Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Definition

Theoretical approach to intra- and intergroup life emphasizing the importance of positive group distinctiveness (i.e., favorable comparisons to other groups).

Introduction

Social identity theory (SIT) was developed by Henri Tajfel in the late 1960s and early 1970s. In part, his theoretical approach recognized that many other approaches to phenomena such as intergroup relations relied heavily on notions of intergroup competition (largely for tangible resources, such as oil or land) and on individual differences (e.g., authoritarianism) as explanations for group life. Underlying SIT is a simple yet elegant idea: at any particular moment in time, humans categorize themselves along a continuum ranging from a completely personal and idiosyncratic identity (emphasizing the aspects that make us distinct from others) to a completely social or group identity (emphasizing the common aspects shared with ingroup members). In essence this distinction captures one’s conceptualization or representation of “me” (self) to “us” (collective). The relative salience of one’s identity (self vs. collective) in the moment is a key factor theoretically driving and shaping one’s thinking and behavior. SIT has arguably become one of the most influential and generative social psychological theories of the twentieth century, giving birth to self-categorization theory (SCT), optimal distinctiveness theory, the common ingroup identity model, and intergroup emotions theory, among others (see elaborations below). Many researchers consider SIT critical to a psychological understanding of intergroup life (e.g., prejudice) and intragroup life (e.g., norm adherence).

Theoretical Underpinnings

The central notion underlying SIT is the recognition that we can identify ourselves solely in terms of our idiosyncratic selves (e.g., “I was born on January 1st,” “I have a weakness for chocolate”) and our collective or group selves (e.g., “We Canadians are passionate about hockey”), to any point in between (e.g., Tajfel 1978; Tajfel and Turner 1979). Group identities are theoretically important for providing a sense of not only who we are but how to think and behave (e.g., Abrams et al. 1990). That is, when identifying as a group member (e.g., Canadian, professor, student), the group in question, on its own and/or relative to other groups, offers information about the relevant norms necessary to navigate the social world and be accepted by others.

According to SIT, a central goal in human life involves striving for positive social identities. That is, we value membership in important, prestigious, and relatively distinct groups. But groups, like people, do not exist in a vacuum. Indeed, distinguishing one’s social group from another group or groups plays a key function in our social lives. Although groups value being different from one another, too much similarity between groups (e.g., Canadians and Americans) can threaten group members and motivate the need to actually or psychologically heighten the differentiation between the groups, particularly among those highly identified with their group (see Jetten et al. 2001). As an illustration, Canadians are known to highlight their differences from Americans (e.g., public health-care system, support for gun control, passion for hockey) under the identity threat of being mistaken for American. These processes are recognized as being natural, and even critical, for functioning and coping with social life. However, there are also negative consequences. As noted by Brewer (2003), “Accentuation of category differences, combined with a need for positive distinctiveness (intergroup social comparison), results in ingroup favouritism. The important point [from an SIT perspective] is that ingroup classification precedes rather than derives from interpersonal processes” (p. 36). The emphasis in SIT is clearly about group processes as fundamental and critical to understanding others, both as group members and also as individuals. Some of the prominence afforded to SIT presumably results from the robust finding that most intergroup outcomes, particularly in lab-based settings, involve ingroup favoritism (i.e., giving praise and/or resources to one’s own group) rather than outgroup derogation or hostility (see Brewer 1979). In fact, trends toward ingroup favoritism (vs. outgroup hostility) are becoming increasingly strong as cultural norms continue to shift toward less overt expressions of anti-outgroup prejudice (Greenwald and Pettigrew 2014). Such trends render SIT increasingly relevant to understanding the psychology of group life.

From an SIT perspective, therefore, we can think of ourselves in terms of our collective (vs. personal) identity, and we do so to keep or gain positive social value. This often necessitates differentiation from other groups, even (or especially) from similar ones, in ways that can favor one’s ingroup at the expense of another. Accordingly, SIT proposes that we strive to keep the balance of intergroup life largely in our own group’s favor. If, after engaging in social comparison with another group, the outcome is positive or favorable (i.e., “we” come out favorably relative to “them”), we continue to invest in and bask in our social group’s evaluative positivity. But sometimes comparisons with other groups reveal the lower status, ability, or value in one’s ingroup. At this point, Tajfel argued that group members seek change in the situation. Some of these strategies can be personal (not group) in nature and in fact can involve leaving the group (i.e., social mobility) where possible. That is, permeability of group boundaries is important in determining how unfavorable social comparisons are handled. Two other relevant factors involve the stability (vs. instability) of the intergroup dynamic, such as whether the relations between groups are considered likely to change or not, and the perceived legitimacy (vs. illegitimacy) of the intergroup relation or dynamic. Here stability refers to the likelihood of change in intergroup relations, whereas legitimacy refers to the justness or fairness of the dynamic. Of course, the psychological element, specifically the perception of intergroup relations, is critical. Other reactions to intergroup comparisons that are unfavorable to the ingroup are decidedly psychological in nature, such as redefining the characteristics along which one’s group is judged or the importance/valence of such dimensions. Consider a situation where one group (e.g., Canadians) scores less favorably relative to another group (e.g., Americans) on a particular dimension of evaluation (e.g., military power). Those in the disadvantaged group, here Canadians, can downplay the relevance of military power and/or make salient other dimensions whereby Canadians score positively in intergroup comparisons (e.g., peacekeeping, hockey prowess). Alternatively we can push back against the source of the negative comparison (e.g., anti-outgroup action). If exiting our ingroup is not possible, or comes with great cost, comparisons may shift from intergroup to intragroup, that is, comparing oneself to less well-off ingroup members to restore positive evaluations more central to the self.

Methodological Practices

As a very broad and comprehensive theory, SIT is adaptable to virtually any research design or context. Yet advocates of the SIT framework popularized several methodological practices that have become associated with SIT. First, SIT researchers were keen to study intergroup life in the laboratory, which poses a series of challenges, including the fact that people come into the lab already as group members (racial, sexual, socioeconomic, etc.), complete with idiosyncratic or ingroup histories and complications. Given the strong emphasis placed on experimental research in social psychology, Tajfel and his followers often randomly assigned participants not only to group conditions but to group categories themselves. This random assignment to novel ad hoc groups became known as the minimal group paradigm (MGP). As observed by Rupert Brown (2010, p. 39), the goal was “…to create groups which had been stripped bare of all of the usual concomitants of group life – face-to-face interaction, an internal group structure, a set of norms, relationships with other groups, and so on.” Groups in this paradigm are “minimal” in that a participant simply learns that they are a member of one group and not the other (an exercise in categorization). Sometimes these random group assignments were allegedly based on performance (e.g., over- vs. underestimating dots on a visual screen) or preference (for artwork by Klee vs. Kandinsky), but the assignment can even be blatantly random and nonetheless instill group categorization in participants’ minds (e.g., Abrams et al. 1990, Hodson et al. 2003, Hodson and Sorrentino 2001).

Following such minimal group assignment, participants typically engage in tasks relevant to the research question at hand. For studies on the influence of ingroup (vs. outgroup) norms, for instance, participants have subsequently made judgments about the nature of stimuli or group norms after learning of the perceptions of other people (see Abrams et al. 1990). Most well known, however, are studies that utilized what came to be known as the “Tajfel matrices” (Bourhis et al. 1994). In such studies, participants are tasked with assigning points or money to other people, who are indicated as being a member of one’s group or an outgroup. Importantly, participants are informed that they will not personally benefit from such distributions, to presumably alleviate concerns about self-interest as a potential confound. That is, to the extent that a participant favors their own group over another, this action presumably reflects an intergroup (not interpersonal) dynamic in action. A particularly clever aspect of the Tajfel matrices is that the researcher can systematically study the distribution of resources or other forms of social value as a function of group status and can also pit different strategies for allocating resources against each other to test their relative prevalence. For instance, participants can allocate an equal number of resources between representatives of the groups (i.e., “parity”). Or participants can adopt allocation strategies that award the most points that their ingroup can receive, regardless of the amount given to the outgroup, known as “maximizing ingroup profit.” But the most psychologically interesting allocation strategy is known as “maximum differentiation,” whereby participants choose a distribution pattern that does not give many resources to the ingroup but critically is particularly harsh toward the outgroup. In essence, maximum differentiation gives a “hit” to the ingroup but an even stronger hit to the outgroup, in ways that maximize the relative intergroup differential in the ingroup’s favor. As a methodological advance, the Tajfel matrices offer a nuanced and practical way to examine intergroup dynamics in the lab. The interested reader is referred to the authoritative text by Bourhis et al. (1994) for further and more specific details on the use of these tools.

Theoretical and Empirical Outgrowths from SIT

One of the chief indices of the importance of SIT in the domain of intergroup relations concerns the vast number of important advances that were generated in its wake. The most direct outgrowth took the form of self-categorization theory (SCT; Turner et al. 1987), spearheaded by John Turner, who, with Tajfel, laid out many of the central tenets of SIT (e.g., Tajfel and Turner 1979). Rather than emphasizing the person-versus-collective identity continuum per se, SCT argued that we think of our identities at varying levels of abstraction, ranging from person, to group, to higher levels (nations or even humanity). Relative to SIT, the SCT approach de-emphasized motivational needs (and concerns with self-esteem) and placed priority on cognitive factors, presumably reflecting the social-cognitive revolution of the 1980s. Central here is an emphasis on the group prototype, or mental representation of the group, and in particular how the perceiver assimilates or contrasts a social target (including oneself) to the prototype. Here the notion of depersonalization does not carry a negative connotation but simply recognizes that the idiosyncratic self becomes de-emphasized when social identities are salient. This depersonalization process is perhaps best reflected in the social identity model of deindividuation effects (Postmes and Spears 1998) that is particularly pertinent to explaining crowd behavior (e.g., rioting). From this perspective, people in crowds do not “lose themselves” (as per classical deindividuation accounts) but rather become sensitive to and adherent to the social norms of the collective. That is, people might not “act like themselves” in crowds, but not because they “lose themselves” but because they fit in with a social categorization and follow those local (group) norms. In such contexts, people shift from their personal identity to a collective identity.

Like SIT, SCT emphasizes the degree to which group differentiation is relevant. SCT emphasizes the notions of comparative fit (differences between groups are perceived to be greater than those within groups) and normative fit (group members are perceived to behave in accordance with our expectations). For example, categorization processes are maximized if the differences between lawyers and athletes are perceptually noticeable, and each group is perceived to have characteristics and behaviors consistent with our understanding of those groups (e.g., being argumentative for lawyers, being health conscious for athletes). These processes are considered fluid and very context dependent.

Several other prominent theoretical approaches also grew from the SIT framework, such as Brewer’s (1991) optimal distinctiveness theory. This approach recognizes that we have competing needs for inclusion (within our social groups/identities), but also needs to differentiate ourselves from others. As such, people seek to define themselves in ways that can meet both needs, in order to avoid feeling over-included (or over-assimilated) by the social identity or over-differentiated (or excluded) at the personal level. The main thrust of this position, therefore, is that the psychological space between the personal and social selves best satisfies our needs. Such middle-ground identities, therefore, are considered the most valuable and invoke the most identification among group members. Brewer’s model thus builds on, and pushes off from, the basic tenets of SIT.

Several theoretical approaches devoted to resolving intergroup tensions build on several of the central tenets of SIT (and SCT). In the interest of brevity, here we discuss the common ingroup identity model (Gaertner and Dovidio 2000), given its prominence in the literature. This approach recognizes the centrality of categorization and social identification in group life. In fact, categorization is seen as relatively inevitable, but the nature or process of categorization is considered malleable. In essence, the common identity approach argues that there are significant benefits to be gained from extending the ingroup boundaries to include members of the outgroup. This process of recategorization redraws group boundaries, capitalizing on our tendency to prefer our ingroups by expanding the metaphorical tent and in turn expanding the scope of those that benefit from this favorability. In reference to SCT, the common ingroup model encourages group members to categorize at a higher, more inclusive, more abstract level. In a clear example of the benefits of recategorization, Dovidio and colleagues (2004, Study 2) exposed White participants to news reports on the threats to Americans (or to White Americans) by an outside terrorist agent (Al-Qaeda). Participants exposed to information stressing threats to all Americans (relative to other conditions) were subsequently more inclusive in their categorizations. Specifically, Blacks were more included into the White Americans’ ingroup. Critically, this recategorization also lowered prejudice toward Blacks. These methodologies are easily adaptable to electronic platforms. For instance, Canadians playing violent video games with an American “teammate” (vs. alone) demonstrated significantly more positive American attitudes after the game, with much of the effect explained by the recategorization of group boundaries (Adachi et al. 2016). Such conceptualization and research effectively capitalize on the tendency to show pro-ingroup attitudes by redrawing the boundaries to include (those formerly considered to be) members of the outgroup.

Another theoretical model that builds on SIT is known as the intergroup emotions theory (Devos et al. 2002). Drawing on SIT, this approach recognizes that people routinely think of not only their personal identity but of their social identity. One’s thinking about the self and ingroup become relatively intertwined. Novel to the approach by Mackie, Smith, and colleagues is the integration of cognitive appraisal approaches into the SIT framework. Instead of examining how a situation affects one as a person and considering the ensuing interpretation and emotional response, the intergroup emotions approach suggests that people also ask how a situation affects their ingroup (and how they feel and act accordingly). For instance, when an outgroup violates ingroup norms (e.g., concerning sex or eating), this can induce disgust reactions, which in turn instigates avoidance of that outgroup. Alternatively, a cognitive appraisal that an outgroup is threatening or powerful leads to fear or anger, which can instigate action tendencies relevant to escaping the situation (if afraid) or attacking the outgroup (if angry). A key contribution of this theory is this incorporation of emotions to the SIT domain, which did not formally propose such emotions in a systematic manner.

Nuances and Conceptual Clarifications

One of the outstanding or unsettled issues underlying SIT concerns the reasons why people invest so heavily in their groups. Proponents of SIT have long argued that self-esteem is critical. Specifically, SIT proposes that the act of discriminating against an outgroup raises one’s self-esteem and also that those low in self-esteem engage in discrimination in order to improve their situation. That is, self-esteem is both a cause and a consequence of discrimination. A comprehensive review of the literature (Rubin and Hewstone 1998) concluded that there is not full or clear support for the self-esteem hypotheses of SIT, with the authors concluding that much of the confusion concerns issues of measurement. In particular, most studies have examined personal self-esteem to the general neglect of social state self-esteem, how one presently feels about one’s social group(s). Relative to personal self-esteem, social esteem is conceptually more relevant to intergroup life. Rubin and Hewstone speculate that the inconsistent findings in the literature probably explain why the self-esteem hypothesis (or hypotheses) has been de-emphasized in more recent SIT and SCT research. Indeed, social identity theorists increasingly emphasize motivations other than self-esteem. For instance, Hogg (2014) argues that people identify with their groups, in large part, in order to provide clarity and reduce uncertainty (i.e., an epistemic rather than an esteem motive). In support of such claims, research demonstrates that certainty-oriented individuals (i.e., those who gravitate toward certainty and what is already known) are particularly likely to discriminate against outgroups when the context is characterized by uncertainty (Hodson and Sorrentino 2001). Clearly more research is needed to fully understand the motivations underlying social identity processes.

Another issue of debate concerns whether or not ingroup identification itself is a necessary predictor of bias toward other groups. Some have argued that identification predicts bias and that SIT has long stipulated this relation (e.g., Gagnon and Bourhis 1996). But others suggest that the actual relation is weak or nonexistent (Hinkle and Brown 1990), and still others have argued that SIT never posited that identification plays this role (e.g., Turner and Reynolds 2001). (Incidentally, Turner and Reynolds also argue that SIT never posited a self-esteem hypothesis, highlighting the lack of agreement, even among proponents of SIT, about the core underlying features of the theory.) These disagreements about the meaning, intentions, and underlying mechanisms of SIT offer fertile ground for future researchers.

Some of the confusion underlying the SIT theory might involve its disregard of individual differences (e.g., Turner and Reynolds 2001). That is, many proponents have argued that person-based factors play little role in explaining intergroup life (see Hodson and Dhont 2015, Table 1). Yet, as argued elsewhere, Tajfel posited a continuum between personal self and collective identity, not a sharp demarcation (e.g., Hodson et al. 2013). Moreover, Tajfel (1978, p. 402) explicitly recognized the importance of individual differences in social life:

If I take part in a race riot or strenuously and vocally support my local team during a football match, or demonstrate against the oppression of my group by another group, I interact with another group – as Sherif wrote – fully in terms of my group identification. However, the truism that marked individual differences will persist even in these situations is still valid. [emphasis added]

Presumably his intention was not to ignore or negate person-based factors (see Hodson et al. 2013, Hodson and Dhont 2015) but rather to (rightly) emphasize the situational factors. Many contemporary theorists call for more integration between person-based and situation-based accounts (Hodson 2009; Hodson and Dhont 2015). For instance, arguments that the identification-discrimination link will be moderated by outside factors (e.g., Turner and Reynolds 2001) would be strengthened by including (rather than excluding) individual differences as potential moderators.

Conclusions

SIT emerged at a critical juncture of research on intergroup relations, urging researchers to consider the more social (vs. personal) aspects of prejudice and discrimination. Moreover, SIT calls on researchers to consider processes of categorization and identification not only as normal but as fundamental to group life. For these reasons, SIT has left an indelible mark on the field that is difficult to ignore or overstate. Few theoretical approaches can successfully claim to be so encompassing, so generative of later prominent theories, nor as relevant to contemporary thinking about social and intergroup life. The underlying principle that social or group identities are relevant to perceiving and thinking about the social world is both appealing and valuable in explaining our social nature.

Cross-References