Download Self - Categorization and Social Identification and more Study notes Social Psychology in PDF only on Docsity!
Self-Categorization and Social Identification:
Making Sense of Us and Them^1
Nils Karl Reimer, Katharina Schmid, Miles Hewstone, &
Ananthi Al Ramiah
Social identity processes are central to social psychological theory and research. Social identity plays an important role in many individual and social phenomena, including health and well-being ( e.g., C. Haslam, Jetten, Cruwys, Dingle & Haslam, 2018), educational achievement ( e.g., Steele, Spencer & Aronson, 2002), and collect- ive action and social change (e.g., van Zomeren, Postmes & Spears, 2008). In this chapter, we provide an overview of social psychological theories on social identity, including social identity theory ( Tajfel, 1978; Tajfel & Turner, 1979) and self-categorization theory ( Turner, Hogg, Oakes, Reicher & Wetherell, 1987). Prior to Tajfel’s and Turner’s seminal writings, two types of theory dominated ex- planations of intergroup relations and intergroup conflict. First, so-called ‘prejudice’ accounts proposed that intergroup conflict results from individual-level variation in a prejudiced personality ( Son Hing & Zanna, 2010). Second, realistic group conflict theory (Campbell, 1965; Sherif, 1966) contended that negative intergroup relations stem from conflicting group goals and competition over resources or power (see Jackson, 1993, for a review). Compared to these explanations, social identity theory and self-categorization theory offered a unique and refined account of intergroup relations that remains important to this day. This chapter is dedicated to these interrelated approaches for understanding intergroup relations. Rather than present these theories chronologically—or focus on differences and similarities across the two theories—we take a more encompassing approach to what is often called the ‘social identity perspective’ (Abrams & Hogg, 2010). Beyond Tajfel’s and Turner’s seminal writings, we focus on more recent efforts to define social identification, and discuss how multiple categorization offer insights into the dynamics underlying (^1) Cite as: Reimer, N. K., Schmid, K., Hewstone, M., & Al Ramiah, A. (2020). Self-categorization and social identification: Making sense of us and them. In D. Chadee (Ed.), Theories in social psychology (2nd ed.). Wiley-Blackwell.
intergroup relations. This chapter is organised into five sections. The first three sections define social categorization, self-categorization and social identification, and the latter two sections examine how these processes are involved in, and help explain, in- tergroup relations. Section 1 introduces social categorization as a fundamental process that underlies self-categorization and social identification. Section 2 defines self-categorization as social categorization applied to oneself, and introduces self- categorization theory as a detailed account of when, how, and why we describe ourselves in terms of social categories. Section 3 defines social identification as a multidimensional process by which we attach value and meaning to our self- categorizations. Section 4 introduces social identity theory as a detailed account of how social identification relates to intergroup relations, focusing on findings from the minimal group paradigm. Section 5 reviews recent advances in understanding how multiple categorization shapes intergroup relations, discussing how more com- plex social identities across multiple categories can improve intergroup relations. We end by drawing some general conclusions, and by discussing implications of the social identity perspective for how we understand intergroup relations.
From Social Categorization to Self-Categorization
As humans, we readily categorize ourselves and others into social groups. That is, we spontaneously divide people according to ascribed categories (e.g., ethnicity or gender), achieved states (e.g., occupation or political affiliation), and other relevant distinctions ( Deaux, 1996)—or, more generally, based on perceptions of similarity, proximity, and common fate ( Campbell, 1958). This process of social categorization is often fast and automatic (Macrae & Bodenhausen, 2000). For example, people recognise another person’s race ( Richeson & Trawalter, 2005) or gender ( Martin & Macrae, 2007) within less than a second of stimulus presentation. Social cat- egorization matters because it helps us navigate a social world marked by group distinctions and hierarchies, but also because it underlies harmful social stereotypes ( Tajfel, 1969). Social categorization acquires further significance when applied to ourselves. As humans, we depend on groups for survival (Brewer, 2007), and thus have developed a fundamental need to belong to social groups (Baumeister & Leary, 1995). This means that we are motivated to categorize ourselves into some groups, and out of others. In the next section, we introduce self-categorization theory as a perspective on how, when, and why individuals categorize themselves into certain social groups.
different football teams. Self-perception and social behaviour, however, should not be understood as either interpersonal or intergroup, but rather as varying along a continuum from purely interpersonal to purely intergroup properties (Tajfel, 1978; Turner, 1982). By considering different levels of self-categorization, self- categorization theory helps us to understand self- and other-stereotyping, as well as ingroup- and outgroup-stereotyping. In both cases, the theory emphasises that individuals ascribe to themselves characteristics associated with their ingroup. In principle, self-categorization thus develops the earlier social identity perspective which we describe in detail later on in this chapter.
When and why do people self-categorize?
Self-categorization theory recognises not only that self-categorization is hierarchical, as outlined above, but also that self-categorization is situational. As discussed, individuals categorize themselves in terms of their personal identity or various social identities. This raises the question of when and why individuals categorize themselves in terms of a collective category (or not). Self-categorization theory states that whether an individual categorizes themself in terms of a collective category depends on the salience of that category in a given situation ( Turner, 1999; Turner et al., 1987). Salience, in this case, is the degree to which a group membership shapes social perception and behaviour in a given situation (Oakes, 1987), and the extent to which group members perceive themselves as similar to other members of their own group, and as dissimilar to members of other groups ( Turner, 1999). In other words, self-categorization depends on the extent to which a particular social category becomes psychologically activated in a given situation ( Veenstra & Haslam, 2000). Whether a specific ingroup–outgroup distinction becomes salient depends on an interaction between its relative accessibility and fit (Oakes, 1987). A category’s accessibility depends on an individual’s immediate social context (situational ac- cessibility), as well as the extent to which an individual tends to make use of a self-categorization across a range of situations (chronic accessibility). For example, a female student’s gender category may become psychologically activated if she finds herself in a room filled with male students (high situational accessibility), but not in a room filled with female students (low situational accessibility). An example of chronic accessibility would be if the same student is part of a feminist campaign group which makes her much more aware of, and think more frequently about, her gender. A category’s fit depends on its comparative and normative fit in a given situation ( Turner et al., 1994). Comparative fit is the extent to which, in a given social context, a person perceives greater similarity between the self and own-group members than between the self and other-group members (the so-called ‘meta-
contrast ratio’). In other words, the smaller the perceived differences between members within one’s own group and the greater the perceived differences between one’s own group and a comparative group are, the greater is the likelihood that a person will categorize themselves in terms of their own group. Normative fit is the extent to which these perceived differences between groups correspond to expected differences in meaning ( S. A. Haslam, Oakes, Reynolds & Turner, 1999; Oakes, Turner & Haslam, 1991; Veenstra & Haslam, 2000). Specifically, one’s own group (the ingroup) and the comparative group (the outgroup) need to comply with one’s stereotypical beliefs about the differences between them (Oakes, 1987). For example, consider a situation in which a group of male and female students are asked to join either the university’s rugby team or its dance troupe. Suppose that all men opt for the former, while all women choose the latter. In this situation, the two groups show high intragroup and low intergroup similarity (comparative fit), and also conform to the expected norm (normative fit). Psychological activation of social categories sets in motion a process of depersonalization by which individuals may come to perceive themselves as inter- changeable representatives of their group ( Turner et al., 1994). As a result, individuals cognitively align themselves with the ingroup prototype (self-stereotyping), shift from personal to group-based perceptions and behaviours, and begin to integrate the group into their self-concept ( Smith & Henry, 1996). This process explains how individuals come to think and act in terms of their group identity (see, for ex- ample, S. A. Haslam et al., 1999; S. A. Haslam, Turner, Oakes, McGarty & Reynolds, 1998, even to the exclusion of their personal identity. Thus far, we have described how, when, and why individuals categorize themselves in terms of social categor- ies. Self-categorization is, however, only one part of social identification. Social psychological research emphasises that people often derive meaning from their self-categorizations and attach emotional significance to it. In the next section, we define social identification as a multi-dimensional process that goes beyond mere self-categorization.
From Self-Categorization to Social Identification
Defining social identification
Social identification varies from individual to individual, and from situation to situation, reflecting the extent to which a group membership is incorporated into an individual’s self-concept (e,g, Branscombe & Wann, 1994; Waddell & Cairns, 1986). In its simplest definition, social identification denotes that an individual perceives their group and their membership thereof as meaningful, desired, and important. Social identification is a broad term that summarises a complex set of processes
evaluative ratings of one’s identity). In contrast, importance is the degree to which an individual perceives a given social identity as important to their self-concept. Ashmore et al. (2004) contended that researchers typically refer to this dimension when using the term ‘strength of identification’. Importance can be subdivided into explicit and implicit importance. Explicit importance is an individual’s subjective sense of how important a given social identity is to their self-concept. Implicit importance assumes that an individual’s identities are hierarchically structured according to their importance for the individual’s sense of self (see, for example, Stryker, 2000). Individuals need not, however, be consciously aware of the implicit importance of their identities. The fourth component, attachment and interdependence, refers to the sense of emotional involvement and oneness with a group. It defines the extent to which individuals perceive members of a given group to receive similar treatment in society (interdependence/mutual fate), the extent to which they are emotionally tied to the group (attachment/affective commitment), and the extent to which they see the self and the group as overlapping (interconnection of self and others). The fifth compon- ent, social embeddedness, describes the extent to which individuals’ social identities are implicated in their everyday lives and social relationships. The sixth component, behavioural involvement, denotes the extent to which individuals’ behaviour and actions are functionally dependent on their social identity. Content and meaning constitute the final component in Ashmore et al.’s ( 2004) framework. This compon- ent reflects both the extent to which a social group’s (stereo-)typical characteristics are seen as reflecting the group members’ traits (self-attributed characteristics), and the ideological beliefs surrounding the group’s experience and role in society, as well as the internalised story of the group and one’s membership thereof (narrative). Ashmore et al. (2004) argued that although these components often covary, they need not converge or correlate for every given category and individual. For example, a person might evaluate a social category positively and consider it very important, while another person might evaluate that category positively without considering it important. Similarly, a person might consider two social categories important, but evaluate one positively and one negatively. While we acknowledge this complexity, we use the summary term ‘social identification’ in the remainder of this chapter—both for the sake of simplicity and to be consistent with earlier research and theories. We also note that although social identification is best under- stood as multidimensional, a single-item measure of social identification has been shown to be reliable, valid, and useful in research practice (Postmes, Haslam & Jans, 2013). In the next section, we discuss social identity theory as a perspective on the consequences of social identification for intergroup relations.
Social Identity Theory
Intergroup bias and the “minimal group paradigm”
Social identification changes the nature of social categorization. We divide others not just into groups, but into groups we are also part of (ingroups) and those we are not part of (outgroups). Social identities determine whom we categorize as “us” and “them”. Dividing others into “us” and “them”, in turn, changes how we treat members of these groups. Intergroup bias refers to “the systematic tendency to evaluate one’s [ingroup] or its members more favourably than [an outgroup] or its members” ( Hewstone, Rubin & Willis, 2002, p. 576). Social categorization into ingroup and outgroup is, by definition, necessary for intergroup bias to occur. But is it also sufficient? Tajfel, Flament, Billig and Bundy (1971) recognised that to answer this ques- tion, they needed to strip social categorization of the regular trappings of group membership. That is, they set out to study social categorization unconfounded by prior attitudes, historical ties, face-to-interactions, and—in contrast to earlier research (Sherif, Harvey, White, Hood & Sherif, 1954/1961)—in the absence of dir- ect competition or conflicting goals. To that end, Tajfel and colleagues developed the ‘minimal group paradigm’ for examining the effects of social categorization in laboratory settings. In the original studies, the researchers randomly allocated participants into one of two groups, ostensibly based on trivial criteria such as their responses in a visual-judgement task (Experiment 1) or their preference for paintings by one of the abstract artists Klee and Kandinsky (Experiment 2). Participants—in this case, adolescent boys—did not know which of their classmates were in which group. Next, the researchers isolated participants into cubicles, and asked them to distribute money between two other participants according to a range of predefined choices. Participants made these choices with complete anonymity, knowing the group memberships but not the names of the other participants. Still, participants tended to favour members of their own group when allocating money between an ingroup and an outgroup member—choosing a fair allocation less often than when allocating money between two outgroup or two ingroup members. As the exper- imental setting ruled out alternative explanations for intergroup bias, Tajfel et al. (1971) concluded that social categorization was indeed sufficient to elicit discrimina- tion. Since the original studies were published, researchers have used this paradigm to uncover bias across many other measures, including, for example, better recall of unfavourable outgroup over ingroup behaviours (Howard & Rothbart, 1980). Researchers have debated various processes that could explain why social categorization leads to discrimination in minimal group studies ( see Hewstone et al., 2002, for a review). Tajfel (1970) first offered a norm-based explanation, before devel-
ate between minimal ingroup and outgroup members when distributing negative outcomes (that is, punishments rather than rewards). This suggests that mere categorization into “us” and “not us” is not sufficient to elicit outgroup-directed discrimination. Sexual prejudice ( Herek & McLemore, 2013), for example, does not seem to be motivated by the desire to advantage one’s ingroup. Instead, sexual prejudice is often motivated by the desire to distance oneself from the outgroup: “I’m not gay... I’m a real man!” (Falomir-Pichastor & Mugny, 2009). Discrimination thus arises from individuals harming outgroup members in the effort to distance themselves (“me”) from a stigmatised outgroup (“them”). As discussed above, ingroup favouritism does not usually necessitate out- group derogation. The two are, however, intrinsically linked in some situations ( Brewer, 1999, 2007). For example, people sometimes perceive an outgroup’s in- terests as being at odds with their ingroup’s—especially when they see the outgroup as a threat to the ingroup ( Sherif et al., 1954/1961; Stephan & Stephan, 2000). In in- tergroup conflict, people might harm the outgroup in order to benefit their ingroup. This relationship tends to be stronger for individuals who strongly identify with the threatened group than for those to whom the threatened group is less important ( see, for example, Bizman & Yinon, 2001; Tausch, Hewstone, Kenworthy, Cairns & Christ, 2007). Discrimination thus results from protecting a valued ingroup (“us”) against the perceived threat from an outgroup (“them”).
Social identity and intergroup relations
Social identity theory has important implications for understanding intergroup pro- cesses beyond artificially created laboratory settings. Tajfel (1981) developed social identity theory to understand intergroup relations by explaining how individuals perceive, and act as a consequence of, their membership in social groups. The theory seeks to understand intergroup relations in real-life settings defined by hierarchies and status differences across groups. The theory contends that individuals have a need to achieve and maintain positive distinctiveness from other groups. This means that individuals strive to demonstrate that the ingroup is, in some form or other, better than the outgroup—or at least different from them (see, for example, Mummendey & Schreiber, 1983). Outside of the laboratory, individuals often at- tach value and meaning to their self-categorizations. As such, social identities help explain how individuals relate to and perceive other social groups. Social identity theory thus argues that when individuals identify with their ingroup and when functional relations between ingroup and outgroup are salient, individuals strive to maintain or attain positive distinctiveness. According to Brewer (2000), this need for positive distinctiveness may lead to intergroup accentuation (individuals minimise differences between ingroup members, and maximise differ- ences between ingroup and outgroup members), ingroup favouritism (individuals
generalise positive affect to ingroup members, but not outgroup members), and social competition (individuals devalue the outgroup when doing so elevates the ingroup). Together, these dynamics can help explain intergroup conflict and group- based status relations in societies. In his theorising, Tajfel (1978) focused on what this means for members of lower-status groups. Below, we discuss how group status influences individuals’ strategies for achieving positive distinctiveness.
Group status and strategies for achieving positive distinctiveness
In their quest for a positive social identity and positive distinctiveness, individuals are bound by, and take into consideration, group status and group boundaries. Intergroup attitudes and behaviour are a function of how low or high the ingroup and outgroup are in their relative status, how stable and legitimate intergroup relations are perceived to be, and whether group boundaries are permeable or impermeable (Ellemers, Wilke & van Knippenberg, 1993; for a review, see Turner, 1999). These factors interact in shaping how members of low-status and high-status groups attain and/or maintain positive distinctiveness. When group boundaries are permeable, low-status group members may engage in social mobility strategies, at least if they do not identify strongly with their group. That is, they may leave or disassociate from the ingroup, and may attempt to join the higher-status outgroup. When group boundaries are impermeable, low-status group members may instead attempt to bolster their social identity by engaging in social creativity or social competition strategies. Social creativity is a strategy employed by low-status groups when they perceive status relations as stable and legitimate, and when group boundaries are impermeable. This strategy involves changing elements of the comparative situation in favour of the ingroup. For example, low-status group members might seek to challenge negative evaluations of characteristics attributed to their group (as in the ‘Black is Beautiful’ movement in the 1970s). Alternatively, they might compare their ingroup to an outgroup that is less valued or of even lower status than the ingroup (so-called ‘downward social comparison’). Social competition (or: collective action) is a strategy employed by low-status groups when they perceive status relations as unstable and illegitimate, and when group boundaries are impermeable. An example would be militant Black South Africans challenging the old Apartheid system in South Africa. In this case, low-status group members tend to perceive the outgroup in less positive terms, and identify more strongly with the ingroup. Social psychological research has extensively tested these predictions of social identity theory. For example, both experimental (Brown & Wade, 1987; Turner & Brown, 1978) and meta-analytic (Mullen et al., 1992; Bettencourt, Charlton, Dorr & Hume, 2001) studies demonstrated that group status can affect the nature of intergroup relations. In particular, researchers observed that members of high-
(see Leonardelli & Brewer, 2001). Subjective uncertainty reduction theory ( Hogg, 2000, 2007; Hogg & Abrams,
- posits that individuals need to feel certain about their social world and to maintain control over their lives—and thus seek to reduce uncertainty in their social perceptions. Affiliating with social groups is one way to reduce uncertainty. Self-categorization and social identification with groups—in particular, if they provide clear social norms for self-definition, social perception and behaviour— give individuals certainty about their place in the social world (see Hogg, 2007, for a review). As a consequence, individuals may evaluate ingroup members more positively than outgroup members because only the former reduce their uncertainty about the social world (see Grieve & Hogg, 1999). Subjective uncertainty reduction theory has particular relevance for situations in which individuals find themselves in new intergroup settings (for example, immigrants arriving in a country, or students arriving at their new university). Similarly, this theory helps explain why members of long-standing groups show ingroup bias when faced with system changes that could destabilize the status of their ingroup (Mullin & Hogg, 1998). Terror management theory ( Greenberg et al., 1990; Pyszczynski, Green- berg & Solomon, 1997) assumes that individuals have an inherent need for self- preservation, and that this need is threatened when we are made aware of the inevitability of our death (so-called ‘mortality salience’). Affiliating with social groups may overcome this mortality threat since such affiliation provides meaning and stability. More broadly, individuals may adopt the group’s cultural worldview, and thus achieve a sense of symbolic immortality (see Castano, Yzerbyt, Paladino & Sacchi, 2002, for a review). Religious beliefs, for example, can ‘buffer’ the adverse effects of mortality salience (see Newheiser et al., 2013). Terror management theory (Greenberg et al., 1990) also argues that individuals hold negative attitudes toward outgroups because they see them as a threat to their ingroup’s worldview ( for a review, see Solomon, Greenberg & Pyszczynski, 1991).
Multiple Categorization
Thus far, we have discussed social identification in relation to a single social cat- egorization. Social identity theory states that how we feel about and act toward others depends on whether we consider them as part of our ingroup. In some situations, the distinction between ingroup and outgroup members rests on a single dimension of categorization. On a football pitch, for example, players tend to think of their own team as “us”, and of the other team as “them”, and try their best to make their own team win. In many situations, however, this distinction rests on multiple social categories. Individuals differ in how they make sense of their mul- tiple group memberships, and often combine and organize their social categories in
complex and differentiated ways. In recent years, researchers have made significant advances in understanding such multiple categorization processes, as well as their consequences for intergroup relations (see Crisp & Hewstone, 2007, for a detailed review). We devote the remainder of this chapter to multiple categorization and its relationship to intergroup attitudes. We review research on crossed categoriza- tion, common ingroup and dual identification processes, ingroup projection, social identity complexity, and social identity inclusiveness and structure.
Crossed categorization
Individuals belong to a range of social groups, some of which cross-cut each other. Imagine, for example, a group of male and female students, half of whom study psy- chology, while the other half studies chemistry. The categories ‘student’ and ‘gender’ are thus cross-cutting each other (if we assume that both male and female students are enrolled in both subjects). The crossed categorization approach assumes that individuals can process information about multiple cross-cutting categorizations, and are able to consider more than one categorization when deciding whether someone is an ingroup or outgroup member (Brown & Turner, 1979; see Brewer, 2000). When two such categorizations are salient, individuals meet people with whom they share both category memberships (double ingroups), share either of the category memberships (mixed groups/partial ingroups), or share neither category membership (double outgroups). Let us assume that, with reference to the previous example, a male psy- chology student is asked to make judgements about his fellow psychology and chemistry students. This individual would perceive other psychology students either as double ingroup members (if they are male) or as mixed ingroup/outgroup members (if they are female). He would perceive chemistry students either as mixed ingroup/outgroup members (if they are male) or as double outgroup members (if they are female). Deschamps and Doise (1978) studied how crossed categor- ization affects evaluations of others. Participants rated targets from each of four combinations of gender and two minimal groups (Experiment 2). In this and similar studies, participants tend to evaluate double ingroups more positively than mixed groups, whom participants evaluate more positively than double outgroups ( Migdal, Hewstone & Mullen, 1998; Urban & Miller, 1998). This corresponds to an additive pattern of evaluation where more shared group memberships lead to more positive evaluations. Researchers have, however, observed other patterns of evaluation (Crisp & Hewstone, 2007). In some situations, a single category dimension dominates target evaluations. That is, all targets that share the dominant category membership are evaluated more positively than all targets that lack this category membership. In other situations, observers may evaluate all targets that share at least one category
European’). Dual identities could overcome some problematic consequences of one-group identities, as reviewed below. The common ingroup identity model has been criticized for its emphasis on recategorization, which risks depriving individuals of valued subgroup identities. This concern is particularly relevant for minority-group members who risk being subsumed in a superordinate category dominated by majority-group members (see Brewer & Gaertner, 2004; Simon, Aufderheide & Kampmeier, 2001). Recat- egorization may thus be an unrealistic means for improving intergroup relations outside the laboratory (Brewer & Gaertner, 2004). Recategorization may further exacerbate hostile intergroup relations by threatening valued identities, especially in intergroup contact where groups differ in size, status, and power (Hewstone et al., 2002). Further, Dovidio, Saguy, Gaertner and Thomas (2012) recognised that focusing on common ingroups can distract from ongoing status hierarchies and thus entrench the status quo. The dual identity model offers a way to reap the benefits of common ingroup identities, while protecting against its problematic consequences. The dual-identity model thus proposes an integrationist view of intergroup relations, while the com- mon ingroup identity model has more in common with an assimilationist approach ( Dovidio, Kawakami & Gaertner, 2000; Dovidio, Gaertner & Saguy, 2009). Ex- perimental research (Ufkes, Calcagno, Glasford & Dovidio, 2016) found that, as predicted, recategorization into a common ingroup identity reduced collective ac- tion intentions among disadvantaged-group members, while recategorization in a dual identity did not. Others have argued that, in some cases, dual identities lead to more positive outgroup attitudes than a superordinate identity alone. This should be the case when the superordinate category is too inclusive to provide individuals with adequate distinctiveness ( Hornsey & Hogg, 2000). Empirical studies have documented positive consequences for intergroup attitudes when both subgroup and superordinate groups are equally salient ( for example, Gaertner, Dovidio & Bachman, 1996).
Ingroup projection model
A problem with both of the preceding models is that individuals tend to expect similarities between themselves and others. That is, they engage in social projection (see Robbins & Krueger, 2005, for a review). Ingroup projection refers to situations in which individuals perceive the defining (often positive) characteristics of the common ingroup as reflecting only those of the former subordinate ingroup ( see Wenzel, Mummendey, Weber & Waldzus, 2003). Ingroup projection may limit the extent to which a common ingroup identity exerts positive effects for intergroup rela- tions. Consider, for example, two subgroups, ‘English’ and ‘Scottish’, nested within a superordinate category, ‘British’. An English person should find it easy to identify as
British if they perceive the category ‘British’ to reflect the characteristics associated with English people. In that case, the subordinate outgroup category ‘Scottish’ is not included in the common ingroup, at least not to the same extent. Identifying as ‘British’ would no longer be sufficient for encouraging positive attitudes towards the subordinate outgroup ‘Scottish’. In majority-minority group relations, ingroup projection may have partic- ularly pernicious effects on intergroup relations when there is no obvious set of features describing the common identity. Drawing on self-categorization theory (Turner et al., 1987), Mummendey and Wenzel (1999) argued that the majority group’s values and attributes are often treated as the default values and attributes of the common identity. What is thought of as an inclusive category might be seen by minority-group members as a projection of the majority group’s values onto them. While ingroup projection might not be problematic for majority-group members, it can represent a symbolic threat (Stephan & Stephan, 2000) to minority-group members. This threat may sow the seeds for intergroup discord (Brown, 2000). Mummendey and Wenzel (1999) propose that the optimal common identity is one that is amorphously defined so as to allow various interpretations of what category membership entails.
Social identity complexity
The multiple categorization approaches discussed above primarily pertain to the categorization of outgroup members in functional relation to the ingroup. For example, in the crossed-categorization approach, the extent to which an outgroup member of one category shares ingroup membership in another category is critical. In the common-ingroup and dual identity models, the extent to which the outgroup is included in a common ingroup is key. Below, we discuss ‘social identity complexity’ as an approach that is explicitly about individuals’ self-categorization, with less reference to comparative outgroups. The social identity complexity approach assumes that individuals have dif- ferent ways of subjectively combining their numerous identities (see Schmid & Hewstone, 2011, for a review). Social identity complexity refers to an individual’s subjective representation of the interrelationships among their multiple identities, where complexity can range on a continuum from low to high ( Roccas & Brewer, 2002). Individuals often perceive their multiple social identities as largely over- lapping. Such individuals regard only people with whom they share all group memberships as ingroup members, and regard all people with whom they share none or only a few group memberships as outgroup members. A person who views their ingroups in this manner espouses a simple identity structure and thus evinces low social identity complexity. A person high in social identity complexity is aware that she belongs to
Ethnicity
Intersection Dominance Merger Nationality
Religion
Ethnicity Nationality
Religion
Ethnicity Nationality
Religion
Low (^) Inclusiveness High
Figure 1. Schematic representations of social identity structures, ordered by social identity inclusiveness ( van Dommelen et al., 2015). Shaded regions represent the groups which a participant has to categorize as “us” to be assigned that structure.
Participants viewed identity cards, each representing a person with whom they shared none, one, two, or all of three group memberships. Participants judged cards as “us” and “not us”, showing whom they included in, and excluded from, their in- group. Participants’ responses reflected various social identity structures that varied in social identity inclusiveness. Figure 1 shows examples of possible social identity structures. A Turkish-Belgian Muslim, for example, may consider only people who share their ethnicity, nationality, and religion as ingroup members (intersection). Someone else may consider all Turks, whatever their nationality or religion, as ingroup members (dominance). Another person may consider anyone who shares their ethnicity or nationality as ingroup members (merger). A person who includes more categories in their ingroup displays a more inclusive identity construal than someone who includes fewer categories in their ingroup. Social identity theory argues that we favour ingroup members over outgroup members in our thoughts, feeling, and actions. Considering more people as “us” should extend ingroup favouritism to former (partial) outgroup members, and should thus reduce prejudice and discrimination (see Gaertner et al., 2016). In line with this prediction, van Dommelen et al. (2015) found more inclusive iden- tities to correlate with more favourable attitudes. Similarly, Reimer et al. (2020) found that South Indians reported more warmth and less social distance for people they had categorized as “us” rather than “not us”. Together, these studies showed that considering social identity inclusiveness and structure helps us to understand how people construct complex social identities from multiple, cross-cutting group memberships. There are, of course, other ways in which social identification can be complex.
Recent research (see, for example, Gaither, 2015; Reddy, 2018) has recognised that people can belong to more than one category in the same category domain (e.g., multi-racial people), that people can hold identities that are not recognised by others, and that people can change how they think about their identities over their lifetime. The social identity literature continues to evolve to address such complexities.
Conclusion
Since their inception, self-categorization theory and social identity theory have been at the heart of social psychological theorizing. They have emphasized—in isolation and in combination—the central role that social identity processes play in a range of social phenomena and, in particular, in intergroup relations. This chapter has provided a comprehensive overview of the so-called social identity perspective, encompassing both theories, and other approaches to social identity related to these theories (for example, optimal distinctiveness theory and subjective uncertainty reduction theory). In so doing, we have paid particular attention to defining self-categorization and social identification, the two concepts on which the social identity perspective rests. We have highlighted that social identification is a multidimensional phenomenon—a complexity that is often overlooked in social psychological research. Researchers interested in social identity phenomena must take this multidimensionality into account, not only in their own research, but also when making comparisons to and drawing conclusions from other research. This chapter has focused on the role of social identity issues in intergroup relations. We have argued that, although social identity theory has significantly advanced our understanding of intergroup relations, self-categorization and social identification are not sufficient to explain negative intergroup attitudes and beha- viour. Any examination of intergroup relations needs to account for the broader social, historical, political or ideological context, and for the interplay between pre- vailing socio-structural conditions (see Tajfel, 1981). We have also drawn attention to an evolving body of research that considers how multiple categorization processes relate to intergroup relations. Although much of social psychological research seeks to understand intergroup phenomena, the underlying motivation of many social psychologists is to find ways of eradicating prejudice, reducing conflict, or improving intergroup relations. Multiple categorization approaches hold particular promise in that respect. Increasing awareness of the multiplicity of social identities often translates into more differentiated perceptions of multiple ingroups, which tend to covary with more positive intergroup attitudes. In sum, this chapter has provided an overview of the theoretical found- ations and key research studies that underpin the social identity perspective in social psychology. By providing a more differentiated account of the social identity
