Ethnographic X-files and Holbraad’s double-bind

This work is licensed under the Creative Commons | © Andrew Apter. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau7.1.021

Ethnographic X-files and Holbraad’s double-bind

Reflections on an ontological turn of events

Andrew APTER, University of California, Los Angeles

The ontological turn asks difficult questions and pushes conceptual boundaries both forward and against the grain, challenging our standard epistemological orientations. That it replaces epistemological with ontological concerns is a position I oppose in the following reflections, inspired by Martin Holbraad’s “double-bind,” one which frames a predicament that not only resonates with my own ethnographic experience but also represents something of an occupational hazard (our X-files) for many of us working on spirit-worlds. In an alternative solution to Holbraad’s double-bind, based on Moore’s paradox and my own exposure to Shango’s wrath in Nigeria, I propose a radical decoupling of Knowledge and Belief to obviate the ontological proliferation of worlds.

Keywords: mystical agency, ontography, possible worlds, Moore’s paradox

Logic isn’t as simple as logicians think it is.

Ludwig Wittgenstein, Letter to G. E. Moore

Like many of my somewhat skeptical peers, I have been equivocating over the relatively recent “ontological turn” in anthropology, wavering between glimpses of real profundity as its “ontographic method” (Holbraad 2012a: 86) proffers access to radically different worlds, and frustration that at the end of the day it represents the latest iteration of the cultural relativism and radical translation debates of the 1960s and 1970s, despite principled claims to the contrary. On the one hand, I appreciate its “quiet revolution” (Henare, Holbraad, and Wastell 2007: 7–10), reversing relations between concepts and things so basic to “representationalist” theories of knowledge, and by taking “things” and “native claims” on their own terms, as “true” by definition, thereby opening pathways to multiple ontologies. I also appreciate its intellectual genealogy, building on the work of worthy luminaries such [288]as Roy Wagner, Marilyn Strathern, Alfred Gell, Bruno Latour, Philippe Descola, and Eduardo Viveiros de Castro. On the other hand, I keep stumbling. Where does this really take us? Does it not implicitly invoke a sociocultural alterity that hardly exists anymore, if it ever did? Is such alterity greater “between” societies than “within” them? Where are the richer historical and sociopolitical contexts? Do native claims that sacramental wine is the blood of Christ, or that Jews are a chosen people, also open up different ontologies? Even if I finally do grasp that Ifá powder and power are the same “thing” in Cuban divination, as well as the “motile” logic which renders its disclosures indubitable (Holbraad 2007), then what? Can I “apply” such motility to other contexts, including Ifá divination in Nigeria or New Jersey? Or to the latest debates in nonstandard logics? In my less generous moments—off the record, of course—I have caricatured the ontological turn as bad philosophy and no ethnography (a charge that may equally characterize what follows!).1

But that is neither true nor fair, as growing research in this vein attests. The ontological turn asks difficult questions and pushes conceptual boundaries both forward and against the grain, challenging our standard epistemological orientations. That it necessitates a commitment to multiple ontologies is a claim that I will challenge in the following reflections inspired by Martin Holbraad’s “double-bind,” one which frames a predicament that not only resonates with my own ethnographic experience, but also represents something of an occupational hazard for many of us working on spirit-worlds. In an alternative solution to Holbraad’s double-bind, based on my exposure to Shango’s wrath in Nigeria, I propose a radical decoupling of Knowledge and Belief to obviate the ontological proliferation of worlds.

Holbraad’s double-bind

In a provocative essay that takes us through the looking glass into the alternative world of Ifá divination in Cuba, Martin Holbraad opens with a discomfiting encounter familiar to many of us after returning from the field, when asked: “Do you think that oracles work?” (2012a: 81, original emphasis). How often have I encountered variations of this question, when asked if I believe in the Yoruba orisha, whether they are real, or might I even channel their power to help someone out? I usually respond that “No, I don’t actually believe in them, but . . . .” And if my interlocutor remains interested, I may describe ethnographic experiences that give uncanny credence to mystical intervention in human affairs. Such “X-files” from the field are by no means uncommon, and are found even among our most right-minded colleagues, archived as private and unsuitable for public release.2 Indeed, the public is always quick to exoticize alterity with “unsolved mysteries” of ghosts, zombies, and aliens, thus one learns to be careful about what one says to [289]avoid looking ridiculous and embarrassing the profession.3 Nor are such queries limited to lay persons, but, as Holbraad points out, come from colleagues as well. Which is why I find Holbraad’s reaction both familiar and refreshing. If he “loves” the question because it rattles the rationalist cage, he also “hates” it because it is so difficult to answer:

What I hate about it is that it tends to put me in a double-bind. If I say that I do not believe in oracles, I provide a quick fix for my interlocutor’s disquiet with what really is a lie. For there is an important sense in which I really do believe in oracles—but if I admit this, I create the conditions for misunderstanding, since the sense in which I trust oracles is very different from the more sensational one in which they are more than likely interested. (Holbraad 2012a: 82, original emphasis)

Holbraad (ibid.; 96–103) proceeds to illustrate the radically different ontological premises which make Cuban divination casts—as “truth-events”— indubitably true, an alembicated exposition that not only illuminates the motile logic of coinciding paths, but, in doing so, also offers a way out of his double-bind.4 For it is with recourse to this alternative ontology, based on a non-representational semantics, and thus not easily grasped by the untrained interlocutor (untrained, that is, in the ontographic method) that Holbraad does believe in oracles. Key to his analysis is the replacement of representational with revelatory notions of truth, such that divinatory disclosures are indubitably true not because they correctly “represent” the client’s situation, but because they manifest colliding pathways or “trajectories of motion,” in this case the mythic “paths” associated with divinatory tosses and the consultant’s problematic “path” in life, converging on a correct verdict and resolution. Truth according to this “motile logic” emerges out of points of intersection rather than correspondence with facts, making Ifa’s revelations true by definition. Moreover, the lesson to be learned from these intersecting pathways is less about expanding our epistemological horizons than exploring the intensional alterity of “other” concepts in other worlds.

Barring systemic scrutiny of Holbraad’s divinatory peregrinations for now, I would like to return to the double-bind itself, and reframe the terms of its solution. Embedded in Holbraad’s familiar predicament are ideas of truth, knowledge, belief, and trust that straddle two worlds, which, if ontologically grounded in radically different ways, nonetheless can be rendered somewhat mutually intelligible. When asked if he believes in oracles, the response “I do not believe in oracles” would be a lie: that is, not true, “in an important sense” that for lay persons is difficult to grasp. In this difficult sense, “I do believe in oracles” is true, a statement that is synonymous with “I trust oracles,” but is likely to be misunderstood. To believe or not believe? The resolution of the double-bind rests on the doubling of worlds and [290]their ontologically associated truth-conditions. In World One (“our” world, for the sake of argument), “I believe in oracles” is false, but in World Two (“their” world, for the sake of argument), “I believe in oracles” is true. However we get from World One to World Two, the journey will lead us out of the double-bind.5

Holbraad’s appeal to truth by way of its negation—a lie—brings epistemology back in, for what can be more epistemological than specifying the conditions under which “x” is true? And with epistemology comes the problem of knowledge: not merely believing that “x” is true, but determining that “x” is true independently of belief. In the unorthodox provocation that follows, I will argue that a radical separation of knowledge and belief offers an alternative solution to Holbraad’s double-bind through the multiple X-files of a singular world. To get there, I return to my worst day of fieldwork in Nigeria.

A series of unfortunate events

Back in the day, and on that ill-fated day, my dissertation research focused on the politics of orisha worship within Yoruba kingdoms, and the factors that “controlled” its ritual variations. Applying Fred Eggan’s method of controlled comparison (Eggan 1954), I compared the centralized military kingdom of Ayede with the decentralized neighboring kingdoms of Ishan and Itaji (in Ekitiland) to map the correlation of political and ritual institutions over time and space, and to develop a framework for analyzing ritual performances as arenas of political negotiation and maneuvering. The “external” data were difficult but not impossible to get: shrines, memberships, altars in relation to lineages, political quarters, core and immigrant town sections, and so on. I realized that the dominant models in the scholarship were far too simplistic, underestimating the composite character of orisha cults and overestimating the lineage as the dominant sociological correlate, but that part was easy.6

Much more difficult was penetrating the domains of meaning, which were hidden and off-limits to noninitiates, safeguarded by an elaborate philosophy of secrecy that I would only gradually and partially assimilate. I was allowed to take pictures and make audio (and, later, video) recordings of festivals and rituals, but whenever I asked about the meaning of a symbol, oratorical reference, particular sacrifice, or choreographic sequence, the answer was invariably “Àsá ni” (It is our tradition), “Mi ò mo” (I don’t know), or, more enigmatically, “Awó ni” (It is a secret). As one educated chief from Ishan assured me, “Ogundele, no matter how many years you spend with the priestesses, they will never tell you anything. They will never reveal their secrets!” I began to despair that my work was destined to remain superficial, externalist, and basically blind. I could feel the allure of what I [291]eventually learned was called “imọ jinlẹ̀,” deep knowledge, and also “history,” ìtàn, as in the knowledge of the ancestors, but I despaired of ever gaining access to it.

Which is why, one early morning in April 1983, I was elated by my one and only “breakthrough” interview. It was with Ìyá Shango, the Shango priestess who had taken me on as something of a “son,” and remained my greatest patron and protector throughout my stay. Her own son, Olu Ibitoye, who became my closest research assistant, brought me to her personal shrine, where I offered schnapps and kola. What followed was simply extraordinary; after obligatory oblations to Shango, Ìyá Shango started to tell me important things. Passwords the priestesses gave each other, as well as hand signals to direct festival choreography; deep incantations to summon the gods, together with which leaves must be used, mixed with blood and palm oil; secret histories of Shango’s Nupe mother, incestuous relations, and more. As I wrote down feverishly in my notebook, I had the sense that finally I was breaking through, that the real research had begun, that I was entering into the world of the cults. The fact that the priestess whispered, with shutters closed in her personal shrine, was to prevent Shango from hearing, and getting angry. Ìyá Shango was taking quite a risk. To show my trust and appreciation, I had brought her a bag of my hair clippings from a haircut, since the hair of an òyìnbó (“European,” “white person”) was a valuable ingredient for juju, and Ìyá Shango was renowned in the medicinal arts.

But my newly gained access was not to be. Later that day, still elated, I drove off on my motorcycle with the master drummer Ajayi to Aiyegunle, about 25 kilometers northeast from Ayede, to document networks between percussionists. When we arrived, just after noon, greeting his friends, indulging in several beers, Ajayi asked if he could borrow my “machine” (motorcycle) for just twenty minutes to go off and greet another friend. Off he went, and as the minutes ticked by I grew more and more anxious. The minutes turned into hours, and I knew something was very wrong. Two hours, three hours, four hours. Finally, as the late afternoon shadows began to lengthen into evening, Ajayi turned up, weaving on my motorcycle blind drunk. Something was wrong; the fenders and handlebars were bent. He had crashed, and had a gash on his forehead. And what was worse, the backpack I had strapped on the back, with my “particulars” and precious notebook, was missing. Ajayi had no explanation other than it had disappeared. He could not explain why he was gone so many hours, other than one thing led to another—he had been drinking with friends and lost track of everything.

I was of course despondent because my breakthrough data were missing. The bike was repaired; Ajayi healed (although my relationship with him never fully recovered), but my secrets were lost forever. I tried to reconstruct them from memory but could not. Several days later a man from the “north” appeared in Ayede with my backpack and particulars, but the notebook was missing. I asked him where and how he got the backpack, but he wouldn’t say, other than that Ifá told him to bring it to me. (I had another powerful medicine in the backpack, called agbónádẹ̀rọ̀—hot becomes cool [lit.“soothing”]—which was also missing, and I think by returning the backpack he was neutralizing any revenge juju from my side.7)[292]

Anyway, I returned to my research, worked closely with the priestesses, gained access to otherwise forbidden spaces (sacred groves, inner shrines, etc.), built up a vast image and recording archive, collected oral histories of struggles and migrations, but never had a comparable interview. In fact, in the eighteen months that followed, none of the priestesses would acknowledge that the “level” of secret knowledge I had glimpsed even existed. And I stopped pushing as well. We all knew what had happened. Shango had punished me for going too far. Twelve months after the terrible loss of notes, I asked Olu if his mother believed that I had been punished by Shango on that unfortunate trip to Aiyegunle, and he said of course, it was perfectly clear to her—which was why she never blamed the drummer Ajayi since he was only an instrument of her deity. I gradually absorbed much of the character of deep knowledge, appreciating the social construction of discursive boundaries, the unstable sphere of interpretations which they sustained, and the political efficacy of its deployment in ritual, which I reworked into the concept of an indigenous hermeneutics of power (Apter 1992). In the end, the loss of that precious notebook may not have mattered that much. But I want to return to a question which the lost notebook raises: “What actually happened?” It is not just a question of ethnographic autobiography—it represents an ontological commitment to the present and the past.

Let me digress for a moment to explain why this question has returned with such resonance in my own professional life. As an anthropologist now among the historians, I am often asked what it is like shifting home departments, from anthropology at the University of Chicago, where we shared a theoretically driven commitment to critical cultural analysis, to history at UCLA, where a strong cadre of social and cultural historians use theory as a tool of illumination rather than an end in itself. To be honest, I have found the transition refreshing, captured in my typified response: “Anthropologists focus on the cultural and political frameworks and rhetorics through which the past is constructed, represented, and experienced (as historical consciousness), whereas historians are more interested in ‘what actually happened.’” I was so socialized into the constructivist perspective that the historian’s ontological commitment to “what actually happened” acquired a radical thrust, as if to say that analyzing constructions, figurations, discourses of the past is easy compared with the heavier task of figuring out the “actual.” Not a naïve invocation of the commonsense “real,” of course, but an admission that at the end of the theoretical day, after our critical investigations of object-construction, the social worlds of history and anthropology are worth getting a handle on. Figuring out what actually happened is much harder than merely exploring the sociocultural dynamics that shape and motivate representations of the real.

So returning to my worst day of fieldwork, what actually happened? Clearly any science of society requires a methodological base-line of empirical description, so where do we begin? We can start with a sequence of events, of social facts. Aspiring [293]anthropologist begins to break down barriers to secret knowledge in an early morning interview, feels excited and pumped up, takes off on his motorcycle with master drummer Ajayi for more research in Aiyegunle; lends Ajayi his Suzuki, which turns out to be a big mistake since Ajayi gets drunk, disappears for four hours, crashes the bike, returns wounded sans notebook and treasured interview data.

Let us assume, for the sake of argument, that we can all agree on a similar version of such “events”: that is, Ajayi, the Ìyá Shango priestess, my sympathetic hosts in Aiyegunle, and my sympathetic friends and associates in Ayede. How, then, do we interpret these events, in terms of meaning and causality? Here is where cultural perspectives diverge, raising well-wrought questions of truth, relativism, method, and mystification. Reduced to essentials, a typical contrast between Western rationalism and African mystification would emerge along the following lines: American anthropologist has a breakthrough interview followed by some really bad luck, owing to his own foolish judgment giving in to Ajayi’s initial request, and to Ajayi’s reckless enthusiasm once he borrowed the bike. Western moral of the story: stand firm, learn how to say no to favors asked, and never let your fieldnotes out of your possession.

Yet, on the other side of the methodological divide, from the so-called “native” point of view, the events were causally connected by the violation of a taboo. The loss of the notebook was no mere bad luck, but was Shango lashing back, angry that his secrets were leaked and rectifying the situation by depriving me of my precious data. From the rationalist perspective of Western social science, this is of course a mystification, for how can we scientifically accord agency to mystical entities? Local natives may believe this happened, but belief and knowledge occupy very different orders of social determination. We “know” that 2 + 2 = 4 even if I believe (mistakenly) that it equals 5. For the objectivist social scientist, mystified beliefs become true only when they are explained with reference to the objective social (political, cultural, etc.) principles that govern their distortion. Cult secrets, from this perspective, are protected by social sanctions such as belief in mystical retribution when secrets are leaked. Much ink has been devoted to exploring this methodological divide, ranging from total relativism (what “happened” is always culturally specific, including the cultural specificity of Western scientific rationality) to scientific reductionism (we demystify “native” thought with reference to objective sociological principles), but this stark methodological opposition is today a caricature; the debates on both sides have become more complex.

Let me muddy the waters a bit. First, it is interesting to note that even at this preliminary stage of my time in the field, I suspected, no, I actually knew, that Shango was striking back. Nobody explained this to me, or suggested it to me. I felt it immediately, to invoke the language of Bourdieu’s embodied knowledge: I could just “tell” since I was already [294]working within a local knowledge-system that inculcated certain social and cognitive dispositions. Recall that I was already using juju, both as a gift from my body but also as a means of gaining social cooperation. Did I believe in it? Not really. Did it work? Yes. Do you among my readers know exactly what I am talking about? Many of you X-filers do.

Thus far, Western social theory can explain my incipient Yoruba cultural verstehen, my developing intuitive grasp of the events from a Yoruba perspective as a process of local socialization, the gaining of cultural competence that comes from working within a set of socially embedded structures of reproduction and calculation. But what happens when these boundaries between observer and observed, and their concomitant worlds, begin to dissolve? If critical anthropology has taught us anything, it is that our methodological appeals to rationality and objectivism are overdetermined by the historical conditions of their genesis—shot through with implicit assumptions of difference, evolutionism, personhood, and transparency that are anything but transparent. So what actually happened? I am not suggesting that what happened on my worst day of fieldwork was simply Shango’s punishment, but something more paradoxical: I know that Shango snatched my fieldnotes and literally erased my morning interview because such data, not properly “given,” had to be taken back. Yet I don’t believe that he did this because I don’t believe that Shango really exists. Here is my version of Holbraad’s double-bind. For him the way out is through ontographic explication: once we clarify the ontology of the world in which Shango strikes back, and the conditions in which “Shango took my fieldnotes” is true, then we can enter into this alternative world, and engage other actual and possible worlds. My solution is much more direct: I know that p is true even if I believe that it is false.8 Applied to Holbraad’s double-bind: I know that oracles never lie (hence always tell the truth) but I don’t believe it. How can we know what we don’t believe?

That is the question.

Knowledge, belief, and the irrational ethnographer

Perhaps it is best to begin by specifying what I don’t wish to convey by this ethnographic paradox. First, I am not extending Rodney Needham’s assault on the mercurial concept of belief itself, which for him generates so much ambiguity and confusion that it is best abandoned as a cross-cultural category—particularly when “ethnographical ascriptions of belief” serve as the basis of our studies (Needham 1972: 9). Rather I am negating what Skorupski (1976: 238) calls an “empirical belief” about the existence of Shango, and, by extension, the category of orisha that he manifests. Belief as such may come under critical scrutiny, and I may push it to questionable extremes, but I do not reject it as a meaningful concept. Second, and this may take some convincing on my part, I hope that I am not merely proposing a pseudoepistemological cover for what boils down to my own psychological ambivalence or denial, in the sense that I “feel” like I know that Shango intervened but can’t bring myself to believe it because I am an atheist and a realist. Rather, I maintain that such psychological denial provided the impetus toward developing a conceptually interesting position that goes beyond ascriptions of inner mental or subjective states. 9 Third, I am not seeking a “half [295]way house” (ibid.: 243) between religion and science, or ontological worlds, such that “I know p in World2 but don’t believe p in World1,” for that only resurrects the ontological divide that I ultimately seek to avoid. Although there are multiple logics and logically possible worlds, as a realist I maintain that anthropologists and those we study exist in one world. Nor, fourth, do I wish to replace different “worlds” with different levels of analysis, disbelieving in p as a mystical agent while knowing it as a Durkheimian social fact. If I remain inclined toward socially grounded illuminations and explanations, in this case they don’t apply because the p which is true (belief as a social fact) is not the same p which is false (mystical agent) in such formulations. And finally, I am not attributing my own paradoxical proposition to the Yoruba discourse community among whom it arose, although similar ambivalences exist. Like Graeber’s discussion of the skeptical framing of fanafody medicine in Madagascar (Graeber 2015), Yoruba attitudes toward the orisha include reflexive forms of bargaining power that are central to the devotee–deity relationship, as propounded in Karin Barber’s landmark essay “How man makes God in West Africa” (1981). But the case of my worst day of fieldwork in Nigeria foregrounds a different propositional attitude, one shared not by my Yoruba colleagues—who knew and believed that Shango lashed back—but among anthropological colleagues with similar experiences who know without believing. We thus frame the problem of Shango’s wrath with the irrationality not of “natives” but of ethnographers.

So how irrational are we? Can the sentence “I know that Shango intervened [by taking back my fieldnotes] but I don’t believe it,” or more simply, “I know p but I don’t believe it” (or even “p and I don’t believe it”) be dismissed as self-contradictory, or as an irrational (dis)belief? If so, then I have reached an aporia, and can go no further. The sentence, however, cannot be so easily dismissed, but belongs to a class of so-called “Moore-paradoxical” sentences—not surprisingly formulated by G. E. Moore—that has spawned a veritable industry of epistemological hair splitting (including within the subfield of epistemic logic, see Hintikka 1962: 64–78, 95–102; Henricks and Symons 2006) which began with an inspired Wittgenstein and shows no signs of slowing down (see, e.g., Green and Williams 2007; Schmid 2014).10 It was in fact Wittgenstein who coined the term “Moore’s paradox” in part II of his Philosophical investigations ([1953] 1972) (variations of which appear in Remarks on the philosophy of psychology based on writings from 1946–47), as posing a momentous logical problem. His immediate excitement is evident in a letter he wrote the day after Moore presented a paper to the Moral Sciences Club in October 1944 praising Moore for his important “discovery” associated with the “absurdity” of the assertion “There is a fire in this room and I don’t believe there is.” He adds, “Pointing out that ‘absurdity’ which is in fact something similar to a [296]contradiction, though it isn’t one, is so important that I hope you’ll publish your paper” (original emphasis).11

Wittgenstein’s ruminations on Moore’s paradox are characteristically complex, subtle, fragmentary, and elusive, but two features stand out in his philosophical framing that help develop my position.12 First, is the insight that Moorean sentences are not paradoxes, but only resemble them as seemingly absurd. For example, in the sentence “There is a fire in this room and I don’t believe there is,” both conjuncts can be true, whereas in a proper paradox they cannot (see also Moore 1993: 209; Sorenson 1988: 15; Williams 1979; Shoemaker 1995: 211). Moore-paradoxical statements “feel” or “sound” like paradoxes, but technically they are not. Thus we can say, somewhat reflexively, that the paradox of Moore-paradoxical sentences is that they are not paradoxical. The point is important because it establishes a philosophical justification for recasting statements like “Shango took back my fieldnotes and I don’t believe it” as nonparadoxical (and thus noncontradictory), with the additional virtue that they are philosophically interesting.13 The second striking feature of Wittgenstein’s take on Moore’s paradox is his rejection of psychological conditions such as states of mind associated with attributions of belief, thereby rejecting Moore’s own appeal to states of mind (belief) implied by assertions (Moore [1944] 1993; Malcolm 1995). As Wittgenstein (1980: 92e) remarked, “It would be asking for trouble to take Moore’s paradox for something that can only occur in the mental sphere,” pushing instead for logical approaches to belief in the changing contexts of different language games.14 Again, my point remains preliminary: Wittgenstein rules out psychological (or mentalist) resolutions of Moorean paradoxes because they side-step the deeper roots of the problem, which hinges on the slippage between supposition and assertion in first-person indicative statements of belief.15 Whether or not Wittgenstein resolved Moore’s paradox, he prepared the ground for nonpsychological solutions.[297]

With Wittgenstein, then, we have established that my way out of Holbraad’s double-bind remains valid: neither logically paradoxical nor psychologically reducible to an irrational state of mind, it provides an opening for a category of ethnographic experience—our classified X-files—that allows “knowledge of” without “belief in.” It is admittedly an extreme formulation that pushes the meanings of “knowledge” and “belief” not only to unusual limits, but also into novel semantic possibilities. If philosophers recognize the complex intersections of knowledge and belief, the lines remain notoriously fuzzy (Hintikka 1962; Stalnaker 2006). For example, in 1963, Edmund Gettier shattered an established epistemological position that all knowledge is justified true belief by proving that rational actors could hold justified true beliefs that nonetheless do not qualify as knowledge, based on possible if highly unlikely scenarios that lead to false conclusions. In Case I of Gettier’s famous three-page article (Gettier 1963: 122), for example, Smith knows that either he or Jones will get a certain job, and that the person who will get the job has ten coins in his pocket; Smith also knows that Jones has ten coins in his pocket without realizing that he himself (Smith) also has ten coins in his pocket. Smith therefore has a justified true belief that Jones will get the job, which turns out to be false, because Smith himself (with his unknowingly held coins) gets the job. The so-called Gettier problems or cases are important because they prove that not all justified true belief is knowledge, relaxing their relations of material coimplication. Taken together, Moore’s paradox and the Gettier problems frame the limits of an epistemological opening between knowledge and belief.16 For Moore, “p and I don’t believe p” gives us knowledge without belief. Conversely, Gettier gives us justified true belief without knowledge.


If nothing else, I have offered an alternative solution to Holbraad’s double-bind in the Moorean sentence, depsychologized by Wittgenstein, of the form “I know p and don’t believe it.” Because it is neither a “true” paradox nor irrational, it is logically permissible even if its meanings strike us as enigmatic, and requires no further commitments to multiple ontologies. Such an epistemological alternative to the ontographic method may miss the point as far as the OTers are concerned, for whom “differences in ‘perspectives’ are to be seen in ontological rather than epistemological terms” (Holbraad 2013: 469), and this in order to gain insight into the possible worlds of radical alterity. The thinking subjects and thoughtful objects encountered by the new ontologists, taken on their own terms, project possible worlds which are actualized “for us” by ethnographic explication—through the kind of ontographic conceptualization that opens up new modalities of being. Perhaps. Yet the conjuring of these possible worlds by the ethnographer’s magic, and, more to the point, the “passage from one possible world to another” (Viveiros de Castro [2002] 2013: 479), remain dubious achievements, quite possibly impossible. In determining the meanings of proper names, as Kripke reminds us, “‘possible worlds’ are stipulated, [298]not discovered by powerful telescopes” (1980: 44, original emphasis). Nor, we might add, by ontological anthropologists.

Which raises the question, what kind of “possible worlds” are the OTers dealing with? At first blush, they appear very different from the arid counterfactual landscapes of possible-world semantics. In an article which for Holbraad (2013: 470) marks the turning point of the ontological turn, Viveiros de Castro ([2002] 2013: 478) invokes Sartre (“le regard”) and Deleuze (1969: 355) to argue that possible worlds are grounded in the structure of the Other as a kind of a priori alterity.17 If I follow him correctly, and at the risk of oversimplification, there seem to be two major moves involved in exploring other ontologically possible worlds. The first is to recognize the philosophical relationality underlying possible worldhood—including “ours”—a kind of transcendental other = x that guarantees the very category of world possibility. This relational condition is not only primordial, but also underscores the varieties of perspectivism (Amerindian and “otherwise”) through which it is empirically manifested (e.g., Viveiros de Castro 1998). The second move into empirically possible worlds involves ethnographic discovery procedures of the ontographic method: taking “native” statements, objects, meanings, and practices that manifest internally the relational (perspectival) concepts of the Other as starting points for anthropological inquiry. It is precisely those statements that seem “strange” or “irrational” to “us,” such as “peccaries are human,” or “powder is power,” which provide openings for the following question: In what possible world are they true? That the question is philosophical, inspiring thought experiments with other possible worlds, brings philosophy into an applied or experimental mode. In a playful (if somewhat unfortunate) reversal of disciplinary registers, Viveiros de Castro proclaims: “If in ‘real philosophy’ imaginary savages are altogether abundant, the geo-philosophy proposed by anthropology conducts ‘imaginary’ philosophy with ‘real’ savages” ([2002] 2013: 488). We may recall not just the noble and ignoble “savages” of early modern political philosophy, but also Quine’s imaginary “native” whose utterance “Gavagai” (“rabbit,” “undetached rabbit part”) established the inscrutability of reference and the indeterminacy of translation (Quine 1960: 29–46).

But the “geophilosophical” crossroads also share a common ground, although possible-world semantics developed with very different goals in mind: namely the search for necessary meanings across all possible worlds rather than those specific to particular ones. Like its ontographic cross-cousin, its basic modus operandi stipulates that the meaning of a sentence specifies how the world would have to be for that sentence to be true (or false). If, for philosophers, possible worlds serve meanings, and for ontographers, meanings serve possible worlds, both approaches share the same stipulative conditions of truth by definition. If such truths, in principle, are conceptually knowable, are they believable? To be sure, the philosopher does not need to believe in the possible world where Richard Nixon died at age four to determine, or even “know,” the meaning of his proper name.18 But what of the possible worlds which the ontographer explicates, and in some sense actualizes [299]“for us”? That the question is important for Viverios de Castro is revealed by his response to Isabella Lepri, a student also working with Amerindians who asked him if, by taking the statement that “peccaries are humans” seriously, he believed that it was true. When he replied in the negative, she responded: “How can you take the [Amerindians] seriously if you only pretend to believe in what they say?” (Viveiros de Castro [2002] 2013: 493). He reflects: “I am convinced that Isabella’s question is absolutely crucial; that all anthropology deserving of the name should answer it; and that it is not at all easy do so very well” (ibid.).

Viveiros de Castro’s solution is to dismiss this question of “our” belief as irrelevant, no more pertinent to Amerindian ideas than “greenness” is to the number two, a category error better avoided by focusing on “what the Indians are saying when they say that peccaries are human” (ibid.: 494). Here is where applied philosophy takes over: “how to transform the conception expressed in a proposition like this into a concept” (ibid.), which, when properly done, provides access to other construals of human–animal relationality in its manifold complexity. My own reply to Isabella Lepri, based not on native testimony but on my worst day of fieldwork in Nigeria, would be: “Peccaries are human and I don’t believe it.” Moore’s paradox lives on in the ethnographic present without proliferating ontological commitments to multiple worlds. It not only resolves double-binds and epistemological conundrums in the field, but may help us rethink our potentially embarrassing X-files, and what we mean by “knowledge.”


Apter, Andrew. 1992. Black critics and kings: The hermeneutics of power in Yoruba society. Chicago: University of Chicago Press.

Barber, Karin. 1981. “How man made God in West Africa: Yoruba attitudes toward the òrìsÀ.” Africa 51 (3): 724–45.

Bascom, William. 1944. “The sociological role of the Yoruba cult-group.” American Anthropologist (N.S.) 46 (1, part 2, Memoires 63): 1–75.

Berliner, David, ed. 2016. “Anthropology and the study of contradictions.” HAU: Journal of Ethnographic Theory 6 (1): 1–27.

Bessire, Lucas, and David Bond. 2014. “Ontological anthropology and the deferral of critique.” American Ethnologist 41 (3): 440–56.

Borges, Rodrigo. 2014. “How to Moore a Gettier: Notes on the dark side of knowledge.” Logos & Episteme 5 (2): 133–140.

Deleuze, Gilles. 1969. Logique du sens. Paris: Minuit.

Eggan, Fred. 1954. “Social anthropology and the method of controlled comparison.” American Anthropologist 56 (5), Part 1: 743–63.

Evans-Pritchard, E. E. 1937. Witchcraft, oracles and magic among the Azande. Oxford: Clarendon Press.

Gettier, Edmund L. 1963. “Is justified true belief knowledge?” Analysis 23 (6): 121–23.[300]

Graeber, David. 2015. “Radical alterity is just another way of saying ‘reality’: A reply to Eduardo Viveiros de Castro.” HAU: Journal of Ethnographic Theory 5 (2): 1–41.

Green, Mitchell, and John N. Williams, eds. 2007. Moore’s paradox: New essays on belief, rationality, and the first person. Oxford: Clarendon Press.

Heal, Jane. 1994. “Moore’s paradox: A Wittgensteinian approach.” Mind (N.S.) 103 (409): 5–24.

Henare, Amiria, Martin Holbraad, and Sari Wastell. 2007. “Introduction: Thinking through things.” In Thinking through things: Theorising artefacts ethnographically, edited by Amiria Henare, Martin Holbraad, and Sari Wastell, 1–31. London: Routledge.

Henricks, Vincent F., and John Symons. 2006. “Where’s the bridge? Epistemology and epistemic logic.” Philosophical Studies: An International Journal for Philosophy in the Analytic Tradition 128 (1): 137–67.

Hintikka, Jaako. 1962. Knowledge and belief: An introduction to the logic of two notions. Ithaca, NY: Cornell University Press.

Holbraad, Martin. 2007. “The power of powder: Multiplicity and motion in the divinatory cosmology of Cuban Ifá (or mana, again).” In Thinking through things: Theorising artefacts ethnographically, edited by Amiria Henare, Martin Holbraad, and Sari Wastell, 189–225. London: Routledge.

———. 2012a. “Truth beyond doubt: Ifá oracles in Havana.” HAU: Journal of Ethnographic Theory 2 (1): 81–109.

———. 2012b. Truth in motion: The recursive anthropology of Cuban divination. Chicago: University of Chicago Press.

———. 2013. “Turning a corner: Preamble for ‘The relative native’ by Eduardo Viveiros de Castro.” HAU: Journal of Ethnographic Theory 2 (1): 469–71.

Idowu, Bolanle. 1962. Olódùmarè: God in Yoruba belief. London: Longmans.

Kripke, Saul. 1980. Naming and necessity. Oxford: Basil Blackwell.

MacIver, A. M. 1938. “Some questions about ‘know’ and ‘think.” Analysis 5 (3/4): 43–50.

Malcom, Norman. 1984. Ludwig Wittgenstein: A memoir. Oxford: Oxford University Press.

———. 1995. “Disentangling Moore’s paradox.” In Wittgenstein: Mind and language, edited by Rosaria Egidi, 195–205. Dordrecht: Kluwer Academic Publishers.

McGinn, Marie. 2011. “Wittgenstein and Moore’s paradox.” In Image and imaging in philosophy, science and the arts, Vol. 1, edited by Richard Heinrich, Elisabeth Nemeth, Wolfram Pichler, and David Wagner, 59–82. Frankfurt: De Gruyter.

Moore, G. E. 1942. “Reply to my critics.” In The philosophy of G. E. Moore, edited by Paul Arthur Schilpp, 535–677. The Library of Living Philosophers, Vol. 4. Evanston, IL: Northwestern University Press.

———. 1944. “Russell’s theory of descriptions.” In The philosophy of Bertrand Russell, edited by Paul Arthur Schilpp, 175–225. The Library of Living Philosophers, Vol. 5. Evanston, IL: Northwestern University Press.

———. (1944) 1993 . “Moore’s paradox.” In G. E. Moore: Selected writings, edited by Thomas Baldwin, 207–12. London: Routledge.[301]

Morton-Williams, Peter. 1964. “An outline of the cult organization and cosmology of Old Oyo.” Africa 34 (3): 243–61.

Needham, Rodney. 1972. Belief, language, and experience. Chicago: University of Chicago Press.

Popper, Karl. 1972. Objective knowledge: An evolutionary approach. Oxford: Clarendon Press.

Quine, Willard V. O. 1960. Word and object. Cambridge, MA: MIT Press.

Schmid, Ulla. 2014. Moore’s paradox: A critique of representationalism. Berlin: De Gruyter.

Shoemaker, Sydney. 1995. “Moore’s paradox and self-knowledge.” Philosophical Studies 77: 211–28.

Skorupski, John. 1976. Symbol and theory: A philosophical study of theories of religion in social anthropology. Cambridge: Cambridge University Press.

Sorensen, Roy A. 1988. Blindspots. Oxford: Clarendon Press.

Stalnaker, Robert. 2006. “On logics of knowledge and belief.” Philosophical Studies: An International Journal for Philosophy in the Analytic Tradition 128 (1): 169-99.

Verger, Pierre. 1957. Notes sur le culte des orisha et vodun À Bahia. Dakar: Institute français d’Afrique noir.

Viveiros de Castro, Eduardo. 1998. “Cosmological deixis and Amerindian perspectivism.” Journal of the Royal Anthropological Institute (N.S.) 4 (3): 469–88.

———. (2002) 2013. “The relative native.” HAU: Journal of Ethnographic Theory 3 (3): 473–502.

von Wright, G. H., ed. 1974. Ludwig Wittgenstein: Letters to Russell, Keynes, and Moore. Oxford: Basil Blackwell.

Williams, John N. 1979. “Moore’s paradox: One or two?” Analysis 39 (3): 141–42.

———. 1994. “Moorean absurdity and the intentional ‘structure’ of assertion.” Analysis 54 (3): 160–66.

Wittgenstein, Ludwig. (1944) 1974. “Letter to G. E. Moore [M.42].” In Ludwig Wittgenstein: Letters to Russell, Keynes, and Moore, edited by G. H. von Wright, 177. Oxford: Basil Blackwell.

———. (1953) 1972. Philosophical investigations. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell.

———. 1980. Remarks on the philosophy of psychology, Vol. 1. Edited by G. E. M Anscombe and G. H. von Wright. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell.

Les X-files de l’ethnographie et la double contrainte de Holbraad: Réflexions sur le tournant ontologique des évènements

Résumé : Le tournant ontologique pose des questions difficiles et repousse les frontières conceptuelles À la fois en avant et À rebours, mettant au défi nos orientations [302]épistémologiques standards. Dans cet essai, je m’oppose À l’argument selon lequel les questions ontologiques remplacent les enjeux épistémologiques. Je m’inspire À cet égard de l’idée de “double contrainte” émise par Holbraad, qui constitue une problématisation résonnant avec mon expérience ethnographique, mais qui représente aussi une part du danger professionnel (nos “X-files”) pour ceux qui comme moi, s’intéressent au monde des esprits. Dans une solution alternative À la double contrainte de Holbraad, qui s’inspire du paradoxe de Moore et sur ma propre expérience de la colère de Shango au Nigéria, je propose un décuplement du savoir et de la croyance afin d’éviter la prolifération ontologique de mondes.

Andrew APTER is Professor of History and Anthropology at UCLA. His books include Black critics and kings: The hermeneutics of power in Yoruba society (1992); The pan-African nation: Oil and the spectacle of culture in Nigeria (2005); Beyond words: Discourse and critical agency in Africa (2007); and Oduduwa’s chain: Locations of culture in the Yoruba-Atlantic (2017), all with the University of Chicago Press. He also coedited Activating the past: History and memory in the Black Atlantic world (Cambridge Scholars, 2010) with Lauren Derby, and is currently working on History in the dungeon: Atlantic slavery and the spirits of capitalism.

Andrew Apter
Departments of History and Anthropology
6265 Bunche Hall
Box 951473
Los Angeles
CA 90095-1473


1. For two signal critiques of the ontological turn, see Bessire and Bond (2014) and Graeber (2015).

2. As Graeber (2015: 35–36) reminds us, even Evans-Pritchard (1937:34) went partly public with his X-files.

3. In 1996 I appeared in an A&E Unsolved Mysteries television episode on Haitian zombies, to weigh in on the question “Are zombies real?” I quickly realized there was no room for anthropological subtleties; the answer was either yes or no. I argued no, contra Wade Davis.

4. For his richer ethnographic case study of Cuban divination along these “lines,” see Holbraad (2012b).

5. Holbraad’s worlds are not to be confused with Popperian Worlds (see Popper 1972).

6. I conducted dissertation fieldwork in Nigeria from September 1982 to December 1984, funded by Fulbright–Hays and the Social Science Research Council. For the dominant models of orisha cult organization at that time, see, e.g., Bascom (1944); Verger (1957); and Idowu (1962). For an important shift beyond from the “lineage” model, see Morton-Williams (1964).

7. The agbónádẹ̀rọ̀, also called ojú mẹ̀jo (“eight eyes”) because of its eight cowries protruding from its leather binding, had been a gift six years earlier from the late Twins Seven-Seven in his maternal natal village of Ogidi, in gratitude for my cutting a ring (with my Swiss Army knife) from his finger after it had grown deeply into his flesh—a procedure that endured into the early morning hours and involved considerable bleeding and ogogoro (local gin). I was told that the juju had returned to Ogidi after Ajayi’s motorcycle accident. But that is another story and potential X-file.

8. Michael Lambek notes that the reverse proposition is more common, “believing something true even when we know it is false” (personal communication). This point adds a new level to the hierarchy of constraints on the knowledge–belief relation and resonates with the Gettier problems discussed below

9. The edited debate by Berliner (2016) raises many related issues pertaining to the significance of contradictory ideas, attitudes, and beliefs of anthropologists in the field, suggesting that multiple selves may be a subjective alternative to multiple ontologies. I am particularly drawn to Richard Shweder’s defense of the law of noncontradiction in this discussion precisely because that is the virtue of Moorean sentences, which preserve the noncontradiction of knowing without believing (see below).

10. According to von Wright (1974: 177–78), the first to formulate “the puzzle about saying and disbelieving” was MacIver (1938). See also Moore (1942: 543; 1944: 204; 1993).

11. Wittgenstein’s letter to Moore was published as M.42 by von Wright (Wittgenstein [1944] 1974). Although von Wright (1974: 177–78) notes that he could not find Moore’s unpublished paper, it was later given to University Library, Cambridge, by his son, Timothy Moore, and published as Moore (1993). In his Memoir, Norman Malcolm recalls Wittgenstein mentioning that “the only work of Moore’s that greatly impressed him was his discovery of the peculiar kind of nonsense involved in such as sentence as ‘It’s raining but I don’t believe it’” (1984: 66). See also Sorensen (1988: 1), who opens his book with this quote and devotes his first chapter to Moore’s paradox as paradigmatic of philosophical “blindspots.”

12. For helpful expositions of Wittgenstein’s position on Moore’s paradox, see Heal (1994), Malcolm (1995), and McGinn (2011).

13. Since “but” is not a logical operator, we can use “and” in its place, as per standard practice.

14. For an elaboration of this strategy, see McGinn (2011). See also Green and Williams (2007: 6–8).

15. For a critical engagement of the “assertability conditions” approach to Moore’s paradox, emphasizing hearers over speakers, see Williams (1994).

16. To my knowledge, Borges (2014) is the first to see that Moore’s paradox and Gettier cases are intimately linked.

17. The article was first published in 2002 as “O nativo relativo,” Mana 8 (1): 113–48.

18. The example derives from Kripke (1980: 40-49).