Naturalism, Normativity, and the Study of Religion

Abstract

This article repudiates the common view that the study of religion, in order to qualify as academic, must be descriptively neutral and naturalistic rather than normative or prescriptive. Following philosophers like John McDowell, John Cottingham, and Tyler Roberts, I claim that such a methodological stance amounts to viewing humans as determined rather than free agents. On the basis of W.V.O. Quine and Donald Davidson’s analysis of translation, I argue that normativity is ineliminable from humanistic scholarship, which is itself inextricable from religious studies. Robert Pippin and Thomas A. Lewis’s readings of Hegel then provide resources to reconcile human freedom and constraint in religion.

1. Normativity and Naturalism

The disciplinary divide between descriptive and prescriptive methodologies in the study of religion is riven by what Thomas A. Lewis has succinctly indicted as a “crucial background assumption”: namely, that “whereas theologians make normative claims, religious studies scholars should refrain from doing so. Rather, scholars in religious studies should distinguish themselves from theologians precisely by striving for some type of distance, neutrality, or objectivity in relation to their subject matter, where this is understood to entail analysis regarding what is rather than claims about what ought to be” (Lewis 2011, p. 169; emphases added).1 Against this assumption that religious studies should refrain from normative claims, I will argue that normativity is an ineliminable (even if often implicit or invisible) element of humanistic description; and that, insofar as religious studies claims to study human agents, it inevitably has humanistic dimensions. The presence and soundness of these elements depend only on the ability of scholars to recognize their subjects’ humanity in the following minimal sense: that human subjects are (just as the scholars that study them) at once naturally or historically conditioned—such that they admit of objective description—and also free agents, such that they are susceptible to normative intervention. And so even if, as Lewis says, prescriptive normativity is most overt in theology, this is not its only site: it is inextricable from humanistic inquiry, I will argue, which is in turn inextricable from the study of religion.2

Although I will elaborate an understanding of normative prescription that is broader than theology proper, I am not concerned to generalize it to all academic discourse tout court. I can clarify my understanding of the descriptive-prescriptive or natural-normative divide by following the philosopher John McDowell in “identifying the natural—as indeed [Wilfrid] Sellars sometimes does—with the subject matter of ‘empirical description’; that is, with the subject matter of a mode of discourse that is to be contrasted with placing something in the normative framework constituted by the logical space of reasons” (McDowell 1994, p. xiv; emphasis added).3 I assume, then, that to do so much as describe a thing empirically—in the sense that is opposed to prescription—is to analyze it according to natural relations. I furthermore assume McDowell’s basically Kantian notion that “to place something in nature on the relevant conception, as contrasted with placing it in the logical space of reasons, is to situate it in the realm of law” (ibid., p. xv)—some might say, in the realm of causal determination that constitutes the objectifiable world4. My main polemical target in this paper is the totalization of the natural, ascendant in religious studies, that McDowell calls “bald naturalism”: the doctrine that “we can reconstruct the structure of the space of reasons out of conceptual materials that already belong in a natural-scientific depiction of nature” (ibid., p. 73), thereby eliminating any need for normative discourse.5 I will argue that insofar as religion is a free human activity—not a purely lawful natural process—its study requires linguistic interpretation, which (as I will further argue) inevitably involves normative dimensions.

McDowell’s work exorcises the temptations of bald naturalism by displaying the “unboundedness” of the normative (more precisely, the conceptual) space of reasons in human experience (ibid., p. 24ff). But my own thesis is not, alas, thus summarily fulfilled by simply observing that all academic discourse (and a fortiori that of religious studies) is engaged in what Robert Brandom calls “the game of giving and asking for reasons” (Brandom 1997, p. 140), and is therefore normative in McDowell’s sense; I will focus on the inevitability of normativity only in certain disciplines, admitting (which is indeed the foil required by my argument) that there do exist bona fide discourses of nature and scientific law.6 Although the people producing these discourses are themselves (one hopes) involved in giving and asking for reasons—no small point to which I will return, since it shows them to be human agents—the technical matters that they discuss might well be devoid of reasons-talk, making reference only to the efficient causes of lawful natural processes (McDowell 1994, p. 70 n.1). That is, while a natural scientist can always be asked questions of the form, “Why does X occur?” (in which case she might produce a theoretical reason for X happening) or “Why do you believe Z to obtain?” (in which case she might produce an evidentiary reason to accept Z as true), reasons will generally not serve as terms in her theory or evidence. Reasons, we might say, are terms in the meta-language of scientific proceedings to talk about theories and their evidence; but reasons are not terms within the languages of the theories themselves. In this paper, by contrast, I attempt to put before us a class of discourse that cannot be purged of reference to the reasons that free agents have for their beliefs and actions. Such discourse would plausibly fall under the rubric “humanistic,” and I will argue that this normative humanistic discourse cannot be expurgated from the study of religion.

2. The Interpretation of Humans

If McDowell’s endgame is to breach the apparent barrier between the natural and the normative, my project here may be viewed as an application and amplification of his insight within the field of religious studies. Consider John Cottingham’s recent call for a “more humane approach” to the philosophy of religion, one that morally involves the philosopher’s own personal commitments: Cottingham proceeds from the claim that although the success of science has inspired naturalistic philosophical approaches emulating it, “neutralist” stances that bracket personal prescriptive commitments are “unstable” in philosophy generally, and especially so in the philosophy of religion (Cottingham 2014, p. 4), in the sense that they cannot be maintained with consistency. I will argue that this critique applies as well beyond the philosophy of religion to religious studies in general.

Before explaining why that should be, I want first to clarify that the claim as I understand it is not of some absolute priority of the normative to the descriptive, but rather of their interrelation. And so we can also endorse the move most resolutely made by modern social scientists and historians from normative doctrine toward interest in actual historical religion as embodied and lived. Indeed, what Cottingham calls his “more humane approach” to the philosophy of religion is based largely on recognition of the importance of these human dimensions in their full-blooded context of culture and praxis, which recognition recommends that “we need to do more than analyse and dissect the truth claims involved [in religious belief]: we need to make a serious attempt to understand the context of culture and praxis that gives life to those claims” (ibid., p. 11). Cottingham thus emphasizes mathesis and askesis—i.e., formation and discipline, respectively—recalling the French classicist Pierre Hadot’s lately popular formulation of “philosophy as a way of life,” in which doctrinal belief forms only part: Spiritual exercises are thus brought to the fore7, opening a potential conversation with thinkers influential in more social scientific quarters such as Foucault with his notion of “technologies of the self”8. From this methodological perspective, interpretation and evaluation of religious thought needs the thick description of concrete life and social formations that social scientists and historians are trained to do.

But by the very same token, understanding religious life also requires evaluation of religious thought beyond a naturalistically value-free account of material processes. Cottingham argues that a posture of value-free academic neutrality actually compromises access to certain kinds of evidence important for robust scholarship on religion. Precisely because human cognition is imbricated within concrete contexts and forms of life—such that belief is tied up with faith and praxis, assent is tied up with institutional commitment and social identity, reason is tied up with feeling and habituation, and so on—description purged of evaluation will fail in its portrayals of religious discourse or practice. This, I will shortly argue, is because belief in religious doctrines is in a mutually constitutive relationship with the forms of life that give them their meaning.9 The point is not utterly to insulate religious thought from examination by non-adherents, but only to widen the field of data under consideration. As Cottingham says, “our human cognitive situation is far more fluid than is implied by ‘flat’ scientific or crudely empiricist models. It is a complex interplay between commitment and withdrawal, affirmation and doubt, yielding and resisting” (ibid., p. 23). On this recognition, a full description of religious phenomena involves not only reading doctrinal formulations, but also understanding how they relate to embodied praxis—and on the other side, not only documenting practices, but also taking into account how agents sometimes ambivalently navigate the claims that the practices and doctrines make upon them. In order to begin to do this, I will now argue, it is incumbent upon the scholar to confront religious utterances and behaviors from a stance that involves more than baldly naturalistic terms abstracted from the normative claims religion might make upon herself.

The most direct way to see this ineliminability of normativity from the description of religious phenomena is to consider what is involved in translation or interpretation. Straight away, of course, some would object to setting interpretation as the relevant goal at all; perhaps what we are after in religious studies ought instead to be explanation.10 The crucial thing to recognize here is that, insofar as we regard our data as intentional action and speech at all—as opposed to mere movement and noise—we are already engaged in interpretation.11 In the words of the philosopher of language whose analysis I will presently exploit: “All understanding of the speech of another involves radical interpretation” (Davidson 2001, p. 125). And this interpetation in turn implicates normativity: McDowell, in his own idiom, says that “our understanding of one another [is] something that must involve ‘space of reasons’ intelligibility” (McDowell 1994, p. 72).

Now, Willard Quine and Donald Davidson have influentially argued that in any attempt at interpretation, we must begin by supposing that we agree on a great many matters with the people we are studying.12 We might assume some agreements only provisionally or hypothetically: for example, I do not believe that the sun is a god; but if I did, then I should call it by the name of a sun-god while chanting and offering it objects; and this might be the only way of making sense of the religious behavior and language of a particular sun-worshipper. Even if we ultimately cannot believe certain things, then, such disagreements can only be picked out against a background of massive agreement about other related matters: for example, that is a sun; this is an object; and I should make offerings to gods that I revere (if any such there be).13 Untethered from some such rudiments of a shared conceptual world, our interpretation would have no place to begin, since anything could mean anything at all. As Davidson says, “Charity is forced on us; whether we like it or not, if we want to understand others, we must count them right in most matters” (Davidson 1984, p. 197). Of course, we can concede to naturalists that all the evidence we have for such concepts and beliefs is empirically observable—specifically, behavioral, either verbal or otherwise, in the context of intercourse with environmental circumstances (Davidson 2001, p. 133).14 But we must furthermore assume many underlying beliefs and at least entertain them as true in order to construe that behavior as the intentional action of conscious human agents, which is what it is to view something as practice or discourse at all rather than mere sound and fury signifying nothing.15

To describe humans as agents is thus to interpret them; and radical interpretation, while assuming at the outset that we know very little about the people we study, only works by the (patently normative) prescription of truth claims.16 As Davidson says, our methodology takes truth as basic and extracts an account of interpretation (Davidson 2001, p. 134), holding belief constant as far as possible while solving for meaning. This is accomplished by assigning truth conditions to alien sentences that make native speakers right when plausibly possible, according, of course, to our own view of what is right. What justifies the procedure is that disagreement and agreement alike are intelligible only against a background of massive agreement. Applied to language, this principle reads: the more sentences we conspire to accept or reject… the better we understand the rest, whether or not we agree about them (ibid., p. 137; emphasis added).

Talk of truth and right belief are the classic marks of epistemic normativity, the prescription of thought that counts as knowledge. And even if one should dispute the normativity of epistemology (as do those who would naturalize it), the act of translation presents a further normative dimension with the necessity of ascribing “our own view of what is right” (per the phrase italicized above) to our interlocutor. Merely to distinguish what someone’s speech can and can’t mean requires some assumption and evaluation of their beliefs by our own standards of truth. We cannot read minds or perform divination; we can only guess what someone could plausibly believe based on what they say and do, with the criteria of fit ultimately being whether we can imagine ourselves saying and doing similar things in similar circumstances under the sway of similar beliefs.

Furthermore, I argue, the evaluation involved in such humanism is not only that of epistemic normativity, concerning what is taken to be true and false; it is normative in the fullest sense, since it requires imagining oneself in another’s shoes, measuring the other’s behavior by one’s own and one’s own by the other’s. The only data supplied by the other are her actions, both verbal and physical; unable to divine her beliefs, we are bound to interpret her behavior in context according to our own standards of appropriate behavior in such contexts. To understand the other, then, one considers not only what must be true and what one ought to believe, but also how one ought to act—which is to say that the normativity involved in interpretation is not only epistemic but also broadly ethical. Quine and Davidson do not (as far as I know) make this explicit, but I think it is entailed by their analysis of translation: Since all of the data available for interpretation of others are ultimately behavioral, we must not only attribute beliefs to our interlocutors but must furthermore hypothesize what difference their beliefs would make for their actions—and initially the only source for such hypotheses is the relationship between our own beliefs and actions. In short, in order to make another’s behavior intelligible, I am bound to imagine how I should behave if I considered certain things true.

I submit this inevitability of imaginative entry into another’s thought and life as the strongest grounding for Cottingham’s “humane or humanistic turn” in the study of religion, which he gives as “a model of understanding that insists that we need to some extent to try to enter into a form of life before we presume to dissect and judge it” (Cottingham 2014, p. 171). On my analysis, such humanistic understanding does not require the adoption of another’s identity as one’s own (an absurdity in any event), or a conversion of the religious outsider to an adherent; rather, the condition of making another human agent’s behavior intelligible is imagining another’s thought in tandem with cognizance of one’s own. This is not an exercise in mind-reading, of course, nor is the other’s mind imagined to float free of their embodied action. The whole point is that to understand another requires examining as much as possible of what they say together with what they do in the context of their material and social environment—which is why scientific study of material culture and social formations remains important—in comparison with one’s own expectations about how to behave and speak in such situations. This normative imbrication of the first-personal with the third-personal qualifies such descriptive-interpretive activity as humanistic in a general, classical sense: In Robert Pippin’s telling, the humanistic tradition at least since Rousseau has been preoccupied with such normative questions17 of both the epistemic and the ethical sort, all of which “are irreducibly ‘first-personal’ questions, and these questions are practically unavoidable and necessarily linked to the social practice of giving and demanding reasons” (Pippin 2009, p. 38).18 To interpret humans—understood as free intentional agents rather than automata—in a scholarly mode just is to engage in humanistic study, which involves the comparison of one’s object of study with one’s own perspective in order to make sense of the other.

Nor should it be thought that privileging the first-personal perspective of the interpreter traps scholars in a solipsistic world or commits us to cultural imperialism. Given those bad options, bald naturalism does indeed seem a palatable third way. Although Quine and Davidson do not (as far as I know) make this explicit either, I would urge that any imputation of one’s own concepts to another ought to involve a potential for self-correction in the face of the other’s concepts: By entertaining foreign claims and practices in the context of their environing circumstances and in comparison with one’s own, one might well come to see how the latter are inadequate for coping with life and even begin to change them. As Thomas A. Lewis says in his recent defense of normativity and self-criticism in religious studies, “Holding normative commitments should not be confused with an unwillingness to revise them” (Lewis 2015a, p. 160). Not only can we maintain revisability as part of a normative stance, we can also see how it is quite diminished by hard-line naturalistic approaches, since they delimit certain conceptual parameters at the outset (namely, those characterizing what qualifies as natural). This is why Tyler Roberts will “suggest that rather than limiting knowledge to the kinds of causes scholars look for when they move to history (for example, historical context) or nature (for example, genes or cognitive mechanisms), we should include the knowledge that results when the humanist moves from freedom to what I have called ‘response,’ knowledge that is existential as well as cognitive, that is articulated and enacted in critical engagement with the behavior in question” (Roberts 2013, p. 116). Accusations that such approaches fail to achieve the academic distance required by good scholarship should be dispelled by Lewis’s espousal of critical normativity with revisability. As Kevin Schilbrack says, “The criterion for what belongs in the academy is not whether one’s inquiries are value-laden—they always will be—but whether those values are open to challenge and critique” (Schilbrack 2014, p. 192). The view of humanistic scholarship that I have laid out requires both normative evaluation and critical response, because to interpret humans as humans just is to be a subject that holds oneself accountable for one’s own thoughts, values, and actions in asking (explicitly or implicitly) what one ought to believe and do in given situations.19

3. Humanism in Religious Studies

Tyler Roberts forcefully applies Pippin’s notion of humanistic scholarship to the special case of academic religious studies in his book Encountering Religion (Roberts 2013, esp. pp. 90ff.). Lest my opposition of naturalism to normativity look like a caricature abstracted from the way any scholar actually proceeds, consider how Roberts astutely reads influential theorists of religion to reveal that one of the decisive fault lines between naturalistic and humanistic approaches in the study of religion is indeed whether human phenomena are analyzed primarily in terms of causal compulsion rather than free agency and responsibility.20 Whereas sound descriptive work is often thought to rely on a view of people as socially and historically determined—echoes of Kant—prescription requires the notion of freedom. When we see the disciplinary divide in this way, it becomes clearer how a scholar located in secular religious studies can resist naturalism: not to favor the supernatural, but rather to maintain human agency in the face of baldly naturalistic causal explanation.21 Leaving aside the many philosophical tomes that have been written to nuance the supposed dilemma of freedom and determinism, I will follow Roberts’s suggestion that we think about religion and religious studies instead in terms of “a complex dynamic of dependence and freedom, a freedom made possible in and through certain practices of inheritance, commitment, and inhabitation” (Roberts 2013, p. 106).

Roberts’s suggestion is well in line with how McDowell proposes we make sense of the human condition with respect to the space of nature and the normative space of reasons. McDowell urges us to recognize that humans are not only animals, creatures of nature, but also have a normative “second nature” imparted through languages and traditions, upbringing and Bildung (McDowell 1994, pp. 87ff). This last term is, of course, borrowed from nineteenth-century German philosophy; and, after Aristotle, it is Hegel that McDowell sees as proposing the best reconciliation of the nature-normativity dichotomy—although, not being a Hegel scholar himself, he defers the details of the solution to scholars like Pippin (McDowell 1994, p. 111). I will now follow the suggestions of Roberts and McDowell to see how recent Hegel scholarship (especially that of Pippin and Lewis) can be used to think about religious behavior as reconciling freedom and determination, thereby incorporating normativity into description of religion.

4. The Hegelian Synthesis of Freedom and Conditioning in Religion

To start with, Pippin’s Kantian reading of Hegel strongly brings out the relationship between freedom and normativity: On this reading, the very possibility of objective rationality rests on the self-constitution of concepts free of determination by nature. That is, for Pippin’s Kant and Hegel both, to have knowledge of objects is to be in the possession of concepts that are prescriptively able to determine what a thing is, which is to say, concepts that are epistemically normative.22 Now, I’ve already made the point that the normativity involved in understanding another is not only epistemic, but also broadly ethical. Similarly, Hegel’s notion of self-determination generalizes Kant’s language of spontaneity beyond epistemology into the practical sphere (Pippin 1989, p. 153). Like Kant’s spontaneity of conceptual understanding—the freedom from external constraint that is tied up with self-consciousness—freedom writ large, for Hegel, is self-conscious self-determination (Lewis 2005, p. 39).

But this self-determination is not arbitrary, even for Kant; and for Hegel, it occurs in accord with the agent’s particular constitution, which is itself a result of history and upbringing. This notion of historically conditioned self-determination is then, as Lewis points out, a confluence of causal determination and freedom.23 We are thus returned from the Kantian a priori realm of pure rational normativity back into the embodied, historical world of social relations, negotiations, and traditions—ostensibly the stuff of social scientific data. Indeed, Hegel had his own anthropology; and while his was a far cry from contemporary social science, it can be seen as a sort of reconciliation of natural history and transcendental philosophy. Lewis usefully mines Hegel’s anthropology to show how it might be possible to think freedom and tradition together, building upon what he calls the "nontraditional" reading of Hegel that emphasizes continuities with Kant and minimizes additional idealist assumptions. This reading has been very influential in philosophical circles and Hegel scholarship in recent decades and stands to be a valuable contribution to religious studies, which still tends to assume that Hegel’s project was a highly metaphysical, cosmically monistic one (Lewis 2015b, pp. 198–201) offering little comfort to attempts at rapprochement between naturalism and the humanities. The nontraditional Hegel offers a form of naturalism that makes the human mind central, escaping between the horns of a putative dilemma between the material world and an ideal world transcending it (Lewis 2005, p. 206).24

According to Lewis’s reading of Hegel’s anthropology, the human first begins to win liberation from immediate physical determination through habits, the cultivated disciplines that allow us to self-regulate in the face of instinctive sensual drives (Lewis 2005, p. 55). Lewis’s discussion of habit brings out why religion is especially important for my discussion: Although habit also includes all “ethical aspects of our way of life” and our “concrete existence” generally (ibid., p. 57), Hegel views religion as the “principal bearer of the most significant appropriated habits” (ibid., p. 187).25 But habitual action is, by definition, not very self-conscious action, and so can’t count as fully free—we often blame the force of habit for our unthinking, non-intentional actions. How can one become free? It requires a further process of development, reflection, critique, and transformation of one’s habits.26 And so while full autonomy is in some way freedom from habit, it is also constituted by habit.27 Freedom entails the ability to be critical of one’s own habits, which in turn entails having appropriated those habits in the first place, such that they are one’s own and help to constitute oneself.28 Religious discourse and practice, as the foremost vehicles of habit and discipline, are therefore simultaneously products of social historical conditioning—and thus subject to naturalistic description as I have conceived of it—as well as constitutive of freedom—and therefore subject to prescription (both as agent and patient, source and target of normative intervention). Put simply, “religion is not mechanically introduced into human beings but lies within them, in their reason and freedom generally” (ibid., p. 307). Albeit historically conditioned, religion admits of engagement within the normative space of reasons, in the Sellars-McDowell idiom.

Although there is some rather arcane Hegelianism that works out the details of this result29, it essentially depends on two very intuitive premises: that we are naturally and historically conditioned, and that we are also free. The only serious question is how to hold these apparently opposing premises together; and I have introduced Hegel to suggest that what might initially look like mutually exclusive or even conceptually contradictory options in this opposition of conditioning to freedom are in fact intelligibly compatible. I’ve drawn this dilemma into relief for the sake of vividly displaying one of the fault lines in contemporary methodological debates, but my overarching picture of simultaneous freedom and conditioning is not particularly exotic; after all, many of our most enduring social theories of religion—from Weber’s and Durkheim’s to Peter Berger’s (which is supposed to be a synthesis of the first two) and all but the most vulgar Marxisms and Freudianisms—present a picture of this sort. Berger (1967, chp. 1) presupposes a dialectic between the social and material products of (individual and collective) human action and the feedback effects of these products on humans: “The two statements, that society is the product of man and that man is the product of society, are not contradictory” (ibid. p. 3). While some might hear my argument as an attack on social scientific methods, it is not so: The thesis of the ineliminability of normativity from religious studies, tied up with the intuitive suppositions that humans are both free and conditioned, is quite consistent with the mainstream of sociological thinking about religion.

What is crucial to see is that these common suppositions apply to all humans: While religion is a paradigm of habituation, science and philosophy and all cultural activity come within the ambit of habit30. All of these social forms must be historically conditioned; and scholars, at least, need to be able to claim freedom and normativity for their own methodologies and concepts. As McDowell plausibly expresses the Kantian view, “the space of reasons is the realm of freedom” (McDowell 1994, p. 5); and scholars are presumably in the reason-giving business31 and thus ipso facto free. Now, as we have seen, those who want to purge normativity from the study of religion need to view religious behavior as more automatic than autonomous—since that’s the only way to think it describable in terms of natural causality without free agency—even as they assume their own scholarly habit to be more free than determined. Not only does such an asymmetrical anthropology look rather presumptuous, I have argued that it undermines the very possibility of interpretation of language. Fairness aside, sheer consistency and intelligibility demand that human scholars recognize the humanity they share with religious people32; and such scholarship must inevitably then be allowed humanistic dimensions, which are irreducibly normative ones. The price of value-free scholarship is either our own humanity or else the ability to make intelligible the humanity of the people we study.