EXCHANGING PERSPECTIVES
The Transformation of Objects into Subjects
in Amerindian Ontologies
Eduardo Viveiros de Castro
At the outset of his reply to Ulrich Beck in this symposium, Bruno Latour
cites a case study of Eduardo Viveiros de Castro and repeats an anecdote he
tells about Amerindians and conquistadors talking at cross-purposes. Latour
deploys the story to illustrate his claim that even the most well-intentioned
and sophisticated peacemakers can get us into worse trouble than we were
in when negotiations began. The problem, he says, is that the likelihood is
low that either side in a communication, let alone a formal negotiation,
knows what the other side thinks is under discussion. Negotiating
contradictory opinions may seem difficult enough, but in cases of deep
enmity, opinions are not what is at stake. The disagreements are
ontological: enemies disagree, as Latour cites Viveiros de Castro saying,
about what world we inhabit. And when peace is achieved, it does not
consist in agreement to a set of opinions or principles; the parties begin,
rather, to live in a different world. The article that follows is not the one to
which Latour refers but a later and related paper, appearing here in English
for the first time. It has already been published, in a somewhat different
version, in Italian and for an anthropological audience. It was not written
for this symposium, in other words, and does not directly respond to either
Latour or Beck; but Viveiros de Castro has revised the article for inclusion
here, and its relevance should be immediately apparent.
—Editor
Common Knowledge 10:3
Copyright 2004 by Duke University Press
463
Symposium: Talking Peace with Gods, Part 1
COMMONKNOWLEDGE464
1. Hypotheses that I have offered previously (“Cosmological
Deixis and Amerindian Perspectivism,” Journal of
the Royal Anthropological Institute, n.s., 4.3 [1998]: 469–88)
are rehearsed here since they ground the argument of this
article. I gave an early version of the present paper, in English,
at the Chicago meeting of the American Anthropological
Association in November 1999, and that version
was subsequently published in Italian as “La transformazione
degli ogetti in sogetti nelle ontologie amerindiane,”
Etnosistemi 7.7 (2000): 47–58. The title of that paper
(a version of which is the subtitle of this essay) pays homage
to Nancy Munn, “The Transformation of Subjects
into Objects in Walbiri and Pitjantjara Myth,” in Australian
Aboriginal Anthropology, ed. Ronald M. Berndt
(Nedlands: University of Western Australia Press, 1970).
2. Claude Lévi-Strauss, Mythologiques, 4 vols. (Paris: Plon,
1964–71).
My subject is the cosmological setting of an indigenous Amazonian model of the
self.1 I will examine two major contexts, shamanism and warfare, in which “self”
and “other” develop especially complex relations. Shamanism deals with the relation
between humans and nonhumans; and in warfare, a human other, an
“enemy,” is used to bring a “self” into existence. I will deliberately use a set of traditional
dichotomies (I mean, in the tradition of modernity) as both heuristic
instruments and foils: nature/culture, subject/object, production/exchange, and
so forth. This very crude technique for setting off the distinctive features of Amazonian
cosmologies carries the obvious risk of distortion, since it is unlikely that
any nonmodern cosmology can be adequately described either by means of such
conceptual polarities or as a simple negation of them (as if the only point of a
nonmodern cosmology were to stand in opposition to our oppositions). But the
technique does have the advantage of showing how unstable and problematic
those polarities can be made to appear, once they have been forced to bear
“unnatural” interpretations and unexpected rearrangements.
Perspectival Multinaturalism
If there is one virtually universal Amerindian notion, it is that of an original state
of nondifferentiation between humans and animals, as described in mythology.
Myths are filled with beings whose form, name, and behavior inextricably mix
human and animal attributes in a common context of intercommunicability, identical
to that which defines the present-day intrahuman world. Amerindian myths
speak of a state of being where self and other interpenetrate, submerged in the
same immanent, presubjective and preobjective milieu, the end of which is precisely
what the mythology sets out to tell. This end is, of course, the well-known
separation of “culture” and “nature”—of human and nonhuman—that Claude
Lévi-Strauss has shown to be the central theme of Amerindian mythology and
which he deems to be a cultural universal.2
In some respects, the Amerindian separation between humans and animals
may be seen as an analogue of our “nature/culture” distinction; there is, however,
at least one crucial difference between the Amerindian and modern, popular
Western versions. In the former case, the separation was not brought about by
a process of differentiating the human from the animal, as in our own evolutionist
“scientific” mythology. For Amazonian peoples, the original common condition of
both humans and animals is not animality but, rather, humanity. The great separation
reveals not so much culture distinguishing itself from nature as nature distancing
itself from culture: the myths tell how animals lost the qualities inherited
or retained by humans. Humans are those who continue as they have always
been. Animals are ex-humans (rather than humans, ex-animals). In some cases,
humankind is the substance of the primordial plenum or the original form of virtually
everything, not just animals. As Gerald Weiss puts it:
Campa mythology is largely the story of how, one by one, the primal
Campa became irreversibly transformed into the first representatives of
various species of animals and plants, as well as astronomical bodies or
features of the terrain. . . . The development of the universe, then, has
been primarily a process of diversification, with mankind as the primal
substance out of which many if not all of the categories of beings and
things in the universe arose, the Campa of today being the descendants
of those ancestral Campa who escaped being transformed.3
The fact that many “natural” species or entities were originally human has
important consequences for the present-day state of the world. While our folk
anthropology holds that humans have an original animal nature that must be
coped with by culture—having been wholly animals, we remain animals “at bottom”—Amerindian
thought holds that, having been human, animals must still
be human, albeit in an unapparent way. Thus, many animal species, as well as
sundry other types of nonhuman beings, are supposed to have a spiritual component
that qualifies them as “people.” Such a notion is often associated with the
idea that the manifest bodily form of each species is an envelope (a “clothing”)
that conceals an internal humanoid form, usually visible to the eyes of only the
particular species and of “transspecific” beings such as shamans. This internal
form is the soul or spirit of the animal: an intentionality or subjectivity formally
identical to human consciousness. If we conceive of humans as somehow composed
of a cultural clothing that hides and controls an essentially animal nature,
Amazonians have it the other way around: animals have a human, sociocultural
inner aspect that is “disguised” by an ostensibly bestial bodily form.
Another important consequence of having animals and other types of nonhumans
conceived as people—as kinds of humans—is that the relations between
the human species and most of what we would call “nature” take on the quality
of what we would term “social relations.” Thus, categories of relationship and
modes of interaction prevailing in the intrahuman world are also in force in most
ViveirosdeCastro•TalkingPeacewithGods:Part1465
3. Gerald Weiss, “Campa Cosmology,” Ethnology 11.2
(April 1972): 169–70.
contexts in which humans and nonhumans confront each other. Cultivated plants
may be conceived as blood relatives of the women who tend them, game animals
may be approached by hunters as affines, shamans may relate to animal and plant
spirits as associates or enemies.
Having been people, animals and other species continue to be people behind
their everyday appearance. This idea is part of an indigenous theory according
to which the different sorts of persons—human and nonhuman (animals, spirits,
the dead, denizens of other cosmic layers, plants, occasionally even objects and
artifacts)—apprehend reality from distinct points of view. The way that humans
perceive animals and other subjectivities that inhabit the world differs profoundly
from the way in which these beings see humans (and see themselves). Under normal
conditions, humans see humans as humans; they see animals as animals, plants
as plants. As for spirits, to see these usually invisible beings is a sure sign that conditions
are not normal. On the other hand, animals (predators) and spirits see
humans as animals (as game or prey) to the same extent that game animals see
humans as spirits or as predator animals. By the same token, animals and spirits
see themselves as humans: they perceive themselves as (or they become) anthropomorphic
beings when they are in their own houses or villages; and, most important,
they experience their own habits and characteristics in the form of culture.
Animals see their food as human food (jaguars see blood as manioc beer, vultures
see the maggots in rotting meat as grilled fish); they see their bodily attributes (fur,
feathers, claws, beaks) as body decorations or cultural instruments; they see their
social system as organized in the same way as human institutions are (with chiefs,
shamans, ceremonies, exogamous moieties, and whatnot).
The contrast with our conceptions in the modern West is, again, only too
clear. Such divergence invites us to imagine an ontology I have called “multinaturalist”
so as to set it off from modern “multiculturalist” ontologies.4 Where
the latter are founded on the mutually implied unity of nature and multiplicity
of cultures—the former guaranteed by the objective universality of body and
substance, the latter generated by the subjective particularity of spirit and meaning—the
Amerindian conception presumes a spiritual unity and a corporeal
diversity. For them, culture or the subject is the form of the universal, while
nature or the object is the form of the particular.
To say that humanity is the original common condition of humans and
nonhumans alike is tantamount to saying that the soul or spirit—the subjective
aspect of being—is the universal, unconditioned given (since the souls of all nonhumans
are humanlike), while objective bodily nature takes on an a posteriori,
particular, and conditioned quality. In this connection, it is also worth noticing
that the notion of matter as a universal substrate seems wholly absent from Ama-
COMMONKNOWLEDGE466
4. See Eduardo Viveiros de Castro, “Cosmological Deixis.”
For a generalization of the notion of “multinaturalism,” see
Bruno Latour, Politiques de la nature (Paris: La Découverte,
1999), and, of course, his contribution to this symposium.
zonian ontologies.5 Reflexive selfhood, not material objectivity, is the potential
common ground of being.
To say, then, that animals and spirits are people is to say that they are persons;
and to personify them is to attribute to nonhumans the capacities of conscious
intentionality and social agency that define the position of the subject.6
Such capacities are reified in the soul or spirit with which these nonhumans are
endowed. Whatever possesses a soul is capable of having a point of view, and
every being to whom a point of view is attributed is a subject; or better, wherever
there is a point of view, there is a “subject position.” Our constructionist epistemology
can be summed up in the Saussurean (and very Kantian) formula, “the
point of view creates the object.”7 The subject, in other words, is the original,
fixed condition whence the point of view emanates (the subject creates the point
of view). Whereas Amerindian perspectival ontology proceeds as though the point
of view creates the subject: whatever is activated or “agented” by the point of view
will be a subject.
The attribution of humanlike consciousness and intentionality (to say nothing
of human bodily form and cultural habits) to nonhuman beings has been
indiscriminately termed “anthropocentrism” or “anthropomorphism.” However,
these two labels can be taken to denote radically opposed cosmological perspectives.
Western popular evolutionism, for instance, is thoroughly anthropocentric
but not particularly anthropomorphic. On the other hand, animism may be characterized
as anthropomorphic but definitely not as anthropocentric: if sundry
other beings besides humans are “human,” then we humans are not a special lot
(so much for “primitive narcissism”).
Karl Marx wrote of man, meaning Homo sapiens:
In creating an objective world by his practical activity, in working-up
inorganic nature, man proves himself a conscious species being. . . .
Admittedly animals also produce. . . . But an animal only produces what
it immediately needs for itself or its young. It produces one-sidedly,
while man produces universally. . . . An animal produces only itself,
whilst man reproduces the whole of nature. . . . An animal forms things
in accordance with the standard and the need of the species to which it
belongs, whilst man knows how to produce in accordance to the standards
of other species.8
Talk about primitive narcissism. . . . Whatever Marx meant by the proposition
that man “produces universally,” I fancy he was saying something to the effect
ViveirosdeCastro•TalkingPeacewithGods:Part1467
5. But see Anne Osborn, “Comer y ser comido: Los animales
en la tradicion oral U’wa (tunebo),” Boletin del Museo
del Oro 26 (1990): 13–41.
6. Animals and other nonhumans are subjects not because
they are human (humans in disguise); rather, they are
human because they are subjects (potential subjects).
7. Ferdinand de Saussure, Cours de linguistique générale
(1916; Paris: Payot, 1981), 23.
8. Karl Marx, Economic and Philosophic Manuscripts of 1844
(Moscow: Foreign Languages Publishing House, 1961),
75–76.
that man is the universal animal: an intriguing idea. (If man is the universal animal,
then perhaps each animal species would be a particular kind of humanity?)
While apparently converging with the Amerindian notion that humanity is the
universal form of the subject, Marx’s is in fact an absolute inversion of the notion.
Marx is saying that humans can be any animal (we have more “being” than any
other species), while Amerindians say that any animal can be human (there is
more “being” to an animal than meets the eye). Man is the universal animal in
two entirely different senses, then: the universality is anthropocentric for Marx;
anthropomorphic, for Amerindians.
The Subjectification of Objects
Much of the Amerindians’ practical engagement with the world presupposes that
present-day nonhuman beings have a spiritual, invisible, prosopomorphic side.
That supposition is foregrounded in the context of shamanism. By shamanism, I
mean the capacity evinced by some individuals to cross ontological boundaries
deliberately and adopt the perspective of nonhuman subjectivities in order to
administer the relations between humans and nonhumans. Being able to see nonhumans
as they see themselves (they see themselves as humans), shamans are able
to take on the role of active interlocutors in transspecific dialogues and are capable
(unlike lay persons) of returning to tell the tale. If a human who is not a shaman
happens to see a nonhuman (an animal, a dead human soul, a spirit) in human
form, he or she runs the risk of being overpowered by the nonhuman subjectivity,
of passing over to its side and being transformed into an animal, a dead human,
a spirit. A meeting or exchange of perspectives is, in brief, a dangerous business.
Shamanism is a form of acting that presupposes a mode of knowing, a particular
ideal of knowledge. That ideal is, in many respects, the exact opposite of
the objectivist folk epistemology of our tradition. In the latter, the category of
the object supplies the telos: to know is to objectify—that is, to be able to distinguish
what is inherent to the object from what belongs to the knowing subject
and has been unduly (or inevitably) projected into the object. To know, then, is
to desubjectify, to make explicit the subject’s partial presence in the object so as
to reduce it to an ideal minimum. In objectivist epistemology, subjects as much
as objects are seen as the result of a process of objectification. The subject constitutes/recognizes
itself in the objects it produces, and the subject knows itself
objectively when it comes to see itself from the outside as an “it.” Objectification
is the name of our game; what is not objectified remains unreal and abstract. The
form of the other is the thing.
Amerindian shamanism is guided by the opposite ideal. To know is to personify,
to take on the point of view of that which must be known. Shamanic
knowledge aims at something that is a someone—another subject. The form of
the other is the person. What I am defining here is what anthropologists of yore
COMMONKNOWLEDGE468
used to call animism, an attitude that is far more than an idle metaphysical tenet,
for the attribution of souls to animals and other so-called natural beings entails
a specific way of dealing with them. Being conscious subjects able to communicate
with humans, these natural beings are able fully to reciprocate the intentional
stance that humans adopt with respect to them.
Recently, there has been a new surge of interest in animism.9 Cognitive
anthropologists and psychologists have been arguing that animism is an “innate”
cognitive attitude that has been naturally selected for its attention-grabbing
potential and its practical predictive value.10 I have no quarrel with these hypotheses.
Whatever the grounds of its naturalness, however, animism can also be very
much cultural—that is, animism can be put to systematic and deliberate use. We
must observe that Amerindians do not spontaneously see animals and other nonhumans
as persons; the personhood or subjectivity of the latter is considered a
nonevident aspect of them. It is necessary to know how to personify nonhumans,
and it is necessary to personify them in order to know.11
Personification or subjectification implies that the “intentional stance”
adopted with respect to the world has been in some way universalized. Instead of
reducing intentionality to obtain a perfectly objective picture of the world, animism
makes the inverse epistemological bet. True (shamanic) knowledge aims to
reveal a maximum of intentionality or abduct a maximum of agency (here I am
using Alfred Gell’s vocabulary).12 A good interpretation, then, would be one able
to understand every event as in truth an action, an expression of intentional states
or predicates of some subject. Interpretive success is directly proportional to the
ordinal magnitude of intentionality that the knower is able to attribute to the
known.13 A thing or a state of affairs that is not amenable to subjectification—to
ViveirosdeCastro•TalkingPeacewithGods:Part1469
9. See especially Philippe Descola, “Constructing Natures:
Symbolic Ecology and Social Practice,” in Nature and Society:
Anthropological Perspectives, ed. Descola and Gísli Pálsson
(London: Routledge, 1996), 82–102; and Nurit BirdDavid,
“‘Animism’ Revisited: Personhood, Environment,
and Relational Epistemology,” Current Anthropology 40,
supp. (February 1999): 67–91.
10. See Pascal Boyer, “What Makes Anthropomorphism
Natural: Intuitive Ontology and Cultural Representations,”
Journal of the Royal Anthropological Institute, n.s., 2.1
(March 1996): 83–97; and Stewart Guthrie, Faces in the
Clouds: A New Theory of Religion (New York: Oxford University
Press, 1993).
11. “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-interpretations.” Marilyn Strathern, Property,
Substance, and Effect: Anthropological Essays on Persons
and Things (London: Athlone, 1999), 239.
12. Alfred Gell, Art and Agency: An Anthropological Theory
(Oxford: Clarendon, 1998).
13. I am referring here to Daniel Dennett’s idea of
n-order intentional systems: a second-order intentional
system is one to which the observer must ascribe not only
beliefs, desires, and other intentions, but beliefs (etc.) about
other beliefs (etc.). The standard cognitive thesis holds
that only humans exhibit second- or higher-order intentionality.
My shamanistic “principle of abduction of a maximum
of agency” runs afoul 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 behaviour.” Daniel Dennett,
Brainstorms: Philosophical Essays on Mind and Psychology
(Harmondsworth, U.K.: Penguin, 1978), 274.
determination of its social relation to the knower—is shamanistically uninteresting.
Our objectivist epistemology follows the opposite course: it considers our
commonsense intentional stance as just a shorthand that we use when the behavior
of a target-object is too complicated to be broken down into elementary physical
processes. An exhaustive scientific interpretation of the world would for us
be able ideally to reduce every action to a chain of causal events and to reduce
these events to materially dense interactions (with no “action at a distance”).14
If in the naturalist view a subject is an insufficiently analyzed object, in the
Amerindian animist cosmology the converse holds: an object is an incompletely
interpreted subject. The object must either be “expanded” to a full-fledged subject—a
spirit; an animal in its human, reflexive form—or else understood as
related to a subject (as existing, in Gell’s terms, “in the neighbourhood” of an
agent). But an important qualification must now be made: Amerindian cosmologies
do not as a rule attribute personhood (or the same degree of personhood)
to each type of entity in the world. In the case of animals, for instance, the
emphasis seems to be on those species that perform key symbolic and practical
roles, such as the great predators and the principal species of prey for humans.
Personhood and “perspectivity”—the capacity to occupy a point of view—is a
question of degree and context rather than an absolute, diacritical property of
particular species.
Still, despite this qualification, what cannot be conceived as a primary agent
or subject in its own right must be traced up to one:
“Social agents” can be drawn from categories which are as different as
chalk and cheese . . . because “social agency” is not defined in terms of
“basic” biological attributes (such as inanimate thing vs. incarnate person)
but is relational—it does not matter, in ascribing “social agent” status,
what a thing (or a person) “is” in itself; what matters is where it
stands in a network of social relations. All that may be necessary for
stocks and stones to become “social agents” . . . is that there should be
actual human persons/agents “in the neighbourhood” of these inert
objects.15
Though there are Amazonian cosmologies that deny to postmythical nonhuman
species any spiritual dimension, the notion (widespread, as is well known,
throughout the continent) of animal or plant “spirit masters” supplies the missing
agency. These spirit masters, equipped with an intentionality fully equivalent
COMMONKNOWLEDGE470
14. Cf. Lévi-Strauss, La pensée sauvage (Paris: Plon, 1962),
355: “La pensée sauvage est logique, dans le même sens et
de la même façon que la nôtre, mais comme l’est seulement
la nôtre quand elle s’applique à la connaissance d’un
univers auquel elle reconnaît simultanément des propriétés
physiques et des propriétés sémantiques.”
15. Gell, Art and Agency, 123.
to that of humans, function as hypostases of the species with which they are associated,
thereby creating an intersubjective field for human/nonhuman relations
even where empirical nonhuman species are not spiritualized. Moreover, the idea
that nonhuman agents experience themselves and their behavior in the forms of
(human) culture plays a crucial role: translating culture into the terms of alien
subjectivities transforms many natural objects and events into indices from which
social agency is derivable. The commonest case is that of defining what to
humans is a brute fact or object as an artifact or cultured behavior: what is blood
to us is manioc beer to jaguars, a muddy waterhole is seen by tapirs as a great ceremonial
house. Artifacts have this interestingly ambiguous ontology. They are
objects that necessarily point to a subject; as congealed actions, they are material
embodiments of nonmaterial intentionality. What is nature to us may well be culture
to another species.
Perspectivism Is Not Relativism
The idea of a world comprising a multiplicity of subject positions looks very
much like a form of relativism. Or rather, relativism under its various definitions
is often implied in the ethnographic characterization of Amerindian cosmologies.
Take, for instance, the work of Kaj Århem, the ethnographer of the
Makuna. Having described the elaborate perspectival universe of this Tukanoan
people of northwestern Amazonia, Århem observes that the notion of multiple
viewpoints on reality implies that, as far as the Makuna are concerned, “every
perspective is equally valid and true” and that “a correct and true representation
of the world does not exist.”16 Århem is right, of course; but only in a sense. For
one can reasonably surmise that as far as humans are concerned, the Makuna
would say that there is indeed only one correct and true representation of the
world. If you start seeing, for instance, the maggots in rotten meat as grilled fish,
you may be sure that you are in deep trouble, but grilled fish they are from the
vultures’ point of view. Perspectives should be kept separate. Only shamans, who
are so to speak species-androgynous, can make perspectives communicate, and
then only under special, controlled conditions.
My real point, however, is best put as a question: does the Amerindian perspectivist
theory posit, as Århem maintains that it does, a multiplicity of representations
of the same world? It is sufficient to consider ethnographic evidence to
see that the opposite is the case: all beings perceive (“represent”) the world in the
ViveirosdeCastro•TalkingPeacewithGods:Part1471
16. Kaj Århem, “Ecosofía Makuna,” in La selva humanizada:
Ecología alternativa en el trópico húmedo colombiano, ed.
François Correa (Bogotá: Instituto Colombiano de
Antropología; Fondo FEN Colombia; Fondo Editorial
CEREC, 1993), 124.
same way. What varies is the world that they see. Animals impose the same categories
and values on reality as humans do—their worlds, like ours, revolve
around hunting and fishing, cooking and fermented drinks, cross-cousins and
war, initiation rituals, shamans, chiefs, spirits, and so forth. Being people in their
own sphere, nonhumans see things just as people do. But the things that they see
are different. Again, what to us is blood is maize beer to the jaguar; what to us
is soaking manioc is, to the souls of the dead, a rotting corpse; what is a muddy
waterhole to us is for the tapirs a great ceremonial house.
Another good discussion of Amazonian “relativism” can be found in a study
of the Matsiguenga by France-Marie Renard-Casevitz. Commenting on a myth
in which the human protagonists travel to villages inhabited by strange people
who call the snakes, bats, and balls of fire that they eat by the names of foods
(“fish,” “agouti,” “macaws”) appropriate for human consumption, she realizes that
indigenous perspectivism is quite different from relativism. Yet she sees no special
problem:
This setting in perspective [mise en perspective] is just the application and
transposition of universal social practices, such as the fact that a mother
and a father of X are the parents-in-law of Y. . . . This variability of the
denomination as a function of the place occupied explains how A can be
both fish for X and snake for Y.17
But applying the positional relativity that obtains in social and cultural terms to
the difference between species has a paradoxical consequence: Matsiguenga preferences
are universalized and made absolute. A human culture is thus rendered
natural—everybody eats fish and nobody eats snake.
Be that as it may, Casevitz’s analogy between kinship positions and what
counts as fish or snake for different species remains intriguing. Kinship terms are
relational pointers; they belong to the class of nouns that define something in
terms of its relations to something else (linguists have special names for such
nouns—“two-place predicates” and such like). Concepts like fish or tree, on the
other hand, are proper, self-contained substantives: they are applied to an object
by virtue of its intrinsic properties. Now, what seems to be happening in Amerindian
perspectivism is that substances named by substantives like fish, snake,
hammock, or beer are somehow used as if they were relational pointers, something
halfway between a noun and a pronoun, a substantive and a deictic. (There is supposedly
a difference between “natural kind” terms such as fish and “artifact” terms
such as hammock: a subject worth more discussion later.) You are a father only
COMMONKNOWLEDGE472
17. France-Marie Renard-Casevitz, Le banquet masqué:
Une mythologie de l’étranger chez les indiens Matsiguenga
(Paris: Lierre and Coudrier, 1991), 29.
because there is another person whose father you are. Fatherhood is a relation,
while fishiness is a intrinsic property of fish. In Amerindian perspectivism, however,
something is a fish only by virtue of someone else whose fish it is.
But if saying that crickets are the fish of the dead or that mud is the hammock
of tapirs is like saying that my sister Isabel’s son, Miguel, is my nephew,
then there is no relativism involved. Isabel is not a mother “for” Miguel, from
Miguel’s “point of view” in the usual, relativist-subjectivist sense of the expression.
Isabel is the mother of Miguel, she is really and objectively Miguel’s mother,
just as I am really Miguel’s uncle. This is a genitive, internal relation (my sister
is the mother of someone, our cricket the fish of someone) and not a representational,
external connection of the type “X is fish for someone,” which implies
that X is “represented” as fish, whatever X is “in itself.” It would be absurd to
say that, since Miguel is the son of Isabel but not mine, then Miguel is not a son
“for me”—for indeed he is. He is my sister’s son, precisely.
Now imagine that all Amerindian substances were of this sort. Suppose
that, as siblings are those who have the same parents, conspecifics are those that
have the same fish, the same snake, the same hammock, and so forth. No wonder,
then, that animals are so often conceived, in Amazonia, as affinely related
to humans. Blood is to humans as manioc beer is to jaguars in exactly the way that
my sister is the wife of my brother-in-law. The many Amerindian myths featuring
interspecific marriages and discussing the difficult relationships between the
human (or animal) in-marrying affine and his or her animal (or human) parentsin-law,
simply compound the two analogies into a single complex one. We begin
to see how perspectivism may have a deep connection with exchange—not only
how it may be a type of exchange, but how any exchange is by definition an
exchange of perspectives.18
We would thus have a universe that is 100 percent relational—a universe
in which there would be no distinctions between primary and secondary qualities
of substances or between “brute facts” and “institutional facts.” This distinction,
championed by John Searle, opposes brute facts or objects, the reality
of which is independent of human consciousness (gravity, mountains, trees, animals,
and all “natural kinds”) to institutional facts or objects (marriage, money,
axes, and cars) that derive their existence, identity, and efficacy from the culturally
specific meanings given them by humans.19 In this overhauled version of the
nature/culture dualism, the terms of cultural relativism apply only to cultural
objects and are balanced by the terms of natural universalism, which apply to nat-
ViveirosdeCastro•TalkingPeacewithGods:Part1473
18. See Strathern, The Gender of the Gift: Problems with
Women and Problems with Society in Melanesia (Berkeley:
University of California Press, 1988) and “Writing Societies,
Writing Persons,” History of the Human Sciences 5.1
(February 1992): 5–16.
19. John Searle, The Construction of Social Reality (London:
Allen Lane, 1995).
ural objects. Searle would argue, I suppose, that what I am saying is that for
Amerindians all facts are of the institutional, mental variety, and that all objects,
even trees and fish, are like money or hammocks, in that their only reality (as
money and hammocks, not as pieces of paper or of string) derives from the meanings
and uses that subjects attribute to them. This would be nothing but relativism,
Searle would observe—and an absolute form of relativism at that.
An implication of Amerindian perspectivist animism is, indeed, that there
are no autonomous, natural facts, for what we see as nature is seen by other
species as culture (as institutional facts). What humans see as blood, a natural substance,
is seen by jaguars as manioc beer, an artifact. But such institutional facts
are taken to be universal, culturally invariable (an impossibility according to
Searle). Constructionist relativism defines all facts as institutional and thus culturally
variable. We have here a case not of relativism but universalism—cultural
universalism—that has as its complement what has been called “natural relativism.”20
And it is this inversion of our usual pairing of nature with the universal
and culture with the particular that I have been terming “perspectivism.”
Cultural (multicultural) relativism supposes a diversity of subjective and
partial representations, each striving to grasp an external and unified nature,
which remains perfectly indifferent to those representations. Amerindian thought
proposes the opposite: a representational or phenomenological unity that is
purely pronominal or deictic, indifferently applied to a radically objective diversity.
One culture, multiple natures—one epistemology, multiple ontologies. Perspectivism
implies multinaturalism, for a perspective is not a representation. A
perspective is not a representation because representations are a property of the
mind or spirit, whereas the point of view is located in the body. The ability to
adopt a point of view is undoubtedly a power of the soul, and nonhumans are
subjects in so far as they have (or are) spirit; but the differences between viewpoints
(and a viewpoint is nothing if not a difference) lies not in the soul. Since
the soul is formally identical in all species, it can only perceive the same things
everywhere. The difference is given in the specificity of bodies.
This formulation permits me to provide answers to a couple of questions
that may have already occurred to my readers. If nonhumans are persons and
have souls, then what distinguishes them from humans? And why, being people,
do they not regard us as people?
Animals see in the same way as we do different things because their bodies
differ from ours. I am not referring to physiological differences—Amerindians
recognize a basic uniformity of bodies—but rather to affects, in the old sense of
dispositions or capacities that render the body of each species unique: what it eats,
COMMONKNOWLEDGE474
20. See Latour, Nous n’avons jamais été modernes (Paris: La
Découverte, 1991), 144.
how it moves, how it communicates, where it lives, whether it is gregarious or
solitary. The visible shape of the body is a powerful sign of these affectual differences,
although the shape can be deceptive, since a human appearance could,
for example, be concealing a jaguar affect. Thus, what I call “body” is not a synonym
for distinctive substance or fixed shape; body is in this sense an assemblage
of affects or ways of being that constitute a habitus. Between the formal subjectivity
of souls and the substantial materiality of organisms, there is thus an intermediate
plane occupied by the body as a bundle of affects and capacities. And the
body is the origin of perspectives.
Solipsism or Cannibalism
The status of humans in modern thought is essentially ambiguous. On the one
hand, humankind is an animal species among other such, and animality is a
domain that includes humans; on the other hand, humanity is a moral condition
that excludes animals.21 These two statuses coexist in the problematic and disjunctive
notion of “human nature.” In other words, our cosmology postulates a
physical continuity and a metaphysical discontinuity between humans and animals,
the continuity making of humankind an object for the natural sciences and
the discontinuity making of humanity an object for the humanities. Spirit or
mind is the great differentiator: it raises us above animals and matter in general,
it distinguishes cultures, it makes each person unique before his or her fellow
beings. The body, in contrast, is the major integrator: it connects us to the rest
of the living, united by a universal substrate (DNA, carbon chemistry) that, in
turn, links up with the ultimate nature of all material bodies. Conversely,
Amerindians postulate metaphysical continuity and physical discontinuity. The
metaphysical continuity results in animism; the physical discontinuity (between
the beings of the cosmos), in perspectivism. The spirit or soul (here, a reflexive
form, not an immaterial inner substance) integrates. Whereas the body (here, a
system of intensive affects, not an extended material organism) differentiates.22
This cosmological picture, which understands bodies as the great differ-
ViveirosdeCastro•TalkingPeacewithGods:Part1475
21. See Tim Ingold, “Becoming Persons: Consciousness
and Sociality in Human Evolution,” Cultural Dynamics 4.3
(1991): 355–78; and Ingold, ed., Companion Encyclopedia of
Anthropology: Humanity, Culture, and Social Life, s.v.
“Humanity and Animality.”
22. The counterproof of the singularity of the spirit in
modern cosmologies lies in the fact that when we try to
universalize it, we are obliged—now that supernature is
out of bounds—to identify it with the structure and function
of the brain. The spirit can only be universal (natural)
if it is (in) the body. It is no accident, I believe, that this
movement of inscription of the spirit in the brain-body or
in matter in general—AI, Churchland’s “eliminative materialism,”
Dennett-style “functionalism,” Sperberian cognitivism,
etc.—has been synchronically countered by its
opposite, the neophenomenological appeal to the body as
the site of subjective singularity. Thus, we have been witnessing
two seemingly contradictory projects of “embodying”
the spirit: one actually reducing it to the body as
traditionally (i.e., biophysically) understood, the other
upgrading the body to the traditional (i.e., culturaltheological)
status of “spirit.”
entiators, at the same time posits their inherent transformability: interspecific
metamorphosis is a fact of nature. Not only is metamorphosis the standard etiological
process in myth, but it is still very much possible in present-day life
(being either desirable or undesirable, inevitable or evitable, according to circumstances).
Spirits, the dead, and shamans can assume animal form, beasts turn
into other beasts, humans inadvertently turn into animals. No surprises here: our
own cosmology presumes a singular distinctiveness of minds but not even for this
reason does it hold communication to be impossible (albeit solipsism is a constant
problem). Nor does our cosmology discredit the mental/spiritual transformations
induced by such processes as education and religious conversion.
Indeed, it is because the spiritual is the locus of difference that conversion
becomes a necessary idea. Bodily metamorphosis is the Amerindian counterpart
to the European theme of spiritual conversion. Shamans are transformers (and
likewise, the mythical demiurges who transformed primal humans into animals
are themselves shamans). Shamans can see animals in their inner human form
because they don animal “clothing” and thus transform themselves into animals.
Solipsism and metamorphosis are related in the same way. Solipsism is the
phantom that threatens our cosmology, raising the fear that we will not recognize
ourselves in our “own kind” because, given the potentially absolute singularity
of minds, our “own kind” are actually not like us. The possibility of metamorphosis
expresses the fear—the opposite fear—of no longer being able to
differentiate between human and animal, and above all the fear of seeing the
human who lurks within the body of the animal that one eats. Our traditional
problem in the West is how to connect and universalize: individual substances
are given, while relations have to be made. The Amerindian problem is how to
separate and particularize: relations are given, while substances must be defined.
Hence the importance, in Amazonia, of dietary rules linked to the spiritual
potency of animals. The past humanity of animals is added to their present-day
spirituality, and both are hidden by their visible form. The result is an extended
set of food restrictions or precautions that declare inedible animals that were,
in myth, originally consubstantial with humans—though some animals can be
desubjectified by shamanic means and then consumed.23 Violation of food
restrictions exposes the violator to illness, conceived of as a cannibal counterpredation
undertaken by the spirit of the prey (turned predator) in a lethal inversion
of perspectives that transforms human into animal. Thus cannibalism is the
Amerindian parallel to our own phantom—solipsism. The solipsist is uncertain
whether the natural similarity of bodies guarantees a real community of spirit.
COMMONKNOWLEDGE476
23. Desubjectification is accomplished by neutralizing the
spirit, transubstantiating the meat into plant food, or
semantically reducing the animal subject to a species less
proximate to humans.
Whereas the cannibal suspects that the similarity of souls prevails over real differences
of body and thus that all animals eaten, despite efforts to desubjectivize
them, remain human. To say that these uncertainties or suspicions are phantoms
haunting their respective cultures does not mean, of course, that there are not
solipsists among us (the more radical relativists, for instance), nor that there are
not Amerindian societies that are purposefully and more or less literally canni-
balistic.
Exchange as Transformation
The idea of creation ex nihilo is virtually absent from indigenous cosmogonies.
Things and beings normally originate as a transformation of something else: animals,
as I have noted, are transformations of a primordial, universal humanity.
Where we find notions of creation at all—the fashioning of some prior substance
into a new type of being—what is stressed is the imperfection of the end product.
Amerindian demiurges always fail to deliver the goods. And just as nature
is the result not of creation but of transformation, so culture is a product not of
invention but of transference (and thus transmission, tradition). In Amerindian
mythology, the origin of cultural implements or institutions is canonically explained
as a borrowing—a transfer (violent or friendly, by stealing or by learning,
as a trophy or as a gift) of prototypes already possessed by animals, spirits,
or enemies. The origin and essence of culture is acculturation.
The idea of creation/invention belongs to the paradigm of production:
production is a weak version of creation but, at the same time, is its model. Both
are actions in—or rather, upon and against—the world. Production is the imposition
of mental design on inert, formless matter. The idea of transformation/
transfer belongs to the paradigm of exchange: an exchange event is always the
transformation of a prior exchange event. There is no absolute beginning, no
absolutely initial act of exchange. Every act is a response: that is, a transformation
of an anterior token of the same type. Poiesis, creation/production/invention,
is our archetypal model for action; praxis, which originally meant something like
transformation/exchange/transfer, suits the Amerindian and other nonmodern
worlds better.24 The exchange model of action supposes that the the subject’s
“other” is another subject (not an object); and subjectification is, of course, what
perspectivism is all about.25 In the creation paradigm, production is causally primary;
and exchange, its encompassed consequence. Exchange is a “moment” of
ViveirosdeCastro•TalkingPeacewithGods:Part1477
24. From the point of view of a hypothetical Amerindian
philosopher, I would say that the Western obsession with
production reveals it as the last avatar of the biblicotheological
category of creation. Humans were not only
created in the likeness of God, they create after His own
image: they “produce.” Ever since God “died,” humans
have produced themselves after their own image (and that
is what culture is about, I suppose).
25. See Strathern, “Writing Societies,” 9–10.
production (it “realizes” value) and the means of reproduction. In the transformation
paradigm, exchange is the condition for production since, without the
proper social relations with nonhumans, no production is possible. Production
is a type or mode of exchange, and the means of “reexchange” (a word we certainly
do not need, for exchange is by definition reexchange). Production creates;
exchange changes.
I would venture a further remark on this contrast: the idiom of material
production, if applied outside the original domain of poiesis, is necessarily metaphorical.
When we speak of the production of persons (social reproduction) or
the production of “symbolic capital” as if we meant the production of subjects
rather than simply of human organisms, we are being no less metaphorical than
when we apply the idiom of praxis to engagements between humans and nonhumans.
To speak of the production of social life makes as much, or as little, sense
as to speak of an exchange between humans and animals. Metaphorical Marx is
not necessarily better than metaphorical Mauss.
I would speculate, further, that the emphasis on transformation/exchange
(over creation/production) is organically connected to the predominance of
affinal relations (created by marriage alliance) over consanguineal ones (created
by parenthood) in Amerindian mythology. The protagonists of the major
Amerindian myths are related agonistically as siblings-in-law, parents-in-law,
children-in-law. Our own Old World mythology (Greek, Near Eastern, or Freudian)
seems haunted, on the other hand, by parenthood and especially fatherhood.
Not to put too fine a point on it: we had to steal fire from a divine father;
Amerindians had to steal it from an animal father-in-law. Mythology is a discourse
on the given, the innate. Myths address what must be taken for granted,
the initial conditions with which humanity must cope and against which humanity
must define itself by means of its power of “convention.”26 If such is the case,
then in the Amerindian world, affinity and alliance (exchange) rather than parenthood
(creation/production) comprise the given—the unconditioned condi-
tion.
The Cannibal Cogito
The analogy between shamans and warriors in Amerindian ethnographies has
often been observed. Warriors are to the human world what shamans are to the
universe at large: conductors or commutators of perspectives. That shamanism
is warfare writ large has nothing to do with violence (though shamans often act
as warriors in the literal sense). But indigenous warfare belongs to the same cos-
COMMONKNOWLEDGE478
26. See Roy Wagner, The Invention of Culture (Chicago:
University of Chicago Press, 1981).
mological complex as shamanism, insofar as both involve the embodiment by the
self of the enemy’s point of view.27 Accordingly, in Amazonia, what is intended in
ritual exocannibalism is incorporation of the subjecthood of a hypersubjectified
enemy. The intent is not (as it is in hunting game animals) desubjectification.
The subjectification of human enemies is a complex ritual process. Suffice
it to say, for our purposes here, that the process supposes a thorough identification
of the killer with its victim, just as shamans become the animals whose
bodies they procure for the rest of their group. Killers derive crucial aspects of
their social and metaphysical identities from their victims—names, surplus souls,
songs, trophies, ritual perogatives; but in order to do so, a killer must first become
his enemy. A telling example is the Araweté war song in which a killer repeats
words taught him by the spirit of the victim during the ritual seclusion that follows
the deed: the killer speaks from the enemy’s standpoint, saying “I” to refer
to the enemy and “him” to refer to himself.28 In order to become a full subject—
for the killing of an enemy is often a precondition to adult male status—the killer
must apprehend the enemy “from the inside” (as a subject). The analogy with the
animist perspectival theory already discussed is clear: nonhuman subjectivities
see humans as nonhumans (and vice versa). Here, the killer must be able to see
himself as the enemy sees him—as, precisely, an enemy—in order to become
“himself” or, rather, a “myself.” It is relevant in this connection to recall that the
archetypal idiom of enmity, in Amazonia, is affinity. Enemies are conceptualized
as “ideal” brothers-in-law, uncontaminated by the exchange of sisters (which
would “consanguinize” them—make them cognates of one’s children—and thus
less than pure affines).
In this idiom of enmity, then, neither party is an object. Enmity of this sort
is a reciprocal subjectification: an exchange, a transfer, of points of view. It is a
ritual transformation of the self (to use Simon Harrison’s term) that belongs entirely
to the “exchange” (not the “production”) paradigm of action—though the
exchange in this case is very extreme. Harrison describes the situation in a
Melanesian context that closely resembles the Amazonian: “Just as a gift embodies
the identity of its donor, so in Lowland warfare the killer acquires through
homicide an aspect of his victim’s identity. The killing is represented as either
creating or expressing a social relationship, or else as the collapse of a social
relation by the merging of two social alters into one.”29 The synthesis of the gift
relates subjects who remain objectively separated—they are divided by the rela-
ViveirosdeCastro•TalkingPeacewithGods:Part1479
27. See Viveiros de Castro, From the Enemy’s Point of View:
Humanity and Divinity in an Amazonian Society (Chicago:
University of Chicago Press, 1992).
28. See Viveiros de Castro, “Le meurtrier et son double
chez les Araweté: Un exemple de fusion rituelle,” Systèmes
de Pensée en Afrique Noire 14 (1996): 77–104.
29. Simon Harrison, The Mask of War: Violence, Ritual, and
the Self in Melanesia (Manchester, U.K.: Manchester University
Press, 1993), 130.
tion.30 The killing of an enemy and its symbolic incorporation by the killer, on
the other hand, produces a synthesis in which all distance is suppressed: the relation
is created by abolishing one of its terms, which is then introjected by the
other. The reciprocal dependence of exchange partners becomes inseparability
here, a kind of fusion.
Ontological predation appears to be the crucial idiom of subjectification in
Amazonia. The relative and relational status of predator and prey is fundamental
to the inversions in perspective that obtain between humans and nonhumans.
Again, the Melanesian context, as Harrison describes it, presents striking parallels
to that of Amazonia: “Aggression is conceived as very much a communicative
act directed against the subjectivity of others, and making war required the
reduction of the enemy, not to the status of a non-person or thing but, quite the
opposite, to an extreme state of subjectivity.”31 Which means, Harrison concludes,
that enmity in these societies “is conceptualised not as a mere objective
absence of a social relationship but as a definite social relationship like any other”
(128). This remark brings to mind a well-known passage from Lévi-Strauss:
Les observateurs ont été souvent frappés par l’impossibilité, pour les
indigènes, de concevoir une relation neutre, ou plus exactement une
absence de relation . . . l’absence de relation familiale ne définit pas rien,
elle définit l’hostilité . . . il n’est pas davantage possible de se tenir en
deçà, ou au delà, du monde des relations.32
“Pour les indigènes,” no difference is indifferent and must immediately be
invested with positivity. Enmity is a full-blown social relationship. Not, however,
a relationship like any other: I would go a bit farther than Harrison and say that
the overall schema of difference in Amazonia is cannibalistic predation. At the
risk of falling into allegorical excess, I would even venture to say that, in Amazonian
cosmologies, the generic attributive proposition is a cannibal proposition.
The copula of all synthetic a priori judgments, in a universe articulated by a
“logic of sensory qualities,” is carnivorous copulation. Let me insist: these predatory
relations are fully and immediately social relations. We are dealing here with
a mode of subjectification, internal to the monde des relations to which LéviStrauss
refers. That world has nothing to do with production and objectification,
modes of action that suppose a neutral relationship in which an active and exclusively
human subject confronts an inert and naturalized object. In the monde de
relations, the self is the gift of the other.
COMMONKNOWLEDGE480
30. See Strathern, Gender of the Gift.
31. Harrison, Mask of War, 121.
32. Lévi-Strauss, Les structures élémentaires de la parenté, 2d
ed. (1949; La Haye: Mouton, 1967), 552–53.
Some Conclusions
Our current notions of the social are inevitably polarized by the oppositions I
have been evoking: representation/reality, culture/nature, human/nonhuman,
mind/body, and the rest. In particular, the social presupposes the nonsocial (the
natural). It is impossible to rethink the social without rethinking the natural, for
in our cosmological vulgate, nature (always in the singular) is the encompassing
term, and society (often used in the plural) is the term encompassed.
The contrast between our basic naturalism and Amerindian cosmologies can
be phrased in the terms of our own polarities. Animism could be defined as an
ontology that postulates a social character to relations between humans and nonhumans:
the space between nature and society is itself social. Naturalism is
founded on the inverse axiom: relations between society and nature are themselves
natural. Indeed, if in the animic mode the distinction “nature/culture” is internal
to the social world, humans and animals being immersed in the same sociocosmic
medium (and in this sense, nature is a part of an encompassing sociality),
then in naturalist ontology, the distinction “nature/culture” is internal to nature
(and in this sense, human society is one natural phenomenon among others).
Animism has society, and naturalism has nature, as its unmarked pole: these
poles function, respectively and contrastingly, as the universal dimension of
each mode. This phrasing of the contrast between animism and naturalism is
not only reminiscent of, or analogous to, the famous (some would say notorious)
contrast between gift and commodity—I take it to be the same contrast, expressed
in more general, noneconomic terms.33 Likewise the distinction that I have made
here between production/creation (naturalism) and exchange/transformation
(animism).
In our naturalist ontology, the nature/society interface is natural: humans
are organisms like all the rest—we are body-objects in ecological interaction with
other bodies and forces, all of them ruled by the necessary laws of biology and
physics. Productive forces harness, and thereby express, natural forces. Social
relations—that is, contractual or instituted relations between subjects—can only
exist internal to human society (there is no such thing as “relations of production”
linking humans to animals or plants, let alone political relations). But how
alien to nature—this is the problem of naturalism—are these social relations?
Given the universality of nature, the status of the human and social world is
unstable. Thus, Western thought oscillates, historically, between a naturalistic
monism (sociobiology and evolutionary psychology being two of its current
ViveirosdeCastro•TalkingPeacewithGods:Part1481
33. “If in a commodity economy things and persons
assume the social form of things, then in a gift economy
they assume the social form of persons.” Chris A. Gregory,
Gifts and Commodities (London: Academic, 1982), 41, as
cited in Strathern, Gender of the Gift, 134.
avatars) and an ontological dualism of nature and culture (“culturalism” and symbolic
anthropology being two of its recent expressions).
Still, for all its being the polar opposite of naturalistic monism, the dualism
“nature/culture” discloses the ultimate referential character of the notion of
nature by revealing itself to be directly descended from the theological opposition
between nature and the supernatural. Culture is the modern name for
Spirit—I am thinking of the distinction between Naturwissenschaften and Geisteswissenschaften;
or at least culture names the compromise between nature and
grace. Of animism, I am tempted to say that the instability is of an opposite kind:
there, the problem is how to administer the mixture of humanity and animality
that constitutes animals, rather than, as is the case among ourselves, how to
administer the combination of culture and nature that characterizes humans.
Amerindian perspectivism might be viewed as a radical polytheism (or
rather, henotheism) applied to a universe that supports no dualism between created
matter and Creator Spirit. I am led to ask whether our own naturalistic
monism is not the last avatar of our monotheistic cosmology.34 Our ontological
dualisms derive ultimately from the fundamental difference between Creator and
creature. Killing off the Creator, as some say we have done, has left us with a
creature whose unity depends on the now-absent God. For God prepared science,
and the transcendence of transcendence has created immanence.35 This
birthmark is visible on all modern efforts to dispose of dualisms. Our monistic
ontologies are always derived from some prior duality—they consist essentially
in the erasure of one of the terms or in the absorption (sometimes “dialectical”)
of the erased term by the remaining one. A genuine monism, anterior and exterior
to the great divide between Creator and creature, seems beyond our reach.
A lesson we can usefully draw from Amerindian perspectivism is that the relevant
conceptual pair may be monism and pluralism: multiplicity, not mere duality, is
the complement of the monism I am contemplating. Virtually all attacks on
Cartesian and other dualisms consider that two is already too much—we need
just one (one principle, one substance, one reality). As far as Amerindian cosmologies
are concerned, it would appear that two is not enough.
My problem with the notion of relativism, or with the opposition between
relativism and universalism, pertains to the concept that underwrites such categories
and oppositions: the concept of representation. And my problem with representation
is the ontological poverty it implies—a poverty characteristic of
COMMONKNOWLEDGE482
34. The question is also posed in Latour, Nous n’avons
jamais été modernes, and in Marshall Sahlins, “The Sadness
of Sweetness: The Native Anthropology of Western Cosmology,”
Current Anthropology 37.3 ( June 1996): 395–
428—to mention only two recent works of anthropology.
35. Amos Funkenstein, Theology and the Scientific Imagination
from the Middle Ages to the Seventeenth Century
(Princeton, NJ: Princeton University Press, 1986).
modern thought. The Cartesian break with medieval scholasticism produced a
radical simplification of European ontology by positing only two principles or
substances: unextended thought and extended matter. Modern thought began
with that simplification; and its massive conversion of ontological into epistemological
questions (questions of representation) is still with us. Every mode of
being not assimilable to obdurate matter has had to be swallowed up by mind.
The simplification of ontology has led to the enormous complication of epistemology.
Once objects or things have been pacified—retreating to the exterior,
silent, and uniform world of nature—subjects begin to proliferate and chatter:
transcendental egos, legislative understandings, philosophies of language, theories
of mind, social representations, the logic of the signifier, webs of signification,
discursive practices, politics of knowledge, and, yes, anthropology of
course.
Anthropology is a discipline plagued since its inception by epistemological
angst. The most Kantian of disciplines, anthropology is practiced as if its paramount
task were to explain how it comes to know (to represent) its object—an
object also defined as knowledge (or representation). Is it possible to know it?
Is it decent to know it? Do we really know it, or do we see it (and ourselves)
through a glass, darkly? There is no way out of this maze of mirrors, mire of guilt.
Reification or fetishism is our major care and scare: we began by accusing savages
of confusing representations with reality; now we accuse ourselves (or,
rather, our colleagues).36
While philosophy has been obsessed with epistemology, ontology has been
annexed by physics. We have left to quantum mechanics the task of making our
most boring dualism, “representation/reality,” ontologically dubious. (Though
physics has questioned that dualism only in the confines of a quantum world inaccessible
to intuition and representation.) Supernature has thus given way to sub-
ViveirosdeCastro•TalkingPeacewithGods:Part1483
36. Polarities and other “othering” devices have had bad
press lately. The place of the other, however, can never
remain vacant for long. As far as contemporary anthropology
is concerned, the most popular candidate for the
position appears to be anthropology itself. In its formative
phase (never completely outgrown), anthropology’s main
task was to explain how and why the primitive or traditional
other was wrong: savages mistook ideal connections
for real ones and animistically projected social relations
onto nature. In the discipline’s classical phase (which
lingers on), the other is Western society/culture. Somewhere
along the line—with the Greeks? Christianity? capitalism?—the
West got everything wrong, positing substances,
individuals, separations, and oppositions wherever
all other societies/cultures rightly see relations, totalities,
connections, and embeddings. Because it is both anthropologically
anomalous and ontologically mistaken, it is the
West, rather than “primitive” cultures, that requires explanation.
In the post-positivist phase of anthropology, first
Orientalism, then Occidentalism, is shunned: the West
and the Rest are no longer seen as so different from each
other. On the one hand, we have never been modern, and,
on the other hand, no society has ever been primitive.
Then who is wrong, what needs explanation? (Someone
must be wrong, something has to be explained.) Our
anthropological forebears, who made us believe in tradition
and modernity, were wrong—and so the great polarity
now is between anthropology and the real practical/
embodied life of everyone, Western or otherwise. In brief:
formerly, savages mistook (their) representations for (our)
reality; now, we mistake (our) representations for (other
peoples’) reality. Rumor has it we have even be mistaking
(our) representations for (our) reality when we “Occiden-
talize.”
nature as our transcendent realm. On the macroscopic side, cognitive psychology
has been striving to establish a purely representational ontology, a natural
ontology of the human species inscribed in cognition, in our mode of representing
things. The representational function is ontologized in the mind but in
terms set by a simpleminded ontology of mind versus matter.
The tug of war goes endlessly on: one side reduces reality to representation
(culturalism, relativism, textualism), the other reduces representation to reality
(cognitivism, sociobiology, evolutionary psychology). Even phenomenology, new
or old—and especially the phenomenology invoked these days by anthropologists—may
be a surrender to epistemology. Is not “lived world” a euphemism for
“known world,” “represented world,” “world real for a subject”? Real reality is the
(still virtual) province of cosmologists, the theorists of quantum gravity and
superstring theory. But listen to these custodians of real reality and it becomes
obvious—it has been obvious, I might add, for more than seventy-five years—
that at the heart of the matter, there is no stuff; only form, only relation.37 There
are “materialist ontologies” on offer as cures for epistemological hypochondria,
but I do not know what to do with them. All I know is that we need richer ontologies
and that it is high time to put epistemological questions to rest. No effort
less strenuous and transformative and dangerously disorienting would make even
disagreement with an animist warrior possible.
COMMONKNOWLEDGE484
37. See Alfred North Whitehead, Science and the Modern
World (1925; New York: Macmillan, 1948).