Kristi's daughter Chloe and her friend Nicole were playing at Kristi's house when the subject of ethnic background arose. “Chloe's a mutt,” Kristi told Nicole.

“What's a mutt?” Nicole demanded, her 12-year-old mind not familiar with the concept.

“It means she's got a lot of different nationalities roaming around her background. Part Greek from her daddy's father. English and Scots from his mother. And just about everything from northern Europe plus a teeny bit of Cherokee on my side.”

Nicole thought about this a moment.

“You're a mutt, too,” Kristi continued.

Now Nicole was interested. “How do you know that?” she quickly demanded.

“Because I know your mother's mother is French and your dad is part English and part German.”

Nicole's face froze, her retort sharp and indignant.

“My father is NOT German! He's art history! He's chair of the department!”

Raised in an academic ghetto, by a father who indeed was chair of the art history department and a mother who was both an art history professor and associate dean of humanities, Nicole's outraged protest reflected her particular worldview, her knowledge of what was important for academics' kids: academic disciplines. Nicole did not care a hoot about ethnic, racial, religious, or national backgrounds. But little Nicole, so seldom party to the social science dinner table conversations that too often bore children of political scientists, illustrates an important concept that all political scientists should incorporate into their analyses of race, ethnicity, and group politics more generally: the concept of moral salience, which we define as the psychological process by which differences between people and groups become deemed ethically and politically relevant (Monroe Reference Monroe2004).

For an academic's kid, disciplines matter. Disciplines are what allocate resources. They are the divisions around which passions cleave. Few in academia even notice, let alone care, if their colleagues are French, or German, or Chinese. Religion's been a non-starter for years among academics, once anti-Semitism (thankfully) became a thing of the past. (Though a religious colleague tells us she is reluctant to talk about her Roman Catholic faith and we suspect Islam is now controversial on many campuses, at least for students.) What is relevant in academia is field, your area of specialization, and perhaps your methodological approach. Faculty brats thus are not raised to think in terms of racial, ethnic, religious, or national prejudice. These differences are simply not relevant for them. They carry no moral salience.

In most discussions of identity in American political science, we find an odd confusion between group politics and identity politics, and a surprising tone in our discussions of differences that implies that the group distinctions that currently dominate American politics exist because they reflect some immutable difference between groups in our society. The difference may exist, and they may even be immutable. But this is not the critical factor. What matters is the political and moral salience we accord these differences. Because this salience is totally an artifact, a social construction that does not lock us into patterns of prejudice, it may prove easier to intervene to change the moral salience attached to differences than it is to change immutable characteristics, such as skin color or biological sex. To be clear, social constructions are not necessarily easy to break. Seminal work by Benedict Anderson (Reference Anderson1983) provides powerful illustration that the very concept of nationalism emerges from a kind of shared hallucination, or a communal agreement that participants simultaneously adhere to in constructing their notion of community or nation. Later path-breaking work on intersectionality, especially critical race theory, demonstrated the ways in which various socially constructed categories based on race or culture intertwine to establish and maintain forms of oppression and inequality in modern social contexts (Crenshaw Reference Crenshaw1991).

But the folly of assuming that the differences we grow up assuming are ethically relevant and condemn us as a society to ongoing prejudice is easily seen if we consider social identity theory.

Social identity theory was formulated by Henri Tajfel (Reference Tajfel1970, Reference Tajfel1981), a Polish Jew who survived World War II and worked in Bristol, England, where he and students such as John Turner (Tajfel and Turner Reference Tajfel, Turner, Austin and Worchel1979; Turner Reference Turner and Tajfel1982) tried to understand the psychological basis of the kind of prejudice and discrimination at the heart of the Holocaust. They argue first that people naturally categorize. We have complex identities and sort ourselves and others into categories, labeling people as members of diverse groups. These groups then are juxtaposed in pairs. We classify people as men or women, young or old, rich or poor, friend or foe, or, in the instance that initially motivated Tajfel's work, the Nazi classification of Jew or Aryan. We eventually identify and associate with certain groups, which Tajfel called in-groups. Doing so plays an important psychological function in bolstering our sense of who we are and how we feel about ourselves. As part of this process, we compare our in-groups with other so-called out-groups, and demonstrate a favorable bias toward the group to which we belong. Tajfel's social identity theory thus roots prejudice, discrimination, and the violence that can result from it in an innate psychological need for distinctiveness, self-esteem, and belonging. We desire our identity to be both distinct from and compared positively with that of other groups. The critical intellectual traction of social identity theory lies in establishing a clear link between the psychological and sociological aspects of group behavior, in effectively linking the micro-level psychological need to distinguish, categorize, and compare ourselves with the broader social phenomenon of group behavior.Footnote ¹

The theoretical claims of social identity theory are substantiated in thousands of experiments, conducted by Tajfel and his students in what became known as the Bristol School of social psychology. The classic experiment takes a group of previously unconnected individuals and randomly assigns them to group A, B, or C. Everyone in group A is then offered option 1 or 2. Option 1 will give all members of group A $5, all members of group B $10 and all members of group C $15. Under option 2, all members of group A will lose $5, all members of group B will lose $10, and all members of group C will lose $15. As we might expect, most members of group A (roughly two-thirds) will take option 1, which gives money to the members of their group. But with a surprising consistency, roughly one-third of members of group A will choose option 2, the option that costs them money. Tajfel's explanation? Under option 2, members of group A do better than all other groups. Their need to do better than others, to feel superior in some way, trumps their need to actually do better in objective terms. This distinction mirrors the common debate in international relations between realists who focus on relative gains and liberals who emphasize absolute gains (see Baldwin Reference Baldwin1993 for an overview of this debate).

This framework thus provides a valuable beginning point for understanding how important both real and perceived differences can become when encounters between individuals are conceptualized as encounters between group members. (For Nicole, the relevant out-group would be non-humanities, as she hears her parents grumble about resources going to the biological or physical science end of the campus.)

Social identity theory has proven a useful tool for analyzing many types of political behavior. It lies at the heart of the classic description of the process by which group identities crystallize into ethnic hostility and genocide (Staub Reference Staub1992), to use as an example the academic subject that currently occupies our attention (Monroe Reference Monroe2008). It suggests (Bar-Tal Reference Bar-Tal1990; Bar-Tal and Teichman Reference Bar-Tal and Teichman2005) how this process of in-group/out-group differentiation can make each group the enemy of the other; groups then limit individual choice by telling members what is appropriate behavior. One contention is that genocide erupts when ethnic identities become reified and boundaries harden into politicized identities, as opposed to less polarizing cultural identities (Petersen Reference Petersen2002, 3–4; Mamdani Reference Mamdani2001, 22). Still other analysts find that merely the creation of a group can result in members becoming caught up in a genocidal dynamic (Fujii Reference Fujii2006), with this process heightened when identities are codified into formalized power-sharing arrangements (Mamdani Reference Mamdani2001).Footnote ²

The power of in-group/out-group identity is not restricted to an analysis of genocide, as the whimsical story about Nicole illustrates. Any discussion of group politics deals with such distinctions. The interesting point for us is the extent to which many of the most frequently discussed political cleavages in America—race, religion, ethnicity—assume immutability about the in-group/out-group distinctions and hostilities. Yet we all know that there is one human race and that—as the news media repeatedly reminded us during the 2008 presidential campaign—a man from Africa can wed a woman from Kansas to produce a child who can become president of the United States of America, just as a Catholic and a Jew can marry, or a German and a Japanese. (Biologically speaking, of course, the concept of “hybrid vigor” actually means that such so-called out-group breeding offers reproductive advantages in the form of minimizing shared genetic inheritance toward certain diseases.)

The critical question is why a particular in-group/out-group combination becomes politically and ethically salient. After an election year in which many of us hoped we could move beyond the traditional discussions of race in the United States to more substantive discussions of the meaning of such differences for all members of society, we might address such questions by thinking about both the theoretical foundation and the experimental support for social identity theory and remind ourselves that while group differences do exist independently of politics, the political and moral salience or these groups is totally socially constructed.

A further body of literature underlines our claim that discussions of differences in American political science need to move away from the existing paradigm, which assumes that existing political cleavages are somehow immutable, and ask instead why certain cleavages and differences are accorded ethical and political salience. The 2008 election should alert us to the extent to which such a conceptual move is critical if we are to recognize the extent to which the human psychology is far more complicated than many traditional political discussions of race might suggest. This literature focuses on the psychological dynamics driving coalitions and alliances.

Recent experimental evidence suggests the possibility that humans categorize race, at least in part, as a byproduct of an unrelated psychological mechanism designed to detect the nature of coalitions and alliances. In a clever study conducted by Robert Kurzban and colleagues (Kurzban, Tooby, and Cosmides Reference Kurzban, Tooby and Cosmides2001), subjects confronted a task where coalitional status was uncorrelated with race. Under such conditions, race effects were quickly “erased,” in the space of an hour's task, in favor of preferred memory for alliance structures. In this study, researchers used the “who said what” paradigm to show subjects an interaction between individuals of different races. These individuals displayed cross-cutting cleavages by wearing T-shirts of different colors designed to mark particular teams or groups. So, for example, white people wore both green and blue shirts, as did black people. When later asked who said what, subjects showed differential ability to notice and remember who said what based on the color of their shirts as opposed to their races, indicating the primary nature of the psychological coalitional-tracking mechanism. These researchers argued that because individuals were less likely to encounter others who looked differently from them in the ancestral past, racial categorization developed as a byproduct of a psychological process designed primarily to detect out-group members because it was so uncommon to encounter members of another race. When individuals did meet such people, they were most often out-group members, signaling their incorporation into a different alliance structure. When it becomes clear that race no longer indicates such coalitional status, race-based categorization can remarkably quickly remand to secondary status in attention and memory.

Another aspect of this study shows that this finding does not result merely as a consequence of clever experimental design. Rather, the Kurzban study also demonstrated that the effect of sex could not be erased in the same way. Gender continues to trump coalitional status regardless of manipulation, thus suggesting that encoding by sex remains a primary psychological adaptation (Kurzban, Tooby, and Cosmides Reference Kurzban, Tooby and Cosmides2001), most likely related to strong reproductive incentives. The implications of the Kurzban experiments for evaluating the impact of race and gender in the 2008 election are intriguing. Pollsters were surprised at the number of voters who admitted to racial prejudice, but nonetheless reported they were voting for Obama because they thought he would be the better person to get the country out of the current economic mess. Is this the “team” aspect that Kurzban, Tooby, and Cosmides found, where people use race as a code for strangers when they are not sure who might hurt them, but then drop their racist views once they conceptualize members of the other race as friends or team players who can potentially help them? And does the resilience of sex as a fundamental category of differentiation explain why more Democratic primary voters said they would not vote for a woman than for a man? The linkages between voting and the deep-seated psychological mechanisms underlying attitudes toward in-group/out-group members need to be examined more closely.

This returns us to the initial question that little Nicole picked up: the political relevance of categorization. The central nature of racial encoding as a marker for in-group versus out-group status results from particular theoretical models of categorization and learning. A classical model of behavioral conditioning, or learning, would argue that any conditioned stimulus—like race—can be associated with any unconditioned stimulus—like fear—simply through repeated associations between the two. Such linkages can be learned vicariously, by watching the reactions of others, such as one's parents. In this way, for example, Nicole can learn to react with outrage to out-groups who are faculty members in the biological sciences by witnessing her parents' reactions to them as “others” who are seen to be hogging scarce university resources. This kind of model not only assumes the equipotentiality that suggests that any two stimuli are equally likely to be associated, but also assumes that fear acquisition becomes easily linked to racial categories in the encoding of out-group status. But the suggestion that our psychological mechanisms of categorization and racial encoding may in fact not be so blunt derives from two interrelated lines of research.

First, Elizabeth Phelps has studied racial stereotyping, prejudice and discrimination using functional Magnetic Resonance Imaging (MRI) technology. The amygdala, the part of the brain associated with emotion in general, and fear in particular, does indeed tend to show greater response to faces of a different race than one's own (Hart et al. Reference Hart, Whalen, Shin, McInerney, Fischer and Rauch2000). In other words, it may not be the linkage between race and fear that causes prejudice and stereotyping, but rather the novelty of difference itself that precipitates a particular kind of reactivity. Indeed, there is evidence that the amygdala shows greater activation to out-group members in general, not just those from different racial groups (Johnson and Fredrickson Reference Johnson and Fredrickson2005; Cunningham et al. Reference Cunningham, Johnson, Raye, Gatenby, Gore and Banaji2004). In this way, faces of other races, because they are visually dissimilar, make some individuals more likely to respond to other racial groups with fear (Ohman and Mineka Reference Ohman and Mineka2001). Indeed, recent evidence indicating that conservatives have a stronger physical response to threat itself, as measured through galvanic skin response and strength of eye blinks in reaction to threatening stimuli, suggests some individuals may in fact be more genetically prepared to experience fear in the face of any threat (Oxley et al. Reference Oxley, Smith, Alford, Hibbing, Miller, Scalora, Hatemi and Hibbing2008). If this process does occur, then unlike the classical conditioning model presented above, racial bias could represent an overgeneralization of a kind of prepared fear response, whereby certain types of novel stimuli are more likely to be perceived as threatening.

Indeed, familiarity does appear to alter the amygdala's response to certain race's faces. For instance, white people do not show the same level of amygdala reaction to famous or familiar black faces, such as Bill Cosby, Tiger Woods, or now possibly Barack Obama. In addition, Phelps et al. (Reference Phelps, O'Connor, Gatenby, Gore, Grillon and Davis2001) finds that inter-racial dating reduces fear activation in the face of out-group members more broadly; when someone is dating a person of another race, they do not demonstrate the same degree of hostile or fearful responses that are generated among those with same-race partners.

A second line of research comes from Susan Fiske's work into person perception and stereotyping. Fiske finds that between 80 to 90% of how individuals respond to others derives from our assessment of two main characteristics: warmth and competence. Particular combinations and interactions between these characteristics result in predictable emotional reactions to individuals and groups. For our purposes, the relevant category emerges from the intersection of judgments that categorize relevant others as low warmth and high competence. This evaluation of others creates envy; the empirical work suggests that this category encompasses Jews, Asians, and rich professionals.

Fiske argues that while prejudice produces distinct brain activation patterns, behavioral manifestations of prejudice need not necessarily result (Fiske et al. Reference Fiske, Xu, Cuddy and Glick1999). Rather, social context affects the link between brain and behavior, just as the constituent elements of prejudice and discrimination remain socially defined (van den Bos et al. Reference van den Bos, McClure, Harris, Fiske and Cohen2007). Fiske and her colleagues make an argument that understanding prejudice requires making sense of the psychological processes by which people come to code and categorize potential threats from dissimilar others.

These lines of research, drawn from psychology, indicate that the human psychological architecture concerning reactions to race may be more complicated than ordinarily conceived in discussions of American politics. In particular, our reactions to out-group members may not derive solely from processes of social learning and acculturation, but also may build on psychological mechanisms designed to respond to out-group members in defensive ways. Racial categorization has often signaled, quickly and easily, such out-group status, and thus becomes an often inaccurate marker of such coalitional membership. Thus, race becomes entrained as part of the process of alliance categorization. Individuals who seek in-group supporters may exclude racial out-group members because of their improper calibration between racial and coalitional status in others. This does not make such a process right or correct, but rather offers an opportunity by which racial prejudice might be overcome by emphasizing the commonality in all humans, noting their shared endeavors toward justice, and emphasizing the political and moral salience of empathy and compassion, instead of hatred and division.

To be clear, the interaction of social and endogenous factors in developing racial stereotyping and discrimination does not render such processes any more immutable than locating its origins in processes of socialization alone. Rather, as with any complex phenomenon, the intersection of endogenous features with particular environmental cues introduces the possibility of intervention and amelioration along a wider variety of avenues (Hart et al. Reference Hart, Whalen, Shin, McInerney, Fischer and Rauch2000; Johnson and Fredrickson Reference Johnson and Fredrickson2005; Cunningham et al. Reference Cunningham, Johnson, Raye, Gatenby, Gore and Banaji2004; Fiske et al. Reference Fiske, Xu, Cuddy and Glick1999; Ohman and Mineka Reference Ohman and Mineka2001; Phelps et al. Reference Phelps, O'Connor, Gatenby, Gore, Grillon and Davis2001; van den Bos et al. Reference van den Bos, McClure, Harris, Fiske and Cohen2007). For those of us who care about ending racism and ethnic and religious prejudice, and believe we can best see the common humanity in others by broadening the boundaries of those included in the community of concern, the scientific evidence provides reason for hope.