Mixing Memory

This was originally posted on the old blog on 1/5/05. I’m reposting it here, with a few editorial ommissions (contextual; references to things from back then that won’t make sense here), because of our recent discussion of religion. Hopefully I’ll be able to post about some of the empirical work over the next few days, including a study on memory and religion published by the authors of the theory described in this post and their colleagues.

A Comprehensive Theory of Religious Cognition

The cognitive science and neuroscience of religion have become hot areas of research over the last several years, so there’s a wealth of theory and data to draw on. So, if you’re interested in work on religion from a cognitive perspective, expect two or three posts in the near future. A common theme that runs through the recent cognitive literature is that, contrary to the way scientists and philosophers have traditionally viewed religion, its cognitive aspects may actually be fairly mundane. As Justin Barrett writes1:

Are god-concepts much different from gorilla-concepts? Is performing a religious ritual a profoundly different action from sending a greeting card to a friend? Perhaps not. When considering the kind of cognitive resources required for representing and acquiring these concepts and actions, the sacred and the profane may be less discriminable than is commonly assumed… Much as language is naturally acquired as a result of cognitive preparedness plus exposure to a typical sociolinguistic environment, ordinary cognition plus exposure to an ordinary environment goes a long way towards explaining religion. (p. 29)

In this post, I want to focus on one particular theoretical framework for understanding religious cognition, as described in Scott Atran’s In Gods We Trust: The Evolutionary Landscape of Religion and Atran and Norenzayan’s “Religion’s evolutionary landscape: Counterintuition, commitment, compassion, communion” (A&N). Most of the quotes will be taken from the latter, because, well, it’s online, so I can cut and paste them. In subsequent posts, I’ll focus more on the cognitive and neuroscientific research.

It’s important to remember that Atran and other psychologists, anthropologists, and neuroscientists who are studying religious cognition are studying ordinary religious beliefs and concepts, and not the complex concepts that you might find in formal presentations of the theologies of various religions. The latter are carefully worked out and studied, and may differ in many ways from everyday religious concepts. For example, theological reprsentations of God may present Him as existing outside of time, whereas everyday God concepts may treat him as experiencing time in ways similar to humans2). While theological concepts of God will have to utilize ordinary cognitive mechanisms, their representational and logical differences may result in the utilization of different mechanisms, or the same mechanisms in different ways. The theoretical and empirical resesarch on religious cognition is not designed to explore the cognitive foundations of theological concepts. Therefore, application of the research I will describe to theological problems would be a mistake. Furthermore, the research I will describe says nothing about the truth or falsity of any set of religious beliefs.

As the title of Atran’s book and paper suggest, he and his colleagues approach religion from an evolutionary perspective. Their primary insight is that religion is “a converging by-product of several cognitive and emotional mechanisms that evolved for mundane adaptive tasks” (A&N, p. 4). They use this insight, along with our knowledge of these “mundance mechanisms” from research on nonreligious cognition to explain four universal properties of religion:

  1. Widespread counterfactual and counterintuitive beliefs in supernatural agents (gods, ghosts, goblins, etc.).
  2. Hard-to-fake public expressions of costly material commitments to supernatural agents, that is, offering and sacrifice (offerings of goods, property, time, life).
  3. Mastering by supernatural agents of people’s existential anxieties (death, deception, disease, catastrophe, pain, loneliness, injustice, want, loss).
  4. Ritualized, rhythmic sensory coordination of (1), (2), and (3), that is, communion (congregation, intimate fellowship, etc.) (A&N, p. 3).

With respect to (1), the belief in”supernatural agents,” many psychologists believe that humans have a “hair-triggered” agency-detection mechanism. This seems to develop in the first year of life, and though it tends to be very overextended in early life (children attribute agency to things like stuffed animals), even adults tend to extend agency to inanimate objects under uncertainty. Thus, A&N write:

From an evolutionary perspective, it’s better to be safe than sorry regarding the detection of agency under conditions of uncertainty. This cognitive proclivity would favor emergence of malevolent deities in all cultures, just as the countervailing Darwinian propensity to attach to protective caregivers would favor the apparition of benevolent deities. (p. 22)

To illustrate this, they give the following example from Lipkind4:

The earth and underworld are inhabited by supernaturals. . . . There are two kinds. Many are amiable and beautiful beings who have friendly relations with humans. . . . The others are ugly and dangerous monsters who cannot be placated. Their woods are avoided and nobody fishes in their pools. (p. 249)

Many religious rituals take advantage of this “hair-triggered” tendency to perceive agency by creating conditions in which agency will be attributed, such as the use of masks, or the detection of religious figures in inanimate objects (e.g., the Catholics who flocked to the Nashville coffee shop that I frequented in high school, because a pastry that resembled Mother Theresa had been discovered there). These intentional and uninentional recreations of the circumstances under which agency is usually detected can serve several purposes. A&N write:

Such manipulations can serve cultural ends far removed from the ancestral adaptive tasks that originally gave rise to those cognitive and emotional faculties triggered, although manipulations for religion often centrally involve the collective engagement of existential desires (e.g., wanting security) and anxieties (e.g., fearing death). (p.24)

In addition to activating ordinary agency concepts, supernatural agents and the worlds in which they exist also create counterintuitive or counterfactual conditions. Humans appear to have an “intuitive ontology,” that contains the ordinary, factual entity-types that we encounter on a regular basis, such as persons, animals, and artifacts. We have certain beliefs about the nature and behavior of these different intuitive ontological categories, and supernatural agents are often designed in such a way that they violate these intuitive beliefs in important, but minimal ways. For instance, most ordinary God concepts contain both counterfintuitive information (e.g., omnipotence) and intuitive information associated with persons (e.g., emotions such as jealousy or anger). The counterintuitive aspects of supernatural agent concepts cause them to command attention, while the connection to established person concepts makes them easier to remember and transmit.

A third important property of supernatural agents is that they tend to be “emotionally powerful.” Thus strong, vengeful, dangerous, and powerful God-concepts are common throughout the world’s religions. Their ability to activate strong emotions makes them both more salient and more memorable. This property also facilitates the transmission of religious stories.

Atran describes his evolutionary explantion of (2), the existence of “costly” and “hard-to-fake” rituals and expressions of belief, by reference to the need for unequivocal and costly displays of commitment due to the potential for deception in human communities. A&N write:

[T]he human metarepresentational ability to deceive and defect has been managed by communicative displays of passionate commitment to omniscient supernatural agents, who unlike humans do not succumb to false beliefs and thus can act as guarantors for future in-group cooperation. Expression of religious prescriptions performatively signals and establishes cognitive and emotional commitment to seek convergence, but it doesn’t specify (the propositional content of) what people should converge to. The truth about religious prescriptions is accepted on faith and communicated through ritual display, not discovered or described as a set of factual or logical propositions. The result of such convergence is to perpetuate a stable community of cooperators who sacrifice for the group in the short run, but benefit from it in the long run.

By making displays of commitment costly and difficult to fake, community members ensure that those who perform those displays are genuinely committed. In turn, individuals who perform such displays are more likely to become more committed, because of the cost of performance. These rituals generally use mundane communicative techniques, but in strictly prescribed ways that are unquestinable in their veracity and appropriateness. Thus, rituals serve both as communicative displays of commitment and as instructions/reminders of what it is proper to believe.

The explanations of (1) and (2) serve as the foundation for the explanation of (3), the activation of existential anxieties by religious beliefs. Atran posits that the activation of existential anxieties, in general, triggers our tendency to detect agency. Thus, religious stories that activate such anxieties, and particularly those that involve supernatural agents who have control over the sorts of events that activate those anxieties (e.g., death, illness, or natural disaster) will be more salient and memorable. In addition, terror management theory5 argues that activating existential anxieties will lead to strengthened activation of one’s own cultural or religious world view, which is designed to alleviate such anxieties, particularly those related to death. In turn, such anxieties will lead to increased animosity toward alternative world views. Thus, the activation of existential anxieties in religious settings also serves to strengthen communal ties, the function of religion highlighted in (2).

Finally, the properties of religious rituals highlighted in (4) also serve to increase the communal involvement of believers. By synchronizing the movements and regulating the sensory environment during the performance of religious rituals, religions are able to create a wealth of common experiences that serve to both strengthen community ties as well as communicating and strengthening the belief in religious truth. These rituals tend to incorporate all the aspects of religious concepts described in the explanations of (1)-(3), namely the activation of emotionally powerful, minimally counterintuitive concepts through the performance of costly and hard-to-fake acts of commitment.

This constitutes the basics of Atran and Norenzayan’s theory of religion. Obviously, I’ve just touched on most aspects of the theory, and there’s much more to it. If the information here has piqued your interest, I recommend Atran’s book, which is a fairly good read, and full of very interesting anecdotes and descriptions of non-European religious practices.

1 Barrett, J.L. (2000). Exploring the natural foundations of religion. Trends in Cognitive Science, 4(1), 29-34.
2 Barrett, J.L. (In Press). Theological correctness: cognitive constraint and the study of religion. Method and Theory in the Study of Religion.
4 Lipkind, W. (1940) Caraj√° cosmography. The Journal of American Folk-Lore, 53, 248-251.
5 Rosenblatt, A., Greenberg, J., Solomon, S., Pyszcynski, T., & Lyon, D. (1989). Evidence for terror management theory: I. The effects of mortality salience on reactions to those who violate or uphold cultural values. Journal of Personality and Social Psychology, 57(4), 681-690. See also Lieberman, J. D. (2004). Terror management theory. American Journal of Psychiatry, 161, 1508.


  1. #1 Ettsem
    December 10, 2006

    I recently came up with a technique for describing consciousness that had sufficient power to bring about the perspective change Buddhists refer to as “enlightenment”.

    The formalism is described on my personal blog, at this URL:


    This leads to a generalized description:


    My question: does a similar mathematic entity exist (in set theory perhaps) that is similar to the “Fuzzy Self-Similar Definition” described in the first article?

  2. #2 Sunil Gangadharan
    December 15, 2006

    Very interesting post on the cognitive aspects of religion. I enjoyed this immensly. Do you know any studies out there about the neurobiology of religion?

  3. #3 bee
    December 16, 2006

    While the logic employed here is admirable, it fails. It fails because it does not really address the question of religion as a phenomena. Rather it seeks to use analogy to map the phenomena of interest, religion, into a domain that it can then dissect. This is done by decomposing religion into a set of beliefs and actions that while assessable to scientific method do not represent religion.

    The logic to decompose agency is a fine example. The example has NOTHING to do with religious belief or religion. It is an exploration of agency. I would encourage the authors not to delve so far afield of the phenomena at hand. To start at birth is reductionist thinking taking to the level of absurdity. We have no way of knowing whether the hypothesis offered was ever acted upon by selective forces. That is the expression we observe may be the by product of other forces.

    I would suggest that if we wish to approach the problem scientifically we should stop “explaining” the phenomena via analogy. Address the specific domain.

  4. #4 tomeg
    December 17, 2006

    Excellent summary. I found it fascinating in the particulars, and satisfying as a whole. I don’t know if or how it might be relevant to discussion, but I consider myself a deeply religious person, and now I much better understand why. I think it odd–and somewhat amusing–that my simply reading items 1-4 has enhanced the satisfaction I feel in my daily life from practicing my faith. What to make of that??