Cognizing the Supernatural & Religion

Over at Religion Bulletin, Kenneth MacKendrick has posted a densely packed piece that should be required reading for all cognitive theorists of religion. It resonated with me, and stung just a bit, because as someone who grants that human brain-minds evolved in a way which results in the natural or spontaneous generation of “supernatural” concepts, I haven’t fully theorized and critiqued my usage of that term. My past practice has usually been to scare-quote “religion” (but not supernatural) as a way to signify recognition that religion is not a natural kind or category but is a contingent social construction with a history.

In concrete terms, I have typically bifurcated my usages — calling supernaturalism before the Neolithic transition “animism” or “shamanism,” and supernaturalism thereafter “religion.” Even this doesn’t quite do, given that religion conceived as something distinct, set apart, private, personal, an institution, a tradition, faith, or a matter of belief, is an altogether more recent development. But at some point, we need to talk-write about whatever it is we purport to study and some kind of shorthand is needed. Thus, while aware of anti-essentialist critiques, I eventually (and perhaps lazily) fall in with Ake Hultkrantz and get on with my “supernatural” business. This has always seemed a natural thing to do, though I now realize that this seeming naturalness may flow, uncritically, from a cognitive or evolutionary approach (which has always been my starting point).

But as MacKendrick disruptively explains, this approach runs the risk of rendering “religious cognition” as something sui generis — untouched by the messy realities of experience and learning:

Cognitive theories usually define religion as having to do with thoughts and practices related to the supernatural. The supernatural is vaguely conceived but focuses on relations with supernatural agencies. The acquisition and transmission of religion is seen as related to how the mind works: our understanding of other minds, bias to detect agency, gravitation toward purpose-based explanations of origins, dualistic notion of body and mind, acceptance of non-natural causality, and memorable attraction to counterintuitive representations. All of these components are thought not only to render the mind susceptible to religious ideas but also to spontaneously generate them. Furthering this view, scholars focusing on cognition in light of an orientation toward the supernatural often build into their theory of religion the notion that religion is an adaptation or byproduct of a process of natural selection. Since in each instance the mind is hardwired for creating, acquiring, and transmitting supernatural representations, religion is deemed as eminently natural (“strong naturalness thesis”). The naturalness of religion readily follows from the adoption of a substantive definition of religion.

However, if we adopt a heuristic definition of religion, even when religion is defined in a similar manner, it does not follow nor could it follow that religion is natural. “Religious cognition” makes about as much sense as “49.8833° N, 97.1500° W cognition.”  The difference between the two is startling, and cognitive theories of religion have yet to consistently clarify the ambiguity.

The problem emerges most clearly when we look at how the strong naturalness thesis re-describes cognitive development in religious terms. For instance, the strong naturalness thesis posits that as theory of mind skills develop they develop in tandem with promiscuous agency detection. Out of this matrix of others in mind is born a special form of thinking, “religious cognition,” the positing of or willingness to accept representations concerning supernatural agencies. When this theoretical impulse to “religionize” cognition becomes systematic it becomes fairly easy to cherry pick contributing elements and identify the origin of religion in evolutionary history as well as in childhood….

Insofar as a cognitive theory of religion engages in a re-description of cognitive development in religious terms it is not really a cognitive theory of religion at all. It would be more accurate to say that the strong naturalist thesis is actually a religious theory of cognition, a sui generis conception of religion – a religious re-description of cognition.

This is a key point, one often missed or ignored by cognitive theorists of religion. I have long been critical of stand-alone cognitive accounts of religion because, had they been tested with actual historical cases or against the ethnographic record, it would be apparent that the supposedly universal neurobiology giving rise to religion is a product of time and place. I just encountered an example of this in Why God Won’t Go Away: Brain Science and the Biology of Belief (2001) by Andrew Newberg and Eugene D’Aquili. The authors provide a perfectly plausible account of the brain functions and neural mechanisms associated with the (mystical) experience of transcendence, but mistakenly assume that transcendence — a dissolution of self or ego into “Absolute Unitary Being” — has always and everywhere been the goal of religion (or what they metaphorically call God). Unsurprisingly, their examples of this allegedly universal religious yearning are drawn almost exclusively from Axial traditions. The authors are unaware of the fact that such traditions arose in response to political, economic, and social conditions that made transcendent experiences attractive and construed them as “religious.” It never occurs to Newberg and D’Aquili that forgetting or dissolving one’s self into something larger might be the result of a social atomism and alienation congruent with new kinds of (Neolithic) societies, inequality, disease, warfare, and slavery. It never occurs to them that hunter-gatherers, living unalienated among kin in tight-knit societies, have little or no interest in transcendence.

But back to MacKendrick’s point. He is rightly critical of free-floating religious cognition, unencumbered by history or socialization:

This clear identification of the origins of religion in natural history as well as developmental history threatens to become a sui generis discourse when it (unnecessarily and illicitly) replaces more compelling, systematic, and historically viable accounts of cognition. In other words, cognitive theories of religion, by relying on a troubling and contested definition of religion without reference to its historical continuities and political implications, create and foster an ideological posturing that, while appearing to be interdisciplinary and scientifically minded, is conceptually anti-historical and ahistoriographical.

For instance, when the history of representation is described as the emergence of myth through the manipulation of symbolic forms, and hence the origin of religion, this account obscures the more accurate and plausible account of development: that the history of representation allows us to chart the emergence of the imagination (not religion). Re-describing the emergence of the capacity to pretend, to act as though the world is as it is not, as religious is profoundly misleading. It lends itself, for instance, to a problematic account of ritual as distinct from pretense and agency detection as distinct from processes related to the development of communicative competence.

It is no wonder, as outlined in a forthcoming essay by Josh Rottman and Deborah Kelemen, that very little (if any) evidence can be found for the existence of “religious beliefs” in early childhood. As [they and others have shown], the vast majority of evidence for the acquisition of religious thoughts and practices emerges only after individuals are socialized into such practices. There is scant support for spontaneous religiosity, a point that would not surprise an historian of religion but seems to threaten an overthrow of several of the primary tenets of popular cognitive theories of religion.

Cognitive theories, in other words, slide too easily into the neural-numinous. This slippage is attributable, in part, to the promiscuous proliferation of “supernatural” agents and agencies, which are imagined by cognitivists to be everywhere. These agents and agencies, in turn, are attributed to the minds of people present at the proverbial-primitive beginning, as a superstitious kind of “proto-religion” that later became religion. MacKendrick is skeptical:

I understand the motive to define religion in this way to be the rather embarrassingly fuzzy idea that supernatural agents are postulated all around the world. Supernatural agency, when globalized in this way, is a bit of a will-o’-the-wisp. It is very beguiling but it can mean almost anything. Guided by Carol Nemeroff and Paul Rozin (2000) I have found it helpful to interpret “the supernatural” as that which “generally does not make sense in terms of the contemporary understanding of science.” Of course what makes sense in terms of contemporary science is itself a moving target, often contested, and inordinately fallible. Given the transitory and culturally contoured nature of “what makes sense” it is even more urgent to be cautioned by historiographical reflection.

Though I have on occasion been guilty of overpopulating the Paleolithic with imaginary agents, this characterization of the supernatural accords with at least some anthropological understandings. As I noted in this post, Plains Indians did not recognize the physical/metaphysical dichotomy that characterizes Western thought, but they “can and did react vehemently to perceptions that are wholly out of the normal range of experience.” These were things that struck them as “mysterious, weird, or miraculous, thrilling or awe inspiring.” Nearly every tribe had an umbrella word to describe such perceptions; for the Lakota (Sioux) it is “wakan” and for the Crow (Apsaroke) it is “maxpe.” There is also the famous Algonkian (Winnebago) “manitou.”

In a later post, I made further amends for (cognitively) overpopulating others’ imaginations with supernatural agents-agencies by considering animism as a form of relational epistemology, one which fits within a larger cosmic economy of sharing. Whatever good this did was probably undone by my further characterization of this as altruistic and perhaps even adaptive.

Here things must come to a rambling end. For his part, MacKendrick closes with this palliative: “unless there is a mindfulness of the ambiguity and ideological history of the term religion, the proliferation of its supposed naturalness will ultimately foreclose upon its richness and explanatory potential.” I agree and would add only that unless there is a mindfulness of the ambiguity and ideological history of the term supernatural, the proliferation of its supposed naturalness will ultimately foreclose upon its richness and explanatory potential.

Did you like this? Share it:

7 thoughts on “Cognizing the Supernatural & Religion

  1. Dominik Lukes

    Thanks Cris, your posts always point in interesting directions. Love the cartoon.

    However, I’d like to return to my comment on your universals post regarding the untenability of ‘supernatural’ as an explanatory category.

    I read with interest Hultkranz’s suggestion for an empirical basis for the concept but I think there are still problems with this view. I don’t see the warrant for the leap from “all religions contain some concept of the supernatural” to “supernatural forms the basis of religion”. Humans need a way to talk about the experienced and the adduced and this will very ‘naturally’ take the form of “supernatural” (I’m aware of McKinnon’s dissatisfaction with calling this non-empirical).

    On this account, science itself is belief in the supernatural – i.e. postulating invisible agents outside our direct experience. And in particular speculative cognitive science and neuroscience have to make giant leaps of faith from their evidence to interpretation. What are the chances that much of what we consider to be givens today will in the future be regarded as much more sophisticated than phrenology? But even if we are more charitable to science and place its cognition outside the sphere of that of a conscientious sympathetic magician, the use of science in popular discourse is certainly no different from the use of supernatural beliefs. There’s nothing new, here. Let’s just take the leap from the science of electricity to Frankenstein’s monster. Modern public treatments of genetics and neuroscience are essentially magical. I remember a conversation with an otherwise educated philosophy PhD student who was recoiling in horror from genetic modification of fruit (using fish genes to do something to oranges) as unnatural – or monstrous. Plus we have stories of special states of cognition (absent-minded professors, en-tranced scientists, rigour of study) and ritual gnostic purification (referencing, peer review). The strict naturalist prescriptions of modern science and science education are really not that different from “thou shalt have no other gods before me.”

    I am giving these examples partly as an antidote to the hidden normativity in the term ‘supernatural’ (I don’t think that saying you’re not using it normatively is enough to remove that connotation) but also as an example of why this distinction is not one that relates to religion as opposed to general human existence.

    However, I think Hultkranz’s objection to a complete removal of the dichotomy by people like Durkheim and Hymes is a valid one as is his claim of the impossibility of reducing it to the sacred/profane distinction. However, I’d like to propose a different label and consequently framing for it: meta-liminal. By “meta-liminal” I mean beyond the boundaries of daily experience and ethics (a subtle but to me an important difference from non-empirical). The boundaries are revealed to us in liminal spaces and times (as outlined by Turner) and what is beyond them can be behaviours (Greek gods), beings (leprechauns), values (Platonic ideals) or modes of existence (land of the dead). But most importantly, we gain access to them through liminal rituals where we stand with one foot on this side of the boundary and with another on the “other” side. Or rather, we temporarily blur and expand the boundaries and can be in both places at once. (Or possibly both.) This, however, I would claim is a discursively psychological construct and not a cognitively psychological construct. We can study the neural correlates of the various liminal rituals (some of which can be incredibly mundane – like wearing a pin) but searching for a single neural or evolutionary foundation would be pointless.

    The quote from Nemeroff and Rozin that ‘“the supernatural” as that which “generally does not make sense in terms of the contemporary understanding of science.”’ sums up the deficiency of the normative of crypto-normative use of “supernatural”. But even the strictly non-normative use suffers from it.

    What I’m trying to say is that not only is not religious cognition a special kind of cognition (in common with MacKendrick), but neither is any other type of cognition (no matter how Popperian its supposed heuristics). The different states of transcendence associated with religious knowing (gnosis) ranging from a vague sense of fear, comfort or awe to a dance or mushroom induced trance are not examples of a special type of cognition. They are universal psychosomatic phenomena that are frequently discursively constructed as having an association with the liminal and meta-liminal. But can we postulate an evolutionary inevitability that connects a new-age whackjob who proclaims that there is something “bigger than us” to a sophisticated theologian to Neil DeGrasse Tyson to a jobbing shaman or priest to a simple client of a religious service? Isn’t it better to talk of cultural opportunism that connects liminal emotional states to socially constructed liminal spaces? Long live the spandrel!

    I cannot recommend Discursive Psychology enough as a framework for the study of religion:

  2. Cris Post author

    Thanks Dominik. I need to give your posts some more thought, and will write more later, but now that I know you are a liminal-theoretical-linguist coming at these issues from the perspectives of Geertz, Turner, Lakoff, and Gould, I think I can address some of your well-stated and multi-perspectival concerns. As for those I probably won’t be able to address, at least to your satisfaction, it probably has to do with the fact that I don’t think science, writ large (or what I call positivism), is just another discourse on a continuum with what I call the supernatural (which I do think I can use without normative connotations for reasons I will later explain). I think there is a difference between the two, which may become fuzzy and perhaps even insensible near the boundaries, but this is in the very nature of boundaries. Without them, there isn’t really much we can say to anyone’s satisfaction. As someone who spends a great deal of time studying evolutionary biology and neurology, I think we know more about the human mind than you seem to think (and far less than scientists would have us believe). I’m not satisfied simply giving up, or getting lost in language, and declaring it all a fictitious mess, or a giant socialized spandrel, as the case may be.

  3. J. A. Le Fevre

    Thanks, another most excellent post.

    As suggested by Warren Zevon:
    But we had to take that long, hard road to see where it would go.
    We took that holy ride, ourselves to know.

    Now if you make a pilgrimage I hope you find your grail.
    Be loyal to the ones you leave with even if you fail.
    Be chivalrous to strangers you meet along the road.
    As you take that holy ride, yourselves to know.
    You take that holy ride, yourselves to know.

    Why does religion look like it does, and so universal yet so diverse? To understand that is to understand ourselves.

  4. Dominik Lukes

    Thanks for the response. I agree fully on the danger of, as you say, “getting lost in language”. You’d be surprised to hear me talk to postmodernists or phenomenological philosophers. There are real things that can be said (provided we are aware of the limitations of the medium).

    And I leave open the possibility that within science, there is a different kind of knowledge (that was, after all, my starting point, I was converted to my stance by empirical evidence).

    Also, I have enormous respect for both neuroscience and evolutionary biology (although I only follow them from a distance).

    My objection (and that is perhaps why I’m a bit more extreme in my formulations than I feel) is the essentialist reductionism that so often comes from these fields. I wrote about my objections in detail here

    Very often neuroscientists (and cognitive psychologists) approach my fields (linguistics, anthropology) as if they were there to be explained away. This is based, more often than not, on an outsider’s view of the evidence (Pinker in his latest book is a great example. Jared Diamond is another.) The reason I like your blog is that you’re neither essentialist nor approach the fields in question as alien territories.

    This is more than just protecting my turf (I hope). Describing the neurological or evolutionary correlates of social phenomena is fine. But I almost never see such explanation contribute to existing understanding. It’s just describing the same phenomenon from a different perspective but not really (most of the time) providing an “explanation”. It broadens our understanding but does not deepen it (if you forgive the wordplay).

  5. Pingback: Religion, if it exists, is negotiation of underdetermined metaphoric cognition [UPDATED] | Metaphor Hacker - Hacking Metaphors, Frames and Other Ideas

  6. Kenneth MacKendrick


    Your generous reception of the essay is much appreciated! I’m quite supportive of your comments. I certainly agree that the concepts “natural” and “supernatural” need to be contextualized. Both seem to readily fall into our laps fully formed and without historical baggage. We should be wary when this happens (Tomoko Masuzawa makes a similar observation about the phrase “world religions”).

    One of the primary issues that I have not seen addressed in much detail concerns a cognitive-theoretic based critique of metaphysics. The critique has been taken up time and time again from numerous philosophical circles, but it has yet to be taken up in terms of a cognitive theory of nature or a cognitive philosophy of nature. What would this look like? It would have to take the form of a philosophy of science, a critical discourse concerning issues of method and theory in the study of “religion” that brings to the foreground concept formation alongside self-reflection in the sciences.

    The most promising path has been developed by Jürgen Habermas. I’m very much in agreement with Habermas’s call for postmetaphysical thinking – a kind of thinking refuses to relinquish metaphysics but is fully cognizant of the dangers of a nostalgic return to metaphysics as well as the radical rejection of metaphysics ( see his book Postmetaphysical Thinking). Habermas, in several places and philosophical conversations, has developed a realist understanding of nature that is rooted in the critique of metaphysical systems without becoming anti-metaphysical (and thus reproducing another metaphysical system). His emphasis on argumentation as the basis of all scientific activity, which is simply an intensification of everyday communicative action, shifts the normal subject-object dialectic of traditional science to a more subject-subject or intersubjective orientation. Once we begin to think in more intersubjective terms it becomes apparent that the boundaries we draw around our precious concepts are embedded in historical contingencies. “Nature” is no more natural than “karma” or “theism” or “dreamworlds.” A sustained critical-reflective emphasis on concept formation is one of the only ways we can address this issue.

    Thanks again for providing the opportunity for a rejoinder!

    Best, Ken

  7. Kenneth MacKendrick

    Dear Dominik Lukes,

    Cris recommended your comments to me – which I liked very much. You might have a look at Ann Taves, Religious Experience Re-Considered. Taves draws on the notion of “specialness” in a way that might be very similar to your meta-liminal. Taves adopts the concept from Jesper Sorensen, A Cognitive Theory of Magic. There is also the very interesting notion of cognitive fault lines, which has to do with how the mind carves up the world. It turns out, according to Ruth Byrne’s The Rational Imagination, that when we are asked to come up with hypothetical alternatives to a described situation, nearly everyone comes up with the same or similar response (e.g. if a driver, taking a non-typical route home is killed in a car accident by a drunk driver, almost everyone says “should have taken the normal route home” as opposed to any other possible alternative “the other driver shouldn’t have been drunk” “should have left work early” etc etc). (not exactly everyone, but there is a remarkable similarity in responses). Thanks for the insights!

Leave a Reply