border patrol at themiscyra

(or: a parable about counterexamples to self-id)
~4256 words (~17 minute read)
12 04 2026

prologue

A problem faces the Amazonians at Themiscyra. An inclusion problem.
Intuitively, their rules for entry are simple: women only. No men, no nonbinary people, and only sometimes genderfluids. But a series of complaints has forced their hand in precisifying this gloss. Their original definition, in terms of gametes and chromosomes, was deemed unfit for the modern age; it unfairly excluded swaths of intersex and trans women. In lieu of these antiquated biological criteria, they placed criteria couched in terms of feminine presentation and social role. These, too, were ruled exclusionary, only now towards gender nonconforming and nonpassing women.

After fumbling through a series of other bad definitions, the Amazons at Themiscyra settled on a standard that saw the most success: they just asked. If entrants say that they are women, then they are allowed in, and if they say anything else, then they are not. But, of course, a few stragglers still posed counterexamples for the Amazonians. Rather than abandoning ship like they had with their prior criteria, they steeled their confidence in self-identification, hoping that the definition could navigate the remaining problem cases after some minor patches. The six stragglers were held at the Themiscyran border hotel, their cases pending litigation from the city. A team of Themiscyra’s best philosophers was assembled and shipped to the hotel, prepared to solve the inclusion problem once and for all.

the cases

#1: bryan and sincerity

Bryan is a man who wants to enter Themiscyra because he heard that they make good beer. He changed the gender marker on his passport to M a few years ago, has he/him/his in his email signature, and identifies as a man in all situations besides this one. But, to the philosophers, he says that he is a woman.

Bryan, it seems, is attempting to trick the system. That is, his identification as a woman is insincere, or deceptive, and thus should not count as gender-determining. To address cases like Bryan’s, the Amazonians add a sincerity condition to their definition of ‘woman’: now, they count women as all and only those people who sincerely identify as women.

But how do we know whether people are being sincere? Is sincerity not subjective? How is such a nebulous concept as ‘sincerity’ to clarify such an important concept as ‘woman’? The Themiscyrans have a three-tiered response:

Every view will have to countenance properties to which we do not have perfect epistemic access. Whether these properties are social, psychological, or biological, there is always a possible gap between what we know (or plausibly could) and the actual features of the world. These blanket epistemological worries grow stronger when we consider who needs to have access to these properties, that is, Themiscyran border officers. Even if it is conceded that in-principle epistemic concerns apply asymmetrically to the cases of sincerity and of one’s social or biological properties, there is nonetheless the practical epistemic question of how Amazonian police access these facts. Are they to inspect people’s genitals? Surveil people’s interactions to see whether they function, socially, as a woman? Perform MRIs to determine the sex of someone’s brain? These options are all either implausible or obvious cases of overreach. Although these alternative views may soothe idealized epistemic worries by tying gender to features more concrete than sincerity, they are placed on a par (at best) to self-id by these practical epistemic concerns.
The question of our access to an individual’s sincerity is not only a challenge for this amendment to self-id, either. The concept of ‘sincerity,’ or ones whose meanings approximate to it, are pervasive in our social institutions in ways whose gravity far outstrips that of gender’s. People, for example, receive guilty verdicts based on sincere confessions, are convicted on the basis of sincere witness testimony, and form binding contracts through sincere assent to its terms. Sometimes, like in cases where the death penalty is being considered, a person’s sincerity is actually a matter of life or death. This all makes a satisfying epistemic story regarding sincerity a desideratum of any worldview, period, regardless of whether or not it’s on board with the concept’s application in this specific instance.
Luckily, there is a fairly straightforward epistemic story to be provided here. There are, speaking broadly, three levels at which we make judgements about people’s mental states (including their intent to deceive).

The first of these is analogous to our sense-perception of external objects: we look at people and subconsciously log clues, like tells in the face and body, which indicate their state of their mind. Someone yelping is taken to be in pain, someone crying is taken to be sad, someone laughing is taken to be happy, and so on. This mode, of course, is not perfect, but it’s important not to understate its reliability. Accurately judging the attitudes of our comrades, just like accurately judging the ripeness of fruit or the pathogen content of old food, was something that was incredibly advantageous for our ancestors, and evolution has therefore equipped us with faculties reliable for this end. For the intents and purposes of normal life, our basic sensory apparatus is enough to tell when fruit has gone off, when meat has gone rotten, and when someone is only telling stories.
In cases where this first option is unreliable, though, we have a second option: to make inferences from circumstantial information to facts about a person’s mental life. While me reading someone’s solemn diary entry is not me witnessing their sadness per se (I haven’t cracked open their skull and looked at its balance of hormones), it nonetheless allows me to infer, with a good degree of certainty, that they are sad. So goes sincerity—from facts about an individual’s behavior we can infer facts about their system of beliefs, and, when this system of beliefs likely does not align with what they profess, they are being insincere. This is what happens when determining mens rea in a court of law; one’s past behavior, statements, and so on, are all data from which an abductive inference regarding the defendant’s intentions is made. An analogous process also what happens when we judge that an apple has expired only from the fact that we bought it a while ago.
At an ultimate level of certainty, there exists direct observation of the concrete phenomena associated with insincerity. This method depends on fairly novel work being done in the empirical sciences, and so is obviously imperfect, but recent research using brain imagining has been able to detect lies with a rate of accuracy verging on 90%. At present, this method is infeasible (especially at a mass scale), but can be regarded as analogous to the role that the karyotype test plays for someone who thinks that gender reduces to chromosomes: it is a nuclear option, one which is in almost all cases unnecessary, that depends on recent or burgeoning advancements in biology. Continuing the analogy with the perception of fruit, this would be like taking a rotten apple to a lab to determine the chemical features of its flesh and the wavelengths of light reflected by its surface. Quite obviously far too much, but its in-principle possibility might comfort some.

All of this is well to the philosophers, and they turn to the next candidate.

#2: penelope and dispositions

Penelope is a citizen of Themiscyra who went on a business trip to Oregon before the rules regarding entry changed. Not knowing that she had to assert her identity as a woman upon entry, Penelope walked past the guards silently. She was seized, and has been held for the past week.

That Penelope should be admitted is obvious: it’s highly implausible (but, I guess, not out of the question) that a trip to Oregon changed her gender. Though, whether or not Penelope should count as a woman on the present definition is a matter of how we understand the meaning of the verb ‘identify’ therein. One possible way is to require that one presently be identifying as a woman to count as a woman: anyone who is not speaking is not right now a woman. Another is to only require that one, in general, be disposed to identify as a woman. This second sense is the more plausible reading. the Themiscyrnan’s original definition is written in the simple present, the marker for the habitual aspect in English, and thus seems to signify a dispositional state rather than an occurrent one. If the first meaning were intended, then the present progressive would have likely (though not necessarily) been chosen by the Themiscyrans instead. To sidestep this issue over interpretation, the Themiscyran philosophers just add a clause to their definition which states the dispositional interpretation explicitly. The philosophers allow in Penelope, write down their new definition, ‘a woman is someone who is disposed to sincerely identify as a woman (when asked for their gender),’ and turn to the next candidate.

#3: lilith and the past (present, and future)

Lilith is a woman who has lost her voice. She has identified as a woman in the recent past, has she/her/hers in her email signature, and has had all of her gender markers changed to F. Despite this, she is currently unable to speak, and thus cannot (be disposed to) identify as anything.

Lilith should be allowed, the Amazonians surmise, and another modification to the law is required. This time, the Amazons have more trouble, though they deliberate upon a preliminary solution after a while. If someone has ever fulfilled their criteria, then they are to be counted as a woman. After hearing of this modification, the honest Lilith musters up the voice to tell the Amazonians that, while she has identified as a woman before, she has also identified as a man.

“If you had asked me before age 25,” she says, “I would have told you that I was a man, and sincerely so. I simply had not realized that I was a woman yet. Proverbially, I was in my ‘egg.’” After stating this, Lilith’s vocal cords give out once again. Now the Amazonians are in a pickle: they can’t flatly use past self-identification, as this would yield the incorrect result that Lilith is both a man (as she so-identified before 25) and a woman (as she so-identified after 25). They return to the drawing board. A few options are considered, and while some might fare well, the Amazonians discard them:

One’s most recent self-identification is gender-determining. So, in spite of Lilith’s insistence that she was merely an enshelled woman before age 20, she was a man regardless. Rather than respecting Lilith’s sense that she had always been a woman, this proposal supposes that she only became one after the started identifying as one, and swapped genders at age 20. This supposition rubs some Amazonians the wrong way: we should, they think, respect Lilith’s sense that she had been a woman all along.
One’s final self-identification (that is, the one most immediately prior to one’s death, not the most recent) is gender-determining. So, supposing that Lilith continues to identify as a woman until she dies, then she is to be presently counted as one. But if her identification changes, (say, at age 50, she begins to identify as non-binary) then all prior self-identifications would have been erroneous, and she would have really been non-binary for her whole life. The presentists among the Amazons balk at this: to them, there isn’t (at least, not yet) any future act of self-identification. This view presupposes that there is a fixed and determinate future wherein this deathbed act of self-identification takes place. The eternalists too find themselves skeptical of the proposal, only because it significantly muddies their epistemology of gender; how are we to know how Lilith will identify on her deathbed?

To balance the strengths of these two options, the Amazonians concoct a view on which Lilith’s identification retroactively changes her past gender. On this view, Lilith’s identification as a woman at 25 reaches into the past, takes acts of identification that were true then, and makes them false now. So, Lilith-at-24 truthfully said “I am a man,” as her revision of her past identity had not yet occurred, but when she does revise her past identity, and says “I was never a man,” the statement “Lilith was a man at 24” is false thereafter. The truth of these statements, then, is indexed to the time at which they were spoken: ‘Lilith is a man at age 24’ is true-spoken-at-age-24, but false-spoken-at-age-25. And, so, Lilith is (attend to the present tense) correct that she was a woman at 24, was (attend to the past tense) correct that she was a man at 24 (though would now be incorrect to say so), and will be (attend to the future tense) correct to say ‘I was always nonbinary (and never a woman)’ if she so-identifies later in her life. Lilith’s entire biography, from the perspective of her at 25, is spent as a female; at 24, as a male; and at 50, nonbinary; and each of these perspectives are each equally valid and truthful at their respective times.

Aside from its verbal clunkiness, this view rests on a strange apparatus of backward causation that some Amazonians find unintuitive and spooky. The majority of Amazonians, though, don’t care about spookiness, and are willing to betray some intuitions for Lilith’s sake. Lilith expresses her gratitude to the philosophers before they turn to the next candidate.

#4: helena

Helena is a professor of philosophy who, one fateful evening, began a pilgrimage to the grave of Immanuel Kant. On their drive to the Königsberg Cathedral, however, they got into a gnarly crash and badly bonked their head on the dashboard. Wires crossed in their brain, and they formed the belief that they are Kant. They came to function as normal otherwise, though, after returning home, they began to write obstinately long sentences, drink a lot of coffee, and feel that they are the final, incorrigible authority on Kant interpretation. As a corollary of this delusional conviction, Helena believes as well that they are a man. They sincerely identify as a man because Kant was a man, and they are Kant, of course! Before their life as Kant, of which they have little memory, Helena happily identified as a cisgender woman.

Helena has kin, the philosophers notice. This area of the border hotel holds, in addition to Helena, people who were hypnotized into believing they are women (who would ever let that happen to them?!), transphobes who believe that they are women only because they have functioning ovaries, and brainwashed Christian fundamentalist women who believe that they’re women only because they’re meek and mild, as (they suppose) women essentially are. Each of these beliefs are false, and they each serve as the bases for the cases of self-identification parallel to Helena’s. The philosophers, having read their post-Gettier epistemology, know how to deal with each of these cases: a no-false-lemmas condition. Each of these persons whose identities are putatively problematic infer their gender from a false lemma, that is, they think that they’re a male or female based on a false belief. A no-false-lemmas clause counts such erroneous self-identification as non-gender-determining.

There is, of course, the question of what the actual gender of these persons is. If their self-identification is not gender-determining, then what determines their gender? Some fallback condition, like social role or biological makeup? Their most recent, non-erroneous self-identification? The philosophers take a moment to ponder this, cannot come upon an answer, and decide to see if the solution will present itself after seeing how other problem cases can be solved.

#5: theo

Theo is a Nordic Pagan, and he has been so for a while. It is his belief that, when the gods made him, they placed his male soul in a female body. This belief serves as a significant basis for his identification as a man, and, thus, as trans. Were he not a hellenic pagan, he would still have gender dysphoria, and would form the belief that he is a man on some other basis (on some secular grounds, like beliefs about prenatal hormone exposure and the sex of his brain, say).

Their no false lemmas clause, now, seems to pose a problem. The Amazonians are not all that syncretic, and are far from prone to admitting the truth of Norse Paganism. But, if this is so, then Theo infers his status as a man from an apparently false lemma (viz. Norse Paganism), and thus should not be counted as a man for the same reason as Helena and her friends. A needle must be thread, here, between false-lemma-borne self-identifications like Theo’s (which we do want to count as gender-determining) and those like Helena’s (which we do not want to count as gender-determining).

The philosophers take a step back and consider what it is that divides the false-lemma cases into kinds like Theo’s and kinds like Helena’s. It is plausible that, if Theo’s epistemic condition improved, and he came to realize (what the Amazonians take to be) his spiritual errors, his self-identification would nonetheless remain the same. The same is not so for Helena: her self identification did change when her false lemma was introduced, and it plausibly would change back if it were removed. This intuitive distinction is characterized (in the precise legal speak the Themiscyran legislature requires) in terms of one’s epistemically ideal counterpart, a hypothetical copy of oneself lacking any relevant false lemmas. Call this counterpart S+. In Theo’s case, Theo+ would not be a nordic pagan, but a worshipper of Artemis, and in Helena’s, Helena+ would believe that she is Helena, and not Immanuel Kant. Rather than taking one’s actual self-identification as flatly gender-determining, one’s gender is determined rather by how one’s plus-self would identify. So, since Theo+ would identify as a man (as actual-Theo does), Theo is counted as a man by the Amazonians, and, since Helena+ would identify as a woman (contrary to how actual-Helena does, though), Helena would count as a woman.

Pesky epistemological issues once again rear their head. Specifically, how do we know how our plus-selves would identify? The Themiscyrans have some balm for these worries:
Firstly, the language used here may be reminiscent of remarkably unpedestrian theoretical entities: ones that are perfectly impartial, have no false beliefs, have perfect imaginative capacities, and so on. The idealizations proposed by the Themiscyrans, however, only have (at worst) a few less false beliefs than their counterparts. Theo+ is different from actual-Theo only in not believing his maleness consists in his possession of a male soul, for example. Theo+, in all other respects, is identical to Theo. In many cases, we simply are our plus selves. The case of Theo and Helena are fairly out-there for this reason, and the requirement to idealize in this way is incurred only by these rare, marginal cases.

Secondly, even in cases where false lemmas are present and one differs from one’s plus-self, is it not that hard to judge the way their plus-self would identify. This is a simple extension of our common ability to assess counterfactuals in the case of beliefs we do not actually have: if you believed this post were cursing you, you’d stop reading; if you believed that the room you’re in were on fire, you’d leave; and if you believed that you would die tomorrow, you’d be spending your time on something cooler than this.

#6: marie-pier

Marie-Pier Côté is a monolingual Québécoise. After a dreary Canadian January, she swore to herself (in French, the only language she knows), that this was the last winter she’ll tolerate in Quebec. She organized her papers, threw herself onto a plane, flew to Asia Minor, and approached the gates of Themiscyra. Upon her arrival, the guards asked her (in English) if she is a woman. She said, “Euhhh, faque, je suis une femme, tsé.” No amount of inquiry yields an answer that the guards have been demanded to expect (these being ‘I am a woman,’ or just ‘woman’), so the guards offer Madame Côté a stay at the border hotel until her case can be analyzed by the philosophers.

Initially, the philosophers propose a simple modification to their definition stipulating that all non-English acts of self-identification synonymous to ones which are gender-determining in English count as gender-determining in their respective languages as well. So, ‘Ich bin eine Frau,’ ‘soy una mujer,’ and ‘je suis une femme’ all count as equally gender-determining, as they all share the same meaning, stateable in English as ‘I am a woman.’ This modification, however, fails in a subtle way. Note, firstly, that the oeuvre of the prior post on self-identification (‘ceci n’est pas une mise en abyme’) was establishing that (despite appearances) no prior understanding of the meaning of ‘woman’ is required to understand self-id’s proffered definition for ‘woman.’ But, surely, in adding a clause to our definition which refers to expressions synonymous to ‘I am a woman,’ we are reintroducing the need to have a prior understanding of ‘woman’’s meaning. Knowing whether ‘femme’ and ‘mujer’ and ‘woman’ are all synonymous requires apprehension of each term’s meaning, as the judgement that they are all synonymous amounts to a judgement about (the identity of) their meanings.

For this reason, the philosophers scrap the initial proposal and explore an alternative. Rather than make explicit reference to synonymy, the philosophers can exploit the fact that a French speaker prone to say ‘je suis une femme’ would also be prone, if they spoke English, to say the synonymous ‘I am a woman.’ Capitalizing on this fact (without introducing the concept of synonymy) requires introducing another counterfactual element to the definition, namely, consideration of what one would say if one could self-identify in English. So, because Madame Côté would be prone to say ‘I am a woman’ if she did speak English, she is counted, on the newly proposed modification, as a woman. There is, of course, the question of how we know what Madame Côté would say, but this question does not seem so hard to answer: multilingual people (whether hypothetical or actual) prone to assenting to a non-English sentence would be prone to assenting to a synonymous English sentence given knowledge of both languages. It is not so hard to judge that someone would be prone to say ‘schnee ist weiss’ and ‘la neige est blanche,’ if they spoke German or French respectively, given that they assent to the English ‘snow is white.’ So goes Madame Côté’s self-identification as a ‘femme’ and her counterfactual identification as a woman.

Briefly, note that the revision made in light of the case of Madame Côté also solves other cases involving the inability to self-identify, like muteness, infancy, a perpetual coma, cognitive disability, etc. In some of these cases, more epistemological issues arise, but the sort of reasoning from analogy involved in Madame Côté’s case (or the sorts of abductive reasoning discussed in Brian’s case) can be applied equally here.

conclusion

In sum, the Amazonians end up with a definition of ‘woman’ reading something like this:

One is a woman iff one’s epistemically ideal counterpart would be disposed to sincerely identify as a woman if they spoke English,

where one’s epistemically ideal counterpart is a hypothetical copy of oneself who lacks all relevant false lemmas. This is somewhat complex and wordy, but the world is fairly complex, and we’ll end up needing to get wordy to capture this complexity anyways. The epistemic concerns have been a theme throughout the post, and, in concluding, I want to flag a few of its aspects. Firstly, I think it is a merit of my view that it can account for the first-person experience of ignorance of one’s own gender. Some people have quite deep and difficult periods of self-discovery where they do not know what gender they are, and my view neatly explains this through the possible gaps between us and our epistemically ideal counterparts. This degree of fallibility when it comes to other people’s genders (or our own), to some, might be unsatisfying, but (for reasons covered in the first section) it seems to involve no more practical trouble than alternative views.