1 HIRSCHFELD, L. A.: RACE IN THE MAKING Introduction To people living in the contemporary world, racial thinking seems a conspicuous and ubiquitous aspect of everyday life. This doesn't seem all that surprising. It is widely believed that the notion of race derives directly from the spontaneous perception of physical variation that is diagnostic of the major partitions of humanity. It is also widely accepted that evidence of these major partitions is frequently encountered, given the rich diversity of modern societies. Thinking racially, then, seems to be directly tethered to unremarkable sensory experience. Still, the idea of race involves more than an awareness of surface difference. In particular, race is thought to capture less obvious variation in humanity. Racial thinking is not simply a catalogue of human difference; it also encompasses beliefs about the very nature of difference. When imaging the world in racial terms, people are also making judgments about the inner and the nonobvious. Some of these beliefs enjov the cachet of modern science (e.g., the genetic basis of the distribution of sickle-cell anemia), but most lack any rigorous footing outside of common sense. Race, of course, is more than a concept. It is a fundamental aspect of social life, one that plays a pivotal role in predicting differential access to resources. In a wide range of social formations---and under disparate conditions of social existence---wealth, power, and opportunity are distributed racially. Answers to questions about who is impoverished, in poor health, politically disenfranchised, or poorly motivated in school invariably mention race (at least in modern secular society). Crucially, race differences are seldom if ever the cause of impoverishment, ill health, political disenfranchisement, or poor academic motivation, nor are these social ills always consistently distributed along racial lines. Other aspects of socioeconomic structure (e.g., class, labor position, economic status, and cultural affiliation) often more appropriately specify the dimension along which such disadvantage lies--- and differences in an individual's or a group's position with regard to these dimensions clearly are often the basis for it. Thus, race does not explain disadvantage so much as explain it away by "distorting" our perception of other material relations (Winant 1994). Expectations about differences in human nature stand in a complex relation to the structural consequences attributed to them. Beliefs about inner racial qualities perpetuate discrimination. They also create compelling (though misguided) explanations for it. That is to say, racial thinking serves both to legitimize and to misrepresent the way power is wielded and opportunity apportioned. To some extent this means that race is a closed system of practice and thought: racial thinking serves as a rationalization for inequitable distribution associated with racial status. Not surprisingly, an idea that warrants so broad a range of expectations is complex. In addition, it incorporates some contradictory aspects. On the one hand, racial thinking often involves contrived taxonomies of difference directly linked to specific cultural, political, and economic traditions. At the same time, racial thinking also involves robust, seemingly self-evident, widely rehearsed, and highly shared beliefs about the meaning and nature of human difference. In this regard, the concept of race appears to vary little by time and place. A distinct disciplinary approach can be identified with each of these contradictory characteristics. Psychologists have examined the way in which the organization of human mental life both captures and creates racial thought. Anthropologists, historians, and other comparative scholars have tracked the emergence and the evolution of racial thinking as a principal political and economic fact of the modern world. These two disciplinary traditions remain largely independent of each other and are often mutually uninformative. This book has two principal goals: to reinterpret the tension between the universal and the specific in race and to provide an account of racial thinking that adequately captures both of these characteristic qualities. This enterprise takes two forms: specifying a model of racial thinking that is consistent with the insights of both disciplinary traditions and empirically exploring how culture and cognition shape the development of racial thinking in the child. I can recast this project as an attempt to answer several questions: Why is it so easy to racialize explanations of human behavior and potential? Why are failings (and successes) in human arenas as varied as the commercial, the cognitive, and the athletic so readily attributed to racial causes? Why is it difficult to see past racial accounts to more structural (e.g., political, economic, and even cultural) descriptions? Why does racial thinking stabilize so quickly and seemingly effortlessly in the minds of children? This book ventures, sometimes indirectly, to provide a framework in which to answer these questions. It is a first step in a jointly psychological and cultural account of why racial ideas come to be widely distributed, easily transmitted, and predictably transformed. In short, it represents the first moves toward an epidemiology of racial representations.1 Whence Race? In view of the remarkable capability of racial thinking to reconcile, explain, and obscure disadvantage, it is no surprise that there has been a major effort by social and biological scientists to understand its psychological, social, and biological sources. It is important, however, to keep in mind that this effort consists in explaining two distinct phenomena. On the one hand, there is racialist thinking---the conviction that human beings (and perhaps other species) can be partitioned into discrete kinds whose reproduction turns on natural processes such as inheritance. On the other hand, there is racist thinking---the widely shared belief that the value and potential of individuals follows from their membership in the various races of humankind. Despite the fact that this latter mode is clearly contingent on the racialist one, it is the tenacity and ubiquity of racism---typically registered and expressed through racial stereotypes and prejudice---that has drawn the sustained and intense interest of social scientists. Somewhat ironically, we now know that one thing that cannot explain a racialist mode of thought is the phenomenon of race itself. Humans are not discriminable into discrete, self-evident biological kinds. Race is "an ideological analysis of social relationships" (Guillaumin 1980, p. 59), not a category of the biological world. It is an artifact of human culture, not a reading of the natural environment. Nor does the natural environment provide reliable evidence for the ideological analysis that is ascribed to it. The idea that race has a natural basis cannot be inferred from human biological variation. Although there is now a considerable body of literature in biological anthropology demonstrating that race lacks a scientific basis (Marks 1995; Molnar 1992; Alland 1971; Brace 1964), at first blush the idea may seem counterintuitive. It is accordingly worth examining the claim in more detail. The strength of our intuitions aside, the human races of both common sense and science simply do not designate biologically interesting populations. Genetic variation between races (to the extent that there is a systematic way to determine racial membership---a tenuous assumption at best) is small compared to variation within racial categories (Net and Roychoudhury 1983; Latter 1980; Lewontin 1972). To claim that racial categories lack a scientific basis is not to say that there are no biological differences among humans, nor is it to claim that there are no biologically grounded differences in human external anatomy. It is merely to say that races as socially defined do not (even loosely) capture interesting clusters of these differences. In large part this follows from the fact that the races as socially defined do not pick out genuine reproductive populations. Difficulties arise even if the racialist claim is restricted to one about continuities in outward appearance (as opposed to a cluster of inner and outer qualities). At least since Boas, most anthropologists have accepted that whereas racial talk is about patterned variation in immutable external traits, racial classification tends to be both context specific and task specific. Summing up decades of anthropological research, Molnar (1992, p. 21) concludes: "The number of races and their boundaries remains a subject of [scientific] dispute partially because of the lack of agreement on which traits identify a person's racial identity. Just what constitutes a race is a difficult question to answer, because one's classification usually depends on the purpose of classification." Nor is this realization limited to anthropologists. In psychology, Klineberg (1935, p. 20) made much the same point six decades earlier: "Granting that physical traits are important, it remains to be decided which trait is most important for purposes of classification; it is clear that the classifications will not be the same if, let us say, skin color is selected by one anthropologist and the shape of the head by another. Since there appears to be no objective way of deciding which trait is preferred, there are a great many different racial classifications, all of them equally subjective and equally arbitrary." To complicate matters, commonsense systems of racial classification encompass beliefs about inner nature as well as outward appearance. In fact, race is commonsensically interesting precisely because putative physical resemblances are supposed to be emblematic of a host of other/ often nonobvious attributes, properties, and competencies. Frequently these involve beliefs about morally laden and valuative characteristics. Although scientists have speculated on the relationship between racial-category membership and such characteristics, more often formal accounts of race have attempted to ground the notion in less charged traits. Perhaps the most widely studied of the supposedly race-relevant biological traits are genetic differences in the agglutinative reaction of blood, the various ABO blood types. The discovery of these types at the turn of the century led many to believe that a scientific foundation for racial classification was at hand: to the extent that the relationship between specific local races (i.e., populations typically defined by some shared geographic origin) and certain blood types was determinable, a scientifically reliable measure of race was possible (Hirszfeld and Hirszfeld 1919; Snyder 1947). Ultimately, however, this line of speculation did not pan out. For one thing, there turns out to be little correlation between blood type and other racially relevant characteristics, such as skin color (Molnar 1992). In fact, there is little correlation between supposedly racially relevant blood types and racial categories. When scientists examined how the ABO blood-group allele is distributed across a large number of populations, it became evident that serum types tend to pick out groups that are racially incoherent (e.g., one cluster consists of an African population, three Asian populations, and one European population) (Lewontin 1987). Moreover, there is little reason to believe that the different bloodgroup genes typically associated with a particular "racial" population form a coherent marker. At least this appears to be the case under conditions of hybridization. Loehlin et al. (1973) assessed blood-group genes among U.S. blacks to determine levels of admixture of European genes. They wanted to know whether variations in admixture correlated with different levels of performance on intelligence tests. They found that they did not. But they also found that the various blood-group genes typically associated with particular racial populations were not inherited as a unit. That is to say, there is little intercorrelation among the blood-group genes even within a single population. Thus, in sharp contrast to the folk notion of racial "blood" (i.e., a unitary physical manifestation of racial heritage), racially relevant properties of blood are independently recombined in reproduction. There is no reason to believe that other traits or qualities putatively linked to race would be any more likely to cluster in inheritance or to map onto the culturally specified populations called races. There are several reasons for this. First, the notion of racial type has never been tied to a particular level of classification in zoology (e.g., species or variety) (Banton 1987), let alone a particular phenomenon. Consequently, it is difficult to imagine how it could be reliably mapped to any biologically grounded property. Second, although the folk notion of race brings to mind images of great genetic differences and great natural distance between groups, the reservoir of genetic variability in any species is actually small compared with what conspecifics share. The sexual reproduction of species-typical architecture could hardly occur otherwise (Tooby and Cosmides 1992). Third, as Loring Brace has long asserted, the argument that distinct human types emerged as responses to adaptive pressures makes little sense: "It has become apparent that the assumption that there is something significant in the association of traits in a single group of people is an assumption which obscures the factors influencing the occurrence and distribution of any single trait. Tlie most important thing for the analysis of human variation is the appreciation of the selective pressures wJiich have operated to influence tJie expression of each trait separately. Since in many important cases the crucial selective factors have no reference to ethnic or population boundaries, obviously an approach which takes population as its unit for study will fail to produce an understanding of whatever is influencing the distribution of the characteristics in question." (Brace 1964, p. 107) When nonbiological factors influence the distribution of a particular property, attribute, or competence, the problem is even more intractable. Consider the long debate about race and intelligence. In later chapters I will discuss the issue in some detail, but for the moment it is worth pointing out that mental ability, although significantly influenced by genetic factors, is clearly sensitive to environmental factors that disproportionately affect members of minority groups.2 These include factors directly linked to minority status (e.g., lower levels of academic preparation due to enforced racial segregation) as well as factors indirectly related to racial status (e.g., poorer health and lower birth weight, which are linked to race via poverty). Most dramatic, even when membership in a racial group is the most important influence, evidence shows that biological arguments simply do not hold. Ogbu (1990) has found that poor school achievement is associated less with minority status itself than with the particular cultural meaning that minority status has. The latter, he argues, derives not from the nature of the group but from the initial terms of the minority's incorporation into the host society. The same ethnic or racial group may excel or do poorly in school depending on the particular meaning group membership has in a particular cultural context. For example, in Japan Buraku ethnics (whose conditions of incorporation in Japanese society parallel those of blacks in North America in important respects) do poorly in school relative to other Japanese. When members of the same group immigrate to the United States and are viewed essentially as undifferentiated Asians, they do better in school than their majority counterparts.3 Studying Race In short, a simple bottom-up approach to the race concept will not do. The idea of race as a complex of beliefs about physical appearance and inner qualities cannot derive from perceptual analysis alone. All this renders race all the more interesting as a widely shared concept. If race is not a genuinely natural phenomenon (i.e., if races as they are socially defined do not pick out biologically interesting or even physically discrete populations), the question "Whence racial beliefs?" demands even more explanation. If people do not discern race, how is it that they come to think about it as they do? Researchers from a remarkably wide range of disciplines---anthropology, biology, history, literary criticism, paleontology, philosophy, political science, and sociology---have all made substantial contributions toward answering this question. Despite this diversity in scholarship, virtually all existing work on race falls under one of two approaches: an avowedly universalist one that interiorizes the source of racial beliefs and a decidedly comparative one that seeks to ground such beliefs in structural features of sociocultural formations. The first approach, adopted principally by psychologists, explores the underlying mental processes that give rise to racial cognitions and biases. The second, favored bv scholars in the comparative and interpretive disciplines (history, anthropology, historical sociology, philosophy, literature), approaches race as the product of specific social and historical forces.4 Each tradition of research provides plausible arguments about the conditions that underlie a belief about the existence of races. Each offers sound explanations for why these beliefs are highly salient. However, each presents claims that directly contradict those of the other. To illustrate these contradictions, let me sketch each approach. The Psychology of Racial Ideas The prevalent point of view in psychology is that racialist thought is a by-product of the way information is organized and processed. In contrast to the comparativists' emphasis on the specific content of racial thought, the psychological approach stresses the consequences of information processing for the way we make sense of intrahuman differences. The generic universalist story has two parts, one situated in the way object categories are generally formed and one situated in the way social categories gain substance. Object categories The signature property of human information processing is a well-developed capacity to sort objects into categories. Not only do these categories facilitate thought by reducing the sheer amount of information to which people need to attend; natural categories (i.e., those that bring together natural objects, such as cats, water, or gold) extend our knowledge of the world by capturing nonobvious similarities between their members. Once we recognize that a particular creature is a dog, we can infer that it shares with other dogs certain food preferences, certain sleeping habits, and an aversion to cats, even in the absence of opportunities to observe this particular dog eat, sleep, or interact with any feline. In recent years cognitive psychologists have shown that such category-based knowledge plays a fundamental role in the way we gain and organize knowledge about the natural environment. Social categories In view of the compelling interest we humans have in other humans, it is not surprising that people also possess a remarkable capacity to categorize and reason about others. A rich and attention-demanding quality of other humans is their elaborate mental life. Thus, it is unremarkable that we recognize commonalities derived from shared mental states and proclivities. There are friendly individuals, angry individuals, depressed individuals, and so on. In short, we classify and label people in terms of transient beliefs and emotions as well as in terms of more stable traits and dispositions.5 This capacity allows both children and adults to extend explanations of human behavior to unknown instances in unfamiliar circumstances. Knowing that humans are psychological creatures---that they act in accord with their beliefs and desires---allows us to predict what individuals will do in virtue of what we can infer about their beliefs and desires independent of the specific situations in which they find themselves. Humans also form knowledge of human types on the basis of outward appearance. Two sorts of appearance, gender and race, have been found to be particularly salient. Psychologists typically attribute this salience to two factors: the propensity to classify together objects that share conspicuous physical similarities and the fact that gender and race have prominent physical correlates. Like other natural categories, gender and racial categories seem to capture nonobvious similarities that can be recruited to extend knowledge. However, in contrast to the benign way other natural categories promote inference, these cognitive simplifications---usually called stereotypes---create patterns of belief that have undesirable social and political effects. Consider how one of these processes, illusory correlation, contributes to the formation of social prejudice. Hamilton and Gifford (1976) describe illusory correlation as erroneous inference about the relationship between two classes of events. According to their model, people overestimate the frequency with which distinctive events co-occur because infrequent events may demand more attention than frequent events, and therefore the conjunction of two infrequent events is even more noteworthy. It is plausible that a process of this sort underlies the formation of racial prejudice. When two things share a distinctive feature (such as being rare), people tend to perceive a correlation between them. Negative events are rarer than neutral or positive events, and minority individuals are rarer than majority individuals. Thus, people tend to attribute negative events or descriptions more frequently to minority-group members than to majority-group members. The content of the events and the specific nature of the minority status is unimportant. What counts is the relative frequency with which a particular type of event is encountered and the prevalence of a particular status in the social environment. Cultural, political, and other social factors influence the formation of racial prejudice by determining which events are rare and which are not. However, the motor that drives prejudice is psychological: it is the impulse to form categories and to calibrate their occurrence with other salient dimensions of the world. The Comparative Study of Race In contrast to psychologists' concern with the mental processes that give rise to racial thinking, historians and other comparativists have turned their attention to the social-structural antecedents of such thinking. Instead of focusing on mental and private embodiment, comparativists have been largely interested in public representations of and doctrines about race. Typically this means an examination of the discursive practices in which European states (and especially their colonial administrations) engaged (and continue to engage) when enumerating and controlling peoples of non-European descent. Race, on this view, is not the inadvertent by-product of the way information is processed; it is the contrived application of a post-Enlightenment impulse to systematize knowledge of natural phenomena in the service of the imperial enterprise. As such, race is not a discovery about the structure of nature but an invention inscribed onto it. Comparativists acknowledge that non-Europeans also think racially, but they interpret such practices as the adoption by colonized and formerly colonized peoples of a historically unique and European ideology about the nature of human difference. Scholars in this tradition do not deny that before the European imperial enterprise humans conceived of themselves and others as belonging to specific groups whose members shared a range of common and fundamental properties. They claim, however, that racial ideology stands apart from these antecedent forms because of the special power relationship with which the modern idea is invested and the special kind of commonality that supposedly is shared among members of the same race. The idea of race is a historically unique phenomenon, arising out of a singular confluence of cultural, political, and economic events and practices that defined a particular set of political relations (generally but not always involving colonized and colonizer). This range of relationships was justified in terms of inborn differences in moral, temperamental, and other essential potentials. Because systems of power differ and notions of potential vary, one of the fundamental lessons of historical and other comparative accounts is that racial thinking is not the same in all contexts. Instead it depends on the particular way domination is defined and the context in which is it transacted. Consequently, rather than a single entity "racial thinking," there are the various and contingent systems that have developed in different cultures and historical epochs. This emphasis on aggregate populations and systems of domination does not mean that comparativists have failed to recognize that racial relations of power are lived by individuals in the context of everyday practice. Nor does it deny that these power relations are interiorized in the individual's psychological experience. In fact, several influential works point out the crucial importance that interiorization plays in imperial racism (Mannoni 1964; Fanon 1968). But these studies differ from the psychological accounts described above in two respects. First, although historically specific systems of racial thinking are sometimes associated with specific mental experiences, such systems are seldom explained by psychological (and particularly cognitive) causes. Second, in comparison with psychological researchers who attribute the development of ideas about race and racism to biases in information processing, interpretive scholars who engage questions of interior states are much more likely to turn to the theories and the vocabulary of psychoanalysis than to those of cognitive psychology. In part this follows from the role that interiorization is supposed to play. For most comparativists, even those concerned with the psychological, the interior landscape provides an understanding of the motives and emotional consequences of racial thinking, not an account of how racial thinking came to exist (see, e.g., Adorno et al. 1950). An Alternative View: Race as a Specialised Belief System Clearly, attention to both information processing and historical factors contributes to our understanding of how it is that people understand race in the way they do. But by itself neither the historical approach nor the psychological approach adequately captures the complexity of racial thinking. Of particular interest is that neither approach alone is capable of answering one fundamental question about racial thought: Why is it that people so readily move from visual categories to inner nature? General psychological processes do not entail this, nor does the instrumental application of race in the service of unequal power and authority. I will argue that only by reconciling the study of mental and structural (i.e. economic, political, and cultural) factors can we advance an adequate answer to this question. Students of both approaches have provided important insights into the nature of racial thinking, and 1 acknowledge the wisdom of both traditions. However, I propose to rethink the two approaches in the following way: Race is indeed a unique sort of belief, in significant measure unlike any other commonsense notion. This uniqueness is due in part to the interaction of historical and cultural particulars, as comparative scholars have long contended. Nonetheless, as psychologists have long maintained, race is an extraordinarily widely encountered notion whose recurrence across varied cultural and historical landscapes derives from deep-rooted psychological processes. I will argue that the two crucial shortcomings of this previous work are the following: The specificity of racial thinking derives as much from the unique manner in which our conceptual system creates and harnesses knowledge of intrinsic human kinds as from historical and cultural conditions. The human conceptual system contributes to the creation of knowledge of human kinds less through the operation of general classifying skills bound to raw sensory perception than through the operation of a special- purpose conceptual device: a domain-specific competence for creating knowledge of and reasoning about human kinds. I will have considerably more to say about domain specificity in chapter 3. Briefly, a domain-specific competence is a cognitive structure dedicated to gaining, organizing, and using knowledge about a particular content area. Traditionally the mind has been conceived as a general-purpose problem solver whose operation can be studied largely independent of particular content. In recent years a sustained challenge to this view has emerged that proposes that the mind is a collection of more special- purpose tools, each targeting a specific problem or content. Commonsense knowledge of the physical world, of the biological world, and of the psychological world have all been interpreted as the products of domain-specific competences. Although each competence is the basis for a distinct body of knowledge, all domain-specific competences share common structural features. Each directs attention to certain sorts of data, specifies the existence of certain kinds of entities, and guides the formation of a certain range of hypotheses. Such devices perform two crucial functions: they simplify the task of imposing order on an often disorderly world and they provide the means to enrich and extend otherwise limited knowledge. I argue that racial knowledge too emerges out of the operation of a special-purpose device. But it is important to underscore what I am not proposing. I am not suggesting that humans possess a special-purpose cognitive competence that targets a racial domain or that is dedicated to the creation of racial cognition. In short, I am not suggesting that race is an innate concept. It would be curious if I were to make such a claim. As I remarked earlier, races are not genuine categories of the natural environment, so the idea that humans have a special-purpose device that targets a racial domain would be bizarre. Humans create races, they don't discover them. It remains an open question how long they have been creating them. It could be that human cognition is prepared to create specifically racial categories. Evolution has prepared humans to believe in the existence of a range of contrived categories, ranging from highly abstract ones, such as the language-specific categories of noun phrase and verb phrase (Pinker and Bloom 1990), to very concrete ones, such as folk species (Atran 1990; Berlin 1992). It is unlikely, however, that humans evolved to pick out what might be called pseudo-races because the historical conditions for discerning pseudo- races (if there are such things) are a fairly modern phenomenon and were absent for most of human history. Indeed, until relatively recently, with the advent of long- distance sea exploration, humans simply did not encounter groups of other humans who differed abruptly in their external race-relevant physical anatomy (van den Berghe 1981). Rather, humans appear to be ready to conceptualize the human world as composed of distinct types---what I call human kinds. A range of human kinds are possible. There are kinds based on common behavioral features, kinds predicated on common physical features, kinds predicated on common emotional characteristics, and so forth. The race concept, I suggest, emerges out of a notion of human kinds predicated on the attribution of common inherent or intrinsic features. Like other domain-specific devices, the human-kind competence is comprehensive. People not only create human kinds (in the sense of sorting individuals into relevant sorts), they assume that category members who might otherwise be dissimilar are fundamentally alike in their nonob-vious and basic natures. As we will see, this is a crucial point. The human-kind-creating competence delivers categories that do not embody difference; they interpret it. This feature contributes significantly to the structural consequences of race. Historical and cultural forces do not so much make race essential to the distribution of power and authority as they enlist race to legitimize that distribution. It is widely accepted that race legitimizes the inequitable distribution of resources by essentializing and naturalizing both similitude and difference. Paradoxically, race relations are seldom about race per sc; usually they are about the social organization of labor and commodities expressed through and justified by supposedly genuine and "natural" human kinds (Fields 1990). Race "discovers what other ideologies have to construct: an apparently 'natural' and universal basis in nature itself (Hall 1980, p. 342). The notion of the naturalness of embodied difference provides race with one of its most powerful conceptual features: race is both material and hidden. As a considerable body of literature by historians and other comparativists has demonstrated, a hallmark feature of beliefs about the nature of race is the expectation of commonalities that are not obvious from visual inspection alone. This turn of thought---that things that are physical may not be as they appear---makes it easy to think about race as fundamental, essential, and basic, on the one hand, and concrete, on the other. In significant measure, this book is about how cognition and culture intersect to construct this apparently natural, universal, and perceptual basis. NOTES 1. Sperber (1985, 1990, 1994) he-is outlined the general theoretical program for an epidemiology of representations. He proposes interpreting beliefs, particularly the widespread beliefs typically associated with culture, as the precipitate of cognitive endowment and constraints on communication and meaning. It may be that social structure and institutions can be recast as part of the distribution of ideas, thereby doing away with an ontology that includes them as elemental parts. I remain agnostic on the question, at least as regards the structural and institutional aspects ot racial thinking 2. The idea that intelligence is inherited is often associated with the notion that there is some unitary phenomenon corresponding to mental ability. There is little reason to believe that this is the case. It is more accurate to say that what is at issue is differences in performance on standardized tests. 3. Race not only lacks the biological basis common sense invests it with; it often lacks the sociological basis too. According to North American common sense, humans are segmentable into discrete groups on the basis of differences in phenotype. Inasmuch as the racially relevant aspects of phenotype are immutable, membership in these groups should not change during an individual's lifetime. However, racial identity is often more labile than this model implies. Researchers in health statistics, who otherwise are committed to the notion that racial categories represent distinct populations, have noted considerable inconsistency in racial identification. For instance, Hahn et al. (1992) found that individuals often are identified as belonging to one race at birth and another at death. Similarly, census followups have found that a nontrivial portion of the population identify as members of one racial group on the initial census report but as members of another racial group on subsequent interviews (Hahn and Stroup 1994). In part this follows from the fact that the processes involved in .self- identification are quite different from (and often produce less "stable" categorizations than) those underlying other- identification. The data-collection techniques used in the census and in other federal research generally do not take this into account. But inconsistencies in racial identification also result from the fact that race and racial identity are often less about membership in distinct groups (based either on common biology or sociology) than about how identity is fundamentally conceptualized and socially enacted. Race is less about groups in the world than about groups in individuals' minds. These conceptual representations underlie a range of practices and discourses. There are differences in the degree to which these practices and discourses are racialized; that is, differences in the degree to which practices and discourse are interpreted in racial terms. To some extent, then, it is incorrect to talk about race at all, even in quotation marks, if we intend by the term distinct and corporate groups of people. Rather we should talk about the way conduct and cognition are racialized. Still, that we can talk about race as a thing is a deeply held notion---indeed, it is almost impossible not to talk about it that wav. For this reason, I will continue to use the term in nominal form throughout the book. 4. In large measure both views evolved in response to an endemic racism permeating both scholarly and folk thinking. In academic thought this has taken two allied forms: an attempt to "scientifically" systematize commonsense intuitions about racial difference and a eugenicist impulse to theorize reproductive strategies with an eye toward "improving" racial stock. Although cloaked in the guise of natural science, both efforts served to justify the prevailing racial hierarchies of nineteenth- and twentieth-century Europe and North America. These were not isolated research activities, taking place on the fringes of the academy Any attempt to deny that racism has powerfully shaped the evolution of both psychology and anthropology risks missing the source and the significance of the countercurrents embodied in modern psychological and cultural theory. For the most part, however, I will not be concerned with this heritage of racism. Instead, I am interested in showing that, in contrast to this third (explicitly racist) view, which made (and continues to make) incredible claims about the innate inferiority (typically) of non-Europeans, both the psychological approach and comparative approach have considerable merit. 5. Although there is dispute as to whether all peoples equally use this capacity to organize trait and disposition knowledge in order to explain everyday behavior, there is no evidence that would provide doubt that everyone can form such categories.