7

Do Races Exist?

In chapter 5 I suggested that the contemporary American conception of race contains certain morally troubling features independent of its use in specifically racist contexts: exaggerated difference and moral distance between those of different races, which discourages an experienced sense of common humanity; an overdrawn and falsely grounded sense of commonality among members of the same race;1 a notion of being trapped in one’s racial fate; encouragement of stereotyping racial others rather than seeing them as individuals; and an implication that, because of their racial membership, some persons have greater worth or ability than others.

In chapter 6 I traced the history of how these features came to be present in classic nineteenth-century racial thought. We can assume neither that contemporary ideas of race are simply carryovers from that period, nor that racial thinking has entirely jettisoned its nineteenth-century forms.2 What remains, then, is to pin down precisely which elements have been preserved and which abandoned or qualified. Once we have done so, we can address the vital question whether there actually are races, and, if so, whether the groups we currently call “races” are actually races.

Contemporary popular racial thinking follows classic racial thought in most respects. It divides the human species into several large ancestral groupings linked historically to certain geographical regions, marked by alleged somatic commonality, and constituting natural, discrete, fixed, and fundamental subdivisions of humankind that can be ranked as superior and inferior in important respects. There seem, however, three major differences between contemporary and classic notions of race. First, racial beliefs are today seldom explicitly avowed. In the nineteenth century almost all white people were explicit believers in race; in contemporary terminology, they were avowed racists. I have argued that blacks were avowed racialists as well.

A second difference is a general softening of the content of racial beliefs. Characteristics currently attributed to the nonwhite races are less noxious and inferiorizing than their nineteenth-century counterparts. Then blacks were regarded as uncivilized and virtually subhuman—“humanoid but not fully human” in Charles Mills’s apt phrase.3 They were seen as like both beasts and children—in neither case capable of directing their own lives, and requiring whites to do this for them. These figurations did not vanish with slavery but continued, revitalized, in the Jim Crow era, and not only in the South. Blacks are still viewed as inferior in certain respects—as violent, lacking in motivation, intellectually deficient—but, except in the far extremes of racism, seldom in such degrading and vicious ways as in classic racial ideology.

In the nineteenth century, Asians (primarily Chinese) were thought of as unfit for self-government, devious, potentially treacherous. The latter two stereotypes have not disappeared, but contemporary racial thought, though still carrying implications of irrevocable alienness, is much less vicious and insulting in its portrayal. Many racial stereotypes of Asian Americans—as high achievers, having strong families, being good in math and science—are at least partially flattering, even if they come with all the moral harms of stereotyping in general,4 and often with negative counterparts as well (“geeky,” lacking independence, not good at humanities subjects).

The same softening applies to the belief in racial hierarchy. In the nineteenth century, for almost all white people, the distance in worth and ability between whites and all other racial groups was quite vast, with blacks generally the least capable, least human, least worthy. Though contemporary racial thinking still carries implications of unequal value and capacity, the degree of inequality has greatly diminished, thanks to challenges by science and by the Civil Rights and Black Power movements, and by the movements for equality of other ethnic and racial groups (Latinos, Native Americans) that were inspired by the latter.5

Inherentism

A third important difference between contemporary and classic racial thought concerns the claim of “inherency”—that certain traits of mind, character, and temperament are inescapably part of a racial group’s “nature” and hence define its racial fate.6 The mindset of racial inherency is that blacks just are intellectually inferior, Asians studious, whites racist, Mexicans lazy, and so on. Classic racial thought located that inherent nature in the groups’ biological endowment, and so may he called “biological inherentism.” Twentieth-century science has waged a multidisciplinary attack on biological inherentism, and contemporary racial thinking has followed mainstream science in largely abandoning it.7 Contemporary racial thought has not, however, abandoned inherentism itself—only its definitively biologistic forms.8

Inherentism can take a qualified or unqualified form. In its unqualified form the characteristic attributed to the group is viewed as present in every member of the group. Qualified inherentism, by contrast, involves attributing a characteristic to a group but allowing that some members do not possess the characteristic; the latter are seen as exceptions, as not typical of the group. For instance, Laverne might think that white people are basically racist but that a few manage to escape this characteristic. The implication that racism is somehow rooted in white people’s nature is what makes her thinking inherentist. Qualified inherentism is similar to one form of what in chapter 1 was called “selective racism.” There the young black male was thought of as “really black” and the black grandmother was thought of as somehow less black. Here the nonracist white person is thought of as “less white” or “less exemplifying of whiteness,” though still acknowledged, in a purely classificatory sense, to be white.

Qualified inherentism must be distinguished from a different intellectual stance with which it might be confused, but which is not inherentist at all. That stance is that on the average a given group possesses a certain characteristic, but those who do not possess the characteristic are no less representative or typical of the group than those who do. For example, Tony says “Asian Americans are, on the average, better at math than other groups” but he in no way implies that Asian Americans who are not good at math are less fully Asian American, or less representative of the overall character of Asian Americans. Tony may well not have adequate evidence for making this generalization; he may, indeed, be drawing on stereotypes of Asian Americans that others hold in an inherentist fashion. Nevertheless, his stance is not itself inherentist; it postulates no natural or necessary link between being Asian Americans and being good at math.

To understand contemporary inherentism, we must first consider the difference between race and culture. Generally, to attribute characteristics to a group’s culture is a way of saying that the characteristics are not inherent, for cultures are human creations and potentially changeable. Mexican American culture may place a strong emphasis, value, and reliance on families, but over time this emphasis could diminish.

Changeable cultural characteristics may be attributed even to a racial group.9 Consider the views of William E. Cross Jr., a theorist of black adolescent identity development.10 Cross plausibly argues that many young blacks develop an “alienated oppositional identity,” a cultural orientation detrimental to their acquiring the skills to negotiate a successful relationship with the white-dominated world. This cultural orientation is rooted in norms and expectations that, Cross argues, are an understandable outgrowth of a history of racism, of a racially segregated existence, of a community in which the young people’s parents and other adults are often unemployed and in which little hope for their own future appears on their horizon. The alienated oppositional identity is not a mere bad habit. It pervades the norms and modes of life of the black youngsters who possess it, which is why it is appropriate to call it “cultural.”

Cross regards this identity as changeable (if not easily). He argues that, in the long run, job programs in urban neighborhoods can help to reverse the hopelessness brought about by parental unemployment that he sees as a source of the oppositional identity, and that, in the short-run, school-based programs recognizing the problem can help change this cultural feature. Cross’s view is clearly not inherentist; he recognizes social and structural features that produce the culture, and looks to the possibility of changing it.

Cultural Inherentism

The retreat from biological inherentism in contemporary racial thought has led some persons to a view one might call “cultural inherentism.” The culturally inherentist mindset is, roughly, “These people (Jews, whites, Asians) just are that way (stingy, racist, studious); it’s part of their culture.” Cultural inherentism is more socially and intellectually acceptable than biological inherentism, which is tainted by association with classic racial thought.11

To be sure, cultural inherentism is not intellectually coherent, since cultural characteristics are not inherent; and it is unlikely to survive sustained reflection by someone who manifests an unacknowledged belief in it. Nevertheless, in popular thought, some racial inherentism survives in a culturalist form.

We must therefore distinguish between cultural inherentism and cultural noninherentism. Only the former is characteristic of racial thinking; the latter, even when one regards the group to which it is applied as a racial group, as Cross does, is not a racial way of thinking about that group’s alleged characteristics.

Some observers, noting the contemporary waning of biologistic racism have suggested that “race” has now come to mean “culture,” which sometimes includes religion, language, or ethnicity.12 Though the word “race” is indeed sometimes used in these ways, the usage undermines any attempt to decide whether a particular usage of the word has drifted so far from its agreed-upon earlier meanings that it has come to constitute a new or different concept. Although there can be no pretense of locating a single true meaning of “race,” I would suggest that inherentist understandings of culture preserve the idea central to the moral and social significance that the notion of “race” has always possessed, and that is of characteristics that are fixed and unchangeable because they are part of a group’s nature.13 Noninherentist notions of culture, which attribute explicitly changeable characteristics, are entirely different.14

One suspects, as with the conceptual inflation of “racism” discussed earlier, that some of the impetus to expand the scope of reference of “race” is a desire to garner for religious, ethnic, cultural, and national groups the sort of moral and political attention generally associated with racially inferiorized or demonized groups.15 An individual can, however, give up her religion and change her nationality (neither easy to do, of course), but she can not change her race. (She can “pass,” but that is a different matter.) Ethnic groups can change their cultural characteristics, as French-Canadians, for example, became much less religious in the 1960s and after.16 Linguistic groups are defined by first language, but are not typically thought of as possessing other distinctive characteristics. Race differs from all these other group-defining characteristics in being an inescapable identity and in implying a range of other inherent characteristics. Of course ethnic, linguistic, religious, and national groups can be the target of discrimination, vilification, inferiorization, and demonization. But this just shows that such depredations are not confined to racial groups, not that those groups are racial.

As we saw in chapter 6, a distinct strand in nineteenth-century European racial thought attributed inherent characteristics to something like the spirit of a people or race. We can call this “metaphysical inherentism,” as it posits an entity which can be understood in a more or less religious way as the source of the allegedly inherent characteristics. This metaphysical source is distinct from a genetic or a cultural one.17 Some current inherentism has something like this metaphysical form.

Uncommitted Inherentism

Biological, cultural, and metaphysical inherentism all postulate a distinct basis for attributing characteristics to a group. At the same time, there can be an “uncommitted” inherentism that takes the form of neither genetic, metaphysical, nor cultural inherentism. Here the subject attributes characteristics to a racial group, and gives every evidence of regarding these characteristics as inherent in the group. Miguel thinks, “That’s just the way these people are”—lazy, stingy, prone to drunkenness, good at dancing, exploitative, and the like. But Miguel has never thought through the implications of his views, and holds no definite opinion as to whether the group is “that way” because of culture, biology, or spirit. I suggest that much inherentist thinking takes this uncommitted form.18 People have soaked up inherentist ways of thinking about racial groups and may not be entirely aware of having done so. More generally, many persons maintain elements of racial thought without reflectively endorsing them. Not entirely aware of their inherentism, they experience no requirement to justify the belief that the racial group in question possesses the characteristics they attribute to it.

Thus uncommitted inherentism is psychologically unstable. If someone presses Miguel to say why he thinks the group is stingy or lazy, he would naturally seek to provide some sort of account—in culture, genetics, or metaphysics. Alternatively, however, he might find all these explanations implausible, and thus be led to question whether he has any basis for his inherentism at all.19

But an uncommitted inherentist might stick to his guns while either rejecting or remaining agnostic about geneticist, cultural, or metaphysical explanations. Miguel might say, “I don’t know what makes them lazy. Maybe it’s their genes; maybe it’s their culture; maybe it’s some metaphysical spirit. Maybe it’s none of these. I have no idea. What I know is, they are lazy.”

If Miguel is further pressed, not so much for a rational basis for his belief but simply for an account of how he arrived at it, he might cite his own experience. Of course it is illegitimate to infer from a small sample in one’s experience to an entire group; and, in any case, people’s interpretation of their own experience—for example, what they perceive as laziness in others—is itself quite often suspect, driven by the very stereotypic generalizations it purports to provide evidence for.20 Nevertheless, just as familiarly, people do engage in such unwarranted generalizing all the time.

So uncommitted inherentism tends to fall either into genetic, cultural, or metaphysical inherentism, or into an abandonment of inherentism entirely; but it can also persist in the face of self-reflection. An individual might even explicitly and sincerely disavow inherentist thinking, but slip back into it when her mental guard is down.21

The Nonexistence of Races: Somatic Variation and Genetic Commonality

Having examined what contemporary belief in races consists in, let us turn to the question whether the groups we call “races”—whites, blacks, Asians and Native Americans—are such. In one sense, the question seems absurd; these are precisely the groups we in the United States mean when we speak of races. But racial thinking requires that each group possess a set of distinct characteristics; if they do not possess these characteristics, they are not races.

One argument frequently given against the popular understanding of race is that populations have moved around so much that no race is “pure.”22 Africans and Europeans, for example, have so intermingled with one another that neither group is any longer a pure race. But this argument assumes that a group that is “pure” would count as a race—for example, that a group of pure (sub-Saharan) African descent would be a race. This concedes too much to race thinking. Races, in the common view, are not merely groups that share certain somatic characteristics and geographical ancestry that distinguish them from other groups; they are also assumed to differ in mental, emotional, and characterological traits expressed in the idea of a “fundamental division of humankind.” Unless the visible characteristics correspond to inherent psychological commonalities within the alleged race, and mark differences among alleged races, the groups defined by those visible characteristics are not races.23

Nevertheless, this point about racial mixture does tell against one component of the racial outlook—that the groups we call races can be sharply distinguished in their somatic characteristics. In fact, neither blacks, whites, Native Americans, nor Asians are sufficiently uniform in physical appearance to be fully distinguishable from the other groups; their differences are a matter of degree, not of kind. This lack of clear somatic boundaries around groups classified as different races is, indeed largely a product of the mixing of geographically defined races.

This is particularly obvious with regard to blacks. American and Caribbean blacks, especially the former, have a wide range of skin colors, hair textures, and size and shape of facial features. Many are of lighter skin than others classified as “white,” “Asian”, and “Native American” and are otherwise physically similar to whites and Native Americans. The reason for this variation is, of course, the peculiar system of classification used in the United States to define “black”—the so-called one-drop rule, according to which an individual with any degree of sub-Saharan African ancestry is counted as black, even if she has a predominance of European ancestry.24 In the context of the one-drop rule, racial coupling between those of European and African ancestry during the period of slavery (consisting primarily, but not only, in slave masters’ sexual relations with their female slaves), helped to create, in subsequent generations, a population of blacks with substantial variation in somatic characteristics. Blacks and Native Americans have also interbred for several centuries, with offspring generally counted as “black.”

Because of this wide variation, many blacks have been able to pass as whites, availing themselves of opportunities denied blacks and avoiding the stigma of black racial identity. If passing is possible—if a black can look white enough to be taken as white by whites—then blacks can not be uniformly physically distinct from whites. Hence, according to the common understanding of race, the group designated as “black” in the United States lacks one requirement for being a genuine race.25 And yet for most Americans the one group most clearly and consistently thought of (by others as well as themselves) as a race are blacks.26

Race, as we have frequently noted, is much more than classification by somatic characteristics. How would we determine if racial groups also differ in deeper characteristics inherent in their nature? One obvious place to look would be genetics. Are the groups we call “races” genetically distinct?

Genetic diversity within any racial group is far greater than genetic difference between races (that is, between the statistical averages of different races).27 Although there is not yet agreement on the precise percentile estimate of genetic commonality and difference, several respected studies agree that approximately 84–85 percent of all genetic variation in the entire human species can be found among a random sample of two people from the same tribal or ethnic group in the same geographical region. This extraordinary fact runs entirely counter to the idea that members of different races are fundamentally different from one another and fundamentally similar to those of the same race.

Of the approximately 15-percent genetic variation not found between “neighbors,” slightly less than half—in other words, 6–7 percent of the total average genetic variation among individuals—is distinctly racial in character.28 To put this another way, genetic differences between two random individuals of different races are only marginally greater than those between two individuals from the same racial group. Moreover, and particularly relevant to the idea of blacks as a race, there is more genetic diversity in sub-Saharan Africa than in the rest of the world combined.29

The amount of genetic material that two random individuals share is, according to one estimate, 99.8 percent of their total genetic material; of two random individuals of different races, the shared amount decreases by only .02 percent.30 So genetically human beings are vastly more alike than different. The groups we call “races” are, as far as all available evidence suggests, only marginally more genetically different from each other than any other random divisions of the human population; they are not races.

This does not mean that there are no statistically significant differences in genetic makeup between racial groups as conventionally understood. (Obviously, genes that control for somatic characteristics conventionally defining races differ significantly among racial groups, although, again, only by a matter of degree.) In addition, molecular polymorphisms are genetic characteristics that are present (or not) in a binary manner; a given individual either possesses the characteristic—for example, a particular blood group, immune type, or enzyme—or does not. Molecular polymorphisms, in contrast to characteristics such as eye color, height, or metabolic rate, are not a matter of degree. Such a blood group allele known as “Duffy Fya” is found in 6 percent of blacks, 42 percent of whites and 90 percent of Asians.31 This degree of racial group difference is significant. If there were a broad range of such genetic differences, they might bear on our ordinary understanding of race; as just mentioned, however, available evidence is that such genetic difference as is found in the Duffy blood group is not broad.

This difference in the presence of a blood group allele, by itself, lends no support to the idea of race for another reason as well: so far as we know, differences in such polymorphisms have no bearing on the mental and behavioral traits that inform our idea of “race.” Indeed, purely genetic differences—differences described at the level of genes and alleles with no link drawn to traits of character, mind, and temperament—are only tangentially related to popular understandings of “race.” Consider the following thought experiment. Imagine two geographically distinct populations numbering in the millions—one located in central Asia and one in Latin America. Research reveals that the two populations are genetically virtually identical. They contain the same proportion of genetic differences of all significant types. But because they have developed distinct cultures, the two groups are behaviorally and psychologically very different. One is very cooperative and nonviolent. The other is aggressive and warlike, constantly threatening its neighbors. Are these two groups of the same race? If race were simply a question of genetics, the answer would be obvious, but at the level of the popular notion of race, it is not obvious, since they are so different in behavior.

Now suppose we imagine a small genetic difference between the two groups—with respect to genes that determine somatic characteristics traditionally associated with race (skin color, hair texture, and so forth). Popular thought would now be much more likely to declare these distinct races, since they differ both somatically and behaviorally. But the judgment would not be based on the marginal genetic difference, which is too minimal to be of significance.

It is not possible to rule out a priori the possibility that some respect in which conventional racial groups do differ genetically will turn out to bear strongly on some important mental or behavioral characteristic; but the burden of proof is with the proponents of this view, and there is little in our current understandings of genetics to support such a view. Regarding this possibility, moreover, the racial group difference in allelic frequency in the Duffy blood group can be positively misleading; it is exceedingly unlikely that any psychologically or behaviorally interesting trait is explained by the presence of such an allele. No one has come remotely close to demonstrating a causal link between psychological traits and the sort of genetic material that has been found in significantly different proportions in different “racial” groups. This is true even of scientists who postulate a genetic underpinning to some humanly important trait, as Herrnstein and Murray do with regard to “intelligence” in The Bell Curve. They infer a genetic substratum, but have not in any way located or even attempted to locate it.32 If there is to be a link between genes and behavior, it will certainly not be in the form of a distinct behavioral or temperamental trait—moodiness, compassion, hyperactivity, laziness—being the direct expression of a distinct gene or allele.

Let us step back a moment and reflect on the bearing of genetics on the popular conception of race. It is relevant not because race is itself a genetic concept. Genetic differences unconnected to psychological traits have little relevance to the commonsense idea of race. It is relevant because racial thought is inherentist and the only plausible account of how psychological characteristics could be inherent is that they are genetically based. This is not to say that racial thinking distinctly affirms the genetic basis of allegedly inherent traits. Nineteenth century American pre-genetics racial thought did affirm a biological basis for traits but, as we have seen, contemporary racial thought is more frequently of a cultural, metaphysical, or uncommited—that is, nongenetic—inherentist form. If the deep, inherent differences racialism claims do exist, however, they can do so only if there is a genetic basis for them. This is why the evidence that the substantial genetic diversity that does exist in the human species as a whole is mostly present within any geographically located population group, but not between racial groups, is so striking.

The Racial Significance of Sickle-Cell Anemia

Statistically interesting differences in some genetically based characteristics among racial groups are not inconsistent with the overwhelming genetic similarity within the human species. A frequently cited difference concerns sickle-cell anemia (a breakdown, expressed in a sickle-shaped formation, in red blood cells). American blacks are much more prone to this disease than are whites of northern European extraction. The reason for this difference is apparently that in the processes of natural selection Africans and other residents of tropical climates have developed a gene that provides immunity to malaria (a tropical disease), but this gene causes sickle-cell anemia.33

Seeing this and other statistical correlations as “racial” in character, however, tends to impose a false commonality that distorts the significance of the correlations. In the sickle-cell case, the source of correlation is not racial but geographical (proximity to the equator). Persons of tropical African descent share this genetic tendency with Arabs of the Arabian peninsula and southern Indians, who would generally be regarded, in our familiar racial schema, as of different races. But the Xhosa tribe of Southern Africa is a nontropical group and does not carry the sickle-cell anti-malarial gene.34 It is not blacks per se but people with ancestors from tropical climates whose grouping is evolutionarily related to a propensity to sickle-cell anemia; and even in this group, by no means all carry the gene. Of course such statistical correlations should be utilized in relevant medical or other practical contexts; American doctors should be more alert to certain symptoms in their African American than in northern European-American patients. But doctors are misled if they see race as the causal factor in such conditions.

In the sickle-cell case the causal and evolutionary factors are known. Frequently, though, researchers are presented with a statistically interesting racial generalization in which the causal links are not known. For example, African Americans are almost twice as likely as whites to suffer from hypertension, or high blood pressure. Here the causes are likely to be a mixture of social (such as poorer health care and greater stress because of societal racism) and perhaps genetic ones. But if one is primed to think in racial terms, one may too quickly ignore the possible social causes, assuming the primacy of an underlying biological one.35

Focusing on genetic differences and similarities may mislead us about the character of human behavioral differences and their relation to the groupings we now call “races.” Notwithstanding the staggering advances in genetic understanding in the past century, represented by the final mapping of the chemical structure of the human genome in the year 2000, we actually know precious little about the relation between genetic features and the wide range of psychological and behavioral characteristics associated with races—intelligence (in its various modes), artistic propensity, industriousness, apathy, violence, secretiveness, compassion, sociability, trustworthiness, athletic ability, and so on.

Popular writing regarding genetics often falls into the error that has been called “geneticization.”36 Consider the following from a New York Times article about recent advances in human genomics: “By contrast with the tiny number of genes that make some people dark-skinned and doe-eyed, and others as pale as napkins, scientists say that traits like intelligence, artistic talent, and social skills are likely to be shaped by thousands, if not tens of thousands, of the 80,000 or so genes in the human genome, all working in complex combinatorial combination.”37 Though overall the article is making the valuable antirace point that all the genes for such important characteristics as intelligence are extremely unlikely to be transmitted in way that would allow for differentiation among races, the passage could easily be read to imply that characteristics such as intelligence and social skills are strongly genetically determined (even if in a complex way).38 But no such correlation has ever been demonstrated. It seems highly likely that human social skills for example, are strongly affected by culture and learning. The current fascination with genetics should not lead us to believe that the contribution of genetic structure to qualities of mind and psyche vastly outstrips that of culture, learning, habituation, and other processes of human social and cultural life.

Science and Race

Natural and social scientists in the twentieth century have disagreed about the scientific value of the concept of “race.” This disagreement cannot be settled by showing (as I have attempted to do) that the groups popularly called races do not possess the features required by the popular concept of race. For particular sciences may have reason to utilize “race” as a theoretical term that does not carry the associations that it carries in popular thought.

But if geneticists or anthropologists do craft a viable meaning for “race,” this will not necessarily tell us anything about whether races, in the ordinary understanding, exist. For example, population geneticists, drawing on an increasingly sophisticated understanding of the structure of DNA and of the processes of genetic transmission, have developed a notion of a “breeding population”—a relatively reproductively isolated human group that differs significantly from others in the frequency of certain genes—and some have proposed calling such a group a “race.”39 The choice of terminology has a historical basis within the biological sciences; “races” have been considered biologically significant subdivisions of the human species. And the definition accords with the assumption in evolutionary theory that breeding populations represent “the primary unit of analysis for evolutionary change.”40

Nevertheless, this definition contains an arbitrariness regarding racial classification that could sow confusion when brought into relation with popular understandings of race. First, genetic characteristics are frequently not “concordant”; that is, they do not vary together. So populations that differ with respect to one characteristic may well not differ with respect to another. For example, scholars of an earlier generation grouped African and New Guinean pygmies together on the basis of body shape, skin color, hair form, and nose shape; but later scholarship has indicated significant differences between the two groups in the hereditary components of their blood.41 Hence selecting distinct genetically significant characteristics as bases for classifying races will result in a hodge-podge of different classification schemes. For instance, if an antimalarial gene is the criterion, Swedes and Xhosas will be one race and Italians, Greeks, Saudi Arabians, and most sub-Saharan Africans another. Possession of the enzyme lactase (enabling digestion of lactose) would place Northern Europeans, North Indians, and west African Fulani in one race, with Japanese, Native Americans, and most other sub-Saharan Africans in another.42 There are many such possible classifications, depending on the genetic trait in question, and the resultant groupings almost never correspond with one another. Moreover, few such classifications are likely to generate groups that correspond to racial groups as traditionally understood.

Within scientific discourse this multiplicity of “racial” classification schemes may not be a problem; for particular purposes, each may contribute to scientific understanding. But such an understanding of race, a contextual one allowing for multiple racial categorizations, is far different from the popular one, in which races are understood as fundamental human subdivisions. Using the same term for these two very different beasts could lead to misunderstanding when the lay public “overhears” conversations among population geneticists.

A deeper problem for the idea of a race as a breeding population is the character and degree of reproductive isolation required for a group to count as a race. Audrey Smedley asks if we are to regard the inhabitants of a small town in Colorado in which a high percentage intermarry as a distinct race from those in a similar town in Vermont, because these are very reproductively isolated from one another.43 Are small, geographically localized ethnic minority groups that set up strong cultural barriers to interbreeding with the wider society—such as the Amish—to be counted as races?44

Finally, this definition of “race” may involve unwarranted assumptions, such as that each reproductively isolated group will in fact possess distinctive frequencies of genes and their allelic variants in relation to significant traits.45 Such an assumption seems implausible if the reproductive isolation is understood literally, as in Smedley’s above example. Such groups are unlikely to be genetically distinct from one another in any interesting way.

If this assumption is jettisoned and the definition of “race” is confined only to those reproductively isolated groups that do in fact possess distinctive gene pools for significant traits, many geographically isolated groups will not be races at all. And the ones that satisfy this condition will still be subject to the contextual multiplicity—the degree of arbitrariness involved in selecting genetic characteristics to count as “significant” for classificatory purposes—that renders this definition far distant from that of race in the ordinary sense.

Scientists in a given field, then, may define “race” in a manner that can be defended by criteria of theory construction in that field, but one must then look closely to see whether that definition is so distant from popular understandings as to provide nothing in the way of support for the idea that races as we ordinarily understand them exist.46

Science is an integral part of society, and scientific findings and concepts can make their way into popular discourse. For this reason, scientific attempts to confine a particular concept such as “race” to a usage internal to a given scientific community may be unsuccessful. The concept may be taken up in popular discourse in a manner that preserves its traditional popular associations—even if the associations are explicitly rejected in the scientific definition. The deeply rooted character of racial thinking in U.S. culture renders it likely that any scientific usage of “race” in the near future will be taken, at least to some significant extent, to support current popular understandings. Even if population geneticists say that races exist only in the sense of a genetically significant, reproductively isolated group, popular thought may appropriate that as affirming that “races exist” in the popular understanding of that idea. Moreover, scientists are not themselves immune from influence by popular thought. It is reasonable to worry that, despite attempts to provide a theoretical meaning for a concept entirely distinct from the popular one, scientists may not be able to shake those associations. They may therefore be led unwittingly to incorporate false popular understandings into their own discourse.

Not all scientists wish to remain aloof from popular concerns and discourse, in defining theoretical terms in a manner wholly distinct from popular usage. Some of them do want to weigh in, hoping that scientific understandings can clarify confusions or misunderstandings in popular thought. To do so requires an engagement with popular vocabularies and understandings. Elazar Barkan has shown that after World War I (and especially after the ascension of Hitler to power in 1933), a group of (primarily) anthropologists in the United States and biologists in Britain had a not insignificant impact in discrediting racial ideologies among the general public in their respective nations.47

Nevertheless, scientific discourse proferred in popular venues has sometimes served to confuse the political core of race, and has thereby masked its full range of moral implications in popular discourse. Audrey Smedley points out that in his attempt to undermine the nineteenth-century connection between biology and personality and cultural characteristics, the German American anthropologist Franz Boas seemed to regard the concept of race that he was attacking as a purely scientific idea. He reasoned, therefore, that decisive arguments against race’s scientific forms would deal a fatal blow to popular appropriation of the concept and the prejudices founded on it.48

What Boas missed, Smedley says, is the folk origins of the concept of race—the way it emerged as a rationalization of domination, exploitation, and slavery. This same error afflicts physiologist and geneticist Jared Diamond’s account of race. Diamond argues that traditional racial classifications, and indeed any classifications grounded in phenotype, are biologically of little or no value. Yet he explains the existence of the popular phenotype-based forms of racial classification as lying in “the body’s signals for differentiating attractive from unattractive sex partners, and for differentiating friend from foe.”49 Whether this account is plausible as an evolutionary explanation of the familiar differentiation of human populations along the lines of the phenotypic differences (skin color, hair texture, and the like), it certainly does not explain why popular thought attributes racial significance to these characteristics. It does not explain race, because race did not arise in human consciousness as a product of sexual selection, but as a way that European colonizers came to see Native Americans and Africans in the process of inferiorizing and dehumanizing them. Like Boas, Diamond sees race as fundamentally generated through processes internal to the development of scientific ideas, rather than as a political and moral idea given legitimacy through science.

Popular ideas of race, confused as they certainly are, remain in place not primarily because of scientific misunderstandings but through the weight of a racialized history and the current legacy of racial depredations. Science has a limited, though far from trivial, role to play in undermining some of the intellectual props of popular racial ideas. But seriously weakening popular attachment to race requires doing battle with the political foundations of race thinking; to that end the achievement of a more just and less racially divided society is the crucial task.