Posted on December 18, 2023

Race Realism: A Moderate’s Manifesto

Bo Winegard, Aporia, June 29, 2023

Since at least the beginning of written history, people have understood that humans are a diverse species. Although some modern scholars contend that race is a relatively recent invention, the documentary record suggests that ancient Greeks and Egyptians recognized that human populations are different from each other in predictable and patterned ways. And natural philosophers have long wrestled with human diversity, trying to understand its causes, its organization, and its consequences.

During the Enlightenment, they elaborated and systematized once vague and inchoate ideas, giving birth to the modern concept of race. Since then, race has been a contentious idea, inspiring furious debate and morally charged dissent. Indeed, most modern intellectuals argue that race is unreal, a classificatory relic from a blinkered past that persists chiefly because it allows privileged groups to protect their unearned advantages.

Paradoxically, many of these same intellectuals argue that racially conscious policies and analysis are the only way to confront pervasive racial injustice. Race may be illusory, but it has bewitched so many people that breaking its spell requires promoting more racial awareness.

The view that race is a pernicious myth, that most ethnic inequalities are caused by racism, and that the way to combat racial injustice is to heighten racial consciousness is so widespread that it is now the orthodox position, and views which contradict it are often either objurgated or suppressed. They are rarely taken seriously or addressed charitably.

Although most people who accept this orthodox position have not considered its theoretical assumptions carefully, and some almost certainly profess allegiance to it merely to avoid social or professional sanction, it is a coherent world view, consisting, inter alia, of these propositions:

(1) Race is a social construct. Although humans vary genetically, they do not vary in predictable, patterned ways that can be classified as races.

(2) Human populations vary primarily only in superficial features, such as skin color, although they may vary in some underlying traits such as blood types or lactase persistence.

(3) Human populations do not vary in psychological traits and tendencies that were shaped by natural, sexual, or social selection.

(4) Human population disparities, which are large and numerous, are caused by pervasive racism.

The primary challenger to this modern orthodoxy is a (more sophisticated) version of the older view: race realismRace realism contends that racial categories carve out real and conceptually interesting variation in the biological world. It therefore argues that race is not an illusion or a distortion of human diversity. It is real just as sexes and species are real.

And because it asserts that race is a real, biological phenomenon caused by evolution, it also asserts that it is plausible that races vary not only in physical traits but also in psychological traits such as cognitive ability and self-control. The view that races differ from each other in psychological traits and tendencies at least partially because of genes is hereditarianism. Hereditarianism is a subset of race realism which consists roughly of the following propositions.

(1) Race is a biological phenomenon. Humans vary genetically in predictable, patterned ways that largely conform to popular racial categories.

(2) Races vary both in physiological and psychological traits and tendencies.

(3) Racial disparities, which are large and numerous, are caused by a combination of variables, including underlying differences in cognitive and behavioral repertoires.

In this article, I hope to convince the reader that not only is hereditarianism reasonable, but also that not taking it seriously is potentially calamitous because it allows implausible and pernicious narratives about pervasive racism to flourish. These narratives undermine confidence in important institutions and promote animosity against whites and other successful races.


Race is not (only) a Social Construct

Since at least the 1940s, intellectuals have raised objections to the concept of race, with more enthusiastic dissenters declaring it a dangerous and divisive myth that encourages invidious prejudices and ethnic factionalism (See, for example, MontaguLivingstone & DobzhanskySesardic, and Winegard, Winegard, & Anomaly for discussion). These dissenters were eventually so successful that their view, once an outlier, is now the mainstream view.

Before challenging their claims, it’s worth noting that a very common strategy among race skeptics is to demolish a caricatured version of “race,” insisting that since that version doesn’t exist, social constructionism is the only viable alternative. But this would be like claiming that since species are not eternal forms in the mind of God, they do not actually exist. Race realists do not believe that races share some kind of mysterious essence, nor do they believe that races are discrete and discontinuous. Instead, race realists posit only that racial categories pick out predictable and patterned biological variation. Race realism is no more committed to mysterious metaphysics than is species realism.

There are, of course, social constructionists who grapple with a more plausible version of race realism, and they have forwarded at least three serious challenges: (1) human variation is almost exclusively clinal; (2) human variation is not correlated; and (3) human variation between groups is very small compared to variation within groups. Unlike straw man arguments about racial essences, these challenges are reasonable, but either false or misleading. Furthermore, current evidence about the relation between genetic variation and self-identified race and ethnicity and the geography of one’s ancestors has refuted the claim that race is merely a social construction.

(1) Variation is predominantly clinal. The claim that human variation is almost wholly clinal and thus inconsistent with the traditional concept of race has been around for many years. In 1962, for example, Livingstone asserted, “There are no races, only clines” (p. 279) Modern scholars often repeat some version of this argument, stressing the gradual blending of human groups into one another. For just one of many available examples, in a recent textbook on physical anthropology that contends that race is not a biologically meaningful concept, the author wrote that because “biological traits generally follow a geographic continuum…,” humans “cannot be subdivided into racial categories.” The assertion that human variation is gradual is largely, though not entirely, correct. But the claim that this invalidates the concept of race does not follow.

Scientists classify many variables that are continuous into discrete categories, understanding that there is fuzziness on the boundaries and that although the categories are not arbitrary, the precise nature of their divisions is. Chronological age is continuous, but that does not mean that the categories “infant,” “adolescent,” and “adult” are “biologically meaningless.” In fact, they are quite meaningful, and, like other reasonable classifications, they provide inferential potential. That is, they increase our knowledge about the individuals who belong to them. If Tony said, “I’m in the car with an infant,” we would know (in a predictive sense) more about the passenger of the car than we would if Tony said, “I’m in the car with a person.”

Because the lines of each division of a continuous variable are somewhat arbitrary, there will be reasonable debate about where one category should end and another should begin. Reasonable people might argue that infancy should end at two and other reasonable people might argue that it should end at three. But these debates don’t vitiate the use of the categories altogether. And they don’t mean that the classifications are anarchical. No reasonable person would suggest that infancy should extend until a person is twelve years old.


Therefore, even if we imagined, arguendo, that human variation was entirely clinal, then it would simply mimic the many continuous variables that we divide into discrete categories for ease of understanding. And it would be socially and scientifically reasonable to categorize it into racial groupings. If, for example, humans from central Africa all the way to Northern Europe created a perfect cline (they do not), then it would not be distortive to divide that cline into different categories, since people at 51 degrees N latitude would be more similar to people at 48 degrees N latitude than they would be to people at 10 degrees N latitude. There would be fuzziness and debate about the boundaries, but the categories would not misrepresent reality.

However, the situation for social constructionists is even worse than this because human variation is not, in fact, entirely clinal. Geographic barriers such as the Sahara Desert and the Pacific Ocean have impeded the free movement of humans, resulting in some discontinuities. {snip}


These discontinuities gave rise to the patterned phenotypic variation that caused humans to notice race differences and to attempt to classify them in the first place. And the classifications they devised were not arbitrary or motivated chiefly by racial animus. Rather, they were motivated by a sincere desire to understand human variation after noticing that human populations are slightly but predictably different from each other.

(2) Variation is not correlated and categories are arbitrary. Jared Diamond, the eloquent and wide-ranging scholar, forwarded a popular version of this argument, writing that “There are many different, equally valid procedures for defining races, and those different procedures yield very different classifications.” So, for example, we might think that a classification based on skin color is objective and indisputable.

But if we chose a different variable such as antimalarial genes, our once obvious and unassailable categories would be upended and rearranged. Swedes would be “grouped with Xhosas but not with Italians or Greeks.” Race, therefore, is a factitious byproduct of our lust for classification with deleterious consequences for minorities, and it should be rejected.

But this would be like claiming that current biological taxonomies are arbitrary or discardable because one could, if one wished, formulate a scheme in which bats and blue jays (flying animals) are in one group and rats and ostriches are in another (non-flying animals). Biological classification is generally not based on one characteristic; and those who contend that race is a real category rely upon evolutionary history, genetic profiles, and phenotypic profiles, inter alia, to make their categorizations, not skin color alone. This does not usually result in obscure or esoteric classifications; and, in fact, the resulting classifications largely correspond to popular intuitions about racial categories because genetic variation, evolutionary history, and phenotypic variation are related to each other. Phenotypic variation reflects underlying genetic variation which in turn was caused by evolution (among other forces).

(3) Human variation between groups is small compared to within. The geneticist Richard Lewontin forwarded the most influential version of this argument in a 1972 article, in which he wrote, “Since such racial classification is now seen to be of virtually no genetic or taxonomic significance either, no justification can be offered for its continuance.” Some variant of this argument remains a virtual catechism among those who deny the importance or validity of racial classification. And although it might sound superficially compelling, the leap from the assertion that variation is greater within than between racial groups to the assertion that race is of no genetic or taxonomic significance is logically unjustified and empirically erroneous.

In a 2003 article, A. W. F. Edwards dubbed this argument “Lewontin’s fallacy,” noting that “the information that distinguishes populations is hidden in the correlated structure of the data.” In other words, it is the pattern, the correlated differences, that are genetically and taxonomically important and that allow various methods of analysis to predict self-identified race/ethnicity from genetic data with remarkable accuracy. For a simile that’s not too misleading, racial variation is like a melody: It is the structure of the notes, not the notes themselves, that is important. Four version of the same melody in different keys are easily recognizable so long as one hears enough of the melody. The significance of race, like the significance of a melody, is in the pattern.


Races Differ Psychologically

Although the orthodoxy is stridently opposed to race realism, it does not deny that human populations (the term it uses instead of race) vary physically. The differences in skin color, hair texture, facial structure, and other physical traits are simply too conspicuous to reject. However, it often denies that human populations vary psychologically, and it vehemently denies that they vary psychologically because of genes. This is an important distinction. Theorists can forward claims about human psychological differences without widespread disapprobation so long as they include the comforting assurance that such differences are almost entirely culturally caused. Hereditarianism must be rejected.

And so it is. Often with a fleet of moral accusations and (plausible) threats of perpetual unemployment, at least in academia. But although these tactics are effective at stifling debate, they do not change reality. The theory and evidence for hereditarianism is strong and growing by the year. Race is real. And race differences in psychological propensities and cognitive repertoires are likely common and heritable, with consequences that are not easy to dismiss.

The best studied race difference is the difference in cognitive ability as measured by various IQ tests and other related instruments. Although there are a variety of such differences across groups, the one that compels the most attention is the difference between Blacks and Whites, especially in the United States. Currently, Blacks and Whites are separated by about 15 IQ points or roughly a standard deviation. As Earl Hunt wrote in his mainstream textbook on intelligence, “There is some variation in the results, but not a great deal. The African American means [on intelligence tests] are about one standard deviation unit…below the White means…”

These differences showed up soon after the invention of reliable tests of cognitive ability and have remained relatively, though not perfectly, stable across time. They are, one might say, intransigent. Optimistic scholars often point to a narrowing of the B-W gap caused by Black gains in the second half of the 20th century, but such gains appear to have paused, and the gap remains quite large.

Of course, the mere existence of a phenotypic difference between groups does not confirm genetic causality. And scholars have debated the etiology of race differences in IQ, often with great fervor, since they became widely known. The two broadest causal theories are called environmentalism and hereditarianism. Environmentalism claims that genes account for little or none of between-group differences (20% or less), while hereditarianism, as previously noted, claims that genes account for significant proportion of between-group differences (21% or above).

Hereditarianism is supported by many lines of argument, including high within-group heritability, transracial adoption studies, persistence of differences, Spearman’s hypothesis, tests of measurement invariance, admixture studies, different proportions of target alleles in mental abilities, and even MRI studies. Some of these are persuasive; some are only suggestive. But all support the hereditarian position that a not insignificant proportion of group variation (20% or more) is caused by differences in genes. Here, I will focus on high heritability and admixture studies, which are possibly the two strongest lines of evidence.


A commonly forwarded environmental variable presumed to account for the IQ gap is socioeconomic status. Data from a variety of studies indicate that the SES (a composite variable that considers parental income, occupation, and educational level) gap between Whites and Blacks is roughly = .66. Thus, as Russell Warne writes, “these socioeconomic status differences account for only 7.3% to 47.6% of the environmental variance” required to explain the gap if the between-group heritability is zero (i.e., is not partially genetic). And this is being excessively charitable since the causal relation between SES and IQ is almost certainly from IQ to SES more than it is from SES to IQ. In other words, the environmental difference in SES is causally related to IQ because IQ causes lower SES and not because SES causes lower (or higher) IQ.

When we encounter highly heritable traits that reliably differ between groups and appear largely unaffected by environmental intervention, it is reasonable to believe that the group differences are at least partially genetic in origin. {snip}

{snip} Admixture studies are more direct and compelling. These use populations whose genetic ancestry varies (i.e., whose genetic ancestry is from different areas of world) to relate the proportions of admixture to outcome variables such as intelligence. The main idea is that individuals who have higher proportion of admixture from an ancestral group should more closely resemble that group than individuals with a lower proportion of admixture. For a non-human example, if a population of dogs hybridized with grey wolves, we would predict that the dogs with higher proportions of admixture (wolf ancestry) would more closely resemble the grey wolves than those with lower proportions of admixture.

In the United States, Hispanics and Blacks are both highly admixed. Hispanic ancestry is 55-70% European (the rest is from the Americas and Africa); and Black ancestry is 15-25% European (the rest is almost completely from Africa). Hereditarianism predicts that since the intelligence of Europeans is higher than Africans or Natives Americans, Blacks and Hispanics with higher levels of European ancestry should have, on average, higher intelligence (as measured by IQ tests).

The best studies have indeed found a positive correlation (r = .23 to .30) between intelligence and European ancestry, consistent with hereditarian predictions. These studies are all limited in various ways, of course, but they do address a common objection: Perhaps the European admixture is correlated with intelligence because it is correlated with skin color; and skin color elicits varying degrees of discrimination. People who have more European ancestry are, on average, lighter skinned; therefore, they face less intense racism than those with less European ancestry (who are darker skinned, on average). In Jordan Lasker and colleagues’ study, for example, they imputed skin color from genetic data and found that “Skin color was not significantly related to cognitive ability in any of our models which included genetic ancestry.”


Race realism is discomfiting because it contradicts contemporary dogmas about human equality. Races, like individuals, are not the same. And because they are not the same, they will create slightly different cultures and have different social outcomes. In the West, vast disparities in income, wealth, and crime are almost exclusively attributed to ubiquitous racism, but race realism rejects this narrative as implausible and divisive. Systemic racism in the West, although often invoked as a nearly omnipotent force, is virtually non-existent. In fact, one of the rare, obvious examples of systemic racism is affirmative action, which actively favors Blacks and other “victims” of oppression in the United States.