1.1 What Is Race?

A natural launching point for an exploration of whether race exists is to ask what race is. That is, what are we talking about when we talk about race? As is standard in philosophy, we can begin by probing what we ordinarily mean by race, that is, how race is used in our everyday discourse. The ordinary concept of race is meant to capture what we commonly mean when we use the word “race” and its cognates. Consider the following statements:

These statements involve what we might call race talk. They refer to categories – black, white, Asian – that we, competent (English) language users engaging in our everyday mainstream discursive practices, take to be racial. The ordinary concept of race captures much of what this mainstream race talk is pointing toward. There are two desiderata for an analysis of the ordinary concept of race. It needs to capture much of the common sense or traditional racial beliefs that prevail in most societies. And it needs to be compatible with a wide range of theories of race, from essentialism to anti-realism. That is, the ordinary concept of race is defined such that various metaphysical positions in the debate are prima facie talking about the same thing.Footnote ³ Hardimon’s (Reference Hardimon2003) account of the ordinary concept of race fulfills both these desiderata.Footnote ⁴ Of course, as we will see in the coming sections, whether or not the race concept necessarily entails substantive metaphysical commitments is open to debate.

Hardimon (Reference Hardimon2003) identifies the “logical core” of the ordinary concept of race to consist of three premises. First, Races are groups of human beings. These groups of humans are distinguishable from other groups of humans by “visible physical features of the relevant kind” (Hardimon Reference Hardimon2003, 442). The relevant visible physical features that distinguish between races are (at least superficially) biological. They should be salient and easily identifiable by the naked eye. The most common candidates for these distinguishing features are skin color, eye shape, nose shape, and lip form, among others. These traits are phenotypes, which are the expressed or manifested result of (as we now know) genetic and developmental mechanisms. Racial groups need not differ from one another in all these features, nor will all members of a racial group have all and only those features associated with their race.

Second, members of a racial group are linked by common ancestry. Different theories of race will vary in how they draw the connection between race and ancestry. Nonetheless, the ordinary concept of race is “something inherited, a property transmitted from generation to generation” (Hardimon Reference Hardimon2003, 446). If two individuals of the same race, R, have biological offspring, that offspring is also a member of R by virtue of his or her ancestry. This lineage can be traced back to a founder group. What makes a lineage racial is that the founder groups are distinguishable from one another by visible physical features which are in turn retained by their descendants. As such, two groups whose members have the same visible physical features but do not descend from the same common founder group are not of the same race.

The third, and final, thesis is that each race is associated with a geographic location. Intuitively, racial groups have distinct geographic origins. Continents, given their size and relative isolation, play the role of specific geographic locations associated with each race: black from (Sub-Saharan) Africa, white from Europe, Asians from (East) Asia, and so on. This fact grounds an important feature of the ordinary concept of race, namely, that races were geographically isolated from one another before the emergence of sophisticated forms of transport. This separation led to reproductive isolation and precipitated the emergence of the racial traits that differentiate racial groups. Hardimon notes we can interpret the visible physical traits that differentiate races as adaptations to local geographic conditions (Hardimon Reference Hardimon2003, 448). However, pre-Darwinian theories of race posited their own explanations of the cause of racial traits (see Section 1.1).

The ordinary concept of race is compatible with a wide range of different conceptions of race. Following Hardimon (Reference Hardimon2003) and Glasgow (Reference Glasgow2009), I draw a distinction throughout this Element between a concept of race and a conception of race. As Hardimon (Reference Hardimon2003, 439) writes,

Glasgow (Reference Glasgow2009) likens the concept/conception distinction to one between meanings and theories (Glasgow Reference Glasgow2009, 23). A concept picks out the subject of one’s discussion whereas a conception is a fleshed-out theory about that subject. For instance, it belongs to the concept of humans, among other things, that humans are mammals. If someone were to point to some X and (literally) claim “this reptile is a human,” they have made an error as to the subject of their speech. Whereas a conception of humans can hold that humans were directly created by God and have an immortal soul; an alternative conception may hold humans are part of an evolutionary lineage stretching back millions of years. These are two rival conceptions of the same concept. They are talking about the same subject but have different theories as to that subject. That is, if two people disagree on their concepts, they are talking about different things. If they disagree on their conceptions of the same concept, they are disagreeing about what features or properties the concept has.

Armed with the concept/conception distinction, we can see the ordinary concept should be conceived of as thin. It does not, for instance, commit to the view that races have essences (i.e., that races have “fixed and immutable” properties) or that race membership is biologically significant across many dimensions. The ordinary concept of race is also silent on whether there is a racial hierarchy and whether moral and intellectual capacities vary by race. The ordinary race concept does, however, purport to refer to something biological. The visible physical features that distinguish races are the product of biological processes. Furthermore, both ancestry and geography are directly or indirectly taken to play a biologically significant role in, at the least, the emergence and perpetuation of (racial) visible morphological traits such as skin color.

However, as Glasgow (Reference Glasgow2009) notes, philosophers of race have proposed other, thicker, accounts of the (ordinary) concept of race. Thicker accounts of the concept of race add conditions to the “logical core” that go beyond what Hardimon (Reference Hardimon2003) identifies. For instance, holding that the concept of race entails the racial purity of the majority of humans (Zack Reference Zack1993); that race is ordinarily understood as both biological and sociopolitical (Outlaw Reference Outlaw1996); or that races are ordinarily taken to have biobehavioral essences (Hirschfeld Reference Hirschfeld1996), among others. Whether or not these views are correct requires detailed philosophical analysis and experimental evidence drawing on the social science of perceptions of race (Glasgow Reference Glasgow2009). Nonetheless, I will rely on Hardimon’s (Reference Hardimon2003) thin conception in subsequent sections because it is fairly intuitive and allows us to examine the widest range of conceptions of race.

To begin (near) the beginning of modern race theorizing, let us first consider the essentialist conception of race.

1.2 Race Essentialism

Race essentialismFootnote ⁵ is a conception of race that holds that race is a natural, fixed, and immutable division of humans into discreet groups. On the essentialist view, race divides humans into biobehaviorally distinct groups. The differences between races are responsible for the social, cultural, and intellectual disparities among racial groups. Consequently, race essentialism often assumes a racial hierarchy. Essentialism is a discredited theory of race with virtually no proponents within the wider scientific and philosophical community working on issues intersecting with race. Nonetheless, an exploration of race essentialism is useful on two counts. First, many ordinary racial ideas were developed at the same time that essentialism about race was dominant (Appiah Reference Appiah, Appiah and Gutmann1996; Hardimon Reference Hardimon2003). Understanding many of the suppositions that come with race talk is not possible except in the context of the emergence of the race concept and race essentialism in the eighteenth century (whether or not it may be possible to liberate it from that context). Second, race essentialism, although false, possesses many features of a conceptual analysis for a genuine kind that sheds light on the shortcomings of latter realist, but nonessentialist, views of race. That is, race essentialism conceives of race as an explanatorily rich category, one that is robustly connected to explaining cultural, psychological, physiological, and other phenomena (see Section 2.6).

An essentialistFootnote ⁶ conception of race has three defining conditionsFootnote ⁷:

1. (1) Humans can be subdivided into (ordinary) races.

2. (2) Individuals from different (ordinary) races are more different from one another than individuals of the same (ordinary) race.

3. (3) Differences between the (ordinary) races explain differences in behavior, culture, psychological traits, and health and vigor.

Race essentialism, at least in its modern form, traces to the emerging taxonomic practices of the eighteenth century. Carl Linnaeus (1707–1778), in his groundbreaking Systema Naturae (1735), provided one of the first systematic taxonomies of human variation. The Linnaean system classified humans into four variants: Europaeus albescens, Americanus rubescens, Asiaticus fuscus, and Africanus nigriculus. Linnaeus took these to be varieties of humans (“homo variat”). The association between human variation and skin color implied by his naming convention – yellowish Asians, blackish Africans, and so on – remained a prominent element of subsequent racial classification. The early editions of the Systema, a work he continually revised, only included skin color and geography as the relevant differences between races. Furthermore, races were not conceived as subspecies, a type of subclassification Linnaeus deployed in the case of plants and animals, but rather as variants of humans.

The tenth edition of the Systema expanded the section on the human genus. In this edition, however, races were identified not only with skin color and geography but also with other physical, moral, and cultural attributes. The four races, corresponding to four continents, were also associated with the four humors (phlegm, blood, black bile, and yellow bile) whose specific proportions in the body were taken to explain temperament and disease. To skin color, other physical features such as hair texture, eye shape, and lip form were added. Additionally, Linnaeus attributed traits ranging from aesthetic judgment to preference for forms of government to racial differences (Müller-Wille Reference Müller-Wille2014).

For instance, Homo sapiens europaeus is white, sanguine, has straight, yellow hair, is inventive, and governed by rites. Homo sapiens americanus is red, choleric, has thick black hair, cheerful of disposition, and is governed by customary right. And so on for each race. With the tenth edition, then, we see much more explanatory significance given to race. Not only is race merely a matter of geography and some climate-related traits, but it is also connected to a host of factors including behavior, health, and mental and cultural capacities and practices. This characterization of race and human variation deeply influenced subsequent developments in anthropology and human biology (Marks Reference Marks2007).

Kenan Malik (Reference Malik1996) and Justin E. H. Smith (Reference Smith2015) crucially highlight the fact that the history of the development of racial theory is not uniform. The contestation of eighteenth-century Enlightenment values played a role in shaping early race theorizing. For some, common humanity, the universality of reason, and belief in progress found expression in their race theorizing in defenses of monogenismFootnote ⁸ and a limited conception of racial difference that attributed variation between races to differences in climate, diet, and lifestyle. No figure better represents the tension between the Enlightenment commitments to universal humanity and the essentialist theorizing of race of the nineteenth and twentieth centuries than Johann Blumenbach.

Johann Friedrich Blumenbach (1752–1840), a leading figure in the development of anthropology, proposed perhaps the most influential racial classification system. Blumenbach’s classification had two major differences from Linnaeus’. First, Blumenbach argued skull shape and size are the defining characteristics of races. Second, Blumenbach’s classification scheme identified five, rather than four, races: Caucasian, Mongolian, Ethiopian, American, and Malay. Both the five-race schema and the use of race categories such as “Caucasian” were widely adopted in subsequent racial theories (Malik Reference Malik1996). Although skeletal craniotomy became a favored tool in the arsenal of essentialist race realism, Blumenbach rejected both polygenism and racial theories that posit deep racial differences (Smith Reference Smith2015). For Blumenbach, racial boundaries are arbitrary and “whatever we might ascribe to racial difference can be accounted for almost exclusively in terms of climate, and color in particular is so superficial that it can easily change over the course of an individual’s life, either through changes in diet or climate, or simply through internal changes in the life cycle, of the same sort as we see in the graying of hair” (Smith Reference Smith2015, 256). For Blumenbach, race is ultimately inconsequential when considered against the naturally more determinative division between species.

Blumenbach’s work represents a path untraveled in the development of racial science. His limited approach was eclipsed by the far more dominant essentialist theories of race. In the history of western philosophy, no less prominent figures than David Hume and Immanuel Kant were proponents of race essentialism. Hume rejected monogenism and embraced a polygenic and hierarchical race theory (Malik Reference Malik1996, 53). Hume held racial differences to be substantial and to be caused by not only physical but also moral factors such as culture, custom, and psychology (Zack Reference Zack2018, 12 f). Furthermore, races are hierarchically placed in relation to several capacities including intelligence.

Immanuel Kant’s writings on human diversity are broadly representative of the emerging essentialist conception that dominated nineteenth-century race theory. Kant outlined in his anthropology a view of race that combined the defining conditions of race essentialism. Unlike Hume, Kant was a monogenist. Nevertheless, Kant took race to be an intermediate category between species and variation (Kleingeld Reference Kleingeld2007, 578). In his 1775 essay On the Different Races of Man, Kant defines races as “those which, when transplanted (displaced to other areas), maintain themselves over protracted generations, and which also generate hybrid young whenever they interbreed with other deviations of the same stockFootnote ⁹” (Kant 1775, 39). The “deviations” that are definitional of races are heritable. They perpetuate even if members of a race migrate away from their geographic regions. And “inter-breeding” between “deviations” produces “half-breeds” with mixed characteristics. The difference between individuals of the same race, on the other hand, is merely variation.Footnote ¹⁰

In an illustrative passage, Kant writes:

Kant’s views in this passage capture condition (ii) of essentialist races. Although there is variation between individuals of the same race, these are not significant. For instance, some white people have blue while others have brown eyes. Notable deviations occur between individuals of different races. This interracial difference is so great that interbreeding between races produces “half-breed” children, while interbreeding between variants of the same race does not require such designation.

Furthermore, Kant’s race theory is hierarchicalFootnote ¹¹ and holds that differences in nonphysical properties such as intelligence, culture, temperament, and so on are a consequence of racial differences. In a hallmark of race essentialist accounts, Kant draws an explanatory relation between climate, geography, physical features, and psychological and behavioral traits. Of the “negro” race, for instance, Kant writes “ … the Negro is … well suited to his climate; that is, strong, fleshy, supple, but in the midst of the bountiful provision of his motherland lazy, soft and dawdling” (46). The tension between commitment to both the contingency of the origins of (racial) physical human diversity and the absence of even the possibility of intellectual achievement among those Kant considers racially inferior is a curious aspect of Kant’s thought. It suggests at the very least that racism played a significant role in Kant’s racial theorizing (Smith Reference Smith2015, 235).

The racial theories of Kant and Hume did not go unchallenged by contemporaries. Johann Gottfried von Herder (1744–1803) rejected Kant’s racialized and hierarchical understanding of human difference. As mentioned above, many enlightenment figures emphasized common humanity and the ability of all humans to reason. Herder, a leading figure of the Romantic reaction to the Enlightenment, emphasized culturalist and national differences to defend the dignity of different “races.” For Herder, reason is embodied in and manifests through culture. And all human cultures manifest reason in their particular way. As such, there is no independent (which is to say, Eurocentric) ground from which to dismiss the capacities and way of life of nonwhite people (Smith Reference Smith2015, 231–233). However, later adopters of the Herder’s approach integrated it seamlessly with the dominant racial and racist thought of the nineteenth century. As Malik (Reference Malik1996) notes, “once it was accepted that different peoples were motivated by particular sentiments, unique to themselves, it was but a short step to view these differences as racial” (Malik Reference Malik1996, 79).

The nineteenth century was the era of what became known as “scientific racism.” Race theories from this period moved further away from some of the shared assumptions of enlightenment figures writing about race and human differences. The belief in common humanity was de-emphasized or outright rejected. Hierarchy became a defining element of racial classification extending beyond what many eighteenth-century theorists proposed (Malik Reference Malik1996). Polygenism gained wider acceptance. Race theory and racism became inextricably linked. The fact that the nineteenth century saw the most intense period of European colonialism and the exploitation of black slave labor in the United States, the Caribbean, and Latin America played a large role in determining which race theories found wide purchase (Smith Reference Smith2015, 229 f). As Zack (Reference Zack2018) argues, racial theories that dominated Western societies in this period served to uphold the social order and licensing practices such as slavery and white supremacy (see Section 3).

Race essentialism had its heyday in the period between the early nineteenth century and the Second World War. The race essentialist underpinnings of Nazi ideology were a death knell for the view in the postwar period. In an influential Statement on Race issued by the newly formed UNESCO in 1950, a distinguished panel of biologists and social scientists were charged with addressing race and racism in light of the events leading up to the Second World War and its aftermath (UNESCO 1950). The Statement recognized that humans can be biologically subdivided into smaller groups, and these groups can be considered races. Nonetheless, the Statement rejected the claim that temperament, personality, character, and other innate mental characteristics were racial traits. Furthermore, developments in human biology and genetics made some claims that were associated with or integral to race essentialism, such as polygenism or racialized psychology, untenable (see Section 2). Nonetheless, a complex set of questions about the relationship between race, biology, and genetics still remained to be answered. In the next section, I discuss contemporary scientific work on human diversity and discuss a range of views defending the biological genuineness of race, namely, biological racial realism.

2.1 Varieties of Biological Racial Realism

Biological racial realismFootnote ¹⁴ holds that races are biologically meaningful categories that capture or correspond to at least one structure of human diversity. There are many ways that race can be biological – genetic, cladistic, ecotypical – depending on what factors are taken to be the biological basis of race. Nonetheless, biological racial realists take the biological basis of race to be more than the physical features (e.g., skin color) that are ordinarily taken to be racial.

Next, let us examine proposals by Robin Andreasen and Phillip Kitcher to defend the biological reality of race on the grounds that races are isolated breeding populations.

2.2 Races as Breeding Populations

Andreasen (Reference Andreasen1998, Reference Andreasen2000, Reference Andreasen2004) defends a cladistic theory of race. The cladistic race concept defines races by means of a single property: shared genealogy. Cladistics is a system of classification that categorizes taxa solely based on genealogy. As such, according to the cladistic race concept, races are clades. One race is distinct from another not in virtue of genetic variation or differences in visible biophysical features (although they may vary along these dimensions). Rather, races are classified on the basis of their distinct genealogies. Andreasen defends a view where “races are ancestor-descendant sequences of breeding populations that share a common origin” (Andreasen Reference Andreasen2004, 425), where breeding populations are “a set of local populations that are reproductively connected to one another and [are] reasonably reproductively isolated from other such sets” (426). Races become distinct when populations branch out and, for whatever reason, become reproductively isolated. This reproductive isolation leads to genetic distance as the separate populations undergo distinct evolutionary pathways (Andreasen Reference Andreasen1998, Reference Andreasen2000, Reference Andreasen2004).

Andreasen constructs a human phylogenetic tree to represent the branching (and isolation) of human breeding populations that form the cladistic races. Consistent with the Out of Africa thesis, the trunk of the tree represents modern humans in Africa some 200 thousand years ago. The tips of the trees are current breeding populations (i.e., cladistic races), with each branching point representing the formation of a new breeding population. On this basis, Andreasen identifies nine cladistic races: New Guinea and Australia, Pacific Islander, Southeast Asian, Northeast Asian, Arctic Northeast Asian, Amerindian, European, Non-European Caucasoid (Andreasen Reference Andreasen2004, 458).

It is important to note that cladistic races are dynamic. That is, it is possible that after mass migration and the breakdown of isolating factors, cladistic races come to cease to exist. If two initially isolated breeding populations begin to interbreed, then they no longer constitute different cladistic races. It is also possible that new forms of isolation emerge that generate branching forming novel cladistic races. Suppose for instance that humans establish a suitably large colony on Mars but, for some reason, this colony is cut off from Earth. If there are no barriers to interbreeding within the Martian colony, “Martian” would constitute a new cladistic race of the human species.

Andreasen argues that there are several virtues of the cladistic account. First, cladistics provides a biological conception of race. After all, evolutionary biology is centrally concerned with reproductive relations, genealogy, and genetic distance. A cladistic account defines race with the same conceptual tools used to define speciation, a core explanatory target of evolutionary biology. Second, cladistic race also avoids two major challenges to a biological conception of race: the genetic and independent variation arguments.

The “genetic argument” against biological race holds that since commonsense racial categories do not track significant patterns of human genetic diversity, they are not biologically meaningful. As discussed in Section 2, Variation among individuals within the same population represents the overwhelming majority of human genetic diversity. Of the remaining fraction of human genetic diversity, variation between individuals of different (racial) genetic clusters is only slightly greater than variation between unrelated individuals within (racial) genetic clusters.Footnote ¹⁵ As such, race is not meaningfully biological. Andreasen (Reference Andreasen2004) argues cladistic race concept sidesteps the genetic argument since it does not conceive of race as a genetic category. Races are defined solely through genealogy, regardless of what share of genetic diversity (if any) is between breeding populations constituting cladistic races.

The “independent variation argument” denies the biological reality of races by targeting the lack of coherence between racial categories and their defining racial traits. That is, traits such as skin color vary independently from traits such as hair texture, and so on. While one or two traits may provide an unambiguous racial classification scheme, the more traits are used the more cross-variance occurs. Consequently, proponents of the independent variation argument maintain that “because there is no non-arbitrary way to choose one classification scheme over another, we ought to abandon biological racial classification altogether” (Andreasen Reference Andreasen2004, 428). Once again, the cladistic race concept is able to avoid this challenge. Biophysical traits play no role in the delineation of cladistic races. As a result, cladistic races do not correspond to the ordinary, or “folk,” racial groups such as “black, Asian, and white.” The aim of the cladistic account is not to license the biological reality of existing racial categories. Rather, it is meant to rescue race as a biological category by severing it from the often incoherent and biased forms of folk racial classification.

However, the supposed virtues of the cladistic account in blocking challenges to the biological reality of race bring into question whether cladistic race is a conception or theory of race at all.Footnote ¹⁶ As Hardimon (Reference Hardimon2003) argues, the logical core of the ordinary race concept, even the relatively thin account he defends, involves more than mere genealogy. Andreasen (Reference Andreasen2004) contends the divergence between cladistic and “common sense” racial categories is not a problem for the cladistic view. After all, as Andreasen (Reference Andreasen2004) notes, “there has always been widespread disagreement over how many racial categories there are and who belongs to what category” (Andreasen Reference Andreasen2004, 437). The existence of multiple incompatible racial classification schemes does not invalidate the fact that any given conception is a conception of race.

Nevertheless, the divergence between the cladistic and ordinary race concepts remains objectionable because the cladistic concept fails to match any of the dominant conceptions of race (Glasgow Reference Glasgow2003, 460). While common ancestry is an intuitively central aspect of the race concept, physical resemblance also plays a central role. If, As Andreasen contends, it is possible for different cladistic breeding populations to be physically indistinguishable from one another, it is hard to see how they can be different races. It seems then that Andreasen is not vindicating the biological reality of races but defending an alternative approach to apportioning humans into smaller groups.

Furthermore, Spencer (Reference Spencer2012) argues that the cladistic race concept has not been shown to provide explanatory or predictive power in cladistics (Spencer Reference Spencer2012, 203). The cladistic race concept has not caught on in the research programs investigating human diversity and population structure and the nine cladistic races have found little purchase within the scientific or broader community. There is a good reason why genetics is a flashpoint of debates around the biological viability of race. And that is because genetic diversity captures a lot of what we care about in broader debates about the usefulness of race as a biological category in, for instance, medicine and psychology. The cladistic account does not connect race to these areas of central concern.

Kitcher (Reference Kitcher and Harris1999, Reference Kitcher2007) defends a biological conception of race that draws on his broader pragmatic approach to science.Footnote ¹⁷ Kitcher argues that it is possible to defend an account of race as a biological subdivision of humans. However, this racial delineation is not a purely biological matter. Crucially, whether a biological conception of race ought to be deployed will partially depend on whether it serves a value we (well-informed, deliberatively democratic agents) endorse in a way that exceeds the costs of racialization.

Kitcher defends a biological account of race that he argues answers to these requirements. On Kitcher’s account, as in Andreasen’s, races are isolated breeding populations. Kitcher grounds his analysis in ordinary biological practice. As noted by Andreasen, evolutionary biologists are centrally concerned with speciation. Kitcher argues an obvious way to begin conceptualizing biological race is to first approach biological species concepts. The dominant biological account of species holds that species are clusters of populations that (1) would freely interbreed in the wild and (2) are separated from other interbreeding clusters by reproductive isolation (Kitcher Reference Kitcher2007, 295). Reproductive isolation is sometimes the result of geographic isolation, but it need not be. Co-located clusters can be reproductively isolated by behavioral, morphological, or temporal factors (e.g., nocturnal vs. diurnal cycles).

Additionally, as part of this general account of species, it is possible to divide species into local varieties or “races.” These subspecies are the result of reduced interbreeding. That is, within a subspecies “it is considerably more probable that members of the population will mate with one another than with outsiders” (296). Subspecies are often distinguishable from one another by differences in phenotypic traits that arose as a result of generations of isolated inbreeding. These subspecies are races. A race on Kitcher’s account is “an inbred lineage, where the inbreeding may initially have resulted from geographical isolation that eventually gives rise to differences in phenotype and to some interference in free interbreeding, even when the geographical isolation is overcome” (Kitcher Reference Kitcher2007, 296).

Kitcher’s account differs from Andreasen’s on two points. First, on Kitcher’s conception phenotypic or genetic distinctness does play a role in racial classification. Second, races can be maintained even in the aftermath of human mass migration between different geographic regions. Kitcher argues this is due to the fact that social factors now play the role of isolating mechanisms once played by geographic separation. In the United States, social and cultural factors sustain racial inbreeding and prevent racial coupling and intermarriage (Kitcher Reference Kitcher and Harris1999, 98). What maintains races in this context is that “these people belong to a different race because they were once labeled – mistakenly, ignorantly, unreasonably – as intrinsically different, for that initial labeling has given rise to the separation of their way of life from that of the labelers” (Kitcher Reference Kitcher2007, 298). As such, race is both socially constructed and biologically real. And it is the social construction that keeps race biologically real.

Kitcher’s defense of the biological reality of race is open to the same criticism as Andreasen’s. Namely, it is not clear that the groupings identified by Kitcher’s account are races as ordinarily understood (Glasgow Reference Glasgow2003). Kitcher’s account goes beyond trimming and adjustment that concepts may undergo as part of a rigorous analysis. For one, the number of populations that have reduced levels of interbreeding far exceeds the number of races proposed by any dominant conception of race. Observable differences in the distribution of allele frequencies occur between dozens of populations that have reduced interbreeding due to, for instance, linguistic barriersFootnote ¹⁸ (Hellwege et al. Reference Hellwege, Keaton and Giri2017). Additionally, as Glasgow (Reference Glasgow2003) notes, in so far as the conception of race Kitcher defends rests on reproductive isolation, different social classes with social barriers to inbreeding may qualify as races (Glasgow Reference Glasgow2003, 470). Once again, there is a divergence between the conception proposed and the concept of race as ordinarily conceived.

Andreasen (Reference Andreasen2005) counters Glasgow’s (Reference Glasgow2003) objection to the cladistic race concept (and by extension to Kitcher’s conception as well). Andreasen (Reference Andreasen2005, 98) argues that Glasgow’s (Reference Glasgow2003) objection rests on a selective interpretation of the ordinary concept of race. Andreasen notes that historically, some racial classification schemes have relied principally on ancestry or genealogy at the expense of morphology. For instance, the one drop rule in the United States prioritizes ancestry in racializing individuals and groups. Furthermore, a study of the history of race conceptions shows that at one time or another anywhere from five to eight racial categories have been recognized in, for instance, the United States (Andreasen Reference Andreasen2005, 100 f).

2.3 Races as Ecotypes

Massimo Pigliucci and Jonathan Kaplan (Reference Pigliucci and Kaplan2003) argue that races are locally adapted ecotypes. Ecotypes are the result of selection pressure for ecologically important traits. The concept of ecotypes was initially developed to capture the genetic adaptation of plants (the latter also applied to animals) to their local environmental conditions (Pigliucci and Kaplan Reference Pigliucci and Kaplan2003, 1163). For instance, skin color is an ecologically significant adaptation. Lighter skin tones are adaptive in areas with low sunlight where the population is consuming a diet low in vitamin D (Pigliucci and Kaplan Reference Pigliucci and Kaplan2003, 1168). The crucial fact about ecotypes is that “adaptive genetic differentiation can be maintained between populations by natural selection even where there is significant gene flow between the populations” (1165). It is also possible for ecotypical traits to emerge independently in reproductively isolated populations undergoing the same selection pressure. Therefore, Pigliucci and Kaplan’s account of race defends the biological basis of race while rejecting the view that races are phylogenetically distinct (contra Andreasen and Kitcher) or undergoing speciation.

However, as Pigliucci and Kaplan note, the ecotype concept of race has “little or nothing” to do with folk racial categories (1161). There are potentially dozens of ecotypes (i.e., local adaptations of smaller populations) including height, bone density, lung capacity, and so on. On the ecotype conception, these would constitute distinct races. Yet these traits are not commonly taken to be racial nor would the greater number of groups that fall under the ecotype conception be consistent with ordinary race. Of course, one can maintain that “race” has a “neutral” meaning that refers to the biological subdivisions of a species. Nonetheless, given the social and historical significance that “race” has, there is a high justificatory threshold to using “race” for a revisionist concept that is quite distinct from the ordinary usage.

So far, we have discussed biological conceptions of race which in one way or another diverge from the ordinary concept of race. Next, I discuss accounts of biological racial realism that are consistent with the ordinary race concept. This view has both minimalist and robust variants, with the view of races as genetic kinds the paradigmatic example of the latter.

2.4 Races as Genetic Kinds

Neven Sesardic (Reference Sesardic2000, Reference Sesardic2010) defends a robust genetic view where races are distinct genetic groupings of human populations. Sesardic (Reference Sesardic2010) disputes the characterization of biological race realism used by its critics in order to dispute the biological reality of race. Sesardic admits that the overwhelming consensus among scientists (geneticists, anthropologists, etc.) as well as philosophers is that races are not biologically meaningful. However, Sesardic argues strong, essentialist assumptions are not necessary to ground a biological notion of race. If reasonable assumptions are made about what kinds of biological categories would count as “racial,” then races can be “rescued” from debunking. Races in his account are not social or political, “racial recognition is not actually based on a single trait (like skin color) but rather on a number of characteristics that are to a certain extent concordant and that jointly make the classification not only possible but fairly reliable as well” (Sesardic Reference Sesardic2010, 155). Sesardic claims that multifocal genetic clustering (that is, methodologies such as the one implemented by Rosenberg et al. Reference Rosenberg, Pritchard and Weber2002) reveals that race is a genetic category. Furthermore, genetic variation between races is what explains variation in hereditary traits, including psychological traits such as intelligence.

The robust view of biological racial realism has few defenders. It is out of step with the scientific consensus on the nature of human diversity (Long, Li, and Healy Reference Long, Li and Healy2009; Rosenberg Reference Rosenberg2011; Biddanda, Rice, and Novembre Reference Biddanda, Rice and Novembre2020). It also draws on a contested view of the role of heredity in the development of complex traits. Sesardic’s attempt to rescue folk races through the identification of races with genetic clusters misapplies the race concept (Hochman Reference Hochman2013). For one, if we specified races as populations with the least in group genetic variation, the resulting (hundreds of) clusters would fail to match any dominant conception of race or one that could reasonably be conceived of as one that picks out the ordinary concept. As Taylor notes, “if we delineated groups that had similar amounts of within-group genetic variation, most of the N groups would be in Africa” (Taylor Reference Taylor2011, 471).

2.5 Minimalist Biological Races

Recently, Michael O. Hardimon (Reference Hardimon2017a, Reference Hardimon and Zack2017b), Quayshawn Spencer (Reference Spencer2012, Reference Spencer2014, Reference Spencer2018), and Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) have defended minimalist or deflationary accounts of race as biological. Their approach, which I call minimalist biological racial realism,Footnote ¹⁹ makes promising departures from previous accounts that defended a biological basis for race. Unlike revisionist biological conceptions of race, minimalist biological racial realism defends a view of biological races that captures the “logical core” of the ordinary race concept. While the view holds that races correspond to at least one structure of human genetic diversity, it rejects a robust interpretation of the genetic basis of race. Minimalist biological racial realism therefore holds that a conception of race consistent with the ordinary race concept is (i) biologically real and (ii) could end up, as a matter of empirical fact, not explaining a wide range of important genetic traits. Let us take each view in turn to see its strengths and, as I will argue, serious weaknesses.

For Hardimon (Reference Hardimon2017a, Reference Hardimon and Zack2017b), it is possible to define a biologically genuine category that fulfills intuitive desiderata for a conception of race. Races, ordinarily conceived, are biological groups that vary from one another in physical characteristics such as skin color, eye shape, and lip form. Races are groups and “we can say that, for any given race R, there is an in-principle answer to the question, what pattern of visible physical characters does it exhibit?” (Hardimon Reference Hardimon and Zack2017b, 151). Hardimon (Reference Hardimon and Zack2017b) provides the following definition of minimalist races: “a (minimalist) race is a group of human beings:

1. (1) which, as a group, is distinguished from other groups of human beings by patterns of visible physical features,

2. (2) whose members are linked by a common ancestry peculiar to members of the group, and which

3. (3) originates from a distinctive geographic location (Hardimon Reference Hardimon and Zack2017b, 150).

Of course, it is trivial to construct categories that vary along some selected property. We could do so for height, weight, or any number of other biological or non-biological properties. Hardimon (Reference Hardimon and Zack2017b) defends the biological genuineness of races on three counts. First, the “patterns of visible physical features” that are the defining characteristic of races are biologically determined. There is a genetic and developmental basis for these varying traits. Second, as discussed above, genetic clustering algorithms such as structure used by Rosenberg and colleagues (Reference Rosenberg, Pritchard and Weber2002) find a structure to human genetic diversity that, at K = 5, yields five continental populations that correspond to minimalist races. Third, there is an explanation both for why minimalist races have an underlying genetic structure and the variation in traits among minimalist races, namely, “biological raciation.” As continental groups, minimalist races are geographically separate from one another and are separately subject to founder effects and genetic drift. Furthermore, the salient racial features such as skin color are plausibly adaptations to the different ecological pressures. Hardimon argues that “minimalist race counts as biologically significant because a number of its visible physical features such as skin color are almost certainly evolutionary adaptations to the climate of the aboriginal home of the minimalist races” (Hardimon Reference Hardimon and Zack2017b, 158). Taken together, Hardimon (Reference Hardimon and Zack2017b) argues these three facts show that minimalist races are a biologically meaningful category.

Hardimon rejects the unsound inferences of the essentialist biological race conception. Minimalist races are not claimed to have normative or further genetic significance. Minimalist race is biological merely because the physical features that classify races – lip form, eye shape, skin color – are biological traits and geographic ancestry is a biologically relevant distinction. Furthermore, physical differences do not tell us anything about underlying genetic diversity other than perhaps about the genetic and developmental source of those differences. That is, there may be a great deal of genetic diversity between populations that show no salient physical differences, and vice versa. The genetic pathways that determine outward traits are a tiny fraction of the overall genome.

Quayshawn Spencer (Reference Spencer2012, Reference Spencer2014, Reference Spencer2018) and Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) defend the biological reality of race along similar lines to Hardimon. Spencer’s account is both revisionist and minimalist. It is revisionist because Spencer takes “race” not be a kind but an entity. Specifically, on Spencer’s account race refers to a set of human populations. The account is minimalist because Spencer does not take race to be robustly explanatory in the way race essentialism does. For Spencer, to say that race is biologically real is to hold that it is “an epistemically useful and justified entity in a well-ordered research program in biology, which I will call a genuine biological entity” (Glasgow et al. Reference Glasgow, Haslanger, Jeffers and Spencer2019, 95).Footnote ²⁰ That is, race has the same status as “monophyletic group, TYRP1 gene, hypothalamus,” which are cases of good scientific classification. These are rivaled by cases of bad scientific classification, such as “gemmule, baramin, and destructiveness organ” (Spencer Reference Spencer2012, 185), which scientists would reject.

Spencer (Reference Spencer2012) argues that the feature that genuine entities share is that they advance long-term scientific progress. The genuine entities promote “epistemic progress in science, such as improving our ability to predict known phenomena, or accurately predicting novel phenomena” (Spencer Reference Spencer2012, 186). Genuine entities play an epistemic role in a well-ordered scientific research program. They are fruitful and lead scientists down exploratory paths that are likely to lead to new discoveries instead of dead ends. In the case of biology, Spencer argues that e is a genuine biological entity if,

1. (i) e is useful for generating a theory t in a biological research program p,

2. (ii) using e to generate t is warranted according to the epistemic values of p to explain or predict an observational law of p, and

3. (iii) p has coherent and well-motivated aims, competitive predictive power, and frequent cross-checks (Spencer Reference Spencer2012, 193).

Spencer argues that each of these conditions is satisfied in the case of race. Condition (iii) is fulfilled by population genetics, which is one of the core research programs of biology. As for the first two conditions, Spencer argues that research into human population structure and genetic clustering algorithms by Rosenberg and colleagues (Reference Rosenberg, Pritchard and Weber2002) and latter researchers vindicate the biological genuineness of race. A species can be subdivided into a number of populations, where K is the number of populations. These divisions are the population structure of the species. As we have seen, at K = 5, we get the human continental populations: Africans, Eurasians, East Asians, Native Americans, and Oceanians. These human continental populations, Spencer argues, satisfy conditions (ii) because, they successfully generate a theory about human population structure in which the “observational law is that humans have K = 5 genetic structure that is largely geographically clustered in the following regions: the Americas, Sub-Saharan Africa, Oceania, Eurasia east of the Himalayas, and Eurasia west of the Himalayas and North Africa” (Glasgow et al. Reference Glasgow, Haslanger, Jeffers and Spencer2019, 99). That is, as Rosenberg and colleagues (Reference Rosenberg, Pritchard and Weber2002) have shown, the human continental populations are the population subdivision obtained by structure at K = 5. Furthermore, Spencer argues that these human continental populations are identical to at least one scheme of racial classification, namely, the US Office of Management and Budget’s (OMB) 1997 racial categories: Black or African, White, American Indian or Alaska Native, and Native Hawaiian or Other Pacific Islander. Therefore, given that human continental populations are biologically real, OMB races (which are identical to these populations) are biologically real.

Crucially for Spencer, even though races are genuine biological entities, “the only metaphysical fact that follows from a [entity] being genuine is that it is real enough to use in ongoing scientific research” (Spencer Reference Spencer2012, 194). Spencer’s biological racial realism is therefore minimalist. The account does not ground the biological genuineness of race in the fact that race explains a vast array of biological, psychological, or other phenomena. Rather, by fulfilling the conditions of realness or genuineness, it could play an explanatory role in scientific research programs such as medical genetics. As Glasgow et al. puts it, “we now know that it’s metaphysically possible for some races to matter in medical genetics because some races are biologically real” (Glasgow et al. Reference Glasgow, Haslanger, Jeffers and Spencer2019, 104). Because OMB race is biologically real, it is metaphysically possible that race matters in medical genetics. It is “real enough” to be explanatorily or predictively useful. However, Glasgow et al. notes:

This is what makes Spencer’s view minimalist. The biological reality of race does not depend on the fact that it robustly explains a wide range of empirical phenomena. In fact, it may turn out, after empirical investigation, that there is no other way OMB races differ phenotypically than in the defining racial traits. But that is neither here nor there on whether races are biologically real. It is possible that races matter explanatorily because they are real, they are not real because they are (robustly) explanatory.

In summary, Spencer and Hardimon defend a minimalist biological racial realism that captures key elements of the ordinary concept of race. What makes their accounts minimalist is that they eschew the inference from biological race to the clustering of other significant biological properties (aside from those that are defining physical characteristics of races). They claim, rather, that it is possible for a minimalist biological race to be useful in biology. The main candidate for the epistemic usefulness of biological race is medicine. In the next section, I discuss criticism of minimalist biological racial realism.

2.6 Are Minimal Races Biological?

Eric Winsberg (Reference Winsberg2022) argues that Spencer’s approach represents the best attempt thus far to ground race in biology, but nonetheless fails. Winsberg argues that the connection Spencer draws between human continental populations, distinguishability by structure-like programs, and the OMB races cannot sustain the conclusion that races are biologically real entities. First, Winsberg argues, human continental populations are not biological entities. As discussed in Section 2.5, Spencer (Reference Spencer2018) drew on two facts to defend the claim human continental populations are biologically real: human continental populations are distinguishable by structure-like programs and we can explain why they are so distinguished by appeal to evolutionary forces such as drift, selection, and mutation. For instance, Spencer (Reference Spencer2018) claims Native Americans are “modified descendants” of Northeast Asians. They are “modified” through changes induced by evolutionary forces. These changes are what explain why Native Americans are a distinct genetic cluster identified by structure.

Winsberg (Reference Winsberg2022) denies that it is the genetic clustering of present-day Native Americans that facilitates structure-like programs’ ability to pick them out is the product of evolutionary forces. Rather, the explanation for why such clustering exists after the age of discovery has to cite “the complex social history of racial segregation that has preserved this clustering over five centuries of colonial and post-colonial history” (Winsberg Reference Winsberg2022, 13). After all, we might add, an “Indian reservation” is not a barrier erected by nature. As such, it is not clear why human continental populations are biological entities if they are (currently) maintained primarily by social mechanisms of isolation and cohesion. They may be real scientific kinds, but they are not biological. More crucially, Winsberg (Reference Winsberg2022) argues that even if we are warranted in holding population subdivisions are real scientific kinds, it does not follow that each and every population subdivision constitutes a real kind. We can hold that color is a real kind without committing ourselves to the claim that green or yellow or wine-dark red are real kinds. Consequently, it is possible that structure-like programs discover real scientific kinds that are integral to the success of population genetics. Nonetheless, it does not follow that any given cluster discovered by structure is a real scientific kind, let alone a biological one.

Furthermore, Winsberg (Reference Winsberg2022) rejects the claim that OMB races are identical to human continental populations. As Winsberg (Reference Winsberg2022) observes, the Census Bureau uses the OMB racial classification scheme to roughly operationalize a (murky) social notion of race prevailing in the United States. How does this highly qualified approach produce categories that end up being identical to races qua biological entities? It is all the more puzzling since Spencer’s claim is not that the OMB races are merely contingently coextensive with biological races. Spencer rather holds that the OMB races refer to the human continental populations, which are in turn biological entities. Winsberg (Reference Winsberg2022) argues the five OMB racial groups are not picked out referentially. The five OMB racial groups are not rigid designators, they do not have fixed references. The social process in which the OMB races were formulated was highly contingent and is even now subject to change. It would be a massive coincidence, one that Spencer leaves unexplained, if the OMB races came to robustly identify biological populations.

2.7 Minimalist Biological Race and the Gerrymandering Objection

The objections Winsberg (Reference Winsberg2022) raises to Spencer are a consequence of a broader issue with minimalist biological racial realism. Namely, their approach attempts to show how our murky, contingent, and contested social notion of race, or at least a conception consistent with one version of it, is also part of the biological structure of the world. I argue that minimalist or OMB races fail to count as genuine biological kinds or entities. Spencer (Reference Spencer2012) and Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) are right that for races to be biologically real they need not fulfill stringent criteria such as being independent from scientific interests or fundamental categories of population genetics. Many accounts ground biological realness (or biological naturalness) in the perspectives, interests, and practices of scientists (see Kitcher Reference Kitcher2007). However, the fact that races according to at least one classification scheme correspond to population structure in humans is not sufficient to establish that races are biologically real or genuine. At least not in the way “hypothalamus” or “TYRP1 gene” are genuinely biological. I argue this is because minimalist biological racial realism is confronted by a dilemma. Either, as a matter of empirical fact, race explains very little or nothing in biology, in which case this gives rise to the gerrymandering objection against minimalist biological races; or race is robustly explanatory in biological sciences, in which case minimalism would have to be abandoned in favor of a robustly realist position.

On the first horn, the minimalist realist is confronted by the gerrymandering objection which holds that minimalist biological races are gerrymandered kinds or entities. The distinction between natural or genuine kinds and properties on the one hand and unnatural or pathological kinds or properties on the other is at the forefront of metaphysics and philosophy of science. David Lewis’ (Reference Lewis1983) groundbreaking “New Work for a Theory of Universals” highlighted the centrality of naturalness in theories of explanation, laws of nature, similarity, and induction, among others. Furthermore, any scientific project involving categorization or classification needs a principle by which to distinguish appropriate from inappropriate classifications, regardless of how pluralist the account (Franklin-Hall Reference Franklin-Hall2015). Classificatory practice in science has to be disciplined by a carving principle that tracks the grooves and joints of nature. As Franklin-Hall (Reference Franklin-Hall2015) writes, “thus we do well or badly, classification-wise, to the extent that our partitions track the kinds embedded in nature itself, and the pathological categories are those that in no way – even but through a glass darkly – match the world’s own” (Franklin-Hall Reference Franklin-Hall2015, 926).

Gerrymandering objections charge that a proposed entity or kind is inappropriately “built up”. To be gerrymandered is one way a kind or entity can lack naturalness or genuineness. There are many accounts of metaphysical naturalness or genuineness, and I do not discuss here the considerable literature on metaphysical naturalness. Nonetheless, there are certain desiderata that are widely held as integral to naturalness. I discuss two core features that minimalist races lack that render them liable to the gerrymandering objection. First, natural kinds or properties are projectible or portable. There is no consensus on how to characterize projectability.Footnote ²¹ The basic idea is that projectability makes inductive inferences permissible. For a natural kind S, we can legitimately infer from X is S to other predicates in relation to which it is projectible (Khalidi Reference Khalidi2018, 1380 f). For instance, in Goodman’s (Reference Goodman[1955]/1983) classic example, we can project from X is an emerald to X is green. One of the telltale signs that an entity or kind is gerrymandered is a lack of projectability or explanatory connection to a wide range of explananda-phenomena.

I argue that biological races (minimally conceived) are not projectible or portable. As such, they do not possess an explanatorily or epistemically privileged status such as realness or genuineness. And this is partly because they have been minimally conceived. By the lights of Hardimon and Spencer themselves, it is possible biological races do not explain a wide range of phenomena, although they maintain it possible that they may do so. I do not claim that there is nothing that minimalist races explain. After all, the fact that minimalist races correspond to genetic clusters at K = 5 in Rosenberg and colleagues’ (Reference Rosenberg, Pritchard and Weber2002) model means that minimalist races can explain facts about the distinctness of those clusters. And racial traits such as skin color are biological traits explained by biological mechanisms. Hardimon and Spencer draw on these facts to ground their realism. But the limited explanatory role is not enough to avoid a gerrymandering objection. Gerrymandered kinds or entities are also capable of explaining some phenomena. Take for instance Fodor’s (Reference Fodor1974, 11) example of a manifestly unnatural kind, the kind is transported to a distance of less than three miles from the Eiffel Tower. Let’s call this kind T. There are things that T predicts and explains. For instance, for goods sold in shops, T explains why they are denominated in Euros. It predicts that they are likely to be more expensive than other goods sold in France. It may explain other things besides. Nonetheless, T has extremely limited portability. It does not figure into robust explanatory relations of the kind sought by sciences. Furthermore, whatever T explains is better explained by other, more natural kinds.

Second, and relatedly, natural kinds or properties contribute to scientific understanding. Scientific fields are interested in or investigate kinds that are maximally mutually explanatory with respect to the target regularities studied by that science. Physical kinds such as charge, spin, charm, field, and so on and neurobiological kinds such as, neuron, neurotransmitter, synapse, and so on have respective robust explanatory connections that facilitate understanding how the regularities studied by a given science (fundamental physics, neuroscience) fit together (Bhogal, Reference Bhogal, Hicks, Jaag and Loew2023). Minimalist races lack these valuable explanatory connections in biology that would justify their biological naturalness or genuineness. Note that, for instance, Kitcher’s revisionist race conception does not face the same challenge. Kitcher (Reference Kitcher2007) begins by explicitly drawing on the standard biological practice in apportioning populations, namely, the division of species and subspecies. On Kitcher’s approach, if “race” is a biological subdivision of the human species, it should be characterized in the same way other biological subspecies are. This approach has the potential to yield explanatory connections between “race” and a cluster of explananda-phenomena in humans as subspecies have to their respective clusters in other species.

The (potential) explanatory value of revisionist biological conceptions of race may have come at the price of abandoning commitment to the ordinary race concept. In so far as the minimalist conception is of an ordinary race concept, it will fail to pick out an explanatorily relevant kind or entity in biology. As Long, Li, and Healy (Reference Long, Li and Healy2009) conclude in their study of race and genetic diversity:

Human genetic diversity tracks patterns of human population dispersal and local population dynamics that are out of sync with the ordinary race concept.

I have argued that minimalist races are liable to a gerrymandering objection. But this is the case only if minimalist races fail to explain a wide range of phenomena. And that is in fact the case. Minimalist races are not projectible or portable; they do limited explanatory work in biology. Some of what they explain is better explained by other subdivisions of humans at a higher or lower grain (Kalewold Reference Kalewold2020). Minimalist races therefore do not advance biological understanding. Race minimally biologically conceived lacks the constraints to count as among the biological kinds along with TYRP1 gene, monophyletic group, and hypothalamus (all of which are robustly explanatorily connected to other genuine kinds and the regularities they underlie or produce). The gerrymandering objection would be weak if it held race to a higher standard than other biologically genuine or natural kinds. However, it is by looking at other biological kinds that we can see how minimalist races fail to secure the kind of value necessary for biological naturalness. To see why, let us turn to a discussion of dimensions of explanatory value in biology.

2.8 Explanatory Value in Biology

Jim Woodward (Reference Woodward2010) discusses three dimensions of explanatory value in the biological sciences; stability, specificity, and proportionality (Woodward Reference Woodward2010, 292). To take just one of these values briefly, Woodward (Reference Woodward2010) develops specificity by (i) drawing on Lewis’ notion of influence and (ii) the one-to-one conception of causation. Specificity is often invoked to account for the privileged role DNA plays in explaining the development of various phenotypes. Woodward (Reference Woodward2010) argues that the best way to conceive of specificity, and the special causal role of DNA, is as a form of fine-grained influence. According to influence, “C will influence E to the extent that by varying the state of C and its time and place of occurrence, we can modulate the state of E in a fine-grained way” (Woodward Reference Woodward2010, 305). Woodward (Reference Woodward2010) illustrates influence with an analogy to a dial on a radio. Moving the dial on a radio will change the stations in a fine-grained way and a given position of the dial is associated with a particular radio station (307). Causal relationships that lack specificity on the other hand are more switch-like. Whether the radio is plugged in or not for instance is not causally specific since it does not have fine-grained influence on the operations of the radio. It is in this sense that DNA is distinct from the other “cellular machinery” in causing various phenotypes. Changes in RNA polymerases, ribosomes, and energy molecules (ATP, GTP) do not have fine-grained influence over the product. However, in the case of DNA, “there are many possible states of the DNA sequence and many (although not all) variations in this sequence are systematically associated with different possible corresponding states of the linear sequences of the mRNA molecules and of the proteins synthesized” (306).

Woodward (Reference Woodward2010) considers a second standard for specificity, which sees specificity as approximating a one cause–one effect relationship. This notion of specificity draws on epidemiology where identifying a one-to-one relationship between a cause and a disease effect is the gold standard of epidemiological research. For instance, cases where being exposed to a particular toxin causes one disease, or a specific gene variant causes one disorder, and so on. Of course, most diseases have a many cause–many effect causal relationship and do not fit the one-to-one standard. Woodward (Reference Woodward2010) amends this notion by marking it not as a criterion of causation, but as a biologically significant type of causal relationship. In this sense of specificity, “C will be a more (rather than less) specific cause (in the one-to-one sense) to the extent that it causes only a few different kinds of effects within a pre-specified range” (311). For example, enzyme activity is more specific if it interacts with only a narrow range of substrates and produces only a limited number of effects. Specificity is an important dimension of explanatory value for biology. Biologists do not merely look for any entity or kind and its relationship to the target they wish to explain. As discussed by Woodward (Reference Woodward2010), it is the specific (causal) explanatory relations, and the entities that engage with them, that are of scientific interest to biologists.

The problem for minimalist or OMB races is that they are at an inappropriate grain to secure explanatory value in biology. In order for minimalist races to have explanatory value in biology, what we would want to see is specific, robust, and proportional explanatory relations between minimalist races and the regularities they (possibly) explain. Consider a contrast case such as enzyme action. The core properties of an enzyme – its size, shape, and configuration – are what explains its derivative properties including its binding, affinities, and so on. In the case of minimalist race, however, what explains biogenetic phenomena that seemingly vary racially, such as the incidence of genetic diseases such as sickle cell disease and lactose intolerance, or even defining physical characteristics such as skin color are not explanatorily connected to race per se. There is no specific explanatory relation between race and these other properties (see the following subsections).Footnote ²²

Additionally, the minimalist biological race concept takes continental barriers to be a core driver of (genetic) distinctiveness between races. Human population genetics research vindicates at least a deflationary view of what that distinctiveness amounts to. Nonetheless, geographic distance as a result of continental barriers is not the only kind of barrier that is sufficient to establish population structure. For instance, linguistic differences between geographically co-located populations are sufficient for reproductive isolation (Hellwege et al. Reference Hellwege, Keaton and Giri2017). Population stratification therefore occurs at a much smaller, and more local, scale than the continental. As Hellwege and colleagues note “allele frequencies change randomly over time as an independent process for each population isolate, ultimately causing observable differences in the frequency of many alleles after several generations of separation and differentiation” (Hellwege et al. Reference Hellwege, Keaton and Giri2017, 2). Sophisticated clustering algorithms are capable of distinguishing between dozens or hundreds of population groups. For instance, Gao and Starmer (Reference Gao and Starmer2007) debuted a clustering program capable of identifying clusters that differentiate between Chinese and Japanese populations (Gao and Starmer Reference Gao and Starmer2007). What population stratification showed was that other clusterings of human populations were also distinct and potentially explanatorily more relevant in explaining medical and other phenomena. They act on populations within and across different putatively biological (ordinary) races.

To motivate this objection, I consider the most widely cited candidate for the explanatory usefulness of race to science – medical genetics.

2.9 The Case of Race and Medical Genetics

The gerrymandering objection to minimalist biological races charges that they lack essential dimensions of explanatory value. For instance, in order to be explanatorily apt in medical genetics, the genetic variation at the level of racial category must correspond to variation in rates of disease incidence characterized at that level (specificity). The racial categories must secure an epidemiological explanatory value that is missing either at higher (species) or lower (local population) organizational levels. As Root (Reference Root2003) argues, this matching of the appropriate grain is one they fail to achieve (Root Reference Root2003). Genetic studies must demonstrate a causal-explanatory relationship between “medically relevant” genes that vary racially and disease-phenomena. Yet, the examples commonly cited by biological racial realists, such as lactase persistence and sickle cell anemia (Spencer Reference Spencer2018), fail to demonstrate the explanatory value of biological race. To see why, let’s look at these cases in turn.

2.10 Sickle Cell Disease

Consider the explanation for the prevalence of sickle cell anemia in sub-Saharan Africa, a commonly cited candidate for a disease in which race is a medically relevant category. Sickle cell anemia (HbSS) is a severe form of the inherited red blood cell disorder called SCD. As the name suggests, patients with SCD have abnormal rigid and sickle-shaped hemoglobin (i.e., red blood cells). Healthy hemoglobin is disc-shaped and flexible, permitting easy movement through blood vessels. Abnormal hemoglobin (Hb) on the other hand has a rigid, sickle-like structure that, among other things, makes them prone to snag on the walls of blood vessels leading to restricted blood flow and other serious symptoms such as pain, blindness, and stroke.

Sickle cell disease is a genetic disorder. Patients with SCD have inherited a gene that codes for abnormal hemoglobin from each parent.Footnote ²³ In the case of sickle cell anemia, patients have inherited sickle (HbS) hemoglobin gene from both parents. Sickle cell trait is a heterozygous form of SCD where the patient inherits a normal hemoglobin gene from one parent and an abnormal hemoglobin (such as HBB or HbC) from the other. While sickle cell anemia is a debilitating disease, individuals with sickle cell trait are usually healthy and do not present with the serious symptoms of other SCD (Piel, Steinberg, and Rees Reference Piel, Steinberg and Rees2017).

The prevalence of a debilitating genetic disorder like sickle cell anemia in any population is striking given the putative selection pressure against it. The explanation for this puzzling phenomenon is now well-known and has to do with the fact that sickle cell trait provides protection from malaria. This in turn explains its relative prevalence in sub-Saharan Africa, the Middle East, the Mediterranean Basin, and India: all regions with high levels of endemic malaria. There are two distinct explanations linking the sickle cell trait to malaria: (1) a mechanistic explanation of how the sickle cell trait neutralizes the plasmodium parasite that causes malaria, and (2) the evolutionary explanation of how the variant responsible for this trait is preserved in the affected populations.

How then does race come into the explanatory picture of sickle cell anemia? The mechanistic explanation of how sickle cell trait protects from malaria does not depend on nor inform facts about race.Footnote ²⁴ The evolutionary explanation of the prevalence of sickle cell disease is tied to the protection from malaria provided by the sickle cell trait. Individuals who carry one copy of the sickle cell allele and one copy of the normal hemoglobin allele have a selective advantage over individuals who have two copies of the normal hemoglobin allele or two copies of the sickle cell allele. The selective advantage arises because individuals with sickle cell trait are more resistant to malaria than those with normal hemoglobin, but they do not suffer from the severe symptoms of sickle cell disease that are associated with having two copies of the sickle cell allele. This results in the selection for balanced polymorphism in malaria-affected populations (Allison Reference Allison2002; Piel, Steinberg, and Rees Reference Piel, Steinberg and Rees2017).

Although sickle cell disease is particularly associated with sub-Saharan Africa, and with African Americans in the United States, it is also present in other regions with endemic malaria. As Williams and Weatherall note, “sickle cell anemia occurs throughout sub-Saharan Africa and in small pockets in the Mediterranean region, the Middle East, and the Indian subcontinent” (Williams and Weatherall Reference Williams and Weatherall2012, 1). The sickle cell gene arose independently at least twice, once in and once out of Africa. Since the selection pressure is due to endemic malaria, it is unsurprising that other malarial geographic regions also show the prevalence of sickle cell trait. Although sickle cell anemia produced by the sickle hemoglobin allele (HbS) is widely identified as a disease affecting “African Americans,” there is considerable heterogeneity in its incidence within the putative minimalist black or OMB “Black or African American” race. As Allison (Reference Allison2002) shows, “among tribes living close to the coast of Kenya or to Lake Victoria, the frequencies exceeded 20%, whereas among several tribes living in the Kenyan highlands or in arid country, the frequencies were less than 1%. These differences cut across linguistic and cultural boundaries and were independent of blood group markers that we documented” (Allison Reference Allison2002, 280). The source of this heterogeneity is the differential prevalence of malaria. Piel and colleagues (Reference Piel, Patil and Howes2010) found “the gradual increase in HbS allele frequencies from epidemic areas to holoendemic areas in Africa is consistent with the hypothesis that malaria protection by HbS involves the enhancement of not only innate but also acquired immunity” (Piel et al. Reference Piel, Patil and Howes2010, 3).

Given that the African ancestors of African Americans were predominantly drawn from malaria-prone regions, they unsurprisingly have higher rates of sickle cell disease and other “genetic variants that confer resistance to malaria are associated with RBC [red blood cell] traits in African-Americans” (Ding et al. Reference Ding, de Andrade and Manolio2013, 1061). However, the social and political history of where the African ancestors of African Americans originated is once again not a fact that is connected to biological races. The evolutionary lineages of local population groups across multiple continents, and the distinct selection pressure that led to the prevalence of sickle cell alleles that are adaptive in those specific malaria-prone environments, specifically and proportionally explain why these populations disproportionately suffer from sickle cell anemia. Minimalist or OMB races do not come into the explanatory picture in a valuable way.

2.11 Lactase Persistence

A second commonly cited example of the relevance of a genetic race concept to medicine is the case of lactase persistence (Spencer Reference Spencer2018, 1028). Lactase persistence refers to the ability to digest lactose in adulthood. It is a genetic trait that has evolved in humans in response to the domestication and exploitation of dairy animals. While lactose intolerance is the norm in most human populations, the frequency of lactase persistence varies widely across different populations, with the highest frequencies found in populations that historically relied on milk as a source of nutrition, such as in northern Europe, and lower frequencies found in populations that did not have a tradition of milk consumption, such as in East Asia and parts of Africa. The absence of lactase persistence (LP) before the bronze and iron ages and the subsequent explosion of LP suggests LP is one of the strongest cases of positive selection in recent human evolutionary history (Evershed et al. Reference Evershed, Davey Smith and Roffet-Salque2022).

The evolutionary explanation of lactase persistence is one example of how human biology and culture have coevolved over time. Although LP selection must be related to milk consumption, the exact mechanism by which LP contributes to inclusive fitness is not fully understood. Evershed and colleagues (Reference Evershed, Davey Smith and Roffet-Salque2022) find that LP is not associated with higher milk consumption. Rather, individuals with LP fare better during contingent adverse events such as famine. The ability to digest milk confers an advantage during these adverse events, driving LP selection (Evershed et al. Reference Evershed, Davey Smith and Roffet-Salque2022). In any case, there is nothing racial (or continental) about the explanation of the rise of LP. It acts on populations on the basis of their dietary choices and other environmental factors, which vary both within and across continents. Nothing at the vast population grain of race plays a specific or robust explanatory role.

2.12 Population Stratification

The upshot of the discussion in previous sections is that race, as a biological population, entity, or kind, is not part of the productive continuity of genetic racial disparities in sickle cell disease or lactase persistence. That is, race qua biological is not robustly explanatorily connected to these genetic traits in a way that secured explanatory value. Of course, this is not to deny that population-level genetic differences can play an explanatory role in medicine. Rather, the population in question is only rarely racial. Given the factors that produce genetic diversity between continental populations – reproductive isolation, selection, genetic drift – also operate at a much finer grain, it would be a massive coincidence if it were racial difference,Footnote ²⁵ as opposed to populational differences at a different grain, that accounted for a large share of epidemiological difference.

This problem is general and not specific to the two medical cases discussed in Sections 2.10 and 2.11. The investigative tool used to investigate the connection between genetics and epidemiological racial disparities, namely, GWAS studies, is limited in its ability to secure the explanatory value or biological races.

Genome-Wide Association Studies (GWAS) are a set of tools drawn from population genetics used to identify statistical relationships between single-nucleotide polymorphisms (SNPs) and phenotypic traits. An SNP is a nucleotide, at a particular locus on a chromosome, that varies across individuals. Genome Wide Association Studies (GWAS) scan the genome of suitably defined groups to identify whether there is SNP variation (e.g., a group of individuals with Multiple Sclerosis and a control group of individuals without the disease). GWAS studies are at the forefront of research into the relationship between genotypes and diseases. The GWAS Catalog includes GWAS studies that have found over 50,000 significant associations between genetic variants and diseases and traits (Tam 2019). Nevertheless, GWAS studies do not indicate whether these associations are causal or spurious, or whether the implicated SNPs are actually involved in the pathogenesis of the diseases with which they are associated.

GWAS studies on epidemiological racial disparities attempt to identify SNPs associated with a disease phenotype within human subpopulations. However, there are limitations to the explanatory value of the statistical associations revealed by GWAS into epidemiological racial disparities. For instance, GWAS studies are constrained by a population stratification (PS) problem (Hellwege et al. Reference Hellwege, Keaton and Giri2017). PS is a significant limitation on the robustness of GWAS’ association identification. PS is a result of nonrandom mating “due to geographic isolation of subpopulations with low rates of migration and gene flow over the course of several generations” (Hellwege et al. Reference Hellwege, Keaton and Giri2017, 2). This isolation is not caused solely by continental barriers but also as a result of linguistic, cultural, or other barriers to free and random mating. In the absence of a random distribution of alleles, GWAS studies need to account for this underlying stratification in order to avoid spurious correlations. As Hellwege and colleagues (Reference Hellwege, Keaton and Giri2017) show, “the differentiation among subpopulations is detectable even when the regional differences are subtle, as has been described in Chinese and Japanese and European populations […] Cultural differences among populations also create stratification, even when populations inhabit the same geographical region” (Hellwege et al. Reference Hellwege, Keaton and Giri2017, 3).

The consequence of population stratification is not that more powerful tools will fail to overcome these challenges and yield meaningful results. Rather, the nature of population stratification reveals that local populations are a biologically significant element of the human genetic mosaic. Of course, arguing that race is meaningfully biological does not preclude the reality of subracial populations and vice versa. However, the fact that criteria such as distinguishability are applicable to a far larger number of population groups implies it would be a massive coincidence if the racial level turned out to be explanatorily apt across a range of explananda-phenomena. Indeed, as Hou and colleagues (Reference Hou, Ding and Xu2023) find in their research on the causal contribution of differing ancestry in admixed (i.e., multiple ancestry) individuals such as African Americans, there is “minimal heterogeneity in causal effects by ancestry.”

Race (as the realist conceives it) figures into far fewer explanatory relations than the more fine-grained local populations of which it is composed. As such, (minimalist or OMB) race is an explanatory orphan in biology. To say that minimalist biological races are explanatory orphans is to hold that they fail to secure an explanatory value that is missing at a higher or lower populational grain. Explanatory accuracy is relatively trivial to obtain for categories. Strevens (Reference Strevens2008), for instance, notes that we can disjunctively create a causal model to explain some target phenomenon which is accurate but nonetheless inappropriate. In his example, he asks us to consider two causal models for Rasputin’s death, one involving his drowning due to being bound and thrown in a river (influviation) and another his poisoning. Suppose that it is the influviation, rather than the poisoning, that is a difference-maker for his death. Then an accurate explanation of Rasputin’s death will cite the fact that Rasputin was bound and thrown in the river as the explanans. We can go further and “take the disjunction of the setups of the two models and form a new model that has the disjunction as its setup: it states that either Rasputin was thrown in a river or he was fed poison teacakes, and so on. The disjunctive model is veridical, since one of these chains of events did occur as claimed, and it entails Rasputin’s death, since both chains of events lead to death” (Strevens Reference Strevens2008, 102). However, the disjunctive model is gerrymandered. In Strevens’ (Reference Strevens2008) account, it lacks cohesion. The ability to generate gerrymandered entities or models that nonetheless are explanatorily accurate is a major impetus for accounts of naturalness or genuineness. Explanations that cite gerrymandered entities might be accurate, but they will lack explanatory value.

In summary, I argue that what matters in establishing the genuineness of biological kinds or entities is whether they participate robustly in biological explanatory practice. The factors that are supposed to secure the genuine biological kindness of minimalist race – adaptiveness, distinguishability, and so on – are readily applicable to larger or smaller population groups that secure dimensions of explanatory value (proportionality, robustness, stability, and so on) that minimalist races lack. Distinguishability, as measured by structure-like programs, is possible for other possible groupings. Indeed, given enough sites, it is possible to detect population structure down to the level of families. Given the discussion in this section, it raises the question of why a race-based model would be favored in biomedical research. As Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019, 115) notes:

And because these different grains will be more explanatorily valuable, (minimalist biological) races do not figure into more robust, more specific, or more stable explanations than explanations that cite other human populations.

Of course, for all I have said, minimalist or OMB races can be real in some other metaphysical sense. They may be socially real (see the next section). Or they may be basic or primitive. For instance, Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) proposes a “basic” racial realism. Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) proposes that race is possibly real in a basic metaphysical sense that is not grounded in either biological or social facts. Race is basic in this sense because it is outside the purview of science (Glasgow et al. Reference Glasgow, Haslanger, Jeffers and Spencer2019, 139). As an illustration of the kind of thing a race could be, Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) gives the example of sundogs, where S is a sundog if it is either a sun or a dog. Even if sundogs play no role in science, they may still nonetheless exist. Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) write that “barring a radical change in the universe, you’ll never see sundogs show up in a biology or sociology textbook. Nevertheless, it sure looks like there are things that are either suns or dogs. Fido is a sundog, because Fido is a dog. So it is plausible to say that sundogs are real, even if they are scientifically irrelevant” (Glasgow et al. Reference Glasgow, Haslanger, Jeffers and Spencer2019, 139). Glasgow calls things like sundogs basic kinds. And race is possibly one such kind.

Glasgow et al.’s (Reference Glasgow, Haslanger, Jeffers and Spencer2019) basic racial realism is not liable to the gerrymandering objection in so far as he readily admits that it is gerrymandered. Furthermore, because it is gerrymandered, as Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) note, it is inappropriate to use in scientific theorizing. The charge I make against minimalist biological racial realism is that minimalist races are also gerrymandered. Their grounding in biology is superficial. The biological races they defend fail to secure explanatory values that are central to the life sciences. After all, it is metaphysically possible that even sundogs could play an explanatory role in science. But they would fail to be good or valuable explanations. Furthermore, Spencer and Hardimon are defending a view of races as biological and not merely scientific. As such, the kinds of explanatory value they secure must be those sought by biological sciences. And as I have argued, it is these values they fail to secure.

In the next section, I turn to social conceptions of race. I begin with accounts of the social construction of race and discuss their main features.

3.1 Mills on Racial Constructivism

In an influential account, Charles Mills (Reference Mills1998) defends racial constructivism, “a view of race as both real and unreal, not ‘realist’ but still objectivist” (Mills Reference Mills1998, 47). It is “unreal” in so far as the essentialist or biological assumptions of nineteenth century race science, which played a prominent role in shaping dominant conceptions of race, are false. Nevertheless, race has an “objective ontological status” due to the intersubjective judgment that underpins it. The fact that people conceive of themselves and others as “raced” grounds the social reality of race.

Mills (Reference Mills1998, 42) introduces a thought experiment to motivate the ontological status of race as both social and real. Mills (Reference Mills1998) asks us to imagine a society in which individuals are assigned a quace (Q1, Q2, Q3, etc.). This assignment is generated randomly by a computer and does not track any morphological or geological facts. Everyone’s quace is registered in official documents such as state IDs, passports, naturalization papers, and so on. However, quace lacks any social significance beyond the mere fact that everyone has a quace. If anyone asked what one’s quace was, one would merely report what was listed in official documents.

Mills (Reference Mills1998) notes that quacial membership lacks “ontological depth.” There isn’t a quace anyone “really” is in the society described by Mills. As Mills puts it, “‘I am a Q1!’ would have no metaphysical ring, no broader historical resonance to it, any more than our declaration of our passport number has any metaphysical ring or broader historical resonance to it” (Mills Reference Mills1998, 42). Race, on the other hand, has deep social and metaphysical significance. If we picture Ellen Craft in her disguise as a white man and ask, what is sheFootnote ²⁹ really? We are drawing a sharp line between appearance and reality. We are claiming that her race is more than skin-deep. It is worth noting that the inquiry does not arise because her racial identity is ambiguous. Although there aren’t any uniform, universally applied norms of racial classification in any society, what draws interest in this case is precisely that there is a definite answer one can give depending on just which norms one appeals to (e.g., one drop rule). And this, Mills argues, is the result of two major divergences between race (in our world) and quace. First, race classification is (generally) based on identifiable morphological features. Second, race, at least in most modern societies, is a vertical system. That is, races are hierarchically placed along several dimensions (social, political, economic, etc.).

The comparison of quace and race is meant to highlight a core thesis of constructivist accounts of race, what I’ll call the contingency thesis. The contingency thesis holds that the racial categories and their hierarchical position are contingent on social and historical facts. The racial categories could have been demarcated differently than they were, or race as it came to be understood could have failed to arise. One line of evidence for the contingency thesis is the diversity of racial classification systems in the actual world. Since the number, type, and boundary of racial categories vary even among societies that are not isolated from one another, it is not a leap to suppose that dramatically different racial classification schemes could have dominated given sufficiently different social conditions. As Malik (Reference Malik1996) and Smith (Reference Smith2015) show, the racial essentialism that came to dominate in the nineteenth century had contested origins and could have failed to launch if it were not met with the historical conditions that facilitated its prominence. Consequently, for racial constructivists, it is from social facts that race draws its metaphysical import. Race is not part of the natural order of things. We do not discover races in the way scientists once discovered, for instance, the circulation of the blood or the temperature at which copper is a superconductor. Rather, race is made and maintained by the social power it has.

3.2 Haslanger on the Hierarchical Foundations of Race

Sally Haslanger (Reference Haslanger2000, Reference Haslanger2012) and Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) defend a critical and anti-racist approach to the metaphysics of race. Haslanger’s project is not merely descriptive, attempting to capture the race concept with a high degree of accuracy. Rather, a critical metaphysical analysis should allow us to achieve the goals of social justice such as combating racism. Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) argues that in determining what race is “the goal is to provide an interpretation of what has plausibly been at issue (though not always clearly at issue) “all along,” as evidenced not only by what we say, but what we do, such as the practices we engage in, the laws we pass, and social scientific explanations of these” (Glasgow et al. Reference Glasgow, Haslanger, Jeffers and Spencer2019, 16). An answer to the question “what is race?” ought to illuminate a host of social and political phenomena that extend beyond what might initially appear to be the narrow theoretical confines of metaphysics. A critical theorist wants an account of race that makes sense of the role of race in our social lives and how to upend the injustices in which it is enmeshed.

Haslanger’s analysis proceeds from the conclusion that there are no human biological groups that correspond to the essentialist or biologically realist conceptions of race. Rather, races are social categories. They are, nonetheless, no less real for it. While Haslanger rejects analyses of race as biological, this is not cause to jettison race talk altogether. A critical theorist still seeks to use race to limn the social realities created by historical and ongoing sociopolitical practices. Race helps explain inequalities and deprivations that are differentially distributed across a society like the United States. The best way to achieve these analytic and normative aims is with a social constructionist approach.

On Haslanger’s critical account, race is embedded within social hierarchies. These social hierarchies are stratified on the basis of real or imagined physical features such as skin color, eye shape, hair texture, and so on that are taken to indicate ancestry and geographic origin. Racialization is the process of assigning evaluative meaning to these real or imagined physical features. Haslanger (Reference Haslanger2000) thus provides the following account:

This account has many explanatory virtues. For one, it helps us see that the different racial classifications in different societies do not undermine the utility of a race concept. What unifies race across these different contexts is that race serves the same sociopolitical role. Namely, racialization creates racial categories and places individuals in those categories based on real or imagined physical “markers” such as skin color or eye form. The racial categories might vary from one society to the next. But their hierarchical role remains.

Take for instance the differences between the slave codes in Virginia and French Saint-Domingo in the eighteenth century. The latter had a category, mulatto, for individuals of “mixed” white and black ancestry. Whereas colonial Virginia had a rule of hypodescent, where the “mixed” offspring were assigned the race of the subordinate parent. This meant in practice that the child of a white slaver and his black slave was racialized as black. Virginia planters were not unfamiliar with the mulatto concept. However, mulatto came to have a far more distinct and socially consequential status in the French colonies along with maroon, octoroon, and other categories of “mixed” racial ancestry. These individuals made up a large share of the free population, could marry into white families, and some even owned slaves.

The explanation for the divergence was in part due to the demographic composition of these societies. Because of the difficulty of living in the French Caribbean posed by diseases such as yellow fever, there was relatively very little migration from France to the Caribbean colonies. As a result, the white population of Saint-Domingo was extremely small at about forty thousand in the late eighteenth century (Hunt Reference Hunt2006, 9). This engendered the need to expand the free population in order to stabilize the social order, one constituted by extraordinarily harsh and deadly living and working conditions for the enslaved population. American colonies and states of the antebellum US, on the other hand, did not have such a lopsided population structure between free whites and enslaved blacks. The relatively large white population meant that managing the enslaved population was manageable even if the social order excluded people of any black ancestry from an intermediate position within the social hierarchy. The economic value of, say, Ellen Craft as a slave exceeded any social value from creating a more stable sociopolitical order through the expansion of a group of free people racialized in an intermediate node between white and black. In those social circumstances, it is more provident for the planter class to racialize Ellen into the same category as William or any other enslaved persons. We see here, therefore, a demonstration of the real basis for racial classification, namely, the sociopolitical forces that prevail in society. And transforming those forces – through egalitarian political action for instance – will have the result of knocking out the base of race.

3.3 Racial Equality

One implication of political constructionist accounts such as Haslanger’s (Reference Haslanger2000, Reference Haslanger2012) and Glasgow et al.’s (Reference Glasgow, Haslanger, Jeffers and Spencer2019) characterization of race is that racial equality is not a coherent outcome of justice. Racialization is always a process that places races in a social hierarchy. Achieving social justice would have the result of robbing social significance from the real or imagined physical features on which racialization depends. An America in which black people enjoy the same opportunities – in education, health care, employment, and housing, among others – on the same terms as white people, and do not disproportionately occupy disadvantaged social roles such as the ghetto poorFootnote ³⁰or higher rates of incarceration, is one in which the defining (hierarchical) boundary between white and black would collapse.

How could this be so? If dark-skinned people with kinky hair come from Africa, and light-skinned people with smooth hair come from Europe, these visible physical features could plausibly serve as a stable and robust basis for racial categories even if membership in these categories does not entail patterns of subordination. However, there is reason to think this is not as implausible as it seems. One of the remarkable facts of the history of racialization is just how malleable and fluid racial categories have been. The varied practices of racialization in non-US contexts only help to illustrate this point. The physical features that seem an obvious and robust basis for the persistence of racial categories are missing in some cases of racialization. In other instances, racialization occurs on the basis of imagined physical differences or seemingly irrelevant facts such as social class. As Telles and Paschel (Reference Telles and Paschel2014) note in their study of racialization in Latin America, “… the relation between color and status with racial identification and the use of mixed-race categories – reflects the country’s distinct sets of historical and contemporary inducements” (Telles and Paschel Reference Telles and Paschel2014, 867). In societies where status and racial identification are intimately linked, equality would undermine continued differential racialization. Racialization knits together real or imagined physical features and social facts that do not have to coincide. But it knits them in iron. This does not mean that the effects of racialization are insurmountable. Nonetheless, undoing them involves a revolutionary reconfiguring of society in much of the world where race has taken on a paramount role in social life.

3.4 Cultural Constructionism

It is safe to say that on the political constructionist accounts we have so far discussed the tale of race is a tale of woe. Oppression, subordination, and hierarchy are the bases of the social reality of race. However, social construction need not take a wholly negative view of the consequences of social construction of race. Chike Jeffers (Reference Jeffers2013) and Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) defend a cultural constructionist account of race which he contrasts with political constructionism. While both are species of social construction, Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) argues that races are both politically and culturally constructed (Glasgow et al. Reference Glasgow, Haslanger, Jeffers and Spencer2019, 55). Races are politically constructed broadly in the way outlined by the accounts of philosophers such as Mills (Reference Mills1998) and Haslanger (Reference Haslanger2000, Reference Haslanger2012). That is, races are created and maintained primarily through differential power relations between racialized groups. Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) adds that a relatively neglected aspect of social construction is that racialization has a cultural component. Outlaw (Reference Outlaw1996) highlighted culture as fundamental to the social construction of race. Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) advances this approach in opposition to views like Haslanger’s that hold that in the absence of differential power relations and hierarchies, races would lose the basis of their social reality.

Political and cultural constructionism agree on the nature of the origins of race. On both accounts, race is the product of sociohistorical factors where groups marked by (real or imagined) physical features are placed in a social hierarchy. Where the two accounts diverge is on the question of what presently maintains the continued social reality of race. For political constructionists, differential power relations are fundamental to the present reality of race. For cultural constructionists, race would survive the end of racism. That is, the different ways of life that accompany the process of racialization would persist and therefore maintain the reality of race. In a world without racial hierarchy, Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) argues, race would entirely be grounded in culture. As such, culture is as fundamental to race as politics.

There are two claims that should be disambiguated. The first claim is that races, as a result of their unique history and social circumstances, come to have distinctive cultures. These cultures are valuable even if the context in which they were engendered is unjust. A second claim is that a group is racial in virtue of its distinctive culture, to be a racial group is to have a distinctive way of life. It is the latter, stronger, claim that is culturally constructionist about race. Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) rejects the essentialist thesis that different races belong to different civilizations. Nonetheless, he argues, it is still the case that people come to have pride in their culture qua race. For instance, a black American child can feel pride in the achievements of ancient African civilization even if he may not be connected to them through close ancestry. Movements such as “Black is Beautiful” answer to a real feeling among black individuals that their physical features, markers of inferiority in racist societies, are intrinsically valuable and aesthetically pleasing. These facts of “racial consciousness” can ground the practices that are sufficient to continue sustaining the social reality of race.

Jeffers’ cultural constructionism makes a valuable contribution to thinking about social race. Social race is not just political, it is also cultural. Moreover, the experience of racialization is not entirely negative. Racial solidarity and cultural forms of belonging are engendered by the act of seeing oneself as part of a racial group. However, the cultural constructionist account has flaws. Crucially, “cultures” are too fluid, amorphous, and numerous to do the work of grounding race. In an infamous 1988 New York Times Magazine interview with James Atlas, the American writer Saul Bellow once asked “who is the Tolstoy of the Zulus, the Proust of the Papuans? I’d be glad to read him” (Atlas Reference Atlas1988). To which question the journalist Raph Wiley much later answered, and in my view perfectly so, “Tolstoy is the Tolstoy of the Zulus – unless you find a profit in fencing off universal properties of mankind into exclusive tribal ownership” (Wiley Reference Wiley1996). In a world rid of racism, I can hardly see how Wiley’s view would not be widely shared. And in that world, a black child may find herself as proud of a Bernini as she is of the Benin Bronzes. Such a world will lack (racialized) cultural barriers that can maintain the social reality of race.

3.5 Social Construction Instead of What?

The title of Ian Hacking’s (Reference Hacking1999) influential book on social construction asked, “The Social Construction of What?” It is even more instructive to ask social construction instead of what? For Hacking (Reference Hacking1999), constructionist accounts of any X are necessarily revisionist. The purpose of identifying some X as socially constructed is to debunk an essentialist conception. That is, the claim that X is socially constructed is “used to undermine the idea that X is essential, even that X has an ‘essence’” (Hacking Reference Hacking1999, 16). There would be no point to the use of “social construction” if what is claimed as constructed wasn’t commonly taken to be inevitable or essential, part of the furniture of the universe (as such, things like money and laws are not on Hacking’s account socially constructed in so far as they are widely understood to be social creations that can be altered through social processes). For instance, there is no social constructionism about money because it is common knowledge that money is the result of sociopolitical and economic human activities. Furthermore, social constructionists often urge that “X is quite bad as it is” and “we would be much better off if X were done away with, or at least radically transformed” (Hacking Reference Hacking1999, 6). Constructionist views therefore frequently come as part of a package that includes normative commitments.

Haslanger (Reference Haslanger2012) agrees with Hacking (Reference Hacking1999) that “the discourse of social construction functions differently in different contexts … ” (Reference Haslanger2012, 113). Some social constructionists defend their view as an analysis of race that rescues the concept from a biological conception and instead emphasizes the social, (and/or) cultural, factors that shape the formation and use of racial categories (cf. Sundstrom Reference Sundstrom2002). To view race as socially constructed is to hold that race does not exist independently of historically contingent social practices. By emphasizing the social and historical origins of race, social constructionists aim to expose the ways in which race is used to perpetuate social inequality and oppression. In any case, a view can hardly be called social constructionist unless it disclaims essentialism. What set of conditions get defined as essentialist is of course a matter of dispute. Since race essentialism has fallen out of favor even among those who maintain a commitment to a biological conception of race, the question arises as to how biological racial realism and social constructionism are related. That is, is social construction a view one must hold instead of a biological conception of race? For many proponents of both social construction and biological racial realism, the answer is no (Kitcher Reference Kitcher2007; Hardimon Reference Hardimon2017a; Glasgow et al. Reference Glasgow, Haslanger, Jeffers and Spencer2019; Hochman Reference Hochman2022).

The relationship between social construction and biological racial realism is more complicated than the one between constructionism and essentialism. Biological racial realism rejects those aspects of essentialism that social constructionists find objectionable, including holding that races are immutable or have rigid boundaries. Yet it still captures what might be objectionable from a social constructivist point of view. Namely, biological racial realism holds that race has independence from human actions and intentions. To claim race is a biological category is to hold that races are part of the nature of things. Race may be socially contingent, but it is not biologically contingent. At least some of our race concepts carve (human) nature at its joints.

However, certain features of the constructionist view suggest that constructionism about race is not necessarily committed to biological anti-realism. Advocates of a biological conception of race such as Kitcher (Reference Kitcher2007, Reference Kitcher2012), Spencer (Reference Spencer2014, Reference Spencer2018), and Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) maintain the compatibility of biological race and racial constructivism.Footnote ³¹ For instance, as discussed in Section 2.2, Kitcher argues that there are conditions in which “races are both biologically real and socially constructed” and that there is no contradiction between the two positions (Kitcher Reference Kitcher2007, 298). Spencer defends an account where the racial categories set out by the 1997 OMB racial classification scheme, a case of a social institution engaged in racialization par excellence, are also biologically meaningful. Spencer’s argument is that at least some of our conceptions of race, the OMB scheme as he defends it, are biological. As such there is no opposition between his biological account and constructivism about race.

Hochman (Reference Hochman2022) argues that the compatibility of social construction with social realist, biological realist, and even anti-realist accounts of race renders the constructivist approach moot. He writes “if ‘social constructionism’ can mean almost anything, then – without further clarification – it means almost nothing. Or rather, we might say that it means whatever the reader believes it to mean, which could be any number of things” (Hochman Reference Hochman2022, 48). Regardless of whether racial constructivism fails to exclude alternative metaphysical pictures, for many constructivists, it is not merely a project for achieving descriptive accuracy about race. Rather, constructivism is partly a project with normative aims (e.g., achieving social justice) (Haslanger Reference Haslanger2000; Mallon Reference Mallon2006). As such, the metaphysical account partly depends on these normative aims. If the race concept is to help us in our normative endeavor of bringing about justice, or undoing historical injustice, one way it can do so is by aiding our understanding of social reality. Race, perhaps, allows us to limn the causal structure of our social world. It enables us to make predictions and identifies sites of intervention for our political ameliorative work.

In the next section, I turn to the putative role of social race in (social) science.

3.6 Race and Social Structural Explanation

Let’s take the explanation of how I ended up in London. Suppose I were to take a train down to London from Leeds.Footnote ³² One explanation could cite the speed at which my body was accelerating toward London, including details about what happened as I stood up to walk around or change seats, among other physical facts specific to my body. Another explanation could cite the fact that there was a train from Leeds to London and that I was on it. The latter explanation strikes us as being more valuable. It is the train that travels from Leeds to London. I am just along for the ride. This type of explanation, where we explain the behavior of an entity by explaining the behavior of a larger entity of which it is a part, is ubiquitous in the sciences.

Haslanger (Reference Haslanger2016) argues that what makes the explanation of the latter type more valuable is that they are, among other things, stable. The fact that I end up in London (the explanandum-phenomenon) is invariant across a range of interventions: whether I speed up or slow down on my way from Leeds, whether I arrive late or on time, and so on. The causal constraints on the realization of the explanandum are those of the containing entity. That is, whether or not I end up in London depends on the causal constraints and capacities of the train.

This, then, is a structural explanation. We explain the behavior of an individual part by explaining the behavior of the containing structure. Haslanger (Reference Haslanger2016) provides a characterization of structures where “structures, broadly understood, are complex entities with parts whose behavior is constrained by their relation to other parts” (Haslanger Reference Haslanger2016, 118). Structures are not mere aggregates. Unlike aggregates, the interaction between the parts of structures constrains their activity. Furthermore, it is not just physical objects that constitute structures. Social structures play a central explanatory role in the social sciences. Haslanger writes, “Structures are important to explanation because they constrain behavior of individual things insofar as they occupy nodes in the structure. The structure does not simply provide background conditions for the events in question [.], for it is the workings of the structure that are sometimes the proper object explanation[…]” (Haslanger Reference Haslanger2016, 121)

To see how, let’s take up the case of the Crafts’ flight from slavery again.

Explanandum: Why does Ellen dress like a man during their escape?

We can provide an explanation in terms of her psychological states and individual choices. Let’s call this an Individual Explanation. For instance, we can cite Ellen’s beliefs, goals, and desires to explain her choice to dress like a man. However, the Individual Explanation fails to capture explanatorily deep facts that strike us as relevant. To capture these facts, we will need to appeal to social structures. The Crafts lived in a racially stratified slave society. It was also a patriarchal society. As such, Ellen and William Craft were embedded as “nodes” within a hierarchical structure constituted by specific race and gender relations. Haslanger (Reference Haslanger2016) writes “my actions are my own, triggered by my thoughts, desires, goals. However, the resources I rely on and the meaning of my action are not mine alone but depend on the structure I’m part of. Illumination of these structures is important for explaining human action” (128). For a white woman to travel unaccompanied with her male slave would be as incongruous as seeing a black couple purchase a first-class train ticket. Crafts’ options for an escape that made use of their particular resources and attributes – Ellen’s light skin tone, William’s carpentry, their standing with their enslavers, and so on – were constrained by the patriarchal and racialized social structures they sought to flee in plain sight.

As Haslanger (Reference Haslanger2016) notes, structural constraints make it so that only some options are genuine possibilities (124). An explanation of why Ellen dressed as a man who does not cite these social structural facts would be worse than one that does. The social structural explanation illuminates the full range of normatively relevant facts. For instance, the intersection between gender and race relations represented by the fact that a white woman would not travel alone with her black slave is explanatorily related to the justifications for lynchings of black men a century later. Furthermore, the social structural explanation itself explains the individual psychological states and individual choices that constitute an alternative explanans. Ellen believes she must dress as a man as a result of the social practices and relations in which she was brought up. “The focus on structural constraints,” then, “provides resources for capturing significant regularities: those whose choices are similarly constrained will tend to act in similar ways, even if their personal histories, psychologies, and attitudes differ” (124).

What kinds of constraints are structural constraints? Ross (Reference Ross2023) proposes that at least some of these social structures impose causal constraints. She defends an interventionist account of how social structures can be causal and explanatory. On the interventionist framework, “to say that X is a cause of Y means that an intervention that changes the values of X, in some background conditions B, will produce changes in the values of Y” (Ross Reference Ross2023, 8). The intervention need not be actually possible. However, much of social science investigates social structures by means of experimental manipulation or by using “natural experiments” in which exogenous factors have changed the experimental variable. Social scientists exploit differences in variables of interest – differential roll out of new policies, disparate impact of economic shocks in different locales, and so on – to identify causal-explanatory factors. To claim that, for instance, school-based nutrition improves learning, is to hold that intervening on school-based nutrition directly changes learning outcomes.

The explanatory challenge of appealing to social structures arises because structures are composed of agents who then make choices we seek to explain by appealing to those structures. Furthermore, structural explanations seem to involve a seemingly mysterious and controversial type of causation, namely, top-down causation. A natural suggestion is that agents and the social structures they inhabit are interdependent. Social structures (causally) constrain agents and agents in turn constitute social structures (that is, social structures depend on agents for their existence). Ross (Reference Ross2023) argues that “in social structural explanations, structure and individual agency are interacting causes with respect to some effect of interest. This framework captures that structure and agency are different properties, that they take on different values, and that they depend on each other in producing outcomes” (Ross Reference Ross2023, 9).

The value of this approach lies in that it accounts for how social structures can be more explanatory, or provide more valuable explanations, than individual agency. For Ross (Reference Ross2023), “[structural] causal constraints have additional features that are not present in standard, run-of-the-mill causes. These constraints can interact with other causal factors and they exert influence on them – they guide, limit, and shape the outcomes produced by other causes” (Ross Reference Ross2023, 6). The additional features of structural causal constraints are that they (1) limit the values of the explanatory target of interest, (2) are often conceived of as separate from or external to the process they limit, (3) are considered relatively fixed compared to other explanatory factors, and (4) structure or guide the explanandum outcome, as opposed to triggering it (Ross Reference Ross2023, 10–11).

What these features together entail is that social structural explanations secure dimensions of explanatory value such as stability and proportionality. As such, explanations that cite the social role of race as a structuring phenomenon will have this value. By determining what options are available to agents in the first place, social structures carve the explanatory space to a greater degree than individual choices. In the example of the Crafts, we see agents who are severely limited in the kinds of choices available to them by the social structure of which they are a part. Although they partly compose this social structure, they do not constrain it. Social structures are relatively fixed and change over long timescales (Ross Reference Ross2023, 10). Agents are for the most part limited to selection among choices that social structures make available. Furthermore, we can see that social race is an explanatory fecund category in the social sciences (Mallon Reference Mallon2018). Unlike in the case of (minimalist) biological races, the racial level is the appropriate grain for securing dimensions of explanatory value such as depth and specific when explaining social phenomena. The conception of race as a social kind, therefore, is able to secure race as a genuine kind in a way minimalist biological racial realism cannot.Footnote ³³ If that is the case, social race is an indispensable element of our explanation of a whole host of phenomena in sociology, economics, and political science.

In the final section, I turn to anti-realist theories of race.

4.1 Mismatch Objection

The mismatch objectionFootnote ³⁷ is a prominent argument against various metaphysical proposals concerning race that exploits the distinction between how race is ordinarily understood and the revisionist accounts many philosophers defend (see Appiah Reference Appiah, Appiah and Gutmann1996; Glasgow Reference Glasgow2003, Reference Glasgow2009; Andreasen Reference Andreasen2005; Mallon Reference Mallon2006; Zack Reference Zack2014). Racial anti-realism holds there are no races because the ordinary or folk concepts of race refer to entities that do not exist. This is partly because there are conceptual constraints to which groupings of humans can be racial. Among the conceptual constraints on what constitutes ordinary races is the fact that races are “relatively large groups of people who are distinguished from other groups of people by having certain visible biological traits (such as skin colors) to a disproportionate extent” (Glasgow et al. Reference Glasgow, Haslanger, Jeffers and Spencer2019, 117). First, races are relatively large groups of people that divide humanity into a handful of groups. From Blumenbach to the United States Office of Management and Budget, R = 5 has been a common and influential division of humans into races. Let’s call this the size constraint. Of course, as we have seen in previous discussions, racial categories are not identical across time and different cultures. Racialization can increase or decrease in granularity depending on factors specific to a society. Nevertheless, as we have already seen it would violate the conceptual constraint of ordinary race to hold, for instance, that there are hundreds of human races.

Second, races are distinguished from one another by morphology. Let’s call this the visibility constraint. This is not to claim that every individual is easily identifiable as a member of a particular race. As in the case of Ellen Craft, it is possible for someone to “pass” as another race than the one to which they are ordinarily taken to belong. Rather, racial groups need to be distinguishable from other racial groups on the basis of visible biological traits such as skin color and eye shape. To motivate this condition, Glasgow (Reference Glasgow2009) asks us to imagine a scenario in which every human spontaneously begins to look roughly like the Dalai Lama. The thought is that intuitively a world of uniform physical appearance among humans is one that is devoid of human races (Glasgow Reference Glasgow2009, 34). It may be ambiguous on the basis of visible traits to which racial group an individual belongs. But it seems conceptually befuddled to claim two groupings of humans that are physically indistinguishable from each other are different races. Armed with these two points, let us see how anti-realists use them to debunk both the biological and social reality of race.

The mismatch objection is that many revisionist philosophical proposals for racial groupings fail to comport with the ordinary concept of race. They fail either on the size or the visibility constraints, or both. Among the naturalistic proposals, mismatches arise because biologically significant groupings of humans are either coarser-grained, encompassing what are traditionally taken to be multiple different races, or are much finer-grained, smaller local ancestry groups and populations, than racial categories. For revisionist accounts drawing on biology, such as populational proposals that hold that races are isolated breeding populations (Andreasen Reference Andreasen1998, Reference Andreasen2000, Reference Andreasen2004; Kitcher Reference Kitcher and Harris1999, Reference Kitcher2007), the objection is that “racial terms are applied to individuals in a way that does not map onto how science applies breeding population terms to individuals” (Glasgow Reference Glasgow2009, 95).Footnote ³⁸ Therefore, there is no biological category that is racial. Biology may subdivide humans, but it does not subdivide them into races.

4.2 Why Not Social Race?

Racial anti-realists reject not only the biological but also the social reality of races (Fields Reference Fields, Kousser and McPherson1982; Fields and Fields Reference Fields and Fields2012; Zack Reference Zack1993, Reference Zack2014; Appiah Reference Appiah, Appiah and Gutmann1996; Blum Reference Blum2002, Reference Blum2010; Glasgow Reference Glasgow2009). Social constructionism about race is open to a version of the mismatch objection. Constructionists and anti-realists share a crucial premise; the racial demarcating line is biologically arbitrary. It has no biological significance. Anti-realists go further and deny that racialization produces races even of a social kind. Anti-realist views grant that racialization, the practice of racial demarcation and differential treatment, is real. However, unlike constructionists, anti-realists deny that the resulting racialized groups are races. Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) argues that racialized groups and social races differ in key respects.

The first difference hinges on the mind-dependence of racialized groups. Racialized groups depend entirely on recognition in order to exist. Suppose we were to wake up one day, for instance, and forget all about our racial categories, then there would be no racialized groups. According to racial antirealists, in the state of amnesia about our practice of racial classification “races” would cease to exist because they depend on our social practice of racialization. Ordinary race, on the other hand, invokes a robust racial demarcation line that is invariant across a range of racialization practices. On the ordinary conception, races did not come to exist in the eighteenth century any more than the flower Scilla italica came to exist when Linnaeus first named and recorded it in his Systema Naturae. A statement such as “the origin of the black race is intimately tied up with the history of slavery and colonialism” is intuitively intelligible as a claim about when these racial classifications were made. However, the claim that the race itself emerged mere centuries ago is strikingly revisionary as an account of race. Revision is precisely what many racial constructivists propose. Racial anti-realists, however, deny that such a revision – if it severs the conceptual connection between race and biophysical difference – would result in the same concept (Zack Reference Zack1993). Such a revised account, anti-realists argue, would no longer be an account of race as ordinarily understood.

The challenge to social race arises because racial constructionists take race to depend on contingent sociopolitical factors. The ordinary race concept, anti-realists argue, takes race to be more robust. As Glasgow et al. (Reference Glasgow, Haslanger, Jeffers and Spencer2019) observes, “For all constructionists, if the relevant social facts, the ones that they think create race, are not in place, then race would disappear. This exposes all versions of constructionism to the different features problem: on the ordinary concept of race, race persists even when the social facts change” (Glasgow et al. Reference Glasgow, Haslanger, Jeffers and Spencer2019, 131). Whether is inequality, oppression, culture, or the act of racialization itself, constructionism holds that race depends on some social factor the disappearance of which would eliminate race. However, anti-realists contend, intuitively we can imagine a world of racial equality, or one where everyone has the same culture yet there are different racial groups, and so on. There is then a mismatch between the feature racial constructionism takes to be constitutive of race and the ordinary race concept. Consequently, the groups racialization produces cannot be the same groups referred to by race.

To conclude, anti-realist views of race deny that race has any reality whatsoever. Although the social practices that uphold race might be deeply entrenched and their consequences far-reaching, race would no more survive their transformation than the belief that certain scars mark a woman out as a servant of the devil. Eliminativists further argue that we can, and ought to, remove race from our common language or use it much the same way we use “witch.”