Steven L. Neuberg and Peter Descioli
People are often prejudiced against foreigners. They are also prejudiced against those who are obese, physically disabled, or elderly. They are prejudiced against people with schizophrenia and young outgroup men. They are prejudiced against those on social welfare and those who are so wealthy that they and generations of their descendants will never need such help. They are prejudiced against gay men, religious fundamentalists, atheists, and members of this or that political party or advocacy group. As the English essayist Charles Lamb wrote in Imperfect Sympathies, humans are a “bundle of prejudices” (Lamb, 1821).
Why? Traditional theoretical approaches posit that prejudices, stereotypes, and discrimination result from a range of proximate processes, including simple ingroup–outgroup categorizations, desires to boost one's self-esteem, authoritarian values, fear of death, and the need to justify oppressive actions against others. We will see, however, that people think about and behave towards others in ways that these approaches cannot predict.
We suggest that understanding prejudices requires a deeper theoretical framework—an evolutionary psychological framework. From this perspective, prejudices, stereotypes, and discriminatory behaviors can be viewed as functionally organized strategies designed to manage the threats posed by the human forms of sociality.
Life was challenging for our ancestors. Food was often scarce, unpredictable, and difficult to extract and secure. Predators and pathogens caused injury, incapacitation, and death. In the face of such challenges, individuals who cooperated with others gained significant reproductive advantages over more solitary, independent individuals (Campbell, 1982; Richerson & Boyd, 1995). Thus, human sociality reflects an evolved set of adaptations to provide safety from danger and to exploit opportunities in challenging environments (Brewer, 1997; Brewer & Caporael, 1990; Leakey, 1978).
Human sociality can also be costly, however. Proximity increases exposure to contagious diseases, and makes people susceptible to theft and violence; cooperation makes contributors vulnerable to others free-riding on their efforts. One general approach to managing the benefits and costs of social life is discriminate sociality—the careful selection of social partners (Kurzban & Leary, 2001). Indeed, our choices of social affiliates are far from random, and tend to favor those presenting cues suggesting they'll provide more benefits than costs. We're more likely to select partners who appear to be kin, cooperative, and trustworthy; who are able to coordinate their efforts with ours and are available for future interactions; and who offer other beneficial traits (for reviews, see Kurzban & Neuberg, 2005; Neuberg & Cottrell, 2008). Selecting some individuals into one's coalition and excluding others constitutes one form of discrimination. As we'll discuss, many cues that heuristically identify individuals as potentially costly social partners also constitute the basis of many contemporary social prejudices.
We focus, however, on two other sets of processes. First, we explore the evolved psychological mechanisms by which individuals (1) identify those who afford fitness threats and opportunities and (2) respond to them in threat-mitigating and opportunity-enhancing ways. These affordance management systems (Gibson, 1979; McArthur & Baron, 1983; Neuberg, Kenrick, & Schaller, 2010, 2011; Zebrowitz & Montepare, 2006), contribute significantly to stigma, prejudices, and discrimination. Second, people create within-group coalitions to counter threats posed by other group members, and we explore the implications of alliance-based processes for prejudices and group-on-group conflict. We then extend these analyses to understand prejudices against foreigners, especially as these prejudices manifest in warfare and issues of immigration and emigration. Last, we discuss the implications of evolutionary approaches for reducing prejudices and intergroup conflict. Throughout, we'll see that by identifying new prejudice phenomena and by anticipating undiscovered nuances in known phenomena, the evolutionary approach poses significant challenges to traditional social psychological and sociological approaches.
An affordance-management view holds that prejudices, stereotypes, and discrimination are responses for managing recurring threats to reproductive fitness over human evolutionary history (Schaller & Neuberg, 2012). Like all affordance-management systems, threat-management systems share a common template: Cues in the environment heuristically (and imperfectly) imply specific threats. These perceived threats, in turn, elicit a suite of functionally relevant cognitions, emotions, and behavioral inclinations designed to manage the threats.
There are several important implications of this perspective, each of which we expand on: (a) Qualitatively different psychological systems are likely to have evolved to manage different threats. (b) Different threats are cued by different kinds of information. (c) Upon activation by cues, these distinct systems engage qualitatively different prejudice syndromes of specific cognitions and beliefs (i.e., stereotypes), emotions (i.e., prejudices), and behavioral inclinations (i.e., discrimination). Thus, to the extent that different groups are perceived to pose different threats, they are likely to be targeted by different prejudice syndromes. (d) Threat management systems are biased to avoid costly mistakes, erring on the side of perceiving greater (rather than lesser) threat; consequently, people err on the side of discriminating against individuals who, actually, may afford no threat at all. And (e) deployment of prejudice syndromes depends on an individual's perceived vulnerabilities to particular threats. People who feel vulnerable to different threats will engage different prejudices and forms of discrimination.
Fitness is the extent to which one's genes are passed into subsequent generations. From this perspective, mechanisms that led our ancestors to be attuned to cues suggesting threats to their (and to their kin's) physical safety, to their ability to acquire necessary resources (e.g., food), and to their health would have been adaptive. Indeed, much of the research from an evolutionary approach has focused on prejudices towards those perceived to threaten others via physical violence or disease (Schaller & Neuberg, 2012).
Because our ancestors benefited from coalitional action, people are also expected to monitor threats both to coalition resources (e.g., access to territory) and the coalition's operational integrity—the social structures that enable coalitions to be effective. Effective coalitions tend to exhibit trust, reciprocity, common values, socialization practices, and authority structures for organizing individual effort and distributing group resources (e.g., Brown, 1991). As a result, people should be wary of those who threaten these group structures (Neuberg, Smith, & Asher, 2000).
How can one anticipate whether someone is likely to threaten safety, health, resources, and the like? One cannot directly perceive another's pathogens or intentions to harm. Rather, people must rely on cues—features of morphology, behavior, or reputation—that correlate (even if only weakly) with threats.
Because threats are qualitatively distinct—threats to physical safety are different, for instance, than are threats to fair trade—the cues implying different threats will also often be distinct. For example, threats of violence are cued by features (imperfectly) implying the capacity to do harm (e.g., prominent upper-body musculature, maleness, presence of a weapon) and the intention to do harm (e.g., angry facial expressions, looming approach, maleness, and outgroup-linked features related to morphology, language, skin color, clothing; e.g., McDonald, Asher, Kerr, & Navarrete, 2011; Navarrete, McDonald, Molina, & Sidanius, 2010; Navarrete, Olsson, Ho, Mendes, Thomsen, & Sidanius, 2009; Sell, Tooby, & Cosmides, 2009). Threats to health are cued by features linked to pathogens (e.g., skin lesions, coughing spasms) and relevant behavioral and physical abnormalities (e.g., nonfunctioning limbs, facial scars, extreme thinness or obesity; Kurzban & Leary, 2001; Oaten, Stevenson, & Case, 2011; Park, Faulkner, & Schaller, 2003; Schaller, Park, & Faulkner, 2003; Tybur, Lieberman, Kurzban, & DeScioli, 2013), and by features that increase contact with pathogens (e.g., lack of normative hygiene and food preparation practices). Threats to group integrity are cued by features suggesting, for instance, an unwillingness to contribute to collective group action (e.g., facial morphologies suggestive of untrustworthiness or membership in another coalition; Yamagishi, Tanida, Mashima, Shimoma, & Kanazawa, 2003; Zebrowitz, Voinescu, & Collins, 1996) or an inability to do so (e.g., physical features and behaviors implying physical or mental disability). In short, different threats are implied by different cues.
If social perception is designed to manage threats and opportunities, then the perceived association between cues and affordances will reflect, to a nontrivial degree, actual associations. Indeed, many stereotypes are meaningfully accurate (Jussim, Cain, Crawford, Harber, & Cohen, 2009; Swim, 1994). Of course, stereotypes are not perfectly diagnostic but rather are statistically associated such that perceivers can make more predictive inferences by using stereotypes (even when this also harms stereotyped individuals). For example, to hold the stereotype that young men are competitive implies that maleness and youth are statistically correlated with competitiveness. Although there has been much research on the content of stereotypes, only recently has an evolutionary approach been used to better make sense of this content (Neuberg & Sng, 2013).
Consider, for example, sex and age stereotypes as traditionally represented by the social psychological literature. People are seen as stereotyping men (and young people) as competitive and agentic, and women (and elderly people) as communal and caring (e.g., Eagly & Steffen, 1984; Hummert, Garstka, Shaner, & Strahm, 1995). Moreover, these sex and age stereotypes are conceptualized as independent of one another. Recent research suggests, however, that people's actual stereotypes are much more nuanced than this.
As predicted from a life history perspective (Kaplan & Gangestad, 2005; Stearns, 1976), the means by which males and females accomplish their major life tasks (e.g., growth, learning, mating, parenting) differ as they age. For example, because of differential parental investment (Trivers, 1972), the sex difference in competitiveness is greater in those years in which mating (relative to parenting) is prioritized. Second, competition tends to be intrasexual—it is directed toward (relatively young) adults of one's own sex. If the task of the social perceiver is to manage the threats and opportunities posed by others, stereotypes should be attuned to these nuances—to the ways in which sex and age interact to drive strategic behaviors, and to the fact that others' strategic behaviors tend to be focused on some targets and not others.
And they are (Sng, Williams, & Neuberg, 2015). Instead of possessing independent sex and age stereotypes, people actually possess interactive “SexAge” stereotypes of the specific forms predicted. Moreover, instead of possessing stereotypes in the form of general traits (e.g., “men are competitive”), people possess directed stereotypes—stereotypes that account for whom stereotyped behavior is directed towards (e.g., “men are competitive towards young men”). Stereotypes are not only more complex than suggested by the traditional literature, but sometimes contradict these previous theories. For example, rather than holding the stereotype that women are less competitive than men, people actually believe—accurately—that women are more competitive towards young women than are young men.
Because social perception relies on imperfect cues, errors are inevitable. Although all errors are potentially costly, some are more costly than others. Social perceivers are biased toward reducing inference errors most costly to reproductive fitness (Haselton & Buss, 2000; Haselton & Nettle, 2006), as illustrated by analogy to a smoke detector (Nesse, 2005). Just as smoke detectors are biased to err on the side of false positives, so too are evolved threat-detection systems designed to err on the side of assuming threats when there are none, rather than missing (potentially fatal) threats. For prejudice, this means people will be biased to overperceive the threats that others pose.
Many people and groups who actually pose no risk may thus be perceived as threatening. For instance, because pathogens often altered body shape and movements, people who are extremely overweight, have limited control over limbs, or are otherwise physically atypical may be (unconsciously) perceived as pathogen risks—even when no actual risk is present. Similarly, because individuals were, ancestrally, indifferent to the welfare of members of other groups, people who bear marks of “outgroupness”—for example, unusual accents or the practice of different rituals—may be perceived as untrustworthy, even if these individuals are actually highly invested in the groups they're entering. “Better safe than sorry” is the operating principle of the contemporary human mind as it perceives such individuals and groups and enacts prejudices against them (Schaller & Neuberg, 2012).
Behaviors that mitigate some threats may do little to mitigate others. Whereas a physical confrontation might help get money back from a cheat, it's unlikely to prevent pathogen transmission from a disease-carrying individual. There are good reasons to expect that different behavioral routines evolved to address different perceived threats.
Emotions play a critical role in driving functional, threat-relevant behaviors. Fear, disgust, and anger serve as alarms that interrupt ongoing activities, focus attention, and activate behaviors to address threats. For example, when we perceive a large object moving rapidly towards us, we feel fear and become aware of danger while physiological systems generate a burst of energy and send blood to the large muscles. This syndrome of responses prepares us to flee or fight, thereby mitigating threats of predation. In contrast, smelling dead flesh leads us to feel disgust, constrict our nasal passages, turn away, and create physical distance—all of which, in combination, reduces our exposure to contagious disease. In sum, different threats elicit different emotional alarms and accompanying functional syndromes of cognitions, physiological responses, and behavioral routines (Izard, 1991; Plutchik, 1980; Roseman, Wiest, & Swartz, 1994; Tooby & Cosmides, 1990).
One would predict that responses to different categories of people will also often be very different (Cottrell & Neuberg, 2005). For example, because many Americans perceive Muslim fundamentalists as threatening physical safety, they show fear and mistrust in encounters with them. Because facial disfigurements are (unconscious) cues for disease, people respond with disgust, implicitly activate disease concepts in memory, and avoid physical contact. Because reciprocity creates a vulnerability to cheating, people show anger toward those seen as taking more than their fair share (e.g., welfare recipients), stereotype them as lazy or selfish, and implement policies such as reducing welfare programs and taxation. Because shared values facilitate coordination and socialization into group norms, deviation from these values elicits contempt, disgust, anger, accusations of immorality, and discriminatory actions to exclude and disempower these individuals. Indeed, research shows links between perceived threats, emotional responses, and functionally related discriminatory behaviors (e.g., Cottrell & Neuberg, 2005; Cottrell, Richards, & Neuberg, 2015; Cottrell, Richards, & Nichols, 2010).
Traditional theoretical approaches to prejudice (e.g., ingroup–outgroup, social identity and other self-esteem-based theories) are unable to explain why there exist qualitatively different prejudices toward different groups. From these perspectives, prejudice has been viewed and operationalized as a general undifferentiated attitude towards groups and their members—as simple valence: We like or dislike others, view them favorably or unfavorably, and so on, and prejudice is assessed using “thermometer” measures of how “warm” or “cold” participants feel toward different groups, aggregated responses to “favorable” versus “unfavorable” statements about groups, and implicit associations between groups and “good” or “bad” stimuli. Yet when measures of specific emotions are assessed, rather than only valence, people show textured feelings and beliefs about groups that can look quite different for different target groups—even for groups that elicit similar reactions on traditional valence measures (e.g., Brewer & Alexander, 2002; Cottrell & Neuberg, 2005; Esses, Haddock, & Zanna, 1993; Fiske, Cuddy, Glick, & Xu, 2002; Mackie, Devos, & Smith, 2000). Such findings, and the fact that these different emotion profiles are predicted by the different threats people perceive these groups as posing, challenge in fundamental ways traditional theoretical explanations of stigma, prejudice, stereotypes, and discrimination.
Evolved threat-management systems can increase reproductive fitness but can also be costly to deploy. Discriminatory behavior is metabolically costly, distracts one from potential opportunities related to other goals, and can result in lost opportunities for future cooperation with targeted individuals. Moreover, discrimination against people can lead to confrontations that cause injury and damage to reputations, relationships, and coalitions. To minimize these costs, threat-management systems are expected to be engaged primarily when the expected benefits outweigh the damage they might cause.
This benefit–cost ratio will be more favorable when perceived vulnerability is high due to features of the environment and the individual's dispositional concerns. Threat-management systems should be sensitive to cues—external or internal—of apparent threat, and engage strenuously in threat mitigation when vulnerability appears to be great but not when vulnerability appears to be low. There is now much evidence for this form of functional flexibility.
Consider, for example, self-protective concerns. Given the costs of hypervigilance, preparation for flight or fight, and fearfully avoiding others, self-protective mechanisms are likely to be engaged only when cues suggest higher-than-usual risk of danger. Indeed, experiments show that activating perceptions of danger alters a wide range of cognitive and affective processes in ways that bias perceivers towards not missing potential threats (Neuberg & Schaller, 2014). For example, increasing people's felt vulnerability to violence increases perceptions of anger in neutrally expressive faces of young outgroup men and increases the likelihood that people identify ambiguously categorizable persons as outgroup members (Maner et al., 2005; Miller, Maner, & Becker, 2010). Moreover, just as being in a darkened room (a cue for humans of vulnerability to physical attack) increases the intensity of the startle response to a blast of noise (Grillon, Pellowski, Merikangas, & Davis, 1997), being in a darkened room increases for North American Whites and Asians the activation of stereotypic links between Black or Iraqi persons and danger concepts such as “hostile” or “criminal” (Schaller, Park, & Faulkner, 2003; Schaller, Park, & Mueller, 2003). It's instructive that this effect holds primarily for those who dispositionally view the world as a dangerous place—who especially view themselves as vulnerable to violence. Moreover, the activation of these threat-linked concepts is specific to outgroup targets (and not to ingroup targets), and to physical-threat stereotypes (and not to other equally negative, but nonthreat, stereotypes, such as lazy and ignorant). These nuances reveal the functionally focused nature of the system.
Similar functional flexibility is apparent in people's responses to cues for contagious disease: For people who feel especially vulnerable to infectious disease, or when people are in circumstances that make salient the presence of pathogens, disease-avoidance prejudice syndromes are prone to activation. For instance, when disease concerns are salient, people focus greater attention to blemished or disfigured faces (Ackerman, Becker, Mortensen, Sasaki, Neuberg, & Kenrick, 2009; Miller & Maner, 2012). Concerns with disease, whether dispositional or situational, also increase negativity toward individuals with cues of disease such as people with asymmetrical faces (Little, DeBruine, & Jones, 2011; Young, Sacco, & Hugenberg, 2011), who are significantly overweight (Kenrick, Shapiro, & Neuberg, 2013; Park, Schaller, & Crandall, 2007), or who belong to unfamiliar outgroups (Faulkner, Schaller, Park, & Duncan, 2004; Navarrete, Fessler, & Eng, 2007; see Schaller & Neuberg, 2012, for a review).
Although less research is available, there is also evidence for functional flexibility in the application of prejudice syndromes within other threat-management systems as well. For instance, when concerns about resource scarcity become salient, people categorize others to exclude more ambiguous-looking individuals (i.e., apparently biracial persons) from their ingroups (Rodeheffer, Hill, & Lord, 2012), and when concerns about economic competition are made salient, prejudices are heightened especially against groups stereotypically viewed as strong economic competitors (Butz & Yogeeswaran, 2011). Moreover, prejudices against groups seen as threatening group values may be particularly pronounced when concerns about socialization practices are salient, as when prejudices against gay men are particularly strong when heterosexuals think about them within the context of socialization domains (e.g., elementary schools, religious institutions; Saad, 1996).
We see, then, that the engagement of threat-management systems and their functional prejudice syndromes are directed specifically toward targets who exhibit cues for specific threats and especially under circumstances in which people perceive their own vulnerability to the threat in question.
The findings briefly reviewed pose a great challenge to traditional theories of prejudice and stereotyping. Those approaches lack the conceptual architecture to a priori account for the nuanced psychology people actually possess: that people apply different stereotypes, prejudices, and discriminatory inclinations toward different groups, based on the specific threats these groups are seen to pose; that these sets of responses are enhanced when people feel themselves to be vulnerable to the particular threats; and that people's responses to apparently quite distinct groups (gay men, obese persons) are nonetheless similar because they are at least partially generated by the same threat-management systems. In contrast, the approach we highlight here impressively predicts these findings.
We have focused on the evolved psychology through which individuals attempt to manage threats posed by other individuals. Some of these threats are inferred from cues that others may belong to out-coalitions—groups of allied individuals working toward common interests, who support one another in disputes against individuals outside their coalition. Indeed, by definition, members of coalitional outgroups will generally be more invested in their own groups than in one's own, and it's thus reasonable for individuals—as individuals—to stereotype members of other coalitions as less trustworthy, more willing to engage them in physical conflict, and so on. We turn now to explore how members of coalitions interact with one another as coalition members and, thereby, create the potential for group-on-group conflict.
For animals that live in groups, there are plenty of opportunities to bang heads with other group members, given limited resources to go around. Hence, social animals use fighting strategies to compete for the group's resources. At the same time, conflict is costly. Fighting risks physical injury and damaging valuable cooperative relationships. Animals thus require strategies that increase access to resources while reducing both the likelihood of injury and damage to cooperative relationships.
There is a large theoretical and empirical literature in evolutionary biology about animal fighting (Arnott & Elwood, 2009; Dawkins, 1976; Maynard Smith, 1982; Parker, 1974). Much of this work centers on the Hawk-Dove game, in which two players choose to either fight (hawk) or flee (dove). Each player receives their highest payoff if they fight and their opponent flees, but the most costly outcome occurs if both fight. This research points to a few broad conclusions. First, evolution favors a judicious mixture of fighting and fleeing rather than all-out aggression or all-out acquiescence. Second, when there are asymmetries in fighting ability, more formidable disputants will fight and weaker disputants will flee (all else equal). Third, animals also use other asymmetries, independent of fighting ability, to decide conflicts, such as which player first discovered or possessed the resource.
Fighting is a coordination game (Schelling, 1960). Although fighters disagree about who should acquire the resource, they also typically share an interest in avoiding the costs of fighting. This implies that fighters will attend to information or signals that might help coordinate their fighting decisions to prevent deadlock and escalation. This includes cues of relative formidability, precedents set by previous fights, and communicative displays of submission and dominance.
In many social animals, the result of these individual strategies is the creation of linear dominance hierarchies (Boehm, 1999; Krebs & Davies, 1993). Individuals learn which group members they can defeat and attribute to them lower status, and which members they cannot defeat and attribute to them higher status. Higher-status individuals can then win disputes merely by displaying dominance, whereas lower-status individuals can avoid the escalation of conflict and further loss of resources by displaying submissiveness. These communicative strategies persist because they reduce conflict costs for both senders and recipients. Low-status individuals are considerably disadvantaged by this coordination scheme, as they are forced to forgo many of the potential benefits of living in the group. We should expect evolution to favor adaptations designed to help individuals avoid this predicament. A small number of social animals, including humans, have evolved a novel adaptation to the problem of being dominated by more powerful individuals—forming alliances.
The original adaptive problem that coalitions are designed to solve is being dominated by a more powerful individual. By teaming up, a few weaker individuals can gain the upper hand against a single more powerful one. The same logic of asymmetric fights applies to the combined power of the coalition, in which sole individuals stand to gain by backing down when they are outmatched by a team.
When one coalition is formed within a group, this creates a new adaptive problem—being dominated by a powerful coalition. The solution to this quandary is to form a coalition in response. In this way, coalitions beget more coalitions until all members are split into teams (Snyder, 1984). Further, small coalitions can increase their power by merging with other small coalitions, which occurs until the group consists of a small number of massive nested coalitions. Individuals can seek cross-cutting alliances with members of other coalitions to bolster their power within their original coalition, creating a complex interlaced network of alliances. Due to this complexity, coalitions are better conceptualized not as fixed and cohesive groups, but instead as arising from interlaced networks of ranked loyalties (DeScioli & Kurzban, 2009, 2011, 2013; DeScioli, Kurzban, Koch, & Liben-Nowell, 2011).
Once coalitions are formed, they have the same problem as individual fighters—avoiding costly fights—and they can apply similar tactics including fighting assessment, fighting displays, dominance and submission signals, and the use of arbitrary asymmetries or conventions. The result is a group-based dominance hierarchy, analogous to individual dominance hierarchies. The existence of group-based dominance hierarchies sets the stage for certain forms of intergroup prejudices and discrimination.
Social dominance theory holds that the evolution of group-based dominance hierarchies explains coalition-based forms of prejudice, discrimination, and oppression (Pratto, Sidanius, & Levin, 2006; Sidanius & Pratto, 1999). From this perspective, coalitions broadcast their memberships, group boundaries, and power advantages over other groups through individual acts of prejudice, institutional discrimination, and legitimizing myths (Pratto et al., 2006). As in individual disputes, both higher-status and lower-status groups gain by using dominance and submission signals because these signals help individuals avoid the costs of violent confrontations (Sidanius & Pratto, 1993). Due to this dominance scheme, individuals in lower-status groups receive a smaller share of the available resources. Moreover, oppression by higher-status groups can become extreme, such as the historically widespread practice of slavery.
From this perspective, then, certain prejudices begin with evolved cognitive adaptations for coalition formation. In supporting an ally, one is exhibiting prejudice and discrimination in favor of one's ally against their opponent. Even supporting family or friends in an argument against strangers is prejudiced, in this sense.
Social dominance theory focuses on three types of group-based hierarchy based on age, sex, and arbitrary sets (Pratto et al., 2006). In the age system, adults as coalitions have disproportionate power compared to children. In the gender system, men allied in coalition have disproportionate power relative to women. In the arbitrary-set system, people construct rival coalitions on arbitrary distinctions such as race, nationality, political ideologies, or religion.
Social dominance theorists argue that, historically, the most damaging prejudices and acts of discrimination have been those used by these arbitrary groups of men to dominate other men (Pratto et al., 2006). They further argue that this gender difference is explained by parental investment theory, which implies high stakes for human male–male competition leading males to fight more intensely than females (McDonald, Navarrete, & Sidanius, 2011; McDonald, Navarrete, & van Vugt, 2012; Pratto et al., 2006). Particularly important in this account are the legitimizing myths—which include unflattering stereotypes about lower-status groups—used to stabilize otherwise arbitrary coalitional alignments to enable dominant groups to oppress weaker groups.
The idea of prejudice based on arbitrary sets fits well with game theoretic models showing indeterminacy and instability in coalition formation (Von Neumann & Morgenstern, 1944). In these models, individuals aim to form teams to gain advantages over other teams, but these motives can lead to many possible partitions of the group. Even after coalitions are formed, individuals often have incentives to switch to a new coalition and these switches can, in some cases, occur indefinitely. Humans appear to be attuned to this game of theoretic logic with psychological mechanisms for tracking coalition membership by adaptively and flexibly alternating among a variety of cues including race, accent, and even tags as arbitrary as shirt color (Kinzler, Shutts, DeJesus, & Spelke, 2009; Kurzban, Tooby, & Cosmides, 2001; Pietraszewski, Cosmides, & Tooby, 2014).
Although coalitions tend to be unstable and shifting, certain special interests among individuals make it more likely that specific kinds of coalitions will form and be relatively stable. Perhaps the most biologically important source of shared interests is family relationships. Kin deeply share fitness interests (Hamilton, 1964), and kin selection favors adaptations for helping kin in conflicts with nonkin. These psychological mechanisms can be viewed as an evolved prejudice—nepotism—for supporting kin against nonkin, and are observed both in humans and nonhuman animals (e.g., Cheney & Seyfarth, 2007; Silk, 2002).
Racial prejudices might be extensions of an evolved kin-based coalitional prejudice. People might perceive racial and ethnic differences as cues indicating low genetic relatedness. Because our ancestors were unlikely to encounter “racially” different individuals within the range of their life experiences, there was little opportunity for race-focused prejudice, per se, to be selected for (Kurzban et al., 2001). There likely did often exist, however, observable differences between competing coalitions—cued by different physical appearances, language or accents, and cultural artifacts and practices—that would enable individuals to evolve a coalitional psychology sensitive to features implying difference. The features denoting “race” may thus serve as super-cues of difference, and be used heuristically by a kin-based coalitional psychology to generate so-called racial prejudices and acts of discrimination).
Age and sex are also potential sources of special interest groups. Life history strategies (Kaplan & Gangestad, 2005) and parent-offspring conflict (Trivers, 1974) can give rise to systematically different evolved preferences for individuals of different ages. For example, children often try to extract more resources than parents want to provide. Parents can potentially work together to suppress their children's extraction efforts, exhibiting a self-serving ageism. For example, the consensus among adults that children should obey their parents can be understood as a strategy for limiting children's demands for resources. For sex, parental investment theory and sexual selection (Trivers, 1972) imply that men and women will have a variety of different preferences. Each sex can potentially ally and conspire to advance its own interests at the expense of the opposite sex, showing self-serving sexism. For example, men and women might disagree about sexual activity outside of long-term relationships, such as prostitution and pornography, due to differences in mating strategies.
Simple ingroup–outgroup coalitional views of race, sex, and age prejudices are likely insufficient, however. Research reveals that coalition-based prejudices are most frequently directed by and toward young men (relative to other sex/age categories). Moreover, the particular stereotypes ascribed to intersectional categories—Sex × Age × Race—are closely linked to the specific threats and opportunities associated with them (Neuberg & Sng, 2013; Sng, Williams, & Neuberg, 2015). An integrative approach that combines both threat-management and coalition dynamics will be especially informative.
Other special interest groups might include those ostensibly based on values—on broad orientations regarding the goals people ought to have and how they ought to behave. Although values themselves appear to be abstract, they often serve as bases for creating or maintaining particular rules, laws, and societal policies that place real, tangible constraints on other people's behaviors—constraints they often wish to avoid. Political parties, for instance, are coalitions that compete over how resources within a society are allocated, the manner in which rule violators should be controlled, and so on. Religious groups, as a second example, are coalitions that compete over similar concerns but also tend to seek to control the sexual strategies pursued by group members (Weeden & Kurzban, 2013). It should not be surprising that activist religious groups often exhibit strong prejudices against one another and engage in extreme forms of conflict (Neuberg et al., 2014).
Foreigners are individuals who have had little or no contact with the focal group. Contact with peoples from different environments can expose individuals to pathogens for which local immune systems are ill-equipped. Indeed, the history of migration shows the virulence of pathogens when entering new populations (Diamond, 1997; Dobson & Carter, 1996; Ewald, 1994). Moreover, foreigners will often be unfamiliar with local hygiene practices, placing residents at risk. Studies show that desires for distance from foreigners and preferences for ingroups are most pronounced in those who feel most vulnerable to disease (Faulkner et al., 2004; Navarrete et al., 2007).
The coalitional reasoning discussed above also applies to prejudices toward foreigners. Whereas ingroup relationships require a balance of cooperation and conflict, there is, by definition, little cooperation with true foreigners who are likely to be seen mainly as competitors. Throughout evolutionary history, interactions with foreigners were conflictual. For most nonhuman primates, intergroup encounters are violent (e.g., Southwick, Siddiqi, Farooqui, & Pal, 1974; Wilson & Wrangham, 2003), and ethnographic studies of human hunter-gatherers similarly show violence between groups (e.g., Chagnon, 1992; Eibl-Eibesfeldt, 1974; Ferguson, 1984; Haas, 1990; Kelly, 1995; Robarchek, 1990). Because males were ancestrally more likely to encounter outgroup individuals (e.g., Goodall, 1986; Hasegawa, 1990), males are expected to exhibit greater group-on-group violence (Carpenter, 1974; Chagnon, 1988; Cheney, 1986; Wilson & Wrangham, 2003). In line with this idea, men perceive intergroup situations as more threatening (Pemberton, Insko, & Schopler, 1996) and hold stronger intergroup prejudices (e.g., Sidanius, Cling, & Pratto, 1991). Moreover, people are especially slow to extinguish learned fearful reactions to outgroup men (Navarrete et al., 2009). In all, there exist strong prejudices against members of outgroup coalitions that are especially held by, and directed towards, men.
This helps explain why immigrants often elicit strong antipathies and sometimes violence from local populations. They are seen as posing multiple threats—to health, resources, physical safety, and values. Importantly, however, not all immigrants are viewed with equal hostility. Those who are subjectively foreign—displaying cues for unfamiliarity—receive greater antipathy (Schaller & Abeysinghe, 2006). In contrast, immigrants who look physically familiar, speak the local language, and act according to local customs are viewed as less threatening. Moreover, in line with the flexibility of threat-management systems, we expect those who arrive during times of relative prosperity will encounter less vulnerable residents and receive less hostility as a result.
This perspective also suggests why immigrant groups elicit less antipathy in subsequent generations. The offspring of immigrants learn the local language, adopt local cultural practices, and live according to local values, and so no longer exhibit these cues for threats. The Irish arriving in the United States in the 1840s were viewed as violent, disease-ridden, resource-grabbing, and allegiant to the Catholic pope, and were stigmatized greatly for it. Today, nearly 200 years later, to be Irish is rarely seen as threatening—and Americans across the ethnic spectrum celebrate St. Patrick's Day.
Of course, foreign groups are often able to get along, at least temporarily. Within societies, alliances form so individuals can more effectively pursue common interests, and this happens at the level of foreign coalitions as well. Moreover, there may be circumstances in which individuals actually become “xenophilic”—in which they prefer outgroup to ingroup members. For instance, in some social species, including humans, females have historically left their home groups to find mates from other groups. The evolutionary logic of female exogamy pertains to incest avoidance, and may help explain why female strangers are stigmatized less than male strangers, and why females are more open to foreigners than are males. Favorable ties might also develop between foreign groups based on trade for rare goods, access to territory, and other mutually beneficial opportunities. That said, the basic evolved inclination is for people to be quite wary of foreigners in their own midst.
To suggest that contemporary prejudices are rooted in an evolved psychology is not to suggest that they are unchangeable. To the contrary. As we've seen, threat-management systems operate in functionally flexible ways, and certain prejudices emerge under some specifiable circumstances and do not emerge under others. Many people in many cultures strongly condemn and oppose prejudice and this anti-prejudice behavior likely also has a basis in human evolved psychology. Evolutionary approaches can inform efforts to reduce harmful prejudices.
The main principle emerging from a threat-management approach is that by reducing vulnerability to particular threats, one can reduce the related prejudices. For example, by changing the geographical frame of reference used by Sinhalese Sri Lankans to estimate their numbers—shifting the perception of being outnumbered by the Tamil to outnumbering them instead—Schaller and Abeysinghe (2006) reduced their prejudices (at least temporarily) and made them more favorable toward peaceful resolutions. Similarly, by providing disease-concerned individuals with hand wipes or flu shots, Huang and colleagues reduced (at least temporarily) their prejudices against immigrants, obese people, and people with physical disabilities (Huang, Sedlovskaya, Ackerman, & Bargh, 2011).
A second critical principle is that different interventions will be required to combat different prejudices. In Huang et al. (2011), reported above, infection-reducing interventions did not reduce all prejudices, but rather prejudices related to contagious disease, and only for individuals who felt most vulnerable to infectious disease. The threat-based approach helps explain why certain interventions succeed and others do not (see Schaller & Neuberg, 2012, for a more comprehensive discussion).
A third idea motivated by an evolutionary approach is that people likely possess adaptations designed to counter the prejudices they confront. That is, just as the capacity for prejudice is a human universal, so is the psychological ability to oppose prejudice. Humans not only oppress and enslave members of rival groups, but some members of dominant groups work to empower and liberate lower-status groups. For example, with the Slavery Abolition Act of 1833, members of a dominant group emancipated oppressed groups throughout the British Empire. A psychological theory of prejudice should account for both the motive to discriminate as well as the motive to oppose discrimination.
One possible explanation derives from the fundamental instability of coalitions. To strengthen their position inside a coalition, individuals can create cross-cutting alliances with individuals outside of their coalition. Coalition members are not immune to infighting and cross-cutting alliances can provide an advantage for disputes within the coalition. Humans might have cognitive adaptations for identifying potential cross-cutting allies, and this could help to explain people's efforts to promote the welfare of individuals in oppressed groups.
Another possibility is that humans have anti-prejudice adaptations designed to diffuse escalating alliance-building. When each individual pursues a prejudiced side-taking strategy, this expands the number of coalitions and subsequent alliance-based obligations, which can lead to expanding and explosive disputes. If, instead, bystanders to others' disputes coordinate on an impartial side-taking strategy, these disputes can be contained. Indeed, moral cognition appears well designed to perform exactly this function (DeScioli & Kurzban, 2013). Moral cognition computes the wrongness of people's actions, providing a basis for side-taking that is independent of their identities, including coalition membership. Moral side-taking involves computing wrongness magnitudes for the actions taken by both sides of a dispute, and siding against the individual who performed the action with the greatest wrongness magnitude. Importantly, moral side-taking strategies do not displace prejudiced side-taking, but rather add to the repertoire of human strategies for choosing sides. An individual's choice of strategy will depend on computations of the costs and benefits of each approach. When these values differ across individuals, they will pursue different strategies, potentially explaining individual and cultural variation in prejudice and anti-prejudice behavior.
The harms of malicious prejudices have plagued human societies throughout history and continue to do so today. From an evolutionary perspective, several prominent approaches to understanding prejudice have key limitations. These accounts posit particular psychological needs or tendencies underlying prejudice, such as group categorization, social identities, self-esteem, authoritarian values, the fear of death, or justification of group standing (e.g., Adorno, Frenkel-Brunswik, Levinson, & Sanford, 1950; Fein & Spencer, 1997; Jost & Banaji, 1994; Schimel et al., 1999; Tajfel, 1969; Tajfel & Turner, 1986). These accounts inevitably lead to questions about why humans have these psychological traits and why they cause these particular behaviors. Moreover, the evidence reviewed in this chapter shows levels of flexibility and complexity in human prejudices, stereotypes, and forms of discrimination that go far beyond what can be predicted by traditional theories.
The evolutionary approach suggests, instead, that the psychological systems underlying prejudices are highly sophisticated computational systems designed to track ancestrally relevant threats (e.g., violence, disease) and opportunities (e.g., cooperation, alliances), and to deploy prejudice—and anti-prejudice—behaviors to manage these threats and opportunities and thereby improve individual fitness. This approach generates nuanced hypotheses, supported by empirical research, well outside the reach of traditional theories.
Critically, the hypothesis that prejudices are designed to benefit the individual does not, in any way, diminish the harms that prejudices cause victims and societies. It is crucial to emphasize, especially for lay audiences, the importance of avoiding the naturalistic fallacy that evolved traits are benign, to be encouraged, or somehow less condemnable. Evolution's products include any number of unseemly and cost-inflicting adaptations, from the viper's poison-injecting fangs to the moose's impaling antlers to human jealousy and homicidal motives (Buss, 2006). The possibility that prejudices are evolved adaptations, if anything, should heighten our concern about these damaging behaviors because they reflect not merely naïve, unschooled biases but instead the operation of evolved, flexible, selfish, and largely unconscious strategies. To outwit such an imposing foe requires theories that are prepared for evolution's most clever and Machiavellian designs.