Practice’, in P. Descola and G. Pálsoon, eds, Nature and Society: Anthropological Perspectives, Routledge, London, pp. 82–102.

24. See A. Seeger, R. DaMatta and E. Viveiros de Castro, ‘A Construção da pessoa nas sociedades Indígenas brasileiras’, Boletim do Museu Nacional, vol. 32, nos 1–2, 1979, pp. 2–19, for a first formulation of the problematic of corporeality in indigenous America. Because it explicitly relied on the Mythologies, that work was developed without the least connection to the theme of ‘embodiment’ that took anthropology by storm in the decades to follow. The structuralist current of Amerindian ethnology, deaf to what Deleuze and Guattari called the ‘at once pious and sensual’ appeal to phenomenological ‘fleshism’ [carnism] – the appeal to ‘rotten wood’, as a reader of The Raw and The Cooked would say – always thought incarnation from the perspective of the culinary triangle instead of the holy Trinity.

25. See Tim Ingold, ‘Becoming Persons: Consciousness and Sociality in Human Evolution’, Cultural Dynamics, vol. IV, no. 3, pp. 355–78

26. Gerhard Baer, Cosmología y shamanismo de los Matsiguenga, Abya-Yala, Quito, 1994.

27. Compare with what Lienhardt says on the heteroclite collection of species, entities and phenomena that served the clan-divinities of the Dinka of Sudan. ‘The Dinka have no theory about the principle upon which some species are included among clan-divinities, and some omitted. There is no reason, in their thought, why anything might not be the divinity of some clan.’ Godfrey Linehardt, Divinity and Experience: The Religion of the Dinka, Oxford University Press, Oxford, 1961, p. 110.

28. See Hallowell A. Irving, ‘Ojibwa Ontology, Behavior, and Worldview’, in Sarah Diamond, ed. Culture in History: Essays in Honor of Paul Rand, Columbia University Press, New York, 1960, pp. 49–82.

29. ‘Human’ is a term designating a relation, not a substance. Primitive peoples’ celebrated designations of themselves as ‘the human beings’ and ‘the true men’ seem to function (meaning pragmatically, not syntactically) less as substantives than as pronouns marking the subjective position of the speaker. It is for this reason that the indigenous categories of collective identity possess this great contextual variability so characteristic of pronouns, marking the self/other contrast through the immediate kinship of an Ego with all other humans, or, as we have seen, with all other beings endowed with consciousness. Their sedimentation as ‘ethnonyms’ appears to be to a large extent an artefact produced through interaction with the ethnographer.

30. Claude Lévi-Strauss, Paroles Données, Plon, Paris, 1984, p. 144. See also Mythologiques IV: L’homme nu. Plon, Paris, 1972, p. 617.

31. See Lévi-Strauss Mythologiques IV, p. 32. The precept is classic, but few of the so-called ‘structuralists’ truly understood how to push this idea to its logical conclusion and thus beyond itself. Might that be because it would bring them too close to Difference and Repetition?

32. I am referring here to Dennett’s notion of the n-ordinality of intentional systems. A second-order intentional system is one in which the observer must ascribe not only (as in the first order) only beliefs, desires and other intentions to the object but, additionally, beliefs, etc. about other beliefs (etc.) The standard cognitive thesis holds that only humans exhibit second- or higher-order intentionality. My shamanistic ‘principle of the abduction of a maximum of agency runs foul of the creed of physicalist psychology: ‘Psychologists have often appealed to a principle known as Lloyd Morgan’s Canon of Parsimony, which can be viewed as a special case of Occam’s Razor: it is the principle that one should attribute to an organism as little intelligence or consciousness or rationality or mind as will suffice to account for its behavior.’ Daniel Dennett, Brainstorms: Philosophical Essays on Mind and Psychology, Penguin, Harmondsworth, 1978, p. 274.

33. As Marilyn Strathern observes of an epistemological regime similar to that of Amerindians: ‘The same convention requires that the objects of interpretation – human or not – become understood as other persons; indeed, the very act of interpretation presupposes the personhood of what is being interpreted.… What one thus encounters in making interpretations are always counter-intepretations’. Marilyn Strathern, Partial Connections, Rowman & Littlefield, Savage MD, 1991, p. 23.

34. See Alfred Gell, Art and Agency: An Anthropological Theory, Clarendon Press, Oxford, 1998.

35. Wagner, The Invention of Culture, p. 51.

36. ‘Nature is different from man: it is not instituted by him and it opposes itself to custom and discourse. Nature is the primordial, which it to say the nonconstructed, the noninstituted’. Maurice Merleau-Ponty, La Nature: Notes, Cours du Collège de France (1956), edited by Dominique Séglard. Seuil, Paris: 1995 p. 20.

37. Peter Sloterdijk, La domestication de l’Être, Mille et une Nuits, Paris, 2000.