Secular shadows: African,immanent,post-colonial

As late as 2006, the anthropologist James Ferguson could express concern about the extent to which Africa didn’t figure in discussions, debates, and understandings of globalization. I say “late” because we had, by that time, been inundated by talk of the global for well-on twenty years, in academic and policy circles alike. And despite some significant anthropological voices in the debates over globalization, this neglect could not, in Ferguson’s view, be placed solely at the doors of neo-liberal think-tanks or various NGOs. Analysts produced this neglect, too, especially those with the most commanding voices, in such fields as sociology, political science, and international relations.

Ferguson argued that this neglect mattered because the usual picture of globalization suggested a world of flows and worldliness: interconnection at all sorts of levels, in meaningful ways. This didn’t really work for Africa, a continent which is, at best, patchily connected to the outside and is, in many corners, gripped by failures and crises. “The ‘global’ we see in recent studies of Africa,” Ferguson writes, “has sharp, jagged edges; rich and dangerous traffic amid zones of generalized abjection; razor-wired enclaves next to abandoned hinterlands” (2006: 48).

Ferguson’s own metaphor is one of global “shadows.” Africa is marked by global shadows, not global flows. And for him the trope of the shadow is not exhausted or even well captured by the long-standing image of Africa as “the dark continent.” Ferguson (2006: 16) actually moves quickly over the question of Africa’s darkness in this sense, even suggesting that, at least in academia, we have moved beyond the prejudice that Africa is “a continent defined by a lack of enlightenment;” inasmuch as it’s a dark, shadowy place, he argues, it’s because it’s a place “where much is unknown, hard to make out, perhaps even unknowable.” What Fergsuon (2006: 16) really wants to highlight in the predication is another sense of the shadow—that of “doubling”, of attachment to, and identity with, something. “A shadow, in this sense, is not simply a negative space, a space of absence”—as it might be in dark-continent imageries—“it is a likeness, an inseparable other-who-is-also-oneself to whom one is bound” (Ferguson, 2006: 17).

The sense of doubling—of reflection and refraction—is helpful for understanding Africa’s place in the world. It builds on long-standing imageries that have emerged out of the colonial encounter. Yet I am not sure we can dismiss the matter of Africa’s “darkness.” Ferguson (2006: 15) says the image lingers “especially in popular and journalistic accounts.” He is being forgiving of academia; it still shapes—or perhaps better, haunts—scholarly accounts, too. Certainly in terms of the debates that occupy scholarly attention, Africa’s place in the world produces more than just “global shadows.”

What I want to suggest in this essay is that Africa produces some pretty heavy “secular shadows” too. As with literature on the global, Africa has often been inconvenient—even apparently irrelevant—when it comes to the literature on the secular. Where is Africa in our analyses and understandings of the secular? ¹

In what follows I consider some of the ways in which various secular formations produce shadows in and for Africa—darkness, doublings, and likenesses vis-à-vis its chief modern other of the West. My considerations, which focus in the main on scholarship produced in the West, are neither systematic nor exhaustive. They are driven by the strong sense that further considerations are necessary: that Africanists need to think both with and against the larger debates taking place in the social sciences and humanities on secular studies.

Africa has certainly been inconvenient to those who still support a classic 1960s version of the secularization thesis. Critics of the thesis are quick to invoke Africa when they need figures of church growth to offset the focus on England or Sweden’s empty pews. This isn’t the kind of inconvenience with which I’m concerned. I do not think that our understandings of the secularization thesis—and still less the secular—are best framed in terms of numbers. That’s not really what matters in terms of secular affects. Rather, Africa—as both a place and an idea—is inconvenient more in terms of how, beyond its invocation in church attendance statistics (or growth of the ummah, or community), it relates to social thought and social analysis. It is in this sense that Africa lurks in the secular shadows, so to speak, passed over and never really considered—or, in line with Ferguson’s interests in “the global,” thought to be a double—like Europe, only slightly off; slanted and enchanted. ² Africa is still often understood as the Europe of a time ago.

It is of course possible that one reason Africa lurks in the secular shadows is because the secular itself is such a shadowy, even phantasmagoric, term. It is possible that “the secular” has not been used much because it’s not descriptively or analytically relevant. At the very least, we should note the secular’s most significant other—“religion”—has long been suspect in the eyes of Africanists. “Religion” is an enlightenment term, a Western term, an imperial term. ³ What scholars have long called “African traditional religion” is no such thing; it is the colonial designation, driven, we might note, by the compartmentalizing logic of the secular—and, often, a “rational” and science-driven Christianity (see Comaroff and Comaroff, 1991). So if religion doesn’t fit, why should anyone then go on to invoke the secular?

More than inconvenience, then, it has been irrelevance that marks our sense of Africa vis-à-vis the secular. As I’ll go on to discuss, some African societies have even been referred to recently as “never secular” (Luhrmann, 2012).

This sense of irrelevance, of a never-land, is certainly evident in the realm of politics and statecraft. One of the most striking aspects of the work done by modernization theorists on Africa during the independence era is how little the secularization thesis figured. In Crawford Young’s (1982) study, Ideology and Development in Africa, for example, the basic idea of a secular political settlement never merits sustained or central discussion; it is only ever mentioned in passing, such as in his discussion of Somalia’s General Mohamed Siad Barre, who showed “no mercy for those who argued an incompatibility between Islamic theology and the secular faith” (1982: 66–67). David Apter’s (1955) early study, The Gold Coast in Transition, doesn’t get to anything like “secularism;” in fact it doesn’t even get to religion. Apter begins the study with an apology for this, especially the omission of Islam and Christianity. He does say that when he speaks of anything “traditional,” though, that should be understood to “subsume religious aspects” (1955: vii). ⁴ Inasmuch as the constitutions of the new sovereign states (such as Ghana) enshrined freedom of religion, secularism simply wasn’t a central topic of interest or action. Since the first period of independence, in fact, constitutions have, if anything, been desecularized in various ways: framed with explicitly religious language or even declarations of faith. In 1996 the preamble of the Zambian constitution was amended to declare Zambia a Christian nation—a designation that was upheld and reiterated in 2013. Ghana’s constitution of 1992 begins: “In the name of the Almighty God.” Ghana’s public sphere has also been infused with a “Pentecostalite” style, driven by the rise of Pentecostal churches as political, economic, and social forces with which all politicians must reckon (see Meyer, 2004). The Nigerian constitution of 1979 declares Nigeria a secular state. Yet, this has come under regular attack from a range of actors, leading one proponent of secularism (Abioje, 2013), who seems to be nearly alone in the wilderness, to decry what he sees as the sorry state of affairs. Many other states—Senegal, Mali, Zimbabwe—also have secular constitutions, yet they too, have come under question, especially in recent years, and in any case Africanists have devoted little time or attention to “secularism in Africa.” The same can’t be said of America, Europe, the Middle East, or, say, South Asia. There are huge literatures looking at secularism in these places.

The seeming irrelevance of the political secular extends to the first generations of African socialist regimes. African socialism was rarely an atheistic socialism—and never successfully so. Julius Nyerere, President of Tanzania, scoffed at the idea of it as something universal in shape, form, and meaning—as opposed to the Word of God. As a Christian, he wrote, the Bible was one thing and socialist tracts another; socialist books are “written by men; wise and clever men, perhaps—but still men” (cited in Young, 1982: 99). The liberation war in Zimbabwe was won by Marxist-Leninist and Maoist guerrillas who drew heavily and explicitly on the power of the traditional spirit mediums. Even in neighbouring Mozambique, where there was a systematic denunciation of “superstition” in a classic socialist-secularist register, it was never straightforward. As Harry G. West (1997; 2001) makes clear, FRELIMO’s model of the “new man,” freed from the shackles of tradition and obscurantist ideas of sorcery, eventually became a liability—so much so that the political heirs of Samora Machel leveraged the ambiguity of their identification by the people as sorcerers. And now in Mozambique we have a full-fledged acknowledgment of (and anthropological wrestling with) enchantment, of what Juan Obarrio (2014) calls “the spirit of the laws.”

Secularism as a political settlement has never been, and is not, a prominent concern in sub-Saharan Africa. Even where it does appear—occasionally in discussions of Nigeria, or South Africa (see Leatt, 2007)—it is often ancillary. As a principle of statecraft, secularism simply isn’t part of the core vocabulary of African modernities. ⁵

Why is this? One reason, I think, has to do with the fact that the European colonial powers found no strong traditions of “world religions” in much of sub-Saharan Africa. There was thus no presumed “religious field” to regulate. As Chidester’s (1996) work makes clear, it took some time for the colonial powers even to recognize “tradition” as having the “religious aspects” about which Apter was eventually to write at the dawn of the independence period. Indeed it is worth noting that where we do see signs of secular statecraft, it is in those areas of sub-Saharan Africa where Islam did have a presence. Islam was religion enough. As we are long used to hearing, the secular and the religious are mutually constitutive; but where there is no “proper” religion—where it’s just “African tradition”—there is nothing “secular” to be constituted.

A related answer might be developed in relation to the evolutionary logic of colonial governance and mission. Secularism follows a social-evolutionary logic; it is seen, by its proponents and architects, as an achievement of civilization. Therefore, secularism as a political project only makes sense in relation to a certain kind of society—those with “world religions,” not just witchdoctors or spirit mediums.

All of this matters. It matters that secularism has seemed irrelevant in this sense. It suggests something about the emplacement of the African political. And inasmuch as secularism is a strand of social evolutionism, its absence here can both produce and reinforce Africa’s ostensible darkness.

It is relatively easy to see that secularism plays a minor role if one focuses on secularism in an explicit sense. And after all, there is so much talk of “religion” in African public spheres. This, though—rather like the focus on numbers—is misleading on its own. Rather, one might argue, drawing on the influential work of Talal Asad (2003: 16), that the secular is best approached indirectly—indeed “through its shadows, as it were”—and that what we have in the Africanist literature is thus not a canon on the secular but a series of discussions and debates—about subjectivity, about ontology, even about statecraft—that otherwise reveal the secular.

There is a lot to be said for this argument. Consider it, for example, in relation to the work of the anthropologist, Elizabeth Hull, who conducted research in a hospital in rural KwaZulu-Natal. Founded by Methodist missionaries, the hospital was taken over by the state in the 1980s, and further transformed in shape and outlook in the post-Apartheid era. Even the chapel, Hull (2009: 124) notes, was eventually transformed into a space for conferences and board meetings. This was never really understood, or spoken about, as “secularization”—certainly not by the nurses Hull (2009: 119) got to know. The more relevant Weberian keywords had to do with “bureaucracy” and “rationalization”—intimately related to the secularization thesis, of course, but not reducible to it. What Hull shows, through her careful ethnography, is how the emergence of new discourses and modes of governmentality, shaped by new ideals—of human rights, of transparency and accountability—reveal a host of what might well be understood as secular formations, formations that have to do with understandings of suffering, health and well-being, workplace equality (and hierarchy), and professional efficiency. What Hull also shows is how the secularization of the hospital allowed for a flourishing of religiosity—not a resurgent Methodism, in this case, but an upstart, born-again faith among the nurses, who held prayer meetings each morning.

Building on Asad’s approach, Hussein Ali Agrama (2012) (one of Asad’s students), has recently drawn attention to “questioning secularism.” By this he means two things: first, and most obviously, that we should never rest easy with our understanding of what secularism is, or does. It changes; it becomes localized; it, in fact, is not even an it—a thing. Yet Agrama also doesn’t want to wholly surrender the concept to particularism. He is interested, after Asad—and, it must be said, many others (not all Asadians)—in approaching secularism “as a general historical phenomenon” (Agrama, 2013). This requires something of a base line. Second, then, and as this kind of base-line claim, Agrama argues that secularism’s “form of power” (as he calls it) is questioning itself; it questions everything, including its own normative presuppositions.

I want to come back to this second point, but it is with a general anthropological questioning of categorization that we might continue. For if we take “secular” in what is often understood to be its original meaning—“of this world”—it quickly becomes clear that the secular/religious binary doesn’t always work well in sub-Saharan Africa. This gets us back to the unease with which many scholars now approach “religion” in African contexts, along with its attendant binaries of immanent/transcendent, natural/supernatural, spirit/matter. One particularly good example of where such binaries fail us comes from the work of Igor Kopytoff (1971), based on research among the Suku, but drawing as well from studies of other Bantu-language groups. As Kopytoff points out, it had been common (and some might argue, still is common) to refer to African attitudes to the dead as worshipful. And yet the idea of “ancestor worship,” he argues, “introduces semantic paradoxes” (1971: 138) where none exist, for the deadness of the ancestors is not the relevant factor. Ancestors may be dead, but they are also elders, and it’s this which gives them their status. “It is striking that African ‘ancestors’ are more mundane and less mystical than the dead who are objects of ‘worship’ should be in Western eyes” (1971: 140). The genius in Kopytoff’s analysis is to break down the distinctions among the core structural-functionalist categories of classification—religion, kinship, politics—and show how, in Suku life, it is all part of something more fluid and porous.

Read today, in light of the burgeoning literature on secularism, Kopytoff’s work might seem like a textbook example of that literature’s purchase and appeal. Because in the process of working to show how neither death nor ideas of the supernatural are apt—ideas assumed to define the category of “religion”—in a sense what he shows us is that the best way to understand this world is in a secular frame, in an immanent frame. Here, of course, immanence has almost the exact opposite meaning assigned to it by Charles Taylor. This is not a “religious” world, Kopytoff tells us, but neither is it a disenchanted world. It is “immanent” because everything is present, everything is part of a single world, a continuum in which there is no break, radical or otherwise. It’s all earth, no heaven. All here, no there. Transcendence is not key, at least not in the way Taylor conceives it.

It is not just in such classic or traditional cases that this particular valence of immanence is relevant. I am more interested in seeing the Suku case as coeval with the Westerners in Kopytoff’s contrasts than Taylor would be; for Taylor, this kind of African world, as I’ll address in more detail below, is only discussed in relation to medieval Europe. But we can see similar kinds of immanence in post-colonial contexts. In Nostalgia for the Future, Charles Piot (2010) brings over thirty years of fieldwork experience in Togo to bear on the question of social transformations. Togo today, after the death of its Cold-War-era dictator, Gnassingbé Eyadéma, and the subsequent withering of the state, has seen the efflorescence of NGOs and Pentecostal churches, each of which produce a logic of immanence, a here-and-now spatio-temporal framing for a national citizenry that wants nothing more than to leave the nation—to realize the promises of modernity that cannot be redeemed by steady progress but only the nearly miraculous event of winning the visa lottery, or facing down local tradition with the inspiration of globally framed charismatic critiques.

In a recent set of reflections on A Secular Age, Simon During questions the comprehensiveness of Taylor’s secular/religious rubric, pointing out that what’s left out of this rubric is the mundane. “Ever since the Enlightenment, the secular has denoted not so much what lies beyond religion’s interest and grasp but what contributes to its intermittent diminution, corruption, marginalization, and undoing. The mundane is the philosophical concept that names what stands outside that division between the secular and the religious” (During, 2010: 113). For During (2010: 113), what the mundane consists of are “those forms of life and experience that are not available for our moral or political or philosophical or religious or social aspirations and projects.” This is not to say the mundane can have no moral or political consequences. During is thinking here in terms of trajectories—and if not teleological, at least “progressive.” The mundane has no such trajectory in itself, or for itself. Citing Theodor Adorno, he refers to “‘being, nothing else, without any further definition or fulfilment,’ [which] might take the place of process, act, satisfaction” (in During, 2010: 116). In the African contexts I’m discussing here this should not be confused with a kind of static, unchanging, traditional life. Perhaps especially in the contemporary case Piot describes, this is not the case: what the Togolose have is, as he tells us, a nostalgia for the future, a shift from their mundane existence back into the slipstream of modernity that life after the Cold War has displaced. One could certainly argue, taking cues from During, that much of Africa is not a religious place, and still less a secular place, but, rather, a mundane place—that which lurks in the shadows, that which provides the shadows, for the West’s conceptions of modernity and enlightenment.

Read against Kopytoff’s characterization of the Suku’s world as mundane, or that captured by Piot on the bare life of the Togolese, During’s reframing of the secular age and its terms allows us to raise a further set of issues—those around what During (2010: 115) pinpoints as the mundane’s atheoretical character, its inability to form a “disinterested and critical attitude toward the world.” In short, the mundane forces the question of critique: the question of, as Agrama puts it, “questioning secularism.” To appreciate how, let me first sketch out some of Taylor’s main arguments and then go on to consider them in relation to a set of earlier arguments within the Africanist literature, arguments that bear a family resemblance.

Taylor’s characterization of the immanent frame is one in which two main things happen. First, and most importantly, we (and by “we” he means those in Latin Christendom, a phrase many of his readers don’t like) come to realize that belief is not the only option, that the religiously saturated world of 1500 has given way to one in which there are other choices, above all those based on scientific materialism and even “exclusive humanism.” Second, and following on from this, enchantment gives way to disenchantment. We gain what he calls “buffered selves,” increasingly impervious to immaterial intrusions and infillings, and buffered selves, moreover, headed up by buffered minds, which become the sole locus of agency, intention, and being. So the disenchanted world, the immanent frame, is one in which it becomes increasingly harder to imagine—much less find—fairies in the garden, or, even, people who really think that they need to be mindful of their dead ancestors’ wishes.

Taylor makes clear from the outset of A Secular Age that he can’t, even in such a long book, do justice to his Latin Christendom alone, and still less that which exists outside of it. To index how the situation might be different, however, he cites an example from the work of anthropologist Birgit Meyer, who relates the story of a Ghanaian woman, Celestine, who, walking home one day with her mother, finds herself accompanied by a man in a white robe. Celestine’s mother does not see the man; he turns out to be an Akan spirit, taking Celestine into his service. This is a porous self in Taylor’s rendering. In a place like Ghana, Taylor writes, this is something different—something akin to the enchanted world of 1500 about which he writes at greater length. The spirit is part of what he calls “immediate reality, like stones, rivers, and mountains” (2007: 12). So this isn’t the immanent frame for Taylor; Celestine lives in a still-enchanted world.

I have already suggested how anthropological work on Africa can be used to understand immanence differently. More directly in relation to this example, Taylor has been criticized for this framing of the frame by Meyer (2012) herself, as indeed by other anthropologists. One of the main criticisms is that, his characterization of Celestine as a “contemporary example” (2007: 11) notwithstanding, it is a denial of coevalness (see Fabian, 1983)—a denial that the West and Africa occupy the same time. Another criticism, in line with my reading of Kopytoff and Piot above, is that it restricts what gets understood as immanent or enchanted or secular in the first place, and how power relations shape them. In this vein, Saba Mahmood (2010: 299) argues that Taylor’s version of the secular “remains blind to the normative assumptions and power of Western Christianity;” it remains blind to, or, at least, seemingly unconcerned with, “Christianity’s enmeshment in an imperial world order” (2010: 287).

It’s here that we can address how the normative assumptions of the secular are the product of contingencies, rather than necessities, of an “immanent frame.” And it is here that a tension—a productive tension, I think—emerges in how we might understand the temporalization of the secular. For on the one hand, Taylor’s metaphor of the frame suggests that what Celestine experienced is something enchanted, extraordinary; on the other hand, as in fact he suggests, in this enchanted world, Celestine’s experience is, or might well be understood as, perfectly ordinary—perhaps even mundane.

Within anthropology, one classic approach to someone like Celestine would be to say that she lives in a “closed system of thought.” This phrase, and the terminology of open and closed systems, is developed in the work of Robin Horton. It has a lot in common with how Taylor describes both the Latin Christendom of 1500 and the Ghana of today, each characterized by what he calls quite flatly, naivety: “forms of immediate certainty” (2007:12) that go unquestioned because people don’t realize there are questions to ask.

Horton’s interest is in tracing what he sees as both the similarities and differences between what he calls African traditional thought and Western science. The similarities are stronger than we might at first imagine. What he wants to dispel above all are two assumptions. The first is that “African traditional thought” is atheoretical. The second is that African traditional thought is blind to causality. In short, what Horton wants to underscore is that Africans can see the big picture, they can think systematically, and they understand cause and effect in the physical, material world. In this Horton is shoring up the arguments of E.E. Evans-Pritchard (1976), in his work on the Azande in central Africa from the 1930s. What Evans-Pritchard set out to do was show how witchcraft, oracles, and magic among the Azande were perfectly rational in their own terms, perfectly logical registers with which and through which to understand the workings of the world. Witchcraft didn’t replace or cloud common-sense; it was, simply, a complement to it, a theoretical outworking of why vis-à-vis the how. “Belief in witchcraft is quite consistent with human responsibility and a rational appreciation of nature” (Evans-Pritchard, 1976: 30).

Evans-Pritchard’s occult apologetics has been the subject of much debate in anthropology and philosophy, a lot of which hinges on the extent to which it seeks to rationalize the world of the other. (Horton has also been regularly criticized.) In this, of course, we might well recognize a project of secularization—peeling back the oddness of magic and other enchantments to reveal a core of reasonable thought. We see this in the quotation just given. It is worth noting, however, that for both Evans-Pritchard and Horton, the “translation” of African thought into a Western idiom is neither total nor, in a sense, wholly desirable. These anthropologists buck their own secularizing, rationalizing tendencies: they would be sometime “antisecularists,” in today’s critical parlance (although of a very different theoretical disposition than those for whom the label is used). ⁶ In a well known and near heretical passage, Evans-Pritchard leaves open the possibility of something other-worldly and other-wise. He relates his one personal experience of Zande witchcraft, seeing “the light of witchcraft” on one of his nightly walks, a light that could not be accounted for and which, followed as it was by the death of an elderly man in a nearby homestead the next day, “accorded well with Zande ideas” (1976: 11). Evans-Pritchard does not distance himself from this possibility, momentarily dropping the critical stance of the British anthropologist. Horton, for his part, includes a personal admission in the essay we’ve been considering: he rather likes the ways in which African traditional thought fosters “an intensely poetic quality in everyday life and thought, and a vivid enjoyment of the passing moment—both driven out of sophisticated Western life by the quest for purity of motive and the faith in progress” (Horton, 1993: 249–250). Horton is no straight secularist; here he sounds more like the contemporary political theorist Jane Bennett—with her commitment to wonder, to being “struck and shaken by the extraordinary that lives amid the familiar and the everyday” (Bennett, 2001: 4)—than a champion for disenchantment, or other soulless modern. (Even if Bennett might not adopt the post-colonial romanticism of the anthropologist.) Here, in fact, Horton’s reflections echo some of those we find in Weber (1946), for whom the “intellectual sacrifice” of religious faith was anathema, but also admirable.

I want to shift now to the proposed difference between African traditional thought and Western science, for it’s here that we get to the synergies with Taylor and debates about secular and post-secular formations. “What I take to be the key difference is a very simple one,” Horton writes. “It is that in traditional cultures there is no developed awareness of alternatives to the established body of theoretical tenets; whereas in scientifically oriented cultures, such awareness is highly developed. It is this difference we refer to when we say that traditional cultures are ‘closed’ and scientifically oriented cultures ‘open’” (Horton, 1993: 222).

Taylor’s characterization of a secular age is, in many ways, hinged on precisely this difference: the assertion that what marks secularity is “openness,” as Horton puts it; options. ⁷ “We live in a condition where we cannot help but be aware that there are a number of different construals, views which intelligent, reasonably undeluded people, of good will, can and do disagree on” (Taylor, 2007: 11). This is not the condition of Evans-Pritchard’s Azande. “A Zande is born into a culture with ready-made patterns of belief which have the weight of tradition behind them. Many of his beliefs being axiomatic, a Zande finds it difficult to understand that other peoples do not share them” (Evans-Pritchard, 1976: 202). Yet for Taylor the openness of a post-scientific-revolution worldview becomes its own kind of closure. Western science actually produces what he calls a “closed world structure” (Taylor, 2007: 560). It tends to push out anything transcendent. “Once one has taken the step into unbelief,” he writes, “there are overwhelming reasons why one will be induced to buy into the official, science-driven story” (2007: 568). Even more than Horton, with his standard-issue measure of romantic appreciation of the African Other’s ways, Taylor issues warnings about the sufficiency of science, of the secular. Yet in outline Horton’s description of what becomes evident in the colonial encounter between Africa and the West is almost exactly the same as Taylor’s description of what becomes evident in the wake of enlightenment within the West. For Horton (1993: 223), “we have two basic predicaments: the ‘closed’—characterized by lack of awareness to alternatives, sacredness of beliefs, and anxiety about threats to them; and the ‘open’—characterized by awareness of alternatives, diminished sacredness of beliefs, and diminished anxiety about threats to them.” This could serve as a précis for the 800-page story in A Secular Age.

At the core of both versions of the story are the same protagonists. Whether in modern Africa or pre-modern Latin Christendom, what upsets the equilibrium is scepticism and doubt. In short, the core of the secular is critique. ⁸

As things stand, critique is probably the strongest candidate for what marks the secular, what those who want to hold on to the secular (especially for political purposes) most vigorously defend. Stathis Gourgouris (2013) and Bruce Robbins (2013) are two of the most convicted proponents of critique’s secularity in this sense. And for each it is the work of Edward Said that provides a point of departure. “Said understood secularism as first and foremost a responsibility to maintain critical distance from the common sense of the tribe” (Robbins, 2013: 248); for Robbins himself, “minimally, secularism means the capacity for self-scrutiny” (2013: 250). Gourgouris offers an even more trenchant defence of secular critique, in an impassioned and testy exchange with Mahmood: “Critique and interrogation—as autonomous, self-altering practices—are the persistent conditions of the secular” (2008a: 443–444). ⁹ Jürgen Habermas (2010: 2) also emphasizes the centrality of critique. In his reflections on “post-secularism,” he marks the secular’s contribution as “the increase in reflection, that is, the ability to step back behind oneself.” And although he comes at it from a different position, this is, again, one of Agrama’s points: secularism is a “questioning power” and as such is animated by “expressions of doubt, scepticism, and suspicion” (2012: 33).

These writers are nailing their colors to the mast. And (Agrama aside) they are taking similar positions. For Robbins and Gourgouris, the positions are defined at least in part by what they see as the reckless tolerance and even crypto-religious sensibilities of those they label “antisecularists.” By this they mean those who, especially in post-colonial studies, seem to countenance conservative religiosities, to defend the Rest from the West’s demanding secular strictures. It’s not that Gourgouris and Robbins have an unswerving allegiance to “secularism,” and still less the liberalism that often accompanies it. It is, rather, that, for them, someone like Mahmood has thrown the baby out with the bathwater, inasmuch as they see her defending (by explaining) the Piety movement in Cairo—the subject of her monograph (Mahmood, 2005)—over and against secular-liberal norms at the expense of what Robbins (2011: 92), for one, calls, with neither irony nor scare quotes, “a narrative of progress.”

As position takers, such critics forge a link between critique and the secular with particular commitment and verve. Yet the centrality of critique is also, by and large, the working assumption of other anthropologists who, like Mahmood, would be labelled antisecularist (at least by the anti-antisecularists). There can be “antisecularists” then, who nonetheless put some stake in the idea that critique is, in fact, secular. Tanya Luhrmann (2012) is one of these anthropologists.

What interests me about Luhrmann’s work is how doubt—as secular critique—becomes pivotal for a certain kind of secular/religious divide. Although her own research is on neo-Pentecostals in the United States, she frames her analysis in relation to a set of non-Western, colonial and post-colonial cases (including Meyer’s on Ghana and my own on Zimbabwe [see Engelke, 2007]), which she refers to (inspired by Joel Robbins) as “never secular” (2012: 372). In doing so, she calls attention to Horton’s work as useful for understanding Western/traditional differences, even if she doesn’t accept all of the implications of his model (chiefly, the presumed connection between Western “openness” and disenchantment). Luhrmann does, however, accept that the neo-Pentecostals she studied live in something like the immanent frame described by Taylor. “One can quarrel with the characterization of ours as a secular age, but one cannot quarrel with the observation that faith is now lived with acute awareness that one can choose not to believe—not only in this specific faith, but in the transcendent at all” (Luhrmann, 2012: 378).

It is not just any kind of doubt or option-taking that matters here. In the never-secular examples of Ghana, Zimbabwe, and elsewhere in the post-colonial world the doubt that’s not recognized as a viable option is the one in “the transcendent at all.” “In never-secular societies, where the reality of the supernatural as a category has not been profoundly questioned, doubt is focused on specific supernatural claims—the reality of non-Christian spirits, the validity of spiritually charged mechanisms, the efficacy of particular prayers” (Luhrmann, 2012: 381). It is true that in my own work, on an African apostolic church, the shift was from ancestors (Shona: midzimu) to the Holy Spirit (Shona: mweya mutsvene). This can be used to suggest that religious change in post-colonial contexts is change within a “closed system” à la Horton. With Kopytoff’s conclusion in mind, however, and the lessons of historians of religion (Chidester, 1996; Masuzawa, 2005), we should not underestimate the extent to which religious change was also change to religion: that is to say, that what happened throughout much of colonial Africa was not a shift within transcendental and supernatural orientations but, rather, a shift to them. ¹⁰ Whatever else it suggests, the ethnographic record on sub-Saharan Africa makes it clear that transcendence and the supernatural can’t be taken as the lowest common denominator of a secular/religious equation. Too often, it is. Yet in Shona, as in many other Bantu languages, and many other languages of erstwhile colonial subjects, there is no word for anything like “heaven” or “hell.”

What transcendence does is make a certain kind of option—the “secular” option—necessarily of a different epistemological and even ontological order. What is it that authorizes this difference? A good argument could be made, I think, that the shift from “traditional” to “secular” would have been easier than that from “traditional” to “religious” (i.e., Christian). In much of sub-Saharan Africa, what Christian missionaries and other religionists had to fight for was the introduction and necessity of the transcendent, of the supernatural. Maybe they would have had an easier cosmological go of it if they hadn’t needed to foist heaven and hell on so much of the globe.

The question of a secular difference, when pegged to scepticism and doubt, can be further appreciated in two essays that stand, I think, as valuable precursors to current debates, by Karin Barber (1981) and Eric Gable (1995). Considering them can also help us wind down the discussion here.

Barber is among those who show how certain Western understandings of religion fail us in Africa. Among the Yoruba—and they are not alone in this—there is an important sense in which humans “make” gods (rather than God making humans). A god (orisa) is “maintained and kept in existence by the attention of humans” (Barber, 1981: 724); in the process of doing so, the humans constantly question, doubt, and occasionally berate the orisa. Yoruba is, in this sense, a culture of critique. And yet while this scepticism might at first blush suggest the Yoruba are good hermeneuts of suspicion, in fact, Barber (1981: 740; 741) argues, the scepticism is “at the heart of the Yoruba devotional attitude” and “in no way questions the existence of spiritual beings as a category.”

Barber’s case could be added to those discussed by Luhrmann; the Yoruba in this reading are never secular. Doubt—and the doom of any given orisa—is not doubt, or doom, of the system. They’re still within—still “closed” in Horton’s sense. Yet Barber (1981: 741) makes another important point, which is that with such a conclusion we’re not asking an important question: “what are the differences of structure—the structure of society and of ideas—which allows something apparently similar to scepticism to play such a different role”? There is not space here to provide her full answer. But what Barber’s work can help us recognize is a point that Agrama, in the contemporary moment, makes more explicitly: secularism’s normative demands force not only particular ways of seeing, but also particular kinds of questions. Barber is “questioning secularism” too.

In Guinea Bissau, Gable found himself troubled by “native skeptics” (1995: 242) amongst the Manjaco. He was troubled because the scepticism seemed integral to their traditions, yet the default language in which this could be discussed would suggest they had simply become colonized moderns. All the men he profiles—who question the spirits, and, in particular, the content and form of a significant initiation ceremony (the kambach)—had, in local parlance, “seen France” (1995: 245). To fully appreciate the Manjaco’s “pragmatic and disenchanted scepticism,” though, he argues we need to take “an ‘anachronistic’ approach to Manjaco consciousness” (Gable, 1995: 242). This does not mean denying them history, not least the ruptures and transformations wrought during the colonial encounter. Rather, we need to move beyond the romance of tradition (something anthropologists are not always good at), especially where it suggests that anything African that looks modern has to be understood as aping the colonizer—as a “bad copy” of something Western (Gable, 1995: 252–253; see also Mudimbe, 1988).

With Gable’s point in mind, we might appreciate that saying the secular is irrelevant is partly commendable because it refuses a totalizing logic, yet is partly remiss, because, in doing so, it reinscribes African difference and forecloses the interrogation of general historical phenomena. If critique can only be secular, then Africa can only be traditional (or failed modern). Yet for the Manjaco (as indeed, following Barber, for the Yoruba), the starting point of “religiosity” is scepticism—questioning—not faith. Ludwig Wittgenstein once said that doubt comes after belief. Just so, we can argue that in other times and in other places, belief comes after doubt, that pragmatism, rather than piety, forms the core of what we dub “religiosity.”

So I am not arguing that we should bring sub-Saharan Africa into line—into the conversations, debates, and arguments taking place in relation to elsewhere. Neither, though, do I come to Robbins’ (2013: 262) conclusion that post-colonial studies “would perhaps be better served either by shunning the secular/religious binary altogether or by returning to its initial ambivalence.” “The loss of this post-colonial context has been a loss for studies of secularism more generally” (Agrama, 2013). As Vincent Lloyd and Ludger Viefhues-Bailey suggest in their introduction to this special issue, it behoves us to consider the ways in which the colonial, the modern, and the secular are related. To be sure, in some instances those relations are either weak or forced. Post-colonial studies of Africa, and the social dynamics and formations they trace, have not lacked the secular; they have often addressed the same or similar issues without recourse to this term—as I have suggested, for example, in relation to Hull’s work on nursing staff in a KwaZulu-Natal hospital. We can’t disentangle Africa from the West or “Latin Christendom.” The particular forms that secularism takes are partially determined in relation to Africa. We have also seen how for prior generations of Africanist anthropologists the debates today over what, if anything, constitutes a secular age would have an uncanny ring. Between Horton in Nigeria in the 1960s and Taylor in Canada in the 2000s, there has been a transposition of registers in which open becomes closed and vice versa. One important point to take away from these observations is a simple one, although not always one that seems to be borne in mind. The secular is a contingent naming. It is a term with purchase but can be exchanged for others. Shadows—secular shadows—need not always be broken up. Sometimes shadows provide a helpful shelter from the heat.