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.

This edition of the Handbook of Social Psychology, like its 2003 predecessor, is organized by theory and levels of analysis, rather than by substantive themes. The earlier volume did not include chapters on specific structural or cultural systems. For this volume, in contrast, we have been asked to write a chapter on gender and race and to address how these systems are theorized by and influence social psychology. We begin by discussing this specific inclusion of structural/cultural systems, and then ask: why gender, and why race?

Although social psychology is typically defined as the study of the relationship between individuals and society, how “society” is treated varies considerably across theories, methods, and empirical applications. Sociological social psychologists tend to be sensitive to the power of social structures, institutions, and organizations. They (we) attend to the dynamics of key social structures and their historical and sociopolitical contexts, and how these shape interpersonal and intra-individual experiences.

Social categorization is core to understanding a social psychology of social structures. The process of social categorization is a principle of social cognition. Human cognitive capacities are limited; we often need to streamline information to manage the demands of everyday interaction. Categorization is one mechanism for organizing, saving, and retrieving information. There are vulnerabilities, however. Categorization is a reduction of information; potentially valuable information is lost. Equally significant, categorization seems always to be accompanied by differential evaluation. Some categories are evaluated as better, others as worse. Gender and race are two central systems of social categorization. Both systems are associated with powerful mechanisms for the allocation of resources, both material and symbolic, and therefore both are core to understanding contemporary societal inequalities.

One could well imagine both greater attention to each system (e.g., separate chapters) as well as more attention to other prevailing systems of social categorization in a Handbook such as this. There is an enormous amount of social psychological research on both gender and race. The scope of this work has mirrored the significance of these systems in society more generally, with considerable growth of work on gender generated by the feminist movement and considerable growth of work on race generated by the civil rights movement. That said, there are also major differences in how each research tradition has fared over the past 60 years since the civil rights movement began; we trace those differences in a subsequent section of the chapter. Treating both systems together in a single chapter highlights the intersections between them, reflecting that societal inequalities do not exist in separate vacuums but almost always work together, sometimes to deepen, sometimes to lessen, inequalities.

Other key systems of social categorization have not received as much social psychological attention. To the extent that theory and research (and space) allow, we include those systems in this chapter. We also speculate about why there is differential attention to gender, especially, with somewhat less on race and dramatically less on socioeconomic systems, which arguably are just as influential to social psychological and sociological dynamics and inequalities. Age, nationality, religion, physical and mental abilities, sexuality, all are used as bases for categorization and the allocation of resources. Importantly, our review reflects a western/northern hegemony of social psychology. Other cultures have developed other principles for categorization and differential allocation of resources. The caste system in India and the hukou system in China are both guided by similar principles, but use different criteria for categorization. We do not address all of these, but rather highlight the social psychological principles and processes that explicate these patterns and hope that the reader herself will pursue these lines of research. Importantly, because categorization may be unavoidable, it is critical to explore how its principles could be harnessed to undermine the parallel processes of differential evaluation and differential allocation of resources.

The chapter is organized in terms of definitions, theories, methodologies, and future directions. We offer specific definitions of the key concepts: gender, race, and intersectionality. The definitions of these terms may seem self-evident, but a closer look suggests otherwise. We also take up other significant social structural and/or cultural systems that organize inequalities. We then turn to an overview of four approaches that have organized extant approaches to gender and to race; this overview is organized historically, as approaches have emerged, dominated, and, in some cases, waned. In the second major section we review how three key social psychological theories – social exchange, social cognition, and symbolic interaction – have addressed gender, race, and intersectionality. We highlight where there is overlap and where there are incompatibilities. In the third section we address key methodological themes: what methodological techniques have been prominent in social psychology generally and how these methods have been used to address dynamics of gender and race. We explore also how reliance on these methods may advance and/or constrain the kinds of questions about gender and race that can or cannot be asked and explored. In the final section, we locate the research on gender and race in current and possible future trends shaping the field of social psychology. Here we address the continuing schism between sociological and psychological social psychologies, and the rise of public sociology and positive psychology. We also highlight a few of the themes throughout the chapter, specifically, cultural and demographic shifts in these systems, the continuing destabilization of these very categories of gender and race, and technological advances that facilitate new methodological and theoretical approaches to gender and race.

Defining Concepts

One important theme in defining gender, race, and other related constructs is the distinction between biological and social components. In the early years of social psychology, the term “sex” was often used to refer to a broad domain that included both biological and social components. In contemporary times, it is conventional to distinguish between these. The history of theorizing about biological and social aspects of race is quite different. Attribution of race to biological factors, once taken for granted, has been complicated greatly. Sociologists and social psychologists have paid much closer attention to the social factors that have created prevailing systems of race. However, developments in genome science have led to discovery of some genetic correlates of race (see Owens & King, 1999; Schwartz, 2001). The contemporary challenge for social and natural scientists is to identify the intersections and mutual influences among these social and biological factors.

Defining Gender

Sex

Sex typically refers to biological characteristics that distinguish females and males: chromosomes, reproductive characteristics, physiological features. For much of human history, sex has been assumed to be both dichotomous and unchangeable, a conviction that has the characteristic of an “incorrigible proposition” (Mehan & Wood, 1975). In interactional practice, individuals assess the sex of others indirectly, through observation of the size and shapes of their bodies, how they move, what they wear, the tone of their voices, e.g. through self-presentations. The sex to which we assign someone is, then, a sex category.

In recent history there has been considerable challenge both to the dichotomous nature of sex and to the immutability of sex. A meta-analysis of the medical literature from 1955 to the new millennium estimates that in as many as 2 % of live births the child is born with reproductive organs, chromosomal structures, and/or secondary sex characteristics that are not uniformly female or uniformly male (Blackless et al., 2000). In earlier days such individuals were typically treated socially as one consistent sex, with discrepancies ignored, hidden, or altered, as possible. Today there is more widespread recognition of the existence of such discrepancies, e.g., the treatment of Caster Semenya by the IAAF, the governing body of track and field competition (Clarey, 2010; Yaniv, 2009). (Semenya is a South African middle-distance runner. After she won the 2009 World Championship in the 800 m with a dramatic performance improvement, the IAAF asked her to undergo a medical examination to determine her biological sex and initially suspended her eligibility for international competition.)

Intersexuality (the presence in one person of physical features that are typical of both female and male) challenges the dichotomy of sex; transsexuality (identification with a gender inconsistent with one’s assigned sex) challenges the immutability of sex. Transsexuals change their sex, either physiologically through surgical and hormonal techniques, and/or socially, through altering their self-presentations. Sex change suggests that sex is mutable. There are variations in this conviction among transsexual communities, however. In that transsexuals have the genetic structure of one sex and the self-presentation (and sometimes physiological) appearance of another, they challenge the assumed dichotomy of two distinct sexes.

Gender

Gender is typically defined as the social and cultural behaviors and characteristics associated with, but not determined by, biological sex. The most comprehensive approach is that gender is itself a social structure, a social system that creates the two categories associated with sex and the differential allocation of resources typically associated with gender. Regardless of whether one endorses this strong view, it is certainly the case that gender is learned, performed and institutionalized, as we detail below.

There are other gender-related concepts of particular relevance to social psychologists: gender identity, gender stereotype, and gender role. Gender identity refers to one’s internal sense of gender, e.g., to one’s self as female, male, or some combination, or neither (in contrast to the terms typically used to categorize sex: woman, man). Gender identity is not necessarily congruent with assigned sex, as our discussion of transsexuality indicates. These incongruities typically cause sufficient anguish that transsexuals change their gender performance and sometimes also their physiological characteristics. The anguish, of course, is an indication of the core role that gender plays in interactional smoothness. Were gender not so core, there would be less reason for anguish (Garfinkel, 1967; although see also Connell, 2009).

Gender stereotypes, like all stereotypes, are generalizations about individuals in particular social categories, in this case, female and male. These generalizations provide the information necessary for satisfactory gender performance. Gender stereotypes are the beliefs about normative female and male behavior, dress, speech style, interactional style, and societal positions.

Gender roles, although often treated as distinct in the literature on gender, are really a subset of gender stereotypes. The concept of role derives from a functionalist view of society in which social behavior functions most effectively when different actors perform different, non-overlapping and presumably complementary roles. Gender roles include behavioral expectations for females and males. There is a substantial history of critiques of the concept of gender role (see Stacey & Thorne, 1985 for a particularly incisive and well-known critique); the core point is that the concept ignores power and inequality. Gender roles characterize the role expectations of women and of men as neutral and complementary, but in reality they are differentiated profoundly by power. To illustrate with an iconic example, the mid-twentieth century marital roles of husband and wife were, according to Parsons (Parsons & Bales, 1955), integral to maintaining a smoothly functioning social order. From the perspective of the twenty-first century, those husbands and wives were not equally empowered; rather, marital roles served to maintain patriarchal structures. For our purposes, we can view gender roles as the behavioral correlate of cognitive gender stereotypes, and extend the critique of roles to the concept of stereotypes. Gender stereotypes also support gender inequality through legitimizing distinctions that reverberate throughout societal institutions.

With each of these concepts, it is important to note that there are always situated, contextual factors that qualify normative expectations of gender and sex. Gender identity, for example, is not necessarily experienced by everyone. Bem’s (1974) early work on androgyny, stresses that in addition to stereotypic patterns whereby men are masculine and not feminine, and women are feminine and not masculine, individuals may be high on both masculinity and femininity, e.g., androgynous, or they may lack a sense of either masculinity or femininity. Gender itself may not be an important part of a given individual’s sense of self. The salience of gender identity also varies markedly across social contexts. Same gender vs. mixed gender groups affect the salience of gender identity (Fiske & Taylor, 2008; Turner & Brown, 2007). Group identity is typically more salient for those in any subordinate social category, a point relevant to both gender and racial identities (Hollander, Renfrow, & Howard, 2011). Similarly, gender stereotypes vary by subtypes within gender, such as those defined by race (Devine & Baker, 1991; Green & Manzi, 2002), class (Lott & Saxon, 2002), or sexuality (Geiger, Harwood, & Hummert, 2006). These variations do not typically lead to modifications in the belief in a dichotomy of gender.

Sexuality

Sexuality is distinct from gender, but the cultural association between gender and sex means that people often infer sexuality from sex category and the gender assumed to accompany sex category. Sexual orientation, the preferred sex of sexual partner, is a component of sexuality, but sexuality is by no means reducible to sexual orientation. Sexuality refers to sexual behavior, eroticism, sexual orientation, and one’s overall inclination to engage in sexual activity (Hollander et al., 2011). Gender stereotypes include expectations for gendered expressions of sexuality, as they do for other forms of gendered behavior. There are cultural variations in these expectations but as with gender, there is an underlying dichotomization to gendered expectations about sexual behavior.

Sexual orientation, although just one component of sexuality, is a critical, indeed some might say, core, component. Like gender, sexual orientation is stereotyped as dichotomous: preference for opposite-sex or same-sex partners. But like gender, sexual orientation can be considerably more varied and fluid than a dichotomy would suggest (Rosenthal, Sylva, Safron, & Bailey, 2011; Rutter & Schwartz, 2011). Individuals may be bisexual, attracted to and/or sexually active with same-sex and opposite-sex partners (Rust, 2000) or asexual, not attracted to and/or sexually active with others (Bogaert, 2012). Moreover, Diamond’s (2008) study of love, desire, and identity among women finds considerable intra-individual fluidity in each construct over the life course. However, sexual orientation is heavily freighted with assumptions of normality, with opposite-sex preferences being normative. Historical context matters: in contemporary times there has been a pronounced shift toward general acceptance of same-sex partnering (at least at the level of civil rights), particularly in the western/northern world (Saad, 2007). Sexual orientation is also highly gendered. Stereotypes about sexuality, both opposite-sex and same-sex, are deeply imbricated with gender. Heterosexuality entails stereotypic femininity for women, stereotypic masculinity for men. Following the logic of dichotomy, same-sex female (or male) couples are often assumed to include one masculine-type partner and one feminine-type partner.

Questions of the etiology of sexual orientation are not the purpose of this chapter. The key point for social psychologists is that the meaning of sexual orientation is socially constructed (see below). Same-sex practices, and their prevalence and social meaning, vary considerably across cultures.

Defining Race and Ethnicity

Race is another key dimension of social categorization, stereotypically based on visible physical characteristics, primarily skin color, body size (weight, height, build), and other physical features. Race is often discussed together with ethnicity; the latter is a less physiologically based concept, referring to people who share a common culture and background. Race is often assumed to be a biological characteristic inherited through biological reproduction. Although there are genetic variations by racial categories, race is largely a social construction. Mayr (2002), among others, has demonstrated that there is more genetic variation within races than between them. Moreover, the physical characteristics that are associated with particular races are often more difficult to identify in practice than in theory. Some Whites have darker skin than some labeled as Black or Chicano; some Asians have lighter skin than many Whites. Nonetheless, as with gender, the science does not necessarily affect the social conception of the system.

Racial classification, unlike gender, is not dichotomous. Historically, three racial groups were identified: Caucasian, Negroid, and Oriental. The very way these terms resonate for a contemporary reader says a great deal about the politicization of racial classification. In the history of the United States, there has been and still is a particular resonance to the Black/White distinction. Members of other racial categories have commented on a degree of invisibility of other racial groups and a tendency of U.S. historians to blur these distinct histories into this particular dichotomy (Fernandez, 2007).

Unlike gender, which has been a primary categorization scheme for centuries, race only became a significant social category with the rise of the scientific revolution, presumably through the spread of colonial exploration and conquest (Marks, 1995). Also unlike gender, what groups of people have been defined by racial distinctions has varied considerably across history. At one time in U.S. history, Irish immigrants were considered to be Black. Individuals from a variety of Asian countries have been defined variously as not White, or Oriental (Espiritu, 2008). Hispanic continues to exist as a sort of hybrid classification, with Census responses for both racial classification and Hispanicity (Kilty & Vidal de Haymes, 2004). Indeed, tracing the racial and sometimes ethnic categories on the Census throughout time is itself a testament to the social construction of race (Omi & Winant, 1986). (Census categories do not necessarily represent popular conceptions of race or ethnicity, but they may be a reasonable indicator, at a temporal lag.)

Native American history highlights another aspect of racial categorization; members of this general category often think of their memberships in terms of nationality, rather than race or ethnicity. The Census treats Native Americans as a racial category, but other units of the U.S. government recognize the national sovereignty of many native groups, demonstrating the structural inconsistencies that can and do exist. Structural inconsistencies in definitions of racial categories, and temporal and historical changes in these definitions, suggest that racial classification, like gender classification, is very real socially and politically.

Paralleling the social construction of gender, there are identifiable racial stereotypes, conceptions that themselves are historically specific. Racial stereotypes have been assessed in a rare longitudinal manner through the use of scales developed originally by Katz and Braly (1933), then administered again in 1951 and 1967, enabling an examination of historical changes in racial stereotypes (Karlins, Coffman, & Walters, 1969). Also paralleling gender, there is variation in the strength of racial identities. Rockquemore and Brunsma (2007) report considerable variation in Black/White biracial identities. Their typology ranges from exclusively White or Black, to a border identity between the two categories, to a protean identity that is contextually sensitive, and a transcendent identity that is, in some sense, raceless.

Politicization of Social Categories: Sexism, Racism, Ethnocentrism

We address in this section the ways in which different social psychological theories incorporate (or not) gender and race into their frameworks. For at least most of these theories, there is some degree of recognition of political bias often associated with categories within these systems, that is, various forms of derogation directed toward members of subordinate categories – women, people of color, disenfranchised ethnicities. Sexism, racism, ethnocentrism are social psychological phenomena, enacted through cognitive, interactional, and structural processes. There has also been some backlash, with members of majority categories sometimes asserting a reverse sexism (toward men), reverse racism (toward Whites), and/or reverse ethnocentrism (toward, say, Anglos). It is certainly the case that any human being learns biases as a part of the cultures and societies in which we are raised and live. Women can be sexist; people of color can be racist. Indeed, one might argue that any member of contemporary U.S. society is sexist, racist, and ethnocentric, also able-ist, age-ist, and so forth, because it would be virtually impossible not to be. Social power relations structure how these biases are expressed and what influences they may or may not have. But individuals have agency; they – we – can actively become conscious of and work against such prejudicial stereotypes and identities (Devine, 1989)

It is a social psychological truism that names and language matter. As Richards (1997) notes, because language provides the terms in which we understand the social world, language is also a key arena in which attempts to alter those understandings are fought. The terms used to describe the social systems we focus on in this chapter have at times been the focus of intense struggles. We attempt to use the terms preferred by the people and groups to which they refer, where that is ascertainable. It is important to note, however, that there is often considerable slippage. “White” is a misnomer, but “Caucasian” is no better. “Black” has different connotations in the U.S. than in, for example, Brazil. “Straight” meant something quite different in the 1950s than it does now. We simply ask the reader to be aware of the insufficiencies of language.

Intersections of Gender and Race

In labeling the two previous sections “Defining gender” and “Defining race and ethnicity,” we do violence to the lived reality of these systems. Neither gender nor race can be adequately understood without analyzing their intersections with each other and with other social positions. Thus much sociological scholarship argues that various forms of stratification need to be conceptualized as matrices of domination (Collins, 1990) or “complex inequality” (McCall, 2005). We use the term “intersectional” (Crenshaw, 1991) as shorthand for this complexity. (See special issue of Gender & Society, 2012 for extended discussion of intersectionality.) Not all social psychological theories focus explicitly on power and/or inequality; therefore, when we turn to specific theories, we see varying degrees of emphasis on gender and race, and also on intersectionality.

“Intersectionality” carries several important connotations. The term can imply the importance of including the perspectives of multiply-marginalized people; a shift from the addition of multiple independent strands of inequality toward a multiplicative analytic, e.g., shifting from a focus on main effects to a focus on interactions; and modeling multiple institutions as overlapping in their production of inequalities (Choo & Ferree, 2010). Empirically, research with this degree of complexity can be extremely challenging. As McCall (2005) notes, multi-group studies must analyze the intersections of the full set of dimensions of multiple categories; if we take a relatively straightforward profile of two categories of gender, three categories of race, three categories of socioeconomic position, we already have 18 different configurations. We discuss these methodological challenges below.

In addition to race and gender, other particularly significant status systems that have been incorporated to some extent in social psychological research include socioeconomic status (social class), sexual orientation, (dis)ability, age, national origin, and religion. Socioeconomic status is particularly significant, but in some ways it is also particularly difficult as a research concept. There can be considerable mismatch between objective and subjective definitions (Bettie, 2003). Many studies indicate that far more people identify themselves as middle class than economic data would suggest (Pressman, 2007; for a real-world example, see Achen, 2011). A further complication is that socioeconomic position is an almost “unspeakable” concept in the U.S. (Hollander et al., 2011). Class variation clearly exists, but talk about class is not normative. This may be related to a distinction between ascribed and achieved social statuses. Race and gender are both traditionally considered to be ascribed statuses, consistent with the view that both are fixed and permanent. The individualism characteristic of the U.S. and other western societies leads to the conviction that social class is an achieved characteristic; that is, any individual can achieve a socioeconomic position to which s/he aspires, through hard work and their own skills. Yet, as is perhaps painfully evident in the larger economic profile of the recession era of the 2008 period forward, socioeconomic standing cannot meaningfully be considered a purely achieved status. Current and historical family circumstances, structural factors, historical societal circumstances, all affect socioeconomic possibilities. Likely for all of these reasons, there is considerably less social psychological research on social class than on gender and race, and less work on the intersections among gender, race, and class, than on intersections between gender and race.

We offer a few examples here to illustrate the value of an intersectional approach. We begin with some examples of intersections in cognition. In defining gender stereotypes, we noted that subtypes exist within any given set of stereotypes. Stereotypes about women and men vary by race, class, sexuality, and various indicators of social locations. Landrine (1985) found that middle-class White students rated White women as more dependent, emotional, and passive than Black women, and rated poorer women as more dirty, hostile, inconsiderate, and irresponsible than wealthier women. Note that both the social characteristics of the observed and of the observer are critical. It is much more common for researchers to document the characteristics of the observed, than of those whose perceptions are measured.

Gender and race intersections may have greater significance for those in subordinate categories. Settles’ (2006) mixed-method study explored Black women’s gender and racial identities and reported that the combined “Black woman” identity is more important to women of color than either “woman” or “Black” alone. Context is critical as well. Bell (2002) considered identities of Black women across different types of institutions – White coed, Black coed, and Black women’s colleges – finding that their gender identities were stronger at Black coed schools and their combined race and gender identities were stronger at White coed schools. The identity or combination of identities that separate one from the dominant group become more central to one’s sense of self in that context.

Other examples of intersections of gender and race derive from interactional perspectives. Pate (2006) shows that “acknowledgment rituals” such as the nod of a head to strangers in public are much more frequent among African American men than among African American women or members of other racial groups. Pate suggests that this pattern reflects the unspoken need for support among this markedly visible racial and gender group, a disproportionate target of racial discrimination.

Conceptions of Gender and Race in Social Psychology

There are historical variations in the recognition of the importance of gender and race. Research on gender, although present to some degree in the early years of social psychology, began to increase markedly in the 1970s and has remained robust to the present day. During these years researchers began to attend to gender-based inequalities, not only to gender difference. The growth in social psychological research on gender coincided with the active feminist movement of the 1970s; it seems reasonable to assume at least some effect of the broader political environment on academic pursuits. Moreover, more women began to earn PhDs – currently one-third of the doctorates in Sociology (Spalter-Roth & Scelza, 2008) and two-thirds of the doctorates in Psychology (Keita, Cameron, & Burrwell, 2006) are earned by women.

Social psychological research on race also intensified during the 1970s, again presumably at least in part due to the civil rights movement. This activity has declined since then, continuing with emphases primarily on attitudes, stereotypes, and other cognitive structures and processes. The marked difference (from gender) is in the demographic profile of the disciplines. In contrast to the ever increasing presence of women scholars in sociology and psychology (now a majority in both disciplines), only a very small proportion (lower than overall population proportions) of the membership in either discipline includes people of color. Moreover, over the past several decades there has been a marked decline in the proportion of published articles in which African Americans are research participants. There is not a one-to-one relationship between the gender and race of the scholar and the focus of their research, but clearly these factors do influence choices of scholarly topics.

In this section we articulate a more comprehensive historical analysis of social psychological theorizing about gender and race, and other social positions where possible. We focus on four distinct theoretical orientations, approaches that mirror historical periods to at least some extent. These four orientations can be viewed as different points on a continuum that ranges from material to structural, in some sense from objective to subjective.

Essentialism

Essentialist approaches are rooted in biologism. Biology is thought to determine the social behaviors of males and females; gender is seen as something literally essential to females and males. Thus essentialism treats sex and gender as equivalent. This approach was prevalent within the social sciences more generally in the first half of the twentieth century (and earlier). Perhaps the best known social psychological manifestation of essentialist approaches to gender is the M-F Test, a tool that measured masculinity and femininity (Terman & Miles, 1936). Test items were based on the assumption that masculinity and femininity are polar opposites and that sex and gender were effectively one and the same: being male meant being masculine, being female meant being feminine, and one individual could not be both. Women and men whose scores did not support these assumptions were often viewed as having homosexual inclinations; thus sexuality too was seen as isomorphic with sex and gender.

This stark form of essentialism sounds dated, but indicators of essentialist thinking are present in contemporary work (and culture). Many studies even today use sex as a variable representing gender. Participants are categorized as females or males, but any differences that result are either labeled or attributed to gender differences. “Sex as variable” research remains common in sociology, indeed throughout the social sciences. This practice ignores critical distinctions between the two concepts. As we have noted, sex category is, for most people, constant and dichotomous. Gender, however, is expressed quite variably and depends greatly on context. Moreover there is considerable overlap between women and men on many characteristics (this is true for sex as well as for gender).

Research on race has also been driven by essentialist assumptions until relatively recently. Although it is more normative to reject essentialist beliefs about race than about gender, Herrnstein and Murray’s The Bell Curve (1994) argued relatively recently that racial differences in intelligence exist and are genetically determined. Although there were numerous articulate critiques of their arguments (see Devlin, Fienberg, Resnick, & Roeder, 1997), The Bell Curve seemed to strike a cultural chord. In the first several months after its release, 400,000 copies of the book were sold, and several thousand reviews and commentaries were written within months after its publication. Indeed, in response to the controversy the volume generated, the American Psychological Association’s Board of Scientific Affairs established a special task force to prepare an investigative report on the research.

Socialization

Socialization is a core social psychological perspective. According to this perspective, individuals learn behavior from their environments through a variety of processes, from modeling and imitation (Bandura, 1977), to the application of rewards and punishments (Cahill, 1986; Thorne, 1995), and/or through intra-psychic processes (Chodorow, 1978). Through socialization, individuals internalize various prescriptions and proscriptions associated with particular characteristics. Gender socialization is one of the most heavily studied of these forms of socialization.

What elements are learned during gender socialization? Gender stereotypes are more than description; they are also prescriptive, hence the attributions of abnormality when individuals do not conform. Gender stereotypes, whether of self or of others, include expectations about personality, bodies (see Chap. 7), occupations, and other role-related behaviors (Deaux & LaFrance, 1998). Whereas Terman and Miles’ M-F Test was an example of essentialist conceptions of personality, Sandra Bem’s Sex Role Inventory (1974) and work on androgyny (Bem, 1975, 1993) illustrate socialization of gender through their assessment of patterns of femininity and masculinity. Her scales treat masculinity and femininity as distinct dimensions, allowing for the possibility of opposition, absence of both, or presence of both (androgyny) in a given individual. The concept of androgyny was in favor for some time, but eventually fell into disuse as paying insufficient attention to power and as reproducing the very gender polarization it sought to move beyond (Pyke, 1985). More recent work has characterized these dimensions not by gender labels but rather by substance, as instrumentality and expressiveness (Deaux & LaFrance, 1998). There is substantial overlap in the distributions of each dimension within most women and men.

The prescriptive implications of gender stereotypes are instantiated in gender roles. The concept of gender roles reflects a profoundly functionalist view of society, as we have discussed above. Despite the apparent contemporary rejection of such a model, the normative power of gender is still strong. Moreover, gender prescriptions legitimize a division between public and private spheres that entrenches women’s subordination and men’s domination (Kimmel, 2006).

Socialization processes and the concept of roles do not work so well for analyzing race and class. Stacey and Thorne (1985) point to this discrepancy, noting the absence of concepts such as “class role” or “race role.” Such concepts would imply that the social order depends on the allocation of social roles to members of different racial or socioeconomic categories, an implication few would accept. The concept of gender role does not sit as uncomfortably, which says a great deal about the underlying resilience of the functionalist argument (and of sexism).

This does not mean there are no tinges of socialization in analyses of race and class. For example, the notion of the culture of poverty, a theory that there is a set of values among the poor that tends to perpetuate their poverty (Lewis, 1969), is certainly present to some degree; this implicates both race and class. But, there has been considerable debate about this concept (Goode & Eames, 1996). Where socialization does seem relevant to race and to class is in the learning of racial and class prejudice by dominant group members, and acculturation to the experience of subordination by racial and socioeconomic minorities. Again, a telling contrast between literatures on race and on gender is that there is little work on how men learn to discriminate against women, and how women come to live with that. Thus, socialization, rather than systematic and structured inequalities, remains the primary explanation of gender difference. Things may be changing. One analysis of introductory sociology textbooks shows that discussions of gender increasingly emphasize both macro and meso level structures and processes, in addition to socialization (Manza & Van Schnydel, 2000). Ferree and Hall (2000) are skeptical, however; they maintain that the apparent increase in sensitivity to structural factors is overstated by Manza and Van Schnydel’s conceptualization and measurement. In summary, gender is analyzed primarily in terms of socialization and difference rather than inequality, and race (and class) are analyzed more in terms of systematic and structured inequalities.

In stressing that gender is learned, not innate, socialization goes beyond essentialism. But in some ways the distinction is weak. In that gender (less obviously, race and class) differences are such a major part of what is learned, the distinction between learned and innate characteristics is almost semantic. If gender socialization is so powerful that gendered behaviors remain throughout the lifespan, then in some sense gender is still conceived as unchangeable, once learned. In moving to two other social psychological perspectives, we see more attention to situational and cross-cultural variations in gender.

Social Constructionist Approaches

Other approaches to understanding human behavior view social realities as created through human action and interpretations of those actions. Constructionism assumes that there is no inherent meaning in objects and behaviors, but rather that human beings create meaning. Social order would not be possible unless there were agreed upon meanings, so a key purpose of the rituals of social interaction is the negotiation of common interpretations of the situations in which we act.

The quintessential social constructionist analysis of gender is Garfinkel’s (1967) classic case study of a transsexual named Agnes. Agnes appeared male at birth, but developed female secondary sex characteristics at puberty and subsequently had sex reassignment surgery. Garfinkel used Agnes’ situation to illustrate the interactive performance of gender. Because Agnes transitioned from male to female, she went to great lengths to appear and behave in ways consistent with female gender stereotypes. Because she did not become female until her adolescent years, she had to be more intentional and self-conscious about gendering her performance than do those identified as female at birth. On the basis of his observations of her behavior, Garfinkel identified some core assumptions that define the performance of gender. The key themes are that gender is dichotomous, invariant, “natural,” and that everyone has a gender. Agnes’ performance, of course, suggests that these assumptions, while normative, are not accurate, but that gender is actually accomplished through everyday behavior. That is, gender is socially constructed.

Kessler and McKenna (1978) go one step further and argue that the idea of two distinct sexes is itself a social construction. As we noted in defining sex and gender, intersexuality and, to some extent, transsexuality, do undermine this assumption of dichotomous sex. To make conceptual room for these identities and experiences, Fausto-Sterling (2000) has proposed a revised classification scheme that includes five sexes (i.e., males, females, true hermaphrodites, female pseudohermaphrodites, and male pseudohermaphrodites). Nevertheless, the cultural resilience of the dichotomous construction of sex means that we categorize intersexuals as “really” one sex or the other. Transsexuals who engage in sex reassignment surgery in some sense reaffirm this construction, altering sex to conform to gender. This would not be necessary if the socially constructed sex/gender system allowed a lack of correspondence between genitals and gender.

There has been a great deal of social psychological research on the performance of gender, conceived by West and Zimmerman (1987) as “doing gender”. Although socially performed, the ways in which gender is “done” are strongly constrained by normative definitions. Gender is such a central dimension of social identity that failures or breaches in this performance can be used to discredit other aspects of an individual’s identity. Goffman’s (1977) work on public bathrooms illustrates this point. Bathrooms, at least in the U.S., are segregated by sex. One could imagine that bathroom segregation occurs because of biological reasons, but the fact that private homes do not segregate bathrooms demonstrates that the segregation occurs for social, not biological, reasons. When individuals enter the “wrong” bathroom, they are breaching gendered norms. This does happen, most often when there is a long line at one category of bathrooms and not in the other. The breach can be repaired if the individual acts embarrassed or in other ways communicates that s/he understands the anti-normative character of the behavior. More serious, however, is when someone whose sex is not clear is identified as being of the wrong sex for the bathroom. This is the basis for advocacy for private, un-sex-identified bathrooms at public institutions (Molotch & Noren, 2010).

West and Fenstermaker (1995a) published an update to the “Doing Gender” article titled “Doing Difference.” Their goal in this article was to extend their analysis of the performativity of gender to other key systems of power with which gender intersects. In other words, they reconceptualize differences of race and class, as well as gender, as ongoing interactional accomplishments. We offer two examples that demonstrate “doing difference”; these examples also underscore the intersectionality of the “Obama moment.” President Obama’s specific configuration of cross-cutting race, gender, and social class locations enables the creation of unique, emergent identities, as well as reflects experiences of both oppression and privilege. Consider two recent events: first, public reaction to a photograph of Arizona Governor Jan Brewer pointing her finger at President Obama when she met him at the airport during his visit to her state (Hennessey, 2012). The event was a heated exchange between political opponents with divergent views on ethnic studies in public schools. It was not taken that way by the public, however, but rather as a racist confrontation: a White woman trying to put a Black man in his place. The race-sex positions of Obama and Brewer allowed observers to find meaning in this image through associations stored in our nation’s collective memory of countless acts of violence directed at men of color. Pundits framed this as unprecedented public behavior toward a sitting president; Governor Brewer framed it as a misunderstanding. Still others framed it as proof that blatant racism exists in the twenty-first century. By contrast, in a second heavily publicized situation, Obama’s location as a man of relatively “modest” means (compared to other politicians), as well as his heterosexuality and masculinity, provided him with advantage. When White police officer Sgt. James Crowley arrested African American scholar Henry Louis Gates, Jr. and ignited a nationwide controversy about claims of racial profiling, Obama was able to reduce racial tensions by inviting both men to the White House to talk it out over beers (Nicholas, 2009). “Doing gender” provided the common ground needed to eclipse racial differences. Although one status may be emphasized over others in these situations, neither situation can be reduced to the workings of only race, of only gender, or of any single social location. Understanding the social meaning of each of these performances requires an intersectional lens.

A subsequent symposium of responses to the “doing difference” article was incisive in critiquing the authors for a perceived insensitivity to social power, history, and structure (see Collins et al., 1995). The authors’ response (West & Fenstermaker, 1995b) offers points that are useful to consider in evaluating a social constructionist approach to intersections of inequalities. They point to the distinction between process and outcome. Social construction focuses on the processes whereby inequalities are, literally, done. Social structures help us identify the outcomes of unequal systems. They highlight as well the accountability to which social actors are held, accountabilities defined by these social structures. Their summary identifies one of the formidable challenges for sociological social psychologists: “The challenge we face – theoretically and empirically – is to describe a system that manifests great interactional variation but, at the same time, rests on far more stable structural and historical legacies… the impact of the forces of social structure and history is realized in the unfolding of those relationships as the sites for the doing of difference…” (Fenstermaker & West, 2002, p. 98).

Social constructionism helps to explain why, despite empirical findings that women and men are more similar than different on most characteristics, they seem to behave systematically differently. There are many examples of social psychological research that demonstrate the self-fulfilling nature of gender expectations. The stereotype that women are naturally more nurturing caregivers than men may produce men who do not feel they are capable fathers (Coltrane, 2004). A second cost of these stereotypes is that agentic and confident women working outside of the home may be viewed as self-interested and unfeminine, and therefore penalized in hiring and promotion decisions (Ridgeway, 1987; Rudman & Glick, 2001). Moreover, individuals often align their performances with what they believe others expect of them. Snyder and Skrypnek (1981) found that men and women applying for a job they believed to be either masculine or feminine emphasized personality traits that confirmed their suitability for that specific job regardless of their sex role identity.

Other studies document how racialized beliefs shape the social construction of race, that is, the self-fulfilling nature of racialized expectations. Harlow’s (2003) study of Black professors finds that even though they are reluctant to point to racial difficulties in the classroom, they often encounter students who view them as less credible than White professors and who are likely to perceive mistakes and to attribute these perceived mistakes to incompetence. Conversely, race becomes an important resource affording these professors more credibility than White professors when (minority) race is central to course material. Together, the discrediting nature of blackness and the responsibility of representing an entire race forces these professors to engage in considerable emotional and physical labor (e.g., suppressing anger, setting extremely high achievement goals) to manage the racial double standard in the classroom.

The expectations held by members of a majority group shape their own behavior, in turn eliciting confirming behavior from members of minority groups. These processes can occur even without the actual application of majority expectations. Behavioral confirmation can occur simply through people’s awareness of expectations others hold for them. Spencer, Steele, and Quinn (1999) show that the awareness that others hold a negative stereotype about one’s group, can, in and of itself, interfere with behavioral performance. They also demonstrate that experimental manipulations that contradict the stereotypes can reduce this effect of stereotype threat. Cognitive and interactional strategies combine to create a social psychological system in which preconceptions about selves and others are performed and entrenched.

While the social constructionist approach is often employed by sociologists to study macro-level phenomena, social psychologists emphasize interaction and the social contexts in which behavior occurs. For example, in her study of young women at Waretown High in California’s Central Valley, Bettie (2003) explores the ways that gender/race/class identity projects among the smokers, cholas, “las chicas,” skaters, hicks, and preps are both performance and performative. Conceptualizing identity projects as performance, Bettie blends the notion that individuals “do difference” (West & Fenstermaker, 1995a, 1995b) with Butler’s (1999) observation that identity projects are constituted by the social and cultural structures in which individuals live and by the social scripts available to them (see also Wilkins, 2008). The expectations of those with whom one interacts, differential opportunities for interaction, and differential consequences of interaction are all important elements in the effectiveness of social construction. At the same time, there are broader aspects of society that constrain or enable the possibilities for individual actors, which brings us to our fourth approach. In focusing on gender, race, socioeconomic position, and others systems of stratification, we explicitly highlight the power of social structure.

Structural Approaches

Gender, race, social class, are all bases for the macro-level allocation of material and social resources and opportunities. At the same time, social structures are always implemented, negotiated, and sometimes redefined through individual action and interaction (Schwalbe et al., 2000). Our task as social psychologists is to consider how social structures shape the cognitive patterns and the micro-level interactional possibilities of social actors, and, in turn, how cognitive and interactional processes can ultimately alter social structures.

Indeed, it is this approach to social psychology that most distinguishes sociological social psychology from psychological social psychology. Considering gender and race forces attention to structure. If we take the processes of socialization, for example, why is it that certain behaviors are rewarded when performed by men and punished (or are at least less effective), when performed by women? If we turn to social construction, do Black actors have the same degree of behavioral choice and control as do White actors? A structural approach turns our attention to social power, to the effects of the pervasive systems of male dominance, White dominance, and other such interlocking systems. Focusing on structure and power can lead to a sense that social change is not possible. The beauty of social psychology is that it recognizes individual agency. Men, Whites, have structural advantages whether or not they desire or consciously enact them. But, as agentic individuals, they can behave self-consciously in ways that undermine those advantages and create advantages for those who do not have the same degree of access. Members of disadvantaged groups can and do also act to contradict stereotypes. As one example, Staples (1994), an African American man, reports a simple strategy he implements when he is walking after dark and encounters women in a public space: he whistles well-known tunes of classical music. This simple strategy resists the association of black men with danger (see section “Resisting social expectations” below). When social groups as a whole act collectively, social structural change is indeed possible. In this important sense, social constructionist and structural approaches are complementary and interdependent. Social structures could not persist without interactional patterns that maintain them. Those interactional patterns occur within structural constraints and possibilities. Explicit recognition of both construction and structure undermines popular assumptions that gender, or race, or class, is natural (Martin, 2004).

Social Psychological Theories

Within the sociological tradition, three primary perspectives offer varying approaches to understanding the relationship between individuals and their societal environments: social exchange theory, social cognition, and symbolic interaction. Each emphasizes a particular set of assumptions about social actors, social interaction, and society, and offers a distinct framework for analyzing complex social situations. We limit this brief review of these major social psychological theories to their treatment of gender, race, and where possible, their intersections with each other and other systems of categorization.

Social Exchange Theory

Social exchange theory is the most widely used of the three perspectives within both sociology and social psychology, possibly because its more formal analytic approach leads it to be viewed as more rigorous or scientific. This theory applies economic models to everyday decision making. For this reason, it is the theory most compatible with individualist societies. Early theorizing by Homans (1961) outlined the fundamental processes of social behavior, guided by the belief that everything that emerges in groups originates in the behavior of individuals. Blau’s (1964) work departs from this assumption, arguing instead that social exchange should be concerned with group properties. Contemporary social exchange theorists attempt to specify the conditions under which exchange takes place. Thus the exchange relationship, conceptualized as a series of transactions between two or more actors, is the unit of analysis (Emerson, 1972a; Thibaut & Kelley, 1959).

Social exchange theory assumes that individuals pursue rewards and attempt to avoid punishments, and considers an individual’s past history of rewards and punishments to be important for predicting future behavior. Exchange theory postulates that interaction takes place when it is mutually rewarding to the parties involved. The resources transferred during social exchanges may be material (e.g., money, property, etc.) or social (e.g., social approval, respect, etc.) (Blau, 1964). Social actors decide whether the exchange in question is a “good deal” or whether they can find a better deal with a different exchange partner. Based on their perceptions of available information, assessments of the costs and benefits of possible alternatives, and estimates of the probability of receiving anticipated outcomes, individuals choose the alternative they believe will be best.

According to exchange theorists, interaction occurs because of individuals’ interdependence; each person depends on others for valued resources. An important distinction between social exchange and the other social psychological theories we discuss in this chapter is its attention to power as a central mechanism of social interaction. Following Emerson (1962), power is conceptualized as an actor’s ability to achieve a favorable outcome at the expense of another and is a quality of a relationship, rather than of an individual. One’s power over another person depends on the value of one’s resources to the exchange partner, together with the availability of alternative suppliers of this resource. Thus power and dependence vary together in exchange relationships. The person who is least dependent on a relationship has the greatest power in it, because that person can more easily abandon the relationship (Homans, 1974). Power is not a static quality of a person, but a dynamic, shifting property of a relationship. Social exchange theory’s explicit attention to issues of power should equip the perspective to explore societal inequalities associated with race and gender. Unfortunately this potential has not materialized.

Most research by social exchange theorists does not attend to gender, let alone race, social class, or other statuses. Studies that do include sex as an independent variable (e.g., through equal inclusion of male and female research participants) do little to move the social exchange tradition beyond “sex differences” and toward a more sophisticated conceptualization of gender. Consequently available studies offer few insights into the origins or persistence of societal inequalities. Three areas of exchange research, however, have paid more attention to gender and race than most, as we discuss in the next section.

Justice and the Distribution of Resources

The devastation wrought by natural disasters, such as Hurricane Irene and the 2011 tsunami in Japan, highlights the difficulties associated with allocating limited resources in high need situations. Resources can be distributed in a number of ways, not all of which may be seen as fair. Fairness is a subjective concept; at its root are perceptions of justice, which are strongly influenced by social norms. To be seen as just, the actual distribution of rewards (or punishments) should match individuals’ expectations about what they deserve and what they are entitled to receive. Space does not allow for a comprehensive review of the justice literature. We limit our discussion here to how gender, race, and other statuses impact preferences for resource distribution.

Eckhoff (1974) identifies five “distribution rules” that are commonly used to allocate resources among individuals or groups. First, resources can be divided equally among individuals. Second, resources can be allocated to individuals on the basis of need, such that those who have the greatest need receive the most resources. Third, an assessment can be made of the contributions each person has made to the group, and resources allocated relative to those contributions, according to the principle of equity. Fourth, resources can be allocated according to status, such that those with higher status (e.g., men, economically privileged, etc.) receive a greater share of the scarce resources. Finally, procedures (such as drawing straws or participating in a contest) can be established such that each person has an equal opportunity to receive a needed resource, although the final outcomes may not in fact be equal.

Which of these rules will apply in a given situation? Many studies investigate this question using experimental paradigms that ask participants to allocate a reward (usually money) between themselves and another person based on their performance on some task. These studies suggest that individuals’ positions in various stratification hierarchies impact their preference for particular distribution rules (Deutsch, 1975; Mikula, 1980). People in advantaged or powerful positions are likely to perceive an unequal distribution as just, while those in disadvantaged positions are more likely to feel it is unjust (Cook & Hegtvedt, 1986; Molm, 2006; Stolte, 1987). Studies conducted outside a laboratory setting similarly find that people in lower social classes are less likely to perceive inequality to be fair (e.g., Robinson & Bell, 1978) and more likely to prefer an economic distribution based on need than are people in upper classes (Form & Hanson, 1985; Kluegel & Smith, 1986). People in lower socioeconomic groups are also less likely to believe that people in general, and themselves in particular, have a fair opportunity to get ahead (Kluegel & Smith, 1986). Kluegel and Smith report that Blacks are far more likely than Whites to express doubts about the fairness of the American stratification system. They also found similar, although smaller, differences between women and men: women, on average, believed the economic system was less fair than did men. Moreover, sex, race, and class interact to produce variation in such beliefs. Wealthier Whites (and, wealthier Blacks to some degree) expressed more faith in the economic system than did poorer Whites and Blacks. Black men perceived more opportunities for Blacks and the poor than did Black women. Despite these differences, faith in the system of economic distribution remained surprisingly high throughout the 1980s. Kluegel and Smith conclude that “most Americans believe that economic inequality is just in principle, and correspondingly, endorse individual and societal equity as just criteria for the distribution of income” (p. 141).

More recently Page and Jacobs (2009), Newman and Jacobs (2010), and McCall and Percheski (2010) have all documented Americans’ increasing concern about rising levels of economic inequality. Over the last 20 years, half to two-thirds of Americans were dissatisfied with levels of income inequality or the rationale that economic prosperity among the most wealthy drives economic prosperity for the whole (McCall & Percheski, 2010). This dissatisfaction is illustrated by the 2011–2012 Occupy Wall Street movement, whose website defines it as “the 99 % that will no longer tolerate the greed and corruption of the 1 %” and who participates in protests aimed at empowering “real people to create real change from the bottom up (Occupy Wall Street, 2012).”

Social Exchange in Intimate Relationships

Intimate relationships, like less personal relationships, are arenas for the exercise of power and control. Early studies conceptualized marriage as an exchange in which women traded household labor for financial support and men traded money and status for a comfortable home life. Researchers noted that husbands tended to have more power (i.e., influence over decision making) than wives, especially when wives were not employed outside the home (Blood & Wolfe, 1960; Turk & Bell, 1972). Later work by England and Farkas (1986) refined the equation of money with power, arguing instead that money is a key determinant of power not because it is a more valuable resource but because it is a transferable resource that can be taken from one situation to another. Given that men’s resources (i.e., money) can more easily transfer to another relationship than can women’s resources (i.e., bearing and looking after the husband’s children), husbands are assumed to have more marital power and to be less invested in the relationship (see Lewis & Spanier, 1979). Due to the division of labor and the continued wage gap, men and women tend to have different types of resources, and consequently, marital power is fundamentally gendered. Numerous studies document how this power translates into other real-world inequities, such as the character of leisure time (Bittman & Wajcman, 2000) or sleep differentials among husbands and wives (Maume, Sebastian, & Bardo, 2010).

Despite women’s increased presence in the formal, paid labor market, numerous studies report that women continue to do the bulk of household chores (Bergen, 1991; Coltrane, 2000; Kessler-Harris, 2003). This difference has been found to occur across racial groups (Cunningham, 2007; Hossain & Roopnarine, 1994) and classes (Tichenor, 2005; Wright, Shire, Hwang, Dolan, & Baxter, 1992) in the U.S., as well as cross-culturally (Fuwa, 2004; Knudson & Wærness, 2008; Sanchez, 1993). Brines (1993, 1994) finds that when the husband is the primary earner and his wife is dependent on him for financial support, housework is distributed according to the predictions of exchange theory (i.e., women do the majority of the housework and men do very little). However, when the division of labor is nontraditional, with the woman acting as the primary bread-winner, the man actually does less housework on average. This finding runs counter to exchange theory predictions. Brines (1994) argues that marriage is an institution through which women and men can display masculinity and femininity. In traditional marriages both women and men are able to fulfill gender expectations. When women out-earn men, the situation becomes more complicated. These couples solve the dilemma by distributing housework even more markedly along traditional lines, rather than having the man compensate by doing most of the housework. This allows both men and women to maintain their adherence to gendered norms, despite their nontraditional financial arrangements. In the language of exchange theory, symbolic exchange trumps material exchange.

More recent studies comparing the allocation of household chores in gay and lesbian couples with that of heterosexual married couples finds that the three types of couples have very different patterns of housework allocation (Kurdek,1993, 2004, 2005). Heterosexual couples tend to follow a pattern of segregation (e.g., each partner does separate tasks, with one partner doing the bulk of the household labor). Gay male couples tend to follow a pattern of balance (e.g., each partner specializes in an equal number of specific and non-overlapping tasks). Lesbians follow a pattern of equality (e.g., partners share tasks either by doing them together or by alternating responsibility). Taken together, these studies of housework allocation highlight how individuals’ gender performances and the gendered nature of some intimate relationships alter processes of social exchange.

Status and Power

Expectation states theory, a distinct offshoot of social exchange theory, incorporates social exchange’s conceptualization of power with insights from symbolic interaction, a perspective we discuss below. Focusing on conditions in which individuals share both collective (i.e., cooperative) and task orientations, expectation states theory holds that individuals form performance expectations, implicit and often unconscious anticipations for their own and others’ abilities to execute the focal task, by assessing observable status characteristics and making comparisons among group members (Berger, Conner, & Fişek, 1974; Berger, Fişek, Norman, & Zelditch, 1977; Correll & Ridgeway, 2003). According to this theory, these expectations act as self-fulfilling prophecies, guiding behavior in two important ways: individuals act consistent with their expectations for self, and individuals allow or deny others the opportunity to act based upon relative expectations for their ability. For example, studies document that students of color often withdraw and fail to engage within the classroom, while White students dominate intellectual tasks – patterns that reflect racially based expectations (Cohen, 1982). Higher expectations ensure individuals more action opportunities. Accordingly these individuals are more likely to offer suggestions (i.e., performance outputs), and other group members are more likely to evaluate these suggestions positively (i.e., agreements and disagreements) and accept them in reaching consensus (i.e., influence and deference) (Moore, 1985; Ridgeway & Walker, 1995).

Expectation states theory argues that individuals look to status characteristics, or characteristics on which individuals hold differently evaluated states, to evaluate self and others’ potential performances. Theorists differentiate between types of status characteristics: task-specific status characteristics emphasize correspondence between an individual’s traits and her expected performance at a specific task (e.g., a feminine person should perform well on a feminine task), while diffuse status characteristics emphasize expectations for an individual’s general competency across tasks – even those not directly connected to the status (e.g., even nonsense tasks such as “contrast sensitivity” or “meaning insight” used in laboratory experiments) (Ridgeway & Erikson, 2000; Troyer, 2003; Wagner & Berger, 1993). One goal of current expectation states scholarship is to identify the characteristics that act as diffuse statuses and the conditions under which they do so.

Considerable empirical work centers on sex category. Lockhead’s (1985) meta-analysis suggests sex category acts as a diffuse status, with men generally displaying higher levels of influence compared to women. Using experimental data from mixed-sex problem-solving groups involving masculine, gender neutral, or feminine skills, Balkwell and Berger (1996) find that: males exhibit more status power activity (e.g., time speaking, chin thrusts, looking while speaking, etc.) than females while completing a neutral task. Males exhibit even higher levels of status power activity while completing a masculine task; while the feminine task reverses the sex difference, the gap is less than the males’ advantage on the masculine task. In other words, the process is not symmetric, as males retain some diffuse advantage of gender as worth/power. Researchers interpret these effects as signs of a gendered double standard and the power of diffuse statuses (Balkwell & Berger, 1996; Foschi, 1992). A wealth of studies documents the effects of other diffuse statuses including race (Cohen, 1982), occupation (Webster, Hysom, & Fullmer, 1998), ethnic accent (Foddy & Riches, 2000), sexual orientation (Childers, 2000; Renfrow, 2005), and physical attractiveness (Webster & Driskell, 1983). Taken together, this research clarifies how micro-interactions facilitate the performance of societal inequalities.

To varying degrees, these three streams of research have incorporated gender and race. So why has the majority of social exchange research neglected these social statuses? To answer this question, we must consider both the perspective’s key assumptions and researchers’ methodological choices. Social exchange theory assumes that all behavior is based on the same exchange principles. Regardless of gender, race, or class, every individual is believed to be guided by rewards and punishments, power and dependence. Although structural position and its effects on the availability of resources is a central part of the theory outlined by Emerson (1972a, 1972b), in practice structure tends to be operationalized narrowly, as a set of relationships among a small number of individual actors. The connection to real-world social structures based on gender, race, or social class is rarely explored.

Moreover, England and Kilbourne (1990) problematize the approach’s assumptions that actors are selfish and pursue only what will benefit themselves, and that rationality, rather than emotion or empathy, guides behavior. Proponents, however, challenge these criticisms, arguing that they result from a misreading of exchange theory, and that in fact such theories can help to explain gender inequality by showing how women and men make choices within different sets of social constraints (Friedman & Diem, 1993).

The experimental paradigm that has become normative in social exchange research is also partly responsible for the neglect of gender and race. When subjects never see or hear each other, the everyday social cues associated with sex, gender, race, social class, and other statuses are artificially removed from the situation, and their effects on exchange behavior go unanalyzed. Other social psychological perspectives, such as the cognitive approach we discuss next, have been much more successful in integrating these statuses into the experimental paradigm.

Social Cognition

Social cognition theorizes the ways in which we think about our social worlds. The perspective’s basic assumption is that thought influences both feeling and behavior. As we noted in the introduction, human thinkers have limited cognitive capabilities and must often act as “cognitive misers” (Fiske & Taylor, 1991). Needs for cognitive speed and efficiency streamline the process through which we take in and synthesize information from social situations. Because it is impossible to incorporate all incoming information in a given situation, we develop systems of categorization that separate information into distinct categories. Categorization directs our attention as we take in information, aids in the storage of new information, and supports the retrieval of information housed in memory.

Cognitive Structures

Information about gender and race must be represented in some mental form. Social cognition emphasizes verbal representations, which provide the basis for our cognitive structures. Research on cognitive structures explores the content of and shifts in gender and racial attitudes, as well as the organization of knowledge about each into social schemas, including stereotypes.

Attitudes have been conceptualized as psychological tendencies “expressed by evaluating a particular entity with some degree of favor or disfavor” (Eagly & Chaiken, 1993, p. 1). Early cognitive models characterized humans as “consistency seekers,” motivated to find or create consistency among their attitudes, feelings, and behaviors (Festinger, 1957). Research results have never fully supported the assumption that knowing someone’s attitudes aids our understanding and prediction of behavior. Kraus’ (1995) meta-analysis of 88 attitude-behavior studies, however, finds significant associations between attitudes and behavior – particularly when studies use self-report measures of behavior, nonstudent samples, and attitudinal and behavioral measures that are equally specific. Current research continues to qualify the relationship between attitudes and behavior (Fishbein & Ajzen, 2010). Improved methodologies may allow social psychologists to better predict behavior today than in the past, but the relationship between attitudes and behavior remains complex (Maio, Olson, Bernard, & Luke, 2003).

Research on gender and racial attitudes has often focused on the attitudes of individuals in positions of power toward those with less power. As Eagly and Mladinic (1989) note, studies of gender attitudes tend to examine respondents’ preferences for gender role accommodation (e.g., Spence and Helmreich’s “Attitudes toward women scale” 1972). Studies of racial attitudes, in contrast, have emphasized general evaluations, often anchoring them to real-world conflicts, as opposed to role-related assessments (e.g., Persson & Musher-Eizenman, 2005). Considerably less scholarly attention has been directed to attitudes toward men or minority attitudes toward Whites (for exceptions, see Eagly & Mladinic, 1989; Stephan et al., 2002); a small but growing body of literature investigates intergroup attitudes among peoples of color (Gay, 2006).

Until recently, studies of the gender attitudes held by power-advantaged groups toward those with less power reported trends toward liberalization and egalitarianism. Cotter, Hermsen, and Vanneman’s (2011) analysis of the most recent General Social Survey data reports only small changes since the mid-1990s, indicating that the liberalization of gender attitudes has stalled. This pattern holds across cohorts and for men and women of all ethnicities, education levels, and incomes. Cotter and colleagues attribute this stall to increased “egalitarian essentialism,” which they describe as endorsement of both feminist equality principles and traditional motherhood roles.

Recent studies of Whites’ attitudes toward Blacks indicate that the norms of egalitarianism have increased with race as well, and that racial prejudice and negative stereotyping of Blacks has declined (Parks & Hughey, 2011). Bobo (2001) summarizes this pattern: “The single clearest trend in studies of racial attitudes has involved a steady and sweeping movement toward general endorsement of the principles of racial equality and integration” (p. 269). Policy studies, however, indicate that this attitude change has not been accompanied by endorsement for policies that would ensure equal opportunity, citing racial resentment as an enduring obstacle to social change (Tuch & Hughes, 1996).

Contemporary societal norms may make the acknowledgement or expression of biased attitudes unlikely (Schuman, Steeh, Bobo, & Krysan, 1997). Sears (1988) argues that symbolic racism, the tendency to indirectly discriminate towards persons of color through implicit assumptions that members of this group do not uphold American values such as the Protestant work ethic, has replaced blatant “old-fashioned” racism, but nevertheless continues to support the racial status quo. Symbolic racism operates at both the individual and institutional levels (Bonilla-Silva, 2006). Similarly, Gaertner and Dovidio (2005) have reported evidence of an aversive racism, in which individuals balance their preference for egalitarianism with the desire to avoid contact with members of another race. Denis (2011) takes an opposite tack, demonstrating that interpersonal contact between Whites and indigenous peoples, even under conditions that should optimize the reduction of prejudice (Pettigrew & Tropp, 2006) does not necessarily eliminate Whites’ sense of superior group position. Denis shows that subtyping (Lamont & Molnar, 2002), ideology-based homophily (McPherson, Smith-Lovin, & Cook, 2001), and political avoidance norms (Eliasoph, 1999) all lead Whites to maintain prejudice and indigenous peoples to avoid challenging the systematic racism that shapes interpersonal relations in these communities.

Research documents a parallel shift in gender attitudes. Swim and Cohen (1997) propose that a modern sexism that is hidden or unintentional now exists in our culture along with the visible, intended, and unambiguous sexist attitudes and behaviors of the past. Glick and Fiske (1996) distinguish between hostile sexism, resentment toward women who transgress traditional gender roles, and benevolent sexism, the paternalistic and ostensibly chivalrous attitudes that seek to “look out for” women’s interests (e.g., Day, 2001). These new conceptualizations of racism and sexism indicate that social psychological assessments must advance beyond the overt measures of the past in order to detect these subtle attitudes.

New computer technologies allow social psychologists to record individuals’ near instantaneous associations with target words or images to assess implicit attitudes, or unconscious attitudes (Nosek, Greenwald, & Banaji, 2007; and see Chap. 11, “Values, attitudes, and ideologies”). This builds on cognitive models that assume that much of our cognitive functioning is automatic and occurs outside of our awareness (Correll, Judd, Park, & Wittenbrink, 2010; Langer, 1989). Studies consistently find evidence of implicit sexist/racist attitudes among individuals who do not believe they are prejudiced (Wittenbrink & Schwarz, 2007). In fact, research finds evidence of such associations where we might otherwise expect to find preference for one’s ingroup. Dunham, Baron, and Banaji (2008), for example, report that Hispanic and Black children do not display an explicit preference for their ingroup, in comparisons with White children. Instead, these children of color display implicit associations favoring Whites over their ingroup (i.e., associating White faces rather than ingroup faces with positive words). Similarly, Jost, Banaji, and Nosek (2004) observe this pattern even among adults: a surprising number of Black adult respondents display White favoritism rather than preference for their ingroup. Social psychologists have consistently documented implicit associations regarding gender (Rudman & Kilianski, 2000), sexual orientation (Banse, Seise, & Zerbes, 2001; Geer & Robertson, 2005), age (Hummert, Garstka, Greenwald, Mellot, & O’Brien, 2002), and other characteristics.

Implicit associations appear to be influenced by contextual factors. Studies have shown that racial bias can be reduced in conditions that emphasize another characteristic, such as sex (Mitchell, Nosek, & Banaji, 2003, and see President Obama example above) or foreign accent (Kinzler, Duponx, & Spelke, 2007). Dasgupta and Greenwald (2001) find that showing respondents photographs of popular Black celebrities before assessing implicit associations resulted in less bias than when individuals did not see such photographs. Furthermore, implicit preference for Whites appears to decrease in the presence of a Black experimenter (Lowery, Hardin, & Sinclair, 2001). These studies suggest that while individuals may not be aware of their implicit preferences, such associations can be interrupted. These studies also document the considerable power of social context.

The concept of implicit associations has faced some skepticism. Tetlock and Mitchell (2008) and Arkes and Tetlock (2004) offer both methodological and conceptual critiques. These scholars suggest that implicit bias studies may actually assess individuals’ awareness of cultural stereotypes rather than their “endorsement” of these beliefs, and they propose that seemingly prejudiced results may in fact reflect processes of statistical discrimination associated with respondents’ awareness of group differences. In short, they are suspicious of methodologies that nearly always find evidence of a hidden prejudice among ostensibly open-minded college students (see Banaji, Nosek, & Greenwald, 2004; Quillian, 2008; and Sears, 2004 for responses to these critiques). Real racial and/or gender differences in societally meaningful variables can and do exist; we argue that it behooves social scientists to consider the historical conditions that have given rise to them.

Attitudes emphasize the content of thought and evaluations of that content; social schemas, in contrast, emphasize the organization of social information. Schemas reduce knowledge through processes of categorization, allowing individuals to make sense of specific cases by treating them as a part of a general category. Schemas act as theories that inform cognitive processes: directing what we attend to and what we ignore in situations, shaping the mental storage of new information, and aiding the retrieval of information stored in memory. While schemas streamline cognitive processes for speed and efficiency, they are implicated in the perpetuation of social expectations associated with gender and race.

We create schemas for both groups and the self. Group schemas are essentially stereotypes. Early work on group schemas, whether based on gender or race, focused on personality attributes. Classic studies by Katz and Braly (1933) and Karlins et al. (1969) (noted above) provided White Princeton students with a list of adjectives and asked them to indicate which traits were typical for specific racial, ethnic, and national groups. Students characterized Blacks as superstitious, lazy, and ignorant, while they viewed Jews as shrewd, intelligent, and industrious. While studies indicate that the content of some group schemas has changed over time (e.g., students were less likely to characterize Jews as shrewd following World War II), many components of racial group schemas endure.

Similarly, studies of gender group schemas consistently report that men are perceived to hold instrumental attributes (e.g., objectivity) more than are women, who are believed to hold expressive attributes (e.g., concern for others’ feelings) (Broverman, Vogel, Broverman, Clarkson, & Rosenkrantz, 1972). Traits are not the only component of gender (or race) group schemas. Carpenter and Trentham (2001) report that schemas include occupational types (e.g., nurse) and interpersonal roles (e.g., grandmother). Research consistently indicates that an awareness of gender as an organizing principle appears early in the lifespan, as toddlers have well developed gender categories and can correctly stereotype everyday activities and objects. Commitment to gender schemas oscillates throughout the life course: gender schemas appear rigid during a childhood consolidation period, but become more flexible in early adolescence before turning rigid again in the early and middle adult years (Signorella & Frieze, 2008) and more fluid again among elderly people. Furthermore, contextual factors, such as having nontraditional parents, may decrease the intensification of adolescents’ endorsement of traditional gender expectations (Crouter, Whiteman, McHale, & Osgood, 2007).

Beliefs about gender and race are represented at different levels, as individuals develop more specific subtypes within each general group (Deaux, 1995). Eagly and Mladinic (1989) note that gender subtypes may help explain an unexpected but consistently replicated finding: attitudes and stereotypes toward women tend to be more favorable than those for men. This finding is inconsistent with other reports that women are viewed less favorably than men in high achievement contexts (Sandler & Hall, 1986), particularly when these women are perceived as taking charge (Ridgeway, 2001). Glick and colleagues (Glick, Diebold, Bailey-Werner, & Zhu, 1997) find that sexist men have subtypes for both “good” and “bad” women. The “housewife” type is viewed as more favorable than the “feminist” type. Furthermore, reported preferences for the global category of women may reflect assumptions of Whiteness and middle class standing, and therefore not hold for lower class women or women of color (Landrine, 1985). Green, Ashmore, and Manzi (2005) recently documented subtypes for men (e.g., nerd and wimp) and women (e.g., bitch and whore) who transgress gender norms. Subtypes often are marked by intersections of gender, race, and other statuses (Stangor, Lynch, Daun, & Glass, 1992).

Self schemas organize knowledge about one’s self, including information about one’s gender and racial identities. Gender self schemas influence cognitive processing, including individuals’ ability to retrieve from memory experiences in which their behavior was consistent or inconsistent with their gender schemas (Markus, Crane, Bernstein, & Sildai, 1982). Moreover, gender self schemas may dictate which situations we choose to enter or avoid. Bem and Lenney (1976) found that both men and women with strong gender schemas avoided situations requiring them to behave in gender incongruent ways. Gender congruency appears be more important for men and boys than for women and girls (McCreary, 1994).

Whereas gender is often assumed to be a dichotomous variable with masculine and feminine poles, multiple racial groups exist, which make possible a multitude of racial identities (see section “Defining race and ethnicity” above). Racial self schemas have received less attention than gender self schemas, perhaps a consequence of social psychology’s reliance on primarily White student samples. The available research, however, examines important patterns across racial identities. Brown (1999) finds that older African Americans and those from lower social classes prefer to self-identify as “Black,” whereas younger and more educated individuals prefer “African American.” Other studies report that ethnic identification among Native Americans has risen in recent years. Nagel (1995) attributes this “ethnic renewal” to increases in opportunities now available to ethnic minorities. These patterns stand in stark contrast to those for White ethnics. Waters (1990) and Alba (1990) report decreases in ethnic identification among Whites, where ethnic affiliation is often embraced or discarded at will. These studies point to considerable variation in racial identifications both within and across groups.

Numerous studies document situational variation in racial identification. In a study on the effects of incarceration on racial identity, Saperstein and Penner (2010) find that the incarcerated are more likely to identify as Black rather than as White, independent of how they viewed themselves before incarceration. A parallel pattern holds for others’ racial perceptions for these individuals. Racial identities thus appear to be more fluid than essentialist perspectives assume.

Cognitive Processes

Cognitive structures focus on how individuals organize social information; cognitive processes address how this social information is used. Drawing from the literatures on attention, memory, and inferences, we briefly trace how each cognitive process informs our social psychological understanding of gender and race.

Attention

Thinking begins with attention, whether directed at a social object in the environment or to one stored in memory. Our attention varies in direction and intensity; some objects are more salient than others. Salience refers to the degree to which an object stands out from others and captures our attention. The schemas we hold for gender and race, together with situational cues, influence what we are likely to focus our attention on, which then has implications for inferences and behavior. Steele and Aronson (1995) suggest that situations making salient both one’s racial identity and cultural expectations associated with race may trigger a stereotype threat effect, which produces a self-fulfilling prophecy. Racist beliefs associating people of color with lower intellectual abilities have a long history in the U.S. To assess the impact of these beliefs in an academic context, Steele and Aronson administered portions of the Graduate Record Examination (GRE) to White and Black college students and found significant differences in performance, with White students outperforming Black students. By changing the instructions of the exercise to indicate that the test does not accurately assess intellectual abilities, the researchers were able to reduce, but not eliminate (Sackett, Hardison, & Cullen, 2004), the performance gap between White and Black students. Stereotypes may bear upon situations even when individuals do not endorse the beliefs (Devine & Elliot, 1995; Gorham, 1999). Research consistently finds stereotype threat effects based on race (Stone, Lynch, Sjomeling, & Darley, 1999), sex (Koenig & Eagly, 2005; Spencer et al., 1999), age (Chasteen, Bhattacharyya, Horhota, Tam, & Hasher, 2005), social class (Croizet & Clare, 1998), and (dis)ability (Quinn, Kahng, & Crocker, 2004).

Memory

Once information has captured our attention, it is then encoded into memory for later retrieval. Both the encoding and retrieval of information are influenced by social schemas. Hitchon and Chang (1995) assessed viewer recall of information communicated in political advertisements and found that respondents tended to recall content related to a female candidate’s family and appearance, while they tended to recall details about a male candidate’s campaign activities. Fyock and Stangor’s (1994) meta-analysis of 26 memory studies concludes that individuals better remember information about social groups that confirms rather than disconfirms a priori expectations. Other studies have found that individuals tend to recall negative information about outgroups but more positive details about ingroups. Moreover, we tend to recall differences between ingroup and outgroup members rather than their similarities. In these ways, memory helps confirm our expectations. Taken together, these observations suggest that schemas direct us to attend to some pieces of information, to encode them into memory, and to then easily recall them – even when the information is not entirely accurate (Cohen, 1981; Kleider, Goldinger, & Knuycky, 2008). Thus the operation of social schemas contributes to the perpetuation of those stereotypes.

Inferences

Individuals draw on information – either information we have stored in memory or that which is available in the immediate situation – to make a variety of social inferences and judgments. Attributions are judgments made to explain the origins or causes of events. Group schemas for gender and race have considerable influence on attribution processes. Comparing student attributions of blame in bias crimes with gay, lesbian, and Black victims, Lyons (2006) finds that individuals display sympathy for Black victims of hate crimes; however, they are less likely to display sympathy for gay and lesbian victims. Furthermore, these latter groups of victims are held more responsible for their victimization. These judgments may be due to individuals’ reliance on widely held lay theories suggesting that sexual orientation is a “choice” whereas race is ascribed and to their greater familiarity with the racial rather than sexual power asymmetries operating within American society. Similarly, Howard’s (1984) study of attributions for male and female victims in rape and robbery scenarios illustrates how gender schemas promote differential evaluations. In general, participants assigned more blame to female victims. This blame resulted from assessments of the female victims’ character (relatively unchangeable factors), whereas male victims were blamed for their behaviors (relatively changeable factors). Inman, Huerta, and Oh (1998) clarify how power relations associated with the race of a hypothetical perpetrator influence whether respondents perceive the perpetrator’s behavior as discriminatory toward another person. White and Black perpetrators can both be seen as discriminatory, depending upon whether participants expect this type of person to be in control of the specific situation. Individuals who are expected to be in control are held more accountable for their behavior, and actions that violate a norm of responsibility (e.g., actions seen as harming the disadvantaged) tend to be perceived as discriminatory. Taken together, these studies suggest that individuals’ lay theories for gender, race, and sexuality affect judgments about responsibility. These judgments can produce real-world consequences, such as offering or withholding help (see Hollander et al., 2011).

Symbolic Interaction

In theorizing the relationship between the individual and society, symbolic interaction is guided by three central premises outlined in the work of Herbert Blumer (1969). First, humans act toward social objects based on the meanings that these objects hold for them. From this perspective, gender and race are significant symbols whose markers are laden with social and cultural meaning. Second, meaning emerges in interactions between individuals or between individuals and social objects. Third, interpretation is central to the processes through which meanings influence interactions (Snow, 2001). Interpretation allows actors to make sense of the situation, to reassess the situation as necessary, to negotiate meanings with others, and to decide among lines of action. Symbolic interaction conceptualizes social actors as agentic, creative beings. While this capacity is often unobservable, instances in which everyday interactions break down provide reminders. The symbolic interactionist frame assumes “[n]either society nor its subparts exist as static entities; rather, these are continuously created and recreated as persons act toward one another” (Stryker & Vryan, 2003, p. 4). It is this negotiation, rather than the individual, that is central to an interactionist analysis. As the examples in the following sections illustrate, the cultural assumptions that undergird the symbols that guide interaction are, in a sense, cognitive stereotypes. Behavioral interactions are guided by these stereotypes in action repertoires motivated both to attain individual goals and to affirm (or disconfirm) broader normative systems.

Gender and Racial Identities

Divergent approaches within the symbolic interactionist tradition tend to emphasize either the structures of identity or the processes of identity construction. Whereas a cognitive approach to identity stresses social categorization, symbolic interactionists concerned with the structures of identity focus on role identities. This focus on role identities follows from the interactionist emphasis on relationships, because role identities are generated through ties to others. In a classic study, Kuhn and McPartland (1954) asked individuals to provide 20 responses to the question: Who am I? Responses included numerous social role identities (e.g., husband, daughter, good wife, etc.), many of which were clearly gendered, that respondents held to be central to their sense of self. Later work by Stryker (1980) suggests that such role identities are organized hierarchically: ties to other people and the emotional strength of these connections impact one’s commitment to a particular identity. These role identities are embedded within social relationships, and consequently, they anchor individuals to social groups, and by implication, also to social structures.

Given the focus on the concept of roles, this branch of symbolic interaction attends to gender more than to race. Role identities associated with gender are defined in part by our positions in social structures and institutions such as the family. This perspective emphasizes the importance of role expectations and performance and the potential for role conflict. Cairns, Johnston, and Baumann’s (2010) study of foodies, or “people with a passion for eating and learning about food”, provides an example of how some women seek to balance the role of “foodie” with gendered expectations associated with being a “good mother.” The motherhood endorsed by these women has clear middle class undertones, illustrating how the meanings associated with social roles, the importance one places on a given role, and one’s sense of success in fulfilling roles are shaped by the intersection of social positions.

Another line of research emphasizes the processes of identity construction through exploring strategies that individuals employ to manage others’ impressions. Drawing on Goffman’s (1959) dramaturgical analysis, these studies highlight the dramatic, performative nature of everyday encounters. In order for social actors to interact with each other, they must first define the situation by establishing who each individual is, what role each is playing, and what expectations each is bringing to the encounter. Consequently, how individuals project information about themselves to others is a chief concern.

As we noted in an earlier section, West and Zimmerman (1987) blend phenomenological and interactionist insights to argue that gender is an accomplishment of everyday interaction – individuals “do gender.” From this perspective, gender is embedded in the mundane, routine activities of everyday life. Ramirez’ (2006) interviews with pet owners from a dog park, for example, describe how men and women use their pets as props in the presentation of gendered selves. From their initial selection of a particular pet based on cultural associations defining animals as either masculine or feminine to the ways these individuals frame their relationship with their pet in conversations with the researcher, informants are doing gender. Scholars note how gender is accomplished in other taken-for-granted cultural rituals, such as engagement proposals (Schweingruber, Anahita, & Berns, 2004) and weddings (Ingraham, 2008). It is through our participation in these activities that individuals (re)produce gender.

Building on the notion that difference is also an accomplishment of everyday interaction, scholars have examined the accomplishment of other identities, such as race (Ahmed, 2007; Wilkins, 2008), class (Bettie, 2003; Dressman, 1997), and sex (Dozier, 2005). For example, Best (2003) examines the ways that both researcher and research participants “do race” through the talk of qualitative interviews. While reviewing a taped interview completed with two young women, one an African American and the other a Latina, Best noted several long breaks in the conversation. She also noted that moments of “narrative bracketing/translating” followed these pauses. This strategy of interrupting a story to define a slang word or interjecting a quick “You know what I mean?” allowed these young women of color to linguistically position the White researcher in relation to the stories they recount. Best is an obvious racial outsider; thus, these young women want to ensure that their stories are actually heard by her. Best concludes, “Their acts of translation – their speech acts – were a way to solidify their racial (and age) identities and to remind me of my own. These young women deployed language in ways that constructed (and sustained) my identity as an outsider and in so doing, discursively sustained my status as ‘White’” (pp. 903–904). Through these methodological reflections, Best argues that both researchers and the people they study actively produce race in the research setting (see also Sprague, 2005; Warren, 2001).

Numerous scholars have begun to apply this analytical lens to the intersection of identities (Bettie, 2003; Higgins & Browne, 2008; Wilkins, 2008). Warren’s (2009) study of the Mapuche people of Argentina describes the intersectional process whereby gender displays, such as wearing traditional garments during public appearances, allow female activists to construct an authentic indigenous identity to promote their movement’s struggle for rights. Women bear primary responsibility for undertaking this identity work on behalf of their community, a gendered pattern that is consistent with other studies on the maintenance of racial and ethnic identities (see Phinney, 1990).

Other scholars are beginning to theorize the “undoing” of gender. If as West, Zimmerman, and Fenstermaker argue, gender and race are done, then they may be “undone” (Deutsch, 2007; Risman, 2009). Social psychological studies might move beyond just describing the production of everyday inequalities and begin to identify strategies that may interrupt these social inequalities.

Identity performances are negotiations based upon social expectations that originate in a shared symbolic universe. However, negotiation has limits. In his study of passing, Renfrow (2004) reports that individuals must stylize their identity performance such that it aligns with social expectations consistent for the identity they are attempting to convey, while simultaneously masking characteristics that may betray the performance. In this way, performances are idealized representations of prevailing stereotypes; individuals are not entirely free to improvise or construct novel performances. Moreover, negotiation does not require that interactive partners be equal partners in the collaboration. Negotiation may reflect structural power asymmetries associated with the actors. Furthermore, the working consensus achieved may be independent of each actor’s private beliefs. Interactionist work characterizes processes of impression management, whereby individuals strategically share and control information about themselves with others during interaction and attempt to shape the information offered by their interaction partners, as smooth, seamless encounters; however, room exists for interactions to break down, stall, or come to an embarrassing halt. Consequently individuals often must direct considerable efforts to repair interactions, particularly when questions arise about who an actor “really” is. Studies document numerous impression management strategies, including the use of cover stories (Roschelle & Kaufman, 2004) and concealment (Thorne & Anderson, 2006). These observations underscore the nature of negotiation: social actors enjoy some flexibility in how they construct and perform their identities, but they must do so in ways that are recognizable and convincing to their social audience.

Behavioral Confirmation

Over 80 years ago, Thomas and Thomas (1928) observed that if people “define situations as real, they are real in their consequences” (p. 572). The implication is that objective differences need not exist for difference to be perceived. Numerous studies have since documented how beliefs – either those held by one’s self or by others – come to bear on real-world behaviors, often reifying preconceived notions of between-group differences and within-group similarities. (e.g., the stereotype threat effects discussed above [Spencer et al., 1999; Steele & Aronson, 1995]). For example, Frederickson and Harrison’s (2005) study of girls in athletics found that girls tended to internalize gender expectations emphasizing physique over performance. These internalized expectations promoted self objectification, or turning a third-party gaze onto their own bodies. This process inhibited the girls’ self-efficacy and athletic performance.

Another line of research documents how social expectations held by others produce and maintain the gendered and racialized order. In the classic study of self-fulfilling prophecies, Rosenthal and Jacobson (1968) provided elementary school teachers with information about their students’ performance on an IQ test at the beginning of the school year and found that at the end of the year the students met teacher expectations. Rosenthal and Jacobson attribute this pattern to the influence of teacher expectations on their own behavior toward students. Subsequent studies have connected teachers’ lowered performance expectations for Black and poor students with a low sense of academic self-efficacy and lack of effort toward school work in these children, and with teachers’ unwillingness to provide academic challenges for these students (Farkas, 1996; Ferguson, 1998). Eccles and Blumenfeld (1985) argue that teachers passively contribute to the beliefs that girls are less likely than boys to excel in mathematics and science. They conclude that teachers fail to “to provide boys and girls with types of information that might lead them to reevaluate their sex-stereotyped beliefs. In this way teachers passively reinforce the sex-typed academic and career decisions made by their students, thus contributing to sex inequity in children’s educational attainment” (p. 80). As these studies show, internalized social expectations can produce self-fulfilling prophecies. Expectations associated with gender and race influence which individuals have access to opportunities and how individuals navigate the opportunity structures available to them. These examples highlight the compatibility of symbolic interaction with social cognition, as both theories can clarify these same processes.

Resisting Social Expectations

Symbolic interaction emphasizes human agency, which guards the perspective against constructing an “oversocialized” social actor (Wrong, 1961). Recent review chapters note that identities can serve as resources for resisting social expectations (see Callero, 2003; Howard, 2000). Anderson’s (2005) observations of male cheerleaders in traditional leagues – those he calls “orthodox” – illustrate how young men committed to hegemonic masculinity attempt to resist cultural stereotypes assuming they are feminine or gay. These men often display “defensive heterosexuality” and engage in “guy talk.” These strategies may be less important for men who are known to be active in other masculine activities, as they often possess “jock insurance,” or greater ability to resist the feminizing influence of a single activity (Pascoe, 2003).

Other studies illustrate how the intersection of positions adds complexity to acts of resistance. Duneier (2000) explores the sexually provocative comments unhoused, Black men selling printed materials on the sidewalks of Greenwich Village use to entangle women into unwanted conversations. Although these men are positioned lower in the social hierarchy along several dimensions (e.g., race, social class, etc.) than are female passersby, they are able to use their position as heterosexual men to dominate women, deflect stigma, and reaffirm their sense of self as men through talk. Santos’ (2009) fieldwork in East Los Angeles tattoo parlors provides another example. Santos characterizes these sites as masculine domains in which men act as cultural gatekeepers by controlling the “Chicana canvas” and setting limits for how women “do Chicana”. However, by taking charge of their bodies and embellishing them as they choose, women reclaim and redefine their Chicana identity. Thus their tattoos “represent a counter-hegemonic choice to achieve autonomy from patriarchal ideals and self-definition against class oppression, heteronormative standards, gender legacies, sexism, and racism” (p. 98). While these studies illustrate successful attempts at the individual level, resisting social expectations can be fraught with difficulties.

Individuals with higher status hold more power with which to enforce their expectations for others, while those with less power may not be able to resist (Cast & Cantwell, 2007; Stets & Burke, 2005). In the context of romantic relationships, for example, individuals’ self-views and spouses’ views of them converge over time (Cast & Cantwell). Where economic and/or relational power imbalances exist, “those with more power are more able to behave in ways consistent with their identity, more able to impose an identity on their spouse, and more able to resist the identity that the spouse, in turn, seeks to impose” (Cast, 2003, p. 185; Cast, Stets, & Burke, 1999). As we noted above in our discussion of the expectation states literature, numerous studies find that an individual’s expectation for an interaction partner based on his or her perceived minority status in systems of race (Cohen, 1972), gender (Pugh & Wahrman, 1983; Ridgeway, Backor, Li, Tinkler, & Erickson, 2009), or sexuality (Childers, 2000; Renfrow, 2005; Webster et al., 1998) – even in cooperative settings in which these statuses are irrelevant – contribute to an unwillingness to allow this partner to contribute to the group’s efforts. Individuals’ ability to resist social expectations must not be overstated. Symbolic interaction concentrates on micro-interactions, and consequently, it often fails to adequately analyze the structural obstacles to resistance.

Together, social exchange, social cognition, and symbolic interaction offer the potential to elucidate the micro-foundations of systems of gender and race. Social exchange theory examines the conditions under which individuals make choices about allocations of resources. Social cognition attends to the structures and processes of thought. Symbolic interactionism examines how meaning is created and shared, and how individuals develop (or resist) a sense of a gendered/racialized self through situated action. Although these theoretical lenses can be applied to and provide insights into the same social phenomena, social psychologists by and large continue to use them independently. Consequently, the full promise of these social psychological theories to explain the complexities of gender, race, and their intersections has not been fully realized.

Methodologies

In considering the social psychological theory and research on gender and race, it is critical to consider how that research is actually done, that is, the methodological choices that guide research. This is considerably more than a technical question. Harding (1987) distinguishes three elements: epistemology, method, and methodology. Epistemology is a theory about knowledge, about who can know what and under what circumstances knowledge is developed. Methods are techniques for gathering and analyzing “data,” information. Methodology delineates the implications of an epistemology for implementation of a particular method (Sprague, 2005). The technical details are located in their social and political context. In earlier years, social scientists tended to see their work as value-free, neutral, objective. Many social psychologists no longer feel that objectivity is possible; therefore, it is critical to engage in methodological reflection.

It is important to ask where research questions come from, whose interests they reflect, whose interests are not represented, what consequences follow from conducting research in certain ways, and with certain populations. Considering the substantive focus of this chapter, we ask which social distinctions are most salient, how questions vary depending on the social category under investigation, or what consequences follow from the prevailing tendency to approach social problems as outcomes of individual rather than societal factors.

As we have noted, social psychology, and sociology more generally, has chosen to explore the differences between some groups that are hierarchically related to each other, especially men and women, and Whites and non-Whites, primarily African Americans. We have seen numerous examples of research on the differences across members of the categories within these systems. We see much less research that considers the relative size of cross-group differences compared to within-group variation. So, for example, many of the stereotypic traits associated with women, on the one hand, or with men, on the other, actually vary more among women and among men, than they do across women and men (see Hyde, 2005; Sprecher & Toro-Morn, 2002).

Also, as we have suggested, questions posed about socioeconomic standing differ considerably from those posed about gender and about race. Indeed, there is much less research overall on social class than on gender or race (Frable, 1997). More generally, researchers ask different questions about the privileged than about those who are subordinated, tending to explore deficiencies of those in subordinate positions. Sprague (2005) cites some telling examples: there is considerable work on women’s lack of self-esteem, but very little on men’s lack of modesty. Moreover, social scientists tend to “study down” (Fine, 1994). Part of the reason for this is that those with considerable social power can usually protect their space such that research cannot be done on them (Fine, 1994). This has begun to change. There is now a body of research on whiteness (Frankenberg, 1993), masculinity (Connell, 1995), and heterosexuality (Herek, 1995; Maynard & Purvis, 1994), and some of their intersections (Robinson, 1996). But there still is very little research on those who are affluent; socioeconomic status has been the most unapproachable system.

Moreover, most social psychological research focuses on individuals, conceptualizing them at a remove from societal context. As Sprague notes, this can lead to a reliance on logical dichotomies that mask important social dynamics. So, for example, Howard and Hollander (1997) maintain that altruism and aggression, two forms of behavior included in virtually all social psychological texts as distinct, are intimately linked through the workings of social power. Power determines who needs help, who has the resources to offer help, and who cannot turn down help. It determines what is even defined as help (e.g., public assistance is help, tax breaks are not) and what is defined as aggression (e.g., stealing is, tolerating hunger across the globe is not).

Social exchange research tends to employ experimental techniques. This is a research paradigm that emphasizes tightly controlled research procedures and discrete, easily measured variables. The experimental procedures typically constrain interactions such that subjects may never see, hear, or encounter each other. These procedures have the advantage of reducing effects of ostensibly extraneous factors, allowing stronger causal claims. This is the key advantage of high internal validity. But they lack external validity; factors that do influence real-world interaction may be considered “extraneous” only in that narrowly constrained context. The process of interaction disappears.

The kinds of tasks used in these experiments also limit the scope of conclusions that may be drawn. Monetary rewards are common in exchange research because money is assumed to be a reward that has equal value for participants. This assumption, however, is insensitive to the socioeconomic positions of those participants. A few dollars may be meaningless to a relatively well-off participant, but be of considerable value to another participant. Indeed, in situations in which people are not participating as part of a course requirement, the very fact of choosing to participate may reflect their need for income. And, symbolic factors may be just as powerful, if not much more powerful, resources in real-world exchange relationships.

The lack of external validity in exchange research is also associated with the lack of attention to relationship history. Exchange histories in real-world cultures can be the cement that allows them to endure (Emerson, 1972a, 1972b). Yet groups used in experimental studies are most often composed of people who have never met. A relationship history is sometimes created by transactions that occur a short time before an experiment begins; this is a far cry from grounded exchange relationships. As we noted earlier, some have argued that these patterns reflect a deep gender bias in social exchange research (e.g., England & Kilbourne, 1990).

Methodological factors underscore the striking absence of race in social exchange research. A key epistemological assumption is that actors come to the marketplace as equals. Although they may have different levels of resources and positions, these are treated as malleable and transferable, an assumption that most often is true neither for women vs. men nor for people of color vs. Whites.

Social cognition uses a greater variety of methods, falling somewhere in the middle between the experimental reliance of social exchange and the field work and ethnographic preferences of symbolic interactionists. One important feature of social cognition is that cognition itself, unlike exchange, cannot be observed. Measures of cognition are necessarily indirect, whether they be self-reports, reports by others, or inferences from behaviors. Social cognitive researchers also use experimental methods, with the same advantages and disadvantages noted in discussing social exchange. The difference is that cognitive researchers are more likely to add elements of social context either within the experimental design, or by adding other methodological components to the overall studies. Social cognitive studies often use survey techniques, asking participants to complete questionnaires, whether they be attitudinal inventories, personality trait assessments, inferences about causal claims, probabilities, or other cognitive judgments. It is common for such studies to combine these assessment techniques with experiments. So, for example, researchers might have participants complete the written instruments a month prior to engaging in an experiment, as a way of measuring cognitive predispositions that can be used as a predictor or correlate of the experimental behavior. If the cognitive measures are assessed well in advance of an experimental procedure, any influence can be viewed as potentially causal.

Studies can be carried out in multiple contexts, allowing cultural comparisons, as evident in the study cited above in which intersections between gender and race led to variations in the prominence of identities across different types of educational institutions, such as coed or single sex, and primarily White or primarily Black colleges (Settles, 2006). Similarly, experiments can be conducted in the “real-world,” outside of laboratory settings. Experiments can be built into survey instruments as well. Field experiments are increasingly common (Bertrand & Mullainathan, 2004; Dutton & Aron, 1974; Paluk, 2009), because they can have more external validity than a lab-based experiment, without significant cost to internal validity.

The extent to which external validity can be achieved depends greatly on the design of the specific instruments and the design of the experiments with which they are combined. As we have seen in our discussion of the M-F scale and of androgyny, a trait assessment can be designed to impose a dichotomy between M and F, on the one hand, or to allow variation in both dimensions, as in androgyny. Deliberately comparative cross-group frameworks can be built in, or avoided, through assessment of within-group characteristics only, or investigated indirectly, through comparing responses to within-group inventories of two or more groups. These choices are in the hands of the researcher. Social schemas influence researchers as strongly as they do others (Hollander et al., 2011). Howard (1988), for example, reconsiders a study that concludes that women are more sexually conservative than men (Hendrick, Hendrick, Slapion-Foote, & Foote, 1985), arguing that this study failed to consider the societal context in which sexual attitudes are formed.

Symbolic interaction research tends to rely on ethnographic methods, with the advantage of greater external validity, but the weakness of lower internal validity, and the consequent inability to make strong causal claims. Some interactionists might go so far as to say that this is not a weakness, arguing that causal assertions about processes that might not be replicable in the real-world are not as meaningful as grounded, embedded research. A number of symbolic interactionists do also use survey instruments, especially to measure identities and to explore the structure of identities. But most also recognize that a deeper understanding of identities may require interviews and conversations. Karp (1986), for example, studied a group of currently upper-middle-class professionals who had been raised as children in socioeconomically deprived circumstances. These upwardly mobile individuals reported having difficulty integrating these different circumstances, and continued to feel uncertainty, inauthenticity, and marginality. More structured methodological techniques would likely not have allowed Karp to detect these complex patterns. Some interactionist analysis entails observations of others who may or may not even be aware of being observed, at least for social psychological purposes. Presidential elections, for example, offer interactionists remarkably rich opportunities to watch the workings of interpersonal dynamics. Goffman’s concept of impression management is an exemplar: when Barack Obama chose to wear – and not to wear – an American flag pin presumably reflected the impressions he sought to communicate.

Experiments can be used by symbolic interactionists, but typically they include some awareness of social context. For example, Word, Zanna, and Cooper (1974) found that White interviewers used different interview styles in interviewing Black as opposed to White job applicants, using a “nonimmediate” interview style for Blacks, entailing a greater physical distance, less eye contact, more speech errors and shorter interviews, in contrast to an immediate interview style for Whites. In a second stage of the study interviewers were trained to work with both styles; applicants subjected to the nonimmediate style, both Whites and Blacks, performed less adequately. This study entailed both experimental methods and contextual sensitivities.

Because language is so critical to interaction, linguistic methods and analyses may also be used by symbolic interactionists. One important feature of interaction is that sometimes it fails, impressions are not successfully created, or relationships are damaged. The study of “repairs” is a hallmark of symbolic interaction. Linguistic analysis is necessary to identify accounts given for behavior, whether they be excuses or justifications.

That symbolic interactionists tend to be much more attuned to the importance of social context, and therefore use methods with strong external validity, does not mean they are necessarily sensitive to the workings of social power. Some scholars treat both negotiation and exchange (a subset of negotiation) as if the actors come to the scene with equal resources, equal opportunities, equal skills. The choice of a method strong in external validity does not in and of itself ensure a full appreciation of societal inequalities.

Current Trends and Future Directions

The current state of the social psychology of gender and race reflects two trends within the parent fields of sociology and psychology. Despite calls for integration from both sociologists (Hollander & Howard, 2000; House, 1977) and psychologists (Ryff, 1987), social psychology remains theoretically, methodologically, and institutionally fragmented. House’s (1977) classic article outlined three faces of social psychology – a psychological social psychology, symbolic interactionism, and psychological sociology – each theorizing in isolation, using unique methodologies to analyze phenomena of common interest, and publishing in separate outlets to different audiences. More than 30 years later scholars continue to ignore House’s impassioned call to abandon intellectual and institutional tradition to establish “new interfaces”. As our review of the largely unintegrated theoretical approaches to the social psychology of gender and race illustrates, these literatures are no exception.

Reflecting on the growth of these different knowledge communities on the centennial of the publication of the first two social psychology textbooks, Gergen (2008) argues that social psychology needs more than to unify its theoretical and methodological traditions. According to Gergen, these social psychological discourses have become so compelling to scholars that they often view orienting concepts as “real” rather than as the social constructs they are. His solution is a call for social psychologists to embrace an ontology of social life, which would free social psychology from the “dualistic distinction between mind and social world” (p. 335). The result, he argues, will be a social psychology “in which prediction and control become secondary, and the uses of research for collaborative transformation of society will flourish. It will be a science centrally concerned with pressing issues of the day, and offering creative options for more viable life forms” (p. 337).

Gergen’s vision for the development of a more critical social psychology dovetails with two contemporary trends: the emergence of public sociology, which seeks to replace disinterested scholarship in the academy with practical research shared with a popular audience (Burawoy, 2005; Clawson et al., 2007; Jeffries, 2009), and the emergence of positive psychology, which seeks to replace a focus on disorder and pathology with a focus on happiness, love, and justice (Seligman & Csikszentmihalyi, 2000). Scholars link these values to early work in each discipline – for example, to work by Durkheim and Du Bois in sociology and Terman in psychology. Sprague (2008) argues, however, that by and large contemporary academics have not fulfilled this potential because they have been “disciplined.” Knowledge discourses determine who can be a knower and what can be known, academic disciplines organize professional activities, and our disciplinary training (re)produces the social construction of the disciplines.

Movement toward a public sociology and positive psychology holds the potential to foster a social psychology with a critical eye and an activist conscience, one whose impact can already be seen in scholarship on gender and race. Studies of stereotype threat, for example, not only seek to document the conditions under which the effect occurs, but many identify conditions that deactivate the threat, enabling the elimination (or at least reduction) of performance gaps associated with gender and race (e.g., Aronson, Fried, & Good, 2002; Good, Aronson, & Inzlicht, 2003). Similarly, numerous experiments attempt to identify conditions that reduce individuals’ reliance on implicit associations (e.g., Lowery et al., 2001; Mitchell et al., 2003). Moreover, important scholarly insights with obvious real-world relevance now are being disseminated to the public through the popular press: so, for example, the finding that being Black is more detrimental to one’s chances of finding a job than having a criminal record (Pager, 2003) and that implicit associations linking people of color with criminality may help account for racial profiling (Greenwald, Oakes, & Hoffman, 2003). Contexts, a relatively young quarterly magazine published by the American Sociological Association (ASA), serves as a vehicle for sharing sociological insights with “educated lay readers.” We should note that not all scholars share our optimism about this trend. Tittle (2004), for example, challenges the notion of a public sociology on the grounds that it is built on false assumptions, that it questions sociology’s cultural legitimacy, and that public sociology is incompatible with scientific practice.

Gender and race remain foundational concepts for social scientists. The ASA sections on sex and gender; race, class, and gender; and race and ethnicity continue to be among those with the highest memberships (among 40 plus sections). (All three have larger memberships than the social psychology section.) This sustained interest yields thousands of scholarly articles each year, many of them by social psychologists.

Future social psychological work on gender and race and their intersections will be shaped by continued technological advances, important demographic shifts in the population, and growing debates about the stability of the concepts themselves. Technological advances over the past decade have facilitated a more sophisticated social psychology. Computer interfaces have become standard tools for maximizing researcher control in laboratory settings. Software applications measuring implicit associations through split-second response latency measures provide just one example of how new technologies continue to advance social psychological understandings of the impact of gender and race. As Correll and colleagues (2010) note, these research tools not only improve our ability to observe the seemingly unobservable, they make possible new conceptualizations of sexism and racism. We expect future technologies to continue to enrich the social psychology of gender and race, both methodologically and theoretically.

In addition, shifts in immigration and intermarriage patterns within the U.S. will continue to move social psychology beyond analyses of the Black-White divide that has long characterized American race relations. Lee and Bean (2004) report that immigrants and their children currently make up 23 % of the population, and since the 1980s approximately 85 % of “legal” immigrants to the U.S. are arriving from Asia, Latin America, or the Caribbean, compared to only 12 % from Europe and Canada. Latinos and Asians are expected to make up approximately one-third of the U.S. population by 2050. This represents a dramatic shift from the largely European immigrations of the past. Along with these new waves of immigration, the rate of racial/ethnic intermarriage continues to increase. Recent estimates suggest racial/ethnic intermarriage hit an all-time high in 2010 at 8.4 % of all marriages (Wang, 2012), up from 5.4 % in 2000 (Lee & Edmonston, 2005). Over 50 % of couples with a self-identified Latino or Asian member marrying during the 1990s included a partner of another race (Lee & Bean, 2004; Waters, 1999). From 1970 to 2000, the number of children born to parents of different races increased from 900,000 to over three million (Lee & Edmonston, 2005). Currently, one in 40 Americans self-identifies as multiracial. Some estimates project that this ratio will be 1 in 5 by the year 2050 (Bean & Lee, 2002).

Research has only begun to investigate how these demographic shifts are influencing social psychological processes. Snipp’s (2003) recent review notes that while the majority of people who report their race as Black or White on the U.S. Census mark a single category, American Indians, Asians, Pacific Islanders, and Latinos frequently report multiracial identities. The literature suggests multiracial individuals have more flexibility in their racial identification (Harris & Sim, 2002; Nobles, 2000). Nevertheless, Doyle and Kao (2007) find consistency in 53–57 % of the self-reported racial identities for most multiracial groups when comparing initial and follow-up longitudinal data. Individuals initially self-identifying as Native American-Whites stand out as the most flexible group, as only about 19 % report the same racial identification over time. In fact, this group displays a greater likelihood of identifying as White in subsequent interviews. Lee and Bean (2004) identify contextual factors such as nativity and generational status, bilingualism in the home, and proximity to non-White communities as important determinants of racial identifications in multiracial children. Taken together, these demographic shifts not only change who will be the focus of social psychological research in the future, but they also point to the need to revise our models of racial identity (Rockquemore & Brunsma, 2007). In pointing to the instability of racial identities and highlighting that many individuals have “ethnic options”, multiracial identifications raise fundamental questions about the meaningfulness of race as an orienting concept.

Similarly, increased scholarly attention to transgender identities and lifeways has begun to reveal a significant destabilization in the concept of gender. Transgender emerged in the 1990s as an umbrella concept widely used to describe individuals who “cross over, cut across, move between or otherwise queer social constructed sex/gender boundaries” (Stryker, 1994, p. 251). Transgender bodies and identities trouble the assumption of a static sexual binary upon which our sex/gender system is constructed.

We are encouraged that many scholars are striving for a more critical social psychology. We urge social psychologists, particularly the next generation of scholars, to continue to pursue research agendas that promote the goals of social justice and fulfill the promise of an integrated social psychology.