Philosophical problems with the gender critical feminist argument against trans inclusion

The Reform of the Gender Recognition Act: Government Consultation (2018) catalyzed a heated debate on transgender rights and trans inclusion in the United Kingdom. I start by explaining what the reforms to the U.K. system of gender recognition propose, why gender-critical feminists oppose them, and how other feminist academics have responded to their arguments. I then offer a more detailed philosophical critique of gender-critical trans-exclusionary feminist arguments. I argue that the gender-critical feminist case against trans women’s access to women-only (or sex-segregated, or single-sex) spaces suffers from a number of fallacies


Introduction
The Reform of the Gender Recognition Act: Government Consultation (2018) catalyzed a heated debate on trans rights and trans inclusion in the United Kingdom, in a context where, as Hines (2018a, p. 1) notes, greater trans visibility and legal protections for trans people have caused something of a backlash.The proposed reforms aim to replace the current medicalized process of gender recognition with one based on self-identification and self-declaration.The most prominent voice opposing the reforms was raised under the banner of gender-critical feminism.Academics from various disciplines, including philosophy and law, have lent their voices to the gender-critical project.In general, gender-critical feminism advocates reserving women's spaces for cis women.A few, though not all, gender-critical feminists make exceptions for some trans women, such as those who have undergone gender reassignment surgery.Apart from seeing the proposed reform as threatening women's spaces, gendercritical feminists, or many of them, are critical of the use of puberty blockers for self-identified trans children.Many applaud the kind of radical feminism championed by Jeffreys (2014) and Raymond (1979).
Gender-critical interventions over the proposed reforms have tended to take the form less of scholarship published in academic outlets than shorter online pieces in websites such as The Conversation, the Oxford Human Rights Hub Blog, and Medium.com,as well as contributions on Twitter.While academics often treat nonacademic online platforms as dissemination fora complementary to their academic output, the use of these platforms by gender-critical academics in the recent debate on gender recognition has tended to replace scholarly production.This is probably due to gender-critical academics' desire that their contributions should reach a wide audience, be timely, and be able to engage their nonacademic support base.It has also been suggested that academic outlets are biased against the work of gender-critical feminists (Stock, 2018c).As I argue, however, that the gendercritical case against trans inclusion is not well made, any reluctance to publish gender-critical work in academic outlets might be legitimate, rather than biased.
I start by briefly explaining what the reforms propose, and why gender-critical feminists oppose them.I then consider in greater detail the arguments made by Kathleen Stock, a philosopher who, as far as I am aware, has developed the fullest gender-critical case against trans inclusion, at least in the U.K. context.In other words, I take as the target of my critique the best in gender-critical feminism.I argue that, judged by a philosophical standard, Stock's claims and arguments are problematic.Among other things, I contend that Stock's key argument against trans inclusion-namely, the risks that inclusion of trans women in women-only spaces poses to cis women-is not developed in a way that sound uses of practical reason could endorse.Finally, as I take Stock's arguments as representative of broader problems in gender-critical thought, I suggest some explanations that help us see these problems as something more than a coincidence of unrelated, accidental missteps.
A terminological clarification: in using the term "trans" and "cis" in this article, I have in mind Serano's (n.d.) definition of "transsexual" and "cissexual.""Transsexual" refers to people who identify with, and live as, the sex that was not assigned to them at birth.This may include undergoing gender reassignment surgery.Everyone else is "cissexual."

Gender-Critical Responses to the Proposed Reforms to the Gender Recognition Act 2004
The centerpiece of proposed reforms to the U.K. Gender Recognition Act is the introduction of a new system for the legal recognition of gender identity.According to section 2 of the Act as it currently stands, a trans person can only obtain a gender recognition certificate after a relatively lengthy process, requiring them to provide evidence of having lived in their acquired gender for at least 2 years, as well as a medical diagnosis of gender dysphoria.It is widely recognized that the process is bureaucratic, and experienced as intrusive and demeaning by many trans people, particularly because the requirement of a diagnosis of gender dysphoria reinforces a view of trans identities as pathological (Baggs, 2018;Davy, 2011, ch. 5;Gender Recognition Reform (Scotland) Bill: consultation, 2019, § § 3.19-3.30;Reform of the Gender Recognition Act: Government Consultation, 2018; Stonewall, 2018, p. 20; U.K. Parliament, 2016, § § 32-44).
The proposed reforms would replace this system with one where a trans person can obtain a gender recognition certificate on the basis of making a statutory declaration (or something like it) about the gender they identify with (Reform of the Gender Recognition Act: Government Consultation, 2018).The Scottish Government has also recently published a draft bill for consultation, which, if enacted, will reform the Gender Recognition Act to introduce a self-declaration gender recognition model in respect of Scotland alone (Gender Recognition Reform (Scotland) Bill: consultation, 2019).Ireland, Malta, Norway, Denmark, Portugal, Belgium, Colombia, Brazil, Argentina, and Chile have already adopted a similar system, and its adoption has been advocated by the Parliamentary Assembly of the Council of Europe (2015).
Gender-critical feminists have argued that the introduction of such a system in the United Kingdom would be damaging to cis women, because it would undermine their safety in women-only spaces, by giving "males," or emboldening "males" to seek, access to such spaces.Part of this argument points to the risk that cis men would impersonate trans women to gain access to such spaces and women's bodies (Auchmuty & Freedman, 2018).Another part is that trans women are not women/females in the relevant sense, because they were not born and socialized as women, and hence lack the distinctive experience of sex-based subordination; on this view, trans women are both not among the proper recipients of the protection offered by women-only spaces, and cast as possible violators of those spaces and the bodies of cis women (Stock, 2018g).As Cooper (2019) puts it, "[t]oilets, changing rooms, girls' youth organisations, hostels, and prisons emerge [here] as the dystopic terrain of [cis] women's vulnerability to enduring predatory male behaviour." The legal response to this argument has been twofold.First, the reforms, by facilitating the acquisition of a gender recognition certificate, will make no difference to the question of access to women-only spaces.This is because those who are not natal females currently need not show a gender recognition certificate to access these spaces (Sharpe, 2018c).This is an effect of sections 7 and 29 of the Equality Act 2010, which, combined, require people who either have undergone gender reassignment (which need not involve any surgical or hormonal intervention) or plan to do so to be treated without discrimination in the provision of goods and services.Service provision here covers trans people's access to spaces reserved for the gender they identify with.As a result, trans people already have a right to access gendersegregated spaces congruent with their gender identity, regardless of their gender presentation, and regardless of whether or not they are in possession of a gender recognition certificate.Second, it is argued that the current law already provides cis women with adequate safeguards from cis men impersonating trans women, or from any malfeasant trans women themselves (Sharpe, 2018c).This is because section 28 of Schedule 3 to the Equality Act 2010 allows service providers to make exceptions to the rule of access without discrimination, if differential treatment is "a proportional means to achieving a legitimate aim." The gender-critical argument against trans inclusion has also been tackled on the philosophical plane.It has been argued that gender-critical feminists make a mistake on at least three points: first, in assuming or arguing that only cis women are in need of the protection of women-only spaces; second, in insisting that trans women's experience of socialization mirrors that of cis males; and third, in not treating gender identity seriously as relevant to determining who counts as a woman (Finlayson et al., 2018).Gender-critical interventions have also been taken to task for their failure to pay credible tribute to decades of feminist, trans and queer philosophical work debating how to define "woman" and how trans women are positioned in relation to that concept (Bettcher, 2018).The gender-critical insistence that the category of woman and/or female must be centered on bodily sex (as a biological category) has also come under criticism (Hines, 2018b).Gender-critical feminists tend to draw a sharp distinction between sex, as a biological reality, and gender, as the social construction of sex.In response, feminist critics of gender-critical feminism insist that we have no unmediated access to biological realities: they too become cognitively significant to us through discourse, including the discourse of biology, which is itself (like any other discursive domain) structured by political values.Thus, defining the concepts of "women" or "females" (as gender-critical feminists do) by reference to biological sex is itself a political choice, rather than one that can claim to neutrally reflect what the world is "really" like (Conaghan, 2018).It has also been argued that moral obligations of solidarity toward trans women require their inclusion in the category of woman and in women-only spaces (Finlayson et al., 2018).

Philosophical Scrutiny of Gender-Critical Arguments
Kathleen Stock has emerged as a particularly lucid voice in the gender-critical camp.Stock's ideas are highly regarded among gender-critical feminists (a look around Twitter is instructive).Stock presents an articulate, relatively comprehensive, and moderate form of gender-critical feminism.For instance, when, on January 30, 2019, Stock distanced herself on Twitter from the anti-trans tactics of some feminist activists, she was widely criticized by Twitter users who would normally support her.For these reasons, this article centers Stock's arguments: As I see it, engaging her arguments has implications for the credibility of the gender-critical project as a whole.
Except for one piece appearing in different versions in The Philosopher and on Medium.com(Stock, 2019b), Stock has not, at the time of writing this article, presented her gender-critical case against trans inclusion in the form of traditional academic outputs.Rather, she has done so in a series of blog-type pieces in Medium.com, as well as articles for The Economist, Quillette, The Conversation, and The Article.This is complemented by bite-sized interventions on Twitter.
Of these outlets, the Conversation (n.d.), Medium.com (n.d.), and the Article (n.d.-b) position themselves as striking a middle ground between "ordinary" journalism and scholarly writing.Stock (2018d) is very clear that her interventions are attuned to the nature of her essays' publication outlets, and that she does not take herself "to be doing substantive academic philosophy in those essays."Despite this disclaimer, Stock's profile, as it appears in connection with her online pieces, always carries the description of Professor of Philosophy.Stock (2018d) also expressly states that in those pieces she is addressing, among others, academic philosophers.And, in clarifying she was not doing "academic" philosophy, she seems to implicitly concede she was doing "philosophy" nonetheless-albeit of the footnote-less, conversational, nonacademic variety.
Thus, I think Stock's arguments against trans inclusion call for a critical evaluation in light of philosophical standards.In carrying out this evaluation, I discuss some of the problems with (a) Stock's case against including gender identity conversion therapy among the kinds of conversion therapies that should attract legal and policy disfavor; (b) Stock's case against including trans women in women's spaces; and (c) Stock's case against including trans women and trans men within the meaning of "female" and "male," respectively.

Begging the Question in the Argument About Gender-Identity Conversion Therapy
Stock has made a case against the understanding of conversation therapy used by the LGBT charity Stonewall in its advocacy against this practice.Her case, however, is question-begging.Stock (2018e) clarifies the meaning of conversion therapy adopted by Stonewall by quoting from the charity's own definition: "Conversion therapy . . .refers to any form of treatment or psychotherapy which aims to change a person's sexual orientation or to suppress a person's gender identity (my italics)."Stock believes that gender identity conversion therapy (as understood by Stonewall) should not be included within the kinds of conversion therapies that reasonable people should object to, or the government outlaw.Stock's (2018e) argument is to the effect that Stonewall's definition of conversion therapy is internally incoherent.It imposes on doctors a twofold requirement to avoid both sexual orientation conversion therapy and gender identity conversion therapy; but she thinks it is impossible for them to comply with one without violating the other.She illustrates this by reference to Margie, a hypothetical 14-year-old who is female-bodied, sexually attracted only to women, and identifies as a boy.
If Margie's self-diagnosis ("I'm a boy") is questioned by the therapist, the therapist can be construed as . . ."converting" . . .a trans child to a "cis" one.If, on the other hand, Margie's selfdiagnosis is affirmed unquestioningly, the therapist is effectively failing to affirm Margie in a sexual orientation of lesbianism; something which also looks like conversion by omission.(Stock, 2018e) The point made in the last sentence is question-begging because it builds (part of) the argument's conclusion into its premises.Specifically, the argument must assume Margie to be (a girl and) lesbian, to reach the conclusion that Margie, if affirmed in her self-defined gender identity, is effectively being converted from lesbianism to heterosexuality.The problem is that, as a matter of stipulation, Stock (2018e) clarifies that in this hypothetical scenario, there is "no prior underlying psychological story to give us the 'real' fact about Margie's transness, or lack of it; nor to tell us why Margie would reliably know that fact."If so, then no conclusion can be reached about whether or not Margie is lesbian (a cis girl desiring women) rather than heterosexual (a trans boy desiring women).The whole argument that affirming Margie's self-diagnosis would result in "converting" Margie's sexual orientation classification then flounders.At most it could be said that, as we do not know, by hypothesis, Margie's "true" gender identity, the therapist, in affirming Margie's selfdiagnosis, runs the risk of "converting" Margie to heterosexuality (not that the therapist is so converting Margie).In the same way, the therapist, in failing to affirm the self-diagnosis, runs the (very real) risk of trying to convert Margie to cissexuality, rather than merely being liable to being "construed" as so doing.Forsaking accuracy, here, means obscuring the reality of the situation, and the argument can be faulted for looking like it is trying to persuade by deploying rhetoric, rather than sound philosophical justifications.In any case, even the conclusion that the therapist runs the risk of "converting" Margie to heterosexuality is unwarranted, for the reasons explained in the next section.

Equivocation in the Argument About Gender-Identity Conversion Therapy
Crucially, Stock's argument also suffers from the so-called fallacy of equivocation (see Internet Encyclopedia of Philosophy, n.d.-c) which obtains when one trades on the ambiguity created by the use of multiple meanings of the same word.The equivocation is on the meaning of "sexual orientation conversion therapy."Stonewall's statement, quoted by Stock, that conversion therapy aims to change sexual orientation is based on the (usual) understanding of conversion therapy in terms of conversion of someone's sexual object choice.On this understanding, if Margie is attracted to women, a doctor would be engaging in sexual orientation conversion therapy only if she were trying to change the object of Margie's sexual attraction from women to men.But by affirming Margie in his gender identity, the doctor is not changing Margie's sexual object choice.
Stock's argument makes it look like a doctor affirming Margie in his gender identity would be converting Margie's sexual orientation only because it equivocates on the meaning of conversion therapy.By the time it reaches its conclusion, the argument has smuggled in an understanding of sexual orientation conversion therapy quite unlike the one initially presented by quoting Stonewall.Under this new understanding, conversion therapy is a matter of clinically reclassifying a patient from one category ("lesbian") to another ("heterosexual"), where the categories and conversion process are designed in such a way that re-classification is entirely independent of (a) changes in the patient's sexual object choice; (b) changes in their experience of sexual desire or sexual identity; (c) an intention on the doctor's part to effect these changes, or to support the patient in their intention to achieve them.This is a seriously implausible understanding of conversion therapy.Even if Stock's argument did not suffer from equivocation, and consistently adopted this understanding of conversion therapy from the start, we would have to reject the argument, because it is built on a concept that is both contrived and implausible.

Secundum Quid and Division Fallacies in the Argument About Trans Women's Exclusion From Women's Spaces
I now turn to Stock's argument against trans women's access to women-only spaces.Stock (2018d) states that her concern about female-only spaces is about legal selfidentification without any period of "living as a woman," prior male socialisation in a way which exacerbates the tendency to violence against female bodies, and the fact that many selfidentifying trans women . . .retain both male genitalia and a sexual orientation towards females.
As she clarifies, if "we think there are good reasons to retain same-sex spaces generally, in terms of protecting females from a small number of malfeasant males, these reasons don't cease to operate when males self-identify as women" (Stock, 2018g).This argument appears confined to trans women who, in accordance with the proposed reforms, will be able to obtain a gender recognition certificate on the basis of self-declaration/identification.In practice, however, the argument commits Stock to the broader proposition that we should protect cis women (and possibly also trans men assigned female at birth) from malfeasant (natal) males by excluding all (natal) males from women-only spaces.This is because of two reasons.First, the pre-reform system, as far as access to women-only spaces is concerned, already works like a self-identification one (Finlayson et al., 2018); so Stock's arguments, if they are valid, are so in the pre-reform system too.Second, if Stock's arguments that trans women lack socialization as women and often retain both male genitalia and a desire for women are valid arguments against access, they would apply even if trans women's access to women-only spaces were granted (as it is not) only upon their obtaining a gender recognition certificate in accordance with a medical diagnosis.
Indeed, Stock (2018c) elsewhere clearly states that she favors a policy that would exclude "all [natal] males from female-only spaces": this clearly covers all trans women, regardless of whether they have a gender recognition certificate and how they obtained it.In a world in which violence is overwhelmingly committed by men, and in which sexual violence is committed overwhelmingly by men and against women, adopting such a policy may seem entirely sensible.But what about people who, despite being assigned male at birth, identify as women?Here the suspicion might arise that the rule perhaps should not be treated as a (categorical) rule, but as a rule of thumb.If so, in treating it like a categorical rule, Stock's argument may be suffering from the so-called secundum quid fallacy (see Internet Encyclopedia of Philosophy, n.d.-a).
Rules are entrenched generalizations: It is in their nature to be suboptimal, that is, over and under-inclusive vis-à-vis their background justification (Schauer, 1991b).Whenever we propose to treat a given prescription as a (categorical) rule, we know that we will have to tolerate some measure of suboptimality, and we are prepared to do so in light of the advantages of treating a prescription as a rule (predictability of outcomes, efficiency in enforcing the rule, etc.).But sometimes those advantages are outweighed by the costs, because the suboptimality that rules carry with them through their insensitivity to particulars is a failure of justice and fairness (Schauer, 1991a).
So, we might be justified if we feel uncomfortable about the deceptively simple statement that the reasons for excluding natal males from women-only spaces "don't cease to operate when males self-identify as women" (Stock, 2018g).It may turn out to be the case that categorically excluding trans women from women-only spaces is suboptimal (unfair and unjust) in ways that we should not tolerate.If so, it might be because Stock's argument for reserving women-only spaces to cis women or natal females suffers from a further fallacy-division.This obtains when one reasons, unsoundly, that what is true of the whole (e.g., folks assigned male at birth) is also true of (some of) the parts that make up the whole (e.g., trans women assigned males at birth) (see Internet Encyclopedia of Philosophy, n.d.-b).
Stock (2018b) argues, "The problem here is male violence . . .[We] have no evidence that self-declared trans women deviate from male statistical norms in relevant ways."Something true of natal males as a class (their proneness to violence against women) is here assumed to be true of trans women as a (sub)class, on the basis of an absence of evidence.The effect of the argument is to treat as evidence of sameness an absence of evidence of difference, despite the fact that the reasonable assumption to make is just the opposite-namely, that there should be a difference in behavioral patterns between natal males as a class (the vast majority of whom are cis men), and trans women as a class.For there is clearly a difference between the experience of a child who is treated by others in way that are characteristic of boys and also feels like a boy, and a child who is treated by others in ways that are characteristic of boys whilst feeling that they are really a girl.(Finlayson et al., 2018) Elsewhere, Stock's (2018c) rationale for keeping trans women out of women-only spaces is different.She states, If the evidence shows (as, in fact, it is already showing) that some males-whether genuinely "truly" trans or just pretending-turn out to pose a threat to females, and it's really hard to tell in advance which ones will, can't we then make a social norm and/or law to exclude all [natal] males from femaleonly spaces . . .?
It is not clear what evidence Stock has in mind here.Prison figures appear to indicate that a high percentage of trans inmates in English and Welsh prisons have convictions for sex offenses, but Sharpe (2018a) showed such figures to be misleading before Stock authored the piece in which she claims that there is evidence of a threat to females.In any case, the best statistical evidence available (published before Stock's remark about the evidence we have) points precisely in the opposite direction: policies that give trans women access to women-only spaces without discrimination do not result in increased risks to cis women (Hasenbush et al., 2019).While Stock (2019d) has, more recently, dismissed these findings on the ground that "plenty of time is needed to see the real impact," her dismissal is far too off-handed.Hasenbush et al.'s (2019) peer-reviewed, quantitative study is clearly the most directly relevant to the question of trans women's access to women-only spaces (unlike, for example, the study cited by Stock and others in Allen et al., 2019, which is, instead, about unisex spaces).To appreciate its relevance, specifically, to the U.K. context, note that in Massachusetts-where the study was carried out-access to sex-segregated spaces is governed by the same self-identification principles and practices as are applicable in the United Kingdom under the Equality Act, as seen in the governmental guidance (Healey, 2016) that explains how these policies are to be operationalized.
Given that the hardest evidence we have indicates that opening women's spaces to trans women does not increase risks to cis women, it seems that in citing evidence in support of her policy to deny trans women access to these spaces, Stock might have in mind individual incidents of trans women (or their impersonators) harming cis women.It is important to note that here Stock does not treat evidence of incidents of violence by individual trans women as evidence of proneness to violence by trans women as a class.Rather, she thinks that natal sex offers the only workable watershed to keep malfeasant trans women (or their impersonators) out of women-only spaces, even if it comes at the cost of keeping out all harmless trans women too, and regardless of how prone to violence trans women may be as a class (relative to natal males as a class or cis men as a class).Effectively, the logic of this argument displaces the question (which elsewhere Stock did seem to treat as significant) of how likely, statistically speaking, trans women assigned male at birth are to commit violence against women.As long as some trans women-Stock (2018c) even concedes they might actually all be cis men impersonating trans women-commit violence against cis women, it is imperative to exclude from women-only spaces all trans women.Admitting trans women "poses unacceptable harm to the original occupants of the category 'woman'" (Stock, 2018g).

Errors of Practical Reason in the Argument About Trans Women's Exclusion From Women's Spaces
This-that admitting trans women into women-only spaces poses unacceptable harms to cis women-ultimately is Stock's justification for trans women's exclusion from women-only spaces.The argument appears consequentialist-that is, roughly, it judges the goodness or badness of a policy admitting trans women to women-only spaces by reference to its consequences, or the consequences it is likely to have (the definition of consequentialism is contested; Raz, 1986, pp. 267-268).Consequentialist considerations are an important part of our everyday uses of practical reason, and are routinely (and appropriately) used as a basis for much policy-making.It does not follow, however, that Stock and other gender-critical feminists are right in maintaining that, as admitting trans women into women-only spaces might result in some trans women (or their impersonators) harassing or assaulting cis women, then no trans women must be admitted.An example will immediately illustrate that this is not how policy choices are made, nor the way to sensibly determine whether a certain course of action is morally sound.
Consider jobs in which workers come into contact with children.We know that, statistically, men are much more likely to be sexually predatory toward children.We also know that it is difficult to tell in advance which men will offend (at least if they do not yet have relevant convictions).Finally, we know that some men who come into contact with children in their work will offend against them.Yet we do not exclude all men from working with children, even if using gender as a watershed would prevent those offenses.Why does the good of minimizing child sexual abuse not lead us inexorably to the conclusion that we must outlaw all male teachers and coaches?Because our practical reason recognizes complexity: We readily see that even the most highly desirable states of affairs (minimizing abuse of children) do not have simple, quasi-mechanistic implications for policy or decision-making, and that they do not justify the indiscriminate suppression of other goods (even less important ones, such as professional vocations).
Some of the considerations we take into account when making these judgments are the relative seriousness of harms that may follow from different courses of action, the likelihood of those harms eventuating, how narrowly tailored a certain policy or course of conduct is to achieving the relevant good, and so on.There will also be considerations having to do with the nature of competing values.Some of them may work like deontological constraints on consequentialist thinking.In other words, practical reason may lead us to decide against a certain policy or course of action because of the kind of thing that it is or instantiates, regardless of the beneficial consequences it might bring about.Such considerations often have to do with some understanding of respect, political "equality," or moral independence, and an appreciation of how the adoption of certain rules in and of itself violates these goods.
These kinds of considerations do not properly feature in Stock's arguments against trans inclusion.Indeed, remember that my rough definition of consequentialism-as a mode of assessing the goodness or badness of different policies or courses of conduct-spoke of consequences that conduct or policy either produces or is likely to produce.Yet, as we have seen, at least in one of its versions, Stock's consequentialist argument ultimately denies that the likelihood of the envisaged harms to cis women matters.Effectively, Stock's argument is asking us to assign more weight to a possible harm to cis women than the certain harm that exclusion will bring to trans women: that of being denied the safety (and other benefits) that women's spaces provide.And it is asking us to do so when the best evidence shows that giving trans women access to women's spaces does not in fact increase risks to space users' privacy or safety (Hasenbush et al., 2019).

Why a Separate but Equal Solution Is No Solution
I argued that the harm of trans women's exclusion must be given due consideration, as must the certainty of its occurrence if trans-exclusive policies are adopted.The proposal of a separate-but-equal solution, which would see parallel services and spaces segregating trans women from cis women (Stock, 2019d), does not give that harm serious consideration.It runs up against the objection that the numbers of trans women are too low to make trans-women-only spaces and services viable.As Finlayson et al. (2018) note, "including trans women in women-only spaces is by far the most practical approach to take under the current circumstances."Even if this objection could be overcome, however, there is another, more fatal one.
This has to do with another kind of harm that occurs ipso facto by virtue of a policy of trans exclusion, whether or not complemented by a "separate but equal" corrective.This is "the harm of being treated as unreal and unintelligible"specifically, unintelligible as a woman (Cooper, 2019).This harm is similar in kind to the illocutionary subordination that attends the performance of hate speech acts (Zanghellini, 2003).(This is, emphatically, not the same as arguing that, in recommending her policy, Stock, who expressly disclaims transphobia, Stock, 2018a, is engaging in hate speech.)Our social world is arranged in a way that makes exclusion from the sex/gender they claim-on the basis of a lack of "authentic" belonging (Serano, 2007)-central to trans subordination.As with other forms of social subordination, trans exclusion has not only material dimensions (Blair & Hoskin, 2018;Hargie et al., 2017;Moolchaem et al., 2015;Movement Advancement Project and GLSEN, 2017;Rondón Garcia & Martin Romero, 2016;Serano, 2013;Stonewall, n.d.;Yona, 2015), but also discursive ones that work in accordance with the logic of so-called performatives.Performatives are utterances that do things with words: specifically, they accomplish something in the act of saying it (Austin, 1975).The classical example is marriage-in the act of declaring a couple married, a celebrant brings about a change in their normative status, provided the celebrant is the right person in the right circumstances.This presupposes a normative background (that is a set of laws, conventions, or other rules) governing all those matters: who qualifies as a legitimate celebrant, what the right circumstances are for the performative to do its work, what marriage status means in terms of spouses' rights and obligations, etc. Butler (1990) famously argued that gender is performative, in the sense that it is called into being through repeated speech acts (starting with the assignation of gendered names at birth): these construct, consolidate, and stabilize genderbased subjectivities, though in a way that is always contingent and open to failure.But the idea of performative is also useful to grasp the discursive dimensions of social oppression.Deadnaming, deliberate misgendering, transphobic and trans-exclusionary speech acts and policies normally work like performative utterances: they secure/reinstall the ideological subordination of trans people by virtue of their consonance with the normative background (reflected across different discursive domains, including the criminal law: see Sharpe, 2018b) that represents and normalizes trans people's relation to the sex/gender they claim as fraudulent, inauthentic, second best.If there is a valid deontological argument in the debate about trans inclusion, I think this comes closest to it.It suggests one reason for objecting to policies such as those recommended by Stock, just for the kind of things that they are-namely, inherently subordinating of trans people.Yet, I am reluctant to treat even this harm as a knock-down argument with conclusive force in all circumstances, regardless, for example, of the nature of who manages and/or funds the relevant single-sex space (the State, local government, activists, volunteers, etc.). 1 I do argue, however, that an adequate consideration of this harm is conspicuous by its absence in gender-critical interventions that recommend a policy of blanket exclusion from women's spaces of all trans women (or all those who have not undergone gender reassignment surgery, or all those who have not obtained a gender recognition certificate in accordance with medicalized procedures).In addition, when it comes to the different question of individuals deciding on different possible courses of action legally open to them, I maintain that the ethically sound choice is to treat the combination of material and discursive harms enacted by trans exclusion as a quasideontological constraint that recommends defaulting to trans inclusion whenever there is no concrete evidence (relating to the specific trans person seeking inclusion) that requires excluding them.As Finlayson et al. (2018) argue, honoring "the subjective sense" of one's gender identity "is, in most contexts, the right political and ethical response." It is important to understand how trans-inclusive policies, courses of action, or linguistic usages do not have a subordinating effect on cis people in the same way in which I argued their trans-excluding equivalent has on trans people.In an exchange on Twitter, Stock has stated that trans women claiming a female identity could be said to deny her reality as a female and lesbian (see Figure 1).In fact, the logic of performatives strongly suggests that trans-inclusive utterances/policies/choices do not deny my reality as a cis (gay) man, or Stock's reality as a cis (lesbian) woman, nor otherwise have any harmful illocutionary effect on those of us that fit into these categories.This is because the normative/conventional background that is needed to activate any (purportedly) harmful (reality-denying) power of trans-inclusive speech acts is missing.Given the normative background that subordinates trans to cis (rather than vice-versa), the idea that trans-inclusive utterances could deny or cause illocutionary harms to cissexual realities is, to me at least, philosophically unintelligible.

Why Trans-Inclusive Practices Do Not
Threaten the Concept of "Female," "Male," "Lesbian," and "Gay" Stock (2018b) has also argued that trans inclusion on the ground of self-identification/declaration threatens "a secure understanding" of concepts intimately related to "woman"namely, "female" and "lesbian."It is hard to see this threat as a real one.After all, conceptually, "trans maleness" and "trans femaleness" presuppose "cis maleness" and "cis femaleness" as their other-namely, the case of female and male for which no transition, no reaching across, is required: the case of femaleness and maleness already on this side of (= "cis") their sex.
Indeed, Stock's own reflections on the concepts of "sex," "male" and "female"-and her very observation that "if we were to lose [the concept of female and male] . . .we would have to reinvent them" (Stock, 2018b)-suggest very convincing reasons why her apprehensions that trans inclusive practices imperil these concepts are not well-founded (even when these practices are based on self-identification/declaration).Specifically, Stock (2019b) states that given "that the human species continue to reproduce, one might well look askance at any conclusion saying that there are no underlying material facts about maleness and femaleness to which humans can have reliable epistemic access."This, as I explain below, refutes her own argument that trans inclusion imperils the concepts of "male" and "female."Stock (2019b) argues, correctly, that "sex [i.e., maleness and femaleness] is not determined by any single, unitary set of essential criteria," and that "there is no single set of features a person must have in order to count as male or female."She goes on to state that: (a) "you do still need to possess some" female (biological) sex characteristics to count as female; (b) that this is "a real, material condition upon sexcategory-membership"; and (c) that "medical professionals [assigning sex]. . .rely upon an established methodology, aimed at capturing pre-existing biological facts" (Stock 2019b).Stock presents (a), (b), and (c) as if they were true without qualification.In fact, they only describe how, for very legitimate reasons, sex is understood and assigned within the discourses of biology and medicine; but our everyday usages of "male" and "female" may well be more capacious.It does not follow, of course, that there is no connection at all between these discursive domains-biology and the everyday.Rather, something like the biological meaning of "male" and "female" refer to the central cases of "male" and "female" as those terms feature in everyday usages.But those usages, if trans-inclusive (as they should be), will also cover, legitimately and usefully, noncentral cases of those selfsame terms.
There really is no good reason to fear that such transinclusive practices will imperil "maleness" and "femaleness" as concepts.It is the very fact that those concepts have and will retain central cases that puts to rest any such fear.What makes something like the biological meanings of "male" and "female" the central cases of everyday usages of those words is "[o]rdinary-life truth seeking, a certain level of which is essential for survival"; this "involves a swift instinctive testing of innumerable kinds of coherence against innumerable kinds of extra-linguistic data" (Murdoch, 1992).Reproduction is a key aspect of human experience: The existence of each of us and the perpetuation of the human species presuppose it.The extra-linguistic reality of the dioecious configuration of human bodies, which is functional to human reproduction, means both that the concept of "female" and "male" are here to stay, and that their central cases will remain well-understood, even after we give up on trans-exclusionary attitudes, practices, and policies.To put it another way: trans-inclusive linguistic usages, policies, and so on, cannot threaten the distinction between the concepts of "male" and "female," which hinges on the nondisposability of the central cases of those concepts.
For similar reasons, it is difficult to agree with Stock that characterizing as "gay" trans men attracted to men, and as "lesbian" trans women attracted to women, "leaves us with no linguistic resources to talk about that form of sexual orientation that continues to arouse the distinctive kind of bigotry known as homophobia" (Stock, 2019d).After all, our linguistic conventions make cissexual womanhood and manhood the central or paradigmatic cases of "womanhood" and "manhood"; cissexual (though not necessarily gender-conforming) lesbianism and male homosexuality the central or paradigmatic cases of "lesbianism" and "male homosexuality," and so on.This will not change.First because of the prevalence of cissexual women/men and cissexual lesbians/ gay men, in terms of sheer numbers, relative to trans women/ men and trans lesbians/gay men.Second, because of the ways in which the concepts of "man," "woman," "gay," "lesbian," "cis," and "trans" sit together with the concepts of "male" and "female," which reference an extra-linguistic reality, of which, as we have already seen, we cannot but take notice.Given these linguistic and empirical facts, a transinclusive use of the terms "lesbian" and "gay" does not carry the dangers Stock (2019d) worries about.
Having recourse to the idea of central and noncentral cases of "male" and "female" is much preferable to Stock's more rigid and narrow understanding of those concepts, which would reserve their use to designate, respectively, only biological males and females.It is politically preferable for all the reasons readily understood by those feminists who embrace trans-inclusive language.And it is descriptively preferable because it enables us to account for trans-inclusive everyday usages of "male" and "female" without having to ascribe them to the speaker's delusions (noncentral cases of a concept are not delusion-based; they are, simply, not the paradigmatic cases of a concept).

Overemphasizing Sex-Based Subordination
I think that the source of some of the problems discussed above, and more generally some of the problems with the gender-critical case against trans inclusion, is in the way structural subordination on the basis of sex dominates gender-critical analysis to the virtual exclusion of everything else.Critical scholars call this way of proceeding "presupposed paranoid structuralism" (Halley, 2008).
Invoking this label carries some risks: It tends to elicit strong and negative reactions because the term "paranoid" immediately raises the specter of mental illness.But that is emphatically not how the term is used in either critical literature or in this article (which is not interested in offering, unlike Elliot and Lyon, 2017, p. 360, a psychoanalytic reading of gender-critical work).Work on paranoid structuralism in critical studies uses "paranoid" not as a "pathologizing diagnosis" (Sedgwick, 2002, p. 126)-although, of course, the central case of "paranoia" is indeed a psychiatric pathology-but as a playful metaphor for the "hermeneutics of suspicion" employed by searing critical work directed at revealing the covert structural determinants of social relations.As Sedgwick (2002) explains, "in the hands of thinkers after Freud, paranoia has . . .candidly become less a diagnosis than a prescription" (p.125).That is, critical scholars are deliberately committed to "paranoid" readings of social reality, treating such "paranoid" commitment as a productive scholarly attitude.It is only when "paranoid structuralism," instead of generating hypotheses that are then tested, is presupposed as an article of faith in critical work that employs it, that it becomes problematic (Halley, 2008).
The concept of presupposed paranoid structuralism, then, is intended to alert those of us doing work directed at unveiling and challenging social hierarchies (i.e., "paranoid" work) to some of the risks inherent in what we do.The risk is that our ideological and methodological investments in the structuralist binaries we care about (males oppressing females; nonqueers oppressing queers, etc.) can lead us to presuppose their validity, instead of taking their validity as a hypothesis that must remain open to falsification.Presupposing, rather than hypothesizing, the validity of structuralist binaries "can lead you," in Halley's (2008) words, "to miss noticing other things that are going on, things that just can't, and even if they can probably should not, be forced into [the explanatory frameworks you construct on the basis of your favorite binaries]" (p.191).
Take the tweet in the figure above.It suggests a lack of appreciation of two crucial points: first of the disanalogy between claiming that trans-exclusionary practices deny trans realities, and claiming that trans-inclusive practices deny cis realities; and second of the ethical imperative to take into account the harm of denying trans realities when navigating the questions of trans recognition and trans inclusion.Such disanalogies, and such needs, are precisely the kind of "things that are going on," but that one can "miss noticing" (Halley, 2008, p. 191) if one inadvertently presupposes the singular centrality, ubiquitous validity, and superior explanatory power of the (natal) male/ (natal) female binary in structuring social relations.Here are some other things we might miss noticing, if we make that presupposition: a.Even assuming that the socialization of trans girls mirrors that of cis boys, the fact that trans girls do not identify with maleness can be expected to make a difference to the outcomes of such socialization (Finlayson et al., 2018).b.It is a mistake to treat "violence and discrimination against trans women . . .as if it were unconnected to that faced by cis women" (Finlayson et al., 2018).c.Saying "Not giving people everything they desire is not a denial of their humanity" (Allen et al., 2019) amounts to an insensitive dismissal of the serious argument that trans exclusion is ipso facto harmful.d.The claim that women "are a culturally subordinated group . . .[while] at best, trans women are a distinct subordinated group; at worst . . .members of the dominant group" entirely discounts the ways in which sex, gender, and cis/trans status intersect.These intersections produce more complex, shifting, and contextdependent power relationships than are captured by the M > F formula. e.A dubious assumption underlies this statement: "[T]he fact that our concept-application [of, e.g., 'woman'] might indirectly convey disadvantage towards some social groups [e.g., trans women] is not itself a reason to criticise the concept use, because the concept use has a further valuable point" (such as "to pick out a distinctive group, relative to recognisably important interests") (Stock, 2019e).The dubious assumption here is that the "valuable point" of a restrictive use of the concept will be lost if the concept is broadened.
The assumption is dubious because even in its broad, inclusive use, the concept retains a readily identifiable central case.Halley (2008) argues, controversially, that much feminist scholarship suffers from presupposed paranoid structuralism.Cossman et al. (2003, p. 635) find Halley's arguments more compelling if confined to radical feminism.Although Catharine Mackinnon does not advocate trans exclusion from women's spaces (Williams, 2015), and although Andrea Dworkin was a trans ally (Stoltenberg, 2020), it seems significant that many (not all) gender-critical feminists are also radical feminists.In any case, I think there is something to gain from entertaining the possibility that presupposed paranoid structuralism is a source of shortcomings in gendercritical arguments, particularly if it has the effect of prompting some gender-critical soul-searching.

Doing Philosophy and Debating Policy in the Age of Social Media and Digital Platforms
The case against trans inclusion in the United Kingdom, as I indicated, has been presented primarily through social media and blog-type or journalistic online platforms lacking the traditional prepublication checks of academic peer review.I think these media of presentation are also partly to blame for the shortcomings in gender-critical arguments.My arguments in this section are less specifically about Stock's own contributions, than about the gender-critical project as a whole.An implication of my argument is also that, to the extent that supporters of trans inclusion make wide use of these digital fora, they too subject themselves to conditions that are less than conducive to reaching genuine understanding.
A distinctive feature of these fora is that they are instances of writing only in a hybrid sense: They are, rather, forms of "secondary orality and secondary literacy," concepts used to "emphasize the immediacy and dialogic nature of digital communications, even though they may occur in replicable, searchable text formats" (Stewart, 2016).Twitter, where much gender-critical discourse is produced, reproduced, and shared, is unlike writing and like speech, having strict word limits on individual tweets, and being designed to encourage to-and-fro exchanges of bite-sized bits of communications: "academic Twitter enables a collapsed space of engagement, wherein the analytic, text-based content of scholarship is shared via often-casual, participatory and dialogic forms of exchange" (Stewart, 2016).Even the Article, the Conversation and Medium.compartake as much of the features of speech as the written word, given the short length of the pieces they publish, their "like," "leave a comment," and "share" (including "share on Twitter") features, and, in the case of Medium.com (which was launched by a Twitter co-founder, Kumar, 2018), the conversational style it accepts.Indeed, the Article (n.d.-a) expressly characterizes itself as a "publisher that is also a social media platform." One might argue that far from being a problem, it is an advantage that the media favored by many gender-critical feminists are more "spoken" than "written" (or, at least, partake as much of speech as of writing).One here might point to Murdoch (1992), who follows Plato in arguing that the spoken word has an edge over the written word because the latter cannot respond, clarify, or restrict the interpretive discretion of a reader in the same way that a speaker can in the context of conversation.Yet, Plato (n.d.) is explicit on the point that speech is no more valuable than writing if "not with any view to criticism or instruction."Here it is useful to invoke Habermas' (1979) distinction between genuine communicative action, where the speakers' only aim is to reach understanding, 2 and linguistically mediated strategic action, where instrumental aims interfere with or replace the goal of reaching understanding.
Needless to say, in real-world face-to-face exchanges, unalloyed communicative action is known only by approximation.But there are very good reasons to think that the distance between the ideal (namely, communicative action) and the real is especially wide in the context of the quasi-spoken digital media used to construct (and respond to) the gendercritical case against trans inclusion.Stock (2019f) herself, discussing the reception of her arguments, has complained about countless "half-arsed takedown attempts" by "online philosophers," crediting, conversely, philosophers she meets offline with "interesting, constructive, and charitable" objections.She also notes that social media siphons "users into paranoid, angry silos" (Stock, 2019d), and that "when reading disembodied words on a screen" it is "easy enough" to engage in "projection" (Stock, 2019a).Why and how do social media and allied platforms have this potential for distorting genuine communicative action?
First, they enable new manipulative communication practices, such as flaming and trolling.The popular support base of gender-critical academics makes ample use of these, though gender-critical scholars are also at the receiving end.Rather than using the quasi-spoken features of social media and allied platforms with a view to genuinely advancing understanding, online activists may exploit these features for strategic aims.Common techniques include drowning a post or blog with irrelevant comments; exposing the blogger to ridicule; deflecting attention from the point she made; forcing her to address spurious objections; pretextually professing a failure to understand, demanding endless further explanations; and so on.Some of these techniques are available in spoken exchanges, but social media and allied platforms magnify their power by enabling "widely-distributed individuals to organize and galvanize around issues of common interest [or] political advocacy" (Stewart, 2016); and by facilitating the use of nonverbal or nonargument-based, but effective, communicative devices, such as memes, gifs, and emoticons.
Another way in which these digital media distort genuine communicative action is by affecting the motivations of the blogger, or micro-blogger, herself.Specifically, they facilitate the interference with genuinely communicative goals (reaching understanding) by noncommunicative, strategic aims.I will discuss three: acquiring influence, career progression, and venting.
In traditional academic communicative practice, one's recognition as an expert is supposed to follow from the credit that accrues to one as a result of the soundness of one's research methods and arguments, judged through peerreview processes.But "in the era of social media there are now many different ways that a scientist can build their public profile; the publication of high-quality scientific papers being just one" (Hall, 2014).Veletsianos and Kimmons (2016) have found, by examining a large data set of education scholars' participation on Twitter, that being widely followed on social media is impacted by many factors that may have little to do with the quality of scholarly work . . .and . . .that participation and popularity may be impacted by a number of additional factors unrelated to scholarly merit (e.g., wit, controversy, longevity; p. 6).
Furthermore, as scholars become more popular on Twitter, they do so exponentially, particularly if they identify online as "Professor" (calling into question the conventional wisdom that academic engagement with social media "leads to new and more egalitarian structures for scholarly dissemination") (Veletsianos & Kimmons, 2016).As Rogers (2018) notes, "[s]ocial media metrics propagate this loop of wellknownness (and constructed celebrity rather than greatness) by keeping score and displaying it in number badges, follower counts, and similar outward indicators."This is also true of such platforms as Medium.comand the Conversation, which allow both bloggers and readers to keep track of the contributions' stats.Social media and allied platforms, in sum, offer us both the opportunity and the temptation to cut (academic) corners in becoming "influencers," unlike traditional peer review processes, which are designed to make influence and recognition a by-product of, and contingent on, genuine communicative action.(The Conversation, while only accepting academic authors, fails to require that its pieces be based on outputs previously published in peerreviewed academic outlets.) A common instrumental reason motivating academic users of social media, particularly heavy users, is their perception that it benefits their career advancement (Veletsianos et al., 2019).Indeed, career advancement appears to be the most significant factor associated with being an academic "super user" of social media (that is, one who frequently posts "updates on their research"), particularly when compared with other motivational factors, such as the belief that communicating research benefits the public, or contributes to scientific discovery (Zhu & Purdam, 2017).Relevantly to the context of the debate on trans inclusion, social media academic "super users" appear "to be based in the humanities and social sciences" (Zhu & Purdam, 2017).In the United Kingdom, where institutions encourage academics to engage with blogging and social media (Kieslinger, 2015), the perception that social media engagement is beneficial for one's career seems well-founded.Zhu and Purdam (2017) point out that in the 2014 Research Excellence Framework (REF), some "impact statements" cited the numbers of a scholar's Twitter followers; and in the United States there have been calls for formally embedding evaluation of the use of online social media in academic career progression processes (Grudz et al., 2011).Digital platforms themselves are unapologetic in marketing themselves to potential academic authors by drawing attention, first and foremost, to the benefits that might accrue to the scholar's recognition and career, rather than the state of our collective understanding of the world.Thus, the Conversation's (nd-a) "Writing, pitching and training guide for academics" opens by giving data about the size of the Conversation's readership, then immediately goes on to state, "The Conversation UK helps raise the profile of you, your work, and your university, positioning you as an expert voice to address a global audience" (emphasis in original).Undoubtedly, the desires for career progression and for recognition have always been one of the drivers behind engagement with traditional scholarly publishing, too.But the potential for these motivating factors' interfering with the quality of academic communication is diminished by the peer-review processes and standards characteristic of traditional academic fora.
Blogs and social media also distort genuine communicative action by facilitating venting.This is partly because they distance interlocutors in a way that lowers benign face-toface inhibitions.This is a well-recognized online phenomenon (Suler, 2004).While the prospect and practice of peer-review processes tend to reduce (though they might not eliminate) ad-hominem attacks, unproductively antagonistic posturing, and the poor argumentation that comes with them, similar checks are not in place in the context of nonacademic digital fora.In the debate on trans inclusion, this has sometimes resulted in a kind of communication that is a far cry from the dialogic, respectful attitude that Habermasian communicative action, let alone empathy-centered feminist forms of communicative ethics, calls for-Admittedly, gender-critical feminists reject the exhortation to be empathic as a "gendered expectation" (Allen et al., 2019), but experience of receiving feedback on drafts of this article suggests they are mistaken.Despite recognizing the potential of social media for encouraging "sneering," "dehumanizing," "insulting," and threatening language (Stock, 2019d), Stock herself, as many others, has fallen prey to online incivility: In a Medium.compiece, she refers to fellow-feminists who do not share her predictions as possibly "quite thick" (Stock, 2019c, emphasis in original).The problem is that "impolite and incensed blog comments can polarize online users based on value predispositions utilized as heuristics when processing the blog's information" (Anderson et al., 2014, p. 383).
Digital incivility is not the only mechanism that may result in compromised thought processes in the blogosphere and social media.First, it has been theorized that the very pacing of "real-time" online communication-in compressing the interval between a post and responses to it-encourages interlocutors to default to heuristics, leading to cognitive bias (Menzies & Newson, 2007, p. 93).Second, use of social media to debate political issues may result in dulling our critical faculties by enabling large-scale, reiterated distribution of disinformation: "people are more likely to be affected by inaccurate information if they see more and more recent messages" (Tucker et al., 2018, p. 40).Third, the quality of our political thinking is diminished by participation in cohesive social networks (Erisen & Erisen, 2012); and scholars using social media find them especially congenial to build just such networks.They do so by "zero[ing] in on shared attributes or outlooks," and creating a sort of "hyper-personal communication," where "computer-mediated intimacy . . .may be stronger than [that] established face-to-face" (Stewart, 2016).This seems consistent with findings that social media "may facilitate greater selective exposure" to "attitude-consistent information about politics" (Tucker et al., 2018, p. 53).Indeed, and SAGE Open perversely, even exposure to opposing views on social media can increase, rather than reduce, polarization for some demographics (Bail et al., 2018).
In sum, the use of social media (and other digital platforms, to the extent that they share some of the same attributes) poses a number of concerns in respect of their ability to facilitate genuine communicative action, particularly when compared with traditional academic publication outlets.This being the case, the enthusiasm with which many gender-critical voices have embraced these media may well be implicated in the problems with their arguments against trans inclusion; likewise, it is unclear that engaging gendercritical voices on that terrain has been the most productive move on the part of those of us who advocate for trans inclusion.

Conclusion
Cooper (2019) has invoked a legal pluralist perspective to argue that it is possible, and may be desirable, for gender as conceived by gender-critical feminists (as "sex-based domination") and gender as conceived in trans-affirming terms (as "identity diversity") to coexist side-by-side in the law.Access to women's spaces is just the kind of policy matter that need not choose between one conception of gender and the other: it can and should be granted on the basis of both.While a compelling feminist case has been made for inclusion (Finlayson et al., 2018), the best feminist case against inclusion suffers from a number of argumentative fallacies (Aristotle, n.d.), and is at odds with well-established and sound uses of practical reason.Many problems in gender-critical thought are consistent with the explanation that paranoid structuralism is too often presupposed in gender-critical work, rather than being treated, productively, as a hypothesis.The nature of the publication outlets favored by gender-critical feminists (social media, blogs, etc.) is also likely to be implicated in generating some of these problems.

Figure 1 .
Figure 1.Twitter exchange between Professor Stock and Professor Sharpe.