Deconstruction’s doubt

This work is licensed under the Creative Commons | © Rosalind C. Morris. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau6.1.032


Deconstruction’s doubt

Rosalind C. MORRIS, Columbia University

Comment on Bubandt, Nils. 2014. The empty seashell: Witchcraft and doubt on an Indonesian island. Ithaca, NY: Cornell University Press.

There is an extraordinary drawing reproduced in Nils Bubandt’s ambitious and dense but beguiling book, The empty seashell: Witchcraft and doubt on an Indonesian island. On one “level” of the manically lined illustration is a monstrous, skirted being with a small head and clearly distinguishable features; beneath, its nipples and genitals seem to show through a sheath and function as the eyes and mouth of another figure. There is nothing in the image—neither perspective nor stylistic markers—that subordinates one “level” to the other. Rather, they vacillate in a sexualized version of the rabbit-duck image so familiar to readers of Wittgenstein’s Philosophical investigations. Bubandt never analyzes the image, but it provides the iconic materialization of his thesis that witchcraft is the “ambivalent aporia, the interminable problem” of life in Buli, North Maluku. The empty seashell of the title and cover image, a shell in which there might or might not be an occupant, provides, by contrast, the de-sexualized figure for doubt. Such doubt, says Bubandt, saturates the consciousness of those who inhabit the world of witchcraft and has led them to embrace the promise of certitude and the elimination of witchcraft offered by Christianity, Suharto’s New Order, or modern technology. Such is the argument of The empty seashell, which asks moderns to relinquish the belief in other people’s belief, and promises to change the way we think about witchcraft.

The discourse that The empty seashell would supplant has itself been revised over the last two decades. During that period, the anthropology of witchcraft has attempted to save itself from charges of residual exoticism by reading witchcraft phenomena as sites at which the transformations of modernity are mediated and made available for resignification (see especially Comaroff and Comaoff 1993, 1999; Geschiere 1997, 2013). It has offered itself as counterpoint to the sensational [512]media accounts of resurgent and vindictive violence against accused witches in many parts of the decolonizing world. For the most part, this has been to the good. Nonetheless, a certain diffidence often afflicts writers on the topic. Thus, for example, following the lethal violence at Marikana, South Africa, in August 2012, when confrontations between striking miners, state police, mining corporations, and unions led to the deaths of forty-four people, analyses of the events were riven by the question of whether and how to discuss the presence of sangomas and the practice of muti magic in the events. Many black South African analysts insisted that muti be taken seriously and they chastised timid white commentators who, wanting to explain the violence in materialist terms, effaced the force of magic in the lives of the migrant laborers. For them, the magic was a not matter of resurgent tradition. Nor was it a phantom projection of anxious modernists. It was a matter of power: visceral, strategically instrumentalizable, and absolutely lethal power. The kind that enables people to go to war, the kind that operates at the point where language fails.

Nils Bubandt’s account of witchcraft in Buli, is, by contrast, a story about discourse, even when this discourse is said to be visceral and even when it is addressed to the lethal effects of cannibalistic beings called gua who menace the society in greedy imitation of normal life. This is not a fault. It is an inevitable function of the phenomenon to which it addresses itself. For, as Evans-Pritchard demonstrated so long ago, witchcraft only becomes visible in the moment of accusation, as a retrospective effect of oracular diagnosis. So, too, the anthropologist learns of witchcraft only in conversation with those who have experienced it. Even when he hears what might be a witch—in the form of a dog on his roof—he has to rely on others to interpret his auditory experience. But there’s the rub—because no one is sure that he has indeed heard a witch. Bubandt reads this fact—the dubitability that afflicts witchcraft discourse—not as an insufficiency of a particular case but as the essential characteristic of all witchcraft in Buli, and perhaps elsewhere: “Witchcraft is not a system of belief, but a condition of doubt” (2014: 237).

What is this doubt that would replace belief as the core of witchcraft? It is doubt about whether illness is caused by witchcraft, about whether one’s neighbor is a witch, and ultimately about whether one is oneself a witch. For people in Buli, according to Bubandt, witchcraft does not explain the world. It is not an alternative mode of reasoning (as Evans-Pritchard had claimed). Nor does it permit anyone to control the forces that assault and wound human beings—though such efforts at instrumentalizing it are continually made. It provides neither certitude nor escape from the anxieties that death bears for the living. To the contrary, it provides an idiom in which the world’s very immunity to explanation is affirmed—with often violent and terrifying consequence. Above all, it provides figures and narrative forms in which to address but also reproduce the opacity and ambivalence of sociality. For, in Buli terms, the interiority of the other is unknown and unknowable, though we must engage others and seek recognition from them to escape solitude and death. We must engage others in forms of giving and reciprocation that demand generosity but that are also likely to provoke avarice—and it is often hard to tell the difference.

Witchcraft emerges in this context as a relentless question about appearances and the problem of knowing what they disclose. Bubandt speaks of this in terms of the [513]relation between absence and presence but he steers clear of any explicit engagement with psychoanalytic theorizations of the Symbolic and this is somewhat perplexing, given the proximity between his arguments about an internally split subject, a symbolic order understood to be rent by the problem of dissimulation, and frequent and explicit recourse to the notion of the “uncanny.” It also makes the analysis of images such as the drawing described above somewhat sexless. What does one make of an image that allows for the interchangeability of consumption and sexual relation? When it comes to questions of sexuality, Bubandt is more structuralist than Derridean—though he is otherwise thoughtful in his efforts to show how a deconstructionist analysis could enable anthropology. He links the doubling of genitalia and mouth in Buli dream imagery (but not the drawing), via homology, to the pairing of sexuality and consumption in a Buli discourse on marriage as the “epitome of tradition.” That tradition, he says, “establish[es] . . . the ideal format for conviviality,” ensuring that sexuality and consumption are made the basis of sociality rather than witchcraft. To be sure, the images of witchcraft assaults often entail the penetration of the genitals of the victim—in terrifying scenes that Bubandt describes without reference to the question of sexual difference or desire—but the fact that witchcraft emerges as a perversion of marriage might have led to a deeper exploration of sexual difference rather than a generic “sexuality.” When gender appears as an object of analysis in this book, it is mainly to describe the organization of political relations, via the homology with a presumptively natural hierarchy between male and female in a system of reciprocity. We might note that this kind of structural analysis of exchange was the object of Derrida’s most powerful critique of structuralism, but this lesson seems to have been less important to Bubandt than the general principle of aporia.

Given the claim that there are ritual traditions such as marriage to manage witchcraft, the desire to escape its psychic predations requires a different explanation. To this end, Bubandt historicizes the problem by describing three episodes over the last century and a half in which Buli residents repeatedly embraced the idea that they could escape witchcraft and be done with doubt: Christian missionization during the late nineteenth and early twentieth century, and again in the 1930s; modernist developmentalism and statism under Suharto; and technologization associated with natural resource-based capitalism. In each, a scenario of felt lack on the part of the inhabitants of Buli becomes the basis for a mistranslation or misrecognition of the foreign modern, and this in turn precipitates an effort to become modern and to thereby enter an order of truth and knowledge, rather than dissimulation and doubt.

In the first, Bubandt renarrates the history of the millenarian movement that swept Halmahera at the end of the nineteenth century. He suggests that Buli conceptions of death left them vulnerable to the Christian promise of resurrection, not least because it resembled their own myths about a return of the ancestors. Overlaying a desire for the restoration of the Jailolo sultanate (an originary moment in Buli historical consciousness), the Christian millennial fantasy summoned incipient desires, but the violent denigration of witchcraft by Calvinist missionaries also reinforced its reality. For, if the old demons required such vigorous prosecution, they must have had a power worthy of contravention.

Haunted and enabled by Marshall Sahlins’ (1987) concept of the “structure of conjuncture,” Bubandt shows how the formal rhyming between otherwise different [514]cosmologies worked to reinforce rather than undermine witchcraft. Thus, the emphasis on the gift inscribed in Protestant Christianity merged with the valorization of reciprocity in Halmahera (Bubandt 2014: 114–15), and the missionaries were themselves willing to use the misunderstanding to their own ends (116). When the hoped for return of the dead did not occur, Christianity appeared as duplicitous as the world of witchcraft.

Again and again, such poignant ironies show themselves. Thus, under the terms of the New Order’s pancasila policy, animism was disavowed and much of Buli tradition assumed an unpatriotic and ultimately criminal status (see, especially, Bubandt 2014: 155–77). But the prohibitions on superstition were indiscriminately applied to both witches and those seeking protection from witches. Thus, the signatories to a petition seeking the arrest of a witch who had ostensibly killed a villager could be prosecuted in the same manner as someone accused of witchcraft. The fact that many officials were themselves “believers” only complicated the matter. To be fair, Bubandt does not speak of “believers.” He writes of the “practical enrollment of many figures of state authority in the reality of the gua” (155), and this circumlocution is symptomatic of the tension that mounts throughout the text as Bubandt insists, again and again, that witchcraft is a discourse of doubt rather than belief. The only belief he will grant is that of Bulians “wishing to believe in the state ideology that in a modern world there where [sic] no devils . . . [who] hoped that those figures who knew from experience about the reality of the gua could be persuaded to use the power of the state to implement a world without witches” (155).

The reader may pause on occasion, despite the rich and patient ethnography of The empty seashell, to ask herself whether doubt is, indeed, the opposite of belief, or whether the insistence on doubt might constitute a legitimation by reversal of the very structure that Bubandt wants to escape. For, is not doubt the condition of possibility of belief? And is not the coimbrication of doubt/belief precisely what the discourses of the Enlightenment science, namely knowledge through revelation of what is, promised to replace? Bubandt invites such questions by virtue of the strenuousness with which he claims to be rethinking the discipline.

In fact, there are two dimensions along which doubt performs its function: an affective dimension, in which doubt is the cause of anxiety, fear, suspicion, and in some cases, rage; and an epistemological dimension, in which doubt is the effect of an incapacity to transcend the aporia that defines the relation between the empirical and the transcendental. These are my terms, or rather Kant’s terms, not Bubandt’s, and I use them to point out what I believe may be a conflation at the heart of The empty seashell, between aporia and doubt on one hand, and between witchcraft and witches on the other.

The empty seashell opens with an ambiguous statement: “I have never seen a cannibal witch” (2014: ix). This first narrative gesture stages anthropology’s epistemological conundrum. An a priori postulate—that there are cannibal witches—confronts an empirical experience that neither proves nor disproves it. The postulation of an irreducible impasse between the empirical and the transcendental, of the incapacity of sensory experience (intuition) to provide the basis of absolute (and not merely general) knowledge was, of course, the philosophical contribution of Immanuel Kant. Bubandt comes to this postulation via Jacques Derrida, although few people paid more attention to the consequences of the Kantian problem for [515]anthropology than did Michel Foucault. Foucault ([2007] 2008), who translated and introduced Kant’s Anthropology from a pragmatic point of view to French audiences, observed that the epistemological problem of the Critique of pure reason was transferred to the very category of the human in Kant’s later writings. Just as intuition cannot ground knowledge of the absolute, so the knowledge of individual human beings and their empirical histories cannot provide knowledge of “the human.” Foucault aptly designated Kant’s “human” an “empirico-transcendental doublet.” But he was not interested in ethnography, and certainly did not speculate on the ways in which this doublet would be implicated in the recurring phenomenon that Bubandt, like James Siegel before him, rightly sees as the originary aporia to which witchcraft responds.

Marcel Mauss ([1972] 2005) extended the Kantian postulate in his efforts to understand magic, and turned not so much to the concept of belief (though this is present in his writings) as to the principle of “a priori synthetic reasoning.” Bubandt doesn’t discuss this, and his expansive theoretical elaborations are almost exclusively devoted to a glossing of Derridean terms, most prominently: aporia, supplementarity, pharmakon, and autoimmunity. (Julia Kristeva’s concept of abjection briefly joins this list.) The Derridean idiom offers itself in radical refutation of structural functionalism. But it is summoned at the point where other, recent analyses of the relationship between witchcraft and modernity (especially those of Peter Geschiere and Jean and John Comaroff) are said to have stalled with the concept of ambivalence: “Instead of modernity being the new social conditions of ambivalence, and witchcraft a partially traditional discourse that allows people to come to terms with these conditions, I suggest the relationship between witchcraft and modernity is the inverse: witchcraft is the ambivalent aporia, the interminable problem; modernity, meanwhile, appears to allow a certain kind of explanation of this aporia and, as such, is mined for answers to it” (Bubandt 2014: 14).

It is certainly true that much structural functionalist analysis of witchcraft attributed to it a pseudo-juridical function, crediting it with explanatory and consoling powers. It is also true that the displacement of witchcraft’s function into the domain of representation, where its capacity to resignify the historical real gives it a renewed relevance in modernizing contexts, does little to mitigate the fundamental presumption of its functionality. But even in those accounts that make the efficacy of witchcraft’s magic contingent on belief (the spell kills because of shared belief), there is often more acknowledgment of doubt and fear than Bubandt implies. Perhaps only Claude Lévi-Strauss (1963) attributed to sorcery the full capacity to ameliorate symbolic and somatic uncertainty. In any case, not all doubt engenders fear, and not all fear is born of doubt. Bubandt is persuasive about the psychic and affective costs of witchcraft, of its incapacity to assuage fear, but his argument about the epistemological basis for witchcraft is not unprecedented.

Already in Witchcraft, oracles and magic among the Azande (1976), Evans-Pritchard acknowledged the doubleness of the epistemological crisis produced when death arrives. That crisis, he shows, arises precisely at the point in which the general principle (termites eat wood and wooden granaries collapse as a result), fails to satisfy the demands for an explanation of the singular event. Not death in general, but this death, of this individual, in this moment and no other. It is not general knowledge (about termites and gravity, about which the Zande are quite confident [516]and correct) but the aspiration to explain the singular that Evans-Pritchard claims witchcraft satisfies. In this sense, it is already conceived as an answer to the aporetic relation between the transcendental and the empirical. For Bubandt, of course, it is not the answer to the aporia but the reproduction and valorization of this aporia that constitutes the essential core of witchcraft in Buli. In this respect he is following fairly directly in the footsteps of James Siegel’s ground-breaking book, Naming the witch (2005), and Bubandt acknowledges as much even as he labors to mark out a territory that is uniquely his own (see, especially, 2014: 53–55). He will provide an ethnography, he says, whereas Siegel has provided a political history. The anxiety of influence is perhaps a problem here, for while it is true that The empty seashell is an ethnography, and an excellent one at that, it is also a political history (and Siegel’s book is, by contrast, not only a political history but an investigation into the entire history of anthropological discourse on witchcraft and sorcery). The differences lie elsewhere.

Toward the end of his book, Bubandt glosses Siegel’s explanation for the explosion of accusatory trials of witches in East Java after the fall of Suharto as a causal one: as the “abrupt effect of an absent state” (2014: 189). He argues further that this was not the case in Buli, where it was the result of a “constant atmosphere of uncertainty that suffused subjectivity and sociality” (189). I trust Bubandt’s ethnography. But, in my reading, Siegel does not posit witchcraft as the effect of the state’s displacement and its vacated capacity to confer recognition on subjects who had come to depend on that very recognition for a source of both subjective and collective identity. To the contrary, he links the crisis of recognition brought on by the collapse of an authoritarian state that had appropriated for itself the function of recognition, to the rise of witchcraft accusations, to efforts to name a witch. Witches and not witchcraft were seen to proliferate after the fall of Suharto. The latter is the condition of possibility of the former though not its cause. Siegel shows how the labor to designate the source of a menace that exceeds the empirical and that fails to explain the singular is inevitably failed. He also shows how, under specific circumstances and in historically specific forms, the accusatory violence aimed at its cancellation could not and did not succeed. The frustration and the frenzy that arose in relation to this failure were born of the misrecognizing belief that eliminating witches could eliminate witchcraft (and death in its singularity). At best, it terminated itself in exhaustion. By eliding the difference between witchcraft and witches, Bubandt himself negates the aporetic space that he has otherwise posited as the core of witchcraft. This occlusion is revealed in the moment that he asks why witchcraft—not accusations or witchcraft events, note—proliferated during the period of his tenure in Buli, despite the fact that it “has no social function, makes no sense, and explains nothing” (14).

Had he made this distinction, Bubandt’s analysis would perhaps not have concluded differently. His claim that witchcraft is not a source of relief from existential doubt would remain, and it is, I believe, an important and correct if not absolutely original claim. But his understanding of the historical events, which are otherwise convincingly described, might be differently inflected. For it remains a question as to whether the people of Buli wanted to get rid of witches while thinking that witchcraft is an irreducible dimension of reality, or whether they wanted to get rid of witchcraft and thereby could think the end of their world as such. Bubandt [517]makes the radical claim that it is the latter that they have repeatedly pursued. A psychoanalytic reader would discern here a death drive in the very place that there is an effort to escape death. Bubandt answers with a description of technological sublimity.

In the last sections of the book, the text accumulates new concepts as quickly as new technologies appear. Not only does Bubandt plumb the idea of the soundscape as the register of social transformation, thereby following John Pemberton’s earlier argument about noise and the dissimulations New Order orderliness (1994), but he also argues that a technological sublime or techgnosis overwhelmed Buli and permitted yet another resignification of gua capacities and a fantasy of escape in the very moment when tradition seemed most doomed. Once again, what promises to liberate the people of Buli is absorbed into the discourse of witchcraft via the discernment of apparent resemblances. Ninety-watt bulbs arrive as means to keep the gua away but also become the figures through which witchcraft itself is metaphorized, the new technology redoubling rather than displacing the hold one (2014: 223), a process not unlike the one I observed in the case of northern Thai mediumship some years ago (Morris 2000).

The detail of a 90-watt bulb is significant. In the end, beyond the theoretical arguments that may or may not persuade readers, the book makes a very powerful claim on the ethnographic imagination, resting as it does on years of repeated and sustained conversations with the residents of Buli. It is characterized by a depth and intimacy that only comes with such enduring engagement, and this is especially evident in the lengthy recollections (and transcriptions) of conversations on witchcraft. It is also evident in Bubandt’s descriptions of the events and forces that have, over the past twenty-five years, transformed this world: from the illnesses, deaths, and marriages (there are very few births or other life-cycle rituals described) of local residents; to the changing place of the national government and its armed forces and police in mediating local disputes during and after Suharto’s reign; to the arrival of mining companies and the influx of cash and mass-produced commodities in local markets. The tight-knit community, long linked to other places through maritime trade, is metamorphosed before our reading eyes, as zinc roofs and tarred roads make their appearance beneath webs of electric wires, while the spasmodic sounds of television sets and broadcast systems fill the air.

Bubandt describes his own changing technological apparatus over the same period (his notebooks are gradually augmented by video cameras), but he makes good use of old-fashioned techniques as well. There are genealogical trees and kinship maps, not to mention diagrams of structural homologies between bodily anatomy and household schema that wouldn’t have been out of place seventy-five years ago. The conventionality of Bubandt’s ethnographic methods and the contemporaneity of his theoretical claims may at first sight appear to be in tension, but I would argue that The empty seashell is an uncommonly fine exemplification of the inherent hospitability that ethnography offers to deconstructionist theory and to the radical empiricism that it demands. As Bubandt’s book shows, this empiricism is nothing more than close reading, and therefore, attentive listening. It is what permits one to disagree with some of his claims, but about his ethnographic accomplishments, there can be no doubt.


Bubandt, Nils. 2014. The empty seashell: Witchcraft and doubt on an Indonesian island. Ithaca, NY: Cornell University Press.

Comaroff, Jean, and John Comaroff. 1993. “Introduction.” In Modernity and its malcontents: Ritual and power in postcolonial Africa, edited by Jean Comaroff and John Comaroff, xi–xxxvii. Chicago: University of Chicago Press.

———. 1999. “Occult economies and the violence of abstraction: Notes from the South African postcolony.” American Ethnologist 26 (2): 279–303.

Evans-Pritchard, E. E. 1976. Witchcraft, oracles and magic among the Azande. Oxford: Oxford University Press.

Foucault, Michel. (2007) 2008. Introduction to Kant’s anthropology. Edited by Robert Nigiro. Translated by Robert Nigiro and Kate Briggs. Los Angeles: Semiotexte.

Geschiere, Peter. 1997. The modernity of witchcraft: Politics and the occult in postcolonial Africa. Charlottesville: University of Virginia Press.

———. 2013. Witchcraft, intimacy and trust: Africa in comparison. Chicago: University of Chicago Press.

Lévi-Strauss, Claude. 1963. “The effectiveness of symbols.” In Structural anthropology, 186–205. New York: Basic Books.

Mauss, Marcel. (1972) 2005. General theory of magic. Translated by Robert Brain. New York: Routledge. Originally published in French in Année Sociologique.

Morris, Rosalind C. 2000. In the place of origins: Modernity and its mediums in Northern Thailand. Durham, NC: Duke University Press.

Pemberton, John. 1994. On the subject of “Java.” Ithaca, NY: Cornell University Press.

Sahlins, Marshall. 1987. Islands of history. London: Tavistock.

Siegel, James. 2005. Naming the witch. Stanford, CA: Stanford University Press.


Rosalind C. Morris
Department of Anthropology
Columbia University
859 Schermerhorn Extension
1200 Amsterdam Avenue
New York, NY 10027-7003