A Onça e a Diferença

Intensive filiation and demonic alliance[]

Eduardo Viveiros de Castro

Museu Nacional, Rio de Janeiro

(inglês provisório; a ser publicado em K. Rodje & C.B. Jensen, Deleuze and Anhtropology)

Life is robbery

A.N. Whitehead (1978: 105)(1)

Deleuze and Anthropology

For my generation, the name of Gilles Deleuze immediately suggests the change in thought that marked the years circa 1968, during which some key elements of our present cultural apperception were invented.(2) The meaning, the consequences and the very reality of this change have given rise to a controversy that still rages. Just like the post-modernity of which it is one of the (anti)foundational dates, ‘1968’ seems never to end. For some, it does not stop to have never occurred; for others, it does not cease to have not begun yet. I place myself among the latter; and thus, I would say the same about the influence of Deleuze and his long-time associate Félix Guattari, the authors of the most radically consistent (conceptually) and most consistently radical (politically) oeuvre created in philosophy during the second half of the last century. Its presence in certain contemporary disciplines or investigation fields is indeed much less evident than one might expect, being felt mostly (when at all) through its effects on the general cultural matrix. This seems to be the case, for example, of social anthropology.

From an anthropological point of view, the novelty which was (and remains) Deleuze’s philosophy was immediately perceived by the countercultural and counterpolitical movements that proliferated in the West during the last three decades, most notably feminism (some versions of it, at least), as well as some currents in experimental art (art de vivre included). Not much later, it was incorporated in the toolkit of certain auto-anthropological projects such as science and technology studies or the sundry disquisitions on the dynamics of late capitalism. On the other hand, the attempts at a more direct engagement between anthropology at large, or allo-anthropology,(3) and Deleuzian concepts have been surprisingly rare. I deem this dearth surprising for two reasons. Firstly, because Anti-Œdipus and A Thousand Plateaus, the two volumes of Capitalism and Schizophrenia, support many of their arguments by a vast bibliography on non-western peoples — from the Guayaki to the Nuer to the Mongolians — developing thereby some hypotheses of great analytical potential for anthropology. Secondly, because the work carried out by some of the most innovative anthropologists in the two last decades — I am thinking, for example, of Roy Wagner, Marilyn Strathern and Bruno Latour — shows evident connections with Deleuze’s ideas. Such connections may be purely virtual (Wagner), certainly partial (Strathern) or explicitly actual (Latour); but, in these three cases, equally real, and deserving of more attention.

It is certainly no accident that the three above-mentioned anthropologists are among the few who could be called post-structuralists (instead of, say, post-modernists) with some propriety. They have managed creatively to take on board the insights of structuralism and move ahead, instead of, like so many of their colleagues, bending backwards and embrace conservative theoretical projects, such as the sentimental pseudo-immanentism of lived worlds and embodied practices, or the macho-positivist truculence of TOEs such as evolutionary psychology or so-called political economy. By the same token, Deleuze’s thought can be seen as a form of extreme deterritorialisation of structuralism, a movement or a style from which he extracted — some would say, into which he introduced — its most radically original insights (Deleuze 1972) so as to pursue, with their help, in other, sometimes very different, directions.(4) After all, what is a multiplicity, if not a structure at long last freed from all complicity with transcendence?

An Anti-Sociology of Multiplicities

In Anti-Œdipus, as is well known, Deleuze and Guattari overthrow the temple of psychoanalysis by imploding its central pillar — the reactionary conception of desire as lack — and replace it with a theory of desiring machines as sheer positive productivity that must be coded by the socius, the social production machine. This theory runs into a vast panorama of universal history, painted in the book’s central chapter in a quaintly archaic style, which may make the anthropological reader wince. It not only employs the venerable savagery-barbarism-civilization triad, as all ethnographic references are treated in a seemingly cavalier way that the same reader might be tempted to call ‘uncontrolled comparison’. Yet if that reader stops to think for a moment, it is probable that she reaches the conclusion that the traditional three-stage topos is submitted there to a far from traditional interpretation, and that the impression of comparative erraticalness derives from the fact that the controls used by the authors are other than the usual ones — of a differentiating rather than collectivizing type, as Wagner (1981) would put it. Anti-Œdipus is indeed the result of a ‘prodigious effort to think differently’ (Donzelot 1977: 28); its purpose is not merely to denounce the repressive paralogisms of psychoanalysis but to establish a true ‘anti-sociology’ (id.: 37).(5) An obviational project like this should certainly appeal to contemporary anthropology. Or at least to that anthropology bent to explore the vast territory that lies beyond the jurisdiction of the three infernal dichotomies that contain the discipline within an iron ring: nature and culture, individual and society, traditional and modern.(6)

A Thousand Plateaus distances itself from AŒ’s psychoanalytical concerns. The project to write a ‘universal history of contingency’ (D. & G. 2003: 290) is carried on in a decidedly non-linear fashion through the crossing of different intensity ‘plateaus’ (a notion inspired in Bateson) corresponding to diverse material-semiotic formations and peopled by a disconcerting quantity of new concepts. The book puts forward and illustrates a theory of multiplicities, the Deleuzian theme that has perhaps achieved greatest repercussion in contemporary anthropology.

Multiplicity is the meta-concept that defines a new type of entity; the well-known (by name at least) ‘rhizome’ is its concrete image.(7) The sources of the Deleuzian idea of multiplicity lie in Riemann’s geometry and Bergson’s philosophy (Deleuze 1966: ch. 2), and its creation aims at dethroning the classical metaphysical notions of essence and type (DeLanda 2002).(8) It is the main tool of a ‘prodigious effort’ to imagine thought as an activity other than that of identifying (recognition) and classifying (categorisation), and to determine what is there for thought to think as intensive difference rather than as extensive substance. The political-philosophical intentions of this decision are clear: it is a matter of severing the link between the Concept and the State. Multiplicities are the tools of a fractal, anexact counter-geometry, a minor mathematics that aims at ‘inverting Platonism’. Thinking through multiplicities is thinking against the State.(9)

The notions of type and entity are, in fact, inadequate to define rhizomatic multiplicities. If there is ‘no entity without identity,’ as Quine famously alliterated, one must conclude then that multiplicities do not qualify for that enviable status. A rhizome does not behave as an entity, nor does it instantiate a type; it is an acentric reticular system of n-1 dimensions, constituted by intensive relations (‘becomings’) between heterogeneous singularities that correspond to events, or extra-substantive individuations (‘hecceities’). Hence, a rhizomatic multiplicity is not truly one being but an assemblage of becomings: a ‘difference engine’, or rather, the intensive diagram of its functioning. Bruno Latour, who in his recent book on Actor-Network theory indicates how much ANT owes to the rhizome concept, is particularly emphatic: a network is not a thing because any thing can be described as a network (2005: 129-31). A network is a perspective, a way of inscribing and describing, ‘the registered movement of a thing as it associates with many other elements’ (Jensen 2003: 227). Yet this perspective is internal or immanent; the different associations of the ‘thing’ make it differ from itself — ‘it is the thing itself that has been allowed to be deployed as multiple’ (Latour 2005: 116). In short, there are no viewpoints on things — it is the things and the beings that are the viewpoints (D. 1968: 79; 1969: 203). If no entity without identity, then no multiplicity without perspective.

A rhizome is not truly one being, either. Multiplicities are constituted by the absence of any extrinsic coordination imposed by a supplementary dimension (n+1: n and its ‘context’, for example); congenitally devoid of unity, they differ constantly from themselves. They evince an immanent organization ‘belonging to the many as such, and which has no need whatsoever of unity in order to form a system’ (D. 1968: 236).(10) Multiplicities are, in short, tautegorically anterior to their own ‘contexts’; like Roy Wagner’s (1986) symbols that stand for themselves, they possess their own internal measure. This turns them into systems whose complexity is ‘lateral’, that is, resistant to hierarchy or to any other type of transcendent unification — a complexity of alliance rather than descent, to anticipate an argument. Emerging when and where open intensive lines (lines of force, not lines of contour; cf. ATP: 549) connect heterogeneous elements, rhizomes project a radically flat ontology (DeLanda 2002), which ignores the distinctions between ‘part’ and ‘whole’.(11) A baroque instead of a romantic conception of complexity, as Kwa (2002) persuasively argued.

Wagner’s fractal person, Strathern’s partial connections, Callon and Latour’s socio-technical networks are some well-known anthropological examples of flat multiplicities. ‘A fractal person is never a unit standing in relation to an aggregate, or an aggregate standing in relation to a unit, but always an entity with relationship integrally implied’ (Wagner 1991: 163, my emphasis).(12) The mutual implication of the concepts of multiplicity, intensity and implication is in fact a point elaborated at length by Deleuze (1968: ch. VI). François Zourabichvili, the most perceptive commentator of this philosopher, observes that ‘implication is the fundamental logical movement in Deleuze’s philosophy’ (2004a: 82); elsewhere, he underlines that Deleuzian pluralism supposes a ‘primacy of relations’.(13)

Not ‘every’ relation will do, though. Multiplicity is a system defined by a modality of relational synthesis different from a connection or conjunction of terms. Deleuze calls it disjunctive synthesis or inclusive disjunction, a relational mode that does not have similarity or identity as its cause (formal or final), but divergence or distance; another name for this relational mode is ‘becoming’. Disjunctive synthesis or becoming is ‘the main operator of Deleuze’s philosophy’ (Zourabichvili 2003: 81), as it is the movement of difference as such — the centrifugal movement through which difference escapes the powerful circular attractor of dialectical contradiction and sublation. A difference that is positive rather than oppositive, an indiscernibility of the heterogeneous rather than a conciliation of contraries, disjunctive synthesis takes disjunction as ‘the very nature of relation’ (id. 2004: 99) and relation as a movement of ‘reciprocal asymmetric implication’ (id. 2003: 79) between the terms or perspectives connected by the synthesis, which is not resolved either into equivalence or into a superior identity:

Deleuze’s most profound insight is perhaps this: that difference is also communication and contagion between heterogeneities; in other words, that a divergence never arises without reciprocal contamination of points of view […] To connect is always to communicate across a distance, through the very heterogeneity of the terms (Zourabichvili 2004a: 99).

Coming back to the parallels with contemporary anthropological theory, it is worth recalling that the theme of separation-as-relation is emblematic of Strathernian anthropology. The conception of relations as ‘comprising disjunction and connection together’ (Strathern 1995: 165; my emphasis) is the basis of the theory of differential relations, the idea that ‘[r]elations make a difference between persons’ (id. 1999: 126; cf. also 1996: 525, and naturally, 1988: ch. 8). To cut a long argument short, let us say that the celebrated ‘system M’ (Gell 1999), the description of Melanesian sociality both as an exchange of perspectives and a process of relational implication-explication, is the prototypical allo-anthropological theory of disjunctive synthesis. From the auto-anthropological standpoint, in turn, it is possible to observe that the subtractive rather than additive multiplicity of rhizomes turns the latter into non-merological, post-plural ‘objects’ (Strathern 1992a), capable of tracing a line of flight from the dilemma of the one and the many that Strathern, with her remarkable sureness of judgment, identifies as anthropology’s characteristic analytical trap:

[A]nthropologists by and large have been encouraged to think [that] the alternative to one is many. Consequently, we either deal with ones, namely single societies or attributes, or else with a multiplicity of ones […] A world obsessed with ones and the multiplications and divisions of ones creates problems for the conceptualization of relationships (Strathern 1991a: 52-53).

To compare multiplicities is different from making particularities converge around generalities, as in the usual case of anthropological comparisons that seek out the substantial similarities underlying accidental differences.(14) Yet it is also different from establishing correlational invariants through formal analogies between extensive differences (oppositions), as in the case of structuralist comparisons, in which ‘it is not the resemblances, but the differences that resemble one another’ (Lévi-Strauss 1962a: 111). To compare multiplicities — which are comparing devices in their own right — is to determine their characteristic ways of diverging, their distances both internal and external; here, every comparative analysis is necessarily a separative synthesis. When it comes to multiplicities, it is less a case of (extensive) relations that vary than of (intensive) variations that relate: it is the differences that differ.(15) As the strange molecular sociologist Gabriel Tarde wrote more than a century ago: ‘The truth is that differences go differing, and changes go changing, and that, as they take themselves thus as their own finality, change and difference bear out their necessary and absolute character’ (1999 [1895]: 69).(16)

Intensive difference, difference of perspectives, difference of differences. Nietzsche observed that health’s viewpoint on illness differs from illness’ viewpoint on health.(17) It was perhaps this observation that inspired Roy Wagner to say about his early relations with the Daribi: ‘their misunderstanding of me was not the same as my misunderstanding of them’ (1981: 20) — my candidate for the best anthropological definition of ‘culture’ ever proposed. Since the difference is never the same, the way is not the same in both directions.(18) The comparison of multiplicities — comparison as production of multiplicity (or ‘invention of culture’) — is always a disjunctive synthesis, like the relations it relates.

Partial Dualities

Deleuzian texts revel in conceptual dyads: difference and repetition; intensive and extensive; nomadic and sedentary; virtual and actual; flows and quanta; code and axiomatic; deterritorialization and reterritorialization; minor and major; molecular and molar; supple and rigid; smooth and striated — the list is long and colourful. Owing to this stylistic ‘signature’, Deleuze has sometimes been classified as a dualist philosopher (Jameson 1997) — a hasty interpretation, to say the least, of the morphogenesis of this philosopher’s conceptual system.(19)

It is interesting to notice how the expositive pace of the two Capitalism and Schizophrenia books, in which dualities abound, is time and again interrupted by provisos, qualifications, distinguos, involutions, subdivisions and other argumentative displacements of the dual (or other) distinctions that had just been proposed by the authors themselves. Such methodical interruptions are exactly this, a question of method and not a recoil of regret after the binary sin; they are perfectly determined moments of conceptual construction.(20) Neither principle nor result, the Deleuzian dyads —one might wish to call them ‘conceptual duplexes’, after Strathern 2005 — are means to arrive elsewhere. The exemplary case here is, once again, the distinction between root-tree and canal-rhizome:

The important point is that the root-tree and the canal-rhizome are not two opposed models; the first operates as a transcendent model and tracing, even if it engenders its own escapes; the second operates as an immanent process that overturns the model and outlines a map, even if it constitutes its own hierarchies, even if it gives rise to a despotic channel. It is not a question of this or that place on earth, or of a given moment in history, still less of this or that category of thought. It is a question of a model that is perpetually in construction or collapsing, and of a process that is perpetually prolonging itself, breaking off and standing up again. No, this is not a new or different dualism. […] We invoke one dualism only in order to challenge another. We employ a dualism of models only in order to arrive at a process that challenges all models. Each time, mental correctives are necessary to undo the dualisms we had no wish to construct but through which we must pass. Arrive at the magic formula we all seek — PLURALISM = MONISM — via all the dualisms that are the enemy, an entirely necessary enemy, the furniture we are constantly rearranging (ATP: 22-23).

Along with this brushing off of the readings that reduce their philosophy to another Great Divide Theory (21), the authors illustrate two characteristic procedures. Firstly, the treatment of concepts in a ‘minor’ or pragmatic key, as tools or vehicles rather than as ultimate objects, meanings or destinations; the philosopher as penseur sauvage. Whence the authors warily realistic attitude towards the dualistic propensities of inertial thinking. In AŒ, they expound a monist conception of desiring production; in ATP, they develop a ‘post-plural’ theory of multiplicities — two pointedly non-dualistic enterprises. Yet they do not suppose that dualisms are a surmountable obstacle through the sheer power of wishful unthinking, like those who fancy that it is enough to call someone else a dualist to stop being such themselves. Dualisms are real and not imaginary; they are not a mere ideological mirage but the modus operandi of an implacable abstract machine of overcoding. It is necessary to undo dualisms because first of all they were made. Moreover, it is possible to undo them for the same reason: for the authors do not think that dualisms are the event horizon of western metaphysics, the absolute boundary that can only be exposed — deconstructed — but never crossed by the prisoners in the Cave. There are many other possible abstract machines.

This takes us to the second procedure. Deleuzian dualities are constructed and transformed according to a recurrent pattern, which determines them as minimal multiplicities — partial dualities, one might say. Every conceptual distinction begins by the establishment of an extensive-actual pole and an intensive-virtual one. The subsequent analysis consists in showing how the duality changes its nature as it is taken from the standpoint of one and then the other pole. From the standpoint of the extensive (arborescent, molar, rigid, striated etc.) pole, the relation that distinguishes it from the second pole is typically an opposition: an exclusive disjunction and a limitative synthesis, that is, an extensive, molar and actual relation itself. From the standpoint of the other (rhizomatic, molecular, supple, smooth) pole, however, there is no opposition but intensive difference, implication or disjunctive inclusion of the extensive pole in the intensive or virtual pole; the duality posed by the first pole reveals itself as the molar echo of a molecular multiplicity at the other pole.(22)

The two poles or aspects are always said to be present and active in every phenomenon or process. Their relation is typically one of ‘reciprocal presupposition’, a notion many times advanced in ATP (pp. 49-50, 73, 97, 235, 554) in lieu of the classic schematims of causality (linear or dialectical), micro-macro reduction (ontological or epistemological), and expressivity (hylomorphic or signifying). From an anthropological standpoint, it is tempting to relate reciprocal presupposition to the Wagnerian double semiotics of invention and convention, in which each mode of symbolization precipitates or ‘counter-invents’ the other, according to a figure-ground reversal scheme (Wagner 1981: ch. 3; 1986).(23) Or still, to the behaviour of certain central analytical duplexes in The Gender of the Gift (Strathern 1988), such as those that preside over the economy of gender or the logic of exchange in Melanesia, in which a pole — cross-sex/same-sex, mediated/unmediated exchange — is always described as a version or transformation of the other, ‘each provid[ing] the context and grounding for the other’, as Strathern summarized in a quite different (precisely!) context (1991a: 72).(24)

The crucial point here is that reciprocal presupposition determines the two poles of any duality as being equally necessary, since they are mutually conditioning, but does not thereby make them into symmetrical or equivalent poles. Inter-presupposition is a asymmetric relation: ‘the way is not the same in both directions’. Hence, as they distinguish the rhizomatic maps from arborescent tracings, Deleuze & Guattari observe that the maps are constantly being totalised, unified and stabilised by the tracings, which are, in turn, subject to all sorts of anarchic deformations induced by rhizomatic processes. Yet, at the end of the day, ‘the tracing should always be put back on the map. This operation and the previous one are not at all symmetrical’ (ATP: 14).(25) They are not symmetrical because the latter operation, tracing or calque, works contrary to the process of desire (and ‘becoming is the process of desire’ — ATP: 334) whereas the other forwards it. Tracing is dangerous because it ‘injects redundancies’ into the map, organizing and neutralising the rhizomatic multiplicity: ‘What the tracing reproduces of the map or rhizome are only the impasses, blockages, incipient taproots, or points of structuration (...) Once a rhizome has been obstructed, it’s all over, no desire stirs; for it is always by rhizome that desire moves and produces’ (ATP: 15). This asymmetrical relation between processes and models in reciprocal presupposition (in which the rhizome is process, while the tree is model) reminds one very much of the distinction between difference and negation developed in Difference & Repetition (D. 1968: 302-ff): negation is real but its reality is purely negative; it is only inverted, extended, limited and reduced difference. Thus, although Deleuze and Guattari warn more than once that it is not the case of establishing an axiological contrast between the rhizome and the tree, the molecular and the molar and so on (ATP: 22, 237), the fact remains that there is always a tendency and a counter-tendency, two entirely different movements: the actualization and the counter-effectuation (or ‘crystallization’) of the virtual (D. & G. 1991: 147-52). The first movement consists in the decay of differences of potential or intensity as these explicate themselves into extension and body forth empirical matters of fact. The second is the creator or ‘implicator’ of difference as such; it is a movement of return or reverse causality (ATP: 476), a ‘creative involution’ (id.: 203), but this does not prevent it from being strictly contemporaneous to the first as its transcendental and therefore non-annullable condition. This last movement is the Event or the Becoming, a pure reserve of intensity — the part, in everything that happens, that escapes its own actualization.

Once again, it seems natural to approximate this asymmetry of inter-implicated processes to certain aspects of Wagnerian semiotics (1981: 51-53, 116, 121-22). The ‘dialectical’ or obviational nature of the relation between the two modes of symbolization belongs as such to one of the modes, the invention-differentiation mode, whereas the contrast between the two modes is, by itself, the result of the other mode’s operation, the conventionalisation-collectivisation one. Moreover, although the two modes operate simultaneously and reciprocally in every act of symbolisation (they operate one upon the other, since there is nothing ‘outside’ them), there is ‘all the difference in the world’ (op. cit.: 51) between those cultures whose controlling context — in the terms of ATP, the dominant form of territorialisation — is the conventional mode and those in which the control is the differentiating mode. If the contrast between the modes is not axiological in itself, the culture that favours conventional and collectivising symbolisation — the culture that generated the theory of culture as ‘collective representation’ — is firmly territorialised on tracing mechanisms, thereby blocking or repressing the dialectics of invention; it must, in the final analysis, ‘be put back in the map’. This, according to Wagner, is what anthropologists ‘do’, or rather, counter-do.

Similarly, the contrast advanced in The Gender of the Gift between gift-based and commodity-based socialities is explicitly assumed as internal to the commodity pole (op.cit.: 16, 136, 343), but at the same time it is as if the commodity form were a unilateral transformation of the gift instead of the contrary, insofar as the analysis of a gift-based sociality forces the anthropologist to recognize the contingency of the cultural presuppositions of anthropology itself and to displace its own commodity-based metaphors (op. cit. : 309). The point of view of the gift on the commodity is not the same that the point of view of the commodity on the gift. Reciprocal asymmetric implication.(26)


If there is effectively one implicative asymmetry that can be said primary within the Deleuzian conceptual system, it resides in the distinction between the intensive and the extensive. The second part of this article discusses the relevance of this distinction to the reinterpretation made in Capitalism and Schizophrenia of two key concepts of classical anthropological kinship theory: alliance and filiation. The choice might be seen as requiring justification. I would argue, then, that the treatment given by Deleuze and Guattari to these two notions manifests with especial clarity an important theoretical displacement that occurs between Œ and ATP; and secondly, that it suggests the possibility of a transformation of the anthropology of kinship so as to align it with the ‘non-humanist’ developments that take place today in other fields of investigation (Jensen 2004). The issue is: how to convert conceptually the notions of alliance and filiation, traditionally seen as the basic social coordinates of hominisation insofar as the latter is effected in and through kinship, into an opening to the extra-human? In other words, how to transform these intra-anthropological operators into cross-ontological ones? If humanity is no longer an essence, what to make of kinship?

After having played a quasi-totemic function in anthropology from the 1950’s to the 1970’s, when they signified two diametrically opposed conceptions of kinship (Dumont 1971), the notions of alliance and filiation, following the general fate of the Morganian paradigm to which they belonged, suddenly lost their synoptic value, receding into the more modest role of analytical conventions, when they did not pass (away) from use to mention.(27) The following pages propose a reflective interruption of this movement, suggesting that some parts of classical kinship theory can be recovered and put back in use. It is certainly not the case of turning back to the statu quo ante and replunge into the analytical formalisms of prescriptive alliance, and even less into the substantialist metaphysics of descent group theory, but of imagining the lineaments of a rhizomatic conception of kinship capable of extrapolating all the possible consequences from the premise that ‘persons have relations integral to them’ (Strathern 1992b: 101). If the theory of descent groups had as its abstract archetype the ideas of substance and identity (the group as a metaphysical Individual), and the theory of marriage alliance the notions of opposition and totalisation (society as a dialectical Totality), the perspective suggested here poaches on Deleuzian philosophy in search of some elements for a theory of kinship as difference and multiplicity (the Relation as inclusive disjunction).

Against Exchange

Anthropological literature is given pride of place in Capitalism & Schizophrenia. From Bachofen and Morgan to Lévi-Strauss and Leach, the first book of the diptych rewrites from scratch the anthropology of kinship. Its main interlocutor and controversial target is Lévi-Strauss’ structuralism, for whom and largely against whom are mobilised a number of references, from Malinowski’s functionalism to Fortes’ structural functionalism, from Griaule and Dieterlen’s Dogon experiment to Meillassoux and Terray’s ethno-Marxism, from Nuer relational segmentarity to Ndembu social dramaturgy.

The general subversionary and liberating effects of AŒ in its time cannot be overstated. But it should be regarded as an epoch-making book also from the restrict viewpoint of kinship theory. In its refusal to take the family as the primary referent of desire, defining the latter as immediately social, AŒ articulated a general philosophical justification (since it applies just as well to the so-called descriptive systems) of the anti-extensionist and anti-genealogist stance defended by many anthropologists then. The argument is important still today, since the popularity of genealogistic or ‘genetic’ interpretations of human relationality has been on the rise, as result of the universal percolation of neo-liberal ontologies. Similarly, extensionist interpretations of semantics are embedded in the many anthropological theories that use the explanatory framework of metaphoric projection to account for personification modes deemed ‘illegal’ in our cosmology.

The thesis of the immediate identity between desiring production and social production fits in with the wider issue of literality in Deleuze’s philosophy, or rather, with his refusal of any distinction between metaphoric and non-metaphoric discourse (Zourabichvili 2004b). In this sense, less than supporting a ‘categorical’ reading of kinship semantics, in the terms of the classic ‘genealogy vs. category’ debate, what is at stake in AŒ is rather a contrast between constitutive-intensive and regulative-extensive conceptions of relationality. A connection to Wagner comes to mind once again. This anthropologist‘s remarks on the tautological character of the incest prohibition (Wagner 1972) — it is impossible to separate relationships, categories and kinship roles, since these aspects are inter-defined — coincide in an intriguing way with Deleuze and Guattari’s arguments over the impossibility of incest:

[I]f relationship is part of the definition of a category [...], then a statement of incest prohibition vis-à-vis categories is the purest and most trivial of tautologies...’ (Wagner 1972: 603).

[T]he possibility of incest would require both persons and names — son, sister, mother, brother, father. Now in the incestuous act we can have persons at our disposal, but they lose their names inasmuch as these names are inseparable from the prohibition that proscribes them as partners; or else the names subsist, and designate nothing more than prepersonal intensive states that could just as well "extend" to other persons ... one can never enjoy the person and the name at the same time (AŒ: 161-62).(28)

The structuralist conception of kinship is grounded, as is well-known, in the transcendental deduction of the incest prohibition as the condition of sociogenesis (Lévi-Strauss 1969). The authors of Anti-Œdipus refuse this conception, under the argument that it is an anthropological generalisation of Œdipal thinking. Their criticism of the Freudian reduction of desiring production to Œdipus is thus extended to what could be called the ‘Maussian reduction’ of social production by Lévi-Strauss, the more eminent contemporary propagator of ‘exchangist notions of society’ (AŒ: 142; 185). Deleuze and Guattari compare disadvantageously The Gift to Nietzsche's On the Genealogy of Morals; the latter, they suggest, should be the anthropologist’s true bedside book (AŒ: 190).(29)

The anti-Œdipal reconstruction of kinship theory made in the first volume of Capitalism and Schizophrenia is, however, interestingly partial or incomplete. AŒ remains riveted to a ‘humanist’ or anthropocentric conception of sociality, and is haunted by the empirico-metaphysical problem of hominisation. The blind spots of this focus only show, of course, from the radically an-Œdipal vantage point of ATP, published one decade later. The first book intended to be a critique of both psychoanalysis and Œdipus; the vocabulary is almost parodistically Kantian: transcendental illusions, illegimate usages of the syntheses of the unconscious, the four paralogisms of Œdipus, and so on. It could thus be argued — somewhat mischievously — that by imagining itself as a sort of Critique of Psychoanalytic Reason, AŒ remains, in a fundamental philosophical sense, an Œdipal book; and worse, dialectically so.(30)

It is this limitation that might explain the systematic interpretation of alliance as a sort of super-conductor for the Œdipal triangle, an argument that poses parenthood as prior to conjugality (the first ‘extends into’ the latter) and alliance as something merely instrumental to the deployment of filiation (AŒ: 71). In other words, the criticism of all ‘exchangist conceptions’ expressed by AŒ is grounded on a counter-theory of Œdipus in which filiation and production, rather than alliance and exchange, are primary. In this and other senses, AŒ is an anti-structuralist book. Yet, if its authors take a distance from the Lévi-Straussian evaluation of human kinship, it is necessary firstly that they accept some of the terms in which the anthropological concept of kinship was formulated by Lévi-Strauss. And it is exactly this what changes, from AŒ to ATP.

Intensive filiation

Against the theme of exchange as the socio-institutive synthesis of opposing (contradictory) interests, AŒ puts forth the postulate that the social machine works paramountly to code the flows of desire. Social production is desiring production in a coded state. Deleuze and Guattari propose an inscriptive conception of society (p. 184) — the task of the socius is the marking of bodies (memory creation) — on the one hand, and a productionist cosmology, on the other: ‘everything is production’ (p. 4). In good Grundrisse style, production, distribution and consumption are defined as different moments of Production seen as a universal process. Inscription is a moment of this Production, the moment of the recording or coding of production, which counter-effects a fetishized socius as the form of the natural or divine Given, a magical surface of inscription and an element of anti-production (the Body without Organs).

In the third chapter of AŒ, the authors engage in a detailed exposition of the primitive territorial machine and analyze its characteristic ‘declension of alliance and filiation’ (AE: p. 171). The basic hypothesis, and crucial analytical decision, consists in supposing that filiation is doubly declensed by the primitive machine; firstly, as a generic and intensive state of proto-kinship; secondly, as a particular and extensive state in complementary opposition to alliance, which appears exclusively on this extensive plane. What is more, alliance appears in order to (as it were) accomplish the task of extensivising and coding kinship, that is, of actualising it. Deleuze and Guattari postulate thus the existence of a filiation prior to incest-as-prohibition; a nocturnal and biocosmic, disjunctive and ambiguous filiation, a germinal implex or influx (p. 162) that is the first character of inscription marked on the full, unengendered body of the earth: ‘a pure force of filiation or genealogy, Numen’ (p. 154).(31)

The analysis here relies centrally on an interpretation of the mythical narratives collected by Marcel Griaule and his team, especially that of the well-known Dogon origin myth of Le renard pâle (Griaule and Dieterlen 1965): Amma the cosmic egg, the placentary Earth, Yurugu the incestuous trickster, the hermaphroditic and ictiomorphic successive Nommo twin-pairs and so on.(32) The narrative works as a kind of anti-myth of Œdipus for Deleuze and Guattari.(33)

It is hardly surprising that this reference myth — a cosmogonic story quite widespread in West Africa (Cartry & Adler 1971: 15) — determines filiation as the original element, and alliance as a supervening dimension whose function is to differentiate lineage affiliations. We are here — as it were? — within a classical ’Africanist’ universe of discourse (Fortes 1969, 1983). What is intense and primordial are the ambiguous, involved, implicated and (pre-)incestuous filiative lineages, which abandon their inclusive and illimitative regime as (being the subject of a ‘nocturnal and biocosmic’ memory) they must ‘suffer repression’ by alliance in order to be explicated or actualized in the physical space of society (AŒ: 155).

It is as if the system of the Dogon, who are the synecdochic Savages at this point of AŒ, was descent-theoretical at the virtual or intensive level and alliance-theoretical at the actual or extensive level. Thus, the authors take totally on board Leach’s criticism of Fortes concerning ‘complementary filiation’ (loc. cit.). They also reach the conclusion — from a famous Lévi-Straussian demonstration of the logic of cross-cousin marriage (1969: 129-33) — that ‘[a]t no time...does alliance derive from filiation’, and that ‘in this system in extension there is no primary filiation, nor is there a first generation or an initial exchange, but there are always and already alliances...’ (AŒ: 157).(34) In the extensive order, filiation assumes an ‘administrative and hierarchic’ trait, whereas alliance, which comes first in this order, is ‘political and economic’ (p. 146). The affine as a sociopolitical persona is there from the start (‘always and already’) to prevent any Œdipal closing to the socius of families, that is, to make sure that familial relations are radically co-extensive to the wider social field (p. 166). Yet there is something before the start: in the metaphysical order of genesis — from the mythical standpoint, precisely (p. 155) — alliance comes second. ‘The system in extension is born of the intensive conditions that make it possible, but it reacts on them, cancels them, represses them, and allows them no more than a mythical expression’ (p. 160).(35) Post-prohibition kinship is, therefore, conceived in terms of a reciprocal pressuposition between alliance and filiation that is actually (politico-economically) ruled by the former and virtually (mythically) by the latter.

The intensive level of myth is thus peopled by (pre-)incestuous filiations that ignore alliance. It is impossible not to remember here the famous last paragraph of The Elementary Structures of Kinship, in which it is written that ‘mankind has always dreamed of seizing and fixing that fleeting moment when it was permissible to believe that the law of exchange could be evaded’ (Lévi-Strauss 1969: 497). Compare, however, this somewhat Freudian conclusion to that other well-know passage in which the author defines myth as ‘a story from the time when humans and animals did not distinguish themselves from one another’ (Lévi-Strauss & Eribon 1988: 193), adding that the one thing mankind has never reconciled itself to is the lack of communication with the other species that people the earth. Not the same nostalgia as the former one, then; in a sense, its very opposite.

Reformulating the problem in the terms of the Deleuzian general conceptual economy, it seems to me that the crucial step in this analysis of the Dogon myth is the determination of intensive filiation as an operator of disjunctive synthesis — the Nommo who is/are one and two, male and female, human and snake; the Fox who is son, brother and husband of the Earth —, whereas alliance is the operator of conjunctive synthesis or pairing:

Such is alliance, the second characteristic of inscription: alliance imposes on the productive connections the extensive form of a pairing of persons, compatible with the disjunctions of inscription, but inversely reacts on inscription by determining an exclusive and restrictive use of these same disjunctions. It is therefore inevitable that alliance be mythically represented as supervening at a certain moment in the filiative lines (although in another sense it is already there from time immemorial) (AŒ: 155).

As we know, disjunctive synthesis is the diagnostic relational regime of multiplicities. On the same page of the passage just cited, we can read that the problem is not one of going from filiations to alliances, but to move from ‘an intensive energetic order to an extensive system’. In this sense,

[n]othing is changed by the fact that the primary energy of the intensive order...is an energy of filiation, and does not as yet comprise any distinctions of persons, nor...of sexes, but only prepersonal variations in intensity (loc.cit.).

Here, one might wish to add that if this intensive order does not know distinctions neither of person nor gender, it does not know either any distinction of species, particularly that between humans and non-humans. In myth all actants are deployed on a single interactional field, which is at one and the same time ontologically heterogeneous and sociologically continuous. There are no humans ‘there’; or rather, everything is ‘human’ there. And of course, there were everything is human, the human is something else entirely (Viveiros de Castro 2004: 16).

We are now in possession of all the elements of the problem: if nothing is changed by the fact that the primary energy is an energy of filiation, could it be possible to imagine an intensive order in which the primary energy be an energy of alliance? Is it really necessary that alliance work always to ordain, discern, discrete and police a prior, pre-incestuous order of filiation? Or could it be conceivable an intense, anŒdipal alliance that comprises only ‘prepersonal variations in intensity’? In a few words, the problem is that of imagining a concept of alliance as disjunctive synthesis.

In order to do it it, we would probably have to take a greater distance from the Lévi-Straussian cosmology than Anti-Œdipus does, at the same time that the exchange concept must be submitted to a properly Deleuzian, or ‘perversive’, interpretation.(36) Minimally, it means abandoning the description of the kinship atom in terms of an exclusive alternative — this woman as either my sister or my wife, that man as either my brother or my brother-in-law — and rephrase it in terms of an inclusive or non-restrictive disjunction: “either...or...or”. The difference between sister and wife, brother and brother-in-law must be taken as an internal difference, ‘indecomposable and unequal to itself’. Just like with the schizophrenic and the male/female, dead/alive disjunctions which s/he confronts, we might say that this woman is indeed my sister or my wife, but she ‘belongs precisely to both sides’, sister on the side of sisters (and brothers) and wife on the side of wives (and husbands) — not both at once to me, ‘but each of the two as the terminal point of a distance over which [s]he glides’ (AŒ: 76).

The point can be rephrased in a language that every anthropologist will recognise (Strathern 1988, 2001). My sister is my sister insofar as she is the wife of another, and vice-versa. It is the cross-sex relationship of my sister/wife to myself that generates my same-sex relationship with my brother-in-law. Hence, cross-sex relations not only engender same-sex relations but also communicate their own internal differential potential to the latter. Two brothers-in-law are related in the same way as the cross-sex dyads presupposed by their relation (brother/sister, husband/wife): not despite their difference but because of it. One of the brothers-in-law sees the conjugal face of his sister in her husband; the other sees his wife's sororal side in her brother. One sees the other as determined by the opposite-sex link that differentiates both of them: each sees oneself as ‘same-sex’ insofar as the other is seen ‘as’ cross-sex, and reciprocally. The two sides of the relating term create, thus, a division that ‘always and already’ is internal to the related terms. All become double, the relater and the related reveal themselves to be interchangeable without thereby becoming redundant; each vertex of the affinity triangle (two triangles, actually, one for each sex as ‘relator’) includes the other two vertexes as versions of itself.(37) This complex duplication is explicitly described by Deleuze and Guattari in a commentary on the analogy between homossexuality and vegetal reproduction in Proust's Sodom and Gomorrah. Something like an atom of gender is suggested:

[T]he vegetal theme ... brings us yet another message and another code: everyone is bisexual, everyone has two sexes, but partitioned, noncommunicating: the man is merely the one in whom the male part, and the woman the one in whom the female part, dominates statistically. So that at the level of elementary combinations, at least two men and two women must be made to intervene to constitute the multiplicity in which transverse communications are established: [...] the male part of a man can communicate with the female part of a woman, but also with the male part of a woman, or with the female part of another man, or yet again with the male part of the other man, etc. (AŒ: 69-70).

Two men and two women at least. Make it a sibling-exchange matrimonial arrangement between two pairs of opposite-sex siblings (two bisexual dividuals, that is), and there you have it: an extensive, structural version of the intensive rhizomatic multiplicity that is gender. But then, as a matter of course, the tracing must be put back on the map, and ‘everything must be interpreted in intensity’ (AŒ: 158). This is the job the little ‘etc.’ at the end of the passage above may be conceived as doing.

Demonic Alliance

The possibility of an intensive interpretation of alliance is only effectively consolidated in A Thousand Plateaus. Although many things change, from AŒ to ATP, the single most important change from the limited viewpoint of this article is introduced in the tenth plateau, ‘1730: Becoming-intense, becoming-animal, becoming-imperceptible...’ It is there that the concept of becoming is developed, a development that carries away all the Deleuzo-Guattarian conceptuality in an astonishing other-becoming.(38)

The chapter begins by an exposition of the contrast established by Lévi-Strauss (1962b) between serial-sacrificial logic and structural-totemic logic; the imaginary identification between human and animal, on the one hand, and the symbolic correlation between social differences and natural differences, on the other. Between these two analogical models, the series and the structure, Deleuze and Guattari introduce the Bergsonian theme of becoming, a type of relation irreducible to serial resemblances as well as to structural correspondences. The concept of becoming designates a relation that has no right of citizenship within structuralism’s theoretical frame, which tends to treat relations as molar logical objects, apprehended essentially in extension, as oppositions, contradictions and mediations. Becoming is a real relation, molecular and intensive, belonging to a truly relational ontology beyond the pale of the more limited epistemological relationality of structuralism.(39) The disjunctive synthesis of becoming is not possible in the terms of the formal combinatorial rules that engender structures; it is born in the far-from-equilibrium fields of real multiplicities (De Landa 2002: 75). ‘Becoming and multiplicity are the same thing’ (ATP: 275).

If serial resemblances are imaginary and structural correspondences are symbolic,(40) becomings are real. Neither metaphor nor metamorphosis, becoming is a movement that deterritorialises both terms of the relation, extracting them from their prior definitional contexts to associate them through a new, necessarily partial connection. The verb to become, in this conceptual sense, does not denote a predicative judgment or a transitive operation: to be implicated in a becoming-jaguar is not the same as ‘becoming a jaguar’. It is the becoming itself that is feline, not its ‘subject’. For as soon as a man becomes a jaguar, the jaguar is no longer there. ‘[I]n his study of myths, Lévi-Strauss is always encountering these rapid acts by which a human becomes animal at the same time as the animal becomes... (Becomes what? Human, or something else?)’ (41). To become, the authors proceed, is a verb with a consistency all its own; becoming is not imitating, appearing, being, corresponding. And — surprise — ‘neither it is producing, producing a filiation or producing through filiation’ (p.263). Neither filiation nor production. We are not in Anti-Œdipus any more.

‘Intensive thinking in general is about production’, remarks DeLanda (2003: 15). Perhaps it is a little more complicated than that. Becoming does play the same axial, cosmological role in ATP that production does in AŒ. Not exactly because ‘everything is becoming’ — that would be a conceptual solecism —, or because there are no other important notions in the book (war machines, segmentarity, ritornello, regime of signs, assemblage; an embarrassment of riches), but because ATP’s anti-representational concept par excellence, in the sense of being the device that pre-empts the work of representation, is the concept of becoming, just like production was AŒ’s major anti-representational concept. Two distinct movements, therefore, whose relation remains to be more precisely determined: production and becoming. Both involve nature; both are intensive and pre-representative; in a sense, they are the same: becoming is the process of desire, desire is the production of the real, becoming and multiplicity are the same thing, becoming is a rhizome, and the rhizome is the process of production of the unconscious. But in another sense — ‘sense’ also in the sense covered by the French sens: direction — they are definitely not the same: between production and becoming, ‘the way is not the same in both directions’, les deux sens. Production is a process in which the identity of man and nature is realised, and nature reveals itself as a historical process of production (‘the human essence of nature and the natural essence of man become one within nature in the form of production...’ — AŒ: 4). Becoming, on the contrary, is an unnatural participation (ATP: 265) between man and nature, an instantaneous (‘these rapid acts...’) or non-processual movement of capture, symbiosis, transversal connection between heterogeneities. ‘That is the only way Nature operates — against itself. This is a far cry from filiative production or hereditary reproduction’ (p.267). Becoming is counter-productive.

‘The Universe does not function by filiation’ (loc.cit.). One could hardly be more direct. The universe, one might note, in all its states, both virtual-intensive and actual-extensive. And if it does not work by filiation (rather than not by something else), I am tempted to conclude that it can only work by alliance. In the book’s first plateau, as a matter of fact, we had already learnt that ‘the tree is filiation, but the rhizome is alliance, uniquely alliance’ (p. 27). And now we learn that

becoming is not an evolution, at least not an evolution by descent and filiation. Becoming produces nothing by filiation; all filiation is imaginary. Becoming is always of a different order than filiation. It concerns alliance. [...] Becoming is a rhizome (p.263).

What has changed, from the affirmative analysis of that intensive, ambiguous, nocturnal filiation of the Dogon myth in Anti-Œdipus to the flat denial of any significant role played by this relational principle in ATP? How did filiation turn from intensive to imaginary? I think that this change reflects the displacement of the analytical focus from an intra-specific horizon to an inter-specific one; from a human economy of desire — world-historical, racial, sociopolitical rather than familial or Œdipal desire, but human nonetheless — to an trans-specific economy of affects, which ignores the natural order of genera and species and its limitative syntheses, disjunctively including ‘us’ (all and any of ‘us’) in the plane of immanence. From the standpoint of the human economy of desire, extensive alliance works to limit intensive and molecular filiation, actualising it under the molar form of the descent group. But from the standpoint of the cosmic economy of affects (of desire as an unhuman force), it is filiation that comes now to limit, with its imaginary identifications, an alliance as real as it is unnatural between radically heterogeneous beings: ‘[i]f evolution includes any veritable becomings, it is in the domain of symbioses that bring into play beings of totally different scales and kingdoms...’ (ATP: 263). Then follows Deleuze’s favourite example of the wasp and the orchid, an assemblage ‘from which no wasp-orchid can ever descend’ — and without which, one might add, no wasp and no orchid as we know them could ever leave descendants. The wasp-orchid assemblage is the origin myth of both the wasp and the orchid tribes: their shared deterritorialization, their common antimemory (ATP: 324).

The deterritorialisation of sexuality begun in AŒ is now completed; the binary organization of the sexes, bisexuality included (cf. the atom of gender) gives way to n sexes that connect to n species at the molecular level: ‘Sexuality proceeds by way of the becoming-woman of man and the becoming-animal of the human: an emission of particles’ (p. 307).(42) If every animal implicated in a becoming-animal is a molecular multiplicity (‘every animal is fundamentally a band, a pack’, p. 264), it is because it defines a multiple, lateral, extra-filiative and non-reproductive sociality that drags human sociality along in a universal metonymyc flow:

We oppose epidemic to filiation, contagion to heredity, peopling by contagion to sexual reproduction, sexual production... Unnatural participations or nuptials are the true Nature spanning the kingdoms of nature (p. 266).

Alliance, then. But then again, not any kind of alliance. As we have seen, Capitalism and Schizophrenia’s first volume postulated two kinds of filiations: one, intensive and germinal; the other, extensive and somatic, posed by and counterposed to alliance seen as an extensive principle playing the role of ‘repressing representation’ of the representative of desire, the germinal influx.(43) Now we can see two alliances appearing: that discussed in AŒ, internal to the socius and even to the male gender (primary, collective homosexuality: affinal alliance as the archetypal same-sex relation sensu Strathern); and another alliance, intrinsic to becoming, as irreducible to imaginary production and metamorphosis (mythical genealogy, filiation to the animal) as to exchange and symbolic classifications (exogamy, totemism).

Every becoming is an alliance, which does not mean, I repeat, that every alliance is a becoming. There is extensive alliance, which is cultural and socio-political, and intensive alliance, which is unnatural and cosmopolitical. If the former distinguishes filiations, the latter confuses species, or rather, counter-effectuates by implicative synthesis the continuous differences that ‘had been’ actualized by the limitative synthesis of discontinuous speciation. When a shaman activates a becoming-jaguar, he does not ‘produce’ a jaguar, he does not ‘affiliate’ himself to the descent of jaguars either; he makes an alliance:

One can say rather that a zone of indistinction, of indiscernibility, of ambiguity, establishes itself between two terms, as if they had attained the point immediately preceding their respective differentiation: not a similitude, but a slippage, an extreme vicinity, an absolute contiguity; not a natural filiation, but a counter-natural alliance… (D. 1993: 100).

One should note that this concise description of becoming cuts right through the middle of the molar contrast between filiation, metonymic continuity and serial resemblance, on the one hand, and alliance, metaphoric discontinuity and oppositive difference, on the other — a contrast characteristic of structuralist theories of kinship. That ‘absolute contiguity’ established by counter-natural alliance, of a differential-tangential kind, differs toto cœlo from the absolute ‘discontiguity’ between filiative lineages established by cultural alliance (exogamy). But it is irreducible as well, it goes without saying, to any identification or imaginary indifferentiation between ‘the two terms’. Therefore, it is not the case of opposing natural filiation to cultural alliance, as in the classical kinship-theoretical models. The counter-naturalness of intensive alliance is counter-cultural or counter-social just as well, insofar as human sociality is necessarily counter-intensive, being generated through the extensivization of ‘the primary energy of the intensive order’. We are speaking of an included middle, a new alliance, a relation which separates.(44)

It is not necessary to leave Africanist ground to find this other alliance. In the section titled ‘Memories of a Sorcerer II’, in the 10th Plateau, the authors evoke the demonic were-animals of the ‘sacred deflowerer type’ studied by Pierre Gordon, and the were-hyenas of some Sudanese traditions described by Calame-Griaule. The latter furnish the occasion for a decisive commentary:

[T]he hyena-man lives on the fringes of the village, or between two villages, and can keep a lookout on both directions. A hero, or even two heroes with a fiancée in each other's village, triumphs over the man-animal. It is as though it were necessary to distinguish two very different states of alliance: a demonic alliance that imposes itself from without, and imposes its laws upon all the filiations (a forced alliance with the monster, with the man-animal), and a consensual alliance, which is on the contrary in conformity with the law of filiations and is established after the men of the villages have defeated the monster and have organized their own relations. This sheds new light on the question of incest. For it is not enough to say that the prohibition against incest results from the positive requirements of alliance in general. There is instead a kind of alliance that is so foreign and so hostile to filiation that it necessarily takes the position of incest (the man-animal always has a relation to incest). The second kind of alliance prohibits incest because it can subordinate itself to the rights of filiation only by lodging itself, precisely, between two distinct filiations. Incest appears twice, once as a monstrous power of alliance when alliance overturns filiation, and again as a prohibited power of filiation when filiation subordinates alliance and must distribute it among distinct lineages. (ATP: 597-598 [n. 21 of p. 273]).

‘This sheds new light on the question of incest’. The authors seem to be referring to the theory of The Elementary Structures, but I believe this observation is also apposite to the manner the issue is dealt with in Anti-Œdipus. Now, it is alliance that has a double incidence, not only to regulate ‘sexuality as a process of filiation’, but also as ‘a power of alliance, inspiring illicit unions or abominable loves’. Its purpose is not that of regulating, but of ‘preventing procreation’: an anti-filiative alliance (p. 271). And, more importantly still, even exchangist, repressive, filiation-producing alliance begins to show some hidden powers — as if it had been conceptually contaminated by that other, intense alliance. ‘It is true that alliance and filiation come to be regulated by laws of marriage, but even then alliance retains a dangerous and contagious power. Leach was able to demonstrate...’ (p. 272).(45) It is interesting to notice how the word ‘power’ (orig. ‘puissance’, not ‘pouvoir’) begins persistently to qualify and determine the notion of ‘alliance’ from this point onwards. The concept of alliance ceases designating an institution — a structure — and becomes a power, a potential — a becoming. From alliance as form to alliance as force, bypassing filiation as substance. We are no longer in in the structural-mythical element of totemism, but neither are we in the serial-mythical element of sacrifice; we are in the real-magical element of becoming (Goldman 2005; Viveiros de Castro 2006).

We are not in the element of the contract, either — of exchange or interest ‘à l’anglaise’ (AŒ: 190). ‘Desire knows nothing of exchange, it knows only theft and gift’ (p.186). But then again, there may be exchange and exchange. There is an exchange that is certainly not ‘exchangist’ in the commodity-economical sense, since it belongs to the ‘theft and gift’ category: debt- or gift-exchange precisely, a movement of double capture in which people shift (counter-alienate) invisible perspectives by moving visible (inalienable) personified things.(46) Gifts may be reciprocal; but this does not make them any the less violent: the whole point of the gift-act is to make the other act, to extract an act from the other, to provoke a response. And in this sense there is no social action that is not an exchange of ‘gifts’, insofar as all action is ‘social’ by being an action upon another action, a reaction to a reaction. Reciprocity means only recursivity; no society intended. Not to mention altruism.(47) Life is robbery.

Amazonian Alliance

The distinction between the two alliances proposed in ATP seems to impose itself, from a ethno-theoretical point of view, when we move over from a West-African(ist) to an Amazonian(ist) landscape. It corresponds closely to the distinction made by the ethnographers of this region between a intensive or potential affinity, of which one could perfectly state that it is ‘ambiguous, disjunctive, nocturnal, demonic’ and an extensive or effective affinity, subordinated to consanguinity (Viveiros de Castro 2001, 2006). In the context of local and cognatic ‘prescriptive endogamy’ prevailing in many Amazonian societies, affinity as a particular relationship is masked or neutralized by consanguinity (or filiation). Terminological affines are seen as types of cognates (namely, close cross-kin), actual affines are attitudinally consanguinised, specifically affinal terms are avoided in favour of their cross-kin alternatives or of teknonyms expressing co-consanguinity, spouses are conceived as becoming consubstantial by way of sex and commensality, and so forth. We can then say that affinity as a particular relation is eclipsed by consanguinity as part of the process of making kinship. As Riviére observed (1984: 70), ‘within the ideal settlement affinity does not exist’. But this may be taken as implying that, if affinity does not exist within the ideal community, it must then exist somewhere else. Within real settlements, to be sure; but, above all, without the ideal settlement, that is, in the ideal outside of the settlement, as ‘ideal’ (intensive) affinity. For as the perspective shifts from local relationships to wider contexts the value distribution is inverted, and affinity becomes the overall mode of sociality.

Marriage alliance is then both locally concentrated and masked in Amazonian socialities. Supra-local relations, on the other hand, are a variable mixture: statistically residual but politically strategic intermarriage; formal friendship and trade-partnership links; intercommunal ritual and feasting; physical and spiritual, actual or latent predatory warfare. This relational complex straddles species boundaries: animals, plants, spirits and divinities are equally engaged in such synthetic-disjunctive relations with humans. All these relationships, whatever their components, manifest the same general set of values and dispositions, as witness the common idiom in which they are expressed, that of affinity. Guests and friends as much as foreigners and enemies, political allies or clients as much as trade partners or ritual associates, animals as much as spirits, all these kinds of beings bathe, so to speak, in affinity. They are conceived either as generic affines or as marked versions — sometimes inversions — of affines.(48) The Other is first and foremost an affine.(49)

This affinity undoubtedly belongs to the ‘second kind of alliance’ mentioned by Deleuze and Guattari. It is hostile to filiation, since it appears above all when and where marriage is not an option (or at least not a preference), and its productivity is not of a kinship-procreative kind; it is, rather, part of a war-machine which is anterior and exterior to kinship as such. This is an alliance against filiation, not in the sense of being a repressing representation of a prior intensive filiation, but in that of being what prevents filiation from functioning as the seed of transcendence (the descent group, the origin, the ancestor). Every filiation is imaginary, say the authors of A Thousand Plateaus. We might add: every filiation projects a State, it is State filiation. Amazonian intensive alliance is an alliance against the State (Clastres 1977).

The ‘affinity first’ configuration — that is, a configuration where the ‘second kind of alliance’ is the condition of possibility of extensive kinship (of marriage alliance as well as parental filiation)—, although perhaps more easily aprehended in the cognatic-endogamous morphologies found among Northern Amazonian peoples like the Trio, the Piaroa, the Jívaro or the Yanomami, is not restricted to them. It is present, under different guises, everywhere in Amazonia and Central Brazil, including those societies that feature descent constructs, house-like corporations, local exogamy rules, Crow-Omaha terminologies, semi-complex alliance systems and what not. I believe it is one of the tell-tale signs of the unity of Amazonian cultures, and probably beyond; for we might be touching here ‘the bedrock’ (Lévi-Strauss 1991: 295) of Amerindian cosmology as a coherent historical entity.

Consider the continent-wide mythological complex analyzed in the Mythologiques, whose subject-matter is the origin of human culture. If we compare the Ameridian myths to our own mythology of cultural origins, the first difference that strikes the eye is the dominance of relations of affinity in the former and that of parenthood in the latter. The central figures of the Amerindian myths of culture are canonically related as affines. One conspicuous character of this mythology,to take an example, is the cannibal father-in-law, the non-human Master of things cultural, who subjects his son-in-law to a series of ordeals before the latter succeds in returning to his fellow-humans with the precious bounty of Culture. The content of this archi-myth (Lévi-Strauss 1971: 503ff) is not too different from the Promethean plot: there is heaven and earth, and a hero stranded in-between the two, and civilizational fire, and the ‘gift’ of women, and human mortality. But the protagonists of the Amerindian myth are wife-givers and wife-takers, not parental and filial figures as those haunting our mythologies, be they Greek, Near Eastern, West African or Freudian. Not to put too fine a point on it, in the Old World humans had to steal ‘fire’ from a divine father, while Amerindians had to steal it from an animal wife-giver, or received it as a marriage counter-gift from an animal wife-taker.

Mythology may be said to constitute the discourse of the Given (see Wagner 1978). It originally grants what must thenceforth be taken for granted, the primordial conditions from/against which humans must define (construct) themselves: it sets the terms of the ontological debt. If such is the case, then in Amerindian thought the debt of the Given does not concern filiation and parenthood, but marriage and affinity; the Other, as we saw, is first and foremost an affine. Please note that I am not referring to the trivial fact that the myths treat affinal relations as ‘already-there’ — they do the same to consanguineal relations, or they may imagine worlds in which pre-humans did not care for the incest taboo, etc.(50) —, but to the fact that affinity constitutes the frame, the sociological armature in which the mythic message is couched. Such frame or setting is always peopled by more than one kind of people; in particular, it is filled with animal affines. It is absolutely essential that they be animal, or, more generally, non-human (future non-humans, that is; in myth everyone is partially human, present humans included, though the way is not the same in both directions).

It is this alliance with the non-human that defines the ‘intensive conditions of the system’ in Amazonia. At this level there is not, strictly speaking, a distinction — necessarily extensive — between alliance and filiation. Or rather, if there are two alliances, then there are two filiations as well. If all production is filiative, not all filiation is (re)productive; if there are reproductive and administrative (representative and state-like) filiations, there are contagious and monstrous filiations as well, those resulting from unnatural alliances or becomings. That is why incest has an intrinsic ‘affinity’ to trans-specific unions: hyper-exogamy and hyper-endogamy flow into each other in the intensive world of myth, the conditioning world of fluent difference that accompanies the actual world as its virtual counterpart.(51)

Production and Exchange

A steadfast refusal of any ‘exchangist’ conception of the socius pervades the two Capitalism and Schizophrenia books. In AŒ, exchange is rejected in favour of production as the general ‘action’ concept, and circulation (to which Maussian exchange is univocally assimilated)(52) is subordinated to inscription. In ATP, as we have seen, production makes room for a different non-representative relation, becoming. If production is inherently filiative, becoming shows an affinity with alliance. What becomes then of the anti-exchangist stance when we pass from production to becoming?

It is well-known though conveniently forgotten that the concept of production in Anti-Œdipus is not exactly identical to the homonym Marxist concept. Deleuze and Guattari’s desiring production or functioning-production cannot be confused with the Hegelian-Marxist, Œdipal-oriented production, dominated by the ideas of need and lack (AŒ: 26ff).(53) The latter concept of production has wide currency in anthropological circles, and it is for its sake and its convenient appurtenances (domination, false consciousness etc.) that, as a rule, ‘exchangist’ positions are criticised within the discipline. Besides, AŒ's desiring production is not that easily distinguishable from a process of generalised circulation; as Lyotard (1977: 15) provocatively argued, ‘[t]his configuration of Kapital, the circulation of flows, is imposed by the predominance of the point of view of circulation over that of production’.

If it is possible, and indeed necessary, to make this disambiguation between necessitated production and desiring production, between labour-production and machinic-production, it can be argued by analogy that it is equally pertinent to distinguish between alliance-as-structure and alliance-as-becoming. The distinction undermines the contractualist concept of alliance, playing with a deliberately equivocal homonymy between the intensive alliance of Amazonian sociocosmology and the extensive alliance of classical kinship theory. Naturally, in both cases the equivocation is a little more than that, since there is a filiation, albeit monstrous rather than reproductive, between the respective homonym concepts. AŒ’s production owes much to the production of political economy, even as it subverts it. Similarly, the potential Amazonian alliance exists in filigree or in counter-light (virtually, as it were) in the Lévi-Straussian œuvre, whose anti-Œdipal and, therefore, (self-)subversive potential must be brought about.

The problem is ultimately that of imagining a non-contractualist and non-dialectical concept of exchange: neither rational interest nor the a priori synthesis of the gift, neither unconscious teleology nor the work of the signifier, neither inclusive fitness nor desire for the Other’s desire; neither ‘contract’ nor ‘conflict’ — but a mode of becoming-other.(54) Alliance is the mode of becoming-other proper to kinship. The machinic and rhizomatic laterality of alliance is ultimately much closer to Deleuzian philosophy than the organic and arborescent hierarchic verticality of filiation. Therefore, the question is that of freeing alliance from the managerial control of (and by) filiation, liberating thus its ‘monstrous’, that is, creative, potencies. As to exchange, it is now tolerably clear that this manifestation of relationality has never been opposed to production, notwithstanding the prevailing dogma. On the contrary, today we are in a position to see that exchange has always been treated by anthropology as the most eminent form of production: production of Society, precisely. The question then is not one of revealing the naked truth of production behind the hypocritical veil of exchange, but rather that of liberating the latter concept from any equivocal function within the repressive machine of filiative and subjectifying production, giving it back to its proper element, the element of becoming. Exchange, or, the circulation of perspectives: exchange of exchange, that is, change.


(1) And "the robber requires justification” (loc.cit.). The quote was plundered from Isabelle Stengers’ capital study of Whitehead (Stengers 2002: 349).

(2) The reader certainly knows to what a small fraction of humankind the pronoun ‘our’ (cultural apperception) refers.

(3) ‘Allo-anthropology’, ‘hetero-anthropology’ — or perhaps, ‘cross-anthropology’. The reader is invited to take the redundancy as a provocation.

(4) If not the century itself, as Foucault famously predicted, it is at least structuralism that seems to be turning Deleuzian of late (which is not the same as saying that Deleuze is turning structuralist). See, for example, the recent re-examination of Saussure’s ideas by Maniglier (2006), which modify dramatically the traditional image of the Ancestor.

(5) In AŒ, ‘the reversal of psychoanalysis [is] the primary condition for a shake-up of a completely different scope [...] on the scale of the whole of the human sciences, there is an attempt at subversion on the general order of what Laing and Cooper had carried out solely on the terrain of psychiatry’ (Donzelot op. cit.: 27).

(6) ‘Infernal dichotomies’ in an analogous if not identical sense to that of the ‘infernal alternatives’ exposed by Pignarre & Stengers (2005: ch. 4).

(7) I say meta-concept because every concept is a multiplicity in its own right, though not every multiplicity is conceptual (D. & G. 1991: 21ff).

(8) DeLanda 2002 (pp. 9-10, 38-40 et passim) is a detailed exposition of the mathematical origins and implications of the Deleuzian concept of multiplicity (also evoked in Plotnitsky 2003). Zourabichvili 2003 (pp. 51-54), in turn, is the best overview of the concept’s properly philosophical connections and its place in Deleuze’s work.

(9) In remembrance of Pierre Clastres (1974). Clastres was (and remains) one of the rare French anthropologists who knew how to make something out of AŒ’s ideas, besides being one of the inspirations for the theory of the war machine developed on the Plateaus 12 and 13 in ATP.

(10) A multiplicity or a rhizome is a system, one must notice, and not a sum of ‘fragments’. It is simply another concept of system, which differs from the arborescent system as an immanent process differs from a transcendent model (ATP: 22). We are not talking post-modernism here.

(11) Or rather, where the whole is only one part among parts. ‘We believe only in totalities that are peripheral. [N.B. ‘lateral’ rather than ‘peripheral’: ‘Nous ne croyons à de totalités qu’à coté’.] The whole not only coexists with all the parts; it is contiguous to them, it exists as a product that is produced apart from them and yet at the same time is related to them’ (AŒ: 42, 43-4; French ed.: 50). About the heterogeneity of the elements connected in a rhizome, it is important to notice that it does not concern a previous ontological condition, or essence of the terms (what counts as heterogeneous, in this sense, depends on the observer’s ‘cultural predispositions’ — Strathern 1996: 525), but an effect of its capture by a multiplicity, which renders heterogeneous the terms that it connects by making them operate as singularities, ‘representatives of themselves’.

(12) The “one man and many men” of Strathern 1991b is not witout evoking a n-1 multiplicity: “(m)any-minus-one men”. If I understand her correctly, in Melanesian æsthetics the unity of a plurality comes from itself (it is a “moment” of itself), not from an exterior/superior principle. Alternatively, one could imagine a “1-n multiplicity”, since, to take another famous Melanesian example, “the greatest possible number that the Iqwaye can and does reach is one” (Mimica 1992: 95).

(13) ‘Pluralism’ as opposed to dualism — cf. ATP: 23.

(14) For an anthropological comparison of multiplicities, see Strathern 2005, specially p. 161: ‘[C]ontrasting types of multiplicity come into view. If we talk of multiple origins in relation to Euro-American works, then multiplicity comes from the way persons are added to one another’s entreprises. If we talk of multiple origins in relation to their Melanesian couterparts, then multiplicity comes from the way people divide themselves from one another’. So perhaps we could see in the former (Euro-American) multiplicity a molar and arborescent, or ‘false’, multiplicity, while the latter would be of the molecular and intensive type, ‘composed of particles that do not divide without changing in nature’ (ATP: 37). Strahern goes on to remark, in the passage just cited: ‘I am not comparing like with like. Quite so.’ I take this as implying a definition — comparing the incomparable — of anthropology.

(15) This could be taken as an acceptable gloss of the canonical formula of myth of Lévi-Strauss (1955). The recurrent presence of this synoptic figure in his Mythologiques and beyond bears out that, for structuralism, not all the ‘differences that make a difference’ are reducible to the reversible proportionalities of the totemic scheme.

(16) The ideas of Tarde, an ultra-Leibnizian thinker who was the greatest metaphysical adversary of Durkheim (a strictly Kantian soul) were rescued from the Land of Lost Theories by Deleuze (1968: 104-5 n. 1; ATP: 240-41, 1988: 147). Recently they have been taken up again by B. Latour and M. Lazzarato, among others. Chunglin Kwa, in an article already mentioned here, observes a ‘fundamental difference between the romantic conception of society as an organism and the baroque conception of an organism as a society’ (2002: 26). This is a perfect description of the difference between the sociologies of Durkheim and Tarde. Against the sui generis character of the social facts proposed by the former, the ‘universal sociological viewpoint’ of the latter states that ‘every thing is a society, every phenomenon is a social fact’ (Tarde op.cit.: 67, 58).

(17) See Deleuze 1969: 202-203. Similarly, in the dialectics of the Master and the Slave, it is the slave who is dialectician and not the master (Deleuze 1962: 11).

(18) Once again Zourabichvili: ‘A meditation on Nietzschean perspectivism puts one in a position to see the positive consistence of the [Deleuzian] concept of disjunction: a distance between viewpoints which is at the same time undecomposable and unequal to itself, since the way is not the same in both directions’ (2003: 79; emphasis in the original).

(19) For a subtler interpretation of Deleuze as a philosopher of ‘immediate or non-dialectical duality’, see Lawlor 2003.

(20) Thus, with the duality between arborescence and rhizome (‘have we not … reverted to a simple dualism?’ ATP: 14), two schemes that do not cease to interfere with each other. Thus, with the two types of multiplicity, molar and molecular, which operate always at the same time and in the same assemblage — there is no dualism of multiplicities but only ‘multiplicities of multiplicities’ (ATP: 38). Thus, with the distinction between form of expression and form of content, in which there is neither parallelism nor representation but ‘a manner in which expressions are inserted into contents, (…) in which signs are at work in things themselves just as things extend into or are deployed through signs’ (ATP: 96). Thus, with the opposition between segmentary and centralized, which must be replaced by a distinction between two different but inseparable segmentations, the supple and the rigid: ‘they overlap, they are entangled’ (ATP: 231, 234). Thus, lastly, with smooth (nomadic, war-machinic) and striated (sedentary, state-like) spaces, whose difference is said to be complex both because ‘the successive terms of the oppositions fail to coincide entirely’ — that is, smooth versus striated is not exactly the same thing as nomadic versus sedentary etc. — and because ‘the two spaces in fact exist only in mixture’ (ATP:524). To summarize, soon after distinguishing two poles, processes or tendencies, the Deuleuzian analysis, on the one hand, unfolds the polarity into further polarities, asymmetrically embedded in the first (bringing about thus a ‘mixture’ de jure), and on the other, it indicates the de facto mixture of the initial poles. And the typical conclusion is: ‘All of this happens at the same time’ (ATP: 246).

(21) Anthropologists are in general much given to this type of knee-jerk deconstruction. See Rival 1998 and Rumsey 2001 for two pertinent examples: both authors protest against a supposed great divide between The West = arborescence and The Rest = rhizome. These two critics show a certain naiveté as they imagine a certain naiveté on the part of the criticized, who knew perfectly well what they were (not) doing: ‘[W]e are on the wrong track with all these geographic distributions. An impasse. So much the better.’ (ATP: 22).

(22) This pattern appears early in the Deleuzian corpus: see his comments on the Bergsonian division between duration and space, which cannot be simply defined as a difference in nature: the division is rather between duration, which support and conveys all the differences in nature, while space presents only differences in degree. ‘There is thus no difference in nature between the two halves of the division: the difference in nature is wholly on one side’ (Deleuze 1966: 23). It is as if each pole would ‘apprehend’ its relation to the other according to its own nature; or said otherwise, as if the relation between the poles belonged necessarily and alternatively to the regime of either pole, either the regime of contradiction of that of the line of flight (ATP: 238), which organize their mutual differences very differently. In any case, that relation cannot be traced from the outside, from a third encompassing pole. Perspectivism (duality as multiplicity) is what dialectics (duality as unity) must negate in order to impose itself as universal law.

(23) Wagner qualifies the reciprocal co-production between cultural convention and invention as ‘dialectical’ (1981: 52; the term is widely used in Wagner 1986), which may confuse a Deleuzian reader. Yet the characterisation of this dialectics, besides being explicitly non-Hegelian, makes it very evocative of the Deleuzian reciprocal presupposition and disjunctive synthesis: ‘a tension or dialogue-like alternation between two conceptions or viewpoints that are simultaneously contradictory and supportive of each other’ (Wagner 1981: 52). ‘Dialectics’ without resolution or conciliation, in short.

(24) In the Melanesian gender-kinship model, ‘each relation can come only from the other’, and ‘...conjugal and parent-child relations are metaphors for one another, and hence a source of internal reflection’ (Strathern 2001: 240). In the same paper we find the remark: ‘cross-sex relations both alternate with same-sex relations, and contain an inherent premise of alternation within’ (op.cit.:227). A fine example of reciprocal asymmetric pressuposition.

(25) In their approach to the smooth/striated contrast, the authors make the same methodological point: although in reciprocal presupposition, ‘the two spaces do not communicate with each other in the same way [...] the principles of mixture [...] are not at all symmetrical’; the passage from the smooth to the striated and vice-versa are ‘entirely different movements’ (ATP: 524).

(26) The same strategy of evoking one dualism only in order to challenge another is employed, for example, by Latour in his counter-critical booklet on ‘factishes’: “The double repertoire of the Moderns does not reside in their distinction of facts from fetishes, but, rather, in the […] subtler distinction between the theoretical separation of facts from fetishes, on the one hand, and an entirely different practice, on the other hand’ (Latour 1996: 42-43). And then: ‘The choice proposed by the Moderns is not one between realism and constructivism; it rather an alternative between this choice itself and a form of practical existence which cannot understand neither the terms nor the importance of such choice’. This is an apt illustration of the Deleuzian concept of disjunctive synthesis: the meta-relation between exclusive (‘the choice’) and inclusive disjunctions is itself an exclusive disjunction, from the standpoint of the former (‘you must choose!’), and and inclusive one, from the standpoint of the latter (“what are you talking about?”).

(27) The effective antagonism between the two kinship conceptions is somewhat debatable (Schneider 1965, 1984).

(28) Also: ‘In reality, global persons — even the very form of persons — do not exist prior to the prohibitions that weigh on them and constitute them’ (AŒ: 70). See Adler & Cartry 1971: 7, for one probable source of the argument.

(29) The difference made in AŒ between Mauss and Nietzsche is perhaps a little overstated. Firstly, ‘exchange’ versus ‘debt’ is not as clearcut a conceptual distinction as the authors (and, earlier, Deleuze 1962: 155) make it appear to be. Secondly, the Nietzschean theory of the proto-historical repression of a ‘biological memory’ necessary to create a ‘social memory’ is also not that obviously antipodal to the nature/culture anthropogenetic paradigm shared by Maussian or structuralist theories of exchange. It is only when Deleuze and Guattari determine becoming as an antimemory, in their second book (ATP: 324), that the terms of problem could be said to change radically. The Mauss/Nietzsche contrast drawn in AŒ has a very complex polemic background which we cannot enter into here. It involves Hegel, Kojève, Bataille, the Collège de Sociologie, and more proximally Lévi-Strauss, Lacan, Baudrillard — among others. The ‘generalised economy’ derived by Bataille from a Nietzschean reading of The Gift is practically not mentioned (but see AŒ: 190). Deleuze and Guattari’s utter contempt for the Bataillean category of transgression (the remark comes from Lyotard) may be involved here. In the essay on Klossowski included in Logic of Sense, however, Deleuze establishes an insightful contrast between exchange, generality (equivalence) and false repetition, on the one hand, and gift-giving, singularity (difference) and true repetition, on the other hand; the contrast, which anticipates much of the theses of AŒ, is partially and somewhat ambiguously connected, via Klossowksi, to Bataille. Deleuze writes that Théodore, the hero of one of Klossowski's novels, ‘knows that true repetition lies in the gift, in an economy of the gift which is the opposite of a mercantile economy of exchange (...homage to Georges Bataille)’ (D. 1969: 334; suspension point in the original).

(30) See AŒ: 74-75, 109-15. The problem is indicated by the authors themselves in the preface to the Italian edition of ATP: ‘Anti-Œdipus had a Kantian ambition...’ (D. & G. 2003: 289). It is not preposterous to conjecture that the authors of ATP would agree, at variance with the authors of AŒ, that any philosophico-anthropological disquisition on the distinctiveness of the human species, regardless of what is to serve as a sign or cause of its election (or curse) — neoteny, language, tool-making, an immortal soul, the incest taboo, a higher-order intentionality and what not — is irremediably committed to Œdipus. The purpose of contemporary anthropology can no longer be that of finding what makes humans ‘different’ from ‘the rest of nature’. (Humans as the West and non-humans as the Rest? Quite so.) Insofar as Nature is concerned, I am afraid it's a case of ‘same difference’. Anthropologists would be better employed in studying the different differences humans are capable of making; the divide between humans and non-humans is not the only one possible, and certainly not a particularly Great one.

(31) One should notice then that the Deleuzian attack on any form of genealogical extensionism does not imply an ‘anti-biological’ stance: ‘It serves no purpose to recall that genealogical filiation is social rather than biological, for it is necessarily biosocial inasmuch as it is inscribed on the cosmic egg of the full body of the earth’ (AŒ: 154). Needless to say, this cosmic egg is quite a different animal from the selfish gene.

(32) The already-mentioned article by Cartry & Adler (1971) on the Dogon origin myth is the main stimulus behind the strategic importance of these materials in AŒ, being cited at crucial junctures of the analysis. The two anthropologists, along with A. Zempléni, read and ‘corrected’ the third chapter of AŒ (cf. Nadaud 2004: 20-21, citing Guattari's correspondence with Deleuze). At the same time, Deleuze and Guattari's general anti-Œdipal thesis was quite obviously a determining influence on Cartry & Adler's text (see op.cit.: 37 n.1).

(33) Or perhaps ‘the myth of AntiŒdipus’. The contrast between expressive and productive conceptions of the unconscious had led the authors to put the flippant question in p. 57: ‘why return to myth?’, referring to the psychoanalitic hijacking of the Œdipus story. A hundred pages later, in the discussion of the Dogon materials, the question receives, as it were, an unexpected answer. The reference to the Pale Fox myth in support of the authors’ arguments is justified by the tantalizing remark: ‘only myth can determine the intensive conditions of the system (the system of production included) in conformity with indigenous thought and practice... [myth] does not express but conditions’ (AŒ: 157; cf. Cartry & Adler op.cit.: 16). These two contrasting usages of the concept of myth in AŒ need a much closer examination than I would be able to provide here.

(34) See also p. 187 for a typically structuralist remark: ‘When considering kinship structures, it is difficult not to proceed as though the alliances derived from the lines of filiation and their relationships, although the lateral alliances and the blocks of debt condition the extended filiations in the system in extension, and not the opposite.’

(35) What would be a ‘mythical expression’, though, since myth ‘does not express but conditions’? Here is where the two usages of ‘myth’ (see note 32) meet, somewhat uncomfortably.

(36) The Lévi-Straussian theory of marriage exchange remains, after all, a much better anthropological proposition than the juralist metaphysics of group-descent it pre-empted. In a sense, The Elementary Structures of Kinship was the first Anti-Œdipus, a radical break with the family-centered, parenthood-dominated image of kinship; or, to put it differently, the complex relationship of Anti-Œdipus to The Elementary Structures is anamorphically analogous to the relationship of the latter to Totem and Taboo.

(37) See R. Wagner’s analysis (1977) of matrimonial exchange for the Melanesian Daribi: the wife-giver patrilineal clan sees the women it provides as an efferent flow of its own male substance; but the wife-taker clan will see the afferent flow as constituted by female substance. At the moment the matrimonial prestations start moving in the opposite direction, perspectives are inverted. The author concludes: ‘What might be described as exchange or reciprocity is in fact a ... intermeshing of two views of a single thing’ (op. cit.: 628). See Gell 1999: 67-68 for a detailed exposition of this idea as developed by M. Strathern, the anthropologist who made the decisive move of defining every exchange as an exchange of perspectives, thereby completely changing the rules of the whole conceptual language-game of ‘exchange’ within anthropology.

(38) Becoming is of course a central Deleuzian concern ever since his essays on Bergson and Nietzsche, not to mention the Logic of Sense (D. 1969). Starting with the joint essay on Kafka (D.&G. 1975), however, ‘becoming’ acquires a specific conceptual consistency that will be fully deployed in ATP: becoming is that which escapes both mimesis (imitation, reproduction) and ‘memesis’ (memory, history).

(39) Note, however, that besides the importance he attributes to the intensive (morphodynamical: Petitot 1988) ‘canonical formula’ in his mythological work (Lévi-Strauss 1985), the founder of anthropological structuralism explicitly recognizes the limitations of the relational vocabulary of extensional logics to account for the transformations that take place in/between myths (id. 1971: 567-68). Moreover, if The Elementary Structures of Kinship can be said to deal with objects resembling, well, structures, the mythic analyses consolidated in the Mythologiques tetralogy — which develops a whole ‘theory of primitive codes’ (AŒ: 185) — and the sequel books are far more evocative of rhizomatic mappings than they are of structural tracings. The relations that constitute Amerindian mythic narratives, rather than forming combinatorial totalities in discrete arborescent logical distribution, in concomitant variation with and dialectical relation to socio-ethnographic realia, appear to instantiate to the point of explicitness the very rhizomatic principles of ‘connection and heterogeneity’, ‘multiplicity’, ‘asygnifying rupture’ and ‘cartography’ that Deleuze and Guattari counterpose to structural models (ATP: 7-13). There is a whole case to be built for this affinity between late, i.e. non-sociologistic, Lévi-Straussian structuralism and Deleuzian differential analytics; in Maniglier 2006 one can glean up some essential elements for the task.

(40) In this sense, both ressemblances and correlations are Œdipal, ‘mythic’ in the negative sense; see AŒ: 83.

(41) In ATP: 262. The Amerindian myths studied by Lévi-Strauss, very far from displaying any ‘indifferentiation’ or originary identification between humans and non-humans, as it is usually formulated, are defined by a regime of infinite difference, albeit (or because) internal to each persona or agent, in contrast to the finite and external differences constituting the species and qualities of our contemporary world (Viveiros de Castro 2001). This explains the regime of qualitative multiplicity proper to myth: the question of knowing whether the mythic jaguar, to stay with this animal, is a block of human affects in the shape of a jaguar or a block of feline affects in the shape of a human is in any rigorous sense undecidable, since mythic transformations describe an intensive superposition of heterogenic states, not an extensive transposition of two homogenic states .

(42) Emphasis removed. The “n sexes” should be written “n-1 sexes” — unity being represented by the male sex which must be subtracted in order for sexuality to reach the state of multiplicity —, just as the n species of animal-becomings are really n-1 species (1 = humankind).

(43) Incest (or what amounts to the same thing, its prohibition) is, in turn, only the retroactive effect of repressive alliance on repressed germinal filiation (AŒ: 164-66).

(44) ‘A becoming is neither one nor two, nor the relation of the two: it is the in-between, the border or line of flight or descent running perpendicular to both’ (ATP: 323; emphasis added).

(45) The reference of course is to ‘Rethinking anthropology’, in which Leach establishes the well-known connection between affinity and ‘metaphysical influence’ (1961: 20). See Viveiros de Castro 2006 for a recent comment on this correlation.

(46) ‘The gift economy, then, is a debt economy. The aim of a transactor in such an economy is to acquire as many gift-debtors as he possibly can, and not to maximize profit, as it is in a commodity economy. What a gift transactor desires is the personal relationships that the exchange of gifts creates, and not the things themselves’ (Gregory 1982: 19; my emphasis). It might be amusing to see what would happen if we gave the verb ‘desire’ in this passage a strict Deleuzo-Guattarian sense. On exchange and perspective, see: Strathern 1988: 230, 271, 327; 1991a: passim; 1992a: 96-100; 1999: 249-56; Munn 1992 [1986]: 16. On double capture, see Deleuze & Parnet 1996: 7-9; Stengers 1996:64 n. 11.

(47) ‘Language can work against the user of it. [...] Sociality is frequently understood as implying sociability, reciprocity as altruism and relationship as solidarity [...]’ (Strathern 1999: 18). ‘An action upon an action’ is Foucault's definition of power, which is close to the Nietzschean ontology of force (a force always refers to another force, not to an ‘exterior’ object; Deleuze 1962: 7-ff). And ‘a reaction to a reaction’ is how Bateson defined schismogenesis.

(48) As an example of inversion, see the Araweté spouse-swapping ritual friends, who are ‘anti-affines’ without thereby being ‘consanguines’ (Viveiros de Castro 1992: 167-178).

(49) It should be stressed that this affinization of ‘others’ occurs in spite of the fact that the vast majority of actual matrimonial alliances take place within the local group. And at any rate, such alliances cannot but accumulate in the local group, since their concentration defines what a ‘local group’ is. By this last remark I mean to imply that the situation does not change much when we consider those regimes that feature village or descent group exogamy, such as those prevailing among the Tukanoan and Arawakan peoples of Northwest Amazonia. Potential affinity and its cosmological attendants continue to mark the generic relations with non-allied groups, enemies, animals, and spirits.

(50) Amerindian myths, of course, also feature ‘Œdipal’ motives, and parental figures, and filio-parental conflicts. Lévi-Strauss even alluded, not without irony, to a Jivaroan Totem and Taboo (1985: Ch. XVI). But it is quite obvious that for him the mythology of the continent, and above all the myths dealing with the origin of culture, revolve around affinity and exchange, not parenthood and procreation.

(51) There are mythological complexes in Amazonia that do feature a pre-cosmological scenario very similar to the ‘intensive filiation’ discerned by Deleuze and Guattari in the Dogon case; Nortwerstern Amazonian origin myths, in particular, must mentioned here (Hugh-Jones 1979, 1993). These myths, however, articulate the same meta-schema of potential affinity as constituting the ontological base-state (Andrello 2006).

(52) Thus AŒ p. 188 repeats the materialist-historical cliché of the 70's to the effect that Maussian-structuralist ethnology is ‘burdened’ by ’the reduction social reproduction to the sphere of circulation’.

(53) See Donzelot 1977, Lyotard 1977, Zourabichvili 2003: 48-51, for three excellent assessments of AŒ's relation to Marxist conceptuality.

(54) If ‘[t]he expression “difference of intensity” is a tautology’ (D. 1968: 287), ‘becoming-other’ is another, or perhaps the same, tautology.


ADLER, A., & CARTRY, M. (1971). La transgression et sa dérision. L'Homme, 11(3): 5-63.

ANDRELLO, G. (2006). Cidade do índio. Transformações e cotidiano em Iauaretê. Sãõ Paulo: Edunesp/ISA/NuTI.

CLASTRES, P. (1977). Archéologie de la violence: la guerre dans les sociétes primitives. Libre, 1: 137-173.

DeLANDA, M. (2002). Intensive science and virtual philosophy. London: Continuum.

DeLANDA, M. (2003). 1000 Years of War: CTheory interview with Manuel DeLanda (with D. Ihde, C.B. Jensen, J.J. Jorgensen, S. Mallavarapu, E. Mendieta, J. Mix, J. Protevi, E. Selinger). CTheory, a127. <www.ctheory.net/articles.aspx?id=383>

DELEUZE, G. (1962). Nietzsche et la philosophie. Paris: P.U.F.

DELEUZE, G. (1966). Le bergsonisme. Paris: P.U.F.

DELEUZE, G. (1968). Différence et répétition. Paris: P.U.F.

DELEUZE, G. (1969). Logique du sens. Paris: Minuit.

DELEUZE, G. (1993). Critique et clinique. Paris: Minuit.

DELEUZE, G. (2002 [1972]). A quoi reconnaît-on le structuralisme?. In L'Ile déserte et autres textes. Textes et entretiens 1953-1974 (D. Lapoujade, Ed.). (pp. 238-269). Paris: Minuit.

DELEUZE, G., & GUATTARI, F. (1972). L'Anti-Œdipe. Capitalisme et schizophrénie. Paris: Minuit.

DELEUZE, G., & GUATTARI, F. (1981). Milles plateaux. Capitalisme et schizophrénie. Paris: Minuit.

DELEUZE, G., & GUATTARI, F. (1983). Anti-Œdipus. Capitalism and Schizophrenia. Preface by Michel Foucault. Translated by R. Hurley, M. Seem, H. Lane. University of Minnesota Press, Minneapolis.

DELEUZE, G., & GUATTARI, F. (2004 [1988]). A thousand plateaus. Capitalism and Schizophrenia. Translated by B. Massumi. Continuum, London.

DELEUZE, G., & GUATTARI, F. (1991). Qu'est-ce que la philosophie? Paris: Minuit.

DELEUZE, G., & GUATTARI, F. (2003 [1987]). Préface pour l'édition italienne de Mille Plateaux. In G. Deleuze, Deux régimes de fous. Textes et entretiens 1975-1995 (D. Lapoujade, Ed.). (pp. 288-290). Paris: Minuit.

DELEUZE, G., & PARNET, C. (1996 [1977]). Dialogues. Paris: Flammarion.

DONZELOT, J. (1977). An anti-sociology. Semiotext(e), II(3): 27-44.

DUMONT, L. (1971). Introduction à deux théories d'anthropologie sociale. Groupes de filiation et alliance de mariage. Paris: Mouton.

FORTES, M. (1969). Kinship and the social order: the legacy of Lewis Henry Morgan. London: Routledge & Kegan Paul.

FORTES, M. (1983). Rules and the emergence of society. London: Royal Anthropological Institute of Great Britain and Ireland.

GELL, A. (1999). Strathernograms, or the semiotics of mixed metaphors. In The art of anthropology: essays and diagrams. (pp. 29-75). London: Athlone.

GOLDMAN, M. (2005). Formas do saber e modos do ser: observações Sobre multiplicidade e ontologia no candomblé. Religião e Sociedade, 25 (2): 102-120.

GREGORY, C. (1982). Gifts and commodities. London: Academic Press.

GRIAULE, M., and DIETERLEN, G. (1965). Le renard pâle. Paris: Institut d'Ethnologie.

HUGH-JONES, S. (1979). The palm and the pleiades. Initiation and cosmology in North-west Amazonia. Cambridge: Cambridge University Press.

HUGH-JONES, S. (1993). Clear descent or ambiguous houses? A re-examination of Tukanoan social organization. L'Homme, 126-128: 95-120.

JAMESON, F. (1997). Marxism and dualism in Deleuze. The South Atlantic Quarterly, 96(3): 393-416.

JENSEN, C. B. (2003). Latour and Pickering: post-human perspectives on science, becoming, and normativity. In D. Ihde & E. Selinger (Eds.), Chasing technoscience: matrix for materiality. (pp. 225-240). Bloomington: Indiana University press.

JENSEN, C. B. (2004). A nonhumanist disposition: on performativity, practical ontology, and intervention. Configurations, 12: 229-261.

KWA, C. (2002). Romantic and baroque conceptions of complex wholes in the sciences. In J. Law & A. Mol (Eds.), Complexities. Social studies of knowledge practices. (pp. 23-52). Durham: Duke University Press.

LATOUR, B. (1996). Petite réflexion sur le culte moderne des dieux faitiches. Le Plessis-Robinson: Les Empêcheurs de Penser en Rond.

LATOUR, B. (2005). Reassembling the social. An introduction to Actor-Network theory. Oxford: Oxoford University Press.

LAWLOR, L. (2003). The beginnings of thought: the fundamental experience in Derrida and Deleuze. In P. Patton & J. Protevi (Eds.), Between Deleuze and Derrida. (pp.67-83). London: Continuum.

LEACH, E. R. (1961 [1951]). Rethinking anthropology. In Rethinking anthropology. (pp. 1-27). London: Athlone.

LÉVI-STRAUSS, C. (1958 [1955]). La structure des mythes. In Anthropologie structurale. (pp. 227-255). Paris: Plon

LÉVI-STRAUSS, C. (1962a). Le totémisme aujourd'hui. Paris: Presses Universitaires de France.

LÉVI-STRAUSS, C. (1962b). La pensée sauvage. Paris: Plon.

LÉVI-STRAUSS, C. (1969). The elementary structures of kinship (R. B. Needham, J.H., & J. R. v. Sturmer, Trans.). Boston: Beacon Press.

LÉVI-STRAUSS, C. (1971). L’Homme nu. Mythologiques IV. Paris: Plon.

LÉVI-STRAUSS, C. (1985). La potière jalouse. Paris: Plon.

LÉVI-STRAUSS, C. (1991). Histoire de Lynx. Paris: Plon.

LÉVI-STRAUSS, C., & ERIBON, D. (1988). De près et de loin. Paris: Odile Jacob.

LYOTARD, J.-F. (1977). Energumen capitalism. Semiotext(e), II(3): 11-26.

MANIGLIER, P. (2006). La vie énigmatiques des signes. Saussure et la naissance du structuralisme. Paris: Léo Scheer.

MIMICA, J. (1992 [1988]). Intimations of infinity: the cultural meanings of the Iqwaye couting and number system. Oxford: Berg.

MUNN, N. (1992 [1986]). The fame of Gawa: a symbolic study of value transformation in a Massim (Papua New Guinea) society. Durham: Duke University Press.

NADAUD, S. (2004). Les amours d’une guêpe et d’une orchidée. In Félix Guattari, Écrits pour l’Anti-Œdipe (textes agencés par S. Nadaud) (pp. 7-27). Paris: Éditions Lignes et Manifestes.

PETITOT, J. (1988). Approche morphodynamique de la formule canonique du mythe. L'Homme, 106-107: 24-50.

PIGNARRE, P., & STENGERS, I. (2005). La sorcellerie capitaliste. Pratiques de desenvoûtement. Paris: La Découverte.

PLOTNITSKY, A. (2003). Algebras, geometries, and topologies of the fold: Deleuze,Derrida, and quasi-mathematical thinking, with Leibniz and Mallarmé. In P. Patton, & J. Protevi (Eds.), Between Deleuze and Derrida. (pp. 98-119). London: Continuum.

RIVAL, L. (1998). Trees, from symbols of life and regeneration to political artefacts. In L. Rival (Ed.), The social life of trees. Anthropological perspectives on tree symbolism. (pp. 1-36). Oxford: Berg.

RIVIÈRE, P. (1984). Individual and society in Guiana: a comparative study of Amerindian social organization. Cambridge: Cambridge University Press.

RUMSEY, A. (2001). Tracks, traces, and links to land in aboriginal Australia, New Guinea, and beyond. In A. Rumsey & J. F. Weiner (Eds.), Emplaced myth: space, narrative, and knowledge in Aboriginal Australia and Papua New Guinea. Honolulu: University of Hawai'i Press

SCHNEIDER, D. (1965). Some muddles in the models: or, how the system really works. In M. Banton (Ed.), The relevance of models for social anthropology. (pp. 25-85). London: Tavistock.

SCHNEIDER, D. M. (1984). A critique of the study of kinship. Ann Arbor: University of Michigan Press.

STENGERS, I. (1996). Cosmopolitiques, tome 1. La guerre das sciences. Paris: La Découverte/Les Empêcheurs de Penser en Rond.

STENGERS, I. (2002). Penser avec Whitehead. Paris: Seuil.

STRATHERN, M. (1988). The gender of the gift. Problems with women and problems with society in Melanesia. Berkeley: University of California Press.

STRATHERN, M. (1991a). Partial connections. Savage, Md.: Rowman and Littlefield.

STRATHERN, M. (1991b). One man and many men. In M. Godelier, & M. Strathern (Eds.), Big men and great men: personification of power in Melanesia. (pp. 197-214). Cambridge: Cambridge University Press.

STRATHERN, M. (1992a). After nature: English kinship in the late twentieth century. Cambridge: Cambridge University Press.

STRATHERN, M. (1992b). Parts and wholes: refiguring relationships in a post-plural world. In Reproducing the future: anthropology, kinship, and the new reproductive technologies. (pp. 90-116). New York: Routledge.

STRATHERN, M. (1995). The nice thing about culture is that everyone has it. In M. Strathern (Ed.), Shifting contexts: transformations in anthropological knowledge. (pp. 153-176). London and New York: Routledge.

STRATHERN, M. (1996). Cutting the network. Journal of the Royal Anhtropological Institute, NS 2(4): 517-535.

STRATHERN, M. (1999). Property, substance and effect: anthropological essays on persons and things. London: Athlone.

STRATHERN, M. (2001). Same-sex and cross-sex relations: some internal comparisons. In T. Gregor, & D. Tuzin (Eds.), Gender in Amazonia and Melanesia. An exploration of the comparative method. (pp. 221-244). Berkeley: University of California Press.

STRATHERN, M. (2005). Kinship, law and the unexpected. Relatives are always a surprise. Cambridge University Press.

TARDE, G. (1999 [1895]). Œuvres de Gabriel Tarde, volume 1: Monadologie et sociologie. Le Plessis-Robinson: Institut Synthélabo.

VIVEIROS DE CASTRO, E. (1992). From the enemy's point of view. Humanity and divinity in an Amazonian society (C. Howard, Trans.). Chicago: University of Chicago Press.

VIVEIROS DE CASTRO, E. (2001). GUT feelings about Amazonia: potential affinity and the construction of sociality. In L. Rival, & N. Whitehead (Eds.), Beyond the visible and the material: the amerindianization of society in the work of Peter Rivière. (pp. 19-43). Oxford: Oxford University Press.

VIVEIROS DE CASTRO, E. (2004). Perspectival anthropology and the method of controlled equivocation. Tipití, 2(1): 3-22.

VIVEIROS DE CASTRO, E. (2006). The gift and the given: three nano-essays on kinship and magic. In S. Bamford and J. Leach (Eds.), Genealogy beyond kinship: sequence, transmission, and essence in ethnography and social theory. Oxford: Berghahn Books. (in print).

WAGNER, R. (1972). Incest and identity: a critique and theory on the subject of exogamy and incest prohibition. Man, 7(4), 601-613.

WAGNER, R. (1977). Analogic kinship: a Daribi example. American Ethnologist, 4(4), 623-642.

WAGNER, R. (1978). Lethal speech. Daribi myth as symbolic obviation. Ithaca and London: Cornell University Press.

WAGNER, R. (1981). The invention of culture (2nd edition revised and expanded). Chicago: University of Chicago Press.

WAGNER, R. (1986). Symbols that stand for themselves. Chicago: The University of Chicago Press.

WAGNER, R. (1991). The fractal person. In M. Godelier, & M. Strathern (Eds.), Big men and great men: personification of power in Melanesia. (pp. 159-173). Cambridge: Cambridge University Press.

WHITEHEAD, A.N. (1978). Process and reality. Corrected edition, D.R. Griffin & D.W. Sherburne (Eds.). New York and London: The Free Press.

ZOURABICHVILI, F. (2003). Le vocabulaire de Deleuze. Paris: Ellipses.

ZOURABICHVILI, F. (2004a). Deleuze. Une philosophie de l'événement. In F. Zourabichvili, A. Sauvargnargues, & P. Marrati (Eds.), La philosophie de Deleuze. (pp. 1-116). Paris: P.U.F.

ZOURABICHVILI, F. (2004b). Deleuze et la question de la littéralité. Unpublished MS.