Culture and Stereotyping Processes: Integration and New Directions

Stereotyping is one of the largest and most enduring research areas in social and personality psychology; many of the processes by which stereotypes are formed, maintained, and applied are now well understood. Yet, little is known about the degree to which stereotyping processes apply outside of North American and Western European contexts. This theoretical paper aims to serve as a starting point for researchers interested in the intersection of culture and stereotyping. We review the nascent literature documenting similarities and differences in intergroup perception across cultural groups and note areas in which the cross-cultural and stereotyping literatures have explored common mechanisms that could be profitably integrated. Finally, we offer suggestions for future research that will greatly improve our understanding of how culturally influenced cognitive tendencies influence the perception of social groups and their members. The goal of this article is to provide a scaffold for the integration of two literatures central to social and personality psychology: stereotyping and culture. Stereotyping and related subjects such as prejudice and intergroup relations are among the topics most dear to the heart of social psychologists. Indeed, fully 40% of researchers profiled at http://www.socialpsychology.org report studying either intergroup relations or stereotyping and prejudice (Social Psychology Network, 2009). At the same time, interest in cultural psychology has been increasing rapidly. Based on the number of social psychological journal articles indexed in the American Psychological Association’s PsycInfo database with the descriptor ‘Cross-Cultural Differences,’ the proportion of social psychological research that included a cross-cultural comparison more than doubled between the 1979–1988 and 1999–2008 periods. Yet with a few notable exceptions (e.g., Terracciano et al., 2005), these research topics have been studied in parallel rather than in concert. We submit that cultural psychology has much to gain from an incorporation of stereotyping and group perception processes into its growing documentation of the effects of culturally influenced beliefs on social cognition and behavior. In the psychological literature, most recent definitions of culture characterize it as a set of shared meanings and practices that are transmitted, communicated, and reinforced among members of a cultural group, and that, in turn, influence the values, beliefs, emotions, and behaviors of their members (Chiu & Hong, 2007; Goodenough, 1999; Lyons & Kashima, 2001; Oyserman & Sorensen, 2009; Triandis, 1996; Triandis & Wasti, 2008). This transmission of a cultural group’s shared knowledge is key to its sustenance. The use of stereotypes, which are a means by which knowledge about social groups is shared among members of a culture, is an important component of a definition of culture as shared knowledge. Social and Personality Psychology Compass 4/8 (2010): 591–604, 10.1111/j.1751-9004.2010.00288.x a 2010 The Authors Journal Compilation a 2010 Blackwell Publishing Ltd Likewise, research in stereotyping and prejudice is incomplete without an understanding of the degree to which stereotyping processes are universal or limited to North American and European contexts. By focusing on stereotyping processes in this article, we are primarily highlighting the potential for cultural comparisons in such mechanisms as stereotype activation, stereotype application and use, stereotype persistence, and the group perception processes that precede stereotyping (such as categorization). We distinguish this focus from work on stereotype content or specific beliefs about a social group. In a historical review of the stereotyping and prejudice literature, Fiske (2000) cites the neglect of cultural comparisons as one of the most important problems to be remedied in future research. We see at least two major reasons why the stereotyping literature suffers from a lack of integration with the cross-cultural literature. First, psychologists have understood for some time that ‘much of what has been taken for granted in the field of psychology as ‘‘basic human psychological experience’’ ... is actually specific to middleclass European-American psychological experience’ (Markus, 2008, p. 653). We simply cannot extend psychological science further until its focus moves beyond these narrow American samples. Second, the study of culture and the study of stereotyping are fundamentally and inherently linked, as culture is the vehicle through which stereotypic knowledge is transferred within a society and across generations. As described above, most discussions of the meaning of culture emphasize that to be considered a culture a group must have a shared set of beliefs or practices and must provide for the continued communication of such beliefs among members. Recent work on stereotype communication has suggested that stereotypes about social groups and categories can be viewed as one of the key beliefs that are shared among members of a cultural group (McIntyre, Lyons, Clark, & Kashima, 2004). When people desire to get closer to each other, to initiate or strengthen their relationships, they are particularly likely to communicate stereotype-consistent (relative to stereotype-inconsistent) information. Perceived consensus about the attributes of the ingroup, as well as attributes of outgroups, serves to demarcate ingroup boundaries and facilitate interpersonal relations (Castelli, Pavan, Ferrari, & Kashima, 2009; Clark & Kashima, 2007; Kashima, 2000; Lyons, Clark, Kashima, & Kurz, 2008; Ruscher, 2001; Ruscher, Cralley, & O’Farrell, 2005). Stereotypes are ‘common ground’ that people can agree upon when discussing a third party. Indeed, the very act of communication among ingroup members appears to strengthen and reinforce stereotypic beliefs (Lyons & Kashima, 2001, 2003). This perspective demonstrates the close theoretical ties between the stereotyping and culture literatures and also helps to account for why and how stereotypes are so readily transmitted and so pervasive. In summary, research on stereotyping processes is incomplete without an investigation of both their universality versus cross-cultural variability and their central role in cultural communication. We see this article as a starting point for scholars interested in addressing the paucity of research at the intersection of culture and stereotyping. In the sections that follow, we briefly review several well-documented phenomena in stereotyping and prejudice, incorporate relevant findings from cultural psychology, and offer predictions about how these phenomena may be expected to show both universality and variability. The Incidence and Application of Stereotypes A stereotype can be defined as ‘a cognitive structure that contains a perceiver’s knowledge, beliefs, and expectancies about a human group’ (Hamilton & Trolier, 1986, p. 133), and 592 Culture and Stereotyping a 2010 The Authors Social and Personality Psychology Compass 4/8 (2010): 591–604, 10.1111/j.1751-9004.2010.00288.x Journal Compilation a 2010 Blackwell Publishing Ltd stereotyping as the application or use of those beliefs in perception, judgment, and decision making (Judd & Park, 1993). But is stereotyping universal? As a starting point toward answering this question, evidence suggests that all societies organize and categorize their members into subgroups, a necessary precondition for within-culture stereotyping (Sidanius & Pratto, 1999). Cultural subgroups are most commonly based on age and gender. However, wealthier societies that have moved beyond subsistence levels of resources, in which resources may be distributed unequally, also are likely to divide individuals into what Sidanius and Pratto term ‘arbitrary-set groups.’ Arbitrary-set groups are those based on social constructions that are specific to the given culture and that account for the unequal distribution of resources among members of the cultural group. Such constructions might include race, caste or class, occupation, religion, geography, and so on. More direct evidence for stereotype universality comes from research that has taken a cross-cultural approach to stereotype content (rather than stereotype development and use). Indeed, this body of work represents the significant exception to the dearth of research at the intersection of culture and stereotyping. A substantial number of papers have documented beliefs about the characteristics of ingroups and outgroups around the world (Best & Williams, 2001; Bond, 1986; Costa, Terracciano, & McCrae, 2001; Cuddy et al., 2009; Krueger, 1996; Leach, Minescu, Poppe, & Hagendoom, 2008; Marin & Salazar, 1985; McAndrew et al., 2000; Poppe, 2001; Terracciano et al., 2005; Terracciano & McCrae, 2007; Walkey & Chung, 1996). Yet, none of these diverse studies has reported that participants were perplexed by questions about stereotypes or were unable to complete the tasks. Moreover, within each study, stereotype content typically shows high interjudge consensus, even for ingroup stereotypes (Terracciano et al., 2005) and even when the content can be shown to be inaccurate (McCrae, 2001). As McCrae and Terracciano (2006) noted, ‘People everywhere find it easy to develop stereotyped ideas of whole nations and agree well enough with each other to believe their views are consensually validated’ (p. 160). Thus far, without exception, inquiries about stereotypes strike study participants as reasonable, in any language. The ability to develop consensual, fixed beliefs about the nature and characteristics of human social groups appears to be universal. However, while members of cultural groups around the world may be able to generate stereotypes readily in response to a researcher’s prompting, they may differ in their tendency to use stereotypes spontaneously. That is, cultural groups may vary in the frequency or degree to which they apply stereotypes. Previous work has established that members of relatively collectivistic cultures, such as those in East Asia, perceive social groups as more coherent, agentic units, compared to members of relatively individualistic cultures such as the United States. To an East Asian – but perhaps not to a North American – it is perfectly normal to think about a social group as enacting a behavior or as having a personality trait (Brewer, Hong, & Li, 2004; Kashima et al., 2005; Menon, Morris, Chiu, & Hong, 1999). Similarly, Zemba and colleagues (Zemba, Young, & Morris, 2006) report that Japanese participants, to a greater extent than American participants, attribute responsibility for a negative organizational outcome (such as a student who is sickened from a contaminated school lunch) first to the organization as a whole and only secondarily to individual managers. We suggest that although all individuals from any cultural group are able to generate stereotypes, as described previously, those who are part of cultural groups that appear especially likely to perceive groups as social agents (i.e., collectivist cultures) may develop and apply stereotypes more readily, compared to societies in which social groups are perceived as less agentic (i.e., individualistic cultures). Stereotypes take on special importance to the degree that cultural norms require individuals to behave in ways that are consistent with group expectations. Culture and Stereotyping 593 a 2010 The Authors Social and Personality Psychology Compass 4/8 (2010): 591–604, 10.1111/j.1751-9004.2010.00288.x Journal Compilation a 2010 Blackwell Publishing Ltd We tested this possibility in recent work involving study participants in the United States and China (Spencer-Rodgers, Williams, Hamilton, Peng, & Wang, 2007). In one study, participants from both cultural groups read descriptions of novel subgroups in a fictional society. They then indicated the likelihood that an individual from each subgroup would possess characteristics that were stereotypical of the group (based on the descriptive information provided) or stereotype-irrelevant, as well as the likelihood that an individual belonged to each subgroup, given that he or she possessed personality traits that were stereotypical of the group or stereotype-irrelevant. Relative to the Americans, the Chinese participants made stronger inferences about stereotypical traits (but not stereotype-irrelevant traits) from group membership and stronger inferences about group membership from stereotypical traits (but not stereotype-irrelevant traits). That is, the Chinese participants stereotyped more readily than the Americans. In two additional studies, participants made trait ratings of a hypothetical work group and the national groups Chinese, Americans, and Kenyans (a relatively neutral outgroup). Results were examined not in terms of the specific content of the trait ratings but in terms of the degree to which the ratings were internally consistent (e.g., a group that was rated high on ‘outgoing’ was also rated low on ‘shy’). For all four groups, the Chinese participants perceived a more coherent group-level personality than did the Americans. Further, for three of the four groups, the effect of culture on personality coherence was mediated by the greater tendency among the Chinese participants to view groups in general as more entitative, cohesive, and unified. In summary, collectivistic cultures appear to be more likely than individualistic cultures to perceive groups as possessing coherent personalities. As a result, members of more collectivistic cultures may be particularly likely to apply group-level characteristics to individuals (i.e., to stereotype). Indeed, the very idea of a stereotype, that an individual is perceived through the lens of his or her social group membership, may not be perceived negatively by members of cultural groups who strongly value their group memberships and more readily adhere to their groups’ norms (Lee & Ottati, 1993; Markus & Kitayama, 1994). Ingroups and Outgroups A significant portion of the stereotyping literature has been devoted to differences in the perception of groups of which one is a member (ingroups) versus not a member (outgroups). There is robust data to suggest that people prefer ingroups to outgroups (Brewer, 2007) and see outgroups as more homogenous than ingroups (Boldry, Gaertner, & Quinn, 2007). We consider these findings in cultural context. Brewer (1999, 2007) argues that the valuing of ingroups is universal, an essential consequence of the fundamental human need to belong (also see Beaupré & Hess, 2003; Bennett et al., 2004; Brewer & Campbell, 1976; Koomen & Bähler, 1996). An unanswered question is whether cultural groups should be expected to vary in the degree to which they show ingroup favoritism. On the one hand, there is evidence that collectivists draw sharper distinctions between ingroup and outgroup members than do individualists when deciding how to resolve a conflict (Leung, 1988; Li, 2002), allocate resources (Leung & Bond, 1984), or select a communication style (Gudykunst et al., 1992), implying that the preferential valuing of ingroups over outgroups might be even stronger among collectivists. On the other hand, papers that report direct comparisons of ingroup favoritism across cultural groups generally show lesser ingroup enhancement among members of collectivist than individualist cultures (Bond & Hewstone, 1988; Cuddy et al., 2009; Heine & Lehman, 1997; Ma-Kellams, Spencer-Rodgers, & Peng, in press; Snibbe, Kitayama, Markus, 594 Culture and Stereotyping a 2010 The Authors Social and Personality Psychology Compass 4/8 (2010): 591–604, 10.1111/j.1751-9004.2010.00288.x Journal Compilation a 2010 Blackwell Publishing Ltd & Suzuki, 2003). Yuki (2003) accounts for these results by arguing that collectivists’ valuation of ingroups is based not on an attachment to the group per se (e.g., me and my family), but rather on the nature of the multiple dyadic relationships between the self and various ingroup members (e.g., me and my father, me and my third cousin). Indeed, evidence suggests that collectivists’ ingroup-favoring behaviors are mediated by specific relationship contexts or anticipated reciprocation by ingroup members (Endo, Heine, & Lehman, 2000; Li, 2002; Yamagishi & Kiyonari, 2000). Thus, for collectivists (more than for individualists), the degree of ingroup favoritism may be particularly sensitive to relational factors (Chen, Brockner, & Chen, 2002) and whether attachment to the ingroup versus specific ingroup members is being measured. Other research provides a dialectical account for the lesser ingroup favoritism sometimes shown by East Asians, relative to North Americans (Ma-Kellams et al., in press). Members of East Asian cultures that espouse a dialectical view of the world are more comfortable with contradictions, compared to North Americans, and show greater acceptance of both the favorable and unfavorable aspects of an object or phenomenon, including the groups to which they belong (Peng & Nisbett, 1999; Spencer-Rodgers, Williams, & Peng, in press). Against the backdrop of cultural variation in ingroup favoritism, however, there is virtually no evidence of country-level differences in the tendency to derogate outgroups, which has been demonstrated to be theoretically and empirically separable from the tendency to favor ingroups (Brewer, 1999). Work on system justification theory suggests that members of cultural subgroups who have relatively lower status within a culture (e.g., ethnic minority groups) sometimes show less outgroup derogation than do members of the dominant majority group (Jost & Burgess, 2000). However, when comparing across participants who represent majority groups in their countries of origin, outgroup derogation has been argued to appear only under specific circumstances of intergroup threat (Fiske, Cuddy, Glick, & Xu, 2002; Hewstone, Rubin, & Willis, 2002; Jackson, 2002; Mikulincer & Shaver, 2001) and thus is more likely to vary as a function of specific group conflicts than as a function of the cultural background of the perceiver. Thus, while our previous research has shown that members of collectivistic cultures more readily stereotype, compared to members of individualistic cultures, we do not hypothesize that members of collectivistic cultures will necessarily be quicker to display outgroup prejudice. Indeed, in our work, we found increased stereotyping among Chinese relative to American participants, but no cultural differences in the valence of outgroup attitudes (Spencer-Rodgers et al., 2007). Along with the tendency to prefer ingroups to outgroups, a similarly robust finding among Western samples is the tendency to perceive ingroups as more diverse and variable than outgroups (the outgroup homogeneity effect, Jones, Wood, & Quattrone, 1981; Linville, 1982; Linville, Fischer, & Salovey, 1989; Quattrone & Jones, 1980). In contrast to these findings, members of collectivist cultures sometimes show ingroup homogeneity effects (Triandis, McCusker, & Hui, 1990; Yuki, 2003). Compared to individualists, collectivists may not feel as strong a need to be differentiated from the groups to which they belong (Triandis, Bontempo, Villareal, Asai, & Lucca, 1988; Tropp & Wright, 2003) and may not ascribe a negative connotation to within-group sameness (Lee & Ottati, 1993). Similarly, Brewer and Roccas (2001) posit that people with collectivist values should be drawn to relatively homogenous ingroups, a result of a stronger motivation for social inclusion. As a consequence, collectivist cultures may show greater ingroup stereotyping, that is, a stronger tendency to perceive ingroup members as sharing common characteristics, and therefore reduced differences in perceived homogeneity between the ingroup and the outgroup (Lee & Ottati, 1995). Culture and Stereotyping 595 a 2010 The Authors Social and Personality Psychology Compass 4/8 (2010): 591–604, 10.1111/j.1751-9004.2010.00288.x Journal Compilation a 2010 Blackwell Publishing Ltd Stereotype Stability One of the biggest challenges noted by stereotyping researchers studying Western participants is how resistant stereotyping is to change. Despite many individuals’ best efforts to avoid applying group characteristics (particularly negative ones) to individual group members, stereotyping persists (e.g., Devine, 1989; Macrae, Bodenhausen, Milne, & Jetten, 1994). The degree to which stereotyping is equally resistant across cultures (e.g., to changing societal norms and individuals’ goals to avoid stereotyping) has gone virtually unstudied. However, there are reasons to anticipate that stereotyping might be more malleable among members of non-Western than Western cultures. We ground this prediction in two core findings from cultural psychology: greater incorporation of situational influences in explaining others’ behavior among members of non-Western cultural groups and greater expectation of change in East Asian cultural groups in particular, relative to Western cultural groups. An increased tendency among members of collectivist cultural groups to incorporate situational and contextual factors (in addition to stable, dispositional properties of the person) in explaining the behavior of others has been well established (Choi, Nisbett, & Norenzayan, 1999; Miyamoto & Kitayama, 2002; Morris & Peng, 1994). These findings have important, understudied implications for stereotyping processes. Indeed, stereotyping can be viewed as a form of attribution; one way to stereotype is to observe the behaviors of a social group (such as a low-status outgroup committing a high rate of thefts) and attribute stable, dispositional characteristics to individual group members (they are dishonest). We do not argue that collectivists’ greater attention to contextual influences on behavior will lead them to abandon stereotyping altogether – indeed, evidence suggests the opposite prediction (Spencer-Rodgers et al., 2007), as previously discussed. Instead, it may be the case that stereotypes will take a more contextualized form when used by collectivists, relative to individualists. For example, even while applying a stereotype of dishonesty to members of a lower-status group, members of a collectivistic culture may recognize that such dishonest behaviors reflect the social role assigned to group members, or situational and environmental demands, rather than deep, stable personalities of individual members of the group (see Costa, Terracciano, & McCrae, 2001, for a similar argument). Fiske, Cuddy, and their colleagues have suggested in a number of recent papers (Caprariello, Cuddy, & Fiske, 2009; Cuddy, Fiske, & Glick, 2007, 2008; Fiske, Cuddy, & Glick, 2007; Fiske, Xu, & Cuddy, 1999; Fiske et al., 2002) that the content of outgroup stereotypes around the world can be predicted from outgroups’ sociostructural position, namely their relative status and the degree to which they are seen as competing with the ingroup. In both Western and non-Western samples, higher-status groups are stereotyped as more competent than lower-status groups, and more competitive, threatening groups are stereotyped as less warm (Cuddy et al., 2009; Eckes, 2002; Fiske et al., 2002; Leach et al., 2008). In complementary thinking on gender stereotypes, Eagly and colleagues have demonstrated that stereotypes of men as agentic and women as communal arise from the social roles they play as breadwinners and caregivers, respectively (Cejka & Eagly, 1999; Eagly, 1987; Eagly & Steffen, 1984, 1986; Eagly, Wood, & Diekman, 2000). What both perspectives have in common is the idea that dispositional characteristics of group members are inferred from the group’s role or position in society. As social roles change, stereotypes can and do change as well (Diekman & Eagly, 2000; Diekman, Eagly, Mladinic, & Ferreira, 2005). We suggest, however, that while these basic sociostructural mechanisms of stereotype content development are likely to hold up across cultures, the 596 Culture and Stereotyping a 2010 The Authors Social and Personality Psychology Compass 4/8 (2010): 591–604, 10.1111/j.1751-9004.2010.00288.x Journal Compilation a 2010 Blackwell Publishing Ltd stereotypes maintained by members of cultural groups that are more inclined to incorporate contextual influences into explanations for behavior (i.e., collectivist cultures) may be more sensitive to societal shifts in a group’s social role, and therefore quicker to change. That is, we predict that stereotypes in collectivist cultures will be less vulnerable to the ‘cultural lag’ described by Diekman et al. (2005), or the delay between when a group’s social role changes and when stereotypes of group members are updated. Zhang and Hunt (2008) showed that Chinese participants, relative to Americans, were less vulnerable to stereotype rebound, or an increase in stereotype use following instructions not to use stereotypes, in evaluating a target (in this case, a gay man). This result occurred in spite of higher levels of antigay prejudice among the Chinese participants. One factor contributing to this phenomenon may be the greater malleability of stereotypes held by members of collectivistic versus individualistic cultures. That is, the Chinese participants in this study were better able than the Americans to adjust their stereotype use to fit a change in social expectations (i.e., a new instruction to avoid stereotypes of gay men). Additional support for the possibility that stereotyping by members of collectivistic cultures – especially those located in East Asia – may be more malleable (compared to those held by members of individualistic cultures) comes from the literature on naı̈ve dialecticism (Peng & Nisbett, 1999; Spencer-Rodgers et al., in press). Members of dialectical cultures (particularly China, Japan and Korea) have been shown to expect more and to better tolerate, change and inconsistency in their environment (Alter & Kwan, 2009; Cheng, 2009; Choi & Choi, 2002; Ji, Nisbett, & Su, 2001; Ji, Zhang, & Guo, 2008; Spencer-Rodgers, Boucher, Mori, Peng, & Wang, 2009a). In one study, for example, Japanese participants expected less consistency between others’ attitudes and behaviors, compared to Australian participants (Kashima, Siegel, Tanaka, & Kashima, 1992). Similarly, Korean participants were less surprised than Americans by people who violated stereotypic expectations, such as a seminary student who fails to help a person in need (Choi & Nisbett, 2000). Extended to the domain of stereotyping, these findings suggest that unlike Westerners, East Asians may maintain outgroup stereotypes yet not be surprised when the behavior of individual group members contradicts those stereotypes. Put differently, stereotypes held by East Asians may be less rigid and more amenable to change (such as if the social role of a group changes over time) or contradiction (such as if a group member behaves in a stereotype-inconsistent fashion). In the West, study participants have repeatedly shown difficulty reconciling their stereotypes of outgroups with the counterstereotypic behavior of individual outgroup members (Kunda & Oleson, 1997; Trope & Thompson, 1997). To resolve this contradiction and maintain the integrity of their original stereotype, Westerners may resort to such strategies as creating new subcategories for counterstereotypical outgroup members (Kunda & Oleson, 1995; Weber & Crocker, 1983), changing the standard of comparison against which the target is evaluated (Collins, Crandall, & Biernat, 2006), or responding with hostility to outgroup members who violate stereotypic expectations (Rudman & Fairchild, 2004). We hypothesize that such phenomena should occur less frequently among East Asians, who are likely to feel less negative tension and dissonance in response to a contradiction between an outgroup stereotype and an individual outgroup member’s counterstereotypic behavior. Indeed, the empirical literature on the dialectical resolution of inconsistent information would suggest that in response to evidence of counterstereotypic behavior, a dialectical thinker might, instead of bolstering his or her stereotype via one of the means described above, attenuate the extremity of the stereotype to account for the new information (Peng & Nisbett, 1999; Spencer-Rodgers, Boucher, Peng, & Culture and Stereotyping 597 a 2010 The Authors Social and Personality Psychology Compass 4/8 (2010): 591–604, 10.1111/j.1751-9004.2010.00288.x Journal Compilation a 2010 Blackwell Publishing Ltd Wang, 2009b; Spencer-Rodgers et al., in press). In summary, we predict that, relative to stereotyping by members of individualist cultures, stereotyping by collectivists is likely to be more malleable in response to changes in a group’s social role or to evidence inconsistent with the stereotype.