Mister D.

This work is licensed under the Creative Commons | © André Iteanu and Ismaël Moya. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau5.1.006


Mister D.

Radical comparison, values, and ethnographic theory


Ismaël MOYA, CNRS, Laboratoire d’ethnologie et de Sociologie Comparative

This article argues that the relevance of Louis Dumont’s work for ethnographic theory today is his radical conception of comparison as an experiment on difference that collapses anthropological analysis and epistemology. The text applies Dumont’s own method— comparison—to his anthropology. In the first part, we follow the trail of Dumont’s ethnographical encounter with the Indian caste system and the radical contrast he drew with Euro-America to provide an insight into his comparative method and his core notions (value, hierarchy, encompassment). In the second part, Dumont’s anthropological strategy is put into perspective with two other radical comparative projects: Marilyn Strathern’s on Melanesia and Eduardo Viveiros de Castro’s on Amazonia.

Keywords: comparison, value, hierarchy, Dumont, Strathern, perspectivism

I don’t have any ideas, comparison provides them.
(Dumont 1991: 8)

In a fascinating book, Pierre Hadot (2004) argues that in ancient Greece, philosophy was much more than an intellectual exercise: it was a way of life.1 Those who practiced philosophy were not isolated individuals responding to one another. They lived in distinct small communities (the schools) characterized not only by a [114]peculiar system of thoughts, but by a global conception of life as well, ruled by specific norms, prohibitions, rituals, bodily exercises, and so forth. For an anthropologist, that ways of thinking matched existential choices in ancient Greece is hardly surprising. It is even a relief to learn that, unlike today, philosophy was not always a sheer intellectual exercise geared to devise abstract conceptual systems or doctrines to explain the universe around us.

Likewise, in line with Hadot’s argument, it may be worth considering that our varied anthropological theories could be associated with contrasted ways of life. At least, this is the impression that those who knew Louis Dumont had, because his conduct was often disconcerting. For example, when some student or colleague addressed him as “Professor”, he often replied: “I am a researcher, not a professor.” This answer was not sheer bashfulness, but a reflection on the nature of anthropology. For him, a professor was someone certain of the truth of what he or she taught and able to comment on all anthropological currents and their relation to philosophy. In short, a person who uses well-defined concepts and talks like a textbook. By contrast, Dumont (1977: viii) argues that anthropological research should be equated to the work of a craftsman who continually re-works ethnographic materials and only reaches provisional results, permanently relativized by the endless richness of social differences.

Dumont’s institutional position matched his personal thoughts. He hardly ever taught a single course. At the École des Hautes Etudes en Sciences Sociales, where he officiated for three decades, nobody did, and allegedly they still do not. The École has no professors but only Directeurs d’études, “Directors of studies,” in the sense of advisors for those who are developing case studies. Dumont ran a weekly seminar presenting his current research to Ph.D. students and colleagues. He did not speak with ease and would rather spell out in front of a dozen students a few pages of a book he had recently encountered or a particular ethnographic point. His presentations were far less spectacular than those of his brilliant colleagues (Lévi-Strauss, Foucault, Barthes, Bourdieu, etc.) who ran seminars in Paris during the same years. In Dumont’s age, professors, rather than researchers, were fairly rare in anthropology. Nowadays, as the institutional situation has changed, they seem to be the overwhelming majority.

What kind of theory, in Hadot’s terms, would match Dumont’s way of life? To answer this question we would need to vet over five decades of anthropological literature which have commented on and criticized abstract notions held as “Dumontian,” such as holism, totality, encompassment, hierarchy, value or ideology,2 to realize that we could add very little.3 However, the purpose of this [115]special section is to highlight that Dumont’s principal intellectual legacy and his main achievement lie less in a theory or a set of concepts and much more in his comparative practice, which he considered the source of all anthropological intuitions. Therefore this introduction aims to unfold Dumont’s radical conception of comparison, an issue largely neglected by his critics and commentators and immediately relevant to the contributions assembled in this section.

Dumont’s comparative project

Louis Dumont was primarily known as a specialist of India and more precisely of the caste system. The contrast he drew between India and Euro-America (Dumont 1980) fueled the anthropological conversation for several decades, especially since the early 1960s. Almost every aspect of his views on India has been vividly debated and challenged by numerous authors (e.g., Marriott 1969, 1976; Leach 1971; Richards and Nicholas 1976; Parry 1979, 1998; Madan 1982; Appadurai 1986, 1988; Béteille 1986; Dirks 1993; Needham 1987; Raheja 1988). He answered some of these critics, particularly those who raised issues concerning caste hierarchy (Dumont 1971, 1980: xi–xlix) and Dravidian kinship (Dumont 1983: 145–71). Yet a misconception still circulates today among many anthropologists, triggered by poor textbooks, of Dumont as the thinker of a grand narrative that reifies the distinction between India and the West, ignores colonialism, and justifies social inequalities. On the contrary, Dumont’s comparison does not aim to intensify the differences that distinguish social formations or civilizations, but repeatedly diminishes the contrasts around which comparison revolves.

In the second half of his life, Dumont moved away from his work on India and strove to understand Euro-American societies and the various facets of the totalitarian regimes they produced during the twentieth century. His work dealt with economic ideology (1977), individualism (1986), and German ideology (1994). Until the end of his life, Dumont never lost touch with the new questions anthropology was confronting. His last two published pieces of work were critiques of the colonial situation in New Caledonia and of the ecological disaster created by the drying of the Aral Sea.

On comparison, as on everything he considered essential, Dumont never produced a methodological text or a systematic formalization since he never ceased to redefine the conceptual repertoire he employed and to refine his analysis and methods. In each new edition of of his books, he added extra chapters, paragraphs, forewords, postscripts, personal commentaries, and notes. Each published work [116]was a step in a research process. What he feared most is that his conceptual framework would become an autopoietic system which might prevent him exploring new conceptual domains. His doubts never left him, and in the final pages of his last book, German ideology (1994), he claimed that the opposition between individualism and holism, which was his big scientific achievement, did not properly fit the German case (ibid.: 194) and therefore was only a heuristic device fit for certain comparative contexts.

Comparison was for Dumont the sole heuristic capable of generating anthropological knowledge. For this reason, he didn’t regard his findings as personal achievements, but saw them as the products of his comparative experiments. We wish to argue that the relevance of Dumont for ethnographic theory (da Col and Graeber 2011) today can be acknowledged from the standpoint of his comparative perspective and by reflexively applying to it its own method, comparison. We shall thus follow the trail of his ethnographical encounter with India and then compare his perspective with two other radical comparative projects: Marilyn Strathern’s on Melanesia and Eduardo Viveiros de Castro’s on Amazonia.4

A man without ideas: The comparative experiment

Dumont’s anthropology did not unfold from a philosophical standpoint, but from ethnographic circumstances: in ancient India, the Brahmans, the priests, were considered superior to the Kings. This superiority was expressed in terms of the relative purity of their castes, that is, in terms of a religious value. It was visible in many ways. Above all, the Brahmans were vegetarians, which made them purer than the carnivorous Kings. However, this subordination did not affect the political power of the latter, who reigned over everyone. This situation triggered Dumont’s curiosity. How could we make sense of the fact that political power was not paramount in ancient India?

To answer this question, Homo hierarchicus (1980) argues that in India, the value of purity was more crucial than political power. Therefore, the Brahmans were superior to the Kings not because they were more “powerful,” but because they were born in castes endowed with the highest purity and status. This social reality conformed to a general principle as all castes are hierarchically ordered according to their relative purity. Since Brahmans are superior to Kings, Dumont concluded that in India caste hierarchy ranked according to purity is distinguished from and more valued than the political power of the Kings that it encompasses.

[117]However, Dumont did not consider the caste system as an empirical reality. In most rural areas only a few castes are present and nowhere in India could one find something like a complete set comprised of all the castes. The caste system was not even an explicit and articulated conception as no Indian regards castes as a coherent string of groups clearly ordered along a line of relative purity. However, the caste system manifests itself in that Indians commonly hierarchize things, people, institutions, and so forth, in terms of their relative purity. Dumont (1980: 34) accordingly stated that, “far more than a ‘group’ in the ordinary sense, the caste is a state of mind which is expressed by the emergence, in various situations, of groups of various orders generally called ‘castes’.” Consequently, a single caste is not an independent social reality. It cannot be understood in itself, but only in relation to other castes which are locally superior or inferior. The minimal unity is therefore not the caste as a basic unit, but a series, ordered in a hierarchy. Dumont generalized this conclusion by stating that a whole cannot be constructed by adding elements. On the contrary, each element must be evaluated in relation to its belonging whole. In other words, to understand the caste system, one needs to adopt the perspective of the whole, in which “certain fixed principles govern the arrangement of fluid and fluctuating ‘elements’” (ibid.).

This initial comparison was only the starting point of Dumont’s work. Comparison is radical, according to him, because it is an experiment that brings into play anthropologists’ own ideas (Dumont 1986: 6).5 Dumont struggled to understand the superiority of the Brahman over the King, precisely because this fact elicited a difference with the idea and practices of his cultural milieu. Compared to India, where hierarchy appeared as unmixed with power, in Euro-America power and hierarchy seemed inseparably mingled. This is, for example, the case all through Euro-America when hierarchical relations are construed as domination. This state of mind was particularly acute, Dumont noted, in the Marxist trends that led French social sciences at the end of the 1960s, mostly under the influence of Althusser, Foucault, and Bourdieu. These authors, according to Dumont, treated the compound made by power and hierarchy as the natural force underlying every social phenomenon. In contrast, for Dumont, the mingling of hierarchy and power was not an objective universal reality, but, seen from India, looked to be a most particular Euro-American idea.

[118]Furthermore, since in India power appeared as limited by caste hierarchy, Dumont mused on whether in Euro-America power was actually as powerful as commonly imagined. This is why in his later work on Western ideology (1977, 1986, 1994), he attempted to understand the limits of power: that is, the constraints that more or less invisibly weigh on political power. One obvious such limitation is constituted by the economy. Following Karl Polanyi’s demonstration of the disembedding of the market economy from society in the nineteenth century (Polanyi 1944), Dumont captured in From Mandeville to Marx (1977) the way in which the economy has gained primacy in Euro-American thought. As Ismaël Moya remarks in his contribution to this special section, one of Dumont’s conclusions was that the “economy” does not refer to any objective domain of social life: it is a construct and not a crosscultural category, contra Polanyi’s (1957) idea of a universal “substantive economy.” According to Dumont, since economics emerged from politics in Euro-America, the two cannot be understood independently from one another. Their reality is only constituted by their relation. However, the autonomy and the claim to primacy of the economic dimension impinges on the political and challenges its power (Dumont 1986: 104–12).

Dumont’s comparative work on India elicited two radically opposed stances on sociality: the Indian social formation manifests the idea that hierarchy is an utmost value that encompasses political power, while Euro-American social life is molded by the contention that power is universal and the source of all hierarchy. Indeed, Dumont thought that anthropology was not to decide which of these alternatives was more “objective.” As both of them were idiosyncratic ideas, he called them values from a comparative perspective. Thus, Dumont’s notion of value is distinct from the economic notions of value and those of moral obligation and subjective judgment informed by a notion of the good and the desirable. This distinction is crucial and is why Dumont’s anthropology of value has very little to do with the recent anthropological current loosely called the “ethical turn”6 or David Graeber’s theory of value (2001),7 although both of them consider that their main concern is the comparative appraisal of values.8

Dumont’s values are neither objective nor subjective facts, but only differences that appear through comparison. Consequently, it is futile to attempt to specify whether they are descriptive notions or artifacts (i.e., the products of a comparative experiment). They overcome the distinction between symbolic and real, [119]representation and reality, action and thought, facts and values. Therefore, values never built up into an overarching standpoint, such, for example, as that afforded by power, from which one could embrace equally all social formations and picture all hierarchical relations as forms of domination. His radical perspective led Dumont to dispense with all conventional anthropological themes like ritual, religion, or ontology and to consider that values only emerge from comparative experiment. Hence his formula: “I don’t have any ideas, comparison provides them” (Dumont 1991: 8).

Nevertheless, let us pause for a moment and focus on Dumont’s assertion, counterintuitive as it is for all those who have gone through an Euro-American style of schooling in which scientific knowledge requires a precise and stable definition of the conceptual apparatus. When ideas develop from comparison only, the concepts deployed are bound to vary in time to express whatever gradually emerges from the contrast. The distinction between power and hierarchy illustrates this movement. The comparison between India and Euro-America led Dumont to disentangle these two notions to propose an intelligible contrast. However, in the two cases, the notions of power and hierarchy are not identical, precisely because in each case their relation is different. However, in both cases, the relation (between power and hierarchy) defines their meaning. In Dumont’s view, in this case, the term value designate synthetically both the relation and the terms elicited by a comparison.9

From such an epistemological position, each value description (i.e., each difference elicited by comparison) is only an analytical starting point and never accounts for the “whole.” Thus, value descriptions inevitably leave blank spots unaccounted for. This is partially why the perception of a particular value changes when observed from a different perspective. For example, on the one hand, Dumont’s comparison contrasts hierarchy in India with power in Euro-America, but, on the other, he states that purity (and not power) orders the caste system. That purity and hierarchy are different points of view on the same difference (i.e., the same value) ensues from the fact that although these terms describe comparative wholes, our appraisal of them is irremediably limited. Analysis must unfold from the relative whole to its parts. In other words, the wider contrast between two social formations constitutes the starting point. Dumont (1980: 212) confessed that he made a bet when he first decided to center his analysis on the distinction between hierarchy and political power, as he was not sure that it would work. The difference always appears at the beginning to be very radical: intensified to a point where it seems in many ways unrealistic. However, progressively, further analysis refines and relativizes this initial difference by drawing on contradictions since, according to Dumont, a value can be never homogeneously dominant. In certain domains, for certain people, each value may be subordinated to other values, which in turn [120]can be more locally subordinated to others.10 For example, hierarchy distinguishes India from Euro-America, but is by no means coextensive with it and does not account for every aspect of Indian society. On the contrary, for Dumont, Indian hierarchy leaves space to the power of Kings that it only encompasses (Dumont 1980: 152–83). Likewise, caste hierarchy is contradicted by the extreme individualistic stand of the renouncer (ibid.: 184–87), a mystic who steps outside his caste to invent or adopt a “discipline of salvation.” In other words, the renouncer relativizes the caste system by being an “individual-outside-the-world.”

As the Indian example shows, a value can never be dominant in all the domains of social life, but conversely values are not globally devalued by the fact that they are at times subordinated or contested. Thus, contrary to what happens when the relation between two values is a matter of choice (or preference), the hierarchical superiority of one value over another does not rule out that inferior value. In encompassment, even when they contradict superior values, inferior values are included in the configuration at subordinated positions. The recognition of logical contradictions is therefore integral to Dumont’s notion of hierarchy (Dumont 1980; Houseman 1984). However, these logical contradictions may not be experienced as such by the members of the cultures considered when, as a function of their context, the diverse alternatives they confront are referred to different values.

Consonant with his notion of values, Dumont’s comparison is a process that relativizes layer by layer the massive contrast initially established. For example, as stated previously, in India, he first deconstructed his own notion of power by distinguishing power from hierarchy. Then, turning back to Euro-America, he tested the distinction only to find that political power was here associated with equality. Back again to India, a new distinction emerged. Whereas in India hierarchy encompasses political power, in Euro-America the relation of power to equality is an exclusive contradistinction. This last formulation accounts for an idea widely spread in Euro-America and also among Marxists, according to which power is consubstantial to hierarchy and absolutely opposed to equality: it is either one or the other. Or to put it otherwise, if power was somehow to disappear, equality would be achieved. This third layer of comparison allowed Dumont to reach a more abstract distinction, that between hierarchical encompassment and exclusive opposition, which may be deconstructed again by other ethnographic examples. In sum, as stated earlier, Dumont’s comparative method only produces provisional results. Repeated reconsiderations sharpen the image fashioned, but never reach a stabilized position, let alone perfection. Each result is nothing but the starting point of a further step that refines, reformulates, and displaces the difference between the two comparative poles that one decided to contrast.

This practice of anthropology does not incite the development of new concepts, but only draws on old and common categories to elicit, through comparison, differences within them. At first, the radical contrast between two social formations is approached by using the common notions (political power, hierarchy, etc.) of the [121]anthropologist’s language. This contrast is then further refined via a juxtaposition with facts pertaining to other familiar categories, and so forth. As a consequence, the notions used by Dumont fall short of being well-defined concepts. Instead, their meaning fluctuates with every new finding obtained through analysis. Momentarily, comparative results may be generalized and formalized, as, for example, when Dumont (1980: 239–45) developed the idea of hierarchy as “encompassing of the contrary.”11 This gives rise to theory as a byproduct of comparison. However, when one aspires to stabilize this theory, the very possibility of comparison may be jeopardized.

Dame M. and Mister D.

Over the last twenty years, postmodernist, neo-Marxist, Foucauldian, and Bourdieuan approaches, among others, occupied most disciplinary debates. Some proponents of these theories deemed Dumont’s ideas unacceptable, sometimes even evil, mostly because they relativized political power and therefore seem to support all kinds of “domination-ism”: capitalism, orientalism, colonialism, essentialism, and so on. Engaging with them (e.g. Appadurai 1986, 1988) would thus inevitably turn into an endless round of hermeneutic quarrels.12

However, a series of anthropological approaches have also come to the forefront of the discipline and they reveal a set of germane elements to Dumont’s thought. What singles them out is the accusation they share, that is, that an epistemological (Great) Divide has been generated between the West and the rest. The closest and most systematic intellectual enterprise of this kind is Marilyn Strathern’s relational anthropology.13 Like Dumont, Strathern has also been widely criticized for not acknowledging the “relationships of power and inequality, particularly with regard to gender and historical relationships between Melanesia and the West” (Street and Copeman 2014: 8).

In our reading, Strathern’s conceptual premises resemble those of Dumont to a certain point, after which their thought takes a diametrically opposed direction. Both authors wrote extensively on a wide range of subjects and have modified some of their ideas over the years. To render our comparison more telling, we have narrowed its focus to Strathern’s work on Melanesia and Dumont’s on India. These two [122]thinkers’ theoretical apparatuses are very articulate and nuanced, yet the comparison that follows radicalizes their contrast in order to make it explicit.

To begin with, Dumont and Strathern elicit differences between the social formation they study and Euro-America. To do so, they both problematize the opposition between the individual and society (Dumont 1980: 4–19; Strathern 1988: 11–15). For Strathern, exchanges and, more broadly, social life as it is practiced in Melanesia are incompatible with the Euro-American notion of the individual. This is so because in Melanesia the agent is not a closed totality, but a “dividual”14 person formed out of relations. In Strathern’s words, “persons are frequently constructed as the plural and composite site of the relationships that produced them” (ibid.: 13). Dumont makes a similar contention when he deconstructs the Euro-American notion of the individual by separating two of its meanings (an “abstention,” in Holbraad and Pedersen’s terms: 2009: 379): the empirical agent and the normative concept (Dumont 1980: 9). Like Marilyn Strathern, Dumont conceives the Euro-American normative individual as an indivisible element. In parallel with Strathern, Dumont (1965: 91) argues that in India the individual in the normative sense does not exist because

the smallest ontological unit (or normative agent) is that in which order (or hierarchy) is still present, i.e. a pair of higher and lower empirical agents, complementary to each other in a particular situation. This consideration (of the smallest unit, thus comparable to an individual) is not essential in itself, but it may be useful for our understanding of the system.

This Indian multiple conception of the person echoes the consubstantiality between exchange partners that Strathern evidences in Melanesia.

However, from this point on, the two authors bifurcate. Dumont is not terribly interested in the Indian person qua agent. When forced to encounter the topic, Dumont defines it (see above) in the light of the Indian hierarchical system, which is his main concern. Since caste hierarchy characterizes India, the relation between two persons is the minimal unit in which a relation can be manifested. Strathern, on the contrary, makes no hypothesis as to why the Melanesian person is dividual. Her only concern is to draw the systematic consequences of this fact. In short, to paraphrase Viveiros de Castro (2004: 7), Strathern’s crucial inquiry is: “What would a world which has such a definition of the person look like?”

In a second move, turning his back to India, Dumont noted that Euro-Americans and Indians tend to conceive differently the social configuration they live in. Indians refer to it in the guise of a hierarchized caste system, while Euro-Americans describe it as a set of individuals. Dumont’s contention is that the former vision refers to an integrated whole while the latter posits a collection of elements. On the contrary, Strathern (1988: 13) asserts that Euro-Americans define society as a whole that mirrors their totalized idea of the individual, “conceptually distinct from [123]the relations that brings them together.” In sum, both agree that Euro-Americans conceive society as a collection of individuals linked by relations external to them. However, their interpretation of such a common conclusion differs. For Dumont, the Euro-American conception of society refers to a collection of individuals; for Strathern, it invokes a totality. “That the English imagined themselves living between different orders or levels of phenomena, in an incommensurate world of parts and wholes, both created and was itself a precipitate of the manner in which they handled perspective” (Strathern 1992: 87). The contrast is quite striking as Strathern invokes here several notions that Dumont would only use to describe India: orders or levels, incommensurability, parts and wholes. In other words, for Strathern (1988: 12–16), since Melanesia has no individuals, there can be no society in the Euro-American sense of the term. Dumont’s position is symmetrically opposed: since Euro-America has individuals, it can have no totality in the Indian hierarchical sense.

From there on, the two authors face quite different problems. Since, for Dumont, India is an integrated hierarchized whole, he can extract the relative value of each element he studies from the ethnography. In sum, in this context, as we have seen, the value of things, that is, their relative position to other things, is “included” in them. In his seminal study of South Indian Dravidian kinship terminology, Dumont (1953) showed that affinity and consanguinity were treated equally, in contrast to Euro-American kinship systems, where affinity is only present at one generation and then transforms into consanguinity at the next. However, as Cécile Barraud remarks in her contribution to this section, the equality in question here is not that associated with the individual in Euro-America, but a form of equality encompassed in the caste system and thus subordinated to hierarchy. In Dumont’s terms, “South Indian kinship presents us with a contrast … something like an island of equality in an ocean of caste” (Dumont 1983: 167). In this example, as everywhere else, Dumont does not face the same problems of scaling as Strathern, because his analysis operates at once as a twofold movement: comparing India and Euro-America and contextualizing each value within the whole to which it belongs.

A clarification is, however, in order: Strathern ([1991] 2004) stated on several occasions that notions of parts and whole are a heuristic and do not “represent” Melanesian sociality. Rather, her goal is to keep her analysis in between the two poles. Cogently, she dispenses locating the Melanesian relations she engages with (e.g., same-sex and cross-sex) within a global social context. However, this strategy confronts her repeatedly with a problem of scaling that could be formulated as follows: under which condition(s) can one compare elements pertaining to different societies and selected from an external point of view? To answer this question, Strathern (ibid.)—following Wagner (1991)—resorts to the fractal paradigm, which posits that each level of certain objects replicates all the others levels in terms of complexity, irrespective of scale. She thus gains tremendous freedom to compare any two elements, which confrontation produces interesting results. Her work thus expands comparison’s possibilities considerably and elicits unexpected analogies: “an imaginative device through which to think about connections if we could dispense with its attendant presumption of integration taking place within a single entity” (Strathern [1991] 2004: 25). The limit of her method is that the comparative endeavors she produces never add up, but remain pure liberating experiences. By contrast, [124]although Dumont never foresaw any conclusive result to anthropology, he expected comparisons to cumulate toward a better understanding of social life, irrespective of all the unpleasant and even immoral revelations that could emerge from it.

Instead of building up a global unity, Strathern tackles recursively along multiple comparative axes the distinction between Melanesia and Euro-America: gender (i.e., anthropology and feminism), persons and things (i.e., gift and commodities), ownership, reproduction, kinship, and so forth. Each axis intensifies a difference by pushing it to its limit.

The differences that plural comparisons measure “between things” now emerge as constitutive of those very same “things,” and can therefore best be thought of as residing “within them.” This … implies also that the plural distinction between things and the scales that measure them also collapses into itself…. So comparisons are things that act as their own scales—things that scale and thus compare themselves.” (Holbraad and Pedersen, 2009: 375)

Dumont also resorts to a recursive comparative strategy, but places it in a different framework. We illustrated how Dumont initially established a radical contrast between India and Euro-America. Afterwards, contrary to Strathern’s intensification, he recursively relativized this contrast by concentrating on encompassed values. This procedure produces successive purposely unimaginative revolutions, in that each stage partially contradicts the previous one. It also creates totalities: the values initially drawn from comparison become wholes when set in relation to subordinated values. Here again, the unity is neither in the part nor in the whole, but in the hierarchical relation that binds them.

This contrast between the two authors is also manifested in their style. Strathern produces many dazzling images, concepts, displacements, and so forth, whereas Dumont always sticks to ordinary language. As his terminology never perfectly fitted his purpose, he often slides from one problematic metaphor to another (value, hierarchy, wholes, etc.) and thus generates a feeling of inaccuracy.

Strathern and Dumont anchor their anthropological practices on comparative differences and achieve a deconstruction of Euro-American theories of knowledge. Both also collapse the distinction between epistemology and comparison. For Strathern, Euro-American knowledge production is held hostage by parts and wholes discourses. Thus, Euro-American anthropologists, no matter what they do, cannot escape the creation of merographic connections. At best, they can adopt a circumventing strategy by rendering visible their analytical process. As exemplified in Strathern’s work, this produces imaginative and aesthetic results (Strathern 1992).

In contrast, for Dumont, Euro-American equality, compared to Indian hierarchy, elicits individuals conceived as wholes and social formations as sets of individuals. True to his relativizing method, Dumont subsequently attempted to discover in Euro-America whether the individual as a value was contradicted by other values. A first reading of some major political philosophy texts (Hobbes, Locke, Mandeville, Marx, Rousseau, Smith, etc.) appeared to show that this was untrue and that the individual claimed to spread a homogeneous shadow over every sector of the ideology. However, further investigation revealed a more nuanced configuration in which the overarching individual was perpetually and irremediably haunted [125]by its opposite (Dumont 1986: 17). According to Dumont, these non individualistic aspects may: (1) manifest the survival of traditional practices and institutions such as the family or gender inequality; (2) result from the application of the very individualistic principle, such as the limitation of economic liberalism by governments; or (3) be the outcome of the interaction between cultures (Dumont 1994: 8–10).

Dumont thus concluded that in Euro-America, when values contradict the superior individual, they are not encompassed by it. They are present but implicit in the midst of equality. His contention is therefore that, in Euro-America, these hidden values opposed to liberty and equality do not manifest mischievous intentions. They are a structural feature that at time evolves into an intensified form, totalitarianism (Dumont 1977: 12–14, 1986: 149–79). This is the most striking and dramatic conclusion of Dumont’s comparative practice.

Finally, as is well known, where anthropology conventionally pictures Melanesian social configurations as fluid, plural, and devoid of centralized authority—all qualities also associated with Strathern’s work—the caste system is linked to hierarchical and pervasive structures, generating contradictions, all attributes lumped in with Dumont’s orientation. In other words, Strathern’s and Dumont’s anthropologies reveal the differences between Melanesia and India, thus blurring the distinction between the methods and the social formation studied.

Our contrast therefore falls short of being conclusive. To refine and relativize it, we require a third point of view on radical comparison, that of Eduardo Viveiros de Castro’s “perspectivism,” which inspired by extension the broader current that now goes by the name of the “ontological turn.”15

Dumont, dervishes, and others

Viveiros de Castro (1998: 469) once defined perspectivism as the conception “according to which the world is inhabited by different sorts of subjects or persons, human and nonhuman, which apprehend reality from distinct points of view.” In perspectivism as in Strathern’s anthropology, a particular form of personhood plays a central role. Unlike Strathern and Dumont, Viveiros de Castro make use of the Western notion of the person as it is, only inverting the characteristics associated with each of its elements (soul and body) and expanding its points of application: “Individuals of the same species see each other (and each other only) as humans see themselves, endowed with human figures and habits, seeing their bodily and behavioral aspects in the form of human culture” (Viveiros de Castro 2004: 6). Consequently, “where our modern, anthropological multiculturalist ontology is founded on the mutual implication of the unity of nature and the plurality [126]of cultures, the Amerindian conception would suppose a spiritual unity and a corporeal diversity—or, in other words, one ‘culture,’ multiple ‘natures’” (ibid.). In short, Euro-American nature is given and unique while culture is constructed and multiple, and Amerindian nature is multiple and constructed while culture is given and unique. Viveiros de Castro thus preserves several Euro-American conceptual oppositions (e.g., nature and culture, body and soul) but makes them whirl, while the distinction between reality and representation is collapsed. In our view, this might be called a “dervish move.”

According to Viveiros de Castro, Amazonians scrutinize multiple juxtaposed worlds (e.g., jaguars’ or peccaries’). As anthropologists do, they strive—with the help of shamans who act as mediators and translators of worlds—to understand what their rules are. The author thus considers that the world of anthropology stands in a relation of continuity to the Indian multiworlded reality. In effect, anthropology, like “indigenous perspectivism, is a theory of the equivocation, that is, of the referential alterity between homonymic concepts. Equivocation appears thus as the mode of communication par excellence between different perspectival positions—and therefore as both condition of possibility and limit of the anthropological enterprise” (ibid.: 5).

Notably, Viveiros de Castro has acknowledged Dumont’s work more than Strathern’s. In a seminal article on Amazonian kinship, Viveiros de Castro (2001) explicitly draws on Dumont’s idea of Dravidian kinship (Dumont 1983) and develops the idea that affinity as a value is paramount in Amazonia. While agreeing with Dumont (contra Needham) that all classificatory relations carry a value asymmetry, Viveiros de Castro parts from him when it comes to conceptions of totality.16 For Viveiros de Castro (2001: 28), any notion of totality should not be rejected indistinctly:

I am not suggesting that we should shun any notions of totality, as if these were wickedly un-Amazonian, just that we be wary of a fallacy of misplaced wholeness. Any cosmology is by definition total in the sense that it cannot but think everything that is, and think it according to a limited number of fundamental presuppositions: holistic approaches are thus amply justified. But from this it does not follow that every cosmology thinks everything that is under the category of totality—that it poses a totality as the “objective correlative” of its own virtual exhaustiveness. Accordingly, I venture to suggest that in Amazonian cosmologies the whole is not (the) given, and it is not the sum of the given and the constructed either. The “whole” is, rather, the constructed, that which humans strive to bring forth by means of a reduction of the Given as the anti-whole or pure universal relation (difference).

In sum, Viveiros de Castro applies to hierarchy the same dervish move that he previously made with the opposition between nature and culture: he retains the parts [127]and the whole distinction, but the latter is constituted by the nature constructed in Amazonian fashion.

In opposition to Dumont, the notion of totality that Viveiros de Castro advocates does not generate hierarchical relations. The categories he inverts are placed at the same level, structurally speaking: one category does not encompass the other and, as he states, there is no whole to incorporate them all. As a consequence, the reversals that Viveiros de Castro performs in his comparative practice do no generate asymmetric relations but equivalent inversions. According to us, this absence of hierarchy renders these ontological configurations elicited by Viveiros de Castro comparable to a Lévi-Straussian mythology (Viveiros de Castro 2008). In her contribution to this section, Aparecida Vilaça comments on Viveiros de Castro’s appropriation of Dumont and also notes that in Amazonia, the coexistence of values produces no moral conflicts but rather a dualism in perpetual (or dynamic) disequilibrium, for example between compatriot and enemy or consanguine and affine oppositions (Viveiros de Castro 2001: 30). In the perspectivist model, if categories have any value, it stands outside of them. Thus Viveiros de Castro can replicate (recursively) the same operation of inversion (e.g., Viveiros de Castro 2001, 2009) in a move that we see as a conceptual whirl.

In sum, while Viveiros de Castro (and other ontological turnees such as Martin Holbraad and Morten Pedersen) radicalizes comparison, it strikingly differs from Dumont and Strathern’s method. Firstly, Viveiros de Castro deals with ontologically framed questions, which, as Laidlaw (2012) remarked, necessarily rely on a “radical alterity.” Secondly, he generates new radical concepts by “transmuting ethnographic exposures recursively into forms of conceptual creativity and experimentation” (Holbraad, Pedersen, and Viveiros de Castro 2014). Finally, he hyperbolizes differences by reversing logical oppositions (nature–culture, given–construct, affinity–consanguinity etc.) while Dumont relativizes radical differences through the incorporation of encompassed values.

Threesome: Mapping

How does one compare comparisons? In other words, to follow the perspectivist inspiration: How does one put in perspective these three perspectives on perspective? Viveiros de Castro, Strathern, and Dumont all share a common ground. While radically distinguishing Euro-America from the social formation(s) they study, all collapse anthropological analysis and epistemology. Yet they all present distinct forms of comparative methods

For example, Strathern argues that the “things” (i.e., the terms of the relation of differences or, as it were, “what” is compared) contain their relation and their own scaling (Strathern [1991] 2004; Holbraad and Pedersen 2009). On the contrary, for Dumont, the meaning of these “things” emerges when juxtaposing their hierarchical relations to other “things” (see above, the definition of the person in India). For the “ontologists,” or at least for Viveiros de Castro, the meaning of things can only appear through an equivocation such as the one “implied in imagining that when the jaguar says ‘manioc beer’ [i.e., what the human see as blood] he is referring to the same thing as us (i.e., a tasty, nutritious and heady brew)” (Viveiros de Castro 2004: 6)

[128]Each orientation collapses different distinctions. Dumont conflates values and facts under the notion of value, which he deems as the source of a comparison free of any value judgments.17 Viveiros de Castro collapses the distinction between reality and representation in the notion of ontology to maintain the equivalent value (continuity) of every form of existence.18 Finally, Strathern collapses the distinction between parts and wholes to escape Euro-American epistemology and shed light on crucial contemporary debates on kinship, reproductive technologies, notions of exchange and production, and so forth.

Each orientation also generates its own obstacles. Dumont invested in multifarious notions of value and hierarchy—a liberty which has won him many critics. Viveiros de Castro generalized the ontological perspective and thus engaged in a recursive conversation with philosophers (especially Gilles Deleuze) who questioned oppositions such as subject–object, given–constructed, nature–culture, and so on. Strathern is caught in issues of scaling and the ways they generate forms of discontinuity. Consequently, her writings are at times quite abstract and easily drift away from their initial concerns as her analysis unfolds.

Amazonia has often been often compared to Melanesia (cf. Gregor and Tuzin 2001), a resemblance matched by the fact that both Strathern and Viveiros de Castro center their work on the person, while Dumont only pays moderate attention to it. Yet conceptually, the ontological approach would be closer to Dumont than to Strathern, inasmuch as they admit the hierarchical nature of categorical oppositions and the necessity of the concept of totality in certain circumstances. Finally, methodologically, Strathern is closer to Dumont than to Viveiros de Castro in that their analysis originates from comparison but does not engage ontologically with the difference it elicits.

Where Dumont first posits a radical difference then relativizes it, drawing on locally subordinated values, Strathern relativizes the difference by combining it with other comparative questions. Ontologists hold on to the radical difference because they consider it as real as it gets.

For heuristic purposes, we may summarize the distinctions outlined above in a naïve chart, as these seem to be back in style nowadays (see Descola 2013):

  Strathern Dumont Ontologists Power
Radical difference Relationality Value Ontology nil19
Comparison practice Different axis Encompassment Equivocation Domination
Result Postplural Hierarchy Metaphysics Liberation

[129]Dumont believed, quite conventionally, that anthropological knowledge was a cumulative enterprise. Strathern seems to adopt a more “leftist” strategy, or should we say behave as a “contemporary artist,” as her knowledge is consumed in the very act of its production. Finally, Viveiro de Castro and the ontologists pursue political goals: to place indigenous people’s ontologies and the Euro-American philosophical tradition on an equal footing.

Values in motion: Hierarchy, power, and dualism

Since modes of comparison are intimately linked to the social formation regarding which they have been created (India, Melanesia, Amazonia, etc.), it is crucial to put these methods to the test away from their context of origin. The contributions in this section aim to engage with Dumont’s comparative project and method in Melanesia, Africa, Amazonia, and Indonesia. As we outlined above, comparison draws on a radical difference between social formations but never promotes symmetry since no social formation is coextensive with the paramount value that characterizes it. On the contrary, Dumont’s method assumes that no single value can be fully hegemonic: a paramount value always coexists with other values that contradict it. Even individualism always combines with holistic values that contradict it in one way or another (Dumont 1983: 17–19, 1994: 3–16).

In this section, Dumont’s Indian findings are neglected yet his comparative approach is not. Taken together, the contributions provide an understanding of the contradictions between values which aims to relativize the overstated contrast which grounds the comparative enterprise Indeed, Dumont’s comparative project does not only focus on the question of what are the differences in our world(s) but also inquires into their becoming. As Iteanu remarks in his contribution, Dumont mostly dealt with changing value configurations in times of historical transition, whether it was the relationship between castes and Vedic texts in India, the emergence of the economic category in eighteenthand nineteenth-century Euro-America (1977), the history of individualism from Stoicism to Hitler (1986), or the transformation of German ideology between 1770 and 1830 (1994). For change puts values in motion, it reveals the contradictions and relations between them. The main question at stake in the contributions gathered in this section is the nature of the relations between values: Is it hierarchy, power, or dualism?

André Iteanu is concerned with the emergence of new configuration of values in Melanesia. He argues that change is a process of recycling values rather than radical creation. According to him, everything flows in Melanesia: objects as well as persons, villages, rituals or ancestors. Over time, relations wither, giving way to a world without limits and direction. People act to slow down and temporarily stabilize this movement: relations must be reactivated for a time by ritual and especially exchanges. In this sociality, the meaningless movement, especially that of the ancestors, has a subordinated value: it is necessary to establish new relations with the outside from which the objects and rituals essential to sociality are elicited. In this configuration, the coming of the “Whites” and Christianity were synonymous with an unprecedented potential expansion of sociality that [130]initially gave rise to high hopes of creating new fruitful relations of exchanges. People abandoned their rituals, converted to Christianity, and embarked on various modern projects. However, their hopes were progressively dashed. The ancestors migrated abroad. They are now as unresponsive as the politicians in the capital, the Whites, and even God. In spite of the people’s creative efforts (letters to the dead, the reconfiguration of mythology to include relations to the Whites, the proliferation of new churches, etc.), all meaningful relations (those in which objects flows) have withered. In the wake of the arrival of the Whites and Christianity, the value that was previously subordinated now holds sway over sociality. Obsolete or subordinated values are recycled to build the new configurations. However, this process in not a mere inversion: the hierarchy gave way to a configuration in which values (openness and relationality) struggle against each other on the same level.

Joel Robbins’ contribution on hierarchical dynamism analyzes the competing relation between individualistic and holistic values generated by the recent conversion to Christianity in Melanesia, Africa, and Amazonia. Following massive conversions, individualism as a value introduced by Christianity became paramount. However, its full realization is limited by the importance of other values, namely relational ones. According to Robbins, the dynamism of social life is driven by this struggle between values. Instead of understanding the contradictions between values as being caused by power relations or failures of the social order, Robbins draws on Dumont’s notion of encompassment of the contrary. Hierarchy is essentially the recognition of the importance of subordinate values. According to Robbins, the dynamism driven by the struggle between contradicting values is inherent in social life.

One could argue that such configurations are peculiar to transitional phases in which remote social formations have recently encountered individualistic values, especially those introduced by a universalist religion. Ismaël Moya deals with a context, Dakar, the capital of Senegal, in which Islam has been present for centuries. In this case, exchange ceremonies of birth and marriage in which women honor kinship relations with lavish gifts are condemned as a local custom (aada) that prevents the full realization of two universalistic and individualistic values: Islam and economic rationality. Yet a comparative experiment allows us to recognize that the relations at stake in exchange ceremonies are highly valued on another level. In Dakar, as in Euro-America, finance is king. However, in the former, money does not only circulate in commodity relations but is involved in every meaningful relation. Neither economics, nor politics or even religion can synchronize the financial relations that compose this sociality: only life-cycle ceremonies (especially birth and marriage) successfully manage to do so. Whereas in the Euro-American configuration, the primacy of the economic dimension impinges on the political (Dumont 1977), the economy in Dakar is subordinated to women exchange’s ceremonies. Moreover, Muslim rituals are systematically articulated to the exchange in birth and marriage: women’s exchange ceremonies combine hierarchically with Islam, the paramount order of value: that of absolute submission of the individual to God.

Aparecida Vilaça’s paper on Christianity in Amazonia, and especially among the Wari’, offers a challenging counterpoint that introduces a perspectivist point [131]of view on the issues at stake in this section. Wari’ sociality is characterized by interspecific transformability (see above). Persons and animals are composed of human (predator) and animal (prey). Wari’ pre-Christian sociality was composed of two movements in opposing directions: on the one hand, the everyday ones of eclipsing alterity and producing kinship relations among humans and, on the other, ritual movement (collective ceremonies, warfare, and shamanic action) in which relations of alterity were objectified and stabilized. The model here is not that of hierarchy but of oscillation, recursive movements in opposing directions which are integral to the perspectivist paradigm. In other words, hierarchy exists but is perpetually reversing. Christianity was initially appropriated in this configuration. For example, Vilaça shows that the Wari’ perceived Christianity as an additional way to eclipse alterity (the animal/prey component) of human selves, that is, to stabilize the human component, as they did before through kinship. However, another aspect of Christianity introduced an oscillation: the presence of the devil, which corresponds to the animal side of the Wari’ and the human pole of the animals. The devil reconstituted dividuality by restoring to the animals the agency taken from them by God (who also was originally dividual before detaching the devil). This configuration progressively transformed after the Christian revivalism of the twenty-first century. Traditional rituals, and especially shamanism, disappeared; the constitutive difference between affinity and consanguinity was replaced by the extension of Christian fraternity to everyone; and the devil shifted from acting through animals to enter the person directly, thereby suppressing his or her animal pole. Rituals are now aimed at the production of identity. From a perspectivist point of view, the changes introduced by Christianity, although not fully realized, tend toward a paralysis of the world, illustrated by the Wari’ interpretation of the Christian afterlife as a dualism made absolute between heaven, a place of superlative humanity composed of individuals deprived of any relation, and hell, a place of superlative animality, composed of endlessly roasting prey.

Cécile Barraud’s contribution deals with Dumont’s fascination with equality in the context of Dravidian kinship categories, in contrast with the hierarchical principle of caste. However, she foregrounds his idea of the equivalent value of the sexes, eclipsed in the anthropological conversation by Dumont’s argument of equal value between affinity and consanguinity. Through a examination of sex distinction in different kinship terminologies, she explores different meanings of the ideas of equality and hierarchy that suggest a comparative perspective on the “equality of the sexes” as distinguished from gender equality.

We would conclude by observing that Dumont’s comparative project, as we understand it, has not been made obsolete by social changes, whether urbanization, religious conversion, or globalization. Instead, it has been clarified and revivified. As we have seen, radical comparison draws on differences, but depends neither on the preservation of an alterity between bounded societies (or wholes, worlds, etc.) nor on the elaboration of generic crosscultural categories. On the contrary, Dumont’s comparative practice focuses firstly on the radical but relative (and thus never absolute) contrast between hierarchies of values but also, secondly, on the dynamic contradictions between values, which are especially intensified by social changes and the interaction between cultures.[132]


We thank Giovanni da Col for his editorial critique and dedication.


Appadurai, Arjun. 1986. “Is homo hierarchicus?” American Ethnologist 13 (4): 745–61.

———. 1988. “Putting hierarchy in its place.” Cultural Anthropology 3 (1): 36–49.

Béteille, André, 1986. “Individualism and equality.” Current Anthropology 27 (2): 121–34.

Celtel, André. 2005. Categories of the self: Louis Dumont’s theory of the individual. Oxford: Berghahn.

da Col, Giovanni, and David Graeber. 2011. “Foreword: The return of ethnographic theory.” Hau: Journal of Ethnographic Theory 1 (1): vi–xxxvii.

Descola, Philippe. 2013. Beyond nature and culture. Translated by Janet Lloyd. Chicago: University of Chicago Press.

Descombes, Vincent. 2014. The institutions of meaning: A defense of anthropological holism. Translated by Stephen Adam Schwartz. Cambridge, MA: Harvard University Press.

Dirks, Nicholas B. 1993. The hollow crown: Ethnohistory of an Indian kingdom. Ann Arbor: University of Michigan Press.

Dumont, Louis. 1953. “The Dravidian kinship terminology as an expression of marriage.” Man 53: 34–39.

———. 1965. “The functional equivalents of the individual in caste society.” Contributions to Indian Sociology 8: 85–99.

———. 1970. “A structural definition of a folk deity of Tamil Nad: Aiyanar, the Lord.” In Religion, politics, and history in India, 20–32. Paris: Mouton.

———. 1971. “On putative hierarchy and some Aalergies to it.” Contributions to Indian Sociology 5: 61–84.

———. 1977. From Mandeville to Marx: The genesis and triumph of economic ideology. Chicago: University of Chicago Press.

———. 1980. Homo hierarchicus: The caste system and its implications. Translated by Mark Sainsbury, Louis Dumont, and Basia Gulati. Chicago: University of Chicago Press.

———. 1983. Affinity as value: Marriage alliance in South India, with comparative essays on Australia. Chicago: University of Chicago Press

———. 1986. Essays on individualism: Modern ideology in anthropological perspective. Chicago: University of Chicago Press.

———. 1991. Homo æqualis II: L’idéologie allemande: France–Allemagne et retour. Paris: Gallimard.

———. 1994. German ideology: From France to Germany and back. Chicago: University of Chicago Press.

[133]Fassin, Didier, ed. 2012. Moral anthropology: A companion volume. Malden, MA: Wiley-Blackwell.

Graeber, David. 2001. Toward an anthropological theory of value: The false coin of our own dreams. New York: Palgrave-Macmillan.

———. 2013. “It is value that brings universes into being.” Hau: Journal of Ethnographic Theory 3 (2): 219–43.

Gregor, Thomas, and Donald Tuzin, eds., 2001. Gender in Amazonia and Melanesia: An exploration of the comparative method. Berkeley: University of California Press.

Hadot, Pierre. 2004. Philosophy as a way of life. Translated by Michael Chase. Oxford: Blackwell

Holbraad, Martin, and Morten Axel Pedersen, 2009. “Planet M: The intense abstraction of Marilyn Strathern.” Anthropological Theory 9 (4): 271–394.

Holbraad, Martin, Morten Axel Pedersen, and Eduardo Viveiros de Castro. 2014. “The politics of ontology: Anthropological positions.” Fieldsights—Theorizing the Contemporary, Cultural Anthropology Online, January 13.

Kapferer, Bruce. 1988. Legends of people, myths of state: Violence, intolerance, and political culture in Sri Lanka and Australia. Washington, DC: Smithsonian Institution Press. http://www.culanth.org/fieldsights/462-the-politics-of-ontology-anthropological-positions.

———. 2010. “Louis Dumont and a holist anthropology.” In Experiments in holism, edited by Ton Otto and Nils Bubandt, 187–208. Oxford: Wiley-Blackwell.

Laidlaw, James, 2002. “For an anthropology of ethics and freedom.” Journal of the Royal Anthropological Institute (N.S.) 8 (2): 311–32.

———. 2012. “Ontologically challenged.” Anthropology of This Century 4. http://aotcpress.com/articles/ontologically-challenged/.

Lambek, Michael, ed. 2010. Ordinary ethics: Anthropology, language, and action. New York: Fordham University Press.

Leach, Edmund R. 1971. “Hierarchical man: Louis Dumont and his critics.” South Asian Review 4 (3): 233–237.

Madan, T. N. 1982. Way of life: King, householder, renouncer: Essays in honour of Louis Dumont. New Delhi: Vikas Publishing House.

Mariott, McKim, 1969, “A review of Homo hierarchicus by L. Dumont.” American Anthropologist 71: 166–75.

———. 1976. “Hindu transactions: Diversity without dualism” In Transaction and meanings, edited by Bruce Kapferer, 109–42. Philadelphia: Institute for the Study of Human Issues.

Needham, Rodney. 1973. Right and left: Essays on dual symbolic classification. Oxford: Oxford University Press.

Otto, Ton, and Rane Willerslev, eds. 2013a. “Value as theory. Part I.” Special Issue of Hau: Journal of Ethnographic Theory 3 (1): 1–171.

———. 2013b. “Value as theory. Part II.” Special Issue of Hau: Journal of Ethnographic Theory 3 (2): 1–243.

[134]Parkin, Robert. 2003. Louis Dumont and hierarchical opposition. Oxford: Berghahn.

Parry, Jonathan P. 1979. Caste and kinship in Kangra. London: Routledge.

———. 1998. “Mauss, Dumont and the distinction between status and power” In Marcel Mauss: A centenary tribute, edited by Wendy James and N. J. Allen: 151–75. Oxford: Berghahn.

Polanyi, Karl. 1944. The great transformation: The political and economic origins of our time. Boston: Beacon Press.

———. 1957, “The economy as an instituted process.” In Trade and markets in the early empires, edited by Karl Polanyi, Conrad M. Arsenberg, and Harry W. Pearson, 243–70. New York: Free Press.

Raheja, Gloria G. 1988. The poison in the gift: Ritual, prestation, and the dominant caste in a North Indian village. Chicago: University of Chicago Press

Richards, J. F., and Ralph W. Nicholas, eds. 1976. “Symposium: The contributions of Louis Dumont.” Journal of Asian Studies 35 (4): 579–650.

Rio, Knut M., and Olaf H. Smedal, eds. 2009a. Hierarchy: Persistence and transformation in social formations. New York: Berghahn.

Rio, Knut M., and Olaf H. Smedal, eds. 2009b. “Hierarchy and its alternatives: An introduction to movements of totalization and detotalization.” In Hierarchy: Persistence and transformation in social formations, edited by Knut M. Rio and Olaf H. Smedal, 1–63. New York: Berghahn.

Robbins, Joel. 2007. “Between reproduction and freedom: Morality, value, and radical cultural change.” Ethnos 72 (3): 293–314.

———. 2009. “Conversion, hierarchy and cultural change: Value and syncretism in the globalization of Pentecostal and Charismatic Christianity.” in Hierarchy: Persistence and transformation in social formations, edited by Knut M. Rio and Olaf H. Smedal, 65–88. Oxford: Berghahn.

Robbins, Joel, and Jukka Siikala. 2014. “Hierarchy and hybridity: Toward a Dumontian approach to contemporary cultural change.” Anthropological Theory 14 (2): 121–32.

Salmond, Amiria J. M. 2014. “Transforming translations (part 2): Addressing ontological alterity.” Hau: Journal of Ethnographic Theory 4 (1): 155–87.

Strathern, Marilyn. 1988. The gender of the gift: Problems with women and problems with society in Melanesia. Berkeley: University of California Press.

———. 1992. “Parts and wholes: Refiguring relationships in a postplural world.” In Conceptualizing society, edited by Adam Kuper, 75–104. London: Routledge.

———. (1991) 2004. Partial connections. Oxford: AltaMira Press.

Street Alice, and Jacob Copeman. 2014. “Social theory after Strathern: An introduction.” Theory, Culture & Society 31 (2–3): 1–31.

Tcherkézoff, Serge, 1987. Dual classification reconsidered. Translated by Martin Thorn. Cambridge: Cambridge University Press.

Viveiros de Castro, Eduardo. 1998. “Cosmological deixis and Amerindian perspectivism.” Journal of the Royal Anthropological Institute (N.S.) 4 (3): 469–88.

[135]———. 2001, “GUT feelings about Amazonia: Potential affinity and the construction of sociality.” In Beyond the visible and the material: The Amerindianization of society in the work of Peter Rivière, edited by Laura M. Rival and Neil L. Whitehead, 19–43. Oxford: Oxford University Press.

———. 2004. “Perspectival anthropology and the method of controlled equivocation.” Tipití: Journal of the Society for the Anthropology of Lowland South America 2 (1). http://digitalcommons.trinity.edu/tipiti/vol2/iss1/1.

———. 2008. “Claude Lévi-Strauss, Œuvres.” Gradhiva 8. http://gradhiva.revues.org/1215?lang=en.

———. 2009. “The gift and the given: Three nano-essays on kinship and magic.” In Kinship and beyond: The genealogical model reconsidered, edited by Sandra C. Bamford and James Leach, 237–68. New York: Berghahn.

Wagner Roy. 1975. The invention of culture. Chicago: University of Chicago Press.

———. 1991. “The fractal person.” In Big men and great men: The personifications of power in Melanesia, edited by Maurice Godelier and Marilyn Strathern, 159–73. Cambridge: Cambridge University Press.

Mister D.: Comparaison radicale, valeurs et théorie ethnographique

Résumé : Cet article suggère que la pertinence des travaux de Louis Dumont pour la théorie ethnographique contemporaine réside dans sa conception radicale de la comparaison comme une expérimentation sur la différence qui abolit la distinction entre analyse anthropologique et épistémologie. Le texte applique donc la méthode de Dumont—la comparaison—à son anthropologie. Dans la première partie, nous suivons le fil de la rencontre de Dumont avec le système des castes indien et le contraste radical qu’il a établi avec l’euro-Amérique pour proposer une perspective sur sa méthode comparative et ses principales notions (valeur, hiérarchie, englobement). Dans la seconde partie, la stratégie anthropologique de Dumont est mise en perspective avec deux autres projets de comparaison radicale: celui de Marilyn Strathern sur la Mélanésie et celui d’eduardo Viveiros de Castro sur l’Amazonie.

André ITEANU is Directeur de Recherche at the CNRS and Directeur d’Études at the EPHE in Paris, France. He has worked for many years with the Orokaiva of Papua New Guinea and with troubled youth in Cergy-Pontoise, a suburb of Paris. He recently edited a volume, La cohérence des sociétés (2010), and translated an Orokaiva autobiography written by Lucien Vevehupa, The man who would not die (2013).

André Iteanu
Centre Asie du Sud-Est - CNRS
190 Avenue de France 75013 Paris

[136]Ismaël MOYA is a former economist who converted to social anthropology under the influence of his ongoing fieldwork in a poor suburb neighborhood of Dakar, Senegal. He is Chargé de Recherche at the Centre National de la Recherche Scientifique (CNRS, France), and a member of the Laboratoire d’ethnologie et de Sociologie Comparative (Université de Nanterre, France). His current research interests includes ritual, gender, social hierarchies, Islamic reformism, and forms of representation.

Ismaël Moya
CNRS, Laboratoire d’ethnologie et de Sociologie Comparative, UMR 7186
Maison Archéologie & Ethnologie René-Ginouvès
21, Allée de l’université
92023 Nanterre cedex


1. This special section is a revised version of a larger collection on Louis Dumont’s comparative anthropology edited by Cécile Barraud, André Iteanu and Ismaël Moya, forthcoming in French at CNRS éditions.

2. Dumont’s legacy in contemporary anthropology has expanded to other regions of the world. Since the 1980s, most writings engaging with Dumont’s orientations have come from scholars working in the Pacific and Southeast Asia (Barraud, de Coppet, Damon, Eriksen, Howell, Iteanu, Mosko, Platenkamp, Robbins, Rio, Tcherkézoff, etc.).

3. Giving a review of Dumont’s work would be redundant since Robert Parkin’s Louis Dumont and hierarchical opposition (2003) has already accomplished this task. That book offers both a general history of Dumont’s work and a comprehensive and heuristic presentation of his core ideas, their origins (Hertz, Needham, etc.), and their reception (debates, critics, and followers). André Celtel’s Categories of the self (2004) discusses Dumont’s conceptions of individualism. More recently, Knut Rio and Olaf Smedal’s Hierarchy: Persistence and transformation in social formations (2009a) presented Dumont’s analytical framework and its critics and confronted it with ethnographies from Oceania, Asia, or the Middle East. Joel Robbins has also reassessed, from an American (cultural) and Weberian point of view, Dumont’s legacy as a theorist of contemporary processes of cultural interaction and transformation (Robbins 2009; Robbins and Siikala 2014). Vincent Descombes’ recently translated The institutions of meaning (2014), starting with the problems raised by the philosophical notion of intentionality, gives a sterling defense of Dumont’s anthropological holism, drawing on Wittgenstein and Peirce. This book also teases out most of the conceptual muddles regarding Dumont’s holism.

4. Our aim is not to scrutinize the relevance Dumont’s ethnography and analysis of India but to explain and clarify his practice of comparison. Although Bruce Kapferer, in his Legends of people, myths of state (1988), does not entirely support Dumont’s argument, he strongly assessed that most dimensions of his Southern Indian analysis apply also to Sri Lanka, and most particularly the notions of totality and the subordination of political power to hierarchy (ibid.: 9). In this introduction, we touch upon these themes only lightly. Following the release of Legends of people, myths of state, Dumont reported (pers. comm.) that he was very impressed by the sharpness of Kapferer’s analysis and understanding of his work (see also Kapferer 2010).

5. Dumont was a student of Marcel Mauss and took much of his inspiration from his work (with the exception of hierarchy). According to Dumont, Mauss’ commitment to ethnographic knowledge enabled anthropology to reach its experimental stage. Dumont (1986: 2) construed the concept of “total social fact” as “a specific complex of a particular society (or type of society), which cannot be made to coincide with any other.” In other words, it implies firstly a stress on difference: facts reacts to the categories, theories, and implicit ideas with which we approach them. It is a comparative experiment that involves the subject in the object (ibid.: 199). And, secondly, the “total” aspect of the fact means that the aim is not to study discrete elements but to compare “wholes.” “How is this [i.e., the whole] to be found? In a sense society is the only ‘whole,’ but it is so complex that however scrupulously we reconstruct it, there is doubt about the result. But there are cases [i.e., total social facts] where consistency is found in less extended complexes, where the ‘whole’ can be more easily kept within view” (ibid.: 194).

6. The anthropology of morality and ethics (e.g., Laidlaw 2002; Robbins 2007; Lambek 2010; Fassin 2012) draws on two meanings of the idea of value: (1) values as a given set of norms (a moral order), somehow exterior to individuals, that determines what they are supposed to do or not to do, what is good or bad; and (2) conceptions of the good that stem from the reflexive work of individuals as ethical subjects.

7. David Graeber critically engaged several times with Dumont’s ideas on value (e.g. 2001: 16–20, 2013: 235–36) without, in our opinion, taking into account in his criticisms the experimental and (radical) comparative aspects of the Dumontian perspective.

8. The “anthropology of value,” as it were, is actually a much larger and more heterogeneous field. See, for example, the diversity of anthropological perspectives on the question of “value” in the fifteen contributions in Hau’s special issue on “Value as theory” edited by Ton Otto and and Rane Willerslev (Otto and Willerslev, 2013a, 2013b).

9. Dumont (1986: 235) also explained that he developed the idea of value reluctantly as his next conceptual “bid” after “trying to sell to the profession the idea of hierarchy, with little success.” He defined hierarchy as the “order resulting from the consideration of value” (ibid.: 279). In other words, what matters for Dumont is not a single value or a set of values but a configuration in which values are hierarchized compared to another configuration.

10. See, for example, Dumont (1970), where he shows that in India, the higher castes revere the higher gods while the lower castes revere the lower gods (the black god) as their dominant gods. Thus the relative value of the gods does not apply evenly to all castes. For the different forms of hierarchical inversions, see Tcherkezoff (1987).

11. This is developed in Michael Houseman’s account of “The hierarchical relation” in this issue.

12. Knut Rio and Olaf Smeldal (2009b: 12–15) addressed these issues.

13. Marilyn Strathern never engaged with Dumont’s ideas. However, she constantly acknowledged her debt to Roy Wagner’s The invention of culture (1975). The latter was influenced by Dumont, at least when he built up the distinction between modern American society and older civilizations. “Les obvious is the inadvertent similarity between Dumont’s homo hierarchicus/homo aequalis contrast and the pointed comparisons I make between ‘relativized’ modern American society and the dialectically balanced social orders of older civilizations” (ibid.: 8). Eduardo Viveiros de Castro, whose ideas on comparison are discussed later on, has also been deeply influenced by Wagner.

14. On this point Marilyn Strathern explicitly follows McKim Mariott (1976), who coined the term about India in his critique of Dumont. Nevertheless, according to us, this academic irony has eclipsed how close Strathern’s and Dumont’s orientations actually are (see also Kapferer 2010: 224).

15. The allusions to the ontological turn are very limited because it is a very diverse current and a more complete consideration of it would have rapidly exceeded the ambition of this introduction. For a definition of the ontological turn and of three different ethnographic strategies within it, see Salmond (2014: 160–68). We mostly discuss here the work of Viveiros de Castro and the people inspired by his “perspectivist” thinking (e.g., Holbraad, Pedersen, and Viveiros de Castro 2014).

16. Dumont’s contention is based on structuralism: practices and classifications in any social formation are imbued with values. One cannot oppose any couple of categories, or move from one to the other, without attributing them unequal value. Dumont reassessed Lévi-Strauss’ idea of structure, but adds value to it, thus transforming it into a hierarchy (a configuration of values). Hence, according to Robert Parkin (2003), the fact that Dumont is more concerned with praxis and agency than Lévi-Strauss.

17. Dumont dared, for example, to propose a nonmoral (as tenable) anthropological reading of Mein Kampf (Dumont 1986: 149–79).

18. However, in Viveiros de Castro’s words, these diverse worlds cannot be kept valueless from the point of view of anthropology. The perspectivist Arawete’s world is a better model for anthropology than the Euro-American, as, in that respect, is the Jaguar’s.

19. Power is universal.