1. The life of concepts

In this paper, we consider a range of trans-inclusive approaches to gender concepts and how they relate to the world, from family resemblance theories to conceptual engineering. In so doing, we shall also examine several concepts that are analogous to gender and other controversial and not-so-controversial concepts, such as “vegan cheese” and “adoptive parents.” Having assessed their merits and demerits, we argue that our gender concepts are not fixed in stone but, rather, evolve alongside human practices and behavior.

A corollary of this view is that conceptual change results from changes in our everyday practices and not the other way around, as conceptual engineers would have it. Our evolving vocabulary for dairy substitutes suggests that there is no objective answer to the question “Is nut milk real milk?” While slogans such as “Nut milk is milk” or “Nut milk is not milk” may in theory be offered in earnest as ontological claims, we must not ignore the things that people do with words when making such assertions. These can range from airing one's moral, social, or political allegiances to attempting to introduce or endorse new ways of speaking.

2. War of the words: neologisms and pronouns

The current controversy over gender-neutral terms, such as “chest-feeding,” is highly emotive, with prominent trans-exclusionary feminists wading in on the argument—including, here, Kara Dansky:

Like several aspects of public disagreements about trans rights, the distress experienced by trans-exclusionary feminists is based on a misunderstanding—or in this case, two. The first is that these neologisms stem from trans people's rejection of biological reality; the second is that they are intended to cancel and replace our ordinary words. In an article titled, “War of words risks wiping women from our language,” journalist Janice Turner unwittingly shows that she is also laboring under this misapprehension. Although she exhibits empathy for the dilemmas that trans people must negotiate, she argues:

Turner contends that trans activists fail to recognize the existence and importance of biological sex: “There is no need for this rancorous divide between trans activists and feminists. Yet peace depends upon an agreement that sex exists, that in certain limited circumstances it overrides gender, and that language to describe biological reality is valid.”Footnote ⁵ Trans people and their allies need not deny that biological sex exists.Footnote ⁶ Indeed, some invest considerable amounts of money to surgically change their natal or “birth” sex. While this is sometimes referred to as “sex assigned at birth,” we reject this terminology on the grounds that it assumes that there is no real biological sex to track and consequently it lends credence to the erroneous view that one can only fully support trans people if one subscribes to an ideology that denies biological reality.

As for language needing to capture biological reality, this is precisely what these neologisms and conceptual extensions do; for example, they indicate that the group of people said to menstruate or bear children include not only cis women but trans men and intersex people. We should not assume, write Ray Briggs and B. R. George, that “there's a single, standard ‘female’ experience, associated with all, and only, cis women” or that, by adding new words, we are rejecting old ones:

The words “man” and “woman” mean much more than human beings born with male or female sexual characteristics. Gender-neutral neologisms such as “chest-feeding” describe our biological bodies as they are—whether it be from birth or post-surgically. Words like “woman,” “man,” “mother,” and “father” have their place, and we oppose attempts to eliminate them from our vocabulary (see §§8–9) while nonetheless supporting the widening of the net in terms of the groups to whom they refer, just as we support the creation of new words to describe people and activities that are not adequately covered by more traditional, binary terminology.

Gender-neutral pronouns are also often decried for various reasons, such as grammatical awkwardness (“they” as singular) or on the grounds that they represent “a fad or an attempt to be cool.”Footnote ⁸ Dennis Baron reminds us that the attempt to institute gender-neutral pronouns is not merely a recent phenomenon:

Gender-neutral pronouns are there to prevent us from using the wrong gendered pronoun when referring to someone. The discomfort of being referred to by the wrong pronoun is not superficial. A 2021 survey encompassing nearly 35,000 LGBTQ youth aged 13–24 across the United States indicated that “transgender and nonbinary youth who reported having pronouns respected by all of the people they lived with attempted suicide at half the rate of those who did not have their pronouns respected by anyone with whom they lived.”Footnote ¹⁰

3. Family resemblance

Some philosophers think that terms such as “man” and “woman” are “family resemblance concepts.”Footnote ¹¹ According to this view, there is no set of necessary and/or sufficient conditions for being of a certain gender. Gender cannot be a mere question of, say, biological sex, natal or otherwise. While the phrase “family resemblance concept” is associated with Wittgenstein, he never actually used the phrase as such himself, and for good reason: it implies that there exists a special subclass of concepts that stubbornly elude analysis. The implication appears to be that there is something special about such concepts that distinguishes them from others, with only precise technical concepts, such as “H₂0,” remaining immune to the open fluidity of everyday language. It is little wonder, then, that philosophers cannot even agree on the definition of “water” or “book” (see §5).

Wittgenstein famously gives the example of the concept of “language”:

Gender terms are no different from any others in having more than one legitimate sense, some more literal than others. Our conceptual pluralism allows for trans men and women to be considered literal (and not merely honorary or allegorical) men and women (§4), terms like “trans man” and “cis man” being subtypes within the umbrella use of the term “man” (see §5). We proceed by exploring how possible analogies with transhood may reflect its conceptual relationship to cishood.

4. The vegan analogy

The conceptual evolution surrounding development of plant-based variants of products traditionally made from animals provides us with an instructive insight into the natural reshaping of concepts alongside changes in eating practices. While meat and dairy for now remain the paradigmatic categories for milk, butter, burgers, and sausages, they are not the only ones. Indeed, the paradigms are already shifting, and recent years have witnessed philosophical and legal debates, the latter typically motivated by the increasingly competing concerns of ministries for the environment and ministries of agriculture.

Whereas in Canada and the US you can buy products labelled as vegan “milk,” vegan “cream,” vegan “cheese,” and vegan “butter,” the European Court of Justice ruled in June 2017 that plant-based foods cannot be sold within the EU using such terms. The words in question include “sausage,” “steak,” “burger,” “milk,” and “cheese.” In 2018, France amended its agriculture bill to prohibit the use of so-called “meat and dairy terms” to describe plant-based products that serve as meat substitutes. These include patties made from soybeans and dairy alternatives made from oats, coconuts, almonds, cashews, hazelnuts, and soy.Footnote ¹³

Coconut milk and coconut cream have been in widespread circulation from long before plant-based diets became fashionable in the West. While their respective terminologies are here to stay, in some countries (including France and the UK), it is now forbidden to market products as soya milk or coconut cheese. In response, brands have been obliged to rename their soya and oat milks as “drinks” and their vegan cheeses as “cheeze,” and so on.

The divergence in laws regulating the description of vegan products has created an interesting duality in the marketing campaigns of vegan brands. Thus, for example, in the US, the Swedish company Oatley brand their drink as “Oat milk.” By contrast, in the UK and the EU, Oatley's slogans entreat the consumer to “ditch milk and switch to oat drink,” promising that “it's like milk but made for humans.” Is Oatley contradicting itself? Or is it merely switching from a loose (ordinary) sense of “milk” to a narrower (legalistic) one? Their website includes the following text, tellingly available only to users outside Sweden: “So what is this oat drink anyway? Milk? No, it's not milk. Milk comes from a cow. It was designed for baby cows.” The company is, of course, attempting to transform what was intended as a legal ban into positive marketing rhetoric. Ultimately, however, it is highly unlikely that people will ask their barista for a drop of “oat drink” as opposed to “oat milk” with their coffee; rather, they will ask for oat milk. Institutions and governments can try to police language all they like, but they cannot so easily alter the vernacular.

Language evolves naturally alongside our behavioral practices; the evolution of new uses and meanings of words cannot be stemmed by legislation and nor should it be. Should one really wish to remain stuck in the past, it is worth remembering that the English word “sausage” and the French saucisse both derive from the Latin salsus, meaning salted. Perhaps France would also like to ban unsalted sausages?

In September 2022, Merriam-Webster added “oat milk” to its dictionary, cementing the popularity of this terminology. The definition offered is “a liquid made from ground oats and water that is usually fortified (as with calcium and vitamins) and used as a milk substitute.”Footnote ¹⁴ It is worth noting that the dictionary deems it correct to use both “oat milk” and “milk substitute” in the same breath. The context makes clear what each use of “milk” refers to. The second of these is, of course, dairy milk, the adjective-noun combination that functions analogously to cis man or woman.

Is there a truth of the matter as to whether oat milk is milk? One may insist on a biological definition of milk as coming from a living mammal and, indeed, this was once the case. However, language evolves, and milk is a concept in flux. Perhaps it is currently ambiguous between a social and biological kind. If so, then used one way it refers to dairy milk and used another it refers to milk of any kind. It is easy to imagine slogans such as “oat milk is milk” or “vegan burgers are burgers.” Exploring what they mean and how they are used is an interesting analogy for the parallel with the vocabulary of gender ideology. Were the barista to ask a vegan customer whether they would like real milk, the latter would answer negatively, and while they may not be sufficiently invested as to take offence, it is not difficult to imagine someone uttering the words “real milk” with a telling sneer.

5. The XYZ analogy

Similar debates have arisen regarding whether water should be defined as “H₂0” or as something like “clear, colourless liquid that fills the oceans and in pure forms is capable of hydrating earth life and quenching thirst.” In Hilary Putnam's famous thought experiment, “Twin Earth” is said to be just like our earth, with the exception that the clear colourless liquid that populates its lakes and oceans is, upon scientific inspection, composed not of H₂0 but of some other elements, which he terms “XYZ.”Footnote ¹⁵

We may envisage people with different intuitive approaches devising the slogans “XYZ is water” and “XYZ is NOT water.” The first slogan aligns with the existing concept of water (which predates by several centuries the scientific discovery that water comprises H₂0), while the second slogan is more dependent on a more technical, physiochemical definition. While the latter may be highly important in the lab, it is of no use in everyday life. After all, were we to discover that scientists were wrong all along, we would not conclude that there had never been any water on earth but only that water was not H₂0 after all. Is either party right or wrong? Is it a question of context? Or does the unexpected discovery create a space in which it is up to us to decide whether to use the word “water” for both liquids, or whether we would prefer to disambiguate between two kinds of water?

Scientists will want to distinguish H₂0 from XYZ, though they may still do so by calling them E-water and TE-water. People may worry that the two have different properties and may also wish to know which of the two they are drinking. But for many people, it simply won't matter whether they are drinking E-water or TE-water.

We similarly distinguish between cis men and women and trans men and women. Some trans people may pass as cis, while others may not. However, “passing” is not a test of anything other than the stereotypes of those who make the judgment. Not only is it logically possible for a cis woman to fail to “pass” as one because she does not satisfy the observer's biased expectation of “what a woman should look like,” such things have actually happened. There are numerous recorded occasions in which someone tried to deny a cis woman entry to women's toilets based on the suspicion that they were trans.Footnote ¹⁶ Similarly, when Liao Mengxue and Tong Zenghuan won gold and silver medals for China in the women's 400-metre relay at the 2019 National Track and Field Championships Finals, their gender was queried with statements such as “They look like men!”Footnote ¹⁷ Surprisingly, such accusations (which were based on how the athletes sounded and appeared) were made by people who argue that what really matters is not what a trans person looks like, but whether or not they have male or female genitals.

6. The adoptive parent analogy

In the 2018 “Issues in Philosophy” series of the Blog of the American Philosophical Society, Sophie Grace Chappell proposes that: “Trans women/men are to women/men as adoptive parents are to parents.”Footnote ¹⁸ As with our own water/H₂0 and vegan cheese analogies, we think that the adoptive analogy is good as far as it goes. As Chappell herself allows, “there are disanalogies of course,” but all analogies come to an end sooner or later. The point of analogies is not to suggest that two different things are identical, but that they are similar in relevant respects. This helps to free us from pictures of how things are that have kept us captive.Footnote ¹⁹

Gender essentialists are held captive by a certain biological picture of what it is to be a man or a woman. They will argue that it is not a mere picture but a scientific fact, and they invariably dismiss anybody who rejects it as a denier of biological reality or worse. Indeed, they will complain that they are being attacked by “woke” people who want to cancel them for merely stating that sex is biologically real. However, most trans people and allies do not deny biological reality. Indeed, Chappell's adoptive parent analogy is intended to highlight the relevance of biology to certain aspects of our concepts of parent, man and woman, while demonstrating that the concepts also have a sociological aspect that cannot be fully reduced to biology. Chappell writes: “Society has found a way for [the adoptive parent] to live the role of a parent, and to be recognised socially and legally as a parent.” However, this is not to downplay the importance that biological parenthood has in certain, largely medical, contexts. Chappell cites examples of blood transfusion, organ donation, tests for inherited illness, and so on. One's concept of parenthood does not suddenly change upon entering a hospital. However, in questions of hereditary disease, it matters which kind of parent you are. The correct question for the doctor to ask in these circumstances is not “Are you their real parent?” but “Are you their biological parent?”.

The same may be said of vegan cheese and water, and the same also applies to trans men and women. There may be medical contexts in which it will matter whether they are trans and, if so, whether they have had gender-affirming surgery. However, doctors have no professional interest in metaphysics or epistemology. Contextualists concerned about the meanings of terms such as “woman” (e.g. Esa Díaz-LeónFootnote ²⁰) are thus incorrect in stating that, in some medical contexts, trans women are men. Rather, what may be relevant in medical contexts is whether someone is trans or cis.

The concept of adoptive parents is not an altogether different concept from that of biological parents. Within the umbrella concept PARENT, there are at least two subtypes: biological and adoptive.Footnote ²¹ Similarly, MAN and WOMAN may also function as umbrella concepts, with “cis” and “trans” as sub-concepts. We might imagine WATER in the Twin Earth scenario functioning similarly. The same may now arguably be applied to CHEESE.

In the Twin Earth water scenario, the linguistic change occurs as the result of a scientific discovery. In the vegan cheese analogy, it is the result of invention spurred by lifestyle and ethical behavioral changes among societal groups. Likewise, the concept of an adoptive (as opposed to biological) parent has grown out of the stream of life whereby someone who raises a child “as their own” is no longer considered a mere guardian but a bona fide parent. These linguistic changes reflect changes not only in human practices and related behavior, but in how we perceive and value the words. The same is true of the concepts of trans men and women, except that trans people have been around for a lot longer than vegan cheese (which, incidentally, dates back to the sixteenth century). Indeed, trans people are a biological reality as much as a sociological one. Their existence does not depend on the conclusions of high-flown philosophical argumentation (including those discussed and put forward here), and to deny it by dismissing it as fashion, ideology, or some form of imposture is to deny a blatant empirical truth. The person who wishes to deny that XYZ is water, that adoptive parents are parents, that plant-based cheese is cheese, and that trans women are women is arguably in the grip of a myopic perspective on how concepts work.

According to Chappell's adoptive parent analogy, biology is crucial to the distinction between cishood and transhood. Just as an adoptive parent is not a biological parent, so too a trans man is—at least in some relevant sense—not biologically male. Chappell's analogy is important because it highlights how trans-inclusivity and biological reality go hand in hand. The power of this analogy is that it also serves as a defence against trans-exclusionary feminists who maintain that trans ideology denies the reality of biological sex, as Chappell's analogy is in fact premised on this reality. We have already articulated our opposition to trans-exclusionary feminists and trans allies who (for very different reasons) reject the sex/gender distinction.

However, the fact that trans women are not natal biological women is no more a reason for denying their realness as it would be to deny the realness of adoptive parents.Footnote ²² As Chappell puts it:

This is not merely a question of good manners: there is a perfectly legitimate sense in which trans women are women, just as there is a perfectly legitimate sense in which adoptive parents are parents, oat milk is milk, and XYZ is water.

We hold that trans women are a subset of women, just as adoptive parents are a subset of parents. However, so too are cis women a subset of women and biological parents a subset of parents. There is no question of one being more real than the other. For some purposes, what matters is whether one is a biological mother or father (an issue that is itself further complicated by the existence of surrogate mothers, sperm donors, etc.); for others, what matters is whether one is a legal parent. One can, of course, be both. Here the analogy with trans people comes to an end, as all analogies sooner or later must. For while not all trans people have had their gender legally reassigned, the situations in which this might make a difference are not the same as those in which it matters who the legal parent is. However, such fine-grained differences in no way detract from the power of the analogy more generally.

Chappell's view differs from ours in one crucial respect. According to her, whether or not a trans woman is literally (as opposed to analogically) a woman “depends on what you want to talk about.” For social purposes, they are literally the same, whereas for medical purposes they are not. While we endorse Chappell's analogy, we disagree with this approach to articulating the situation. Chappell is correct in stating that there are different (but equally legitimate) senses in which one can be a woman, just as there are different senses in which one can be a parent. Which sense is the most appropriate changes from one context to another. So far so good. However, for Chappell's position to hold, there must be a sense of “man” or “woman” in which trans men and women are literally men and women, respectively, and a sense in which they are (only) allegorically so. From our perspective, however, a trans woman who has not had gender affirming surgery is not an allegorical woman in the medical context any more than an adoptive parent is an allegorical parent in the parallel case. The fact that she is not a biological woman (allegorically or literally) does not preclude her from literally being a (certain kind of) woman. Trans and cis men and women are kinds of men and women, and no particular kind is more real than another.

Chappell considers her adoption analogy to be a relatively “conservative proposal” for conceptual engineering revisionism or engineering (see §7). From our standpoint, however, the real strength in her analogy lies not in any kind of amelioration of our gender concepts; rather, it plays the far more important role of elucidating how our everyday concepts of man and woman already function for many people, demonstrating that there already exist ordinary senses of “man” and “woman” that are trans-inclusive. In sum, it does not propose a change of use but, rather, a change of aspect-perception, as Chappell invites us to see trans men and women as we see adoptive parents.

7. The naturalized citizen analogy

Let us now consider another pair of analogies offered by Chappell, namely that of (a) honorary membership of a group and (b) naturalized citizenship. Chappell relates an apocryphal story of a Provost's dog that was deemed an “honorary cat” for the purposes of the statutes of King's College Cambridge so as to allow the Provost to keep a dog in the grounds of a college whose statutes only permitted cats. Should we likewise proclaim XYZ, plant-based milk, adoptive parents, and trans women as mere “honorary” members of the respective sets to which they are said to belong? While Chappell stops short of claiming that this is “the right way to think about our gender concepts,” she allows it as a “possibility” that “is still on the table.”Footnote ²⁴ While we agree that an “honorary” status would pave the way for debates about regulations to be liberated from pointless metaphysics, we do not consider the statement “trans women are women” to be a legal fiction.

The next and final analogy we wish to consider is that of naturalized citizenship. It is tempting to regard a naturalized citizen as an honorary citizen as opposed to a real one. If so, they would be akin to an honorary cat or, perhaps, the holder of an honorary degree. In the case of honors bestowed on people, there exists the belief (illusory or otherwise) that the person in question deserves the honor. This desert does not, however, detract from the notion that an honorary degree is not a real degree and may not function as one in the eyes of, say, an employer. Should we really think of citizenship in the same terms? Or is a naturalized citizen more like an adopted child?Footnote ²⁵

Chappell uses the example of Scottishness, though any nationality would have served equally well. On one side of the debate on Scottishness are the hardcore nationalists who claim that one can only be born Scottish; one cannot become Scottish. Even within this position, there are two camps: (1) the culturalists who think birthplace is sufficient and (2) the racialists who insist on parentage as either the main or a necessary additional criterion. This further division is of interest insofar as it separates innate biology from any external features that may be present at birth.

On the other side of the debate are those, like Chappell herself, who believe that it is sufficient to identify as Scottish. In Chappell's own words, “a Scot is anyone who identifies with Scotland enough to make their home here, and to seek to join in the project of building a new nation together.” According to Chappell, this stance makes it possible for her “to decide one day that from now on I will identify as Scottish.”Footnote ²⁶ We are not convinced that a single act of identifying as a member of any nation is anywhere near sufficient for counting as a member of said group, honorary or otherwise. Indeed, there is reason to think that Chappell herself would agree with this, since she goes on to point out that, by 2015, she “had been living continuously in Scotland for seventeen years, and for another nearly three-year period a decade earlier than that,” and that she had and continues to have “every intention of living in Scotland for the rest of my life.”Footnote ²⁷ While the specifics are not being presented as necessary and/or sufficient conditions for being Scottish, we might say that living as a Scot is part and parcel of authentically identifying as one. There are pragmatic reasons why no nation grants citizenship simply on the basis of declarations, regardless of how earnest they might be. By contrast, it is unproblematic to grant someone recognition of their gender upon declaration. This is not merely down to trust but to a deep disanalogy between naturalization (which is something one earns) and gender (where there is no question of earning).

Chappell's aim, of course, is to draw an analogy between being Scottish and being a cis woman and identifying as Scottish and being a trans woman. Indeed, according to her own analogy, we ought to regard her trans Scottish and no less a real Scot for it than someone who is cis Scottish (namely Scottish by birth). While the analogy is helpful, we fear that it is also misleading in equal measure. First, irrespective of its merits or demerits in the case of nationality, the notion of identifying as a man or woman is (as we have already argued in §2) unhelpful. An additional concern with the Scottish analogy is that, while trans Scots are by definition not Scottish by birth or culture, there is no obvious parallel in the case of trans men or women. Indeed, not only is it an open question whether at least some trans people are born trans; there is also empirical evidence to support the idea that one's gender is at least in part a matter of innate disposition.Footnote ²⁸ Moreover, the testimonies of trans children (and, indeed, Chappell's own autobiographical writingFootnote ²⁹) suggest that whatever the mix of nature and nurture may be that gives rise to one's real gender—be it of a cis or trans variety—this is something that is generally set during early childhood, if not from birth.Footnote ³⁰

These concerns aside, we agree with Chappell's contention that questions of gender, like those of parenthood and nationality, are at least partly determined by “how inclusive we want to be” (it would be absurd to think that trans rights hinge on either party offering the correct analysis of such analogies).Footnote ³¹ A helpful analogy here is that put forth by Naomi Scheman in her landmark paper “Queering the center by centering the queer: Reflections on transsexuals and secular Jews,” in which cis women are likened to those who are Jewish by heritage, while trans women are likened to those who convert to Judaism.Footnote ³² Neither kind of person, she points out, is “really Jewish” or any more authentically Jewish than the other. To us, this appears to hit the nail on the head.

What the citizenship, Jewishness, and parenthood analogies all share is an insight into how our evolving practices shape our language and concepts, which in turn shape not only the way in which we think about things but what it means for anyone to authentically be included within a particular category. This is part and parcel of the stream of life. But what of attempts to tinker with the stream of life? What of conceptual engineering, which takes a “top–down” approach (from concepts to practices) as opposed to our “bottom–up” approach (from practices to concepts)?

8. Conceptual engineering

In her abstract to “Gatekeepers, engineers, and welcomers” (2021), Chappell describes her project as a form of “conceptual (re-)engineering.” Conceptual engineering is a fashionable philosophical movement that focuses on the conscious tinkering with concepts with the aim of improving them for a variety of reasons: metaphysical, pragmatic, moral, sociopolitical, and so on. In this section, we highlight some of the perils of this approach and how we think they might be best overcome. We do not intend that what follows is a knock-down argument against the very project of conceptual engineering, which probably does have several valuable aspects. Our more modest aim is to caution that it should not be pursued at the expense of ignoring what people ordinarily do with words. We thus disagree with the strong ameliorative thesis that “no matter what topic a philosopher is concerned with, they should assess and ameliorate the meanings of central terms.”Footnote ³³

Psychologically speaking, the phenomenon of conceptual engineering is a compelling one. In recent decades, tech companies have managed to persuade people to conceive of everything from VHS players to computers as things capable of thinking, believing, and wanting. Now, with the rise of electronic assistants such as Alexa and Cortana, they even have us talking of artificial intelligence (AI) “understanding” or “misunderstanding” us. As Nietzsche put it:

We no more discover that machines think or understand than we discover that Pluto is not a planet. Tech companies have not discovered that machines can “think”; rather, they have engineered the concept through multimillion-dollar campaigns to get us to unthinkingly refer to computers as “thinking.” But once the newer use of “thought” has entered vernacular language, it is futile to resist it. All one can do is distinguish between different senses of the term, just as one might with “parent,” “milk,” or “man.” Some will inevitably be etymologically parasitic on others, but this does not render the original senses any more legitimate than the newer ones. No correct use of a term is any more legitimate than any other.

Wittgenstein argues that the meaning of words is tied to the ways in which we use them, and our linguistic usage is ultimately grounded on our behavioral practices and the forms of life that underlie them: in the beginning is the deed, and the deeds in question include the making of vegan ice cream, soya burgers, aubergine steaks, mock duck, meaty not-dogs, oat lattes, vegan milkshakes, plant-based shrimp, and even “faux gras.”Footnote ³⁵ Which precise terminology will eventually stick will largely be determined by pragmatic concerns, but it is bound to share the vocabulary of the non-vegan varieties of the corresponding foods.

Meanings change alongside our practices and social agreements, and concepts are constantly reformed. In the words of Cassie Brighter:

By studying other periods and cultures, we can learn how differently gender concepts evolve. Such evolution is not purely semantic; it is also deeply intertwined with the things we choose to do with words and the multifarious purposes that underlie our linguistic behavior. Harmful linguistic practices cannot be reduced to the use of independently derogatory (or otherwise problematic) terms. It is what people do when they say things like “trans women are not women” that matters. As Nikki Ernst argues:

In the section that follows, we consider the pitfalls of a hugely popular and well-intentioned attempt to engineer the term “woman” for social good.

9. Gender eliminativism

It has become fashionable to ask what any given concept should be and then propose to modify or “engineer” it accordingly (see §8). By contrast, we maintain that—in the case of gender, at least—this approach puts the proverbial cart before the horse. Concepts typically evolve from the stream of life, whose ebb and flow affects the course of ordinary language. It is through social, cultural, and behavioral change that words come to change their meaning, and not the other way around.

In 1997, the OED introduced the term “cisgender” as an adjective designating “a person whose sense of personal identity and gender corresponds to his or her sex at birth” and “transgender,” negatively, as a person whose abovementioned sense does not correspond to their sex at birth. It's not a bad starter for ten. But to paraphrase J. L. Austin, the dictionary should always be the first word, but not necessarily the last, on any matter.

One potential problem with the OED's definitions of “cisgender” and “transgender” is its grouping of personal identity and gender as two things that go hand in hand such that we have one unified sense of them. The term “personal identity” has a troubled philosophical history which we shall not go into here; but what the OED is gesturing at is the idea that gender is something with which one identifies. This thought has been the cause of much unnecessary trouble and confusion, not least when the definition further implies that there is some kind of wedge between one's gender and one's sense of it, as though to say that being cis or trans is not a matter of actually having a certain gender but merely of identifying as having it.

It is important here to distinguish between three senses of the term “gender”: (1) grammatical, (2) natural, and (3) psycho-sociological. It is in the context of the first of these that the now famous trope of “the three genders” first arose. Thus, in the 1824 edition of his magnum opus English Grammar, Lindley Murray writes that “there are three genders, the masculine, the feminine, and the neuter.”Footnote ³⁸ According to the OED, grammatical gender refers to “each of the classes (typically masculine, feminine, common, neuter) of nouns and pronouns distinguished by the different inflections which they have and which they require in words syntactically associated with them.”

In the English language, grammatical gender was formerly closely associated with natural distinctions of sex—namely, natural gender.Footnote ³⁹ Thus, in its 1997 edition the OED still characterized the psycho-sociological use of “gender” as follows: “The state of being male or female as expressed by social or cultural distinctions and differences, rather than biological ones; the collective attributes or traits associated with a particular sex, or determined as a result of one's sex” (our emphasis). By contrast, the 2020 edition of the OED relates that, “in most European languages, grammatical gender is now only very loosely associated with natural distinctions of sex.” It is common to think that these natural distinctions have been replaced by “socially constructed” characteristics. For example, the World Health Organization states:

There are instances in which one can refuse to use a word (e.g. the “N” word). Race and gender abolitionists might belong to this school. For example, Toni Morrison rightly claims that the notion that “race” is something distinct from the human race is a fiction created by racists.Footnote ⁴¹ There is thus a very important sense in which we all need to acknowledge that there is no such thing as race. The same cannot be said of gender. Sally Haslanger writes that “both race and gender are real, and both are social categories.”Footnote ⁴² However, there is a deep asymmetry between “race” and “gender.” Gender eliminativists such as Haslanger are not interested in explicating everyday concepts. While they do not necessarily wish to change or altogether abandon our everyday uses of “woman” and “race,” they nonetheless believe that it is theoretically useful to engineer the concept of a woman so that it encapsulates the very oppression and harm we aim to eradicate.

Haslanger thus pursues an “analytical” approach to questions such as “what is gender?” and “what is race?” according to which we should be less interested in what race and gender are than in what we (should) want them to be.Footnote ⁴³ At the most general level, her stated task is “to develop accounts of gender that will be effective tools in the fight against injustice.”Footnote ⁴⁴ Our own view is that any account of X which parts ways with our everyday concepts is not really an account of X at all.Footnote ⁴⁵ Haslanger would not be remotely fazed by this. While she acknowledges that she appropriates everyday terms for rhetorical reasons, she concludes that “if someone else is determined to have those terms, I'll use different ones.”Footnote ⁴⁶ To us, this entails that she doesn't actually wish to abolish gender at all, but something quite different which she has chosen to term “gender.” As Ernst puts it, her analysis “does not aim to trace the concept we are actually using when we talk about ‘women’, i.e., the operative concept. Instead, this analysis aims to develop a concept of woman that serves as a tool for feminist theory or activism to talk about a certain class of oppressed people for liberatory political purposes.”Footnote ⁴⁷ If this is what is being abolished, then she is not in fact interested in abolishing our everyday concepts of man and woman at all.Footnote ⁴⁸

Gender eliminativists or abolitionists are on a spectrum: from a radical elimination of gender, to one that resembles an anti-essentialist stance, such as Julia Serano, for whom “there is no such thing as a ‘real’ gender—there is only the gender we experience ourselves and the gender we perceive others to be.”Footnote ⁴⁹ Judith Butler adopts a more nuanced approach, maintaining that sex is a social construct that is indistinguishable from gender. In the early 1990s, Butler appears to want to dispense with that which they call “identity categories,” such as woman, man, homosexual, or heterosexual. While they characterize them as “temporary idealizations,” they seem compelled to concede that these are “necessary errors” and that “a double movement” is needed: “To ameliorate and rework this violence, it is necessary to learn a double movement to invoke the category and, hence, provisionally to institute an identity and at the same time to open the category as a site of permanent political contest.”Footnote ⁵⁰

In contrast to all gender eliminativists as well as gender-suspicious minimalists, such as Butler, we defend a biological realism that leaves space for a concept of gender that is distinct from that of biological sex. We are opposed to the elimination of gender concepts, for both pragmatic and existential reasons. The pragmatic perspective is succinctly expressed by Lori Watson, who sees gender elimination as a false hope for the elimination of gender inequality:

We agree that gender eliminativism is not a viable solution to “the false sex dualism and binary gender hierarchy we all live in.”Footnote ⁵² Not only is it an option, it is safe to say, that most cis people would reject, but it is also one that many trans people would reject. Indeed, one usually “transitions” to a particular other gender. But were the elimination of gender a possibility, should we opt for it? Would it not simply leave us with the bland alternative that we are all, and only, persons? This would suppress—negative gender norms notwithstanding—a whole range of vivid, lived, and enjoyed differences in human life.

The problem is not in the gender differences, but in our acceptance of them and in our acceptance of the reality that conceptual boundaries are porous and transitable. In this paper, we have sought to demonstrate that the concept of gender is no different from any other concept in being open. Concepts are not bounded by necessary and sufficient conditions but rather encompass overlapping and criss-crossing similarities, none of which is all-defining and all of which are susceptible to open-ended, vigorous and sensitive extension. For such extensions to occur through our naturally everyday interactions, we must continue to alter the conditions for gender perception. This can only be achieved through the normalization of LGBTQ+ people. However, as Katy Steinmetz rightly says, “visibility is important. But visibility without empathy is just spectacle.”Footnote ⁵³

We began this paper with a consideration of the relationship between labels and reality. How do words acquire their meanings? Who gets to decide what words such as “man” and “woman” mean? We proceeded by way of analogy to demonstrate how changes to everyday practices affect the ways in which words are used across time. Focusing on gender terms in particular, we argued that words like “man” and “woman” have no fixed meanings, let alone meanings that are determined by purely biological facts. For this reason, we conclude that arguments that proceed by way of necessary and sufficient conditions for being a woman (e.g. “adult human female”) will inevitably reach a cul-de-sac. However, we equally reject the view that we should engineer concepts to suit our moral, social, or political needs. Concepts emerge and develop from within the stream of life. Their meanings arise as a result of the various ways in which we use words—that is, from the multitude of things we do with them. One of these things is to extend the meaning of a term through both linguistic and non-linguistic practices. This may in itself be regarded as a form of conceptual engineering, but isn't. It is neither bottom–up nor top–down; real conceptual changes stem not from the theoretical armchair but from the streets.