Abstract
Theorists and researchers in the psychology of religion have often focused on the mind as the locus of religion. In this article, we suggest an embodied cognition perspective as a new dimension in studies of religion as a complement to previous research and theorizing. In contrast to the Cartesian view of the mind operating distinctly from the body, an embodied cognition framework posits religion as being grounded in an integrated and dynamic sensorimotor complex (which includes the brain). We review relevant but disparate literature in cognitive and social psychology to demonstrate that embodied cognition shapes the way that people represent the divine and other spiritual beings, guides people’s moral intuitions, and facilitates bonding within religious groups. Moreover, commitments to a religious worldview are sometimes manifested in the body. We suggest several promising future directions in the study of religion from an embodied cognition perspective.
For many people, religion stands defiantly as the conceptual foil of the material world. At the core of many religious, ideation is the belief in a parallel reality inhabited by immaterial agents (e.g., deities and spirits); the possibility of transcending the physical body both during and after this life; and a belief in causes, effects, and forces outside the physical world. Unlike mundane knowledge of material categories and processes, religious ideas seemingly cannot be acquired through first-hand experience, and they are not logically deduced from physics or principles of nature outside the realm of religion. Indeed, religious beliefs and practices often seem to defy a natural explanation and may involve nonintuitive moral commitments (e.g., fasting) or may require the costly contribution of resources—amounting at times to sacrificing one’s life without a verifiable reward (e.g., sacred wars, suicide bombings). Thus, it remains a puzzle for many to comprehend how religious ideations have emerged, how individuals in religious communities become so tightly bound, or how beliefs are translated into behaviors.
In this article, we contribute what we think is a missing dimension in the ongoing quest to explain the psychology of religion. Although many researchers and theorists have focused on religious cognition as a mental process, we draw on recent empirical evidence in the subfield of embodied cognition to shed light on the role that the whole human body—as a structured, integrated, sensorimotor system—plays in laying the psychological foundations of religion.
Theoretical Perspectives
Previous perspectives
Theories of religion abound ranging from religion as divine revelation to religion as a human construct (Pals, 1996). Explaining the emergence of counterintuitive religious beliefs and the related constellations of costly religious practices has been a special challenge for social scientists. Some, such as Freud (1928/1961), have sought to reduce religion and religious cognition to a pathological fallacy—a coping mechanism that creates an illusory state of comfort thereby helping to alleviate societal pressures and to regulate base instincts. Viewing religion more graciously, William James (1902/2002) argued that religious beliefs are shaped following strong emotional–arousal states that are construed by the individual as transcendental. The inability to integrate such states in existing knowledge structures leads individuals to hypothesize the existence of a God, spirits, or a higher self.
Sociologist Emile Durkheim (1912/1995) has famously argued that religion is also functional in that religion binds individuals together in communities, thus allowing for the transmission of shared beliefs, morals, and sentiments from one generation to the next. Further, anthropologist Clifford Geertz (1973) argued that those shared beliefs seem to take on an “aura of factuality” (p. 90). It is through religion, others argue, that people find meaning (Park, 2005) and resolve existential anxieties (Jonas & Fischer, 2006).
More recently, several scholars in the cognitive science of religion (CSR) have drawn on cognitive and evolutionary psychology, and they have developed a naturalistic explanation of religion theorizing that humans share a tendency to believe in the existence of spiritual agents or an immaterial, parallel reality (Andresen, 2001; Boyer, 2001; Pyysiainen & Anttonen, 2002). CSR theories are often based on a modular view in which the mind is composed of a host of fitness-enhancing computational modules, each evolved to solve one of the problems recurrent in people’s ancestral environment. These specialized problem solvers are preset to detect a narrow range of the available environmental input, to process the information according to a preset algorithm, and to rapidly work out a fitness-enhancing solution. For example, all humans are said to have an agency-detection module that appears to be activated when there are new movements, sounds, or anomalies in the environment. Agency-detection favors erring on the side of false positives to facilitate the escape from predators and, thereby, to enhance survival (Guthrie, 1993). Indeed, agency may be erroneously detected in moving dots, dancing two-dimensional shapes (Heider & Simmel, 1944), or rolling marbles (Barrett & Johnson, 2003). Therefore, a “bump in the night” without a known source may sometimes be attributed to a supernatural agent.
The mental representations of supernatural agents are generated via a set of other innately endowed computational modules that are typically involved in generating mundane ordinary knowledge about the world. For instance, the folk psychology module that people use to infer the mental states of others is the same one that people draw on to consider whether a friend is sad or to reckon whether God is angry. In much the same way, the form and modes of religious rituals are explained as incidental outcomes of the workings of religious–mental algorithms subserving primarily ordinary psychological functions. One hypothesis proposed by anthropologists McCauley and Lawson (2002) is that all religious rituals are grounded in an innate understanding of intentionality (e.g., God wants to accept or heal) and, subsequently, take the form of an agent influencing a patient via an instrument to carry out those intentions (e.g., a priestly agent baptizes a child with water or lays his or her hands on the sick).
Religion as embodied cognition
A common thread in each of the foregoing theoretical perspectives is the underlying assumption that the mind is the locus of religiousness. However, this assumption may be limited by people’s own philosophy of mind and historical positioning (Snibbe & Markus, 2002). Although most psychologists would explicitly reject Cartesian views of the mind as being separate and distinct from the body (Stanovich, 1989), theory often seems to reflect a Western bias in emphasizing religiousness as being seated in the mind (cf. Cohen & Rozin, 2001; Lilliard, 1998). Consequently, we believe, the role of the body in shaping religiousness is generally understudied and certainly not well understood.
As one example, in the CSR model of religion, the human mind is sometimes metaphorically thought of as a computing machine. The machine has an important processing core and a periphery of input and output devices. Knowledge is conceived of as coded packets of information devoid of their antecedent sensory input format and their prospective motor output format. Thus, religion is sometimes viewed as being hooked up to a computational core (e.g., a human mind or a central processor) and only trivially linked to the input device (e.g., a human eye or a digital camera) or output device (e.g., a muscle or a printer).
Many psychologists have begun moving away from this dualistic Cartesian view of the mind. The embodied approach to cognition obviates the need for an epistemic divide between the bodily and the mental (Barsalou, 2008; Gallese & Lakoff, 2005; Glenberg, 2010; Niedenthal, Barsalou, Winkielman, Krauth-Gruber, & Ric, 2005). Embodied cognition posits, instead, a more parsimonious model that assigns a cognitive status to the sensorimotor system as well, by stretching its functional territory past the boundaries of the brain. As the organism attempts to purposively engage its physical or sociocultural environment, the upstream flow of sensory information that is reciprocally regulating and being regulated by the downstream flow of motor efference is captured and systematically channeled into the formation of “knowledge” structures that reference the body, the situational goal, and the environment. This emerging multimodal knowledge structure is at once conceptual and bodily, and it is deployed in nontrivial ways to serve even the most abstract of people’s cognitive capacities, such as algebraic problem solving (Landy & Goldstone, 2007), reading comprehension (Kaschak & Glenberg, 2000), inference and prediction (Barsalou, 2009; Oztop, Wolpert, & Kawato, 2005), mental imagery (McNorgan, 2012), as well as social cognition (Cohen & Leung, 2009; Meier, Schnall, Schwarz, & Bargh, 2012; Soliman & Glenberg, 2014). Research in embodied cognition posits that human cognition is not merely the result of abstract mental activities involving amodal information processing; instead, human thinking, feeling, and behavior depend to some extent on bodily experiences and embeddedness in the environment (Neisser, 1993). We believe that this more integrative approach can be useful in understanding the various dimensions of religion and religious experience (see also Barsalou, Barbey, Simmons, & Santos, 2005).
Facets of Religious Experience
Religion is a multidimensional construct (Atran & Norenzayan, 2004; Saroglou, 2011), and the body can inform different facets of religion, including beliefs about spiritual beings, moral intuitions, and the solidarity of religious communities. Further, commitments to a religious worldview can, in reverse, influence the sensorimotor system.
Embodied cognition informs beliefs about spiritual beings
One distinctive feature of religion is its assertion of the existence of God, spirits, or other supernatural agents. The supernatural is usually characterized as operating by forces and constraints different, at least in part, from those regulating the natural, physical, and material world (Boyer, 2001; Guthrie, 1993). For example, supernatural beings may be omnipresent (i.e., not constrained by space and time), omniscient (i.e., all-knowing or having special knowledge), immortal (i.e., not subject to death or destruction), or eternal (i.e., having no origin or end). Religious individuals have often depicted spiritual beings and the divine in literature, art, and artifacts (Barsalou et al., 2005; McDannell, 1995); however, supernatural agents and the spiritual world are often immaterial, invisible, and outside the range of human sensation and perception. That being the case, how can a person’s cognitive system integrate representations of supernatural beings into its conceptual repertoire in the first place?
Motor simulation and representations of divine action
One line of research points to the possibility that people use knowledge of their bodies and sensorimotor capacities and limitations to scaffold their understanding of the supernatural or divine. Despite theological teachings that God is not bound by space and time, people do seem to interpret God’s actions through the filter of their own bodily experiences—experiences that are, of course, constrained relative to the capacities of an omnipresent, omniscient deity. For instance, in one study, college students read vignettes portraying God as intervening miraculously in the world. Although the participants generally agreed that God was omnipresent, their memories of the vignettes showed a reliable trend toward an anthropomorphic characterization of God’s actions and the attribution of human physical constraints. Participants falsely recalled that God must answer prayers sequentially and must move to the location where the miracle was needed. This was equally true for both Christian and Hindu participants (Barrett & Keil, 1996).
Although the mechanism underlying these results is not clear, we suggest that an anthropomorphic representation of God is rooted in the implicit knowledge of the human body (see also Epley, Waytz, & Cacioppo, 2007). When reasoning about God in action, people might be using their own perceptual, motor, and emotional apparatus to imagine the actions of God. Research outside the study of religion supports this possibility, showing that people often engage in implicit motor simulation of behaviors when they read about the actions of others. For instance, when people read sentences describing actions involving particular body parts (e.g., kicking), the motor cortex corresponding to the referenced body part (e.g., legs) is activated. The phenomenon was discovered and then confirmed through various measures, including functional magnetic resonance imaging and transcranial magnetic stimulation (Buccino et al., 2005; Hauk, Johnsrude, & Pulvermuller, 2004; Postle, McMahon, Ashton, Meredith, & DeZubicaray, 2008; Pulvermuller, Harle, & Hummel, 2001; Tettamanti et al., 2005). In addition, when people read words describing certain actions in a particular direction (e.g., closing a drawer), actual movement in the incongruent direction (e.g., arm moving toward body) was interfered with, whereas movement in the congruent direction (e.g., arm movement away from body) was facilitated (Glenberg & Kaschak, 2002).
The results of these studies indicate that reading and thinking about an action is not simply an abstract manipulation of symbols but often involves actual implicit motor simulation. If this motor simulation process applies when people read about, think about, or imagine God’s actions, then it follows that the simulation of God’s actions might be limited by—or extensions of—one’s own human bodily capability.
A recent brain-imaging study may provide partial support to the simulation account of supernatural agents hypothesized here (Kapogiannis et al., 2009). Participants read sentences describing either a disengaged deity or an engaged, caring, and protective God. The mirror neuron system (Rizzolatti & Craighero, 2004) that exploits one’s own body representations to make sense of the perceived or imagined actions of others was found to be more active when the participants read the latter sentence type. This result is suggestive of the deployment of the participants’ sensorimotor system during semantic processing of the sentence content, but it does not indicate whether this deployment is ancillary to comprehension or constitutive of it. Participants could have been actively and deliberately imagining the reported God behavior to scaffold an ongoing primary process of abstract comprehension. Alternatively, their sensorimotor simulation could have been implicitly and automatically running to construct the primary semantic sense of what it means for a “God” to be “protective.”
Discerning among these possibilities could be done by rerunning the study while introducing a visual working memory load alongside the reading task (e.g., memorizing spatial layout of object arrays for later recall). If the differential mirror neuron system activity replicates and survives this manipulation, then this would be indicative of an automatic implicit process of sensorimotor simulation that is robust to the taxing of conscious working memory resources. Alternatively, if the mirror neuron system activity attenuates, then this would indicate that its differential involvement was primarily driven by a controlled visualization process consciously used as a strategy to aid comprehension.
Motor simulation effects might also partially explain why the emotional and behavioral dispositions of the self are projected onto the conception of the divine and other supernatural beings (Epley, Converse, Delbosc, Monteleone, & Cacioppo, 2009; Epley et al., 2007; Roberts, 1989). The attribution of an emotional state (e.g., anger) or a cognitive state (e.g., intention) to another individual has been consistently shown to engage one’s own somatosensory and motor mirror circuits (Decety & Chaminade, 2003; Keysers, Kaas, & Gazzola, 2010; Lawrence et al., 2006; Schulte-Ruther, Markowitsch, Fink, & Piefke, 2007). The attribution of emotional states has been interpreted by these researchers as an attempt to reproduce the perceived or imagined somatic correlates of the other’s mental states using one’s own motor repertoire. This process taps into one’s own sensorimotor repertoire that is not initially neutral but, instead, carries the frequency signature of one’s own sensorimotor–mental history. When humans tap into their sensorimotor system to reconstruct the bodily correlates of the mental states of others, their response is biased to be consistent with their own general tendencies and dispositions more so than may be objectively supported by the situation (Calvo-Merino, Glaser, Grezes, Passingham, & Haggard, 2006; Casile & Giese, 2006; Gutsell & Inzlicht, 2010). We suspect this bias may be even more pronounced when attempting to attribute an emotional or cognitive state to an invisible spiritual agent.
Could the body also mediate emotional and mental influences when they go in the reverse direction? Research has shown that reading or hearing about God’s authoritarian nature leads to increased aggression (Bushman, Ridge, Das, Key, & Busath, 2007), whereas belief in God’s benevolent nature has been shown to increase prosociality (Johnson, Li, Cohen, & Okun, 2013). Effects such as these may be mediated, in part, by bodily processes.
Semantic processing and comprehension of emotion words and phrases entail an implicit activation of motor states congruent with the valence conveyed by the text. Indeed, when participants cannot engage relevant facial muscles, it is sometimes difficult to process emotion-laden text. Participants who had their cheek muscle (i.e., zygomaticus major), upper lip, and nose-wrinkling muscle (i.e., levator labii) blocked by tightly holding a pen laterally in between their lips while sparing the activity of the frowning muscles (i.e., corrugator supercilii) were selectively impaired in processing concrete and abstract emotion words related to happiness and disgust, respectively, but not anger (Havas, Glenberg, & Rinck, 2007; Niedenthal et al., 2005). Havas et al. (2007) had women read angry, sad, and happy sentences before and 2 weeks after they received a cosmetic injection of Botulinum toxin that induces muscular denervation of the frowning muscle (i.e., corrugator supercilii). Reading times were selectively longer after the injection for angry and sad, but not for happy, sentences. In another set of studies, participants generating words related to disappointment reduced their posture height relative to participants generating words related to pride (Oosterwijk, Rotteveel, Fischer, & Hess, 2009). It may be that reading about God’s authoritarian or benevolent nature induces bodily facial, gestural, and postural responses in the listener/reader.
Whereas simulated actions might be qualitatively similar to the subjective experience of an action that people can imagine themselves performing, that action might deviate quantitatively. For example, God’s actions might be comprehended as occurring with faster movement across space than is otherwise physically possible. It is possible that the portrayal of God’s anger is simulated too, and perhaps accentuated in magnitude relative to human anger, thus potentially having a significant impact on human thoughts, feelings, or behavior.
Spatial perception and representation of a divine hierarchy
The supernatural, by definition, is not subject to physical constraints on space and time; however, people often represent supernatural beings as “up” or “down” (Goodenough, 2001; Meier, Hauser, Robinson, Friesen, & Schjeldahl, 2007). Aristotle has been attributed with introducing a hierarchy of ontologies that proceed downward from God to angels, humans, animals, plants, nonliving objects, and finally the Devil (Brandt & Reyna, 2011); sanctification involves the repositioning of humans into an upper position of this vertical ontological space (Pargament & Mahoney, 2005), whereas dehumanization entails a descent from humans to animals or, in extreme cases, to the Devil’s lowest position (Haslam, 2006).
The notion of psychological mapping of ontologies on vertical space has some empirical support (Schubert, 2005). Other research has shown that participants do indeed process God-related words or images faster when these are placed at the upper rather than at the lower portions of a computer monitor display, whereas the reverse pattern is obtained with Devil-related words and images (Meier et al., 2007). The visuospatial underpinnings of this effect were recently highlighted in a study showing that priming participants with words related to God and Devil caused automatic shifts of attention to the upper and lower portions of the visual field, respectively (Chasteen, Burdzy, & Pratt, 2010). Notably, neither religious belief nor degree of religiosity moderated the effect. This could be partly explained by a ubiquitous spatial representation of God and Devil—or good and evil—regardless of theological or philosophical beliefs, suggesting that religious beliefs can be grounded in embodied cognition.
Embodiment grounds moral intuitions
One function of all religions is to direct moral behavior, and religions often provide oral or written moral codes that are held to be divine commandments specifying what is moral and immoral. Research suggests that these moral codes, too, are not simply imperatives with which people may mentally agree or disagree, but, instead, moral intuitions can be undergirded by people’s sensory or motor repertoire.
Right-handed fluency
One example of embodied moral intuitions is the left–right distinction observed in religious texts. The Bible has many positive connotations regarding the right side of the body, such as “the Lord’s right hand is full of righteousness” (Psalm 48:10), “His right hand will spread out the heavens” (Isaiah 48:13), “Jesus sits at the right hand of God” (Ephesians 1:20; Romans 8:34), and “God spares the righteous on the right side” (Matthew 25:33). Although this association is relatively weak in modern day Christian doctrine, in Islam there is still a strong association and belief that the left side is the side where the Devil resides. Most Muslims do not use their left hand for daily activities (other than to clean after defecation), and it is prohibited to touch the scriptures with one’s left hand (Masud & Ajmal, 2012).
The metaphorical use of the word “right” seems to be more than a linguistic embellishment. The metaphors that people use are typically grounded in bodily experiences that later scaffold their understanding of more abstract concepts (Lakoff & Johnson, 1980). As one example of this phenomenon, Boroditsky and her colleagues (Fuhrman & Boroditsky, 2010; Lai & Boroditsky, 2013; Saj, Fuhrman, Vuilleumier, & Boroditsky, 2014) have provided experimental evidence that people use bodily experience in space to understand the abstract concept of time. Time is among the mental constructs that are so influential in people’s lives, yet the concept of time lacks a direct and succinct experiential dimension that people can easily capture through perception and action. Space, however, is a richly experiential concept and is directly grounded in the way people move their bodies as they navigate their environment, arrange and rearrange it, and interact with its social and physical constituents. Thus, people’s language often reflects reliance on the more experiential concept of space to confer meaning onto the more abstract notion of time. People metaphorically borrow attributes and characteristics from their experience in the former domain and map it to express and give structure to their understanding of the latter (e.g., “I’m falling behind schedule,” or “thinking ahead of our time”).
In a similar manner, research suggests that the conceptual link between the right side of the body and beliefs about what is moral and good is at least partly grounded in the fluency experience associated with movement of the right side of the body. Casasanto (2009) has shown that abstract concepts—such as attractiveness, intelligence, honesty, and happiness—are associated with the right side of space in right-handers and with the left side of space in left-handers. In tasks in which participants needed to choose between two products, two job candidates, or two aliens, right-handers tended to prefer the ones on the right side, whereas left-handers preferred the ones on the left (Casasanto, 2009). This effect has been demonstrated in children as well (Casasanto & Henetz, 2011).
The researchers explained that people manipulate the world with more ease on the side of their dominant hand, and acting in accord with this fluency can generate positive affect. Fluency and positive affect, in turn, can lead to judgments of truth, safety, and beauty (Reber & Schwarz, 1999; Song & Schwarz, 2009). Fluency and feelings of ease can be associated with cognitive processing, oral articulation, and bodily movements (Reber, Schwarz, & Winkielman, 2004; Slepian & Ambady, 2012; Topolinski, Maschmann, Pecher, & Winkielman, 2014); furthermore, these various modes of fluency often have equivalent effects. Because right-handedness is typically predominant in the population, the religious association between the right side and morality may be grounded in positive affective experience related to right-handed fluency.
Physical cleanliness
Beliefs about cleanliness and purity are often promoted by religious groups. For example, religious adherents may celebrate baptism in water, institute ritual washings, and use purity metaphors in religious texts (e.g., “create in me a pure heart”; Psalm 51:10). The links among cleanliness, purity, and personal hygiene were found to exceed the level of mere language use and have been associated with actual physical cleansing. Participants who were asked to recall past immoral (vs. moral) events preferred hand sanitizer over a pencil as a gift, and washing their hands after thinking about their immoral behaviors attenuated the tendency to compensate for immoral behaviors through helping others (Zhong & Liljenquist, 2006).
Similar embodied beliefs and attitudes were observed in other modalities as well. For instance, participants instructed to lie using their mouths (i.e., via voicemail) or hands (i.e., via e-mail) were willing to pay more for mouthwash and hand sanitizer, respectively (Lee & Schwarz, 2010a), relative to those in the control conditions. Participants reciprocated trust in economic games and donated to charities more so in a clean-scented room than in a neutral room (Liljenquist, Zhong, & Galinsky, 2010). Furthermore, moral disgust appears to be linked with physical disgust (Chapman & Anderson, 2013). Indeed, activating physical disgust also accentuated harshness in judging subsequent moral transgressions (Zhong, Strejcek, & Sivanathan, 2010) whether the sensory stimuli used to induce disgust was gustatory (Eskine, Kacinik, & Prinz, 2011; Rozin, Haidt, & Fincher, 2009), visual (Mortensen, Becker, Ackerman, Neuberg, & Kenrick, 2010), or olfactory (Schnall, Haidt, Clore, & Jordan, 2008).
These effects are not merely driven by semantic association between cleanliness and morality but, instead, appear to be due to an overlap in brain structures related to moral and gustatory disgust (Borg, Lieberman, & Kiel, 2008; Moll et al., 2002, 2005). In addition, research has shown that the mere simulation of cleansing can produce the same effects (Cohen & Leung, 2009). The act of cleansing seems to not only remove the guilt but other emotional residuals as well, such as cognitive dissonance and luck (Lee & Schwarz, 2010b; Xu, Zwick, & Schwarz, 2011), suggesting again that the original cleansing effect may not be driven by a mere semantic association between purity and morality.
In sum, moral intuitions are not simply driven by religious or scriptural moral codes but, instead, may be grounded in human physical experiences and natural physiological reactions. Although right-handedness and purity concerns may be emphasized to a greater degree in certain religious populations, these effects may generalize to the wider population and seem to reflect universal human experiences.
Embodied rituals facilitate communal bonding
Religion is not just about mental assent and beliefs about God, spirits, and right and wrong (Atran & Norenzayan, 2004; Saroglou, 2011). For example, the Five Pillars of Islam involve four behaviors (i.e., reciting ritual prayers, giving alms, fasting, pilgrimage to Mecca) but only one belief (i.e., there is no deity other than Allah). There are many nonbelieving Jewish individuals who cherish their genetic heritage more than their beliefs (indeed, Judaism has been characterized as a religion of descent rather assent; Morris, 1996). Although theory of mind develops in every typical human being, there is variation in the extent to which indigenous people prioritize the mind, and many view one’s actions as the center of religious concern (Lilliard, 1998). Even among Christians, for whom beliefs and mental assent are often highly esteemed, believers are also expected to gather periodically and participate in collective sensorimotor activities. Group singing, swaying, kneeling, praying, reciting, applauding, and dancing can all require the congregation to synchronize their actions in space and time.
Such synchronous interpersonal motor behavior is likely to engender feelings of bonding, and research seems to affirm this possibility. Anshel and Kipper (1988) randomly assigned participants to either joint articulatory activities (i.e., joint singing or joint poetry reading) or to equivalent passive tasks (i.e., joint music listening or movie watching). After the initial activity, participants in the former groups collaborated more in a dyadic economic game that allowed participants to freely make either selfish and personally profitable choices or partner-oriented and personally less-profitable choices. These results have been replicated and extended in a number of studies. For example, participants who sang a song delivered via headphones while perceiving only the articulatory movements of their partners allocated more funds to a public resource in an economic game relative to those in an asynchronous condition—and that at the expense of their own personal benefit (Wiltermuth & Heath, 2009). On their way out of the laboratory, participants who had drummed together helped a confederate pick up more pencils that the latter had apparently accidentally dropped than did participants who had not engaged in synchronous drumming (Kokal, Engel, Kirschner, & Keysers, 2011). Similarly, 4-year-olds spontaneously opted to help their partners in a problem-solving task after they had jointly sung and drummed more than after jointly gesturing and talking (Kirschner & Tomasello, 2010). After either synchronous or asynchronous hand tapping with a confederate, the confederate was placed by apparent random selection in a stressful situation that required laborious work. Afterward, participants who had moved in synchrony with the distressed confederate were more likely to stay and help the confederate with difficult tasks. In addition, participants rated the confederate as more similar in personality to themselves, and these similarity ratings mediated the relation between motor interaction condition (i.e., synchronous or not) and the amount of help offered (Valdesolo & DeSteno, 2011).
The quality of the prosocial outcome also appears to rely on the degree of synchrony between the actors and the effort required to achieve synchrony. The degree to which an experimenter managed to synchronize finger tapping with that of the participant was found to correlate with the latter’s feeling of affiliation with the experimenter (Hove & Risen, 2009). When experimentally manipulated, the ease and degree to which two participants were able to achieve synchronous swinging of pendulums or rocking chairs led to different ratings of how friendly and harmonious the interaction was and how well the participants formed a team in future activities (Marsh, Richardson, & Schmidt, 2009).
In addition to prosocial behaviors, memory for the experimenter’s face and utterances as well as task-relevant stimuli were found to be better in a surprise recall task administered to participants after synchronous (vs. asynchronous) bodily movement with that same experimenter (Macrae, Duffy, Miles, & Lawrence, 2008; Miles, Nind, Henderson, & Macrae, 2010). This finding is consistent with evolutionary psychologists’ argument that good memory for individuals and their past actions enhances the formation of groups of reciprocators and leads to the exclusion of free riders (Cosmides, 1989).
Other research suggests that interpersonal synchrony may induce a sensorimotor representational overlap among the bodies of the coactors (Glenberg & Soliman, 2013). After dyadic joint activity requiring interpersonal synchronous movement (vs. an equivalent solo activity in the presence of a passive partner), participants were asked to report the location of a vibration applied to one of two possible locations on their own bodies while perceiving a flash at one of two homologous body parts of their (former) partner. As predicted, participants in the joint condition showed a significant reduction in the efficiency of processing somatosensory information referencing their own bodies (i.e., the vibration location) while perceiving visual information referencing an incongruent location on the body of their partner (i.e., the flash). Presumably, the joint synchronous task induced a bodily overlap that prompted the somatosensory circuits of the participant to instantiate parameter values representative not only of his or her own body state but also that of the partner. Thus, the spatially incongruent flash and buzz stimuli (which are objectively on two different bodies) simultaneously and strongly referenced two incongruent body locations in the participant’s somatosensory circuits, leading to the processing cost and efficiency reduction. The induction of this interpersonal bodily overlap (among complete strangers) required just 5 min of sharing a synchronous laboratory task and lasted for around 5 min past the task.
These results raise the following question: How strong would be the body overlap and its physiological consequences among long-term partners who share a considerable portion of their lives interacting? Konvalinka et al. (2011) have recently provided empirical evidence for the prolonged bodily coupling among long-term partners. Physiological markers were continuously recorded from individuals participating in a Spanish fire-walking festival involving bare-footed walking on glowing coals. The recordings tracked the arousal state of the fire walker, a related spectator (spouse or friend), and an unrelated spectator. Analysis revealed a selective strong coupling of the arousal states of the fire-walkers and their relatives; their hearts were beating in synch as the former crossed the fire trail. Arguably, the long-term relation induced a prolonged bodily representational overlap among the relatives, which led them to share not only the sensory states of one another’s bodies (i.e., how hot the glowing coal is/might be) but also their physiological emotional states. They have—literally—become “one” at the neuro-somatic representational level as well as at the peripheral physiological level.
Studies in which researchers investigated the effects of synchrony and joint-body schemas might partly explain why a periodic collective activity, such as a pilgrimage to Mecca or attending church on Sunday, is likely to lead to prosocial behavior (Malhotra, 2010; Okun, O’Rourke, Keller, Johnson, & Enders, 2014). It is not surprising that active participation in religious rituals is emphasized as being part and parcel of nearly every religious faith tradition. One intriguing hypothesis is that churches with more synchronous religious rituals (e.g., vibrant singing and dancing) may facilitate the formation of more ethnically diverse congregations in which people whose languages and cultures are different from each other may join together more readily as one. Another possibility in applied research would be to examine whether religious conflicts can be mitigated via synchronous interfaith activities.
The embodiment of religious worldviews
We have argued that embodied knowledge informs representations of spiritual beings, grounds moral intuitions, and facilitates group bonding. Conversely, we contend that religious beliefs and practices can also influence embodied cognition. Religions often provide meaning and direction in life by elucidating a religious worldview: what exists, what can be known, what should be valued, how one should behave, and what the immediate and ultimate goals of life ought to be (Johnson, Hill, & Cohen, 2011). A worldview is intended to guide the religious adherent as he or she navigates through life events and frames the bits and pieces of personal experiences into meaningful wholes.
Although many religions have systematic theologies intended to shape a devotee’s worldview via semantic propositions, it is likely that these beliefs, values, and goals do not simply remain in believers’ minds as semantic propositions or as a network of ideological concepts but, instead, become embodied in their sensory motor apparatus. In other words, as an individual begins to think, feel, and behave in accord with the teachings of a particular religious worldview through repeated social and physical interactions, the original semantic propositions may “incarnate,” becoming embodied and a part of one’s procedural memory just as a seasoned soccer player follows the rules of the game without consciously acknowledging where the hands should be positioned, which way to run, or the importance of driving the ball into the net.
Empirical support for the sensorimotorization of religious worldviews comes from several studies demonstrating differences between individuals from holistic religious groups versus secularists. Holism involves a belief in the interconnectedness of all things in the world, and it stresses the illusory nature of its apparent phenomenological separation and the segregation among beings. Causality is distributed and is best understood as a wave propagation rather than atomistic collision, and the world is represented as being in a continuous state of flux (Choi, Koo, & Choi, 2007; Varnum, Grossmann, Kitayama, & Nisbett, 2010).
Holistic beliefs appear to uniquely shape the sensorimotor apparatus of the practicing Buddhist (Colzato, Hommel, van den Wildenberg, & Hsieh, 2010; Colzato, van den Wildenberg, & Hommel, 2008). When required to visually attend to local patterns composing a global pattern in a Navon task (e.g., a “T” made up of “Vs”), Buddhist Taiwanese participants, more so than atheist Taiwanese participants, were distracted by the global pattern, and they found it more difficult to suppress the global percept and, therefore, were less able to selectively attend and respond to the local constituents. Likewise, when paired with a coactor in a joint task that required each of the dyad members to respond to only one aspect of a common stimulus, the Buddhist participants were not able to ignore the aspects of the stimulus assigned to a partner. Trials with stimuli that referenced the partner affected the slow responses of the Buddhist participants more so than the atheist participants, indicating a stronger dyadic representational body merger for the Buddhist participants.
In contrast, Calvinism teaches a modular view of life in which every sphere has its own characteristics and responsibilities. Dutch Calvinists and atheists were exposed to a rapid succession of visual stimuli and were required to report all instances of a prespecified stimulus type (Colzato, Hommel, van den Wildenberg, & Hsieh, 2010). In this attentional blink paradigm, participants typically failed to detect and report the second instance of the target stimulus in the stimuli flow where the second target stimulus was presented following a very short time lag after the first target stimulus. This task is intended to demonstrate a global versus local processing style; people with local processing style overallocate attentional resources to the first stimulus, becoming deficient of attentional resources to allocate to the second target stimulus and, therefore, perform worse than people with a global processing style. Consistent with the modular worldview of Calvinism, the magnitude of the attentional blink (missing the second target stimulus) was significantly higher for Calvinists relative to matched atheists (although the first instance of the target stimulus was equivalent). The frequency of daily prayers correlated positively with the magnitude of the blink, suggesting that sensorimotor simulations of a religion-compatible perspective on the world could contribute to the tuning of the emergence of these perceptual dispositions. Researchers obtained similar results in action control tasks, such as the Simon task, which measures the ability to inhibit conflicting responses (Hommel, Colzato, Scorolli, Borghi, & van den Wildenberg, 2011).
These studies demonstrate that worldviews are not merely the sum of propositional theological systems but that religious teachings may actually become embodied in a practicing believer’s sensory-motor system.
Future Directions
We have reviewed previous research in support of our argument that embodied cognition shapes the different dimensions of religious experience, including the way people represent the divine and other spiritual beings, moral intuitions, and feelings of belonging within religious groups. Additionally, and in the reverse direction, research suggests that religious beliefs can become incarnate in the bodies and behaviors of religious adherents. We believe that new directions for research and many interesting questions and predictions can be generated taking an embodied cognition approach in the study of the psychology of religion.
Discourse analysis and the construction of shared knowledge have been a research priority in the social sciences (e.g., Berger, 1967). One interesting extension of this research and theory will be to investigate how reading scriptural texts may translate into whole body phenomena (and not mere beliefs), thus exploiting the action–sentence compatibility paradigm previously discussed (Glenberg & Kaschak, 2002). For example, participants could be requested to judge whether a certain sentence (e.g., “God lifted the burden of worry from his shoulders”) is grammatically correct by moving their dominant hands either upward or downward to push one of two vertically aligned response buttons. An anthropomorphic representation of the spiritual agent that automatically mobilizes relevant motor commands from the listener’s/reader’s motor repertoire to make sense of the described action would be expected to incur processing cost (i.e., greater latencies) when responses required hand movements incongruent with the direction of motion implied by the sentence.
Using experiments, researchers could also examine the ability of environmental stimuli to induce illusory, biomechanically impossible body states sometimes reported as visions or out-of-body experiences (e.g., James, 1902/2002). In principle, the neural sensorimotor representations of the body are inherently malleable and can extrapolate beyond past experiences, evoking a direct, bodily awareness of unnatural qualia. For example, applying a slight vibration to the distal tendon of the bicep typically engenders an illusory feeling of forearm extension (Goodwin, McCloskey, & Matthews, 1972). The illusory extension can exceed the possible range of angular displacement permitted by the elbow joint and could be accompanied by feelings of anatomically implausible body–part relations (e.g., the back of the hand touching the back of the forearm; Craske, 1977; Lackner & Taublieb, 1983). Studies in nonclinical human populations indicate that individuals can also identify with a rubber hand, can be induced to experience a supernumerary limb, and can even swap their bodies with a manikin and undergo an out-of-body experience to view and interact with their (genuine) bodies from a second-person perspective (Botvinick & Cohen, 2014; Guterstam, Petkova, & Ehrsson, 2011; Newport, Pearce, & Preston, 2010). Experimental paradigms such as these may be used to better understand the reported visions, perceptions of spiritual agents, or out-of-body experiences experienced by religious adherents.
More studies are also needed in which researchers investigate the links among somatic experience, morality, and group solidarity. As one example of a future direction, research by Xygalatas et al. (2013) has highlighted the need to better understand how painful religious rituals may affect prosociality. Although few people will endure being pierced with spears, as in Xygalatas’s sample, many religious individuals subject themselves to other less painful rituals, such as kneeling, fasting, or prostrating themselves in prayer. How might focusing on one’s body for either pleasure or pain affect spiritual inclinations, moral intuitions, and values? Indeed, because much of the emphasis in the study of religion has been on religious beliefs, there is a paucity of psychological research and theory regarding religious ritual (Whitehouse & Lanman, 2014). It is conceivable that the synchronous activities that the followers of a given religion perform serve the same function in fostering group cohesion. Synchrony may induce an overlap in group members’ bodily representations, which, in turn, may engender coupled physiological arousal states and contingent feelings of compassion and affiliation. However, the cascade of mechanisms and processes contributing to this phenomenon awaits further examination.
If religious worldviews can, in some ways, become incarnate, then understanding religion from an embodied perspective may also help explain why religion is often associated with better physical health and longevity (McCullough, Hoyt, Larson, Koenig, & Thoresen, 2000; Newberg & Waldman, 2009). An embodied perspective becomes especially important as increasing numbers of people are embracing different forms of spirituality and religious practice (e.g., Buddhism) that emphasize the interconnectedness of the body and environment (Barsalou et al., 2005) or more holistic approaches to healing (Linders & Lancaster, 2013).
Certainly more work is needed to ascertain whether one’s religion leads to the development of a more holistic mindset or whether, instead, people gravitate toward a particular religion deemed to be congruent with preexisting worldviews. Along those same lines, it will be important to understand how, and to what extent, aspects of a sensory-motorized holistic versus modular worldview might influence an array of human interactions, including education, consumption, jury decisions, or organizational behaviors. For instance, are religious people with a modular perceptual bias more inclined to hold stereotypes and strong categories? Would these same individuals achieve greater success in careers or academic disciplines that favor modular cognitive processes?
Many other questions remain: How might embodied mechanisms shape the varieties of spiritual experiences? Do some religions capitalize on the embodied approach more than others? How does contemporary use of space, art, or music facilitate—or impede—embodied religious experience? Can an embodied perspective account for similarities across otherwise diverse belief systems in terms of the constraints and affordances of the human body? We see these as some of the many exciting possibilities for future research as psychological researchers move toward a more inclusive view of the mind as a sensorimotor, whole body phenomenon.
Conclusions
Theologians, philosophers, anthropologists, and sociologists have all had much to say in the way of explaining religion, how it originates, why it persists, and what function religion might serve. Most recently, evolutionary psychologists have taken a turn at understanding religion by focusing on religion in terms of the modularity of the mind and how religious beliefs relate to and solve ordinary social problems common to all individuals. This has been a fruitful approach. However, the body and mind are inextricably linked (at least during this life); yet, very little theory and research have been advanced taking this more integrative approach. Here, we have offered suggestions about how the body, as a sensorimotor system, composed of both brain and body, also informs religious beliefs and practices—and, in turn, how religious beliefs might become embodied. We do not claim that all religious experience, or that every facet of religion, can be accounted for from an embodied cognition perspective. However, researchers investigating the links between embodied cognition and religion have the potential to provide an important complement to current theories of religion in psychological science.
Footnotes
Declaration of Conflicting Interests
The authors declared that they had no conflicts of interest with respect to their authorship or the publication of this article.