NOTE-394: Haslanger — Gender and Race: (What) Are They?

Contribution

The paper does two things. It names a method: besides asking what our concept of gender is (conceptual) or what kind our gender talk tracks (descriptive), one can ask what concept a theory needs and define the term to fit (analytical). And it applies the method to give the first unified definitions of man, woman and racialized group as social positions in a hierarchy, with a non-hierarchical genus (gender, ethnicity) defined above them so that a just society’s categories can be described in the same terms.

Key insight

Gender is to sex as race is to “color”: the social meaning of a body as it is read. What all women share is not a body, a psychology or a role but a kind of position: being subordinated because one is taken to be female. Defining the category that way answers the objection that women share nothing, and it turns the objection that any definition is normative into a statement of the definition’s purpose.

Assumptions

• The analytical method is legitimate. “It is up to us to decide what in the world, if anything, they are” (p. 34). Two conditions govern appropriating an ordinary term: a semantic one, that central functions of the term remain (p. 35, n. 6), and a political one that the paper does not state in general.
• A critical framework. Four concerns fix the purpose: explaining persistent inequality, sensitivity to similarity and difference (intersectionality), tracking gender and race in phenomena not obviously about them, and taking the agency of women and people of color seriously (p. 36). The paper addresses the first two.
• Oppression as structural. After Frye and Young: “an enclosing structure of forces and barriers”, sustained by “the normal processes of everyday life” (pp. 39–40). The paper says a satisfactory theory of oppression is not supplied.
• A dominant ideology exists, as a placeholder for “background”, “hegemony” or “habitus” (n. 14).
• “Matters of established fact” about race (p. 43): there are no racial genes for the morphologies and cultural patterns associated with races, and racial classification varies by context (Brazil and the US). Asserted, without citation.

Key results

Definitions, with the arguments for adopting them.

• Three projects (pp. 32–34). Conceptual (reflective equilibrium on our concept), descriptive (the kind our usage tracks, possibly social), analytical (what concept serves our legitimate purposes). The paper’s is analytical, linked to the descriptive claim that gender and race talk tracks social kinds “typically obscured by the manifest content” of everyday concepts (p. 34).
• Woman and man, tentative (p. 39). “S is a woman iffdf S is systematically subordinated along some dimension … 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.” Man is the privileged counterpart.
• Woman and man, refined (p. 42). Three conditions: (i) S is regularly and for the most part observed or imagined to have such features; (ii) having them marks S within the dominant ideology as someone who ought to occupy positions that are in fact subordinate; (iii) S’s satisfying (i) and (ii) plays a role in S’s systematic subordination along some dimension. Plus functions as a woman in context C (pp. 42–43), so that, for example, a Black man targeted for violence as male may in that context function as neither (p. 41).
• Racialized group (p. 44). A group G is racialized relative to C iff its members are all and only those observed or imagined to have bodily features presumed in C to show ancestral links to a region, marked by the ideology in C for positions that are in fact subordinate or privileged, and subordinated or privileged in part because of it. Italians, Germans and the Irish have been racialized in the US and are not now (p. 45). Note 17 lists four qualifications (passing, Brazil, groups not marked by body type, scattered members).
• Ethnicity (pp. 45, 50–51) is the non-hierarchical counterpart of race, an “idealized conception”.
• Commonality answered (p. 45). Women share a kind of position, sexually marked subordination, whose content varies. Non-oppressed females, if any, are not counterexamples: the category is political by design (pp. 45–46). “It is part of the project of feminism to bring about a day when there are no more women” (p. 46).
• Normativity accepted (p. 46). Choosing which facts are significant on explicit values is “both inevitable and important”; the definition is more likely to offer “a negative ideal”.
• Terminology as politics (pp. 46–48). Group terms carry prescriptive force; to ask what one should be called is to ask what norms one should be judged by. The paper asks readers to “refuse to be gendered man or woman, refuse to be raced” (p. 48), and says whether the shift is useful depends on context.
• Gender as genus (p. 50). A group is a gender relative to C iff its members are regularly observed or imagined to have features presumed in C to be evidence of reproductive capacities, these mark them in C’s ideology as motivating some aspect of their social position, and that plays a role in their having it. Man and woman are its hierarchical species; others could be non-hierarchical (p. 50; diagram, p. 51).
• Conclusion (pp. 51–52). Gender and race are real and social, not chosen, hierarchical as we know them, and contingent. “The world itself can’t tell us what gender is” (p. 52).

Claims

id	claim	strength	support
C1	”What is gender/race?” admits conceptual, descriptive and analytical readings, and the analytical one is legitimate	moderate	methodological argument, pp. 32–35; the boundary between revising a concept and changing the subject is admitted unclear (p. 34)
C2	Appropriating an ordinary term is warranted when its central functions remain and the politics are acceptable	weak	semantic condition stated (p. 35); political condition left general
C3	Woman (man) is best defined as sexually marked subordination (privilege)	moderate as a proposal; not a discovery	definitions pp. 39, 42; justified by the four concerns (p. 36)
C4	A racialized group is one marked by presumed ancestral-geographic bodily features for subordination or privilege	moderate as a proposal	definition p. 44; qualifications n. 17
C5	There are no racial genes, and racial classification varies by context	assertion here	”matters of established fact”, p. 43, uncited
C6	The commonality problem is answered: women share a kind of position	moderate	p. 45; rests on the definition chosen
C7	The normativity problem is not an objection to a critical project	moderate	p. 46; values-first epistemology of §II
C8	Gender and ethnicity, as non-hierarchical genera, let the theory describe a just society’s categories	moderate as a proposal	p. 50; whether to build new genders or abolish gender is left open
C9	Gender and race are real and social	follows from the definitions	pp. 51–52

Concepts

• Analytical project — defining a concept by “what work we want these concepts to do for us” (pp. 33–34); later literature calls this ameliorative.
• Manifest and operative concepts — the concept users think they apply and the one their practice tracks (n. 5, citing Haslanger 1995).
• “Color” — the contextually variable physical markers of race, by analogy with “sex” for gender, and including the markers of Whiteness (n. 7).
• Marked — taken, within the dominant ideology, to have features that make one’s position appropriate (pp. 40, 42).
• Functioning as a woman in C — satisfying the conditions in a context, which can come apart from being a woman (pp. 42–43, n. 15).
• Focal analysis — explaining norms, symbols and identities as gendered by their relation to the social class (p. 37, n. 9).

Connections

• Ritchie 2018 (LIT-486, NOTE-389). Quotes the tentative definition (p. 39) and principle (iii) (p. 38) accurately, and reads both as identifying women with a node in a social structure of subordination relations, with the marker as an extra requirement on occupiers. On her terms the groups defined here are feature social groups, not organised wholes. Ritchie’s n. 30 says her framework, unlike this paper’s view, can accommodate self-identification as a marker, citing Jenkins’s charge that this account is trans-exclusionary; that charge is not examined here.
• Ásta (LIT-491). Thanked in n. 1 as Ásta Sveinsdóttir. Her conferralism places gender in conferred constraints and enablements, a rival on the same ground; the book is filed unread.
• Durkheim (LIT-484). Not cited. Durkheim’s social fact is a way of acting external to individuals and constraining them; this paper’s social positions are a way of being placed. Both locate the reality of the category in social relations rather than in individuals’ properties.
• The group-agency and group-mind works (LIT-391, LIT-401, LIT-159, THEORY-043). No connection. Nothing here treats a gender or a race as an agent or a subject.

Bearing on the record

• No THEORY document cites or needs it. It does not bear on THEORY-043.
• It is the first held statement of social construction as a claim about what a category is, the topic blurb’s “social kinds” and “social construction”.
• No ML instruction. Nothing belongs in the anthology.

Note of 2026-10-16: THEORY-667 now cites LIT-480 as a source.

Limitations

• The definitions are stipulated for a purpose and defended by that purpose; the paper does not claim they capture ordinary meaning (p. 52).
• The theory of oppression and the notion of dominant ideology are placeholders (pp. 39–40; n. 14).
• The empirical claims about race are uncited (p. 43).
• The definitions are intersectional only in that the content of a position can vary with race and class (p. 39); there is no account of intersectional categories as categories.
• Ethnicity is left vague (n. 19).

Open questions

• What is the political condition on appropriating a term, stated generally (p. 35)?
• Should a feminist antiracism construct non-hierarchical genders and ethnicities or abolish them (pp. 50–51)? Left for “another occasion”.

Corrections

• No dossier existed, so there is no seeded skim to correct. The request’s citation is right in every detail.
• Date. Crossref gives a full date only for the Wiley online record, 17 December 2002. The printed issue is March 2000, which is the first appearance.
• Slips in the paper’s references. Note 3 cites “Stich 1998”; the reference list has Stich 1988 (Synthese 74). The text cites “Guess 1981” (p. 36) for the reference list’s “Geuss, Raymond (1981)”. The reference list spells Frye’s first name “Marylin” for 1983 and “Marilyn” for 1996.