Descola: Beyond nature and culture: Forms of attachment

Publisher’s note: This article is chapter 13 of the forthcoming translation by Janet Lloyd © by The University of Chicago. All rights reserved. Published 2013. The University of Chicago Press, Chicago 60637. The University of Chicago Press, Ltd., London. Originally published as Philippe Descola, Par-delà nature et culture (Paris: Editions Gallimard, 2005) © Editions Gallimard, Paris, 2005. We are very grateful to the University of Chicago Press for permission to publish this chapter.

Beyond nature and culture

Forms of attachment

Philippe Descola, Collège de France

Translated by Janet Lloyd


Modes of identification broadly schematize our experience of things, distinguishing between parcels of ontological properties distributed in accordance with the arrangements of existing beings, arrangements whose structural characteristics we have examined above, each in turn. It is a distribution of beings according to their attributes, the principles according to which socio-cosmological collectives are organized, the dominant regimes of knowledge and action, and the boundaries of identity and otherness. Each of these forms of identification defines a specific style of relations with the world. Long-established expressions of these relations are to be found in geographical regions, many of which are immense, and over very long periods. Yet we cannot use those styles as criteria for distinguishing between singular collectives with contours limited both in time and in space—the kind that historians, ethnologists and sociologists usually choose to investigate. Rather, we should regard those stylizations of experience as what are usually called “world views,” “cosmologies” or “symbolic forms,” all of these being terms of vague epistemological status yet that constitute a handy intuitive way of synthesizing under a simple label (such as “the modern West,” or “shamanistic societies”), “families” of practices and mind-sets that seem to display affinities despite the diversity of their concrete manifestations. However, within those great archipelagos marked out by a shared mode of identification one comes across numerous kinds of collectives that consider themselves to be very different from one another (and that are, indeed, perceived as different by those who study them). This is not only on account of their different languages, institutions and, more often than not, the discontinuity of their territories, but also because the interactions within them present remarkable contrasts. For even when the ontological distribution of existing beings and the ways that they come together are based on identical principles, the links that they weave between one another, the ways that they affect one another and the manner in which they treat one another can all vary through and through. It is thus primarily the general form of the local relations that structure the connections between entities that are all distinguished by the same process of identification, which makes it possible for collectives to differentiate themselves from one another and for each to display the singularity of their own particular ethos, of which any observer soon becomes aware.

Like modes of identification, relational modes are integrating schemas; that is to say they stem from the kind of cognitive, emotional, and sensory-motor structures that channel the production of automatic inferences, orientate practical action, and organize the expression of thought and feelings according to relatively stereotyped patterns. A relational schema becomes dominant in a collective when activated in a whole range of very different circumstances in relations with humans or nonhumans. The effect of this is to subject all relations to its particular logic, either by limiting their field of application or by subordinating this to the achievement of the ends that the dominant schema embodies. But, unlike modes of identification, dominant relational modes are also identifiable thanks to the fact that in many cases they express the greatest possible difference from those in action in the immediate neighbourhood. It is as if each collective would concentrate its greatest efforts on whatever it judged to be capable of distinguishing it most effectively from the collectives surrounding it and with which it coexists: namely, the styles of interaction and behaviour that its human members are led to adopt in the course of daily life. However, the nature and the limits of a collective of this kind are never fixed a priori since it is, on the contrary, the area covered by the dominant relational schema that establishes them in the first place. A collective defined in this way does not necessarily coincide with a “society,” a “tribe,” or a “class,” all of which are misleading terms to use because of the substantive closure that they imply. Rather, it is characterized primarily by the discontinuity that is introduced all around it on account of the ostensible close presence of other principles for the schematization of relations between existing beings. Its existence is thus positional, not intrinsic, and is revealed through comparisons.

When seen as dispositions that bestow form and content upon the practical links between myself and a human or non-human alter, relational schemas can be classified according to whether or not that alter is or is not equivalent to me on an ontological level and whether the connections that I establish with it are or are not mutual. So numerous are the kinds of relations that can be established between the entities that fill the world that it is clearly not possible to summarize them all. So let us concentrate here upon no more than one group of six types of relationships which appear to play a preponderant role in the connections that humans establish between one another and also with non-human elements in their environment. Whether they are identified by the words that I use or are given other names, these relations have for many years attracted the attention of the social sciences, some of them even to the point of becoming key concepts. The relations in question are those of exchange, predation, gift, production, protection and transmission. These relational modes that come to modulate all modes of identification may be divided into two groups. The first is characterized by potentially reversible relationsbetween terms that are similar. The second is characterized by univocal relations that are founded upon connections between non-equivalent terms. The first group covers exchange, predation and gift; the second covers production, protection and transmission.

Giving, taking, exchanging

The relations at work in the first group correspond to three formulae that ensure the movement of something valuable between two terms of the same ontological status, terms that may themselves actually contain that value and therefore circulate in such a way that one may be led to disappear physically as a result of being absorbed by the other. The first relationship, that of “exchange,” appears as a symmetrical one in which any agreed transfer from one entity to the other requires something in return. The other two are asymmetrical. In the one, entity “A” takes something of value from entity “B” (perhaps its life, its body or its interiority) without offering anything in exchange: “predation” is what I call this negative asymmetry. In the other, entity “B” offers something of value to entity “A” (maybe even itself) without expecting any compensation: I call this positive asymmetry “gift.” At least two of the terms that I use to qualify these relations have a long anthropological history, so I need to specify their meaning in relation to previous definitions.

As is well known, Lévi-Strauss ascribes a crucial role to exchange in the developing and functioning of social life. The prohibition of incest is a rule of reciprocity in that it instructs a man to renounce a woman for the benefit of another man who, in turn, rules out his use of another woman who thus becomes available for the first man. The prohibition of incest and the exogamy that is the positive side to it would therefore simply be a means of instituting and guaranteeing reciprocal exchange, which is the basis of culture and a sign of the emergence of a new order in which the relations between groups are governed by freely accepted conventions. But culture does not play a totally innovating role here. According to Lévi-Strauss all it does is codify universal mental schemas that pre-exist the norms that bring them into play. Among these categorical imperatives that are written into the architecture of the mind before the emergence of the symbolism that makes it possible to express them, one finds “the notion of reciprocity regarded as the most immediate form of integrating the opposition between the self and others; and the synthetic nature of the gift, i.e. that the agreed transfer of a valuable from one individual to another makes these individuals into partners and adds a new quality to the valuable transferred” (Lévi-Strauss 1969 [1967]: 84). The pre-eminence of reciprocity and gift thus results from the fact that those two means of founding and maintaining the social link are a legacy of human phylogenesis, a reminder of the function of natural predispositions in the structuring of the “being together” that is organized by culture.

It is no doubt not necessary to go as far as Lévi-Strauss and postulate an innate neural basis for reciprocity and gift, in order to agree with him that those two relational schemas do indeed orientate many forms of human behaviour. Besides, there is nothing new about the idea. The suggestion that reciprocal exchange and gift constitute the true cement of all social life seems to be a leitmotif in Western political philosophy, in which it is hard to sort out how much of the idea stems from empirical observation and how much from a moral ideal regarding the most desirable way of ensuring that a collective of equals sticks together. Although he does not explicitly acknowledge it, Lévi-Strauss is thus positioned along the main line of development from a tradition recorded as early as Antiquity. Aristotle, for instance, declares that reciprocity in relations of exchange “is the bond that maintains the association,” and Seneca declares that gift “constitutes the chief bond of human society.”1 However, this venerable precedent should not deter us from asking two questions. Is it legitimate to associate reciprocity and gift within the same set of phenomena? And is it certain that every collective considers those two values as the basis of its social life?

In answer to the first question, we must briefly look back to Marcel Mauss’ famous Essay on the gift and consider this text’s influence on Lévi-Strauss (Mauss 1950).2 Although critical with regard to certain aspects of this essay, Lévi-Strauss does confirm the conception of the gift that Mauss presents there, namely “a system of total prestation,” characterized by the three obligations of giving, receiving, and giving back. Lévi-Strauss does not challenge that definition, but he does criticize the way in which, he claims, Mauss explained the reciprocity involved in the exchange of gifts and counter-gifts, namely by resorting mainly to a local theory centred on the Polynesian notion of hau, a mysterious force that resides in the given gift, which forces the gift-receiver to reciprocate. Lévi-Strauss claims that Mauss allowed himself to be mystified by a deliberate interpretation put forward by a group of native specialists instead of endeavouring to discover the underlying realities of exchange where they could be found, that is to say “in the unconscious mental structures that may be reached through institutions” (Mauss 1950: xxxix). If exchange plays a founding role in social life, according to Lévi-Strauss that is because it constitutes an absolutely primitive phenomenon, “a synthesis immediately given to, and by, symbolic thought” (Mauss 1950: xlvi). What is paradoxical about this famous critique is that Lévi-Strauss seems not to have realized that Mauss’ characterization of the gift, with which he himself does not disagree, was in truth derived from another theory, every bit as native as that of the hau, but also truly Western. For Mauss implicitly echoes his own cultural tradition when he interprets the gift as resting upon the obligations of giving, receiving, and giving back. As Denis Vidal has shown, this is a common interpretation that goes back to the well-known ancient image of the Three Graces. These constitute a most precise allegory of the three obligations that surround gifts, as Seneca makes perfectly clear: “some would have it appear that there is one for bestowing a benefit, another for receiving it, and a third for returning it” (Seneca 1935: 13). Despite his extensive classical culture, Mauss never mentions this line of thought on the theme of gifts, which commentaries on the Three Graces have high-lighted from Chrysippus right down to Pico de La Mirandola. But it seems unlikely that that unacknowledged source did not affect his conception of the nature of gifts (the three obligations). It may also have affected his desire to see restored the values associated with it, namely generous behaviour, in particular the energeticism ofprominent figures, that testifies to a reputedly more authentic sociability as illustrated by archaic societies (Vidal 1991).3

In view of its antecedents, it thus seems reasonable to question whether this concept of the gift bequeathed by the Ancients, which anthropology then took to its heart in the wake of Mauss, really does match the practice that it claims to characterize. For, unlike exchange, the gift is above all a one-way gesture that consists in abandoning something to someone without expecting any compensation other than that, possibly, of gratitude on the part of the receiver of the gift. For if the notion is given its literal meaning, reciprocal benefaction is never guaranteed where a gift is concerned. To be sure, reciprocation is a possibility that one may well hope for, either as a tacit wish or an out-and-out calculation, but the realization of such a wish remains independent of the actual act of giving, which would, ipso facto lose its meaning if it was conditioned by an imperative to obtain something in compensation. Alain Testart is thus quite right to draw a clear distinction between exchange and gifts: the former consists in handing something over in return for something else; the latter, in doing so with no expectation of reciprocity (Testart 1997). Thus, the presents that I receive from those close to me on the occasion of my birthday cannot in any way be regarded as a deferred return in exchange for the presents that I gave them on their birthdays, for there is no obligation inherent to the custom according to which these gracious transfers take place that would make the present given to me conditional upon the gifts that I offered. It is true that one may say that one is “much obliged” by the gift that one receives. But, contrary to what Mauss claims, the fact that one is “obliged” in no way makes a counter-gift obligatory, at least not in the sense in which an initial favour might be accompanied by a compelling clause such as those that stem from a contract or responsibility and which, ignored, might well lead to sanctions. In the case of a present, the obligation to repay one’s benefactor in some way is purely moral. If one evades it, one may eventually be despised, lose face or be labelled stingy by the gift-giver, although for him there can be no recourse to any means of obtaining reciprocity for something freely given and that he would never even think of demanding. If the gift gives rise to any obligation, it is, strictly speaking, neither obligatory nor obliging (Testart 1997: 43).

In this respect, a gift is profoundly different from an exchange. Every gift constitutes an independent transfer by reason of the fact that nothing can be claimed in return. There are, of course, societies where it is customary to respond to a gift with another gift, as in the potlatch of the Indians of the north-west coast of America, which is very much to the fore in the Essay on the gift: the riches offered during a ceremony provided the opportunity of presenting some appropriate counter-gift in the course of some later ceremony. Yet no-one was, strictly speaking, obliged to honour a gift with a counter-gift. In societies that had placed generosity at the pinnacle of their values, not to do so certainly meant that one’s honour was seriously sullied and one’s access to the highest spheres of political prestige would be compromised wherever that political prestige was founded above all on one’s reputation for liberality. But even if, as in many other societies too, the fear of being discredited no doubt constituted a powerful motive to respond with a counter-gift, that was not the same as the obligation to repay that would have been implied in a contractual and quasi-legal way by the fact of accepting the original gift.

Exchange, in contrast, requires as a necessary condition that something be obtained in return. Regardless of whether or not it is equal in value to the thing received, it is this return that represents both the purpose and the means of the exchange, whether this be immediate or deferred and whether or not it be a commercial deal. For even when its nature is not explicitly stipulated or when the time allowed for repayment is not specified, some kind of reciprocation can always be demanded: each party only gives away the goods that he has in exchange for other goods. In this sense, as Testart also notes, the essence of exchange lies in two transfers in opposite directions, transfers that are intrinsically linked, since each of them results from an obligation the raison d’être of which lies in the other. The whole operation is thus a closed system, which may, of course, be inserted into a whole series of similar transactions, but each of which is formed by an independent combination of two elementary mirrored operations (Testart 1997: 51). Unlike a gift, which is a single transfer that may eventually prompt a counter-transfer but for motives other than the principle of liberality that made it possible in the first place, each of the two transfers that an exchange involves is both the cause and the effect of the other. A reciprocal relationship is inherent to this kind of deal and is peculiar to it: I give to you so that you give to me, and vice-versa.

This is why the all-too vague notion of reciprocity should be set aside when analysing transfer relations. Literally, reciprocity only designates what happens between term “A” and term “B” and subsequently between term “B” and term “A.” There is no overt indication of the nature of the obligations that link the two terms. Thus, a gift may be reciprocal when it is followed up by a counter-gift, although reciprocity, even so, does not constitute an intrinsic characteristic of this type of transaction, given that a gift in return is not obligatory or binding as it is in an exchange. On the other hand, exchange does necessarily imply reciprocity, since it is precisely the obligation to respond with a counterpart that defines it. I shall therefore be using the word “exchange” to refer to what Lévi-Strauss sometimes means by “reciprocity”: namely, a transfer that requires something in return and, contrary to the use established by Mauss, I shall use the term “gift” to refer to an accepted transfer with no obligation to provide a counter-transfer. It is difficult altogether to avoid the cinematic illusion that leads to characterizing transfers of things or persons by the directions in which they move (and I am aware that I myself did that in my initial definition of exchange, predation and gifts). But if one takes into account the form of the transfers according to the obligations that they impose, it becomes possible largely to correct that distortion of perspective. In any case, it encourages one to distinguish between phenomena arbitrarily grouped under the same rubric simply because they involve the circulation of things between particular terms. It then becomes impossible to continue to set on the same level the exchange of goods in all its different forms, the exchange of signs in language, the exchange of women in marriage alliances, the exchange of deaths in a vendetta, or the sequence of gifts and counter-gifts.

Whether understood in the general sense that Lévi-Strauss lends it or in the more specific sense that I give it, exchange is certainly present in all societies and takes such diverse forms that it is not hard to understand the point of view of those who have wished to see it as the principal “bond that maintains association.” Gifts are also common at every latitude. From the Stoics down to Mauss, many authors have hinted at their nostalgia for a hypothetical golden age in which this disinterested practice was more widespread and have expressed their desire to see a generalization of benevolent practices that would give some idea of the consideration that people have for one another. All the same, no moralist has ever seriously thought that gifts could become the key regulatory institution in any real society. So what grounds do I have for saying that gifts might constitute the integrating relational schema in certain collectives? How can one even think of undermining the pre-eminence of exchange by suggesting that patterns of behaviour founded on the principle of agreed gracious transfers may have been adopted as an ideal norm by certain human communities? These rhetorical questions prompt a reminder: a mode of relations does not become dominant because it has successfully supplanted schemas of interactions that do not accept its logic. It does so because it provides the most effective cognitive model for a simple and easily remembered synthesis not only of many patterns of behaviour but also, and above all, of those recognized to be the most distinctive of the collective not only by its members but also by outside observers. Just as with modes of identification, no relational schema is hegemonic. The most that can be said is that one or other of them acquires a structuring function in certain places, even if it is not always possible to put a name to it, when, in an immediately recognizable manner, it orientates many attitudes vis-à-vis both humans and non-humans. It is not the case that exchange disappears when the ethos of gifts dominates, for in truth it is simply encompassed by the latter.

If one accepts this, one has to agree that there are plenty of collectives that do seem to place the logic of the gift at the heart of their practices. Without anticipating the ethnographic case studies that will be discussed in the next chapter, I would like to draw attention to the importance granted to the action of “sharing” in recent studies devoted to hunter-gatherer societies, in order to characterize both their internal relations and those that they maintain with plants and animals. Bird-David has played a decisive role in this domain, at least at the terminological level, by forging the expression “a giving environment” to synthesize the conception that the Nayaka of Tamil Nadu have of their forests. Just as humans share everything between them with no thought of obtaining anything in return, similarly the environment unstintingly hands out its liberalities to the Nayaka, so that the human and the non-human components of the collective find themselves integrated into one and the same “cosmic sharing economy” (Bird-David 1990). The idea and practice of the gift constitute among the Nayaka a habitus so deeply rooted that it is inconceivable not immediately to give someone whatever he asks for. When, very exceptionally, the Nayaka do not want to part with something, “rather than disrupt the ongoing sense of sharing—the rhythm of everyday social life—they hid it away or avoided people” (Bird-David 1999: 72). For Bird-David, the values of sharing and the gift are typical of hunter-gatherer societies in general, a point of view that Ingold takes up and elaborates. His view is that what characterizes the so-called “sharing” relations of this kind of society (both with other humans and with non-humans) is simply “trust,” that is to say a particular combination of autonomy and dependence. According to him, to place my trust in a person is to act vis-à-vis him or her in the expectation that he or she will behave toward me in the same favourable spirit as I am manifesting, and will continue to do so for as long as I do nothing to limit his or her autonomy, that is to say his or her option of acting differently. This is thus a situation of dependence that is freely entered into and that places a high value on my partner’s choice to adopt toward me the same attitude that I adopt toward him. In short, “any attempt to impose a response, to lay down conditions and obligations that the other is bound to follow would represent a betrayal of trust and a negation of the relationship” (Ingold 2000: 70; italics original). The difference between on the one hand the giving or sharing and, on the other, exchanging could not be expressed more clearly. The former is unconditional: even to suggest that it is not condemns it immediately to vanish and give way to the latter. Disinterested trust is then replaced by a tacit or contractual obligation.

Hunter-gatherers—to use the current term—do not constitute the only kind of collectives that are characterized by a high value set upon sharing. Those are also the terms that Joanna Overing and some of her disciples use to analyse indigenous sociability in Amazonia, when this is apprehended at the level that seems to them the most significant, namely within the framework of a local group with consanguineous kinship relations. It appears that this domestic and village sphere is marked above all by relations of mutual trust that are confirmed by productive cooperation, daily and festive commensality, an affectionate solicitude for others and a constant flow of gifts and counter-gifts. This is a kind of moral economy based on intimacy, free of calculation and ambiguities, the effect of which is to render those within it so consubstantial that they consider themselves to be of the same species. Within this “aesthetic of conviviality,” sharing plays a central role in that it testifies to a disposition to open oneself up to others with generosity and compassion and thus, through concrete acts, expresses the ethical insistence on unreservedly helping one another that informs the whole of social life.4 Further examples are really not necessary, for our account need not ratify all the ethnographical interpretations that we have mentioned. At this stage in our enquiry all we need do is recognize that there are at least some very diverse societies which, as anthropologists confirm, are animated by an ideology of sharing, here understood as the pre-eminence of the role played by reciprocal gifts in interpersonal relations.


The opposite of a gift is seizing something, making no offer of anything in return. It is an action that creates no more obligations for its perpetrator that the gift does for the gift’s recipients. If one wishes to underline the illicit and generally condemned aspect of the operation, it may be called theft, seizure or wrongful appropriation. However, the term “predation” seems preferable here, in that it conveys the idea that an appropriation of this kind may not result from a desire to harm or some fleeting need. Instead, it may be prompted by the fundamental constraint to which the lives of animals are subjected. Every animal needs to replenish its sources of energy at regular intervals, by consuming some prey, a body originally distinct from itself but that the animal in question ends up assimilating in such a way that it becomes a part of its own organism. Humans are not excepted from that imperative, for they have obeyed it for tens of millennia, even if, thanks to the development of livestock-raising and the deferred consumption of products already transformed by agriculture and herding, the evolution of techniques of subsistence has by now succeeded in partly blurring the memory of the intrinsic link between the capture and the ingestion of prey. Nevertheless, predation remains a central mechanism in the preservation of living creatures, an elementary way in which animals are related to their environment, so much so that René Thom has constructed a mathematical model of the “predation loop,” which seems applicable to many biological processes (Thom 1990: 222-231, 526-530). Predation is thus a phenomenon of productive destruction that is indispensable for the perpetuation of individuals: far from being an expression of gratuitous cruelty or a perverse desire to annihilate others, it on the contrary transforms the prey into an object of the greatest importance for whatever creature ingests it. Indeed, it is the very condition of that creature’s survival.

But is it legitimate to transpose a biological phenomenon to the social sphere and claim that collectives have simply converted predatory patterns of behaviour into a dominant relational schema? First, we should bear in mind (indeed, how could we forget?) that the primacy of exchange and sharing is not accepted by all those who have reflected upon the foundations of political existence. Hobbes was by no means alone in emphasizing, following Plautus, that man’s original condition is to be a wolf to his fellows, for his egoistic awareness of his own interests constantly leads him to try to dispossess others. And although Hobbes’s pessimism has, not without reason, above all been interpreted as an unconscious naturalization of competitive interpersonal relations within a nascent market economy, it cannot be reduced solely to that. For there is no denying that violent appropriation and the destruction of others are not the doubtful privileges of individuals fashioned by a bourgeois society. Traces of both are to be found in every period, at every latitude, and it would be as ridiculous to deny this predatory propensity as to claim it to have been the dominant characteristic of human nature up until such time as the latter was pacified by institutions introduced by the social contract. Predation is a disposition that, among others, is a legacy of our phylogenesis, and if certain collectives have adopted it as their own particular ethos, this means not that they are more savage and primitive than others, but simply that they have found it a paradoxical means of incorporating the deepest kind of otherness while remaining faithful to themselves.

My ethnographic experiences among the Achuar led me, some years ago, to come to that conviction and to apply to the Achuar the notion of “predation” in order to explain a style of relating to both humans and non-humans, based on capturing principles of identity and vital substances reputed to be necessary for the perpetuation of the self. This predatory attitude was evident not only in warfare and its rituals but also in many aspects of daily life; and it was not peculiar to the Jivaro groups. I noticed very soon that signs of it were detectable here and there among other indigenous Amazonian societies, in total contrast to the philosophy of equal exchange by which the Amazonian form of social life had long been exclusively defined. However, it was not at all my intention to substitute one hegemonic relationship for another, for it was also perfectly clear that some of the peoples in this vast cultural region did, for their part, adhere fully to the obligations that exchange imposed (Descola 1990, 1992, 1993). At about the same time and in parallel fashion, Viveiros de Castro was developing his own thoughts about the ontological foundations of cannibalism and warfare in the Tupi world and this led to a model of “the symbolic economy of predation” on the basis of which he set out to elucidate the sociological peculiarities of Dravidian kinship in Amazonia. Far from being symmetrical, as in other Dravidian systems, here the opposition between affinity and consanguinity seems to be characterized by a hierarchical reversal that is dynamically inspired by a diametric structure. Although masked at the level of a particular local group by the behaviour-patterns and values associated with consanguinity, in relations with other local groups affinity seemed to predominate and was itself subordinated to a more totalizing relationship for which it provided a specific code: namely, cannibalistic predation on enemies (Viveiros de Castro 1993).

With the passing of time, and despite a few minor divergent interpretations, the idea that, for many Amazonian societies, predation constituted the cardinal schema for relations with “others” has become widely accepted (See Taylor 1993, 1996; Fausto 2001; Surrallés 2003). But it has also encountered resistance and given rise to many misunderstandings. Without going into the ethnographic details, which we shall be examining in the next chapter, I should make it clear that predation is above all a disposition for incorporating otherness, both human and non-human, because this is reputed to be indispensable for a definition of the self: in order truly to be myself, I must take possession of another being and assimilate it. This can be done by means of warfare, hunting, real or metaphorical cannibalism, the seizure of women and children or by ritual methods of constructing the person and mediating with ideal affines, in which violence is confined to the symbolic level. Predation is not an unbridled manifestation of ferocity or a deadly impulse set up as a collective virtue. Even less is it an attempt to reject as inhuman some anonymous “other.” It constitutes recognition that without the body of this other being, without its identity, without its perspective on me, I should remain incomplete. This is a metaphysical attitude that is peculiar to certain collectives, not a troubled exaltation of violence that some ethnologists might be guilty of, as they project their own fantasies upon the Amerindians.


The trilogy of the gift, exchange and predation seems to present affinities with the distinction that Marshall Sahlins draws between the three forms taken by reciprocity in tribal societies: “generalized” reciprocity qualifies altruistic transfers within the local group and requires no automatic reciprocation; “balanced” reciprocity corresponds to a direct exchange of equivalent values within the tribal group; “negative” reciprocity consists in trying to obtain something for nothing, sometimes in a dishonest or violent fashion, and is a feature of intertribal relations (Sahlins 1968: 82-86; who takes over this typology from Service 1966). But the resemblance is no more than superficial, in the first place because neither the gift nor prédation involve reciprocity. In both cases what is involved is a unilateral operation: the gift comes unaccompanied by any binding obligation to return the favour; and a predatory act is unlikely to imply that the perpetrator ardently hopes for a reciprocal response. Such a response is always possible and in fact often takes the form of reprisals, but it is certainly not constitutive of the intention that prompted the action. Besides, the three kinds of reciprocity affect the direction of movements and the balance sheet, positive, negative or equal, of the passage of objects, not the causes and obligations inherent to each of those kinds of transfers; and it is precisely those causes and obligations that make it possible to distinguish exchange, predation and gift as classes of heterogeneous phenomena. Finally, the typology of reciprocity describes modes of circulating goods that are to be seen everywhere in operation and that, when combined within a single society, are differentiated from one another above all by their position along a continuum defined by the greater or lesser spatial and kinship distance between the agents in those transfers. In contrast, in the sense in which I understand them here, exchange, predation and gift are general relational schemas that concern far more than the circulation of goods, since one or other of them may come to structure the ethos of a collective in a distinctive fashion.

Producing, protecting, transmitting

The relations in the first group allow for reversibility of movement between the terms: this is indispensable for an exchange to take place and it remains possible, though not always desired, in predation and gift. In contrast, the relations in the second group are always univocal and operate between terms set in a hierarchy. This is particularly clear in the case of production. The genetic antecedence of a producer over his product does not allow the latter, in return, to produce its producer (even if it may help to support him), and this places the product in a situation of dependence vis-à-vis the entity to which it owes its existence, at least initially. Marx dispels any doubt about the matter. Production is both a relationship that humans weave among themselves according to well-defined forms in order to procure jointly their means of existence (the relations of production); and it is also a specific relationship to an object that one creates for a particular purpose. In the famous pages of his Introduction to a critique of political economy, Marx stresses that “the act of production is, therefore, in all its aspects an act of consumption as well” (Marx 2010 [1897]: 277). This is because, in the first place, the individual who develops his faculties by producing something expends energy in this operation and consumes raw materials, which are the means of production; this is “productive consumption.” But consumption is also, in an immediate fashion, a production, in the sense that all consumption, whether of food or other things, contributes toward creating the body and the conditions of subsistence of the subject who produces it; this is “production geared to consumption.” “In the first [productive consumption], the producer transforms himself into things; in the second [consumptive production], things are transformed into human beings” (Marx 2010 [1897]: 277). Although an identity is established between production and consumption, this is only possible thanks to a mediating movement between the two terms: “Production furthers consumption by creating material for the latter, which otherwise would lack its object. But consumption, in its turn, furthers production, by providing for the products the individual for whom they are the products” (Marx 2010 [1897]: 278). However, this extremely original dialectical parity between objectivizing production and subjectivizing consumption fades away a few lines further on, when Marx forthrightly reaffirms the primacy of production over consumption. In effect, consumption is simply a particular moment in production since, once the individual who has produced an object returns it to himself by consuming it, he is acting as a productive individual who is thereby reproducing himself; in consequence, “production forms the actual starting point and is, therefore, the predominating factor” (Marx 2010 [1897]: 282).

Marx’s position is indicative of the more general tendency of modern thought to regard production as the element that determines the material conditions of social life and as the principal way for humans to transform nature and, by doing so, transform themselves. Whether or not one is a Marxist, it is now commonly thought that the history of humanity is primarily founded on the dynamism introduced by a succession of ways of producing use-value and exchange-value of the materials that the environment provides. But it is fair to question whether this pre-eminence ascribed to the process of productive objectivization applies generally to all societies.5 To be sure, humans have always and everywhere been productive; everywhere they have modified or fashioned substances intentionally in order to procure themselves the means of existence, thereby exercising their capacity to behave as agents who impose specific forms and purposes upon matter that is independent of themselves. But does this mean that this kind of action is everywhere apprehended in accordance with the model of a relation to the world known as “production,” a model so paradigmatic and familiar to us that we have become accustomed to use it to describe extremely heterogeneous operations carried out in very diverse contexts?

It seems hardly necessary to recall first that the idea of production by no means suffices to define the general manner in which many hunter-gatherers conceive their subsistence techniques. That is why some specialists of those societies now prefer to use the term “procurement” rather “production,” the better to underline that what we call hunting and gathering are primarily specialized forms of interaction that develop in an environment peopled with intentional entities that are comparable to humans (Bird-David 1992: 40; Ingold 2000: 58-59). But the inadequacy of the notion of production is also obvious when it comes to accounting for the way in which great non-Western civilizations conceptualize the process by which things are engendered. Francois Jullien shows this clearly, in the case of China, in his commentary on the oeuvre of Wang Fuzhi (Jullien 1989). For this seventeenth-century neo-Confucian scholar, who systematizes a fundamental intuition of Chinese thought, the whole of reality can be conceived as a continuous process resulting from the interaction of two principles, neither of which is more fundamental or more original than the other: for example, yin and yang, or Heaven and Earth. From this stems a logic of a mutual relationship with no beginning and no end that excludes any external founding agent, any need for a creator-agent as an initial cause or prime mover and any reference to some transcendent “otherness.” The process of rest alternating with movement that is given dynamism by the primacy of movement, acts in a totally impersonal and unintentional manner; “So order cannot be imposed from outside by a deliberate act of some subject or other implementing a certain plan . . . it is inherent in the nature of things and stems totally from their continuous development” (Jullien 1989: 85). In short, the world is not produced by the intervention of an actor with a plan and a will. It results solely from its own internal propensities (lishi), which manifest themselves spontaneously in a permanent flux of transformations.

This self-regulated process is a far cry indeed from the heroic model of creation as developed in the West and proclaimed as an unquestioned fact on the twofold authority of the biblical tradition and Greek thought. The idea of production as the imposition of form upon inert matter is simply an attenuated expression of the schema of action that rests upon two interdependent premises: the preponderance of an individualized intentional agent as the cause of the coming-to-be of beings and things, and the radical difference between the ontological status of the creator and that of whatever he produces. According to the paradigm of creation-production, the subject is autonomous and his intervention in the world reflects his personal characteristics: whether he is a god, a demiurge or a simple mortal, he produces his oeuvre according to a pre-established plan and with a definite purpose. Hence the abundance of craftsmanship metaphors that are used to express the origin of this type of relationship. In the Psalms, the creator is compared to a well-sinker, a gardener, a potter, and an architect. In the Timaeus, the demiurge creates the world, fashioning it as a potter would. He carefully composes the mixture that he is about to work on; he turns it on his wheel to form a sphere, then he rounds it off and polishes the surface (Plato 1929: 61-65). Here, the image of fabrication, poiesis, is central; and so it remains in the modern conception of the relationship of a producer and that which he produces. What also remains is the idea of the absolute heterogeneity between them: the creator, craftsman or producer possesses his own plan of the thing that he will bring into existence and gives himself the technical means to realize his intended purpose by projecting his will upon the matter that he manipulates. In the same way, just as the creator and his creation are incommensurable in christian dogma, in the Western tradition there is no ontological equivalence between the producer and whatever he brings into being.

Nothing could be more alien to the manner in which the Indians of Amazonia conceive their relations with the entities upon which they feed. For the Achuar, for example, it would be meaningless to speak of “agricultural production” or “hunting production,” as though the aim of those activities were to bring into being a consumable product that would be ontologically dissociated from the material from which it came—even if such operations may come to be quantified and assessed vis-à-vis the potential productivity of resources, as by myself in the past (Descola 1994 [1986], Ch. V-IX). Achuar women do not “produce” the plants that they cultivate: they have a personal relationship with them, speaking to each one so as to touch its soul and thereby to win it over, favour its growth and help it to survive the perils of life, just as a mother helps her children. Achuar men do not “produce” the animals that they hunt. They negotiate with them personally, in a circumspect relationship made up, in equal parts, of cunning and seduction, trying to beguile them with misleading words and false promises. In other words, here it is the relations between subjects (humans and non-humans) that condition the “production” of the means of existence, not the production of objects that conditions (human) relationships.

In Amazonia, even the production of artefacts seems not to fit into the classic model of the demiurge-craftsman. This is what is suggested by Lucia van Velthem’s studies of the wickerwork of the Wayana of the northern Para, who, like some of their Carib and Arawak neighbours, are noted for the diversity and refinement, both technical and aesthetic, of the objects that they plait (Van Velthem 2000, 2001). Wickerwork is a masculine activity that is both valued and prestigious, complete mastery of which is only acquired quite late in life, at least in the creation of the most difficult pieces such as the great katari anon, the carrying basket that is entirely decorated by plaited motifs that differ on each of the external and internal sides. However, the Wayana do not regard the fabrication of baskets as a virtuoso fashioning of a raw material, but rather as an incomplete actualization, in slightly different forms, of the bodies of animal spirits that they reconstitute using plant fibres that are assimilated to human skin. Their baskets, receptacles, trays, mats and the plaited containers in which they press manioc are thus, as Van Velthem puts it, “transformed bodies” (Van Velthem 2001: 206). Each has an anatomy—a head, limbs, breasts, a trunk, ribs, buttocks and genitals—and the motifs that adorn them are stylized representations of the being of which they constitute a transmutation. The designs on the inner sides of baskets even represent that being’s internal organs: the point is, in this way, to evoke the predatory capacity of assimilation of the animals’ spirits which, however, is rendered inoffensive in the artefact by virtue of its incompleteness. For the fibre body differs from the threatening body of the prototype of which it is an actualization, given that it is not recomposed altogether identically, and on that account it lacks the intentionality of the original. However, that is only so in the case of domestic basketry, the daily use of which makes it necessary somehow to “devitalize” it. Objects woven for ceremonial use are said, on the contrary, to be complete materializations of the bodies of animal-spirits. The most expert of the basket-makers are even credited with the ability to recompose in their handiwork the non-visual characteristics of the prototype, such as its movements, sounds and smells. This ontological mimetism allows these objects to function, in their turn, as agents of transformation. They are used extensively in healing rites, since they possess properties identical to those of the entities of which they are reincarnations. Far from being apprehended as the production-creation, out of inanimate material, of a new thing informed by the art and purpose of an autonomous agent, the work of a Wayana basket-weaver is regarded as something that can make a veritable metamorphosis possible, that is to say it can produce a change in the state of an entity that already exists as a subject and that preserves all or part of its attributes throughout this operation.

As a way of conceiving action on the world and a specific relationship in which a subject generates an object, production thus does not have a universal applicability. It presupposes the existence of a clearly individualized agent who projects his interiority on to indeterminate matter in order to give form to it and thus bring into existence an entity for which he alone is responsible and which he can then appropriate for his own use or exchange for other realities of the same type. Now, to return to our two examples: the production model does not correspond either to the concept of a continuous autopoietic process as expressed in Chinese thought, nor to the priority that, in Amazonia, is granted to reciprocal transformation over fabrication ex nihilo. For this reason, anthropologists are perhaps unwise when they succumb to the convenient temptation to use the familiar language of production to interpret the very diverse phenomena by means of which a reality, whether or not of a material nature, comes to be instituted. To speak of the “production” of a person, of social links, of a subject or the difference between the sexes outside the Western context in which, for several millennia, this notion has encompassed an altogether singular relationship, is at best, in most cases, an abuse of language that leads to false parallels.


Protection, too, implies the non-reversible domination of the protector over the one who benefits from that protection. But although it is never reciprocal, the relationship may certainly be reversed in the course of time. The care that parents devote to their children right up to the dawning of adulthood will perhaps be repaid by their children when they grow old. It also sometimes happens that a protector is himself protected by someone more powerful, in particular in relations of patronage that sometimes take the form of a hierarchical chain of dyadic links of clientship. And, finally, frequently protection is mutually profitable in that it guarantees the protector not only the gratification brought by real or supposed gratitude of the person protected, but also the possibility of enjoying help from the latter and also whatever advantages stem from the situation in which he is placed. But even where there is a reciprocal interest, the relationship remains inegalitarian, for it is always founded on the fact that the offer of assistance and security by which it is manifested stems from the initiative of the party who is in a position to make that offer. A child who is a minor is no more able to refuse the protection of its parents than a citizen is able to refuse that of the State or than pandas are in a position to refuse the protection offered by their ecologist defenders.

In relations with non-humans, protection becomes a dominant schema when a group of plants and animals is perceived both as dependent on the humans for its reproduction, nurturing and survival and also as being so closely linked to them that it becomes an accepted and authentic component of the collective. The most complete model of this is probably the extensive kind of herding that pastoral societies practise in Eurasia and Africa. Of course, some of the herded animals are consumed by the humans, either directly or indirectly, but it is seldom this utilitarian function that is foremost in the herdsmen’s idea of their relations with the animals that they tend on a daily basis. They commit themselves above all to take charge of the animals, to help them and watch over them and to offer them care in every domain of life, since the control that wild animals possess over their destiny is here passed over to humans. The latter must therefore see that the animals are fed, if only by choosing the best pastures and water-holes for them. They also have to ensure that the animals reproduce, providing a collective of descendants in the most favourable conditions; and they do this by selecting the reproducers, organizing fertilization and aiding the new-born. Furthermore, they must defend the herd against predators and care for any diseased animals. Although the term “production” is sometimes used to designate this way of making it possible for the animals to live, it hardly seems suitable, since the direct action exerted upon the animals is of an entirely different order from the work of a craftsman or worker fashioning an artefact out of inorganic material. Whatever the degree of standardization achieved by selection, each animal remains different, with a character of its own and its own whims and preferences. So the idea of protection is the one better able to suggest the mixture of constant attention, individualized control and well-meaning forms of constraint that define the relationship between the herdsman and his animals. In fact, in some cases, such as Nilotic societies, those duties take on the appearance of a total subjugation of humans to the mission of satisfying their animal partners. As Evans-Pritchard wrote, “The cow is a parasite on the Nuer, whose lives are spent ensuring its welfare” (Evans-Pritchard 1940: 36).

East Africa is also where the famous “cattle complex” of nomadic herdsmen has been best described. What this expression implies is a super-valuation of the herd, the effect of which is to make all utilitarian aims seem to disappear, producing a situation in which this sole source of wealth comes to play a mediating role in social relations generally: humans are named after the beasts that they control; the possession of those beasts provides both the means of exchange and the principles upon which social aggregation and transmission are based. As Evans-Pritchard, again, observes in connection with the Nuer, “They tend to define all social processes and relationships in terms of cattle” (Evans-Pritchard 1940: 19). Godfrey Lienhardt asserts the same of the Dinka, adding that these neighbours of the Nuer “conceive their own lives and the lives of cattle on the same model” (Lienhardt 1961: 16). Such interdependence between the domesticated animals and the human society cannot be reduced to the classic interpretation of fetishism, which regards relations with non-humans as the basis of inter-human relations. Here, the interdependence indicates that the animals are indeed full members of the collective, and so are not just a socialized segment of nature serving as a metaphor or idiom for relations between humans that are external to it. Lienhardt furthermore emphasizes that the Dinka do not anthropomorphize their animals but, on the contrary, seek at every level to imitate the characteristics and behaviour of their cattle, which is why these constitute the best possible substitutes for humans. In other words, the relational schema here seems to be twofold: the humans’ protective attitude toward the livestock is combined with relations of a different nature between the humans themselves; and these, paradoxically, are copied from those that structure the world of the cattle. The organization of the herd, the competition between bulls and the relations between the male and female animals serve as models for thinking about political and spatial organization, about the bellicose nature of men and about the relations between the sexes. This is why, despite the exorbitant role played by cattle-raising among the herders of East Africa, protection does not play the role of a general principle of action that structures all the interactions between humans and non-humans, however fully the latter are integrated within the collective.

To find clearer illustrations of what Haudricourt calls “the pastoral treatment” of humans, we need to turn to the ancient Mediterranean civilizations: the Roman world, for example, where under the imperious but protective “crook” of the paterfamilias, a little cohort of dependent beings would develop. Within the order and relative safety of the living conditions guaranteed by their master’s authority, women, children, slaves and flocks all found the means to contribute to the common prosperity, by fulfilling their respective roles. Virgil, better than anyone, sketched in the ideal picture of this agricultural Arcadia, in which a diligent labourer, through wise management of his dependents, “provides sustenance for his country and his little grandson and . . . for his herds of kine and faithful bullocks” (Georgics, Book II: 514-515). Two of the four books that make up the Georgics are devoted to care for the animals, the measures to take in order to protect them against danger, the services that they render and the benefits that they provide. Thanks to their virtues and their sense of duty, the fat oxen, the powerful bulls, the mettlesome chargers and the industrious bees all behave like responsible citizens under the enlightened supervision of their owner, just as he himself flourishes on his land in Campania under the aegis of Augustus, the defender of the power and prosperity of the State. Under Virgil’s pen, submission is sweet, and even slavery is almost tolerable, given that all that is required is a little obedience and a lot of hard work in return for the security that one’s master offers!

The mutual benefits that protection is believed to procure are often part of a long chain of dependence that link several ontological levels by a series of duplications of asymmetrical relations. Just as humans take care of the animals and plants from which they derive their subsistence, they may themselves be protected by another group of non-humans: the deities, who derive from their patronage the most substantial of advantages, namely their own raison d’être. These deities, who are in some cases hypostases of a plant or an animal particularly important to the local economy, are thus seen as founding ancestors and guarantors of the humans’ well-being. At the same time, they are regarded as the condition (or even the direct creators) of the effective domination over the non-humans that the humans use and protect. This is clearly illustrated by the example of the Exirit-Bulagat, a group of Buryat herders in Cisbaikalia (Hamayon 1990, Ch. XII). For these people, herding, although adopted only relatively recently, has become their dominant activity and the motor of their social and ceremonial life; so much so, indeed, that they claim to have been engendered by a heavenly bovine, Buxa nojon, or Lord Bull, the tribe’s principal deity upon which minor ancestral figures depend. Several times each year, mares are sacrificed to the Lord Bull and to the ancestors to ensure good luck and wealth and to persuade them to protect the herds and ensure their growth. The victims are sacrificed to the deity in order that it will then fill their meat with grace and so too the humans when they consume it. Furthermore, every herdsman has in his herd a “consecrated” bull that must never be maltreated or mounted or sold or castrated. This is the embodied emissary of the Lord Bull and of his inexhaustible virility and it guarantees and promotes the fecundity of the domesticated animals. The care that the humans lavish upon this bull represents, as it were, a service that they render to their bull-ancestor to thank it for its fertilizing power, while the souls of the sacrificed mares that are offered up to this deity, as a substitute for human souls, constitute a propitiatory gift motivated by the hope that the deity will concede its benefits to the humans and its protection to their herds. By distinguishing between these oblations—on the one hand a restitution of life through care for the sacred animal and, on the other, a restitution of souls through the death of the mares—the humans avoid having to repay the bull the debt that they have incurred toward its hypostasis. In this way, protection can become the all-encompassing value of a system of interactions that combines two asymmetrical relationships: a relation of predation in which one takes the lives of non-human dependents without allowing them any direct recompense, and a gift-giving relation in which one offers protected non-humans to non-human protectors so as to encourage the latter to perpetuate the domination that they allow the humans to exercise over the former (For a fuller analysis of the case of the Exirit-Bulagat, see Ch. XV).


As I understand it, transmission is above all what allows the dead, through filiation, to gain a hold over the living. We may owe many things to those who have preceded us: material goods and land received as bequests, prerogatives that are inherited—responsibilities, hereditary statuses and functions, and symbolic attributes such as a name or the possession of certain kinds of knowledge; and also physical, mental or behavioural characteristics reputed to be inherited. The extent of this material and immaterial patrimony through which we are indebted to previous generations varies enormously from one civilization and social situation to another, but it largely depends not only on the quantity of transmitted items, but also and above all on the importance ascribed to the very phenomenon of transmission, understood as an accepted dependence upon more or less distant ancestors. In every collective, things pass down from one generation to another in accordance with precise and recognized norms. However, it is only in certain circumstances that this ceding process acquires the form of a veritable debt owed by the living to the dead, the former considering themselves to be debtors of the latter with regard to more or less everything that conditions their existence. This includes the order and values according to which they live, the means of subsistence placed at their disposition, the differential advantages that they may enjoy and even their very persons, in as much as a person is formed by principles, substances and in some cases a destiny that stems from one’s direct parents and those who, in the past, engendered them. In order to transmit such things, real and fictitious genealogies are certainly needed, genealogies that go back quite a long way and explicit indications stipulating what each individual has the right to receive by way of the identity, privileges and obligations that are transmitted through these channels stretching down from the past. It seems reasonable to assume, however, that initial conditions are less important than the institutional consequences that the preponderance of some specific schema of relations with others may gradually have acquired. The depth of genealogies, the rules unequivocally confirming a maternal or paternal filiation, the segmentation into different descent groups that act in the manner of moral persons, and the legitimation of rights stemming from particular ancestral groups: all these are mechanisms that anthropologists for a long time failed to recognize to be by no means universal. Now however, it seems that they should be regarded as the means that certain collectives employ in order dogmatically to perpetuate the sovereignty that the dead exercise over the living through relations of transmission.

The clearest expression of this relationship can be found where the dead are converted into ancestors to whom a cult is devoted. These are close ancestors, not the distant and more or less mythical figures also conventionally called “ancestors,” who are sometimes placed at the origin of clans or tribes but are nevertheless not accorded any direct influence over the destiny of those they have created. Immediate ancestors are individualized, named, often given material form on domestic or lineage altars, and nothing that concerns the living eludes their meddlesome jurisdiction. West Africa is one of the places that such ancestors favour most, as can be seen from the example of the Tallensi of Ghana.

The patrilineages of the Tallensi, which are localized, set in segmented hierarchies and enjoy a relative political independence, trace their identity and solidarity back to the cult that each devotes to a group of agnatic masculine ancestors, which goes back twelve generations at the most. In his Oedipus andJob in West African religion, Fortes describes the despotic domination that the ancestors exert over the living, a domination that is analogous in its content to the absolute authority of a father over his sons. Just as a man has no economic rights, legal status or ritual autonomy as long as his father is living, similarly, the members of a lineage depend upon the ancestors for access to land, the exercise of political responsibilities and the well-being of each one of them (Fortes 1959). A son does not, in any case, succeed his father in his rights and privileges until such time as he has executed the funerary rites that turn his father into an ancestor, thereby transforming his own subordination to a living individual into an authority delegated to him by earlier generations. The power of the ancestors manifests itself at two levels. Collectively they require that the living should conform to the moral precepts and respect the values upon which their socio-political organization is founded. Every death is thus interpreted as a sanction organized by the ancestors on account of a misdeed that an individual or even his father or an agnatic relative may have committed, in many cases inadvertently. But every human being is also flanked by a specific ancestor deemed to watch over him/her provided he/she submits to its will, as this is revealed by a diviner. Hence the importance of the ancestor cults, which take the form of sacrifices, prayers, and libations. As Fortes explains, “their solicitude is gained not by demonstrations of love, but by proofs of loyalty” (Fortes 1959: 50). The prerogatives that may be enjoyed within the framework of a lineage, the possessions at one’s disposal, and the quota of happiness or misfortune allocated to each individual are all fixed by the ancestors who extend over the living a cloak of justice as impossible to question and as terrible and inaccessible to human understanding as that which the suffering Job eventually credited to his God.

Among the Tallensi, as in other West African civilizations, the cult addressed to the ancestors is thus not so much a way of honouring them and thanking them for all that they transmit; rather, it is an attempt to conciliate them and dispel their anger, an attempt that one can never be sure will be crowned with success. The movement between the generations is strictly one-way, for what one’s forebears have given can never be returned to them, starting, for all of us, with the lives we have received from our parents. Nor, in the present case, is it a matter of the ancestors acting with liberality, since they have no choice but to transmit, in their turn, whatever they themselves received, and by doing so they commit their descendants to a spiral of dependence from which they can never free themselves. The debt that the living inherit is thus passed on inexorably from one generation to the next, as the indebted members of the collective join the mass of the dead and so become creditors; they can now make their descendants pay, just as they had to, for the right to existence and all that makes this possible, in return for unswerving obedience to the power that they hold. For the ancestors, this constitutes a precious guarantee of survival of a sort. Like Aeneas, fleeing from Troy, every man carries his father on his back, but also his father’s father perched on his father’s back and so on, in a by no means metaphorical pyramid, the weight of which crushes the freedom of movement of the living. In these collectives, the burden, both vital and deadly of the ancestors, is perpetuated by a filiation that cannot be rejected and it is fair to apply to them what Pierre Legendre says of the subject of transmission in general: “The genealogical institution functions against the background of the subject’s distress” (Legendre 1985: 80). However, in Africa, that distress does not encompass a tragic dimension such as that which pervades the particular destiny of an individual faced by capricious gods or the hermetic purposes of the Christian god, since all concerned, including the ancestors, share the fate of dependency upon earlier generations, for better or for worse. So, despite the reference to Oedipus in the title of Fortes’ book, it is unlikely that a Tallensi would lament using the words that Sophocles puts into the mouths of the chorus surrounding Laius’ son: “Not to be born comes first by every reckoning” (Sophocles 1994: 547).

Transmission is not only a relationship within the segments of a collective that links in a chain of dependency on the one hand living people destined to become ancestors and, on the other, the dead who live on and whose power and will are felt in all circumstances. It is also what distinguishes one collective, with all its elements, from another. For some collectives claim as the principal source of their contrastive identity the fact that they have their own particular groups of ancestors from whom stem both their legitimacy as an autonomous social body and also all the attributes attached to it. The latter range from the right to live in a particular space and exploit its resources—echoes of which are to be found in the notion of a fatherland—to a consciousness of sharing certain hereditary physical and moral properties. This use of transmission in the definition of collectives and their properties is altogether specific to certain regions of the world and should not be confused with the universal phenomenon of ceding certain material and immaterial assets from one generation to the next. In Indian Amazonia, for example, nowhere does one find the kind of hold exerted by the ancestors that exists in Africa and in China. In Amazonia, the very idea of an ancestor seems incongruous. The recently dead are supposed to disappear as soon as possible from the memory of the living and if anything of them does remain for a while, it is in the form of more or less malicious spirits whose company is to be shunned. Moreover, genealogies seldom go back further than the grandparents’ generation and descent groups, in the rare cases where these exist, control neither access to the means of subsistence nor the devolutions of the latter; and they may anyway concern only a fraction of the population, as is the case of the Sanumâ of Brazil. No cult is addressed to the dead and if there is anything to be inherited from them, it will be, not so much a meagre physical patrimony (their objects are usually destroyed) but, rather, symbolic attributes: names, songs, myths, the right to make certain garments, or to wear certain ornaments.6 In short, the dead are excluded from human collectives and have no power over them.

This is probably a feature of the animist regime in general. So when Ingold criticizes the use of the model of a genealogical tree and the primacy of ancestrality as a means of explaining the relations of indigenous peoples to one another and to the space that they occupy, he finds most of his examples in the kind of collectives that I call animist: the Chewong of Malaysia, the Nayaka of Tamil Nadu, the Ojibwa, the Cree and the Yup’ik of North America, and so on. According to Ingold, the image of a rhizome, which he borrows from Gilles Deleuze and Félix Guattari, is far more appropriate for characterizing the reticular relations that these people, who are indifferent to unilinear filiation, maintain with the various components of their environment (Ingold 2000: 140-146). Ingold’s point is fair enough provided one does not go to the other extreme and declare that all reference to transmission and ancestrality should be banned if the aim is accurately to restore the idea that “indigenous peoples” have of themselves (Ingold 2000: 140).7 For there is no reason to exclude the Tallensi or the Malgaches from the benefit of autochthony and a distinctive identity, or to believe that they have succumbed to the Western perversity of the genealogical principle and a cult of lares. By choosing to ascribe considerable importance to their dead and all that these transmit and control, some collectives have made the matrix relationship to the ancestors the main lynch-pin of the precepts and values that organize their common lives. Others have preferred to ignore that dimension of human life and instead to base their individual and collective identity on a dense and shifting network of multi-polar relations with a mass of entities both contemporary and of the same status as themselves. Just because anthropology has for a long time tended, when interpreting the practices of animist collectives, to adopt as a standard the institutions of the ancestor-worshippers, there is no reason why the contribution made by the latter type of collective to the diversity of world-states should be considered suspect or allowed to fall into oblivion.

The relational modes that we have just considered fall into two groups: the first covers potentially reversible relations between substitutable terms, since the latter are situated at the same ontological level (exchange, predation and gift); the second covers one-way and irreversible relations between non-substitutable terms, since these are intrinsically hierarchical (production, protection and transmission). The former are characteristic of the symmetrical or asymmetrical movement of something of value between subjects of equal status whose identity or essence is not transformed by the actualization of the relationship that links them. Meanwhile, the latter imply a connection of a genetic, spatial or temporal order between the agents and the objects of an action by means of which the disparity of their respective positions is either created or maintained.

Figure 10
Figure 10: The distribution of relationships depending on the type of relations that exist between the terms involved

A place in this inventory could no doubt be found for many other relational modes. Most, though, can be included either as the complement to one of the relationships that we have considered here or at least to one of its dimensions. There is no protection without dependence, no liberality without gratitude, no exchange without obligation. As for domination and exploitation, the absence of which might well be criticized in view of the role that they have played in history, those can be fitted into other relationships, as one of their components: domination is inherent to protection and transmission, exploitation manifests itself in the relations of force that are established at the time of dictating the conditions of production or exchange. Moreover, unlike the relations that we have picked out, only rarely are exploitation and domination seen as what they are by those whom they concern. More often, they affect the appearance of a relationship involving an exchange of services that to some extent masks their fundamental inequality: payment in exchange for work, protection against forced labour, prosperity in return for subservience. But that is really beside the point. For, I repeat, my intention is not to examine all the relations that occur between existing beings and that are given an institutional form. Rather, it is simply to mark out a few major schemas of action that structure the lives of collectives, in order to examine how compatible or incompatible they may be with the modes of identification picked out earlier. So this typology makes no claim other than to group together a few of the elementary structures that make up the great variability of ways of intervening in the world; and that variability is so rich that it would not be possible to propose any more than a rough syntactical sketch of them.

Although relational schemas are based on specific cognitive mechanisms, such as schematic induction, analogical transposition from one domain to another, or the influence of affects upon memorization, they are not categorical imperatives written into the architecture of the human mind. Rather, they should be considered as objectivised properties of all collective life. They are properties that are embodied in mental, affective and sensory-motor dispositions by means of which behaviour patterns stabilize in distinctive forms of interaction. Giving something or oneself to another, taking from another, receiving from another, exchanging with another, but also appropriating another, protecting him, producing him or placing oneself in his dependence are all actions inherent to the phylogenetic evolution of social primates. They are actions that all humans perform both within the family unit and also in wider contexts. They provide a register of combinations upon which all collectives draw, selecting (we do not really know why) one field of relations rather than another, in order to orientate their public behaviour. But none of these practical schemas, on its own, dictates the ethos of a collective. Rather, each schema constitutes an indeterminate ethical landscape, a style of mores that one learns to cherish and by which one differentiates oneself from one’s neighbours: a style of mores that colours one’s daily attachments to beings and things, with underlying nuances. However, this does not rule out other types of relations to others, ones that individual idiosyncrasy, the unpredictability of feelings, and the arbitrariness of conventions all make it possible to express more discretely in less stereotyped situations.


Aristotle. 1926. Nicomachean ethics. Translated by H. Rackham. Cambridge, Massachusetts: Harvard University Press.

Bird-David, Nurit. 1990. “The giving environment: Another perspective on the economic system of hunter-gatherers.” Current Anthropology, 31: 189–196.

———. 1992. “Beyond the hunting and gathering mode of subsistence: culture-sensitive observations on the Nayaka and other modern hunter-gatherers.” Man (N. S.) 27: 19–44.

Descola, Philippe (1986) 1994. In the society of nature: A native ecology in Amazonia. Translated by Nora Scott. Cambridge: Cambridge University Press.

———. 1990. “Cosmologies du chasseur amazonien.” In Pour Jean Malaurie: 102 témoignages en hommage à quarante ans d’études arctiques, edited by Sylvie Devers, 59–64. Paris: Plon.

———. 1992. “Societies of nature and the nature of society.” In Conceptualizing society, edited by Adam Kuper, 107–126. London and New York: Routledge.

———. 1993. “Les affinités sélectives: alliance, guerre et prédation dans l’ensemble jivaro.” L’Homme 126–128: 171–190.

Evans-Pritchard, Edward E. 1940. The Nuer: A description of the modes of livelihood and political institutions of a Nilotic people. Oxford: Clarendon Press.

Fausto, Carlos. 2001. Inimigos fiéis. Historia, guerra, xamanismo na Amazônia. Säo Paolo: Editora da Universidade de Säo Paolo.

Fortes, Meyer. 1959. Oedipus and Job in West African religion. Cambridge: Cambridge University Press.

Jullien, François. 1989. Procès ou création. Une introduction à la pensée des lettrés chinois. Paris: Editions du Seuil.

Godelier, Maurice (1996) 1999. The enigma of the gift. Translated by Nora Scott. Cambridge: Polity.

Gregory, Chris A. 1982. Gifts and commodities. London: Academic Press.

Hamayon, Roberte. 1990. La Chasse à l’âme. Esquisse d’une théorie du chamanisme sibérien. Nanterre: Société d’ethnologie.

Ingold, Tim. 2000. The perception of the environment: Essays in livelihood, dwelling and skill. London and New York: Routledge.

Legendre, Pierre. 1985. L’Inestimable objet de la transmission. Etude sur le principe généalogique en Occident (Leçon IV). Paris: Fayard.

Lienhardt, Godfrey. 1961. Divinity and experience: The religion of the Dinka. Oxford: Clarendon Press.

Overing, Joanna and Alan Passes, eds. 2000. The anthropology of love and anger: The aesthetics of conviviality in Native Amazonia. London and New York: Routledge.

Marx, Karl. (1897) 2010. A contribution to the critique of political economy. Translated by N. J. Stone. Charleston, South Carolina: Forgotten Books.

Mauss, Marcel. 1950. Sociologie et anthropologie. Précédé d’une Introduction à l’oeuvre de Marcel Mauss par Claude Lévi-Strauss. Paris: P.U.F.

Plato. 1929. Timaeus. Translated by R. G. Bury. Cambridge: Harvard University Press.

Sahlins, Marshall. 1968. Tribesmen. Englewood Clifs: Prentice-Hall.

Seeger, Anthony, Roberto Da Matta and Eduardo Viveiros de Castro. 1979. “A construçào da pessoa nas sociedades indigenas brasileira.” Boletim do Museu nacional 32: 2–19.

Seneca. 1935. De beneficiis. Translated by John W. Basore. Cambridge, Massachusetts: Harvard University Press.

Service, Elman. 1966. The hunters. Englewood Cliffs: Prentice-Hall.

Sophocles. 1994. Oedipus at Colonus. Translated by Hugh Lloyd-Jones. Cambridge, Massachusetts: Harvard University Press.

Surrallés, Alexandre. 2003. Au coeur du sens. Perception, affectivité, action chez les Candoshi. Paris: Editions du C.N.R.S. Fondation de la M.S.H.

Taylor, Anne Christine. 1993. “Les bons ennemis et les mauvais parents. Le traitement de l’alliance dans les rituels de chasse aux têtes des Shuar (Jivaro) de l’Equateur.” In Les Complexités de l’alliance, edited by Elisabeth Copet-Rougier and Françoise Héritier-Augé, 73–105. Paris: Editions des Archives contemporaines.

———. 1996. “The soul’s body and its states: An Amazonian perspective on the nature of being human.” Journal of the Royal Anthropological Institute 2 (2): 201–215.

Testart, Alain. 1997. “Les trois modes de transfert.” Gradhiva 21: 39–49.

Velthem, Lucia van. 2000. “Fazer, fazeres e o mais belo feito.” In Os Indios, Nós, edited by J. P. de Brito, 174–180. Museu Nacional de Etnologia

Velthem, Lucia van. 2001. “The woven universe: Carib basketry.” In Unknown Amazon: Culture in nature in ancient Brazil, edited by Colin McEwan, Cristina Barreto and Eduardo Neves, 198–213. London: The British Museum Press.

Vidal, Denis. 1991. “Les trois Grâces ou l’allégorie du Don: Contribution à l’histoire d’une idée en anthropologie.” Gradhiva 9: 30–48.

Viveiros de Castro, Eduardo. 1993. “Alguns aspectos da afinidade no dravidianato Amazônico.” In Amazônia: etnologia e historia indigena, edited by Manuela Carneiro da Cunha and Eduardo Viveiros de Castro, 149–210. Säo Paolo: NHII-Universidade de Säo Paolo.

Philippe Descola holds the chair of anthropology at the Collège de France where he heads the Laboratoire d’Anthropologie sociale and is also a professor at the École des Hautes Études en Sciences Sociales, Paris. Besides publishing over a hundred articles and book chapters, he is the author of, among others, In the Society of Nature (Cambridge, 1994), The Spears of Twilight (New York, 1996), Par-delà nature et culture (Paris, 2005), La Fabrique des images (Paris, 2010), Diversité des natures, diversité des cultures (Paris, 2010), and L’écologie des autres. L’anthropologie et la question de la nature (Paris, 2011). He is a fellow of the British Academy and a member of the American Academy of Arts and Science.


1. See Aristotle, Nicomachean ethics, V.5.6: “But in the interchange of services Justice, in the form of Reciprocity, is the bond that maintains the association” (Aristotle 1926: 281). Seneca, De beneficiis I.IV: “What we need is a discussion of benefits and the rules for a practice that constitutes the chief bond of human society” (Seneca 1935: 19).

2. On this point, see, Année sociologique, republished in Mauss 1950: 145-279.

3. Maurice Godelier addresses another criticism to Mauss (and to Lévi-Strauss). He claims that they did not draw the consequences from the fact that a thing given is not alienated by reason of having been given, since the giver continues to be present in that thing and through it exercises pressure on the receiver, not to give it back but, himself, to give it away. According to Godelier, such an enigma only becomes comprehensible if the things that one gives are defined by those that one does not give, chief among the latter being sacred objects that represent collective identities and their temporal continuity; it is because such objects exist that exchange is possible: “The formula for social behaviour is thus . . . to keep in order [to be able] to give and to give in order [to be able] to keep” (Godelier 1999 [1996]: 36, modified).

4. See, for example, Overing & Passes (2000), in particular the Preface and the Introduction.

5. As Chris Gregory saw clearly in his study of the circulation of goods in Melanesia, see for example (Gregory 1982: 31-32).

6. Although published over twenty years ago, the critique of the application of the African model of transmission to the Amerindian context proposed by Seeger, Da Matta & Viveiros de Castro (1979) remains totally up to date.

7. “I believe that a relational model, with the rhizome rather than the tree as its core image, better conveys the sense that so-called indigenous people have of themselves and of their place in the world” (Ingold 2000: 140).