The Power of Powder
Life-force and motility in Cuban divination



Along with ‘totem’ and ‘taboo’, the Oceanian term ‘mana’ is one of the few words to have crossed over from the language of ethnographic report to the vocabulary of anthropological analysis – and, partly by that virtue, has also managed to lodge itself in the general lexicon of Euro-American intellectual discourse (Bracken 2007). In this sense, the term has become something of an icon of itself. Having translated mana variably, and with vague traction, as ‘life-force’, ‘sacred power’, or even ‘energy’1, it was always its transgressions that most impressed anthropologists: its peculiarly double universality, one might say, of semantic breadth (‘mana is everywhere’, said the native) coupled with geographical diffusion (‘mana-terms are everywhere’, replied the anthropologist). Commenting on Marcel Mauss’s General Theory of Magic (2001 [1902-3]), where the ‘category’ of mana was presented, with Kantian overtones, as the condition of possibility for all magic, Claude Lévi-Strauss diagnoses the sources of anthropological fascination with mana in

the apparently insoluble antinomies attaching to [it], which struck ethnographers so forcibly, and on which Mauss shed light: force and action; quality and state; substantive, adjective and verb all at once; abstract and concrete; omnipresent and localised.

(Lévi-Strauss 1987: 63-4)

Like magic, in other words, mana presents itself as an aberration, since it systematically transgresses distinctions one could be expected to consider axiomatic. It is, to use a mana-term more contemporary than ‘energy’, excessive.

The classic anthropological response to mana has been defensive.2 If, as Lévi-Strauss shows, the problem mana presents is that it transgresses distinctions that are deemed axiomatic, then the strategy for its solution must be to confirm the authority of the axioms by showing how they are nevertheless equipped to ‘explain’ mana. Axioms are, after all, axioms, and should therefore trump any putative transgressions ethnographic exotica such as mana might throw their way.

Lévi-Strauss’s own weapon of choice, for example, is the semiotic distinction between signifier and signified. Having noted the ‘apparently insoluble antinomies’ that Mauss identified in mana, Lévi-Strauss goes on to chastise him for nevertheless attempting to find meaning in them – albeit a meaning that is so ‘singularly ambiguous’ (Mauss 2001: 132) as to provide the premise for ontologically transgressive ‘magical propositions’ (156) . If mana consists of a ‘series of fluid notions which merge into each other’ (134), that can only be because, taken in itself, it has no meaning at all. A ‘symbol in its pure state, [and] therefore liable to take on any symbolic content whatever’ (Lévi-Strauss 1987: 64), mana is the ‘floating signifier’ par excellence. In this sense, says Lévi-Strauss, terms like mana are analogous to words such as ‘thing’ or ‘stuff’ – contentless forms that allow us to speak of things about which we may know little (see also Sperber 1974, Boyer 1986). And by this virtue, he argues, they stand at the very beginning of the evolution of human knowledge. At a loss on how to allocate his ‘signifier-totality’ to an as yet unknown ‘signified’, man at the dawn of language finds solace in the sweeper-like quality of mana-terms, whose role is to ‘soak up’ the ‘signifier-surfeit relative to the signifieds to which it can be fitted’ (1987: 62). The implication being, of course, that Polynesians, and other people for whom mana-terms are more salient than words like ‘thing’ are for us, are closer to the dawn of knowledge – which is to say, not quite yet in its bright of day.

Ingenious as it is, the argument is somewhat perverse. For one thing, whatever mana may be, it certainly is not a thingamajig or a whatyoumaycallit to those who invoke it. Lévi-Strauss’s provocation, in this respect, is no more convincing for being intentional. But more importantly for present purposes, we may note also that by mustering the semiotic machinery of signifiers and signifieds, Lévi-Strauss flatly contradicts the transgressive quality of mana in the very effort to account for it. For even if the logical and ontological excesses of mana can be thought pacifically as a ‘float’, such a float traverses the axiomatic division between signifiers and signifieds, which could only be a variant of mana’s other famous ‘antinomies’ – abstract versus concrete, quality versus thing, and so on.

Explaining mana away in this way may be no crime. Maybe we should just admit that the transgressions of mana are there to be pacified, like the savages who first presented us with them were. In this paper, however, I wish to explore the obverse possibility. Rather than assuming that the task must be to uphold the axioms mana transgresses by showing how they can explain the transgression, my aim is to explore how, in its very transgressions of our axioms, mana may offer occasion to change them – how it may trump them. In other words, if the ‘problem’ of mana—its ‘apparently insoluble antinomies’—is a function of its relationship to our assumptions, who’s to say that it is not those assumptions that require analytical attention rather than, as it was always assumed, mana itself? Who is to say mana could not take the role of axiom in our analytics? (Certainly not the Polynesians.)

My focus here will be on mana’s transgression of one (putative) axiom in particular: namely, the commonplace assumption that ‘things’ must necessarily be thought of as ontologically distinct from ‘concepts’. Such a move is possible just because the ‘universality’ of mana cuts systematically across this distinction: it is both a thing and a concept (e.g. see Mauss 2001: 134-5). So, the question is this. If, as was so widely commented in the literature, mana is both, say, a stone and ritual efficacy—both thing and concept, as we would say—then might thinking through it provide us with an analytical standpoint from which we would no longer need to make this distinction? Might there be a frame for analysis in which mana does not register as an ontological anomaly, as it does when we say—surprised—that it is both thing and concept?

Taking a cue from earlier writers who considered mana as just an Oceanian version of a much wider phenomenon, this argument will be made from an otherwise parochial ethnographic standpoint, namely that of Cuban Ifá, a male diviner cult of West African origin that I have been studying in inner-city Havana since 1998. As we shall see, Ifá diviners’ seemingly nebulous appeal to the notion of ‘aché’, on which I focus, displays apparent ‘anomalies’ that are analogous to the ones anthropologists have associated with mana.


Aché is relevant in one way or other to all aspects of Ifá worship – as well as Santería, the other main Yoruba-based cult in Cuba, with which Ifá largely shares its cosmology (cf. Brown 2003). Its ‘universality’, in this sense, renders it as much a ‘mana-concept’ as mana itself. By way of illustration consider El Monte, the classic monograph on Afro-Cuban religion by Lydia Cabrera (2000), in which aché is mentioned eleven times, and characterised differently in each one of them. Sometimes Cabrera writes of aché in the abstract as ‘grace’3 (ibid: 16), ‘magical power’ (ibid: 99), ‘all the powers, force, life, the secret of the earth (ibid: 103), or ‘luck’ (ibid: 301). But, elsewhere, aché appears concretely as ‘Orula’s [ie the patron deity of Ifá] grace [kept by the Ifá priest] in his saliva’ (ibid: 106), a ‘powder that belongs exclusively to a deity’ (ibid: 481), or, yet more specifically, as iyefá, the white powder full of virtues which is spread on to Orula’s divining-board’ (ibid: 494). Furthermore, aché appears as something with which deities are born (ibid: 314), or it may inhere in plants (ibid: 113), or be invested on idols through consecration rites (ibid: 103). But rituals themselves may ‘accumulate’ aché through the presence of plenty of initiates (ibid: 108); indeed aché is also the kind of thing that initiates themselves can ‘have’ or ‘give’ (ibid: 108).

Here I focus on the role of aché in the cosmology and practice of Ifá divination. This is hardly arbitrary, since it is mainly as diviners that Ifá initiates (called ‘babalawos’) are distinguished from other Afro-Cuban cult practitioners, including practitioners of Santería, and divination furnishes the basic organising principle for other aspects of worship. Furthermore, in terms of the present argument, it is in divination that abstract and concrete senses of ‘aché’ come together most clearly.

Asked how aché relates to divination, babalawos’ initial response is most often that aché is the power or capacity (in Spanish usually ‘poder’ or ‘facultad’) that enables them to divine in the first place: ‘to divine you must have aché’, they say. In fact, conducting the séance, as well as other rituals, such as consecration, is also said to ‘give’ the babalawo aché, which he may also ‘lose’ if he uses his office to trick people or do gratuitous evil through sorcery. The importance of aché as an enabling condition or force is enshrined in the liturgy of the divinatory séance, with babalawos invoking it by name as part of the various Yoruba incantations that need to be chanted for a successful divination. But while these invocations were usually explained to me in rather vague terms—sometimes as appeals for the aché of ritual ancestors, other times as solicitations of the aché of nature, or of the deities, or of Orula (the patron deity of divination)—there were evidently also senses in which aché was understood much more precisely, to refer specifically to the secret powders that are an indispensable ingredient in just about all Ifá ceremonies, including divination. I shall describe only the uses to which aché-powders are put, and not their ingredients, which babalawos guard closely, since—and this is, really, the point—powders are a principal source of their divinatory powers.

Unlike spirit-mediums and other seers, babalawos divine only with the help of certain consecrated paraphernalia. The method is basically similar to other geomantic systems found throughout sub-Saharan Africa and elsewhere. Most importantly, at initiation, babalawos receive a divining-board and a number of palm nuts (ikines) of which they use 16 in order to divine for ceremonial occasions from then on. This is done by clutching all 16 nuts with both hands, and then separating most of them off with the right hand so as to leave either one or two nuts in the left. If only one remains, the babalawo marks two lines with his middle and ring finger on a layer of aché-powder, also called iyefá, which is spread on the surface of his divining board. If two nuts are left, he marks a single line with his middle finger. The process is repeated until eight (single or double) marks are made on the board, arranged in two columns of four (referred to as ‘legs’). In what we might call a random way, this yields one of 256 possible divinatory configurations, referred to either in Yoruba as oddu or in Spanish as signos (signs) or letras (verses). Each oddu is connected to a series of myths that are interpreted in various ways during the latter parts of the divinatory séance so as to give pertinent advice to consultants (see Holbraad 2003, 2008b).

Practitioners emphasise that all the paraphernalia involved only work, as it were, provided they are properly consecrated during the ceremony of initiation, and this most crucially involves ‘charging’ or ‘loading’ (cargar) each item with aché-powder in secret ways. When I asked what would happen if one were to conduct a divination with ‘jewish’ equipment (as yet to be consecrated objects are called), babalawos dismissed the idea: Orula ‘does not speak’ with such objects, they said. The notion that Orula literally speaks in divination is important in this context. For the consecration of the palm nuts, in particular, is, in fact, conceptualised as the birth of Orula himself, or, better, the birth of an Orula, since each babalawo has his own, so there are many. As a general principle in the polytheistic cosmology of Ifá and Santería, while deities may be thought of as transcendent beings who reside in the temporal beyond of mythical time and the spatial one of nature or the sky, they are also given a radically immanent role in the form of idols (Bascom 1950, Peel 2003: 94). Unlike Christian icons and the like, these idols, which usually are stones (otá) placed inside decorated pots (soperas), are not taken to ‘represent’ the deity, but rather to be it and are, hence, fed with blood, spoken to and generally taken care of in ritual contexts. So, in this sense, Orula simply is the palm nuts (which, by the way, when not being used to make Orula ‘speak’ in divination, are kept in a clay pot on a shelf in the babalawo’s home).

Connected to this is the idea that the divinatory configurations that the babalawo marks on the powder—the oddu—are, also, considered to be divine beings. The exploits of each individual oddu feature in countless myths, wherein oddu may appear as kings, warriors, tradesmen, animals, etc., or, even, as guises of Orula himself. In fact, on this score, there is a certain amount of ambiguity insofar as babalawos often speak of these divine personages as ‘paths’ of Orula (caminos)—rather than minor deities in their own right—by analogy with the ‘paths’ that other popular deities (orishas) have, each orisha being basically a multiplicity of different ‘aspects’ or ‘avatars’ in Yoruba cosmology (e.g. Bolívar 1994: 27-64). Be that as it may, the point is that there is nothing ‘arbitrary’ about these ‘signs’, as they are often referred to in Spanish (signos). When the babalawo marks the eight single or double marks on the powder of his divining board, the oddu in question is said, literally, to ‘come out’ (salir), or, more actively, to be ‘drawn’ (in the gambling sense – sacar). And there is no mistaking its potency: crouching around the divining board, the babalawos and their consultants are in the presence of a divine being, a symbol that stands for itself if ever there was one (sensu Wagner 1986).


So, aché is excessive like mana: power and powder, abstract and concrete, concept and thing. But I would argue that this does not index an antinomy, as Lévi-Strauss would have it. Nor, for that matter, are we faced with an ordinary ambiguity to be compared, for example, with the English concept of love, which can hardly be said to have a single meaning that incorporates filial and erotic senses. For, while babalawos certainly distinguish between the different senses in which they use the word ‘aché’—no one is confused about the difference between power and powder—they also assume (and, if invited, explicitly draw) a clear logical connection between the two senses. To have the power of aché as a diviner, one must be properly consecrated as a babalawo and this, most crucially, involves receiving and knowing how to use the consecrated equipment, charged with the powder of aché. No powder no power, so to speak. Conversely, the secret knowledge required to prepare aché powders and use them for Ifá is possessed only by babalawos – a term they translate from Yoruba as ‘father of secrets’ (Menéndes 1995: 51). In other words, preparing and using these powders is within the power of babalawos exclusively – so ‘no power no powder’ also.

It is worth being clear about the logical status of this mutual implication, since it goes to the heart of the ‘anomaly’ aché poses – the heart of the problem about concepts and things. For there are two ways of glossing this implication that make it appear quite unanomalous, so to speak, though clearly unusual – or even ‘irrational’, as the old phrase has it (Sperber 1985). The first would be to gloss the implication in causal terms: no powder no power and vice versa, because each is a necessary causal condition for the other. The problem, then, would simply be one of showing why certain people ‘believe’—to use a hackneyed word—in such seemingly strange causal sequences. We can all understand how, say, gunpowder was a necessary condition for the power of the Conquistadors, and how, conversely, the gunpowder’s power was predicated on its Spanish makers’ privileged knowledge of how to produce it. Why might Cuban diviners posit an analogous relationship between their power and their powder, given that no causal efficacy seems ‘actually’ to be involved?

The problem with this causal gloss is that it does violence to the ethnography. For, like all causal sequences, the circular one proposed here is cast in terms of logically contingent relations between discrete elements. On such a view, aché-powders and aché-power are first posited as logically independent from one another and then related ‘externally’ by what philosophers call ‘physical’ necessity. This hardly tallies with what practitioners say. As shown by their reaction to my suggestion that I might start divining with jewish equipment, for them the notion of powerless powder, and of powderless power, is not just untenable as a matter of fact, but rather inconceivable as a matter of principle. A babalawo who hasn’t been properly consecrated with aché powder just isn’t a babalawo. And powder that hasn’t been prepared properly by a babalawo—father of these secrets—just isn’t aché. In other words, the relationship between power and powder is, philosophically speaking, ‘internal’: each is defined in terms of the other.

This suggests an alternative gloss on the mutual implication, in terms of logical, rather than physical, necessity. On this view, the slogan ‘no powder no power’ (and its converse) is to be taken as what philosophers call ‘analytic’, of the same order as statements like ‘the Queen is Head of State’, or ‘2 + 2 = 4’. Just like 2 isn’t 2 unless added to itself it gives 4, so a babalawo isn’t what he is (i.e. doesn’t have the power of aché) unless he has been consecrated with the powder, and vice versa. The question, then, would be what it is about the meaning of power and powder in this context that makes them mutually definable in this way. Why, in other words, do Cuban diviners consider these two concepts mutually constitutive while we, presumably, do not?

Persuasive though it may seem—inasmuch as it reflects the internal relationship of power and powder in Ifá—this analysis in terms of logical necessity is just as inadequate to the ethnography as the causal account, though for opposite reasons. The problem here is that by treating the relationship between powder and power as an analytic implication, one effectively ignores its irreducibly practical character. True, for a babalawo it is enough to contemplate what his power means to know that it requires his consecration with aché powder (and vice versa). But, the point is that this logical operation presupposes a practical one, since it is only because aché powder is actually efficacious that it can be used to produce babalawos ‘with aché’ (ie with power). Babalawos only count as such, provided they have been consecrated with powder, not because the meaning of powder logically implies their power (like the meaning of ‘Queen’ implies being Head of State), but because the powder itself gives them power. (And vice versa for powder.) In other words, the difference between the relationship of powder and power and that described by an ordinary analytic statement is that while the latter states a conceptual identity, the former implies a real transference (of powder that gives power and of power that gives powder).4

At this point my account of this dilemma will appear contradictory. It seems as if what I’ve done is complain of the causal analysis that it distorts the logical implication that binds power and powder, only to reject the logical analysis on the grounds that it distorts the causal character of the implication. Which is it: logical or causal? But I would argue that this appearance of contradiction is a consequence of the Procrustean character of the analytical choices we seem to have at our disposal. Indeed, we may note here that the distinction between causal connection and logical identity, which renders aché contradictory, is corollary to the distinction between things and concepts. It is precisely because aché does not fall tidily into either of the latter categories that it does not lend itself to analysis in terms of either of the former relations. If aché powder and aché power were ‘things’, then they could be conceived of as discrete variables: which, as we have seen, they cannot be. If they were concepts, then their internal connection would be a purely deductive matter, whereas, as we have seen, it is not. So, properly speaking, aché is neither thing nor concept but rather a bit of both: an indiscrete thing and a concept that literally transfers itself. Clearly, for as long as we insist on thinking in these terms, we can only articulate aché as a paradox, as Lévi-Strauss and others did.

The strategy of this paper is based on the idea that this negative analytical predicament actually prescribes a positive methodology that may lead to an analytical resolution. If aché is to be taken as both what we call a concept and what we call a thing, then it follows that the connection between the two sides of its ‘double aspect’, so to speak, is not arbitrary. That is to say, while the logical status of aché is still obscure due to its apparent ‘excess’, we do know one thing: that its abstract meaning as ‘power’ is internally related to its concrete nature as powder. So the meaning of aché (the ‘concept’) is literally constituted by the things to which it would otherwise be assumed simply to ‘apply’. Its intension is modified by its extension, if you like, by what one might call a relation of ‘hyper-metonymy’ (imagine a crown that didn’t just signify royalty, but actually constituted it – a ‘magical’ crown, then).

On this hypothesis, which—as we have seen—is motivated by the ethnography, the task is to understand aché in its very excess. And the opportunity for doing so is there, in the ethnography of the thing itself, in powder. For, with the axiom of concept versus thing discarded, the ethnography of things like powder can no longer be assumed to be about ‘interpreting’ them in terms of the meanings the people we study ‘attach to’ them. Things may carry their own context within themselves, as Marilyn Strathern has put it, writing of other things (Strathern 1990). So the method aché dictates for itself could be captured by the old phenomenological injunction: ‘back to the things themselves’. But only with the proviso that this is not because somehow the level of things and of people’s ‘practical’ engagement with them is more ‘primordial’ than the level of theoretical manipulation of concepts, as some anthropological versions of phenomenology would sometimes have it. Rather, because the internal relationship between concepts and things implies that in some important sense things just are concepts. So, our job is to think through them, as it were, rather than about them (cf Henare et al 2007). In particular, this method allows us to determine the logical status of the ‘hyper-metonymy’ relating powder to power in Ifá, by attending to the role of the ‘thing itself’—powder—in Ifá divination.

So, what makes powder power in Ifá? The beginning of a reply may be given by making explicit what divinatory power consists in for babalawos. As we saw, the difference between unbaptised and consecrated equipment is that Orula only speaks through the latter, and this is because he is deemed to be the latter: when he speaks, he is 16 palm nuts being cast into oddu-configurations, marked on the powder of the divining board; and those marks themselves are Orula’s oddu, not just their representations. So, the divinatory power of the babalawos can fairly be glossed as their ability to render Orula and his oddu immanent during the consultation: an otherwise transcendent deity, who is imagined to have lived and died many generations ago in ‘Africa’, and who now is deemed to reside in natural features, the sky, or mythic time as what one might call an ‘ex-human’, is made temporarily present during the séance. Divinatory power, then, most crucially involves the capacity to engender what we might call ‘ontological leaps’ on the part of the deity: from transcendence to immanence or, more simply, from radical absence (the ‘beyond’) to presence. (These leaps are the polytheistic counterpart to the epistemological ‘leaps of faith’ associated with monotheism, though even God—in Christianity—needed His Son as a one-off ontological transgression, or at least He thought we did).

Before going on to explain how aché-powders may be said to condition such leaps, as the logic of ‘no powder no power’ would imply, it is necessary to comment on the problem they present – the ‘problem of transcendence’. The problem, which babalawos have the power to solve, amounts to the danger that Orula and the rest of the orishas might remain in a state of transcendence, permanently separated from humans in the ‘beyond’. Such a state of affairs would render all aspects of Ifá worship impossible, including not only divination, but also initiation, consecration, sacrifice, and magic, all of which are premised on the idea that deities and humans can enter into relations with each other (Orula’s speech to the consultant through the oracle, mortals asking the deities for divine favour, feeding them blood, bringing divine power to bear on personal affairs in sorcery, etc.).

My argument depends on the idea that, despite its theological twist, this problem is familiar at a more abstract philosophical level as the classic problem of ‘individualism’. The question, which goes back at least to the social contract theorists of the 17th and 18th centuries, is how from a position of an aggregate of individuals, separate and self-contained units that are transcendent with respect to one another, we might arrive at a position where these units are formed in relation to each other – how, in human terms, ‘society’ is created out of ‘individuals’. A peculiarly anthropological dissolution of this question goes back to at least Mauss, and has in recent years become particularly associated with Melanesia, due to the work of Marilyn Strathern (1988, 1995; see also Dumont 1970, Marriott 1979). That is to say that the question of how relations might be engendered out of individuals is itself arbitrary, for one could perfectly well ask how individuals are engendered out of relations. Indeed, anthropologically speaking, phenomena like magic, or gifting, or certain kinds of tribal leadership, or caste, or hunting and shamanism, or affinity, or, even, anthropological analysis itself, will be endlessly misunderstood unless we perform exactly this kind of analytical reversal, viewing relations as logically primitive, and the terms they relate (‘individuals’) as derivative effects.

While I cannot comment on gifting, caste, affinity and so forth, I would suggest that, with regard to Ifá divination, the choice between giving priority to relations over individuals or vice versa is, at least, problematic and, ultimately, false. Crudely put, if one were to say that the problem of transcendence in Ifá is not really a problem because deities and humans mutually constitute each other in the relationships that divination implies, one would be denying the very condition that leads clients to the diviners in the first place, namely that the deities are transcendent most of the time, so that the diviners’ powers are necessary in order to elicit them into relation. In this sense, relation and transcendence are symmetrical in Ifá cosmology, so a choice of giving priority to one over the other must be false (cf. Højer 2005). Therefore, the analytical question is how relation and transcendence might themselves be related, other than by antinomy.

Powder gives us the answer, and to see this we may pay attention to its role in divination. As we saw, spread on the surface of the divining board, powder provides the backdrop upon which the oddu, thought of as deity-signs, ‘come out’. In this most crucial of senses, then, powder is the catalyst of divinatory power: i.e., the capacity to make Orula ‘come out’ and ‘speak’ through his oddu. Considered prosaically, powder is able to do this due to its pervious character, as a collection of unstructured particles – its pure multiplicity, so to speak. In marking the oddu on the board, the babalawo’s fingers are able to draw the configuration just to the extent that the ‘intensive’ capacity of powder to be moved (displaced like Archimedean bathwater) allows them to do so. The extensive movement of the oddu as it appears on the board, then, presupposes the intensive mobility of powder as the medium upon which it is registered. Of course, physically speaking, this is always the case – movement presupposes movement. Even if the babalawo marked the oddu with a pencil on a piece of paper, the lead would only leave a mark provided the paper particles reacted accordingly. But the point is that powder renders the motile premise of the oddu’s revelation explicit, there for all to see by means of a simple figure-ground reversal: oddu figures are revealed as a temporary displacement of their ground, the powder.5

This suggests a logical reversal that goes to the heart of the problem of transcendence. If we take seriously babalawos’ contention that the oddu just are the marks they make on aché-powder—as we must if we are intent on thinking through these things—then the constitution of deities as displacements of powder tells us something pretty important about the premises of Ifá cosmology: that these deities are to be thought of neither as individual entities nor as relations, but rather as motions. Indeed, beyond the role of powder, this accords with the mechanics by which the oddu get determined in the first place, since, as we saw, Orula himself is a plural object: 16 palm nuts that only ‘speak’ through motion – the diviners’ cast. This also accords with the otherwise perplexing idea that all orishas manifest themselves as one of a number of ‘paths’, as I mentioned.

If the oddu of Orula, as well as the orishas more generally, just are motions (or ‘paths’), then the apparent antinomy of giving logical priority to transcendence over relation or vice versa is resolved. In a logical universe where motion is primitive, what looks like transcendence becomes distance and what looks like relation becomes proximity. Motions through and through, the deities are never divorced from humans, stuck in the ‘beyond’ of transcendence – to say so would be to place limits on the logical priority of motion. Conversely, humans’ relations with motile deities cannot be taken for granted, as the Melanesianist image would have it, for there is no guarantee that the deities’ movement will be elicited in the right direction, as it were. The relation, then, is potential, and it is just this potential—the potential of directed movement—that aché-powder guarantees, as a solution to the genuine problem of the distance deities must traverse in order to be rendered present in divination.

The notion of potentiality here, and particularly that of potential relation, is closely akin to arguments presented by Eduardo Viveiros de Castro and Marcio Goldman, on the role of ‘virtuality’ in Amazonian and Afro-Brazilian religious cosmologies respectively (Viveiros de Castro 1998a, 1998b, 2002, 2005, 2007; Goldman 2003, 2005, nd). Drawing on the Deleuzian conceptions of ‘virtuality’, ‘difference’, and ‘becoming’ (eg Deleuze 1994), their analysis of these cosmologies turns on a figure-ground reversal that is closely analogous to the one powder just performed for us here. Here I want to pause to compare in some detail the ‘motile’ analysis that I have offered for my Afro-Cuban material with Viveiros de Castro and Goldman’s parallel ‘virtualist’ analyses of Amazonian and Afro-Brazilian cases, since, as we shall see, such a comparison serves to sharpen the analytical purchase of my argument considerably. Their point (and I paraphrase wildly across texts for the sake of brevity) can be summarised like this.

Anthropologists tend to describe the cosmologies they study as systems of classification. The assumption is that cosmologies are populated by different entities (gods, ancestors, spirits, and so on) that relate to one another in different ways (hierarchically, genealogically, temperamentally, or whatever). Cosmologies can be ‘charted’ by placing these entities in relation to each other in conceptual space—or, indeed, on paper—according to their differences, characterising their relations in the spaces that are provided—notionally or graphically—between them. Now, try imagining a figure-ground reversal on such a chart, as in an Escher sketch. Cosmological elements now feature not as self-identical marks relating to each other externally in space (the ‘scheme’, the ‘paper’), but are rather extended across the spaces that previously divided them. What was assumed to be a scheme of entities, now appears as a field of relations, so that the differences that previously distinguished one cosmological element from another ‘extensively’ now become ‘intensive’ characteristics of those elements themselves, now conceived of as ‘self-differentiating’ relations. Showing that such a ‘plane of immanence’ (2007: 155) underlies pan-Amerindian notions of myth, spirits, and shamanism, Viveiros de Castro writes:

[T]he actants of origin myths are defined by their intrinsic capacity to be something else; in this sense, each mythic being differs infinitely from itself, given that it is posited by mythic discourse only to be substituted, that is, transformed. It is this self-difference which defines a ‘spirit,’ and which makes all mythic beings into ‘spirits’ too. […] In sum, myth posits an ontological regime commanded by a fluent intensive difference which incides on each point of a heterogenic continuum, where transformation is anterior to form, relation is superior to terms, and interval is interior to being.

(2007: 158)

While Viveiros de Castro does not add the priority of motion over rest to his list, it is clear that such a logical reversal, which I have argued is necessary to make sense of the ‘intrinsic capacity’ of Ifá deities to move from transcendence to immanence, is confluent with his argument. Of course, a full discussion of these analogies and their possible breakdowns would involve far-reaching comparisons between Amerindian ‘animism’ and Afro-American ‘polytheism’ – ‘spirits’ versus ‘deities’, so to speak. Here, I shall only make use of the analogy with Viveiros de Castro’s argument on virtuality to make two points—one positive and one negative—that may help to sharpen my own argument on motility. Both points pertain to the question as to how Viveiros de Castro’s use of the concept of ‘intensive difference’ fairs in relation to the problem of deities’ transcendence: which, as we have seen, aché has the power to solve.

Firstly, it will be noted that the point of analogy between the virtual and the motile lies precisely in the idea of potentiality, which is a common corollary of both (see also Holbraad 2008a, 2008b). As we have seen, if aché forces us to conceive of the oddu of Ifá as motions, it also allows us to think of the oddu as having the potential to become immanent in divination, so as to enter into relations with the babalawos that invoke them. In this sense, being motions, oddu are also ‘potential relations’. Analogously, if for Viveiros de Castro spirits are virtual in the sense that they are ‘self-differential’, then they too should be construed as potential relations inasmuch as their self-difference just amounts to their inherent potential to ‘be something other than themselves’ (Lévy-Bruhl’s 1926: 76). Indeed, more than just analogous, these two senses of ‘potential relation’ stand in a relationship of logical implication. For the oddu’s potential to enter into relations with humans is premised on what Viveiros de Castro calls ‘self-difference’. Oddu do not simply ‘travel’ from the beyond of mythical transcendence to the here of the divining board, for their ‘motion’ is not one of a self-identical entity. As we have seen, the capacity of oddu to reveal themselves in divination implies a transformation, which resembles the one Viveiros de Castro envisages for Amerindian spirits. Be they conceived as ‘paths’ of Orula or as deities in their own right, the oddu are ‘posited’ as characters that reside somewhere in the beyond as variable mythical guises, only to be ‘substituted’ during the divinatory séance, first as configurations of the palm-nuts and then as ‘signs’ on the aché powder. In other words, oddu can relate to ‘others’ just because they can ‘other’ themselves, inasmuch as their motion from transcendence to immanence is premised on their capacity to ‘self-differentiate’.

The upshot of this is that the motion of the oddu as they ‘come out’ on the divining board should not be conceived in spatial terms at all, but rather in ontological ones. Aché, then, is the space in which ontological transformations happen, and its role on the divining board as a ‘register’ (registro6) is also ‘ontological’ through and through. In the motile universe of Ifá, the very act of registration on the surface of the divining board—as the babalawo’s fingers move through the powder to reveal the oddu—is not an ex post facto representation of an already pertaining state of affairs, but rather an act of ontological transformation in its own right, for it is in this act that the oddu is ‘substituted’ as an immanent presence in the séance.

Indeed such an analysis of aché as the premise/catalyst of transformation can arguably be generalised in Ifá, beyond the immediate context of divination. For present purposes the most pertinent ethnographic evidence for this has to do with the role of aché, as conceived by practitioners on this broader cosmological scale. Marcio Goldman’s characterisation of the notion of ‘axé’ in Candomblé, the Afro-Brazilian cousin of Cuban Santería, is pertinent:

[Candomblé cosmology involves] a kind of monism that supposes a single essence that diversifies into various modalities that constitute all that exists or that can exist in the universe. This essence, which is clearly similar to the Melanesian notion of mana […], is referred to in Candomblé as axé. The diversification of axé is initially manifested in the divinities themselves, the Orixás, since each of them incarnates a specific modality of the general essence. In turn, each thing or being that exists in the world – stones, plants, animals, human beings, etc… – ‘belongs’ to one of these Orixás to the extent that they share with him this essence, which is at the same time both general and individual.

(Goldman nd: 1-2, my translation from the original Portuguese)

Although babalawos in Cuba have not given me such a concerted cosmogony regarding aché (their varied accounts tend to focus on the generative role of ‘major’ orishas at the beginnings of time), Goldman’s synthesis does reflect babalawos’ common observation that the orishas and their worldly ‘belongings’ ‘have aché’. Indeed, from the point of view of such statements, Goldman’s paradoxical appeal to the notion of ‘essence’ (eg “both general and individual”) is perhaps unnecessary. Building on my earlier argument about aché in divination, an alternative analysis would posit aché not as diversified essence, but as the premise of diversification itself.7 Orishas ‘have aché’ precisely inasmuch as they are able, qua motions, rather than entities or relations, to ‘become’ the various elements of the world: stones, plants, animals, humans… These, in turn, ‘belong’ to the orishas just in the sense that they are them: they are varied outcomes of the orisha’s motile becoming, and hence ‘have aché’ also.

This, arguably, is the significance of the ritual requirement in Ifá (and Santería) that all consecrated items be physically ‘loaded’ with aché-powders. This includes not only babalawos’ Orula deity and the divinatory objects that go with it (see above) but, also, all the other deities practitioners receive as ‘loaded’ idols at different stages of their initiatory career (Holbraad 2008a), as well as the initiates themselves, who are ‘marked’ with aché-powders at various parts of their body during initiation. Just as the powder babalawos use on their divining boards is powerful as a surface on which Orula’s oddu can ‘come out’, so consecrated idols and initiates are powerful (‘have aché’) as conduits that render the presence of the relevant orishas immanent, particularly as and when this is required in ritual. In light of our earlier analysis of the role of powder as the pervious ‘ground’ on which deities manifest as immanent ‘figures’, it makes sense that powder should also be the ‘active ingredient’, so to speak, of consecration. Admittedly, its role as motile ground here is—literally—not as graphic. Powder is not itself marked, but rather is either ‘loaded’ in small portions into secret cavities of the idol-deities, or used to ‘mark’ the bodies of neophytes. One is tempted to say that the power of these pinches of aché-powder is metonymic, though only on the proviso that this is a ‘hyper-metonymy’ in a strict and pertinent sense (see also above). Unlike ordinary metonymy in which a part comes to stand, symbolically, for the whole (e.g. Crown for King), the pinches of powder that are used in consecration do not merely ‘stand for’ the whole from which they are partitioned but, rather, they reconstitute it as wholes in their own right. This follows from a second, ‘prosaic’, property of powder. As a pure multiplicity of particles, powder is not only pervious, as we saw, but also ‘partible’: even pinches of powder constitute wholes, inasmuch as no qualitative difference (other than quantity!) distinguishes them from the wholes from which they were detached (cf. Reed 2007). So, even though in consecration powder does not actually display motility as it does in divination, it does retain literally the property (viz. perviousness) that would allow motile deities to be rendered immanent, and thus is powerful in the same sense – albeit, of course, in principle.

Be that as it may, it is clear that in consecration powder is power in the same sense as it is in divination: namely, as a catalyst for the ontological transformation of the orishas from a state of transcendence in the ‘beyond’ to a state of immanence in the consecrated items. This brings us back to a second point of comparison with Viveiros de Castro’s notion of ‘self-difference’. For it should be noted, that the kind of ontological transformation that is at stake in Ifá is in significant ways different from the ones Viveiros de Castro had in mind in the Amerindian context. One could visualise the contrast in terms of a spatial metaphor of ‘horizontal’ versus ‘vertical’ transformations (Hugh-Jones 1996, Pedersen 2001). Viewed in this way, Viveiros de Castro’s account of spirits is ‘horizontal’ inasmuch as it is essentially anarchic. Spirits are conceived as endlessly multiplying, like ‘forests of mirrors’, and their multiplicity is irreducibly qualitative inasmuch as it implies ‘modulations’ of ‘form’ (2007: 161) – of ‘ontic’ form, we might say, mutating from species to species, from animal to human, minuteness to monstrosity. Ifá, on the other hand, in accordance with its imperial associations in West Africa (eg see Bascom 1991, Peel 2003), presents the question of deities’ potential for transformation in irreducibly vertical or, as practitioners also put it, ‘hierarchical’ terms. What I have in mind here is not primarily the much debated ranking of deities in the form of a systematic ‘pantheon’, whose historical evolution, as David Brown has argued convincingly, is heavily bound up with Christian and other ‘theologising’ influences both in Africa and in Cuba (Brown 2003: 114-128). Rather, what makes Ifá ‘vertical’ is the cosmological premise upon which such rankings are conceived, namely the idea that deities can be characterised by their degree of ‘distance’ from the human world. Brown writes:

[R]ankings along the spirit–matter continuum can be flexible and ambiguous. For example, Echu is, in fact, “everywhere and sees everything”: he is a “warrior” of the highly “material” street and forest; he is also right up there as “God’s secretary”; at the same time, he is not merely a “mischievous” orisha but is, in one road [viz. ‘path’], Alosí, the Devil, who manifests himself in this world […]. Obatalá [sky deity, patron of peace], too, defies ontological confinement. Numerous relatively more “material” and more “spiritual” Obatalás, who are “younger” warrior types or “older” sage types, respectively, populate this continuum

(2003: 127, references omitted).

So if the orishas are as multiple as Amazonian spirits, each of their ‘paths’ taking a different ontic form, their multiplicity is nevertheless ‘vertically’ distributed. They differ from the Amazonian paradigm in that their ontic transformations also imply shifts in what one might call ontological status, since their multiple becoming is inflected hierarchically as a ‘continuum’ of relatively proximate and relatively distant ‘manifestations’. Indeed, the distinction between shifts of ontic form and of ontological status allows us to conceptualise the difference between (horizontal) shamanism and divination as the respective modes of divine disclosure in Amazonia and in Ifá. The shamanic ability to ‘call spirits’, explains Viveiros de Castro, is a mater of ‘vision’: where non-shamans just see animals in the forest, for example, shamans see spirits (Viveiros de Castro 2007: 159-160). This makes sense, since the ‘problem’ that spirits present to humans is that there is more to them than meets the non-shamanic eye – qua intensive differences, they are always more than the sum of their ontic snap-shot appearances (eg the forest animals, which are just a form of their becoming). The problem with the orishas, on the other hand, is not so much that they are invisible but, rather, that they are not fully ‘here’ in the first place. After all, insofar as the orishas are visible at all, it takes no special powers to see them. The idol-deities in which they manifest ceremonially, the natural features of which they are patrons, the devotees whose bodies they possess during Santería rituals, as well as the aché powder that Orula marks during divination – all these concretions are there for all to see. The problem is how to elicit the deities’ presence in these concrete forms – how to elicit immanence, having posited transcendence. One might say that if the shaman’s task is to see what is present, the diviner’s is to render present what is seen.

It follows that, while Viveiros de Castro’s notion of intensive difference is, as already shown, pre-supposed by the idea of motility, such a notion is nevertheless insufficient on its own to account for the peculiar verticality of Ifá deities’ ontological transformations – the problem of transcendence (see also Holbraad 2004, Holbraad and Willerslev 2007). True, such a notion is, in this respect, an improvement on the Melanesianist concept of the ‘relation’, which precludes the possibility of ontological distance altogether. Nevertheless, the distance admitted by the ‘potential’ relations of the virtual is not of a kind that allows us fully to make sense of deities’ transcendence. A matter of ontic form, rather than ontological status, the potentiality of virtual spirits is that of transforming themselves horizontally into what they are not (‘becoming-other’), whereas the vertical axis of transcendence to immanence implies transformations that are also constituted as shifts between orders of otherness (‘becoming-other-kinds-of-other’, if you like).

The idea of motility, I argue, is able to capture these differences between difference. By distributing virtual becomings across a continuum of motion, with its peculiar capacity to ‘self-scale’in terms of the formal relations of ‘distance’ and ‘proximity’ (which, as we have seen, the notion of direction implies), we effectively add a second dimension to the concept of ‘becoming’ itself. Perhaps, the clearest way to express this is in terms of the structuralist distinction between ‘syntagmatic’ and ‘paradigmatic’ relations. Virtual continua relate differences paradigmatically. Motile ones relate them syntagmatically, which is to say that they relate them ordinally, in sequences that provide them direction in terms of asymmetrical (positional) relations of ‘before’ and ‘after’. So, no less ‘intensive’ than their virtual counterparts, motile differences are nevertheless more sophisticated from a logical point of view, in that they are able to render two dimensions of difference—paradigmatic ‘form’ and syntagmatic ‘status’ (or ‘position’)—at once. Both dimensions are needed in order to articulate the problem of transcendence, which, as we have seen, is so central to Ifá cosmology. Motile deities’ transformations allow them to enter into relations with humans. And the fact that these transformations scale themselves as changes of ontological status shows that deity–human relations are not given as cosmological fait accompli, but rather have to be accomplished by eliciting the deities from the relative ontological distance of transcendence to the relative proximity of immanence.


So, the answer to the question as to why aché powder is power is that, in Ifá, powder provides the condition under which deities—ex-human, mythical, of the ‘beyond’—can manifest themselves immanently and, thus, enter into relations with the everyday world of the living. If deities’ moves to immanence are a function of their motility, then aché powder is an essential ingredient for eliciting such moves, since it allows them to be articulated as such – articulated quite literally, as we have seen, on the surface of the divining board as a series of intensive motions (inward displacements) of powder that reveal the ‘figures’ of Orula’s oddu.

By way of conclusion, however, we may note that the analytical dividends of this argument go beyond the immediate concern with aché to the broader question that motivated it, namely the relationship between concepts and things. To see this, consider the strategy of the argument itself. As a response to earlier failures to account for mana’s systematic transgressions of the ontological distinction between concepts and things, my suggestion from the outset was that an ethnographic analysis of the ‘excess’ of aché might provide a conceptual frame within which such transgressions may no longer register as logical absurdities. Driven ethnographically by the logical connection practitioners of Ifá draw between power and powder, I proposed to experiment with the idea that the conceptual properties of aché (as power) could be delineated with reference to its concrete characteristics (as powder) – thus, methodologically, revoking the axiomatic distinction between concepts and things. This line of inquiry led to the analytics of ‘motility’ which, as we saw, render sensible the otherwise absurd-sounding claim that powder is indeed power and vice versa.

One may want to wonder at the circularity of this argument. After all, if the ethnographic analysis of aché had to begin from the stipulation that concepts and things may be identical, then how can it also purport to show it? The circularity involved, however, is arguably virtuous. Rescinding the ontological distinction between concepts and things in order to show that the concept of power is identical to powder would be viciously circular if all it had produced were a confirmation of its own premise. However, the approach has offered more than that. Proceeding from a stipulative identification of concepts with things, it has yielded the analytics of motility. The circle is virtuous precisely because motility does not merely presuppose a collapse of the concept/thing divide but, rather, provides its logical justification. So, if the initial stipulation allowed us, like Wittgenstein’s ladder, to get to the concept of motility, then that concept in turn allows us to discard the ladder of mere stipulation, and accept a novel logical framework that denies the axiomatic dichotomy of concepts versus things.

For the conclusion can only be this. If the motility of powder dissolves the problem of transcendence versus immanence for babalawos, then motility also dissolves the problem of concept versus thing for us. And this, because the latter problem is just an instance of the former. After all, the notion of transcendence is just a way of expressing the very idea of ontological separation. And ontological separation is what a non-motile logic posits at the hiatus that is supposed to divide concepts from things. Motility, on the other hand, turns on the idea that ontological differences do not amount to separations at all, but rather to intensive and ‘self-scaling’ transformations. Thus, just like in a motile logical universe powder can be power, deities can be marks on the divining board, and so forth, so concepts and things can also be each other. All it takes is to stop thinking of concepts and things as self-identical entities, and start imagining them as self-differential motions.


1 Such terms of translation are effective inasmuch they are the closest modern Euro-Americans come to transgressing their own axioms – asking, say, a Foucauldian about ‘power’ is comfortably comparable to asking a Polynesian about mana. However, the notion that such terms, in themselves, may also be theoretically illuminating is misleading, since they do ‘no more than provide difficult native concepts with an equally mysterious gloss’ (Viveiros de Castro 1998a: 79; see also Keesing 1984).
2 In a longer version of the present paper (Holbraad 2007) I trace in some detail the development of the anthropological debate about mana, from the 19th century onwards.
3 Translations from the Spanish text are mine.
4 Hubert and Mauss express exactly this point when they write that the idea of mana ‘not only transforms magical judgements into analytical judgements but converts them from a priori to a posteriori arguments, since the idea dominates and conditions all experience’ (Mauss 2001: 156). Saul Kripke put the possibility of a posteriori analyticity on the philosophers’ table almost a century later, though not much to our use here since mana-terms like aché are anything but ‘rigid designators’ – as the terms of a posteriori analytic truths, for him, must be (eg ‘water is H2O’ – 1980: 48-9).
5 I have made a parallel argument regarding the role of money in Ifá cosmology (Holbraad 2005).
6 The act of consulting the oracle is commonly referred to as ‘looking at one’s self with Orula’ or ‘registering one’s self with Orula” (mirarse con Orula, registrarse con Orula).
7 See Keesing 1984 and 1985 for a critique of the tendency in Melanesian ethnography to view mana as a “diffuse substance”, as opposed to “a process or a state” (1985: 203). The tendency, he argues, is characteristic “of European, not native, theologians” (ibid.), by which he means anthropologists.


Bascom, W. R. (1950), ‘The focus of Cuban Santeria’, Southwestern Journal of Anthropology 6: 64-68
——— (1991), Ifa Divination: Communication Between Gods and Men in West Africa. Bloomington: Indiana University Press
Bolívar A., N. (1994), Los Orishas en Cuba. La Habana: PM Ediciones
Boyer, P. (1986), ‘The “empty” concepts of traditional thinking: a semantic and pragmatic description’, Man (N.S.) 21: 50-64
Bracken, C. (2007), Magical Criticism: the Recourse of Savage Philosophy. Chicago: University of Chicago Press
Brown, D. (2003), Santería Enthroned: Art, Ritual, and Innovation in an Afro-Cuban Religion. Chicago: The University of Chicago Press
Cabrera, L. (2000), El Monte. Miami: Ediciones Universal
Deleuze, G. (1994), Difference and Repetition (trans. P. Patton). London: Athlone Press
Dumont, L. (1970), Homo Hierarchicus: the Caste System and its Implications. London: Weidenfeld & Nicolson
Goldman, M. (2003), ‘Observações sobre o sincretismo Afro-Brasileiro’, Kàwé Pesquisa. Revista Anual do Núcleo de Estudos Afro-Baianos Regionais da Universidade Estadual de Santa Cruz I (1): 132-137
——— (2005), ‘Formas do saber e modos do ser: observações sobre multiplicidade e ontologia no candomblé’, Religião & Sociedade 25, 102-20
——— (nd), ‘Serie, estrutura, devir: Lévy-Bruhl, Deleuze e o Candomblé’, paper presented in round table dicussion on “A Antropologia a Favor da Diferença, ou, Para que Servem os Autores ‘Menores’”, at XXVIII Encontro Anual da Associação Nacional de Pós-Graduação e Pesquisa em Ciências Sociais, 2004.
Henare, A., Holbraad, M. & Wastell, S. (eds.) (2007), Thinking through things: theorising artefacts ethnographically. London: Routledge
Holbraad, M. (2003), ‘Estimando a necessidade: os oráculos de ifá e a verdade em Havana’, Mana 9(2): 39-77
——— (2004), Response to Bruno Latour’s ‘Thou shall not freeze-frame’ (available on-line:’s_%22Thou_shall_not_freeze-frame%22_(Martin_Holbraad), accessed 20 January 2009)
——— (2005), ‘Expending multiplicity: money in Cuban Ifá cults’, Journal of the Royal Anthropological Institute 11(2): 231-54
——— (2007), ‘The power of powder: multiplicity and motion in the divinatory cosmology of Cuban Ifá (or mana again)’, in A. Henare, M. Holbraad & S. Wastell (eds.), Thinking through things : theorising artefacts ethnographically. London: Routledge, pp 189-225.
——— (2008a), ‘Relationships in motion: oracular recruitment in Cuban Ifá cults’, Systèmes de Pensée en Afrique Noire 18: 219-264
——— (2008b), ‘Definitive evidence, from Cuban gods’, in The Objects of Evidence, M. Engelke (ed.), Special Issue of JRAI (N.S.): S93-S109
Holbraad, M. & Willerslev, R. (2007), Transcendental perspectivism: anonymous viewpoints from Inner Asia, Inner Asia 9: 191-207
Hugh-Jones, S. (1996), ‘Shamans, prophets, priests and pastors’, in Nicholas Thomas & Caroline Humphrey (eds.) Shamanism, History and the State, Ann Arbor: The University of Michigan Press, pp. 32-75
Højer, L. (2005), ‘The anti-social contract: enmity and suspicion in Northern Mongolia’, Cambridge Anthropology 25(1)
Keesing, R. (1984), ‘Rethinking mana’, Journal of Anthropological Research 40: 137-156
——— (1985), ‘Conventional metaphors and anthropological metaphysics: the problematic of cultural translation’, Journal of Anthropological Research 41: 201-217
Kripke, S. (1980), Naming and Necessity. Oxford: Blackwell
Lévi-Strauss, C. (1987), [1950] Introduction to the Work of Marcel Mauss (trans. F. Barker). London: Routledge & Kegan Paul
Lévy-Bruhl, L. (1926), How Natives Think (trans. L.A. Clare). London: George Allen & Unwin Ltd.
McKim, M. (1979), ‘Hindu Transactions: Diversity without Dualism’ in B. Kapferer (ed.) Transaction and Meaning: Directions in the Anthropology of Human Issues. Philadelphia: Institute for the Study of Human Issues, pp.109-142
Mauss, M. (2001), [Hubert & Mauss 1902-3], A General Theory of Magic (trans. by Robert Brain). London: Routledge Classics
Menéndez V., Lazara (1995), ‘Un cake para Obatalá?!’, Temas 4 38-51
Palmié, S. (2002), Wizards and Scientists: Explorations in Afro-Cuban Modernity and Tradition. Durham: Duke University Press
Pedersen, M. (2001), ‘Totemism, animism and North Asian indigenous ontologies’, JRAI 7(3) 411-427
Peel, J.D.Y. (2003), Religious Encounter and the Making of the Yoruba. Bloomington: Indiana University Press
Reed, A. (2007), ‘”Smuk is king”: the action of cigarettes in a Papua New Guinea prison’, in A. Henare, M. Holbraad & S. Wastell (eds.), Thinking through things : theorising artefacts ethnographically. London: Routledge, pp 32-46
Sperber, D. (1974), Rethinking Symbolism. Cambridge: Cambridge University Press
——— (1985), On Anthropological Knowledge. Cambridge: Cambridge University Press
Strathern, M. (1988), The Gender of the Gift: Problems with Women and Problems with Society in Melanesia. Berkeley: University of California Press
——— (1990) ‘Artefacts of history: events and the interpretation of images’. In J. Siikala (ed.) Culture and History in the Pacific. Helsinki: The Finnish Anthropological Society, pp. 25-44
——— (1995), The Relation. Cambridge: Prickly Pear Press
Viveiros de Castro, E. (1998a), ‘Cosmological Perspectivism in Amazonia and Elsewhere’, 4 lectures delivered 17 February – 10 March at Dept. of Social Anthropology, University of Cambridge
——— (1998b), ‘Cosmological deixis and Amerindian perspectivism’, JRAI 4(3): 469-88
——— (2002), ‘O nativo relativo’, Mana 8(1): 113-148
——— (2005) ‘The gift and the given: three nano-essays on kinship and magic’, in Sandra Bamford & James Leach (eds.). Genealogy beyond kinship: sequence, transmission, and essence in ethnography and social theory. Oxford: Berghahn Books
——— (2007), ‘The crystal forest: notes on the ontology of Amerindian spirits’, Inner Asia 9(2): 153-172
Wagner, R. (1986), Symbols That Stand for Themselves. Chicago: University of Chicago Press

  • About

    Martin Holbraad teaches anthropology at University College London and does ethnographic research on Afro-Cuban religion and the cosmology of socialism in Havana. He has coedited two volumes on the theorisation of artefacts in anthropology - Thinking Through Things (Routledge 2007) and Technologies of the Imagination (Ethnos, Special Issue, 2009) - and his monograph titled Cuban Divination and Anthropological Truth is in preparation.

    His essay 'The Power of Powder' appears in Bedeutung Magazine Issue 3/ Life & Death, available here for purchase.