Psychologists finding religious belief: Building bridges between developmental cognitive science and cultural psychology

Introduction

Researchers in the cognitive science of religion (CSR) have propagated the idea that religious beliefs exist because they rest upon universal pre-given cognitive architecture (e.g. Barrett, 2007; Boyer, 2001; Cohen, Lanman, & Whitehouse, 2008; Guthrie, 2002; McCauley, 2011; McCauley & Lawson, 2002; Spelke & Kinzler, 2007). That is, humans are supposedly endowed with many domain-specific modules that allow, for example, theory of mind (e.g. Bering, 2006, 2011), agency detection (e.g. Barrett, 2004; Barrett, Richert, & Driesenga, 2001), and the retention of religious ritual (e.g. McCauley & Lawson, 2002; Pyysiäinen, 2002). Such modules are taken to conjunctively work to produce religious phenomena, which in turn lead to religious belief. Since a central presupposition is that capacities for religion are hard-wired, much of the work in this area involves examining children with the hopes of assessing them at a young enough age to preclude the meaningful influence of enculturation (e.g. Barrett, 2012; Keleman, 2004; Spelke, Breinlinger, Macomber, & Jacobson, 1992; Spelke & Kinzler, 2007).

There are two interrelated challenges to the CSR. First, we have shown that the phenomenon of religious belief is quite different from the kind of phenomena CSR is studying (Cresswell, 2014). To fully understand belief, researchers must apprehend socially constituted experiential realities that communities of believers live (Hampson, 2005; Reber, 2006; Slife & Reber, 2009, 2012). To address these realities, we have spelled out elsewhere an approach to religion that falls into more synchrony with scholars arguing that religion is deeply entwined with cultural practices (e.g. Geertz, 1977). We do not seek to restate this theory of religion in our article, but rather to outline the development of religious cognition and provide theoretical insights into how CSR can be improved. Consequently, we can only summarize and allude to the previous work and refer readers to the original source for a more thorough discussion of religious phenomena, per se. Cresswell (2014) spelled out a “pragmatic cultural psychology of religion” resting upon the work of Mikhail Bakhtin and William James (1981, 1982). It draws upon James’ claims that belief is always about something and one cannot talk about religious belief without considering what culturally specific systems constitute the content of belief. He also argues for a deeply experiential approach to religious belief that includes the experiences that shape religious belief as personally sacred. Bakhtin compliments James nicely with his work on language and how religious language constitutes the religious beliefs people know and experience. Because of humans’ saturation in dialogue and alterity, Bakhtin adds the important observation that we are always oriented towards an other—even when alone in religious practices like confession (see Poole, 2001). It is in this context that language practices, we have argued, shape the realities people believe in and experience. CSR bypasses such culturally linguistically constituted lived realities by focusing on individualist mechanistic processing.

A second challenge is the presupposition that enculturation of religious cognition does not happen until a child is well into the use of lexicon. CSR is predicated on the notion that innate mechanisms form the basis of cognitive architecture whose operation precedes the acquisition of lexicon and concomitant enculturation (e.g. Sperber, 1994, 1996). However, socio-linguistic phenomena are integral to psychological development from birth (Cole, 2003; Fogel, 1993; Wertsch, 1985). These challenges indicate that CSR could be enhanced if provided with the theoretical articulacy to grasp the constitutive role of language in the formation of religious cognition.

The objective of this article is to offer a theoretical approach to the ontogenetic development of religious cognition that (1) takes into account the ontogenetic role of language practices in the formation of religious cognition and (2) offers a view of cognition that is compatible with (1). Our paper attempts to accomplish this objective by first outlining a sociocultural approach to the development of religious cognition. Drawing on Mikhail Bakhtin’s notion of speech genres and current theorists in cultural psychology ¹ allows us to articulate how religious cognition is cultivated in early childhood. We then specify what CSR stands to gain by taking the forgoing seriously before turning to dynamic systems theory (DST), which offers an approach to cognition that naturally incorporates a sociocultural approach (Elman et al., 1998; Noë, 2009; Prinz, 2009; Thelen & Smith, 1996). Such work should lead to theory that enables the study of religious belief as it emerges in life, and provide a generative theoretical framework that allows researchers in CSR to proceed in harmony with a culturally informed psychology of religion.

Culture, language, and development of religious cognition

Cognition is generally understood to be about thinking. It includes higher order thoughts such as knowing what are one’s religious beliefs (Harré, 2002; Thagard, 2005). It also includes a discussion of the kinds of tasks or activities of which we are not always aware. As such, cognition is about thinking in a broad sense that encompasses activities ranging from reflective thought to dynamic embodied dispositional behavior.

Studying religious cognition is a difficult task, as we cannot simply peer into the heads of others and physically see their religious beliefs. Adding to this complication is the way that cognition involves meaningful symbols, which are notoriously difficult to pin down in a scientific manner (Harré, 2002). Cognition is often about symbolizing the world around us inside the mind (Boden, 2006) and cognitive scientists generally discuss these symbols as mental representations: “To address the crucial questions about the nature of mind, the psychological experiments need to be made interpretable within a theoretical framework that postulates mental representations and procedures” (Thagard, 2005, p. 8). Consequently, “[h]ere is the central hypothesis of cognitive science: Thinking can best be understood in terms of representational structures in the mind and computational procedures that operate on those structures” (Thagard, 2005, p. 10). Harré (2002) notes, however, that symbols are central to cognition, yet presuming that mental representations capture this centrality may be premature. He argues that cognitive tasks include a central role for language as the most definitive form of symbols. This approach makes room for a pragmatic cultural psychology of religion because it allows for a discussion of religion as a phenomena constituted in shared linguistic practices. Bakhtin (1990) and James (1982) both treated religion as meaningful and compelling by virtue of its shared symbolic nature. Relying on inner representations may actually bypass the way that religious cognition is deeply entwined with sociocultural experiences of learning language.

Harré’s (2002) claim is highly resonant with cultural psychology, which also argues that the symbols we learn in practicing language are deeply tied to our thinking and this includes our dispositional activities. Cultural psychology is predicated on the notion that cognition is constitutionally entwined with culture (see Valsiner, 2012). A cultural approach often turns to language as a primary consideration when establishing a case for culturally constituted cognition (see Burkitt, 2013; Cresswell & Teucher, 2011). Participating in language, for example, entails narratives that shape concepts cognition depends upon (Larraín & Haye, 2014). Language must thereby be understood as more than simply the use of lexical communication used to express thought as it constitutes cognition. As such, any type of religious cognition and what such belief entails is constituted in a rich notion of language. Bakhtin’s (1990) approach to religion highlighted how languages people practice shape the world as they apprehend it. As such, when people talk about religion, they jointly constitute religious realities populated with religious concepts. As people use religious words, they are shaping a world that ontologically includes these notions. Bakhtin offers a cultural view of religion that locates it in our shared practices.

Furthermore, language is deeply entwined with the body (Burkitt, 2013; Cresswell & Baerveldt, 2011; Cresswell & Teucher, 2011). Bakhtin (1986) discussed the notion of speech genres to articulate how there are generic ways that communities linguistically embody what one ought to do and say. He pointed out that, without fully articulating an ideology through lexicon, generic conventions are tacitly enacted and continually re-established by members of a community. A member of a community comes to express oneself in the generic style of the community and feels compelled to express that style. Therefore, religious belief and the concomitant experience entailed in such cognition are constituted through, and bound up in language practices. Cresswell (2014) addresses how Bakhtin (1990) and James (1982) both emphasize the experiential quality of religion. Bakhtin does so through his discussion of language as embodied. As people participate in language practices involving religious belief, they do so in a deeply embodied way. To use religious words that shape a world that ontologically includes religious notions include the deeply experiential qualities of such language. The world and religious beliefs entwined with it have deeply compelling force by virtue of the deeply embodied quality of language, making religious beliefs socially constituted phenomena that are personally experienced.

To highlight how CSR could be improved, we will attend to the development of religious belief. Specifically, we will turn to the ontogenetic development of religious belief and how religious cognition is shaped through language practices in infancy. Researchers are advancing sophisticated theories about how participation in dynamic interactions shapes an infant’s cognition (e.g. Cole, 2006; Fogel, 1993). Central to the cultivation into the embodied practice of language is “co-regulation” (see Fogel, 1993). Co-regulated interactions between a caregiver and infant unfold in a dialogical back and forth interchange of trial and error, where the actions of one party are entwined with the actions and reactions of another. Infants engage in a rhythmic flow of dialogue turn-taking with vocalizations, such as babbling and cooing (Hsu & Fogel, 2002). Caregiver’s reciprocate to infants’ vocalizations with affirming babble/coo versions of their own, and pauses to encourage responses.

It is important to maintain sight of how caregivers’ actions are enactments of generic language practices within a community that tacitly enculturates infants’ cognition (see Burkitt, 2013; Cresswell & Teucher, 2011; Goldstein & Schwade, 2008). Although these interchanges take place with infants expressing a pattern of sounds, rather than fully formed words, it does not mean that they are not participating in language practices. Since these language practices are embodied, infants not only begin to differentiate among words and sounds but also begin to feel differences in tandem with them. Greenspan and Shanker (2007), for example, demonstrate how infants express elementary global states that correspond with physiologic nervous system arousal. An infant cries out loud and the emotion is totalizing as if its entire being is brought to bear in the cry. An infant’s affect involves global sensations of comfort or discomfort, pleasure or pain, and so forth. In the course of co-regulated interactions, emotion is produced and subtly transformed. When a caregiver regulates an interaction with excited faces, gestures, and affirming sounds, an infant will begin to tailor the global totalizing emotion into a more refined repertoire of emotions. Such co-regulation enables infants to follow along with a caregiver’s more sophisticated emotionality through tacit tutoring into more sophisticated emotional expressions.

It is crucial to remember that infants participate not only in the emotion of caregivers but also the language through which they use to participate in the emotion of caregivers also entails the language of religious belief. Participating in the word “no” can provide an example of how co-regulation works. A caregiver verbalizes the word “no” to an infant who reaches for a dangerous or off-limits object. Significantly, “no” is also physically enacted. The concept of “no” becomes a bodily experience by virtue of a caregiver removing the object from an infant’s reach or removing an infant from the area in which the object resides. Through continual interactions, a pattern will emerge around the use of “no” as the concept begins to enact various types of restriction. The experience is also bodily insofar as a caregiver accompanies “no” with emotional expression. When a caregiver uses “no,” we are arguing that it is not employed in a vacuum as it is expressed in conjunction with other words and actions. For example, “no, we shouldn’t eat that” with an act of taking something from a 4-month-old’s mouth is a common kind of action involving embodied language. This investment of emotion in conceptualization and categorization in language tacitly entwines bodily feelings with the development of language. How does this developmental process connect with religious belief, more specifically?

Consider the notion of promiscuous teleology as described by Keleman and Rosset (2009). While these theorists are definitely not cultural psychologists, we are interested in the phenomenon. Promiscuous teleology is teleological in that a child believes a general super-agent causally underlies happenings in the world through purpose and design. It is promiscuous in that a child invokes any kind of non-specific super-agent as the primary cause. As we have elaborated more extensively elsewhere, Bakhtin wrote about how language is filled with expressions oriented to others as if they were present even when they are not (Bakhtin, 1990; see Cresswell, 2014). He articulates how we are always addressing another person in a way that invokes an unseen agent because language is always addressed to another. He points out that language includes what we ought to do in light of something like an omnipresent other. This kind of addressivity is implicit in the “we” evoked in the illustration of “no.” As “no” is enacted, it is done in a constellation of language so oriented to an ever-present other that humans address. That which CSR proponents call promiscuous teleology could just as easily be the enculturation of religious cognition in religious language practices. Even in the simple enculturation of the word “no” is the rudimentary alterity and addressivity that constitutes a sense of an ever-present other (see Poole, 2001).

In sum, a cultural-linguistic approach to the ontogenetic development of religious cognition bypasses a classic binary between cognition and body, demonstrating that how children come to think about the world cannot be teased apart from how they come to experience the world. It is through co-regulation into generic language practices that an infant acquires religious cognition such as promiscuous teleology. The result is that we cannot think of mind only in terms of mental representations, because it is fundamentally bound to the world. The flow of life with others is central to our cognition and an adequate account of cognition needs to be careful not to slide into too much of an individualist paradigm. Cognition may be emergent in minds (plural) and not in any individual mind in isolation, leading CSR into a realm that considers the sociocultural constitution of religion. A transition that cultural psychology would call for in CSR is to cease (1) talking about cognition employing symbols represented solely inside individual heads to (2) cognition emergent in language (symbols) as a much more sociolinguistic phenomenon integrated with the world. Does CSR have anything to gain from paying attention to cultural psychology?

Why CSR should pay attention to cultural psychology

Thagard (2005) is a generally accepted authority in the area of cognitive science and he outlines criteria for evaluating the efficacy of an approach to cognition. One criterion is what he calls “representational power.” This criterion asks us to consider if a given approach to cognition bypasses important functions or information that minds process. CSR stands to gain representational power if it takes the constitutive role of language practices seriously. Spelke and Kinzler (2007) write about the capacity for religious belief in terms a “number of separable systems of core knowledge. New, flexible skills and belief systems build on these core foundations” (p. 89; see also Sperber & Hirschfeld, 2004). In this view, religious belief is facilitated because of mechanisms that enable one to hold information about super agents as they are conceived in one’s milieu (see also Barrett, 2007).

As we have outlined elsewhere, CSR tends to bypass central components to religious cognition such as the content of belief and embodied experience entailed therein (Cresswell, 2014). What one believes in and how such beliefs are experienced with compelling verisimilar objectivity has little relevance in CSR because the mechanisms underlying the possibility for belief are treated as primary. When we argue that, to function and succeed in life, one must be fluent in the language practices of one’s community, we are putting forth the claim that religious belief means to be fluent in the religious language practices of one’s community. It amounts to participating in a shared reality constituted in language. Belief is a kind of language practice that is missed if it is bypassed in favor of the operation of core knowledge because the very things that belief entails are missed. That is, religious belief is entwined with the constitution of the world people jointly enact. Promiscuous teleology, for example, is taken by Keleman and Rosset (2009) to be about a general non-specific super agent. When we consider what matters to people who are religious is the development of such belief into a particular content, Christian parents would tacitly socialize a child’s promiscuous teleology to a specific belief in Jesus Christ. This content is central for understanding religion, because it is part of the reality experienced and thought about by those within a religious system. Rectifying the bypassing of it could seriously enhance representational power.

Thagard (2005) establishes another criterion as “computational power,” which refers to how well an approach deals with high-level thinking and activities like problem solving (planning, decision making, explaining) and learning. Because of the complexity of human sociocultural contexts and the complexity of our everyday lives, humans need to solve complex problems. We have to explain why the world works in the way that it does and learn to anticipate what will happen in the future to successfully get along. An approach to cognition should account for how humans learn insofar as humans dynamically adapt to life. To articulate how CSR has room to enhance the computational power—if it takes a cue from cultural psychology—we will describe two common examples in CSR.

One example is a system that enables theory of mind (e.g. Barrett, 2012; Barrett & Richert, 2003; Bering, 2006, 2011). It is speculated that children develop socially by being able to solve problems about anticipating others’ thoughts and acting on the basis of a mechanism that enables them to imagine the perspective of another person. Researchers in the CSR point out that we never actually have access to the subjectivity of someone else and so we are stuck with something like an imaginative representation of another mind. This capacity can be generalized beyond humans that we concretely deal with from day to day. The logic is that the ability to imagine what is in the mind of one’s mother can be generalized to presuppose what would be in the mind of one’s deceased ancestor. On the basis of this mechanism, humans are supposedly enabled to generate representations of minds that may not be embodied.

We argued that it is through co-regulation into religious language practices that an infant learns the experience and concomitant conceptual organization of ontological givens. Religious cognition is about problem solving in the context of participation in a set of affective practices about how the world is experienced. It is not about a subjective representation based on maturing cognitive capacity, because it is the formation of cognition in concert with others. In dynamic interactions with others, children form a way of seeing the world before they express themselves in what we normally would think of as language. A child is brought into participating in socially constituted cognition that shapes the world through the give and take of life as caregivers speak with a child. Instead of treating others’ cognition as something that we need to guess to figure out how to act, a potentially more computationally powerful approach is offered by way of showing how we participate in cognition with others (for discussion, see Robbins & Aydede, 2009). Other minds are not a problem on a culturally informed account and we have a potentially more parsimonious account of religious cognition as it plays out in higher order cognition.

Another example is a system enabling the capacity to detect agency (e.g. Barrett et al., 2001; Boyer, 2001; Guthrie, 2002; McCauley, 2011). Drawing on evolutionary psychology, researchers have argued that, of the vast array of stimuli in an environment, things that humans need to pay attention to are signs of agency. That is, if one is going to engage in a high level cognitive process of deciding what to notice, it is advantageous to notice the movement of a potential agent because it may be predatory in nature. CSR has adopted the position that humans are endowed with an agency detection device and that it is overly sensitive as, from an evolutionary perspective, false positives are the safest route to take. As such, humans are considered predisposed to signs of agency in the world around them. Along with the capacity to imagine disembodied minds, a predisposition toward agency can exacerbate this hypersensitive tendency and so set human cognition up to be optimally prepared for belief in a super-agent.

Taking religious language practices as constitutive of religious cognition means that mature functioning is to experience and conceptualize the world in a generic communal manner. It is participation in the way that a community shapes the world when members talk about it and is cultivated in children from infancy. As such, the development of religious belief is about coming to apprehend categories of experience in a generically similar way as others. What matters in the study of belief is that children learn to enact the particular world lived in a community. Such claims offer generative potential in reference to religious cognition. Understanding higher level thinking, like problem solving, requires that the linguistically shaped categories of experience are addressed because such experience is what determines stimuli (see Thompson, 2007). That is, that which counts as stimuli that could be considered agentic (or not) is shaped by religious language practices. When selecting features of the environment to think about, we need to realize that the environment is already shaped by religious language practices, making it deeply meaningful if not experientially sacred. Computational power can be seriously enhanced if the constitutive role of language practices is meaningfully taken into account because it offers insight into what the hyper-active agency detection device is predicated upon. That is, it offers insight into how such a device is predicated upon a richer and more culturally constituted approach to religion that we see in authors like Bakhtin (1990) and James (1981, 1982).

Another criterion for evaluating approaches to cognition is identified by Thagard (2005) as “practical applicability.” This criterion is about how well an approach to cognition can be used to improve the life of people. A discussion of representational and computational power above already illuminates how CSR has room to be more practically applicable. Neglecting language practices cuts off the discipline from effectively addressing religious experience and the content of belief—leaving a science effectively mute to religiosity as it is practiced. Understanding the world, as it is constituted by people enacting religious language practices, both enhances computational power and turns researchers to the practical doings of life. Thagard notes practical applicability also illuminates how people learn and an example will illuminate how this CSR could stand to be improved on the bases of this account.

Consider how religious cognition is taken to be transmitted over generations. Researchers in CSR have claimed that humans have attractor mechanisms that prepare them for some kinds of representations over others (e.g. Atran & Norenzayan, 2004; Barrett, 2008; Barrett, Burdett, & Porter, 2009; Gregory & Barrett, 2009; McCauley & Lawson, 2002; Pyysiäinen, 2002). They describe how super-agents tend to have a mix of properties where some are counterintuitive and some are not. For example, an ancestor spirit has a mind and this is normal to people but they do not have a body, which is counterintuitive. What researchers in CSR have claimed is that we have cognitive architecture that puts limits on the degree to which a super-agent can be counterintuitive. If a super-agent has no counterintuitive properties, it is not remembered as a viable super-agent and if it has too many counterintuitive properties, then it also dropped from memory. The idea is that people are endowed with mechanisms that are primed to store information about super-agents when it fits within these parameters. As such, humans are supposedly endowed with a capacity to imagine others’ minds and an overactive agency detection device that capitalizes on this capacity produces a representation of super-agents. Added to this is a mechanism that specifies what sorts of religious representations will be retained and transmitted.

Turning again to the language practices and the notion of co-regulation can illuminate how there is space for growth in this domain. We argue that it is through co-regulation into religious language practices that an infant learns the experience and concomitant conceptual organization of ontological religious givens. A cultural approach specifically focuses on the micro-dynamics of human interaction that specify learning. It provides us with an articulation of how learning of religious cognition occurs in its emphasis of the embodied give and take of life as caregivers speak with a child. That is, it allows us to address how religion amounts to socio-dialogical phenomena (Cresswell & Teucher, 2011). A child is brought into participating in socially constituted cognition that shapes their world through such participation in religious languages practices along with others. The potential for explaining how we learn religious cognition does not rely on supposed attractor positions or specifying a counter-intuitive detection mechanism. It gets us away from the problem of trying to define what counts as counter-intuitive without reference to normativity and language. A linguistically culturally informed approach offers a simpler approach that is more practically viable because it incorporates normativity.

Reframing how CSR approaches such systems in light of the linguistic constitution of cognition can enhance its representational power, computational power, and practical applicability. It is possible to capitalize on these opportunities in CSR if researchers re-tool how they approach their work by transitioning to more powerful underlying theory that resonates with pragmatic cultural psychology. Specifically, by reformulating research in ways that is not reliant on self-contained processing modules.

Moving forward with DST

Interpreting development through a lens that includes culture and language is not unprecedented in cognitive science (e.g. Maturana & Varela, 1980; Robbins & Aydede, 2009; Thompson, 2007; Varela, Thompson, & Rosch, 1992). Developmental researchers have also critiqued the notion of predetermined maturation of domain-specific modules (e.g. Elman et al., 1998; Lewis, 2000; Thelen, 2005; Thelen & Smith, 1996). A main challenge starts from the vague way that psychological phenomena are often classified as arising from “interactions” among nature and nurture. “X causes Y” arguments fail to delineate actual complex interactions between X and Y and the contributions of extraneous qualities that effect the outcome of Y (Brown & Stenner, 2009). In other words, the dichotomous thinking of nature versus nurture places a naïve restriction on understanding development by resisting the significance of interdependent complexity.

DST seeks to overcome facile binaries and enable a broader theory of cognitive development (Thelen, 2005). It has led researchers to question the linearity of development (Elman et al., 1998; Lewis, 2000; Thelen, 2005). There are, of course, milestones and plateaus that are normatively stable across developing children, but there are also discrete differences in how individual children arrive at the same abilities. Thelen and Smith (1996) demonstrated that infants develop abilities through diverse trajectories; with each infant having a unique constellation of experiences. They showed how a biologically programmed pattern of maturation located in the brain is not an impetus for the acquisition of various abilities, rather, patterns of behavior emerged through ongoing and interactive attempts at a task.

Consider the example of walking, where continual interactions among various dynamic systems all contribute to the acquisition of effective bipedal locomotion. Environmental systems include such factors as the physics of gravity, a caregiver’s response to attempted walking (which often stems from larger system like the importance society places on walking), and a caregiver’s teaching an infant how to walk. An infant’s immediate efferent systems also come into play, including body weight and proportions, nervous system and metabolic processes, a child’s motivation to walk, and previous experiences with walking behaviors.

Each of the components intertwines into a multidirectional flow that creates a unique developmental path to a common developmental milestone. Significantly, an infant’s system will not autonomously negotiate muscle movement and attempts at equilibrium necessary for walking without first being put in a position, usually by an adult, to practice these competencies. It is the lived experience of being held in a walking position—at first by the trunk of the body for stability, then by the arms and hands as internal systems begin to negotiate balance and coordination, and muscle strength—that stimulates an infant’s interior systems to organize to meet exterior conditions.

Importantly, Thelen and Smith (1996) observed the development of the capacity to walk as interdependently entwined with an array of environmental contingencies in an infant’s milieu. This argument illuminates how dispositional qualities of cognition are instilled through dynamic interaction. Figure 1 expresses, in a simplified manner, what we are arguing for in reference to developing the ability to walk. The body is coupled to the world insofar as somatosensory nerves are entwined with the dynamic forces and happenings in the world. As gravity pulls on a child, her nerves and neurons fire as they are coupled with it. This coupling could be conceived as a starting condition in the recursive system. In relation to the world-body-brain coupling, neurons fire in response to stimulate dynamic movement such as musculoskeletal coordination and attempts at balance. These responses are like finishing conditions that may not result in being optimally balanced, which means that the response is not effectively coupled to the initial conditions. A recursive cycle with different world-body-brain configurations of responses occurs. Eventually, children move to a somewhat effective coupling with a digit-first gate and this involves their system organizing itself effectively to the dynamic demands of the world. Improved coupling with the environment and increased musculoskeletal sophistication emerges into the heal-strike gait that demarcates skillful human walking. OPEN IN VIEWER

DST also provides an approach to the emergence of cognition as a phenomenon emergent in dynamic interactions among internal systems of an organism—such as the nervous, endocrine, and musculoskeletal systems—and that organism’s external environment—such as religious language practices enacted by caregivers and physics like gravity. As systems entwine, an infant begins to self-organize based on the constraints imposed by others (Thompson, 2007; Van Orden, Holden, & Turvey, 2003). For example, a young child’s desire for something would be an instance of self-organization on the basis of neural functioning, caregivers’ guidance, objects, and events in the environment. Many other factors can contribute to cognition, altering the course of its organization and function over time. These can include the history of interactions the infant has experienced that accumulate in healthy neurobiological development, the manner in which a caregiver introduces concepts in language, the cultural-historical significance of objects in the environment, and so forth. There is no hierarchical value in this paradigm as each component of the various systems works in tandem to produce a resulting ability or behavior. This also contributes to the non-linearity of psychological development, as small changes in any of the systems can lead to larger subsequent changes in the self-organizing process. Since this process is dependent on the participation of multiple systems, which can include biological, temporal, and ecological ingredients, teasing these systems apart as distinct causal mechanisms becomes an ineffectual task.

Consider, for example in Figure 2, how the foregoing would apply to the development of promiscuous teleology emerging in tandem with the practice of “no.” If a caregiver consistently says “no” when an infant reaches for a particular object, an infant will begin to organize itself in the presence of such patterns around this behavior. Internal systems could include an infant’s physical (e.g. somatic and autonomic nervous systems) and emotional feeling (e.g. central nervous systems) in tandem with a caregiver’s visual and audible actions in recurrent use of the word “no.” External systems encompass factors like the cultural significance of the object being reached for. If it is constituted as dangerous, then “no” may be more forceful or panicked accompanied by arousal (e.g. sympathetic nervous system arousal). Other external factors include a caregiver’s own conception of how an infant should be interacted with. A caregiver may gently remove an infant from the vicinity of a prohibited object, whereas another caregiver may use physical intervention to stop an infant from touching it. A caregiver furnishes a dynamic context that shapes what an infant can do and feel in a way that co-constitutes an infant’s cognitive dispositional experience, and an infant’s system organizes to integrate these experiences presented to it. OPEN IN VIEWER

When caregivers’ responses to an infant’s actions become predictable, an infant may infer these responses to other actions and “test” them accordingly (Thelen & Smith, 1996). An infant who is consistently removed from crawling under the dining table may test the generalizability of a dispositional self-organization by crawling under the coffee table (see Figure 2). If a caregiver is unresponsive in the latter case, an infant will learn there are some behaviors that will elicit a “no” and others that will not. Thus, it is through predicting, testing, and reformulating actions that an infant learns to organize her own dispositions and so navigate an environment efficiently and effectively. If a context does not support an action, an infant will reorganize behaviors in a way the context supports, such as crawling under the coffee table instead when it learns under the dining table is restricted by a caregiver.

For a caregiver, such dynamic interactions already include an orientation to the ever-present other by virtue of the structure of how language is used (see Cresswell, 2014). The nuance of the above includes the presupposition of what looks like promiscuous teleology in the expression of language that is oriented to an other. As an infant progressively self-organizes in conjunction with those around him, it will include this self-organization in line with such an ought and promiscuous teleology. The word “no” can be highly variable in its usage and linguistic richness such as tacit teleology is part of it. It is through its embodiment that the concept “no” and its emotionality becomes differentiated in tandem with higher order cognition like thoughts about an unseen agent. While an infant may have unique and variable encounters with “no,” a consistent caregiver dialogue serves to shape these experiences into a perceptual concept that resonates with those in a community. In this way, religious cognition is developmentally adaptable. It allows an individual to cognitively self-organize in concert with religious language practices and so offers a unique theory of cognitive development that includes higher level cognition as well as concomitant dispositional cognition.

CSR often maintains that the development of religious cognition is biologically pre-determined and occurs on a relatively stable maturation trajectory (e.g. Barrett & Kurzban, 2006). As an individual develops over time, additional aspects of a religious cognition blueprint are taken to become operational allowing for more complex functions. These functions are asserted to be domain-specific in that particular mechanisms were biologically programmed over the course of evolution to process and organize specific information (Cosmides & Tooby, 1994; Spelke et al., 1992). Although this strong nativist stance acknowledges influences from sociocultural factors, these influences are defined as marginal supplements to an already fixed path of cognitive development—sociocultural factors are taken as a stimulus input to cognitive mechanisms that already have the aptitude to perceive and decipher incoming information.

In contract, DST provides a theory of cognition that fits well with the notion of language practices outlined above. This approach to cognition presupposes a rich conception of religion that we see in Bakhtin (1990) and James (1982). A central issue that drove the need to consider language is the way in which CSR could improve its representational power, computational power, and practical applicability. To get at the phenomenon of religious cognition in a way that realizes these considerations, we argued that researchers can draw on DST to account for how, to function in life, one must be fluent in the language practices of their community. We also argued that mature functioning is to experience and conceptualize the world the same generic way as members of a community, that is, it means to fully participate in the religious world as it is constituted in language.

DST presupposes the necessity of caregivers using language to shape cognition and, consequently, religious cognition. We established how dynamic co-regulation implied that development of religious cognition is a sociocultural phenomenon enabled by religious language. Learning language involves developing in a way that is generic to a community, such as acting and talking in ways that are accepted as legitimate and sensible. Participation thereby means enacting language in a generic way that shapes the perceptual categories, emotion valence, and ideological compellingness generally demarcating a community. The strong form of this claim is the Bakhtinian idea that one has developed to the point of shaping the world of experience in the same way as others. The dynamic self-organization that is specified in DST amounts to participation as spelled out above. Self-organization is finding resonance with one’s milieu and this activity means to enact the same world of experience as others, which incorporates the richness of religious experience we see in James. It means that one embodies the same kind of experience as others and to organize oneself to categorize and feel about the world as one ought. Participating in a community by learning its religious language to shape ontological givens is the same thing as self-organization.

We also argue that researchers need to account for how it is through co-regulation into generic language practices that an infant learns the experience and concomitant conceptual organization of ontological givens. A cultural approach relies heavily on the dynamics of interpersonal relationships and what we referred to as co-regulation. As an infant acts, caregivers respond and in this give-and-take, we get the development of abilities to participate in a community. DST treats co-regulation as central because it points out that all behavior happens in the context of dynamic relations. Acquiring dispositions and enacting higher mental functions requires interrelations among many parts. By arguing that an isomorphic approach to developing behavior is unnatural, DST takes an approach toward religious cognition, which immediately presupposes co-regulation.

Conclusions and future directions

It is beyond the scope of this article to thoroughly rework previous research by reinterpreting CSR findings through the lens of DST. We will, instead, offer prescriptive illustrations in reference to theory of mind (e.g. Bering, 2006), agency detection processes (e.g. Barrett et al., 2001), and the retention of religious ritual (e.g. McCauley & Lawson, 2002; Pyysiäinen, 2002).

We described how one system that is taken to enable religious belief is theory of mind (e.g. Barrett, 2012; Barrett & Richert, 2003; Bering, 2006, 2011). On the basis of this mechanism, humans are supposedly enabled to generate representations of minds that may not be embodied. DST starts from a place that does not consider the central problem to be one of reading other minds that one cannot access. Because psychology is a matter of self-organization through interpersonal action, one coordinates jointly with others to establish psychological activities in situation. That is, people come to a situation with a wide repertoire of potential to organize themselves in many different ways and dynamically do so to functionally couple with others. What looks like theory of mind is actually finding language practices that resonate effectively with others. Instead of predicating CSR on research that presumes theory of mind, researchers should look to the dynamic activities as they are embedded in an interpersonal context. The central agenda would include looking at the micro-dynamics of interactions that illustrate how belief is jointly accomplished. It would involve careful consideration of the dynamic interacting systems ranging from biological systems like the parasympathetic system to grand narratives that all jointly come to play (see, for example, Xygalatas, 2012).

A second system that comes to bear when researchers in CSR study belief is the capacity to detect agency (e.g. Barrett et al., 2001; Boyer, 2001; Guthrie, 2002; McCauley, 2011). Humans are taken to be overly predisposed to signs of agency in the world around them. DST points out, in contrast, that developing cognition includes developing predispositions to act and then adjusting performance in relation to how well (or not) a disposition fits with the milieu. Human cognition is constantly shaping and reshaping as they dynamically attune to this milieu. As such, an immediate startle response to a supposed sign of agency—like a rustling shrub—is a system initially readjusting to the milieu. Research could involve looking at such moments when startle responses are coupled with ineffectual attunements to the environment. It would involve, for example, considering the moments when one’s attunement to religious cognition being practiced are not aligned or mismatched. Religious cognition could be understood as the dynamic attunement to others in the context of moments when systematic dispositions do not fit and misalignment would occur. Consequently, research would involve considering language and the grand narratives that play a role in the cultivation of dispositions that can be violated.

A third system that furnishes CSR with an explanation pertains to how beliefs can be transmitted over generations. Researchers in CSR have supposed that humans have attractor mechanisms preparing them for some kinds of representations over others (e.g. Atran & Norenzayan, 2004; Gregory & Barrett, 2009; Barrett et al., 2009; McCauley & Lawson, 2002; Pyysiäinen, 2002). DST would approach this phenomenon in terms of the capacity of a person to adjust to a milieu. If an idea is present that falls too far from the parameters that someone is accustomed to, it will require too much for them to adjust. Beliefs are transmitted through generations because successive generations are accustomed to those that are practiced. It is in this context that DST theorists talk about the natural drift of a population: cognitive structure (i.e. self-organization) changes slowly over time via small systemic changes that eventually amount to large-scale change initially too far afield to fit within a person’s scope of possible self-organization (Thompson, 2007). Future research could involve the same kind of analysis as above that carefully watches for break-downs in dynamic systems and readjustment to fit the changing milieu. This work could be conducted through, again, careful observation of beliefs as they are expressed and embodied in language.

In conclusion, combining a strong form of cultural psychology that attends to language practices with dynamic systems theory offers substantial potential to enhance CSR. This article marks one in a series spelling out “pragmatic cultural psychology of religion”. In terms of understanding religion and religious belief, our work alludes to James’ claims that belief is always about concrete life and it is deeply embodied. He and Bakhtin highlight how one cannot talk about religious belief without considering what culturally specific systems constitute the content of belief. Taking the constitutive role of language and culture seriously enables an approach to cognition that does not bypass the content of belief or the experienced ontological givens entailed in religiosity. An implication is that co-regulation precludes the necessity to concern oneself with finding an initial blueprint or design. Hard-wired maturing mechanisms are not needed when we give primary place to a cultural account. To look to the cognition entailed in religious cognition is to look to the language used by believers. We can move away from approaches that rely upon some sort of primal cause—genetic or otherwise. We can get to an approach that enables us to see cognition in a cultural light and so get at the phenomena of religious cognition in a way that offers more representational power, computational power, and practical applicability. In short, researchers in CSR can move forward in a way to better find belief.