Jump to ContentJump to Main Navigation
What Is Race?Four Philosophical Views$

Joshua Glasgow, Sally Haslanger, Chike Jeffers, and Quayshawn Spencer

Print publication date: 2019

Print ISBN-13: 9780190610173

Published to Oxford Scholarship Online: June 2019

DOI: 10.1093/oso/9780190610173.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 20 October 2021

Tracing the Sociopolitical Reality of Race

Tracing the Sociopolitical Reality of Race

(p.4) 1 Tracing the Sociopolitical Reality of Race
(p.iii) What Is Race?

Sally Haslanger

Oxford University Press

Abstract and Keywords

The concept of race has a troublesome history. It has been used to divide societies and subordinate groups in unjust ways. It has also been a source of pride and strength for the subordinate (as well as, unfortunately, for the dominant). Historically it has also carried assumptions of naturalness: races are natural kinds that exist independent of human thought and activity. In recent years, however, the naturalness of race has been challenged and replaced with the idea that race is socially constructed. This raises many important philosophical questions: How should one inquire into the concept of race when there is such broad controversy over what race is? What are the relevant phenomena to be considered? How should such an inquiry take into account the social stakes (e.g. the potential impact of maintaining or rejecting the concept of race)? Is it possible for concepts to evolve, or is conceptual replacement the only option? In Chapter 1, the author takes up these methodological questions and positions herself as a critical theorist considering what role the concept of race has in the sociopolitical domain. She argues that there is a meaningful political conception of race that is important in order to address the history of racial injustice. This is compatible with there being different conceptions of race that are valuable in other contexts and for different purposes (e.g., for medical research, cultural empowerment).

Keywords:   race, racism, social construction, critical theory, conceptual amelioration, conceptual engineering, conceptual pragmatism, racial justice

1.1. Methodological Preliminaries

The question before us is: What is race? When we pose questions of the form “What is X?” there are a variety of ways we might go about answering them. For example, if, pointing to a small wiggly thing in the corner, I ask, “What is that?” I will probably want someone to help me figure out the species of insect it belongs to, as determined by entomology. If you tell me it is a silverfish, I might also pose a question about the kind of thing (e.g., “What is a silverfish?”). Plausibly I am asking how silverfish, as a group, are classified: what features something must have to count as a silverfish, what to expect of silverfish, and how they are related to other sorts of creepy crawly things.

These sorts of questions seem to presuppose that we have a well-developed science that will provide us with empirically based answers. However, sometimes our “What is it?” questions take us beyond what science has figured out. For example, if in the seventeenth century someone pointed at a burning log and asked, “What is that?” a straightforward answer would be “Fire.” But if the speaker already knew that and proceeded, “But what is fire?” the question is probably attempting to probe features of fire that aren’t apparent from our ordinary familiarity with it; and it would (and did) take substantial empirical research and future scientific theory to reach any answers.

In the sorts of contexts just considered, it would be, at the very least, odd to answer the questions by consulting our linguistic intuitions.1 Our judgments about when to use the term ‘silverfish’ (p.5) don’t tell us what a silverfish is. However, there are a variety of “What is X?” questions that many philosophers seem to think can be answered by discovering the meaning of the term(s) substituted for X, as determined by our disposition to apply the term(s) in question (e.g., ‘knowledge,’ ‘moral worth,’ ‘justice,’ ‘a person,’ ‘causation’). In some of these cases, one might think that this a priori methodology is warranted because the boundaries of these kinds depend in some way on us and our practices. Perhaps moral worth, justice, personhood, and the like, don’t exist independently of our judgments of what counts as moral worth, justice, and personhood. So, of course, we should at least begin by investigating our judgments and putting them in order. (This is more plausible in some cases than in others; e.g., the answer would have to be more complicated in cases such as ‘causation’ or ‘intrinsic property.’)

But the idea that (some) philosophical kinds somehow “depend on us” is not entirely clear; nor is it clear why our a priori (linguistic) reflections should be sufficient to provide an adequate theory of them—for example, “What is a sheriff?” Even if you are a competent user of the term ‘sheriff,’ you may not be able to tell me what a sheriff is. A full answer would presumably require information about the jurisdiction of sheriffs, what their responsibilities are, how they are chosen, etc. as determined by law. We might need to consult experts in civics to get answers (and the answers will depend on what country we are in). We can’t just depend on common sense or linguistic intuitions. But surely what counts as a sheriff depends on us—there are no sheriffs outside of a humanly constructed system of government.

In the case of ‘sheriff,’ there will be a well-defined role specified by statute, and someone who knows the relevant statutes will know the answers to our questions. But there are also social phenomena that in some sense “depend on us” but are not stipulated or planned by us. Such social phenomena range from macro-scale economic depressions, globalization, urbanization, and gentrification, to more local social practices and relations (e.g., within a town, religious congregation, or family). These phenomena call for explanation, and the social sciences (broadly construed) endeavor to provide theories that enable us to understand them, usually identifying kinds of institutions, economic relations, cultural traditions, social meanings, and psychological predispositions, to do so. The kinds in question are social kinds, in the sense that they are kinds of things that exist in the social world (and so, in some sense, depend on us). But we discover these kinds through empirical inquiry, just as we discover chemical kinds through empirical inquiry.

For example, accounts of gentrification often make reference to the “urban pioneer,” sometimes characterized as artists and “bohemians” who take (p.6) advantage of low rents in poor neighborhoods. When single people who share rent enter a neighborhood, businesses (such as cafés and pubs) take interest, and landlords see opportunities to raise rents, which drives out the locals. Urban pioneers are a functional kind that identifies a particular role in an evolving real estate market. The term ‘pioneer’ is chosen due to the perceived parallel with pioneers who “settled” the western United States, displacing the local population. If someone were to object to the term ‘pioneer’—perhaps thinking that it carried an overly positive connotation—this would not undermine the explanatory claims.2 The adequacy of explaining gentrification by reference to singles moving into an urban neighborhood does not depend on our linguistic intuitions about applying the term ‘pioneer’ to them. The choice of terminology was intended to illuminate a parallel; if the terminological choice doesn’t work, then another term could be used as a substitute.

However, insofar as philosophical kinds such as justice and personhood “depend on us,” it is not in the sense that we stipulate what they are (like sheriff), or in the sense that they serve in explanations of social phenomena (like urban pioneers). Rather, it is something along these lines: the adequacy of our theory is not to be judged simply by reference to “the facts,” but also by its responsiveness to our prior understandings. In the case of sheriff, you might think that there aren’t any independent facts we’re trying to accommodate. Oversimplifying, we simply create sheriffs and then talk about them. In the case of urban pioneer, the prior understandings of ‘pioneer’ are not crucial to the explanation provided by the theory. But in the case of justice, there is something we are aiming to understand that is not simply constituted by what we say, but at the same time, our conclusions cannot float completely free of the discursive tradition in which we are aiming to understand it.

How might we explain this? Note that in the philosophical cases, we are not situated as anthropologists trying to understand the social life of the “natives.” Nor are we legislators specifying new practices. We are seeking an understanding of practices in which we are currently engaged as participants. The practices are not fully understood, however. And they are open-ended, revisable, possibly self-defeating. In making sense of them, we are making judgments about how to better understand what we are doing, and how then to go on. This is not primarily a linguistic exercise: we aren’t just deciding how (p.7) to use existing terminology, but how to collectively orient ourselves toward the world and toward each other. Language provides tools to achieve this. But language is a practice within practices and is itself a proper target of philosophical inquiry: meanings are not simply constituted by what we believe, yet we are situated within a tradition of linguistic practices that have already shaped our meanings and our world; so ignoring those practices would be a mistake. We are situated inquirers, and the question is how we should go on, given where we have been, where we are now, and where we are trying to go (Lear 1986).

1.2. What Is the Question?

The question arises, then, what sort of question is at issue when we ask, “What is race?” Is it an empirical question that we should answer using the methods of biology? Or should we use the methods of empirical sociology or history? Is it a question about what ‘race’ means? And how might one determine the meaning of ‘race’? Do we get to stipulate the meaning? Are we seeking a philosophical tool for explanatory purposes? Or is the question best understood as arising for us as participants in racializing practices?

I don’t think there is one right way to pose the question, “What is race?” In fact, I think it is useful to ask different versions of the question in order to understand the phenomenon, and different forms of the question, raised in different contexts, will call for different answers.3 In my own work (e.g., Haslanger 2012, Ch. 7), I have explored the question as a critical theorist. There are multiple ways of characterizing critical theory, but for our purposes here, I will draw on Tommie Shelby’s characterization of a social critic:

There is also the discourse of the social critic, which is identical with neither everyday discourse nor scientific discourse. Social critics don’t merely systematize common sense or popularize scientific findings. Social critics seek to inform, and possibly shape, public opinion with clear and careful thinking, well-established facts, and moral insight. They will of course draw on and engage both common sense and (p.8) scientific thought, but they do so without taking a slavish attitude toward either . . . [In the context of debates over race and racism,] the principal role of the philosophical social critic, as here conceived, is to shed light on the most fundamental conceptual and normative issues that race-related questions raise.

(Shelby 2014, 63)

Plausibly, all inquiry is situated. Inquiry begins with questions, and all questions have presuppositions. And any serious effort to answer a question relies on a method that is taken to have at least some epistemic credentials. I’ve been suggesting that certain forms of philosophical inquiry are situated in an additional sense; that is, the project is not simply a descriptive or explanatory project, but aims to shape or guide our thinking and acting. Social critics take this even a step further: we are situated as critics of ordinary social practices and offer tools and understandings that are designed to improve them (Fraser 1989; Marx 1843). The social critic embraces the normative dimension of philosophical theorizing, and also relies crucially on empirical research. The idea of race is already embedded in our customs, practices, and institutions, and facts about its role in our lives are crucial to the critical project. Such empirical information and normative concerns are also important, on some accounts, for adjudicating linguistic meaning, and so, in particular, for understanding what ‘race’ means.

1.3. The Semantic Strategy

Quine (1953) has taught us that if we are engaged in an ontological debate about the existence of some kind of thing, say, races, we should semantically ascend. In other words, instead of asking directly whether races exist, we should ask whether the term ‘race’ picks anything out in the world, and if so, what. This is an especially helpful move if parties to the debate don’t agree on what the term ‘race’ means, for if, say, a racial realist and a racial anti-realist have different understandings about what ‘race’ means, then the conflict between them may be only apparent.4 It may be true, for example, both that biological races don’t exist and social races do exist.

(p.9) Ron Mallon (2006, 527) considers this “semantic strategy,” widespread in the race debate, and argues, however, that it is not helpful. The problem is that different parties to the debate seem to adhere to different theories of meaning. So the controversy is just pushed up a level. For example, the realist and anti-realist, it would seem, don’t agree on how we might determine the meaning of the term ‘race.’ When the realist claims that ‘race’ refers to a social kind, and the anti-realist says that ‘race’ does not refer to anything, they haven’t established a basis for debate because they are committed to different theories of meaning or reference. As a result, there is a risk that the race debate just collapses into a debate in the philosophy of language. Given the unlikelihood that we will be able to settle on a theory of reference any time soon, it looks like the race debate is left hanging.

However, Mallon suggests that there are important questions that should be asked and whose answers shouldn’t depend on a metaphysical or semantic theory. His proposal is that we take up a normative approach to race. The important question isn’t the metaphysical one (i.e., whether races exist or not) or the semantic one (i.e., what ‘race’ means, if anything). Rather, the question is normative: how we should think and talk when it comes to matters of race.

While there is (or should be) a wide basis of metaphysical agreement on the expanded ontological consensus, there is profound disagreement over the practical and moral import of ‘race’ talk. Resolving this disagreement requires a complex assessment of many factors, including, the epistemic value of ‘race’ talk in various domains, the benefits and costs of racial identification and of the social enforcement of such identification, the value of racialized identities and communities fostered by ‘race’ talk, the role of ‘race’ talk in promoting or undermining racism, the benefits or costs of ‘race’ talk in a process of rectification for past injustice, the cognitive or aesthetic value of ‘race’ talk, and the degree of entrenchment of ‘race’ talk in everyday discourse. The point is that it is on the basis of these and similar considerations that the issue of what to do with ‘race’ talk will be decided, not putative metaphysical or actual semantic disagreements.

(Mallon 2006, 550)

I am sympathetic with Mallon’s (2006) suggestion that resolving the issue depends on normative and empirical considerations.5 However, as he frames (p.10) it, the questions we must answer still concern “race talk.” But recall that, by hypothesis, the different parties to the race debate disagree about what ‘race’ means because they embrace different accounts of meaning. What counts, then, as “race talk?” It can’t be identified by talk about race, for we don’t agree on what race is, or even whether there is such a thing as a race. And we can’t just consider talk that includes linguistic items pronounced as English speakers pronounce ‘race,’ for we use that sound also for boat races, running races, and the like. Although Mallon is right that we need to ask a wide range of epistemic and moral questions of the sort he lists, his characterization of the task retains too much of the semantic strategy. What’s at issue isn’t just our talk and thought, but racial structures and practices of all sorts—linguistic, cultural, medical, political, juridical. We begin our theorizing already situated in these practices. What we are trying to do is understand how they work, what is salvageable (if anything), and how to go on.

Consider a comparison with the notion of moral worth. We begin with our practices of distinguishing the worth of an action from its consequences. We seem to be prepared, at least sometimes, to commend an action as good, even if it has unfortunate consequences, and to condemn an action as bad even if it has good consequences; yet we don’t have a clear idea of what moral worth is. For example, if I bring my new neighbor a bouquet of flowers, not knowing that she has severe allergies, and she suffers as a result, my action was nevertheless kind and thoughtful, and seems to have moral worth, even though it had bad consequences. When we ask, “What is moral worth?” we consider a full range of cases, the presuppositions and effects of this practice of attributing moral worth, and what function it has. The point is not to look at “moral-worth-talk,” since the language of ‘moral worth’ is rather rare in common parlance. We are attempting to capture a set of practices of moral evaluation. After careful scrutiny, we may find that the feature that seems to distinguish worthy actions isn’t as valuable as we thought, or the worthy feature is more rarely present; and this justifies a revision to the practice. If we are consequentialists, we may find that the practice isn’t justified at all and we may recommend discontinuation.

As mentioned before, ideas of race are “woven into” many of our everyday practices (i.e., racial distinctions seem to play a role in so much of what we do, where we go, with whom we associate, in what resources are available to us and what is required to access them). This is not to say that race is explicitly and intentionally functioning in these practices. But our lives are shaped by a racial geography. As in the example of moral worth, we begin by collecting a full range of apparent examples, consider their presuppositions and effects, (p.11) and consider what function they have. What is it, if anything, about these practices that makes them “racial”? In the contemporary “post-racial” climate, some will no doubt argue that there is nothing specifically “racial” about them (e.g., they are to be understood in terms of class). But there is also plenty of evidence that racial distinctions, racial assumptions, and racial identities continue to structure our lives together (and apart).

1.4. Representational Traditions: ‘Water’ as an Example

Laura and François Schroeter (2015) offer an account of meaning that not only seems to be compatible with the spirit of Mallon’s suggestion, but also situates our linguistic activities within our broader social practices. They focus on the example of ‘water,’ and suggest that to determine what ‘water’ means, we should undertake an inquiry into what water is. But how do we do this? We cannot assume from the start that this is a task for the chemist, for when the chemist says that water is H2O, she may be using the term in a technical sense, in which case it would not provide an account of what the ordinary person means by ‘water.’ (Note that the same might be said of the biologist’s use of ‘race.’) But neither can we just undertake reflection on linguistic usage or common sense. Before you explicitly reflect on the question of what water is, your own assumptions about the topic are bound to be heterogeneous, incomplete, and partially contradictory—and this heterogeneity is only exacerbated when you take your whole community’s views into account. Thus justifying an answer to a ‘what is x?’ question is nothing like slotting some missing values into an implicitly grasped formula. Your goal in rational deliberation is to find some principled way of prioritizing and systematizing your own and your community’s commitments about water, so as to identify the appropriate normative standards for evaluating the truth and acceptability of beliefs about the topic. (2015, 430)

The broad idea is this: when we deliberate about what X [water, race, freewill, moral worth . . .] is, we have to start with something. In the sorts of cases we are considering, we can take ourselves to be situated within a broad representational and practical tradition concerned with X. We are not starting from scratch and stipulating the meaning of theoretical terms. And we may assume that the tradition has a certain epistemic ambition, so we may “take (p.12) our words and thoughts to represent genuinely interesting and important features of the world—not just whatever happens to satisfy our current criteria” (Schroeter and Schroeter 2015, 436). So scientific inquiry, although not definitive, is relevant, since it discloses some parts of the world that are important for many of our purposes. But where do we begin? The Schroeters (2015, 426) give a sample of inputs to deliberation in the case of water (the examples are theirs):

  • Particular instances: there’s water in this bottle, in Port Phillip Bay, Lake Michigan, etc.;

  • Perceptual gestalts: the characteristic look, taste, odor, tactile resistance, and heaviness of water;

  • Physical roles: water’s rough boiling point, its transformation into steam, its role as a solvent, the fact that it expands when it freezes, etc.;

  • Biological roles: water’s necessity for the survival of plants and animals; how it’s ingested; the effects of water deprivation; etc.;

  • Practical roles: the roles water plays in agriculture, transport, washing, cooking, surfing, etc.;

  • Symbolic roles: water is strongly associated with cleanliness and purity, it plays an important role in many religious rituals, etc.;

  • Explanatory roles: water has a non-obvious explanatory structure, which explains many of its characteristic roles; water is composed of H20;

  • Epistemology: water is easy to spot but hard to define; our beliefs about water may be mistaken or incomplete; observation of instances of water grounds induction to unobserved cases.

Our aim is to answer to the “What is X?” question. The project is not semantic but meta-semantic; that is, we are not trying to find what the X-term means. Rather, we are trying to determine what the kind X is. The inputs just considered help us narrow down the kind so we can investigate it further. As we proceed, we may find that some of our background beliefs are false and our theoretical efforts misguided. It is only the result of our investigation that gives us the meaning of the term. But what do we do with these inputs? How do we balance various considerations? Schroeter and Schroeter (2015) propose that

. . . ideal epistemic methods for answering ‘what is x?’ questions hinge on rationalizing interpretation of one’s representational traditions. You need to diagnose the most important representational interests at (p.13) stake in a representational tradition with ‘x’, and you should identify the correct verdict about the nature of x as the one that makes best sense of those interests. (2015, 430)

A rationalizing interpretation, on their view, is not determined by reports of beliefs and intentions of participants in the tradition, nor is it a causal explanation of the tradition:

From the deliberative perspective of a rational epistemic agent, the interests that are relevant to adjudicating ‘what is x?’ questions are those that help justify or rationalize that tradition. Ideal methods for adjudicating ‘what is x?’ questions don’t simply construe representational practices as meeting psychologically or causally fixed representational interests. Our interpretive methods construe them as meeting representational interests that help make sense of our practices—that help construe them as having a point or rationale. (2015, 435)

In the case of ‘water,’ there are at least two candidates. One set of interests served by our attitudes toward and talk of water are explanatory, another set is practical. These two interests may come apart; for example, our practical interests do not require that we identify water with H2O, for liquids that are mostly H2O but contain other ingredients (harmless trace chemicals, fluoride) are fine for most purposes (drinking, bathing, swimming, etc.). However, scientific inquiry enables us to explain the properties of water—and how it can actually serve our practical interests—by reference to its chemical structure. This divergence of possible interpretations of what’s at stake in the tradition leaves us with two candidate answers to “What is water?” and so two candidates for the meaning of ‘water.’ Water is H2O, or water is the watery stuff found in lakes and rivers (etc.). It might appear that this leaves the term ‘water’ as ambiguous, or perhaps with no determinate meaning.

On the Schroeters’ (2015) view, there is a best interpretation of the representational tradition, where the scope of that tradition is determined by commitment to de jure sameness of reference and shared linguistic and epistemic practices (428). (We all take ourselves to be referring to the same thing in our thought and talk and are engaged in talking and thinking together.) What I mean is not just a function of what I think water is, or any old interpretation of our representational tradition: I can get the meaning wrong if I don’t do adequate justice to the interpretive task. For example, if I decide that, given our interests and collective uses of the term, water is the alcoholic beverage also (p.14) known as ‘beer,’ I would be wrong. I would have failed to capture a reasonable interpretation of our representational tradition. But I could also be wrong if I miss what is worth talking about:

As rational epistemic agents, we normally take our words and thoughts to represent genuinely interesting and important features of the world—not just whatever happens to satisfy our current criteria. When asking about the nature of water (or free will, color, etc.), we don’t assume that we (or our community as a whole) already implicitly know the right answer. (2015, 436)

We postulate ambiguity or opt for an error theory only as a last resort.

The Schroeters’ view seems to me to provide a kind of middle ground between adopting the “semantic strategy” and moving entirely to normative considerations. Recall the previous suggestion that in undertaking at least certain kinds of philosophical inquiry, what we are doing (roughly) is interpreting an indeterminate tradition and deciding “how to go on” with our practices. In doing philosophy, we are both interpreting and recommending. However, they seem to suggest that there is one “best” way to rationalize the representational (and practical) tradition, and so just one way to go on. I find this implausible and unnecessary.6 Different communities may highlight different parts of the tradition because of what is important to them, what practices they are committed to, what questions they ask, and how the world around them pushes back (e.g., what else they come to know).

1.5. Representational Traditions: ‘Race’

Does the Schroeters’ model give us resources to make progress in understanding what race is? The case of race is clearly more complicated: there are substantive disagreements about the different roles the idea plays in our representational tradition, and the tradition has clearly changed over time. This is something we should take into account. Moreover, there are significant differences in how the idea of race functions in different cultures; for (p.15) example, which racial group one belongs to may differ depending on the country one is in, and the background beliefs about races may differ. In the case of water, there is a basic human interest in being able to refer to the stuff in question, and at least most languages will have some way of talking about it. This is much less obvious in the case of race. So the idea that there is a single best interpretation of what race is—across languages and cultures—is not entirely plausible.

For example, the United States has relied—sometimes implicitly and sometimes explicitly—on a rule of hypodescent (i.e., assuming a racial hierarchy, the child of individuals of different races is assigned the “lower” race of the two parents).7 However, social scientists have found a variety of other rules for assigning race in the case of “mixed” offspring. In some societies (such as Hawaii, at least before statehood), “mixed” offspring are fully included as members of both races (Davis 1995, 116). In other societies, children of differently raced parents constitute a separate group that, depending on the case, may be considered inferior to, superior to, and or between the racial groups of the parents. In parts of Latin America, the race of “mixed” offspring does not depend simply on ancestry, but also on “economic and educational achievements”:

Whites are at the top of these class structures and unmixed blacks and Indians on the bottom. Blacks are defined as only those of unmixed African descent. Although the many rungs on the long status ladder are indicated by terms that describe the highly variable physical appearance of mulatto and mestizo individuals, this racial terminology can be quite misleading. These are actually class systems in which lifestyle is much more important than racial ancestry or physical traits. “Money whitens” as the phrase goes, and a person who rises in educational and economic status is identified by whiter racial designations.

(Davis 1995, 119)

And finally, in some cases, there is a possibility of assimilation, so that after some number of generations, “mixed” offspring can become members of the superior or dominant race. Thus, at least currently in the United States, individuals who are, say, one-eighth Asian and seven-eighths White may count as White with an “Asian” heritage (Davis 1995, 120). The practices of racial identification are also evolving.

(p.16) So when I suggest we consider “our representational tradition,” what do I have in mind? For the purposes of our discussion I will be focusing on what race is in the United States, keeping in mind that the goal is to provide an interpretation of what has plausibly been at issue (though not always clearly at issue) “all along,” as evidenced not only by what we say, but what we do, such as the practices we engage in, the laws we pass, and social scientific explanations of these. Given the history of the United States, the representational tradition draws upon some historical uses of the term and practices in Europe as well. In the following I’ve made a start on the relevant inputs to deliberation about what race is within this tradition. Inputs include both ideas that I take to be broadly shared in the United States, ideas from both natural and social sciences, and normatively relevant ideas, as the model recommends, though some will be controversial. I will use the term ‘raceus’ to designate what I take to be outputs of deliberation about this representational tradition. I hope that in interpreting our own tradition we will gain insight into related ones.8 I also use an initial upper-case letter for the names of purported races.

  • Particular instances: When we say that Martin Luther King, Jr., is Black, Hillary Clinton is White, Che Guevara is Latino, Sacagawea is Native American, and Aung San Suu Kyi is Asian, we are classifying each as belonging to a different race. Everyone belongs to at least one race, possibly more than one. The criteria for racial membership varies depending on context and is not consistent: the US government relies primarily on self-identification; epidemiologists and demographers sometimes rely on self-reports, but also on birth certificates, mother’s birth certificate, death certificates, doctor’s (or other’s) attribution of race (Root 2001, 2003, 2009). Generally, however, one’s racial designation is confirmed or disconfirmed by facts about whether one’s ancestry derives from a particular geographical region or regions. It is possible for someone to belong to a race without knowing that they do, e.g., an illiterate Kayin peasant from Myanmar is racially Asian, even though she may know nothing about Asia (p.17) as a continent or US racial practices. There is disagreement about whether Latino(a)s are a race (Gracia 2007).

  • Perceptual gestalts: Members of different races can usually be distinguished by physical features such as skin color, hair texture and color, eye shape; it is sometimes difficult to identify the racial makeup of mixed-race people, so perceptual gestalts are fallible, and some individuals do not have the distinctive features associated with their race (and so may intentionally or unintentionally “pass” as a member of a different race).

  • Biological roles: People inherit their race (though the criteria for inheritance have been contested and variable over time, and seem to differ, depending on the race at issue). Race is correlated with differences in life expectancy, various diseases, etc. Historically, it was thought (and some, but not all, people continue to believe) that one’s race is part of one’s nature, and at least in the case of some races, it is passed along to biological offspring (though Whiteness, apparently, is not always passed on, though Blackness is, according to the system of hypodescent!). Scientific research suggests that there isn’t a meaningful biological basis for racial distinctions, sufficient to postulate racial “natures” or essences, though it is currently a matter of controversy whether there are minor biological differences among groups roughly corresponding to the most commonly assumed racial groups (Black, White, Asian, Native American, Pacific Islander) (see Spencer’s Chapter 3 in this volume; also Andreasen 2000, 2004; Kitcher 2007).

  • Historical roles: Attributions of race have played a major role in world history. For example, the trans-Atlantic slave trade and European colonization across the world were justified on the basis of beliefs about race. The most common racial divisions have been based (roughly) on appearances that differ between continents, but there has not been unanimity on what races there are (Bernasconi and Lott 2000; Herzog 1998, 288ff), and arguably we are in a historical moment when those of (apparent) Arab descent, having been White, are being re-racialized as non-White (Jamal and Naber 2008). US federal and state law has restricted the civil rights of members of non-White races, and there have also been attempts at legal remedies, e.g., affirmative action. Race continues to be a matter of heated social, political, and legal debate. The history of science reveals ongoing scientific attempts to justify claims about racial differences, especially in intelligence and character.

  • Practical roles: Race is a significant factor in the organization of social life, e.g., in patterns of association, housing, religion, employment, crime, (p.18) athletics. It is also an important part of many people’s identity and sense of solidarity with others, and contributes to shaping their life plans and political views.9

  • Symbolic roles: Race is strongly associated with cultural norms, artistic traditions, and forms of life; historically it has been associated with character traits and degrees of moral worth.

  • Explanatory roles: Race is used to explain a broad range of differences between social groups, including educational attainment, patterns of arrest and incarceration, health outcomes, social history, etc. It is also used to explain different interests, cultural and artistic tendencies, and political affiliation. These explanations vary in their form. Some purport to be biological explanations, others sociological explanations.

  • Epistemology: A person’s race is usually taken to be evident based on widely accepted perceptual gestalts; however, race has been hard to define, and many assumptions about race have been undermined by scientific inquiry; our beliefs about race may be mistaken or incomplete; nevertheless, observation of racial regularities grounds induction to unobserved cases.

Given these inputs (I don’t mean for these to be exhaustive—this is just a sample), the task is to provide an interpretation of our representational tradition.10 What are we doing when we divide humans into different races? What interests are being served? Is there an interpretation that rationalizes or justifies the tradition? If not, then should we reject the idea of race completely?

One might argue that, as in the case of ‘water,’ there are several different reasonable ways to go here. A first option is to note that the representational tradition concerning race includes a history of drawing distinctions between groups of people on the basis of certain bodily features (skin color, hair texture, eye shape, and the like) and postulating racial “natures” underlying these observable differences to explain further cultural and behavioral differences (p.19) and to justify unjust treatment of non-Whites. The tradition was in the material and cultural interests of Whites and continues to play a role in many Americans’ thinking about race. The Schroeters are clear, however, that the best interpretation of the representational tradition must capture what we have been thinking and talking about “all along.”

The method we have sketched precisely aims at determining what’s interesting and important relative to the subject’s own past representational tradition. So from the point of view of a rational epistemic agent, these pragmatic meta-cognitive methods are ideally suited to getting us closer to the truth about the interesting and important topics that we were thinking and talking about all along. (2015, 436)

It is reasonable to claim that our linguistic forebears were thinking and talking about races distinguished by racial natures or essences. Yet at this point we know that there are no racial “natures,” (i.e., a set of properties that a member of a race has necessarily, by virtue of which they are a member of the race, and that explains their characteristic behavior and abilities).11 If the point or purpose of the tradition was to attribute racial natures to humans, and there are no such racial natures, the representational tradition has failed and we should give it up. In short, we should be anti-realists about raceus, more specifically, error theorists.

One need not think that our representational tradition is invested in racial natures, however, in order to account for the inputs to deliberation about race. Michael Hardimon (2017; also 2003) has argued for a minimalist account of race according to which:

A race is a group of human beings

(C1) that, as a group, is distinguished from other groups of human beings by patterns of visible physical features,

(C2) whose members are linked by common ancestry peculiar to members of the group, and

(C3) that originates from a distinctive geographic location. (2017, 31)

(p.20) Hardimon argues that there are groups that satisfy the minimalist race concept, and so racesus exist. Given the simplicity and plausibility of Hardimon’s conditions, it would seem that his account is an excellent candidate for an interpretation of the inputs regarding race. Should we take this view to be sufficient and the task to be complete?

Joshua Glasgow (2009) rejects a minimalist view such as Hardimon’s. He maintains that according to the “ordinary concept of race” (i.e., the one that has the most currency in the contemporary United States), “while an individual’s particular race might depend on social factors, each racial group is, as a conceptual matter, defined only in terms of its visible, biological profile” (2009, 123) (see also Alcoff 2005). The condition Glasgow isolates, however, is not satisfied: human appearance falls on a broad spectrum, and the supposed visibly notable and biologically relevant clumping that would be required by the condition does not occur.12 Glasgow concludes that “since these groups’ putative distinctiveness is not, as a point of fact, legitimated by the biology, there are no races” (2009, 123). So the term is vacuous, and statements employing the term are false.

I agree with Glasgow (2009) that there are no existing human groups that meet the condition that there are inherited visible features that demarcate the races. Note, however, that an error theory about raceus has substantial costs, given that we are attempting to give an interpretation of the inputs described earlier. Not only would we have to claim that our attributions of race to individuals are false, but that the historical, symbolic, explanatory, practical, and epistemic roles of race are all founded on illusion. We would need to give up the idea that race explains certain group differences (from artistic traditions to health outcomes); we would need to give up the idea that race provides reasons for certain practical, historical, and symbolic choices. We could potentially replace these claims with the suggestion that false racial beliefs about racial groups explain the broad range of racial formations. But this is to take a substantive stand on difficult explanatory questions in the history and sociology of racial practices, racial institutions, and the like.

Although false beliefs about racial groups may be the best explanation of early forms of racial hierarchy (though I find even that questionable, given the economic and other forces at work), it is implausible that such beliefs are the best explanation of ongoing racial injustice, including the perpetuation (p.21) of economic and political injustice, social segregation, and cultural stigma (Haslanger 2016, 2017a). For example, the waning of racial essentialism is not sufficient to undermine the legacy of economic deprivation because belief in racial essentialism—or racial naturalism more generally—is not what runs the economy; nor does correcting false beliefs about race correct the legacy of centuries of legal and political wrongs. So an error theory about race needs to be supplemented with alternative explanations of apparently racial phenomena, including our abilities to deliberate about race, perform induction on racial regularities, and find meaning in racial identification.

Consider an analogy. For centuries, philosophers have attempted to provide necessary and sufficient conditions for being a person. Candidate necessary conditions have included the following:

  • X is a person only if X has a soul.

  • X is a person only if the stages of X are psychologically continuous.

  • X is a person only if X occupies a continuous living human body.

There are compelling examples suggesting that accepting any of these conditions fails to capture what we mean by person due to a mismatch between what the condition requires and the cases we judge to be persons. Should we conclude that there are no persons and be error theorists about person-talk? Surely not! For ordinary cases, we do fine in judging whether an individual is a person or not. There are hard cases and our practices in these cases are contested, but we usually have rules or laws for settling them, at least for the time being. The question remains whether these rules or laws for the controversial cases are appropriate or justified, and they are open to revision based on actual cases that need to be settled. Our practices are evolving: new technologies and medical discoveries have forced us to answer questions that never had to be faced before. Such open-endedness and revisability do not show that there are no persons, that person-talk is vacuous, or that we should discontinue practices that rely on a background idea of persons.

In discussing the example of person, I am relying on the methodology I have recommended: given the function of language to guide and coordinate our ongoing practices, we should investigate what something is by providing an interpretation of the representational (and practical) tradition that helps us make sense of the new and challenging cases. Admittedly, there may be circumstances in which the tradition cannot or should not be sustained. But holding fixed one condition on the application of a concept that actually serves multiple purposes is not a sufficient reason to reject or eliminate (p.22) the concept. Communication in a context does not require a rigid representational tradition that anticipates every empirical discovery and every technological change. We make do with rough overlapping understandings of phenomena that concern our shared practices, and update as life goes on.

A third option for interpreting the inputs sketched in the preceding would be to turn to a social constructionist account of raceus. When the representational tradition was historically postulating racial natures, the point was to provide an explanation of the striking observable differences in human appearance, behavior, and culture found through voyages of exploration and conquest. The explanations that were ready to hand at the time were based on biblical interpretation or neo-Aristotelian biology (Stocking 1994). These explanations of human differences, we have found, are faulty. But there were and continue to be differences among the groups that were then designated as races that call for interpretation and explanation. A better approach looks to social formations.

A social constructionist account (e.g., sociopolitical account or cultural account, to be discussed more fully in the next section) proposes that the conditions for being a member of a racial group are to be given in social terms, rather than in physical, biological, or other non-social terms.13 Consider, for example, slaves. Aristotle seems to have thought that there were natural slaves (Politics, Bk. 1): natural slaves are individuals who are incapable of sufficient practical reason to lead an autonomous life. Natural slaves, on his view, were justifiably owned by others and were better off as a result. But the idea of a natural slave is badly mistaken. Slaves are a social category, that is, to be a slave is to be owned by someone according to the laws or customs of one’s social milieu. A social constructionist account of slaves is an improvement on the naturalistic account that defines slaves in terms of their cognitive capacities. There are many other cases in which social constructionists have challenged naturalistic accounts, for example, of sex, gender, sexual desire, disability, parenthood, family, and race.

How does attention to social relations and social structures help us understand the observable differences among human groups? There is overwhelming evidence that differences between racial groups in educational attainment, health outcomes, incarceration rates, and the like are due to the looping effects of social structures that impose a racial hierarchy. Many (p.23) of the great achievements and cultural traditions of different races are also a product of living within such structures (Taylor 2016). This social structural hierarchy is partly a product of a history of false beliefs about races and racial natures, but such beliefs are systematically linked to cultural and material factors that are equally important in accounting for the systematic nature of racial differentiation and racial injustice; false beliefs are a small part, maybe even an eliminable part, of what sustains the system. For example, it is insufficient to explain racial differences in educational achievement simply in terms of false beliefs about the abilities or “natures” of Whites and people of color. Additional factors include the racial patterns of wealth and poverty, patterns of housing segregation, the dependence of school funding on property taxes, the expense of university education, hiring discrimination, and the social meaning of intelligence and education.14 Such social phenomena do not depend entirely on the psychological states of individuals (Epstein 2015; Haslanger 2017b).

This second, social constructionst, approach gains further support from the parallels with other scientific advances. Early explanations of many natural phenomena have been rejected over time and have been supplanted by better explanations without disrupting our representational traditions. Hippocrates was aware of and treated cancer, though he thought it was caused by an excess of black bile (thought to be one of the four humors) in the body; it is plausible that Hippocrates is part of our representational tradition concerning cancer (he is credited with the origin of the word karkinoma), in spite of the fact that some of his core assumptions about cancer have been thoroughly rejected. The idea that empirical hypotheses about the nature of a kind are analytically entailed by our use of a term would make scientific inquiry difficult (this is an old point made by Quine, Putnam, Kripke, and many others). If we could not substantially revise our understanding of kinds, then as we develop new hypotheses about a phenomenon, we would not be improving our understanding of a poorly understood kind, but investigating a new kind, thus obscuring the dynamics of inquiry.

One might argue, however, that shifting from a “natural” to a “social” kind is more than meaning can bear. But shifts across different categories of explanation are not uncommon. For example, medical conditions that were once (p.24) thought to have been the result of God’s punishment, or evil thoughts, or anxiety, have been shown to have straightforward physical causes. Various caste, class, and ethnic divisions have been thought to be established by divine law or nature, but are now understood in terms of the workings of social systems. For example, monarchs were once thought to gain their political legitimacy from God. To be a monarch is to have sacred power, invested in the family lineage. This explanation of a monarch’s legitimacy was eventually rejected, and yet we did not give up the idea of a monarch. Instead, alternative social accounts of monarchy were supplied.

A social constructionist account of raceus will also face challenges in accommodating some of the inputs to deliberation listed in the preceding. For example, it is commonly thought that race is inherited. But social position is only inherited metaphorically; one usually occupies a similar social position to one’s parents, not by virtue of “blood” but by virtue of social conditions and pressures. I mentioned earlier, however, that there are multiple ways of “tracing” race through ancestry, hypodescent being only one of them, even in the United States. This suggests that a commitment to the idea that race is inherited is not a fixed point. Moreover, the use of ancestry to track race is a phenomenon that an error theorist will also need to explain. Any account of raceus—whether realist or anti-realist, naturalist or constructionist—will need to include details that make sense of or explain away the complexities of representational and practical tradition. So there is much work to do.

1.6. What Is Race?

Even if the representational tradition concerning race allows for a social scientific analysis of the explanatory interests being served, two questions remain:

  1. (i) How exactly should a social constructionist capture what race is?

  2. (ii) Are our current interests served by continuing with the representational tradition concerning the term ‘race,’ or should we replace race with another term, e.g., ‘racialized group’?

I will consider (i) in this section and (ii) in the next.

Two forms of social constructionism about race have been proposed in the literature. One is the sociopolitical account, the other is the cultural account. The two accounts agree on many points, for example, that the current dominant racesus emerged in a particular historical context of White racial domination; that members of races are “marked” as having a particular appearance; (p.25) that the “marking” is taken to be evidence of where, geographically, the group mostly lived at a key moment in time; that racial groups function differently within the contemporary sociopolitical structure, and are positioned on a hierarchy. The primary differences between the two accounts are (a) the cultural account requires that races, as a group, share a culture,15 whereas the sociopolitical account does not, and (b) the sociopolitical account takes the sociopolitical hierarchy to be a defining feature of race, whereas the cultural account does not. I defend the sociopolitical account. Chike Jeffers (2013) elaborates and defends the cultural account.

In my earlier work (2012, Ch. 7), I argued that critical theorists should adopt the following core account of race, and use this to explicate other racial phenomena, such as racial identities, racial norms and traditions, racial narratives, racial oppression, racial justice, and the like.16

Social/Political Race (SPR): A group G is racialized relative to context C iffdf members of G are (all and only) those

  1. (i) who are observed or imagined to have certain bodily features presumed in C to be evidence of ancestral links to a certain geographical region (or regions)—call this “color”;

  2. (ii) whose having (or being imagined to have) these features marks them within the context of the background ideology in C as appropriately occupying certain kinds of social position that are in fact either subordinate or privileged (and so motivates and justifies their occupying such a position); and

  3. (p.26) (iii) whose satisfying (i) and (ii) plays (or would play) a role in their systematic subordination or privilege in C, that is, who are along some dimension systematically subordinated or privileged when in C, and satisfying (i) and (ii) plays (or would play) a role in that dimension of privilege or subordination.17

The idea is that racesus are racialized groups, that is, those groups demarcated by the geographical associations accompanying perceived body type, when those associations take on evaluative significance (or social meaning) concerning how members of the group should be viewed and treated, and the treatment situates the groups on a social hierarchy.

Thus, to say that Martin Luther King, Jr., is Blackus is to say that he is a member of a group that meets these conditions, and in particular, that he is marked in the United States as having relatively recent ancestry from Africa, and this situates him as subordinate in the social hierarchy of the United States. Moreover, to say that Whitesus have higher educational achievement than Latinxus is to say that a group that is marked as having recent ancestry (p.27) in Europe and that is situated as privileged, as a result, has higher educational achievement than those marked as having recent ancestry in Latin America and who are disadvantaged as a result. This claim reveals a correlation between certain forms of social subordination/privilege and outcomes. It does not itself make a causal claim. A relevant causal claim might be this: those who are marked and privileged as Whiteus have higher educational achievement because of their racially marked privilege. This is not a tautology, nor is it a vacuous explanation: a group with racial privilege could have educational success due to other causes. However, the explanation is far from being complete, for we would want to know how the privilege is more specifically related to the achievement.

The proposed SPR account also helps us explain certain aspects of racial meanings, artistic traditions, and cultural norms. On my view, races are distinct from ethnicities. An ethnicity is a cultural grouping—involving shared language, customs, social meanings, cultural formations—that typically (but not always) relies for its existence and coherence on geographical and genealogical connections, and sometimes carries (defeasible) presumptions about appearance. So Germans, Italians, Basques, Armenians, Berbers, Croats, Fula, Hausa, Gujarati, Icelanders, Kurds, Luo, Manchu, Mongols, etc., are ethnicities. Ethnicities are often positioned hierarchically within a society or broader sociopolitical formation. On my account, the hierarchical positioning of an ethnic group within a broader society (or broader political formation) is a process of racializing the group. The ethnicity may predate the racialization, and will (hopefully) continue after racialization has ended. Moreover, multiple ethnic groups may be racialized together as a single race; this may result in what Yen Le Espiritu (1992) calls pan-ethnicities. So, Asians are considered a racial group, but include many different ethnicities (e.g., Bamars, Bengalis, Gujarati, Han Chinese, Hindustani, Hmong, Hui, Japanese, Kashmiri, Khmer, Konkani, Korean, Manchu, Marathi, Mongols, Napali, Sinhalese, Tais, Telebu, Tibetens, Uyughur, Vietnamese, Zhuang, to name a few). Such ethnic groups do not share a form of life, and may have long-standing conflicts over land, religion, and politics (see also Alcoff 2000 on the different ethnic groups considered Hispanic or Latinx). The cultural differences between the ethnic groups does not prevent them from forming a race, however, because racialization is not in the first instance a matter of identity or shared culture, but of an imposed (ascribed) position in a sociopolitical formation.

The Africans who were forcibly brought to the United States came not as “blacks” or “Africans” but as members of distinct and various ethnic (p.28) populations. As a result of slavery, “the ‘Negro race’ emerged from the heterogeneity of African ethnicity” (Blauner 1972, 13). . . . Diverse Native American tribes also have had to assume the pan-Indian label in order to conform to the perceptions of the American State. . . . Similarly, diverse Latino populations have been treated by the larger society as a unitary group with common characteristics and common problems. . . . And the term ‘Asian-American’ arose out of the racist discourse that constructs Asians as a homogeneous group. Excessive categorization is fundamental to racism because it permits “whites to order a universe of unfamiliar peoples without confronting their diversity and individuality (Blauner 1972, 113).”

(Espiritu 1992, 6)

The development of a pan-ethnicity may emerge, for example, among Asian immigrants in the United States, but is an accomplishment, not a given. And such pan-ethnic identities do not necessarily extend to communities of origin; “group formation is not only circumstantially determined, but takes place as an interaction between assignment and assertion. . . . In other words, panethnic boundaries are shaped and reshaped in the continuing interaction between both external and internal forces” (Espiritu 1992, 7). Thus, Asian American may be a pan-ethnicity because Asian immigrants to the United States, and their descendants, form a sense of shared Asian American culture. This suggests that there are three relevant types of groups: ethnicities, pan-ethnicities, and races. Ethnicities have distinctive cultures. Races typically consist of people from multiple cultures. Pan-ethnicities emerge when multiple groups are racialized and treated as one group, and form an identity and way of life as a result.18 So Hmong, Japanese, Khmer, and Korean are ethnicities. They are all treated as Asian in the United States, and Asian Americans form a pan-ethnicity. Some individuals living in Asia may come to see themselves as Asian in response to the racialization of Asians in the United States (and elsewhere), so there may be a group larger than just Asian Americans who are members of the pan-ethnicity; we might call these the pan-ethnic Asians. But this does not make the group of people living, or with recent ancestry, in Asia who are “marked” as Asian, a pan-ethnicity. The large heterogeneous group does not have a shared culture. Asian is not recognized as an identity by those living outside of a process of Asian racialization; nonetheless, it is, or has been historically considered, a race. Plausibly also, Black is (p.29) not a pan-ethnicity, even if African American, or Diasporic African, is. This is a central difference between the SPR account I support from a cultural constructionist account that takes shared culture to be a defining feature of race (Jeffers 2013).

I agree with Jeffers that often identities and cultural practices associated with races (e.g., “African American” or “Asian American” or “Latinx”) offer creative (and protective!) resources for those who have been racialized (2013, 422) that go well beyond a response to oppression; and a pan-ethnicity such as “White” offers other creative opportunities, in addition to resources for domination—or even more often, escape from subordination. Pan-ethnic groups share at least some minimal culture. How this works will vary from context to context. At one time the ideology invoked racial essences to justify the differential treatment of different “colored” groups, and conceptions of the essence or spirit of a people was a basis for identity; ideology has also linked racially marked people with cultural traditions, histories, and talents. People so marked have shared experiences, and some have bonded together in celebration and resistance. This has resulted in racially identified artistic movements, cultural norms, and forms of association. I do not claim that racially inflected culture is all about the position of the group in a hierarchy. Culture is dynamic and relatively autonomous from, and so not determined by, economic, political, or historical factors with which it is always manifested (Sewell 2005). But such pan-ethnicities are not races, or so I would argue. An individual ethnic Hmong living in China or Laos is, I would maintain, Asianus, even if there is nothing distinctively “Asian” about Hmong culture, and she does not identify as Asian (and maybe has not even heard of the designation). That she counts as Asianus is clear, however, by how Hmong are viewed and treated within the United States (Fadiman 2012), and how she would be viewed and treated if she came here.

Although I believe that SPR is a reasonable interpretation of the representational tradition concerning race, there are also reasons to resist it. This is to be expected, given that I embrace the idea that we can reasonably draw different conclusions about what is crucial to our representational and practical traditions, depending on the questions we ask and the purposes we bring to the inquiry. For example, some definitely take their racial identity to be an important part of who they are, and it is offensive to them to regard it as a response to racial subordination or privilege. It is important to note, however, that it does not follow from SPR that a racial identity must focus on facts of subordination or privilege, and nothing I have said entails that it is wrong or illegitimate to embrace an identity or (p.30) the distinctive way of life that has emerged with the pan-ethnicity. In fact, I believe that many forms of racial identity are important, valuable, and in some cases even inevitable responses to racial hierarchy. As I see it, a racial identity is a kind of know-how for navigating one’s position in racialized social space (Haslanger 2012, Ch. 9). The apt content for a racial identity, then, may be positive, affirming, and empowering, even if the racialized social position one occupies is oppressive.

There is a key normative difference, I think, between the sociopolitical account of race and the cultural account that becomes clear when one asks why hierarchy is built into race according to the SPR. Why not say that races are groups who are “marked” by reference to ancestry and geography, where this marking has implications for the group’s social position, without claiming that the social positions in question need be arranged hierarchically? If I drop the hierarchy condition, then the account comes much closer to the cultural account, on the assumption that those who occupy the same social position are likely to share some non-trivial practices that would amount to at least a thin “way of life.” Jeffers argues that we should adopt an account of race that does not have the result that race is eliminated once racial hierarchy is eliminated. He suggests:

From the cultural perspective, though, a situation in which racial groups persist but in a state of equality rather than socioeconomic and Eurocentric cultural hierarchy, respecting and mutually influencing each other while remaining relatively distinct, is a coherent and admirable goal.

(Jeffers 2013, 421)

I worry, however, about the extent to which we should embrace cultural groups marked by ancestry and appearance in the long run (of course in the short run, they are necessary to achieve justice). Currently, ethnic groups carry a presumption of shared ancestry, appearance, and geography, but this is merely a presumption. At least many cultural groups (understood as groups sharing a way of life, a language, a religion, a set of common practices) have porous boundaries: one can marry into them, convert, immigrate, look very different from other members, not originate where other members originated. Jeffers emphasizes the benefits of racial cultural unity, but not the costs of racial segregation. As I see them, the costs include tendencies to cultural norming and authenticity tests of those with a “marked” racial appearance (this results in the arguably slurring racial terminology of ‘oreo,’ ‘banana,’ ‘twinkie,’ ‘apple,’ ‘coconut,’ and ‘egg’). It also suggests that those without the right physical and (p.31) ancestral credentials don’t belong in the culture, shouldn’t participate in the way of life, and are suspect when they build strong alliances with and take up the practices of those who satisfy the racial conditions. Living, myself, in a mixed-race (Black-White) and cross-cultural (Jewish-Christian) family, I may be overly influenced by the huge contemporary challenges posed by racial (ancestry and appearance-based) membership criteria in cultural practices and cultural communities. These challenges could—and I think should—subside under conditions of justice. I find problematic the idea that a just world is one in which cultural groups can restrict their membership on racial grounds. I embrace, instead, a model of multiple coexisting cultures that are mutable, flexible, and creatively tolerant around issues of ancestry and appearance.

Clearly there is more to be said about the ways in which SPR does or does not make sense of our representational tradition concerning race. I believe, however, that it is an excellent candidate, given the Schroeter-style method, for determining at least one thing race is, and so at least one thing we can mean by ‘race.’

1.7. Going On: The Normative Dimension of Racial Classification

I think it is unquestionable that SPR captures an important set of social groups. They are those groups that have been racialized. Drawing on the Schroeters’ methodology, I have also argued that there is a good case to be made that SPR is a reasonable interpretation of our ongoing representational tradition and social practices with respect to the idea of race. However, I do not want to be committed to there being a single best interpretation of that tradition, nor do I think that how we should go on with our representational practices depends entirely on what our past practices have endeavored to identify as an important matter of shared concern. Even if the best interpretation of the tradition shows that it is semantically permissible to use the term ‘race’ along the lines that SPR suggests, that does not settle how or whether we should continue to use the term. In other words, even if we can isolate a set of social groups that are reasonably considered races, we could still decide not to use the term anymore, or to use a new term.

So the question remains whether our current interests are served by continuing with tradition of using the term ‘race.’ For example, some theorists have chosen to reject the term ‘race’ because of its problematic history in justifying racial injustice, and have opted instead for terminology that echoes but does not maintain the term (e.g., ‘race’ is replaced by ‘racialized group’; Blum (p.32) 2002). As mentioned at the start, I enter this debate as a social critic, and believe we can criticize our past practices and recommend changes to them. This includes changes to our linguistic practices.

On my view, this is a practical and political issue that is best answered by well-informed activists at a specific historical moment. As Mallon suggested, there are empirical and normative considerations that matter, for example, “the epistemic value of ‘race’ talk in various domains, the benefits and costs of racial identification and of the social enforcement of such identification, the value of racialized identities and communities fostered by ‘race’ talk, the role of ‘race’ talk in promoting or undermining racism, the benefits or costs of ‘race’ talk in a process of rectification for past injustice, the cognitive or aesthetic value of ‘race’ talk, and the degree of entrenchment of ‘race’ talk in everyday discourse” (2006, 550, also quoted earlier). How we go on also depends on the sources of solidarity that unify and empower a movement, and the importance of consistent demographic information across time and domain. These are clearly not questions that can be addressed a priori, and depend enormously upon context and moment (Shelby 2005).

To say that the issue is best addressed by well-informed activists, however, is not to relinquish philosophical input. Suppose we find reasons to think that the racialization of groups is a bad thing and that society would be better if we were to acknowledge and respect ethno-cultural differences but cease to think and act in racial terms. (I think there are compelling reasons of this sort, and briefly discussed this in the previous section.) It would be unrealistic, I think, to suggest that we can achieve such a society simply by ceasing to use racial terminology, by becoming “color blind,” or by denying that races are real. This is because racialization has caused tremendous social and economic harms, and reparative justice is required. But how can we go on, if on the one hand, it would be wrong to continue our current racial practices, and on the other, it would also be wrong to ignore the legacy of what’s been done?

One strategy mentioned earlier is to employ a new term for the groups that have been racialized. But there are two risks here. First, most neologisms don’t catch on. Second, racial identity has a deep and pervasive grip on Americans. It is very difficult to cast off an identity without offering another in its place, for identities shape our relations to others, the practices we engage in, and the possibilities we imagine. A second strategy is to offer a debunking account of race. Debunking accounts aim to shift our understanding to reveal how our prior thinking is false or misguided. The point is to disrupt our ways of thinking, to motivate a new relationship to our practices. This is the sort of account I think SPR provides. The hope is that if we can see that what we are (p.33) tracking with our racial classifications is something captured by SPR, then we will begin to see the importance of disrupting race and organizing ourselves on different terms.

Note, however, that debunking accounts are employed strategically; whether they are apt is highly sensitive to contextual factors. The goal, recall, is to challenge our investment in certain unjust practices whose injustice is occluded or masked. The debunking attempts to highlight features of the practices that make it hard for those of goodwill to continue enacting them. Yet there are different kinds of racial practices, and people engage in them with different degrees of awareness. Some practices we enact routinely, mindlessly. Others we enact in spite of knowing they harm us or others, for they define the broad shape of life in our social milieu. And others are recuperative practices that offer counter-hegemonic understandings and opportunities. Because debunking has an epistemic and political aim, it may not be necessary if the harm or wrong of the practices are transparent, or if the practices have already been turned toward justice (Botchkina 2016). In defending the SPR account, I offer it as an option to be taken up, or not, as a tool in moving forward toward racial justice.

1.8. Conclusion

It will become clear to the reader that my methodology for answering the question “What is race?” is different from that of my coauthors. According to all of them, we should be seeking an understanding of what we are ordinarily talking about when we talk of race, and with caveats mentioned earlier (i.e., that it isn’t all about our talk), I agree with that. But how do we determine what that is?

In answering the question “What is race?” there are semantic constraints on us. It would not be reasonable to answer, “Race is a type of furniture.” But the semantic constraints don’t determine how we must go on. There are different epistemic and pragmatic standards that may guide our interpretation of the representational tradition. And there are normative considerations about what practices we should continue and the best route for maintaining or discouraging them. I have argued that the SPR account is semantically permissible, and that in some contexts it is morally and politically valuable, depending on the practices that are being targeted and the epistemic position of those engaged in them. One of the important functions of language is to highlight features of the world that matter for coordination; the function of (p.34) SPR is to highlight—in the relevant cases—how our racializing practices and identities contribute to injustice.


Bibliography references:

Alcoff, Linda. 2000. “Is Latina/o Identity a Racial Identity?” In Jorge Gracia and Pablo DeGreiff (eds.), Hispanics/Latinos in the U.S.: Ethnicity, Race and Rights. New York: Routledge, pp. 23–44.

Alcoff, Linda. 2005. Visible Identities: Race, Gender and the Self. Oxford: Oxford University Press.

Anderson, Elizabeth S. 1995. “Knowledge, Human Interests, and Objectivity in Feminist Epistemology.” Philosophical Topics 23(2): 27–58.

Anderson, Elizabeth S. 2010. The Imperative of Integration. Cambridge, MA: Harvard University Press.

Andreasen, Robin. 2000. “Race: Biological Reality or Social Construct?” Philosophy of Science 67(3): 666.

Andreasen, Robin. 2004. “The Cladistic Race Concept: A Defense.” Biology and Philosophy 19(3): 425–442.

Appiah, K. Anthony. 1996. “Race, Culture, Identity: Misunderstood Connections.” In A. Appiah and A. Gutmann (eds.), Color Conscious: The Political Morality of Race. Princeton, NJ: Princeton University Press, pp. 30–105.

Appiah, K. Anthony. 2002. “The State and the Shaping of Identity.” In The Tanner Lectures on Human Values, vol. 23, ed. Grethe B. Peterson. Salt Lake City: University of Utah Press, pp. 234–299.

Bernasconi, Robert, and Tommy L. Lott. 2000. The Idea of Race. Indianapolis and Cambridge, MA: Hackett.

Blauner, Robert. 1972. Racial Oppression in America. New York: Harper & Row.

Blum, Lawrence. 2002. “I’m Not a Racist, But . . .”: The Moral Quandary of Race. Ithaca, NY: Cornell University Press.

Botchkina, Ekaterina. 2016. “Issues in Objectivity and Mind-Dependence.” Ph.D. thesis, MIT.

Botchkina, Ekaterina, and Jerome Hodges. 2016. “Objectivity and Conceptual Change.” Unpublished manuscript.

Davis, F. James. 1995. “The Hawaiian Alternative to the One-Drop Rule.” In Naomi Zack (ed.), American Mixed Race. Lanham, MD: Rowman and Littlefield, pp. 115–131.

DuBois, W. E. B. 1991[1987]. “The Conservation of Races.” In Philip S. Foner (ed.), W. E. B. DuBois Speaks: Speeches and Addresses 1890–1919. New York: Pathfinder, pp. 73–85.

Epstein, Brian. 2015. The Ant Trap. Oxford: Oxford University Press.

Espiritu, Yen Le. 1992. “Ethnicity and Panethnicity.” In Espiritu, Asian American Panethnicity. Philadelphia: Temple University Press, pp. 1–18.

(p.35) Fadiman, Anne. 2012. The Spirit Catches You and You Fall Down. New York: Farrar, Straus, and Giroux.

Fraser, Nancy. 1989. “What’s Critical about Critical Theory? The Case of Habermas and Gender.” In Fraser, Unruly Practices. Minneapolis: University of Minnesota Press, pp. 113–143.

Garfinkel, Alan. 1981. Forms of Explanation. New Haven, CT: Yale University Press.

Glasgow, Joshua. 2009. A Theory of Race. New York: Routledge.

Gooding-Williams, Robert. 1998. “Race, Multiculturalism, and Democracy.” Constellations 5: 18–41.

Gracia, Jorge J. E. 2007. Race or Ethnicity? On Black and Latino Identity. Ithaca, NY: Cornell University Press.

Haslanger, Sally. 2012. Resisting Reality: Social Construction and Social Critique. Oxford: Oxford University Press.

Haslanger, Sally. 2014. “Studying While Black: Trust, Opportunity, and Disrespect.” DuBois Review 11(1): 109–136.

Haslanger, Sally. 2016. “What Is a (Social) Structural Explanation?” Philosophical Studies 173: 113–130.

Haslanger, Sally. 2017a. “Racism, Ideology, and Social Movements.” Res Philosophica 94(1): 1–22.

Haslanger, Sally. 2017b. “Failures of Individualism: Materiality.” Presented at the First Annual Critical Social Ontology Workshop, St. Louis, MO.

Haslanger, Sally. Forthcoming. “Going On, Not in the Same Way.” In Alexis Burgess, Herman Cappelen, and David Plunkett, (eds.), Conceptual Engineering and Conceptual Ethics. Oxford: Oxford University Press.

Hardimon, Michael. 2003. “The Ordinary Concept of Race.” Journal of Philosophy 100(9): 437–455.

Hardimon, Michael. 2017. Rethinking Race: The Case for Deflationary Realism. Cambridge, MA: Harvard University Press.

Herzog, Donald. 1998. Poisoning the Minds of the Lower Orders. Princeton, NJ: Princeton University Press.

Hochman, Adam. 2017. “Replacing Race: Interactive Constructionism about Racialized Groups.” Ergo 4(3): 61–92.

Irwin, Neil, Claire Cane Miller, and Margo Sanger-Katz. 2014. “America’s Racial Divide, Charted.” New York Times, August 19. https://www.nytimes.com/2014/08/20/upshot/americas-racial-divide-charted.html

Jamal, Amaney, and Nadine Naber. 2008. Race and Arab Americans before and after 9/11: From Invisible Citizens to Visible Subjects. Syracuse, NY: Syracuse University Press.

Jeffers, Chike. 2013. “The Cultural Theory of Race: Yet Another Look at Du Bois’s ‘The Conservation of Races,’” Ethics 123(April 2013): 403–426.

Kapusta, Stephanie. 2016. “Misgendering and Its Moral Contestability.” Hypatia 31(3): 502–519.

(p.36) Kendig, Catherine. 2011. “Race as a Physiosocial Phenomenon.” History and Philosophy of the Life Sciences 33(2):191–222.

Kitcher, Philip. 2007. “Does ‘Race’ Have a Future?” Philosophy and Public Affairs 35(4):293–317.

Lear, Jonathan. 1986. “Transcendental Anthropology.” In Philip Pettit and John McDowell (ed.), Subject, Thought and Context. Oxford: Oxford University Press, pp. 267–298.

Mallon, Ron 2003: “Social Construction, Roles, and Stability.” In Frederick F. Schmitt (ed.), Socializing Metaphysics: The Nature of Social Reality. Lanham, MD: Rowman and Littlefield, pp. 327–353.

Mallon, Ron. 2006. “‘Race’: Normative, Not Metaphysical or Semantic.” Ethics 116(3): 525–551.

Marx, Karl. 1843. Letter to Ruge, Kreuzenach, September. http://marx.eserver.org/1843-letters.to.arnold.ruge/1843.09-ruthless.critique.txt

McPherson, Lionel, and Tommie Shelby. 2004. “Blackness and Blood: Interpreting African American Identity.” Philosophy and Public Affairs 32(2): 171–192.

Mills, Charles W. 1997. The Racial Contract. Ithaca, NY: Cornell University Press.

Mills, Charles W. 1998. Blackness Visible. Ithaca: Cornell University Press.

Mills, Charles W. 2017. Black Rights/White Wrongs: The Critique of Racial Liberalism. Oxford: Oxford University Press.

Omi, M., and H. Winant. 1994. Racial Formation in the United States. New York: Routledge.

Quine, W. V. O. 1953. “On What There Is.” In Quine, From a Logical Point of View. Cambridge, MA: Harvard University Press, pp. 1–19.

Risjord, M. W. 2000. Woodcutters and Witchcraft: Rationality and Interpretive Change in the Social Sciences. Albany: State University of New York Press.

Root, Michael. 2001. “The Problem of Race in Medicine.” Philosophy of the Social Sciences 31(1): 20–39.

Root, Michael. 2003. “The Use of Race in Medicine as a Proxy for Genetic Differences.” Philosophy of Science 70(5): 1173–1183.

Root, Michael. 2009. “Measurement Error in Racial and Ethnic Statistics.” Biology and Philosophy 24: 375–385.

Rosen, Gideon. 2015. “Real Definition.” Analytic Philosophy 56(3): 189–209.

Saul, Jennifer M. 2012. “Politically Significant Terms and Philosophy of Language: Methodological Issues.” In Sharon L. Crasnow and Anita M. Superson (eds.), Out from the Shadows: Analytical Feminist Contributions to Traditional Philosophy. Oxford: Oxford University Press, pp. 195–214.

Schroeter, Laura, and Francois Schroeter. 2015. “Rationalizing Self-Interpretation.” In Chris Daly (ed.), The Palgrave Handbook of Philosophical Methods. New York: Palgrave-Macmillan, pp. 419–447.

Sewell, William. 2005. “The Concept(s) of Culture.” In Gabrielle M. Spiegel (ed.), Practicing History: New Directions in Historical Writing after the Linguistic Turn. New York and London: Routledge, pp. 76–95.

(p.37) Shelby, Tommie. 2005. We Who Are Dark. Cambridge, MA: Harvard University Press.

Shelby, Tommie. 2014. “Racism, Moralism, and Social Criticism.” DuBois Review 11(1): 57–74.

Spencer, Quayshawn. 2014. “A Radical Solution to the Race Problem.” Philosophy of Science 81(5): 1025–1038.

Sundstrom, Ronald. 2002. “Race as a Human Kind.” Philosophy and Social Criticism 28(1): 91–115.

Stocking, George Jr. 1994. “The Turn-of-the-Century Concept of Race.” Modernism/Modernity 1(1): 4–16.

Taylor, Paul. 2004. Race: A Philosophical Introduction. Cambridge: Polity Press.

Taylor, Paul. 2016. Black Is Beautiful: A Philosophy of Black Aesthetics. Hoboken, NJ: Wiley-Blackwell.

Zack, Naomi. 2002. Philosophy of Science and Race. New York: Routledge.


(1.) Some parts of this section are repeated and developed more fully in Haslanger (forthcoming).

(2.) Metaphors and analogies can play an important and even ineliminable role in theorizing and can aid in explanation. My claim here is only that the choice of terminology for the functional kinds in the proposed mechanisms of gentrification (specifically the influx of singles) is not essential to the success of the model for some purposes (though it may be for others).

(3.) I embrace an “eretetic” approach to explanation that takes explanations, and theories more generally, to be answers to questions. So the first task of any theoretical project is to clarify the question being asked. Apparent disagreements can sometimes be resolved by noting that the parties to the disagreement are answering different questions (see Garfinkel 1981; Risjord 2000; Anderson 1995).

(4.) I assume for the purposes of this discussion that a racial realist believes that at least some statements involving the term ‘race’ are both truth-apt and true. Anti-realists disagree. Anti-realists may hold that all statements involving the term ‘race’ are not truth-apt (they are “non-cognitivists” about race talk), or they may hold that race talk is truth-apt, but false (they are “error theorists” about race). In this book, Spencer, Jeffers, and I are all realists (though we disagree about what makes race talk true); Glasgow is either an anti-realist error theorist or a “basic” realist that allows ‘race’ to refer, but to uninteresting groups. (Other anti-realists include Appiah [1996] and Zack [2002]; Blum [2002]; Hochman [2017].)

(5.) For a parallel argument concerning gender terms such as ‘woman,’ and ‘man,’ see Saul (2012) and Kapusta (2016).

(6.) Botchkina and Hodges (2016) defend a view similar to theirs, but that allows for multiple reasonable interpretations. Moreover, the Schroeters’ view is more individualistic than my own. On their view, a primary normative constraint is to provide a rationalizing self-interpretation, i.e., to make sense of one’s own beliefs and practices (linguistic and otherwise). I see this as a more collective project. See also Haslanger (forthcoming) for an elaboration of the idea that conceptual amelioration through such reflection is possible.

(8.) I have found it challenging to judge which instances of the word ‘race’ should include the subscript ‘us’. My goal has been to leave the subscript off when we are considering candidate inputs to the deliberation—allowing that they may or may not be aptly considered a core part of the phenomenon and may be simply associations or related phenomena—but adding ‘us’ when drawing conclusions about the US phenomenon we’re aiming to track. I’m not sure I’ve been wholly consistent in this because it isn’t obvious, to me at least, what occurs as part of deliberation and what occurs as a result. My apologies for any confusion this may cause.

(9.) There is a broad literature on the content of racial identities, and African American or Black identity in particular. A helpful philosophical snapshot may be found in Gooding-Williams (1998); Appiah (2002); Shelby and McPherson (2004); Shelby (2005); Gracia (2007); Kendig (2011).

(10.) As I read Quayshawn Spencer’s (2014) argument for a modest racial naturalism, he could agree with the Schroeters’ approach that has us trying to make sense of a representational/practical tradition; he chooses to place great weight on the federal discourse around race as regimented by the census. I don’t think his choice of emphasis takes sufficient account of the many functions of race in our ordinary discourse and places too much weight on the role of the state; but as I make clear in my reply, our priorities are different.

(11.) I prefer the term ‘nature’ to ‘essence’ in this context because of the complexities in the historical and contemporary use of ‘essence,’ though ‘essence’ is the more commonly used for this postulation. Think of something like the nature of a tiger—each tiger has a set of properties necessarily by virtue of which it counts as a tiger (tiger is its kind), and this set of properties explains its behavior and abilities, e.g., it is by nature a feline, a carnivore, etc.

(12.) An extension of this argument is also relevant to discussion of Quayshawn Spencer’s (2014) minimal naturalism (see Chapter 3 for details).

(13.) Social constructionism about race takes many forms. For other examples, see Omi and Winant (1994); Mills (1997, 1998); Gooding-Williams (1998); Sundstrom (2002); Mallon (2003); Taylor (2004); Alcoff (2005).

(14.) There is a huge social science literature on racial health gaps, educational achievement gaps, etc. For a glance at the numbers in 2014, see Irwin et al. (2014). An important approach to explaining this is offered in Mills (2017, Ch. 7). See also Anderson (2010) and Haslanger (2014).

(15.) A cultural constructionist view does not require that every member of the group participates fully in the culture; rather, a group does not count as a racialus group unless it represents a particular form of life. DuBois is often taken as offering a paradigm of the cultural account, suggesting that a race is “a vast family of human beings, generally of common blood and language, always of common history, traditions and impulses, who are both voluntarily and involuntarily striving together for the accomplishment of certain more or less vividly conceived ideals of life” (DuBois 1991[1987], 75–76), also quoted in Jeffers (2013, 405). In other words, the set of conditions that make a group a racialus group may include reference to a form of life, but the conditions for being a member of a racialus group may not include this condition; e.g., the condition could simply be that one’s parents are a member of the group. So, for example, the Jewish people have a particular form of life, but not all Jews are observant. Nevertheless, one is a member of the Jewish people by virtue of being born of a Jewish mother (or in some forms of Judaism, a Jewish mother or father, or by conversion), not by virtue of observing the practices of Judaism.

(16.) Note that the term ‘iffdf’ is sometimes used to indicate that the biconditional is offering a definition of a word or a concept, I intend it here to indicate that I’m answering a “What is X?” question or “What is it to be X?” question, i.e., to give what is sometime called a “real definition.” See, e.g., Rosen (2015).

(17.) There are several aspects of this definition that need further elaboration or qualification. First, the definition does not accommodate contexts such as Brazil in which membership in “racial” groups is partly a function of education and class. This is because my project here is to capture what race is in the contemporary United States, i.e., raceus. However, a related racial phenomenon can be found in other representational/practical traditions and another version on which appropriate “color” is relevant but not necessary might be captured by modifying the second condition:

  1. ((ii*)) having (or being imagined to have) these features—in combination with factors such as economic and educational status—marks them within the context of C’s cultural ideology as appropriately occupying the kinds of social position that are in fact either subordinate or privileged (and so motivates and justifies their occupying such a position).

The first condition already allows that the group’s members may have supposed origins in more than one region (originally necessary to accommodate the racialization of “mixed-race” groups); modifying the second condition allows that racialized groups may include people of different “colors” and may depend on a variety of factors. Second, I want the definition to capture the idea that members of racial groups may be scattered across social contexts and may not all actually be (immediately) affected by local structures of privilege and subordination. So, for example, Black Africans and African Americans are together members of a group currently racialized in the US, even if a certain ideological interpretation of their “color” has not played a role in the subordination of all Black Africans; there are parallel phenomena in the case of other races. So I suggest that members of a group racialized in C are those who are or would be marked and correspondingly subordinated or privileged when in C. Those who think (plausibly) that all Blacks worldwide have been affected by the structures and ideology of White supremacy do not need this added clause; and those who want a potentially more fine-grained basis for racial membership can drop it.

(18.) See also Alcoff (2000) on ethno-race, and Gooding-Williams (1998).