Virtual Christianity in an age of nominalist anthropology

An anthropology, but of what?

In what may be a small irony, the anthropology of Christianity can be seen as the child of the anthropology of Islam. As originally put forward by Joel Robbins in the first essay to call for an anthropology of Christianity, ‘What is a Christian? Notes toward an Anthropology of Christianity’, the anthropology of Islam could serve as an exemplar for what an anthropology of Christianity would look like. As Robbins presents it, the anthropology of Islam is a metaphorical place where ‘people working in different geographic areas publish in the same fora, read one another’s work, recognize the relevance of that work for their own projects, and seek to develop a set of shared questions to be examined comparatively’ (Robbins 2003: 192); this is what Robbins himself wishes to midwife through his call for an anthropology that would be centered on communities that understand themselves as being Christian.

Now, Robbins does not spend too much time in his essay imagining what those shared questions would be (though he does suggest that a few exemplars might be found in the special issue that the essay was an introduction for), but he does make an odd move in suggesting the one question that should not be given too much attention: Robbins suggests that anthropologists of Christianity should not tarry by spending too much time on the issue of what kind of object an ‘anthropology of Christianity’ would be addressing in the first place. Such a move, he says, is unnecessary to the venture.

While it may seem obvious that an anthropology of Christianity would be focused on Christianity, Robbins notes that such a project is theoretically problematic, though he states that this should not delay us. Again referring to the anthropology of Christianity’s Islamic sire, Robbins observes that the initial call for an anthropology of Islam by Abdul Hamid el-Zein ended paradoxically with a denial that ‘Islam’ was any one object; as El-Zein stated in closing his call for an anthropology of Islam,

… neither Islam nor the notion of religion exists as a fixed and autonomous form referring to positive content which can be reduced to universal and unchanging characteristics. Religion becomes an arbitrary category which as a unified and bounded form has no necessary existence. ‘Islam’ as an analytical category dissolves as well. (El-Zein 1977: 252)

Robbins makes note of this because he was concerned that an anthropology of Christianity could be smothered in its crib by what Robbins calls ‘object dissolving critique’, that is, arguments animated by the same logic that El-Zein closes his essay with. As Robbins observes:

[t]here are many kinds of Christianity, and when the number of different kinds is multiplied by the number of different situations in which they have been spread and the number of different cultures to which people have adopted them, it is hard to escape the conclusion that at best we are dealing with Christianities rather than with Christianity, and that at worst these Christianities really have rather little in common with one another. (2003: 193)

This wealth of difference, if taken as a denial of any commonality, would prevent the formation of the kind of comparative community that Robbins longs for. This community was all the more necessary because, as Robbins saw it then, ‘anthropologist of Christianity’ was a ‘stigmatized’ identity (2003: 195) because of the distasteful nature of Christianity to most anthropological sensibilities (also see Harding 1991). Already spurned as crypto-Christians, this possibility of difference could keep anthropologists of self-professed Christian populations from cohering around Christianity as a (phantasmatic) common issue.

However, Robbins tells us, the anthropology of Islam shows that some common foundation need not be laid down before an anthropology of Christianity can begin construction:

What the history of the anthropology of Islam teaches us is that the intellectual obstacles that confront any effort to establish the anthropology of a world religion can be overcome. What is more, they can be overcome even in the absence of fully developed theoretical arguments for why disciplinary strictures against any devaluation of local variation should be relaxed in order to allow comparison to take place in these cases. (2003: 194)

As an instituting gesture, there is something to this openness that should be respected, concerned with what can be crafted later on rather than with policing at the outset. However, what I would like to take the time to point out is that this odd elision is not particular to Robbins. I argue that there is a recurrent refusal or inability to define exactly what Christianity is. ¹ This is in part because of the diversity of ‘Christianities’ out there, but also, I claim, because of the unconscious influence of a kind of nominalist ontology in contemporary anthropology; and I also argue that, oddly enough, it is the intellectual tools being turned to in contemporary nominalist anthropology that will get us back to a realism that can allow us to talk meaningfully about the anthropology of Christianity as having a single object, without doing violence to the wealth of ‘Christianities’ out there as well. I argue that we should take the object of anthropology to be Christianity in its virtual form, as a multiplicity which is predicated on and produces difference, even as that difference is still comprehendible.

The best way to make this case, though, is to start by showing that Robbins’s demurral is met by other instances of declining to give an object for the anthropology of Christianity, or in some cases by an attempt that falls apart when looked at in broad daylight. With that I’d like to turn to Fenalla Cannell’s introductory essay to The Anthropology of Christianity (2006), the other candidate for an instituting call for an anthropology of Christianity. Like Robbins, Cannell too sees Christianity as suffering under a kind of interdict: Christianity is ‘the repressed’ of anthropology, where the religion acts as an ‘anxiety’-provoking object whose potential ethnographers are stigmatized by a suspected belonging, or at least a susceptibility to conversion (Cannell 2006: 4). The fraught relation ‘between anthropology and Christianity’ is something that must be worked through, she suggests, and in the introduction to the volume Cannell presents herself as engaging in that very therapeutic work.

And it is here, at the very cusp of her working through the fraught relationship between Christianity and anthropology, that something curious begins: an elision as to what Christianity is, which seems to be fittingly the kind of symptomatic omission that one might expect when attempting to articulate the repressed. It is not that she does not try. ‘In considering this question’, Cannell tells us ‘it is necessary to reach some provisional working definition of the term Christianity itself. This is something more difficult than it might first appear’ (2006: 5). The difficulty, Cannell tells us, lies is the multifold nature of Christianity that Robbins also observed. Cannell notes that there are ‘diverse ways’ to ‘balance these models of what Christianity does against the specificities of local interpretations’ (2006: 6), as exhibited by the contributors to her volume; Cannell’s own approach is to deny that Christianity is an ‘arbitrary construct’, but merely insist that it is a ‘historically complex one’ (2006: 6). ‘It is not impossible to speak meaningfully about Christianity, but it is important to be as specific as possible about what kind of Christianity one means’ (2006: 6). This is because ‘Christianity is built on a paradox’ with a ‘central doctrine’ that is a vision of the incarnation and resurrection that points simultaneously to both the spirit and flesh, leading to a core ambivalence (2006: 7). Cannell suggests that ‘a recognition of the centrally paradoxical nature of Christian teaching allows us to move some ways further in conceptualizing… local encounters with missionary Christianities’, pointing to how ‘the unorthodox position remains hanging in air’ (2006: 7).

Now, as we will see near the end of this essay, this observation that other potential readings of Christian practice and thought tend to linger on, even in the absence of institutional endorsement, has a great deal of merit, but it is not by any means a ‘definition’ of what the object of an anthropology of Christianity would be, and certainly not one that can assuage another set of anxieties – anxieties about whether Christianity is an entity that can hang together enough to be an object of study. It at best tells us one of the features of what we are addressing, but this paradoxically is a predicate that cannot be given to ‘Christianity’ alone nor, would one imagine it, can it be the only predicate that Christianity bears. We also cannot hang a definition of Christianity on a ‘central doctrine’, for to do so is to take the side of a certain orthodoxy, since the position that Cannell takes up, though recurrent, is not uniform throughout Christian history. Christianity may be historical, it may be paradoxical, but these terms do not tell us what is the nature of the object we are interrogating.

What is interesting is that Cannell and Robbins are not alone in this inability or demurral to name an object; numerous subsequent authors addressing the possibility of an anthropology of Christianity have also either declined or been unable to give us a picture of what kind of object Christianity is, at least in the eyes of an anthropology orientated around it. An instance of the failure to specify what the object of an anthropology of Christianity might be seen in Gil Anidjar’s ‘The Idea of an Anthropology of Christianity’ (2009). In a sustained encounter with Talal Asad, Anidjar, who works primarily in literature and religion, gives another call for the creation of an anthropology of Christianity, drawing on Asad’s work on the anthropology of Islam to create a concept that is adequate to dealing with Christian diversity. Andijar, however, does not himself produce such a concept, being content to observe that it must be expansive (covering doctrine, spaces of power, and peoples). In any event, Andijar insists that a concept of Christianity by nature must be a polemical concept, having only enough consistency to allow us to ask the better question of what it is that Christianity does politically, which is his true concern. Again, we end without a concept.

We can say something similar about Tomlinson and Engelke’s observation (2006) that Christianity frequently has a concern with meaning; this claim, while backed by the ethnographic material presented in the accompanying edited volume (Engelke and Tomlinson 2006), is at best a claim about probability, about the likely or common behavior of an object, rather than its nature, especially since, as these ethnographic sketches testify, meaning can fail at times but Christianity still endures. Indeed, Tomlinson and Engelke, in asserting that Christianity ‘is not a stable, singular object’ (2006: 19), seem to take away as much as they give when it comes to identifying exactly what the object of an anthropology of Christianity is.

Alongside Robbin’s and Cannell’s demurral, and Tomlinson and Engelke’s equivocation, stands Michael Scott’s (2005) refusal. Here Scott attempts to articulate what should be the program of an anthropology of Christianity, charting a middle-ground between what he sees as an essentialism regarding what ‘Christianity’ might be taken to be on one hand, and Robbins’s ‘object-dissolving-critique’ on the other. He does this by allowing for emergent properties as Christianity (again, whatever that is) interacts with what he calls ‘indigenous religions’, and the latter ‘appropriate portions of Christianity expediently and transiently without far-reaching consequences for their indigenous models of reality and moral order’ (2005: 117). Because aspects of Christianity can be selectively peeled off, especially as Christianity is never encountered whole cloth, Scott sees Christianity as being constituted by nothing less than diverse ‘logical trajectories’ resulting from different responses to a ‘problematic’ – a problematic that not all Christians, Scott wishes to emphasize, feel compelled to take up in the first place (2005: 118, 119). This seems to be again an argument that the anthropology of Christianity has no real object, that Christianity in its end is, contra Joel Robbins, a thing of ‘shreds and patches usable in a piecemeal fashion’ (Scott 2005: 104; cf. Robbins 2004: 3; also see Lowie 1920: 441); Scott says otherwise, but he provides no conceptual frame to take up Christianity in his way without it becoming purely ephemeral. We will see later on, as with Cannell, there is much to recommend in Scott’s project, though the choice between multiplicity and having a concept of Christianity as an analytic object is a false one.

While I am at it, I should observe that the first set of review articles regarding the anthropology of Christianity did no better in giving a definition of Christianity (Bialecki et al. 2008, Lampe 2010), content as they were merely to document refrains in the efflorescence of literature that appeared over the last decade. The Bialecki et al. piece never even broaches the question in the first place, and the Lampe piece ends with the refrain that in Christianity ‘there is no clear and widely shared explanation of what constitutes Christianity or how to understand it’, a conscious echoing of El-Zein’s position regarding Islam (Lampe 2010: 82). Again, absences or elision, but no definition of an object.

One last possibility for grasping an anthropology of Christianity might be the hope that while one has not been explicitly stated, an unspoken consensus on this question might have been reached by the various articles and ethnographic monographs that could be seen as constituting this project. As practiced, the anthropology of Christianity is suffused by two concerns: one, roughly, that might be thought of as language ideological (see Bialecki and Hoenes del Pinal 2011) and another that might be classed as Foucauldian-informed interest in Christian practice as a form of subjectification (Bialecki n.d.). Indeed, it seems that Foucault is over-represented in the anthropology of Christianity. However, given the portability of these models to other concerns, it seems a stretch to say that these concerns tell us something specific about what the object of an anthropology of Christianity is. Foucauldian and semiotic analysis are not by any means particular to Christianity (whatever that is) or Christian objects (whatever they may be) alone.

This recurrent demurral/refusal/failure is important because the anthropology of Christianity has not had the same good fortune as its father. In some ways, the anthropology of Christianity’s luck has been the inverse of the anthropology of Islam, which began its life by denying its own possibility, only to have that impossibility brushed aside as unproblematic. In contrast, the anthropology of Christianity is assailed by doubts regarding its consistency as an ethnographic object. While acknowledging the worth of the ethnography that has come out of it, Chris Hann (2007) has questioned the value of an anthropology of Christianity as an overarching project, saying that it obscures comparative thought, both within various types of Christianity, but also between Christianity and other forms of religiosity. ‘Why’, Hann asks, ‘demarcate one world religion as a suitable domain for comparison’ (2007: 46)? Hann also faults the anthropology of Christianity for an idealist bent, though he does note that it cannot be described as having a ‘crudely predictive idealism’ (2007: 405). Interestingly enough, he echoes Scott in suggesting that rather than focus on Christianity, we focus instead on problems as the ‘key entities’ (2007: 406), something that we will return to later; but what we should focus on now is Hann’s rejection of tradition as a ‘suitable domain’, indicating that in Hann’s eyes there is no integrity, no real object, no phenomenon in itself that needs to be charted. More recently, John Comaroff has taken Hann’s objections and repackaged them as an accusation that the anthropology of Christianity is ‘reductionistic, incoherent in defining its subject matter, contradictory in the claims it makes about that subject matter, and unreflective in its idealism’ (2010: 529). For Comaroff, it is a retreat to the culture concept by way of religion conceived of as ‘immaterial’ and ‘ahistorical’, and is part of habits that ‘give anthropology a bad name’, a false solution to a larger disciplinary crisis (2010: 529).

There have been other criticisms as well. From its first moments, there has been a concern that the anthropology of Christianity has been too narrow in its purview; Brian Howell worried that studies of Pentecostalism had overshadowed other forms of Christianity (Howell 2003: 235), a complaint that Hann has joined in his repeated observations that Orthodox Christianity has made scant appearance in the literature (2007; Hann and Goltz 2010). Taken altogether, all this forms quite a bill of particulars, and a dizzying one, if one includes the qualifications, hedging, and elisions made by some of the proponents of an anthropology of Christianity themselves. The anthropology of Christianity takes as its object something with no fixity and yet is in need of being rescued from essentialism. It must be spoken about at once in careful historical terms and yet is ahistorical, immaterial and suffused by idealism yet overrun by Foucauldian subjectification and regimes of power; finally, it is too ethnographically diverse yet too centered on a single exemplar.

This is traditionally the moment when we would start asking which of these accounts and critiques have merit, and which do not; in short, separating goats from sheep. I would like to ask a different question, though. What if we were to start out with the postulate that, law of the excluded middle be damned, everyone was right, even when their descriptions of the still mysterious object of the anthropology of Christianity seemingly contradict each other? I would like to do this by taking what seems to be a sidestep into the question of anthropological ontology, a sidestep that may seem unrelated but will in the end allow us to answer the question that has stood at the center of our discussion so far: what is the object of the anthropology of Christianity?

Nominalist anthropology and the exemplary assemblage

Now, ontology is an issue that has received a great deal of attention recently in the anthropological literature (Henare et al. 2007; Keane 2009; Magnus 2010; Laidlaw 2012; Pedersen 2012) but not always in the sense that I wish to invoke it. Today in socio-cultural anthropology, discussions of ontology are usually at the level of mapping ethno-ontologies, of articulating alternative, non-western modes of imaging being – the most striking instance of this being the work on shamanic perspectivalism (Viveiros de Castro 1998; Willerslev 2007). While this has been an exciting field which promises to open up new lines of inquiry not only for anthropology generally but also for the anthropology of Christianity (Vilaca 2011), it is not what I have in mind. What I am referring to now is not anthropological articulations of others’ imaginings of what sorts of beings count as entities that exist and what their properties are; rather, I am talking about anthropology’s own ontology, what the discipline considers to be ‘real’ and, conversely, what it considers not to be real as well.

My claim here is that, as of late, anthropology has become increasingly nominalist in its ontology. By nominalism I mean the ontological position that rejects the existence of abstract objects, of universals, or of both of them altogether (see Rodriguez-Pereyra 2008); it is in short a claim that only individuals exist, and is a form of thought usually juxtaposed against realism, a contrasting stance towards ontology which grants that in some way universals, abstractions, and categories can be said to have an existence beyond their mere crafting and use by human beings in ways that have performative social effects. A full charting of the reasoning and evidence behind this claim of a contemporary nominalist anthropology, and an enumeration of the inevitable exceptions to it, would require at least an article-length exegesis in and of itself, so only a quick sketch is really possible here. Nominalist strands of thinking in the discipline go back as far as Edward Sapir’s sparring with Alfred Kroeber, a debate that turned over whether the ‘superorganic’ could be considered to be an entity in its own right, or merely a mirage caused by the conglomeration of linguistic, cognitive, and psychological properties locatable at the level of the individuals that comprise the interrogated category of ‘the social’ (Kroeber 1917; Sapir 1917). However, during most of the 20th century, the direct or indirect influence of Durkheim’s and Mauss’s social theory on anglophone social science, and the particular influence of the concept of culture on American anthropology, kept the discipline from slipping into the nominalism that constitutes Anglo-American ‘common sense’ (see Dumont 1992: 1–2; Rabinow 1988). Accounts differ as to what the engine of a shift towards nominalism was. In one account the nominalist turn was another pendulum swing in anthropological theory, a counter-reaction to universalist and essentialist trends in mid-century social theories (Zenner 1994). Other accounts see it as the outgrowth of the conceptualizing of anthropological abstractions and universals themselves. In this narrative the shift was a result of a critical turn, not unlike an immunological disorder, in which anthropological theory was directed against its own reasoning apparatus and categories that were previously taken to be the social science analog of natural kinds were placed into question, now suspected of being artifacts of, if not the discipline itself, then at least of broader social-science engineering (Rabinow 1988). Finally, there is the narrative in which this shift is at least partially symptomatic of large-scale transformations in the political economy of the West (Kapferer 2005). Either way, the result of this turn has been a suspicion of the totalities that compromised the units of comparative analysis, and indeed of recognizability, of an earlier anthropology, including the social, the notion of the ethnographic site, and even the idea of culture (though the latter concept often has some ideational violence visited upon it as part of the rejection process: see Brightman 1995). As put by Matti Bunzl, contemporary anthropology is infused with ‘a desire to challenge all essentialisms and question all generalizations; the ethnographies of today are often simultaneous exercises in total deconstruction and absolute empirical specificity’ (2008: 57). Back to (just and only) things in themselves, to misuse Husserl.

This shift has not been all at once, and one can intuit in the literature gaps between the programmatic abandonment tout court of outward affiliation with larger classificatory totalities that run against a nominalist ontology, and the still under process creation of new modes of forming ethnography that escape nominalism in practice as well as in thought; this might account for the increasingly bricolage-like use of theory in anthropology today as conceptual stop-gaps (Knauft 2006). However, despite the lag between disaffiliation and actual transformations in practice, and the degrading effect that nominalism has on theory itself, one can see in anthropological theory some effects of nominalism. One long-running sign is the growing interest, running from the mid-80s to the present, in theoretical tools whose vigor is dependent on their being deployed in localized, circumscribed manners; a leading indicator here was the shift from modes such as Marxian analyses, predicated on large-scale abstractions such as labor and ideology, to that of a Foucauldian interest in institutional and quasi-institutional arrays stabilized by tactical deployed practices of knowledge and power (even if, in the hands of some of the less agile practitioners, these analytic Foucauldian categories were unthinkingly endowed with near-metaphysical significance and scope).² The interest in ‘person-centered’ ethnographies in psychological anthropology, and anthropological accounts of individual biographies in socio-cultural anthropology, is another nominalist tell. A more recent index of the shift has been the adoption of Latour-derived actor-network theory outside the confines of science-studies, where the analytic was originally crafted; predicated on enchained and individual actants, it is a view of causality perfectly suited for those skeptical of larger abstractions and universals.

But perhaps the most developed nominalist mode of organizing and accounting for ethnographic data is the idea of the assemblage. Unlike actor-network theory, ‘a conceptual apparatus somewhat more domesticated to classical theory’, assemblage theory is predicated on material and temporal flux (Marcus and Saka 2006: 102). The assemblage, we are told in the introduction to Ong and Collier’s sizable and influential edited volume Global Assemblages: Technology, Politics, and Ethics as Anthropological Problems, is characterized by the ‘heterogeneous, contingent, unstable, partial, and situated’ (Collier and Ong 2004: 12). As observed by Marcus and Saka, the allure of the term assemblage is the way in which it brings together the heterogeneous and the ephemeral (Marcus and Saka 2006: 102). The appeal of this to a nominalist anthropology is obvious. It is a denial of not just essentialism but of even any kind of long enduring solidity. Everything must be taken as it is, a chance and passing agglomeration, resistant to being captured by any of the universals or abstractions that is an anathema to contemporary anthropological thought.

While the Ong and Collier piece might seem to be a natural place to look at how the assemblage functions, I’d like to turn instead to the works of Jarrett Zigon instead (2010, 2011a, 2011b). There are two reasons to look to Zigon. First, as a theoretically accomplished anthropologist who has recently placed the concept of the assemblage in the forefront of his thought, he shows us what assemblages look like in the era of nominalist anthropology. The second reason is that, as opposed to other anthropologists of his generation who have also invoked the assemblage in the study of religion (see e.g. Rudnyckyj 2010), Zigon has relied upon assemblage theory in an ethnographic scenario where it might make sense to describe oneself as, if not engaging in, then at least being in dialogue with the anthropology of Christianity, and yet he has declined to do so; this despite the fact that his citation pattern shows that he must have some awareness of the sub-field (writing often on the subject of morality, he has been deeply engaged, though not always uncritically, with the works of Joel Robbins [see Zigon 2009a, 2009b]). A discussion of Zigon’s use of assemblage theory will show how a certain kind of anthropological nominalism is interfering with a definition of Christianity necessary to make the project of an anthropology of Christianity more cogent; I also claim that an analysis of what assemblage theory is (and also is not) for Zigon will show us a way to work back to a conception of Christianity.

In his work under consideration here, Zigon’s ethnographic object is Russian Orthodox church-sponsored attempts to rehabilitate heroin addicts, primarily as encountered at The Mill, a residency program situated in the rural periphery of St Petersburg. Zigon’s project is not just how it is that these heroin users win a coveted rare spot in the drug treatment facility, and then struggle to comply with its disciplines, all in order to at least partially escape the pull of addiction and live what they call a ‘normal life’; Zigon also wishes to make a contribution to the theory of morality and ethics. To do this, in his ethnographic accounts he is sensitive to the diverse sources of the discourses and practices that constitute The Mill, which are taken not just from various historical strata of orthodoxy but also from what Zigon portrays as more unlikely sources, such as the Soviet language of the ‘new man’, neoliberalism, and secular western therapeutic techniques. Zigon is also attentive to the sensibilities that the patients themselves bring to the facility, as well as to their pre-existing senses of habitus, and to the new bodily dispositions that disciplinary techniques are attempting to inculcate.

Zigon attends to these diverse discourses, practices, and dispositions to make the point that ethical and moral systems are not a totality; stitched together, they form what he calls a moral and ethical assemblage, a nonce-structure that denies any whole. Extrapolating globally from his findings, he claims that

the theory I outline here denies the common philosophical and social scientific assumption that a moral totality – in either universalist or relativist terms – exists anywhere in the world, and rather sees all particular social contexts defined not by one morality and its ethics, but rather by a unique local moral and ethical assemblage constituted by the various aspects. (Zigon 2010: 5)

Further, each of the discourses and practices that constitutes the assembly is never encountered in complete form, even as a subsumed portion of the assemblage. Only ‘various aspects’ of the constituting discourses and practices are encountered as they are activated in moments of what Zigon has earlier called ‘moral breakdown and ethical demand’. Totality, if it is experienced at all, is an illusion brought about as an aftereffect of an enunciation or action that follows a moment of breakdown, where it seems that these disparate strands have come together to produce the emergent, and now self-definitional, act.

In this focus on foreshortened temporalities and distinct composite elements, we can see how this is line with the idea of the assemblage as found in contemporary anthropology. Further, we can see in the logic of the assemblage some of the reasons why Zigon would decline to see his project as part of an anthropology of Christianity, even if his object of study was a Russian Orthodox religious rehabilitation center. First, anything like Christianity is unlikely to be encountered in a way where it is not stapled together with a host of other autonomous entities. Therefore, why let only one component part play a determining role, or even an autonomous one, in the conversation? And in fact, in a note that is only a slight tangent from our point, Zigon argues that despite the Russian Orthodox intentions and control over the treatment center, it is a neoliberal sensibility and telos that ends up shaping the forms of subjectivity that the center inculcates; orthodoxy has no force or directionality of its own that is orthogonal to the ‘stronger discourse’ of neoliberalism.

Second, we can see that since the whole is an illusion and is never encountered, it appears that the constituent part of the assemblage, the part that we are tempted to call ‘Christian’, is itself an assemblage, historically formed, heterogeneous and unstable. If this is the judgment of nominalism, and if nominalism is the sensibility of the age, we can see why the inhibition to invoke universals or abstractions can stand as a block to giving Christianity a form wide enough to encompass all the divergent manifestations that comprise it. We are left in a position very close to Michael Scott’s, as discussed in the section on clamant conceptions of an anthropology of Christianity: bits of Christian material shorn off, repurposed and denuded of whatever energies or directionality they might (theoretically) have had before their disaggregation.

Virtual realism

I want to be careful about stating what I’m claiming here, of course. I am not stating that everyone writing on the anthropology of Christianity is animated by a nominalist instinct, nor that what they have put forward is not a useful contribution to the field; rather, I’m merely stating that a general tendency to be skeptical regarding abstractions and universals, along with the multifarious nature of Christianity itself, is working in such a way to inhibit articulating what it is that the putative object of an anthropology of Christianity could be. To substantiate this, I’ve used the anthropological apparatus of ‘the assemblage’ as an exemplar of the kind of reasoning that this tendency to be skeptical of abstract and universals engenders, and shown in Zigon’s work how the logic of the assemblage, as it is used in contemporary anthropology, mitigates against giving Christianity any coherence as an object. What is not being claimed is that either the anthropology of Christianity on one hand, or nominalist anthropology or the assemblage on the other, is intellectually deficient or ethnographically manqué, merely that they are both, vis-à-vis each other, inimical concepts. They both seem to have value, at least in the eyes of contemporary anthropology, in as much as they are ‘growth stocks’, recent anthropological developments that show continued thriving – only not together (as it may seem at this point in the discussion), at least not if one wishes to think rigorously.

However, I think that there is an element in the assemblage that itself forms assemblage theory that can allow us to think through this impasse. Let me illustrate this by returning to Zigon. As Zigon notes (2011a: 16), the concept of assemblage has a history that transcends that of Ong and Collier; the assemblage was crafted by Gilles Deleuze and Felix Guattari in their book A Thousand Plateaus (1987). However, that reference to the idea’s native space is the only moment that Deleuze and Guattari, singularly or together, appear in Zigon’s book. Zigon’s assemblage is what Ong and Collier have made of it, heterogeneous and ephemeral, but not necessarily what Deleuze has made of it. Working back to Ong and Collier themselves, we see that in the introduction they refer us to the same work, supplementing it with a reference to Deleuze’s Foucault (1988); and yet Deleuze appears as more of a name invoked than as a metonymic indication of the specifics of a technical means of thinking something through. Indeed, this is part of a larger pattern with discussions of the assemblage; as Marcus and Saka have noted, ‘none of the derivations of assemblage from Deleuze and Guattari of which we are aware of is based on… a technical and formal analysis of how this concept functions in their writing’ (2006: 103). It is evocative power of the assemblage, and of Deleuze and Guattari’s writing in general, which seems to be appealing, rather than any formal engagement with the specificities of their thought.

Of course, I hesitate a bit to bring up Deleuze. While, as we have seen, Deleuze has been granted space in anthropology for a while now, his recent enumeration as one of the ‘four main themes of particular significance’ in the American Anthropologist sociocultural anthropology year in review list (Hamilton and Placas 2011; see also Biehl and Locke 2010) makes the coin embossed with the name ‘Deleuze’ seem so common that one cannot but help suspect that its metal is being adulterated. But then this is the point. The solution to this is not to shy away, but to take up Deleuze in the ‘technical and formal’ manner that Marcus and Saka suggest. When one does this, what is striking is that despite Deleuze’s interest in the heterogeniety, ephemerality, and becoming that has made him such an object of interest to contemporary nominalist anthropology, Deleuze himself is not a nominalist.

First, as Manuel DeLanda has observed, Deleuze is comfortable talking about kinds of totalities in a way that nominalist anthropology would balk at; there are moments in Deleuze’s works where he talks about ‘“society as a whole” and specifically, of a virtual multiplicity of society’ (DeLanda 2002: 195; see also Deleuze 1994: 186, Deleuze and Guattari 1977, 1987). This is of course the kind of overarching entity that nominalism is trying to escape from. But this use of the word ‘virtual’ brings us to an important qualifier. Because if we want to say that Deleuze is a realist instead of a nominalist, we have to grant that he is not a realist in the same way that, say, Plato or the mythical mid-20th-century cultural anthropologist might be thought of as a realist. As John Rajchman has said, ‘[w]e might call Deleuze a “realist” of a peculiar sort – a realist about virtualities that can’t be forecast or foreseen, that have another relation to thought’ (2000: 62).

It appears, then, that we have to have an understanding of what the virtual is if we are to grasp what might be the gap between Deleuze and nominalist anthropology, and what difference a more technically informed reading of Deleuze might bring to the problem of an anthropology of Christianity. Like a single thread, the idea of the virtual runs either explicitly or implicitly through most of his authored or co-authored works, an outstanding consistency for an author who was more than willing to use different language in different works to do what appears to be the same conceptual labor. The concept of the virtual was originally taken from Bergson, where it stood as a sort of metaphysical memory, an immaterial, condensed compendium of the past that existed coeval with the present (and in some ways encompassed it). As reworked by Deleuze, the concept of the virtual continued to have an air of this (Deleuze and Parnet 2007: 148–52), but it serves more as a way of speaking about an unquantifiable field of generative potential in being and thought, a potential intelligible yet specifically undeterminable in advance of development, a potential that is always threatening to run off at times in different and disparate directions, a potential which serves to constantly bring new ‘actual’ entities into being. Described by Deleuze as ‘real without being actual, ideal without being abstract’ (Deleuze and Guattari 1996: 156), it is a means of granting a sort of ontological status of real to both the this potentia and the objects they engender, and to demarcate this potentia as having its own characteristic apart from those of the ‘actual’ world.

While the virtual has been taken up as a way of speaking about material and ideational processes in general, based on Deleuze’s endorsement of a univocal ontology of pure difference it is clear that, however else the Deleuzian virtual can be discussed, Deleuze himself also intended it to be a means of discussing thought; this can be seen most clearly in Difference and Repetition (1994), which contains Deleuze’s most detailed working through of what the virtual is, and the relation that it has with the actuality that the virtual engenders. A review of this gives us ways to think about kind, nature, and degrees of potentiality inherent in the concept of the virtual, yet without thereby framing it in a deterministic manner.

The virtual in Deleuze is not fixed; as he puts it, ‘[i]deas are by no means essences’ (1994: 187). This is important because essentializing the idea would mean thinking in terms of the virtual as a model for the actual, privileging representation and typological thought in a way that would undermine his insistence on the ontological primacy of difference. What Deleuze is trying to get away from is any moment where the virtual idea, historically produced and mutable, might be mistaken for the Platonic idea (or the caricature of the idea of culture), transcendent and ahistorical. Deleuze escapes this in four different ways.

First, for Deleuze, the virtual idea is not unitary; as he expressed it, ‘[a]n Idea is an n-dimensional, continuous, defined multiplicity’ (1994: 182), meaning that it is composed of multiple elements which differ qualitatively as opposed to quantitatively, with elements of the idea having their own separate axis on which they can act even as they tend to fuse into concepts that are contiguous in intellectual space. In short, aspects of the idea, even while conjoined, are independent of each other, and the difference between elements must be thought of qualitatively. This sense of play and fusion suggests some kind of transformation within the virtual, which brings us to the second way in which the idea escapes fixture. For Deleuze, the virtual is a process, or at the very least always in motion, meaning that the virtual idea is purposefully unfixed while at the same time not indeterminate as it brings the actual into being. This can be seen in a competing definition of the idea (Deleuze, whatever else he was, was not uncharitable in producing definitions, often giving multiple ones for the same concept). In this second definition, he painted the idea as ‘a multiplicity constituted of differential elements, differential relations between those elements, and singularities corresponding to those relations’ (Deleuze 1994: 278). In this sequence we can sketch out the processual unfolding of the idea, the series through which open aspects of the idea are assembled, through which these aspects are put in differential relation to each other, and finally through which they produce singularities that can be seen as being an expression of those differential relations. Aspects are initially empty; purely virtual, indeterminate, they are given determination by their relation to other equally ‘empty’ aspects. For Deleuze, these relations are differential relation, differential in the sense of the term as it existed in infinitesimal calculus. The term differential is important here, as it points to the variable yet reciprocal relations that aspects of the virtual idea have with one another and through which they are given determination, allowing for at times incremental variance in play in the relations that constitute an idea, while at other times also allowing for vast difference. In other words, the difference between aspects of the idea need not be utter difference of negation found in either Hegel or structuralism. Rather, there can be gradations of difference between aspects of the idea. This concept of gradations of difference means that the virtual idea is in a sense topological, capable of undergoing a great deal of torsion while still expressing the same set of relations, much the way that a coffee cup and a donut are topologically the same shape despite obvious extensive differences between them.

The way in which the concept of ideas undergoing a degree of torsion, of a mutability of the idea as it simultaneously corresponds to what seems to be different forms, allows us to understand a third manner in which the idea is not an essence. As progressive determinations of these differential relations occur, they indicate the singularities which will correspond to them, in the same way that an equation corresponds to its solution. It is for this reason that Deleuze refers to the virtual as problem, in that the various indeterminate virtual aspects are given a specific determinate content as they are individualized from the ephemeral virtual to the actual, the realm where objects are capable of being thought of as discrete, extensive, and quantifiable. It is because of this transition from the virtual to the actual, which with each repetition produces something new due to the different circumstances that the operation is carried out in, that Deleuze says that virtual ideas as problems ‘belong on the side of events, affections, or accidents rather than on that of theorematic essences’ (1994: 187).

Finally, there is another way in which the virtual idea differs from the Platonic idea. Much like the actual has its history, virtual ideas have history as well, being both ‘made and unmade’ and capable of being put in different relations with other virtual ideas over time (1994: 187). To be more exact, though, rather than having a history, the virtual, in that it points to the creation of something new, is a break from history, and rather than saying that it is in time, the virtual occurs in the ‘dead time’ that belongs neither to eternity nor to time. This temporal oddness does not mean that the virtual is not informed in a way by the actual that it forms. Much as the actual is the result of an individualization over time, the virtual is itself attached to, addressing, and predicated upon actual entities, creating a resonance between the actual and virtual as they go through their vicissitudes. It is affects and percepts, intensities that are nothing more than actuality experienced not as extensivities but as force (see Deleuze 1988b: 91–2), that trigger the further actualizations of the virtual, which along with memory affect the mode through which a narrow swath of the virtual idea will be actualized yet again. Viewed synoptically, then, the actual and the virtual run alongside each other as a double series that is separate and yet resonant, ‘echoing each other without resembling each other’ (1994: 189). The actual, however, constantly obscures the virtual, with the extensive produce of virtual processes in effect covering up the other divergent lines inherent in the virtual that happened not to be actualized; and while there is always a way to ‘ascend’ back to the virtual from the actual, a process that Deleuze calls counter-effectuating (Deleuze and Guattari 1996: 157, 159), Deleuze speaks of seeing the trace of the virtual in what has been actualized as if it were no easy task.

Diagramming the assemblage

Although this presentation has to, by necessity, boil it off to the point of being flavorless, this is a functioning enough presentation of Deleuze’s virtual to meet our proximate needs: an understanding of the difference between anthropological assemblage theory and Deleuzian assemblage theory. This, though, is itself just a way-station to grasping what might be the object of an anthropology of Christianity. Let us start then with a first question: What does the virtual mean for those who would be using the assemblage as an analytic tool in the anthropology of religion? It does nothing to impeach their ethnographic description of the state of affairs on the ground, but it suggests that they have not given Christianity the autonomy it deserves.

In his book-length exploration of Deleuzian assemblage theory, DeLanda makes two observations that are relevant for our discussion here. The first is that assemblages can be, and often are, component parts in other assemblages (DeLanda 2006: 21), meaning that, like a matryoshka doll, there can be endless encompassing assemblages, each predicated on the smaller constitute assemblages that they subsume. This is interesting in light of the second aspect of assemblage theory that I wish to foreground: assemblages are characterized by more than their heterogeneous mixture but come in what we might call ‘phylums’. DeLanda has observed that while the assemblage is in a way nominalist in that we cannot speak about ‘species’ of assemblages but only individual assemblages, that does not mean that there is not a certain anatomy to assemblages writ large; assemblages are marked by certain ‘topological invariants’, spaces of possibility that limit, but do not exhaust, the potentiality of assemblages:

The notion of the structure of a space of possibilities is crucial in assemblage theory given that, unlike properties, the capacities of an assemblage are not given, that is, they are merely possible when not exercised. But the set of possible capacities of an assemblage is not amorphous, however open-ended it may be, since different assemblages exhibit different sets of capacities. (DeLanda 2006: 29)

DeLanda’s name for this topology that ‘structures the space of possibilities associated with the assemblage’ is the diagram (DeLanda 2006: 30), a name that he takes from Deleuze’s short book on Foucault (Deleuze 1988a), the same book that inspired Collier and Ong. DeLanda wishes to underscore the fact that each assemblage, while individual, is associated with a diagram that is by no means necessarily unique to it:

Thus, while persons, communities, organizations, cities and nation-states are all individual singularities, each of these entities would also be associated with a space of possibilities characterized by its dimensions, representing its degrees of freedom … [i]n other words, each of these social assemblages would possess its own diagram. (DeLanda 2006: 30)

It should be no surprise to hear that the diagram is another way to speak about the virtual, having an autonomy from the actual (or as DeLanda calls it, the diagram is ‘mechanistic-independent’; DeLanda 2006: 31). This, combined with the earlier observation that assemblages can have other assemblages as constituent parts, has effects for how assemblages operate – and specifically for how we must think about assemblages that constitute Christian material. If it holds that the assemblage has a virtual image in the sense of the diagram of a virtual idea, then it holds too that the constituent parts of the assemblage, as potential assemblages in their own right, have their own autonomous virtuality as well, and that whatever work the larger subsuming assemblage’s virtuality or diagram does, the virtual image of the subsumed assemblages does the same for the constituent assemblages that comprise its parts.

What does this mean for our discussion of Zigon? Not to put too fine a point on it, if Christianity is a thing of shreds and patches, then at least in the eyes of a reading of assemblage theory that takes the conception of the virtual seriously, then it is an assemblage, and if it is an assemblage it is not just a thing of shreds and patches but itself a recursively embedded entity formed by a virtual image that is a manifold, an abstract topology, with various axes granting it various degrees of freedom and certain capacities – axes that are independent of each other even as it is that conjunction of various singular axes that gives the specific characteristics of various Christianities as they are actualized. In short, if one wishes to include Christian elements in an assemblage, it seems that rather than allowing for an escape from a realist account of Christianity, it demands it; and that rather than having Christianity vanish in the expression of an assemblage, Christianity rather insists on the level of the virtual, and therefore as a virtuality it has the capacity to always have the larger assemblage actualize its diagram in a new way in which the Christian element of the admixture might spin the assemblage’s expression in new directions. And since the ‘event’ of the virtual is always occurring, this Christian torque is always in potential ready to leave its mark. This means that Zigon is right to focus on the nonce aspect of the assemblage as he does in his description of how it effects moral and ethical action, but wrong to see it as by force always allowing the dead hand of neoliberalism to win every throw of the dice. Like all virtualities, Christianity insists, and insists in determined, but undeterminable, ways.

Virtual Christianity in the age of nominal anthropology

The idea of a virtual Christianity has an importance beyond our discussion of Zigon, though; in fact, it brings us back to the issue that started this article. I would like to suggest that this acknowledgment of a virtual aspect of various actualized Christianities gives us a task for those interested in an anthropology of Christianity, and it does so in a way that not only affects those interested in assemblage theory with religious elements, but also those who have struggled with what the object, goal, or warrant of the anthropology of Christianity might be. Following the logic of the discussion above, it seems to me that the proximate goal of an anthropology of Christianity is to engage in the work of counter-effectuating virtual Christianity, with the further goal of grasping how virtual Christianity’s nature as a virtual multiplicity allows it to be actualized in differing manners at differing moments, becoming part of larger social assemblages in diverse ways.

This act of counter-effectuating means working back from various actualized Christianities and, with an eye towards the specific local affects and precepts that ran through the virtual to create those actualizations, attempting to intuit the virtual multiplicity that engendered that actuality. This can be done through working back from ‘solution’ to ‘problem’, to see how various forms of Christian practice are the ‘result’ of multiple problems. Under this approach, each problem could be thought of as axes or elements of the virtual multiplicity, whose differing solutions account for the shifts in the topology of the virtual. This is not surprising. Not only have we seen again and again a call for the anthropology of Christianity to work at the level of problems, from people such as Chris Hann and Michael Scott, but in point of fact the anthropology of Christianity has been from its very inception dominated by various ‘problems’. Examples of these problems are the tension between the fact of mediation and the desire for immediacy and presence in Christian-inflected forms of communication (Keane 2007; Engelke 2007); there is also the problem of collective as opposed to individual accountability that runs through Joel Robbins’s works; we must also remember the issues of Christian ‘territorialization’, of who even ‘counts’ as a Christian in any local instantiation (Garriott and O’Neill 2008), and of how indigenous entities (Meyer 1999) or folk-ontologies (Scott 2007; Mayblin 2012) with a history apart from Christianity will be threaded into new Christian assemblages. This list of problems is not exhaustive by any means, but it shows how the contours of this field can be seen as being shaped by actualizations of virtual problematics.

What has not been thought through is to ask how each of these problems sketches out various continuums of potential, and how selection of a tranche from this continuum resonates with other tranches selected from other problems/continuums without limiting the formal independence of all these axes/continuums from each other at the level of the virtual. In short, because we have thought of Christianity in a nominalist way, rather than as a real, though virtual, object, we have not charted what we might call the mutagenic capacity in its diagram, the way in which Christianity may be many things – an unlimited number of things – but not everything or anything, and how shifts and transformations in one axis of the virtual anatomy, the diagram of the assemblage, interact with other axes. In short, these problems are perhaps found in every instance of Christianity, but the solutions vary, and the privileging of one solution over another may vary not only from region to region but between various moments as each Christianity is always in a state of becoming.

It is for this reason that we should be wary of Hann’s suggestion that anthropologists of Christian populations should not privilege intra-Christian comparisons but ground themselves solely in larger comparative endeavors instead. While such extra-Christian comparative efforts should be encouraged, privileging them over intra-Christian comparisons only makes sense if Christianity is not a domain where comparative work would yield particularly productive results, which is what one would expect if Christianity in effect did not exist, if Christianity had no ontological basis. Here, we have argued that it does. As we’ve seen, Christianity takes the appearance of differing actualized Christianities because it exceeds any particular actualized instance of it. It is a multiplicity, subject to play at the joints, that is brought into being in different circumstances. It is that continual possibility to have some other tranche of the virtual actualized that gives it what Cannell called its paradoxical nature, this sense of potentia that is never seemingly exhausted or conquered, popping up again and again like Marx’s revolutionary mole.

I would further argue that it is the play of divergence that, in a way not yet consciously articulated, has foregrounded the virtual aspect of Christianity; this is why, despite the recurrent Foucauldian refrain in the anthropology of Christianity, the charge of idealism recurs with both Hann and Comaroff, and this also is why John Comaroff is in an odd way right to describe the putative object as ‘timeless’. To the degree that we are touching on the virtual, we are touching on the aspect of Christianity that is ‘real without being actual, ideal without being abstract’, and on the aspect of Christianity as a virtual object that is situated not in moments but always between them; potential subsists even when not actualized, floating on between times, but not within them. Of course, we slip out of time to the virtual, located neither in eternity nor in the moment, only so that we can throw ourselves back into the swim of social processes at the level of the sensible; the goal is not to merely sketch out Christianity in a virtual state but to have a sense for the range and complexity of actualized elements from it, so that we can grasp how these actualized elements themselves can be folded into larger assemblages. This again is something that has been already occurring in the anthropology of Christianity, perhaps most explicitly in Simon Coleman’s thinking through Pentecostal and Charismatic Christianity as a ‘part culture… worldviews meant for export but often in tension with the values of any given host society’ (2006: 3; see also Coleman 2010: 800). Whatever else these actualizations are, they are rarely, if ever, total systems, and must entwine themselves, sometimes agonistically, with other elements to make larger assemblages; it is this latter phenomenon that is the true object of concern, but can only be addressed if we can grasp the pluriform variety of insistent Christian becomings first.

I would also say that while this does not excuse it by any means, this task of counter-effectuating back to the virtual explains how it is that the ethnographic interest in Pentecostal and Charismatic Christianity is due to more than its global exponential growth. Remember that the virtual does not thunder directly into the actual, but that various intensities, which Deleuze in his work with Guattari (1996) calls affects and percepts, are threaded through the virtual after they are conditioned by habit and memory (Deleuze 1994: 70–128); after actualization both the intensities and virtualities are covered up by the actualities that follow. Deleuze does suggest that there are moments where affects and percepts can be ‘monumentalized’, where they can be caught up in an aesthetic mode of expression in a way that does not obscure them, but rather displays them. This opens up a whole new line of argument that cannot be done justice here, but it should be noted that the importance of the aesthetic to Pentecostalism has already been observed by Birgit Meyer (2010, 2011). While all forms of religiosity (and, indeed, all forms of human life) are conditioned by the affective and the perceptual, there does seem to be a recurrent thematic in Pentecostalism to foreground the production of, and to luxuriate within, heightened affective states that suggests that Pentecostalism is ripe for counter-effectuating. Under this line of thought, Pentecostalism and Charismatic Christianity, while in no way exhausting a Christian virtual, are the sort of low-hanging-fruit of the anthropology of Christianity when it comes to ways to see at least from one position the Christian virtual in sensible form.

Now, caveats: three about the theory used to inform the anthropology of Christianity in this article, and one about the anthropology of Christianity itself. First, the conception of a Christian virtual should not be taken to mean that Christian material cannot escape Christian virtual fields of belonging; as the long anthropological history of religion has shown, syncretisms happen, and Christian material can be deterritorialized and reterritorialized in the same way that ‘a club is a deterritorialized [tree] branch’ (Deleuze and Guattari 1987: 172). Wrenched away from context and placed as an element in another virtual multiplicity, it would be doing entirely different work, and we should be alert for that, but only if we can see that there has been a break with the Christian virtual. Repurposing occurs, but so do different actualizations of the virtual Christian form, and we should only assume that we are not dealing with the latter when the Christian topology cannot be found, despite the presence of some Christian elements (an excellent example of this might be secularism, which could be considered a radical reterritorialized form of certain self-erasing features of Christianity).

But we still do not have a grasp yet of when that happens, because the work of identifying a Christian virtual has only begun, and much of what passes as debates about what might as well be the ‘essence’ of Christianity in the anthropology of Christianity literature has only been charting the range of variability in both within and between Christianities, often cast in a polemical tone that has done little to further debate but instead has been more orientated towards policing what can and cannot be thought by social scientists. Thinking about the full play of possible Christianities within and between ethnographic scenes is something that has not yet really occurred, and we cannot be sure we are dealing with deterritorialization and not differentiation until we have imagined a virtual Christianity adequate enough to think through the range of Christian actualities, including historical actualities, such as the odd early burgess-shale type efflorescence of Christian forms like Docetism, Patripassianism, and Marcionism. It is because of this that I have to offer the second caveat: this article no more offers a firm definition of what the object of an anthropology of Christianity is then did Robbins, Cannell, Scott, Engelke-Tomlinson, or even Andijar. What this article does do is to enlist Deleuze in the effort to scrape together an ontology that would make something similar to a definition possible. Here we come to the third caveat. We should be clear that it is in this definitional work that the idea of the virtual serves us. Christianity is no more, or no less, ‘virtual’ than any other entity; it is just that due to nominalist anthropological presumptions, we need the virtual to see various Christianities as in some way products of a single differentiating field.

Which brings us to the final caveat. There will most likely be those who will be unhappy with the way that Deleuze was used here; there will be complaints that Christians are not proper Deleuzian nomadic subjects, that Christianity is too molar, too Oedipal, too much locked into the logic of transcendence to be thought through in this fashion, too part of an older order to participate in the question of how the new is created (see Bialecki 2010: 710); Deleuze’s long-standing antipathy to religion, and particularly to Christianity, proves this, they will say. There are two answers I have to those complaints. First, to the degree that it is true, and there are certainly forms of Christianity that Deleuze would find politically and aesthetically objectionable, it should be remembered that Deleuze spent as much time and energy on that which hinders a certain kind of freedom as he did on that which participates in freedom. However, we can think of his obvious fondness for Kierkegaard, or his paean to belief in Cinema 2 (Deleuze 1989: 172–3) to see that there were moments where a certain kind of Christian ethos was not automatically inimical to him. And that brings us to the second point, that just as we don’t know what a virtual Christianity might be, we don’t have a full inventory of what actual Christianities are; to say that this long-standing yet rapidly mutating field all falls under a single ban, it being too molar or too monological, without attending to its variations and transformations, is in a sense to let one’s Deleuze get in the way of one’s Deleuze, and to forget that in the last moment, at least in anthropology, it is theory which must be adequate to reality – and not reality which must be adequate to theory – as we encounter the real.