Cognitive schemas, abstract and organized packages of information, are the cognitive version of identities. Self-schemas include organized knowledge about one's self, the cognitive response to the questions of identity. These include the characteristics, preferences, goals, and behavior patterns we associate with ourselves. Group schemas
(analogous to stereotypes) include organized information about social positions and stratification statuses, such as gender, race, age, or class. Because the social positions we occupy have immediate consequences for our sense of self, group schemas play a major part in processes of identification. Self and group schemas illustrate both advantages and disadvantages of categorisation systems. They allow us to summarize and reduce information to key elements; thus, they also entail losing potentially valuable information. And, categorisations are almost always accompanied by systems of evaluation of some categories as better or worse. Schemas are not just perceptual phenomena; they can serve as explanatory devices and justifications of social relationships (Tajfel 1981). Thus, social identities are embedded in sociopolitical contexts.

Social identity theory focuses on the extent to which individuals identify themselves in terms of group memberships (Tajfel & Turner,
1986). The central tenet of social identity theory is that individuals define their identities along two dimensions: social, defined by membership in various social groups; and personal, the idiosyncratic attributes that distinguish an individual from others. Social and personal identities are thought to lie at opposite ends of a continuum, becoming more or less salient depending on the context.
Deaux (1993), however, argues for an interplay between the two, suggesting they are not easily separable. Social identities provide status and enhance (or not) self-esteem. Because people are motivated to evaluate themselves positively, they tend to evaluate positively those groups to which they belong and to discriminate against groups they perceive to pose a threat to their social identity.

Empirical support has relied heavily on studies using the minimal group paradigm (Tajfel 1970), whereby people are classified into distinct groups on the basis of an arbitrary and trivial criterion under conditions free from other factors usually associated with group memberships. Under these minimal conditions, people do discriminate in favor of in-groups in allocation of various rewards. The most sociologically relevant recent studies have extended this tradition to socially meaningful groups and situations. Simon et al (1997), for example, demonstrate that being in a numerical minority (a predictor of identification in this tradition) does not lead to identification unless the in-group-out-group categorisation is situationally meaningful. The more positive, and more personally important, aspects of the self are likely to be bases on which a person locates her- or himself in terms of collective categories (Simon & Hastedt 1999), demonstrating the relationship between categorization and evaluation. This points toward more successful attainment of a positive social identity for those in dominant social groups. This process is a challenge for members of stigmatised, negatively valued groups, who may attempt to dissociate themselves, to evaluate the distinguishing dimensions of in-groups as less negative, to rate their in-group as more favorable on other dimensions, or to compete directly with the out-group to produce changes in the status of the groups. Much of this research accords considerable agency, both cognitive and material, to social actors. One relevant line of research explores the psychological consequences of identifications with ethnic in- and out-groups. Fordham & Ogbu
(1986), for example, suggest that academic failure among
African-American students represents a desire to maintain their racial identity and solidarity with their own culture. High-achieving
African-American children develop a "raceless" persona, but at the cost of interpersonal conflict and ambivalence; adoption of "raceless" behaviors and attitudes do have negative psychological consequences for African-American students (Arroyo & Zigler 1995). Direct impression management strategies intended to counter negative evaluations of their in-group also increase, one of many

indicators of the interdependence of cognition and interaction. The focus on psychological consequences of identification speaks also to the interconnectedness of cognition and emotion. Thus, for example, individuals' prejudices may shape not only their own identifications but also their categorisations of others. Racially preducice individuals do not appear to be more motivated to make accurate racial categorsations both in-group and out-group, than do non-prejudiced individuals (Blascovich et al 1997); accurate categorisations maintain clear boundaries between groups.

Strong identification with a group need not, in principle, be correlated with out-group hostility. Only under conditions of intergroup threat and competition are in-group identification and out-group discrimination correlated (Branscombe &Wann 1994, Grant &
Brown 1995). Social identity theory maintains that it is ingroup identification that causes out-group bias. Realistic conflict theory
(LeVine & Campbell 1972), on the other hand, maintains that out-group threat and hostility lead to in-group identification. In a study of
Black South Africans' ethnic identifications before and after South
Africa's transitional election in 1994, Black African identification was related only to attitudes toward Afrikaans Whites, not whites in general or English Whites (Duckitt & Mphuthing 1998). Longitudinal analyses suggest that attitudes affected identifications, more consistent with realistic conflict than social identity theory, a useful caution to overly cognitive approaches to identification.