Distinguishing ontologies

This work is licensed under the Creative Commons | © Anne-Christine Taylor. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 DOI: http://dx.doi.org/10.14318/hau3.1.016

Book Symposium

Distinguishing ontologies

Comment on LLOYD, G. E. R. 2012. Being, humanity, and under-standing. Oxford: Oxford University Press.

Anne-Christine TAYLOR, CNRS, Musée du Quai Branly


Geoffrey Lloyd’s latest book is an essay on comparative ontology, how to conduct such a comparison and why. Apart from its intrinsic merits—acute analytical insight, vast erudition lightly worn, clarity of purpose, and an engaging style of writing—the book is timely, given the prominence the issue of ontologies has gained in recent works both in anthropology and philosophy. Lloyd draws on three main sources to develop his arguments: ancient Greek metaphysics, classical Chinese political and scientific writings, and contemporary anthropological accounts of “exotic” (mainly Amazonian) ways of world-making. On the one hand, he is dealing with textual sources referring to ontological premises more or less explicitly stated; on the other, with implicit ontologies inferred by anthropologists from observed practices (including discursive ones) and reported in terms that allow them to be understood by Western readers. But are these different brands of ontology, treated as equivalent, really comparable?

The two anthropological examples Lloyd deals with are taken from the work of Philippe Descola and Eduardo Viveiros de Castro, exemplifying two quite distinct approaches, as Lloyd himself acknowledges. What Lloyd presents as Descola’s (2003) ontologies are, in fact, no such thing; they are models—in classic structuralist fashion—of entities that have no empirical existence, namely ideal-types of the kind of worlds which would be generated by the strict application of rules of composition of principles of identity and difference (the building blocks of any ontology) along two axes, “physicality” and “interiority.”1 The four ontologies produced in this manner are thought experiments, since no actual society or cultural group actualizes any of them in a pure and exclusive way; all empirically observable societies combine to some degree elements proceeding from the application of the four modes of identification, though often one of these modes comes to acquire more weight than the others in shaping areas of practice, core institutions, and mental habits. Descola’s four models are in fact heuristic devices for identifying the dynamics generated by the actualization (usually at the level of a single collective, as identified both by its members and outsiders, or at the level of cultural areas) of ontological premises corresponding more or less closely to those set out in one of the ideal-types. Their main purpose, as Descola himself stresses, is to help resolve a specifically anthropological problem, namely the compatibility or co-occurrence as well as the incompatibility of kinds of practices, institutions, and ideologies, as evidenced by comparative ethnographical studies. Why, for example, is sacrifice (as usually defined) so uncommon in cultures of animist orientation? Why, conversely, is hierarchical ordering of the components of the world of such paramount importance in South and East Asian societies as well as in medieval Europe?

Viveiros de Castro’s perspectivism is something else. If Descola’s models are infra-ontologies, Viveiros de Castro’s late brand of perspectivism is an über- or hyper ontology: as he defines it, a full-blown but implicit metaphysics embedded in indigenous Amazonian practices–including, of course, discursive ones, though the latter cannot unproblematically be compared to explicit Greek metaphysical discourse. Viveiros de Castro’s late brand of perspectivism (as set out principally in Métaphysiques cannibales [2009)]) is a brilliant exercise in transliteration or translation (the importation into the language of Western metaphysics, and working out in this context, of ontological premises shared by a range of Amazonian societies) elaborated as a war machine against modern naturalism, for political purposes rather than primarily for broad comparative anthropological ones. In spirit, his method is closer to François Jullien’s (2000, 2006) playing off of “China” against Greece than to Descola’s structuralist model-making, and it carries with it the same vulnerability to the criticism leveled at Jullien—of unduly systematizing differences and disregarding commonalities, of crossing the line between fiction and the game of science, in short, of inventing rather than translating indigenous ontologies (not “this is how it is, so far as historical and ethnographic evidence allows me to understand it,” but rather “this is how it could be, and should be for the purpose”). Viveiros de Castro is evidently aware of these risks, but seems to be willing to take them because of the benefits they offer in terms of fire power against modern naturalism.

So, we already have two very different things included under the label of ontology: (1) models for detecting and unfolding the ontological assumptions underlying areas of practice in given societies and conferring a distinctive spin to the core institutions and ways of life of these collectives; (2) an elaborate transposition of a postulated metaphysics inferred from observed ethnographic data, and above all from the misunderstandings that arise when this ontology comes into collision with modern naturalism. Then we have classical Chinese ontologies, foregrounding process (rather than Being) and the complex interplay of elementary forms of process, and of course Ancient Greek explicit—and profuse— metaphysics. In both these cases Lloyd bases his arguments on texts produced by a specialized class of intellectuals for specific purposes; “advisors” on how best to manage the flow of forces constitutive of the world in the case of classical China, and fiercely competitive philosophers intent on devising and publicizing irrefutable arguments in the case of Ancient Greece. He does not deny the existence in Classical China and Greece of more basic and wide-spread ontologies in the rest of the population—ontologies in a sense closer to the one favored by anthropologists in their usage of the term. But he does not care to distinguish clearly between them (i.e., explicit metaphysics and ontologies anthropological-style), and indeed seems to incline to the view that anthropological-style ontology develops out of the trickle-down of explicit metaphysics. Thus, in his discussion of Greek naturalism he anticipates a possible objection by Descola by stating that the relation between explicit philosophical naturalism and the ontology of the common Greek man is no different from that obtaining between scientific naturalism and the ontology shared by the rest of the population in the seventeenth century European world, who only gradually took to naturalism.

I don't think that Descola would agree with this trickle-down conception of ontology; scientific naturalism would, in his view, be the outcome of the slow installation of anthropological naturalism rather than its source. By the same token, Greek naturalism might not be as naturalist (in Descola’s sense) as it appears. Agreed, Greek metaphysicians may have carved out (or rather transformed the sphere previously governed by deities into) a separate domain (physis) governed by principles distinct from those presiding over human practice, as such at least a forerunner of the Nature set up by naturalism à la Descola. Assuming that some version of this view was shared by most Greeks as opposed to a handful of philosophers, the crucial point would be whether the workings of this version of physis correspond more closely to naturalist principles or to analogist ones, as evidenced, for example, by medical, ritual, and iconographic practices (rather than by metaphysical discourse). Beyond this, in a context favorable to the proliferation of competing metaphysics, the occurrence of a naturalist view of the world would not be entirely unexpected; if the game, as Lloyd himself suggests, was one of argumentative one-upmanship, the practitioners of this game would draw on the virtualities of metaphysical fiction inherent to the four modes of identification, given that all humans can and do use them to varying degrees in different circumstances. I would even venture to suggest, for arguments sake, that competitive metaphysics in the Greek style, based on the imaginative “blow up” of the alternative ontologies lurking behind a given lived world, might well be a typically analogist move.

Do these sharp variations in the kinds of thing that Lloyd gathers under the label ontology make any difference to his general argument? Not really, if the point is to celebrate the mind-stretching and purgative effect of taking unfamiliar ontologies seriously, and to devise the best means for engaging with the philosophical issues they involve and profiting from what they have to offer. Whether we are dealing with explicit first-hand metaphysics or implicit second-hand ones as conveyed by anthropologists and historians is of little relevance in this respect. (But, in that case, why not also consider alternative ontologies as developed in science-fiction, the nineteenth century novel, or for that matter in art in general?) If, however, the point is to tease out the logic underlying sets of practices in human groups, and further to seek for commonalities and differences between the dynamics thus brought to light, which is what anthropologists and social scientists usually try to do, then it becomes important to distinguish these various sorts of ontologies and not treat them as equivalent. For an anthropologist, explicit metaphysical propositions such as those put forth by Ancient Greeks and Chinese must of course be treated as ethnographic material rather than as transparent expressions of Greek or Chinese “deep” ontologies; they are things to be accounted for and related to underlying, implicit ontological premises—indeed the problem they raise is to understand how and why explicit metaphysical speculation becomes a common practice in these settings, a question to which Lloyd himself brings some wonderful answers. But his book as a whole does not deal fully enough, at least to my mind, with the epistemological and even political issues involved in lumping together these various levels of the process of world-making.


Descola, Philippe. 2006. Par delà nature et culture. Paris: Gallimard.

Jullien, François. 2006. Chemin faisant, connaître la Chine ou relancer la philosophie. Paris: Seuil.

Jullien, François and Thierry Marchaisse. 2000. Penser d'un dehors. Paris: Seuil.

Viveiros de Castro, Eduardo. 2009. Métaphysiques cannibales. Paris: Presses Universitaires de France.



Anne-Christine Taylor
Musée du Quai Branly 222
Rue de l'Université 75343
Paris cedex 07


1. Here I must admit to a special interest: I happen to be married to Philippe Descola, a fact which puts me in a complicated situation whenever I want, as a fellow anthropologist, to comment publicly on his work. I trust HAU’s readers to suspend disbelief and assume I am discussing Descola’s hypotheses with due objectivity.