The dominant view among race scholars is that “race” is modern (Alcoff Reference Alcoff2005; Goldberg Reference Goldberg2002; Hannaford Reference Hannaford1996; Holt Reference Holt2002; Malik Reference Malik1996; Mallon Reference Mallon2017; Mills Reference Mills1998; Omi and Winant, Reference Omi and Howard2015; Smith Reference Smith2015). In recent years, this view has been increasingly challenged. According to the continuity thesis, “race” had a real purchase before modernity (Bartlett Reference Bartlett2001; Heng Reference Heng2018; Lampert Reference Lampert2004; McCoskey Reference McCoskey2012; Ramey Reference Ramey2014; Thomas Reference Thomas2010; Whitaker Reference Whitaker2015). Does this view have any merit? Is “race” modern, or is it pre-modern?
In this paper, I argue that race scholars have not been able to agree on an answer to this question partly because the question is ambiguous. There is not one question race scholars are answering when they locate “race” in historical time, but at least six. These are the questions race scholars are really answering: (1) Is the concept of race modern? (2) Is there a modern concept of race that is distinct from earlier race concepts? (3) Are “races” themselves modern? (4) Are racialized groups modern? (5) Are the means and methods associated with racialization modern? (6) Are the meanings attached to racialized traits modern? Sometimes it is clear which question race scholars are answering when they locate “race” historically, but often it is not. The last three questions tend not to be asked at all, even though the work of some race theorists is best interpreted as answering these questions.
I will offer my own answers to these six questions in this paper. My hope, of course, is to convince the reader that these answers are correct, and thus to offer a viable solution to the debate about “race” and periodicity. However, this is a secondary aspiration, and the success of the paper does not turn on whether my particular answers are accepted. My primary goal for the paper is to show that disambiguating the question “is ‘race’ modern?” is helpful for thinking about ‘race’ and racialization historically.
It is worth noting from the outset that just as some race theorists have been questioning the modernity of “race,” some historians have been questioning periodization itself. As leading continuity theorist Geraldine Heng (Reference Heng2018) explains, these historians,
working against the grain, have attempted to disrupt the narrative of scientific, economic, and demographic transformation separating modern from premodern time in the West… they have argued for the legitimacy of complex, nonlinear temporalities: temporalities in which multiple modernities have recurred in different vectors of the world moving at different rates of speed within macrohistorical time. One position… [sees] human history as punctuated by scientific and technological “efflorescences” that, coupled with labor specialization and intensive market orientation, have driven both Schumpeterian and Smithian growth and change in various societies and various eras, thus muddying the monomythic simplicity of a radical break favoring the West in modernity’s singular arrival (p. 21).
The periodization of history into ancient, medieval, and modern periods has been maintained despite the critiques to which Heng alludes (e.g. Abu-Lughod Reference Abu-Lughod1991; Goldstone Reference Goldstone2002). This is perhaps unsurprising, given the entrenched structuring of both university history departments and professional history journals according to the tripartite model.
While Heng is a staunch critic of periodization, she still names her recent book The Invention of Race in the European Middle Ages (2018). In light of her critical stance, we can view Heng’s periodization as merely conventional. That is, she uses the tripartite division of history without believing that history comes neatly parceled into three coherent units of time. While it may be useful to eventually move away from the tripartite model, I will also adopt a conventionalist attitude toward periodization in this paper. Specifically, I will adopt the conventional dating of the beginning of modernity at 1492, the year in which Columbus arrived in the “New World” and the Jews were expelled from Spain.
In response to the first question, which asks whether the concept of race is modern, I will answer in the negative: the concept of race is late medieval. It appears to have emerged for the first time at around the middle of the fifteenth century, in Spain. In response to the second question, which asks whether there is a modern concept of race that is distinct from earlier race concepts, I will answer in the affirmative: the concept of race changes enough for it to be useful to distinguish between modern and pre-modern race concepts. If my answers to the first two questions are correct, it should already be clear why asking whether “race” is modern needs disambiguation. If we simply ask whether the concept of race simpliciter is modern, we collapse questions one and two. Because questions one and two have different answers—or so I will argue—this is a recipe for confusion.
The question “is ‘race’ modern?” is ambiguous in another way. “Race” is not only a concept; it is also a purported kind. The third question asks whether “races” themselves are modern. How we answer this question will depend on how we define “race” and whether there are any human races. If “race” is defined biologically, and there are human biological races, then races would be transhistorical groups. However, racial naturalism is increasingly rejected by scientists, at least in the West (see Wagner et al., Reference Wagner, Yu, Ifekwunigwe, Harrell, Bamshad and Royal2017). I find the arguments against racial naturalism convincing, and for the purposes of this paper I will assume that “race” is not a biological kind.Footnote 1
The non-existence of biological races does not necessarily mean that we cannot sensibly ask whether races themselves are modern. As W. E. B. Du Bois wrote in The Conservation of Races (1897), the “races… perhaps transcend scientific definition” but “nevertheless, are clearly defined to the eye of the Historian and Sociologist” (p. 110). Analysis and evaluation of this claim have been central to the establishment of “race” as a topic of philosophical investigation. In The Uncompleted Argument: Du Bois and the Illusion of Race , Kwame Anthony Appiah (Reference Appiah1985) argues that Du Bois didn’t go far enough in his racial skepticism. Appiah argues that like biologists, historians and sociologists should reject race as an illusion. In response, Paul Taylor (Reference Taylor2000) criticizes Appiah for ushering “Du Bois into the light mainly to make visible what appear to him to be blemishes” (p. 103). Taylor makes the case that Appiah’s critique of Du Bois’s racial realism fails, and that Du Bois’s “sociohistorical racialism is a viable conceptual option” (2000, p. 128). Indeed, the view that race is a social kind is now dominant among race scholars (Diaz-Leon Reference Diaz-Leon2015; Haslanger Reference Haslanger2012; Mallon Reference Mallon2017; Mills Reference Mills1998; Omi and Winant, Reference Omi and Howard2015; Sundstrom Reference Sundstrom2002; Taylor Reference Taylor2000).
The third question can thus be interpreted as asking, “are social races modern?” My answer to this question will be the most controversial of all of my answers. Elsewhere I have argued that when a social approach to race is applied consistently, and taken to its logical conclusion, it has unacceptable theoretical consequences (Hochman Reference Hochman2017, Reference Hochman2019b). I make the case that when race is defined as a social kind, too many groups count as races and race loses its historical specificity.Footnote 2 So I will argue that races themselves are not modern, because “races” do not exist.
The third question should be replaced by the fourth question: “are racialized groups modern?” The terms “race” and “racialized group” are usually used as synonyms (e.g. Goldberg Reference Goldberg1992; Haslanger Reference Haslanger2000; Mallon Reference Mallon2018). However, there is an alternative tradition that distinguishes these concepts. In this tradition, “racialized group” is a category adopted by race skeptics as an alternative to “race” (see Blum Reference Blum2002, Reference Blum2010; Darder and Torres, Reference Darder, Torres, Darder, Baltodano and Rodolfo D2003; Hochman Reference Hochman2017, Reference Hochman2019b; Miles Reference Miles1988, Reference Miles1989; Small Reference Small1994). “Race” is understood as a failed biological category, whereas a racialized group is conceptualized as a group that has been misinterpreted as a biological race. In response to the fourth question, I will argue that racialized groups first formed in the mid-fifteenth century, in Spain. If this is right, and we adopt the convention that modernity begins in 1492, or indeed any later date, then racialized groups are not modern.
The fifth question asks whether the means and methods associated with racialization are modern. When race scholars find deep continuities between modern racist practices and earlier forms of discrimination they are tempted to claim that “race” is pre-modern. What they actually show, I will argue, is that some of the means and methods associated with racialization are pre-modern.
Similarly, when race scholars find historical continuities in the social meanings that are attached, for example, to skin color, they are tempted to claim that “race” is pre-modern (Hahn Reference Hahn2001; Whitaker Reference Whitaker2015). Skin color is, after all, the paradigmatic “racial” trait. However, I will argue that continuity theorists working on skin color are not well understood as answering the general question, “is ‘race’ modern?” Rather, they are best interpreted as answering the sixth question: “are the meanings attached to racialized traits modern?” They show that some of the meanings attached to some racialized traits are pre-modern in origin, although this is not to say that the significance attached to those traits is exactly the same throughout history.
These questions about the history of “race” and racialization are important for theoretical and political reasons. On the theoretical side, it is crucial to distinguish between the six questions, because asking whether “race” is modern is insufficiently nuanced. There are both historical continuities and discontinuities relevant to racial concepts and racialized practices (see Hanchard Reference Hanchard2018). My six questions can help us to capture both the relevant continuities and the relevant discontinuities.
These theoretical issues feed into those of a political nature. A primary motivation for studying “race” is to help us understand and fight racism. If we want to know the historical conditions under which racism flourishes, then we need to know the history of racialization. As Heng (Reference Heng2018) notes, “Racial formation has been twinned with conditions of labor and capital in modernity such as plantation slavery and the slave trade, the rise of capitalism or bourgeois hegemony, or modern political formations such as the state and its apparatuses (we think of David Theo Goldberg’s magisterial The Racial State), nations and nationalisms (Étienne Balibar’s chapters in Race, Nation, Class), liberal politics (Uday Mehta), new discourses of class and social war (Foucault of the 1975–6 Collège de France lectures), colonialism and imperialism (the work of many of us in postcolonial studies), and globalism and transnational networks (Thomas Holt on race in the global economy)” (pp. 16–17). However, if racialization precedes these allegedly modern institutions, then racism cannot depend on them. As a consequence, their abolition would not necessarily abolish racism. For reasons such as this, understanding the history of “race” and racialization may be helpful in the fight against racism.
IS THE CONCEPT OF RACE MODERN?
Let us consider each question in turn, beginning with the first, which asks whether the concept of race is modern. This question is often conflated with the second question, which asks whether there is a modern concept of race that is distinct from earlier race concepts. As Denise McCoskey (Reference McCoskey2006) writes, “So powerful and persistent is the “scientific” version of race, with its unyielding reference to the body’s surface and, above all, to skin color, that, as many scholars have argued, we tend to confuse it with the concept of race itself” (p. 250). If race is understood to be a modern scientific concept, then there is no interesting debate to be had about the possible pre-modern origins of “race.” However, if we do not foreclose the debate by definitional fiat, we can meaningfully ask whether there was a concept of race before modernity.
As David Nirenberg (Reference Nirenberg, Eliav-Feldon, Isaac and Ziegler2009) explains, when it comes to the modernity of the race concept, “there is room for doubt, and the scholarly expression of that doubt tends to cluster around Spain in the late Middle Ages” (p. 241). Most historians of race, if they find the Middle Ages significant at all, focus on the Purity of Blood statutes of fifteenth-century Spain, which led to the expulsion of all Jews from the country in 1492. This date, also marking the arrival of Columbus in the “New World,” is often conventionally considered to mark the beginning of modernity. Were the Purity of Blood statutes (first issued in 1449) premised on a concept of race? If they were, then there was a pre-modern concept of race.
For Ivan Hannaford (Reference Hannaford1996)—author of the only history of the idea of race in the west from antiquity to modernity—“it is unhistorical to perceive the concept of race before the appearance of physical anthropology proper, because the human body, as portrayed up to the time of the Renaissance and the Reformation, could not be detached from the ideas of polis and ecclesia” (p. 147). While it is true that in fifteenth-century Spain, beliefs about Christian, Muslim, and Jewish bodies were attached to ecclesiastical ideas, there was a biological turn during this period that can arguably be understood as the beginnings of racial thought.
This biological turn had to do with the crisis of classification that followed large-scale Jewish conversions to Christianity after the summer of 1391, during which Jews were attacked and robbed throughout Spain (Roth Reference Roth1992). “By the middle of the fifteenth century,” writes Norman Roth (Reference Roth1992), “a substantial converso class had been created. All doors were open to them as Christians, including those of the Church” (p. 22). It was not long before the Cristianos nuevos (new Christians) ascended to positions of power and prestige barred to them and their ancestors as Jews—positions that Cristianos viejos (old Christians) fought ruthlessly to preserve.
Cristianos viejos, also calling themselves Cristianos de natura, complained that it was impossible to tell a “natural Christian” from a “Jewish or Moorish convert.” “Failing in their efforts to enforce a legal distinction between old and new Christians, converso-phobes,” explains Roth,
evolved the first medieval example of true anti-Semitism, as opposed to anti-Jewish sentiment based on objections to Jewish religious tradition. According to the notorious doctrine of limpieza de sangre (purity of blood), Jews and Jewish converts to Christianity constituted a race. Jewish blood irreconcilably corrupted its possessor down to the fourth generation, despite intermarriage with old Christians. This doctrine removed the sincerity of converts and their descendants from consideration and made opposition to Jews a biological issue. Since “Jewish blood” remained to the fourth generation, such people were to be barred from holding public or ecclesiastical office or from studying in universities (1992, p. 23 emphasis added).
Did the Purity of Blood Statutes constitute racial thinking, as Roth believes? I can think of two ways of resisting this conclusion, but neither of them is convincing. First, one could identify “race” exclusively with the scientific race theory to come, but this would be pure scientism. Second, one could identify “race” with color-coded, continental groupings rather than the religious “bloodlines” (the Jewish and Moorish “razas”) of fifteenth-century Spain. However, this presents us with a false dichotomy. As George Fredrickson (Reference Fredrickson2002) explains, we can “trace the origins of the two main forms of modern racism—the color-coded white supremacist variety and the essentialist version of antisemitism—to the late medieval and early modern periods” (p. 46). Just as neither version of racism is more racist than the other, color-coded “races” are no more “racial” than the “razas” of fifteenth-century Spain or Jews in Nazi Germany. It would be arbitrary to conceptually privilege one type of racialization over the other. As Roth argues, the Purity of Blood Statutes did appeal to a concept of race, just not one that involved color-coding.
There is now widespread agreement that, as Lynn Ramey (Reference Ramey2014) writes, “It is not necessary to have the word ‘race’ to have the concept of race” (p. 26).Footnote 3 However, it turns out that the concept of race emerges at the same time as the term is first applied to human groups. As Nirenberg (Reference Nirenberg, Eliav-Feldon, Isaac and Ziegler2009) explains,
…words like raza…were already embedded in identifiably biological ideas about animal breeding and reproduction in the first half of the fifteenth century…the sudden and explicit application of this vocabulary to Jews coincides chronologically (the 1430s) with the appearance of an anti-converso ideology…which sought to establish new religious categories and discriminations, and legitimate these by naturalizing their reproduction (p. 252).
Perhaps for the first time in history, the potency of baptismal water was seriously brought into question. While this was a complex and ambivalent development, Jewish and Muslim identities began to be understood as “in the blood.” And blood, it turned out, was thicker than baptismal water. Jews and Muslims were not only labelled as “razas,” they were understood to be major human groups and lineages, formed due to reproductive isolation, in which group membership was biologically inherited. This marks a break with ancient and earlier medieval approaches to group membership. A change in religion, or a change in climate, could no longer change one’s bodily essence. It is unclear what we might call this new approach to group identity, where membership is determined by “blood,” if not by the name of “race.” So in response to our first question, which asks whether the concept of race is modern, we should answer in the negative. “Race” is not a modern concept.
IS THERE A MODERN CONCEPT OF RACE THAT IS DISTINCT FROM EARLIER RACE CONCEPTS?
Our second question asks whether there is a modern concept of race that is distinct from the earlier race concepts. As we have seen, a concept developed in the late Middle Ages which was recognizably racial. This concept contained the four following components:
(1) Races form due to their reproductive isolation.
(2) Races constitute major human lineages.
(3) Races are major biological groups in the present.
(4) Race is biologically inherited.
This may seem, at first glance, sufficiently similar to modern conceptions of race to conclude that there is not a uniquely modern concept of race. However, such a conclusion should be resisted.
The following two claims distinguish modern from pre-modern concepts of race:
(5) All human populations belong to at least one race.
(6) Race is the taxonomic level below species in humans.
Modern ideas about “race” are relational in a way that their medieval counterparts were not. “Race” came to be seen as something that not only some populations had: every population was thought to be a “racial” population (5). There may be racially mixed populations, but not raceless populations. “Race” also came to be seen in a particular relation to the species concept—namely, as the taxonomic level below it (6).
These were genuinely modern innovations. Before 1492, there was no attempt to divide the entire human species into biological lineages (5). Even in the early seventeenth century, Sebastian de Covarrubias defines “raza” in his dictionary as follows: “the caste of purebred horses, which are marked by a brand so that they can be recognized…. Race in [human] lineages is meant negatively, as in having some race of Moor or Jew” (as cited by Nirenberg Reference Nirenberg, Eliav-Feldon, Isaac and Ziegler2009, p. 251). Only some groups were believed to be “racial” groups.
And before 1492, “race” was not considered to be the taxonomic level below species (6). There was no attempt to describe what we would now call human biological races or subspecies in antiquity or the Middle Ages. Aristotle was very clear about this. In Parts of Animals (1984) he observed that the human species is not divisible into smaller groups:
By ‘common’ I mean those which belong to all animals; by ‘to a genus’, those of animals whose differences from one another we see to be matters of degree—Bird is a genus. Man is a species, and so is everything not differentiated into subordinate groups. In the first place the common attributes may be called analogous, in the second generic, in the third specific” (p. 1005).
A genus was seen as a broad class, divisible into species. The human species could not, according to Aristotle, be divided into subordinate, or subspecific groups.
We can see the early development of modern race theory, and the first attempt to racialize the world, in François Bernier’s (Reference Bernier, Bernasconi and Lott1684) New Division of the Earth. While Bernier tells us that everyone (not just the Moors and the Jews) is raced (5), he does not distinguish between the concepts of “race” and “species” (6). His paper bears the full title Nouvelle Division de la Terre par les differentes Especes ou Races d’hommes qui l’habitent, envoyée par un fameux Voyageur à M. l’Abbé de la Chambre. This lack of distinction between species and race (found in the extended title and evident throughout his discussion) demonstrates that the theoretical status of “race” was still unclear.
In 1950, Walter Scheidt argued that Immanuel Kant produced “the first theory of race which really merits that name,” (p. 372) and subsequent scholarship has reinforced this view (see also Bernasconi Reference Bernasconi and Bernasconi2001; Eze Reference Eze and Eze1997; Larrimore Reference Larrimore2008; Sloan Reference Sloan2014). While Claude-Olivier Doron (Reference Doron2012) emphasizes Comte de Buffon’s role in the formation of the modern race concept, he still acknowledges that “Kant… gave Buffon’s claims a decisive epistemological status” (p. 91).
What is so unique about Kant’s anthropological work to single him out as such an important figure in the history of “race?” It is not his influence, although he was influential (Bernasconi Reference Bernasconi and Bernasconi2001). Nor is it his radical break with earlier race theory, although his theory was unique in some respects (Mensch Reference Mensch2013). It is, rather, that Kant offered a division of the world’s human populations into four “races” (5) and he also rigorously distinguished between “race” and “species” (6). Even if confusion remained about whether the “races,” or some of them, could be separate species,Footnote 4 Kant gave “the concept sufficient definition for subsequent users to believe that they were addressing something whose scientific status could at least be debated” (Bernasconi Reference Bernasconi and Bernasconi2001, p. 11). Indeed, Kant’s student Johann Gottfried von Herder (Reference Herder, Bernasconi and Lott1828) wrote a polemical reply, arguing that “there are neither four nor five races… The colors run into one another” (p. 26). The race debate had begun, and race skeptics have been making the same argument as Herder ever since.
There is enough of a difference between pre-modern and modern race concepts to justify a distinction between them. And in doing so, we have our first example of how asking “is ‘race’ modern?” without qualification gets us into trouble. If we are asking whether the concept of race is modern, then the answer is “no,” but if we are asking whether there is a modern concept of race that is distinct from the earlier race concepts, then the answer is “yes.”
Consider the following claims:
Race is a modern concept and a product of colonial encounters (Golash-Boza Reference Golash-Boza2016, p. 130).
Race is a modern category invented by white male scientists in the “era of modernity” (Brewer Reference Brewer2006, p. 513).
...‘[R]ace’ is a modern construct dependent on the language of science (Dikötter Reference Dikötter2015, p. ix).
These theorists are all right in one sense, and wrong in another. They are right that there is a distinctively modern race concept that was a product of colonial encounters, and dependent on the language of science. However, they are wrong to identify this as the race concept. There was, as I have argued, a race concept that evolved earlier, in medieval Spain. So we should distinguish between our first question, “Is the concept of race modern?” and the second question, “Is there is a modern concept of race that is distinct from the earlier race concepts?” The answer to the first is “no,” while the answer the second is “yes.”
ARE “RACES” THEMSELVES MODERN?
Question three asks whether “races” themselves are modern. If we simply ask whether “race” is modern, it is unclear whether we are asking about the concept or the purported kind. Yet many race theorists are interested in when “races” themselves were first formed. How one answers this question will depend on whether one believes that “races” actually exist. In this section, I will argue that there are no races, and that the third question should be rejected, before arguing in the following section that it should be replaced with the fourth question, which is about “racialized groups” as distinct from “races.”
Note that when race scholars ask whether “races” themselves are modern, they are not talking about race as a biological category. Most race theorists reject racial naturalism, so that question is not one that they would ask. Among the remaining race naturalists, it is universally accepted that so-called “racial boundaries” have been increasingly blurred since modernity (e.g., Hardimon Reference Hardimon2017; Spencer Reference Spencer2018a, Reference Spencer2018b). Nobody believes that biological races were formed in modernity.
Yet most race theorists do believe that “race,” defined as a social kind, is a product of modernity (e.g. Goldberg Reference Goldberg2002; Holt Reference Holt2002; Mallon Reference Mallon2017; Mills Reference Mills1998; Omi and Winant, Reference Omi and Howard2015).Footnote 5 The view that race is a social kind is often labelled “social constructionism about race,” and I will be using this phrase to refer to the position.Footnote 6 Social constructionists believe that the modern scientific and taxonomic project of describing “biological races” had the unintended effect of creating “social races.” The descriptive project was a failure, but the very act of labelling groups as races and treating them in different—and often horrendous—ways created a new social kind, which we call “race.”
While this is the dominant account of the history of racial formation, it is not beyond doubt. We can thus ask whether “races,” as social groups, are indeed modern, or whether there were “social races” before modernity. To answer this question, we need a sense of what it means to say that “race” is real, but social. Ronald Sundstrom (Reference Sundstrom2002) offers a helpful account. He writes that, “A kind is real when its members are unified—held together by a number of shared properties. The properties that unify a kind vary with the domain…. A biological kind is unified by some significant biological relation.… Similarly, a human [or social] kind is real when unified by a significant social relation” (p. 95). So in order to answer our third question, which asks whether “races” themselves are modern, we need an account of “racial” social relations.
It is common for such accounts to focus on essentialism as essential to racial social relations (e.g. Fredrickson Reference Fredrickson2002). Consider Heng’s (Reference Heng2018) definition: “‘Race’ is one of the primary names we have—a name we retain for the strategic, epistemological, and political commitments it recognizes—that is attached to a repeating tendency, of the gravest import, to demarcate human beings through differences among humans that are selectively essentialized as absolute and fundamental, in order to distribute positions and powers differentially to human groups” (p. 3).
Are social races modern according to this definition of race? Heng (Reference Heng2018) argues that there were social races as early as the thirteenth century. Her focus is on Jewish-Christian relations in thirteenth-century England, starting with the execution of the law (set out in the Fourth Lateran Council of 1215) that Jews and Muslims be set apart from Christians by a difference in dress and ending with the forced expulsion of all Jews from England in 1290:
In England…the Jewish badge, expulsion order, legislative enforcements, surveillance and segregation, ritualized iterations of homicidal fables, and the legal execution of Jews are constitutive acts in the consolidation of a community of Christian English—otherwise internally fragmented and ranged along numerous divides—against a minority population that has, on these historical occasions and through these institutions and practices, entered into race (Heng Reference Heng2018, p. 31).
The treatment of the thirteenth-century English Jewry foreshadows the twentieth century’s holocaust. Being a Jew in this period and place was much more than following a certain religious doctrine. It meant being positioned in a very particular location in the social hierarchy with economic, political, and legal consequences. Jews were essentialized, although imagined Jewish essences could transform though conversion (Resnick Reference Resnick2000). On Heng’s definition of race, thirteenth-century English Jews do indeed count as a racial group.
If we were to stop the discussion here, we would be answering “no” to the third question, which asks whether “races” themselves are modern. However, I have argued elsewhere that there is a problem with the definition of race above (Hochman Reference Hochman2017, Reference Hochman2019b). I argue that, on this definition, “race” loses its conceptual specificity. It creates false positives—allowing too many kinds of groups to count as races. We essentialize all sorts of groups in order to differentially distribute power between them. Ethnic groups are often essentialized, and so unracialized but essentialized ethnic groups would count as de facto races. Men and women are also essentialized and would belong to different races on the definition above. The same would be true of gay people, under certain conditions—a consequence that Heng (Reference Heng2018) recognizes and appears to accept. The idea of a “gay race” goes beyond the sort of counterintuitive position that some philosophers enjoy defending. Instead, it amounts to a reductio ad absurdum. A social kind approach to race, applied consistently and taken to its logical conclusion, has unacceptable consequences. When “race” loses its conceptual specificity, it loses its value as a conceptual resource (Hochman Reference Hochman2017, Reference Hochman2019b).
Social constructionists might reply by arguing that the definition above only produces these counterintuitive consequences because it does not mention the body, or biology. A “race” is not simply a social group, it is a social group misinterpreted as a specific kind of biological group. As creators of racial formation theory Michael Omi and Howard Winant (Reference Omi and Howard2015) put it, “race is a concept which signifies and symbolizes social conflicts and interests by referring to different types of human bodies” (p. 110). This sounds right at first, but on reflection, it would seem to entail that males and females count as separate “races.” However, couldn’t a more nuanced definition solve the specificity problem?
I do not believe that it could. As Heng has argued, canonical race theory has been blind to its own radicality. “So tenacious has been scientific racism’s account of race, with its entrenchment of high modernist racism as the template of all racisms, that it is still routinely understood, in everyday life and much of scholarship, that properly racial logic and behavior must invoke biology and the body as their referent” (Heng Reference Heng2018, p. 19). That is a mistake. Once race is redefined as a social kind, we can no longer assume that biologization is necessary for racial formation. In other words, social constructionists cannot simply help themselves to criteria of biologization in their definitions of social race. If biologization is not necessary for the formation of groups with the social properties and relations constructionists identify as “racial,” then such biologization is not definitional of race. What Heng shows with her example of Christian-Jewish relations in thirteenth-century England is that it is possible to produce the sort of social groups that constructionists call “races” without the biologization usually associated with “race.” Yet when we strip definitions of race of these unjustified references to biologization, a social kind approach to race is susceptible to a reductio argument.
Social constructionists could try to avoid the reductio by using more nuanced social criteria. This is exactly how Phila Msimang (Reference Msimang2019) defends social constructionism in his response to my critique. Msimang offers what he calls the minimalist account of social race. He maintains that, “The characteristics of a social race are that they are stereotyped groups,” about which there is a “presumption that racial identity is fixed” and a “belief that race is a genealogical and heritable kind of group belonging” (Msimang Reference Msimang2019, p. 15). This shows how constructionists can avoid the implication that the sexes are races (sex does not fit the genealogy condition, and the presumption that sex is fixed is weakening). However, Msimang’s account does not avoid the reductio. The thirteenth-century English Jews would still count as a race, as they saw themselves as the chosen people (thus meeting the stereotyping condition) and they understood Jewishness as fixed and matrilineal (meeting the fixity and genealogy conditions). Gay people would also count as a race in certain contexts, because stereotypes about gay people are common, and homosexuality has often been understood as fixed and heritable. Many other sorts of groups that are surely not races meet Msimang’s three conditions of stereotyping, fixity, and genealogy as well, including unracialized ethnic groups. Msimang’s minimalist account of social race fails to save race as an analytic category.
Social constructionists might respond by looking for alternatives to a social kind interpretation of their position. For instance, they could defend the view that social races exist as socially ascribed groups. However, social constructionists have avoided this position because of their legitimate concerns about reification. Constructionists do not want to say that races are simply those groups that have been labelled as races, because this would be akin to the claim that witches are simply people accused of being witches. There must be something more, they believe, to being a race. The claim that race is a social kind is a way of capturing this “something more.” However, as I have argued, when race is defined socially, it loses its conceptual specificity and its usefulness as an analytic category. As such, we ought to reject social definitions of race altogether. In response to our third question, “are races themselves modern?” we should reject the question itself. Races are not anything, because races do not exist.
ARE RACIALIZED GROUPS MODERN?
If races don’t exist, is our race-talk tracking anything at all? Appiah (Reference Appiah, Appiah and Gutmann1996) has argued that while there are no races, there are racial identities. However, this formulation does not quite capture the way in which groups have been racialized. I believe that race-talk is best understood as tracking racialized groups (see also Blum Reference Blum2002, Reference Blum2010; Darder and Torres, Reference Darder, Torres, Darder, Baltodano and Rodolfo D2003; Hochman Reference Hochman2017, Reference Hochman2019b; Miles Reference Miles1989; Small Reference Small1994). Lawrence D. Bobo (Reference Bobo, Koenig, Soo-Jin Lee and Richardson2008) explains that, “like most social scientists, I have grown comfortable in the thought that we study race without actually believing in race” (p. x). However, the concept of racialization offers a way out of this apparent paradox or tension. Social scientists study race concepts, racialization, racialized groups, and racism, but not “race” itself. The third question—which asks whether “races” themselves are modern—should be replaced by the fourth, which asks whether racialized groups are modern.
What, then, is the difference between a “race” and a racialized group? For most race theorists, there is no such distinction—the terms are used as synonyms (e.g. Goldberg Reference Goldberg1992; Haslanger Reference Haslanger2000; Mallon Reference Mallon2018). The distinction I wish to draw is therefore prescriptive, rather than descriptive. That distinction is as follows: “race” is a purported biological or social kind, whereas a racialized group is a group which is misunderstood to be a biological race.Footnote 7 More specifically, a racialized group is a group understood to be a major biological entity and human lineage, formed through reproductive isolation, in which membership is transmitted by biological descent.
I define “racialized groups” as groups subject to a specific form of biologization because this offers the kind of conceptual and historical specificity that is lost on a social definition of “race.” Recall that the thirteenth-century English Jews appear to count as a “race” according to a social-kind approach to race. Let’s see whether they were a racialized group according to my definition of the term. Given the numerous Christian beliefs about Jewish bodies in the Middle Ages—that they had horns, an identifying stink, etc.—this is not an unreasonable question to ask. However, in doing so, it is important to distinguish between biologization in general and the very specific form of biologization that is racialization. As I will argue, the thirteenth-century Jews were biologized, but they were not racialized.
Consider the medieval myth of Jewish male menstruation. As Irven Resnick (Reference Resnick2000) explains, while Christian beliefs about Jewish male menstruation focused on the body, Jewish physiology was not believed to be the cause of the bleeding. The cause was believed to be mystical rather than biological. Jewish male menstruation was understood to be the effect of a curse for alleged deicide, and the imagined bleeding was thought to cease upon conversion to Christianity. So long as Jewish bodies could transform through conversion, Jews could be biologized, but they could not be racialized. Racialization is a very specific form of biologization, according to which membership in a biological lineage—not one’s religion or environment—determines one’s physical traits.
So thirteenth-century Jews were not a racialized group. When did racialization first take place? Above I argued that the concept of race first emerges in the late Middle Ages, in Spain, and that this concept was applied to the Jews and the Moors. If a racialized group is a group understood to be a biological race, then racialization first took place in the mid-fifteenth century, in Spain. The labelling of the Jews (and also the Moors) as “razas,” and naturalizing their reproduction as “in the blood,” can be usefully understood as an instance of racialization. In response to the fourth question we should answer in the negative. Racialized groups are not modern—they are late medieval.
We can see the conflation of “race” and racialized group—the third question with the fourth—in James Thomas’ (Reference Thomas2010) article The Racial Formation of Medieval Jews: A Challenge to the Field. Racial formation theory endorses a realist social ontology of race (Omi and Winant, Reference Omi and Howard2015). Racial formation is supposed to be the process or processes through which social races are formed (which is usually—but not always—understood to involve biologization). Thomas claims, as the title of his article indicates, that racial formation took place in the Middle Ages. However, he does not appear to argue that medieval Jews were actually a race. He writes, for instance, that “the denial of all New Christians in fifteenth-century Spain the right to public office for all time, present and future, was predicated on the belief that Jews were a race that, even through conversion, could not be cured of their innate Jewishness” (Thomas Reference Thomas2010, p. 1752). The key point here is that fifteenth-century Spanish Jews were understood to be a race, not that they in fact were a race. Thomas is not arguing that racial formation—the formation of races—is medieval. Rather, he is arguing that racialization is medieval.
Where racial formation is supposed to form “races,” racialization is best understood as the process or processes through which racialized groups are formed.Footnote 8 This distinction is seldom recognized. Just as most race theorists use “race” and “racialized group” as synonyms, they use “racial formation” and “racialization” synonymously as well (e.g. Goldberg Reference Goldberg1992; Haslanger Reference Haslanger2000; Mallon Reference Mallon2018). However, as I noted above, there is a tradition of race skeptics who use the term “racialization” as an alternative to race-realist concepts such as “racial formation” (see Blum Reference Blum2002, Reference Blum2010; Darder and Torres, Reference Darder, Torres, Darder, Baltodano and Rodolfo D2003; Hochman Reference Hochman2017, Reference Hochman2019b; Miles Reference Miles1988, Reference Miles1989; Small Reference Small1994).
Maintaining a distinction between “racialization and “racial formation” is crucial. The racialization that Thomas describes in fifteenth-century Spain did not bring a social kind (“race”) into existence, so it did not constitute racial formation. There is too much continuity on the social level to maintain that racialization created a new social kind called race. In both the English and the Spanish cases, Jews were made to wear identifying badges, they were the subjects of homicidal fables, they were legally executed, and they were all eventually expelled from their countries. The central difference, from a social perspective, was that in the thirteenth-century, Jews could still convert. However, there were already social barriers to conversion, and as Robert Stacey (Reference Stacey1992) explains, “By the middle of the thirteenth century in England, there was clearly an irreducible [though non-biological] element to Jewish identity in the eyes of many Christians, which no amount of baptismal water could entirely eradicate…. Through baptism, converts from Judaism became Christians, but this did not mean that they had entirely ceased to be Jews in the eyes of their brothers and sisters in Christ” (p. 278).
The similarities between the treatment of Jews in thirteenth-century England and fifteenth-century Spain are striking. It would be difficult to maintain that in the former they were an ethno-religious grouping but that in the latter they were a totally new social kind—a “race.” Racialization was a new, biological, interpretation of identity. This biologization had consequences—some of the most terrible consequences of world history—but the formation of a new social kind was not one of them.
If my analysis is good, then we should reject the third question, which asks whether races themselves are modern, along with race-realist concepts such as “racial formation.” We should ask, rather, whether racialized groups are modern. I have argued that they are not: that racialization first took place in the middle of the fifteenth-century, in Spain. If we adopt the conventional dating of modernity as beginning in 1492, then racialized groups are pre-modern in origin.
ARE THE MEANS AND METHODS ASSOCIATED WITH RACIALIZATION MODERN?
Let us move on, then, to the fifth question: “are the means and methods associated with racialization modern?” Continuity theorists who are ostensibly answering the third question, “are ‘races’ themselves modern?” can be read as answering, among other questions, the fifth question. If this is right—and I will make the case that it is—it further demonstrates the value of disambiguating the question “is ‘race’ modern?” While the work of continuity theorists is certainly taken seriously by many, others will dismiss it as presentist and anachronistic (Jordan Reference Jordan2001; Seth Reference Seth2005). Yet if we interpret the continuity theorists as investigating the origins of the means and methods associated with racialization, rather than as discovering pre-modern “races,” the relevance and value of their work should become clear.
Discontinuity theorists—who believe that “race” is modern—argue that the means and methods associated with racialization are unique to modernity. “Race” is modern, discontinuity theorists argue, because it is intimately linked with the industrialization and global capitalism that they identify as uniquely modern, along with the rise of the modern nation state (Holt Reference Holt2002). It was the brutal exploitation by Europeans of, in particular, sub-Saharan Africans and Native Americans that partially financed the industrialization associated with modernity, while the transatlantic slave trade paved the way for global capitalism. As for modern nation states, these are, as David Theo Goldberg (Reference Goldberg2002) argues, racial states: states ordered around racial divisions and hierarchies, aided by law, bureaucratic apparatuses such as the census, and invented histories.
The work of continuity theorists challenges the dominant view that “Modernity and Race helped bring each other into existence” (Taylor Reference Taylor2004, p. 23). Thomas (Reference Thomas2010) explains that, “By and large the study of race and racism within the social sciences seeks to understand the development and maintenance of racial classificatory systems through a model of imperialist global economic expansion and development, rather than seeing it as a historically complex process over a much larger period of time” (p. 1752). While I have argued that continuity theorists such as Thomas are mistaken in their belief that racial formation occurred before modernity (because there are no races, only racialized groups), they have, I believe, successfully demonstrated that some of the means and methods associated with racialization are pre-modern.
To illustrate with an example, let us consider the claim that “race” is modern because of its connection to the modern nation state. “Race,” writes David Theo Goldberg (Reference Goldberg2002), “is not a premodern condition but a quintessentially modern one masquerading in the guise of the given and the ancient, bloodlines and genetic pools. States have acquired their modernity more or less and partially through racial assumption, through being drawn into the terms and forms, shapes and spaces, temporalities and rhythms of racial world ordering and world racial definition” (p. 240). If we follow Goldberg, “race” must be modern, because it developed hand-in-hand with modern nation states.
In the quote above, the question “is ‘race’ modern?” appears as a question not about a concept or a kind, but a “condition.” It is difficult to evaluate Goldberg’s claim without knowing what it means for “race” to be a “condition.” Nevertheless, in medieval Europe the church acted in many ways like a modern nation state. “The church’s bid for an overarching authority and uniformity,” writes Heng (Reference Heng2018), “importantly furnished medieval societies with an array of models on how to consolidate unity, power, and collective identity across internal differences. A church with universalist ambitions in effect sought to function like a state, a state without borders” (p. 32). If the medieval church acted like modern nation states, and modern nation states are racialized states, then when it comes to racialization, the so-called “dark ages” can shed light on the present.
Continuity theorists, I have argued, are mistaken to claim that thirteenth-century English Jews were a “social race.” They were not even a racialized group, but were rather an ethno-religious group subject to structural persecution. However, the ways in which thirteenth-century English Jews were persecuted informed the racialization of later medieval and modern periods. Jews were already made to wear identifying badges in thirteenth-century England; laws were enforced that restricted the sorts of work they were allowed to perform; there were state-endorsed executions; and finally, in 1290, all Jews were exiled from the country. There is a striking continuity in the treatment of the Jewish populations in thirteenth-century England, fifteenth-century Spain, and twentieth-century Germany. It was in fifteenth-century Spain that Jews were first racialized, but this racialization did not appear out of thin air. Rather, it was built upon previous historical developments. While continuity theorists want to show that “races” themselves are pre-modern, what they have actually shown is that some of the means and the methods associated with racialization are not entirely modern.
ARE THE MEANINGS ATTACHED TO RACIALIZED TRAITS MODERN?
Moving on to our sixth question, also about racialization, we ask, “are the meanings attached to racialized traits modern?” Continuity theorists can not only be interpreted as asking whether the means and methods associated with racialization are modern, they can also be interpreted as asking whether the meanings attached to racialized traits are modern.
Much of the work answering question six focuses on skin color. “The study of race in the Middle Ages,” notes Cord Whitaker (Reference Whitaker2015), has “largely focused on the deployment of blackness in literature and history and the dynamics it registers, creates and reifies” (p. 8). Blackness is typically considered to be the quintessentially racial trait, so this is unsurprising. In Before Color Prejudice (1991), Frank Snowden argues that despite “the association of blackness with ill omens, demons, the devil, and sin, there is in the extant records no stereotyped image of Ethiopians as the personification of demons or the devil” (p. 107). Nevertheless, the negative associations attached to Blackness (which were inconsistently applied to dark-skinned people in antiquity) may be relevant to the development of later color prejudice and, later still, color-coded racism.
Regardless of the level of skin color prejudice in antiquity, Whitaker notes that, “the logic of race, with regard to whiteness and blackness, is already deeply ingrained in Latin Christian culture by the late Middle Ages” (2015, p. 5). Indeed, there are numerous historical sources demonstrating that as early as the thirteenth century, medieval societies were preoccupied with skin color, coding dark skin as “evil” and white skin color as “good” (Hahn Reference Hahn2001; Lampert Reference Lampert2004; Ramey Reference Ramey2014).
White supremacist ideology proper may be modern (Mills Reference Mills1998). However, it partly grew out of an earlier medieval discourse on skin color. That discourse is not well understood as racist, because until around the middle of the fifteenth century there was no concept of race. It was, rather, colorist. Whitaker takes this colorism to have a bearing on “whether race matters to the Middle Ages” (2015, p. 7). However, what this work on skin color shows is not that “race matters to the Middle Ages,” but rather that some of the meanings attached to some racialized traits are pre-modern in origin.
Consider the following example from fourteenth-century English literature. In The King of Tars (Chandler Reference Chandler2015), a sultan converts to Christianity and his skin color miraculously changes from dark to light. Whiteness, in this fable, is an outward sign of goodness. Thomas Hahn (Reference Hahn2001) considers The King of Tars and other earlier examples in medieval literature in his cautious defense of the continuity thesis. He concedes that “Black and white as terms of racial difference may have been less stable in the twelfth century than they have been in the bitterly contested discourses of the last centuries” (Hahn Reference Hahn2001, p. 24). However, what tales such as The King of Tars suggest is not that the connection between skin color and race was still unstable at the time it was written. Rather, The King of Tars suggests that there was no concept of race in this period. “Race” is meant to be heritable, biological—it cannot be altered through spiritual transformation.
The King of Tars is evidence that dark skin color was coded as evil and light skin color as good before skin color was racialized. Skin color can only be racialized if it is understood to be biologically heritable: that is a part of what it means for a trait to be racialized. Continuity theorists have been tempted into claiming that “race” itself is, in some sense, pre-modern, because the associations we have with a currently racialized trait—skin color—are pre-modern. However, those associations were not “racial” at the time.
It would be a mistake to dismiss medieval interpretations of skin color as irrelevant to the history of racialization. What The King of Tars actually shows is that the meanings associated with racialized traits (such as skin color) started forming before the concept of race emerged, and before racialization first took place. This answers our sixth and final question. No, the meanings attached to racialized traits are not exclusively modern.
Race scholars have been unable to agree about whether “race” is modern partly, I have argued, because the question itself is ambiguous. They have been answering at least six different questions. Some are interested in whether the concept of race is modern (1), others in whether there is a modern concept of race that is distinct from earlier race concepts (2). Continuity theorists are interested in whether there were actual social races before modernity (3). I have argued that social definitions of race lead to conceptual inflation and should be abandoned. The third question is therefore wrongheaded. We should ask, instead, whether racialized groups are modern (4). From this perspective, the work on the historical origins of social races is misguided, but still valuable: misguided because there are no races, only racialized groups; valuable because this literature answers the following two questions: “are the means and methods associated with racialization modern?” (5), and “are the meanings associated with racialized traits modern?” (6).
I have answered the six questions as follows: (1) no, the concept of race is not modern, it is late medieval; (2) yes, there are modern concepts of race that are distinct from the race concept used in fifteenth-century Spain; (3) no, races are not modern, because there are no races, only racialized groups; (4) no, racialized groups are not modern: the Jews and the Moors were racialized in Spain in the mid-fifteenth century; (5) no, the means and methods associated with racialization are not exclusively modern, as illustrated by the example of Christian-Jewish relations in thirteenth-century England; and (6) no, the meanings attached to racialized traits are not modern either, as illustrated by medieval discourse on skin color.
If I am right that there are racialized groups, but no actual races, then only three of the six questions that race scholars have been answering are actually about “race.” The other three are about racialization, the process or processes that produces racialized groups, not “races.” One of the three questions about race (“are ‘races’ themselves modern?”) is misguided, so that leaves only two questions about whether “race” is modern—those about race as a concept. So race scholars are not only answering different questions, half of these questions are not even directly about race, but are rather about racialization and racialized groups.
Given that race scholars have been answering not one but at least six questions, it is not surprising that there is still no consensus about whether “race” is modern. The only way in which the debate about periodization and “race” can be settled is if we disambiguate the question. My particular answers to the six questions may be controversial. The aim of this paper has not been to necessarily convince the reader that my answers are correct (though I have certainly attempted to do so). It has been, more modestly, to demonstrate that a more nuanced approach to the history of “race” and racialization can move the debate in the right direction.
I would like to thank Jean-Philippe Deranty, Andrew Dunstall, Andrew Gillett, Frances Olive, Vanita Seth, and two anonymous reviewers for their helpful comments on an earlier version of this paper. I am also grateful for feedback from audiences at the University of Sydney’s HPS Research Seminar Series, Macquarie University’s Markers of Authenticity Seminar Series, and the symposium on “The History and Philosophy of ‘Race’” held at Macquarie in 2016. This work was funded by a Discovery Early Career Researcher Award (DE190100411).