Cristine H. Legare and Rachel E. Watson-Jones
Rituals are universal features of human behavior (Boyer & Liénard, 2006; Whitehouse, 2000, 2004). The ethnographic record is rife with evidence for exotic and seemingly unusual ritual behavior (Humphrey & Laidlaw, 1994; Whitehouse, 1995). Consider the ritual cycle of pig slaughter of the Tsembaga of New Guinea, as described by Rappaport (1967, 1984). The Tsembaga ritual cycle centers around periodic warfare between groups who compete over resources and retaliate over transgressions. The timing of warfare is closely related to the size and spread of pig herds. Alliances with other neighboring groups, usually through extended kin networks, are formed to aid in battle. There are a number of rituals performed prior to the beginning of the warfare to inform the ancestors of the intention to fight. If an amicable agreement cannot be reached through negotiations and tensions escalate, “fighting stones” are hung, indicating that debts will be repaid to ancestors and allies who will be compensated for their assistance in the fight. Hanging the stones also indicates that a number of taboos must be followed throughout the period of warfare. For example, taboos against eating particular kinds of animals and plants take effect and group members are no longer permitted to engage with members of the enemy group (even looking at the enemy is prohibited).
Pig slaughter is a key feature of the ritual cycle. Pigs are highly valuable and are never slaughtered outside of ritual contexts. At the beginning of the ritual cycle, two pigs are killed as offerings to the ancestors and are cooked overnight. On the morning of the battle, the warriors consume one of the pigs, and taboos against engaging in social and sexual intercourse with women take effect. Men cover their bodies with the ash from the fire to encourage the spirits to “come into their heads where they burn, informants say, like fires, imbuing [the warriors] with strength, anger, and the desire for revenge” (Rappaport, 1984, p. 134). The black ash also masks their faces, resulting in anonymity on the battlefield. Fighting may continue for weeks or months, but it is often interrupted by various ritual performances and mounting casualties.
Fighting typically ends through a truce between the warring groups. If a truce is reached, both groups return to their region and plant a rumbim (a local bush) and slaughter more pigs to offer to the ancestors for their assistance in the fight. After removing the ash from their bodies, all of the men place their hand upon the rumbim before it is planted in the ground. This action solidifies each individual's connection to the group and the communal land. The truce period denotes a time of repayments of debts to ancestors and allies who helped during the war, and many of the taboos remain in effect. The truce period typically remains in effect (and the rumbim remains in the ground) until there are enough pigs to sacrifice for the pig festival (kaiko). When there are sufficient pigs for the festival, the rumbim is uprooted and taboos are lifted. During the approximately yearlong pig festival, the Tsembaga host and give gifts to ally groups. During these visits, the men dance together in mass dances that last all night. The number of men from ally groups who come to dance indicates the amount of support the Tsembaga can anticipate in future fighting efforts. At the conclusion of the pig festival, the majority of the group's pigs are slaughtered and some of the meat is offered to ally groups through a fence that is ceremonially torn down at the conclusion of the kaiko. If a truce is not reached, and one of the groups is conquered and their land appropriated, the survivors take up residence with neighboring ally groups (in which case, particular rituals are enacted to secure their membership in the new group).
What function, if any, do rituals like these serve in human social groups? “The problem of ritual is the familiar ‘rationality problem’ in a new guise—old wine in a new bottle” (Sax, 2010, p. 4). Ritual is often interpreted in both popular scientific discourse and in ritual studies as action that is ineffective, irrational, or purely conventional. Rituals often represent sacred beliefs, express inner states of feeling and emotion, symbolize theological ideas or social relations, and invoke psychophysical states (Csordas, 2002; Ruffle & Sosis, 2003; Sax, Quack, & Weinhold, 2010; Shore, 1996). Yet ritual serves important social functions in human culture.
Rituals are socially stipulated group conventions that are opaque from the perspective of physical causality (Legare & Souza, 2012). Rituals are the result of “a positive act of acquiescence in a socially stipulated order,” and thus are not the product of individual innovation. “The peculiar fascination of ritual lies in the fact that here, as in few other human activities, the actors both are, and are not, the author of their acts” (Humphrey & Laidlaw, 1994, p. 5). Even when rituals are explained in the context of a certain belief (e.g., engaging in a ritual action will result in a desired outcome), there is often not an expectation of a direct causal connection between the ritual actions and outcomes (Schoejdt et al., 2013). Because humans are expert intention-readers, seeing someone engage in a detailed course of (ritual) actions gives the impression that features of the action sequence (i.e., repetition, number of steps, time specificity) have the potential to produce the intended outcome, even if the underlying mechanism responsible for the outcome is imperceptible, supernatural, or simply unknowable (Legare & Souza, 2012, 2014).
The recurrent features of ritual have been difficult to define due to the complexity and diversity of ritual forms (Rappaport, 1999). Thus, the diversity of ritual across the globe has made it difficult to establish robust generalizations about the causes and effects of features of rituals on social cognition and behavior. The historical separation between the disciplines of psychology and anthropology has also resulted in ritual becoming the exclusive domain of anthropology (Bruner, 1996). Because ritual has been primarily studied from an anthropological lens, until recently, rituals have also been studied with almost exclusively qualitative methods. While this has provided substantial insight into the diversity of ritual forms, using only qualitative methods has limited the establishment of strong causal inferences about the impact of ritual on human cognition and behavior (Rossano, 2012).
New experimental research on the social function of ritual provides fresh insight into the relationship between ritual and the evolution and ontogeny of social group cognition. Over the course of human history, the ratio of kin to nonkin has increased. With this increase in nonkin within social groups, rituals have allowed groups to remain cohesive, while reducing the need for physical and social intimacy and proximity. We argue that although the capacity to engage in ritual is psychologically prepared, rituals are a culturally inherited, behavioral trademark of our species. The structures and functions of rituals have been selected for and transmitted through a process of cultural evolution.
The first objective of this chapter is to describe the social functions of ritual within human groups. We propose that ritual aids in solving the adaptive problems associated with group living by: identifying group members, ensuring their commitment to the group, facilitating cooperation with coalitions, and maintaining group cohesion. Findings from a variety of social scientific disciples provide evidence that rituals facilitate coordinated and cooperative group action, one of the greatest challenges of group living. We also provide a psychological account of how the structure of ritual facilitates high-fidelity cultural transmission over time. Next, we examine evidence that the threat of social exclusion and loss of status motivates engagement in ritual throughout development. In the final section, we provide a psychological account of the ontogeny of ritual cognition. Prior work examining ritual and group processes has focused on adult samples (Sosis, 2000, 2003, 2005; Sosis & Alcorta, 2003; Sosis & Bressler, 2003) and mathematical modeling (Henrich, 2009). We review recent research examining the mechanisms by which children learn the rituals of their group and the cues children use to interpret the behavior of group members (Herrmann, Legare, Harris, & Whitehouse, 2013; Watson-Jones, Legare, Whitehouse, & Clegg, 2014). Combining theory and findings from these lines of inquiry promises to open up new avenues for research on ritual and the evolution and ontogeny of social group cognition.
Living in cohesive groups has helped solve the adaptive problems faced by humans (Buss, 1990; Buss & Kenrick, 1998). Living in groups decreased predation risk (Shultz, Noe, McGraw, & Dunbar, 2004; van Schaik, 1983), allowed for coordinated caretaking of offspring (Hawkes, 2014; Kaplan, Hill, Lancaster, & Hurtado, 2001), and facilitated technological innovation (Reader & Laland, 2002). Our larger-than-average primate brains (Byrne & Whiten, 1988; Dunbar, 1998) and species-specific cultural complexity (Boyd, Richerson, & Henrich, 2011) are adaptations to the demands of group living (Cosmides & Tooby, 1992; Kurzban & Neuberg, 2005). Individual fitness benefited from psychological mechanisms that facilitated coordinated problem solving and increased social cohesion (Dunbar & Shultz, 2007), for example, the capacity to understand the intentions of others, to track social relationships, and to form coalitional alliances all aid in cooperation with ingroup members (Brewer, 2007; Dunbar & Shultz, 2010; Kurzban, Tooby, & Cosmides, 2001; Tomasello, Carpenter, Call, Behne, & Moll, 2005). The evolution of cooperation also selected for tactical deception within social exchanges, in which an individual strategically changes the perception of another for his or her own benefit (McNally, Brown, & Jackson, 2012; McNally & Jackson, 2013). Large-scale sociality was facilitated by the same behavioral predispositions that allowed for the evolution of small-scale sociality (Jordan et al., 2013).
Cooperation with kin, as well as nonkin, is a core feature of human social group living (Mathew, Boyd, van Veelen, 2013; Rekers, Haun, & Tomasello, 2011; Wobber, Herrmann, Hare, Wrangham, & Tomasello, 2014). There are reasons to think that cooperation among kin versus nonkin relationships may operate differently, however. Individuals have adaptations to cooperate with those who share their genes, based on the principles of inclusive fitness, and thus the closer the genetic relatedness, the more cooperation (helping behavior) individuals engage in, all else equal (Hamilton, 1964). Psychological adaptations for tracking exchange relationships, such as reciprocal altruism and mutualism, may account for the evolution of cooperation with nonkin group members (Axelrod & Hamilton, 1981; Trivers, 1971). Individuals track exchange relationships over extended periods of time, thus allowing for selective cooperation with ingroup members.
Living in large groups introduces additional adaptive problems, problems different from those involved in reciprocal dyadic exchange, such as coordination of group members for collective action, minimizing free-ridings, increasing group commitment to joint goals, and preventing the defection of group members to rival groups. Thus, the ability to engage in cooperation is a necessary but not sufficient prerequisite for participation in goal-directed coalitional alliances (Tooby, Cosmides, & Price, 2006). The adaptive problems presented by living in large groups of nonkin in turn required the evolution of psychological mechanisms to solve them (Chudek & Henrich, 2010; Chudek, Zhao, & Henrich, 2013). We propose that rituals solve adaptive problems associated with group living by (a) identifying group members, (b) demonstrating commitment to the group, (c) facilitating cooperation with social coalitions, and (d) increasing social group cohesion.
To illustrate how ritual functions within a social system, we reference the ritual cycle of pig slaughter of the Tsembaga of New Guinea, as described by Rappaport (1967, 1984). In the following section, we use examples of the Tsembaga ritual warfare cycle to demonstrate the social functions of ritual. We will also describe how the same behavior can serve multiple functions within the ritual context.
Through providing practical and psychologically powerful markers of group membership, rituals allow identification of ingroup members. This provides important information about who is more likely to cooperate and less likely to free ride (Cosmides & Tooby, 2013; McElreath, Boyd, & Richerson, 2003). Humans can keep track of approximately 150 of their group members (Dunbar, 1992), although other estimates are somewhat higher (McCarty, Killworth, Bernard, Johnsen, & Shelley, 2000), and the preference to interact with ingroup members may be evolutionarily stable (McElreath et al., 2003). Rituals provide a demonstration of shared beliefs and behaviors. Recognizing that another person shares the same behavior and values as one's self indicates that he or she is likely to be a trustworthy reciprocator. Thus, markers of group membership facilitate cooperative interactions because they provide a marker of one's “behavioral type” (McElreath et al., 2003, p. 127). Rituals often involve special communication systems or “languages,” and thus may act in much the same way that accent acts for identifying group members in both children (Kinzler, Dupoux, & Spelke, 2007) and adults (Pietraszewski & Schwartz, 2014a, 2014b). Identifying group members is also crucial for determining whom to imitate and for tracking status hierarchies (Henrich, 2009).
An example of how rituals are used as identity markers comes from the Maring-speaking people to which the Tsembaga belong. As Rappaport (1984) notes, “The rituals surrounding the rumbim provide an additional criterion for distinguishing the Tsembaga from adjacent groups” (p. 19). For the Tsembaga, a distinguishing feature of their rituals from neighboring groups is the timing of their planting rituals. “It is on the basis of their coordination of some of these rituals and their joint and exclusive participation in others that we may distinguish the Tsembaga as a single congregation distinct from all others” (p. 19). These rituals identify the members of the group that can be trusted in future interactions. Markers of group membership can also be exploited by those who seek to gain the benefits of group membership without contributing to group-specific goals. Thus rituals that demonstrate commitment to the group act as powerful mechanisms that ward off exploitation against potential free riders.
Actions that might be considered costly, in terms of the ability to perform them and the time it takes to perform them, operate both as reliable signals that convey the signaler's commitment to the group or its beliefs, and as credibility-enhancing displays, which foster the cultural transmission of these commitments to others, including children. Consistent with costly signaling theory (Irons, 2001; Zahavi, 1975), rituals serve as hard-to-fake or honest signals of group commitment. For example, the greater the amount of costly rituals within a group, the longer that group will last (Sosis & Bressler, 2003). Similarly, Monsma (2007) demonstrated that ritual participation and resource donation are positively correlated.
Related to costly signaling accounts of ritual, Henrich (2009) has proposed that costly rituals act as credibility-enhancing displays (CREDs). CREDs provide evidence of an individual's commitment to in-group values. CREDs are important because verbally expressed beliefs and commitments are open to deception. Thus, humans may have evolved cognitive mechanisms that privilege behavioral commitment over verbal commitment. Rituals, as CREDs, provide salient evidence of behavioral commitment to groups. When rituals are costly to perform, in terms of time, energy expenditure, pain, and sacrifice, they act as signals of commitment to group values (Lanman, 2012; Whitehouse, 1996; Xygalatas et al., 2013).
For the Tsembaga, the importance of demonstrating commitment to the group is evident in the ritual taboos that are enforced at the onset and through the duration of warfare. For example, warriors engage in taboos that entail a high personal cost; they are prohibited from drinking any liquids during a battle, are required to consume salted pork, and are not allowed to engage in social or sexual intercourse with women. Similarly, community members also observe a variety of food restrictions (e.g., marsupials may not be trapped and eels may not be eaten). The competing group formally becomes the enemy through observing ritual taboos, such as prohibitions against entering their territory, speaking to a member of the enemy group, eating food grown on their land, or even looking at them. These taboos require group members to demonstrate their commitment to their local group by incurring personal cost by not engaging in typical activities during a period of intergroup conflict.
Continued protection and cooperation between allied groups is also encouraged through the costly ritualized slaughter of pigs during the kaiko. The slaughter of the group's pigs sends a signal to ancestors and allies that the group is willing to incur a cost (in terms of pork) for their assistance in the fight.
Rituals contribute to cooperative behavior with ingroup members (Sosis, 2000, 2005; Sosis & Alcorta, 2003; Sosis & Bressler, 2003) by signaling group commitment. Cooperation must be conditional and involve mutualism for group action to provide a benefit to the individual (Cosmides & Tooby, 2013). Evidence for this comes from research conducted with men living in an Israeli kibbutz where it was found that religious males who engaged in public religious rituals were more likely to cooperate in an economic game than secular males (Ruffle & Sosis, 2003). Adherents of a Brazilian religious tradition called Candomble who reported greater religious commitment were more likely to behave generously in an economic game and were also more likely to be the recipients of cooperation from other group members (Soler, 2012).
Free-riding (reaping the benefits of attaining a group goal without contributing to the outcome) and defection are two potential problems associated with collective action (Cosmides & Tooby, 2013). Because the ecological and social environment is in many ways opaque and uncertain, to determine the best behavior for any given situation, humans also use social learning biases, such as conformity to the most common behavior witnessed within a group. Conformist transmission stabilizes cooperation and punishment in social groups (Chudek & Henrich, 2011; Henrich & Boyd, 2001). Much of the Tsembaga ritual cycle, such as strengthening ties with allied groups through meat sharing, is oriented toward promoting cooperation and strengthening coalitions. Through their ritual participation, allies demonstrate that they share norms of reciprocation with the Tsembaga, and thus can be recruited in future cooperative endeavors. The Tsembaga also have means of detecting potential defectors within the ritual cycle. Before the first day of fighting, the men sacrifice two pigs to the spirits and engage in a divination ritual involving the “smoke woman” who will name the members of the enemy group who may be easily killed in the fight the next day. The members of the enemy group that are named often coincides with the shaman's “fight packages” (bags containing “exuviae”—traces of human skin and hair—of an enemy male or his father). Rappaport (1984) reports that it is often the case that the man whose exuviae is given has provoked some antagonism from his ingroup members. This usually occurs because the man is suspected of being a sorcerer and because the man “has departed sufficiently from certain approved modes of behavior to arouse covert, but not general, animosity” (p. 131). The fighting packages may provide a means for groups to punish members who do not adhere to the norms of the group, are greedy, and are likely to free ride and also functions to deter others in the group that might be tempted to free ride. Finally, by collectively seeking out and punishing free riders, group cohesion may be increased. Group cohesion is an essential aspect of collaborative problem solving that results in the achievement of group goals.
Beyond demonstrating commitment to the group and allowing for the identification of ingroup members, rituals function as mechanisms of social group cohesion, which in turn fosters the longevity of social groups. Classic ethnographies and sociological theory posit that rituals promote interpersonal bonding (Durkheim, 1915; Turner, 1969) and shared beliefs (Geertz, 1973). How ritual is connected to belief is another interesting avenue of research that is increasingly being explored using quantitative methods, a full discussion of which lies outside the scope of this chapter. Recent evidence suggests that rituals may provide a mechanism by which the self becomes “fused” with other group members through shared experiences (Atkinson & Whitehouse, 2011; Swann, Gomez, Seyle, Morales, & Huici, 2009; Swann, Jetten, Gomez, Whitehouse, & Bastian, 2012). Individuals who are highly fused with their group(s) can experience a feeling of “oneness” with the group that promotes acting for the group the same as one would act for one's self (Swann et al., 2012). Rituals also increase group cohesion because they involve shared experiences that require personal sacrifice (Atkinson & Whitehouse, 2011; Whitehouse, 1995, 2000, 2004; Whitehouse & Lanman, 2014). Simply engaging in synchronous movement (even synchronous singing) increases cooperation, self-reported feelings of connection to group members, and increased trust of group members when playing economic games (Wiltermuth & Heath, 2009).
Collective activities, with both group members and closely allied groups, feature prominently in Tsembaga ritual. Commitment to group goals is demonstrated through the shared sacrifice of observing the many taboos enacted during wartime. The massed dancing performed at the kaiko is a synchronous activity that could also amplify group cohesion. The combination of continued ethnographic and experimental research promises to elucidate the mechanisms through which ritual activity increases group cohesion. Possibly as a result of the group commitment rituals display, these kinds of activities are passed from generation to generation with high fidelity.
For cultural groups to maintain cohesion over time, there must be mechanisms for high-fidelity transmission of group beliefs, values, and practices (Liénard & Boyer, 2006). We propose that rituals facilitate high-fidelity imitation and resist individual innovation because they are socially stipulated and not interpretable from the perspective of physical causality (Legare & Souza, 2012, 2014). This makes them ideally suited to high-fidelity cultural transmission (Legare & Herrmann, 2013).
The causal opacity associated with many of the ritual elements of the Tsembaga ritual cycle contribute to high-fidelity reproduction of the cycle intergenerationally. For example, even though many of the rituals within the Tsembaga cycle were “elaborate and exotic” (Rappaport, 1984, p. 176), ritual participants could often not offer any clear explanation for the significance of the ritual procedures. Anthropologists examining ritual often remark on ritual participants' inability to articulate why the ritual they are performing is done in the specified manner, only that they must be done in the way they were done before (Bloch, 2005; Boyer, 2001; Sperber, 1975; Whitehouse, 2012). The causal opacity of ritual also invites rumination of meaning based on symbolism, supernatural beings, and metaphor (Whitehouse, 2004, 2013).
How might rituals have evolved? One possibility is that collective ritual is not the outcome of an adaptive capacity but instead is the by-product of evolved cognitive architecture and therefore an indirect consequence of its evolution by natural selection (Boyer, 2001). According to Liénard and Boyer (2006, p. 825), “a collective ritual typically activates the hazard-precaution system. Given this system and its input format, a pattern of interaction that activates them may well become attention demanding and intuitively compelling. In this view, rituals can be considered highly successful cultural ‘gadgets’ whose recurrence in cultural evolution is a function of (1) how easily they are comprehended by witnesses and (2) how deeply they trigger activation of motivation systems and cognitive processes that are present in humans for other evolutionary reasons.” Psychological mechanisms adapted for group living, such as selective social learning biases, which evolved through natural selection, may have been coopted by a process of cultural evolution. The behaviors that emerged from group living were then selected for by an ongoing process of cumulative cultural evolution (Liénard & Boyer, 2006).
Rituals are cultural adaptations to the problems of group living that are built upon reliably developing features of our social group cognition. This raises compelling questions about the process by which the elements of rituals were aggregated and honed so as to address these adaptive problems. Are rituals culturally evolved to have this adaptive fit (like blow guns and kayaks) or are they genetically evolved cognitive mechanisms like cheater-detection mechanisms or pregnancy sickness?
We argue that the organizational complexity of rituals is the result of selective cultural evolution and not from selective genetic evolution. For example, the rituals of various human groups are unlikely to be equally effective at promoting solidarity, cohesion, and cooperation. Instead, rituals likely vary within and between groups in how successfully they solve the adaptive problems of social groups. This variation in efficacy and cultural success provides the raw materials upon which different groups of individuals can pick and chose, presumably favoring those that are more effective at achieving social goals.
Evolved cultural learning biases, such as conformity bias, operating over generations, could adapt the form of ritual to local environmental challenges. For example, divination rituals may allow hunters to effectively randomize their hunting strategies (Moore, 1957) and overcome the gambler's fallacy (Henrich et al., 2001). This could be achieved by merely copying more successful hunters within the group. However, intergroup competition also shapes rituals, producing collective rituals that foster solidarity and success in intergroup competition (Henrich, 2009). This process, carried out over many generations, is known as cultural group selection (Richerson & Boyd, 2005). The cultural evolution of ritual may operate similarly to how language has evolved. Humans have evolved the cognitive machinery for language, but the lexicons of all languages continue to evolve through processes of cultural selection. Indeed, recent detailed quantitative work shows that languages vary substantially in their complexity and communicative efficiency (Deutscher, 2005).
For rituals to provide utility for social group functioning, individuals must be motivated to engage in collective behaviors that promote their inclusion and status within the group. In the next section, we examine the threat of social exclusion as a motivational mechanism underlying ritual cognition.
Collective group rituals often concern addressing, averting, and mitigating danger. Addressing perceived threat is also a common theme in many individual ritualized behaviors (Boyer & Liénard, 2006). As demonstrated by the Tsembaga ritual cycle, ritual is often associated with violence, misfortune, and dangerous activities. Magical rituals are thought to provide a means of coping with the stress of dangerous circumstances and activities. They are also thought to provide a sense of control over the uncontrollable (Malinowski, 1925/1948). Indeed, recent evidence indicates that engaging in group-specific rituals helps ease the stress of dangerous circumstances. For example, Sosis (2007) found that psalm recitation was successful in helping Israeli women cope with the stress of war. Interestingly, the most powerful component associated with the palliative coping benefits of ritual was the sense of power and community associated with psalm recitation in this population (Sosis & Handwerker, 2011). On the other side of the coin, Legare and Souza (2014) have recently provided evidence that the perception of a ritual's efficacy (its ability to bring about the desired outcome) is increased when primed with randomness (lack of control).
Many collective rituals involve prescriptive and rigid behavioral patterns geared toward averting perceived threat. Perceived threats are thought to activate mental security systems, such as the “hazard precaution system” (Boyer & Liénard, 2006), designed to signal an alarm to direct resources toward coping with the threat (Szechtman & Woody, 2004). The activation of mental security systems results in security-related behavior, of which ritual may be a part. In collective rituals, fear of potential danger of not following the ritual rules (i.e., moral threat, social exclusion, or negative outcomes) may activate the hazard-precaution system (Liénard & Boyer, 2006).
In general, implied threats to fitness (e.g., avoid snakes, spiders, large carnivores, dangerous humans, strangers, social exclusion, contamination) have been found to result in stronger adherence to in-group normative ideologies (Navarrete & Fessler, 2005; Navarrete, Kurzban, Fessler, & Kirkpatrick, 2004). Environmental and social cues that were recurrently associated with threats to fitness are likely to result in coalitional thinking and the implicit goal to foster alliances. This is because conspecific aid can be useful in addressing most threats to individual fitness. Perceived threats prime coalitional thinking and due to psychological systems geared toward enabling coordination with social groups, people endorse a stricter adherence to ingroup ideologies (Navarrete & Fessler, 2005). We propose that increased endorsement of ingroup ideology is used as a means of strengthening group bonds and increasing affiliation with group members.
Due to the importance of group membership, selection has favored individuals who engage in affiliative behaviors as a means of promoting inclusion within a group (Chartrand & Bargh, 1999; Lakin & Chartrand, 2003; Lakin, Chartrand, & Arkin, 2008; Pickett, Gardner, & Knowles, 2004). Thus, the threat of social exclusion, or ostracism, may be especially likely to result in increased affiliative efforts (Williams, 2007; Williams & Nida, 2011). Individuals who had mechanisms to anticipate and address the threat of ostracism had an advantage over those who did not possess such mechanisms. Consistent with error-management theory (Haselton & Buss, 2000), an ostracism-detection system of this kind may be geared toward overdetection, because misperceiving the threat of ostracism when it is not present is much less costly than not perceiving it when it is (Kerr & Levine, 2008; McKay & Efferson, 2010; Spoor & Williams, 2007). Simply maintaining group membership is important, yet possibly more important is achieving status within a group, as high-status individuals typically garner more resources and reproductive opportunities (Betzig, 1986; Buss, 2012). Rituals provide evidence of affiliation with social groups; they display investment in social group values and endorsement of social norms and, in some cases, may increase status within a group.
Individuals are thus motivated to participate in and accurately reproduce group specific rituals. This motivation need not be conscious and deliberate. Indeed, much research indicates that individuals engage in affiliative behaviors without conscious awareness of doing so (see Chartrand & Lakin, 2013, for a review). People unintentionally mimic the actions of others; they engage in behavioral matching or automatic mimicry. Automatic mimicry increases positive affect between interaction partners; participants who had been mimicked by a confederate, as compared to those in a control condition, reported liking their partner more and perceived the interaction to have gone more smoothly (Chartrand & Bargh, 1999). Further, when given the conscious or nonconscious goal to affiliate, participants displayed increased automatic mimicry (Lakin & Chartrand, 2003). Individuals also increase automatic mimicry following social exclusion from ingroup members (Lakin et al., 2008).
Automatic mimicry may thus have a social function, most obviously for coordination and communication purposes. For example, it facilitates the coordination of action by allowing vital social affiliation cues to be transmitted between group members (Lakin, Jefferis, Cheng, & Chartrand, 2003), and may serve as social glue. The propensity to engage in behavioral matching promotes affiliation, and vice versa, and results in a virtuous circle of automatic mimicry and prosocial attitudes that contributes to cooperation among group members (Heyes, 2013).
Despite convergent evidence across social scientific disciplines for the function of ritual in social group cognition and behavior and for the motivational mechanisms underlying ritual participation with adults, the process by which rituals are learned and come to influence group attitudes has not been studied from a developmental perspective until very recently. In the following section, we examine new experimental research on the ontogeny of ritual cognition.
The development of ritualistic behavior has important implications for understanding the ontogeny of cultural learning in childhood (Herrmann et al., 2013; Watson-Jones et al., 2014) as well as for informing our understanding of the evolution of social cognition in humans (Brewer, 2007; Caporael, 1997; Kurzban & Neuberg, 2005; Richerson & Boyd, 2005). To understand the ontogeny of ritual cognition, we must first examine the development of cognitive systems that support social categorization and social group cognition. Social group cognition develops early in human ontogeny and is developmentally privileged (Killen & Rutland, 2011). Young children view social categories as having a stable, unchanging psychological essence (Gelman, 2009; Gelman, Heyman, & Legare, 2007; Hirschfeld, 1996; Rhodes, 2012; Rhodes & Gelman, 2009). Our propensity for social categorization is so strong, in fact, that simply placing individuals into arbitrary groups creates ingroup biases among adults (Billig & Tajfel, 1973; Diehl, 1990; Tajfel, 1970; Tajfel, Billig, Bundy, & Flament, 1971; Tajfel & Turner, 1979, 1985) and children (Abrams & Rutland, 2008; Dunham, Baron, & Banaji, 2008). For example, when children experience novel social groups (i.e., based on T-shirt color) they have expectations for ingroup reciprocity, positive behavioral attributions for the ingroup, and preferences for in- over outgroup members (Dunham, Baron, & Carey, 2011). There is also evidence that children preferentially interact with ingroup members (Kinzler et al., 2007).
To coordinate behavior for cooperative efforts, children must learn and adhere to the norms and conventions of their social groups through a process of imitation (Kalish, 2005) and social learning (Heyes & Frith, 2014). Even young children tacitly accept status assignments, rules, and prescriptions and expect others to do the same (Diesendruck & Markson, 2011). They also readily engage in normative protest when rules are violated (Rakoczy, Warneken, & Tomasello, 2008). By the age of 4, children attribute conventional knowledge selectively to ingroup members (Diesendruck, 2005). Young children placed within groups expect group members to behave in conventional ways (customs, traditions, and etiquette) and can differentiate conventional from moral rules (Killen & Rutland, 2011; Smetana, 2006; Turiel, 1998).
Even infants expect members of social groups to act similarly (Powell & Spelke, 2013) and are more likely to imitate members of an ingroup than an outgroup (Buttelmann, Zmyj, Daum, & Carpenter, 2013). New research on high-fidelity imitation in early childhood indicates that imitation has evolved social functions, such as encoding normative behavior (Kenward, Karlsson, & Persson, 2011; Keupp, Behne, & Rakoczy, 2013), affiliation (Over & Carpenter, 2012), and detecting ostracism (Lakin et al., 2008; Over & Carpenter, 2009; Watson-Jones et al., 2014).
Evolved selective social learning mechanisms are attuned to detect social conventionality and promote high-fidelity imitation, a mechanism of cultural transmission (Legare, Wen, Herrmann, & Whitehouse, 2015). A growing body of research has demonstrated that as highly specialized cultural learners, children are well equipped to engage in high-fidelity imitation, a potential indicator of group affiliation through conformity (Herrmann et al., 2013). For example, there is now substantial evidence that young children readily overimitate or overcopy the behavior of others (Nielsen & Tomaselli, 2010; Over & Carpenter, 2009, 2012). Overimitation is a uniquely human predisposition; even when it is obvious that some actions are causally irrelevant to retrieve a reward from a puzzle box, children still faithfully copy all of the actions of a demonstrator, as compared to chimpanzees, who omit obviously irrelevant actions to retrieve the reward (Horner & Whiten, 2005).
Overimitation may thus be an adaptive human strategy facilitating more rapid social learning of instrumental skills than would be possible if copying required a full representation of the causal structure of an event. As a social learning strategy, overimitation may be so adaptive that it is employed at the expense of efficiency (Flynn & Whiten, 2008; McGuigan, Whiten, Flynn, & Horner, 2007; McGuigan & Whiten, 2009; Whiten, Custance, Gomez, Teixidor, & Bard, 1996; Whiten, McGuigan, Marshall-Pescini, & Hopper, 2009). The tendency to overimitate is consistent with the “copy-when-uncertain” social learning strategy (Toelch, Bruce, Newson, Richerson, & Reader, 2014). This proposal is akin to that of error management theory (Haselton & Buss, 2000) in which, in this case, the costs of not imitating with high fidelity in an uncertain situation outweigh the benefits of the reduced effort entailed in imitating with low fidelity. Children infer from the purposeful and intentional nature of an action that they are supposed to copy it (Horner & Whiten, 2005). The underlying logic of these arguments is consistent with dual inheritance theory. For example, according to the costly information hypothesis, unless the world is at least somewhat uncertain (or opaque), natural selection would not favor imitation (Richerson & Boyd, 2005).
Overimitation has also been interpreted as overattribution of causal efficacy to redundant elements or automatic causal encoding (Lyons, Young, & Keil, 2007; Lyons, Damrosch, Lin, Macris, & Keil, 2011). This interpretation has been challenged by accounts of imitation that emphasize the social and normative function of imitation (Kenward et al., 2011; Nielsen, 2012; Over & Carpenter, 2012). In fact, despite substantial psychological evidence for the early developing and sophisticated capacity to reason causally (Baillargeon, Li, Gertner, & Wu, 2011; Carey, 2009; Gopnik & Schulz, 2007; Keil & Wilson, 2000; Legare, 2012, 2014; Legare, Gelman, & Wellman, 2010), much of what people need to learn and interpret is not based on understanding physical causality and instead is based on social conventionality.
Young children are thus highly sensitive to social and contextual cues to high-fidelity imitation (Herrmann et al., 2013; Watson-Jones et al., 2014). For example, children are sensitive to cues to consensus and synchrony, potential markers of conventionality (Chudek, Heller, Birch, & Henrich, 2012; Claidière & Whiten, 2012; Corriveau, Fusaro, & Harris, 2009; Corriveau & Harris, 2010; Pasquini, Corriveau, Koenig, & Harris, 2007). Children have also been shown to conform to a group consensus in purely social situations, where no new instrumental knowledge can be gained (Schmidt, Rakoczy, & Tomasello, 2011). For example, they disguise their correct opinions in order to conform to a group consensus (Haun & Tomasello, 2011). Based on these early developing capacities, research by Henrich and colleagues have argued for an early developing “norm psychology” that supports reasoning about the conventionality of behavior (Chudek et al., 2013; Chudek & Henrich, 2010), an essential prerequisite for ritual cognition.
Additionally, children, as cultural learners, are also sensitive to credibility-enhancing displays (Henrich, 2009). For example, a child witnessing, and adhering to, the ritual taboos of the Tsembaga may implicitly use information about social categorization within their ingroup to determine that this is a social convention adhered to in times of turmoil when alliances and group affiliations are salient and important.
We propose that children and adults imitate ritual actions with high fidelity as a means of ingroup affiliation and that threats to group membership or social exclusion amplify motivation to engage in collective rituals. There is evidence that young children are highly sensitive to the threat of ostracism (Over & Carpenter, 2009; Watson-Jones et al., 2014) and that following an experience of social exclusion from their ingroup, children imitated an ingroup ritual with higher fidelity than children excluded by outgroup members or than children included by in- or outgroup members. These studies demonstrate that young children may use “affiliative imitation” as a behavioral strategy to reaffiliate with social group members when faced with the threat of social exclusion.
In sum, early developing social cognitive capacities provide the foundation for the development of ritual cognition. Young children are adept at using social and contextual cues to determine which actions are conventional and attempt to imitate these actions with high fidelity (Herrmann et al., 2013; Watson-Jones et al., 2014). Children are also highly motivated to imitate ritual as a means of affiliation with group members.
Despite the fact that ritual has been understudied from a psychological and an evolutionary perspective, convergent developments in cognitive science (Legare & Souza, 2012, 2014; McCauley & Lawson, 2002; Rossano, 2012), social psychology (Norton & Gino, 2014; Vohs, Wang, Gino, & Norton, 2013), and cognitive and evolutionary anthropology (Atran & Henrich, 2010; Boyer & Liénard, 2006; Bulbulia, 2004; Henrich, 2009; Humphrey & Laidlaw, 1994; Ruffle & Sosis, 2007; Shore, 1996; Whitehouse, 2011) have opened up new directions for research on ritual. New experimental research on the function of ritual in human social behavior provides fresh insight into the role of ritual in cultural transmission and the development of social group cognition.
Rituals serve four core functions within social groups that help address the problems of coordinated and cooperative group action associated with the ultra-sociality of our species: They (1) provide reliable markers of group membership, (2) demonstrate commitment to the group, (3) facilitate cooperation with social coalitions, and (4) increase social group cohesion. The social stipulation and causal opacity of rituals make them ideally suited to high-fidelity cultural transmission over time. We have also provided evidence that the threat of social exclusion and group affiliation motivates engaging in ritual.
Finally, we have provided a cognitive developmental account of the psychological foundations of ritual behavior. Examining the ontogeny of ritual cognition increases our understanding of the emergence of social group cognition in general and provides unique insight into high-fidelity cultural transmission over time. We propose that the capacity to engage in ritual is a distinctly human predisposition, a psychologically prepared, culturally inherited, species-specific behavior.