Early in January this year, the French radical feminist Marguerite Stern – known in particular for her street postering campaigns – introduced the project “Femelliste”, described in its manifesto as “a movement that fights to maintain women’s sex-based rights”. Two years earlier, the British women’s rights activist Kellie-Jay Keen (aka Posie Parker) had posted a stream to her YouTube channel with the title “The future is #femalism”. The hashtag #Femalism has 2.5 million views on TikTok, and others have used variations of the term. Something interesting is happening here: some women appear to be trying to do feminism without the word “feminism”. Why?

One obvious answer is that the word has implications these new feminists want to get away from. Putting the “female” back into “feminism”, as signalled by words like “femalism” and “femelliste”, is a way of signalling a commitment to a distinctively sex-based movement. This is a preoccupation that feminism had, and then lost. The neologisms are a way of trying to put it back.

Within feminism today, it’s possible to find at least four different answers to the question of who feminism is for. One is “everyone”, most prominently defended by the black feminist Gloria Jean Watkins (aka bell hooks) in her book Feminism is for Everybody (2000). Another is “anyone hurt by the gender system”, which may be argued to be everyone, and so collapse into the first answer, or may leave out a limited number of people (e.g. alpha males) thought to benefit from the gender system rather than being hurt by it (this is roughly the view put forward in the discipline of Gender Studies). A third answer is “anyone who identifies as a woman” (or girl), which is perhaps the most common answer among progressives at this time. This captures some of the common-sense idea that feminism is for women, but opens up the category of “woman” by making it a matter of gender identity rather than sex. A final answer, and the one Stern, Keen, and others have given, is that feminism is for females, those members of the biological sex category that typically produces large, immobile gametes (eggs).

These answers differ in the size of the constituency they deliver to feminists: everyone; almost everyone; about half the population (the third and fourth answers deliver constituencies of about the same size, because tracking gender identity replaces the females who identify out of the group with the males who identify in).

Is it possible to adjudicate this disagreement about who feminism is for? I think it is; or at least, that it’s possible to make some progress. Those who say that feminism is for everyone tend to mean that feminism is good for everyone, men and women alike. But that can be true without it being the case that feminism is for everyone, in the sense of being about everyone. So we can set that answer aside. Those who say that feminism is for anyone hurt by the gender system choose to focus on a broader problem than those who say that feminism is for women or that feminism is for females, but there’s not necessarily any incompatibility between these projects. We could all agree that there’s a “gender system” which produces negative effects for men who depart from paradigmatic masculinity, as well as for women, gay and bisexual people, intersex people, trans and nonbinary people, and anyone else “gender non-conforming”. We could also agree that each of these groups will be affected in different ways, that some are almost exclusively hurt by the system while others experience a mixture of harms and benefits, and yet others are almost exclusively benefited. We could divide up the intellectual and political labour by working on different aspects of the “system”, with some focusing exclusively on the harms to female people. There’s no reason why we couldn’t have both (the narrower) women’s studies and (the broader) gender studies. So there’s not necessarily a disagreement here about which projects are valuable, only perhaps a disagreement about which projects people should choose to contribute to, and which project deserves the word “feminism” (assuming that one deserves it more than the others).

For this reason, I’ll focus on the disagreement between those who agree that feminism is for women, but disagree about what a woman is, with one group tracking gender identity and the other group tracking sex. I’m a gender-critical feminist, which means I’m with Stern and Keen in thinking that feminism is for women understood as females – but, like Julie Bindel in her (2021) book Feminism for Women, I depart from Stern and Keen in that I am not willing to cede the word “feminism” to the gender identity project. Here, though, my concern is not with the word but with the project. So my question will be, should gender identity replace sex as the core concept of feminism? First I’ll respond to some common objections to using sex as a core concept, then I’ll consider whether gender identity is up to the job.

Two objections to thinking that feminism is about sex, and so for females, have been prominent in feminist history. The first is that this is to understand the feminist project in terms of a fixed constituency, and involves providing a definition of that constituency that is essentialist and universalist. (Try to spit these words when you say them, to get the flavour of how they’re both currently viewed inside feminist circles.) The second works by counterexample: some of the people mistreated in the way feminists typically care about are not female, so feminism can’t be about sex, and so for (only) females. Let me take these in turn.

For a definition to be essentialist is for it to make use of a single necessary and sufficient condition, that condition naming a property that all things must have to be a member of the category. For example, “a shape with exactly three straight sides” is an essentialist definition of triangle. Any shape with exactly three straight sides is a triangle, and any shape with fewer than three or more than three straight sides – or any number of non-straight sides – is not a triangle. For a definition to be universalist is for it to make a claim about the things it defines that is true for all of them, at any time. Our definition of triangle is not simply the definition of some triangles, or triangles now. It’s the definition of all triangles now, and all triangles ever.

Feminists forty/fifty years ago worried about essentialist definitions of “woman” because they associated them with the idea of biological determinism, thinking that whatever property was named as the essential one would slip from being merely descriptive (what she is) to being normative (what she should be, or should do). For example, if “woman” means “person capable of motherhood”, that might slide into “person who should be a mother”, person who, if not a mother, is somehow aberrant, abnormal, a failure. They worried about universalist definitions of “woman” because they worried about exclusion. To continue with the same example, if “woman” means “person capable of motherhood”, then people not capable of motherhood, including infertile women and post-menopausal women, are not women. But we generally think that infertile and post-menopausal women are women, so this definition is exclusionary: it doesn’t include some people that it should include.

(As a quick aside, some feminists responded to worries about essentialism and universalism by simply refusing any definition of “woman” at all. If we don’t define “woman”, the reasoning went, then we won’t end up essentialising or excluding anyone. This may be helpful in explaining why contemporary feminists seem so supremely unconcerned about legal proposals for sex/gender self-identification. If “woman” has no definition, then there can’t be any falsehoods about who is or isn’t one, and there can’t be any boundary policing about who feminism is and isn’t for. But this move alone can’t justify feminism being about gender identity rather than sex – it works against sex, but not for gender identity. The flipside of there not being any falsehoods is that there aren’t any truths, so it’s also not the case that feminism is about gender identity, and so for anyone who identifies as a woman.)

Still, there’s a big difference between defining “woman” as “person capable of motherhood”, and defining “woman” as “adult human female”. The latter is a definition in terms of sex, and it’s essentialist and universalist. But the essential property can’t easily slide from descriptive to normative. She is an adult human female; she, the adult human female, should be someone who doesn’t go about casually murdering her enemies. Similarly, there’s no meaningful sense in which a frog should be a frog. It just is one. If there are normative questions, they come later; about what she should do, or be like.

There’s more room to argue about whether the universality of the definition makes it exclusionary in a bad way. “In a bad way” is important here, because categories are inevitably exclusionary in the sense that they all include some things and exclude others. The category “frog” excludes “dragon lizard” and includes “Australian green tree frog”. Defining “woman” as “adult human female” will then inevitably exclude Amy, who is an adult human male, and include Bella, who is an adult human female. The question is not whether this is exclusionary full stop, which is unavoidable, but whether it excludes something that should be included.

This takes us to the second objection to thinking feminism is about sex, which works by counterexample. Even feminists who think that feminism is about sex and so for females should surely care about people being mistreated because they are (wrongly) assumed to be female. Individuals with the intersex condition 5-alpha reductase deficiency (5-ARD) are male, but may appear female. Some male transsexual and transgender people “pass” as female in their daily lives. In both of these quite different cases, the individuals may be subject to mistreatment typically directed at females, for example sexual harassment. This appears to be a counterexample to the claim that feminism is about sex. If feminists should care about these people, then feminism has something to do with sex, but that “something” is not so strict as to make feminism only for and about females. That opens the door – at least a crack – to saying that feminism is about gender identity instead.

However, just because mistreatment on the basis of sex casts a slightly skewed net, catching some people not of the targeted sex and missing some people of the targeted sex, doesn’t mean the mistreatment isn’t because of sex. That is, the reason why people who look female are targeted is that people assume looking female is a reliable guide to actually being female. It’s not looking female that is the problem – someone widely known to be male but who looks female (the conservative trans YouTube content-creator Blair White may be a good example here) is likely to be subject to different mis/treatment than someone thought to be female who looks female. Societal expectations about how persons should be are attached to sex, and they’re different for male and female people. A feminism for and about females may be good for some males, because it reduces mistreatment they are also subject to. But that doesn’t mean feminism has to be for and about some/any males.

Perhaps it’s not that sex is particularly objectionable as a core concept, but just that gender identity is better. Let’s consider some reasons to think that it is. The first is based in a claim about consistency. Here’s one version: feminists think that women are oppressed, and that this must end. But if you recognise that oppression is bad for women, then you should be able to recognise that it is bad for anyone. So feminists should not, as a matter of consistency, contribute to the oppression of others. They do that when they don’t accept transwomen’s claims to be women. (A version of this argument appears in Katie Kirkland’s 2019 paper “Feminist Aims and a Trans-Inclusive Definition of ‘Woman’”.) This consistency argument moves us from sex to gender identity, because if transwomen are accepted as women, and feminism is for all women, then feminism must be about gender identity. (If gender identity is universal, meaning that everyone has one, then it can be only about gender identity; if it isn’t, then this argument will only suffice to establish that feminism, on pain of inconsistency, must be about sex and gender identity).

Here’s another version, relating to the answer discussed earlier that feminism is for anyone hurt by the gender system. Second-wave feminists were concerned with the imposition of gender – by which they meant, the socialisation of females into femininity – on the basis of sex. They want this to end, so that women have more freedom. But if you recognise that the imposition of gender on the basis of sex is constraining of freedom for women, you should recognise that for trans and nonbinary people as well. So feminists, as a matter of consistency, should not contribute to the constraining of trans and nonbinary people on the basis of their sex. They should accept trans people’s self-identifications, just as they want everyone else to accept women’s self-identifications as more/other than “feminine”. Again, this moves us from sex to gender identity, because if transwomen are accepted as women, then feminism must be at least in part about gender identity.

The apparent persuasiveness of these two versions of a consistency argument depends on an elision of the distinction between sex and gender identity, and the suppression of a disagreement over what is in fact oppressive. It’s true that those who want an end to their group’s oppression should not actively contribute to the oppression of others (unless there is an unavoidable conflict of interests between the two groups). Let’s accept for the sake of argument that it’s oppressive not to accept people’s claims about their own gender identity. Still, it’s not clear that maintaining a sex/gender identity distinction contributes to that oppression. Feminists can accept transwomen’s claims to be women in the sense that that is the term they use to name their gender identity. But it’s not their sex. One term, “woman”, is being used in two very different senses. It could be true that transwomen are women understood as a gender identity, and false that transwomen are women understood as a sex. Then we can be consistently anti-oppression while yet maintaining that feminism is about sex, and for females. Similarly, feminists might recognise that more people than just women have an interest in being free from gendering on the basis of sex. They might try to pull those other people in as allies to the feminist movement, or they might make themselves allies to those other people’s political movements. But there’s nothing to say that they must, on pain of inconsistency, absorb anyone constrained by gendering into the feminist movement. Strategic alliances may be more than sufficient for political purposes.

Another consideration in favour of gender identity rather than sex, and the one I suspect to be driving the move from sex to gender identity in the public understanding of feminism’s constituency and project, is simply that gender identity does the same job as sex, as well as being more inclusive. It doesn’t miss anything out, and it adds something further in – so it’s better for everyone. If this were true, it would be hard to resist. If there’s no cost to women of adding in to feminism those who are not female but identify as women, and it would be good for those people, then we might as well do it. More good is better than less good! So; is it true?

Gender identity replacing sex is not more inclusive, it’s just differently inclusive. If feminism is about sex then it includes females who do not identify with/as women, namely transmen and female nonbinary people. If it’s about gender identity then it excludes those females, and instead includes males who do identify with/as women, namely transwomen. This is a cost to those who are women understood as a sex but not women understood as a gender identity, because they experience a disproportionately high rate of sexual assault, and that is something feminism is paradigmatically concerned with. Gender identity replacing sex also leaves out all the women who don’t have a gender identity. That is at least all those women who lack the cognitive capacity to form a self-conception in terms of masculinity, femininity, both, or neither (a point made by Elizabeth Barnes in her 2020 paper “Gender and Gender Terms”). But it may be a much larger number of women: radical and gender-critical feminists, for example, say that they do not have gender identities. (If “gender identity” is like sexual orientation, then this is silly: just because “heterosexual” is more common, doesn’t mean it’s not an orientation. But if “gender identity” is like pregnancy then it’s not: the fact that some people are pregnant doesn’t mean that everyone else is a “non-pregnant woman” or a “dormant reproducer” or whatever novel term we might like to come up with).

The common view that gender identity does the same job as sex and is more inclusive, then, is wrong. But suppose that we decide we care more about the inclusion of transwomen than about the exclusion of females without “woman” gender identities, and so are happy to make the switch – that is, we’re happy to be differently inclusive instead of more inclusive. Are the interests of the women who both groups overlap on (females who identify as women) equally well-served by this switch? I don’t think so. There are various paradigmatically feminist concerns relating directly to sex that can’t be captured as a matter of gender identity, like sex-selective abortion, child marriage, female genital mutilation, and abortion. In delivering some males to feminism, gender identity introduces substantially more difference into feminism’s constituency, which undermines existing commonalities. Feminist language will shift so as not to presume a woman’s sex, and this is likely to end up occluding sex differences that matter. If feminism is for women and “woman” is a gender identity, then feminism is a mixed-sex project. This does not serve the great majority of women. (In the recent UK census, the number of transwomen was found to be ~0.1%).

One final argument that might support the shift from sex to gender identity is associated with the radical feminist legal scholar Catharine MacKinnon. MacKinnon said in one interview that “how one becomes a woman is not … our job to police”, and that “To be a woman, one does have to live women’s status. Transwomen are living it”; described herself as “aggressively indifferent” to the question of whether some women/females can have a penis; and said that “We all have a lot to do beyond policing ‘women’”. These remarks may seem surprising from a radical feminist, who would normally have a vehemently sex-based analysis. But MacKinnon has a distinctive and strongly socially-constructionist view of sex and gender, which is more amenable to accommodating transwomen than many other radical feminists’ accounts.

MacKinnon thinks that the core problem of sex inequality is sexuality, which has been socially constructed as the eroticisation of power: male dominance, and female subordination. Unlike many of the second-wave feminists, MacKinnon does not think that there is first sex, the natural, and then gender, the cultural, imposed upon its basis. Rather she thinks that there is class first, a separation of human persons into two categories, the dominators and the dominated, by way of sexuality. In her 1982 paper “Feminism, Marxism, Method, and the State: An Agenda for Theory”, she wrote that “sex as gender and sex as sexuality are thus defined in terms of each other, but it is sexuality that determines gender, not the other way around”. Society creates sexuality; sexuality creates both sex and gender. All of this is social. In Chapter 19 of her 2017 book Butterfly Politics, MacKinnon targets prostitution and pornography as central culprits in the maintenance of sex inequality, saying “Women have to be seen as sexual things so that their sexual use can be normalised as what women are for. Those qualities, displayed, are called femininity”. She argues that pornography makes sex inequality sexy, and that makes it much harder to address.

While MacKinnon does not talk in the chapter about transwomen – and indeed a lot of what she says can be read as a straightforward analysis of the sexual subordination of females – we can nonetheless extract a kind of argument from it. If the core problem of feminism is sex inequality, and this is a matter of sex, sexuality, and gender being constructed from an initially arbitrary segregation of persons into two social categories, then there’s room to think this segregation doesn’t track what we think of as sex (namely biological sex). And if it doesn’t track what we think of as sex, then it might track gender identity instead. Whether MacKinnon would go all the way to identity alone is unclear, but it is easy enough to imagine her agreeing that certain males are “the dominated” rather than “the dominator” when it comes to the politics of sex, for example the males in prison who are raped, and the males working in prostitution. (In a 2017 symposium she said “I also regard raped men as among my sisters”). If being a woman is a matter of systematic sexual subordination, then autogynephilic transwomen who desire the sexually submissive social role (“systematic” here means not just in the bedroom, but full-time) may count as women for MacKinnon. (Andrea Long Chu expresses such a view of women, and such a desire, in Females).

Still, this is only an argument for shifting from sex to gender identity in understanding who feminism is for if you buy MacKinnon’s analysis. There’s a lot to be said for the centrality she assigns to sexuality, and to prostitution and pornography as drivers of continuing sex inequality. But her account falls down at the social construction of everything. She offers no account of the incredible coincidence that the two social categories of dominator and dominated track millions-of-years-old reproductive differences. If we modify her account to reintroduce sex as the thing sexuality and gender are constructed out of, we lose the easy route to the inclusion of some males as “women”.

Although this survey has not been exhaustive, I am inclined to conclude that the arguments against feminism being about sex and for females have been overstated; and that the arguments for feminism being about gender identity and for anyone who identifies as a woman do not hold up under scrutiny.

Holly Lawford-Smith is a philosopher, scholar, researcher, author and associate professor in political philosophy, University of Melbourne