by Miroslav Imbrišević
Thy wish was father, Harry, to that thought.
–Shakespeare, Henry IV
Making Language Better
If we can improve our language, by making concepts (roughly: the meaning of words) more precise or by naming something so far unnamed, we should do so. This is what motivates “conceptual engineering,” a relatively new branch of philosophy. 
The engineers of language and thought don’t just restrict themselves to making words more precise, they also want to make them “better.” The latter is called ‘conceptual ethics’. It’s not just about what words mean; it’s about what words should mean.
Conceptual engineering has been taken up by some feminist philosophers. A central concept in feminist philosophy is ‘woman’. Ordinarily it means “adult human female,” but some feminists would like to include transwomen under the term ‘woman’. This view is now widely accepted in academic feminism. If you dare to question this, you will be considered “transphobic,” as Kathleen Stock, a philosophy professor at the University of Sussex, has experienced.
Subsuming transwomen under the class ‘women’ doesn’t makes the term more precise, nor does it name a new phenomenon. We do have names to refer to people who are not happy with their biological make-up. Formerly we called them ‘transsexuals’, now they are ‘trans people’ (‘transwomen’ or ‘transmen’). For these feminist philosophers, conceptual engineering has a moral/political aim. In this case, to minimise the exclusion of trans people in society. This is a worthy aim, but it is difficult to achieve it by changing the meaning of words.
By including transwomen under the umbrella ‘woman’, these well-meaning philosophers suggest that there is no real difference between the type of women in expressions like ‘young women’, ‘German women’, ‘married women’, ‘happy women’, ‘single women’, ‘tall women’ and ‘transwomen’ (or ‘trans women’). They are all women. The aim here is to shape reality; to change how we view the world.
Natural Kinds and Social Kinds
The class of women – by which I mean, adult human females – can be understood as a natural kind; that is, as something that is part of nature. There is a material, biological reality to it. The class of transwomen, on the other hand, is a social kind; that is, something we find in society. It is a notion that is socially constructed. We invented it. It relies on the idea that some people have a gender identity which can be in conflict with their sex. Male-bodied persons wish they were female or believe that they are female, and many want to express this through their gender presentation, which might include body modification. The natural kind term ‘woman’ refers to a material reality (sex), the social kind term ‘transwoman’ refers to a psychological reality (attempting to disregard your sexed body). For this reason, the expression ‘gender dysphoria’ is misleading. It should be called ‘sex dysphoria’ instead. So, the term ‘gender dysphoria’ may need to be “re-engineered.”
The social category ‘woman’ has a biological foundation: women, understood as a natural kind. The social kind supervenes upon the natural kind; that is, the social category has an underlying material basis: being of the female sex. There is nothing similar with respect to being a transwoman. In this case, one social kind (transwoman) supervenes on another social kind (woman). Transwomen, in this sense, represent a kind of supervenience squared; a supervenience of supervenience. And because the concept ‘transwoman’ is free-floating, without a tether (a female sexed body), there is a fundamental difference between women and transwomen.
Take the concept of marriage. We now accept that same-sex attracted people can get married. Our linguistic (and legal) practice has changed and with it the concept of marriage. But ‘marriage’ is a social kind term, something we created by agreement, and we can extend/alter its meaning through further agreement. Contrast this with the concepts: ‘tiger’, ‘water’ or ‘woman’. These three are natural kind terms and their concepts are not open to radical revision through our linguistic practice, because they are tied to how things actually are in the world. There are facts about tigers which we cannot alter. For example, we cannot simply decree that it would be good to class lions among the tigers. Admittedly, they have something in common: they are both big cats, but ‘tiger’ and ‘lion’ are distinct concepts, as tigers and lions are distinct species.
Sometimes we get the facts about a concept wrong. We used to think that whales and dolphins were fish, but now we know that they are mammals. Our conceptions of whales and dolphins are now more precise. Similarly, we may in future find out more about the endocrinology of women and men, which will make our concepts ‘women’ and ‘men’ more precise. But we will not find out that biological males who self-identify as women are female, because we would be confusing social kinds with natural kinds. These biological males may outwardly present as stereotypical women or have some body modifications done, but this does not make them into a woman, if what we mean is an adult human female. To put it differently, the direction of fit for concepts like ‘tiger’, ‘water’ or ‘woman’ is from world to mind. Contrastingly, the direction of fit for concepts like ‘marriage’, ‘game’ or ‘good manners’ is from mind to world. What we think about marriage over time will determine and constitute what marriage is.
Humans and mammals in general, come in two reproductive classes or “shapes”: female and male (known as sexual dimorphism). This is a fact of nature, independent of the societies we live in. The slogan “transwomen are women” equates a social kind with a natural kind, which is a category mistake.
Alternatively, if the term ‘woman’ in the above slogan referred exclusively to the social role females stereotypically play, then it would be expressing something trivial: transwomen play, or attempt to play, the same – stereotypical – social role as many women. Does this make them into women?
Trans activists and their supporters in feminist philosophy deny that there is such a clear-cut dichotomy between the male and female of the species. For their political purposes it would be ‘better’ if people believed that ‘sex is a spectrum’ or that ‘sex is socially constructed’. People with ‘sexual development disorders’ (DSD) are roped in to support the thesis. The reasoning goes something like this: biological sex is not a clear-cut criterion for determining who is a woman, because there are intersex people, who don’t fall neatly into the male/female binary. From this ‘refutation’ of the biological basis for being a woman, trans activists conclude that transwomen belong in the class ‘women’. But this reasoning is faulty. The existence of intersex people doesn’t prove that you can be a woman without being of the female sex. It also doesn’t prove that there is a third sex (or many more sexes); being trans would then be another (fourth?) ‘sex’ which is beyond the binary.
Intersex people are being used by trans activists for their propaganda purposes. They claim that it is a common condition, and this has been uncritically perpetuated by many organisations and in the media. The original source for these claims is Anne Fausto-Sterling, who stated that 1.7% of the population are intersex. But being intersex is only one of the many conditions within the DSD range. You can only arrive at this high a figure if you (wrongly) assume that all people who have DSDs are like intersex people. But most people with DSDs do fall into the female/male dichotomy. According to Leonard Sax ‘the true prevalence of intersex is seen to be about 0.018%, almost 100 times lower than Fausto-Sterling’s estimate of 1.7%’ (see also here).
Sally Haslanger’s Account
There are several accounts, attempting to include transwomen under the concept of ‘woman’. Let’s look at one prominent one. Sally Haslanger proposes that someone is a woman, if they are systematically subordinated in certain respects (economic, political, legal, social, etc.), and if they are targeted “for this treatment by observed or imagined bodily features presumed to be evidence of a female’s biological role in reproduction.” Conversely, you are a “man” if you are systematically privileged along some dimension (economic, political, legal, social, etc.), and if you are targeted “for this treatment by observed or imagined bodily features presumed to be evidence of a male’s biological role in reproduction.”
Haslanger’s definition of ‘woman’ is metonymical. It focuses on one consequence of being female: experiencing oppression. But it isn’t that one consequence of being a woman that makes you a woman. Being a woman precedes or underlies the consequence of oppression. And Haslanger admits that the discrimination women suffer is based on being perceived to have a certain body. So, it isn’t the oppression that makes you a woman, but rather, the target of that oppression: having a female body.
If transwomen are perceived to have a female body and face discrimination because of this, then they are women for Haslanger. Under this concept the Queen of England and other women who don’t experience discrimination would not be women. And men who are not privileged wouldn’t be men. But let’s leave this aside.
The discrimination transwomen experience, based on their perceived/imagined body, isn’t uniform. Those who “pass” as women will face some of the same oppression as women, but those who don’t pass (or don’t wish to pass) experience a different form of discrimination. They are treated (oppressed, discriminated against, mocked) differently, based on being (perceived to be) male-bodied. So, the metonymical understanding of ‘woman’ is limited to those who pass. This means that only some transwomen are women.
Those who don’t pass or don’t wish to pass as women are perceived to be male-bodied. They are not systematically privileged as men, but systematically discriminated against because of their male bodies. So, on Haslanger’s account they are neither men nor women. That is an odd side-effect of trying to be inclusive.
Three Types of Conceptual Engineering
There appear to be three ways we can engineer a concept. We can do it from scratch (e.g. ‘sexual harassment’, ‘gaslighting’, or ‘mansplaining’); we can change the meaning of an existing term: ‘woman’ (through expansion, limitation, clarification, etc.); and we can alter parts of an existing term. An example for the third way of engineering is the expression ‘sex assigned at birth’. The term ‘assigned’ suggests an arbitrariness of the process and/or that newborns come with a fully formed gender identity, which has not been correctly recognised. But all of this is misleading. Sex is normally observed and registered at birth by medically qualified people. And this is a fairly reliable process. Sex is not “assigned.”
Conceptual engineers distinguish between semantic amelioration (improving informational content) and epistemic amelioration (improving our knowledge or grasp of reality). The first way of engineering, creating new concepts like ‘gaslighting’, is often successful in both respects. The second way of engineering, improving the concept ‘woman’, aims for a semantic amelioration for the benefit of transwomen, but not necessarily for women. As an epistemic amelioration it fails, because it erases the difference between sex and gender. The third way, changing part of a phrase as in ‘sex assigned at birth’, fails semantically and epistemically. The same holds for the terms ‘cis woman’ and ‘non-trans woman’.
At first glance, it may appear that the law is also changing the meaning of words. Someone who is biologically male can be legally recognised as a woman. The law permits that the sex on the birth certificate (and other documents) be changed to “female.” But the law (here in the UK) does not actually engage in conceptual engineering. In this context, it doesn’t change the meaning of words, it merely changes the legal status of a person. The law cannot make a man into a woman, but it can prescribe that (for the purposes of the law) it will treat a biological male as if he were a biological female. This is known as a legal fiction, an ancient device going back to Roman law. In practice this means that the law (or a court) allows statements to be made, which are strictly-speaking false, and everyone involved knows this. Lawyers aren’t worried about this because fictio legis neminem laedit: a legal fiction doesn’t injure anyone. 
The law creates legal fictions when novel circumstances or societal change may lead to pressure to fit new phenomena into a pre-existing framework, because it is presumed that this will result in some social benefit and it would accord with the purposes of the law. Here, we have some overlap in purpose between legal fictions and conceptual engineering for moral/political change. However, there is an important difference. The law is open about treating biological males as if they were women or female, whereas the conceptual crusaders want to effect a change in the way we think and in the way we view the world.  They want us to believe that a transwoman is a woman, in the same way that a tall woman is a woman.
Two Types of Woman
The trans-friendly feminist wants people (but particularly women) to accept that the concept ‘woman’ has two sub-sets: women and transwomen. This is odd (a paradox?), because women become a sub-set of their own class. The conceptual engineers use a little trick to hide the paradox: they re-label the recalcitrant sub-set ‘woman’ by giving it a prefix. The class ‘woman’ then consists of the sub-sets ‘cis women’ and ‘trans women’ (also popular among the crusaders is the pairing: ‘non-trans women’ and ‘trans women’).
The re-engineering of the term ‘woman’ has had a ripple effect on other concepts, as well as giving rise to new expressions. The meaning of the term ‘lesbian’ has (allegedly) changed, once again introducing two sub-sets into the definition. Now a lesbian is: a female or a transwoman who is sexually attracted to cis women and transwomen. The ‘and’ conjoining cis women and transwomen in the dependent clause is significant. It suggests that lesbians are naturally attracted to trans “lesbians.” If we replaced the ‘and’ with ‘or’, it would at least give lesbians the freedom to choose.
This new understanding of the concept ‘lesbian’ has given rise to a neologism: the “cotton ceiling” (invented by the trans porn actress and activist Drew DeVeaux). This metaphor alludes to the “glass ceiling” of old and to women’s cotton underwear. Transwomen find it hard to convince lesbians (i.e. same-sex attracted females) that they should also be attracted to transwomen, particularly if they retain their male genitalia. So-called “trans lesbians” are frustrated by not being able to break through the cotton ceiling. There is wide agreement that the glass ceiling needs to be broken, but the cotton ceiling is designed (engineered) to make lesbians feel guilty for their alleged moral failings. Just like the glass ceiling needs to be broken, so does the cotton ceiling. (Note the echoes of the incel movement here.) In this example, there is also a normative element (a value judgement) engineered into the concept, in addition to the semantic and epistemic elements: i.e. You are a bad lesbian, if you don’t want to help trans lesbians to break the cotton ceiling.
Finding a solution to the problem of the “cotton ceiling” has required further conceptual engineering in the form of new coinages. Many transwomen retain their male genitalia and often refer to their penis as a “lady dick” or a “girl dick.” This is supposed to help them in their struggles with the “cotton ceiling.” 
There is no doubt that coining the term ‘sexual harassment’ in the 1970’s has helped women to articulate their oppression. Using the terms ‘lady dick’ or ‘girl dick’ may make some “trans lesbians” feel better about themselves, but it is at the same time a form of gaslighting (another useful coinage to counter hermeneutical injustice) of lesbians. It suggests to them that they should not just be attracted to females but also to transwomen. Note that the conceptual ethics at work favours the perspective of “trans lesbians” to the detriment of lesbians. It suggests to the latter that they are wronging “lesbians with penises,” if they don’t find them attractive.
But let’s look at it from a different perspective. If you don’t accept that transwomen are women, but are trans-identified males instead, then the thought of having sex with a “trans lesbian” would mean to have intimate relations with the opposite sex. And, politically, it would mean that you are having sex with your oppressor. (Another problem for Haslanger?) Why should lesbians agree to this?
Some transmen who give birth wish to be recorded as the father of the child on the birth certificate. The meaning of the term ‘father’ would then include: ‘person giving birth to a child’. This would have seemingly paradoxical consequences: The birth-giver would be ‘father’ (legally) and ‘mother’ (biologically) at the same time. This would make the concepts of ‘father’ and ‘mother’ less precise and might actually cause confusion. So far, British and French courts are resisting these attempts at conceptual engineering.
Some transwomen claim to experience period pains. They describe their symptoms as being moody and wanting chocolate. This is a re-conceptualisation of the term ‘period pain’. The experience may be real enough for them, but it is a non-veridical experience. It doesn’t map onto reality, which involves shedding the lining of your uterus. This is also metonymical in that it takes certain symptoms to be the cause of the experience.
There is a Price to be Paid
You would think that the conceptual engineers in philosophy would give us an ungendered reality. But what we actually get is a transgendered reality. And some gender-critical feminists would argue that if you dig deeper, you’ll find a male perspective underneath the transgendered reality. Breaking through the “cotton ceiling,” for example, evokes images of sexual violence.
The feminist conceptual re-engineering of the word ‘woman’ fails women. It doesn’t consider how such engineering infringes the rights of women (for example in the contexts of sport or prison), and it doesn’t consider the ripple effects this will have on other concepts.  The law is a much better – and proven – driving force for change in society. Think of the abolition of slavery, giving women equal rights, the de-criminalisation of homosexuality, introducing same-sex marriage, etc. Admittedly, the law is often slow to recognise the necessity for societal change, but good things take time and deliberation.
Some feminist philosophers, in their enthusiasm for improving concepts, have overlooked that language is a network of interconnected terms. The lesson is: if you want to tinker with language, you need to look at all related concepts, and you need to consider the impact this will have on people. The conceptual crusaders in academia (and trans activists) will not have to pay the price for their tinkering. It is female athletes, women prisoners, and candidates for female-only shortlists who will bear the brunt of making concepts “better.”
As a playful exercise, let me try a bit of conceptual re-engineering myself. I would suggest that a “transman” is a male who is transitioning (or has transitioned) to being a “woman,” either in appearance and/or self-identifying as such. The converse would apply to the term ‘transwoman’. A transman would then fall within the extension of ‘man’, and a ‘transwoman’ would come within the extension of ‘woman’. The advantage of my re-conceptualisation is that it is more coherent, more precise, and it is less likely to cause cognitive dissonance – conflicting attitudes, beliefs or behaviours – among language users. Now the slogan ‘transwomen are women’ would actually (literally) be true. But, most importantly, my terms don’t cause the kind of ripple effects that would stretch related concepts to breaking point (like ‘lesbian’ or the invention of the ‘lady-dick’, etc.). Wishing you all happy engineering!
 The German-born philosopher Rudolf Carnap is considered to be an early exponent – or perhaps further back: Socrates?
 But if people start to believe the legal fiction (transwomen are women), this may harm or disadvantage others: women.
 The exemptions in UK legislation with regard to trans people confirm the legal fiction. If there were no difference between women and transwomen, we wouldn’t need exemptions. For this reason, trans activists want to abolish the exemptions. It may actually be useful if the law stated clearly, in every instance, that it had created a legal fiction. This would stop people from believing that the fiction is true: for example, that you can change your sex.
 Recently Joanna Harper has coined the term ‘cement ceiling’ in response to World Rugby’s ban on transwomen athletes at the elite level: https://www.nytimes.com/2020/10/26/sports/olympics/world-rugby-transgender-women.html?smid=em-share ; Harper’s expression suggests that World Rugby want to exclude transwomen from the sport altogether, but this is misleading.
 There is another ripple effect when it comes to statistics. Many police forces in Britain have succumbed to conceptual engineering. They record sexual crimes (like rape) in accord with the gender self-identification of the perpetrator, regardless of whether they have a Gender Recognition Certificate or not. This will distort statistics – more ‘females’ will be sex offenders. But language is also affected. Contrary to the definition of rape in English law (non-consensual penetration with a penis), a ‘woman’ could then rape another woman.