Title: A Generic Ontological Finickiness
-Date: 2025-09-14 11:00
+Date: 2025-09-14 13:00
Category: commentary
Tags: Eliezer Yudkowsky, categorization, sex differences
-Status: draft
On Twitter, [Eliezer Yudkowsky wonders](https://lightbrd.com/ESYudkowsky/status/1961782023573709003) why male-to-female transsexualism triggers more reactionary public opprobrium than female-to-male.
Yudkowsky proposes four potential explanatory factors, not mutually exclusive: (1) that modern liberal Society doesn't care about gatekeeping male roles the way its forbearers did, (2) that MtF triggers intuitions in men for the repression of specifically male homosexuality, (3) that FtM transitions are more effective on the merits of passability with existing technology, and (4) that MtF threatens sex-segregation conventions whose purpose is to protect females from males (as of, _e.g._, bathrooms and chess tournaments).
-I agree that to the extent that MtF faces more or different opposition than FtM, all four explanatory factors seem broadly plausible. I seem to disagree with Yudkowsky on the relative importance of the four factors, but the details are probably better relegated to a footnote.[^four-factors]
+I agree that to the extent that MtF faces more or different opposition than FtM, all four explanatory factors seem broadly plausible. I seem to disagree with Yudkowsky on the relative importance of the four factors and am cautiously skeptical of his reading of history, but the details are probably better relegated to a footnote.[^four-factors]
[^four-factors]: In [footnote 7 to "The Categories Were Made for Man to Make Predictions"](http://unremediatedgender.space/2018/Feb/the-categories-were-made-for-man-to-make-predictions/#the-categories-were-made-for-note-7) (February 2018), I mentioned that "trans men tend to pass better, and because insofar as the intended purpose of many sex-segregated social contexts is to protect females from males, biologically-female trans men aren't perceived as a threat: cis men are assumed to be able to take care of their own interests"—precisely anticipating Yudkowsky's theories #3 and #4. I hadn't previously considered theories #1 and #2.
- Yudkowsky seems to put much less weight on theory #4 than I do, saying he mentions it "for completeness" and that he finds it "dubious as a complete account, because of how FtM used to produce more outrage back when men were higher-status." I'm cautiously skeptical of Yudkowsky's read of history, which would greatly benefit from specific examples of alleged FtM panic. [Deborah Sampson](https://en.wikipedia.org/wiki/Deborah_Sampson) impersonated a man in order to serve in the American Revolutionary War, and successfully petitioned for back pay after being honorably discharged. Cases of females living as men like those of the surgeon [James Barry](https://en.wikipedia.org/wiki/James_Barry_(surgeon)), coachman [Charley Parkhurst](https://en.wikipedia.org/wiki/Charley_Parkhurst), or politician [Murray Hall](https://en.wikipedia.org/wiki/Murray_Hall_(politician)) seem to have been met with more curiosity than horror upon posthumous discovery.
+ Yudkowsky seems to put much less weight on theory #4 than I do, saying he mentions it "for completeness" and that he finds it "dubious as a complete account, because of how FtM used to produce more outrage back when men were higher-status." Yudkowsky's claims of alleged anti-FtM panic in history would greatly benefit from specific examples; it's not hard to find cases that imply a more nuanced picture of historical attitudes towards transmasculinity.
+
+ For example, [Deborah Sampson](https://en.wikipedia.org/wiki/Deborah_Sampson) impersonated a man in order to serve in the American Revolutionary War, successfully petitioned for back pay after being honorably discharged, and gave lectures about her military service. Other cases of females living as men, like those of stagecoach driver [Charley Parkhurst](https://en.wikipedia.org/wiki/Charley_Parkhurst) or politician [Murray Hall](https://en.wikipedia.org/wiki/Murray_Hall_(politician)) seem to have been met with more curiosity than horror upon posthumous discovery. [An 1880 obituary of Parkhurst in the _San Francisco Call_](https://timesmachine.nytimes.com/timesmachine/1880/01/09/98876579.pdf) described his career in glowing terms (and with masculine pronouns) before expressing outright admiration: "[t]hat a young woman should assume man's attire and, friendless and alone, defy the dangers of the voyage of 1849 to the then almost mythical California—dangers over which hardy pioneers still grow boastful—has in it sufficient of the wonderful." [A 1901 obituary of Hall in _The New York Times_](http://archive.today/2025.09.14-170351/https://www.nytimes.com/1901/01/19/archives/murray-hall-fooled-many-shrewd-men-how-for-years-she-masqueraded-in.html) expresses astonishment rather than condemnation: "How a woman could for so many years impersonate a man without detection, deceiving even her physician and some of the cleverest men and women in New York with whom she frequently came in contact, though the secret must have been known to at least two others—the wives—is a mystery quite as inexplicable as the character that accomplished the feat."
The occasion of the present post is that Yudkowsky goes on to say:
Yudkowsky acknowledges that his theories #1–4 aren't mutually exclusive. (Theory #2, for example, only tries to explain anti-MtF sentiment in men.) But _the same is true of theory #0_. Yudkowsky is correct to point out that theory #0 can't be the whole story of public anti-trans sentiment, because that would imply equal amounts of anti-MtF and anti-FtM sentiment, which isn't what we observe. But that only means that the philosophers are outnumbered by more political actors (like the gender-criticals and conservatives), which I don't think anyone would doubt—least of all the philosophers.
-The problem here is not just that no one said theory #0 was the whole story. Theory #0 _wouldn't_ be the whole story even if the people described by theory #0 were objectively correct in their views, because it's possible for humans to arrive at correct beliefs for bad reasons. (The correctness of the theory of evolution isn't the whole story for why people believe it: some people express pro-evolution sentiments to express resentment of their repressive religious upbringing. But evolution still actually happened.)
+The problem here is not just that no one said theory #0 was the whole story. Theory #0 _wouldn't_ be the whole story even if the people described by theory #0 were objectively correct in their views, because it's possible for humans to arrive at correct beliefs for bad reasons. (The correctness of the theory of evolution isn't the whole story of why it receives public support: some people express pro-evolution sentiments to express resentment of their repressive religious upbringing. But evolution still actually happened.)
-It's not clear why Yudkowsky would argue against the position that theory #0 by itself explains most public anti-trans sentiment (without pointing to any examples of anyone who thinks that!), unless his goal were to minimize the gender-political relevance of the of the philosophy of language—to tar anyone expressing alarm about inaccurate descriptions as rationalizing bigotries held for less high-minded reasons.
+It's not clear why Yudkowsky would argue against the position that theory #0 by itself explains public anti-trans sentiment (without pointing to any examples of anyone who thinks that!), unless his goal were to minimize the gender-political relevance of the of the philosophy of language—to tar anyone expressing alarm about inaccurate descriptions as rationalizing bigotries held for less high-minded reasons.
But the philosophy of language clearly _is_ relevant to what humans are doing when they argue about gender politics. [The cognitive function of categorization is to group decision-relevantly similar things into the same category in order to make similar decisions about them.](https://www.lesswrong.com/posts/esRZaPXSHgWzyB2NL/where-to-draw-the-boundaries) As a matter of AI theory, that's what human brains are doing when we use words, whether or not anyone knows it. (In the time of Aristotle, people didn't know it.)
In deriding the "generic hardness of ontological categories producing alarm about generic boundary violations" as something that "people reach[ ] for [...] afterward as a rationalization", Yudkowsky seems to portray philosophical arguments about accuracy of representations as somehow opposed to culture-specific sensitivities to specific misrepresentations—as if objecting to some lies more than others implies that it's hypocritical to have a philosophical account of lying.
-But a generic theory of deception is compatible with a policy that doesn't punish all deceptions equally: indeed, it's commonplace for punishments of deception to depend on who was misled about what. If you try to sell [pyrite (a.k.a. "fool's gold")](https://en.wikipedia.org/wiki/Pyrite) as gold, people will get angry at the attempted fraud. If you try to sell gold as pyrite, people will be confused but not angry.
+But that's wrong. Policy arguments and epistemic arguments are deeply intertwined, because a policy presupposes the existence of some reality that the policy is about. A generic theory of deception is compatible with a policy that doesn't punish all deceptions equally.[^kerr] Indeed, it's commonplace for punishments of deception to depend on who was misled about what. If you try to sell [pyrite (a.k.a. "fool's gold")](https://en.wikipedia.org/wiki/Pyrite) as gold, people will get angry at the attempted fraud. If you try to sell gold as pyrite, people will be confused but not angry.
+
+[^kerr]: In ["Challenges to Yudkowsky's Pronoun Reform Proposal"](/2022/Mar/challenges-to-yudkowskys-pronoun-reform-proposal/) (March 2022), I quoted Barra Kerr's article ["Pronouns are Rohypnol"](https://fairplayforwomen.com/pronouns/) and summarized Kerr's views as "saying that a pronoun convention that groups together females, and a minority of males who wish they were female, affects our cognition about that minority of males in a way that's disadvantageous to Kerr's interests (because she wants to be especially alert to threats posed by males), such that Kerr refuses to comply with that convention in her own speech."
+
+ Clearly, this is not "a generic ontological finickiness" of being "outraged by things being classified into the wrong intrinsic category." Kerr is not a philosopher standing up Truth for no other reason than its Truth. She's outraged specifically by men pretending to be women. But so what? She wasn't claiming to be a philosopher!
Nevertheless, pyrite is not gold.
People who claim that pyrite is gold because they define _gold_ as whatever sellers identify as gold, are lying or confused.
+All of this continues to be true if gold used to be cheap and pyrite used to be expensive.
+
If someone develops a new chemical treatment that alters pyrite to have some but not all of the properties of gold, and they trademark the altered substance as "trans gold" (two words), then someone could validly make a case on the empirical merits that there should be a sense of the word _gold_ such that trans gold is gold: "Make it yellower, increase the density, decrease the Mohs hardness, and I think most metallurgists would shrug and say 'good enough.'"
Fine. But it has to _actually_ be good enough. The shrug has to be genuine, not coerced: a shrug of "I can't be bothered to pay attention to this", rather than of accepting a social convention that it's none of your business.[^privacy]
In the comments on Twitter, [a user proposed a theory #5](https://lightbrd.com/pelvis_man/status/1961786999515545677) to Yudkowsky's puzzle, that FtMs are "indistinguishable from guys in the personality and mood and thought process and demeanor department. not so with mtf, they are more like a third different gender". [Yudkowsky replied](https://lightbrd.com/ESYudkowsky/status/1961797535938535858), "That's #3."
-(Theory #3 verbatim was that "[m]odern medtech makes it easier to go FtM, and the lack of FtM panic in modern times compared to earlier times is because in modern times the aspiring FtM can use T to actually grow a beard. [...] That is: The uncanny valley does not need to be symmetrical or fair with respect to perceptual cues or current medical technology. Maybe if you grow a beard, cut the hair, and remove the breasts, that just works to cross the valley.")
+(Theory #3 verbatim was that "[m]odern medtech makes it easier to go FtM, and the lack of FtM panic in modern times compared to earlier times is because in modern times the aspiring FtM can use T to actually grow a beard. (And acquire male psychology (anger), but this is not as visible.) That is: The uncanny valley does not need to be symmetrical or fair with respect to perceptual cues or current medical technology. Maybe if you grow a beard, cut the hair, and remove the breasts, that just works to cross the valley.")
-But theory #3's focus on visibility and perceptual cues is missing the commenter's point. If the medtech got much better, if it were cheap and easy to swap into a real female body, I would absolutely transition—and then that table at RustConf would have six or seven trans women. Even if there were no uncanny-valley effect, even if nothing looked visibly out of ordinary to the naked eye, The joint gender and assigned-sex-at-birth composition of that table would still _not be a coincidence_. Again, why?
+But theory #3's focus on visibility and perceptual cues is missing the commenter's point. If the medtech got much better, [if it were cheap and easy to swap into a real female body, I would absolutely transition](/2021/May/sexual-dimorphism-in-the-sequences-in-relation-to-my-gender-problems/)—and then that table at RustConf would have six or seven trans women. Even if there were no uncanny-valley effect, even if nothing looked visibly out of ordinary to the naked eye, The joint gender and assigned-sex-at-birth composition of that table would still _not be a coincidence_. Again, why?
It still feels ugly to say even now—it's not something I _want_ to be true—but if I'm being honest, I have to put a decent chunk of probability on the hypothesis that women on average have less of whatever [cognitive repertoires](/2020/Apr/book-review-human-diversity/) make people to want to spend their lives writing Rust code.[^implied-misogyny]
If the goal is to be a true sex change, then it should, right?
-But I don't think most trans women actually _want_ that! I think we like being the kind of status-seeking gynephilic systematizers that we largely are. [Without prematurely committing myself to any particular policy decision](/2021/Sep/i-dont-do-policy/), I think this has a lot of implications for how we should plan our lives and how the rest of Society should relate to us!
+But I don't think most trans women actually _want_ that! I think a lot of us like being the kind of status-seeking gynephilic systematizers that we often are. [Without prematurely committing myself to any particular policy decision](/2021/Sep/i-dont-do-policy/), I think this has a lot of implications for how we should plan our lives and how the rest of Society should relate to us!
Is this a "generic ontological finickiness", "a deep generic horror of things being described as other than what they are"?
It depends on what you mean by [_generic_](https://en.wiktionary.org/wiki/generic). I certainly don't spend an equal amount of time blogging about all possible [misrepresentations-via-suboptimal-categorization](https://www.lesswrong.com/posts/onwgTH6n8wxRSo2BJ/unnatural-categories-are-optimized-for-deception). That's obviously not possible and it's not even clear what it would mean. I selfishly focus on the topics that are important to me.
-In another sense, I suppose it is "generic" in the sense that I think [there are general laws of cognition](https://www.readthesequences.com/No-One-Can-Exempt-You-From-Rationalitys-Laws) governing [language](https://www.readthesequences.com/Mutual-Information-And-Density-In-Thingspace) and [thought](https://www.readthesequences.com/Conditional-Independence-And-Naive-Bayes), which I seek to understand and apply in my work in the hopes of improving the accuracy of shared maps. The methodology is general even if the choice of topic is idiosyncratic and personal.
+In another sense, I suppose it is "generic" in the sense that I do think [there are general laws of cognition](https://www.readthesequences.com/No-One-Can-Exempt-You-From-Rationalitys-Laws) governing [language](https://www.readthesequences.com/Mutual-Information-And-Density-In-Thingspace) and [thought](https://www.readthesequences.com/Conditional-Independence-And-Naive-Bayes), which I seek to understand and apply in my work in the hopes of improving the accuracy of shared maps. The methodology is general even if the choice of topic is idiosyncratic and personal.
Yudkowsky writes that he could only dream of a world where human beings cared that much about exact descriptive accuracy.
projects / Ultimately_Untrue_Thought.git / commitdiff