From Noûs 34, 1 (2000): 31-55.


Reprinted by permission of the author and Blackwell Publishers.


Visit Noûs at:


Gender and Race: (What) Are They? (What) Do We Want Them To Be?

Sally Haslanger



If her functioning as a female is not enough to define woman, if we decline also to explain her through ‘the eternal feminine,’ and if nevertheless we admit, provisionally, that women do exist, then we must face the question: what is a woman?

—Simone de Beauvoir, The Second Sex


I guess you could chuckle and say that I’m just a woman trapped in a woman’s body.

—Ellen DeGeneres, My Point...and I Do Have One


The truth is that there are no races:  there is nothing in the world that can do all we ask race to do for us.

—Kwame Anthony Appiah, In My Father’s House


It is always awkward when someone asks me informally what I’m working on and I answer that I’m trying to figure out what gender is. For outside a rather narrow segment of the academic world, the term ‘gender’ has come to function as the polite way to talk about the sexes. And one thing people feel pretty confident about is their knowledge of the difference between males and females. Males are those human beings with a range of familiar primary and secondary sex characteristics, most important being the penis; females are those with a different set, most important being the vagina or, perhaps, the uterus. Enough said. Against this background, it isn’t clear what could be the point of an inquiry, especially a philosophical inquiry, into ‘what gender is’.

But within that rather narrow segment of the academic world concerned with gender issues, not only is there no simple equation of sex and gender, but the seemingly straightforward anatomical distinction between the sexes has been challenged as well. What began as an effort to note that men and women differ socially as well as anatomically has prompted an explosion of different uses of the term ‘gender’. Within these debates, not only is it unclear what gender is and how we should go about understanding it, but whether it is anything at all.

The situation is similar, if not worse, with respect to race. The self-evidence of racial distinctions in everyday American life is at striking odds with the uncertainty about the category of race in law and the academy. Work in the biological sciences has informed us that our practices of racial categorization don’t map neatly onto any useful biological classification; but that doesn’t settle much, if anything. For what should we make of our tendency to classify individuals according to race, apparently on the basis of physical appearance? And what are we to make of the social and economic consequences of such classifications? Is race real or is it not?

This paper is part of a larger project, the goal of which is to offer accounts of gender and race informed by a feminist epistemology. Here my aim is to sketch some of the central ideas of those accounts. Let me emphasize at the beginning that I do not want to argue that my proposals provide the only acceptable ways to define race or gender; in fact, the epistemological framework I employ is explicitly designed to allow for different definitions responding to different concerns. It is sometimes valuable to consider race or gender alone or to highlight the differences between them; however, here I will begin by exploring some significant parallels. Although there are dangers in drawing close analogies between gender and race, I hope my discussion will show that theorizing them together can provide us valuable resources for thinking about a wide range of issues. Working with a model that demonstrates some of the parallels between race and gender also helps us locate important differences between them.

1       The Question(s)

It is useful to begin by reflecting on the questions: ‘What is gender?’, ‘What is race?’ and related questions such as: ‘What is it to be a man or a woman?’[1], ‘What is it to be White? Latino? or Asian?’ There are several different ways to understand, and so respond to, questions of the form, ‘What is X?’ or ‘What is it to be an X?’ For example, the question ‘What is knowledge?’ might be construed in several ways. One might be asking: What is our concept of knowledge? (looking to apriori methods for an answer). On a more naturalistic reading one might be asking: What (natural) kind (if any) does our epistemic vocabulary track? Or one might be undertaking a more revisionary project: What is the point of having a concept of knowledge? What concept (if any) would do that work best?[2] These different sorts of projects cannot be kept entirely distinct, but draw upon different methodological strategies. Returning to the questions, ‘What is race?’ or ‘What is gender?’ we can distinguish, then, three projects with importantly different priorities: conceptual, descriptive, and analytical.

A conceptual inquiry into race or gender would seek an articulation of our concepts of race or gender (Riley 1988). To answer the conceptual question, one way to proceed would be to use the method of reflective equilibrium. (Although within the context of analytic philosophy this might be seen as a call for a conceptual analysis of the term(s), I want to reserve the term ‘analytical’ for a different sort of project, described below.)

In contrast to the conceptual project, a descriptive project is not concerned with exploring the nuances of our concepts (or anyone else’s for that matter); it focuses instead on their extension. Here, the task is to develop potentially more accurate concepts through careful consideration of the phenomena, usually relying on empirical or quasi-empirical methods. Paradigm descriptive projects occur in studying natural phenomena. I offered the example of naturalistic approaches to knowledge above: the goal is to determine the (natural) kind, if any, we are referring to (or are attempting to refer to) with our epistemic talk. However, a descriptive approach need not be confined to a search for natural or physical kinds; inquiry into what it is to be, e.g., a human right, a citizen, a democracy, might begin by considering the full range of what has counted as such to determine whether there is an underlying (possibly social) kind that explains the temptation to group the cases together. Just as natural science can enrich our ‘folk’ conceptualization of natural phenomena, social sciences (as well as the arts and humanities) can enrich our ‘folk’ conceptualization of social phenomena. So, a descriptive inquiry into race and gender need not presuppose that race and gender are biological kinds; instead it might ask whether our uses of race and gender vocabularies are tracking social kinds, and if so which ones.

The third sort of project takes an analytical approach to the question, ‘What is gender?’ or ‘What is race?’ (Scott 1986). On this approach the task is not to explicate our ordinary concepts; nor is it to investigate the kind that we may or may not be tracking with our everyday conceptual apparatus; instead we begin by considering more fully the pragmatics of our talk employing the terms in question. What is the point of having these concepts? What cognitive or practical task do they (or should they) enable us to accomplish? Are they effective tools to accomplish our (legitimate) purposes; if not, what concepts would serve these purposes better? In the limit case of an analytical approach the concept in question is introduced by stipulating the meaning of a new term, and its content is determined entirely by the role it plays in the theory. But if we allow that our everyday vocabularies serve both cognitive and practical purposes, purposes that might also be served by our theorizing, then a theory offering an improved understanding of our (legitimate) purposes and/or improved conceptual resources for the tasks at hand might reasonably represent itself as providing a (possibly revisionary) account of the everyday concepts.[3]

So, on an analytical approach, the questions ‘What is gender?’ or ‘What is race?’ require us to consider what work we want these concepts to do for us; why do we need them at all? The responsibility is ours to define them for our purposes. In doing so we will want to be responsive to some aspects of ordinary usage (and to aspects of both the connotation and extension of the terms). However, neither ordinary usage nor empirical investigation is overriding, for there is a stipulative element to the project: this is the phenomenon we need to be thinking about. Let the term in question refer to it. On this approach, the world by itself can’t tell us what gender is, or what race is; it is up to us to decide what in the world, if anything, they are.

This essay pursues an analytical approach to defining race and gender. However, its analytical objectives are linked to the descriptive project of determining whether our gender and race vocabularies in fact track social kinds that are typically obscured by the manifest content of our everyday race and gender concepts.[4] Although the analyses I offer will point to existing social kinds (and this is no accident), I am not prepared to defend the claim that these social kinds are what our race and gender talk is ‘really’ about. My priority in this inquiry is not to capture what we do mean, but how we might usefully revise what we mean for certain theoretical and political purposes.

My characterization of all three approaches remains vague, but there is one reason to be skeptical of the analytical approach that should be addressed at the outset. The different approaches I’ve sketched differ both in their methods and their subject matter. However, we come to inquiry with a conceptual repertoire in terms of which we frame our questions and search for answers: hence, the subject matter of any inquiry would seem to be set from the start. In asking what race is, or what gender is, our initial questions are expressed in everyday vocabularies of race and gender, so how can we meaningfully answer these questions without owing obedience to the everyday concepts? Or at least to our everyday usage? Revisionary projects are in danger of providing answers to questions that weren’t being asked.

But ordinary concepts are notoriously vague; individual conceptions and linguistic usage varies widely. Moreover, inquiry often demonstrates that the ordinary concepts used initially to frame a project are not, as they stand, well-suited to the theoretical task at hand. (This is one reason why we may shift from a conceptual project to an analytical one.) But precisely because our ordinary concepts are vague (or it is vague which concept we are expressing by our everyday use of terms), there is room to stretch, shrink, or refigure what exactly we are talking about in new and sometimes unexpected directions.

However, in an explicitly revisionary project, it is not at all clear when we are warranted in appropriating existing terminology. Given the difficulty of determining what ‘our’ concept is, it isn’t entirely clear when a project crosses over from being explicative to revisionary, or when it is no longer even revisionary but simply changes the subject. If our goal is to offer an analysis of ‘our’ concept of X, then the line between what’s explication and what’s not matters . But if our goal is to identify a concept that serves our broader purposes, then the question of terminology is primarily a pragmatic and sometimes a political one: should we employ the terms of ordinary discourse to refer to our theoretical categories, or instead make up new terms? The issue of terminological appropriation is especially important, and especially sensitive, when the terms in question designate categories of social identity such as ‘race’ and ‘gender’.

Are there principles that determine when it is legitimate to appropriate the terms of ordinary discourse for theoretical purposes? An answer, it seems to me, should include both a semantic and a political condition (though in some cases the politics of the appropriation will be uncontroversial). The semantic condition is not surprising: the proposed shift in meaning of the term would seem semantically warranted if central functions of the term remain the same, e.g., if it helps organize or explain a core set of phenomena that the ordinary terms are used to identify or describe.[5] Framing a political condition in general terms is much more difficult, however, for the politics of such appropriation will depend on the acceptability of the goals being served, the intended and unintended effects of the change, the politics of the speech context, and whether the underlying values are justified. We will return to some of these issues later in the paper once my analyses have been presented.

2       Critical (Feminist, Anti-Racist) Theory

In an analytical project we must begin by considering what we want the concept in question for. Someone might argue, however, that the answer is simple: our concepts must do the work of enabling us to articulate truths. But of course an unconstrained search for truth would yield chaos, not theory; truths are too easy to come by, there are too many of them. Given time and inclination, I could tell you many truths— some trivial, some interesting, many boring—about my physical surroundings. But a random collection of facts does not make a theory; they are a disorganized jumble. In the context of theorizing, some truths are more significant than others because they are relevant to answering the question that guides the inquiry. (Anderson 1995.)

Theorizing—even when it is sincerely undertaken as a search for truth—must be guided by more than the goal of achieving justified true belief. Good theories are systematic bodies of knowledge that select from the mass of truths those that address our broader cognitive and practical demands. In many contexts the questions and purposes that frame the project are understood and progress does not require one to investigate them. But in other contexts, e.g., especially when debate has seemed to break down and parties are talking at cross-purposes, an adequate evaluation of an existing theory or success in developing a new one is only possible when it is made clear what the broader goals are.

With this sketch of some of the theoretical options, I want to frame my own project as a critical analytical effort to answer the questions: ‘What is gender?’, ‘What is race?’ and the related questions ‘What is it to be a man?’ ‘...a woman?’, ‘...White?’ ‘...Latino?’ etc.’ More specifically, the goal of the project is to consider what work the concepts of gender and race might do for us in a critical—specifically feminist and antiracist—social theory, and to suggest concepts that can accomplish at least important elements of that work. (Guess 1981.) So to start: why might feminist antiracists want or need the concepts of gender and race? What work can they do for us?

At the most general level, the task is to develop accounts of gender and race that will be effective tools in the fight against injustice. The broad project is guided by four concerns:


    (i)            The need to identify and explain persistent inequalities between females and males, and between people of different ‘colors’[6]; this includes the concern to identify how social forces, often under the guise of biological forces, work to perpetuate such inequalities.

  (ii)            The need for a framework that will be sensitive to both the similarities and differences among males and females, and the similarities and differences among individuals in groups demarcated by ‘color’; this includes the concern to identify the effects of interlocking oppressions, e.g., the intersectionality of race, class, and gender. (Crenshaw 1993.)

(iii)            The need for an account that will track how gender and race are implicated in a broad range of social phenomena extending beyond those that obviously concern sexual or racial difference, e.g., whether art, religion, philosophy, science, or law might be ‘gendered’ and/or ‘racialized’.

(iv)            The need for accounts of gender and race that take seriously the agency of women and people of color of both genders, and within which we can develop an understanding of agency that will aid feminist and antiracist efforts to empower critical social agents.


In this paper I will begin to address the first two concerns, though the fourth will become relevant later in the discussion. Let me emphasize, however, that my goal in this paper is not to provide a thoroughgoing explanation of sexism and racism, if one wants by way of explanation a causal account of why and how females have come to be systematically subordinated throughout history, or why and how ‘color’ has come to be a basis for social stratification. My goal here is in some ways more modest, and in other ways more contentious. Prior to explanation it is valuable to provide clear conceptual categories to identify the phenomenon needing explanation, e.g., categories that identify the kind of injustice at issue and the groups subject to it. In the case of racial and sexual subordination this is not as easy as it may seem. In the first place, the forms of racial and sexual subordination are tremendously heterogeneous and it would help to have accounts that enable us to distinguish racial subordination and sexual subordination from other sorts. But further, we must be cautious about treating familiar demarcations of ‘color’ and ‘sex’ as purely natural categories, as if the question at hand is simply why one’s ‘color’ or sex—where we take for granted our familiar understandings of these terms—has ever seemed to be socially significant. At least at this stage of the inquiry we must allow that the criteria for distinguishing ‘colors’ or ‘sexes’ differ across time and place, and that the boundaries are at least partly political; but in spite of this variation, we are still dealing with an overarching phenomenon of racial and sexual subordination.

3       What Is Gender?

Even a quick survey of the literature reveals that a range of things have counted as ‘gender’ within feminist theorizing. The guiding idea is sometimes expressed with the slogan: ‘gender is the social meaning of sex’. But like any slogan, this one allows for different interpretations. Some theorists use the term ‘gender’ to refer to the subjective experience of sexed embodiment, or a broad psychological orientation to the world (‘gender identity’[7]); others to a set of attributes or ideals that function as norms for males and females (‘masculinity’ and ‘femininity’); others to a system of sexual symbolism; and still others to the traditional social roles of men and women. My strategy is to offer a focal analysis that defines gender, in the primary sense, as a social class. A focal analysis undertakes to explain a variety of connected phenomena in terms of their relations to one that is theorized as the central or core phenomenon. As I see it, the core phenomenon to be addressed is the pattern of social relations that constitute the social classes of men as dominant and women as subordinate; norms, symbols, and identities are gendered in relation to the social relations that constitute gender.[8] As will become clearer below, I see my emphasis as falling within, though not following uncritically, the tradition of materialist feminism.[9]

Among feminist theorists there are two problems that have generated pessimism about providing any unified account of women; I’ll call them the commonality problem and the normativity problem. Very briefly, the commonality problem questions whether there is anything social that females have in common that could count as their ‘gender’. If we consider all females—females of different times, places, and cultures—-there are reasons to doubt that there is anything beyond body type (if even that) that they all share (Spelman 1988). The normativity problem raises the concern that any definition of ‘what woman is’ is value-laden, and will marginalize certain females, privilege others, and reinforce current gender norms (Butler 1990, Ch. 1).

It is important to note, even briefly, that these problems take on a different cast when they arise within a critical analytical project. The emphasis of an analytical project is not on discovering commonalities among females: although the empirical similarities and differences between females are relevant, the primary goal is an analysis of gender that will serve as a tool in the quest for sexual justice (see section II). Moreover, a critical project can accept the result that an effort to define ‘what women is’ carries normative implications, for critical projects explicitly embrace normative results; the hope is that the account’s implications would not reinforce but would help undermine the structures of sexual oppression. However, we will return to these issues below.

Given the priority I place on concerns with justice and sexual inequality, I take the primary motivation for distinguishing sex from gender to arise in the recognition that males and females do not only differ physically, but also systematically differ in their social positions. What is of concern, to put it simply, is that societies, on the whole, privilege individuals with male bodies. Although the particular forms and mechanisms of oppression vary from culture to culture, societies have found many ways—some ingenious, some crude—to control and exploit the sexual and reproductive capacities of females.

The main strategy of materialist feminist accounts of gender has been to define gender in terms of women’s subordinate position in systems of male dominance.[10] Although there are materialist feminist roots in Marxism, contemporary versions resist the thought that all social phenomena can be explained in or reduced to economic terms; and although materialist feminists emphasize the role of language and culture in women’s oppression, there is a wariness of extreme forms of linguistic constructivism and a commitment to staying grounded in the material realities of women’s lives. In effect, there is a concerted effort to show how gender oppression is jointly sustained by both cultural and material forces.

Critiques of universalizing feminisms have taught us to be attentive to the variety of forms gender takes and the concrete social positions females occupy. However it is compatible with these commitments to treat the category of gender as a genus that is realized in different ways in different contexts; doing so enables us to recognize significant patterns in the ways that gender is instituted and embodied. Working at the most general level, then, the materialist strategy offers us three basic principles to guide us in understanding gender:


    (i)            Gender categories are defined in terms of how one is socially positioned, where this is a function of, e.g., how one is viewed, how one is treated, and how one’s life is structured socially, legally, and economically; gender is not defined in terms of an individual’s intrinsic physical or psychological features.


(This allows that there may be other categories—such as sex—that are defined in terms of intrinsic physical features. Note, however, that once we focus our attention on gender as social position, we must allow that one can be a woman without ever (in the ordinary sense) ‘acting like a woman’, ‘feeling like a woman’, or even having a female body.)


  (ii)            Gender categories are defined hierarchically within a broader complex of oppressive relations; one group (viz., women) is socially positioned as subordinate to the other (viz., men), typically within the context of other forms of economic and social oppression.

(iii)            Sexual difference functions as the physical marker to distinguish the two groups, and is used in the justification of viewing and treating the members of each group differently.


(Tentatively) we can capture these main points in the following analyses:


S is a woman iffdf S is systematically subordinated along some dimension (economic, political, legal, social, etc.), and S is ‘marked’ as a target for this treatment by observed or imagined bodily features presumed to be evidence of a female’s biological role in reproduction.[11]

S is a man iffdf S is systematically privileged along some dimension (economic, political, legal, social, etc.), and S is ‘marked’ as a target for this treatment by observed or imagined bodily features presumed to be evidence of a male’s biological role in reproduction.


It is a virtue, I believe, of these accounts, that depending on context, one’s sex may have a very different meaning and it may position one in very different kinds of hierarchies. The variation will clearly occur from culture to culture (and sub-culture to sub-culture); so e.g., to be a Chinese woman of the 1790’s, a Brazilian woman of the 1890’s, or an American woman of the 1990’s may involve very different social relations, and very different kinds of oppression. Yet on the analysis suggested, these groups count as women insofar as their subordinate positions are marked and justified by reference to (female) sex. (Also Hurtado 1994, esp. 142.) Similarly, this account allows that the substantive import of gender varies even from individual to individual within a culture depending on how the meaning of sex interacts with other socially salient characteristics (e.g., race, class, sexuality, etc.). For example, a privileged White woman and a Black woman of the underclass will both be women insofar as their social positions are affected by the social meanings of being female; and yet the social implications of being female vary for each because sexism is intertwined with race and class oppression.

There are points in the proposed analysis that require clarification, however. What does it mean to say that someone is ‘systematically subordinated’ or ‘privileged’, and further, that the subordination occurs ‘on the basis of’ certain features? The background idea is that women are oppressed, and that they are oppressed as women. But we still need to ask: What does it mean to say that women are oppressed, and what does the qualification ‘as women’ add?

Marilyn Frye’s account of oppression with Iris Young’s elaborations provides a valuable starting point (Frye 1983; Young 1990). Although these ideas are commonplace within certain intellectual circles, it is useful to summarize them very briefly here. There are of course unresolved difficulties in working out a satisfactory theory of oppression; I’m afraid I can’t take on that further task here, so I can only invoke the rough outlines of the background view with the hope that an adequate account can at some point be supplied. Nonetheless, oppression in the intended sense is a structural phenomenon that positions certain groups as disadvantaged and others as advantaged or privileged in relation to them. Oppression consists of, ‘an enclosing structure of forces and barriers which tends to the immobilization and reduction of a group or category of people.’ (Frye 1983, 11.) Importantly, such structures, at least as we know them, are not designed and policed by those in power, rather,


...oppression refers to the vast and deep injustices some groups suffer as a consequence of often unconscious assumptions and reactions of well-meaning people in ordinary interactions, media and cultural stereotypes, and structural features of bureaucratic hierarchies and market mechanisms—in short, the normal processes of everyday life. (Young 1990, 41.)


Developing this concept of oppression, Young specifies five forms it can take: exploitation, marginalization, powerlessness, cultural imperialism, and (systematic) violence. The key point for us is that oppression comes in different forms, and even if one is privileged along some dimension (e.g., in income or respect), one might be oppressed in others.[12] In fact, one might be systematically subordinated along some social axis, and yet still be tremendously privileged in one’s overall social position.

It is clear that women are oppressed in the sense that women are members of groups that suffer exploitation, marginalization, etc. But how should we understand the claim that women are oppressed as women. Frye explains this as follows:


One is marked for application of oppressive pressures by one’s membership in some group or category...In the case at hand, it is the category, woman....If a woman has little or no economic or political power, or achieves little of what she wants to achieve, a major causal factor in this is that she is a woman. For any woman of any race or economic class, being a woman is significantly attached to whatever disadvantages and deprivations she suffers, be they great or small....[In contrast,] being male is something [a man] has going for him, even if race or class or age or disability is going against him. (Frye 1983, 15-16.)


But given the diffusion of power in a model of structural oppression how are we to make sense of one’s being ‘marked’ and the ‘application’ of pressures? In the context of oppression, certain properties of individuals are socially meaningful. This is to say that the properties play a role in a broadly accepted (though usually not fully explicit) representation of the world that functions to justify and motivate particular forms of social intercourse. The significant properties in question—in the cases at hand, assumed or actual properties of the body—mark you ‘for application of oppressive pressures’ insofar as the attribution of these properties is interpreted as adequate, in light of this background representation, to explain and/or justify your position in a structure of oppressive social relations. In the case of women, the idea is that societies are guided by representations that link being female with other facts that have implications for how one should be viewed and treated; insofar as we structure our social life to accommodate the cultural meanings of the female (and male) body, females occupy an oppressed social position.

Although I agree with Frye that in sexist societies social institutions are structured in ways that on the whole disadvantage females and advantage males, we must keep in mind that societies are not monolithic and that sexism is not the only source of oppression. For example, in the contemporary US, there are contexts in which being Black and male marks one as a target for certain forms of systematic violence (e.g., by the police). In those contexts, contrary to Frye’s suggestion, being male is not something that a man ‘has going for him’; though there are other contexts (also in the contemporary US) in which Black males benefit from being male. In examples of this sort, the systematic violence against males as males is emasculating (and may be intended as such); but there are important differences between an emasculated man and a woman. On the sort of view we’re considering, a woman is someone whose subordinated status is marked by reference to (assumed) female anatomy; someone marked for subordination by reference to (assumed) male anatomy does not qualify as a woman, but also, in the particular context, is not socially positioned as a man.

These considerations suggests that it may be useful to bring context explicitly into our account. Recent work on gender socialization also supports the idea that although most of us develop a relatively fixed gender identity by the age of three, the degree to which the marked body makes a difference varies from context to context. In her study of elementary school children, Barrie Thorne suggests:


Gender boundaries are episodic and ambiguous, and the notion of ‘borderwork’ [i.e., the work of contesting and policing gender boundaries] should be coupled with a parallel term—such as ‘neutralization’—for processes through which girls and boys (and adults…) neutralize or undermine a sense of gender as division and opposition. (Thorne 1993, 84.)


Thorne’s study is motivated by a recognition that gender is a well-entrenched system of oppression. However, her comments here are intended as an antidote to two problematic tendencies in speaking of girls and boys, men and women: first, the tendency to over-generalize gender differences based on paradigm or stereotyped interactions; second, the tendency to view individuals (specifically children) as passive participants in gender socialization and, more generally, gendered life.

In some respects, Frye’s and Thorne’s approaches appear to be in tension with one another. Frye is keen to highlight the structural facts of sexist oppression: like it or not, your body positions you within a social hierarchy. Thorne, on the other hand, examines how oppression is lived, enforced, and resisted at the micro-level. There are important advantages to both: without a recognition of oppressive structures and the overall patterns of advantage and disadvantage, individual slights or conflicts can seem harmless. But without a recognition of individual variation and agency, the structures take on a life of their own and come to seem inevitable and insurmountable. But can both perspectives be accommodated in an account of gender? The idea seems simple enough: there are dominant ideologies and dominant social structures that work together to bias the micro-level interactions, however varied and complex they may be, so that for the most part males are privileged and females are disadvantaged.

Although an adequate account of gender must be highly sensitive to contextual variation, if we focus entirely on the narrowly defined contexts in which one’s gender is negotiated, we could easily lose sight of the fact that for most of us there is a relatively fixed interpretation of our bodies as sexed either male or female, an interpretation that marks us within the dominant ideology as eligible for only certain positions or opportunities in a system of sexist oppression. Given our priority in theorizing systems of inequality, it is important first to locate the social classes men and women in a broad structure of subordination and privilege[13]:


S is a woman iff


                (i)            S is regularly and for the most part observed or imagined to have certain bodily features presumed to be evidence of a female’s biological role in reproduction;

              (ii)            that S has these features marks S within the dominant ideology of S’s society as someone who ought to occupy certain kinds of social position that are in fact subordinate (and so motivates and justifies S’s occupying such a position); and

            (iii)            the fact that S satisfies (i) and (ii) plays a role in S’s systematic subordination, i.e., along some dimension, S’s social position is oppressive, and S’s satisfying (i) and (ii) plays a role in that dimension of subordination.


S is a man iff


                  (i)            S is regularly and for the most part observed or imagined to have certain bodily features presumed to be evidence of a male’s biological role in reproduction;

                (ii)            that S has these features marks S within the dominant ideology of S’s society as someone who ought to occupy certain kinds of social position that are in fact privileged (and so motivates and justifies S’s occupying such a position); and

              (iii)            the fact that S satisfies (i) and (ii) plays a role in S’s systematic privilege, i.e., along some dimension, S’s social position is privileged, and S’s satisfying (i) and (ii) plays a role in that dimension of privilege.


These accounts are, however, compatible with the idea that (at least for some of us) one’s gender may not be entirely stable, and that other systems of oppression may disrupt gender in particular contexts: a woman may not always function socially as a woman; a man may not always function socially as a man.[14] To return to a previous example, when systems of White supremacy and male dominance collide, a Black man’s male privilege may be seen as so threatening that it must be violently wrested from him. In an effort to accommodate this variation, we can add:


S functions as a woman in context C iffdf


                  (i)            S is observed or imagined in C to have certain bodily features presumed to be evidence of a female’s biological role in reproduction;

                (ii)            that S has these features marks S within the background ideology of C as someone who ought to occupy certain kinds of social position that are in fact subordinate (and so motivates and justifies S’s occupying such a position); and

              (iii)            the fact that S satisfies (i) and (ii) plays a role in S’s systematic subordination in C, i.e., along some dimension, S’s social position in C is oppressive, and S’s satisfying (i) and (ii) plays a role in that dimension of subordination.


And mutatis mutandis for functioning as a man in context C.

 It is important to note that the definitions don’t require that the background ideology in question must use (assumed) reproductive function as itself the justification for treating men or women in the way deemed ‘appropriate’; (assumed) reproductive features may instead simply be ‘markers’ of supposedly ‘deeper’ (and morally relevant?) characteristics that the ideology supposes justifies the treatment in question. (Appiah 1992, 13-15.)

Although ultimately I will defend these analyses of man and woman, I’ll argue below that there are reasons to modify the broader materialist strategy in defining gender. In short, I believe that gender can be fruitfully understood as a higher-order genus that includes not only the hierarchical social positions of man and woman, but potentially other non-hierarchical social positions defined in part by reference to reproductive function. I believe gender as we know it takes hierarchical forms as men and women; but the theoretical move of treating men and women as only two kinds of gender provides resources for thinking about other (actual) genders, and the political possibility of constructing non-hierarchical genders.

4       What Is Race?

One advantage of this account of gender is the parallel it offers for race. To begin, let me review a couple of points that I take to be matters of established fact: First, there are no racial genes responsible for the complex morphologies and cultural patterns we associate with different races. Second, in different contexts racial distinctions are drawn on the basis of different characteristics, e.g., the Brazilian and US classification schemes for who counts as ‘Black’ differ. For these reasons and others, it appears that race, like gender, could be fruitfully understood as a position within a broad social network.

Although suggestive, this idea is not easy to develop. It is one thing to acknowledge that race is socially real, even if a biological fiction; but it is another thing to capture in general terms ‘the social meaning of color’. There seem to be too many different forms race takes. Note, however, that we encountered a similar problem with gender: is there any prospect for a unified analysis of ‘the social meaning of sex’? The materialist feminist approach offered a helpful strategy: don’t look for an analysis that assumes that the meaning is always and everywhere the same; rather, consider how members of the group are socially positioned, and what physical markers serve as a supposed basis for such treatment.

How might we extend this strategy to race? Transposing the slogan, we might say that race is the social meaning of the geographically marked body, familiar markers being skin color, hair type, eye shape, physique. To develop this, I propose the following account.[15]

First definition:


A group is racialized iffdf its members are socially positioned as subordinate or privileged along some dimension (economic, political, legal, social, etc.), and the group is ‘marked’ as a target for this treatment by observed or imagined bodily features presumed to be evidence of ancestral links to a certain geographical region.


Or in the more elaborate version:


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


                  (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);

                (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

              (iii)            whose satisfying (i) and (ii) plays (or would play) a role in their systematic subordination or privilege in C, i.e., 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.[16]

In other words, races are those groups demarcated by the geographical associations accompanying perceived body type, when those associations take on evaluative significance concerning how members of the group should be viewed and treated. As in the case of gender, the ideology need not use physical morphology or geography as the entire basis for ‘appropriate’ treatment; these features may instead simply be ‘markers’ of other characteristics that the ideology uses to justify the treatment in question.

Given this definition, we can say that S is of the White (Black, Asian...) race [in C] iff Whites (Blacks, Asians...) are a racialized group [in C], and S is a member.[17] On this view, whether a group is racialized, and so how and whether an individual is raced, is not an absolute fact, but will depend on context. For example, Blacks, Whites, Asians, Native Americans, are currently racialized in the US insofar as these are all groups defined in terms of physical features associated with places of origin, and insofar as membership in the group functions socially as a basis for evaluation. However, some groups are not currently racialized in the US, but have been so in the past and possibly could be again (and in other contexts are), e.g., the Italians, the Germans, the Irish.

It is useful to note a possible contrast between race and ethnicity. I don’t have a theory of ethnicity to offer; these are some preliminary comparisons. One’s ethnicity concerns one’s ancestral links to a certain geographical region (perhaps together with participation in the cultural practices of that region); often ethnicity is associated with characteristic physical features. For our purposes, however, it might be useful to employ the notion of ‘ethnicity’ for those groups that are like races as I’ve defined them except that they do not experience systematic subordination or privilege in the context in question.[18] Ethnic groups can be (and are) racialized, however, and when they are, one’s membership in the group positions one in a social hierarchy; but (on the view I’m sketching) the occurrence of this hierarchical positioning means that the group has gone beyond simply being an ethnic group and functions in that context as a race. In short, we can distinguish between grouping individuals on the basis of their (assumed) origins, and grouping them hierarchically on the basis of their (assumed) origins, and the contrast between race and ethnicity might be a useful way to capture this distinction.

5       Normativity and Commonality

So what, if anything, is achieved by adopting the above analyses? Are they the tools we need? Let’s first consider the problems of commonality and normativity, and begin with gender.

Remember, the problem of commonality questions whether there is anything social that all females can plausibly be said to have in common. If we ask whether females share any intrinsic (non-anatomical) features such as psychological make-up, character traits, beliefs, values, experiences or, alternatively, whether there is a particular social role that all females have occupied across culture and history, the answer seems to be ‘no’.

On my analysis women are those who occupy a particular kind of social position, viz., one of sexually-marked subordinate. So women have in common that their (assumed) sex has socially disadvantaged them; but this is compatible with the kinds of cultural variation that feminist inquiry has revealed, for the substantive content of women’s position and the ways of justifying it can vary enormously. Admittedly, the account accommodates such variation by being very abstract; nonetheless, it provides a schematic account that highlights the interdependence between the material forces that subordinate women, and the ideological frameworks that sustain them.

One might complain, however, that there must be some women (or rather, females) who aren’t oppressed, and in particular, aren’t oppressed as women. Perhaps there are; e.g., some may ‘pass’ as men, others may be recognizably female but not be subordinated in any way linked to that recognition. I’m not convinced that there are many cases (if any) of the latter, but I’ll certainly grant that there could be females who did not satisfy the definition that I’ve offered. In fact, I believe it is part of the project of feminism to bring about a day when there are no more women (though, of course, we should not aim to do away with females!). I’m happy to admit that there could be females who aren’t women in the sense I’ve defined, but these individuals (or possible individuals) are not counterexamples to the analysis. The analysis is intended to capture a meaningful political category for critical feminist efforts, and non-oppressed females do not fall within that category (though they may be interesting for other reasons).

But this leads us directly from the commonality problem to the normativity problem. The normativity problem raises the challenge that any effort to define women will problematically privilege some women and (theoretically) marginalize others, and will itself become normative. One worry is that bias inevitably occurs in deciding which experiences or social roles are definitive; a second worry is that if someone wants to be a ‘real’ woman, she should conform to the definition of women provided, and this will reinforce rather than challenge male dominance.

On the account I’ve offered, it is true that certain females don’t count as ‘real’ women; and it is true that I’ve privileged certain facts of women’s lives as definitive. But given the epistemological framework outlined above, it is both inevitable and important for us to choose what facts are significant on the basis of explicit and considered values. For the purposes of a critical feminist inquiry, oppression is a significant fact around which we should organize our theoretical categories; it may be that non-oppressed females are marginalized within my account, but that is because for the broader purposes at hand—relative to the feminist and antiracist values guiding our project—they are not the ones who matter. The important issue is not whether a particular account ‘marginalizes’ some individuals, but whether its doing so is in conflict with the feminist values that motivate the inquiry. And as far as I can tell, not focusing our theoretical efforts on understanding the position of oppressed females would pose just such a conflict.

The question remains whether my definition of woman helps sustain gender hierarchy by implicitly offering a normative ideal of woman. Given that women on my definition are an oppressed group, I certainly hope not! Instead, the definition is more likely to offer a negative ideal that challenges male dominance.

I won’t defend here my account of racialized groups against an extension of the normativity and commonality complaints, for I would simply repeat the strategy just employed. Although there are interesting nuances in adapting the arguments to apply to racialized groups, I don’t see anything peculiar to race that would present an obstacle to developing the same sort of response.

6       Negotiating Terms

Let me now turn to summarize some of the advantages of the proposed definitions. At this point we could bracket the terminological issues and just consider whether the groups in question are ones that are important to consider given the goals of our inquiry. I hope it is clear from what I’ve already said how the analyses can help us identify and critique broad patterns of racial and sexual oppression (MacKinnon 1987), and how they accommodate the intersectionality of social categories. But a further and, I think, more interesting question is whether it is useful to think of these groups in these terms: Does it serve both the goal of understanding racial and sexual oppression, and of achieving sexual and racial equality to think of ourselves as men or women, or raced in the ways proposed?

By appropriating the everyday terminology of race and gender, the analyses I’ve offered invite us to acknowledge the force of oppressive systems in framing our personal and political identities. Each of us has some investment in our race and gender: I am a White woman. On my accounts, this claim locates me within social systems that in some respects privilege and in others subordinate me. Because gender and racial inequality are not simply a matter of public policy but implicate each of us at the heart of our self-understandings, the terminological shift calls us to reconsider who we think we are.

This point highlights why the issue of terminological appropriation is especially sensitive when the terms designate categories of social identity. Writing in response to a NY Times editorial supporting the terminological shift from ‘Black’ to ‘African-American,’ Trey Ellis responded:


When somebody tries to tell me what to call myself in all its uses just because they come to some decision at a cocktail party to which I wasn’t even invited, my mama raised me to tell them to kiss my black ass. In many cases, African-American just won’t do.[19]


The issue is not just what words we should use, and who gets to say what words to use, but who we take ourselves to be, and so, in some sense, who we are. Terms for social groups can function as descriptive terms: it may be accurate to say that someone is a woman when she satisfies certain conditions. However, terms for social groups serve other rhetorical purposes. Typically the act of classifying someone as a member of a social group invokes a set of ‘appropriate’ (contextually specific) norms and expectations. It positions her in a social framework and makes available certain kinds of evaluation; in short, it carries prescriptive force. Accepting or identifying with the classification typically involves an endorsement of some norms and expectations, however, not always the socially sanctioned ones. The question whether I should be called a ‘woman’ or a ‘wommon’, ‘White’ or ‘Euro-American’, is not just a matter of what words to use, but what norms and expectations are taken to be appropriate; to ask what I should be called is to ask what norms I should be judged by. (Haslanger 1993, esp. 89-91).

Although ‘identifying’ someone as a member of a social group invokes a set of ‘appropriate’ norms, what these norms are is not fixed. What it means to be a woman, or to be White, or to be Latino, in this sense, is unstable and always open to contest. The instability across time is necessary to maintain the basic structure of gender and race relations through other social changes: as social roles change—prompted by the economy, immigration, political movements, natural disasters, war, etc.—the contents of normative race and gender identities adjust. The flexibility across contexts accommodates the complexity of social life: what norms are assumed to apply depends on the dominant social structure, the ideological context, and other dimensions of one’s identity (such as class, age, ability, sexuality). But this instability and flexibility is exactly what opens the door for groups to redefine themselves in new ways. One strategy is for the group to adopt new names (‘African-American’, ‘womyn’); another is to appropriate old names with a normative twist (‘queer’); but in some cases the contest is over the meanings of the standard terms (‘Ain’t I a woman?’). Because individuals are so deeply invested in gender and, at least in the US, race categories, it remains of crucial importance to be and to be perceived as a ‘woman’ or a ‘man’ and as a member of one of the standard races. But even so, (although this is something of an exaggeration) it is possible to view our gender and race vocabulary as, in effect, providing terminological place-holders marking space for the collective negotiation of our social identities.

Given the normative force and political potential of identifying someone (or self-identifying) in racial or gendered terms, how do we evaluate a terminological appropriation of the kind I’m proposing? For example, isn’t there something disingenuous about appropriating race and gender terminology because it is used to frame how we think of ourselves and each other in order to use them for new concepts that are not part of our self-understandings?

This latter question is especially pressing because the appropriation under consideration intentionally invokes what many find to be positive self-understandings—being Latina, being a White man—and offers analyses of them which emphasize the broader context of injustice. Thus there is an invitation not only to revise one’s understanding of these categories (given their instability, this happens often enough), but to revise one’s relationship to their prescriptive force. By offering these analyses of our ordinary terms, I call upon us to reject what seemed to be positive social identities. I’m suggesting that we should work to undermine those forces that make being a man, a woman, or a member of a racialized group possible; we should refuse to be gendered man or woman, refuse to be raced. This goes beyond denying essentialist claims about one’s embodiment and involves an active political commitment to live one’s life differently (Stoltenberg 1989). In one sense this appropriation is ‘just semantics’: I’m asking us to use an old term in a new way. But it is also politics: I’m asking us to understand ourselves and those around us as deeply molded by injustice and to draw the appropriate prescriptive inference. This, I hope, will contribute to empowering critical social agents. However, whether the terminological shift I’m suggesting is politically useful will depend on the contexts in which it is employed and the individuals employing it. The point is not to legislate what terms to use in all contexts, but to offer resources that should be used judiciously.

7       Lingering Concerns, Promising Alternatives

There is, nonetheless, a broader concern one might have about the strategy I’ve employed: Why build hierarchy into the definitions? Why not define gender and race as those social positions motivated and justified by cultural responses to the body, without requiring that the social positions are hierarchical? Wouldn’t that provide what we need without implying (implausibly) that women are, by definition, subordinate, men, by definition, privileged, and races, by definition, hierarchically positioned?

If we were to remove hierarchy from the definitions, then there would be two other benefits: first, by providing a place in our model for cultural representations of the body besides those that contribute to maintaining subordination and privilege, we could better acknowledge that there are positive aspects to having a gender and a race. And second, the accounts would provide a framework for envisioning the sorts of constructive changes needed to create a more just world. The suggestion that we must eliminate race and gender may be a powerful rallying call to those who identify with radical causes, but it is not at all clear that societies can or should avoid giving meanings to the body, or organizing themselves to take sexual and reproductive differences into account. Don’t we at least need a concept of gender that will be useful in the reconstructive effort, not only the destructive one?

Consider gender. I am sympathetic to radical rethinkings of sex and gender. In particular, I believe that we should refuse to use anatomy as a primary basis for classifying individuals and that any distinctions between kinds of sexual and reproductive bodies are importantly political and open to contest. Some authors have argued that we should acknowledge the continuum of anatomical differences and recognize at least five sexes (Fausto-Sterling 1993). And if sexual distinctions become more complex, we would also need to rethink sexuality, given that sexual desire would not fit neatly within existing homosexual/heterosexual paradigms.

However, one can encourage the proliferation of sexual and reproductive options without maintaining that we can or should eliminate all social implications of anatomical sex and reproduction. Given that as a species there are substantial differences in what human bodies contribute to reproduction, and what sorts of bodies bear the main physical burdens of reproduction, and given further that reproduction cannot really help but be a socially significant fact (it does, after all, produce children), it can seem difficult to imagine a functioning society, more specifically, a functioning feminist society, that doesn’t acknowledge in some way the difference between those kinds of bodies that are likely able to bear children, and those that aren’t. One could argue that we should work towards a society free of gender in a materialist sense—one in which sex-oppression does not exist—while still allowing that sexual and reproductive differences should be taken into account in a just society. (Frye 1996; Gatens 1996.)

 I will not debate here the degree to which a just society must be attentive to sexual and reproductive differences. Whether we, as feminists, ought to recommend the construction of (new) non-hierarchical genders or work to abolish gender entirely is a normative issue I leave for another occasion. Nonetheless, at the very least it would help to have terminology to debate these issues. I propose that we use the definitions of man and woman offered above: it is clear that these dominant nodes of our current gender structures are hierarchical. But borrowing strategies employed before, we can define gender in generic terms under which the previous definitions of man and women fall,[20] thus allowing the possibility of non-hierarchical genders and breaking the binary opposition between man and woman.


A group G is a gender relative to context C iffdf members of G are (all and only) those:


                  (i)            who are regularly observed or imagined to have certain bodily features presumed in C to be evidence of their reproductive capacities[21];

                (ii)            whose having (or being imagined to have) these features marks them within the context of the ideology in C as motivating and justifying some aspect(s) of their social position; and

              (iii)            whose satisfying (i) and (ii) plays (or would play) a role in C in their social position’s having one or another of these designated aspects.


I offer this analysis as a way of capturing the standard slogan: gender is the social meaning of sex. Note, however, that in imagining ‘alternative’ genders we should be careful not to take for granted that the relevant biological divisions will correspond to what we consider ‘sex’.[22] (Alternative groupings could include: ‘pregnant persons,’ ‘lactating persons,’ ‘menstruating persons,’ ‘infertile persons,’ (perhaps ‘homosexuals,’ depending on the story given about physical causes)). Neither should we assume that membership in a gender will constitute one’s personal or psychological identity to any significant degree. Recall that on the accounts of gender and race I am proposing, both are to be understood first and foremost as social groups defined within a structure of social relations; whatever links there might be to identities and norms are highly contingent and would depend on the details of the picture. For example, we might imagine that ‘after the revolution’ gender is a component of one’s overall social position because, for example, there are legal protections or medical entitlements granted to individuals classified as having a certain sort of ‘sexed’ body; but this need not have broad implications for psychological identity or everyday social interactions, for the ‘sex’ of bodies might not even be publicly marked.

Turning briefly to race, the parallel issue arises: Do we need a concept of non-hierarchical ‘races’ in order to frame and debate different visions of a ‘racially’ just society? It would seem that we have the terminological resources available without a further definition: let races be, as previously defined, those hierarchically organized groups that are defined (roughly) by physical features and (assumed) geographical origins, and call those that aren’t hierarchically organized (in the context in question) ‘ethnicities’. Admittedly, ethnicity as we know it does have implications for social status and power, so my proposal is to employ the term for a somewhat idealized conception.

As in the case of gender, the question arises whether it ought to be part of an anti-racist project to recommend the preservation of existing ethnic groups or the formation of ‘new’ ethnicities. And more generally, we need to ask whether a feminist anti-racism should treat genders and ethno-racial groups in the same way over the long term. Should we seek, e.g., to eliminate all genders and ethno-racial groupings; to preserve and proliferate them; to eliminate gender but not ethnicity (or vice versa)? These questions deserve careful attention but I cannot address them here.

Because the structure of definitions has become quite complex, it may help at this point to provide a diagram:





hierarchical genders                              non-hierarchical genders



men      women                                [other (new) genders]


Ethno-racial groups: defined (contextually) by

(assumed) physical features, ancestry, and geographical origins



hierarchical ethno-racial groups  non-hierarchical ethno-racial groups



     [races]                              [(idealized) ethnicities]


8       Conclusion

On the accounts I’ve offered, there are striking parallels between race and gender. Both gender and race are real, and both are social categories. Neither gender nor race is chosen, but the forms they take can be resisted or mutated. Both race and gender (as we know it) are hierarchical, but the systems that sustain the hierarchy are contingent. And although the ideologies of race and gender and the hierarchical structures they sustain are substantively very different, they are intertwined.

There are many different types of human bodies; it is not the case that there is a unique ‘right’ way of classifying them, though certain classifications will be more useful for some purposes than others. How we classify bodies can and does matter politically, for our laws, social institutions, and personal identities are profoundly linked to understandings of the body and its possibilities. This is compatible with the idea that what possibilities a human body has is not wholly a function of our understandings of it. Our bodies often outdo us, and undo us, in spite of the meanings we give them.

Within the framework I’ve sketched, there is room for theoretical categories such as man, woman, and race (and particular racial groups), that take hierarchy to be a constitutive element, and those such as gender and ethnicity that do not. As I have suggested before, I am willing to grant that there are other ways to define race or gender, man or woman, that are useful to answer different questions, motivated by different concerns and priorities. I’m sure we need several concepts to do all the work needed in understanding the complex systems of racial and gender subordination.

In short, (speaking of my analyses) I’m less committed to saying that this is what gender is and what race is, than to saying that these are important categories that a feminist antiracist theory needs. As I’ve explained above, I think there are rhetorical advantages to using the terms ‘gender’, ‘man’ and ‘woman,’ and ‘race’ for the concepts I’ve defined, but if someone else is determined to have those terms, I’ll use different ones. To return to the point made much earlier in characterizing analytic projects: it is our responsibility to define gender and race for our theoretical purposes. The world itself can’t tell us what gender is. The same is true for race. It may be as Appiah claims that ‘there is nothing in the world that can do all we ask race to do for us’ (Appiah 1992, 45), if our project inevitably inherits the concept’s complex history; but we might instead ask ‘race’ to do different things than have been asked before. Of course, in defining our terms, we must keep clearly in mind our political aims both in analyzing the past and present, and in envisioning alternative futures. But rather than worrying, ‘what is gender, really?’ or ‘what is race, really?’ I think we should begin by asking (both in the theoretical and political sense) what, if anything, we want them to be.


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

Appiah, K. Anthony and Amy Gutmann. (1996) Color Conscious. Princeton, NJ: Princeton University Press.

Appiah, K. Anthony. (1992) In My Father’s House. New York: Oxford University Press.

Butler, Judith. (1990) Gender Trouble. New York: Routledge.

Crenshaw, Kimberlé. (1993) Beyond Racism and Misogyny: Black Feminism and 2 Live Crew. Words that Wound, ed., M. Matsuda, C. Lawrence, R. Delgado, and K. Crenshaw. Boulder, CO: Westview, 111-132.

Fausto-Sterling, Anne. (1993) The Five Sexes: Why Male and Female Are Not Enough. The Sciences 33:2: 20-24.

Frye, Marilyn. (1996) The Necessity of Differences: Constructing a Positive Category of Women. Signs 21:4: 991-1010.

Frye, Marylin. (1983) The Politics of Reality. Freedom, CA: Crossing Press.

Gatens, Moira. (1996) A Critique of the Sex-Gender Distinction. Imaginary Bodies. New York: Routledge, 3-20.

Gates, Jr., Henry Louis. (1992) Loose Canons. New York: Oxford University Press.

Geuss, Raymond. (1981) The Idea of a Critical Theory. Cambridge: Cambridge University Press.

Haslanger, Sally. (1993) On Being Objective and Being Objectified. A Mind of One’s Own , ed., L. Antony and C. Witt. Boulder: Westview, 85-125.

Haslanger, Sally. (1995) Ontology and Social Construction. Philosophical Topics 23:2: 95-125.

Hennessy, Rosemary and Chrys Ingraham, eds. (1997) Materialist Feminism. New York: Routledge.

Hoy, David. (1994) Deconstructing Ideology. Philosophy and Literature 18:1.

Hurtado, Aída. (1994) Relating to Privilege: Seduction and Rejection in the Subordination of White Women and Women of Color. Theorizing Feminism, ed., Anne Hermann and Abigail Stewart. Boulder, CO: Westview Press, 136-154.

MacKinnon, Catharine. (1987) Difference and Dominance: On Sex Discrimination. Feminism Unmodified. Cambridge, MA: Harvard University Press, 32-45.

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

Ortner, Sherry. (1996) Gender Hegemonies. Making Gender. Boston: Beacon Press, 139-172.

Riley, Denise. (1988) Am I That Name? Minneapolis: University of Minnesota Press.

Roscoe, Will. (1996) How to Become a Berdache: Toward a Unified Analysis of Gender Diversity. Third Sex, Third Gender, ed., Gilbert Herdt. New York: Zone Books, 329-372.

Scott, Joan. (1986) Gender: A Useful Category of Historical Analysis. American Historical Review 91:5:1053-75.

Spelman, Elizabeth. (1988) The Inessential Woman. Boston: Beacon Press.

Stevens, Jacqueline. (forthcoming) Reproducing the State. Princeton: Princeton University Press.

Stich, Stephen. (1988) Reflective Equilibrium. Analytic Epistemology, and the Problem of Cognitive Diversity, Synthese 74: 391-413.

Stoltenberg, John. (1989) Refusing To Be a Man. New York: Meridian Books.

Thorne, Barrie. (1993) Gender Play. New Brunswick, NJ: Rutgers University Press.

Young, Iris. (1990) Justice and the Politics of Difference. Princeton: Princeton University Press.


[1]I use the terms ‘man’ and ‘woman’ to distinguish individuals on the basis of gender, the terms ‘male’ and ‘female’ to distinguish individuals on the basis of sex.

[2]See Stich 1998. Stich uses the term ‘analytical epistemology’ for what I would call a ‘conceptual’ rather than an ‘analytical’ project.

[3]Cf. Appiah and Gutmann 1996, pp. 30-105. Appiah doesn’t consider an analytical approach to race except rather elliptically on p. 42.

[4]On the distinction between manifest and operative concepts, see my 1995, esp. p. 102.

[5]It is important to keep in mind that what’s at issue is not a criterion for sameness of meaning, but the boundary between what could count as a revisionary project and a new project altogether.

[6]We need here a term for those physical features of individuals that mark them as members of a race. One might refer to them as ‘racial’ features, but to avoid any suggestion of racial essences I will use the term ‘color’ to refer to the (contextually variable) physical ‘markers’ of race, just as I use the term ‘sex’ to refer to the (contextually variable) physical ‘markers’ of gender. I mean to include in ‘color’ more than just skin tone: common markers also include eye, nose, and lip shape, hair texture, physique, etc., and it is presumed of the markers of race that the features are inherited through an ancestry that can be traced to a particular geogrpahical region. Although the term ‘people of color’ is used to refer to non-Whites, I want to allow that the markers of ‘Whiteness’ count as ‘color’.

[7]There are at least four different uses of the term ‘identity’ that are relevant in considering the issue of gender or racial ‘identity’; here my comments about ‘gender identity’ are admittedly superficial.

[8] Very roughly, feminine norms are those that enable one to excel in the social position constituting the class women; feminine gender identity (at least in one sense of the term) is a psychological orientation to the world that includes the internalization of feminine norms; and feminine symbols are those that encode idealized feminine norms. What counts as a ‘feminine’ norm, a ‘feminine’ gender identity, or a ‘feminine’ symbol is derivative (norms, symbols, and identities are not intrinsically feminine or masculine), and depends on how the social class of women is locally constituted.

[9]For a sample of materialist feminist work, see Hennessy and Ingraham 1997.

[10] Some theorists (Delphy, Hartmann) focus on the economic exploitation of women in domestic relations of production; others (Wittig) focus on sexual and reproductive exploitation under compulsory heterosexuality; others (MacKinnon) focus on sexual objectification.

[11]These analyses allow that there isn’t a common understanding of ‘sex’ across time and place. On my account, gendered social positions are those marked by reference to features that are generally assumed in the context in question to either explain or provide evidence of reproductive role, whether or not these are features that we consider ‘sex’.

[12]On the importance of disaggregating power and oppression, see Ortner 1996.

[13]This proposal depends on the claim that at least some societies have a ‘dominant ideology’. Others have employed the notions of ‘background,’ ‘hegemony,’ ‘habitus,’ for the same purpose. Rather than debating what is the preferred notion, I’m happy to let the term ‘dominant ideology’ serve as a placeholder for an account to be decided upon later. Given the strategy of my accounts, however, we must be sure to allow for multiple ideological strands in any society. See Geuss 1981, Hoy 1994.

[14]We noted before that on a materialist account sex and gender don’t always coincide. I’m making here a further claim: one may be gendered man or woman without functioning socially in that gender every moment of one’s life.

[15]On this I am deeply indebted to Stevens forthcoming, Ch. 4, and Omi and Winant 1994, esp. pp. 53-61.

[16]There are aspects of this definition that need further elaboration or qualification. I will mention four here.

First, on my account, those who actually have the ancestral links to the specified region but who ‘pass’, do not count as members of the racialized group in question. This is parallel to the case of a female functioning socially as a man or a male functioning socially as a woman. Because the goal is to define race and gender as social positions, I endorse this consequence of the definitions.

Second, as it stands the definition does not accommodate contexts such as Brazil in which membership in ‘racial’ groups is partly a function of education and class. It excludes privileged (‘Whitened’) members from the subordinate races they might seem—considering only ‘color’—to belong to, and subordinated (‘darkened’) members from privileged races, because they don’t satisfy the third condition. But it cannot handle the inclusion of the ‘Whitened’ members in the privileged group or the ‘darkened’ members in the subordinated group because they don’t satisfy the first condition. However, we could take the definition to capture a strong version of racialization, and develop another version on which appropriate ‘color’ is relevant but not necessary by modifying the second condition:


ii*) having (or being imagined to have) these features—possibly in combination with others—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 depend on a variety of factors.

Third, need racialized groups be ‘marked’ by actual or assumed body type? What about Jews, Native Americans, and Romanies? (Romanies are also interesting because it isn’t entirely clear that there is a supposed place of origin, though I take ‘no place of origin’ to be a factor in their racialization, and to serve as the limit case.) I would suggest that that there are some (perhaps imagined) physical features that are regarded as salient in contexts where Jews and Native Americans are racialized, though not every member of the group need have those features if there is other evidence of ancestral links. However, ultimately it might be more useful to allow racial membership to be determined by a cluster of features (such as physical appearance, ancestry, and class) weighted differently in different contexts.

Finally, 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. 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.

[17]As in the case of gender, I recommend that we view membership in a racial/ethnic group in terms of how one is viewed and treated regularly and for the most part in the context in question; though as before, one could distinguish being a member of a given race from functioning as one by considering the degree of one’s entrenchment in the racialized social position (not on the basis of biology or ancestry).

[18] We may want to allow there to be kinds of social stratification between ethnic groups that falls short of the kind of systematic subordination constitutive of race. My account remains vague on this point. Clarification might be gained by plugging in a more sophisticated account of social hierarchies. The body is also relevant: are ethnicities distinguishable from races by the degree to which they are perceived as capable of assimilation?

[19]Trey Ellis, Village Voice, June 13, 1989; quoted in H. L. Gates 1992, ‘What’s In a Name?’, p. 139. Gates quotes the passage differently, leaving out ‘black’ before ‘ass’. Although he adds Ellis’s conclusion, he robs the quote of its self-exemplifying power by the alteration.

[20]Thanks to Geoff Sayre-McCord for suggesting this approach.

[21]It is important here that the ‘observations’ or ‘imaginings’ in question not be idiosyncratic but part of a broader pattern of social perception; however, they need not occur, as in the case of man and woman, ‘for the most part’. They may even be both regular and rare.

[22]I leave it an open question whether groups that have been identified as ‘third genders’ count as genders on my account. Some accounts of gender that purport to include third genders pay inadequate attention to the body, so cannot distinguish, e.g., race from gender. See, e.g., Roscoe 1996.