Challenges from 4e Cognition to the Standard Cognitive Science of Religion Model

Abstract

Embodied, enactive cognition, which is also embedded or emplaced cognition and extended cognition through tools, including language, presents various challenges to the standard model of the cognitive science of religion. In its focus on unconscious brain mechanisms, the standard model downplays or eliminates religious meaning as epiphenomenal or illusory. It often denies that religion, once present, is adaptive or admits as adaptive only costly signaling. It regards humans’ perceptions of their environments as representations, mistaking an environment as determinate before cognition occurs. This support for indirect perception makes no sense given its emphasis on the need for sensing possible threats to survival. As brain mechanisms of individuals do all the heavy lifting, the model regards culture and its influence as nonexistent or insignificant. This stance denies how the social constitutes a huge part of our embodied preobjective and tacit engagement with the world, as well as socio-cultural realities, including religion, as self-organizing systems. The neglect of embodiment extends to its take on supernatural agents as allegedly disembodied minds. The standard model overlooks how ordinary rituals promote bonding through group presence, synchrony, and endorphin production and how some rituals increase knowledge of a particular natural environment, thus overlooking how religion can be adaptive.

1. Introduction

Embodied, enactive cognition, which is also embedded or emplaced cognition and extended cognition through tools such as a cane or realities less immediately perceptible such as a language or a tradition, presents various challenges to the standard model of the cognitive science of religion (CSR—when I use the abbreviation CSR in the remainder of this article, I will mean the standard model unless otherwise specified). 4e cognition embraces a holism with respect to the human organism, the human body—or as religionist William Poteat puts it, the human “mindbody”, where the parts of the body, including brain, nervous system, viscera, appendages, and so on, act or function as an integrated reality where its properties are different than the sum of the properties of its parts. According to the Oxford Handbook of 4E Cognition, 4e cognition maintains that cognition is shaped and structured by dynamic interactions between the brain, body and both the physical and social environments. From that work’s categorizations, I will take a strong approach with the following stipulations: (1) A cognitive process is strongly embodied by bodily processes if it is partially constituted by (essentially based on) processes in the body that are not in the brain; (2) A cognitive process is strongly embodied by extrabodily processes if it is partially constituted by extrabodily processes; (3) A cognitive process is strongly enacted if it is partially constituted by the ability or disposition to act (Newen et al. 2018, pp. 4–5; emphases in the original). The human organism comes embedded or emplaced in an environment with affordances, to use James J. Gibson’s term (Gibson 1979). Its cognition—always involving meaning, in the sense of both knowledge and value or disvalue—consists of a mutual constitution, determination, definition, and specification with its natural–social environment, which includes other organisms. No dualism pertains between the natural and the social, though interactions with the environment may involve different proportions of social and nonsocial engagement. The value or disvalue aspect of experience always involves affect or emotion and an aesthetic quality or sense. We can characterize dynamic cognition with concepts such as interaction, interconnectedness, intimacy, and intercorporeality with respect to other sentient organisms. Causation here is nonlinear, and some internalization of one’s environment pertains; perception is in some sense direct. With regard to nonlinearity, Maurice Merleau-Ponty observes,

The ‘psychophysical’ event is therefore no longer of the type of ‘worldly’ causality, the brain becomes the seat of a process of ‘patterning’ which intervenes even before the cortical stage, and which, from the moment the nervous system comes into play, confuses the relations of stimulus to organism. The excitation is seized and reorganized by transversal functions which make it resemble the perception which it is about to arouse.

(Merleau-Ponty 1962, p. 75; emphases Merleau-Ponty’s)

While we can, of course, discern particular events, 4e cognition recognizes that the passage of time is continuous with ever-present connections with the past and the future. Tacit knowledge figures heavily in the connectedness with past experience. 4e cognition envisions a complex interplay between prereflective and reflective experiences, in which culture, including religion, constitutes a reality other than the sum of individual experiences.

By way of contrast, the standard model of the cognitive science of religion tends to separate the human mindbody into parts relatively external to each other, with the primary focus on the brain and postulated brain mechanisms that cause humans to believe in supernatural agents and teleology when usually neither agents nor telos are present. The resultant cognition lacks meaning in a positive sense, for the information and attendant value or disvalue are illusory and, for many holding the standard model, not adaptive. Such cognitive experience may then be reduced to an epiphenomenon of the brain mechanisms. As Jeppe Sinding Jensen puts it, “conscious mental activities are ‘nothing but’ physical brain processes” (Jensen 2013, p. 242) as if regarded from a third-person perspective. CSR plays on evolution in positing that religion arises from evolved brain mechanisms not at all selected for by religious behavior (religion as a spandrel). CSR divides on whether religion, once present, is adaptive, with many answering in the negative in light of the downplaying of religious meaning, while some regard it as facilitating bonding, specifically through costly signaling. The standard model pictures the environment and other sentient beings as largely external to the perceiver, with linear causation and the relationship in perception being one of stimulus and response. As Merleau-Ponty relates, this model permits between the organism “and other objects only external and mechanical relationships, whether in the narrow sense of motion received and transmitted, or in the wider sense of the relations of function to variable” (Merleau-Ponty 1962, p. 73). Finally, the standard model banks on discrete events of perception by individuals, rather than acknowledging complex socio-cultural developments involving both prereflective and reflective activities to explain religion.

As I develop the contrast between 4e cognition and CSR, I will address the standard model’s positions regarding the role of brain mechanisms,1 perception as representational, its downplaying of culture, its denial of embodiment and the nonagential with respect to the supernatural,2 and whether and how religion is adaptive.

2. The Role of Brain Mechanisms

Given these fundamental differences with respect to the standard CSR model, 4e cognition presents multiple challenges to that model. The first challenge I will take up is CSR’s separation of the brain from the rest of the body as it focuses on brain mechanisms, sometimes conceived as within separate “modules”. The positing of brain modules represents a further move supporting separatism and gainsaying holism. Recent research contravenes this thesis, showing flexibility and multitasking among brain areas. Additionally, even when certain areas are especially engaged, many other brain areas play a part in the integration involved in all sentient experiences (Fuchs 2018, pp. 46–52; Jensen 2013, p. 246). Psychiatrist and philosopher of phenomenology Thomas Fuchs attributes to such modularity the “localization fallacy” (Fuchs 2018, pp. 46–52, 246). With this separation of the brain, aggravated by positing modules, CSR promotes a dualism reminiscent of mind/body dualism. What Mary Midgeley writes about contemporary philosophy of mind applies to CSR, as she notes the irony of “the persistence in their thinking of the shades of the Enlightenment ghost they thought they had routed. For when they discourse about the ‘mind–body’ relation, they rarely consider anything in that ‘body’ below the level of the neck. Flesh and bones (and, unsurprisingly, women’s minds) are still relatively neglected subjects in the field” (Midgley 1997, pp. 66–67). Similarly, Fuchs indicates that the brain, after toppling the “Cartesian Ego”, “immediately do[es] the same as the Ego in Descartes, to putatively imagine, to perceive, and to decide. Nevertheless, brains think or decide just as little as bodiless Egos” (Fuchs 2018, p. 45). Fuchs finds here a “mereological fallacy”, following M. R. Bennett and P. M. S. Hacker’s terminology (Bennett and Hacker 2003, pp. 71–72), where a part is confused with the whole (Fuchs 2018, pp. 43–46). I also see in such dualism an instantiation of what Alfred North Whitehead labeled “the fallacy of misplaced concreteness” (Whitehead 1925, pp. 50–55, 58; 1929, p. 35). The fundamental problem lies in treating the brain as a person or agent who does whatever a human being does instead of acknowledging the whole embodied person acting in conjunction with its environment (Fuchs 2018, pp. 43–44, 67–69). (A parallel category mistake happens when consciousness in itself—as separated from integral embodiment—allegedly decides to act (Fuchs 2018, p. 226)). For it is the sentient, living body or mindbody who acts, believes, and feels—who dances in a rain ritual, who believes in her ancestor spirit, or who fears punishment from the storm god. The brain is an organ of the whole body, albeit the single most important one. Jensen finds CSR unable to accommodate meaning, for “it is difficult to see how meaning could have extensions in space or time, and it certainly does not obey the laws of gravity” (Jensen 2013, p. 245). In contrast with the standard model, Fuchs answers: “Unlike physical mechanisms, consciousness is not analyzable into distinct spatiotemporal components; it covers space, time, and the body” (Fuchs 2018, p. 60). Viewing the brain as computer-like and perceiving it as computational is related to this problem, for a computer is not an organ of a body. This computational focus causes CSR to miss the aesthetic dimension of cognition, what William James and John Dewey, recognizing both particulars and ambience, referred to as the “much-at-once” quality or “pervasive quality” of an experience, respectively (James 1911; Dewey [1930] 1988, p. 246; Johnson 2007, pp. 72–77, 89), and to some extent to downplay the affective dimension.

Given CSR’s commitment to innate brain mechanisms, theorizing about and studying young children is a high priority. Interpreting data on whether children believe that inanimate natural objects have an intended purpose, Deborah Keleman concludes that children are natural teleologists and “tentatively” answers “yes” to the question posed by her article title, “Are children ‘Intuitive Theists’?” (Kelemen 2006). Given that very young children assume that other people perceive what they perceive even when not positioned to do so and that various experiments flesh that out, Justin L. Barrett theorizes that such children ascribe to agents’ superpower (Barrett 2012, pp. 74–77), superknowledge (Barrett 2012, pp. 82–96), and perhaps immortality (Barrett 2012, pp. 113–17). He further hypothesizes that they are pre-disposed to believe in a monotheistic God. The elephant in the room is the inculturation in the idea of God that young children in the West have received from adults before cognitive scientists test them. We have no evidence that children invent de novo concepts of a supernatural agent with extraordinary capabilities. Moreover, Jensen calls attention to brain plasticity (Jensen 2009, p. 148; 2013, p. 250; Duque et al. 2010, p. 138), which raises the possibility that particular cultures may shape brain structures of young children, thus complicating the distinction between innateness and development.

I do grant some propensity to attribute agency to human perception and intuition—though I find the term mechanisms to overstate matters. However, several studies call into question the strength of children’s tendencies to invoke supernatural agential or teleological explanations. Kelemen mentions parenthetically a study by Frank C. Keil but does not describe nor engage his results (Kelemen 2006, pp. 101–2; Nikkel 2015, p. 624). Keil studied children in grades kindergarten, 2, and 4 on the categorization of life forms, posing questions about a “thing” that could enter and cause harm to a human body. Subgroups were offered alternative descriptions of the thing: (1) functional or teleological, where the thing must get inside bodies and use parts of them, or it will not survive (Keil 1992, pp. 123–24); (2) simple mechanical, where the thing makes abrasions; (3) intentional, attributing desires and goals to the thing; (4) artifactual where the thing originates from human design; or (5) no description at all (Keil 1992, pp. 124–25). “Children at all ages thought that the ‘teleological’ thing did not know what it was doing any more than the mechanical one, attributing knowledge roughly three times as much to the intention/desire entity”, Keil concludes (Keil 1992, p. 126). I think that if the human propensity to ascribe intentional agency were as universally strong as CSR holds, at least the younger children would impute such intentionality to the “functional/teleological” thing (Nikkel 2015, p. 624). (In another context, indigenous people sometimes blame evil spirits for harm to human bodies; that context involves a whole religious system involving some reflection).

Kelemen, in a footnote, cursorily discusses a 1932 study of children from the animistic Manus culture. She disputes Margaret Mead’s negative conclusions about spontaneous animistic thinking by Manus children, citing Mead’s use both of drawings and of questions on supposedly malintentioned artifacts (namely, a canoe and a pencil) (Kelemen 2006, p. 111, n. 1; Nikkel 2015, p. 624). This characterization, however, fails to appreciate the scope of Mead’s fieldwork, which analyzed 32,000 spontaneous drawings and numerous responses to ink blots. Mead observed and transcribed copious interactions of children in play and in conversation with other children of various ages and with adults, including children’s responses to adult explanations, serious illness, accidents, natural disasters, and animals. Mead identified her glass wind chimes as a ramus, the Manus term for a charm intended to cause others to give the possessor desired objects. Despite manifold opportunities to invoke supernatural agency, such references were rare (Mead 1932). Though adult Manus normally attributes chance events to a spirit, including blaming a spirit when losing a small object and regarding cyclones as punishment by an angry spirit, no child was observed acting in this manner (Mead 1932, p. 181). Only one (of 41) children spontaneously spoke of ghosts, even though boys at age 5 or 6 are assigned a ghost guardian. The boy who did broach the topic of his guardian had faced the death of his father 2 years ago, and this talk was “regarded as aberrant by his companions” (Mead 1932, p. 182). Though “children hear many reports of ghostly activity”, children identified only six of their drawings as of ghosts, and these manifested “no distinguishing ghostly attributes” (Mead 1932, p. 183). Parents warn children about being eaten by tchinals—or land devils, if they play at a “slightly distant islet”. Nevertheless, children exhibit “but slight real belief”. Children invoked tchinals in just three circumstances: (1) a happy-go-lucky game of capture on a single occasion (Mead 1932, p. 183), (2) designating the issue of a failed attempt to draw a human as a tchinal, (3) adding “of a tchinal” to their interpretation of an ink blot baring “little resemblance to the object named” (Mead 1932, p. 184). The drawings identified as a tchinal never included any of the features the traditional appearance of a tchinal (Mead 1932, p. 183). Additionally, adult Manus believe that viewing one’s reflection in fresh water will result in the capture of some of one’s soul stuff by a water demon. However, since children will look at their reflections regardless of warnings, parents will not take children to the mainland (Mead 1932, p. 183).

Mead asserts that “Manus children not only show no tendency towards spontaneous animistic thought but that they also show what may perhaps legitimately be termed a negativism towards explanations couched in animistic rather than practical cause and effect terms” (Mead 1932, p. 186). Kelemen grants that Mead’s study might count against children being natural animists but claims it does not count against her theory of children as natural theists (Kelemen 2006, p. 111, n. 1). Theism in comparison with animism involves wider scope and higher abstraction: that is, deities in ancient polytheism are responsible for a whole realm like the sky or sea, while nature spirits are localized. Mead did not ask the children how things originated in nature. Still, the Manus children were free to invoke a god or goddess, including a Lady or Lord of the animals, to explain random events, ill health, storms, or how animals or humans came to exist.

Mead even turns the tables by suggesting that Western children are inclined more to “animistic” or supernatural thinking than Manus children from an animistic culture. Regarding Manus culture, Mead cites a simple language lacking metaphors that personify natural or artifactual objects, the fact that young children learn “to make correct physical adjustments” to the environment because failure brings “severe punishment”, and the fact that children do not hear tribal myths, nor undergo explicit religious instruction, nor participate in religious ceremonies. In contrast, English proffers a wealth of personalizing or “animistic” metaphors, while parents share stories and poems that personify animals, forces of nature, and artifacts. Furthermore, Western children take part in religious ceremonies and receive religious instruction (Mead 1932, pp. 187–89; Nikkel 2015, pp. 624–26). Finally, Western children learn beliefs and practices involving popular nonreligious characters with extraordinary powers. Mead’s research indicates that the study of other tribes not influenced by Western culture or adult theistic concepts would advance the cognitive science of religion. To her credit, Kelemen endorses “further research” on how well the “description” of children as intuitive theists “really holds across individuals and cultures” (Kelemen 2006, p. 109; Nikkel 2015, p. 626).

E. Margaret Evans studied three different age groups of children and their parents, from fundamentalist and non-fundamentalist communities, on belief in creationism, evolution, or spontaneous generation as explanations for the origin of species. Referencing E. Mayr, Evans clarifies that “unlike Lamarckian evolution or creation, spontaneous generation explanations are nonteleological in that they do not directly invoke purpose or an underlying design” (Evans 2000, p. 236). In the open-ended question segment of the study, the children were asked how the first instantiation of these three species occurred: an Asian sun bear, a tuatara lizard, and a human (Evans 2000, p. 242). In contrast to the closed or forced response segment, this segment would be amenable to spontaneous thinking. Significantly, she finds that among the youngest children, 5- to 7-year-olds, “no significant difference between creation and spontaneous generation” beliefs obtained for non-fundamentalist households (Evans 2000, pp. 246–47; Nikkel 2015, pp. 626–27). Barring specific instruction—as with fundamentalist households, this age group would be the least likely to be impacted by their culture. Both Evans’ and Mead’s research casts doubt on the hypothesis that children are natural teleologists.

Concerning adults, the standard model holds that brain mechanisms or devices cause incorrect theological thinking in the vein of anthropomorphizing. I will not take the time to look at studies supporting this claim. I will just note that I find theologically incorrect language in setting up the scenarios subjects consider, thus biasing subjects to respond incorrectly, thereby weakening the studies’ conclusions (see Nikkel 2015, pp. 527–31).

3. Perception as Representational Rather than Direct

The standard model of CSR regards humans’ perceptions of their environments as representations. This contrasts with the position of 4e cognition, in accord with Edmund Husserl, Merleau-Ponty, and Gibson, as well as that of pragmatism, which holds that perception is nonrepresentational and, in some sense, direct. Of course, different species perceive the world differently. Yet for each perceiver, it engages and perceives its lifeworld—with its ambience, objects, other perceivers, and events—as directly relevant to its behavioral possibilities rather than as a once-removed representation. The meaning of representation involves bringing back in a derivative way something that is not present to the perceiver—something stands in or for something else. The representational view mistakes everything of an environment as already fully determinate before a sentient organism’s mutually constitutive interaction with it. Whether this engagement or interaction includes touching, receiving reflected light (and not receiving light that is absorbed), receiving sound waves, and/or other modes, the embodied human or other animal perceives a meaningful whole.

Were the representational views correct? What candidates might represent or stand for what is actually there? The neural correlates of a perception count as one such candidate. However, as Fuchs recognizes, “there is no homunculus in the brain who would be able to grasp neuronal representations as representations, to see them as pictures or to read them as traces of memory (Fuchs 2018, p. 42; emphases Fuchs’). Furthermore, one would need a second body to have an objective, third-person-like apprehension of one’s own neurons. Yet if a sentient being were to perceive neurons as neurons, this experience would be, for most intents and purposes, meaningless. So, this apprehension would, at best, provide a weak approximation of the immediacy of what an animal perceives. Here, we have a dualism between the objective reality of neurons and the supposedly subjective (and epiphenomenal) experience of the object. On the contrary, Fuchs insists that “a meaningful connection cannot be ascribed to functional, rule-consistent procedures without there being someone who understands this connection (Fuchs 2018, p. 38; emphasis Fuchs’).

As another candidate, one could propose that the components and properties of something, someone, or some event represent the whole of what is perceived. But phenomenologist Walter Hopp provides the antithetical example of a scratched hood: “We can only perceive that the hood is scratched if we see the hood and the scratch, but neither the hood nor the scratch represents the state of affairs of the hood’s being scratched” (Hopp 2020, p. 27; emphasis Hopp’s). This candidate runs up against the holistic phenomenon of realities typically being different than the sum of their constituent parts and their individual properties.

One could even speculate that what someone perceives constitutes representations of the activities of fundamental forces, subatomic particles, and photons. This harkens back to the distinction between primary qualities like size, shape, position, and motion versus secondary qualities like color, texture, and smell by Locke and other Enlightenment philosophers. Our representations then fail to capture much of reality, including the huge amount of empty space within the things we perceive. Of course, the binding within and between atoms and molecules forming our bodies and much of our environment means that our lifeworld perception of the solidity of much of what we encounter and our cognition that we cannot pass through solid things are quite accurate. Here again, we find a dualism between the objective and the subjective.

In addition to the general problems described above, CSR’s support for representationalism and indirect perception creates a problem with respect to a key premise of CSR. That is, CSR emphasizes the uncertainty of the natural environment and the paramount need to sense possible threats in the evolutionary struggle to survive. It thus seems strange to favor a theory of perception that is indirect and involves a two-step negotiation between objective reality and its decoding or interpretation rather than a mutual constitution of cognition between mindbody and the environment.

4. The Downplaying or Dismissal of Culture

With its extremely strong individualism, where the brain mechanisms of discrete individuals do all the heavy lifting, CSR regards culture and its influence as nonexistent or insignificant. Culture becomes just an abstraction for the sum of different persons’ brain processes. To that point, Ilka Pyysiainen maintains that “‘cultures’ are merely abstract summaries of the thoughts and behaviors of individuals, not identifiable wholes that can be empirically explored” (Pyysiainen 2003, p. 53). In critiquing the seminal figure of sociology, Pyysiainen finds no explanation in Durkheim for how the collective consciousness of a society “emerges from individual consciousnesses. Nor is there any compelling evidence for the existence of any such entity” (Pyysiainen 2003, p. 74). Jensen quotes Pyysiainen’s dismissal of a constitutive role for social communication and “extended” material objects in cultural meaning:

As meaning does not reside in isolated objects, such as works of art, texts, or buildings, but is the property of a conscious cognitive process, symbolism is best viewed as a cognitive mechanism that participates in the construction of knowledge as well as in the functioning of memory. It is not an instrument of social communication or a property of phenomenon that can be considered apart from this mechanism.

(Jensen 2013, p. 246; Pyysiainen 2003, p. 41)

For his part, Pascal Boyer contends that it is “confusing to say that people share a culture, as if culture were common property. We may have strictly identical amounts of money in our respective wallets without sharing any of it” (Boyer 2001, pp. 35–36). Jensen labels Pyysiainen’s and Boyer’s views of culture as “eliminativist” (Jensen 2009, p. 138; 2013, pp. 242–46). Boyer’s Minds Makes Societies attempts to disabuse social scientists of the opinion that social norms exist outside of people’s heads (Boyer 2018). While Scott Atran’s views are not as stark as Pyysiainen’s and Boyer’s, the causative influence of culture is minimal (Atran 2002, pp. 200–34). As Joseph Bulbulia puts it, “external culture provides only a triggering and shaping function” for Atran (Bulbulia 2009, p. 158), reducing “religious culture” to “cognitive noise” (Bulbulia 2009, p. 167). Atran, along with Ara Norenzayan, dismisses any evolutionary approach to religion and culture that relies on “machinelike [sic] patternings of collective norms (worldviews)” (Atran and Norenzayan 2004, p. 6).

Of course, every individual appropriates and interprets culture(s) somewhat differently. However, that fact does not preclude the reality and causative influence of culture. CSR’s stance denies how culture and the social constitute a huge part of our embodied preobjective and tacit engagement with the world. Merleau-Ponty remarks on how the influence of a society’s culture runs deep, most of it prereflective:

It is as false to place ourselves in society as an object among other objects, as it is to place society within ourselves as an object of thought, and in both cases the mistake lies in treating the social as an object. We must return to the social with which we are in contact by the mere fact of existing, and which we carry about inseparably with us before any objectification.

(Merleau-Ponty 1962, p. 362)

Philosopher of science and epistemology, Michael Polanyi, emphasizes “the tacit dimension”, as our cognition always involves attending from subsidiaries to a comprehensive focal meaning (1966). Since explicit knowledge relies on, is rooted in, tacit knowledge for understanding and application, Polanyi declares that “a wholly explicit knowledge is unthinkable” (emphasis Polanyi’s 1969, p. 144; see also, Polanyi 1966, p. 61).

Explicit knowledge for instances of cultural and social knowing (not to mention instances of cognition of nature, not dualistically separated from cultural knowledge) constitutes the tip of the cognitive iceberg, for this knowledge depends upon an incredible amount of background, tacit, implicit, contextual, prereflective knowledge. (This reality is the primary reason for computer programming’s difficulty in mimicking human language [See for example (Dreyfus 1992)]). From a Polanyian perspective, any use of even a single word or concept involves a tacit coefficient as the speaker relates it to the particular context, including tacit knowledge of its application in countless other previous contexts. Note that what at one point is reflective, such as language, other symbols, and thinking (themselves dependent upon a prereflective substantive connection with embodied meanings), often becomes tacit—more or less prereflective—as subsidiary elements in new cognitions. While I have just focused on linguistic meaning, the prereflectivity and tacitness of much of human culture, including religion, applies even more so to the obviously material aspects, including ritual, architecture, moral behavior, and institutional practices.

CSR’s seeming picture of the origin and development of religion imagines that the intuitions of individuals (or individual brains) in discrete encounters with their environment do the important work, and the rest more or less falls into place. However, evolutionary psychologist Robin Dunbar notes how religion could only develop after humans were able to share individual experiences or intuitions with their group through language and adequate mentalizing capacities. Indeed, Dunbar stipulates that communal religion requires fifth-order intentionality or mentalizing, such that people agree that a transcendental realm exists, populated with intentional agents that affect us (Dunbar 2022, pp. 112–23). One does not need to buy into all the details of Dunbar’s scheme to accept the crucial nature of mentalizing for organized religion. Dunbar cites a study measuring peoples’ “mentalizing skills, the effectiveness of their agency detection mechanism [sic], their schizotypal tendencies and their religious beliefs and behaviors (religiosity)” (Dunbar 2022, p. 120; Wlodarski and Pearce 2016). This study “suggests that mentalizing positively influences religiosity quite independently of agency detection and schizotypal thinking, both of which are extremely closely correlated” (Dunbar 2022, p. 121). The importance of mentalizing and orders of intentionality for the existence of religion indicates the necessity of reflection beyond intuitive, prereflective processes involving agency detection or anthropomorphism to explain religion—and its meaning.

I am willing to grant that humans have some tendency toward teleology in their reflections. However, it is selective and contextual, relative to particular societies engaging in their natural–social world. After all, no animistic society attributes agency to all the potential candidates in their natural environment. Calling prereflective propensities to attribute agency “mechanisms” or “devices” is not helpful or accurate, as they do not come to fruition for communal religion sans reflection. Moreover, humans have prereflective and reflective capacities to cognize nonteleological causation, indeed, even strong tendencies to do so in various contexts. As Manuel Vasquez observes, the stipulations of CSR “underdetermine religious practices and artefacts”, as it ignores human creativity in engagement with particular places that both open up and constrain possibilities for individuals and their group (Vasquez 2011, p. 198; emphasis Vasquez’s). Religious meaning with its beliefs and practices arises not incidental to brain mechanisms but cognizant of particular environments, and, I would add, thus probably adaptive through increasing knowledge of a natural environment and its self-organizing systems and other realities. I will say more about this in the upcoming section concerning whether and how religion is adaptive.

Speaking of self-organizing systems, CSR ignores evidence that human socio-cultural realities, including religion, instantiate complex, dynamic, self-organizing systems with regulative feedback loops (Rappaport 1979, p. 151) shaped by attractors. These are not as tightly integrated as biological organisms (Rappaport 1979, pp. 152–53), and they always involve some contestation between attractors. As each socio-cultural happening is not exactly the same as any previous one, some creative change occurs as people interact with their environments. Sometimes, change becomes great enough to engender significant disequilibrium, resulting in substantial or even revolutionary change—or the death of a system. Justin L. Barrett, though accepting much of the standard model (see, for example, Barrett 2017, pp. 198–99), recognizes that the standard model’s work with individual psychology can become disconnected from studying group level behavior, which necessarily involves culture. A complex relationship pertains between individual cognitive differences and cultural differences (Barrett 2017, pp. 200–2). While CSR does not deny individual differences in the strength of purported mechanisms, it does tend to assume that for a large majority of people, their potency is similar, despite some exceptions. Barrett cites evidence that minimally counterintuitive concepts—influential for the standard model—“are more memorable for adolescents and young adults than comparable intuitive concepts, but the opposite may be true for older adults” (Barrett 2017, p. 199). Dunbar underlines that those on the autistic spectrum, with decreased mentalizing abilities, score dramatically lower in tests for belief in a personal god (Dunbar 2022, pp. 121–23; cf. Barrett 2017, p. 200). Exactly how such individual differences play out within religions or other socio-cultural systems is unpredictable. Yet, as participants engage differently, appropriately, and interpret religious culture, they partake in similarities not explained by the sum of individual brains. The primacy of individual brains and of discrete events for CSR avoids the holistic reality that religious actors form part of a cultural system extending across a society and enduring across time, embedded in places, continually changing while usually maintaining sufficient stability. A cultural system includes narratives, rituals, and other regulatory means for how culture is transmitted to future generations.

5. Denying Embodiment and the Nonagential Vis-a-Vis the Supernatural

CSR’s neglect of embodiment extends to its take on supernatural agents as allegedly disembodied. This assumption projects onto religious practitioners a mind–body dualism. Boyer asserts the irrelevancy of anything but the unaccompanied mind: “anthropologists know that the only feature of humans that is always projected onto the supernatural is the mind (Boyer 2001, p. 144; emphasis Boyer’s). In a similar vein, psychologist Paul Bloom maintains that “a spirit, ghost, or deity” has an essentially discarnate nature (Bloom 2004, pp. 212–13). Unfortunately for the standard model, understanding supernatural agents as discarnate runs counter to the history of religions. As K. Mitch Hodge argues, folk bring a social stance with expectations of social affordances and structures of meaning to their engagement with parts of their environment, typically understanding nonhuman spirits and deceased human spirits as embodied (Hodge 2018, pp. 16–23). Animistic spirits can best be understood as embodied in a force or feature of nature, such as the wind or mountain, for example. These have not been understood as immaterial souls free to move around and take charge of some other part of nature. Early theorist of religion E. B. Tylor reports that hunter-gatherer cultures, though often imagining spirits as “vaporous”, most definitely do not conceive them as “immaterial”. He highlights how some tribes ensure that an opening exists in a container holding a spirit so that it can escape (Tylor 1871, pp. 410–13). When the Toraja of Indonesia are going to sacrifice a water buffalo, they warn the spirits to keep away to escape injury (Eyre et al. 2001). Other animistic beliefs involve a spirit having an animalized or anthropomorphic body, such as shamanistic beliefs in a spirit animal or the Toraja belief that each grain of rice, their staple crop, is a little yellow person (Eyre et al. 2001). Furthermore, individual gods and goddesses not confined to a locale have been viewed as having bodies for most of religious history. As some ancient religions developed, certain animistic beliefs yielded to a god or goddess who controlled a part of nature, like Poseidon and the oceans. However, such ancient gods and goddesses were blatantly anthropomorphic in body. Of course, these embodied spirits do not endure all the limitations that human and animal bodies do. Their bodies may well be hidden or invisible to ordinary vision. In some cases, they may appear in some form other than their proper bodies. Not only did tribal and ancient people picture deities as embodied, but they also believed that actual divine bodies bore some analogy to their representations.

What of the Western monotheisms, where theologians typically hold the immateriality of God? In truth, the total disembodiment and immateriality of God in learned Jewish, Christian, and Islamic theology issued from a long process strongly influenced by Greek philosophy, particularly Platonic and Aristotelian. Francesca Stavrakopoulou demonstrates that for ancient Israelites, God had a distinctly male body (not to mention a female consort) (Stavrakopoulou 2021). Various Hebrew biblical passages specifically refer to God’s body. Ancient Judaism eventually prohibited the imagist representation of God, though the historical reality of that happened much later than depicted in the biblical narrative. Even so, this did not repudiate the corporeal reality of the divine, only that one could adequately portray it in an image: the magnificence of God and the divine body compared to human or animal bodies would tolerate no visual representations. Despite theologians’ repudiation of divine corporeality, nonelite practitioners of Western monotheisms have often assumed that God has a body (for example, Cook 2024, pp. 140–41).

Another consequence of CSR putting all its eggs in the basket of disembodied supernature agents is its neglect of embodied religious beliefs and rituals relating to nonagential supernatural powers, specifically transpersonal realities, and nonpersonal magical forces, these latter representing a kind of physical causation. Some indigenous religions have featured a transpersonal force or energy, which is not a conscious being with intentions but rather pervades the universe and is present in everything to a greater or lesser degree. Native American concepts of orenda, manitou, or wakan, paired with an adjective meaning great or large; the Mayan Ch’ol, especially associated with blood and sacrifices that empower the sun god; and the Polynesian concept of mana fall under this category. These Native American concepts took on meanings of a personal supreme being under accommodationist pressure from European colonizers. Nonpersonal concepts of divine or ultimate reality, like the Dao (the Way), Tian (Heaven), or Sunyata (Emptiness), are common in East Asian religions. The Dao of Daoism (as the source of and reflected in nature) and Tian of Ruism (commonly regarded as having meanings of cosmic order) bear analogies to the idea of the universe itself being a self-organizing, holistic reality. For Daoism, the Dao is the source of qi, the life force or energy. I grant that the Dao and Tian have met construal as intentional agents during Chinese religious history, especially in popular religion (Clark 2007; Clark and Winslett 2011; Puett 2004). Still, the influential thinkers of Daoism and Ruism (Confucianism) understand them in nonpersonal ways, which complicates the relative strengths of more prereflective processes versus more reflective ones. Concerning Mahayana Buddhism, though celestial boddhisattvas, as well as the historical Buddha, were understood as divine beings, they did not possess ultimate status in their own right, nor did they function as creator deities. Instead, they were understood as embodiments of the ultimate reality of Sunyata, the Buddha-nature, the dharmakaya (truth body), or Nirvana, which functions as the ultimate source (Nikkel 2015, p. 436).

I now turn my attention to magic. James Frazer famously distinguished between magic and religion, assigning them to distinct stages of prehistory, with magic pertaining to nonpersonal or nonintentional causation and religion involving appealing to spirits or deities (Nikkel 2015, pp. 433–34). Scholars of religion have corrected Frazer in that the two types per Frazer overlap (pre) historically. Some beliefs and practices combine both. Although some CSR scholars downplay and some even explicitly eliminate any nonintentional elements in ritual and magic (Lawson and McCauley 1990), such elimination defies the evidence. Some indigenous ritual practices, including using charms and amulets, entail the belief that, if the procedure is correctly carried out, a certain magical outcome will happen, or more likely will happen, quite aside from any intentions of a supernatural agent. The ground-breaking anthropologist E. E. Evans-Pritchard observed the Azande’s practice of divination through poisoning a chicken to determine whose witchcraft had brought great harm upon a person (Evans-Pritchard 1937, pp. 282–312), as well as their multifarious explanations for why the ritual sometimes failed (pp. 338–351). Azande believe that witchcraft itself is a physical substance residing within the witch and possessing supernatural powers (pp. 9, 21–23), sometimes causing harm without the witch’s intention or knowledge (p. 121). While Evans-Pritchard uses the phrase “poison oracle”, he does not mean that a deity or spirit is addressed or answers. Rather, the officiant addresses the benge, the poison, which gives its answer through the conventions of the ritual. Evans-Pritchard raises this question: “Do addresses to medicines and attributions of psychic action to them justify us in concluding that Azande personify them?” He responds: “Magical action is sui generis, and is not explained by the presence of spirits residing in medicines nor by the attribution to them of personality and will” (p. 465; see also, p. 441). The poison oracle, healing rituals, and other traditional Zande rituals eschew invocation of ghosts or gods, while explanations for failure forgo the intentions of supernatural agents or of the ritual participants (pp. 286–87, 332–33).

In a contemporary counter to the standard model, Harvey Whitehouse endorses both “quasi-mechanical” and “agent-driven” supernatural causation (Whitehouse 2007, p. 280). Also, Richard Sosis presumes that the recitation of Psalms by contemporary Israelis constitutes “magico-religious” responses to the threat of Hamas rocket attacks, which entail belief in not just divine intervention but in more automatic processes as well (Sosis 2007, pp. 903–11). Lastly, I would offer an analysis of two traditional categories for magic, first articulated by Frazer, namely, sympathetic or imitative magic and contagious magic. Both types carry an embodied logic, even if scientifically inaccurate. In contagious magic, bodies come in contact with an animal or person or part of the body of one, expecting that the association will bring positive or negative effects to a practitioner or to the person targeted. Similarly, imitative magic anticipates that bodily engagement with part of one’s environment will cause similar effects in another, but similar, part of one’s environment. To suppose that these types of magic require a supernatural agent adds an unnecessary complication while downgrading the intelligence of the purported agent by requiring an “object lesson” rather than a simpler communication.

6. Religion as Adaptive?

In my estimation, the minority of standard-model cognitive scientists of religion who see religion as promoting bonding is correct to part with those who see no adaptive effects of religion. However, they do not go far enough. Costly signaling gets highlighted, while the benefits of more ordinary rituals are overlooked. Dunbar, casting his net more broadly, identifies the primary function of religion as community belonging and bonding. Better bonding promotes cooperation, which loops back to strengthen bonding (Dunbar 2022, pp. 73–74). An indirect benefit of this bonding is better psychological and physical health. The key to bonding for Dunbar is ritual, which “form[s] the bedrock on which most, if not all, religions rest” (Dunbar 2022, p. 129). While not denying various theories that account for the immense importance of ritual (Dunbar 2022, pp. 131–36), Dunbar avows that they overlook “something more fundamental”, namely, “that most rituals probably trigger the endorphin system” (Dunbar 2022, pp. 136–37). Dunbar mentions the “considerable evidence that behaviors like singing, dancing, and emotional storytelling” lead to endorphin uptake, which in turn enhances bonding (Dunbar 2022, p. 137). The positive feelings surrounding endorphin uptake reduce feelings of pain. In cases of pain-inducing rituals, the endorphins temper the pain, while the sharing of painful experiences in themselves tends to increase bonding (Dunbar 2022, pp. 137–40). Dunbar highlights one more significant element of communal religious rituals: they are done in synchrony (Dunbar 2022, pp. 144–48). Synchrony “exaggerat[es] the magnitude of the endorphin effect, though it is not entirely clear how or why it does so” (Dunbar 2022, p. 147). Lest anyone suspect that just synchronous ritual itself, even if nonreligious, boosts bonding just as well, Dunbar cites a study indicating that religious meaning in synchronized rituals “gives added value and significantly increases the bonding aspect of rituals” (Dunbar 2022, pp. 147–48).

Religious rituals can also be adaptive by providing knowledge about natural systems. As suggested above, animists bring social expectations to the forces of nature they perceive as intentional, even as these embodied extraordinary beings are perceived as being in social relationships with social affordances of their own. These forces of nature are usually self-organizing systems. As such, these systems’ behavior is unpredictable, though never simply random, following fractal patterns centering around attractors, at least approximating nonlinear quadratic equations. This is to say that they behave in somewhat similar ways to sentient biological systems, so it should surprise no one that indigenous folk attribute intentionality to some forces of nature. Of the abundant self-organized systems in nature, not all are typically understood as controlled by intentional agents, even by cultures believing in animistic spirits. That hunter-gatherer societies, as well as other religious individuals and groups, interpret these patterns as conducive to magical influence is not surprising either, particularly with regard to sympathetic magic rituals. Though I do not believe that rituals have direct causative effects on natural phenomena, interactions with perceived social agents and nonintentional forces increase knowledge of nature, as well as contribute to the already mentioned adaptive effects of shared meanings and rituals that increase social bonding and cooperation.

For example, in his first book, Pigs for Ancestors, Rappaport documents how the religious ritual system of Tsembega Maring farmers of Papua New Guinea sustains an ecosystem that appears to allow humans and animals to thrive (Rappaport 1968). This system involves raising pigs in former gardens, clearing the ground, and facilitating new planting. The ritualistic sacrifice of pigs, where ancestral spirits are invoked, limits porcine damage to the ground and can also provide extra protein in times of stress or celebrate connections and alliances both within and beyond the tribe (Rappaport 1968, pp. 1–4, 87; 1984, this second edition attempts to rebut criticisms of the original with two addendums). Jorge Legoas P. provides another example with the Cabañuelas practices of Peruvian Andean small-scale agricultural communities. The full sweep of these practices includes an initiatory main “payment to the earth” for the year; reading signs of the frost, wind, rain, clouds, and/or mountains, which are understood as agents, in order to forecast weather and engage in appropriate agricultural activities—this is the heart of Cabañuelas, and tossing coca leaves in order to discern the will of Mama Coca and a secondary practice of reading molten lead. Legoas P. emphasizes that these practices involve negotiation to create a hybrid social network (Legoas P. 2022, pp. 987, 990, 1005–6). As one might guess, climate change is making farming practices more difficult for these peoples (Legoas P. 2022, p. 988). And no one claims that their forecasting matches or exceeds modern scientific weather predictions. However, my sense is that these practices have, through the generations, resulted in agricultural success at a much better rate than chance.

7. Conclusions

The standard model of the cognitive science of religion rests on dualisms, primarily of mind versus body—with the brain on either side of the binary depending upon context—and culture versus nature and biology, opting for one side of this latter dualistic coin as it subsumes culture under nature by reducing it to individual brain mechanisms. This dualistic reliance goes so far as to allege an innate dualism while managing to transcend it with its own ontology of reductive physicalism (Nikkel 2015, pp. 237–38).

As argued above, the mind/body dualism plays out in stipulating that supernatural agents are cognized as disembodied, not to mention the opinion that afterlife beliefs all entail mind/body dualism, despite the reality that most afterlife beliefs in the history of religions entail some sort of body—even if vaporous, or shadowy, or “spiritual”—and that most believers in an afterlife today imagine having some sort of body in heaven or whatever be the locale of life after death for them. Unsurprisingly then, advocates of the standard model, as well as some other CSR scholars, posit an “innate” dualism, also using such terms as natural, intuitive, common-sense, or folk dualism (Bering 2006, p. 453; Bloom 2004; Boyer 2001, p. 144; Bering and Bjorklund 2004, p. 228; Cohen and Barrett 2008, p. 43; Sosis and Kiper 2013, pp. 264–65; Pinker 2002, p. 126; Slingerland 2008, pp. 284–95). They believe that infants distinguishing between intentional agents and objects entail this dualism. However, the ability to distinguish between animate and inanimate realities suffices to explain this ability. They assume that fictional accounts of people switching bodies, accounts of spirits possessing bodies, or so-called out-of-body experiences require an innate dualism. I will first deal with the latter.

Those having out-of-body experiences report seeing, hearing, and, of course, having a sense of motion. Further, according to neuroscientist Antonio Damasio, they have background feelings of their bodies (Damasio 1994, pp. 150–55, 235–44; see also Damasio 1999, pp. 110, 285–87), even when their perception of their embodied location is displaced. The fundamental problem with presuming to have a body different from one’s proper body entails dualism, which is the absence of imagining or conceiving a disembodied, immaterial soul between incarnations. As with very young children, the innate dualist turns folks into tacit Platonic or Cartesian philosophers. I regard this as phenomenologically incorrect thinking. One can come up with such an explanation only with a very reflective process. Philosopher Mark Johnson writes evocatively of our capacity to imagine being an other-than-human body living in that body’s milieu:

Part of the awe we feel observing animals in the wild—the trotting coyote, the soaring eagle, the playful porpoise, the aggressive jay, the magnificent diving whale is that we too feel some of the sense of trotting and soaring, playing and diving. Shamans in aboriginal cultures observe animals closely by empathetically “becoming” the animals, and ritual practices in a wide range of aboriginal religions employ the movements of animals to achieve an ecstatic experience, an experience of being in the body of a very different kind of being.

(Johnson 1987, p. 566)

A much more parsimonious explanation than innate dualism for our capability to follow accounts of religious possession, of “out-of-body” experiences, or of exchanging bodies in popular culture is that we humans are highly intelligent embodied beings who naturally imagine in bodily ways (Nikkel 2015, p. 239).

The standard model, contrary to purported human innate dualism, opts for the mindless materialistic side of the dualistic coin. It thereby favors a metaphysics of reductive physicalism through its focus on brain mechanisms, consigning phenomenal experience and meaning to illusion or epiphenomenon (even as it creates a dualism between the brain and the rest of an organism’s body). Rather than allowing for a meaningful lifeworld of an organism through its engagement with affordances of and mutual constitution with its environment, CSR opts for second-hand representations, a position interconnected with reductive physicalism and the focus on isolated individuals’ brain mechanisms.

Regarding socio-cultural realities, the reduction of the human being—the human mindbody—and culture to individual brain mechanisms disregards the profound tacit or prereflective influences, as well as the reflective influences grounded in our prereflective embodiment, of culture as ongoing self-organizing systems. Furthermore, the reductive focus on individuals’ brain mechanisms allows rituals involving costly signaling to constitute the only significant cause of the adaptiveness of religion. This ignores the crucial nature of individuals gathering as a group during rituals—often involving synchronous speaking, singing, and/or movement—and the attendant increase of endorphins that facilitate bonding and cooperation.

The standard model of the cognitive science of religion has made significant contributions to the academic study of religion in calling attention to human tendencies to attribute intentionality to the supernatural through agency detection, anthropomorphism, and teleological cognition and to human belief in life after death. It has also conducted tests and experiments with respect to these tendencies in various contexts. Taking seriously 4e cognition’s challenges can greatly enhance the cognitive science of religion through more broadly and holistically studying religious cognition in terms of the full embodiment and embeddedness of human beings in particular natural and social environments and of the creative activity of the human individual and of self-organizing religious culture—culture incorporated into the human mindbody. Doing so enables the recognition of meaningful connections between humans and their social–natural environment, which expedites discovering ways that religion can be adaptive.