On Multiple Ontologies and the Temporality of Things

Webb Keane, University of Michigan
An interesting book called Thinking Through Things (2007), edited by Amiria Henare, Martin Holbraad, and Sari Wastel (see earlier discussion on this site, starting on Dec. 14, 2006), proposes that we think of ethnographic otherness in terms of multiple ontologies. Thus Holbraad reports that Cuban Ifa diviners tell him a certain red powder is power. He argues that in order to take this assertion seriously, we–those of us who are not Ifa diviners–must understand that their red powder exists within a radically different ontology from ours. It’s not that there is one thing, powder, which diviners interpret in a certain way that differs from ours. Rather, in this alternative world, that’s what the powder is.
Now in certain respects this is an appealing version of a familiar, if highly unstable, ethnographic move. If Azande say witches exist, the traditional ethnographer’s first responsibility is to take them at their word (though the very act of writing tends to sabotage that epistemic stance). And certainly, against the Eurocentric self-certainties of nearly all other academic disciplines, this responsibility to alterity grounds one of anthropology’s distinctive contributions to knowledge and its ethics.
But Henare et alia would, I think, argue that the Azande example doesn’t go far enough, since at the end of the day, the usual treatment turns on reducing their statements about witches into nothing more than different interpretations of a reality the Azande share with those of us who are not Azande. In contrast, the editors assert that the claim about multiple ontologies is not just a matter of different interpretations or epistemologies. Ifa powder really is power.
Now there are various questions one might raise about this assertion. We might, for instance, ask for a more precise analysis of the semantics and pragmatics of the word the authors translate as “is,” a notoriously difficult linguistic and philosophical problem. We might also wonder how the Cuban diviners in my old neighborhood in New York managed to converse with me and otherwise handle relations among the distinct ontologies they seem to inhabit. But what I want to point to here is a consequence that seems to follow for any attempt to make sense of the temporality.
The notion of radically different ontologies, if I read the authors correctly, seems inadvertently to render the very temporality of things incomprehensible, and to confine them within static realities. In material terms, history often reflects people’s capacity to respond to the things around them in new ways. They see new possibilities in what had always been there. The world may surprise them. When people see new possibilities, it is often because they have discovered something new about their material surround. As things enter new contexts, they enter into new human purposes, afford new kinds of actions, and suggest new projects (see Keane 2003, 2008). If this is so, it seems to follow from two characteristics of material things. First, material things cannot be reduced to whatever happens to be found in concepts. Henare et al’s invitation to rethink the relationship between “concepts” and “things” is certainly full of promise. But if we simply collapse the distinction, we are likely to overlook this: at any given moment there remains something about them that is unknown or unnoticed. Second, it is because things in their very materiality exceed any particular concepts, times, and projects, that they persist across different concepts, times, and projects. They enter into quite distinct concepts, times, and projects.
From this perspective, then, the red powder of which Ifa diviners speak cannot be confined to that singular ontology in which red powder really “is” power, and “is not” whatever someone who is not an Ifa diviner thinks it is. In order to get at the historicity of people and their things, we need to understand things in ways that do not reduce them to some stable essence, to the particularity of a certain context.
If there were an essence—which I think is implicit in the idea of multiple ontologies–then things would never suggest anything new, beyond what is already known about them given the terms of any particular ontology. They would be like the US Constitution, according to Antonin Scalia: no one would ever get past the original intent of their creators. Things would have no futures, since there would, in principle, be no link between one ontology (the present) and another (the future or the past). Indeed, we might go so far as to say, people would have no futures either, to the extent they would always already be full masters of any conceptual universe they might find themselves within. There could never be any surprises, material or conceptual. But the importance of things for people lies, in part, in the ways they may contribute to new futures. They suggest new possibilities and, given certain novel conjunctions and shifts in experience, can steer people’s attention to new aspects of the world. The postulate of multiple ontologies, in its most radical form, seems to erect insuperable barriers between different parts of people’s present worlds, as well as between their pasts and possible futures.
Keane, Webb 2003. Semiotics and the Social Analysis of Material Things. Language and Communication, 23: 409-425.
__________. 2008. The Evidence of the Senses and the Materiality of Religion. Journal of the Royal Anthropological Institute, 14: 110-127.


  1. Dear Webb,
    It’s a privilege to have such a thorough engagement with our TTT argument from you. In many ways you capture the remit and agenda of the argument from ontological difference that we advance, and the worries you raise are clearly the ones that most need to be addressed to make the argument stick. Crucially, you are of course absolutely right to place the moves we make in the context of anthropology’s constitutive investment in alterity, and to trace it back to the famous E-P take on Zande oracles and such.
    It is also very helpful of you to extend the usual counterargument from incommensurability (viz. if alterity is to be construed as ontological divergence, then how is communication across the ontological divide – including the task of anthropological analysis itself – possible?) to the question of temporality. The analogy, of course, is of the order of ‘the-past(-and-future)-is-a-different-country’. It follows that whatever response one would want to propose to the questions you raise would also be likely to count as a response to the (usual) worry about incommensurability. So here goes one such.
    You put your misgivings about the ontological take on alterity in a number of ways, but it does not seem to me unfair to say that your worries boil down to a charge of undue ‘essentialism’. To use my example from Cuban Ifa divination (incidentally, the powder babalawos use is not red – more greyish; does that count as a thing-driven ‘surprise’ in your terms?!): if we say that the statement ‘powder is power’ has ontological rather epistemic weight, we are inter-defining the essences of powder and power in a way that seems to preclude the possibility that either of them might change. (In fact, you are more worried about the possibility of powder – qua thing – being thus onto-locked, although I’m not sure why you are more worried about things being onto-locked than concepts. It seems to me that all the points you make about the mutability and historicity of things could be made also about concepts, to buy into that distinction for this purpose alone).
    It is not completely clear to me why you draw this conclusion about the ‘onto-locking’ (let’s call it) of things, but I will hazard a guess as to the underlying syllogism: If the notion that powder is power is taken in epistemic terms (an ascription of a concept to a thing), then we indeed have a free rein to posit change. The powder may be power today and stop being it tomorrow (e.g. predication a la Aristotle). But if we posit the identification in ontological terms, then change seems precluded by fiat. If the definition of what powder is in Ifa is that it is power, then a powder that stops being powerful tomorrow thereby also stops being powder. Hence, if I’m not mistaken, your keenness (Keaneness?) to maintain the idea that ‘things in their very materiality exceed any particular concepts’ – a rather particular semiotic ideology, you may agree, inasmuch as the ‘uncanny’ character (to coin a phrase) of ‘sheer materiality’ casts a large metaphysical shadow of its own. (I of course have no reason to suppose that it has much purchase on Ifa at any rate – whether it does is an open ethnographic question at least, though I do take some steps to close it (in the negative) in my own chapter in TTT).
    It is important to make clear that this distinction between epistemic and ontological takes captures very well the virtue of the latter, as I see it, when it comes to dealing with alterity. I have argued this point in a number of ways (note, the big arguers of the case are Viveiros de Castro very explicitly and Roy Wagner somewhat more implicitly, and I draw on both in my own writing, as we did in TTT – now I’d also add Terry Evens to the list). Here I just note one point in its favour, which turns directly on the distinction as I’ve ascribed it to you. If, as you suggest, things are to be understood as ‘exceeding’ concepts, then one can always raise the (epistemic) question: Which concepts ‘fit’ them best? On this view, then, alterity becomes a matter of different people (or, for you perhaps, same people at different times) ascribing different concepts to the same things – same inasmuch as the different concepts are ‘about’ the same thing, as the concept ‘power’ is about ‘powder’ when babalawos say ‘powder is power’. In other words they are disagreeing about the nature of powder (and, I would add, about the nature of power too). And if this is so, then it is also in principle sensible to ask who is right in this dispute – is powder really power, or do they just think it is? By contrast, if things are to be taken as coterminous with concepts, then this kind of epistemic judgement becomes a category mistake – a silly question. If powder is defined as power, then to wonder whether it really is power is a mistake rather like that of wondering whether bachelors really are unmarried men – a misunderstanding rather than a disagreement. And a basic advantage of this (there are a number of other ones) is that it robs the anthropologist from the option of seeing her/his analysis as an answer to the question of why people might ‘believe’ such strange (ultimately false as far as we can possibly be concerned) notions as ‘powder is power’ (E-P’s ‘twins are birds’ would do just as well, just to evoke the kinds of ‘apparently irrational beliefs’ approaches the ontological move is designed to foreclose).
    Thus framed, your critique becomes a baby/bathwater issue: you are claiming that in discarding the bathwater of ‘apparently irrational beliefs’ (and all the colonial chauvinism that it supports, I’d add, following Viveiros and also Vassos Argyrou) we are also loosing the baby of change – the irreducible ‘openness’ that characterises the trajectories of things (like powder) in the world and in time.
    But it seems to me that this worry is more a function of the assumptions as to how to think about change that you build into your critique, than it is a shortcoming of our ontological move. Definitely: if change is to be parsed metaphysically as some version of Aristotelianism, which I take minimally as any analysis that posits change as a shift in the relationships between subjects (content, thing) and predicates (form, concept), then we have a problem. But to assume that change can only be conceived in such terms would be odd, particularly when dealing with an approach that makes an explicit virtue of collapsing these kinds of ‘Aristotelian’ distinctions. Put simply, if, as you fairly acknowledge, our whole argument relies on refusing to take the distinction between concepts and things as axiomatic, then you can hardly expect us to assume that a take on change that relies on just that distinction (sheer materiality and so on) would itself be carried as axiomatic.
    Now, I don’t need to point out to you that a number of alternatives to Aristotle (or versions thereof) are available in the philosophical market – dunno… Bergson, Deleuze, Latour, all very fashionable, and we are certainly ‘in’ with that in some sense, admittedly (though that raises problems of its own, which I won’t address here, but which some of my work since has tried to do). I should also say that, again, my own chapter in TTT is, among other things, an explicit attempt to advance an alternative to such subject/predicate takes on change. The whole point of my ‘powder is power’ argument (rather than its use as an illustration of alterity in the introduction to the book) is to yield what I call a *motile* analytic. One way to capture the peculiar features of that analytic (forced upon me by the practice of Ifa) is to say that in it change becomes not a ‘disturbance’ of things but their constitutive feature – powder is powerful because it *is* change, rather than being ‘subject to’ change, and similarly for the deities that it is charged with making appear during divination. So in a sense my own analytic is even more sensitive to change than one based on the irreducible ‘openness’ of ‘excessive’ materials – though that comparison is one worth pursuing further I think.
    But I want to make only a point that comes straight out of the Introduction to TTT, to which your commentary is, I think, primarily directed. Certainly in one sense you are right: In that text we are focusing on the problem of alterity as such, and to do so we present the powder-is-power palaver in a pretty static way. What do we do about the fact that people like Cuban babalawos insist on such apparently weird things as powder is power?, is our question. What would we do about the equally demonstrable fact that what people like Cuban babalawos might insist on might change over time (along with the things’ about’ which they appear to be insisting) is a further question, which we do not address, and which lies at the heart of your critique. The issue, then, is whether our answer to the first question precludes an answer to the second one. I think it doesn’t at all. To see this, consider our answer to the first question.
    We deal with the strange claim ‘powder is power’ by casting it ontologically. To invoke a comment you make in passing about the verb ‘is’ and its philosophical baggage, we take the ‘is’, here, as ‘definitive’ rather than ‘predicative’. The strangeness of the notion, then (its alterity), comes down to its divergence from our initial assumption that whatever powder might be it is *not* power –a ‘commonsense’ assumption, let’s say, invoking E-P again. So when Cubans say that it is, they are propounding a definition of something quite different – another powder, rather than another view about powder (that’s the move). So what do *we* do, now, as anthropological analysts? Well, we do what I try to do in my chapter, namely *change* our concepts of what *counts* as powder *and* power in such a way as to arrive at a position from which we can say, along with Cuban diviners: yes, powder (thus conceptualised) is indeed power (thus conceptualised). My ‘motile logic’ is the result of such a process of ‘re-conceptualisation’ (I call the process ‘ontography’ to underline its ontological stakes: what *is* powder, power and their corollaries, rather than what must Cuban diviners ‘think’ it is, and why they might think this – the ‘apparently irrational beliefs’ heritage).
    But the point is that such an approach would be a complete non starter if I didn’t assume, contra Aristotle et al, that my conceptualization of powder and power in my analysis can *change* and, furthermore, that such a change counts as an ontological change in as much as it pertains to what *counts* as powder/power rather than to what I think about it (in philosophical terms, again, it is a change in the ‘intension’ of the concepts involved, rather than a change in their ‘extension’ that would leave their ‘referents’ (things) indifferent). So the whole approach is premised on the idea that the possibility of change is at the heart of these things we call ‘things-cum-concepts’. Furthermore, the whole approach is, in this very way, designed to show that the worry about incommensurability (and ipso facto about trans-temporal incommensurability) is misplaced. Only if you assume that concepts as such do not change, but only their relationships to the things they predicate changes (Aristotle etc.), does the worry about incommensurability carry force. Our whole approach is premised on the opposite.
    In sum: I don’t need objects to ‘exceed’ concepts in order to make sense of the demonstrable fact that they are mutable. And this is because the concepts which they lock within themselves (in your manner of speaking) are themselves mutable – they are excessive unto themselves, as it were. And this is not an immutable metaphysic I’m proposing – just a method for undoing such things! But that’s for another post…
    Thanks again Webb!
    PS. Sorry I can’t resist this kind of thing: things-cum-concepts are not onto-locking; they’re onto-leaking. 🙂

  2. Sorry, I forgot to state: as with my exchange with Danny in an earlier entry on this, the comments above are made on my own behalf only – not my co-editors’, Ami Henare (now Salmond) and Sari Wastell, whose views on these matters may well be different.

Leave a Reply