It is a testament to the power of the “strong program” image that most commentators on our working paper read Matt May and me to be optimistically praising its emergence in the sociology of religion, despite our statements to the contrary. Of course, a writer criticizing readers is bad form, and truth be told, we deeply appreciate the commentators’ willingness to discuss a working paper whose positions and prose are not yet entirely solidified. Our original title had “a critical engagement” as its subtitle; leaving it out probably didn’t help communicate our intent. If we add to this the positive connotations of the term “emerging,” we can certainly understand how commentators saw us as identifying a wave we were preparing to surf.

Undoubtedly there is, as Bryan Turner suggested, a code that gives the idea of a “strong program” a positive normative charge. Let’s take pause to understand what this is about. What we call the “strong program” in the sociology of religion refers to a perspective that focuses on religion as an autonomous phenomenon that has causal impact, rather than something that is determined by non-religious factors. Apart from the clearly normative binary of strong/weak, what is the attraction of this image?

First, for people of faith, the autonomy of at least some religion is a necessary but not sufficient condition for the reality of the supernatural, and thus is a logical analytic goal. Indeed, as Asad and others have argued, the carving off of a domain of social reality as “religious,” autonomous, and separate from other, “secular” domains was precisely a mechanism by which the early modern Church was able to maintain a space for religious authority vis-à-vis encroaching secular authority. Likewise, Courtney Bender has recently argued that the residual categorization of religious experience as ineffable, pre-cultural, and inexplicable extends from attempts of early twentieth-century scholars to carve off a domain of human experience that would not be susceptible to scientific analysis. We should not be surprised that this is an enduring motivating interest in the scientific study of religion.

Second, at least since Kant, the idea that human beings give form to the world, rather than simply being determined by it, has been one enduring basis of the idea of human freedom. And for scholars who, regardless of whether they have faith, see the concept of human freedom as a cornerstone of human dignity and morality, the irreducibility of religion is an important image. Christian Smith’s work on “moral, believing animals,” for example, clearly works in this direction, as do Charles Taylor’s writings on the self and religion.

Finally, and perhaps most importantly, the idea of a phenomenon’s autonomous reality provides a time-honored foundation of legitimacy for a discipline’s professional activity. If there is a domain of knowledge dealing with X, it is most obviously in the interests of specialists in that domain to underline and drive home the reality and importance of X. Ferdinand de Saussure’s Course in General Linguistics has become the seminal text in linguistics precisely because it succeeds in portraying language as an irreducible formal system of signs beneath the messy details and disorder of actual speech. Emile Durkheim sought to create a foundation for sociology in turn-of-the-century France by arguing that society was a reality sui generis that needed a new discipline to study it. Talcott Parsons sought to do the same in the U.S. context through his thesis that scholars from different disciplines and countries had simultaneously and independently converged on the “voluntaristic theory of action,” in which values and norms were irreducible. And most recently, Jeff Alexander has largely succeeded (if we judge by the burgeoning numbers in the ASA Culture Section) in creating a foundation for cultural sociology by arguing that culture is an autonomous phenomenon that has causal power.

But I would like to suggest that the “strong program” is actually a weak model for where we should be going in the sociology of religion, for one negative and one positive reason. First, while a healthy sub-discipline probably does depend on studying a phenomenon that actually exists, the politics of representation also needs to be taken into account. In his description of the religious inclinations (or disinclinations) of various social classes and strata, Max Weber argued that there was an elective affinity between the position of intellectuals (such as priests, theologians, and scholars) and the rationalization of religion. Of course, in sedentary societies there will always be “religiously musical” individuals who become specialists in thinking through and logically organizing ideas regarding the supernatural. But they also thereby create a role for themselves as theological interpreters, and thus have a rational self-interest in emphasizing the importance of logically coherent religious thought. This rational self-interest becomes a political interest insofar as it simultaneously dis-empowers people who do not engage in rationalized religious practices. When having a “moral order” is considered a fundamental component of human nature, then those whose religious practices (or lack of them) appear eclectic and inconsistent become less-than-human “others.” When “true” religion is considered autonomous and disinterested, then people whose religion is oriented towards practical interests and engaged in everyday life are portrayed as insincere and vacillating, and their religious practice as inauthentic and unsustainable. We sociologists of religion need to soberly realize that our structural position is going to lead us time and again to emphasize the sui generis reality, coherence, and irreducibility of our subject matter; and need to have enough self-reflexivity to realize that this may unduly impact our analysis, and in ways that in turn may unduly impact people.

Second, arguments about the autonomy of religion should not dominate our research, even though they legitimately remain of interest for some, including those who feel their faith threatened by science, or for those interested in neo-Kantian arguments that underline humans’ freedom by pointing to their form-giving capacity. Beyond these topics, the “autonomy of religion” issue remains of limited interest in the larger debates of the discipline. Indeed, while the main point picked up in discussion of the working paper was the assertion of vitality in the sub-discipline, I think John Evans is right to suggest that the trend line depicting articles on religion in major journals should be read as flat. At a minimum, given the growing public interest in religion over the past two decades, I think we need to ask why this upward trend is not more impressive.

Instead of trying to continually prove that religion matters, we should take the “stronger” starting point that “of course religion matters,” and simply concentrate on what it is and how it is involved in contemporary social and political issues. Not “why does it still exist?” but “how does it exist?” “how does it relate to its ‘others’?” “how does it affects people’s lives?” and, of course, “who creates it?” “who has the control of its means of production?” “who has an interest in its moving in this direction or that?” In such an approach, religion can plausibly be either cause or effect (or non-causal), and either good or bad (or neutral). Such a robust engagement of the problems of modernity is what will make the sociology of religion a vital subfield and contribution to our social world.