The incomplete materialism of French materialist feminism

The incomplete materialism of French materialist feminism

Alison stone

According to one important and influential line of feminist interrogation of the category of sex, we only believe that there are two biological sexes because our thought and perception are constrained by the two-gender social system under which we currently live. [1] The French materialist feminists – Christine Delphy, Monique Wittig, Colette Guillaumin and Nicole-Claude Mathieu, among others – are among the earliest and best-known exponents of this line. In this article I will take issue with their position on sex, by way of an initial reconstruction of the history of the English-speaking feminist reception of French materialist feminism. I will use this reconstruction to bring out two key elements of French materialist feminism: (1) its proposal that gender can and should be abolished; (2) its – related – denial that sex division is a biological reality. I will then suggest that this latter denial damages the claim of French materialist feminism to be materialist, and that – contrary to the French materialists’ claims – it is possible to affirm the biological reality of sex division and still pursue the abolition of gender. This is possible, I will suggest, if we adopt a cluster-based understanding of sex; some strengths and potential limitations of this cluster-based understanding will be considered in conclusion.

The specificity of french materialist feminism

Critics of the category ‘French feminist thought’ that emerged in the 1980s have observed that it is an eminently Anglo-American construction, in which the so-called ‘holy trinity’ of Irigaray, Kristeva and Cixous rank as the canonical figures. [2] Among those excluded from this construction are the French materialist feminists. Toril Moi has suggested that, ironically, ‘these [materialist] feminists have become less frequently translated and less well-known [than Irigaray et al.] precisely because of their relative similarity [to Anglophone feminism]: they have … been perceived as lacking in exotic difference’. [3] The French materialist feminists were perceived to be ‘relatively similar’ in two particular ways. First, like many Englishspeaking socialist feminists of the 1970s, their account of women’s subordination focused on the exploitation of women’s labour within the home.

Second, and more relevantly here, the French materialist feminists made use of the concept of gender, a concept that was also central to anglophone feminism of the 1970s and 1980s. The French materialists insisted that women’s subordination was caused by social arrangements and not biology, and, being social, could be removed. Thus it appeared that the French materialists adhered to the same sex/gender distinction that English-speaking feminists did, with both groups (apparently) holding that there are biological sex differences between males and females but that these do not cause the gender division between men and women, which is social in origin.

Actually, though, the fact that the French materialists used the concept of gender obscured the fact that they understood their conception of gender to differ from – and to radicalize – the prevailing anglophone conception. Delphy claims that most feminists who use the concept of gender accept that because there are two sexes, there must be two genders, which means that these feminists can only aim to redefine the genders non-hierarchically but not to abolish gender altogether. [4] In contrast, Delphy maintains that the gender division is necessarily hierarchical so that feminists must aim to abolish gender, and hence need to show that this division is entirely independent of, not necessitated by, biology (otherwise, we may assume, the division could not be abolished any more than biology can).

Because Delphy (and other French materialists) reconceive gender division as completely independent of sex difference, they rename this reconceived division, largely avoiding calling it a ‘gender’ (genre) division. Nicole-Claude Mathieu renames gender ‘social sex’ (sexe social), while Delphy (in earlier work) renames the genders ‘sex-classes’ (classes de sexe). [5] This seems puzzling, since their talk of ‘social sex’ and ‘sex-class’ might lead us to suppose that Mathieu and Delphy are discussing sex and not gender. This supposition would be mistaken. To see this, let us focus on Delphy’s concept of ‘sex-class’, which belongs within her broader account of women’s subordination as presented in Close to Home (1984). For Delphy, this subordination rests on men’s appropriation of women’s economic, sexual and reproductive labour within the home. [6] This relation of exploitation divides human beings into two genders. One is made feminine (féminin) or masculine (masculin) – produced as a woman (femme) or a man (homme) – by one’s position as victim or beneficiary of this exploitative relation.

If being a woman is being a victim of exploitation, how do women differ from other exploited groups such as proletarians? Delphy’s answer is (1) that the exploitative relation that produces women has a distinctive form: it (a) occurs in the family and (b) involves those exploited having to labour for their upkeep rather than for a wage; (2) that the labour that is being exploited here includes sexual and reproductive labour (in accordance with the traditional, unwritten marriage ‘contract’ whereby wives are obliged to have sex with and bear children for their husbands). So the economic–gender hierarchy is simultaneously a sexual hierarchy in which those who are exploited are sexually objectified. This is one reason why Delphy speaks of ‘sex-classes’. A second reason why she uses this term is to indicate that what it is to be a woman (or a man) is to belong to a particular exploited (or exploiting) class, within an exploitative relation that is a social reality and not a consequence of biology. Thus, by renaming gender ‘sex-class’, Delphy intends to highlight how gender, as a fundamentally economic division, is entirely independent of sex (whereas, she thinks, speaking of ‘gender’ would disguise this independence, given the normal view that because there are two sexes there must be two genders). [7]

Other French materialists share the same underlying commitment to a concept of gender. Mathieu speaks of ‘social sex’ in preference to ‘gender’ for two main reasons: (1) to stress that this division primarily oppresses women (i.e. those who are ‘the sex’, whose sex is marked); and (2), most importantly, to stress that the division into two categories (men and women) is a social division, and one that gives rise to the mistaken perception that these categories are grounded in nature – that is, to the belief in two sexes. In contrast, Mathieu claims, the usual conception of gender is that it is overlaid onto a pre-existing sex, so that talk of ‘gender’ carries with it uncritical belief in biological sex. [8] Nonetheless, since Mathieu talks of ‘social sex’ so as to stress that the man/woman division is completely social, her ‘social sex’ is still actually a reconception of, not an alternative to, the concept of gender.

The French materialists, then, sought to radicalize anglophone conceptions of gender by emphasizing gender’s total independence of sex. However, so far the French materialist standpoint still seems to resemble that of anglophone feminism in that both groups appear to accept that there are two biological sexes. For instance, Delphy remarks that it is almost always females who are exploited in the family, and who are made into women through undergoing this exploitation. [9] This remark seems to presuppose that females (and males) exist biologically.

Elsewhere, though – more prominently in her later work, particularly the article ‘Rethinking Sex and Gender’ (1993) – Delphy denies the biological reality of sex. She and other French materialists argue that we only believe that there are biological sex differences, and only perceive human beings as members of two different sexes, because society’s gender division constrains and limits our thought. [10] Whenever any social division into two classes or categories of human beings has been instituted, we begin to assume that certain of people’s biological features make them into members of these two classes, and to perceive those features in that light. Because in this case the two classes are sexual subjects and objects, we find salient those of people’s biological features which are relevant to sex and reproduction (the internal and external genitals, sex hormones, etc.), and we perceive these features as dividing people into two biological sexes. In short, gender (or ‘sex-class’) determines sex, or to be precise determines our belief in and perceptions of sex, which for the French materialists is all that there is to sex.

The French materialists are not denying that some people have, among other properties, breasts, vaginas, wombs, and so on, while others have penises, testicles, and so on. Rather, the French materialists are claiming that society and not nature makes these features salient, and that society and not nature causes us to define those with breasts and vaginas (etc.) as members of one sex, those with penises and testicles (etc.) as members of another. All this illuminates a third reason why Delphy and Mathieu use terms such as ‘sex-class’ and ‘social sex’: they do so because, in their view, gender-class membership determines what sex one is perceived to be and so, effectively, what sex one is. (Moreover, this means that Delphy’s claim that it is almost always females who are exploited and thereby made into women in the domestic mode of production, while males almost always exploit and become men, must be qualified. She herself comments that it ‘remains to be proven … that women are (also) females, and that men are (also) males’. [11] As this comment hints, Delphy should, consistently, say that it is almost always those who have certain bodily features – wombs, breasts, and so on – who are exploited and are thereby made into women and into females, but who are not females prior to being exploited.

Thus, the French materialists again differed from most anglophone gender feminists in that the former denied, while the latter affirmed, that sex difference is a biological reality. This difference went largely unnoticed in the initial English-speaking reception of French materialists (perhaps because that reception focused almost exclusively on Delphy’s Close to Home, in which, as I mentioned, her questioning of the category of biological sex was not yet so well developed). In any event, during the 1990s, many anglophone feminists themselves came to believe that sex is just as much of a ‘social construction’ as gender, due largely to the influence of Judith Butler (herself significantly influenced by Monique Wittig). By a circuitous route, anglophone feminism has, as it were, caught up with French materialism.

But should the French materialist feminists’ denial of biological sex be accepted? I will argue not, and specifically that their position that gender determines sex undermines their claim to be materialist. I will go on to suggest that feminists can accept that sex is a biological reality and follow the French materialists in pursuing the abolition and not merely the restructuring of gender.

How to have sex without gender

Delphy explains, as follows, in what the French materialists take their ‘materialism’ to consist. Materialism, she says, is the view that: (1) ‘the way in which life is materially produced and reproduced is the base of the organization of all societies’;12 (2) the social relations under which the production and reproduction of life takes place are always relations in which one class exploits the work of another;13 (3) historical change results from changes in these relations of exploitation;14 and (4) what ideas and beliefs people hold depends on these exploitative social relations, which people’s ideas reflect. [15] Evidently, then, ‘materialism’ for Delphy means ‘Marxism’, which she is extending so as to analyse women’s oppression as consisting at root in material exploitation. Thus, for Delphy, ‘materialist’ feminism holds that the material exploitation of women by men is primary and that ideas about men’s and women’s ‘natures’ or proper roles, and the belief that there are two sexes corresponding to the two genders, are secondary consequences of this exploitation. [16]

It may be argued that ‘materialism’ as Delphy defines it is not fully materialist. The Italian Marxist Sebastiano Timpanaro argues that the traditional selfunderstanding of Marxism as ‘materialist’ in respect of its recognition of the primacy of economic life (an understanding which Delphy evidently endorses) makes Marxism into a theory that is not completely materialist. [17] This is, Timpanaro argues, because Marxism so understood takes it that humans relate to nature solely by working on it actively, which ignores how natural environments condition human life and how we human beings are conditioned by our own biological nature, a nature which gives us strictly limited powers and makes us insurmountably subject to ill health, old age and death. To be fully materialist, Marxism must acknowledge these facts, Timpanaro claims. Extending his claims, we might say that part of our biological nature as limited, finite beings is that we reproduce sexually – we are neither immortal nor capable of asexual reproduction – so that to be fully materialist, French materialist feminism would have to acknowledge this biological reality of sexed reproduction and, concomitantly, of the sex division. (Ironically, Timpanaro himself does not mention sex difference among the limiting aspects of human nature.)But, of course, Delphy and her co-workers would deny that sex division is a biological reality. In their view, we only believe in this division because there is a socially instituted gender division. One of the clearest statements of the materialist feminist argument on this point comes from Monique Wittig. Wittig denies that any physical property or group of physical properties in itself makes someone male or female. Rather, certain physical properties – those which enable heterosexual sex and reproduction – are only believed to make people sexed against the background of particular assumptions about gender – namely, assumptions that women’s role is to have reproductive sex with men and bring up men’s children. These assumptions derive from the material relation of women’s sexual, reproductive and economic exploitation by men within the family. Wittig says:

What we believe to be a … direct perception [of someone’s sex] is only a sophisticated and mythic construction … which reinterprets physical features (in themselves neutral … ) through the network of [social] relationships in which they are perceived. [18] According to Wittig, because our social arrangements produce an expectation that women and men should have heterosexual sex and reproduce, we come to find the properties that enable heterosexual sex and reproduction salient and to classify people in terms of just these properties; hence, we come to regard people as ‘sexed’. She sums up: ‘It is oppression that creates sex and not the contrary.’ [19] There are problems with Wittig’s argument. To see these problems, we first need to recognize two features of this argument. (1) Wittig’s argument presupposes that people do have certain physical properties – the genitals, sex glands, sex hormones, and so on – which, together, really do enable people to engage in reproductive sex. More precisely, (most) people have either the set of physical properties which enables them to play the ‘male’ role in reproductive sex (that is, to be able to get erections, ejaculate, etc.) or the set of physical properties which enables them to play the ‘female’ role (to be able to ovulate, enclose penises vaginally, etc.). (2) Part of Wittig’s argument is the claim that there is a social norm specifying that men and women should have reproductive sex. But for it to be possible for reproduction to become a social norm, the sets of properties that enable reproductive sex must occur together fairly regularly – regularly enough that this co-occurrence cannot be accidental but must be caused by the properties encouraging one another’s presence. (For instance, having XX chromosomes encourages a human embryo to form ovaries, which, in turn, tend to secrete relatively high quantities of oestrogen, which, in turn, encourages the growth of female genitals.) Unless these properties encouraged one another’s presence in this way, a norm prescribing reproduction could not become as firmly and widely established as it has.

Now, according to one account from philosophy of biology, when certain biological properties tend to occur together, and when this happens non-accidentally (because these properties encourage one another’s presence), and when the co-occurrence of these properties has important causal effects (in this case, enabling individuals to play one or the other role in reproductive sex), then those properties form a cluster. The initial point of this account from philosophy of biology is to specify that things belong to a natural kind when they have most of the properties in a given cluster. [20] But to apply this analysis within the context of sex, individuals belong to either the male or female ‘kind’ – or sex – when they have enough of the properties in either the ‘male’ or ‘female’ clusters. [21] (On this cluster model, individuals do not have to have all of the properties in the relevant cluster to be male or female – for instance, one can be female while having had a mastectomy or while being a post-operative male-to-female transsexual who lacks XX chromosomes.) [22] So Wittig’s claim that there is a social norm prescribing reproduction, when put together with her presupposition that some sets of bodily properties enable reproductive sex, already implies that these properties are clustered. And given that these properties are clustered and, specifically, form two clusters each relating to a different reproductive role, (most) individuals are male or female, depending on whether they have more of the properties in one or the other cluster. So contrary to what Wittig claims, her own arguments imply that most people are biologically male or female, regardless of what gender divisions society erects. [23]

Delphy would be worried by my criticisms of Wittig.

She argues that if there really is a biological division between the two sexes, then people will inescapably be led to classify one another by sex and will then, in any possible society, use the sex division as a basis for assigning different social roles, and different types of work and activity, to differently sex-classified people. And in practice it is unlikely that any such division of labour could ever be other than hierarchical. Delphy’s worry, then, is that if two sexes exist then there must be two hierarchically organized social genders as well. [24]

I would suggest that this worry need not arise if we understood sex, as I have very briefly proposed, to be a matter of having enough of the properties in one or the other cluster. On this view someone can be female or male by having enough but not all of the properties in the relevant cluster and, in fact, while having some of the properties of the other sex as well. Consider again here a post-operative male-to-female transsexual who has the XY chromosomes that form part of the ‘male’ cluster. This does not prevent that person from being female, as long as they have enough of the properties in the ‘female’ cluster and do not also have enough of the properties in the ‘male’ cluster to be male. [25]

Now, if we had this understanding of sex in the background when we classified one another by sex, then our classifications would not imply that all females should be expected to act in one set of ways that is systematically different from how all males should act. This implication would not follow, because our classifications would be based on recognition that females (and males) differ greatly from one another in terms of which, or how many, they have of the properties in the relevant cluster. (This would contrast with the current situation, in which our classifications tend to be based on a misplaced expectation that all females – or males – should have all of the relevant properties or should invariably have one single supposedly sex-defining property, such as XX chromosomes.) Additionally, if a cluster-based conception of sex were in the background, then our classifications would be based on recognition that females and males can be more or less similar to one another physically (because females often have some of the properties of males, and vice versa). In this situation, then, we would have no reason to expect all females to act alike and all males to act alike. We would therefore have no reason either to allocate all females to one position within a division of labour and males to another. Thus, if this cluster-based understanding of sex came to prevail, then sex could be, and be acknowledged to be, a biological reality without this having to lead, as Delphy fears it would, to an undesirably gender-divided society. Contrary to Delphy and her co-workers, it is possible to affirm the biological reality of sex and to pursue the abolition of gender.

Materialism of sex, idealism of nature

At least two worries might be raised concerning the cluster-based model of sex. First, this model might seem to imply that some individuals are more female (or more male) than others in virtue of having more of the properties in the relevant cluster. But potentially the idea that some people are more female (or male) than others is in itself normative, suggesting that it is desirable to be as female (or male) as possible. And if so, then sex classifications would lead to a gender division – for if a sharp sex dimorphism were considered desirable, then it would also be considered desirable to differentiate the sexes further by constraining them to act in systematically different ways, with those people who are less female (or male) than others being expected to conform to this gender-dimorphic pattern as best they can.

However, even if the cluster model does indeed imply that some individuals are more female or male than others, this model does not in itself suggest that being highly female or male is desirable. Only against the background of a two-gender social system does it make sense to regard being highly female or male as desirable. Within this system, where all individuals are expected to conform to one of two opposed genders, being highly female or male appears desirable because it seems to enhance people’s ability to realize their assigned gender. But were we to eliminate the two-gender system (as I have suggested we can, in principle), the cluster model would cease to carry these invidious normative implications.

In any case, the cluster model need not be interpreted as implying that some individuals are more female or more male than others. Instead, the model can be interpreted as stating that anyone who has enough properties from the relevant cluster crosses a threshold into belonging to that sex, where all those who cross this threshold26 are equally as female or male as one another (irrespective of whether they have, say, all of the properties of their sex, most of these properties, or just some of them). As Natalie Stoljar says of this interpretation of cluster concepts, ‘once an individual satisfies enough of the features in the concept, it is fully fledged; there is no room for satisfying the concept to a greater degree.’ [27] A second worry about the cluster model is that it treats sex in humans as the same as sex in other animal species. Certainly, the particular ranges of properties that are relevant to being a human female or a human male may well differ, in whole or part, from those relevant to being a female or male of another species: for instance, female birds have ZW rather than XX chromosomes, and male birds ZZ rather than XY. Nonetheless, on the cluster model sex in humans is the same kind of thing as sex in other species: namely, for humans and other animals, being sexed consists in having enough of one of the clusters of biological properties that are relevant to reproductive sex. On this view, there is nothing ontologically distinctive about sex in humans.

Yet this view has difficulties from a feminist perspective. Because feminists have wanted to insist that gender norms can be changed or removed altogether, they have generally maintained that gender – and the social, cultural and historical dimension of human life of which gender is part – is irreducible to humanity’s natural or biological side. But if we also maintain that human sex difference is the same kind of reality as sex difference in non-human animals, then we seem to be picturing human beings as split between a uniquely socio-cultural side and a still animal, sexed side. Even if it is accepted in this picture that the two ‘sides’ interact, this interaction appears powerless to alter the mode of being of human sex, which seemingly remains the same as that of animal sex.

Let me tentatively suggest that we could relieve the cluster model of these uncomfortable implications by combining it with elements of German Romantic and Idealist philosophies of nature. Arguably, the German Romantics and Idealists think that the cultural and historical domain that is unique to humanity is essentially characterized by normativity – that is, this domain is organized around epistemic, ethical, aesthetic and other values which operate as norms guiding various human activities. On one line of interpretation, the German Romantics and Idealists take it that this norm-guidedness makes human cognitive, moral and cultural activities irreducible to the causal order of nature. [28] But we need to complicate this interpretation, given that the German Romantics and Idealists think that nature itself contains a normative dimension. According to Hegel, all natural things develop rationally – that is, in accordance with norms of rationality – and so have a rational, conceptual side that is irreducible to their causal, material side. For the early German Romantics, nature is a self-organizing whole that develops, through self-differentiation, into an infinite multiplicity of things and processes. [29] On this view, nature’s way of endlessly, creatively, unfolding parallels the process by which works of Romantic literature unfold in an unfinished, fragmentary way, which reveals that nature as a whole, and the many component natural things, are developing according to aesthetic norms.

On both these views – Hegel’s and that of the early German Romantics – the norm-guided character of human cognitive, artistic and cultural activities is a further development and realization of the normguidedness within nature. As such, humans are only able to develop a cultural side because they – in common with other animals – are organisms whose organization is norm-guided (as part of nature’s overall norm-guided self-organization), a norm-guidedness that can be developed into further cultural forms. Even if our cultural activities affect and change our natural, organic side, then, they cannot alter its basic mode of being as natural and organic, for these cultural activities could not exist unless our organic nature persisted as the soil out of which they are constantly regenerated. Humans, qua cultural, must also be natural and organic. Moreover, for Hegel and the German Romantics, an essential part of the character of higher organisms is that they reproduce sexually and are differentiated into two sexes with different roles in reproduction30 (which need not imply that every single higher organism must be unambiguously sexed). On these grounds one might suggest, following Romantic and Idealist Naturphilosophie, that sex in humans must be the same kind of thing as sex in (other) animals because humans, although unique in respect of their cultural activities, can only be unique in this respect because they retain an organic, and sexed, nature in common with other higher animals.

The French materialist feminists conceived their account of sex as materialist in that it treated gender as a socio-economic reality which is not merely independent of biology but, indeed, constructs sex out of non-sexed biology. I have argued that a more completely materialist feminism would recognize sex to be a biological reality, according to a cluster model of sex which allows that gender can be eradicated. But, ironically, it may be that to defend this cluster model we need to combine it with German Idealist philosophies of nature – philosophies whose ‘idealism’ consists in their attribution to nature of not only materiality but also rational or aesthetic normativity (i.e. ‘ideality’). A fully materialist feminism may require an idealist approach to nature.


My thinking about the ontology of sex has benefited from discussions with Stella Sandford.

1. ^ Versions of this position can be found in: Judith Butler, ‘Variations on Sex and Gender: Beauvoir, Wittig,

Foucault’, in Feminism as Critique, ed. Seyla Benhabib and Drucilla Cornell, University of Minnesota Press,

Minneapolis, 1987; Anne Fausto-Sterling, Sexing the Body: Gender Politics and the Construction of Sexuality, Basic Books, New York, 2000; Suzanne Kessler and Wendy McKenna, Gender: An Ethnomethodological Approach, University of Chicago Press, Chicago, 1985; Suzanne Kessler, Lessons from the Intersexed, Rutgers University Press, New Brunswick NJ, 1998; Georgia Warnke, ‘Intersexuality and the Categories of Sex’, Hypatia, vol. 16, no. 3, 2001, pp. 126–37; as well as in the French materialist feminists discussed here.

2. ^ The publication of the anthology New French Feminisms, ed. Elaine Marks and Isabelle de Courtivron (University of Massachusetts Press, Amherst, 1980) was the founding moment in the emergence of this category.

3. ^ Toril Moi, Introduction to French Feminist Thought, ed. T. Moi, Blackwell, Oxford, 1987), p. 6. More recent anthologies of and books on ‘French feminism’ recognize the need to be more inclusive; e.g. Dani Cavallaro, French Feminist Theory: An Introduction, Continuum, London, 2003; Kelly Oliver, ed., French Feminism Reader, Rowman & Littlefield, Lanham MD, 2000; Kelly Oliver and Lisa Walsh, eds, Contemporary French Feminism, Oxford University Press, Oxford, 2004.

4. ^ Christine Delphy, Close to Home: A Materialist Analysis of Women’s Oppression, ed. and trans. Diana Leonard,

Hutchinson, London, 1984, pp. 24–5.

5. ^ Leonard and Adkins (‘Reconstructing French Feminism: Commodification, Materialism and Sex’, in Sex in Question: French Materialist Feminism, ed. Leonard and Adkins, Routledge, London, 1996, p. 16) argue that because the French materialists use terms (‘sex-class’, ‘social sex’) other than ‘gender’, they oppose the concept of gender. In contrast, I think that they use these terms as part of a programme of conceiving gender in a new, improved way – as something fully independent of sex and hence eradicable. Delphy is most explicit about this (Close to Home, pp. 25–7). However, I shall argue – with reference to Mathieu – that the same commitment to a reinterpreted concept of gender is effectively, if less explicitly, present in the other French materialists too.

6. ^ Men’s economic exploitation of women in the family constitutes what Delphy calls the ‘domestic mode of production’, which is patriarchal in that it benefits men and which, Delphy claims, exists alongside and entwined with the capitalist mode of production.

7. ^ In later work Delphy argues directly for retaining the term ‘gender’ (genre): see ‘Rethinking Sex and Gender’ (1993), in Sex in Question, p. 36.

8. ^ See Mathieu, ‘Sexual, Sexed and Sex-Class Identities [Identitée sexuelle/sexuée/de sexe]’ (1989), in Sex in Question, pp. 59, 68.

9. ^ Delphy concedes that some younger brothers and male agricultural workers are also exploited by male heads of household. (Her account of the domestic mode of production is based on her studies of French agricultural families.)

10. ^ Sometimes the French materialists follow Simone de Beauvoir in holding that only women are seen as ‘the sex’, defined by their genitals as members of a class, whereas men are seen as sexually unmarked. But elsewhere the materialists claim that all individuals are seen as sexed because of the existence of a gender division.

For instance, Guillaumin claims that, given gender, people’s reproductive organs are treated as the ‘sign’ or ‘mark’ of their (sex-)class membership. ‘The Practice of Power and Belief in Nature’ (1978), in Sex in Question, p. 107.

11. ^ Close to Home, p. 24.

12. ^ Ibid., p. 159.

13. ^ Ibid., pp. 212–13.

14. ^ Ibid., p. 159.

15. ^ Ibid., pp. 159, 212–13. According to Guillaumin, the idea of natural sex is just ‘the mental form taken by certain determined social relationships’ (‘Practice of Power’, p. 74).

16. ^ However, suppose we ask Delphy why it is almost always those who have wombs, breasts, etc. who are exploited. Plausibly, this is because of the prevalence of certain beliefs about what tasks/social positions it is appropriate for people who have those bodily properties to do/occupy. That is, certain beliefs or ideas look like preconditions, not merely consequences, of the exploitative social relation. Delphy objects that granting ideas this much status is ‘idealism’, but arguably a viable feminist analysis just has to be more ‘idealistic’ than Delphy wishes. For unless we grant ideas and beliefs this preconditional status, then it looks as if it must be the sheer fact of possessing wombs, breasts, etc. which tends to render their possessors liable to exploitation.

In short, we would have returned to the very biological determinism from which Delphy seeks to distance herself.

17. ^ Sebastiano Timpanaro, On Materialism (1970), trans.

Lawrence Garner, New Left Books, London, 1975, ch. 1.

18. ^ Monique Wittig, The Straight Mind and Other Essays, Beacon Press, Boston MA, 1992, pp. 11–12. Wittig stresses that the domestic mode of production is not only patriarchal but also, equally fundamentally, heteronormative. This mode of production, under which women must work for their husbands, needs women to be tied into marital relationships and therefore requires women and men to be heterosexual, so that they will be motivated to contract heterosexual marriages.

19. ^ Ibid., p. 25.

20. ^ For this account, see Richard Boyd, ‘How to Be a Moral Realist’, in Essays on Moral Realism, ed. Geoffrey Sayre-McCord, Cornell University Press, Ithaca, 1988), pp. 96–9.

21. ^ This account of the clusters of properties involved in sex might seem circular, since it says that someone is male (or female) if they have enough of the ‘male’ (or ‘female’) cluster of properties. But I am speaking of the ‘male’ (or ‘female’) cluster only as a shorthand for all the relevant properties – e.g. XY chromosomes, penis, prostate gland, etc. – so that listing all these properties would remove the circularity. However, I have also said that part of what makes these properties a cluster is that they enable those who have them to play the ‘male’ (or ‘female’) role in reproduction. Again, this circularity can be removed if we spell out all the elements of that role, e.g. production, storing and ejaculation of sperm, capacity for erection, etc. Let me stress, I am not claiming here that what it is to be female or male is to be able to play the relevant role in reproductive sex. (This would exclude those who are too old or young to reproduce or who are infertile, and would, problematically, imply that males and females ought to realize their natures by engaging in reproductive, heterosexual activity.) Rather, to be female or male is to have enough of the relevant cluster of properties, where part – but only part – of what makes these properties cluster is that, together, they generally support particular kinds of reproductive activity in people of suitable age (unless causes of infertility are present). So one could be female or male by having enough of these properties while not having enough, or the right combination of, properties to be able to contribute to reproduction.

22. ^ I explain this cluster-based approach to sex more fully in my An Introduction to Feminist Philosophy, Polity Press, Cambridge, 2007.

23. ^ To clarify: Wittig accepts that some individuals have breasts, vaginas, etc. while others have penises, testicles, etc., but she denies that these biological properties as such suffice to make those individuals sexed. Yet, contrary to her intentions, Wittig’s own claims imply that individuals’ possession of those properties does make them biologically sexed.

24. ^ Delphy, ‘Rethinking Sex and Gender’, pp. 37–40.

25. ^ Those with enough properties from both clusters to be both female and male are hermaphroditic (e.g. they might have one ovary and one testis, produce both egg and sperm cells and develop internal and external genitals which mix male and female attributes).

26. ^ Unless they also cross the threshold into the other sex, in which case they are hermaphroditic.

27. ^ Natalie Stoljar, ‘Essence, Identity, and the Concept of Woman’, in Philosophical Topics, vol. 23, no. 2, 1995, p. 285. Stoljar contrasts this to the view that those who have crossed the threshold into femaleness/maleness (or into ‘womanness’, the subject of Stoljar’s interest) can be female/male to greater or lesser degrees.

28. ^ For example, Wayne Martin argues that for Fichte the fact that cognition is guided by a norm of representing outer reality makes cognition irreducible to the causal order of nature; see W. Martin, Idealism and Objectivity: Understanding Fichte’s Jena Project, Stanford University Press, Stanford, 1997, ch. 2.

29. ^ See Frederick Beiser, The Romantic Imperative: The Concept of Early German Romanticism, Harvard University Press, Cambridge MA, 2003. Here I follow Beiser in assuming continuity between early German Romanticism and the Idealism of the earlier Schelling and of Hegel.

30. ^ See, for instance, Schelling, First Outline of a System of the Philosophy of Nature, trans. Keith R. Petersen, SUNY Press, Albany, 2004, pp. 36, 40–42.









S Diffusion Populaire 21 ter, rue Voltaire, 75011 Paris, [archive]