At the risk of facing a lot of potential backlash for this comment, I am just disappointed to see this discussion being upvoted so high in LW. I don’t think it is relevant or important for the vast majority of people. English happens to use gendered pronouns. Most people can be categorized into two categories and gendered pronouns refer to those categories. The fact that there are exceptions to the rule does not mean that those categories are not meaningful. Redefining the categories won’t change the perception of most people. Changing the language or the definitions won’t solve that most people will still perceive the world as divided roughly 50%/50% men and women because it is. And yet, most people will choose to treat trans people with respect and use whatever pronouns they prefer if they think it is important for them.
I don’t understand why you don’t think it is relevant. LessWrong is a site founded by Eliezer Yudkowsky, and dedicated to reasoning better. Is it not important to know when the founder (who is still an influential figure) of a community dedicated to a principle is being a poor example of that principle?
Maybe you are right, if we frame this issue on “Eliezer not following good rules” and not so much on “What pronouns we should use”, which in fact is what the OP intended I think.
I think OP is fairly clear about it being about the former, rather than the latter? E.g.
At this point, some readers may be puzzled as to the mood of the present post. I agree with Yudkowsky’s analysis of the design flaw in English’s pronoun system. I also agree that not misgendering trans people is a completely reasonable thing to do, which I also do. I’m only disputing the part where Yudkowsky jumps to declaring his proposed “simplest and best protocol” without acknowledging the ways in which it’s not simple and not unambiguously the best.
Many observers would consider this a very minor disagreement, not something anyone would want to spend 12,000 words prosecuting with as much vitriolic rhetoric as the target audience is likely to tolerate. If I agree with the problem statement (pronouns shouldn’t denote sex, that’s dumb; why would you define a language that way), and I don’t disagree with the proposed policy solution (don’t misgender trans people in public), why get so hung up on the exact arguments?
(I mean, besides the fact that it’s arguments that matter rather than conclusions, as a completely general principle of correct cognition.)
I guess for me, the issue is that this is a question where I need the correct reasoning in order to make extremely impactful social and medical decisions. Let me explain.
That would be easier to believe if Zack ever posted anything here that wasn’t overtly or covertly (1) about trans issues and (2) on the “trans Xs are not real Xs” side of the usual battle lines.
Zack is clever and insightful and I agree with most of the specific things he says. But I am wary of the fact that pretty much all he ever does here is to point out errors made in support of one particular side of a live sociopolitical debate.
That’s not necessarily a bad thing. For instance, it might be a good thing if the following things are true (which I suspect Zack thinks they are, and which it’s certainly not inconceivable that they might be):
That a lot of rationalists are systematically making a bunch of politically-motivated errors.
That there’s a single political motivation for a lot of these.
That rationalists collectively are making a lot fewer similar errors “on the other side”.
That this particular class of error is at the top of a list ordered by some combination of (1) intrinsic importance and (2) Zack’s ability to say relevant things.
Even so, the situation concerns me a bit.
[EDITED to add:] Concerns me because (perhaps counterintuitively) it’s possible for one’s thinking to be made worse overall by a process of correcting errors, if the choice of which errors has enough bias in it, so someone who corrects errors but only when they point a particular way is not necessarily doing us a favour. The point of the preceding bullet points is that they’re more likely to be doing us a favour if the bias they introduce points the opposite way from some preexisting bias.
In the present instance, I’m pretty sure Zack is right that Eliezer doesn’t really have (or at least shouldn’t have) such difficulty understanding how someone might have a stronger “this person doesn’t look like a she” sense than a “this person doesn’t look like an Oliver” sense; and that it’s better to have accurate information about the likely consequences of any given intervention (such as asking everyone to say “she” rather than “he” when referring to you).
But the final portion of Zack’s post feels off to me in very much the same way as Eliezer’s original comment does. Zack tells us that Eliezer’s wrongness on this point matters because of the possible harm some people in situations like Zack’s might suffer if they ask everyone to use particular pronouns for them while not appreciating the likelihood of awkwardness because the people doing so will often deep down not perceive them as being of the gender they want to be referred to as. (My apologies for the cumbersomeness of that sentence.) For this to be a good explanation of Zack’s 12k-word article, there would somehow need to be a non-negligible chance that Zack’s writing it leads to that not happening, and I can’t see at all how that’s plausible. What sequence of events would do that?
It seems to me that anyone contemplating asking other people to change what pronouns they call them by already knows that there will be awkwardness and already knows that the pronoun-change will not necessarily go along with an actual change in their instinctual judgement of the other person’s sex or gender. How’s any of that going to be changed by what Zack wrote here? Even if it completely changes your idea of Eliezer’s attitudes, that’s not going to make much difference to what you want people to call you unless most of your interactions with other people are specifically with Eliezer.
(Also: it seems to me hard to reconcile the stated motivation—save some gender-dysphoric people awkwardness by warning them that getting others to use different pronouns won’t actually change how the others see them—with the fact that a substantial chunk of the post is actually trying to argue that getting others to use different pronouns is a “typographic attack on [their] brains” that does to some extent change how they see the person they’re referring to with those pronouns. Or with the recurring “sucks to be you” motif which is clearly presenting the matter as a conflict between the interests of those seeking to be called by different pronouns and those of the people they’re asking to use those different pronouns.)
if Zack ever posted anything here that wasn’t overtly or covertly [...] pretty much all he ever does here
I dispute this. I think the relevant test is, “Is it plausible that I could have produced the text of this post in the counterfactual universe where I wasn’t on a gender-political crusade?”
(I find this counterfactual very easy to imagine, because from my perspective, the crusade feels like a defensive effort in response to cultural trends that stabbed me in the face after moving to Berkeley in 2016, not something I would have done if everyone else hadn’t “shot first”. I (of course) wouldn’t necessarily expect strangers on the internet to uncritically accept my self-report about this, but for some third-party-visible data, you can look at some pre-2016 posts on my real-name blog and imagine the same author writing more longform Less Wrong posts.)
When I ask this question about my Less Wrong non-linkposts over the last two years, I get:
So, as an estimate, I’d say it’s about a third of what I do here? It’s true that many of the posts on my Yes list clearly wouldn’t have happened if not for the crusade which happened in my actual history, but I’m claiming that the relevant test is whether they could have happened in the counterfactual where I had some other reason to get intensely interested in the philosophy of language.
As a separate illustration of the proposed methodology here, consider a post about how erasure codes work that I published in 2015 in honor of my employer at the time (RIP) integrating an erasure coding feature into our object storage product. Causally, the reason I happened to write about erasure codes in particular was because it was relevant to my employer; if I had worked somewhere else, I wouldn’t have written that exact post. But it doesn’t seem fair to accuse me of being a biased shill for Big Object Storage, because if you cut off the parts at the beginning and end talking up our product, the post is visibly the kind of innocent technical exposition I “would have done anyway” on some topic.
If it helps, I think I’m almost done?? (I need to do one more memoir-megapost telling the Whole Dumb Story about how I wasted the last six years of my life on this, but after that, I think my grievance with the so-called “rationalists” will be totally played out and I won’t be tempted to pollute this website with it anymore.)
(2) on the “trans Xs are not real Xs” side of the usual battle lines. [...] someone who corrects errors but only when they point a particular way is not necessarily doing us a favour
I strongly agree that selective argumentation is a problem, which is why I do put effort into searching for and publishing arguments on the other side of the usual battle lines, which you likely haven’t seen because I usually don’t linkpost or crosspost object-level stuff from my secret (“secret”) blog to Less Wrong unless I have a specific rationale for why it’s relevant to Less Wrong.[1]
So, my output certainly isn’t balanced around the usual battle lines (I definitely put out a lot more stuff that would code as “anti-trans” rather than “pro-trans” if you orthogonally project into the subspace of the usual battle lines, and it’s fine for you to notice that), but I also don’t think I should be optimizing for that kind of balance; I think I should be running a search of argument-space relevant to the questions I’m legitimately interested in, and I think I’m doing an okay (not perfect) job of that.
top of a list ordered by some combination of (1) intrinsic importance and (2) Zack’s ability to say relevant things.
It’s probably mostly (2). I don’t think I have any illusions about the issues I’m obsessed with being particularly important in the grand scheme of things. I do think they’re easy to get right, and that if people who are much smarter than me are repeatedly getting them wrong, that’s a warning sign for much deeper pathologies that could affect the things that do actually matter.
there would somehow need to be a non-negligible chance that Zack’s writing it leads to that not happening, and I can’t see at all how that’s plausible. What sequence of events would do that?
I think that person’s life would potentially be different if the Yudkowsky of 2016 was willing to say some of the things that the Yudkowsky of 2009 said that made a big difference to my life (even though—especially though—I didn’t want to hear it at the time). If Yudkowsky changed his mind between 2009 and 2016 because of new evidence, that would be one thing. If, on the other hand, he changed what he says in public because of new political incentives, that would be another thing altogether. (Lots more to say on this in a forthcoming memoir-megapost.)
anyone [...] already knows [...] How’s any of that going to be changed by what Zack wrote here?
clearly presenting the matter as a conflict between the interests of those seeking to be called by different pronouns and those of the people they’re asking to use those different pronouns
I think the conflict of interests is already there whether or not I write about it!
I argue that the present post is relevant because it takes a direct shot at Yudkowsky’s reputation, “Sexual Dimorphism” was relevant as commentary on the Sequences, the review of Charles Murray’s Human Diversity was relevant for recapping some of our core ideas in an novel context, the review of The Origins of Unfairness was relevant for being game theory, and “Blegg Mode” was relevant for being Sequences fanfiction. That’s five linkposts in three years, which I think is reasonable. If I were sharing everything from that blog here, that would definitely be a problem.
(Replying to note that Zack’s pushback on my “Zack pretty much only posts about gender-related things” seems to me entirely fair, and to apologize for posting what I did without actually checking the numbers. Obviously Zack’s comment deserves an actual proper reply, but I’ve been busy. Soon, I hope.)
So, first of all, I agree: not everything you write on LW is explicitly or implicitly part of what you call your gender-political crusade, and it was unfair of me to say that it is. I think I still think that said crusade is a sufficiently big part of your posting here that it’s reasonable to consider the OP to be somewhat “about” pronoun issues and not merely about a bit of bad thinking on Eliezer’s part that merely happens to involve pronoun issues. Backtracking a bit to see why that question was ever relevant: mukashi was saying OP doesn’t deserve all its upvotes because LW doesn’t need extensive discussions about pronouns; I think the fact that OP is inter alia about an apparent motivated-reasoning error of EY’s is sufficient explanation for why it’s here. I can’t agree with mukashi’s criticism. (If they’d said something more like “this topic is inflammatory and we could use less of it here even if it turns out to be leading EY to make errors of reasoning”, there’d be a stronger case though I don’t know whether I’d agree.)
I indeed hadn’t seen the UGS posts you mention that are “on the other side” (when one does the usual 1d projection, which I agree is as always a potentially dangerous simplification) -- I read your blog only occasionally. Actually, no, looking again I think I had seen the “Schelling point” one before and forgotten it, but not the “super-proton” one.
Your documented case of Yudkowsky-triggered transition is intriguing. Seems like the sort of thing where the difference he made was more “it happened a couple of weeks earlier” rather than “it happened and would otherwise not have happened”, though, don’t you think? And if the point here is that EY is a public figure, then there’s still at least one highly-attenuating step between your post and such changes. I dunno, it still just doesn’t feel plausible to me that your main reason for writing what you wrote was to save gender-dysphoric people from the awkwardness-for-them that might ensue if they started asking people to use different pronouns. If you say that really was your purpose—well, as I said before, I think I feel about it roughly the same way as you feel about Eliezer claiming not to know what it feels like to have a strong association of a particular pronoun with a particular person.
The “everybody knows” pathology you cite Zvi as describing is a reasonable thing to be concerned about. It seems to me that our case is, at most, an extremely non-central instance of what Zvi describes. Almost all of Zvi’s post is about cases where someone says “everybody knows X” when X is in fact false; in this case we agree that X is true (it’s something like “if you ask people to use non-obvious pronouns when referring to you, awkwardness may ensue”). But he does mention in passing the possibility of using “everyone knows X” to discourage telling people X (in cases where X is revealing a fraud and the person saying “everyone knows” is trying to suppress knowledge of the fraud; I mention this just because I’m sure you wouldn’t want the rather serious accusation you’re throwing at me to go unnoticed and I’d like to acknowledge that I noticed it). So I guess the question is: among people on LW who might read your 12k-word piece, how much underestimation do you actually think there is of the awkwardness that might ensue if they ask others to use nonobvious pronouns for them?
I completely agree that any conflict of interest between pronoun-requesters and pronoun-requestees is there regardless of whether you write about it. I’m not sure why you think that needs pointing out. The observation I was making is that it’s hard to reconcile (1) your stated motivation of making life less unpleasant for gender-dysphoric people by saving them from the awkwardness that might ensue if they ask others to use pronouns for them that match their internal gender-perception, with (2) the fact that a large fraction of what you actually wrote postulates an adversarial relationship between those people and the people they’re making the request of, and complains of the harm the former are inflicting on the latter.
(If the argument at the end of your post were “gender-dysphoric people need correct argumentation on this to make correct decisions because they might not ask people to use different pronouns if they appreciated the harm they were doing to those people by making the request” then there wouldn’t be that inconsistency. But it’s not, it’s “gender-dysphoric people need correct argumentation on this to make correct decisions because they might not ask people to use different pronouns if they appreciated the awkwardness they were bringing to themselves by making the request”.)
the difference he made was more “it happened a couple of weeks earlier” rather than “it happened and would otherwise not have happened”, though, don’t you think? [...] there’s still at least one highly-attenuating step between your post and such changes
The marginal impact of any one bullet on the outcome of a war is very small, but it would be very odd to therefore proclaim that it’s implausible that a soldier’s main reason for having taken a shot is to win the war. Of course, it’s true that one shot won’t make much of a difference, even if it hits. The soldier knows that, but fights anyway, because a tiny impact is nevertheless more than zero impact. Or maybe, because his decision to shoot is logically correlated with that of other soldiers. Or maybe—to die with dignity.
It’s the same thing with culture wars. (And with … culture-steering and knowledge-creation efforts that are hopefully doing something a little more sophisticated and productive than the usual one-dimensional war.) Nothing I write is going to have a huge impact on the world, because I’m very small in comparison to the world. I know that, but I write anyway. With dignity.
using “everyone knows X” to discourage telling people X (in cases where X is revealing a fraud and the person saying “everyone knows” is trying to suppress knowledge of the fraud
Yes, that’s exactly what I meant.
your main reason for writing what you wrote
Is that this is the continuation of an argument between me and Yudkowsky that has gone on for years. (Long, dumb story for a future memoir-post.) You can’t expect me to let him have the last word!
So I guess the question is: among people on LW who might read your 12k-word piece, how much underestimation do you actually think there is of the awkwardness that might ensue if they ask others to use nonobvious pronouns for them?
A lot, but mostly because people haven’t thought about the question, rather than because they’d necessarily get the wrong answer if prompted to spend five minutes thinking about it as measured by an actual, physical clock.
I have a dumb personal anecdote to explain where I’m coming from here. Back in the late ’aughts, there was a period when I tried using my first-and-middle-initials (“Z.M.”) as a nickname: my reasoning was, I wanted a gender-neutral byline (I didn’t like how “Zack” marked even my writing as male, even if I couldn’t expect people to not notice what sex I am in real life), and I wanted my byline to be the same as what people called me in real life, and I didn’t want to pick a new name unrelated to my legal name.
In retrospect, this turned out to be a terrible idea that caused my a huge amount of completely pointless identity-crisis emotional pain before I eventually ended up reverting it—partially because “Z.M.” never really “felt like a name”, even to me, and partially because of the backwards-compatibility problem (where I wasn’t comfortable being known by different names to different people, and I wasn’t bold enough to nag everyone who already knew me to switch, especially for something that didn’t really feel like a name, even to me).
(Also, “Zachary” is an order of magnitude more common than “Zoë” and “Zelda” put together, so the gender-neutral rationale almost certainly never held up in practice—but, you see, it was the principle.)
I think I would have made better decisions if I had read a careful 12,000-word blog post arguing that nickname changes are actually hard (especially if you anticipate not being comfortable being known by different names to different people, and aren’t bold enough to nag everyone who already knows you to switch) and that not all possible pairs of initials equally “feel like a name” to many native English speakers, even if using initials as a name isn’t uncommon for some pairs of initials.
(I’m actually still not sure what’s going on there psychologically! Why does “Z.M.” sound terrible, but “A.J.” or “J.T.” work? Does there need to be a J; is that the rule? Just “Z.” (zee) would have worked better …)
It’s not that I couldn’t have anticipated these points in advance, if I had spent five minutes with an actual, physical clock thinking of ways in which changing nicknames might be a bad idea. I just—didn’t think it through; I hadn’t considered the possibility that an idea that appealed to my ideological whimsy might be different from what I was actually happy living with.
The reason this dumb anecdote is relevant is because I think all the factors that caused me to underestimate the awkwardness of asking for a nonstandard nickname for gender-feeling-related reasons in 2007 (and therefore end up inflicting a lot of pointless identity-crisis emotional pain on myself) are substantially worse for people at risk of underestimating the awkwardness of asking for nonobvious pronouns for gender-feeling-related reasons in 2022. At least my dumb decisions of 2007 took some initiative on my part; no one pushed me.
I … don’t think this is true in the current year. If you don’t already see why, it’s probably more useful for me to explain at memoir-length rather than comment-length.
hard to reconcile (1) your stated motivation of making life less unpleasant for gender-dysphoric people
Hm, I don’t think I meant to come off as that altruistic. (I don’t think most gender-dysphoric people thinking under the distribution of ideologies in today’s Society would say I’m correctly advocating for their/our interests; a lot of them think I’m a traitor.) In my mind, the point was to explain my personal stake in getting this topic right. I’m open to wording suggestions if there’s some way to make my selfishness come through more clearly.
I think it’s a fair critique that it seems implausible for Zack to be simply motivated by wanting to help other gender-questioning people. (You don’t see many other gender-questioning people do the same as what Zack is doing, after all.) However, I don’t think this implies that one has to go all the way to “What pronouns we should use” as being what this post is about. (Heck, people on the anti-trans side of that question also don’t do what Zack is doing, showing that this motivation can’t explain Zack’s behavior either.)
The post is explicitly about pronouns, so I’m not sure what you mean by “I don’t think this implies …”.
(I am not suggesting that Zack’s purpose is to get everyone to refuse to call people by the pronouns they prefer; I take him at his word when he says he almost always goes along with people’s requests and thinks that others should generally do likewise. If anything I wrote sounds as I think he wants to change that, then I screwed up and I apologize.)
(I am not suggesting that Zack’s purpose is to get everyone to refuse to call people by the pronouns they prefer; I take him at his word when he says he almost always goes along with people’s requests and thinks that others should generally do likewise. If anything I wrote sounds as I think he wants to change that, then I screwed up and I apologize.)
Hm, then I’m not sure what you meant by
“What pronouns we should use”
as a topic.
Like if you see Zack taking a position on “What pronouns we should use”, and you don’t see him taking a position that pronouns should be used in accordance with biological sex, then what position do you see him taking?
I think we’re somehow at cross purposes. To whatever extent that’s my fault, please accept my apologies.
The phrase “what pronouns we should use” wasn’t mine. mukashi proposed two options for what Zack’s post is really about: “Eliezer not following good rules” and “what pronouns we should use”. I wouldn’t pick either of those exact phrases myself, and took them as gesturing at two broader possibilities: “about general principles of thinking” and “about gender issues in general and pronouns in particular”, and either way I couldn’t agree with your statement that it’s definitely “the former rather than the latter”: I think there’s as much of “the latter” as of “the former” in it.
More precisely, what I think is that Zack’s purpose is something like to correct any cognitive errors he sees that tend to encourage the idea that trans Xs are real Xs or should be thought of as such. Eliezer is a mere target of opportunity; pronouns happen to be the specific issue; Zack’s post is “about” both, though not exactly about Eliezer not following good rules, I think, and not exactly about what pronouns we should use.
I think Zack’s position on pronouns is that we should generally refer to people with the pronouns they prefer, that we should take care not to let that manipulate us into thinking that those pronouns are correct in any sense beyond social convenience, and that it would be better if fewer people whose self-image and what-Zack-considers-actual-sex diverge asked others to use what-Zack-considers-less-accurate pronouns for them. (More precisely, each bit of those is a thing I think to be Zack’s position, but I may have some parts wrong and there’s a very good chance that the whole thing therefore fails to be an accurate summary.)
The phrase “what pronouns we should use” wasn’t mine. mukashi proposed two options for what Zack’s post is really about: “Eliezer not following good rules” and “what pronouns we should use”.
Oh, my bad, I hadn’t realized you and mukashi were different people.
More precisely, what I think is that Zack’s purpose is something like to correct any cognitive errors he sees that tend to encourage the idea that trans Xs are real Xs or should be thought of as such. Eliezer is a mere target of opportunity; pronouns happen to be the specific issue; Zack’s post is “about” both, though not exactly about Eliezer not following good rules, I think, and not exactly about what pronouns we should use.
So I think there are two things to say here.
First, yes, Zack seems to be motivated partly by a point like this, more so than by general correctness (judging by the difficulty I’ve had getting him to engage with various critiques that are orthogonal to the whole trans issue). It seems like a general thing to me, that one cannot really get people to summon the energy to care about getting things correct if they don’t happen to be about topics that they care about.
But secondly, my impression is that this is less stuff that he wanted to write about, and more stuff that he felt forced to write about because people kept dismissing other of his concerns with “trans women are women”. If there’s a cognitive error that entirely shuts down discussion about a topic, then that seems like something worth addressing?
I think Zack’s position on pronouns is that we should generally refer to people with the pronouns they prefer, that we should take care not to let that manipulate us into thinking that those pronouns are correct in any sense beyond social convenience, and that it would be better if fewer people whose self-image and what-Zack-considers-actual-sex diverge asked others to use what-Zack-considers-less-accurate pronouns for them. (More precisely, each bit of those is a thing I think to be Zack’s position, but I may have some parts wrong and there’s a very good chance that the whole thing therefore fails to be an accurate summary.)
I don’t really like the phrasing “what-Zack-considers-actual-sex”. It seems to me that Zack’s position on “actual sex” is quite popular, even among trans people. Specifically, Zack and many others seem to favor something along the lines of taking a number of socially relevant sex characteristics, and defining actual sex to be the first principal component of those characteristics. (Essentially, the thingspace-cluster definition.)
This position has a number of challenges, which Zack mostly seems to bite the bullets on. Zack regularly points out that e.g. trans people cherry-pick what characteristics they use for defining the principal component, but that’s disagreements within this conception of sex; as long as the conception remains popular, it seems a bit sketchy to blame it on him.
I wasn’t intending “what-Zack-considers-” to come with an implied “wrongly”, and I’m sorry if it sounded as if I was. What I intended was merely for it not to come with an implied “rightly”, since pretty much everything in this area is controversial. I agree that something like (what I take to be) Zack’s understanding of “sex” is both reasonable and widely held.
At the risk of facing a lot of potential backlash for this comment, I am just disappointed to see this discussion being upvoted so high in LW. I don’t think it is relevant or important for the vast majority of people. English happens to use gendered pronouns. Most people can be categorized into two categories and gendered pronouns refer to those categories. The fact that there are exceptions to the rule does not mean that those categories are not meaningful. Redefining the categories won’t change the perception of most people. Changing the language or the definitions won’t solve that most people will still perceive the world as divided roughly 50%/50% men and women because it is. And yet, most people will choose to treat trans people with respect and use whatever pronouns they prefer if they think it is important for them.
I don’t understand why you don’t think it is relevant. LessWrong is a site founded by Eliezer Yudkowsky, and dedicated to reasoning better. Is it not important to know when the founder (who is still an influential figure) of a community dedicated to a principle is being a poor example of that principle?
Maybe you are right, if we frame this issue on “Eliezer not following good rules” and not so much on “What pronouns we should use”, which in fact is what the OP intended I think.
I think OP is fairly clear about it being about the former, rather than the latter? E.g.
That would be easier to believe if Zack ever posted anything here that wasn’t overtly or covertly (1) about trans issues and (2) on the “trans Xs are not real Xs” side of the usual battle lines.
Zack is clever and insightful and I agree with most of the specific things he says. But I am wary of the fact that pretty much all he ever does here is to point out errors made in support of one particular side of a live sociopolitical debate.
That’s not necessarily a bad thing. For instance, it might be a good thing if the following things are true (which I suspect Zack thinks they are, and which it’s certainly not inconceivable that they might be):
That a lot of rationalists are systematically making a bunch of politically-motivated errors.
That there’s a single political motivation for a lot of these.
That rationalists collectively are making a lot fewer similar errors “on the other side”.
That this particular class of error is at the top of a list ordered by some combination of (1) intrinsic importance and (2) Zack’s ability to say relevant things.
Even so, the situation concerns me a bit.
[EDITED to add:] Concerns me because (perhaps counterintuitively) it’s possible for one’s thinking to be made worse overall by a process of correcting errors, if the choice of which errors has enough bias in it, so someone who corrects errors but only when they point a particular way is not necessarily doing us a favour. The point of the preceding bullet points is that they’re more likely to be doing us a favour if the bias they introduce points the opposite way from some preexisting bias.
In the present instance, I’m pretty sure Zack is right that Eliezer doesn’t really have (or at least shouldn’t have) such difficulty understanding how someone might have a stronger “this person doesn’t look like a she” sense than a “this person doesn’t look like an Oliver” sense; and that it’s better to have accurate information about the likely consequences of any given intervention (such as asking everyone to say “she” rather than “he” when referring to you).
But the final portion of Zack’s post feels off to me in very much the same way as Eliezer’s original comment does. Zack tells us that Eliezer’s wrongness on this point matters because of the possible harm some people in situations like Zack’s might suffer if they ask everyone to use particular pronouns for them while not appreciating the likelihood of awkwardness because the people doing so will often deep down not perceive them as being of the gender they want to be referred to as. (My apologies for the cumbersomeness of that sentence.) For this to be a good explanation of Zack’s 12k-word article, there would somehow need to be a non-negligible chance that Zack’s writing it leads to that not happening, and I can’t see at all how that’s plausible. What sequence of events would do that?
It seems to me that anyone contemplating asking other people to change what pronouns they call them by already knows that there will be awkwardness and already knows that the pronoun-change will not necessarily go along with an actual change in their instinctual judgement of the other person’s sex or gender. How’s any of that going to be changed by what Zack wrote here? Even if it completely changes your idea of Eliezer’s attitudes, that’s not going to make much difference to what you want people to call you unless most of your interactions with other people are specifically with Eliezer.
(Also: it seems to me hard to reconcile the stated motivation—save some gender-dysphoric people awkwardness by warning them that getting others to use different pronouns won’t actually change how the others see them—with the fact that a substantial chunk of the post is actually trying to argue that getting others to use different pronouns is a “typographic attack on [their] brains” that does to some extent change how they see the person they’re referring to with those pronouns. Or with the recurring “sucks to be you” motif which is clearly presenting the matter as a conflict between the interests of those seeking to be called by different pronouns and those of the people they’re asking to use those different pronouns.)
I dispute this. I think the relevant test is, “Is it plausible that I could have produced the text of this post in the counterfactual universe where I wasn’t on a gender-political crusade?”
(I find this counterfactual very easy to imagine, because from my perspective, the crusade feels like a defensive effort in response to cultural trends that stabbed me in the face after moving to Berkeley in 2016, not something I would have done if everyone else hadn’t “shot first”. I (of course) wouldn’t necessarily expect strangers on the internet to uncritically accept my self-report about this, but for some third-party-visible data, you can look at some pre-2016 posts on my real-name blog and imagine the same author writing more longform Less Wrong posts.)
When I ask this question about my Less Wrong non-linkposts over the last two years, I get:
Yes (9): “Comment on ‘Deception as Cooperation’”, “Feature Selection”, “Communication Requires Common Interests or Differential Signal Costs”, “Message Length”, “Maybe Lying Can’t Exist?!”, “Algorithmic Intent”, “Optimized Propaganda With Bayesian Networks”, “Comment on ‘Endogenous Epistemic Factionalization’”, “Zoom Technologies, Inc. vs. the Efficient Markets Hypothesis”
No (4): “Blood Is Thicker Than Water”, “Reply to Nate Soares on Dolphins”, “Unnatural Categories Are Optimized for Deception”, “Philosophy in the Darkest Timeline” (this one was almost a Yes, but the “cult” jabs and neighbor’s playing-dumb behavior push it over the edge)
So, as an estimate, I’d say it’s about a third of what I do here? It’s true that many of the posts on my Yes list clearly wouldn’t have happened if not for the crusade which happened in my actual history, but I’m claiming that the relevant test is whether they could have happened in the counterfactual where I had some other reason to get intensely interested in the philosophy of language.
As a separate illustration of the proposed methodology here, consider a post about how erasure codes work that I published in 2015 in honor of my employer at the time (RIP) integrating an erasure coding feature into our object storage product. Causally, the reason I happened to write about erasure codes in particular was because it was relevant to my employer; if I had worked somewhere else, I wouldn’t have written that exact post. But it doesn’t seem fair to accuse me of being a biased shill for Big Object Storage, because if you cut off the parts at the beginning and end talking up our product, the post is visibly the kind of innocent technical exposition I “would have done anyway” on some topic.
If it helps, I think I’m almost done?? (I need to do one more memoir-megapost telling the Whole Dumb Story about how I wasted the last six years of my life on this, but after that, I think my grievance with the so-called “rationalists” will be totally played out and I won’t be tempted to pollute this website with it anymore.)
I strongly agree that selective argumentation is a problem, which is why I do put effort into searching for and publishing arguments on the other side of the usual battle lines, which you likely haven’t seen because I usually don’t linkpost or crosspost object-level stuff from my secret (“secret”) blog to Less Wrong unless I have a specific rationale for why it’s relevant to Less Wrong.[1]
Specifically, “Self-Identity Is a Schelling Point” is a novel consideration in favor of the self-identity criterion (the politically-clean generalization of which for Less Wrong was “Schelling Categories, and Simple Membership Tests”), and “On the Argumentative Form ‘Super-Proton Things Tend to Come In Varieties’” is rejecting an argument from Yudkowsky that supports my position that “dysphoria” is more than one taxon, because I think it’s a bad argument. (It seems plausible that he published that as a “concession” to my agenda in order to get me to stop emailing him—which makes it especially important that I published a criticism rejecting it, because I don’t want concessions; I want a discourse that actually gets the right answer.) I also want to write up a “pro-trans” argument based on the idea in Stuart Armstrong’s “Declustering, Reclustering, and Filling in Thingspace”, but I haven’t gotten around to it yet.
So, my output certainly isn’t balanced around the usual battle lines (I definitely put out a lot more stuff that would code as “anti-trans” rather than “pro-trans” if you orthogonally project into the subspace of the usual battle lines, and it’s fine for you to notice that), but I also don’t think I should be optimizing for that kind of balance; I think I should be running a search of argument-space relevant to the questions I’m legitimately interested in, and I think I’m doing an okay (not perfect) job of that.
It’s probably mostly (2). I don’t think I have any illusions about the issues I’m obsessed with being particularly important in the grand scheme of things. I do think they’re easy to get right, and that if people who are much smarter than me are repeatedly getting them wrong, that’s a warning sign for much deeper pathologies that could affect the things that do actually matter.
I think statements from public figures like Yudkowsky do affect transition decisions on the margin. (And that my reponses to those statements have effects on a much smaller margin, in proportion to my much smaller megaphone.) We actually have documentation of one case of this! Soon after a 2016 Facebook post in which Yudkowsky claimed that “for people roughly similar to the Bay Area / European mix, I think I’m over 50% probability at this point that at least 20% of the ones with penises are actually women”, he published a followup post reporting, “Just checked my filtered messages on Facebook and saw, ‘Your post last night was kind of the final thing I needed to realize that I’m a girl.’”
I think that person’s life would potentially be different if the Yudkowsky of 2016 was willing to say some of the things that the Yudkowsky of 2009 said that made a big difference to my life (even though—especially though—I didn’t want to hear it at the time). If Yudkowsky changed his mind between 2009 and 2016 because of new evidence, that would be one thing. If, on the other hand, he changed what he says in public because of new political incentives, that would be another thing altogether. (Lots more to say on this in a forthcoming memoir-megapost.)
Zvi Mowshowitz has written about how the claim that “everybody knows” something is typically used to silence attempts to tell the thing to people who don’t know. I think that applies here. In general, people don’t know things! People especially don’t know things that no one talks about on the grounds that everyone allegedly knows!
I think the conflict of interests is already there whether or not I write about it!
I argue that the present post is relevant because it takes a direct shot at Yudkowsky’s reputation, “Sexual Dimorphism” was relevant as commentary on the Sequences, the review of Charles Murray’s Human Diversity was relevant for recapping some of our core ideas in an novel context, the review of The Origins of Unfairness was relevant for being game theory, and “Blegg Mode” was relevant for being Sequences fanfiction. That’s five linkposts in three years, which I think is reasonable. If I were sharing everything from that blog here, that would definitely be a problem.
(Replying to note that Zack’s pushback on my “Zack pretty much only posts about gender-related things” seems to me entirely fair, and to apologize for posting what I did without actually checking the numbers. Obviously Zack’s comment deserves an actual proper reply, but I’ve been busy. Soon, I hope.)
(Terribly belated reply, sorry.)
So, first of all, I agree: not everything you write on LW is explicitly or implicitly part of what you call your gender-political crusade, and it was unfair of me to say that it is. I think I still think that said crusade is a sufficiently big part of your posting here that it’s reasonable to consider the OP to be somewhat “about” pronoun issues and not merely about a bit of bad thinking on Eliezer’s part that merely happens to involve pronoun issues. Backtracking a bit to see why that question was ever relevant: mukashi was saying OP doesn’t deserve all its upvotes because LW doesn’t need extensive discussions about pronouns; I think the fact that OP is inter alia about an apparent motivated-reasoning error of EY’s is sufficient explanation for why it’s here. I can’t agree with mukashi’s criticism. (If they’d said something more like “this topic is inflammatory and we could use less of it here even if it turns out to be leading EY to make errors of reasoning”, there’d be a stronger case though I don’t know whether I’d agree.)
I indeed hadn’t seen the UGS posts you mention that are “on the other side” (when one does the usual 1d projection, which I agree is as always a potentially dangerous simplification) -- I read your blog only occasionally. Actually, no, looking again I think I had seen the “Schelling point” one before and forgotten it, but not the “super-proton” one.
Your documented case of Yudkowsky-triggered transition is intriguing. Seems like the sort of thing where the difference he made was more “it happened a couple of weeks earlier” rather than “it happened and would otherwise not have happened”, though, don’t you think? And if the point here is that EY is a public figure, then there’s still at least one highly-attenuating step between your post and such changes. I dunno, it still just doesn’t feel plausible to me that your main reason for writing what you wrote was to save gender-dysphoric people from the awkwardness-for-them that might ensue if they started asking people to use different pronouns. If you say that really was your purpose—well, as I said before, I think I feel about it roughly the same way as you feel about Eliezer claiming not to know what it feels like to have a strong association of a particular pronoun with a particular person.
The “everybody knows” pathology you cite Zvi as describing is a reasonable thing to be concerned about. It seems to me that our case is, at most, an extremely non-central instance of what Zvi describes. Almost all of Zvi’s post is about cases where someone says “everybody knows X” when X is in fact false; in this case we agree that X is true (it’s something like “if you ask people to use non-obvious pronouns when referring to you, awkwardness may ensue”). But he does mention in passing the possibility of using “everyone knows X” to discourage telling people X (in cases where X is revealing a fraud and the person saying “everyone knows” is trying to suppress knowledge of the fraud; I mention this just because I’m sure you wouldn’t want the rather serious accusation you’re throwing at me to go unnoticed and I’d like to acknowledge that I noticed it). So I guess the question is: among people on LW who might read your 12k-word piece, how much underestimation do you actually think there is of the awkwardness that might ensue if they ask others to use nonobvious pronouns for them?
I completely agree that any conflict of interest between pronoun-requesters and pronoun-requestees is there regardless of whether you write about it. I’m not sure why you think that needs pointing out. The observation I was making is that it’s hard to reconcile (1) your stated motivation of making life less unpleasant for gender-dysphoric people by saving them from the awkwardness that might ensue if they ask others to use pronouns for them that match their internal gender-perception, with (2) the fact that a large fraction of what you actually wrote postulates an adversarial relationship between those people and the people they’re making the request of, and complains of the harm the former are inflicting on the latter.
(If the argument at the end of your post were “gender-dysphoric people need correct argumentation on this to make correct decisions because they might not ask people to use different pronouns if they appreciated the harm they were doing to those people by making the request” then there wouldn’t be that inconsistency. But it’s not, it’s “gender-dysphoric people need correct argumentation on this to make correct decisions because they might not ask people to use different pronouns if they appreciated the awkwardness they were bringing to themselves by making the request”.)
The marginal impact of any one bullet on the outcome of a war is very small, but it would be very odd to therefore proclaim that it’s implausible that a soldier’s main reason for having taken a shot is to win the war. Of course, it’s true that one shot won’t make much of a difference, even if it hits. The soldier knows that, but fights anyway, because a tiny impact is nevertheless more than zero impact. Or maybe, because his decision to shoot is logically correlated with that of other soldiers. Or maybe—to die with dignity.
It’s the same thing with culture wars. (And with … culture-steering and knowledge-creation efforts that are hopefully doing something a little more sophisticated and productive than the usual one-dimensional war.) Nothing I write is going to have a huge impact on the world, because I’m very small in comparison to the world. I know that, but I write anyway. With dignity.
Yes, that’s exactly what I meant.
Is that this is the continuation of an argument between me and Yudkowsky that has gone on for years. (Long, dumb story for a future memoir-post.) You can’t expect me to let him have the last word!
A lot, but mostly because people haven’t thought about the question, rather than because they’d necessarily get the wrong answer if prompted to spend five minutes thinking about it as measured by an actual, physical clock.
I have a dumb personal anecdote to explain where I’m coming from here. Back in the late ’aughts, there was a period when I tried using my first-and-middle-initials (“Z.M.”) as a nickname: my reasoning was, I wanted a gender-neutral byline (I didn’t like how “Zack” marked even my writing as male, even if I couldn’t expect people to not notice what sex I am in real life), and I wanted my byline to be the same as what people called me in real life, and I didn’t want to pick a new name unrelated to my legal name.
In retrospect, this turned out to be a terrible idea that caused my a huge amount of completely pointless identity-crisis emotional pain before I eventually ended up reverting it—partially because “Z.M.” never really “felt like a name”, even to me, and partially because of the backwards-compatibility problem (where I wasn’t comfortable being known by different names to different people, and I wasn’t bold enough to nag everyone who already knew me to switch, especially for something that didn’t really feel like a name, even to me).
(Also, “Zachary” is an order of magnitude more common than “Zoë” and “Zelda” put together, so the gender-neutral rationale almost certainly never held up in practice—but, you see, it was the principle.)
I think I would have made better decisions if I had read a careful 12,000-word blog post arguing that nickname changes are actually hard (especially if you anticipate not being comfortable being known by different names to different people, and aren’t bold enough to nag everyone who already knows you to switch) and that not all possible pairs of initials equally “feel like a name” to many native English speakers, even if using initials as a name isn’t uncommon for some pairs of initials.
(I’m actually still not sure what’s going on there psychologically! Why does “Z.M.” sound terrible, but “A.J.” or “J.T.” work? Does there need to be a J; is that the rule? Just “Z.” (zee) would have worked better …)
It’s not that I couldn’t have anticipated these points in advance, if I had spent five minutes with an actual, physical clock thinking of ways in which changing nicknames might be a bad idea. I just—didn’t think it through; I hadn’t considered the possibility that an idea that appealed to my ideological whimsy might be different from what I was actually happy living with.
The reason this dumb anecdote is relevant is because I think all the factors that caused me to underestimate the awkwardness of asking for a nonstandard nickname for gender-feeling-related reasons in 2007 (and therefore end up inflicting a lot of pointless identity-crisis emotional pain on myself) are substantially worse for people at risk of underestimating the awkwardness of asking for nonobvious pronouns for gender-feeling-related reasons in 2022. At least my dumb decisions of 2007 took some initiative on my part; no one pushed me.
I … don’t think this is true in the current year. If you don’t already see why, it’s probably more useful for me to explain at memoir-length rather than comment-length.
Hm, I don’t think I meant to come off as that altruistic. (I don’t think most gender-dysphoric people thinking under the distribution of ideologies in today’s Society would say I’m correctly advocating for their/our interests; a lot of them think I’m a traitor.) In my mind, the point was to explain my personal stake in getting this topic right. I’m open to wording suggestions if there’s some way to make my selfishness come through more clearly.
I think it’s a fair critique that it seems implausible for Zack to be simply motivated by wanting to help other gender-questioning people. (You don’t see many other gender-questioning people do the same as what Zack is doing, after all.) However, I don’t think this implies that one has to go all the way to “What pronouns we should use” as being what this post is about. (Heck, people on the anti-trans side of that question also don’t do what Zack is doing, showing that this motivation can’t explain Zack’s behavior either.)
The post is explicitly about pronouns, so I’m not sure what you mean by “I don’t think this implies …”.
(I am not suggesting that Zack’s purpose is to get everyone to refuse to call people by the pronouns they prefer; I take him at his word when he says he almost always goes along with people’s requests and thinks that others should generally do likewise. If anything I wrote sounds as I think he wants to change that, then I screwed up and I apologize.)
Hm, then I’m not sure what you meant by
as a topic.
Like if you see Zack taking a position on “What pronouns we should use”, and you don’t see him taking a position that pronouns should be used in accordance with biological sex, then what position do you see him taking?
I think we’re somehow at cross purposes. To whatever extent that’s my fault, please accept my apologies.
The phrase “what pronouns we should use” wasn’t mine. mukashi proposed two options for what Zack’s post is really about: “Eliezer not following good rules” and “what pronouns we should use”. I wouldn’t pick either of those exact phrases myself, and took them as gesturing at two broader possibilities: “about general principles of thinking” and “about gender issues in general and pronouns in particular”, and either way I couldn’t agree with your statement that it’s definitely “the former rather than the latter”: I think there’s as much of “the latter” as of “the former” in it.
More precisely, what I think is that Zack’s purpose is something like to correct any cognitive errors he sees that tend to encourage the idea that trans Xs are real Xs or should be thought of as such. Eliezer is a mere target of opportunity; pronouns happen to be the specific issue; Zack’s post is “about” both, though not exactly about Eliezer not following good rules, I think, and not exactly about what pronouns we should use.
I think Zack’s position on pronouns is that we should generally refer to people with the pronouns they prefer, that we should take care not to let that manipulate us into thinking that those pronouns are correct in any sense beyond social convenience, and that it would be better if fewer people whose self-image and what-Zack-considers-actual-sex diverge asked others to use what-Zack-considers-less-accurate pronouns for them. (More precisely, each bit of those is a thing I think to be Zack’s position, but I may have some parts wrong and there’s a very good chance that the whole thing therefore fails to be an accurate summary.)
Oh, my bad, I hadn’t realized you and mukashi were different people.
So I think there are two things to say here.
First, yes, Zack seems to be motivated partly by a point like this, more so than by general correctness (judging by the difficulty I’ve had getting him to engage with various critiques that are orthogonal to the whole trans issue). It seems like a general thing to me, that one cannot really get people to summon the energy to care about getting things correct if they don’t happen to be about topics that they care about.
But secondly, my impression is that this is less stuff that he wanted to write about, and more stuff that he felt forced to write about because people kept dismissing other of his concerns with “trans women are women”. If there’s a cognitive error that entirely shuts down discussion about a topic, then that seems like something worth addressing?
I don’t really like the phrasing “what-Zack-considers-actual-sex”. It seems to me that Zack’s position on “actual sex” is quite popular, even among trans people. Specifically, Zack and many others seem to favor something along the lines of taking a number of socially relevant sex characteristics, and defining actual sex to be the first principal component of those characteristics. (Essentially, the thingspace-cluster definition.)
This position has a number of challenges, which Zack mostly seems to bite the bullets on. Zack regularly points out that e.g. trans people cherry-pick what characteristics they use for defining the principal component, but that’s disagreements within this conception of sex; as long as the conception remains popular, it seems a bit sketchy to blame it on him.
I wasn’t intending “what-Zack-considers-” to come with an implied “wrongly”, and I’m sorry if it sounded as if I was. What I intended was merely for it not to come with an implied “rightly”, since pretty much everything in this area is controversial. I agree that something like (what I take to be) Zack’s understanding of “sex” is both reasonable and widely held.