Replacing Race: Interactive Constructionism about Racialized Groups

2.1. Normative Arguments against Social Constructionism

Social constructionism is commonly interpreted as an alternative to racial naturalism. “To say that race is a social construction”, writes Blum, “is at least to say that it is not a biological kind, a ‘natural kind’” (2010: 304). It does appear that most scholars assume that social constructionism is incompatible with racial naturalism. However, many race naturalists endorse some version of social constructionism about race (Hardimon 2013; Kitcher 1999; Outlaw 1996; Relethford 2009; Spencer 2015). As Jeremy Pierce notes, “Someone could very easily hold that there are races that are natural kinds without denying that there are also social kinds called races” (Pierce 2014: 28). This is precisely the view that Michael Hardimon defends. “The concept SOCIALRACE makes it possible to represent the race-that-is-a-social-phenomenon as a social phenomenon”, he explains. “Those who deny and those who affirm the reality of biological race can use it in good intellectual conscience” (Hardimon 2013: 26). Social constructionism and racial naturalism are compatible—‘biological race’ and ‘social race’ just need to be distinguished.

The compatibility of social constructionism about race and racial naturalism does not present a metaphysical problem for social constructionists, but it does present some quite serious practical problems. Most social constructionists want their view to do something it cannot—entail a rejection of racial naturalism. The claim that there are social races does not rule out the possibility that there are also biological races. Social constructionists need to give race two definitions: one social, the other biological. To defend the claim that race isn’t biologically real one must endorse some definition of biological race, and argue that the science does not support the existence of human biological races according to that definition. As studies have shown, the debate about the reality of biological race is far from over (Lieberman et al. 2004; Morning 2011; but see Wagner et al. 2016). By referring to ‘race’ simpliciter, social constructionists talk past race naturalists.

While some, like Hardimon, carefully distinguish between ‘social race’ and ‘biological race’, almost all social constructionists talk and write about ‘race’ simpliciter. This creates an ambiguity about the meaning of the term, an ambiguity which is especially problematic when it comes to public understanding of ‘race’. There is widespread agreement that ‘ordinary people’ or non-specialists understand race as a biological category. The continued use of race as an analytic concept by social constructionists is likely to reinforce racial naturalism in the public imagination, even though it is the intention of most constructionists to distance ‘race’ from biology. As the sociologist Robert Miles explains, the use of the race concept “is particularly problematic when ‘race relations’ sociology reports its findings and conclusions to politicians, the media and the public in a historical context in which the nineteenth century idea of ‘race’ continues to be reproduced” (1988: 10). Even if current conceptions of race are more flexible and to some extent less tainted by racism than those of the nineteenth century—which is in part the result of anti-racist work done by social constructionists—race is still generally understood as a biological concept. Unless social constructionists who are also anti-realists about biological race are able to distance ‘race’ from biology in the public imagination, they will give an academic stamp of approval to racial naturalism by their continued use of what has historically been a biological concept. In other words, the language used by social constructionists may inadvertently spread racial naturalism, even though most social constructionists do not believe that race is biologically real.[3]

2.2. A Metaphysical Argument against Social Constructionism

The arguments above are normative. They are not about whether race is real. Rather, they are about whether we ought to use the concept of race, and so only support eliminativism about race. The distinction between the normative and the metaphysical is not a sharp one, because what we include in our ontology is driven, in part, by normative considerations (Kincaid, Dupré, & Wylie 2007). However, there is a distinction to be made between normative/eliminativist and metaphysical/anti-realist arguments. Eliminativists about race urge us to eliminate race from our speech, thoughts, and practices regardless of our metaphysical position on race. The anti-realist about race believes that race is not real according to certain standards, and may or may not endorse eliminativism. Anti-realists about biological race believe that race is not real according to biological standards. Anti-realists about social race believe that race is not real according to the standards of the humanities and social sciences. The standards are, of course, norms. Normativity therefore plays a role, but the conclusion is ontological, metaphysical.

Turning now to a metaphysical line of critique, I will argue that social constructionism suffers from a specificity problem. That is, on a social constructionist or a social kind approach, race loses its historical and conceptual specificity, and thus its value as a conceptual resource. It would be a mistake, I believe, to conclude that social race is simply not what we thought it was. I argue that the right conclusion is to reject a social metaphysics of race, because ‘social race’ does not meet the standards of conceptual value we should expect in the humanities and social sciences.

The specificity problem is best approached from a historical perspective. There is a broad consensus that race is a modern concept and that ‘social races’ are products of modernity (Bernasconi 2001; Doron 2012; Goldberg 2002; Holt 2002; Mills 1998; Omi & Winant 2015). However, this consensus is being called into question by scholars of the pre-modern (Hahn 2001; Heng 2011a, 2011b; Isaac 2004; McCoskey 2012; Nirenberg 2009; Ramey 2014; Thomas 2010; Whitaker 2015). What many of these pre-modernists are trying to demonstrate is that there were social groups with the properties social constructionists identify as ‘racial’ before modernity. This has implications for how ‘social race’ is defined. For instance, if modern forms of biologization do not play a necessary role in the formation of so-called social races, then such biologization cannot play a necessary role in the definition of ‘social race’. However, if we strip ‘social race’ of these modern biological associations—as I argue we must—the category becomes vague and transhistorical, losing its value as a conceptual resource.

To begin, we will need to get clear on what it means to think of race as a social kind. Ronald Sundstrom offers an account of such kindhood:

There are different, but largely overlapping accounts of the shared properties and social relations that make ‘race’ a social, or a human kind: Sundstrom (2002a) lists disparities in income, health, education, and housing; Sally Haslanger (2012: 132) emphasizes economic, political, and legal inequality; Paul Taylor (2000) and Esa Diaz-Leon (2015) focus on joint histories and shared experiences as the bonds that tie people together racially.

From a social constructionist perspective, race could thus be defined as a social kind, distinguished on the basis of real or imagined differences, which is used to differentially distribute power and privilege between groups who—as a result of their classification—have shared histories, experiences, and opportunities. Notice that I did not specify whether the differences used to distinguish between the social kinds are biological. What sort of kind ‘race’ is, and how ‘races’ are distinguished, potentially come apart. Pierce explains this point well when he writes that

This is the standard view among social constructionists. Race is social kind, not a biological kind, but it has biological inclusion criteria.

The standard view offers a way to locate race in historical time. Race is a social kind which is the product of a modern taxonomic project, a project which was used to justify the exploitation of those ‘races’ deemed inferior. This project was supposed to describe race as a biological kind, but it was in fact instrumental in creating race as a new social kind (Mallon 2015). By identifying racial formation with the modern taxonomic project, social constructionism piggy-backs on the consensus that biological race is a modern concept. However, if the social kind existed before the taxonomic project, the origins of ‘race’ are thrown into question.

One of the strongest cases for race before modernity has been made by Geraldine Heng, who argues that modern forms of biologization are not necessary for the creation of the kind of social groups that social constructionists call ‘races’. She shows that such groups existed before the modern era (Heng 2011a, 2011b). Heng’s focus is on Jewish-Christian relations in thirteenth-century England, which in many ways foreshadowed those in Weimar Germany. She argues that “the Jewish badge, expulsion order, legislative enforcements, surveillance and segregation, ritualised 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 2011b: 336; emphasis in original). Heng demonstrates, I believe decisively, that the thirteenth-century English Jewry meet the definition of a ‘social race’ before modernity.

Most social constructionists locate race only in the modern period. Those committed to this position could not respond to the above argument by simply adding biological criteria to their definitions of race. To legitimately add biological criteria as necessary to the concept, social constructions would need to demonstrate that the social kind that they call ‘race’ is only produced under particular conditions of biologization.

As it happens, almost all social constructionists already include biological criteria in their definitions of race. For example, some have stipulated that race has an ancestry component (Gracia 2005; Taylor 2004). This seems like common sense. However, Heng’s example of Christian-Jewish relations in thirteenth-century England shows this to be mistaken. Jewishness was not yet defined as ‘innate’, as it would be in fifteenth-century Spain with the Purity of Blood Statutes (Nirenberg 2009). ‘Social race’ is not necessarily determined by ancestry, because until the fifteenth-century Jews—who, if we follow Heng, fit the description of a social race centuries earlier—could change their group membership by conversion (see Stacey 1992).

Social constructionists might accept that racial formation does not necessarily involve modern forms of biologization, but insist that race is modern for some other reason. For example, they could tie race to the creation of the modern nation state (Goldberg 2002), or to the development of white supremacist ideology (Mills 1997). I do not believe that such an approach is successful. 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, “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” (Heng 2011b: 337). If this is right, then identifying race with the creation of the modern nation state appears artificial: groups fitting the description of ‘social race’ could be, and were, produced before the rise of the modern nation state.

A similar argument applies to the identification of race with white supremacist ideology. As Cord Whitaker notes, “the logic of race, with regard to whiteness and blackness, is already deeply ingrained in Latin Christian culture by the late Middle Ages” (Whitaker 2015: 5). There are numerous historical sources that show 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 2001; Lampert 2004; Ramey 2014). White supremacist ideology became more virulent, more pervasive, more global, and more thoroughly institutionalized in modernity, but it is not modern in origin, and so does not cut history into pre-modern non-racial and modern racial periods.

My point is not that all is continuity—that being a member of a so-called ‘social race’ was the same in the thirteenth century as it was in the nineteenth century. That would be preposterous. My point is rather that the sort of social relations that social constructionists identify as kind-making, as race-making, were present before modernity. Once we start adding extra criteria—‘races’ must be biologized in a certain way; they are only produced in modern nation states, etc.—we have already given up on a social kind approach in favor of an approach which seeks to cut history into racial and pre-racial times. Even if a cutting approach worked—slicing history at ‘the right time’—it would be artificial to identify ‘race’ with some uniquely modern phenomenon (e.g., the modern nation state) if the sort of social groups that social constructionists call ‘races’ were present before modernity.

Of course social constructionists could simply accept that race is not modern, and that it is not necessarily produced under conditions of modern biologization. They could define race, as I suggest above, as a social kind distinguished on the basis of real or imagined differences, which is used to differentially distribute power and privilege between groups who have shared histories, experiences, and opportunities. This definition of race has, I think, a ring of plausibility. However, on close inspection it has extremely counterintuitive consequences. It would no longer be possible to distinguish between race and ethnicity, as race has lost its biological content. Race would be conflated with racism, as hierarchy is built into the definition of race. Worse still, men and women would count as ‘races’. Men and women are differentiated on the basis of real or imagined difference, power is differentially distributed between them, and they each have shared histories, experiences, and opportunities. Even children and adults would count as different ‘races’ on the definition above. While we may be willing to accept that race is a vague concept, most would be unwilling to accept this variety of vagueness.

Heng is perhaps unique among social constructionists as she sees and accepts some of these kinds of counterintuitive consequences. She suggests, for instance, that in certain circumstances homosexuals can be conceptualized as a race (Heng 2011a: 319). I believe that this ought to count as a reductio ad absurdum against social constructionism about race. If race cannot be distinguished from other group concepts, such as ethnicity or sex, or from forms of prejudice, such as racism or homophobia, then the idea loses its conceptual value. We can just refer to structural discrimination or persecution instead. There has been structural discrimination between groups for all of recorded history. On a social definition ‘race’ does not only lose its conceptual specificity, it also loses its historical specificity—it becomes transhistorical.[4]

The conclusion that we should draw from this is not merely eliminativist: that social race is real but we ought not to talk about it because of the specificity problem. Rather, we should draw an anti-realist conclusion: that social race is not real. The specificity problem shows that ‘social race’ has little conceptual value, and so attempting to construct a social metaphysics of race is a wrongheaded project.

The specificity problem does not arise on a biological definition of race. On the standard biological account, here expressed by Claude-Olivier Doron, two conditions needed to be met before race theory proper could emerge:

It was the joining of what Doron calls the classificatory and the genealogical style of reasoning in the eighteenth century (by, for example, Buffon and Kant) in attempts to divide the entire human species into major biological sub-groups that makes race a modern concept.[5] To claim that the idea of race is modern is to claim that it only emerges as a relatively stable and coherent scientific concept during the period we call modernity (Bernasconi 2001; Doron 2012). This is not to downplay how the idea of race was mobilized politically and socially, nor to dismiss pre-modern conceptions of difference as irrelevant. Identifying race as a biological concept will only sharpen our theoretical tools. It will allow us to trace proto-racial theory—which was missing one or more of the key elements of ‘race’—back to antiquity (Isaac 2004; McCoskey 2012).[6] It will also allow us to trace pseudo-racial theory—which conflates ‘race’ with other concepts, such as ‘nation’ and ‘population’—to the present day (Gannett 2013; Hudson 1996; Millstein 2015). Defining race biologically is good historical practice.

If we ought to define race biologically, but there are no biological races, then we must conclude that race doesn’t exist. I now turn to anti-realism about race.

3.1. Objections and Clarifications

Some have worried that an ontology of ‘racialized groups’ would entail that certain identities are not genuine. “Blum’s substitutionism”, writes Joshua Glasgow, “would require that those who care about being, say, black have to stop thinking of themselves as members of a race … and instead conceive of themselves as members of erroneously racialized groups, so that their identities are, in some sense, fraudulent” (Glasgow 2009: 150). In response, Blum argues that,

Identifying with a racialized group (or indeed with more than one) does not entail buying into a lie. It does, however, raise the following puzzle. If the groups we call ‘races’ are actually racialized groups, could it be that ‘race’ and ‘racialized group’ can be used synonymously, with the consequence that race is real after all? Indeed, the term ‘racialization’ is typically understood by social constructionists to refer to the race-making process, the process by which social races are formed (e.g., Haslanger 2012: 237).

One way of cashing this out theoretically would be to adopt a causal-historical approach to reference. If the meaning of our terms “just ain’t in the head”, as Hilary Putnam argued, then perhaps we could be radically mistaken about what race is and still achieve reference when we talk about it (Putnam 1973). That is, when we talk about race—even when we are thinking of races as scientifically valid biological groups—we might really be talking about racialized groups. This would indeed be a consequence of Putnam’s causal-historical theory of reference. If Putnam was right, maybe we can re-conceptualize ‘races’ as ‘racialized groups’ but keep on using the term ‘race’, despite my arguments against the category.

The strategy of trying to resolve metaphysical disputes by appealing to a putatively determinative theory of reference—the ‘flight to reference’—has, I take it, been discredited (Bishop & Stich 1998; Mallon, Machery, Nichols, & Stich 2009; Stich 1996). There is no single way in which terms refer. For some terms a causal-historical approach works well (e.g., star, atom). For other terms a descriptive approach appears to be correct (e.g., phlogiston, natural place). Which theory of reference ‘works’ for any individual example is, according to Stich, determined by principles of rational ontological inference like those described by Larry Laudan’s project of normative naturalism (Donovan, Laudan, & Laudan 1988) and more or less arbitrary factors: implicit previous agreements about the properties necessary to a posited entity, and social and political factors both internal and external to the relevant field.

What this means for the metaphysics of race is that we cannot settle debates about racial ontology by appealing to a theory of reference (Mallon 2006). A theory of reference may well describe how the term ‘race’ refers or fails to refer if a consensus is someday reached. However, if the flight to reference is a mistake, then we should not appeal to a theory of reference in order to forge such a consensus. The sort of considerations I have been discussing are overriding, and do not hold the “correct account of race hostage to issues in the philosophy of language and metaphysics about which there is little agreement” (Mallon 2006: 548).

If we were to accept a causal-historical approach to reference as determinative, and identify racialized groups as races, this would blur a very useful distinction. As Blum (2010) argues, racialization should be understood as the process by which racialized groups—rather than ‘races’—are formed. This approach gives us a way of conceptualizing the groups commonly called ‘races’ without reifying them. And it does not suffer from the central problem with unreconstructed anti-realism: that conventional ‘racial’ categories would have to be abandoned, along with the possibility of targeting these groups in public policy. The anti-realist racialization theorist is able to recognize the reality of the groupings, but reconstruct our understanding of them. And at the same time, she can save the intuition that race is a biological concept. By defining race biologically, she can be very clear about what it is she is claiming does not exist.

The concept of the racialized group is more disruptive than the idea of social race, and this seems like an advantage. As J.L.A. Garcia explains, while a ‘race’ is something one has, “racialization is something that is done to a group, by some social agent, at a certain time, for a given period, in and through various processes, and relative to a particular social context” (Garcia 2003: 285). This suggests that the language of racialization is less likely to perpetuate false beliefs about the naturalness of racial classification than standard race-talk. Indeed, Blum (2004) has discussed the benefits of teaching the idea of racialization as a part of anti-racist education.

It may appear that I am favoring a descriptive over a causal-historical approach to reference, and thus taking the flight to reference myself—the very move I am arguing against. That would be a misunderstanding of my position, and of the flight to reference. The critique of the flight to reference is only against the idea that there is a single true, and thus determinative, theory of reference. In some specific cases, something like a descriptive approach will be preferable, but that is not because a descriptive approach offers the correct account of reference relations. In other cases we should favor a causal-historical approach. For instance, a causal-historical approach may be preferable when changing the meaning of a term will not cause too much confusion. That is manifestly not the case with ‘race’. Given the normative and metaphysical arguments against ‘race’, and those in favor of replacing ‘race’ with ‘racialization’ and ‘racialized group’, appealing to the causal-historical approach as a determinative theory of reference seems like a last-ditch attempt to save the category of race, and one which is unlikely to convince those who are not already committed to this position in the philosophy of language.

Before introducing interactive constructionism as a framework for understanding racialization, I will consider another possible objection: that my version of anti-realist reconstructionism also suffers from a specificity problem. I have already argued, following Robert Bernasconi and Doron, that ‘race’ emerges as a relatively stable and coherent scientific concept in the second half of the eighteenth century. However, the term ‘race’ and its cognates had already been used for hundreds of years. So even if we can, with some precision, locate race as a scientific concept historically, this leaves open the following question: when did racialization first occur?

We can answer this question by identifying racialization as taking place when the term ‘race’ is applied to human 'breeding groups', or even when groups are understood in a certain way without being explicitly labelled as races. The earliest example of which I am aware of the term ‘race’ being used to describe human breeding groups was in fifteenth-century Spain. As David Nirenberg 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” (Nirenberg 2009: 252). If we think of racialization as taking place when a human group is called a ‘race’, then we can date racialization back to the fifteenth century.

While such an approach is appealing, it does not allow for what we might call ‘covert racialization’: racialization that takes place without the explicit naming of a group as a ‘race’. For instance, when media outlets report attacks on Muslims as racist, this may have the effect of racializing Muslim people, even when they are not explicitly called a race. Perhaps, then, racialization can be understood to take place when groups are understood in a certain (biological) way. This would not necessarily involve fully-fledged race theory. For example, the race/species distinction as we know it—a distinction considered central to the development of a scientific concept of race—had not yet been drawn in fifteenth-century Spain. Nevertheless, the labelling of the Jews (and also the Moors) as a ‘raza’, and naturalizing their reproduction as ‘in the blood’, can be usefully understood as an instance of racialization (Nirenberg 2009).

To see this, consider a key difference for Jews in thirteenth-century England and fifteenth-century Spain. Jews in thirteenth-century England were not racialized: they were a religious group who were subject to structural persecution. Jewishness was, to some extent, biologized, in that it was associated with bodily characteristics, but Jews could convert, and many if not most of those characteristics were understood to disappear through conversion (Resnick 2000). In fifteenth-century Spain, on the other hand, Jewishness was deemed to be ‘in the blood’ and conversions were nullified. We can mark this difference as one of racialization.

Whether we take the explicit naming of a group as a race as determinative of racialization, or a certain biologized understanding of the group as also important, racialization appears to have first taken place in fifteenth-century Spain. The anti-realist racialization approach does not, therefore, suffer from the lack of conceptual and historical specificity that we find with a social kind approach to ‘race’. Nevertheless, there will likely be trouble cases—particularly because racialization comes in degrees—and more work on this subject would be welcome.

4.1. The Framework

The reason we need a theory of racialization is straightforward. Social constructionists have racial formation theory to account for the way ‘races’ are created, inhabited, transformed, and destroyed. Anti-realists can’t endorse racial formation theory because it assumes that race is real. Anti-realists need a theory of racialization, a theory which describes the formation of racialized groups, rather than ‘races’.

I take, as a useful metaphor for racialization, the idea of constructive interactionism from DST. Recall that DST is a biological framework for understanding development, inheritance, and evolution. It proposes that every trait is the result of interactions between a wide range of developmental resources. Developmental systems theorists make the case that development occurs in a niche in which the resources necessary for ontogenesis are “contingently but more or less reliably reassembled for each life cycle” (Oyama et al. 2001: 1). It is the interaction between the many elements of the system that is responsible for development.

DST rejects metaphysical dichotomies such as gene/environment, nature/nurture and biology/culture in favor of a concern with developmental systems as a whole. Developmental systems theorists argue that these dichotomies have outlived their usefulness (Griffiths & Hochman 2015). DST is sometimes misinterpreted as the idea that all developmental factors (genetic, epigenetic, ‘environmental’, etc.) are the same or just as important as each other. The idea is rather that no interactant has a metaphysical privilege. Genes, for instance, do not exclusively control development, and they are not the only developmental resource that is inherited (Griffiths & Hochman 2015).

Drawing on DST, interactive constructionism is the view that racialization is a process involving a wide range of interactants, which are jointly involved in the production of racialized groups. On the interactive constructionist approach, racialized groups emerge out of the ongoing interaction between a number of factors: administrative, biological, cultural, economic, geographic, gendered,[10] historical, lingual, phenomenological, political, psychological, religious, social, and so on. Interactive constructionism rejects the metaphysical privileging of any of these interactants as the key to what racialized groups ‘really are’. The importance of any given interactant will depend on the context, and can change over time.[11] None of these interactants are ‘racial’. Yet together, in interaction, they can produce racialization. And just as groups can undergo a process of racialization, they can be deracialized over time. Racialization is not an all or nothing affair. As Blum (2002) points out, groups can be partly racialized.

Interactive constructionism offers a way of conceptualizing human biological diversity without reinstating race as a valid biological category. How a person is racialized depends, in part, on the interpretation (by a particular group and at a particular time and place) of biological traits such as skin color, hair type, eye shape, and so on. Contra racial naturalism, biological traits alone cannot determine racialization because human biological diversity is comprised of characteristics that are predominantly smooth in their distribution across geographic space and have different distribution patterns depending on which trait is selected (Barsh 2003; Zack 2002). In technical terms, we would say that human biological diversity is composed primarily of discordant clines. So interactive constructionism includes biology as a factor in racialization, but human biological diversity is not understood as racial.

On a systems approach to racialization we should expect there to be interactions between (non-racial) biological difference and other factors in the racialization process. We should also expect biological difference to be more or less important at different points in history, and in different places. This is exactly what we find. Kant’s racial taxonomy was color-coded (Kant 1777/2000). However, once we move from the eighteenth to the nineteenth century, ‘race’ was conflated with ‘nation’. As Nicholas Hudson writes, “Pseudoscientific studies of ‘race’ … focused on racial differences between ever smaller groups. British, Irish, Saxons, Aryans, Jews, even ‘Americans’ (that is, citizens of the United States) were all, at one time or another, vaunted or condemned as ‘races’” (Hudson 1996: 258). It would be very difficult to classify people into these groups by their visible biological features. For a period, then, visible biological difference played a less important role in racialization. However, as more light-skinned groups were elicited into the ‘white race’ in the early to mid-20^th century, visible biological difference became—once again—central to race thinking (Carter 2007). The importance of visible biological differences in the racialization process waxes and wanes in interaction with other factors: social, political, economic, and so on.[12]

This sort of interaction can also be observed on the individual level. Anyone who has ‘passed’ knows something about how this works. How you dress, how you talk, how you act, and what you know, all play a role in whether you are able to pass as being from a racialized group other than the one you are usually associated with. For many of us ‘passing’ is almost impossible because we have biological traits that are strongly associated with a particular racialized group. This is not because race is biologically real, but rather because people occupy different locations on the various clines of human biological difference (some of which are racialized, such as skin color). Whether you can ‘pass’ depends not simply on your visible biological make-up, but on a complex interaction between this factor and others, such as how you dress, talk and act. This is one more reason why we should want a systems approach to racialization.

Interactive constructionism makes breadth central. Most feel quite comfortable including historical analysis, political analysis, and cultural analysis under the rubric of ‘the social’, but less so biological analysis, or even, for that matter, psychological analysis.[13] Individual mental processes uphold social and political forces, so it is wrong to think that we have a full explanation once we deal with the social, historical, political and biological factors behind racialization. As Michael Rustin puts it, “a crucial means by which structures are upheld is through irrational mental processes, and … this dimension needs to be recognised and confronted as such” (Rustin 1991: 71). We therefore need to take seriously the role that psychology plays in racialization. “Without some sense of the emotional intensity of race’s appeal”, argued Patrick Wolfe, “we cannot begin to account for it, let alone do anything about it” (2002: 53). And without the conceptual resources of psychology we cannot hope to account for this appeal.

At the risk of sounding imperialistic, most work done on the topic of ‘race’ could be absorbed into an interactive constructionist framework. As Ron Mallon has argued, there is now widespread agreement about the facts relevant to the debate about the nature and reality of race. This makes the idea of an overarching framework plausible:

If there is a broad consensus about the facts, then the factual claims made by anti-realists, social constructionists, and race naturalists should be able to fit together in one overarching framework. By including both the biological and the social in the metaphysics of racialization, interactive constructionism provides such a framework.

4.2. Objections and Clarifications

I will now consider three possible objections to interactive constructionism about racialized groups, with the aim of clarifying the view. First, one might argue that race naturalists and social constructionists are already interactionists, and that interactive constructionism does not bring anything new. For instance, some race naturalists accept that social forces play a role in shaping biological ‘races’ (Kitcher 1999; Outlaw 1996; Spencer 2015), and perhaps all social constructionists believe that biological diversity is a part of the raw materials out of which ‘social races’ are formed.

However, these are weak forms of interactionism, according to which many factors influence race, but race is essentially biological or social. Race naturalists define race as a biological kind, giving a metaphysical privilege to the biological. They may claim that social factors influence our biology, but if race is a biological kind, then its reality (or lack thereof) is ultimately to be determined on biological grounds. What race is, and whether it is real, is still an entirely biological issue. In this way, race naturalists are committed to privileging biological factors, and their interactionism can only be of a weak variety.

Similarly, social constructionism confers a metaphysical privilege to the social, and so can only involve a weak form of interactionism. If race is a social kind, its reality is ultimately social. Biology can only play a peripheral role for the social constructionist, just as social forces can only play a peripheral role for the race naturalist. Just as the race naturalist must accept that biological races could be produced without social forces playing a role, social constructionists must accept that there could be social races that are not created under modern conditions of biologization, a consequence which results in the specificity problem described above.

One social constructionist does seem to endorse a strong form of interactionism, and this presents a puzzle. In A Realist Metaphysics of Race, Pierce argues that “the biological component of the social construction matters. Social constructions based on existing features of the world, however arbitrarily put together, do not thereby generate groups of things classified according to the construction. They just give them social significance” (2014: 89). Pierce calls his view a social kind approach, but I am not sure that this best describes his position. Just as chemical kinds are composed of chemical properties and relations, social kinds are composed of social properties and relations. Pierce does not argue that race is a social kind, but rather that social practices pick out pre-existing but biologically arbitrary groups and make them socially significant. His might be called a social significance view, rather than a social kind view. Because Pierce does not metaphysically privilege the social, he can avoid some of my criticisms of social constructionism. However, if the distinction between racial naturalism, social constructionism, and anti-realism is a metaphysical one, his view no longer fits neatly into any of these categories. Pierce’s view may actually be closest to interactive constructionism about racialized groups, and our main disagreement may turn on issues in philosophy of language surrounding reference.

Let us consider a second line of objection. Social constructionists may insist that we privilege social factors by considering them as constitutive of group membership. If we do not offer a metaphysical privilege to either the biological or the social, as interactive constructionism suggests, how can we determine group membership in cases where social and biological factors point in different directions? Consider, for instance, African Americans with high proportions of recent European ancestry who grow up and live in typical social positions as African Americans, and who are identified by themselves and others as such. Race naturalists may identify these individuals as white. Social constructionists may argue that this is a mistake, that we should privilege social factors in determining their ‘race’. What does the interactive constructionist say in such a case? Does the interactive constructionist approach entail that we need to weigh the biological and social factors equally? This seems to leave group membership underspecified.

This criticism misses the mark. Just as DST does not claim that all developmental factors are just as important as each other, but rather that no single factor has a metaphysical privilege, the interactive constructionist rejects any a priori privilege of the biological or the social. Interactive constructionism does not entail that biological, social, and other factors make equal causal contributions to the formation of racialized groups, whatever that may mean. On an interactive constructionist approach, there are no universal criteria for group membership. To return to the example above, if a person has a high proportion of recent European ancestry, but also has recent African ancestors and is racialized as black, it does not make sense on an interactive constructionist account to ask, “but what is she, really?” Interactive constructionism does not prescribe how we ought to classify people. Rather, it offers descriptive framework for understanding how racialization works.

Let us consider a third and final objection. While interactive constructionism rejects racial naturalism, it still includes a role for biology in racialization. Some may worry that, by including biology in the picture, we risk opening the door to hereditarian explanations of differences between racialized groups. This worry would apply much more directly to the race naturalist, who believes that there are biological races. The almost universal response of race naturalists (or at least those in philosophy departments) has been to sanitize ‘race’, removing from the concept the racist associations of the past, and to cite works demonstrating the poor scientific standing of hereditarianism (e.g., Gould 1996; Kaplan 2015; Nisbett 2005; Nisbett et al. 2012; Suzuki & Aronson 2005). The interactive constructionist rejects racial ontology, and so has an easier time with the objection. There cannot be inborn psychological or moral differences between ‘the races’ if there are no races. Any hereditarian reading of interactive constructionism would be a misinterpretation of the view. Racialized groups do not form valid biological categories, about which one could logically ask these sorts of questions.

While it may seem that the social constructionist who is also an anti-realist about biological race avoids these sorts of misinterpretations, this is not the case. Social constructionism is often misinterpreted by hereditarians as the view that there are absolutely no statistical differences in any biological traits (even skin color) between conventional ‘racial’ groups (e.g., Sesardic 2013). Nobody in the debate actually seems to hold this view (Hochman 2016b). Nevertheless, the social constructionist straw-man is derided as absurd, and it is against this background of apparent absurdity that not only racial naturalism, but also hereditarianism, is advanced.

It is probably impossible to insulate any view on this topic from misinterpretation. Interactive constructionism is unlikely to be misinterpreted in the way social constructionism is misinterpreted. However, by including biology and psychology as interactants in the racialization process, it may risk being misunderstood along hereditarian lines. The point, however, is not that there are biological races with innate psychological differences, but rather that there are racialized groups with vague and arbitrary biological inclusion criteria, and that there are psychological explanations for the practice of racialization itself (Clarke 2003; Dalal 2002; Hochman 2013b; Machery & Faucher 2005; Smith 2011). Interactive constructionism offers no support or intellectual basis for hereditarianism.