Is religion a valid category of scholarly inquiry? In this post, I briefly set out three distinct approaches to the study of religion: criticizing religion, upholding religion, and disaggregating religion. Although I cannot make the full case here, I sketch a preliminary defense of the third approach, in the context of recent debates in political theory.

Criticizing religion

By “criticizing religion,” I mean not the critique of the beliefs or practices of self-described religious individuals or groups but rather the critique of the concept of religion as a scholarly category. According to a number of scholars (often influenced by Foucauldian or post-colonial thought), the category of religion is deeply implicated in the history and practice of western statism and imperialism. The only appropriate scholarly stance towards this object is one that is critical and skeptical. There are two main reasons for this suspicion.

The first is that religion is (as Daniel Dubuisson once wrote) “the West’s most characteristic concept, around which it has established and developed its identity.” There is a deep and significant continuity between the Christian defense of the “true religion” and the nineteenth-century invention of “world religions.” The universalism of the concept masked an internally hierarchical classificatory scheme, which, as Tomoko Mazuzawa has shown, was crafted within the crucible of the missionary and colonial encounter. Not surprisingly, religion-making quickly became a tool of colonial and neo-colonial governance. The first European missionaries had thought that the native peoples of the Americas or Africa had no religiononly “cultures” or “superstitions.” Once colonized, it was discovered that they did have religion after all. Religion-making involved singling out certain social activities and cultural practices as “religious” and de-politicizing them (see Hinduism in India). Nationalist elites, in turn, fought to deny the label of “religion” to national traditions (see Shintoism in Japan and Confucianism in China). Such struggles were political through and through, and it is impossible to discern any common core or essence to all the world religions, as W. C. Smith pointed out in 1962. Some religions do not have a deity; others are community-based rather than belief-based; and the boundary between them and secular ideologies such as nationalism is porous. Religion is a concept created by modern scholars and superimposed on a variety of different phenomena for a variety of motives. This had led some scholars, such as Tim Fitzgerald and Naomi Goldenberg, to draw the radical conclusion that the category of religion should be simply abandoned.

The second ground for skepticism is that religion has served to define the western idea of the secular, and remains deeply entangled with it. For critics, this is visible in three different ways. First, the secular state defines what religion is; religion is privatized as a faith whose object is the supernatural, and differentiates it from the natural and the rational which are the jurisdiction of the state. Such distinctions, however, only serve to shore up the arbitrary power that the state has to demarcate the sphere of its own sovereignty. Second, secular modernity is itself rooted in a distinctive Protestant anthropology, the ethos of which neatly maps onto a modern liberal subjectivity that encourages the individual to cultivate the autonomy and discipline required to relate to her beliefs and ends in the right way. Talal Asad and Saba Mahmood have shown, for instance, that liberal secularism requires individual beliefnot communal observance, rituals or embodied pietyto be taken as the essence of religiosity. Third, the secular separation between a privatized, individualized sphere of religion and a public, social, rational sphere of politics has obscured the way in which the state, the nation, and the law operate as the modern sacred. The modern sovereign state is grounded in a distinctive political theology that mobilizes the structural categories of metaphysics and theology to bolster and consolidate the higher identity of secular citizenship. According to William Cavanaugh’s provocative thesis, the fact that Christianity is construed as a religion, whereas nationalism is not, ensures that the Christian’s public and lethal loyalty belongs to the nation-state. Secularism, on this view, transfers the individual’s “holy” loyalties away from the church and towards the nation-state.

Even this cursory summary of this first approach is, I hope, sufficient to convey the force and depth of the critique of religion. If religion is a western concept that has been used to shore up the authority of the colonial and sovereign state, through shifting, arbitrary demarcations between “religion” tradition, culture, reason, and the nation, then scholars should be wary of treating it as a stable, coherent object of academic study.

Upholding religion

The implications of the first approach have been resisted in many quarters. Here I focus on two objectionswhich, for ease of presentation, I call the “anthropological” and the “normative” objection. (Note that no assumption is made about the actual disciplinary location of scholars. Much of the interesting work done in anthropology, sociology, and religious studies departments, for example, draws on both the first and the second approach).

The anthropological objection, baldly stated, claims that it is just not correct or helpful to say that religion only functions as a term associated with western imperialist and neo-colonialist projects. Like most abstract concepts, the concept of religion is a construction projected onto the world, but one consequence of this large-scale projection may well be that our world has genuinely come to exhibit it. Nor does the origin of a concept in itself discredit its uses: the concept may well have been forged in the crucible of missionary and colonial encounters, but its meanings and uses have further proliferated in non-colonial and post-colonial settings, in ways that escaped, distorted, and subverted the original discourse. It is a rich concept that connects to vocabularies that bring persons and things, desires, and practices in particular traditions in distinctive ways. Not surprisingly perhaps, the work of those scholars who have most influenced the “critical religion” school, such as Asad and Mahmood, explicitly aims to provide a thick anthropological description of the complexity and diversity of religious experienceit brings out, for example, the distinctive modes of Muslim religious subjectivity.

From this perspective, it would be a mistake to think that conceptual imprecision is in itself an obstacle to scholarly inquiry. It may well be true that the boundaries between religion and other cognate concepts such as tradition, culture, ideology, faith, reason, and so forth are porous and fuzzy. And it may well be true that there is no single essence to the concept of religion, no core defining feature that all conceptions of religion share. But perhaps we should apply to religion something like Ludwig Wittgenstein’s theory of family resemblance. What Wittgenstein said about the concept of “game” also applies to the concept of “religion”: the fact that we cannot identify a single feature that all religions exhibit does not mean that the concept of religion is meaningless. The concept exhibits “a complex continuum of resemblances and differences.” In sum, the “anthropological” objection urges us to take both ordinary language and ordinary experience seriously.

The normative objection is slightly different. The worry here is the hard-won liberal right to religious freedom will not elicit much respect if the existence of religion itself is radically questioned. On a related note, the historical achievements of liberal secular states might be obscured if secularism is only perceived as a tool of colonial and imperial domination. Secularism and freedom of religion, in liberal thought, are connected to moral values such as individual autonomy, moral self-determination, freedom of conscience as well as more political ideals such as non-theocratic constitutional democracy, separation of church and state, and non-establishment of the dominant religion. If we get rid of the concept of religion, can we adequately express and protect those values?

As it happens, one group of influential political philosophers has tried to do just that. For “egalitarian theorists of religious freedom” (as I call them), there is nothing special about religion as such. Religious beliefs, in particular, are only one sub-set of a larger class, which political philosophers, following John Rawls, call “conceptions of the good.” On this view, a liberal state protects, not only freedom of religion, but more generally freedom of conscience; it disestablishes, not only dominant religion, but also any controversial conception of the good. So neutrality, for liberals such as Rawls and Ronald Dworkin, is “generalized non-establishment.” Egalitarian theorists of religious freedom offer a compelling response to the “critical religion” school. They suggest that, even though historically the liberal state may have defined itself in relation to a problematic construal of religion, it has since broadened the scope of its moral concern, and is therefore less vulnerable to charges of ethnocentrism. Yet we may wonder whether analogizing religion with “conceptions of the good” is satisfactory. I do not think it is, although I cannot make the full case here. Instead, I shall merely sketch an alternative strategy.

Disaggregating religion

Let me get back to the starting point of our inquiry: Is religion a valid category of scholarly research? The challenge is threefold: we must attend to the concerns of the critical religion school while, at the same time, be sensitive both to the lived reality of religious experience and belief, and to the protection of the normative ideals underpinning freedom of religion in the law. How would a strategy of “disaggregating religion” help here?

First, different disciplines will have different answers to the question of what the concept of religion is for; and as a result they will work with different conceptions of religion. Anthropologists, for example, legitimately study a key dimension of human experiencethe religious experiencein its diversity and unruliness. Theologians adopt a more content-based approach: they study the texts and dogma which capture our fundamental dependence on a greater order of things. Intellectual historians and students of political thought study religious traditions as coherent, inter-generational, scholarly bodies of thoughtthe “Islamic” or the “Christian” or the “Confucian” tradition, for example. These (and others) are all legitimate uses of the concept of religion. So the starting point of the disaggregation strategy is to accept that, as religion is indeed not a “thing” but a term of art (it has different meanings in different contexts), different dimensions of it will be appropriated in different ways in different disciplines.

Yet the strategy of “disaggregating religion” has a further, more portentous implication for normative political theory. The specific question which political and legal philosophers (should) ask themselves is this. Which concept of religionif anydo we need in the law of the liberal democratic state? How can we best identify and protect the normative values which historically have underpinned freedom of religion, non-establishment and the like? Egalitarian theorists of religious freedom, from this perspective, are right to seek to identify the normative grounds on which certain beliefs and practices call for state protection (or state restraint, in the case of non-establishment). They argue, for example, that exemptions from general laws can be justified on the ground of conscience; and that what the state should not “establish” are conceptions of the good. A standard line of criticism against this egalitarian strategy has been to argue that it rests on a contested understanding of religion. Critics have pointed out thatby analogizing religion with individual conscience or conceptions of the goodliberalism reveals its Protestant, individualistic bias, and is unable to capture the fullness of the religious experience. The point has been made with particular acuity by legal scholar Winnifred Fallers Sullivan in her book The Impossibility of Religious Freedom. She shows that the First Amendment Free Exercise clause is quite incapable of capturing the popular, unruly, ritualized religiosity that she sees at work in the baroque funerary displays in a Florida cemetery. Sullivan demonstrates that religion is too complex, too comprehensive, and too multi-faceted to be adequately captured by the law of the liberal state. This is an important critique, which draws on what I have called the first (critical) and the second (anthropological) approach to religion.

The disaggregating approach helps us answer this critique. First, as religion is not a “thing” but a term of art, it is perfectly legitimateas suggested abovethat different dimensions of it are appropriated in different disciplines and areas of life. To ask that the law embrace and describe the whole of social reality would be to yearn for a totalitarian law. Consider an analogy with marriage. The law does not describe and regulate the full experience of marriage: the fact that, in law, marriage is a contract does not mean that marriage is, or should be experienced as, a contract. Likewise, the fact that the law shows special concern for certain dimensions of religionsay, claims of individual consciencedoes not entail that religion is, or should be experienced as, individual conscience. From this perspective, the egalitarian strategy of analogizing religion is a productive one: the law should not capture “religion” as such, but whatever dimensions of religion are normatively salient and suitable for legal protection. So critics confuse an empirical inquiry (describing what religion is) with a normative inquiry (accounting for its status in the law).

Second, however, we may ask whether the normative analogy used by egalitarian theorists of religious freedom is the right one. Reformulated in this way, Sullivan’s critique has bite. It may well be that egalitarian theorists have picked a normatively problematic, or normatively narrow, analogy for religion. It may well be, for example, that, in their exclusive focus on the claims of individual conscience, egalitarian theorists have neglected the normative importance of other dimensions of lived religion, such as the centrality of traditional collective rituals to the moral lives of believers. Hence the proposed shift from an analogizing to a disaggregating strategy. This involves identifying a plurality of normative analogies for religion. Religion should be disaggregated into a number of different values which relate to the law of the liberal state in different ways. We pick out, from the complex notion of “religion” that we have inherited, distinct elements and values that democratic law has good reason to protect. The disaggregated approach does not claim that it captures the fullness of the lived, anthropological reality of religion. But nor does it reduce the normative inscription of religion in the law to a narrow (often ethnocentric) set of concerns and values.