Cosmological Deixis and Amerindian Perspectivism Author(s): Eduardo Viveiros de Castro Reviewed work(s): Source: The Journal of the

Royal Anthropological Institute, Vol. 4, No. 3 (Sep., 1998), pp. 469488 Published by: Royal Anthropological Institute of Great Britain and Ireland Stable URL: http://www.jstor.org/stable/3034157 . Accessed: 03/07/2012 18:19
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Royal Anthropological Institute of Great Britain and Ireland is collaborating with JSTOR to digitize, preserve and extend access to The Journal of the Royal Anthropological Institute.

http://www.jstor.org

COSMOLOGICAL DEIXIS AND AMERINDIAN PERSPECTIVISM

EDUARDO VIVEIROs DE CASTRO

Rio dejaneiro MuseuNacional, Cambridge King'sCollege,

the ideas in Amazonian This studydiscusses the meaning of Amerindian'perspectivism': see both themselves cosmologies concerningthe way in which humans,animals and spirits of the classicalcategories of a redefinition and one another.Such ideas suggestthepossibility or point of view. of 'nature','culture'and 'supernature' based on the concept of perspective The study argues in particular that the antinomy between two characterizationsof of which would deny the attributes indigenousthought- on the one hand 'ethnocentrism', humanityto humans fromother groups,and on the other hand 'animism',which would extend such qualities to beings of other species - can be resolved if one considers the difference between the spiritual and corporalaspectsof beings.
... ... la reciprocite deperspectives de la pensee mythique ouij'ai vu le caractere propre (L6vi-Strauss1985: 268)

Introduction This articledeals withthataspectofAmerindian which has been called thought its 'perspectivalquality' (Arhem 1993): the conception, common to many by different peoples of the continent, according to whichtheworld is inhabited sortsof subjectsor persons,human and non-human,which apprehendreality fromdistinct concept pointsofview.This idea cannotbe reducedto our current of relativism it seems to call to mind.In fact, it (Lima 1995; 1996), which at first and univeris at rightangles,so to speak,to the oppositionbetweenrelativism to the termsof our epistesalism.Such resistance byAmerindian perspectivism of the and transportability mologicaldebates casts suspicionon the robustness whichtheypresuppose.In particular, as manyanthropoloontological partitions gistshave alreadyconcluded (albeit for other reasons), the classic distinction between Nature and Culture cannot be used to describedomains internalto non-Western cosmologies without firstundergoinga rigorous ethnographic critique. of the predicates Such a critique,in the presentcase, impliesa redistribution subsumedwithinthetwoparadigmatic setsthattraditionally oppose one another underthe headingsof 'Nature' and 'Culture': universal and particular, objective and subjective, physicaland social,factand value, the givenand the instituted,
J. Roy.anthrop. Inst.(N.S.) 4, 469-488

470

EDUARDO VIVEIROS DE CASTRO

necessityand spontaneity, immanence and transcendence, body and mind, and humanity, animality among many more. Such an ethnographically-based reshuffling ofour conceptualschemesleadsme to suggest theexpression, 'multinaturalism', to designateone of the contrastive features of Amerindian thought in relation to Western 'multiculturalist' cosmologies. Where the latter are of the unityof natureand the plurality foundedon the mutual implication of of body and cultures - the firstguaranteedby the objective universality of spiritand substance,the second generatedby the subjectiveparticularity meaning- the Amerindianconceptionwould suppose a spiritualunityand a corporeal diversity. Here, culture or the subject would be the form of the whilstnatureor the objectwould be theformof the particular. universal, mustbe This inversion, to be morethanspeculative, perhapstoo symmetrical of developed by means of a plausible phenomenological interpretation which determinethe constitutive Amerindiancosmologicalcategories, condiwe can call 'nature'and 'culture'.Clearly, tionsof the relational contexts then,I between Nature and Culture must be subjected to thinkthatthe distinction critique, but not in orderto reachtheconclusionthatsuch a thingdoes not exist whichdo not exist).The flourishing too manythings (thereare already industry of criticisms of the Westernizing character of all dualisms has called for the of our conceptually abandonment dichotomousheritage, butto datethealternativeshave not gone beyond the stageof wishfulunthinking. I would prefer to on our own contrasts, themwiththe distinctions gain a perspective contrasting in Amerindian actually operating perspectivist cosmologies. Perspectivism were thenumerousreferences in The initialstimulusforthe present reflections to an indigenoustheoryaccordingto which the way Amazonian ethnography humans perceiveanimalsand othersubjectivities thatinhabittheworld - gods, the dead, inhabitants of othercosmic levels,meteorological spirits, phenomena, - differs even objectsand artefacts fromtheway plants, occasionally profoundly in which thesebeingssee humansand see themselves. in normal conditions, humans see humans as humans,animals as Typically, animalsand spirits howeveranimals(predators) and (iftheysee them) as spirits; see humans as animals(as prey)to the same extent thatanimals(as prey) spirits see humansas spirits or as animals(predators). By the same token,animalsand spiritssee themselvesas humans: they perceive themselvesas (or become) anthropomorphic beingswhen theyare in theirown houses or villagesand they in the formof culture- theysee theirown habitsand characteristics experience theirfood as human food (jaguars see blood as manioc beer,vulturessee the meatas grilledfish, maggotsin rotting etc.),theysee theirbodilyattributes (fur, or culturalinstruments, feathers, claws,beaks etc.) as body decorations theysee theirsocial system as organizedin the same way as human institutions are (with chiefs,shamans,ceremonies,exogamous moieties,etc.). This 'to see as' refers to percepts and not analogically to concepts,althoughin some cases the literally rather thanon the sensoryaspectof emphasisis placed more on the categorical the phenomenon. In sum, animals are people, or see themselvesas persons. Such a notion is formof each speciesis virtually alwaysassociatedwiththeidea thatthemanifest

EDUARDO VIVEIROS DE CASTRO

471

usually human form, a mere envelope (a 'clothing')which conceals an internal or to certain the beings the of trans-specific visible to species particular eyes only such as shamans. This internalformis the 'soul' or 'spirit'of the animal: an materito humanconsciousness, identical formally or subjectivity intentionality mask. behind an animal schema concealed in a human bodily let us say, alizable, between an anthropomorphic At first sightthen,we would have a distinction essence of a spiritualtype,common to animatebeings,and a variablebodily thanbeing of each individual characteristic speciesbutwhich rather appearance, This notionof is insteada changeableand removableclothing. a fixedattribute the - spirits, 'clothing'is one of the privilegedexpressionsof metamorphosis beaststhatturninto otherbeasts, dead and shamanswho assume animal form, processin turnedintoanimals- an omnipresent humansthatare inadvertently world' (Riviere1994: 256) proposedbyAmazonian transformational the 'highly 1 ontologies. can be seen in various and cosmological transformism This perspectivism but in general it is only the object of short South American ethnographies, It can also be found, elaborated.2 and seems to be quite unevenly commentaries of NorthAmerica value, in thefarnorth and maybewitheven greater generative populationsof otherpartsof the and Asia, as well as amongsthunter-gatherer In South America,the cosmologiesof theVaupes area are in thisrespect world.3 1996; Reichel-Dolmatoff highly developed (see Arhem1993; 1996; Hugh-Jones 1985), but otherAmazonian societies,such as the Wari' of Rondonia (Vila?a 1992) and theJurunaof the Middle Xingu (Lima 1995; 1996), also give equal emphasisto the theme. does not usually Some general observationsare necessary.Perspectivism otherbeings);theemphasisseems to involveall animalspecies (besidescovering role such as the a keysymbolicand practical be on those specieswhich perform speciesof preyforhumans- one of the central and the principal greatpredators inverdimension,of perspectival dimensions,possiblyeven the fundamental statusesof predator and prey(Arhem sions refers to the relativeand relational 1993: 11-12; Vila?a 1992: 49-51). On the otherhand, however,it is not always to each individual clear whetherspiritsor subjectivities are being attributed animal, and there are examples of cosmologieswhich deny consciousness to animals (Overing 1985: 249 sqq.; 1986: 245-6) or some other post-mythical (Baer 1994: 89; Viveiros de Castro 1992a: 73-4). spiritual distinctiveness 'masters'('mothersof Nonetheless,as is well known,the notionof animalspirit the game animals', 'mastersof the white-lipped peccaries',etc.) is widespread endowedwithintentionclearly thecontinent. These spirit masters, throughout as hypostases of the animal species alityanalogous to thatof humans,function field for with which they are associated,therebycreatingan intersubjective relations even whereempirical animalsare not spiritualized. human-animal Amerindian above all, thatifthereis a virtually universal We mustremember, between humans and notion,it is thatof an originalstateof undifferentiation withbeingswhose form, name Mythsarefilled animals,describedin mythology. in a common mix human and animal attributes and behaviour inextricably identical to thatwhich definesthe present-day context of intercommunicability, between 'culture' and 'nature',which intra-human world. The differentiation is not a mythology, themeofAmerindian Levi-Strauss showed to be the central thehumanfrom theanimal,as in our own evolutionist processofdifferentiating

472

EDUARDO VIVEIROS DE CASTRO

mythology. The originalcommon conditionof both humansand animalsis not animality but ratherhumanity. The great mythicalseparationreveals not so much culturedistinguishing itself from naturebutrather naturedistancing itself fromculture:the myths tellhow animalslostthe qualitiesinherited or retained by humans (Brightman1993: 40, 160; Levi-Strauss1985: 14, 190; Weiss 1972: 169-70). Humans are thosewho continueas theyhave alwaysbeen: animalsare nothumansex-animals. In sum,'thecommon pointofreference for ex-humans, all beingsofnatureis nothumansas a speciesbutrather humanity as a condition' (Descola 1986: 120). - betweenthehuman speciesand the human conditionThis is a distinction which should be retained.It has an evidentconnexionwith the idea of animal clothinghidinga common spiritual 'essence' and with the issue of the general meaningof perspectivism. For themoment, we maysimply note one of itsmain thepasthumanity ofanimalsis added to their corollaries: present-day spirituality in orderto producethatextended hiddenbytheir visibleform setoffoodrestrictions or precautionswhich eitherdeclare inedible certainanimals that were mythically co-substantial with humans,or demand theirdesubjectivization by shamanisticmeans before they can be consumed (neutralizingthe spirit, transubstantiating the meat into plant food, semantically reducingit to other animalsless proximate to humans),underthethreat of illness,conceivedof as a cannibalcounter-predation of thepreyturnedpredator, undertaken bythe spirit in a lethalinversion of perspectives whichtransforms the human intoanimal.4 It is worth pointing out that Amerindianperspectivism has an essential relation withshamanismand withthevalorization of the hunt.The association between shamanism and this 'venatic ideology' is a classic question (for thatthisis Amazonia,see Chaumeil 1983: 231-2; Crocker1985: 17-25). I stress a matter of symbolic notecologicalnecessity: horticulturists such as importance, the Tukano or theJuruna(who in any case fishmore than theyhunt) do not differ much fromcircumpolarhuntersin respectof the cosmologicalweight on animalpredation, ofanimalsand thetheory conferred spiritual subjectivation intentionalities accordingto which the universeis populated by extra-human endowedwiththeir own perspectives. In thissense,thespiritualization ofplants, or derivative seemsto me to be secondary meteorological phenomenaor artefacts in comparison withthespiritualization ofanimals:theanimalis theextra-human of the Other,maintaining relations withotherprototypical prototype privileged of alterity, such as affines figures (Arhem 1996; Descola 1986: 317-30; Erikson 1984: 110-12). This huntingideology is also and above all an ideology of therelations betweenhumans shamans,in so faras itis shamanswho administer sincetheyalone are capableof and thespiritual componentoftheextra-humans, are capable of assuming the point of view of such beings and, in particular, is relativism as public to tell the tale. If Westernmulticulturalism returning as cosmic thenAmerindianperspectivist shamanismis multinaturalism policy, politics.

Animism
is reminiscent ofthenotion The reader will have noticedthatmy'perspectivism' of 'animism' recentlyrecuperatedby Descola (1992; 1996). Statingthat all of non-humansalwaysreferto the social domain, Descola conceptualizations

EDUARDO VIVEIROS DE CASTRO

473

where the differnature:totemism, threemodes of objectifying distinguishes thatis, ences betweennaturalspeciesare used as a model forsocial distinctions; in character betweennatureand cultureis metaphorical wherethe relationship where and markedby discontinuity (bothwithinand betweenseries); animism, social life'organizethe relationsbetween categoriesstructuring the 'elementary betweennatureand a socialcontinuity species,thusdefining humansand natural and social characterof human dispositions foundedon the attribution culture, ofWestern typical isticsto 'naturalbeings'(Descola 1996: 87-8); and naturalism, betweennature, thedomain whichsupposes an ontological duality cosmologies, areas separated by of necessity,and culture, the domain of spontaneity, of societies in The 'animic mode' is characteristic metonymicdiscontinuity. of natureand of its focusof the objectification which animalsare the 'strategic socialization'(1992: 115), as is thecase amongstindigenouspeoples ofAmerica, reigning supreme over those social morphologieslackingin elaborateinternal or combinedwith But thismode can also be foundco-existing segmentations. exist,the Bororo and theiraroelbope totemism,wherein such segmentations dualismbeingsuch a case.5 which I cannot discuss here as fullyas it These ideas formpartof a theory would merit. I merely comment on the contrast between animism and angle from the original one. naturalism but from a somewhat different (Totemism, as defined by Descola, seems to me to be a heterogeneous rather thancosmological:itis not a system classificatory phenomenon,primarily of relations betweennatureand cultureas is the case in theothertwo modes, but rather of purelylogicaland differential correlations.) whichpostulates thesocial character Animismcould be defined as an ontology of relationsbetween humans and non-humans:the space betweennatureand axiom: relations societyis itselfsocial. Naturalismis foundedon the inverted betweensociety and naturearethemselves natural. Indeed,ifin theanimicmode is internalto the social world, humans and the distinction'nature/culture' animals being immersedin the same socio-cosmicmedium (and in this sense ontology, the then in naturalist 'nature'is a partof an encompassingsociality), is internal distinction to nature(and in thissense,humansociety 'nature/culture' is one natural phenomenon amongst others). Animism has 'society' as the and has 'nature':these poles function, unmarkedpole, naturalism respectively as the universaldimension of each mode. Thus animism and contrastively, naturalism are hierarchical and metonymical structures them (thisdistinguishes from which is based on a metaphoric correlation betweenequipollent totemism, opposites). In Western is natural:humans naturalist the nature/society interface ontology, are organismslike the rest,body-objects in 'ecological' interaction with other all ofthemruledbythenecessary bodies and forces, laws of biologyand physics; express, naturalforces.Social relations, and thereby 'productive forces'harness, or instituted thatis,contractual relations betweensubjects, can onlyexistinternal to human society.But how alien to nature- this would be the problem of - aretheserelations? thestatus ofthe naturalism ofnature, Giventheuniversality humanand socialworldis unstableand,as thehistory thought shows, ofWestern itperpetually oscillatesbetweena naturalistic monism ('sociobiology'beingone of itscurrent dualismof nature/culture ('culturalism' avatars)and an ontological of this latter dualism,forall expression).The assertion being its contemporary

474

EDUARDO VIVEIROS DE CASTRO

of the notion of nature,by that,only reinforces the finalreferential character revealing itself to be thedirect descendant oftheoppositionbetweenNature and Culture is the modernname of Spirit- let us recallthe distinction Supernature. - or attheleastitis thename and Geisteswissenschaften betweenNaturwissenschaften of the compromise between Nature and Grace. Of animism,we would be is located in the opposite pole: there the temptedto say that the instability and animality constiproblem is how to administer the mixtureof humanity tutinganimals,and not, as is the case amongstourselves,the combinationof a cultureand naturewhich characterize humans; the point is to differentiate 'nature'out of the universal sociality. of differences and qualities However,can animismbe definedas a projection model internal to the humanworldonto non-humanworlds,as a 'socio-centric' areused to map theuniverse? in whichcategories and social relations This interin some glosseson thetheory: 'iftotemicsystems pretation by analogyis explicit model society afternature,then animic systemsmodel nature aftersociety' is to avoid any undesirable (Arhem 1996: 185). The problemhere,obviously, proximity with the traditional sense of 'animism', or with the reduction of classifications' but equally the 'primitive to emanationsof social morphology; of the relation between problemis to go beyondotherclassicalcharacterizations societyand naturesuch as Radcliffe-Brown's.6 Ingold (1991; 1996) showed how schemes of analogicalprojectionor social modelling of nature escape naturalistreductionism only to fall into a 'really natural' nature from nature/culture dualism which by distinguishing to be a typical cosmologicalantinomy 'culturally constructed' naturerevealsitself The notion of model or metaphorsupposes a faced with infinite regression. are constitutive betweena domain whereinsocial relations previousdistinction and metaphorical. and literal and another where they are representational as human sociality Animism,interpreted projectedonto the non-humanworld, but the metaphor of a metonymy. would be nothing to be resolved, is thatofknowing therefore, Amongstthequestionsremaining use of categories to whetheranimismcan be describedas a figurative pertaining thedomainofnon-humansand their thehuman-socialdomainto conceptualize with the former. Anotherquestion: if animismdepends on the attrirelations faculties and sensory to animals,and the same form butionof human cognitive between humans and of subjectivity, then what in the end is the difference to be animals?If animalsarepeople,thenwhydo theynotsee us as people?Why, ifanimismis a way of objectifying naturein precise,the perspectivism? Finally, whichthe dualismof nature/culture does not hold,thenwhatis to be done with the centrality of this oppositionto South the abundantindicationsregarding Americancosmologies? Arewe dealingwith 'totemic illusion',ifnot just another withan ingenuousprojection of our Western dualism? Ethnocentrism In a well-knonv essay, ceases Levi-Strauss observedthatfor'savages' humanity at the boundaryof the group,a notionwhich is exemplified by the widespread of auto-ethnonym meaning'real humans',which,in turn,impliesa definition as somehow pertaining to the domain of the extra-human. Therefore, strangers of theWest but a naturalideological would not be the privilege ethnocentrism

EDUARDO VIVEIROS DE CASTRO

475

illustrates the universal inherent to human collectivelife.Levi-Strauss attitude, withan anecdote: reciprocity of thisattitude
of America,whilstthe Spanish were In the GreaterAntilles,some yearsafter the discovery whetherthe nativeshad a soul or not, dispatching inquisitionalcommissionsto investigate theseverynativeswere busy drowningthe whitepeople theyhad capturedin orderto find whetheror not the corpseswere subjectto putrefaction out,after lengthy observation, (1973: 384).

derived thefamous whichLevi-Strauss The generalpointofthisparable(from and foremost the man who believesin barbarism') moral: 'The barbarian is first consideredthatonlythe is quite simple:the Indians,liketheEuropean invaders, are on the otherside groupto which theybelongincarnates humanity; strangers culturefrom of the borderwhich separateshumans fromanimals and spirits, As matrix and conditionforthe existenceof ethnocennatureand supernature. trism,the nature/culture oppositionappearsto be a universalof social apperception. of vindiwas writing these lines,the strategy At the timewhen Levi-Strauss of savageswas to demonstrate thattheymade the same catingthe fullhumanity distinctionsas we do: the proof that they were true humans is that they consideredthattheyalone were the true humans. Like us, theydistinguished are alwaysthe others. culturefromnatureand theytoo believedthatNaturvdlker The universality of the culturaldistinction betweenNature and Culture bore witnessto theuniversality ofcultureas humannature.In sum,theanswerto the question of the Spanish investigators (which can be read as a sixteenth-century versionof the 'problemof otherminds') was positive:savagesdo have souls. has changed. The savages are no longer ethnocentric Now, everything but rather cosmocentric;insteadof havingto prove thattheyare humans because they distinguishthemselvesfrom animals, we now have to recognize how inhumanwe are foropposing humans to animals in a way theynever did: for themnatureand cultureare partof the same sociocosmicfield.Not onlywould Amerindiansput a wide berthbetween themselvesand the Great Cartesian Divide which separated humanity fromanimality, but their views anticipate the fundamental lessonsofecologywhichwe areonlynow in a positionto assimilate to concede predicates of (Reichel-Dolmatoff 1976). Before,the Indians' refusal humanity to othermen was of note; now we stress thattheyextendsuch predicates far beyond the frontiers of their own species in a demonstration of 'ecosophic' knowledge(Arhem1993) which we should emulatein as faras the limitsof our objectivism permit. Formerly, it had been necessary to combatthe assimilationof the savage mind to narcissistic animism,the infantile stage of naturalism, showingthattotemismaffirmed the cognitivedistinction between cultureand nature;now, neo-animismrevealsitselfas the recognition of the universaladmixture of subjectsand objects,humans and non-humansagainst modernhubris, the primitive and post-modern 'hybrids', to borrowa termfrom Latour(1991). Two antinomiesthen,which are, in fact,only one: eitherAmerindiansare ethnocentrically 'stingy'in the extension of theirconceptof humanity and they 'totemically' oppose natureand culture;or theyare cosmocentric and 'animic' and do not professto such a distinction, being models of relativist tolerance, a multiplicity of pointsofview on theworld. postulating

476

EDUARDO

VIVEIROS DE CASTRO

I believethatthesolutionto theseantinomies7 lies notin favouring one branch over the other,sustaining,for example, the argumentthat the most recent ofAmericanattitudes is thecorrect characterization one and relegating theother to theouterdarkness ofpre-post-modernity. Rather, thepointis to show thatthe 'thesis'as well as the 'antithesis' are true(both correspond to solid ethnographic intuitions), but thattheyapprehend thesame phenomenafromdifferent angles; and also it is to show that both are false in thattheyreferto a substantivist conceptualization of the categoriesof Nature and Culture (whetherit be to or negatethem)which is not applicableto Amerindian affirm cosmologies. The firstpoint to be considered is that the Amerindianwords which are in thosesupposedlyethnoas 'human being'and whichfigure usuallytranslated centric self-designations do notdenotehumanity as a natural species.They refer rather to the social conditionof personhood,and theyfunction(pragmatically when not syntactically) less as nouns than as pronouns. They indicate the position of the subject; they are enunciativemarkers,not names. Far from a semanticshrinking of a common name to a propername (taking manifesting thesewordsmove in theoppositedirection, 'people' to be thename ofthetribe), to perspective goingfromsubstantive (using'people' as a collective pronoun'we have that people/us'). For this very reason, indigenouscategoriesof identity enormous contextual of scope thatcharacterizes variability pronouns,marking local group,all humans,or even all contrastively Ego's immediatekin,his/her their as 'ethnonyms' seems largely beingsendowedwithsubjectivity: coagulation to be an artefact of interactions Nor is it by chance thatthe withethnographers. ofAmerindian whichentertheliterature are not self-desigmajority ethnonyms nations,but rathernames (frequently pejorative)conferred by other groups: is primordially notto theones in the ethnonymic objectivation appliedto others, position of subject.Ethnonymsare names of thirdparties;theybelong to the notto the category of 'we'. This, bytheway,is consistent witha of 'they' category avoidance of self-reference on the level of onomastics: widespread personal names are not spokenby the bearersnor in theirpresence;to name is to externalize,to separate(from)the subject. Thus self-references such as 'people' mean 'person', not 'member of the human species',and theyare personalpronounsregistering thepointofview of the subjecttalking, not propernames. To say,then,thatanimalsand spirits are to non-humansthe capacpeople is to saythattheyare persons,and to attribute ities of conscious intentionality and agencywhich definethe position of the as thesoul or spirit withwhichthesenonareobjectified subject.Such capacities has a humansare endowed. Whatever possessesa soul is a subject,and whatever soul is capable of havinga point of view.Amerindian souls, be theyhuman or animal,are thus indexicalcategories, cosmologicaldeicticswhose analysiscalls or substantialist as fora theory not so much foran animist psychology ontology of the signor a perspectival pragmatics (Taylor1993a; 1993b;Viveirosde Castro 1992b). would be a subject;or Thus, every beingto whom a pointofview is attributed whereverthereis a point of view thereis a subjectposition.Whilstour better, the constructionist can be summed up in the Saussureanformula: epistemology - the subject being the original,fixed condition theobject pointof viewcreates whence the point of view emanates- Amerindianontologicalperspectivism creates the is activated view whatever proceedsalongthelinesthatthepoint of subject;

EDUARDO VIVEIROS DE CASTRO

477

or 'agented'bythepointofviewwill be a subject.8 This is whyterms such as wari' (Vilaca 1992), dene(McDonnell 1984) or masa(Arhem1993) mean 'people', but theycan be used for- and therefore used by - verydifferent classes of beings: used by humans they denote human beings; but used by peccaries,howler monkeys or beaverstheyself-refer or beavers. to peccaries, howlermonkeys As it happens,however, these non-humansplaced in the subjectperspective do not merely'call' themselves'people'; theysee themselvesanatomically and as humans. The symbolicspiritualization of animalswould implytheir culturally imaginary hominization and culturalization; thustheanthropomorphic-anthroof indigenousthoughtwould seem to be unquestionable. pocentriccharacter is at issue. Anybeingwhich different However,I believethatsomething totally vicariouslyoccupies the point of view of reference, being in the position of as a memberof the human species.The human bodilyform subject,sees itself and human culture - the schemata of perceptionand action 'embodied' in specificdispositions- are deictics of the same type as the self-designations discussed above. They are reflexive or apperceptive schematisms by which all and constitutive humanpredicates subjectsapprehend and notliteral themselves, projected metaphorically (i.e. improperly)onto non-humans. Such deictic 'attributes' and move with it (Brightman are immanentin theviewpoint, 1993: - enjoy the same prerogative 47). Human beings- naturally and therefore see themselves as such.9It is not thatanimalsare subjectsbecause theyare humans in disguise,but rather thattheyare human because theyare potentialsubjects. This is to say Culture is theSubject's nature; it is the formin which everysubject experiencesits own nature.Animismis not a projectionof substantive human qualities cast onto animals,but ratherexpressesthe logical equivalence of the reflexive thathumansand animalseach have to themselves: relations salmon are to (see) salmon as humans are to (see) humans,namely, (as) human.10 If,as we have observed,the common conditionof humansand animalsis humanity not thisis because 'humanity'is the name forthe generalformtakenby animality, the Subject. Multinaturalism Withthiswe may have discardedanalogicalanthropocentrism, but only apparto adoptrelativism."1 For would thiscosmologyof multiple ently viewpoints not implythat'every perspective is equallyvalid and true'and that'a correct and true representation oftheworld does not exist'(Arhem1993: 124)? But thisis exactly the question: is theAmerindian in fact perspectivist theory a multiplicity of representations of the same world? It is sufficient asserting to considerethnographic evidenceto perceivethatthe oppositeapplies: all beings see ('represent') theworld in thesameway- whatchangesis theworldthatthey see. Animals impose the same categoriesand values on reality as humans do: their worlds, like ours, revolve around hunting and fishing,cooking and fermented drinks,cross-cousinsand war, initiationrituals,shamans, chiefs, is involvedin fishing spirits. 'Everybody and hunting; everybody is involvedin feasts,social hierarchy, chiefs,war, and disease, all the way up and down' (Guedon 1984: 142). If the moon, snakes and jaguars see humans as tapirsor white-lipped peccaries(Baer 1994: 224), it is because they, likeus, eat tapirsand peccaries,people's food. It could only be thisway,since,being people in their

478

EDUARDO

VIVEIROS DE CASTRO

own sphere,non-humanssee things as 'people' do. But the thingsthat theysee aredifferent: whatto us is blood, is maize beerto thejaguar; whatto thesouls of the dead is a rotting corpse,to us is soakingmanioc; what we see as a muddy the tapirssee as a greatceremonialhouse. waterhole, of subjectiveand partialrepre(Multi)cultural relativism supposes a diversity and unifiednature, each striving to graspan external which remains sentations, to those representations. Amerindian perfectly indifferent thought proposesthe or phenomenological unity which is purely opposite: a representational or deictic, One pronominal indifferently appliedto a radically objectivediversity. is multinaturalist, for a single 'culture', multiple 'natures' - perspectivism is not a representation. perspective of are a property is not a representation because representations A perspective 12 The ability whereasthepointofview is locatedin thebody. themind or spirit, to adopta pointofview is undoubtedly a powerofthesoul, and non-humansare between subjects in so far as they have (or are) spirit;but the differences lies not in the soul. viewpoints(and a viewpointis nothingifnot a difference) identicalin all species,it can onlysee the same things Since the soul is formally - the difference of bodies. This permits is given in the specificity everywhere answersto be found for our questions: if non-humansare persons and have themfromhumans?And why,being people, do souls, thenwhat distinguishes theynot see us as people? Animalssee in thesameway as we do different thingsbecause theirbodies are - as faras that different fromours.I am not referring to physiological differences is concerned, of bodies - but rather Amerindians recognizea basic uniformity to whichrender or capacities thebodyof every affects, dispositions speciesunique: what it eats, how it communicates, where it lives,whetherit is gregariousor and so forth. The visibleshape of the body is a powerfulsign of these solitary, in affect, differences althoughit can be deceptivesince a human appearance could, forexample,be concealingajaguar-affect. Thus, what I call 'body' is not a synonym fordistinctive substanceor fixedshape; it is an assemblageof affects orwaysofbeingthatconstitute a habitus. Betweentheformal ofsouls subjectivity and thesubstantial oforganisms thereis an intermediate materiality planewhich is occupied by the body as a bundle of affects and capacitiesand which is the originof perspectives. between bodies, however,is only apprehendablefrom an The difference exterior viewpoint, byan other, since,foritself, every typeof beinghas the same form(the genericformof a human being): bodies are theway in which alterity is apprehendedas such. In normalconditions we do not see animalsas people, and vice-versa, because our respective which they bodies (and the perspectives of the subject, allow) are different. Thus, if 'culture'is a reflexive perspective objectifiedthroughthe concept of soul, it can be said that 'nature' is the if Culture is the viewpoint which the subject takes of other body-affects; thenNature istheform Other thatis,as theobjectfor Subject'snature, ofthe as body, a subject.Culturetakesthe self-referential formofthepronoun'I'; natureis the formof the non-personor the object,indicatedby the impersonal pronoun 'it' (Benveniste1966a: 256). the bodymakesthe difference, thenit is easily If,in the eyes ofAmerindians, understoodwhy,in the anecdote told by Levi-Strauss, the methodsof investigationintothe humanity of the other, employedby theSpanishand theinhabi-

EDUARDO VIVEIROS DE CASTRO

479

For theEuropeans,the issue was showed such asymmetry. tantsof theAntilles, theothers possesseda soul; fortheIndians,theaimwas to find to decidewhether the out what kindof body the othershad. For theEuropeansthe greatdiacritic, is thesoul (are Indianshumansor animals?); in perspective, marker ofdifference The Europeans fortheIndiansitis thebody (are Europeanshumansor spirits?). neverdoubted thatthe Indians had bodies; the Indians neverdoubted thatthe Europeans had souls (animals and spiritshave them too). What the Indians thebodies ofthose'souls' were capableofthesame wantedto knowwas whether affects as theirown - whethertheyhad the bodies of humans or the bodies of consisted and protean.In sum: European ethnocentrism spirits, non-putrescible in doubting whetherother bodies have the same souls as they themselves; in doubting whether other souls had the same Amerindianethnocentrism bodies. thought (1994; 1996), the statusof humansin Western As Ingold has stressed ambiguous: on the one hand, humankindis an animal species is essentially is a domain thatincludeshumans; on the other amongstothers,and animality whichexcludesanimals.These two statuses is a moralcondition hand,humanity notionof 'human nature'.In other and disjunctive co-existin the problematic words, our cosmology postulates a physical continuityand a metaphysical theformer betweenhumansand animals, makingofman an object discontinuity an objectforthe'humanities'. Spiritor mindis forthenatural sciences,thelatter in general, itdistinitraisesus above animalsand matter our great differentiator: guishes cultures,it makes each person unique beforehis or her fellowbeings. it connectsus to the restof the The body,in contrast, is the major integrator: which,in turn, (DNA, carbonchemistry) substrate living, unitedby a universal links up with the ultimatenatureof all materialbodies.'3 In contrastto this, and a physicaldiscontinuity continuity Amerindianspostulatea metaphysical the latter in resulting in animism, betweenthebeingsofthe cosmos,theformer a the spirit or soul (here not an immaterial substancebut rather perspectivism: but a system organism while the body (not a material reflexive form)integrates, of activeaffects) differentiates.

Thespirit's many bodies

in Amazonian The idea that the body appears to be the greatdifferentiator to the extent cosmologies- thatis, as thatwhich unitesbeingsof the same type, thatit differentiates them fromothers- allows us to reconsidersome of the of the regionin a new light. classicquestionsof the ethnology in Amazonian of corporeality Thus, the now old theme of the importance societies (a theme that much predatesthe current'embodiment'craze - see For example,itbecomes possible foundations. Seegeretal. 1979) acquiresfirmer - be they to gain a betterunderstanding of why the categoriesof identity expressedthroughbodily personal,social or cosmological- are so frequently The universal food practices and body decoration. through idioms,particularly symbolicimportanceof food and cooking regimesin Amazonia - fromthe to the Piro idea thatwhat 'raw and the cooked' of Levi-Strauss, mythological fromwhite people is 'real food' makes them different literally (i.e. naturally) (Gow 1991); fromthe food avoidanceswhich define'groups of substance'in Central Brazil (Seeger 1980) to the basic classification of beings accordingto

480

EDUARDO VIVEIROS DE CASTRO

their eating habits (Baer 1994: 88); from the ontological productivity of commensality, similarity of diet and relative condition of prey-objectand predator-subject (Vilaca 1992) to the omnipresence of cannibalism as the 'predicative'horizon of all relationswith the other, be they matrimonial, or bellicose(Viveiros de Castro 1993) - thisuniversality demonstrates alimentary thatthe setof habitsand processesthatconstitute bodies is precisely the location fromwhich identity and difference emerge. The same can be said ofthe intensesemioticuse ofthebody in the definition of personal identitiesand in the circulationof social values (Mentore 1993; Turner 1995). The connexion between this overdetermination of the body of itsvisiblesurface)and the restricted recoursein theAmazonian (particularly - thatis,a situation socius to objectscapableofsupporting relations whereinsocial such as thosecharacteristic exchangeis not mediatedbymaterial objectifications of giftand commodityeconomies - has been shrewdly pinpointedby Turner, who has shown how the human bodytherefore mustappearas the prototypical social object. However,theAmerindian emphasison the social construction of thebodycannotbe takenas theculturalization ofa natural substract but rather as the productionof a distinctly human body,meaningnaturally human. Such a process seems to be expressing not so much a wish to 'de-animalize'the body its culturalmarking, to particularize through but rather a body stilltoo generic, it fromthe bodies of otherhuman collectivities as well as from differentiating those of otherspecies. The body,as the siteof differentiating perspective, must be differentiated to the highest degreein ordercompletely to expressit. The human body can be seen as the locus of the confrontation between and animality, but not because it is essentially animal by natureand humanity needs to be veiled and controlledby culture(Riviere 1994). The body is the instrument and at the same timetheobjectpar subject'sfundamental expressive thatwhich is presented excellence, to the sightof the other.It is no coincidence, of bodies,theirmaximalparticuthen,thatthe maximumsocial objectification is at the same time the larizationexpressedin decorationand ritualexhibition momentof maximumanimalization (Goldman 1975: 178; Turner1991; 1995), when bodies are coveredby feathers, colours,designs,masksand otheranimal to the animal prostheses. Man ritually clothedas an animal is the counterpart intoan animal,revealsto himself transformed supernaturally naked.The former, formand the 'natural'distinctiveness of his body; the latter, freeof its exterior itself ofspirit. The model as human,showsthe'supernatural' revealing similarity is the human spirit, of spirit but the model of bodyis the bodies of animals;and iffromthe pointof view of the subjectculturetakesthe genericformof '' and natureof 'it/they', then the objectification of the subject to itselfdemands a i.e. embodiesit- whilstthe ofbodies- whichnaturalizes singularization culture, - which oftheobjectimpliescommunication atthelevelofspirit subjectification culturalizes it. Put in these terms, the Amerindian nature,i.e. supernaturalizes distinction of Nature/Culture, beforeit is dissolvedin the name of a common animichuman-animal mustbe re-readin the lightof somaticperspecsociality, tivism. It is important ofas given to notethat theseAmerindian bodiesarenotthought as made. Therefore, but rather an emphasison the methodsforthe continuous fabrication ofthebody(Viveiros de Castro 1979); a notionofkinship as a process of activeassimilation of individuals(Gow 1989; 1991) throughthe sharingof

EDUARDO VIVEIROS DE CASTRO

481

of - and not as a passive inheritance bodily substances,sexual and alimentary of memory which inscribesit in the flesh essence; the theory some substantial thetheory which situates (Viveirosde Castro 1992a: 201-7), and more generally knowledge in the body (Kensinger 1995: ch. 22; McCallum 1996). The AmerindianBildunghappens in the body more than in the spirit:there is no a redefinition of itsaffects 'spiritual' changewhichis nota bodilytransformation, between body and soul is while the distinction and capacities.Furthermore, as an obviously pertinentto these cosmologies, it cannot be interpreted and sites discontinuity (Townsley1993: 454-5). As bundlesof affects ontological bodies 'are' souls,just, incidenrather than material organisms, of perspective, as souls and spirits'are' bodies. The dual (or plural) conception of the tally, between the human soul, widespread in indigenousAmazonia, distinguishes site of of an individual'shistory, soul (or souls) of the body,reifiedregister the and a 'true soul', pure, formalsubjectivesingularity, memoryand affect, abstract markof a person (e.g. McCallum 1996; Viveirosde Castro 1992a: 201which inhabitthe 14). On the otherhand,the souls of the dead and the spirits but equallytypesof bodies, endowed with entities, universeare not immaterial - affects - sui generis. Indeed, body and soul, just like nature and properties or ontological entities self-subsistent to substantives, culture, do not correspond perspectives. to pronounsor phenomenological provinces, but rather thangiven character of the body,a conceptionthat The performative rather itself'culturally' in order for it to be 'naturally' requires it to differentiate a possihas an obvious connexionwith interspecific metamorphosis, different, by a way byAmerindian cosmologies.We need not be surprised bility suggested yetat the same time of thinking which positsbodies as the greatdifferentiators statestheirtransformability. Our cosmologysupposes a singulardistinctiveness of minds, but not even forthis reason does it declare communication(albeit or denythe mental/spiritual problem) to be impossible, solipsismis a constant transformations induced by processes such as education and religious is the locus of difference it is precisely because thespiritual conversion;in truth, that conversionbecomes necessary(the Europeans wanted to know whether Indians had souls in order to modifythem). Bodily metamorphosisis the In the conversion.14 to theEuropean themeof spiritual Amerindian counterpart our cosmology threatens sameway,ifsolipsismis thephantomthatcontinuously - raising ourselvesin our 'own kind'because theyare thefearof notrecognizing of minds- thenthe possinot like us, giventhe potentially absolutesingularity of metamorphosis of no longerbeing able to expressesthe oppositefear, bility betweenthe human and the animal,and, in particular, the fearof differentiate - hence the withinthebodyoftheanimalone eats"5 seeingthehumanwho lurks of food prohibitions potency and precautions linkedto the spiritual importance of animals,mentionedabove. The phantomof cannibalismis the Amerindian derivesfromthe uncertainty equivalentto theproblemof solipsism:ifthe latter of a real community as to whetherthe naturalsimilarity of bodies guarantees prevailoverthe thentheformer of souls might suspectsthatthesimilarity spirit, real differences of body and thatall animals thatare eaten might,despite the shamanistic efforts them,remainhuman.This, ofcourse,does to de-subjectivize such as not prevent us havingamongstourselvesmore or less radicalsolipsists, and moreor therelativists, societiesbe purposefully variousAmerindian northat less literally cannibalistic.'6

482

EDUARDO VIVEIROS DE CASTRO

The notion of metamorphosis is directly linked to the doctrineof animal 'clothing',to which I have referred. How are we to reconcilethe idea thatthe withthethemeof the 'appearance' bodyis the siteof differentiating perspectives and 'essence' which is always evoked to interpret animismand perspectivism (Arhem 1993: 122; Descola 1986: 120; Hugh-Jones1996; Riviere1994)? Here seems to me to lie an importantmistake,which is that of taking bodily 'appearance'to be inert and false, whereasspiritual 'essence' is activeand real(see the definitive observations of Goldman 1975: 63). I arguethatnothing could be theIndians'mindswhen theyspeakofbodies in terms of'clothing'. further from It is not so much thatthebodyis a clothing but rather thatclothing is a body. We are dealingwithsocietieswhichinscribe efficacious meaningsonto theskin,and which use animal masks (or at least know theirprinciple)endowed with the to transform theidentities ofthosewho wearthem,ifused powermetaphysically in the appropriate ritualcontext.To put on mask-clothing is not so much to butrather to activate the conceal a humanessencebeneathan animalappearance, powers of a different body.17The animal clothesthatshamansuse to travelthe cosmos are not fantasies but instruments: theyare akinto divingequipment,or masks.The intention when donninga wet suitis space suits,and notto carnival like a fish,to breatheunderwater, to be able to function not to conceal oneself under a strangecovering. In the same way, the 'clothing' which, amongst 'essence' of a human type, is not a mere disguisebut animals,coversan internal their distinctive and capacitieswhich equipment, endowed with the affects defineeach animal.'8 It is truethatappearances can be deceptive(Hallowell 1960; is thatin Amerindian Riviere1994); but my impression narratives whichtakeas a themeanimal 'clothing'the interest lies more in what these clothesdo rather thanwhattheyhide. Besides this,betweena beingand itsappearanceis itsbody, which is more thanjust that- and theverysame narratives relatehow appearances are always 'unmasked' by bodily behaviourwhich is inconsistent with them. In short:thereis no doubt thatbodies are discardableand exchangeable to humans. and that'behind' themlie subjectivities which are formally identical But the idea is not similarto our oppositionbetweenappearanceand essence; it the objectivepermutability of bodies which is based in the merelymanifests subjective equivalenceof souls. Anotherclassic theme in South Americanethnology which could be interwithin thisframework is thatofthesociologicaldiscontinuity betweenthe preted distinction livingand the dead (Carneiro da Cunha 1978). The fundamental betweenthe livingand the dead is made by the body and precisely not by the which prevailsas differentiator over the spirit;death is a bodily catastrophe common 'animation' of the living and the dead. Amerindian cosmologies dedicateequal or greater interest to theway in whichthedead see reality as they do to the vision of animals,and as is the case forthe latter, theyunderlinethe radicaldifferences vis-a-vis the world of the living.To be precise,being definifromtheirbodies,the dead are not human.As spirits definedby tively separated theirdisjunctionfroma human body,the dead are logicallyattracted to the bodies of animals;thisis whyto die is to transform intoan animal(Pollock 1985: 95; Schwartzman1988: 268; Turner1995: 152; Vilaca 1992: 247-55), as it is to transform intootherfigures ofbodilyalterity, such as affines and enemies.In this ifanimismaffirms a subjectiveand social continuity betweenhumans manner, establishesan objective and animals, its somatic complement,perspectivism,

EDUARDO VIVEIROS DE CASTRO

483

discontinuity, equallysocial,betweenlive humansand dead humans.'9 Having examined the differentiating component of Amerindianperspeca cosmological 'function'to the transtivism,it remains for me to attribute specificunityof the spirit.This is the point at which, I believe, a relational definition could be givenfora category, whichnowadayshas fallen Supernature, into disrepute(actually, ever since Durkheim),but whose pertinence seems to me to be unquestionable. domains Apartfromitsuse in labellingcosmographic of a 'hyper-uranian' type, or in defininga third type of intentionalbeings in indigenouscosmologies,which are neitherhuman nor animal (I occurring referto 'spirits'),the notion of supernature may serve to designatea specific relational contextand particular which is as distinct phenomenological quality, fromthe intersubjective relations thatdefinethe social worldas fromthe 'interwiththebodies of animals. objective'relations Followingthe analogywiththepronominal set (Benveniste1966a; 1966b)we 'I' of culture(the generator can see thatbetweenthereflexive of the conceptsof soul or spirit)and the impersonal'it' of nature (definerof the relationwith or the somaticalterity), thereis a positionmissing, the 'you', the second person, othertakenas othersubject, whose pointofview is the latent echo ofthatofthe thesupernatural context. An 'I'. I believethatthisconceptcan aid in determining abnormal contextwherein a subject is captured by another cosmologically dominantpoint of view,whereinhe is the 'you' of a non-humanperspective, an objectification is theform as Subject, of the Supernature oftheOther implying human I as a 'you' for this Other. The typical'supernatural' situationin an Amerindian world is the meetingin the forest betweena man - alwayson his own - and a being which is seen at first merelyas an animal or a person,then revealsitself or a dead personand speaksto the man (the dynamics as a spirit of thiscommunication are well analysedbyTaylor1993a).20 These encounters can be lethalforthe interlocutor who, overpowered bythe non-humansubjectivity, passes over to its side, transforming himselfinto a being of the same species as thespeaker:dead, spirit or animal.He who respondsto a 'you' spokenbya nonhuman acceptstheconditionofbeingits'second person',and when assumingin his turnthe positionof 'I' does so already as a non-human.The canonicalform of thesesupernatural encounters, then,consistsin suddenlyfinding out thatthe otheris 'human',thatis,thatitis the human,which automatically dehumanizes and alienatesthe interlocutor and transforms him into a preyobject,thatis, an animal. Only shamans,multinatural and office, beingsby definition are always thevariousperspectives, capableoftransiting callingand beingcalled 'you' bythe animal subjectivitiesand spiritswithout losing their condition as human subjects.2' I would conclude byobserving that Amerindian perspectivism has a vanishing point,as it were,where the differences betweenpointsof view are at the same time annulled and exacerbated: which thus takeson the character myth, of an absolutediscourse.In myth, every speciesofbeingappearsto othersas itappears to itself(as human), while acting as if already showing its distinctive and definitive nature (as animal,plant or spirit).In a certainsense, all the beings which people mythology are shamans,which indeed is explicitly affirmed by some Amazoniancultures(Guss 1989: 52). Mythspeaksofa stateofbeingwhere bodies and names, souls and affects, the I and the Other interpenetrate, in thesame pre-subjective submerged and pre-objective milieu- a milieuwhose

484

EDUARDO

VIVEIROS DE CASTRO

end is precisely whatthe mythology setsout to tell.

NOTES A shorterversion of this article was presented as a Munro Lecture at the Universityof Edinburghearlierthisyear.The articleis theresultof an extendeddialoguewithTania Stolze Lima, who, in parallelwith and synchronousto its earlierversion (published first in Portuguese),has written a masterful articleon perspectivism in Jurunacosmology(Lima 1996). PeterGow (who, togetherwith Elizabeth Ewart, translatedmost of the article into English), Aparecida Vilaca, Philippe Descola and Michael Houseman made invaluable suggestionsat various stages in the elaborationof the materials I presenthere. Bruno Latour (1991) was an indirect but crucialsource of inspiration. I read an essayby FritzKrause (1931, After thisarticlehad reacheditspresentform, mentionedbyBoelscher 1989: 212 n.10) which advancesideas strikingly similarto some developed here. I This notion of the body as a 'clothing'can be foundamongstthe Makuna (Arhem 1993), the Yagua (Chaumeil 1983: 125-7), the Piro (Gow, pers. comm.), the Trio (Riviere 1994) the Upper Xingu societies (Gregor 1977: 322). The notion is verylikelypan-American, havingconsiderable symbolicyield forexample in North-westCoast cosmologies (see Goldman 1975 and Boelscher a question I cannotconsiderhere. 1989), ifnot of much wider distribution, 2 For some examplessee amongstmanyothers: Weiss 1969: 158; 1972 (Campa); Baer 1994: 102, 119, 224; Renard-Casevitz 1991: 24-31 (Matsiguenga); Grenand 1980: 42 (Wayapi);Viveiros de Castro 1992a: 68 (Arawete);Osborn 1990: 151 (U'wa); Jara1996: 68-73 (Akuriyo). 3See for example, Saladin d'Anglure 1990; Fienup-Riordan 1994 (Eskimo); Nelson 1983; McDonnell 1984 (Koyukon,Kaska); Tanner 1979; Scott 1989; Brightman 1993 (Cree); Hallowell 1960 (Ojibwa); Goldman 1975 (Kwakiutl);Guedon 1984 (Tsimshian);Boelscher 1989 (Haida). See also Howell 1984; 1996; and Karim 1981,fortheChewong and Ma'Betisek of Malaysia; forSiberia, Hamayon 1990. See Arhem 1993; Crocker 1985; Hugh-Jones1996; Overing 1985; 1986; Vila?a 1992. of totem and 5Or, as we may add, the case of the Ojibwa, where the co-existenceof the systems forthe generaloppositionbetweentotemism manido (Levi-Strauss1962a: 25-33) servedas a matrix and sacrifice withinthe framework (Levi-Strauss1962b: 295-302) and can be directly interpreted of a distinction betweentotemismand animism. 6 See Radcliffe-Brown 1952: 130-1, who, amongstother interesting arguments, distinguishes of species and naturalphenomena (which 'permitsnatureto be thoughtof processes of personiftcation as if it were a societyof persons,and so makes of it a social or moral order'), like those found of naturalspecies, like amongstthe Eskimos and Andaman Islanders,fromsystems ofclassification those found in Australiaand which compose a 'systemof social solidarities'between man and as well as the nature- this obviously calls to mind Descola's distinctionof animism/totemism of manido/totem contrast exploredby Levi-Strauss. I The uncomfortable tension inherentin such antinomiescan be gauged in Howell's article where the Chewong are describedas being both 'relativist' and (1996) on Chewong cosmology, - a double mischaracterization, I believe. 'anthropocentric' 8 'Such is the foundation It does not expressa dependencyon a predefined of perspectivism. whateveraccedes to the point of view will be subject ...' (Deleuze 1988: subject; on the contrary, 27). 9 'Human beings see themselves as such; the Moon, the snakes,thejaguars and the Mother of Smallpox,however,see themas tapirsor peccaries,which theykill' (Baer 1994: 224). 10If salmon look to salmon as humans to humans- and thisis 'animism'- salmon do not look do humansto salmon (theylook likespirits, human to humans (theylook likesalmon),and neither or maybebears;see Guedon 1984: 141) - and thisis 'perspectivism'. Ultimately, then,animismand to totemism thanDescola's model allows for. perspectivism may have a deeper relationship 11The attribution of human-likeconsciousness and intentionality (to say nothingof human denominated bodily form and cultural habits) to non-human beings has been indifferently or 'anthropomorphism'. 'anthropocentrism' However, these two labels can be taken to denote radicallyopposed cosmological outlooks. Westernpopular evolutionismis veryanthropocentric, but not particularlyanthropomorphic. On the other hand, 'primitive animism' may be if sundryotherbeings not anthropocentric: characterized as anthropomorphic, but it is definitely

EDUARDO VIVEIROS DE CASTRO

485

besides humans are 'human', thenwe humans are not a special lot. 12 'The point of view is located in thebody,saysLeibniz' (Deleuze 1988: 16). 13 The counterproof of the spiritin our cosmologieslies in the factthatwhen of the singularity itwith is out of bounds - to identify to universalizeit,we are obliged- now thatsupernature we try the structure and functionof the brain.The spiritcan only be universal(natural) if it is (in) the body of unequivocal examples of spiritpossession in the complex of Amerindian 14 The rarity The classical shamanismmay derivefromthe prevalenceof the themeof bodilymetamorphosis. illuminatedfromthis problem of the religiousconversionof Amerindianscould also be further and angle; indigenous conceptions of 'acculturation'seem to focus more on the incorporation than on spiritual sex) rather embodimentof Westernbodily practices(food, clothing,interethnic assimilation (language,religionetc.). 15 The traditionalproblem of Western mainstreamepistemologyis how to connect and have to be made); theproblemin Amazonia universalize(individualsubstancesare given,relations is how to separate and particularize(relations are given, substances must be defined). See Brightman (1993: 177-85) and Fienup-Riordan(1994: 46-50) - both inspiredby Wagner's(1977) - on thiscontrast. ideas about the 'innate'and the 'constructed' 16 In Amazonian cannibalism, of the subjectwhat is intendedis preciselythe incorporation in verymuch the same way as that hyper-subjectivized, aspect of the enemy (who is accordingly not itsdesubjectivization as is the case describedby Harrison [1993: 121] forMelanesian warfare), with game animals. See Viveirosde Castro 1992a: 290-3; 1996: 98-102; Fausto 1997. 17 PeterGow (pers. comm.) tellsme thatthe Piro conceive of the act of putting on clothesas an animatingof clothes. See also Goldman (1975: 183) on Kwakiutl masks: 'Masks get "excited" duringWinterdances'. a body coveringbut also refers to theskilland 18 "'Clothing" in thissense does not mean merely to carry out certaintasks'(Rivierein Koelewijn 1987: 306). ability 19Religions based on the cult of the ancestorsseem to postulatethe inverse:spiritualidentity of death,the livingand thedead are similarin so faras theymanifest goes beyondthebodilybarrier and spiritualpossession on the same spirit.We would accordingly have superhumanancestrality on the other. one side, animalizationof the dead and bodilymetamorphosis 20 This would be the truesignificance of appearances'theme:appearances of the 'deceptiveness deceive because one is nevercertainwhose point of view is dominant,thatis, which world is in injunctionnot of thisidea to thefamiliar withotherbeings.The similarity force when one interacts appearance. to 'trust just anotherdeceitful is, I fear, yoursenses' ofWesternepistemologies 21 As we have remarked, animals, work consistsin de-subjectivizing a good partof shamanistic thatis in transforming them intopure,naturalbodies capable of being consumed withoutdanger. theminto In contrast, thefactthattheyare inedible;thistransforms whatdefinesspirits is precisely eaterspar excellence, i.e. into anthropophagousbeings. In this way, it is common for the great and it is understandable manifest themselves, formsin which spirits to be the preferred predators thatspirits and predatoranimals should see us as thatgame animals should see humans as spirits, game animals and thatanimals takento be inedible should be assimilatedto spirits(Viveirosde Castro 1978). The scales of edibility of indigenousAmazonia (Hugh-Jones1996) should therefore include spirits at theirnegativepole. REFERENCES en el tr6pico ecologia alternativa humedo Arhem,K. 1993. Ecosofia makuna. In La selvahumanizada: Fondo FEN colombiano (ed.) F. Correa. Bogota': Instituto Colombiano de Antropologifa, Colombia, Fondo EditorialCEREC. Amazon. In 1996. The cosmic food web: human-naturerelatednessin the northwest (eds) P Descola & G. Palsson. London: Routledge. anthropological perspectives Nature andsociety: Quito: Abya-Yala. de losMatsiguenga. yshamanismo Baer,G. 1994. Cosmologfa Paris: Gallimard. de linguistiquege'ne'rale. Benveniste, E. 1966a. La naturedes pronoms.In Problemes Paris: 1996b. De la subjectivitedans le langage. In Problmes de linguistique ge'ne'rale. Gallimard. discourse. Vancouver:Univ. of British Haida social and mythical within: Boelscher,M. 1989. Thecurtain Columbia Press. Univ. of California R. 1993. Grateful relationships. Berkeley: prey: RockCreehuman-animal Brightman,

486

EDUARDO VIVEIROS DE CASTRO

Press. e os outros. Sao Paulo: Hucitec. Carneiroda Cunha, M.M. 1978. Os mortos Paris:Ecole le chamanisme chezlesYagua du nord-est Chaumeil,J.-P 1983. Voir, savoir, pouvoir: pe'ruvien. des Hautes Etudes en Sciences Sociales. and shamanism. natural Tucson: Univ. of cosmology, symbolism, Crocker, J.C. 1985. Vital souls:Bororo Arizona Press. Paris:Minuit. Deleuze, G. 1988. Le pli: Leibnizetle baroque. Paris:Maison des etpraxis dansl'e'cologie desAchuar. domestique: symbolisme Descola, P 1986. La nature Sciences de i'Homme. In Conceptualizing society (ed.) A. Kuper. 1992. Societies of natureand the natureof society. London: Routledge. en a l'approvisionnement: chasse,alliance et familiarisation Erikson,P 1984. De l'apprivoisement Amazonie amerindienne. Techn. Cult.9,105-40. entreos Parakandda Amaz6nia oriental. Fausto, C. 1997. A dialeticada predacdoe familiarizacdo Thesis, Museu Nacional, Univ. of Rio de Janeiro. and passages:rule and ritualin Yup'ikEskimooral tradition. Fienup-Riordan,A. 1994. Boundaries Norman: Univ. of Oklahoma Press. an introduction New York:WileytoKwakiutl thought. religious Goldman, I. 1975. The mouth ofheaven: Interscience. Gow, P 1989. The perversechild: desire in a nativeAmazonian subsistenceeconomy.Man (N.S.) 24, 567-82. and history in Peruvian Amazonia.Oxford: Clarendon Press. 1991. Of mixed blood: kinship Chicago: Univ. of Gregor,T. 1977. Mehinaku:thedramaofdailylIfein a BrazilianIndian village. Chicago Press. de l'univers desIndiens du Haut-Oyapock d l'etude waydpi: ethno-e'cologie Grenand,P 1980. Introduction (GuyaneFran,aise). Paris: SELAF/CNRS. In The to the Tsimshianworld view and its practitioners. Guedon, M.-F. 1984. An introduction thepresent (ed.) M. Seguin. Vancouver: Univ. of British Tsimshian: images ofthepast,viewsfor Columbia Press. and narrative in theSouth American rainforest. Berkeley: symbol Guss, D. 1989. To weaveand tosing:art, Univ. of CaliforniaPress. and world view.In Culture in honor in history: essays behavior, Hallowell, A.I. 1960. Ojibwa ontology, ofPaul Radin (ed.) S. Diamond. New York: Columbia Univ. Press. du chamanisme d l'dme:esquisse d'unethe'orie Nanterre:Societe sibe'rien. Hamayon, R. 1990. La chasse d'Ethnologie. and the Manchester:Univ. Press. ritual Harrison,S. 1993. The maskofwar:violence, sefin Melanesia. and cosmos: Chewong ofpeninsular Malaysia.Oxford: Univ. Press. Howell, S. 1984. Society 1996. Nature in culture or culture in nature? Chewong ideas of 'humans' and other anthropological perspectives (eds) P Descola & G. Palsson. London: species. In Natureand society: Routledge. Hugh-Jones,S. 1996. Bonnes raisons ou mauvaise conscience? De l'ambivalence de certains Amazoniens enversla consommationde viande. Terrain 26, 123-48. in human evolution.Cult.Dyn. 4, Ingold, T. 1991. Becoming persons:consciousnessand sociality 355-78. In Companion humanity, culture encyclopedia ofanthropology: 1994. Humanity and animality. andsociallife (ed.) T. Ingold. London: Routledge. In Redefining 1996. Hunting and gathering as ways of perceiving the environment. nature: culture and domestication (eds) R.F. Ellen & K Fukui. London: Berg. ecology, delKumu: ecologia entre losAkuriyo de Surinam. Quito: Abya-Yala. y ritual Jara,F. 1996. El camino London: AthlonePress. concepts ofliving things. Karim,W-J. 1982. Ma'betisek to live:theCashinahuaofeastern Peru. Prospect Heights: Kensinger,K. 1995. How realpeopleought WavelandPress. Indians Dordrecht:Foris. oftheTrio ofSurinam. Koelewijn, C. with P Riviere 1987. Oral literature das Motiv der Hille und das Prinzip der Form. Ethnol. Krause, F. 1931. Maske und Ahnenfigur: Stud. 1, 344-64. modernes. Paris:EditionsLa Decouverte. Latour,B. 1991. Noun n'avonsjamaiste' C. 1962a. Le tot6misme Paris: PressesUniversitaires de France. aujourd'hui. JAvi-Strauss, 1962b.La pense'e Paris: Plon. sauvage.

EDUARDO VIVEIROS DE CASTRO

487

In hisAnthropologie structurale deux.Paris: Plon. 1973 [1952]. Race et histoire. Paris: Plon. 1985. La potierejalouse. of Rio Thesis, Museu Nacional, University Lima, T.S. 1995. A partedo cauim: etnografiajuruna. de Janeiro. em uma cosmologia tupi. sobre o perspectivismo 1996. 0 dois e seu multiplo:reflex6es Mana 2:2, 21-47. McCallum, C. 1996. The body that knows: from Cashinahua epistemology to a medical of lowland South America.Med.Anthrop. Q. 10:3, 1-26. anthropology on theKaska symbol turmoil:centering and systematic McDonnell, R. 1984. Symbolicorientations of dene.CanadjAnthrop.4, 39-56. vital substance,and knowledge Mentore, G. 1993. Temperingthe social self: body adornment, Anthrop. 9, 22-34. among theWaiwai.J.Archaeol. Raven.Chicago: Univ. of Chicago Press. tothe Nelson, R. 1983. Makeprayers animals: In Signifying Osborn, A. 1990. Eat and be eaten: animals in U'wa (Tunebo) oral tradition. world in thenatural human (ed.) R. Willis. London: Unwin Hyman. meaning Overing,J. 1985. There is no end of evil: the guiltyinnocentsand their falliblegod. In The ofevil(ed.) D. Parkin.London: Basil Blackwell. anthropology Soc.Amer. deathand dominationin a 'non-violent'society.J. 1986. Images of cannibalism, 72, 133-56. Pollock, D. 1985. Personhood and illnessamong the Culina of westernBrazil. Thesis, University of Rochester. andfunction A.R. 1952 [1929]. The sociologicaltheoryof totemism.In Structure Radcliffe-Brown, London: Routledge& Kegan Paul. inprimitive society. Man G. 1976. Cosmology as ecological analysis:a view fromthe rain forest. Reichel-Dolmatoff, (N.S.) 12, 307-18. and metaphors Amazon. In Animalmyths 1985. Tapiravoidance in the Colombian northwest America in South (ed.) G. Urton. Salt Lake City: Univ. of Utah Press. del'e'tranger. Paris:Lierre& Coudrier. unemythologie F.-M. 1991. Le banquet masque': Renard-Casevitz, Riviere,P. 1994. WYSINWYIG in Amazonia.JASO 25, 255-62. space and social Saladin d'Anglure,B. 1990. Nanook, super-male;the polar bear in the imaginary world animals: meaning in thenatural human timeof the Inuit of the Canadian Arctic.In Signifying (ed.) R. Willis.London: Unwin Hyman. of Chicago. Schwartzman, S. 1988. The Panara of theXingu National park.Thesis, University Scott, C. 1989. Knowledge construction among the Cree hunters: metaphors and literal 75, 193-208. Soc.Ame'r. understanding.J. Seeger,A. 1980. Corporacdo e corporalidade:ideologia de concepcdo e descendencia. In his Os e nos.Rio de Janeiro:Campus. tndios Cree oftheMistassini ideology and modeofproduction homeanimals:religious Tanner,A. 1979. Bringing hunters. StJohn's: Memorial Univ. of Newfoundland. langageet croyancedans la pensee achuar.L'Homme stupefiants: Taylor, A.-C. 1993a. Des fant6mes 126-8, 33/2-4,429-47. identity, mourningand memoryamong theJivaro.Man 1993b. Rememberingto forget: (N.S.) 28, 653-78. G. 1993. Song paths:theways and means ofYaminahuashamanicknowledge.L'Homme Townsley, 126-8,33:2-4, 449-68. In Beyond T. 1991. 'We are parrots, twinsare birds':playof tropesas operationalstructure. Turner, thetheory in anthropology Univ. Press. (ed.) J. Fernandez.Stanford: oftropes metaphor: among the and sociality 1995. Social body and embodied subject:bodiliness,subjectivity, 10, 143-70. Kayapo. Cult.Anthrop. comogente:formas do canibalismo Wari'(Pakaa-Nova).Rio de Janeiro:Editora Vilaca,A. 1992. Comendo da UFRJ. Viveirosde Castro,E. 1978. Alguns aspectosdo pensamentoyawalpiti(AltoXingu): classificac6es e transformac6es. Bol. Mus. nac.26, 1-41. 1979. A fabricacdo do corpo na sociedade xinguana.Bol. Mus. nac.32, 2-19. society. Chicago: in an Amazonian 1992a. Fromtheenemy's and divinity point ofview:humanity Univ. of Chicago Press. do canibalismo 1992b.Apresentacdo to A. Vilaca. In Comendo como gente:formas Wari'Rio de Janeiro:Editoralda UFRJ.

488

EDUARDO VIVEIROS DE CASTRO

1993. Alguns aspectos da afinidadeno dravidianatoamaz6nico. In Amaz6nia:etnologia e hist6ria indigena (eds) E. Viveirosde Castro & M. Carneiro da Cunha. Sao Paulo: Nucleo de Historia Indigenae do Indigenismo(USP/FAPESP). 1996. Le meurtrier et son double chez les Arawete:un exemple de fusion rituelle.Syst. Pens.AfrNoire14, 77-104 of the innate: a semiotic Wagner,R. 1977. Scientificand indigenousPapuan conceptualizations In Subsistence andsurvival: in the rural critiqueof the ecological perspective. ecology Pacfic(eds) T. Bayliss-Smith & R.G. Feachem. London: Academic Press. Weiss, G. 1969. The cosmology of the Campa Indians of eastern Peru. Thesis, Universityof Michigan. 1972. Campa cosmology. Ethnology 11, 157-72.

Deiis cosmologique et perspectivisme amerindien

Re'sume' du 'perspectivisme' Cet article discute la signification am6rindien,c'est-a-direles idees qui concernentla facondont les humains,les animauxet les espritsse percoivent eux-memeset se percoivent les uns les autres dans les cosmologies amerindiennes.Ces idees suggerentla les categories sur la base possibilitede redefinir classiques de 'nature','culture'et 'supernature' des concepts de perspectiveou de point de vue. Larticle soutientplus particulierement que l'antinomieentredeux caracterisations de la pensee indigene- d'une part l"ethnocentrisme' selon lequel les attributs a d'autres de l'humaniteseraientrefusesaux humains appartenant groupes,et d'autrepartl"animisme',qui appliquerait ces qualites humainespar extensiona des entreles a d'autresespeces - peut etreresolue si l'on considerela diff6rence etresappartenant et corporelsdes etres. aspectsspirituels King'sCollege, Cambridge CB2 1ST