Religion and reducing prejudice

Abstract

Drawing on findings from the study of prejudice and prejudice reduction, we identify a number of mechanisms through which religious communities may influence the intergroup attitudes of their members. We hypothesize that religious participation could in principle either reduce or promote prejudice with respect to any given target group. A religious community’s influence on intergroup attitudes will depend upon the specific beliefs, attitudes, and practices found within the community, as well as on interactions between the religious community and the larger social environment in which it is embedded. Basing our proposals on findings from the literature on prejudice formation and prejudice reduction allows us to outline useful directions for future studies of religion and prejudice.

In this paper, we propose a set of simple hypotheses about ways in which religions might tend to promote or reduce prejudice. Our method is to build on established findings from the study of prejudice reduction generally, and apply these findings to the study of religion and prejudice. Drawing on the general study of prejudice reduction allows us to offer a number of well-supported hypotheses about the diverse roles that religious communities may play in shaping intergroup attitudes.

To take one example, an important finding from prejudice research is that intergroup contact tends to reduce prejudice. In light of this finding, we predict that religious communities will tend to reduce prejudice between groups if they encourage social contact between them, particularly under certain facilitating conditions (e.g., equality of status between the groups, successful cooperation, affirmation of positive distinct identities, etc.). Conversely, we predict that religious communities will tend to increase prejudice to the extent that they discourage contact between social groups, or encourage contact under problematic conditions (e.g., inequality of status between groups, unsuccessful cooperation, failure to affirm positive and unique subgroup identities, etc.). In practice, religious communities vary widely in their messages regarding intergroup contact. Religious communities may discourage contact between some social groups, while facilitating and encouraging contact between other social groups. The contact hypothesis may therefore help to partially explain variation in religious influences on intergroup attitudes.

In this paper we identify several such hypotheses, drawing on findings from the study of prejudice reduction. Specifically, we predict that religious communities will tend to reduce prejudice to the extent that they do the following: (a) Explain important differences in the social world by pointing to differences in circumstances, instead of by reference to innate or essential group characteristics; (b) Promote inclusive and pluralistic theologies; (c) Oppose prejudice-supporting ideologies; (d) Model ways of categorizing the social world which have been shown to reduce prejudice, such as emphasizing positive common ingroup identities, while also affirming distinct subgroup identities; (e) Encourage intergroup contact, cooperation, and friendship. These are just a few of the many possible hypotheses that might be identified by drawing on prejudice-reduction research. We hope that these and related hypotheses will be taken up by future researchers, and will be used to develop valuable insights into the diverse ways in which religious communities can shape intergroup attitudes.

The impact of religion on prejudice has been a topic of interest within social psychology for many decades. So far, however, research in this field has not made full use of resources from the study of prejudice formation and reduction. Linking these bodies of research more systematically could lead to significant advances in understanding. One advantage of the approach we propose is that it supports a fine-grained study of the possible relationships between religion and prejudice. Historically, social psychological research in this area has tended to treat religions generically (for instance with respondents being identified as “Christian” or “Jewish”), with broad conclusions being drawn about the influence of religiosity on intergroup attitudes. However, the range of beliefs, attitudes, and behaviours found across religious communities is profoundly varied (Esposito, Fasching, & Lewis, 2014; Juergensmeyer, 2006; McKim, 2012, pp. 5–7; Meister, 2010; see also Burch Brown, 2013), and it is reasonable to expect this inter- and intrareligious diversity to result in significant variation in intergroup attitudes. Our approach provides a more fine-grained route into exploring the diversity of influences likely to result from the practices of different groups.

This paper necessarily works within some chosen limits. First, because the literature on prejudice is so rich and well developed, the possible directions for this project are many. In the space available, we discuss several important hypotheses, but many others could have been explored instead. We hope that readers will not be limited by our omissions, and will find further hypotheses based on their own expertise. Second, the paper adopts a theoretical focus rather than an empirical one. We do not attempt to apply the hypotheses to practical cases. For researchers interested in operationalizing these hypotheses in religious contexts, a useful resource may be Stausberg and Engler (2013). Moreover, this paper does not attempt to review the contemporary literature on religion and prejudice, which would be a substantial undertaking on its own. (A recent overview can be found in Rowatt, Carpenter, & Haggard, 2013.) Instead, in the space available, we have focused on developing our positive proposal. Finally, it is worth noting in advance that although our ultimate aim in this paper is to identify ways in which religious communities might reduce prejudice, we also give considerable attention to ways in which religions might contribute to prejudice formation. One reason for this is that findings to date have tended to show a positive association between religious participation and prejudice against certain groups (Batson, Schoenrade, & Ventis, 1993; Hall, Matz, & Wood, 2010; Johnson, Rowatt, & LaBouff, 2012; Whitley, 2009). Moreover, one way for religious communities to reduce prejudice is by challenging those social practices that tend to cause it, including their own. Thus we have given some extended attention to the topic of prejudice formation, although the overall aim of the paper is to identify avenues towards prejudice reduction.

Our hope is that the relatively straightforward model offered here will attract the attention of fellow researchers and will stimulate new directions for the field. There are at least three reasons to think that researchers interested in prejudice may benefit from exploring the influences of religion. The first reason is that religion may sometimes play an important role in establishing social hierarchies and social divisions (Fredrickson, 2009; Harvey, 2011). This makes religion a potentially important factor for understanding prejudice and prejudice reduction. The second reason is the converse of the first. It is that religion may sometimes play an important role in challenging social hierarchies and social divisions (Harris, 1999; Harvey, 2011; Morris, 1984). Both of these ideas are discussed in some depth in the section entitled “Reducing Prejudice”.

A third reason has to do with the contemporary global context. Nearly a third of the world’s population (2.18 billion people) self-identify as Christian, and a quarter (1.57 billion people) as Muslim, with estimates suggesting that the world’s Muslim population will rise rapidly to 30% by 2050 (Pew Research Center, 2015a). Many other people identify as Hindu (15%), Buddhist (7.1%), adherents of folk religions (5.9%), and Jewish (0.2%). Only 16% of people are religiously unaffiliated (including atheists and agnostics), the majority of whom live in China (Pew Research Center, 2015a). Thus religion plays an important role in the lives of communities across the world, and many intergroup conflicts involve religious identities in some way (Appleby, 2000; Pew Research Centre, 2015b). Researchers who wish to make a contribution to understanding these conflicts cannot afford to ignore religion. Having a sophisticated understanding of diverse religious influences may be crucial for finding paths towards reducing conflict and promoting better intergroup relations in many communities across the world (Appleby, 2000; Johnston & Sampson, 1995; Smock, 2006).

We begin with a brief overview of past evidence concerning relationships between religion and prejudice. We then draw on Social Identity Theory to help explain why religion could sometimes lead to the formation of prejudice. Finally, we consider five ways in which religious communities might reduce prejudice. Our hope is that these hypotheses can be used to identify fruitful directions for future research.

Evidence of Relationships Between Religion and Prejudice

Research into relationships between religion and prejudice began with the study of race relations in America. Over a number of decades, this research has shown that religious identification and participation can at least sometimes be associated with increased prejudice towards a range of outgroups (Altemeyer, 2009; Batson et al., 1993; Hall et al., 2010; Johnson et al., 2012; Rowatt, LaBouff, Johnson, Froese, & Tsang, 2009; Whitley, 2009). For instance, a recent meta-analysis examined the results of 55 studies carried out in the US since the Civil Rights Act of 1964 and found significant—though declining—correlations between levels of religious participation and identification, and levels of overt and covert racial prejudice (Hall et al., 2010). Religious identification in the US during this period has also been positively correlated with prejudice against a range of other minority target groups, such as gay and nonreligious people (Hood, Hill, & Spilka, 2009; Hunsberger & Jackson, 2005; Jackson & Hunsberger, 1999; Whitley, 2009). During the same period, by contrast, atheism and low levels of self-reported religiosity have been associated with lower levels of racism and prejudice towards the same target groups in the US (Hall et al., 2010).

Recent research has also found that activating religious concepts can increase prejudicial attitudes. For instance, subliminally priming U.S. undergraduates with Christian words increased covert racial prejudice and negative affect towards African Americans (Johnson, Rowatt, & LaBouff, 2010). Religious priming amongst U.S. undergraduates can also lead to increased derogation towards value-violating outgroups, and increased in-group favouritism (Johnson et al., 2012). Outcomes may be sensitive to specific primes used. Preston, Ritter, and Hernandez (2010) found that priming for “Christianity” increased helping behaviours towards ingroup but not outgroup members, whereas priming for “God” was associated with universal helping behaviour. Priming (either subliminally with religious terms, or supraliminally with images and contextual cues) activates neural networks, which in turn influence cognitive and affective processing; thus priming can reveal on-line causal influence of religious concepts. These studies show that activating religious concepts can in some cases increase prejudice.

Researchers have sometimes used such findings to draw broad conclusions about associations between religiosity and prejudice. For instance, on the question of whether religious persons are less tolerant than nonreligious persons, Batson writes “Based on existing research, the answer is very clear: in spite of what religions preach about universal brotherhood, the more religious an individual is, the more intolerant he or she is likely to be” (Batson, 2013, p. 89). Similarly, Altemeyer writes that “The more one goes to church, the more likely one will be prejudiced against a variety of others” (Altemeyer, 2009, p. 18).

There are reasons to be cautious, however, in drawing general conclusions from this research (Gries, Su, & Schak, 2012; Olson, 2014). For one thing, most social psychological studies of religion and prejudice have focused on Christians in North America and Europe since the 1960s; and most have examined either racism from Whites towards Blacks (for review see Hall et al., 2010), or attitudes towards groups perceived as challenging traditional Christian values, such as gays and lesbians, feminists, and atheists; or members of non-Christian religions such as Judaism or Islam (Duckitt & Sibley, 2007; Hunsberger & Jackson, 2005; Jackson & Hunsberger, 1999; Whitley, 2009). Traditional approaches have treated religiosity as a unitary construct, and explained variation by reference to individual personality variables (e.g., Altemeyer & Hunsberger, 1992), rather than mapping variation between religious communities. As more recent research suggests (e.g., Saroglou & Cohen, 2011; Tsang & McCullough, 2003) this traditional approach is not designed to reflect the variability of religious communities, and may lead researchers to overgeneralize on the basis of distinctive religious and cultural contexts.

The few studies focusing on East Asian religions and prejudice suggest a different pattern of intergroup attitudes, with religiosity at least sometimes being associated with greater intergroup tolerance (Clobert, Saroglou, Hwang, & Soong, 2014). East Asian religiosity amongst individuals in Japan, South Korea, and Taiwan has been found to predict reduced levels of prejudice towards numerous target groups, compared with nonreligious individuals, including reduced ethnic prejudice (e.g., against Africans), reduced antigay and antiatheist prejudice, and reduced prejudice towards members of other religions (e.g., interreligious prejudice against Muslims; Clobert et al., 2014). Relatedly, supraliminally priming Christians in Belgium with images of Buddhist monks in meditation was found to decrease implicit prejudice towards ethnic outgroups, suggesting both that interreligious priming can influence intergroup attitudes, and that subjects associated Buddhism with values of openness and tolerance (Clobert & Saroglou, 2013). Preliminary as these findings may be, they suggest that research thus far may not be representative of the range of possible influences of religion, and they support the view that some forms of religious belief and practice may reduce intergroup bias.

Relatedly, when dimensions such as right-wing authoritarianism (RWA), fundamentalism, and conservative ideologies are controlled for, religion ceases to be predictive of racial prejudice (e.g., Newheiser et al., 2013; Rowatt, LaBouff, Johnson, Froese, & Tsang, 2009). A representative random survey by Rowatt et al. (2009) found that “general religiosity” was only negligibly associated with racial prejudice, although it was positively correlated with prejudice towards gays and lesbians. Rowatt et al. (2009) propose that these findings offer support for the selective intolerance hypothesis, according to which religion is associated with prejudice towards groups perceived as behaving inconsistently with religious values. On this hypothesis, we should expect for prejudicial attitudes to reflect the specific value and belief systems of a religious community, and also to change over time as a reflection of changes in values and perceptions of outgroups. The selective intolerance hypothesis is consonant with the approach that we take in what follows, and suggests that intergroup attitudes are likely to vary across religious communities and over time, following changes in perceived value-consistency of outgroups.

Evidence also suggests that associations between religion and prejudice in the US may be changing, alongside changes in broader social norms. For example, associations between “extrinsic religiosity” and racial prejudice appear to be declining (Hall et al., 2010). Whereas “intrinsic religiosity” involves valuing one’s religious faith for its own sake, “extrinsic religiosity” involves valuing religion for benefits that are not strictly spiritual, such as comfort and security in times of difficulty, and a respected place in a social community (Allport, 1950, 1954, 1966; Batson, 2013; Kirkpatrick & Hood, 1990). Extrinsic religiosity has historically been associated with higher levels of both overt and covert prejudice (Batson, 2013). This association appears to be mediated by social-cognitive motivations for social acceptance and conformity (Hall et al., 2010). Motivations for social acceptance may lead individuals to agree with prevalent social attitudes (such as racial attitudes), and also to accept received religious practices and beliefs. Intrinsic religiosity, on the other hand, has been negatively associated with overt prejudice, but neutral with respect to covert prejudice. This effect is commonly said to be mediated by social desirability (Batson, Naifeh, & Pate, 1978; Kirkpatrick & Hood, 1990; Sedikides & Gebauer, 2010). However, it could also reflect variables not well-distinguished within the Social Desirability Scale, such as an individual’s motivation to meet moral standards they personally value. Whatever the underlying mechanisms, evidence shows that these associations can shift. For instance, the correlation between racism and extrinsic religiosity declined from the 1960s to 2008 (Hall et al., 2010). In a society where racism has become unacceptable while religion remains popular, motivations for social acceptance may lead people to become both religious and overtly racially tolerant. This finding demonstrates that relationships between religion and prejudice may change with social context (for detailed review of extrinsic, intrinsic and “quest” dimensions of religiosity, see Batson, 2013).

For similar reasons, caution should be exercised in generalizing from priming experiments. An advantage of priming studies is that they may reveal on-line influences of religious concepts on group attitudes. Nevertheless, the inferences they warrant are limited, because results inevitably reflect social and cultural experiences of subjects. For instance, words like “Bible,” “Jesus,” and “prayer” are conceptually associated with conservative social values for many people in the US, reflecting current cultural patterns. If a correlation exists between religion and prejudice in a given cultural context, then regardless of the underlying causes, familiarity with these social patterns is likely to be reflected in the semantic networks that are activated. In other words, results will reflect cultural and social learning, and not simply a general effect of “religiosity.” Even if there is a general effect of religiosity, it is not clear that these methods can isolate it or rule out alternative, culture- and socialization-based hypotheses. Priming experiments may reveal religion’s causal influences in particular cultural contexts. However, without extensive cross-cultural comparisons, they cannot warrant general conclusions about the effects of religiosity as such.

Finally, one of the strongest reasons for taking a broader view has to do with the importance of interpreting immediate group attitudes within the context of larger processes of social change (Guimond, Sablonniére, & Nugier, 2014; Tajfel & Turner, 1979; Wright, 2009). Historically, minority religious communities have played central roles in progressive social movements opposing racism, segregation, and prejudice. In the U.S. Civil Rights movement, religious leaders like Dr. Martin Luther King, Jr. and church networks like the Southern Christian Leadership Conference (SCLC) played a major part in mobilizing Black and White communities against racism (Dorrien, 2011; Harris, 1999; Harvey, 2005, 2011; Morris, 1984). Other examples include the role of the Quakers and religious leaders like John Woolman and William Wilberforce in the opposition to slavery (Dorrien, 2011); the Gandhian movements in South Asia, which drew ecumenically on Hinduism, Buddhism, and Christianity in constructing practical philosophies of nonviolent resistance to social oppression (Atack, 2012); the egalitarian Islamic pacifist movement of Pakistani leader Abdul Ghaffar Khan (Appleby, 2000); the Catholic Workers movement and South American liberation theology (Appleby, 2000; Dorrien, 2011); and the cooperation between progressive Black and White churches in resistance to South African apartheid (Appleby, 2000). Many social movements directed towards overcoming group-based injustice and discrimination have been organized through religious communities—often starting with the minority group’s own religious communities and leaders—and have drawn on religious ideas and values in making the case for resistance to social oppression (Dorrien, 2011; Harris, 1999; Harvey, 2011; Morris, 1984).

The fact that such communities have often been opposed by other religious groups advocating social hierarchy and segregation underlines the argument of this paper—that there is a high degree of internal variation within religious faiths and between communities, and that the cumulative social messages of different religious communities will often lead to significantly different group attitudes. Strongly egalitarian theologies and religious practices may historically be the exception rather than the rule. Nevertheless, these minority voices have sometimes had profound influences on the larger culture. Such considerations give support to recent methodological shifts towards treating religion as a multidimensional construct (Atran & Norenzayan, 2004; Saroglou, 2011; Saroglou & Cohen, 2011; Tsang & McCullough, 2003).

Social Identity, Religion, and Prejudice

In this section we draw on a social identity perspective and the concepts of realistic and symbolic threat to explain why religion might sometimes contribute to the formation of prejudice. In the subsequent sections, we will turn towards hypotheses concerning possible ways that religions may reduce prejudice.

Religious communities often profoundly shape the ways in which their members conceptualize the social world (Hood, Hill, & Spilka, 2009; Paloutzian & Park, 2013; Saroglou, 2013; Ysseldyk, Matheson, & Anisman, 2010) and religion can play an important role in shaping the formation of individuals’ social identities (Verkuyten & Yildiz, 2007; Ysseldyk et al., 2010). According to social identity theory, an individual’s self-concept is shaped in part through subjective identification with a range of social groups. Tajfel defines a person’s social identity as his “knowledge that he belongs to certain social groups together with some emotional and value significance to him of the group membership” (Tajfel, 1972, p. 292). This subjective identification is supported by the individual’s self-categorization (J. C. Turner, Hogg, Oakes, Reicher, & Wetherell, 1987; J. C. Turner, Oakes, Haslam, & McGarty, 1994), through which she conceives of herself as belonging to various social groups (carpenter, Quaker, Christian, academic, family member, African American, and so on) and assigns some emotional significance or meaning to these groups. Human beings have a need for positive group membership (Tajfel & Turner, 1979). Strong and positive identification with social groups is associated with many psychological benefits, including improved coping, self-esteem, social support, and clinical outcomes such as protection against depression (Haslam, Jetten, Postmes, & Haslam, 2009; Hughes, Kiecolt, Keith, & Demo, 2015; Paloutzian & Park, 2013).

Religious identification may be experienced as a particularly valuable form of group membership, because of the meaningfulness of religious experience, and the cognitive and emotional rewards of participating in a community organized around important moral ideals and values like love and commitment (Graham & Haidt, 2010; Hood et al., 2009; Parmagent, 1997; Saroglou, 2011, 2013; Ysseldyk et al., 2010). For many people, religious identification provides valued cognitive and affective benefits, such as a sense of spiritual, moral, and social purpose; a sense of meaning; feelings of connection and belonging; emotional bonding; and cognitively and emotionally rewarding ways of understanding existence and one’s place within it (Graham & Haidt, 2010; Hood, Hill, & Spilka, 2009; Parmagent, 1997; Saroglou, 2011). Religious participation may provide particularly rewarding social bonds, leading to feelings of social identification (Ysseldyk et al., 2010). From a sociofunctional perspective, religion binds people together in moral communities (Graham & Haidt, 2010).

In some cases, rewarding social bonds and social identification might develop in ways that support prejudice reduction. For instance, as noted previously, religious communities played important roles in both the abolitionist and U.S. civil rights movements. Members of these religious communities often reported feelings of community pride, purpose, solidarity, interpersonal bonds, and social identification associated with their religious community’s role in these social movements (Harris, 1999; Harvey, 2005, 2011). A person might form a strong social identification with a religious community on the grounds that the community is committed to ideals of egalitarianism, universal love, pluralism, inclusivity, and so on.

On the other hand, in part because of its importance to social identity and to moral community, religious identification might also sometimes be related to outgroup prejudice, for instance either through realistic or symbolic conflict. In religious identity threat, perceived challenges to a religion or religious community may be experienced as threatening to an individual’s values, worldview, culture, family, moral community, and social group (Ysseldyk et al., 2010; Ysseldyk, Matheson, & Anisman, 2011). For instance, social and religious diversity may be experienced as symbolically threatening if it is perceived as altering or eroding the religion’s values, tradition, and culture, or weakening the religion’s unity (compare with national identity threat; Yogeeswaran & Dasgupta, 2014). Some religious communities place strong emphasis on the importance of right belief (orthodoxy). For members of these religions, interaction with nonbelievers may be perceived as undesirable or threatening, for instance if it is thought that it might lead to weakening of belief (and, in turn, of social and moral bonds; Hood, Hill, & Williamson, 2005). For reasons we explain in what follows, teachings that separate and discourage contact between social groups are particularly likely to increase in-group favouritism and out-group derogation. This suggests that a strong emphasis on right belief, combined with an expectation that belief might be weakened by exposure to different views, may make contact with religious outgroup members especially threatening. Diversity may also be threatening if subgroups feel that their distinct positive identities are being suppressed through pressure towards assimilation (Holoien & Shelton, 2012; Hornsey & Hogg, 2000; Plaut, Thomas, & Goren, 2009) or if there are anxieties about stigmatization (Hyers & Hyers, 2008).

Finally, from an integrated threat theory perspective (Stephan & Renfro, 2002) prejudice may arise if social groups organized through religion compete for material goods of various kinds (such as territorial control of sacred places); and symbolic threat may arise towards religious group members if they are perceived as advocating worldviews, practices, and values that are in tension with those of other members of society. It has also been suggested that interreligious prejudice may be driven by perceived threats to freedom (Cottrell & Neuberg, 2005). Evidence suggests that perceived dogmatic language triggers aversive reactions, including threat and state reactance, which is conceptualized by Brehm (1966) as an aversive motivational state involving resistance to perceived attempts at persuasion (Cottrell & Neuberg, 2005; Quick & Stephenson, 2008). Contact with religious outgroups may trigger state reactance, along with associated aversive appraisals related to distrust and anger at perceived threat (Cottrell & Neuberg, 2005; Gervais, Shariff, & Norenzayan, 2011). This initial discussion illustrates just a few of the possible ways in which associations between religion and prejudice might develop.

Reducing Prejudice

In the rest of this paper, we turn our attention to prejudice reduction. We start from the simple thought that insofar as religious communities promote the kinds of beliefs, attitudes, and behaviours that the empirical literature has shown tend to be associated with reduced prejudice, then participation in these communities will tend to be prejudice-reducing. We then turn to the empirical literature to identify more specific hypotheses. This simple approach is attractive because it should help researchers to develop well-theorized, empirically supported, and fine-grained accounts of the diverse influences of religion on intergroup attitudes.

We have found it useful to structure our more specific hypotheses by drawing on Bar-Tal and Halperin’s integrative model of barriers to peacemaking (Bar-Tal & Halperin, 2011; Halperin & Bar-Tal, 2011; see also Bar-Tal, 2013). On this model, factors through which religious communities can influence the intergroup attitudes of their members fall into a number of mutually influencing categories:

• Generalized worldviews.

• Society-specific social beliefs, including ideological social beliefs and circumstantial beliefs particular to a given group situation.

• Intergroup emotions, such as intergroup bonding and feelings of affinity, or emotions like hate, disgust, and anxiety.

• Intergroup behaviour.

These categories interact with each other over time to influence intergroup attitudes. For example, generalized worldviews, together with society-specific social beliefs, will shape attitudes towards outgroup members. These cognitive factors may in turn influence intergroup emotions, such as feelings of affinity and empathy; and cognitive and affective factors together may influence subsequent behaviour. For instance, they may influence willingness to participate in intergroup contact and dialogue (Halperin & Bar-Tal, 2011; Halperin, Crisp, & Husnu, 2012; Smock, 2006); to create opportunities to listen to outgroup members explain their experiences (Smock, 2006); to apologize for ingroup wrongdoing (Cehajic-Clancy, Effron, Halperin, Liberman, & Ross, 2011; Smock, 2006); or to challenge the ingroup’s collective narrative (Bar-Tal, Sharvit, Halperin, & Zafran, 2012). This integrative model offers a useful framework from which to conceptualize different aspects of a religion’s influence on intergroup attitudes. With this model in view, we now explore five broad hypotheses concerning the ways in which religious communities may reduce prejudice amongst their members.

Generalized Worldviews

Modelling malleable instead of fixed styles of social explanation

Our first hypothesis is that religious communities will tend to reduce prejudice if they model styles of social explanation associated with greater openness and tolerance. These include (a) promoting the belief that “all people can change”; and (b) teaching that important differences between social groups are often due to circumstances, rather than essential natures.

“Natural kinds” social reasoning explains group similarities by reference to a common inner nature or essence of individuals involved (Brewer, Hong, & Li, 2013; Yzerbyt, Judd, & Corneille, 2004). Explaining social differences by reference to natural kind conceptions of race, gender, and other traits tends to reinforce existing social hierarchies, since observed characteristics are taken to be fixed rather than a result of context. Evidence suggests that natural kinds approaches to social explanation tend to be associated with increased levels of prejudice (Brewer et al., 2013; Chiu, Hong, & Dweck, 1997; see also Andreychik & Gill, 2015; Paloutzian & Park, 2013).

Agency-based social explanation, by contrast, tends to support a view of the members of a group as capable of change, and of the qualities of the group as malleable (Chiu et al., 1997). Group similarities are taken to reflect shared or common circumstances, beliefs, values, and experiences. Thus early Black activists in America often sought to demonstrate that the circumstances in which Blacks had lived were responsible for characteristics like illiteracy and lack of skills, which Whites at the time generally explained through appeal to fixed conceptions of racial characteristics (Dorrien, 2011). Black activists sought to gradually change the views of Whites by showing that inequality in attainment was the result of social circumstances instead of fixed traits (Dorrien, 2011).

Research shows that providing subjects with evidence that “all people can change” and that social realities reflect circumstances rather than fixed traits, can significantly reduce prejudice (Dweck, 2012; Halperin & Bar-Tal, 2011; Halperin, Russell, Trzesniewski, Gross, & Dweck, 2011). In a recent study conducted with three socially distinct and politically divided groups of Palestinians and Israeli Jews, for instance, Halperin et al. (2011) found that for all groups, having subjects read articles depicting social groups in general as malleable significantly improved positive attitudes towards the political outgroup, increased willingness to engage in intergroup contact, and increased willingness to problem-solve and consider compromise. One explanation for this effect is that if conflict-supporting behaviour is conceived as arising from fixed traits, then the perceived possibility of positive change (and therefore the anticipated value of intergroup contact and communication) may be significantly reduced (Halperin & Bar-Tal, 2011; Halperin et al., 2011). We hypothesize that religious communities may reduce prejudice by teaching that social, psychological, and behavioural differences are often the result of differences in circumstances, and reflect malleable characteristics rather than fixed inner nature.

Promoting inclusive and pluralistic theologies

Our second hypothesis is that religious communities will tend to reduce prejudice if they develop inclusive and pluralistic theologies. An important body of research into religion and prejudice has focused on distinguishing between flexible (“open” or “questing”) and closed or rigid styles of religiosity (Batson et al., 1993; Batson & Stocks, 2005). Individuals with flexible styles of religious faith report seeing their religion or spirituality as an ongoing exploration, rather than a set of fixed answers; are willing to explore different religious viewpoints, acknowledge doubts, and see their faith as an evolving journey (Batson, 1976). Research has found that individuals who describe their faith as an ongoing exploration of open questions tend to be more comfortable with ambiguity, and also show lower levels of outgroup prejudice. By contrast, individuals who think of their religion as offering settled answers tend to be less comfortable with ambiguity, and to have higher levels of outgroup prejudice (Batson, 1976, 2013). The relationship between fundamentalism and prejudice appears to be mediated by a need for closure and by preferences for consistency (Brandt & Reyna, 2010, 2014; Hill, Cohen, Terrell, & Nagoshi, 2010; Shen, Yelderman, Haggard, & Rowatt, 2013; see also Mavor, Louis, & Laythe, 2011).

This suggests that religious communities will tend to reduce prejudice if they guide members to develop flexible styles of religious faith (Brandt & Reyna, 2010). For instance, Williams (2013) hypothesizes that religious communities can support young people (in the case of his study, British Muslims) in developing integratively complex ways of thinking about social identitities. Of particular significance for religious communities, he suggests, is evidence showing that interpersonal environments play a substantial role in shaping cognitive styles (Suedfeld, Leighton, & Conway, 2006; Tetlock, 1986). Building on Tetlock’s (1986) hypothesis that people will think in integratively complex ways when faced with values which they perceive as similarly important and conflicting, Williams hypothesizes that a community may be able to encourage complex thinking amongst members by providing social opportunities which encourage young people to practice thinking flexibly and exploring multiple perspectives.

Comfort with accepting multiple perspectives may help to explain why forms of East Asian religiosity are associated with reduced prejudice (Clobert et al., 2014). It is common in Buddhism and Hinduism, for instance, to teach that all truths are partial, that all human experience of the world is illusory, that the truth cannot be fully known, that right teachings may be relative to times and places, and that many different and apparently inconsistent realities and perspectives on the world may simultaneously be true (Atack, 2012; Clobert et al., 2014; Gries et al., 2012; Holmes, 2014). Many Asian religions place less emphasis, if any, on the importance of “right belief” (orthodoxy), while placing greater emphasis on the importance of forms of “right practice” (orthopraxy) associated with spiritual enhancement (Heim, 1995; Knitter, 2003). In many Asian contexts it is also common for a single individual to participate in multiple religions (multiple religious participation) and to self-identify as belonging to multiple religions (multiple religious identification; Gries et al., 2012; Holmes, 2014). Clobert et al. (2014) hypothesize that Eastern religiosity’s association with lower levels of prejudice might have to do with greater comfort in exploring multiple perspectives, and reduced need for closure and consistency.

Openness to diverse perspectives and reduced emphasis on closure is also associated with various traditions within Christianity and Islam. This perspective is expressed through branches of mysticism, Quaker faith and practice, and mystic traditions or apophatic theology, amongst others. Mystic traditions characteristically conceive of the divine as beyond ordinary realms of perception. They encourage spiritual practices aimed at increasing awareness of the limits of human perception and comprehension. Spiritual wisdom, according to these traditions, involves appreciation of the divine as ineffable, or beyond finite human perception and knowledge. Ideals of open exploration (in Anselm’s phrase, of “faith seeking understanding”) are core values and tenets in many religious traditions. They are often advocated as central parts of spiritual development and exploration (Knitter, 2003). (In our view, these considerations support treating “questing” or open styles of faith as genuine religious orientations rather than an expression of agnosticism; cf. Hall et al. [2010].) We hypothesize that reduced prejudice will be associated with these more pluralistic and inclusive religious orientations.

Society-Specific Social Beliefs

Opposing prejudice-supporting ideologies

Our next hypothesis focuses on prejudice-supporting ideologies. We hypothesize that religious communities will tend to reduce prejudice if they challenge ideologies that justify strong social hierarchies and divisions between social groups. Prejudice-supporting ideologies can be usefully understood with reference to social dominance theory (Sidanius & Pratto, 1999). According to social dominance theory, human beings have evolved in communities characterized by group-based social hierarchies. Within these hierarchies, dominant groups tend to benefit from disproportionate access to both symbolic and material positive social value (e.g., status, good education, jobs, healthy environments) while subordinate groups suffer from a disproportionate share of negative social value (such as lower status, lower quality education and jobs, and less healthy environments). Sidanius and Pratto (1999) propose that group-based social hierarchies generally develop along age, gender, and “arbitrary-set” characteristics, with arbitrary-set characteristics including traits like religion and race, and varying widely depending on culture. The development of hierarchies takes place in part through a combination of direct teaching and indirect social learning (Sidanius & Pratto, 1999). Social dominance theory posits that prejudice arises partly because of the importance of group affiliation in human societies for determining access to positive social value, and protecting against negative social value. Sidanius and Pratto propose that ideological frameworks can either legitimize and strengthen group-based social hierarchies, or they can delegitimize and weaken these hierarchies. Thus we propose that a central way that religions may shape intergroup attitudes is through either legitimizing or delegitimizing hierarchy-supporting ideologies.

The importance of this hypothesis can be powerfully illustrated through historical examples. In the American south prior to the Civil War, White churches widely promoted a theology according to which racial hierarchies were conceived as part of God’s social order (Fredrickson, 2009; Harvey, 2005; Irons, 2009). Hierarchical ideals were reflected in sermons, Sunday schools, and church organization. For instance, in mixed congregations, Whites led services and sat in the main areas of the church while Blacks were restricted to the balconies, with their worship overseen by Whites (Harvey, 2011). Racist ideologies also played an important part in religious worship. For instance, a common theme of sermons was that the relationship between the races should mirror the relationship between God and humankind; in the same way that the church is the servant of God, and is guided and governed by him, men should guide and govern women, and Whites should guide and govern Blacks (Harvey, 2011; Irons, 2009). In the antebellum south, many White evangelicals believed that teaching Christianity to Black slaves would help them to accept slavery, by reinforcing ideals of social order (Fredrickson, 2009; Harvey, 2011).

This example suggests that religion can sometimes serve as a vehicle for establishing social hierarchies, and these hierarchies may shape social identities in important ways (Fredrickson, 2009; Harvey, 2011; Irons, 2009). Groups may develop identities in which an important feature of their positive self-concept is their superiority to other social groups (Fredrickson, 2009; Irons, 2009). Dominant groups might be motivated to develop theologies that justify and make moral sense of their ongoing advantage, for instance by interpreting social hierarchies as part of the divine order (Harvey, 2011). This illustrates some of the complex ways in which religious identity, practice, and theology may evolve in interaction with social hierarchies, potentially leading to increased prejudice.

On the other hand, religion has also been used to challenge social hierarchies. For instance, Black Christians in the American south identified egalitarian messages within Christianity and used them in the public sphere to challenge racial inequality (Harris, 1999; Harvey, 2005, 2011; Morris, 1984). During the Civil Rights movement, religious communities such as the Southern Christian Leadership Conference (SCLC), led by Reverend Martin Luther King, Jr., played pivotal roles in mobilizing Black and White Americans to oppose racist hierarchies in the south. King’s philosophy of nonviolent resistance drew on egalitarian interpretations of Christian love, along with Gandhian examples, which themselves reflected diverse religious traditions (Washington, 1986). In addition to an egalitarian commitment to universal love, abolitionists and civil rights leaders often employed biblical imagery and ideals to convey moral ideas, and to convince Blacks and Whites of the possibility of social transformation (Washington, 1986). Religious communal support and spirituality have been frequently cited by African Americans as sources of psychological resilience and positive self-concept in the face of racial oppression (Harris, 1999). Organizationally, churches have been an avenue through which communities have trained and developed leaders and mobilized for collective action (Harris, 1999; Morris, 1984)—an often key capacity in relation to larger social aims of establishing equal relationships and reducing prejudice (Guimond et al., 2014; Tajfel & Turner, 1979; Wright, 2009). These examples illustrate ways in which religion has been used within large social movements to resist social hierarchies. They also highlight the potential importance of religion for the study of prejudice more generally, since religion has in some important cases served as a vehicle through which communities have opposed prejudice-supporting ideologies.

Within the existing literature on religion and prejudice, the concept of prejudice-supporting worldviews has most often been explored using three constructs: social dominance orientation (SDO), right wing authoritarism (RWA), and fundamentalism (F) or religious fundamentalism (RF; Altemeyer & Hunsberger, 1992, 2004; see also, Johnson et al., 2011). In some of the literature, these constructs are interpreted as personality variables, but they may be thought of in a number of ways, including as characterizations of prejudice-oriented ideologies (see Guimond, Dambrun, Michinov, & Duarte, 2003). SDO measures the strength of a subject’s desire to be deferred to by others, and the strength of their support for the view that some groups are superior and should be higher in the social hierarchy (Sidanius & Pratto, 1999). RWA measures the extent to which an individual is inclined to defer to expert authorities (authoritarian submissiveness) and to adhere to established social norms (conventionalism; see Mavor, Louis, & Sibley, 2010). F measures the extent to which an individual believes that her organizing worldview is the only true path, and the extent to which her sense of purpose and meaning is organized around this commitment. Researchers have found that individuals scoring highly on any of these three measures also tend to score highly in measures of prejudice towards outgroups (Altemeyer & Hunsberger, 1992). Duckitt and Sibley find that RWA especially predicts prejudice towards groups that are seen as threatening or dangerous, while SDO especially predicts prejudice towards derogated groups, reflecting motivation for group dominance (Duckitt & Sibley, 2007, 2010).

Although SDO is not a religious variable, research into religion and prejudice has often used RWA, SDO, and F together to explain patterns of religion-related prejudice. For instance, one method has been to measure the extent to which these three constructs predict observed patterns of prejudice in a population, and then to measure whether religious identity or affiliation predicts prejudice over and above that predicted by RWA, SDO, and F (e.g., Newheiser et al., 2013).

However, a number of issues arise in relation to this kind of study. One is that religious communities are likely to influence the development of members’ worldviews. If religious communities are sometimes responsible for promoting the attitudes measured by RWA, SDO, and F, then the most important question may not be whether religion is associated with prejudice beyond that predicted by these measures. Instead, the key question may have to do with the role of religious communities in generating prejudice-supporting worldviews, or in challenging them. It could be misleading to say that these constructs moderate or mediate the relationship between religion and prejudice, if in fact the relationship is a very direct one, with some religious communities actively teaching the ideologies in question, and others actively opposing them.

Moreover, an important discussion is currently ongoing in the literature concerning problems in the construct validity of RWA, and the relationship between RWA, fundamentalism, and social dominance orientation (e.g., Mavor et al., 2011). RWA is a composite construct, measuring three attitudinal clusters (aggression, submission, and conventionalism). In treating RWA as a unitary construct, the relationships between subconstructs are problematically obscured, resulting in unreliable results concerning the scale as a whole (Mavor et al., 2011). For instance, emphasizing that most respondents fall towards the middle of the RWA scale, and that subjects may receive similar scores on RWA but differ widely in relative scores for each contributing construct, Mavor et al. (2011) reanalyse findings from Laythe, Finkel, Bringle, and Kirkpatrick (2002) in order to distinguish between roles of the three attitudinal clusters of aggression, submission, and conventionalism. They find that aggression and submission correlate strongly with both antigay and racial prejudice while conventionalism correlates strongly only with antigay prejudice. This suggests that relationships of interest may in some cases be hidden within the scale, rather than being revealed by it.

A related problem uncovered by Mavor et al. (2011) concerns statistical artifacts resulting from suppression effects due to obscured relationships between subconstructs. Suppression effects in some cases undermine empirical findings. For instance, evidence from Hall et al. (2010) suggested that once authoritarianism is controlled for, fundamentalism might be correlated with reduced prejudice. However, this finding turns out to be a statistical artifact, which disappears once relationships between subconstructs are accounted for. Finally, further validity problems arise in using RWA together with F, because the scales share certain items, and some items make direct reference to certain prejudices (e.g., homophobia). This undermines the claim that they are predictive of these prejudices. Indeed it might be generally argued that RWA and SDO are useful for measuring prevalence of distinct styles of prejudiced worldviews, but are not suitably interpreted as explaining these prejudices. Thus significant methodological challenges need to be worked through in using RWA, SDO, and F effectively in the study of religion and prejudice. Nevertheless, it seems clear that one important way in which religion can influence prejudice is through promoting or opposing prejudice-supporting ideologies. It would be valuable for future research to explore underlying factors (such as existing social status) that influence communities to develop egalitarian or inegalitarian theologies and religious practices.

Positive common ingroup identities, multiple categorization

One of the most fundamental ways in which religious communities influence intergroup attitudes is by developing and communicating ideas about social groups, including conceptions of the different groups that make up society, and causal explanations of the qualities and characteristics of these groups. Our next hypothesis is that religious communities will tend to reduce prejudice if they model strategies of social categorization which (a) include outgroup members and ingroup members as equals within a positive common ingroup identity, while (b) keeping salient each subgroup’s positive unique identity (Gaertner & Dovidio, 2014; Hornsey & Hogg, 2000; Plaut et al., 2009). For instance, a religious community may teach that all people, both within the faith and outside of it, are fundamentally equal, share a common humanity, and are united in one moral community through divine universal love, but that each community has something distinctive and valuable to offer (Atack, 2012; Dorrien, 2011; Hashmi, 2003; Smock, 2006; Washington, 1986). Such styles of religious social categorization are likely to reduce prejudice.

Conversely, we hypothesize that a community will tend to increase prejudice if it draws sharp and rigid social boundaries—for instance between those whom God favours and those whom God rejects—and if it discourages members from thinking in terms of positive identities that they share with members of the outgroup. Finally, we hypothesize that a community will fail to decrease prejudice if it solely emphasizes common ingroup identities without positively affirming social groups’ distinct identities (Brewer, 1991; Holoien & Shelton, 2012; Hornsey & Hogg, 2000; Plaut et al., 2009).

The two approaches on which these hypotheses are based are the common ingroup identity model (Gaertner & Dovidio, 2014) and multiple categorization models (Brewer, 2000; Crisp & Hewstone, 2007). The common ingroup identity model posits that intergroup prejudice may be reduced if individuals conceptualize ingroup and outgroup members in terms of a positive, shared identity. However, focusing solely on a common identity may induce identity threat by reducing individuals’ sense of having a valuable, unique identity (Brewer, 1991; Hornsey & Hogg, 2000; see also debates on multiculturalism vs. difference-blind/colour-blind approaches to diversity in e.g., Holoien & Shelton, 2012; Plaut et al., 2009; Yogeeswaran & Dasgupta, 2014). According to the optimal distinctiveness model, the formation of an individual’s social identity is driven by a desire for belonging and assimilation, on the one hand, and a desire for positive distinctness and differentiation on the other (Brewer, 1991). For this reason, it is now thought that the most promising recategorization strategies involve multiple categorization (or dual-representation) models, in which distinct, positive ingroup identities are kept salient alongside positive social categories that are shared with the outgroup (Brown & Hewstone, 2005).

A potential limitation is that if groups see themselves as prototypical or as excelling with respect to the normative standards of a shared category, then making this category salient may reinforce outgroup prejudice (Kunst, Thomsen, & Sam, 2014; Mummendey & Wenzel, 1999). Another important limitation is that communities may not always be knowledgable enough to identify positive shared group identities. For instance, many religious communities may be unsure of how to understand nonreligious ethics and moral identity. In the US, 53% of people agree that belief in God is necessary for morality (Pew Research Center, 2014). In most countries this figure is much higher, reaching over 90% of agreement in Ghana, Nigeria, Indonesia, Pakistan, the Philippines, Jordan, Egypt, and El Salvador (Pew Research Center, 2014). Religious people may doubt the reliability of nonreligious moral systems that are based on motivations like principled commitments to justice or compassion, rather than fear of God. This may mean that communication and mutual education is required between groups before shared positive identities are discovered and affirmed. However, both religious and nonreligious ethical systems often advocate similar ideals, such as values of intergroup harmony and social justice, love, respect, compassion, and broad messages of social inclusion. Affirming these and other core values (Schwartz & Huismans, 1995) may provide a basis of shared moral community for religious and nonreligious individuals. For instance, theists and nontheists might conceive of themselves as united by shared commitments to social justice and universal love.

Hypotheses based on social categorization theory are important for the study of religion and prejudice, because religious communities often devote substantial portions of their sacred texts, sermons, and social discourses to explaining appropriate relationships between different social groups, or between the religious group and the broader society. How a religious community conceptualizes the social world will have a substantial effect on the intergroup attitudes of its members.

Behaviour

Contact, cooperation, and intergroup friendship

Our previous hypotheses focused on generalized worldviews and society-specific social beliefs. Our final hypothesis focuses on behaviour. We hypothesize that religious communities will tend to reduce prejudice if they encourage contact, friendship, and cooperation across social groups. Conversely, we hypothesize that religious communities will tend to increase prejudice if they discourage contact, friendship, and cooperation across social groups.

Intergroup contact is one of the most widely tested and strongly confirmed methods of prejudice reduction. We aim here only to give an overview of key findings from this extensive literature. Evidence robustly demonstrates that intergroup contact is associated with reduced prejudice. An important meta-analysis reviewing studies from 38 countries, and involving a quarter of a million subjects, found strong correlations between intergroup contact and reduced prejudice (Pettigrew & Tropp, 2006, 2011). Mere contact with outgroup members was associated with reduced prejudice under a wide range of conditions (effect size r = −.20; Pettigrew & Tropp, 2006). However, intergroup contact most effectively reduces bias if a number of facilitating conditions are met. For instance, evidence shows that contact is most effective if participants

• have equal status;

• have approval from authority figures;

• are cooperating towards shared goals;

• are successful in achieving their goals;

• are not in competition with one another;

• have equal but differentiated tasks;

• are given positive affirmation of distinct group identities; and

• if members of the target group convey that they are typical of their social group, rather than exceptional.

Support for these conditions can be found in: Aronson and Patnoe (1997); Bettencourt, Brewer, Rogers-Croak, and Miller (1992); Bettencourt et al. (1999); Brewer and Kramer (1985); Brewer and Miller (1984); Brown and Adams (1986); Cohen and Lotan (1995); Dovidio et al. (1997); Gaertner and Dovidio (2014); Sherif (1966).

One form of contact strongly associated with reduced prejudice is cross-group friendship (R. N. Turner, Hewstone, Voci, Paolini, & Christ, 2007). Effects are particularly moderated by the amount of time spent with outgroup friends, and self-disclosure to friends (Davies, Tropp, Aron, Pettigrew, & Wright, 2011). Cross-group friendships are also associated with reduced prejudice across extended friendship networks; individuals whose friends have outgroup member friends tend to have reduced prejudice (Paolini, Hewstone, Cairns & Voci, 2004; Swart, Hewstone, Christ, & Voci, 2010; R. N. Turner, Hewstone, Voci, Paolini, & Christ, 2007; R. N. Turner, Hewstone, Voci, & Vonofakou, 2008). Contact that disconfirms stereotypes can also lead to more generalized, open-minded attitudes and flexible thinking (Crisp & Hewstone, 2007; Crisp & Turner, 2011; Gocłowska, Crisp, & Labuschagne, 2013). Mechanisms mediating the effects of intergroup contact include reduced anxiety about future interactions, increased liking through familiarity, and increased empathy and perspective-taking (Pettigrew & Tropp, 2008).

Prejudice-reducing effects are greater if groups are initially primed with positive affirmation of their distinct group identities (see Brown & Hewstone, 2005). If members are criticized for having a distinct group identity, or if characteristics of their group are criticized, then triggering of intergroup anxiety can lead to an increase in prejudice (Islam & Hewstone, 1993; Stephan & Stephan, 2000). There are also conditions under which contact will backfire. If co-operators fail to achieve their goal, then the overall effect can entrench prejudice, rather than reduce it; and if groups have identical tasks, then this can sometimes lead to intergroup anxiety, causing the individuals’ sense of positive group identity to be threatened, leading to defensiveness and hostility (Stephan & Stephan, 2000).

Reduced prejudice can also follow from merely imagining positive contact with outgroup members (Crisp, Stathi, Turner, & Husnu, 2008; Crisp & Turner, 2009; Miles & Crisp, 2014; R. N. Turner & Crisp, 2010). For instance, reading a story about intergroup friendship increased positive attitudes towards refugees amongst young children in Rwanda (Cameron, Rutland, Brown, & Douch, 2006; see also Paluck, 2009). Prejudice-reducing effects from imagination-based exercises are greatest when participants imagine interactions that are positive, relaxed, and comfortable (Stathi & Crisp, 2008; see also Blair, Park, & Bachelor, 2003). Television shows portraying charismatic outgroup members and positive intergroup contact have been found to reduce prejudice (Schiappa, Gregg, & Hewes, 2005; see also Paluck & Green, 2009a). Prejudice reduction can also result from perspective-taking exercises, such as imagining being a member of the target group (Batson, Early, & Salvarini, 1997; Batson, Eidelman, Higley, & Russell, 2001). For instance, writing an essay from the perspective of an outgroup member going through an important life event can reduce subsequent stereotyping of members of that group (Galinsky & Moskowitz, 2000).

Interventions involving contact, imagined contact, and perspective-taking have been utilized within the context of interfaith dialogue and religious peacebuilding, sometimes in high-conflict circumstances (e.g., Abu-Nimer, 1996a, 1996b, 2008; Gopin, 1997; Johnston & Sampson, 1995; Lederach, 1995; Paolini et al., 2004; Smock, 2006; Swart et al., 2010). These interventions have often drawn on a mixture of techniques, including imagination-based exercises and perspective-taking, affirmation of positive distinct identities, and ingroup bonding, followed by structured dialogue with outgroup members, cooperative volunteering, and so on (Smock, 2006). These interventions have also often involved facilitating opportunities for members of different groups to listen to one another’s experiences in a structured setting (Smock, 2006). An interesting direction for future research would be for social psychologists to collaborate more extensively with interfaith groups to examine the relative effectiveness of different contact-based techniques in diverse settings.

Finally, leadership and perceived norms can both play influential roles. Examples set by leaders can substantially influence a community’s beliefs, attitudes, and behaviours (Landis, Hope, & Day, 1984; Levy, Stroessner, & Dweck, 1998). Moreover, people’s attitudes are often influenced by their beliefs about what is normative for their group; being told that the majority of members of their group hold a belief makes people more likely to report that they also hold this belief (Crandall & Stangor, 2005; Levy et al., 1998; Stangor, Sechrist, & Jost, 2001). A religious community’s traditions are often considered sources of moral and spiritual guidance, reflecting the cumulative moral knowledge of that community. This suggests that religious groups might be able to reduce prejudice by raising awareness of prejudice-reducing norms from the community’s traditions, such as norms of intergroup fellowship and cooperation.

A number of potential objections might be raised regarding the findings discussed in this section. One is that the contact hypothesis has most often been tested in lower conflict settings, and further research is needed to understand the effects of contact in higher intensity conflict situations, and across a wider range of cultural contexts (Paluck & Green, 2009b). It is also worth emphasizing a second concern, which relates to a potential “sedative” effect of intergroup contact. Religion has sometimes been used as a vehicle for collective action, and this is one important way in which it may support reduced prejudice (Harvey, 2011; Morris, 1984). However, intergroup contact may have a “sedative” effect on collective action (Cakal, Hewstone, Schwär, & Heath, 2011; van Zomeren, Postmes, & Spears, 2008). By reducing threat and increasing liking between groups, it may make disadvantaged groups less likely to resist unjust social relations. It may also have a sedative effect by reducing perceptions of relative deprivation, and by reducing the strength of ingroup identification (Cakal et al., 2011). These hypotheses, therefore, should be regarded only as a starting point. More sophisticated developments would integrate contact theory with theories of collective action (Cakal et al., 2011; van Zomeren et al., 2008; Wright, 2009).

Conclusion

In this paper, we have proposed that religious communities can in principle either promote or reduce prejudice amongst their members. Their influence will depend upon the specific beliefs, attitudes, and practices of the community, and on their interaction with the broader society in which they are based. We have hypothesized that religious communities will tend to reduce prejudice to the extent that they do the following.

Generalized World-Views

• 1. Teach that differences between social groups are often driven by differences in circumstances, instead of innate or essential group characteristics.

• 2. Teach pluralistic and inclusive theologies.

Society-Specific Social Beliefs

• 3. Challenge prejudice-supporting ideologies, and support mobilization around egalitarian social messages.

• 4. Draw attention to positive common ingroup identities, while also making salient positive distinct identities of different subgroups.

Behaviours

• 5. Foster intergroup contact, friendship, and cooperation; model sympathetic perspective-taking; and create opportunities to imaginatively explore the experiences of outgroup members.

In relation to each of these hypotheses, it is likely that guidance from religious leaders and the tradition’s perceived norms may be particularly influential. In outlining these hypotheses, our aim has been to indicate possible directions for future research. We hope that readers will test these hypotheses, and will identify further hypotheses by drawing on their own expertise in prejudice research. There are many reasons for prejudice researchers to develop sophisticated understandings of the diverse influences of religion on intergroup attitudes. First, as we have seen, religion may sometimes be a vehicle through which social hierarchies are enforced. Second, religion has also played important roles historically in movements to challenge prejudice and unjust social hierarches. Finally, the vast majority of people around the world today identify with a religious faith. A sophisticated appreciation of how religious groups can reduce prejudice may be important for improving intergroup relations in many communities around the world.