Legal Fictions: Changing Sex by Changing Gender

by Miroslav Imbrišević

A couple of years ago the question ‘What is a woman?’ entered the public imagination. Germaine Greer said in 2015 that transgender women are not women, and the novelist Chimamanda Ngozi Adichie said in 2017: “A trans woman is a trans woman”. But the controversy about who falls under the category ‘woman’ doesn’t just originate with trans activists and feminist theorists – the law (in the UK) must also take some of the blame.

It may come as a shock to people to find out that the law sometimes relies on falsehoods – legal fictions – in order to promote justice. This is reflected in the legal maxim: fictio legis neminem laedit – a legal fiction doesn’t injure anyone. This means in practice that the law (or a court) allows statements to be made which are strictly false, and everyone involved knows this. This differs from outright lying, where someone is being deliberately deceived.

In company law, for example, we treat a corporation (in some respects) as if it were a natural person (a human being). This makes it possible for people to enter into contracts with a corporation or to take the corporation – a legal person – to court. A corporation has rights and duties, just like a natural person.[1]

We find legal fictions in Roman law, in the common law, but also in continental legal systems. There is broad agreement about their usefulness, but there also is the odd dissenting voice. Jeremy Bentham calls them: ‘the most pernicious and basest sort of lying’.

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. The law then treats somebody or something ‘as if’ it were something else (or the opposite). An early example can be found in the Lex Cornelia (81 BCE) in Roman law. Roman citizens who died in captivity lost their status as free citizens – they were slaves. This meant that they did not have the capacity to make a will. According to the Cornelian Law such citizens should be treated as if they had died free men; in this way their will would be valid.

In 2004 the UK government introduced new legislation (the Gender Recognition Act, GRA) to help people suffering from gender dysphoria, so that their lives may go better. This legislation allowed people to change their gender legally and to obtain a Gender Recognition Certificate (GRC). But the Gender Recognition Act relies on a legal fiction:

Where a full gender recognition certificate is issued to a person, the person’s gender becomes for all purposes[2] the acquired gender (so that, if the acquired gender is the male gender, the person’s sex becomes that of a man and, if it is the female gender, the person’s sex becomes that of a woman).

There is a little bit of confusion in the above quote, because we normally use ‘male’ and ‘female’ to refer to sex (biology) and ‘man’ and ‘woman’ refer to the gender role (social presentation). By ‘acquired gender’ the law makers mean the gender role of ‘man’ or ‘woman’ (which is usually associated with ‘male’ or ‘female’ sex). One of the requirements before applying for a GRC is to have lived ‘in your acquired gender’ for two years, meaning that the lawmakers expected you to have adopted another gender presentation and changed your first name to accord with that gender during this time.

Although someone was born as a boy, having male sex markers and registered as such, they can apply for a new birth certificate (once they get a GRC) which will state their sex to be ‘female’ (and vice versa for people who transition to be recognised as a man). This is a legal fiction. The law treats a transwoman for all intents and purposes ‘as if’ she belonged to the female sex. But the transperson’s sex at birth hasn’t changed, neither has it been assigned wrongly. The midwife ascertains male or female sex markers, rather than a gender identity. The latter wouldn’t have been formed at the time of birth anyway, and if it were innate, as some claim, it couldn’t be detected in a newborn.

Why create this legal fiction? We associate a particular sex (female) with a particular gender (woman). If the law permitted a change of gender (woman), but insisted that the sex of birth (male) could not be altered, this would perpetuate the pain of people who suffer from gender dysphoria, or rather, body dysphoria.[3]

The legislation gives transpeople the option to change their gender, to be recognised and treated as a woman (or man) in society. But the legislation cannot change the biological facts – irrespective of any gender re-affirming surgery the transperson may have. Though some people claim that it isn’t just the gender role which is a social construct, it is also your biological sex. Consequently, some transwomen claim to be literally (i.e. biologically) women. A variation of this view is that one’s ‘gender identity’ determines one’s sex. But then the notion of ‘sex’ as biology (and distinct from ‘gender’ as the social role) becomes meaningless. Gender and sex appear to be the same thing.

But the law in the UK still makes a distinction between women (born female) and transwomen (born male). And by doing so the law acknowledges the legal fiction contained in the GRA from 2004. A transwoman (with a GRC) acquires a similar (but not identical) set of legal rights and duties as a woman. For example, transwomen are not subject to abortion legislation, because they cannot give birth[4] – but transmen are.

Subsequent legislation confirms that the 2004 GRA created a legal fiction. The rights of transwomen are constrained by the UK 2010 Equality Act. At present the ‘exemption clause’ in the Equality Act protects sex-based rights: it permits female-only spaces. This means that ‘discrimination’ (i.e. drawing a distinction) can be lawful when it has a legitimate aim – here, to protect someone on the grounds of their sex. For example, it would be lawful to exclude a transwoman from a group counselling session for female victims of sexual assault (Explanatory Notes to the Equality Act 2010: 157). This illustrates that the makers of the Equality Act acknowledge (implicitly) that the GRA relies on a legal fiction. If transwomen were literally (i.e. biologically) women, if there were no difference between them, then the exemption clause wouldn’t make any sense.

Such an exemption also holds for sports which use male and female categories. Transpeople may be excluded from competing in their newly acquired gender in order to insure ‘fair competition’ or to guarantee ‘the safety of competitors’. This again is an acknowledgement of the legal fiction created in the GRA.

This doesn’t mean that all transwomen are really men. It means that those who socially and/or medically transition, and thus acquire some markers of being a woman, are transwomen. They have changed their gender, but not their sex at birth.[5]

Some trans activists and organisations in the UK are now demanding the removal of the exemption clauses from the Equality Act. If this were to happen, it would turn fiction into reality. Then all sex-based protections and female-only spaces would have to be open to transwomen. Transwomen athletes could compete in the female category. Female prisons would have to accommodate all transwomen prisoners, etc.

Rather than promoting the legal fiction that transwomen are literally women, we should try to create a third category for transwomen (or transmen) in those areas of public life, where their demands would disadvantage women and clash with the hard fought-for rights of women. Keeping the sex-based exemptions envisioned in the Equality Act is important because it reminds us that the 2004 GRA relies on a legal fiction. You can change your gender (presentation), but you cannot change your biological sex.

Miroslav Imbrisevic is a philosopher, formerly of Heythrop College/University of London. His background is in legal and political philosophy, but recently he has started working in philosophy of sport as well.


[1] But in US law a corporation does not have the right against self-incrimination.

[2] The phrase ‘for all purposes’ is short for ‘to/for all intents and purposes’, going back to English law in the 16th century. It means: ‘in every practical sense’.

[3] Many, but not all, US states permit a change in sex designation on your birth certificate.

[4] Advances in medical science may soon allow transwomen to give birth though – by creating a ‘birth sack’ or some such.

[5] There are transpeople who present as men (which is in accord with their sex at birth), but claim that their gender identity is ‘woman’. They don’t appear to experience gender or body dysphoria. It isn’t clear in what sense they are ‘trans’ (-itioning).






36 responses to “Legal Fictions: Changing Sex by Changing Gender”

  1. Peter Smith

    So we have become legally fixated on legal fictions. The Red Queen’s advice to practice believing impossible things before coitus is beginning to make sense.

  2. Excellent paper. Concise, accurate and articulate; and timely.

  3. Alan Henness

    Excellent. Very clear and helpful.

    The GRA certainly mashes terms up a bit but tries to sidestep the disjoin between sex and gender by using the term ‘acquired gender’, but the sleight of hand takes place on the Gender Recognition Certificate itself where it declares the holder’s ‘gender’ to be female or male, ie it assigns a gender as if it was a sex. The Registrar then uses that to swap the sex recorded on the birth certificate to the sex the person is not.

    I think I’m right in saying that the only people who could possibly be regarded as having a ‘legal gender’ is someone with a GRC? No other official document uses that term as far as I’m aware (other than, unfortunately, the Equality Act 2010 when referring to ‘gender pay gap’ – but I suspect the drafters baulked at calling it the ‘sex pay gap’!).

  4. Oh, dear! I guess I committed a faux pas. A previous essay denied there were enough differences between the sexes for gender to be worth talking about. I replied with examples of what I thought were significant gender characteristics that the majority of people, heterosexuals, experienced in themselves and their partners and took delight in. No one agreed or disagreed. Now a second essay appears serving the interests of the same small minority of people who welcome sexual differences being blurred. Clearly that small minority is in the majority here, and I am out of place. Or has philosophy in general been captured by the gender-blurry? Is sexual difference no longer to be celebrated? Well, perhaps, but not here. I get it.

  5. I have no idea what you are talking about.

  6. And the essays don’t “serve” anyone’s interests. They are either written by our regular contributors or sent to the magazine as submissions. If they are well enough written and argued, they get published. I think you’re spending way too much time inside your own head.

  7. JVP

    I agree — excellent paper. However, I don’t know how laws (or philosophy) can be precisely written without always pairing the word ‘gender’ with some other word for disambiguation. For example, not “gender” but “gender presentation” or “gender stereotype” or so-called “gender identity.” As if gender is always gender-qua-something.

  8. ombhurbhuva


    You write-

    There is a little bit of confusion in the above quote, because we normally use ‘male’ and ‘female’ to refer to sex (biology) and ‘man’ and ‘woman’ refer to the gender role (social presentation).

    No we don’t unless we are in the habit of speaking in that irritating manner of ‘in quotes’ which calls into question what we ascribe to a view. For me the initial encounter with man/woman, male/female came within my family. That is my normal. The new normal is a complete fabrication out of gender dysphoria. And did we actually think it would stop at a well meaning attempt to make people afflicted by this condition more comfortable? In general a healthy society should stop caring so much about hurt feelings and just tell the truth.

  9. Miroslav

    Thank you Barry!

  10. Miroslav

    Thank you Alan! I think you’re right. Only poeple with a GRC have a ‘legal gender’.

  11. Alan Henness

    Have you seen this supposedly academic (and well-funded) project?

    The Future of Legal Gender – A critical law reform project

    I tried to get an explanation of ‘legal gender’ from them but failed. As Descartes said, “First, define your terms…”

  12. Thanks for that link Alan!

  13. “And it’s really even wrong to call this transition because gender identity is legal fiction”

  14. Great article! I saved it on Wayback Machine, just in case.

  15. ombhurbhuva:

    Yes we do. There has been a long history of differentiating between biology and gender in our language use. Examples: When we fill out many kinds of government or medical forms, they want us to fill in ‘male’ or ‘female’, not man or woman. (I think this is what Miroslav was getting at with the quote from the GRC Act). We make the distinction all the time when we refer to chairman or chairwoman, not chairmale or chairfemale, and fireman, policemen, not firemale or policemale. Women sometimes complain about men ‘mansplaining’ things to them. They never say ‘malesplaining’, because it’s meant as a stereotype of a masculine character trait, not a biological trait. When we commit the ultimate sin of using the word ‘mankind’ in reference to all humans, we do not literally mean only biological males. Even if one were a chauvinist caveman (not a cavemale) and had a man cave (not a male cave) and believed all humans were once divided into strict roles of Man the Hunter, and Woman the Gatherer, one would still never say, Male the Hunter and Female the Gatherer.

    There may be exceptions, but in general we tend to reserve the words ‘male’ and ‘female’ for biological contexts, and ‘man’ and ‘woman’ for gender/social contexts. And don’t get me started on the French language where nearly every noun has a gender (not a sex).

  16. davidlduffy

    “The hallmark of a legal fiction, according to Fuller, is that it is not intended to deceive”. By contrast, it seems to me that the GRC is supposed to have performative force (switching from a legal to a more philosophical vocabulary). Rather than the judicial type of legal fiction, it is more like the idealizations made in universal declarations or bills of rights. Injecting a relatively recent concept like gender into the existing body of law can’t be that easy – the main advantage being that only a small proportion of the population are affected, until one gets into distributional tails, like in elite sports. ISTM that the core “theory” is the idea that where existing laws referred to the sex of an individual, they were referring to a legal concept rather than a biological concept all along. The argument that opposite-sex marriages unproductive of children were still legally recognised, so this should mean SSM have the same legal status does not involve legal fictions, though it could have (eg required one person to be the bride). Does the analogous argument for gender transition require it?

  17. ombhurbhuva

    It might be useful to try and trace the history of the present muddle about gender. It used, not long ago, to be a term in grammar. There were masculine, feminine, and neuter nouns so called by convention. They might as well have beeen called x, y, and z nouns. Some literal minded people were puzzled by the masculine and feminine appellation. How was ‘patria’ feminine? How is ‘cailin’ (girl) masculine in Irish? Is this early Celtic gender confusion? In other words there was an analogical extension from ‘feminine’ to female. Sex as Vilfredo Pareto often remarks is a very powerful residue.

    As with the conventional ascription of gender in nouns the notion of the conventional applied to male and female arose. Each society has its folk ways, roles, duties and so on. True and useful; however to go on from there and to equate gender (masculine/feminine) as convention with sex (male/female) as convention is a category error. Of course it is also true that male and female conventional roles arise out of sexual differences.

    So where are we now? Essentially this: I like what is conventionally supposed to be feminine so therefore I am a woman. On the other hand you are the rightful King of France.

  18. Thank you David. Not sure what you are asking here: “Does the analogous argument for gender transition require it?” Require what?

  19. Miroslav,
    I think David is asking the following: Recognizing that rights accrue to a Sae0Sex Marriage does not require generation of a legal fiction. The baseline argument is that marriage is a contract between two people, and that the motives behind the agreement of the partners to such a contract ought not be restricted in such a way as to prohibit the pursuit of happiness the partners are engaged in by way of the marriage. The counter argument is that marriage is something more and other than a contract, – which, had it been decided by the SCOTUS (as dissenting justices wanted) would redefine the marital laws of every state in the union as legal fictions. This is congruent with a certain trend in Fundamentalist religious thought that has shadowed the history of American law since at least Dred Scott – namely that the real social glue here has been determined by Biblical doctrine, and that the Constitution is itself mere legal fiction – a point of view Michael Pence essentially endorsed in a campaign speech some years ago.

    At any rate, assuming that the Constitution is our foundational social contract, and that state marital laws, all contractual in nature, are intended to legitimate promises and benefits to the participants, granting them legal recourse in the event of breach, it follows that marital contracts cannot be ‘legal fictions,’ and the question of the right of the participants in contractual agreement does not require the participants to be legally defined except as parties to the contract.

    So, I think David is asking, is it possible to make an argument legitimating gender transitioning in a manner that doesn’t require construction of identity as a legal fiction, but rather derives from some other established legal principle; for instance, possibly based on the implicit right of self-identity/ self-expression, such as we find in the right to free speech in the First Amendment.

    I think you have largely answered that question as far as legal developments in Great Britain are concerned. In America, the matter has different problems to consider, but in general, I would suggest to David that such is not really possible. The problems include (but are not limited to) the following: First, gender expression is largely a matter of dress and behavior. At this point in history, laws prohibiting trans-gender dress or behavior have lost all force, except in certain situations. In other words, except in these situations, the problem of trans-gender dress or behavior no longer has any legal status. There is still social tension concerning it, sometimes leading to public confrontation, even violence in some localities; but the legal right to dress and act as one chooses in such matters is largely assumed.

    However, the situations where the matter is not so clear are precisely the ones giving us political head-aches. Should men trans-gendering as women be granted the right of access to women’s locker-rooms? Allowed to enter women’s sports as competitors? Demand to be addressed as women rather than as men. Etc. I think there the lesson from Britain is clear – Yes, ‘woman,’ as a legal fictional relabeling of the given individual would eventually prove necessarily. So far, trans-sympathetic progressive, including Barack Obama, have tried to maneuver the political debate around that eventuality, emphasizing it’s derivation from the right of self-expression. But such a foundation provides no criteria for adjudication in situations where the presumed right of self-expression conflicts with others’ rights, such as privacy or assured security in a locker-room. Only the legal construction of gender as a legal fiction would accomplish that.

    However, this is not the kind of legal fiction the American legal system likes. It would probably necessitate a costly licensing process; potentially invasive policing procedures; unprofitable law suits. the keeping of voluminous records in both public and private sectors… legally requiring educational courses for employees in trans-gender etiquette would be the least of it.

    Most issues that have developed such requirements and procedures and record keeping – such as the privilege to drive an automobile, or the rights of women in the workplace – have developed fairly organically over time, or have presumed rights or privileges already in existence, and only requiring further enforcement. But this, as I say, would be to create an entirely new legal class. Regardless of the political resistance to this, institutional inertia would weigh against it.

    Attenuating the problem, of course, is that many trans-activists have adopted extreme positions in these matters, and are not really arguing for privileges or even rights – they demand total acceptance of their self-identity claims, individually, culturally, institutionally. Legal adoption of their positions would only be the start of it.

    Let me end on a side note: It should be noted that as long as trans-gendering is the issue, and not trans-sexuality (the claim that one is not of the biological sex one is born with but somehow the opposite), the issue was for many centuries resolved by way of transvestism. An interesting aspect of the current twist in these problems is that ‘transvestite’ is now a term that can only be applied to men; for the simple reason that women have been dressing and acting “like” men for so many decades now that it is hardly noticed anymore (remember “tom-boys”” whatever happened to them?)..

  20. davidlduffy

    Dear Miroslav.

    EJ below has already expanded in a useful fashion, but I was alluding to the ideas that legal fictions:

    1) are a work-around that can be usually replaced by appropriate legislation
    2) are supposedly seen by all involved as a fiction – as opposed to an incorrect belief on the part of the legislature or judiciary

    As I understand it, the UK law arose following Goodwin v UK and I v UK in 2002, where the European Court of Human Rights considered the existing legislation infringed rights of transsexuals to respect for their private and family life, and to be able to marry as they would. The ECHR’s argument was something like: a) gender reassignment as a medical procedure is well accepted (and NHS funded); b) one’s legal status following this was ambiguous in the UK; c) privacy around these matters was being undermined by this ambiguity revealing information to employers etc. The claim that such privacy protects one from discrimination specific to being transgendered was made by Goodwin, but wasn’t taken up the ECHR. Some of the same kind of considerations re privacy and discrimination come up in:

    So, just taking the alteration of birth certificates and other records so that others cannot discern that you are presenting yourself as a sex different from that at birth. I can’t see an easy alternative that would maintain privacy (per my point 1), and I don’t know if all involved see it as a fiction, but rather as the case that person X now lives the social and sexual life of a woman or man, and this is being recognised and supported by the law. Now, the 2004 legislation was seen as “well beyond any requirement of European law”, but the choices made do reflect the practicalities of gender reassignment – as a medical treatment they take some years, surgery can be unsuccessful, maybe someone does want the possibility of backing out and of having children “naturally”.

  21. Thank you David. Will give this some thought.

  22. Marc Levesque

    Enjoyed, got me thinking, and a lot of good comments.

    “Though some people claim that it isn’t just the gender role which is a social construct, it is also your biological sex”

    I think that’s partly true. Biologically I see male and female as two peaks on a continuum so I think there’s convention involved in a binary framing of the sexes.

  23. Alan Henness

    What is the parameter that defines this continuum?

  24. Marc Levesque

    I’d say maleness and femaleness, for each individual it would include all the biological factors that count towards that.

  25. Alan Henness

    How do you believe that could be measured and quantified? And can it be done without resorting to regressive, demeaning, sexist stereotypes?

  26. Alan Henness

    This ‘masculine/feminine’ score sheet allegedly comes from Ernst & Young. Is this the sort of thing you mean?

    See: Women At Ernst & Young Instructed On How To Dress, Act Nicely Around Men

  27. Marc Levesque

    I mean factors at the biological level, up to and including developmental biology, meaning factors like genetics, chromosomes, hormones and their interactions throughout development. I don’t mean anything like what Ernest and Young seem to be suggesting.

  28. Alan Henness

    That’s good, but what factors, what measures, what scale? What is it that gives ‘two peaks on a continuum’?

  29. Marc Levesque

    “That’s good, but what factors, what measures, what scale? What is it that gives ‘two peaks on a continuum’?”

    I’m glad for your questions. I haven’t been very clear. I’ve imagined it for a while without ever really trying to explain it or put into words. And I’m not sure I understand what you mean or expect by factors, measures and scale in relation to a continuum.

    Conceptually, I think the two peaks can be seen as biological types, male and female, with individuals mapped according to how their relevant biological characteristics relate to each type. The distributions are continuous and there won’t be a sharp cutoff between the two.

    A few weeks ago I listened to a talk on natural kinds between Daniel Kaufman and Massimo Pigliucci, and Massimo who’s a biologist seemed to refer to a similar idea as the one I’m trying to convey here. The whole talk was really interesting:

  30. Alan Henness

    The problem is of definition. Imagine these two peaks of yours drawn out on a chart. The vertical axis is some measure of the strength or quantity. The horizontal axis is the parameter of the continuum: that which is changing as you move along the axis. Your peaks are drawn there, with some distance between your two peaks. But they are measures of something: something that gives a count or other measure on the vertical axis and the thing being measured along the horizontal. Without numbers on these axes, your two peaks are no more than meaningless squiggles. For them to be meaningful, they have to mean something: they seem to be showing a distribution, but of what? What are they showing? What is it that these peaks represent? If you can’t say what the axes are, it’s difficult to see how the chart shows anything that can be measured and if you can’t measure it, how can you differentiate peaks?

    It’s too late at night to start watching that long video – I’ll try to take a look tomorrow, but I hope it doesn’t try to claim that chromosomal and endocrine variations give rise to something between male and female: they give rise to developmental differences within the two sex classes.

  31. Marc Levesque

    I gather you can’t see what the kind of thing I’m aiming for but I think I have a better idea of how you’d like things expressed and comments are about to close so maybe it’s best I think about it some more for now.

    “I hope it doesn’t try to claim that chromosomal and endocrine variations give rise to something between male and female: they give rise to developmental differences within the two sex classes.”

    Massimo’s a solid biologist, and I’m not sure how he’d respond to that, but can you tell me what you mean biologically by sex classes.

  32. Marc Levesque

    An example marker would be testosterone, if you graph it (on the x axis) by individuals (on the y), male and female individuals will form two peaks with a fair amount of overlap between the two clusters.

  33. Marc Levesque

    I meant to say “if testosterone *levels* are graphed on the x axis…”