Descola’s Beyond nature and culture, viewed from Central Brazil

This work is licensed under the Creative Commons | © Marcela Coelho de Souza. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau4.3.029


Descola’s Beyond natureand culture, viewed from Central Brazil

Marcela COELHO DE SOUZA. Universidade de Brasília

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

Philippe Descola’s book (2013) takes off from a small clearing in the lower reaches of the Kapawi, a river in upper Amazonia: the house of Chumpi, an Achuar man distressed by his wife’s accident with a snake. Chumpi’s interpretation of this event—a vengeance sent by one of the “mothers of game” for Chumpi’s abusive massacre of a herd of wooly-monkeys—seems to have precipitated for the author one of those revelatory experiences that continues to reverberate throughout one’s subsequent endeavors as an anthropologist. It was an ethnographic moment that he, as Marilyn Strathern says of her own moment, “has never been able—wanted— to shake off ” (Strathern 1999: 6).

So let me start from a clearing (and moment) of my own: this was a recently opened small camping site by the Suiá-Miçú river in Mato Grosso (Central Brazil), where a team formed by a bunch of Kĩsêdjê youngsters, three elders, one a shaman, his son (an ascending political and community leader), and I had just hanged our hammocks to spend the night. My hammock and the shaman’s were hung on the same trunk: that of an impressive turetxi—the yellow ipê (Tabebuia spp.)—a tree the Kĩsêdjê consider one of the most enthusiastic singers inhabiting their forests. Pointing to it, Ntoni, the shaman, told me that its “owner” (kandê)—“that is, his soul,” he explained in Kĩsêdjê—who was there quietly hidden behind (not inside) the trunk, would not have died even if we had felled it: on the contrary, it would take vengeance of the deed by making us sick. Now, this was not a difficult idea for me to understand, accustomed as I was to many similar statements made by this shaman and other Kĩsêdjê friends, all of them well in line with the picture of Amerindian worlds we've been getting from descriptions old and new, but especially those already informed by the themes and concepts of Amerindian animism and perspectivism.

What was not so easy to grasp was this correspondence: owner ≈ soul. It suggested an analogical proportion of the sort “the owner is to the thing owned as the soul is to the body,” but before pondering what this could mean for my understanding of the category of “owner,” I needed to understand what this implied regarding Kĩsêdjê concepts of soul and body.

I have been dodging the issue of the Gê soul—yes, the Kĩsêdjê are a Northern Gê speaking people, of the same linguistic stock as those Central Brazilian societies that were so important to Lévi-Strauss, societies whose myths on the origin of fire take us through the passage from nature to culture in Le cru et le cuit— for some time. But now, before the impossible challenge of commenting Descola’s book, maybe I should take the opportunity to tackle this other daunting task and try a little bit harder to grasp the meaning of the Northern Gê karõ, “soul,” “image,” “shadow.” My conviction is that the only way I can productively relate to this book is by refraining from a global evaluation that would require me either to embrace the project as a whole or to criticize it, as if it were all of a piece. This is not to suggest Descola’s project lacks in consistency—all I mean is that it does not lack in inconsistencies either, and that to these I feel I can relate. Specially those, as the ones I will try to explore here, that are not inconsistencies between different parts of the system but between systematizing as a procedure and the procedures or knowledge practices it systematizes.

I will try then to relate to Descola’s book through a specific question. If the equivalence soul ≈ owner posed by my Kĩsêdjê teacher seems to lead to a conceptualization of the soul/body duality (as far as it does evoke such duality to me) very much at odds with that which informs Descola’s system—I mean the interiority/ physicality axis—is it possible to find in this same system, despite this incongruence, something that would allow me to better understand my ethnographic moment? I hasten to warn the reader I won't be able to give an answer to this here, for my aim is far more humble: I want to pose this question to my other teacher, that is Descola, connecting it with a parallel one regarding the history of his own argument—to its history and prehistory, because I will summon here a double ancestor to us all, americanists and (post)structuralists alike: the Bororo-Lévi-Strauss twins.

No need, I hope, to document the impact of Bororo thought on Lévi-Strauss’ thinking. The better-known facet of this history, Lévi-Strauss’ musings over the complexities of Bororo dual (diametric and concentric, symmetrical and asymmetrical) and triadic interlocking segmentary structures, will not be my focus here though. What I have in mind is the aroe/bope duality, described by Jon Crocker as two antithetical principles underlying Bororo cosmology. The first concept, aroe, refers to the formal or nominal essences of living beings: for every physical thing has another “modality of being,” which is their essence, soul, or name, existing primarily in a underworld that is also the dwelling of human (Bororo) souls (also aroe) after death (Crocker 1985: 33). These “spirits of classification,” as “representatives of categorical form,” are opposed and complemented by the bope, the “spirits of metamorphoses” that represent the principle of all organic transformations—growth, death, and decay—and is manifested in these processes as well as in the raka, “blood,” “which animates every being and enables it to accomplish those actions appropriate to its kind” (36). If the ensuing dialectics generates a system whose complexity we can no more than glimpse at here, what is important to retain is perhaps the idea of complexity itself.

So on the one hand we have the aroe, spatially organized into moieties further segmented in four divisions (oriented by the cardinal points) each, articulating a “symbolic geography” in terms of which localized corporate groups are differentiated in the eight clans that constitute the ideal Bororo village plan. These clans are in this way associated with eight fixed sets of aroe, in a relationship that Crocker calls totemic “only in the analogic sense established by Lévi-Strauss” since there is no idea of descent or mystical connection between humans in clans and the species whose aroe are associated with them (Crocker 1985: 34). On the other hand, if every species or kind of creature have their aroe classified in one of these eight classes, they also have their bope aspect, “since all things have a double aspect, being at once process (bope) and form (aroe)” (180). There are creatures, however, that manifest this aspect in such a degree that they come to be identified with bope; “so tangibly filled with appearances and habits otherwise thought unique to the bope that a clear distinction cannot be drawn between them and the suprareal principle” (180). These animal species (vultures, alligators, coral snakes, boa constrictor, among others), as well as the sun, the moon, rain, and thunder, are said to be “Themselves bope” in Bororo parlance—they are “metonyms for the bope” (180). But there are other species that are more like the aroe (all the while having their own bope aspect, as everything else in the world does): aquatic birds, macaws, the harpy eagle—they are sometimes said to be “Themselves aroe.” These positions in a continuum (see table in Crocker 1985: 190) must not be confused with the clear-cut distribution of the aroe of each and every species in the eight nominal classes (for a sample, see Crocker 1985: 269). But some “confusion” (in fact, complexes series of inversions and conversions) is produced by the fact that those classes themselves, or rather the moieties into which they are divided, are said to be one more aroe—the Tugarege— and the other more bope—the Exerae. Despite this, however, the “totemic system” of the aroe remains distinguishable from that other system, concerning the protocols involved in food consumption that affect species considered to be, like humans themselves, bope ure, foods of the bope (Crocker 1985: 139ss). These species may be consumed by humans (they are in fact the best tasting of all game), but only after they are offered to the bope through their shaman, the bari, and only if its procuring and preparation is conduced with the utmost care, according to precise rules: if these are broken, illness and death might ensue. It is to this second “system” that Lévi-Strauss refers in La pensée sauvage when, discussing the “system of transformations” of the “so called-totemism,” he mentions the Bororo as an example of a system traditionally seen as totemic, but where alimentary prohibitions obey a different logic, forming a “extra or para-totemic” system not reducible to the totemic or clanic one (Lévi-Strauss 1962: 132).

The Bororo are absent from Descola’s Beyond nature and culture (2013)—but not from the path that brought the author from Chumpi’s house in the Kapawi back to Paris. They appear, specifically, in an important article (Descola 1996), written while the book was in gestation, as an instance of totemic systems, taken then as now by the author as a mode of identification “rather exceptional outside Australia” (1996: 96) but that could be often combined with animic systems “which allow the expression of a relation of reciprocity with at least a fraction of non-humans” (1996: 96). The specificities of such (Bororo) combination are laid down in a note:

This is the case, for instance, among the Bororo of Eastern Brazil, who make a clear distinction between, on the one hand, aroe species (the jaguar and most felidaes, aras, aquatic birds, the harpy eagle, etc.) that are associated with totemic classifications, social order and nominal essences and, on the other hand, bope species (vultures, deer, the tapir, the capybara, peccaries, catfish, etc.) which embody the life processes, both positive and negative, and which exchange vital energy with humans in a complex system of reciprocity (see Crocker 1985). (Descola 1996: 100n4)

At this point in his journey from the Kapawi to Paris, Descola had already moved away a little bit from his initial contrast, laid down in his 1992 chapter in Kuper’s Conceptualizing society, between two schemes for the objectification of nature he then termed “animic” and “totemic,” defining the first as, “among other things, the belief that natural beings posses their own spiritual principles and that it is therefore possible for humans to establish with these entities personal relations of a certain kind” (Descola 1992: 114). At this first stopover, when animic systems were for the first time (as far as I know) named as such by the author, they appeared as a “symmetrical inversion of totemic classifications” (114)—whereas in our book, it is the symmetrical inversion regarding naturalism that is emphasized. And totemism was back then still understood by Descola in the Lévi-Straussian sense: as a logic of classification, that would “make use of empirically observable discontinuities between natural species to organize, conceptually, a segmentary order delimiting social units” (114). This means that animism made its (re-)appearance in the anthropological stage as an inversion of totemism taken as a classificatory procedure: the contrast (here in 1992) was between a system that treats plants and animals as mere signs or privileged operators of taxonomic thought and another that “treat[s] them as proper persons, as irreducible categories”; in the latter, “the relation of plants and animals to humans is not metaphorical, as in totemism, but at the most, and only in certain cases, metonymic” (114). Through this move, Descola laid the ground for specifying the modalities of social treatment of others proper to an animist world: what kinds of relationships humans and nonhumans may establish, such as reciprocity or predation.

A few years later, in 1996, at that second stopover from the Kapawi to Paris I have already referred to—when he evokes the Bororo as a combination of animism and totemism—the scenery has already changed a little, for we have now animism, totemism, and naturalism, as three distinct and equipollent modes of identification that, combined with specific modes of relation and modes of classification, should account for the “social objectification of nonhumans” in any society or culture. Among these “schemes or schemata of praxis,” the modes of identification would “define the boundaries between self and otherness as expressed in the treatment of humans and nonhumans, thus giving shape to specific cosmologies and social topographies” (Descola 1996: 87). As for their variety, to the pair animism/totemism as defined in the previous article, a third one—naturalism—is now explicitly added (of course, it was already there in the background). Another development regards the addition of a third mode ofrelation—protection—to the first two (predation and reciprocity). The latter are still exemplified, as in 1992, by the Jivaro and Tukano cases in Amazonia: among the first predation encompasses reciprocity, among the second we have the reverse. While protection, in its turn, may still occur in some corners of this endangered and magnificent forest, as a dominant mode of relation, says Descola, it is seldom associated with the animic systems that predominate here (because of the prevalent hunting ethos of Amazonian peoples): typical of interactions with domesticated species, it would prevail for instance among the Buryat of Siberia, where the conjunction of herding and hunting could account for the animism cum protection combination.

It is in terms of the combinatorial pairing of modes of identification and modes of relation, then, that the Bororo case appears in 1996. Totemism is here said to be incompatible with predation or reciprocity: how could one have a (social) relationship with animals or plants, or nonhumans of whatever kind, which are there only as food for thought? They would be mere “signs that a society uses to conceptualize its segmentation and, as such, they cannot constitute the terms of social relations with humans” (Descola 1996: 96). If relations with them, social or otherwise, exist, then they belong to other registers: be it the “literal,” practical one, in which predation as in hunting appears as a “quite mundane activity of food procurement,” as it happens among the purely totemic Australians; or, and here’s the general answer, in impure contexts, in which totemism appears combined with animic systems. Reciprocity thus becomes possible, for people like the Bororo, between humans and “at least a fraction of nonhumans” (a fraction, that is, distinct from that other fraction formed by their totemic species).

This whole construction is clearly unbalanced. In accordance with the Lévi-Straussian sense of totemism Descola sustains up to this point, he maintains, “the modes of relation between humans and non-humans typical of totemic systems are necessarily dichotomised” (1996: 95). We could then observe, with Viveiros de Castro ([1998] 2012: 88–89), that totemism remains less an ontology or a mode of identification to be put side by side with animism, than a form of classification (the only way, perhaps, naturalistic thought is able to conceptualize “kinship” between humans and nonhumans as something more than religious illusion and imaginary identification, rationalizing the connection as symbolic and metaphoric). To go beyond this and achieve a coherent matrix of all possible ontologies, something had to be done to totemism.

It is now time to disembark finally in that beautiful Paris, where balance is everything. Here, Descola’s inimitable mastery of the tools and style of the founder of his chair in the Collège allows him to surprise us indeed. Instead of demoting totemism as an ontology and dispossessing it from the dignity of a specific mode of identification capable of creating, in fact, a whole new world, irreducible to that of their animist congeners, he leaves Lévi-Strauss behind and reconceptualizes the totemic scheme as a full-blown ontology. All the while, Descola discovers, in the same movement, a fourth continent (or “great archipelago” [Descola 2013: 309])— analogism—which indisputably and forcefully puts on the anthropological map a whole series of phenomena and configurations that—at least from the point of view of an animist (or perspectivist) Amazonia, to be sure—it was perhaps harder to visualize and not so easy to identify.

In Beyond nature and culture, the distinction between modes of identification and modes of relation is given a more explicit and reasoned formulation (2013: 112ss), and the definition of the first as configuring different ontologies is carefully spelled out: the modes are polyvalent, universal dispositions that “come to have a public existence in the form of ontologies that favor one or another of them as the principle according to which the regime of existing beings is organized” (247). They are derived from an elegant simple matrix that generates a fourfold typology from a conventional distinction between interiority and physicality (soul, or mind, and body, in its Western version) as dimensions or axes along which humans may see themselves in continuity or in discontinuity regarding nonhumans: if the interiority of nonhumans is similar to those of humans, but their physicality is not, we have animism; the inverse case give us naturalism as similar physicalities combined with dissimilar interiorities; if the interiority and physicality of nonhumans are identical to those of humans, we have totemism; its inverse case is analogism, with dissimilar interiorities and dissimilar physicalities all the way down… . The modes are thus not exclusive since they “can each tolerate a discreet presence of other emerging modes,” which in their turn “introduce nuances and modifications in the expression of the locally dominant schema, thereby engendering many of the idiosyncratic variations that are customarily called cultural differences” (167). Notwithstanding these possible, perhaps inevitable, idiosyncratic mixes, each ontological mode “prefigures” a specific kind of collective: it is here that most aspects of what we normally call “social systems” are considered (chapter 11). These specific collectives, moreover, are composed by different kinds of subjects, produced by their respective modes of identification—subjects that will confront distinct epistemological and metaphysical problems. It is here that (as discussed in chapter 12) theories of knowledge or epistemologies arise.

Modes of relation as “means of specifying the general form of the links between … beings” are on a par with modes of identification as “means of specifying the properties of existing beings,” in the sense that modes of relation, as schemes of practice, are not specified by modes of identification in the manner the kinds of collectives and kinds of subjects, sociologies, and epistemologies seem to be. So, what the possible and impossible connections between modes of relation and modes of identification are remains an open question. It may be approached in two ways: by observing how one relational schema is able (or not) to structure interactions in a variety of ontologies; or by observing whether an ontology is able (or not) to accommodate more than one relational schema (and which ones) (2013: 336). These are the objects of chapters 13 and 14, respectively. The last chapter—and I believe readers of La pensée sauvage would not be surprised by this—takes up the problem of change, under the title “Histories of Structures”…

Let us see how all this looks like from my own small clearing in Central Brazil. I said I wanted to do this by means of a detour through the Bororo, so I could pose my question to Descola starting with a not very interesting typological problem: where have the Bororo gone, in the big picture of Beyond? I suppose they would still be classed as animists, not only because of the obvious social, cultural, and historical continuities that link Amazonia and Central Brazil and of the mediating position the Bororo seem to occupy here, but also on account of the social transactions between humans and the bope, regarding the food of the bope, conduced by the bari (bope shamans). It was on the basis of such relations that both Lévi-Strauss and Descola had pointed to the hybrid character of Bororo society—but in somewhat different ways. For Lévi-Strauss, this mixing was just a particular instance of a series of transformations leading us to challenge a hard-and-fast opposition between strictly totemic systems (founded on the homology between two systems of differences) and other systems (like castes) in which homologies are displaced by direct analogical relations across the two series (natural and cultural) between pairs of terms. Although the opposition resists—in the end, it is maintained by Lévi-Strauss under the guise of the totemism/sacrifice contrast—it also vacillates, becoming less typologically diacritical at the level of whole sociocultural systems. The opposite seems to be the case with Descola’s conceptualization, however. In the new architecture proposed by the book, we would need to find a niche where the combination of animism and totemism would take the form it assumes among the Bororo. A few questions would arise; for instance, if a combination of ontologies or modes of identification is indeed what we have here, which would be the dominant system that would “tolerate a discreet presence of other emerging modes” (2013: 167)? How can one determine such a thing? In raising the typological question, my point is not so much to situate the Bororo in the monumental architecture of Beyond nature and culture (although I am of course curious about how Descola would respond to this). What is unsettling to my mind, what my ethnographic moment makes me unable (or unwilling) to shake off, is the way Ntoni’s lesson—the soul is the owner of the (tree) body—seems to subvert the orthogonal double contrast that allowed Descola to give balance to his (now four) ontologies; a subversion that points, I suspect, to the same difficulty posed by Bororo “totemism.” Let me see if I can persuade the reader of this.

As the very compressed account given earlier should have suggested, Bororo cosmology is not easily analyzed as a composite formation that could be separated into two discrete parts of an encompassing (and hierarchical—remember the “tolerated discreet presence”…) whole. Remember their exogamous moieties: divided by their different aroe; being, simultaneously, one of them more aroe and the other more bope; and, at the same time, again, each being composed of beings that have, as anything that lives—i.e., grows and decays—their bope aspect. Not one or two modes of identification, nor even a hybrid composite or an intermediate compromise: for every move of encompassment or mediation here is bound to generate a countermovement that reveals that third to be an entre-deux: not a term in itself, but the way each term provides on the other a perspective that opens it up, allowing perspectives to be exchanged … (Lima 2008). The duality aroe/bope, as a duality between logically heterogeneous terms (stability and change, being and becoming, synchrony and diachrony, etc.) is clearly in this like that implicit triadic scheme that reveals all such dual structures to be animated by the perpetual disequilibrium that defies and redefines hierarchy through a contrahierarchical principle we would like to call, with Tania Lima, a “reciprocity of asymmetric perspectives” capable of capturing the “fractal dimensionality of systems that are recouped by points of view irreducible to a hierarchical assemblage of parts and wholes” (Lima 2008: 249). This fractal dialectics seems to point to an original mode that obviates the very parameters of Descola’s quadripartite matrix (interiority/physicality, continuity/discontinuity) and the contrast humans/nonhumans across which such parameters operate. If we could read this as a “presence” of one mode in the other, this presence would not be “discreet” at all, for it would profoundly change the nature of both modes. They could not be distributed in different provinces of social practice. What emerges, then, is perhaps a fifth mode, but one that I would still hesitate to call a “mode of identification.”

It is something like this that Ntoni’s explanations evoked to me. A vertiginous sense that, first, binary oppositions like soul/body, owner/owned, visible/invisible, active/passive, continuous/discontinuous, being/becoming, et cetera cannot be stabilized as axes capable of generating a closed group of structural transformations; and, second, that they cannot even be chained in analogical proportions like my tentative “the soul is to the body as the owner to the owned” without creating some serious disproportions that, from another point of view, lead to other proportions that cannot not encompass nor be encompassed by the first. My proportional analogy, “the soul is to body as the owner is to the owned” of course introduces a concept we should be careful not to take for granted here: namely, the body. It also introduces a figure, the owner in its relation to what is owned that is equally dangerous to take for granted, considering the variegated ways it has been identified in recent Amazonian ethnography (Seeger 1981; Lima 2005; Costa 2007; Fausto 2008; Cesarino 2010; Guerreiro 2012; Soares Pinto 2014). Those two concepts (the body and the owner) have been ethnographically linked by assertions, made by some peoples in Amazonia, that identify certain bodies—which often comprise vegetable body-like structures like tree trunks—as owners, i.e., magnified forms of personhood that would contain or embody other subjectivities or persons. Without entering here into a discussion that would need a whole article (or book) to develop, I believe is quite safe to say that the inverted formula presented by Ntoni—identifying the soul as the external owner of the body—can be taken as an expression of the positional or perspectival sense in which concepts of body (and soul) need to be understood in Amazonia (Viveiros de Castro 2001, [1998] 2014; Lima 2002).

It is perhaps a coincidence that Lévi-Strauss, in his comments to Crocker’s exposition of the Bororo dialectics of aroe/bope at the level of the self—after remarking that “ce qui frappe dans la pensée Bororo, c'est, partant de differénces, cette volonté éperdue d'identification”—asks the ethnographer about those “mythes si singuliers où an individu rencontre un autre individu qui lui est totalement identique, qui est un autre lui-même” (Lévi-Strauss 1977: 180). The coincidence being that it was exactly through such an oneiric experience that Ntoni came to know himself as a shaman: a shaman of a particular kind, a me khatwãn khêrê, “a person without spirit” (for his spirit lives elsewhere than in his body), that contrasts with the other type of shaman, the wajanga (someone whose body has been contaminated by the agency of other souls/bodies) (Coelho de Souza, forthcoming). Like the soul-owner of that turetxi in whose trunk our hammocks were hung that was hidden behind (not inside) the body of the tree, Ntoni’s soul was also an inside turned outside… . It seemed clear to me that I could not see this soul as possessing the body in the sense of maintaining with it a relation based on some sort of identity by extension (since inclusion was off the table); instead, what seemed important was that this soul could be separated from it. “Ownership” here meant a “separative” relation, a separation without which one could not appear as the objectification of the other, and without these objectifications, none of them would appear at all.

Perhaps what I am doing here is not much more than subscribing to Viveiros de Castro’s reservations regarding the “parallel” schemes of totemism (continuity in physicality and in interiority) and analogism (discontinuity in both axes), as far as they suppose mutually independent definitions of “physicality” and “interiority,” whereas in the two “crossed” schemes, animism (continuity in interiority, discontinuity in physicality) and naturalism (the inverse), the values of those axes appear as relationally determined by an internal contrast, in which one pole would appear reciprocally as figure to the other, its background (Viveiros de Castro 2009: 48–49n2). But the fact remains that my ethnographic moment, and the echoes it evokes from my own (virtual) voyages around Central Brazil and Amazonia, make it very difficult for me to engage with Bororo and Kĩsêdjê combinations in terms of whatever complex hierarchical arrangements we could devise between “modes of identification” whose identification itself would depend on a stable and independent definition of dimensions like interiority and physicality. For it is those dimensions themselves that the arrangements evoked here seem devised to relativize: to unfold in a series of versions and inversions of themselves.

I am not trying (or I am trying not) to criticize Descola for using “ethnocentric” concepts as body and soul; nor am I simply attempting to assess whether the way the Kĩsêdjê or the Bororo as persons and collectives make known to themselves what is inside or outside them, what is proper to themselves (what is Same) and what is Other, can in fact ever find a place in his quadripartite scheme—since the operation of their concepts (of soul and body) by successive displacements destabilizes the very axes that define that space. My point is rather that, as modes of conferring intelligibility to experience—as knowledge practices—the conceptual schemes operating here seem to contrast with, or transform, in potentially interesting ways, the mode Descola’s conceptualization exemplifies.

By the way, Marshall Sahlins’ comments (and Almeida’s diagrams) in the colloquium on the “French ontological turn” published by Hau earlier this year (Kelly 2014) may correspond to a different, but perhaps complementary, series of displacements—produced there by the way the four ontological modes, in being differentially mapped onto different societal arrangements, are made to differ from their original definitions (regarding the form of the anthropos, for instance). These displacements are not equivalent, among themselves or with respect to Amerindian ones, but together they all suggest to me, in the end, a more positive way of relating to the imposing architecture of Beyond nature and culture. It is true that Descola’s response to those comments (reaffirming and marking boundaries, sorting out compatibilities and incompatibilities among his four modes of identification, their respective types of collectives and kinds of subject, and other schemes of practice like the modes of relation) tends to sound somewhat protective, suggesting, together with his avowedly baffled recognition of the “ontodiversity” of Papua New Guinea (Descola 2014: 298), as something of an anomaly, a certain attachment to the ontologies he identifies and oppositions that generate them as substantive. But nothing forbids us to read his interiority/physicality opposition—following his own reading of Lévi-Strauss’s nature/culture “dichotomy”—as a mere (“methodological”) device, “the aim of which is to enable us to reject any substantive opposition between these two domains” (Descola 2009: 108). As a “philosophical fiction,” it would then enable us (again) to see how types and bombs (Latour 2009) need each other to work their effects in this entre-deux mode of making knowledge that is anthropology.


Cesarino, Pedro. 2010. “Donos e duplos: relações de conhecimento, propriedade e autoria entre Marubo.” Revista de Antropologia 53 (1): 147–97.

Coelho de Souza, Marcela. Forthcoming. “Cross-twins and outcestous marriages: How kinship (under)determines humanity for the Kĩsêdjê (Central Brazil).” In The culture of invention in the Americas, edited by Pedro Pitarch and José Antonio Kelly.

Costa, Luiz Antonio. 2007. “As faces do jaguar: Parentesco, história e mitologia entre os Kanamari da Amazônia Ocidental.” PhD thesis, Universidade Federal do Rio de Janeiro.

Crocker, Jon Christopher. 1985. Vital souls: Bororo cosmology, natural symbolism, and shamanism. Tucson: University of Arizona Press.

Descola, Philippe. 1992. “Societies of nature and the nature of society.” In Conceptualizing society, edited by Adam Kuper, 107–26. London: Routledge.

———. 1996. “Constructing natures: Symbolic ecology and social practice.” In Nature and society: Anthropological perspectives, edited by Philippe Descola and Gisli Pálsson, 83–102. London: Routledge.

———. 2009. “The two natures of Lévi-Strauss.” In The Cambridge companion to Lévi-Strauss, edited by Boris Wiseman, 103–17. Cambridge: Cambridge University Press.

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

———. 2014. “The grid and the tree: Reply to Marshall Sahlins’ comment.” Hau: Journal of Ethnographic Theory 4 (1): 295–300.

Fausto, Carlos. 2008. “Donos demais: Maestria e domínio na Amazônia.” Mana. Estudos de Antropologia Social 14 (2): 329–66.

Guerreiro Jr., Antonio R. 2012. “Ancestrais e suas sombras: Uma etnografia da chefia kalapalo e seu ritual mortuário.” PhD. thesis, University of Brasilia.

Kelly, John D., ed. 2014. “Colloquia: The ontological French turn.” Hau: Journal of Ethnographic Theory 4 (1): 357–60.

Latour, Bruno. 2009. Pandora’s hope. Cambridge, MA: Harvard University Press.

Lévi-Strauss, Claude. 1962. La Pensée Sauvage. Paris: PLON.

———. 1977. L'identité. Paris: Presses Universitaires de France.

Lima, Tânia Stolze. 2002. “O que é um Corpo?” Religião e Sociedade 22 (1): 9–20.

———. 2005. Um peixe olhou para mim: O povo Yudjá e a perspectiva. Rio de Janeiro: Editora UNESP/ISA.

———. 2008. “Uma história do dois, do uno e do terceiro.” In Lévi-Strauss: Leituras brasileiras, edited by Ruben Caixeta de Queiroz and Renarde Freire Nobre, 209–64. Editora UFMG.

Seeger, Anthony. 1981. Nature and society in Central Brazil: The Suyá Indians of Mato Grosso. Cambridge, MA: Harvard University Press.

Strathern, Marilyn. 1999. Property, substance and effect: Anthropological essays on persons and things. London: The Athlone Press.

Soares Pinto, Nicole. 2014. “Entre as Teias do Marico: Parentes e pajés djeoromitxi.” PhD diss., University of Brasilia.

Viveiros de Castro, Eduardo. (1998) 2012. Cosmological perspectivism in Amazonia and elsewhere. HAU Masterclass Series 1: 45–81.

———. 2001. “GUT feelings about Amazonia: Potential affinity and the construction of sociality.” In Beyond the visible and the material: The amerindianization of society in the work of Peter Rivière, edited by Laura M. Rival and Neil L. Whitehead, 19–44. Oxford: Oxford University Press.

———. 2009. Métaphysiques cannibale: Lignes d'anthropologie post-structurale. Paris: Presses Universitaires de France.



Marcela Coelho de Souza
Department of Social Anthropology
Universidade de Brasília
Instituto de Ciências Sociais Campus Universitário Darcy Ribeiro
70.910-900 Brasília DF, Brazil