Ontologies, anthropologists, and ethical life

This work is licensed under the Creative Commons | © Webb Keane. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 DOI: http://dx.doi.org/10.14318/hau3.1.013

Book Symposium

Ontologies, anthropologists, and ethical life

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

Webb KEANE, University of Michigan


What is ontology? There seem to be two versions in play in Geoffrey Lloyd’s book, what we can call strong and weak ontology. Lloyd is often talking about what I am calling the weak form: “ontologies are… comprehensive accounts of whatever there is” (Lloyd 2012: 39). Thus he refers to a “picture” the Greeks held, and remarks that where Jesuits saw things, Chinese saw events (ibid.: 23). Such expressions treat ontology as a perception; the language of visuality suggests a world that different people share, but view in various manners: if Jesuits see something one way, Chinese another, they nonetheless seem to be looking in the same direction. Moreover, “the Chinese spoke not of elements but of phases.” Thus we have a view manifested in words. Notice the double mediation: it’s a verbal representation of an idea about the world. In all these cases, ontology turns out to be, or at least our access to it is by way of, a set of propositions or assertions. Lloyd’s ancient Greek and Chinese examples are weak ontology; that is they are theories or interpretations of reality. A representation has the structure of what Brentano (1973) called “intentionality”: it is about something. This “aboutness” seems to require that there be some kind of object other than the representation itself that the representation is about, however opaque or ultimately unknowable that object might be.

But the ethnographers Lloyd discusses, Eduardo Viveiros de Castro and Philippe Descola, seem to have something stronger in mind. Of Viveiros de Castro’s perspectivism, Lloyd (2012: 21) writes, “different members of the human race have such different experiences, perceptions, and ways of interacting with their environment that we should think of them as living in different worlds.” Now on first blush, this may seem to be merely a variation on, say, the linguistics of Whorf (1956: 158) when he suggests “Hopi ‘duration’ seems to be inconceivable in terms of space or motion,” or the cultural analysis of Sahlins (1985: 110) in asserting that “Cook was welcomed as Lono.” But the language of ontology suggests something more profound is involved. The claim seems to be that the different worlds in which people live are not just theories, pictures, or performative acts, nor world views, habituses, epistemes, or cultures. As what I will call strong ontology, perspectivism seems to be entering the domains marked out by Martin Heidegger’s (1962) Being-in-the-World or W. V. O. Quine’s (1969) ontological commitments. Strong ontology would seem to be a fundamental reality independent of any representation of it, and sufficiently self-contained as to yield no significant overlap with any other reality.

In its weak form, ontology poses no special observer paradoxes. Greek descriptions of reality are potentially legible to non-Greeks. But strong ontology does not seem to rest there. Lloyd (2012: 21) reports that Viveiros de Castro says all living beings see the world in the same way; what changes is the world that they see. But on what grounds can we make such a claim that it is the world and not our construal of it that differs? What knowledge of the world do I have independent of how I see it, or how they tell me they see it, such I can affirm, with Viveiros de Castro, where the difference lies between how people see things, and what it is that they see? After all, to the extent ethnographers adhere to the proposition that truth claims are context-dependent, then we must also deny ourselves the god’s-eye perspective, what Nagel (1986) called “the view from nowhere.” If, on the other hand, we stake our ethnographic authority on some objective stance autonomous of any context, then we seem to be committed to a strong but ultimately positivist ontology, granting access to a unified reality shared by anyone at all who is able to enter into that objective stance. This seems paradoxical: what version of strong ontology is consistent with the objectivism implicit in the idea that there is some viewpoint that enables the ethnographer to tell us confidently that multiple ontologies exist and are irreducible to one another?

The symptom of the paradox is that it turns out to be hard to speak ethnographically about a strong ontology, some fully inhabited reality distinct from other equally fully inhabited ones, without falling back on weak ontology, that is on indigenous theories and representations. Thus in places Lloyd (2012: 20) reads Descola as giving us “a concept, a theory,” which focuses on whether beings differ in terms of either physicality or interiority, or neither, or both. This seems to presuppose that there is some pre-theoretical physical experience that is construed in the theory, or weak ontology, such that souls or interiority are potentially contrastive to something they are not, such as bodies or exteriority. The theory rests on a distinction between what is or is not available to perception (and potentially anybody’s perception, or so it would seem).

So too, Lloyd sometimes treats Viveiros de Castro’s perspectivism as a kind of weak ontology, calling it an indigenous anthropology, a theory about people. But this sounds quite a bit like what we find in familiar accounts of culture. Somewhat stronger, but still close to conventional views of culture is this: “conscious beings construct their world as they interact with it” (ibid.: 21). The verb “construct,” however, tends to smuggle in a host of unstated arguments, and plays so central a role in the taken-for-granted of contemporary anthropology that it can be hard to reflect on. Does construct mean to formulate a proposition or to construe an interpretation? To build something sui generis from scratch? To assemble something from pre-existing bits and pieces that other people are also working with to different ends? Is the result ideas? Material stuff? Cultures? Governments? Rules? Kinfolk? Habits? Structures of feeling? Intuitions? Perceptions? Desires? Much hangs on this easily uttered word.

Referring to perspectivism, Lloyd points out that we have only humans’ descriptions of both jaguars and humans. If this is right, then it would seem, first, that perspectivism is asymmetrical, skewed toward the human as the giver of accounts, and second, that it is manifested as a set of propositions, a represent-ation. But it’s one thing to say “Those people tell me about a different reality” (weak ontology) and quite another that “Those people inhabit a different reality” (strong ontology). The first assertion is the familiar stuff of ethnography. The second seems to encounter Ludwig Wittgenstein’s (1953: 223) lion—“If a lion could talk, we could not understand him.” In that case, fieldwork would end up being not much more than the ethological observation of opaque creatures. Yet anthropological proponents of strong ontology seem to understand shamans; moreover, their readers are unsurprised at this.

Let’s be clear: anthropologists certainly know people who do inhabit drama-tically different worlds of practice and thought; moreover, our vocation requires us to open ourselves to the challenges they pose. But we do not enter into those worlds primarily via the didactic virtuosity of indigenous metaphysical theorists. Those worlds are not inhabited first and foremost as talk. So how do we enter them? As Lloyd (2012: 97) astutely comments, we should “remind ourselves of our childhood.” I take this to be a call to take seriously our own practices: fieldwork begins with our practical ability to learn how to live with other people, and continues with our taking up the ethical demands that ability entails.

The ability to live with other people is not just a solution to an epistemological problem. It brings us to the question of what’s at stake in posing ontological questions at all. For in our academic homes, the attempt to grasp strong ontologies is a special professional concern for a small handful of philosophers, theologians, and theoretical physicists—so why should they also be of interest to anthropologists and, more to the point, to their interlocutors in their gardens and forests? Let’s start with weak ontologies, which are the main concern of Lloyd’s book (and, one suspects, of all anthropological accounts of ontology, wittingly or not). After all, representations form the raw materials with which Lloyd, as well as Descola and Viveiros de Castro are working. Most accounts of weak ontologies or represent-ations of reality seem to beg the question of why anyone should even have a theory of reality. Claude Lévi-Strauss (1966), of course, did have an answer to that question, the classificatory intention, a universal cognitive propensity. I will suggest something different, that ontologies, as something people might be able to talk about, are a response to the ethical demands of social life (which may include relations with animals and other entities). But first, consider Lloyd’s discussion of the Greeks and Chinese.

Lloyd offers some very useful insights into the practices and social relations within which any given weak ontology emerges and thrives. The Greeks entertained multiple theories of reality. This was the outcome of a certain agonistic spirit. Because speakers were competing against one another for prestige and authority over domains of knowledge, they strove to differentiate themselves from their competitors. The result was radical disagreements about the composition of reality. Yet these ontological disagreements depended upon a shared set of practices. The quest for virtuosity encouraged arguments and competition among theories carried out before a potentially skeptical audience (Lloyd 2012: 83). This had a number of consequences. Because speakers had to persuade listeners, they had to render their accounts explicit (ibid.: 59). What the speakers had in common was the very condition of possibility for their divergences. Their way of life didn't determine the content of the ontology, but rather its semiotic modality, its production as an object of theoretically and discursively available knowledge. The radical differences among ontologies were a response to a condition that Greek metaphysicians shared, the demand that you give your opponent a verbal account of things, and that it be novel (ibid.: 57). Thus a certain way of life accounts for two distinctive features of Greek ontologies: their explicitness and their multiplicity.

The Greeks inhabited a certain reality, no doubt, but not the one portrayed by the content their theories of reality. As Lloyd points out, the larger Greek population was uninfluenced by elite theories (ibid.: 70). Not only did the majority of Greeks ignore the theories, even the theorists did: “you could not live by their consequences, not in the case of Parmenides at least” (ibid.: 94). This suggests that the Greeks held a practical reality in common. Otherwise they would have been unable, or at least unmotivated, to argue with one another. If we want to speak of a Greek ontology in the strong sense, of an inhabitable reality, we should not be looking at their representations of reality but at the preconditions for making such representations in the first place, and the consequences that follow from them.

Lloyd’s descriptions of Greece and China suggest we should not focus entirely on the manifest content of ontological assertions. We should be asking who is being addressed, to what end, with what real or imagined consequences, and through what semiotic modalities. (Diagrams? Mathematical formulae? Institutions? Laws? Architecture? Rituals? If via speech, is it poetic? Propositional? Narrative? Argumentative? The sounds of mantra?) We should wonder: What is at stake when a shaman explains reality to an ethnographer? How literal-minded should we be when we engage in talk that seems to be about ontology?

There is no doubt that anthropologists should take seriously that there are dramatically different ways of being in the world. But if practices construct a way of being in the world, it is not obvious that the outcome should be a coherent, totalizing theory that can be described as so many beliefs (weak ontology). And when we do encounter such theories, we should wonder why they're there. As I argued long ago (Keane 1995), they are likely to be precipitates of practices of objectification whose ground may not lie in the purposes of portraying reality at all, even if that’s what they look like. When we listen to assertions about jaguars or atoms or souls, we should consider what’s being done in this kind of talk.

So what’s going on in depictions of reality? Many things, no doubt, but in most cases not simple truth claims for their own sake. Lloyd (2012: 8) tells us that Greeks defined the human within a triadic relationship, located between gods and animals. But the point does not seem to have been to describe reality pure and simple. Rather, by depicting those beings amidst which they found themselves, they were guided to what their relations to them should be. If gods are dangerous, they are to be mollified. If animals are kin, it may be wrong to kill them (ibid.: 12). But I would suggest that it is the problem of killing animals that induces questions about their nature, not vice versa.

We can say something similar of Lloyd’s account of China. When Chinese philosophers addressed real or imagined rulers, it was to provide ethical guidance to those who governed—if they were speaking truth, it was because they were speaking truth to power. Thus the production of Chinese ontologies took place with reference to a real or imagined audience, across a hierarchical distribution of temporal power, and intellectual authority (ibid.: 81–82). The character of that audience was crucial; it constituted the demand for theories and shaped the talk that conveyed them.

How to define the boundaries of the human also worried nineteenth century Europeans faced with the death of God and the birth of Darwinism. That’s the context that led Ivan Karamazov from an ethical worry to an ontological assertion, returning him to an ethical conclusion: if there is no God, then everything is permitted. Ethical implications underwrite even an apparently clear case of ontological difference, the familiar example of the Eucharist (see ibid.: 107–8, n. 11). What is wine for me is Christ’s blood for you. (Notice that this has never been a consistent and untroubled assertion; it was contested enough that the Fourth Lateran Council of 1215 was compelled to take steps to reassert it, adding the concept of transubstantiation to help explain it.) At the end of the day, the stakes are the questions “What shall I do?” and “with what consequences?” If, and only if, the wine is Christ’s blood, then drinking it is a sacrament, a practice geared to my salvation. And worries about the ethical implications of abortion laws, new reproductive technologies, artificial intelligence, environmentalism, robots, and animal rights, for example, are driving new efforts to define the human.

The Greeks, Lloyd tells us, were concerned with “such matters as the honesty of witnesses, the accuracy of reports, the diagnosis of intentions” (ibid.: 100)—that is, with ordinary everyday practices taken as signs that enable the ethical evaluation of persons and events. Similarly, Mesopotamian scribes took celestial phenomena to be signs leading to predictions. These predictions “were usually expressed in the form of conditionals, if so and so (the sign), then so and so (the outcome, called the verdict)” (ibid.: 48–49). Here is a clear instance of what I've called semiotic ideology (Keane 2007): the quest for signs is shaped by assumptions about what might count as a sign. As the Greek and Mesopotamian examples show, the quest for signs is characteristically undertaken in response to the question what shall we do and how shall we live? Semiotic ideology may retroactively give power to ontological presuppositions, but it is not ontology that drives that quest for signs. It is in the consequences for ethical life within the world, not in some set of beliefs about the world, that it makes sense for us to say people truly inhabit their respective realities. And it is in interacting with others’ response to that question, that we might come to say we know them.


Brentano, Franz. 1973. Psychology from an empirical standpoint, translated by A. C. Rancurello, D. B. Terrell, and L. McAlister. London: Routledge.

Heidegger, Martin. 1962. Being and time, translated by John Macquarrie and Edward Robinson. New York: Harper and Row.

Keane, Webb. 1995. “The spoken house: Text, act, and object in eastern Indonesia.” American Ethnologist 22 (1): 102–24.

———. 2007. Christian moderns: Freedom and fetish in the mission encounter. Berkeley: University of California Press.

Lévi-Strauss, Claude. 1966. The savage mind. Chicago: University of Chicago Press.

Lloyd, G. E. R. 2012. Being, humanity, and understanding. Oxford: Oxford University Press.

Nagel, Thomas. 1986. The view from nowhere. Oxford: Oxford University Press.

Quine, W. V. O. 1969. Ontological relativity and other essays. New York: Columbia University Press.

Sahlins, Marshall. 1985. Islands of history. Chicago: University of Chicago Press.

Whorf, Benjamin Lee. 1956. Language, thought, and reality: Selected writings of Benjamin Lee Whorf, edited by John B. Carroll. Cambridge, MA: MIT Press.

Wittgenstein, Ludwig. 1953. Philosophical investigations, translated by G. E. M. Anscombe. London: Blackwell.



Webb Keane
Department of Anthropology
University of Michigan
101 West Hall
Ann Arbor, MI 48109