Conflating and misgendering: why World Athletics (and other sports governing bodies) should jettison the competitive labels ‘Women’s’/‘Men’s’

ABSTRACT Martínková et al provide an overview of a tendency to use gender terms in key sports contexts, including eligibility criteria and testing, where gender is unintended. They argue that to avoid conceptual confusion and aid clarity, we should disentangle gender and sex, acknowledging that often gender talk should be interpreted as talk of sex. One of their recommendations is that the labels of competitive categories ‘women’s’/’men’s’ should change to ‘female’/’male’. I first make their argument against gendered labelling more precise by showing that important yet neglected moral and practical reasons support their abandonment. I then argue that in the case of WA regulations, those moral reasons also cut against Martínková et al.’s relabelling proposal ‘female’/’male’. I sketch a testosterone-based proposal which circumvents the problem and which WA itself should accept. More generally, I argue we should be more mindful of the risk that competitive category labels unnecessarily harm athletes.


Introduction
make the observation that gender and sex are often unhelpfully conflated by sports organisations and sports researchers. In particular, they claim that competitive eligibility regulations routinely sort athletes by sex, yet the resulting competitive categories are labelled 'women's' and 'men's'. The problem according to Martínková et al. is that the latter are gender terms and thus inappropriate; the more appropriate sex-related labels 'female' and 'male' should instead be used to describe the categories. Additionally, the verification procedures that determine which sex category athletes should be allowed to take part in are often misdescribed as 'gender verification' procedures, instead of being called what they in fact are -sex verification processes. Martínková et al. argue that avoiding this tendency to misapply gender terminology would aid conceptual clarity. My aim here is twofold. Firstly, I argue that relinquishing gender labels for sports categories receives support from other grounds that Martínková et al. do not discuss: most prominently, gendered labels have the morally problematic consequence of needlessly misgendering certain trans athletes, so there are good moral grounds to reject such labels. Additionally, abandoning gendered labels for sports categories would have practical and dialectical advantages: it would avoid some unnecessary antagonization of opposing parties in the debate on the eligibility of trans athletes and athletes with differences of sex development (hereafter 'DSDs'), thus making the prospect of a solution less distant; and it would also help sports governing bodies state their current policies in a less philosophically objectionable form.
My second aim is to critically scrutinise Martínková et al.'s suggestion that the current 'women's'/'men's' labelling is best replaced by the labels 'female'/'male'. As a case study, I focus on the controversial World Athletics (hereafter 'WA') policies governing the eligibility of trans athletes and athletes with DSDs. In this context, I argue contra Martínková et al. that in the specific case of these WA policies, their proposal faces similar pitfalls to the ones affecting the gendered labelling it is intended to replace: while it does not misgender certain athletes, it disrespects certain athletes by missexing them. I sketch an alternative labelling proposal based on testosterone levels that circumvents the problem.
While I argue that this proposal is superior both to the gender-and sex-based labelling proposals, in the sense that it doesn't disrespect athletes' gender identity and sex, it should be emphasised that mine is a claim about the labelling of the categories given a set of eligibility rules, and not a claim about the acceptability of the eligibility rules themselves. One can claim, as I do here, that renaming competitive categories in athletics by using testosterone-related terminology is the best labelling option, given the current regulations in force, while remaining neutral on the acceptability of the underlying regulations themselves. In other words, my paper can be read as arguing for a conditional claim:if one thinks that the eligibility criteria set by WA regulations are broadly correct, then one ought to jettison the current 'women's'/'men's' labelling in favour of testosterone-based labelling. Thus, my argument is compatible with the view of opponents of current regulations -for example, those who believe that the sport categories themselves ought to be defined in terms of gender, with all athletes participating in the competition of the gender they identify (e.g., Gleaves and Lehrbach (2016), Ivy and Conrad (2018)). My argument points to a need for change in category labelling that those who endorse current regulations (presumably, WA itself) should recognise and act upon, and is primarily directed towards such sports governing bodies.
But from the specific case of WA, a more general lesson should be drawn, one that is relevant to any sports governing body: my argument can be read as drawing attention to the risk for any set of eligibility regulations of disrespecting certain athletes when the labels used to describe competitive categories differ significantly from the actual eligibility criteria that delineate the categories themselves, particularly when the labels concern characteristics of an athlete, such as their gender and sex, deserving special respect.

Martínková et al. (2022) on the conflation of sex and gender
Before proceeding, a brief sketch of Martínková et al.'s views is needed. Their paper provides a useful overview of the history of the distinction between sex and gender within and outside of sports, to which I refer the reader for a fuller appreciation. But for the purposes of this paper, and in a nutshell: the distinction emerged originally from John Money's research into intersex conditions -now called 'differences (or variations) of sex development'and has been influential particularly within feminist movements from the second half of the twentieth century.
In brief, sex can be understood as a characteristic that is determined by biological or anatomical features (chromosomes, genitalia, and other bodily markers). Sex categories include 'female' and 'male', with some taking these to be exhaustive, while others argue instead that people with DSDs motivate the further category 'intersex' and/or a spectrum view. 1 By contrast, gendersometimes glossed as 'the social expression of sex' -can be understood as determined by self-identification or other social factors, such as norms of behaviour, social roles, upbringing, that may bear some connection to sex, but that are not fully determined by it. Gender categories include 'woman' and 'man' ('girl' and 'boy' for younger humans) and gender pronouns include 'she'/'he' and 'her/him'. The need for further gender categories beyond 'woman' and 'man' is increasingly recognised. While sex and gender are not unrelated, those who endorse the distinction guard against simple inferences which traditionally commanded widespread assent: for example, the inference from 'X is female' to 'X is a woman', and vice-versa; and the inference from 'X is male' to 'X is a man', and vice-versa.
The distinction is theoretically and politically significant for at least two reasons. First, in breaking the automatic inference from 'female' to 'woman', it has lent force to the feminist observation that the underprivileged social roles to which women are disproportionately consigned are not an inevitable fate, in the way one might think one's sex is. In other words, once we recognise that being a woman is merely a social expression of being female, space opens up for thinking of ways of being a woman that are different from traditional ones (being a housewife, raising children, doing domestic chores) in which being female was traditionally expressed. In de Beauvoir's famous words 'one is not born but becomes a woman ' (1949, 330) through upbringing and social norms -norms that with sufficient political will can be changed for the betterment of women's social standing. 2 Second, the distinction opens up the necessary space for trans identities to exist and be legitimised. If being female can in principle come apart from being a woman, then it is possible to have female anatomy and simultaneously be a man; or to have male anatomy and be a woman. These and other transgender identities, including non-binary identities (neither woman nor man), are not possible on the traditional picture that takes sex and gender to be binary and essentially interchangeable, and that takes their subcategory pairs (female/male and woman/man) to be exhaustive.
Martínková et al. advance compelling arguments for the view that in sports, gender terminology is often used when sex is intended. In particular, it is clear that what goes by the name 'gender verification' -the processes used to ensure that only eligible athletes are allowed to compete in what are routinely called 'women's' events -makes its determinations by examining anatomical features of the athletes in question, and not social roles, selfidentification, or norms of behaviour to which athletes subscribe (140-142). This constitutes for Martínková et al. conclusive reason to believe that the relevant competitive categories are sex categories, not gender categories.
After arguing that gender talk is all too often inaccurately used when sex is intended instead, Martínková et al. make the following recommendations (145): • In contemporary sport, the gender category should be the sex category • 'Gender verification' should be called 'sex verification' and gender should not be relevant to sex verification procedures. • 'Men's'/'women's' categories should be replaced by 'male'/female' categories, and this change should be reflected in related language in eligibility regulations, competition names, and in sports research and journalism • Depending on eligibility rules, the male category can include athletes who are women, and the female category can include athletes who are men. 3 With this overview of Martínková et al.'s contribution in hand, I will first argue in section 3 that their critical point against gendered terminology can be strengthened in two important ways. Firstly, I claim that there are good (but somewhat neglected) moral reasons to jettison the gendered labelling of categories, since doing so brings significant moral gains, in the form of avoidance of a certain type of harm. Secondly, and relatedly, avoiding these harms removes an exacerbating factor in the vexed debate over the inclusion of trans athletes and athletes with DSD. These demonstrate that misapplying gender terminology in particular sports contexts is a matter to be taken even more seriously than Martínková et al.'s piece might initially lead us to believe. In section 4 I will show that similar moral considerations cut against Martínková et al.'s proposal to replace the gendered labelling of sports categories with the labels 'female'/'male', by focusing on current World Athletics policies. In section 5 I sketch an alternative labelling proposal based on testosterone levels that does not have the moral costs that genderand sex-based labels incur. To repeat, my argument does not endorse the current WA regulations based on testosterone, but merely claims that given such regulations, those who propose them must avoid the moral harm of disrespecting athletes' protected characteristics, most prominently gender and sex, by ensuring that category labels better reflect the actual eligibility criteria.

The moral downsides of inaccurate gender terminology
Why is it important to avoid the conflation of sex and gender in sports? Martínková et al. address this question twice in their paper, the first time quickly, and the second slightly more thoroughly: In those cases [sports and medicine], it is vital to distinguish between sex and gender, so that we know what we are talking about. (135) And: The proposed terminological distinction between sex and gender (along with associated vocabularies) may help the sport community to speak more appropriately about athletes and their eligibility for the sex category. Whether we accept theories that say sex and gender mutually influence each other or not, or whether or not sex is socially constructed, the vocabulary with respect to sex and gender needs to be respected. Sports federations need to be clear what their sport categories are about. Whatever type of categorization eligibility officials decide to use, they must use the terms that reflect their verification procedures; and this is not always happening currently. Further, it would be desirable if some consistency might be established between sporting codes and organisations, so that the language and message is consistent throughout. (145) Thus, for Martínková et al., the primary aims that accurate terminology would fulfil are those of conceptual clarity and consistency. These are admirable aims. But I argue that there are important moral aims that are served by correcting sports terminology and that are not appropriately emphasised by their discussion. In particular, imposing an inaccurate gender framing unnecessarily risks harming certain groups of athletes by misgendering them. Misgendering is the act of using gender terms that exclude people from their gender category, and is a harm particularly felt by trans people (Kapusta 2016) and some people with DSDs.
To illustrate, consider the following three types of case. Firstly, athletes of any sex may identify as non-binary, that is, as neither belonging to the gender category 'woman' nor to the gender category 'man'. However, insofar as sports competitions provide no category beyond those labelled 'women's' and 'men's', the participation of non-binary athletes is conditional on their being misgendered.
Secondly, consider a track athlete with DSDs who identifies as a woman, who is not androgen-insensitive, and whose testosterone levels exceed the threshold required for taking part in the 'women's' competition. Suppose that in terms of performance, she is perfectly capable of competing in the 'men's' category. Again, her participation is conditional on her being misgendered as a man.
Thirdly, consider an athlete with male anatomy who has recently started to identify as a woman. She is not taking testosterone-suppressants that might allow her to comply with the eligibility criteria for taking part in the 'women's' competition, and in any case prefers to take part in the more competitive and challenging 'men's' category, for which she is eligible. She can only do so by being misgendered. 4 In all three cases, the fact that gender terminology is used to label the categories gives rise the implication that all participants belong to the gender corresponding to the label. For the three athletes described, this implication is false. Because it is false, it is an instance of misgendering. The misgendering occurs in virtue of the fact that the categories bear the labels 'women's' and 'men's'. The perpetrator of the misgendering is the sports governing body that sets eligibility criteria, and labels the resulting categories in such a way that the participation of athletes in the category for which they are eligible leads to a mismatch between the gender identity of the athlete and the gendered label of that category. If the categories bore labels that did not refer to gender, but more accurately reflected actual eligibility criteria, no misgendering would occur.
But, one might object, is misgendering all that concerning? Is it correct to say that it's harmful? A critic might concede that misgendering may well prove offensive to those on the receiving end, but claim that to describe it as a harm is an exaggeration. 5 While we ought not exaggerate harms, there is evidence that the consequences of misgendering damages its targets in a way that is best categorised as morally harmful. For example, misgendering has been understood (Kapusta 2016, 504) as a type of microaggression: an 'everyday verbal, nonverbal and environmental slight, snub or insult, whether intentional or unintentional, that communicates hostile, derogatory or negative messages to target persons based solely upon their marginalized group membership'. Theses slights 'may on the surface appear quite harmless, trivial, or be described as small slights', but 'have a powerful impact upon the psychological well-being of marginalized groups) (Sue 2010, 3). Individual microaggressions cause negative emotional or behavioural reactions, butmore importantly -continued and cumulative exposure to these acts can bring long-term serious harms, including 'stress, anxiety, depression, high blood pressure, insomnia, substance abuse, eating disorders, social withdrawal, suicidal ideation, and post-traumatic stress disorder' (Friedlaender 2018, 8).
Additionally, the specific microaggression of misgendering brings further distinct harms. It undermines the self-respect and identity of social groups who are often already the targets of oppression and gender policing. In addition to contributing to broader social gender oppression, in contradicting the gender with which the target identifies, misgendering may also hinder our collective understanding of gender identity -a wrong dubbed by Miranda Fricker (2007) 'hermeneutical injustice'.
These harms are bad enough when one is misgendered by an individual. But when the perpetrator of misgendering is a sports governing body or institution, matters are exacerbated, for two reasons: firstly, the misgendering occurs in a visible and public setting -a sporting competition which may include hundreds, thousands or even millions of viewers, and many competitors and sports officials, thus magnifying the wrong; secondly, and relatedly, the misgendering is legitimised in the eyes of spectators, since it is perpetrated by an organisation with a substantial degree of power, influence and authority in its area of operation. 6 I have argued that misgendering of the kind triggered by gendered sports category labels genuinely cause harm to misgendered athletes. But suppose a critic deemed the above arguments unconvincing, and still thought that the gendered labelling at most offended rather than harmed misgendered athletes. Nonetheless, there is good moral reason for sports governing bodies to move away from the misgendering labelling.
By way of analogy, consider an imaginary society where members of a particular group can be denoted by two terms: 'squeltz' and 'meep'. Suppose that 'squeltz' is deemed acceptable to all, but that 'meep' is considered by many squeltzes to be a slur. However, the matter is contested: some members of society believe, despite squeltzes' public claims to the contrary, that 'meep' is perfectly acceptable. Suppose you are in charge of a government economic task force that is devising new tax policy. You and your advisors have agreed that the income of squeltzes should be taxed at a higher rate than the current one -a policy that, understandably, is unlikely to be welcomed by squeltzes. One of your advisers drafts policy to this effect, using the term 'meeps' to pick out the relevant group, and labels the relevant tax bracket 'meeps'. You point out that the term is deemed to be a slur and will be perceived as harmful and offensive by squeltzes. He replies that in his view, there is nothing wrong with 'meep': squeltzes might think it is harmful and might find it offensive, but they are mistaken. And squeltzes do not have a right to avoid being called 'meeps'.
I submit that in this case, it would be morally problematic to agree with your advisor to phrase the policy using the controversial term 'meep'. This is so even if you and everyone on the task force happened to hold the view that the term 'meep' was unproblematic. Simply knowing that some squeltzes found the term offensive would suffice to make it compelling from a moral viewpoint to avoid using 'meep'. The reason is that there is a readily available alternative, 'squeltz', which nobody considers problematic, and which could be readily used instead. So to use 'meep' instead of 'squeltz' in the tax policy is to cause offense gratuitously. There may well be cases where some group's taking offense is a price that must be paid in order for some important social aim to be achieved. But where the offense is unnecessary and easily avoidable, not avoiding offense is morally disrespectful.
An additional, pragmatic benefit also arises from utilising the term 'squeltz' instead of 'meep'. Suppose that tax policy has already been contested by squeltzes in the past, and that efforts of devising a policy agreeable to squeltzes have been particularly vexed. Squeltzes' willingness to agree to the tax policy is thus an important element of the policy's success. To use the term 'meep' in policy will predictably have the effect of further antagonising squeltz, and making the acceptance of the policy a more distant prospect.
I submit that the situation is analogous in sports: even if sports governing bodies believed that there is simply nothing wrong with misgendering athletes, the mere fact that misgendering is perceived as offensive by its recipients in combination with the ready availability of alternative phrasing that would not misgender athletes provides compelling moral reason for this misgendering to be avoided. To insist on phrasing policy with misgendering language would constitute a failure of respect to misgendered athletes. In addition to this moral case, avoiding misgendering would have the abovedescribed practical benefit: not further antagonising athletes who already deem current eligibility regulations problematic for other reasons.
Before moving on to examine how the above considerations apply in a concrete case, a connected point deserves clarification. Some might hold the view, strongly contested by trans communities, that 'woman' means something like 'adult human female', and 'man' means something like 'adult human male '., 7,8 Those who hold this view may insist that sex and gender are coextensive. Does my argument presuppose that this stance is mistaken? Could one resist the need to change the labels of competitive categories by taking this view on sex and gender?
The answer to both questions is no. The moral reasons I offered for moving away from gendered labels should be persuasive even to those who hold that sex and gender are co-extensive. Even if in their eyes, gender and sex terms are interchangeable, defenders of this view still have moral reason to abandon the gendered labelling of competitions, since, as discussed above, policymaking organisations have a compelling moral reason not to cause harm or offense to one's constituents unnecessarily. Because there are nonmisgendering alternative labels available, to continue using gendered terms would be, morally speaking, the wrong course of action.

A case study: world athletics' eligibility regulations
In the light of previous discussion, let's examine current eligibility regulations by World Athletics (WA). I focus on WA because their regulations, which have in recent years been the subject of much debate, are especially prominent, and because by examining them a point of criticism of Martínková et al.'s argument arises. Discussion will help bring out three important points: (i) the extent to which WA conflates sex and gender in its regulations; (ii) that by its own regulations, WA is already committed to avoiding misgendering, and should therefore unquestioningly adjust its regulations' phrasing; (iii) that for WA, there is a better alternative to Martínková et al.'s proposal of replacing the gender terms 'women's'/'men's' with 'female'/'male'. Let's proceed in order.
The WA regulations currently governing trans athletes and athletes with DSD are 'Eligibility Regulations for the Female Classification' (World Athletics 2019a) -hereafter 'ERFC' -and 'Eligibility Regulations for Transgender Athletes' (WA 2019b) -hereafter 'ERTA'. The ERFC concern athletes with DSDs who are sensitive to androgens and whose blood testosterone level is, as a result of their DSDs, at least 5 nmol/L. Of an athlete falling in this category, these regulations state: (a) She must be recognised at law as either female or intersex (or equivalent); (b) She must reduce her blood testosterone level to below five (5) nmol/L for a continuous period of at least six months (e.g. by use of hormonal contraceptives); and (c) Thereafter she must maintain her blood testosterone level below five (5) nmol/L continuously (i.e. whether she is in competition or out of competition) for so long as she wishes to maintain eligibility to compete in the female classification in Restricted Events at International Competitions (or to set a World Record in a Restricted Event at a competition that is not an International Competition) (WA 2019a, 4) Moving to the ERTA, they require the following of a transgender athlete who wishes to compete in the female category: 3.2.1 she must provide a written and signed declaration, in a form satisfactory to the Medical Manager, that her gender identity is female; 3.2.2 she must demonstrate to the satisfaction of the Expert Panel (on the balance of probabilities), in accordance with clause 4, that the concentration of testosterone in her serum has been less than 5 nmol/L continuously for a period of at least 12 months; and 3.2.3 she must keep her serum testosterone concentration below 5 nmol/L for so long as she wishes to maintain her eligibility to compete in the female category of competition. (World Athletics 2019b, 5-6) Sex/gender conflation is observable even in these very brief (yet crucial) passages. For example, in the first passage quoted from ERFC, clauses (a)-(c) use the feminine gender pronoun 'she', thus betraying an expectation that any relevant athlete will have the gender identity 'woman'. The appropriate gender pronoun 'they/them' would best replace 'she/her' throughout, since it remains neutral on the expected gender of any athlete covered by these regulations. In the passage quoted from ERTA, clause 3.2.1 requests a declaration that the athlete's gender identity is female: but 'female' is a sex term and not a gender identity, so the request should be either for a declaration of gender identity 'woman', or -more plausibly -for a declaration of female sex. There are several other passages in WA's regulations where such conflation is evident. For example, on the very first page of ERFC we read: Because of the significant advantages in size, strength and power enjoyed (on average) by men over women from puberty onwards, due in large part to men's much higher levels of circulating testosterone, and the impact that such advantages can have on sporting performance, it is generally accepted that competition between male and female athletes would not be fair and meaningful, and would risk discouraging women from participation in the sport.' (WA 2019a, 1, my emphases). 9 In sliding from the gender terms 'men' and 'women' early in the sentence to 'male' and 'female' later in the same sentence, and then back to the gender term 'women', WA conflates sex and gender, thus implicitly endorsing the view that the two are interchangeable. As discussed above, this view is considered problematic because it does not allow for the possibility of trans athletes (e.g. male women or female men). It also suggests that the gender categories 'men' and 'women' are exhaustive, thus neglecting non-binary athletes.
The quoted passages, then, show conclusively that WA regulations conflate sex and gender. WA is in ample company -Martínková et al. point to several sports governing bodies whose literature exhibits a similar conflation (135). What is particularly interesting is that in its own regulations, WA explicitly expresses an intention not to question athletes' gender identity, and suggests that doing so is required to respect their dignity. The ERFC state: In no way are they [i.e., the ERFC] intended as any kind of judgment on or questioning of the sex or the gender identity of any athlete. To the contrary, the IAAF [now WA] regards it as essential to respect and preserve the dignity and privacy of athletes with DSD (World Athletics 2019a, 2).
Similarly, the ERTA state: In no way are they [i.e., the ERTA] intended as any kind of judgment on or questioning of the gender identity or the dignity of any Transgender athlete.
WA are thus in this puzzling situation: their eligibility regulations conflate sex and gender, and the labels of their competitions misgender the athletes whose eligibility is at issue. And yet WA simultaneously proclaim that they do not intend to question the gender identity or disrespect the dignity of the relevant athletes. But since misgendering clearly involves questioning or disrespecting gender identity, WA are failing in their own intentions. The solution is simple: in order to bring WA's actions into alignment with its own intentions, the regulations should be rephrased and WA should relinquish the gendered labelling of its categories of competition. Given its commitments, the need for this change should be uncontroversially accepted by WA.
I mentioned earlier that by avoiding misgendering, a pragmatic advantage would follow -trans athletes and athletes with DSD would not be unnecessarily antagonised. But a further advantage of a dialectical nature would also accrue to WA: WA's regulations would be immune from one common criticism, which seeks to afford all trans women access to the same competitive categories as cis women -a move which WA believe ought to be resisted. 10 Trans women are understandably concerned with being afforded the same rights that cis women enjoy, including access to safe spaces (e.g. women's shelters), facilities (changing rooms, bathrooms) and, in sports, access to the competitive category labelled 'women's'. But if WA were to eliminate sex/ gender conflation from its regulations by shedding their gendered labelling of competitive categories, then WA would no longer (falsely) imply that all athletes who take part in 'women's' competitions are women. What is called the 'women's' category is not in fact currently open to all and only womenin particular, it is not open to all cis women since some cis women, those with DSDs and resulting high testosterone levels, do not comply with ERFC. Therefore, a trans woman's exclusion from the competitive category when she does not meet ERTA's requirements could not be understood as denying trans women a right to access that is afforded to all cis women. Declining a trans woman access to this competitive category would be legitimate by WA's lights (assuming she does not meet the relevant criteria), on the grounds that the categories do not reflect gender, but rather something else, such as sex and/or testosterone levels. In other words, the adequately amended regulations would not be denying to trans women a right to participation in a particular competitive sport category that cis women automatically enjoy. Relinquishing the gendered framework would help WA circumvent this charge, thus clarifying and strengthening the philosophical standing of WA's regulations. 11 What should WA change its labels to? Martínková et al. recommend that the the labels of sports categories be changed from the current gender terms 'women's'/'men's' to the sex terms 'female'/'male'. I agree that by their own lights, WA should move away from gendered terminology, but I believe that what exactly should replace the current gender terms is a matter requiring further discussion. There may well be better alternatives in the offing, certainly for WA, and potentially for other sports governing bodies, depending case-by-case on the precise nature of their eligibility rules.
At least for WA, the 'female'/'male' labels do not reflect the related eligibility rules fully accurately (though they reflect WA's eligibility criteria more accurately than the gendered labels to be jettisoned). Why? Because the 'female'/'male' categories will still trigger false implications regarding some athletes' sex. Example 1: on Martínková et al.'s proposal, WA regulations would have an athlete who is intersex take part in the 'female' category if they have a relevant DSD and suppress their testosterone level (see clause (a) of the ERFC quoted above, which explicitly allows for athletes who are legally recognised as intersex); or in the 'male' category otherwise. Example 2: a trans woman athlete with male biology who reduces her testosterone levels below the 5 nmol/L threshold for the relevant period is eligible to compete in what Martínková et al. propose to rename the 'female' category. So just as the gendered labels give rise to misgendering, sex labels give rise to what could be called 'missexing'. While Martínková et al.'s sex categories are undoubtedly progress (sex, unlike gender, is at least a key property that features in WA's eligibility criteria), labels that do not give rise to false implications concerning protected characteristics, including sex, would be preferable. Could this be achieved? If so, how?
Here I sketch one proposal. To understand this, let's first summarise which athletes are currently allowed by ERTA and ERFC to compete in the 'women's' category: (i) Cisgender women athletes without testosterone-affecting DSDs, or who are insensitive to increased testosterone level caused by DSDs (e.g. polycystic ovary syndrome); (ii) Athletes legally recognised as female or intersex who have DSDs that increase testosterone, who are sensitive to testosterone, but whose testosterone level is reduced below 5 nmol/L for the relevant period; and (iii) Trans women athletes (i.e. natal males) whose testosterone level is reduced below 5 nmol/L for the relevant period.
All other athletes can only compete in the 'men's' category. My proposal is prompted by a commonality between athletes in (i), (ii) and (iii): levels of testosterone below the threshold of 5 nmol/L. My recommendation, then, is for WA to relabel the categories 'Low-Testosterone ' and 'High-Testosterone' (or, for short 'Lo-T' and 'Hi-T'). 12 The advantages of this proposal are dual. First, it retains the advantage that sex has over gender: testosterone is (unlike gender, but like sex) a property which actually features in the eligibility criteria of ERTA and ERFC. This makes the labels at least partially informative. Second, it avoids reference to sex or gender, thus avoiding the problematic consequences of misgendering and missexing.
This proposal arguably does not give rise to false implications. All athletes who take part in the 'Lo-T' competition have a low blood level of testosterone, where 'low' is understood as <5 nmol/L. (Note that even females with DSDs who are insensitive to testosterone have a level of at most 4.8 nmol/L, with a 99.99% confidence limit (Handelsman, Hirschberg, and Bermon 2018)). All athletes who are excluded from the 'Lo-T' category have testosterone levels above 5 nmol/L, since the normal male range extends from 6.9 to 34.7 nmol/L. Thus, this proposal is superior, in the context of WA regulations, to Martínková et al.'s 'female'/'male' proposal and to the current gendered labelling 'women's/'men's'.
Nonetheless, an objection might be that here, too, there is the possibility of false implication. This could come about in two ways: first, might there not be a male athlete with exceptionally low levels of testosterone, beneath the 5 nmol/L threshold? This is unlikely, but perhaps possible in principle. In this case, they would be forced to compete in the 'Hi-T' category, which would falsely suggest a higher level of testosterone. Conversely, there is nothing in the regulations that excludes athletes who are eligible to compete in the 'Lo-T' category from competing in the 'Hi-T' category (though stronger competition is often a disincentive). So if such an athlete chose to compete in the 'Hi-T' category, again the implication would arise that this athlete has testosterone levels above 5 nmol/L, which is not true.
Two replies here are in order. First, note that these cases are far-fetched, more so than the realistic cases of misgendering and missexing discussed above. Second, even if false implications arose in these very rare cases, they would not be implications that risked causing harm or offense to the relevant athletes. The reason misgendering and missexing are so damaging is that gender and sex form an important part of people's conceptions of themselves. To be ascribed the wrong gender or sex thus can cut particularly deep. By contrast, it is not part of people's self-conception that their blood testosterone level are above or below a certain threshold: most people are completely unaware of what their blood testosterone level even is. So if any false implication arises from this alternative labelling, it is not one that risks causing harm or offense to the relevant athlete.

Conclusion
I have argued for three claims. First, Martínková et al.'s argument against gendered labels in sport can be strengthened by appeal to moral and practical reasons: gendered labels misgender several athletes, unnecessarily antagonise them, and hinder the articulation of current eligibility regulations in their most defensible form. Second, at least in the case of WA regulations (but potentially also for other governing bodies' regulations), the moral reasons to avoid unnecessary harm and offense that cut against gendered labelling similarly cut against Martínková et al.'s proposed relabelling of the categories to 'female'/'male'. Third, relabelling the categories in terms of testosterone levels does not give rise to misgendering or missexing, and is thus preferable to a labelling system expressed in gender or sex terminology.
A specific and a general conclusion can be drawn. The specific conclusion is that if WA's eligibility criteria are deemed acceptable, WA has a moral obligation to make a change, even by its own lights: it should avoid the unnecessary misgendering that the current labelling of categories entails and change its labels. More broadly, our discussion reveals an important desideratum for all sports regulations: no matter what eligibility criteria are set in sports, we should ensure that the labelling of the resulting categories avoids needlessly harming athletes by disrespecting their protected characteristics. 13 Notes 1. Estimates of the number of people with DSD range from 0.018% (Sax 2002) to as much as 4% (Sanders, Carter, and Goodacre 2011). This variation partly stems from disagreement over which conditions should be subsumed under the category DSD. 2. Note, however, that not all feminists endorse the sex/gender distinction. For a prominent exception, see Butler (1999). 3. I have paraphrased their recommendations to make them more concise. Note that Martínková et al. speak of male and female 'sub-categories' rather than 'categories'. Their rationale is that the relevant category is sex (as opposed to gender), and that since male and female are two types of sex, it is best to call them 'sub-categories'. I depart from this usage partly for simplicity, and partly because their rationale here is not compelling. Virtually all categories can be subsumed under a category of greater generality. It is not inappropriate to describe a certain geometric figure as falling under the geometric category 'square', even though 'square' is strictly speaking a sub-category of the category 'quadrangle'. 4. A reviewer observes that the second case and third cases are not realistic. The second, because it is unlikely that the woman in question would be able to compete at the highest level against natal males. Perhaps, as the reviewer helpfully suggests, the example gains plausibility if the athlete is read as competing at lower levels. The third case is allegedly unrealistic because there are no trans women who want to compete in men's events, but who do not do so over a desire not to be misgendered. By way of reply, we should bear in mind that a change of gender identity does not necessarily (and certainly doesn't immediately) entail a change of gender presentation, so it does not seem unlikely that some athletes who compete in the 'men's' category identify as women. (Again, perhaps this is more plausible at grassroots level). This possibility seems realistic, and that is all that is required for the case to be probative. At any rate, my argument that gender labels gratuitously bring the harm of misgendering to some athletes is clinched as long as at least one of the three cases is plausible. I think that even critics will find the first case realistic. Thus, my point here is clinched even if one harbours strong doubts about the plausibility of the second and third cases. 5. Thanks to Jim Parry for pressing me to consider this point. 6. I should emphasise that these are exacerbating and not intrinsic factors: an athlete who is made to participate in a competition that has a gender label different from their own gender suffers the harm of misgendering, regardless of whether many (or any) spectators, competitors or officials themselves misgender the athlete. Thanks to a reviewer for pressing me on this point. 7. See Bogardus (2020) and Byrne (2020Byrne ( , 2021 for recent articulations of this view. For a countervailing view, see Dembroff (2021). 8. I am grateful to Jon Pike, whose remarks prompted me to clarify this point. 9. A nearly identical passage with the same gender/sex conflation is found in the ERTA (WA2019b: 2). 10. Examining whether this move is philosophically defensible will lead us too far afield, so I will not discuss this matter here. I emphasise that my argument does not rely on endorsing this move. 11. Again, I am not suggesting that the ERTA and ERFC would then be perfectly acceptable. These regulations have been criticised by authors on various grounds, and especially on their reliance on the controversial premise that testosterone levels are a key driver of athletic performance (among many others, see, e.g. Karkazis et al. (2012), Pielke, Tucker, and Boye (2019) for criticism). I am simply claiming that the amended regulations could be defended from the charge that, because all trans women are women, they should automatically be allowed to compete in the same category which cis women (typically) take part in. Recall the point in section 1 that my argument is of a conditional form and does not require commitment to or endorsement of WA regulations: rather it is directed to those who endorse WA eligibility criteria, and points to an improvement that, by their own lights, they ought to make. 12. A reviewer notes that discussion in the debate has refocused from an athletes' current testosterone levels to whether the athlete has at some point undergone male puberty. I bracket this issue for present purposes, since I aim to engage with WA regulations as they currently stand. Should WA change the demarcation of the categories to have it depend on whether an athlete has undergone male puberty, then the labels of the categories should change accordingly.
13. Thanks to Jim Parry and Jon Pike for pressing me to develop my argument in significant ways, and to two anonymous reviewers and the editor for comments that substantially improved this paper. Thanks also to the audience at the 2022 British Philosophy of Sports Association Annual Conference .

Disclosure statement
No potential conflict of interest was reported by the author(s).