Working Papers Series #7
ISSN 2045-5763 (online)

Can the Thing Speak?

Martin Holbraad
University College London

© 2011 Martin Holbraad
Creative Commons License
Open Anthropology Cooperative Press

Forum discussion in the OAC network.
Download as PDF, EPUB, MOBI.


It may appear that the last thing that the study of ‘things’ needs right now is another manifesto, as the echo of Spivak’s 1980s subaltern radicalism (1998) in my title may suggest. As archaeologist Severin Fowles has recently observed (2008, 2010), the rise of ‘the thing’ in social theory at the turn of our century has emancipatory tonalities that echo the emancipation of ‘the native’ (or the ‘subaltern’) a generation earlier. If for too long things, under the guise of ‘material culture’, had ‘hibernat[ed] in the basements of museology’, as Tim Ingold puts it (2007: 5), their study in recent years has been all about achieving their visibility: making the thing manifest or, in Peter Pels’ phrase, allowing it to ‘speak back’ (Pels 1998: x).

To see why these are more than echoes of expression, consider the analogy of purpose. Notwithstanding their variety, late 20th century arguments tagged as ‘post-colonial’ and valorised as ‘de-colonizing’ can also be characterised as emancipatory (sensu Argyrou 2002). This insofar as they typically take the form of what I will call ‘widening the circle of the human’. The move turns on a basic diagnosis of the colonial condition as, in one way or other, a deficient attribution of humanity to the colonial subject (the native, the subaltern): a denial of its history, its agency, its subjectivity, its rationality, in short, its human dignity. The response, then, takes the form of a more equitable distribution of these attributes, a move to globalise the sense of justice which they express, in a kind of extension of the global-political dominion of the categorical imperative. The colonised subject is elevated, its subjectivity recognised, its voice heard. The conceptual mould of the agenda, if not its historical precedent, is perhaps the emancipation of slaves, from relative object-commodities to (relative…) subject-persons (cf. Guyer 1993).

An analogous agenda, argues Fowles, is pursued in the more recent literature on the rise of the thing (material culture studies, thing-theory, ANT, speculative realism, post-phenomenology etc.). Here too polemical writing has been motivated in large part by a diagnosis of a deficit of humanity – an obvious one when it comes to things, of course, though all the more powerful for it. And the remedy too has been various species of widening the circle of the human. ‘Agency’ has been the most vocal term, perhaps due to its relative neutrality, though its corollaries of personhood, history, voice, freedom and responsibility, and other dignities of the kind are never far off in the emancipatory agenda. Indeed, the political tenor of the move is certainly evident in these writings, as is its post-colonial aesthetic. Fowles cites, among others, Bruno Latour, who calls for a ‘democracy extended to things’ (including a ‘parliament’ of them); Danny Miller, who renounces the ‘tyranny of the subject’ and ‘the corpse of our imperial majesty: society’ in favour of a ‘dialectical republic in which persons and things exist in mutual self-construction and respect for their mutual origin and mutual dependency’; and fellow archaeologist Bjørnar Olsen, who calls his colleagues to ‘unite in a defence of things, a defence of those subaltern members of the collective that have been silenced and “othered” by the imperialist social and humanist discourses.’ (Latour 1993: 12, Miller 2005: 45, 37, Olsen 2003: 100, all cited in Fowles 2008).

Now, the faint sarcasm of calling all this an agenda for ‘emancipation’ is really more of an irony, since I have subscribed to this agenda myself, along lines that are not dissimilar to the ones Fowles describes – particularly in the volume Thinking Through Things, which I co-edited and co-introduced with Ami Salmond (nee Henare) and Sari Wastell (2007), as well as in a couple of single-authored publications related to it (Holbraad 2005 and more explicitly 2009). In the latter part of the present paper I revisit those arguments in some detail, in an effort to clarify what I have since come to see as the somewhat confused way in which they bundle together the two parallel agendas of Fowles’s analogy. As I shall argue, however, this is worth doing, not in order to recoil from the agenda of emancipating the thing, but to move it forward. In a nutshell, I want to show that while the approach set out in Thinking Through Things (henceforth ‘TTT’) is offered partly as a way of emancipating things as such, the weight of its argument ends up subsuming this task to that of emancipating the people for whom they are important.1 If things speak in TTT, they do so mainly by ethnographic association with the voice of ‘the native’ – a kind of anthropological ventriloquism.

Hence the question: might there nevertheless be a sense in which things could speak for themselves? And what might their voices sound like? Suitably reconsidered and improved, I argue, the approach of TTT is indeed able to articulate answers to these questions, complementing the anthropological concern with native voices with what in the Conclusion I shall call a ‘pragmatological’ (cf. Witmore forthcoming) engagement with the voices of things – voices which, to anticipate my core suggestion, stem from the contingent material characteristics that make things most obviously thing-like. In order to prepare the ground for this argument, we may begin by fleshing out, with reference to the recent literature on things, the guiding distinction between emancipating things ‘by association’ with persons as opposed to emancipating them ‘as such’ – a pretty tricky distinction, as we shall see, and subject to all sorts of caveats.

Emancipation as the entanglement of persons and things

In line with Fowles’s analogy with writings in post-colonialism, the past twenty years’ or so literature on the rise of the thing could be plotted as a trajectory of increasingly (self-consciously) ‘radical’ attempts to dislodge or even erase the line that divides things from people. Consider, just as an illustration, the shift from proposing that things acquire ‘biographies’ and a ‘social life’ of their own through their complex involvement in the lives of the people who engage with them (Appadurai 1986), to saying that the very distinction between people and things (or humans and non-humans) should be eliminated from the way we think about such engagements (Latour 1993, cf. Pinney 2005). Or the difference between suggesting that people and things emerge out of each other dialectically (Miller 1987, 2005) and claiming that in certain contexts they are best conceived as being identical (Strathern 1988, 1990). Such differences may be said to correspond to two broad stages on the axis of radicalism, which, following Haraway (1991, cf. Webmoor & Witmore 2009), I shall tag as ‘humanist’ and ‘posthumanist’ respectively. The distinction turns on contrasting stances to the ontological division between humans and things. Humanist, then, would be approaches that seek to emancipate the thing in terms of this division, while posthumanist would be ones that do so by going beyond it. The move from one towards the other, I argue, can also be understood as a move from emancipating things by association, i.e. by letting some of the light of what it is to be human shine on them too, to emancipating them as such, i.e. showing that they can radiate light for themselves – though in a way that, as we shall see, is not altogether satisfactory. Let us explore this with reference to some of the most influential contributions to the literature.

Danny Miller’s introduction to his edited volume Materiality (2005, cf. Miller 1987) presents a transparent example of what I’m calling a humanist approach, as well as of the emancipation of things ‘by association’, with reference to the role in the lives of humans, that such approaches tend to imply. Miller is fully cognizant of the importance to anthropological discussions of materiality of ‘philosophical resolution[s] to the problematic dualism between people and things’ (Miller 2005: 41), and includes as an example his own preference for theorizing the relationship between people and things in terms of the forms that emerge out of a Hegelian dialectical processes of objectification, rather than through the ‘mutual constitution of prior forms, such as subjects and objects’ (2005: 9). The job of the anthropologist, he argues however, cannot be simply (or complexly) to reinvent such philosophical wheels, not least because the people he or she studies ethnographically so often have a much more ‘commonsense’ understanding of things, including all sorts of ways of distinguishing them from people, spirits and so on. Ultimately, Miller is saying, the role of an anthropology that is seriously committed to reflecting ethnographically on the world in which we live, and to theorising what it is to be human, must recognise and ‘respect’ (2005: 38) material objects and the implicit as well as explicit ways in which they give form to people’s lives. Its aim, through strategic combinations of dualism-busting philosophical models and ethnographic sensitivity and empathy, must be to show the myriad ways in which ‘the things people make, make people’ (ibid).2

It is perhaps not entirely clear how Miller squares the circle (not to say wheel) of the contrasting demands of a philosophical impulse to overcome dualism and an anthropological one to dwell on the myriad forms in which it may play itself out ethnographically.3 Still, what he makes abundantly clear is that his heart lies with the messiness of the ethnography, and the ‘vulgar’ study of ‘the way the specific character of people emerges from their interaction with the material world through practice’ (Miller 2007: 26), as he and his students at UCL have been doing for some time. If he is interested in emancipating the thing from the ‘tyranny of the subject’, that is because doing so gives us a more profound understanding of what it is to be human. Material culture studies may displace an anthropology obsessed with the imperium of the social, but only to replace it with a better anthropology humble enough to recognise the ways in which things also so pervasively contribute to our humanity. Which is exactly the kind of stance I have in mind when talking of humanism and its emancipation of things by association.

Alfred Gell’s argument in Art and Agency (1998) provides another example of this approach, though a less straightforward one. Certainly, the idea for which Gell’s landmark book most often gets cited, namely that things can be understood as possessing agency in the same sense as humans do, may well appear as an attempt to emancipate things ‘as such’. In contrast to, say, Miller, the flag of emancipation (if such it is) is here pinned not on things’ role in making human beings what they are (although this is a central concern for Gell too), but rather on the extent to which things may themselves be more like humans than we might assume. Insofar as things (e.g. cars, bombs, effigies) can be construed as indices of a prior intention, as they so often are (e.g. an intention to make us late for work, Pol Pot’s desire to kill, the blessing of a benign deity), they themselves become something akin to humans, and thus could be said to be emancipated as such rather than by association.

Nevertheless, as a number of discussions of Gell’s argument have tended to show, there is some ambiguity as to how far agency really attaches to things themselves in his scheme. Indeed, in reading the book, one is never quite sure how seriously Gell wants us to take the, after all, rather scandalous notion that things can be ascribed with intentions. Part of the problem is that in his analysis Gell tends to treat as equivalent ascriptions of agency that, ethnographically speaking, vary rather vastly in the degree to which they are taken seriously by those who engage in them. Broadly put, if swearing at one’s car for failing to start is meant to be a phenomenon of the same order as praying to an effigy, then one wants to know whether the latter ascription of agency is supposed to be taken as lightly as the former surely should be (which makes Gell look rather dismissive of devotees who take their prayers and effigies very seriously indeed), or whether the agency of the car should be imagined as being as weighty as that of the effigy (in which case Gell would look like a bit of a New Age mystic). Indeed, when it comes down to it, it does seem that Gell’s scheme is slanted towards the former option. As James Leach has argued, a close reading of Gell reveals that agency for him is only ever an indirect attribute of things, its origins lying ultimately with a human agent, whose intention the thing in question only indexes – hence, for example, the significant distinction Gell makes between the ‘secondary’ agency of indices and the ‘primary’ agency of the intentions they are abductively surmised to index (Leach 2007, cf. Gell 1998: 17-21). Things, for Gell, cannot really be agents, if by that we mean anything more than the kind of attribution of agency involved in swearing at a car for making us late. As Miller puts it in his own critique, ‘Gell’s is a theory of natural anthropomorphism, where our primary reference point is to people and their intentionality behind the world of artefacts’ (Miller 2005: 13). Indeed, Gell’s emancipation of things by conferring them with agency turns out to be more similar to Miller’s than may at first appear. Where Miller raises the profile of things by making them operative in the making of human beings, Gell does so by making them operative in acts of human agency.

So, in sum: humanist approaches, which leave the ontological distinction between things and people unmodified, cannot but emancipate things by association. The whole point about the common sense distinction between people and things is that the former are endowed with all the marks of dignity, while the latter are not. So if you want to emancipate the thing while leaving the ontology untouched, then all you can do is find ways to associate it more intimately with the person.

Post-human approaches, by contrast, can be seen as taking up just that challenge: they propose a different ontology of people and things and thus precipitate a re-definition of their properties (i.e. rather than merely a re-distribution of them across the person/thing divide). This tack does indeed raise the hope of an emancipation of the thing ‘as such’, although one immediately has to add the proviso that ‘the thing’, following its ontological re-constitution, is no longer the thing as we ordinarily know it.

Think, for example, of Latour’s denial of human/ non-human purification in favour of the flat ontology of the Actor Network. All the ‘entities’ that modernist purification takes as ‘people’ and ‘things’ are refashioned analytically ‘hybrid’ knots of mutually transformative relations. Each element of which these relations are composed (itself a relation – hence the network’s fractal structure, à la Strathern 2004 [1991]) is an ‘actant’ inasmuch as it has a transformative effect on the assemblage (i.e. the contingent and analytically localised aspect or moment of the Network.)

So agency for Latour is not the effectuation of a human intention (e.g. as it is for Gell). It is a property of networks of relationships (hybrid ones, involving all the elements that a modernist ontology would want to distinguish from one another) that emerges as and when the elements they involve make a difference to each other. The classic and much cited example being Latour’s discussion of the gun debate in the USA (e.g. see 1999: 180). The responsible agents are neither the guns themselves (as the anti-gun campaigners argue) nor the people who use them (as the gun-lobby would have it – ‘guns don’t kill, people do’). It is the hybrid assemblage, or ‘collective’, which gun users and guns form together: the ‘person-with-gun’.

There can be no doubt that, thus ontologically revised or redefined, things are indeed emancipated ‘as such’. The new kind of analytical entity that Latour proposes, the hybrid assemblage of humans and non-humans in mutual transformation, is an agent in as serious a sense one might wish to take that term: its very constitution is defined by its ability to act as such. Indeed, the bold political philosophy that Latour has been building on the back of his move to networks of things-and-people in recent years is testimony to this: ‘political representation of nonhumans seems not only plausible now but necessary, when the notion would have seemed ludicrous or indecent not long ago’, he writes, and raises the prospect of a ‘parliament of things’ (Latour 1999: 198).

Yet, in terms of the framework of the present argument, there is also a significant irony involved in Latour’s tack of emancipation. In order for him to avoid emancipating things ‘by association’ to humans, as per Miller or Gell, Latour ends up defining them, in a revisionist move, as associations (assemblages, collectives, networks), thus binding them to humans by ontological fiat. This, however, begs a question: to what extent and, if at all, how does the dignity conferred on the actants of a Latourian network rub off on the things a pre-Latourian metaphysic would call ‘things’? Does the Latourian revision of the constituents of the world get us any closer to answering our question of whether the thing can speak? Of course, from a Latourian point of view, these questions are either meaningless or foolish. There is no ‘thing’, other than in the modernist chimera. To raise the very question – Can the thing speak? – is to engage in an act of purification. One should rather bite oneself and ask, Can the thing – I mean the actor network! – speak? (Answer: yes.)

Yet, I want to suggest that something important is lost in this act of analytical (because ontological) censorship. Far be it from me to propose any kind of return to modernist ontology – not even for the sake of an anthropological commitment to vulgar common sense à la Miller. Indeed, I am not even sure at this stage of thinking about the matter whether the sense of dissatisfaction I express here points to a principled flaw in Latour’s analytic or an accidental feature of the way Latourian analyses tend to get done. Still, so often when reading such analyses one gets the impression that all the qualities that seem peculiar to ‘things’ as one ordinarily conceives of them – I mean the aspects of things we would ordinarily tag is their ‘material’ qualities, such as those studied by material scientists – somehow get muted, lost in the Latourian translation. I am not saying they don’t get a mention, or that they do not play a significant role in Latour’s often highly sophisticated empirical analyses, as well as those of his followers. For example, Latour’s refutation of the technological determinism of saying that guns kill people does not stop him from emphasising the particular forms of agency that a gun’s technological characteristics – the mechanics of detonation, velocity, accuracy and so on – contribute to the man-with-a-gun assemblage. What I am saying is that the net effect of Latour’s ontological amalgamation of such characteristics with the people they act to transform renders them (or at least tends to render them) corollaries of projects and concerns that a lay non-Latourian account would interpret as irreducibly human: what is important about Boyle’s air-pump is its contribution to modernity (1993a), the significant thing about sleeping policemen is that their concrete curvature participates in the patrolling of traffic (Latour 1993b), what the elements that make for a gun’s firing power do is they engender the potential to kill (1999).

All this may indeed be a contingent function of the particular questions on which Latourian analyses have been put to work.4 Nevertheless, one can make the principled point that Latour’s prime ontological revision, namely the ‘symmetry’ of treating the entities that a modernist metaphysics purifies as persons ‘or’ things as hybrid relations of persons ‘and’ things (see also Viveiros de Castro 2002), renders any interest in those aspects of things one would ordinarily view as distinctively thing-like considerably harder to pursue. Qualities one would call ‘material’ are, as such, always in deep ontological entanglement with the (also) human projects that they help constitute, so one wonders whether in practice, let alone in principle, a Latourian take on things could at all let one disentangle them and allow them to be explored as such. One suspects that with the metaphysical bathwater of ‘materiality’ (as opposed, that is, to ‘humanity’) goes also the baby of ‘materials’ as a legitimate analytical concern.

This way of putting it shows how close this worry comes to one expressed recently by Tim Ingold (2007). Fed up with what he sees as perversely abstract and intractably abstruse debates about ‘materiality’ in recent years, Ingold urges anthropologists to ‘take a step back, from the materiality of objects to the properties of materials [… –] a tangled web of meandrine complexity, in which – among myriad other things – oaken wasp galls get caught up with old iron, acacia sap, goose feathers and calf-skins, and the residue from heated limestone mixes with emissions from pigs, cattle, hens and bears’ (Ingold 2007: 9). Ingold, we may note, makes no secret of the fact that his manifesto for a renewed focus on materials is itself metaphysically motivated, and bound up with a particular way of viewing the relationship between humans and things. Inspired by Gibson as well as phenomenology, Ingold sees humans and things as submerged on an equal ontological footing in ‘an ocean of materials’ (2007: 7). He writes:

Once we acknowledge our immersion, what this ocean reveals to us is […] a flux in which materials of the most diverse kinds – through processes of admixture and distillation, of coagulation and dispersal, and of evaporation and precipitation – undergo continual generation and transformation. The forms of things, far from having been imposed from without upon an inert substrate, arise and are borne along – as indeed we are too – within this current of materials. (ibid.)

One might say that Ingold’s tactic for emancipating the thing involves a kind of inverse humanism (for this is not materialism as we know it), in which, rather than raising things to the power of the human, humans and things alike are factorised down to their primordial material denominator: Life on Earth (ibid). Nevertheless, my point here is that Ingold’s plea for materials can be taken independently of the theoretical agenda from which it may flow, and heeded as a powerful reminder of a whole terrain of investigation that any attempt to take things seriously – even to emancipate them in the terms developed here – cannot afford to ignore.

Indeed, it is with Ingold’s plea for materials that I want to cut to the chase of what asking for things that speak could mean. The problem is one of, if you like, wanting to have one’s cake and eat it. Eating the cake, in this case, is taking fully on board the post-human (e.g. Latourian) point that a proper emancipation of the thing must eschew any principled distinction between it and humans as a starting-point. Having the cake is finding a way nevertheless to credit the Ingoldian intuition that a full-hog emancipation of the thing must place those characteristics that are most think-like or ‘thingy’ (the designation is purely heuristic, with no metaphysical prejudice!) at the top of its agenda. Asking whether the thing can speak, then, is to ask for it to speak on its own terms – in its own language, if you like. Any interesting answer to this question, I suggest, would have to start form the rather blatant observation that it would be a shame if such a language – call it ‘thingese’? – turned out to have no sonorities of what we take to be the most obvious distinguishing feature of so-called things, namely their material characteristics. It is in answer to this question that a critique of the argument of TTT may be useful.

Rethinking through things

Plotted onto the trajectory of increasingly radical attempts to erase the human/thing divide, TTT should probably be placed at the far posthumanist extreme. Indeed, were one permitted to compound this already horrible term, the argument of TTT is post-posthumanist, in that it takes on board the Latourian suggestion that the distinction between people and things is ontologically arbitrary, but adds (contra Latour among others) that, this being so, the solution for emancipating the thing must not be to bind it to an alternative ontological order (e.g. that of the Actor Network), but rather to free it from any ontological determination whatsoever. TTT, in other words, operates within the economy of the literature announcing and articulating the rise of the thing, and its self-conscious polemic purports to offer a corrective even to the most extreme proponents of this (otherwise) common emancipatory goal. Let me indicate briefly how our attempt to emancipate the thing was supposed to work

As put forward in the Introduction of TTT, the argument involved two key claims – one critical and one positive. The critical move, which took off directly from Strathern (see above), went as follows. If in any given ethnographic instance things may be considered, somehow, also as non-things (e.g. an artefact that, ethnographically speaking, is a human being, as per Melanesian gifts, or a river that is a spirit), then the notion of a ‘thing’, anthropologically speaking, can only have a heuristic, rather than an analytical, role. So attempts to analyse the things we call objects, artefacts, substances, or materials in terms of their objectivity, substantiality or, as has become most popular, their ‘materiality’, are locked in a kind of ethnographic prejudice – they are, to use the dirty word, ethnocentric. And this goes also for attempts theoretically to emancipate things by attributing them with all sorts of qualities earlier shacklers would take to belong only to humans, such as sociality, spirituality, and again, most popularly, agency. In other words, if what a thing may be is itself an ethnographic variable, then the initial analytical task must not be to ‘add’ to that term’s theoretical purchase by proposing new ways to think of it – e.g. as a site of human beings’ objectification (Miller), an index of agency (Gell), an on-going event of assemblage (Latour), or what have you. Rather it must be effectively to de-theorise the thing, by emptying it out of its many analytical connotations, rendering it a purely ethnographic ‘form’ ready to be filled out contingently according only to its own ethnographic exigencies. Treating the thing as a heuristic (i.e. just as a tag for identifying it as an object of study) was indeed, then, a way for us to allow it to speak in its own terms – which in ethnographic principle may be as varied as there are things to listen to – from behind the clamour of social theoretical attempts to theorise such a thing as the thing as such. Things do speak, ran the thought, but the problem is how to hear them past all the things we say about them.

If half of the way towards addressing this problem is to empty out the notion of ‘thing’ of its contingently a priori metaphysical contents – thing-as-heuristic –, the other half is to formulate a way of allowing it to be filled by (potentially) alternative ones in each ethnographic instance. This can be seen as the second and positive emancipatory move of the TTT argument, which is captured by a complementary methodological injunction: ‘concepts = things’. The move is complementary in that it follows directly from the issue that motivates the heuristic approach in the first place, namely the possibility – and in so many instances the fact – that the things we call ‘things’ might not ethnographically speaking be things at all, or not in the way we might initially assume them to be. For note that the things-as-concepts injunction is determinedly not proposed as some new theory of the thing. The idea is emphatically not to propose some kind of revisionary metaphysic, to the effect that, where people have so often assumed things and concepts to belong to opposite ontological camps, we should all from now on recognise them as belonging to the same one (viz. the kind of approach Latour and Ingold advance in different ways, as we have seen). To the contrary, the ‘things = concepts’ formula is offered as a further methodological clause for side-stepping just such theoretical prescriptions. In particular, it is supposed to foreclose a very real danger when it comes to thinking anthropologically about the different ways in which things may feature ethnographically, namely that of parsing them as different ways in which people may think about (represent, imagine, socially construct) them. This is to parse ethnographic alternatives to our metaphysic of things in terms of it – in fact, in terms of what nigh all-thing emancipators consider its crassest version, namely the idea of inert and mute things invested with varied meanings only by human fiats of representation. It is, in effect, to raise the erasure of things to the power of a necessity for thinking of them.

So the ‘concepts = things’ clause is meant to placate just this danger. Put very simply: instead of treating all the things that your informants say of and do to or with things as modes of representing the things in question, treat them as modes of defining them. The immediate advantage of this way of parsing the issue is that it renders wide open precisely the kinds of questions that lie at the heart of the emancipatory agenda, namely questions about what kinds of things ‘things’ might be. Instead of merely offering sundry ways of confirming the base metaphysic of mute things invested with varied meanings by humans, the things-as-concepts tack holds up that very ethnographic variety as a promise of so many ways of arriving at alternative metaphysical positions – whatever they might be. If every instance anthropologists would deem a different representation of a thing is conceived as a potentially different way of defining what such a thing might be, then all the metaphysical questions about its character qua ‘thing’, what materiality might be, objectification, agency – all that is now up for grabs, as a matter of ethnographic contingency and the analytical work it forces upon us.

As we did in the Introduction to TTT itself, let me illustrate the approach with reference to my own chapter in the book, in which I elaborate an analysis of aché. Aché is a mana-type term that Afro-Cuban diviners use to talk both about their power to make deities appear during divination, and about a particular kind of consecrated powder that they consider as a necessary ingredient for achieving this. The terminological coincidence, I argued, corresponds to an ontological one: a diviner’s power is also his powder and the powder (qua consecrated) is also his power. Now, this is obviously a counter-intuitive suggestion, of the order of ‘twins are birds’ (Evans-Pritchard 1956, cf. Holbraad 2010). If we know what powder is at all, we know that it is not also power in any meaningful sense (it’s just powder!), and much less can we accept that power (a concept with proportions as grand as Nietzsche or Foucault) might also be just powder (of all things!). Hence the classical anthropological type of question: why might Cuban diviners ‘believe’ such a crazy idea? For as long as our analysis of aché remains within the terms of an axiomatic distinction between things and concepts, we cannot but ask the question in these terms. We know that powder is just that dusty thing there on the diviner’s tray (see below). So the question is why Cubans might ‘think’ that it is also a form of power. How do we explain it? How do we interpret it?

Alternatively, we could treat the distinction between concepts and things merely as a heuristic device, as per TTT’s first move. This would allow us to ask questions about that powder that we would intuitively identify as a ‘thing’, without prejudicing the question of what it might be, including questions of what it being a ‘thing’ might even mean. Answers to such questions, then, would be culled from the ethnography of all the data we would ordinarily be tempted to call people’s ‘beliefs’ about this powder, including the notion that it is also power. As per the second move of the TTT method (concept = thing), we would treat such data as elements of a conceptual definition of the thing in question. So: Cuban diviners do not ‘believe’ that powder is power, but rather define it as power. Note, crucially, how this way of setting up the problem raises the metaphysical stakes. Since our own default assumption is that powder is not to be defined as power (it’s just a dusty thing, we assume), the challenge now must be to reconceptualise those very notions and their many ethnographic and analytical corollaries (powder, power, deity etc. but also thing, concept, divinity etc.) in a way that would render the ethnographically-given definition of powder as power reasonable, rather than absurd. It is just this kind of analytical work I attempted to carry out in my chapter in TTT (I shall cover more of that ground later).

At the time we presented this mode of analysis in TTT, I for one imagined it as having cracked the problem of the thing’s emancipation as I have been outlining it here. Taken together, I thought, our argument’s two key moves effectively opened up the space for things themselves, as one encounters them heuristically in any given ethnographic instance, to dictate their own metaphysics – to dictate, if you like, the terms of their own analytical engagement. Just what I have in mind when asking for an approach that allows things to speak for themselves, in their own language! Yet, to see why I may have been wrong, one needs only to contemplate how the prospect of things speaking in the ‘own’ language in this TTT-sense measures up to the Ingoldian caveat, namely that a proper emancipation of things ‘as such’, whatever that may mean or involve, should place their material characteristics centre stage – that things should speak in thingese, and that thingese should somehow be an expression of things’ peculiarly material qualities. In the sometimes flamboyantly programmatic pronouncements of the TTT Introduction, nothing is in fact made of such qualities, and certainly their role in ‘thinking through things’ is left largely unspecified.

In fact, it is indicative that this first dawned on me (at any rate) when faced with a searching question by an archaeologist in a conference at which my co-editors and I presented our argument (see also Holbraad 2009). Being himself consigned to working with things without the benefit of rich ethnographic information about them, he admitted, he found himself at a loss as to how archaeologists might deploy our approach to any effect. Notwithstanding our claim to have found a way to let things speak for themselves, our argument seemed at most a method for allowing the ethnography of things to speak on their behalf – to set, indeed, the terms of their analytical engagement. If what motivates the whole approach is, as explained above, the fact that in varied instances people speak of or act with things in ways that contradict our assumptions about what a thing might be; and if, furthermore, it is just those ways of speaking and acting around things that are supposed to provide the ‘content’ of their potentially alternative metaphysics; then how might archaeologists, for whom, what people might have said or done around the things archaeologists call ‘finds’ is so often the primary question? If anyone ever needed a way of letting things speak for themselves that is the archaeologist, for whom things are so often all he has to go on. Our unproblematised reliance on, and unabashed love for, ethnography in our way of ‘thinking through things’ is of no huge help. The clue is in the book’s subtitle: ‘theorising artefacts ethnographically’.

These misgivings go to the heart of the problem I wish to tackle here, and are tellingly connected to another worry that as a social anthropologist I have had myself (privately!) about the TTT argument, namely the fact that the analytical experimentations it seeks to promote seem in one way or other to be wound around ethnographic phenomena one might broadly call ‘magical’ or even ‘animist’ in one sense or other. Cigarettes that make Port Morsby inmates’ thoughts fly out of prison, Maori and Swazi legal paraphernalia that have metaphysical efficacy, shamanic costumes that transport Mongols to legions of skies, and family chests and photographs that contain their life force, divinatory powder that is the power to reveal deities: these are the things contributors to our volume thought through, along with the people who ‘informed’ them ethnographically about them. In line with the archaeologist’s comment, I suspect that this ‘magical realist’ tenor of the chapters is not accidental. The leverage for thinking out of the metaphysical box that so entranced us as editors was owed, at least to a large extent and at the first instance, to the chapters’ ethnographic magic, to coin a phrase, rather than the specifically ‘thing-like’ character of their subject-matter.

It emerges, then, that TTT’s claim to offer an emancipation of the thing along the lines I have been discussing is open to a critique that is analogous to the one advanced earlier in relation to Latour. Latour, we saw, emancipates the thing by entangling it ontologically with persons – subsuming both under the terms of his revisionary ontology of networks comprising people-and-things. TTT does something similar, though now at the level of analytical methodology. It emancipates the thing by entangling it heuristically with all that the people concerned with it say and do around it, subsuming things and their ethnographic accounts under the terms of our revisionary methodology. Indeed, just as a Latourian might object that to demand an emancipation of the thing ‘as such’ is flatly to deny the significance of Latour’s ontology of networks, so we might want to contend that that same demand merely contradicts our methodological injunction of concept = thing. As far as TTT is concerned, things as such just are what our ethnographic descriptions of them define them to be. Still, if this is emancipation by ethnographic ‘association’, the Ingoldian bugbear remains: what of materials and their properties?

Yet, I want to argue that the force of this line of critique pertains more to the rhetoric of the TTT argument than to its substance. Suitably reconsidered, the methodological approach of TTT is indeed able to give ‘voice’ to material characteristics, making analytical virtue of them as such. The fact that this prospect remained mute in the way we pitched the argument when we wrote it relates directly to the guiding homology with which I began this paper, between the postcolonial agenda of emancipating the native and the thing-theoretical one of emancipating the thing. With particular reference to Viveiros de Castro’s ongoing project of de-colonizing anthropological thinking by using ethnography to subvert its most domineering (because ontological) presuppositions (see Henare et al 2007: 8-9, cf. Viveiros de Castro 1998, 2002), ours was pitched above all as an attempt to put the ethnography of things at the centre of such an endeavour. If for Viveiros de Castro the emancipation of the native in anthropology is a matter of opening up space for her ‘conceptual self-determination’ (2002) within it, then the TTT argument amounted mainly to the addendum that the ethnography of (people’s engagement with) things is a prime site for pursuing this goal. In other words, whatever emancipation TTT might offer to things was rhetorically subsumed under the older (but surely no less pressing) political agenda of emancipating the native. Indeed, TTT’s two-step methodology reflected this directly. The ‘thing-as-heuristic’ move opened up ‘things’ as a locus of ontological self-determination, while the ‘concept = thing’ clause allowed the ethnography of what natives do and say around them to provide it with ontologically variable contents.

What, then, of the substance of this argument? Might it, albeit inadvertently, provide a way for things to speak, not as proxies for ethnographic natives, but for themselves? In a longer, fully written up version of this paper I plan to use three examples of anthropological and archaeological analyses of things in order to explore the question concretely.5 Here I limit myself to making the argument from first principles, and illustrating it briefly with reference to my Cuban powder-is-power case.

It all depends, of course, on what one takes ‘things that may speak’ to mean – what counts as a thing that speaks for itself? It is on this point that I think the homology with Spivak’s question about the subaltern is most instructive. We have already seen that attempts to transpose the humanist agenda of postcolonial emancipation onto things by including them in ‘the circle of the human’ provide only half-hogged emancipations, ‘by association’. But we have seen also that there is an alternative to humanism in the struggle to de-colonise anthropology – not least in the rhetoric of TTT itself, as well as in the work of Viveiros de Castro. Captured by Viveiros’s slogan of ‘conceptual self-determination’, this is the project of constructing an anthropology that opens spaces for natives to set the terms of our anthropological engagements with them, positing them as producers of concepts rather than, say, consumers of ours. Rather than worrying about how far natives might (or should) be considered as humans, agents, subjects and so on, we should be asking what concepts of humanity, agency, subjectivity and more our anthropological engagement with them might yield, and be fully prepared to be surprised by what we find (sensu Strathern 2005). It is this notion of emancipation, then, that I propose to transpose onto things: things can speak insofar as they can set the terms of their anthropological engagement by acting as the originators (rather than the objects) of our anthropological conceptualisations. Things can speak if they can yield their own concepts.

This way of putting the matter already gets us much closer to seeing why TTT might after all be suited to stage such a move. Bracketing for this purpose the underlying postcolonialist concerns to which it was put to use, the ‘concept = thing’ formula speaks directly to the problem at hand. All one needs to do is read the formula backwards (in school we called this ‘symmetry of equality’): ‘thing = concept’. Indeed, the thought is in a pertinent sense the reverse (though not the opposite) of the one advocated explicitly in TTT. If there the formula ‘concept = thing’ designated the possibility of treating what people say and do around things as manners of defining what those things are, here its symmetrical rendition ‘thing = concept’ raises the prospect of treating the thing as a manner of defining what we (analysts now, rather than natives) are able to say and do around it. At issue, to coin another phrase, are a thing’s conceptual affordances.

Indeed, thinking of the present argument as a symmetrical reversal of the one made in TTT also allows us to flesh the thought out in Ingold’s direction, towards the question of materials and their properties. As I noted when outlining the TTT position, the promise of conceptual experimentation that it holds up is grounded in ethnographic contingency. Having emptied the notion of ‘the thing’ of any conceptual presuppositions of what may count as one, we fill it back up with alternative conceptualisations drawn from the contingent ethnographic data we find around it. One way of describing the procedure, it strikes me, would be as a form of ‘empirical ontology’, where ‘empirical’ denotes its ethnographic grounding. So we may ask: what is the equivalently empirical grounding of the reverse procedure that I seek to articulate here for things? Following through on the symmetry of our reversal-strategy, the answer can be found only in the material characteristics of the thing itself. What was empirical about (ethnographically driven) concepts that defined things must now be so about (let’s say, ‘pragmatographically’ driven) things that now define concepts. With what other ‘stuff’ can things feed their conceptualizations than the very stuff that makes them what they are, as heuristically marked ‘things’? The data that make a (conceptual) difference, in this case, are no longer what we hear and see people say and do around things, but rather what we hear, see, smell, taste and touch of the thing as we find it (heuristically) as such.

The difference from Ingold, however, is that, in line with his phenomenologically inclined vitalism, he is content to revel at this material and sensuous level of things, to explore their mutual ‘enmeshment’ with people and other organisms, as well as their ‘affordances’ for them in the broader ecology of living. By contrast, in raising the question of the conceptual affordances of materials and their properties, my interest is not in the ecology of their material alterations but rather in the economy of their conceptual transformations: how their material characteristics can dictate particular forms for their conceptualization. At issue, if you like, is not the horizontal traffic of materials’ enmeshment in forms of life, but rather what one might imagine as a vertical axis of materials’ transformation into forms of thought – mainly for fun, I’d call this the ‘intensional vertizon’ of things (to mark its orthogonal relationship to phenomenological notions of things’ ‘intentional horizon’ in, say, Husserl). Simply put, this vertizonal movement would be what ‘abstraction’ would look like were it to be divorced from the ontological distinction between concrete (things) and abstract (concepts). Indeed, this is just what the ‘thing = concept’ clause of our analytical method would suggest. Where the ontology of things versus concepts would posit abstraction as the ability of a given concept to comprehend a particular thing, external to itself, in its extension, the heuristic continuity of ‘thing = concept’ casts this as a movement internal to ‘the thing itself’ (to echo Husserl again): the thing differentiates itself, no longer as an instantiation ‘of’ a concept, but a self-transformation as a concept.

I am of course aware that this way of thinking takes us into deep philosophical waters which I am incompetent to chart (although one may note with pleasure that this is exactly as it should be: one would hardly hope for the scandalous idea of things that speak to have tamer philosophical implications). Indeed, in my amateur understanding, there is a line in Western philosophy, which runs from Heraclitus through Leibniz and up to Deleuze, that deals with many of the relevant problems. Still, adopting a distinctively anthropological slant with reference to Marilyn Strathern’s notion of ‘partial connections’ (1991), Morten Pedersen and I have elsewhere tried to articulate in some detail the analytical implications of things’ capacity for vertizonal transformation – we called this form of self-motion ‘abstension’, to indicate the intensive (as opposed to extensive) character that abstractions acquire when they are thought as self-differentiating transformations of things-into-concepts (see Holbraad & Pedersen 2009). Rather than cover this ground again for present purposes, however, I close by showing what this kind of analytical movement looks like with reference to the example of aché which I began to discuss earlier. (Indeed, in retrospect, it seems remarkable that this line of argument was pasted over, not only in the Introduction to TTT, where the notion of a powder that is power is used as an illustration as we saw, but even in my own chapter in the book, where the actual analysis of this material is carried out.)

If, as I have argued, the problem with TTT is that it emancipates the thing only by associating it in ethnography with an ontologically emancipated native, then my analysis of aché in my TTT chapter is certainly an instance of this. We have already seen, for example, that the very problem that article was devoted to solving – what might a powder that is also power be? – was ethnographically driven: it was not powder that told me it is power, it was my diviner informants. And certainly, a host of ethnographic data serve to frame and develop the problem itself, as well as parts of its analytical solution. Crucially, for example, since what powder might be in this instance depends on the notion of power, part of my attempt to articulate the question involves developing its various dimensions ethnographically. In a nutshell, I provide an account of Afro-Cuban divinatory cosmology based on informants’ responses, to show that for diviners power consists in the ability to render otherwise absent divinities present during the divinatory ceremony, and that this power manifested in divination as the ‘signs’ the diviners mark with their fingers on the powder that is spread in the surface of their divining board, which are called ‘oddu’, and said to ‘be’ divinities in their own right. On the basis of this ethnographic information, I go on to show that the notion of a powder that ‘is’ power emerges as a solution to an age-old theological conundrum, familiar in the anthropology of religion (e.g. Keane 2007): apparently transcendent deities are rendered immanent on the surface of the divining board, allowing those present in the divination to relate to them directly. Conceptualising powder as power, then, requires us to understand analytically how Afro-Cuban divination effectively solves this ‘problem of presence’, to recall Matthew Engelke’s book on a related conundrum (2007). And it is to this question that powder, finally, speaks:

Powder gives us the answer […]. 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, ie the capacity to make [deities] ‘come out’ and ‘speak’ […]. 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 (to be 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. [In this way] 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. […] 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 […], 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 [not as] entities, but rather as motions. […] If the oddu […] just are motions […], 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. [So, qua motions, the deities have inherent within themselves the capacity to relate to humans, through 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. (Holbraad 2007: 208-9)

It in not an accident that the content of this analysis (i.e. the relationship between transcendence and immanence) is recursively related to its form (i.e. the relationship between analytical concepts and ethnographic things) – in the article itself I made much of this. Here, however, I want only to focus on the latter question, to draw attention to the work powder does for the analysis, by virtue specifically of its material characteristics. If ethnography carries the weight of the analytical problem, in this argument, it is the material quality of powder that provides the most crucial elements for its solution. If deities are conceptualised as motions to solve the problem of presence, after all, that is only because their material manifestations are just that, motions. And those motions, in turn, only emerge as analytically significant because of the material constitution of the powder upon which they are physically marked: its pervious quality as a pure multiplicity of unstructured particles, amenable to intensive movement, like the displacement of water, in reaction to the extensive pressure of the diviner’s fingers, and so on. Each of this series of material qualities inheres in powder itself, and it is by virtue of this material inherence that they can engender vertizonal effects, setting the conceptual parameters for the anthropological analysis that they ‘afford’ the argument. As an irreducible element of the analysis of aché, it is powder that brings the pivotal concepts of perviousness, multiplicity, motion, direction, potential and so on into the fray of analysis, as conceptual transformations of itself, as per the ‘thing = concept’ clause. In that sense, I submit, it speaks for itself – louder, in fact, than any other element of the analysis presented.

Conclusion: anthropology and/or pragmatology

By way of conclusion, it may be worth clarifying a little how I see the dividends, as it were, of the kind of amplification of things’ voices in anthropological analysis that I have sought to articulate. In particular, it is important at this point to be rather precise about the degree and manner in which this way of sourcing anthropological conceptualizations in things counts as a way of emancipating them ‘as such’. Indeed: one might be tempted to object that, whatever the merits of the case I have sought to make for things speaking ‘as such’ in our analyses, their emancipation in this way nevertheless remains unavoidably circumscribed by the human-oriented agendas to which these analyses – anthropological after all – are directed. Sure (the objection would go): powder may be operative in the analysis of my Cuban example, providing the material source for my conceptual abstensions, as I called them, of such analytical ingredients as perviousness, multiplicity, intensive motion, and so on. Still, these ingredients are part of a longer recipe, so to speak, which includes not only things like powder, but also divinities, diviners, their clients and so on. And what this analytical recipe is meant to cook is an argument about Cuban practitioners of divination – that is people, my informants – and how we may best conceive of their notion that powder, in a divinatory context, is a form of divine power. While part of our answers to such questions, in other words, might be driven by things ‘as such’ in the manner I have indicated, their anthropological significance is nevertheless a function of their association, in the economy of anthropological analysis, with people and the ethnographic conundrums they pose to us. So the aforementioned archaeologist’s bemused complaint, it seems, remains after all: could things really speak without their association to human (in this case ethnographically talkative) subjects?

The correct response, I would suggest, is to bite the bullet. Anthropological examples such as the one on Afro-Cuban powder indeed do not demonstrate that things can speak of their own accord, and seem bound to continue to render them subservient to the analysis of the human projects into which they enter. Arguably, however, this line of scepticism is contingent squarely on the anthropological – by which I mean also human-centric – character of the example. Indeed, while admittedly staying within the economy of undeniably anthropological analyses, what I have ventured to argue is that such analyses may involve an irreducibly thing-driven component or phase – one we might call ‘pragmatological’, borrowing somewhat subversively a term coined, tellingly, by the archaeologist Christopher Witmore (forthcoming). Indeed, while the analytical difference things can make pragmatologically might in this instance be gauged with reference to the anthropological mileage they give, the very notion that things might make such a difference of their own accord, ‘as such’, does, it seems to me, ultimately raise the prospect of pragmatology as a sui generis field of inquiry.

Allow me, then, to indulge in a final and absurdly programmatic speculation. Might one imagine a thing-centric discipline called pragmatology in which things’ material properties would form the basis of conceptual experimentations that would be unmediated by, and run unchecked from, any human projects whatsoever? I have to admit that my own conception of what such a discipline might look like is hazy to say the least… Certainly, notwithstanding my earlier comments, I don’t think archaeology would be enough to provide a model, if only because archaeology shares the anthropo-centric slant of social anthropology, its problem being mainly that its otherwise thing-oriented methodology suffers from a deficit of human conformation. Theoretical physics may come considerably closer, since so much of it apparently takes the form, precisely, of radical conceptual experimentations in the service of understanding the material forms of the universe. Still, this also has problems, partly due to physicists’ still encompassing demand for causal explanation (a demand that is certainly distinct, and possibly incompatible, with our pragmatological concern with conceptualization). At any rate, there is no reason to limit our putative pragmatology to physicists’ takes on matter, to the exclusion of those of, say, chemists, biologists, engineers, or, indeed, artists, sculptors or musicians. In fact, I suspect the closest one might get to the kind of inquiry pragmatology could involve would be an inverse form of conceptual art – construed, of course, very broadly indeed. If the labour of the conceptual artist is supposed to issue in an object that congeals in concrete form a set of conceptual possibilities, the work of the pragmatologist would be one that issues concepts that abstend in abstract form a set of concrete realities. Pragmatology, then, as art backwards.


Versions of this paper has been presented at departmental seminars in Aberdeen, SOAS and UCL, as well as at the Things and Spirits conference in Lisbon (September 2010) and the TAG 2010 conference at Bristol. I thank James Leach, Ed Simpson, Victor Buchli, Ricardo Roque and Joao Vasconcelos and Dan Hicks for their respective invitations, and to the participants in each of them for their valuable comments. I am also grateful to Lise Philipsen for illuminating conversations during writing.


Appadurai, Arjun, 1986. ‘Introduction: commodities and the politics of value’. In: The Social Life of Things: Commodities in Cultural Perspective, A. Appadurai (ed.). Cambridge University Press, Cambridge, 3-63

Argyrou, Vassos, 2002. Anthropology and the Will to Meaning: a Postcolonial critique. London: Pluto Press

Engelke, Matthew, 2007. A Problem of Presence: Beyond Scripture in an African Church. Berkeley: University of California Press

E.E Evans-Pritchard, 1956. Nuer Religion. New York & Oxford: Oxford University Press

Fowles, Severin, 2008. ‘The perfect subject: Postcolonial object studies’. Paper at Annual Conference of Theoretical Archaeology Group, May 24, Columbia University, New York.

———, 2010. ‘People without things’. In: An Anthropology of Absence: Materializations of Transcendence and Loss, M. Bille et al (eds.). New York: Springer, 23-41

Gell, Alfred, 1998. Art and Agency: An Anthropological Theory. Oxford: Clarendon Press

Guyer, Jane, 1993. ‘Wealth in people and self-realisation in equatorial Africa’. Man N.S. 28(2): 243-265

Haraway, Donna. (1991) Simians, Cyborgs and Women: The Reinvention of Nature, Free Association Books: London.

Harman, Graham, 2009. Prince of Networks: Bruno Latour and Metaphysics.

Henare, A., M. Holbraad & S. Wastell, 2007. ‘Introduction’. In Thinking Through Things: Theorising artefacts ethnographically, Wenare et al (eds.). London: Routledge, 1-31.

Holbraad, Martin, 2005. Expending multiplicity: money in Cuban Ifá cults. Journal of the Royal Anthropological Institute (N.S.) 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 et al (eds.) Thinking Through Things: Theorising Artefacts Ethnographically, London: Routledge, 189-225

——–, 2009. ‘Ontology, ethnography, archaeology: an afterword on the ontography

of things’. Cambridge Archaeological Journal 19(3): 431-441

———, 2010. Ontology is just another word for culture: against the motion. Debate & Discussion (from GDAT 2008, S. Venkatesan (ed.)). Critique of Anthropology 30(2): 179-185

Holbraad, Martin & Morten A. Pedersen, 2009. ‘Planet M: the intense abstraction of Marilyn Strathern’. Anthropological Theory 9(4): 371-94

Ingold, Tim, 2007. ‘Materials against materiality’. Archaeological Dialogues 14(1): 1-16

Keane, Webb, 2007. Christian Moderns: Freedom and Fetish in the Mission Encounter. Berkeley: University of California Press

Latour, Bruno, 1993. We Have Never Been Modern. Translated by C. Porter. London: Prentice Hall.

———-, 1993b. ‘On technical mediation: philosophy, sociology, genealogy’. Common Knowledge 2: 29–64

———–, 1999. Pandora’s hope: Essays on the reality of science studies. Cambridge, MA: Harvard University Press

Leach, James, 2007. ‘Differentiation and encompassment: a critique of Alfred Gell’s theory of the abduction of creativity’. In: Thinking Through Things: Theorising artefacts ethnographically, Wenare et al (eds.). London: Routledge

Miller, Daniel, 1987. Material Culture and Mass Consumption. Oxford: Basil Blackwell

———–, 2005. ‘Materiality: an introduction’. In Materiality, D. Miller (ed.). Durham & London: Duke University Press, 1-50

———–, 2007. ‘Stone age or plastic age?’. Archaeological Dialogues 14(1): 23-7

Olsen, B. 2003. ‘Material culture after text: re-membering things’. Norwegian Archaeological Review 36(3):87–104

Pedersen, Morten A., 2011. Not Quite Shamans: Spirit Worlds and Political Lives in Northern Mongolia. Ithaca: Cornell University Press

Pels, Peter,1998. ‘The spirit of matter: On fetish, rarity, fact and fancy’. In Border Fetishisms: Material Objects in Unstable Places, Patricia Spyer (ed.), New York: Routledge, 91-121.

Pinney, Chris, 2005. ‘Things happen: or, from which moment does that object come?’. In: Materiality, D. Miller (ed.). Durham & London: Duke University Press, 256-72

Spivak, Gayatri Chakravorty. 1998. “Can the Subaltern Speak?” in Marxism and the Interpretation of Culture. Eds. Cary Nelson and Lawrence Grossberg. Urbana, IL: University of Illinois Press, 1988: 271-313

Strathern, Marilyn, 1990. ‘Artefacts of history: events and the interpretation of images’. In Culture and History in the Pacific, J. Siikala (ed.). Helsinki: Transactions of the Finish Anthropological Society, 25-44

———–, 2004 [first published 1991]. Partial Connections (Updated edition). Walnut Creek: Altamira Press

———–, 2005. Kinship, Law and the Unexpected: Relatives Are Always a Surprise, Cambridge: CUP

Tilley, Christopher, 2007. ‘Materiality in materials’. Archaeological Dialogues 14(1): 16-20

Viveiros de Castro, Eduardo, 1998. Cosmological Perspectivism in Amazonia and Elsewhere. Lecture series delivered at Dept. of Social Anthropology, University of Cambridg, 17 February – 10 March

———–, 2002. And. Manchester: Manchester Papers in Social Anthropology.

Webmoor, Tim & Christopher Witmore, 2008. ‘Things are us! A commentary on human/things relations under the banner of a ‘social’ archaeology’. Norwegian Archaeology Review, 41(1), 53-70

Witmore, Christopher, Forthcoming 2011. “The realities of the past: Archaeology, object-orientations, pragmatology.” In B.R. Fortenberry and L. McAtackney (eds) Modern Materials: Proceedings from the Contemporary and Historical Archaeology in Theory Conference 2009. Oxford: Archaeopress.


1While my comments here speak to the argument as presented by all three of its authors, I do not claim that Salmond and Wastell would agree with the retrospective critique I develop here (although they may well do so, at least in part).

2 Chris Tilley (whose own dualism-busting efforts draw mainly on phenomenology), puts it most simply, in defence of the notion of ‘materiality’: ‘The concept of materiality is required because it tries to consider and embrace subject-object relations going beyond the brute materiality of [things] and considering why certain [things] and their properties become important to people.’ (Tilley 2007: 17)

3On this point Miller, for one, resorts to a rather elaborate metaphor about philosophical wheels and the anthropological vehicles they help along which does not, to my mind, express very clearly the relationship between the two analytical demands (see Miller 2005: 43-46).

4There may exist out there a Latourian analysis of an assemblage of actants consisting only of the things we’d call things, though the prospect seems more speculative at present – see Harman 2009.

5 My favourite candidates are Pedersen’s forthcoming analysis of shamanic costumes in Mongolia (Pedersen 2011), in which he claims that these artefacts ‘provide an analysis of themselves’, Strathern’s commentary on Battaglia’s analysis of Sabarl pick-axes (Strathern 1991), in which she argues that these artefacts ‘contain their own contexts’, and the archaeological debate about skeuomorphism, where, I want to argue, materials analyse each other through translating prior forms into novel contents (material analysis as concretions of abstractions).