When someone makes a completely wrong assumption and a completely wrong deduction about you
But, dfranke didn’t do these things. He made a completely correct assumption based on his knowledge of LessWrong’s population, or on his prior for a random sample of the population. He didn’t know anything about the downvoter except that they downvoted—he had to guess at the rest of their characteristics, he chose (again, possibly not entirely consciously) to guess at their gender in order to express himself the way he wished to.
If he had known he was speaking of a transgendered downvoter, you would be justified in being angry. As he did not, you should not be angry. Note that in the past, commenters have been corrected on their usage of male gendered pronouns when explicitly referring to other posters who do not appreciate that practice, and these corrections have been upvoted—as I believe Alicorn may have mentioned.
If you wish to criticize the practice of using gendered pronouns in common communication, you may do so; LessWrong is already partial to this argument, but it’s not a community norm, so indignation is not the correct response.
“Correct” clearly is not, in shokwave’s statement, intended to refer to its truth value. There are other forms of correctness, such as obeying the rules.
I’m slightly saddened that you were so rapidly upvoted for lazily mis-applying Eliezer cites as a substitute for doing your own thinking.
I’ll take the cites one by one.
The first cite (which you badly paraphrased) concerns people who fail to follow Bayesian rationality and who attempt to excuse this by saying that they followed some social-custom-based rationality. His point in this article is that rationality, true rationality, is not a matter of social rules.
You have mis-applied this. True rationality does not preclude having false factual beliefs on occasion. It is possible, for example, for the evidence to temporarily mislead an ideally rational person.
The second cite is an improvement on the first since you are providing a quote but still does not apply in the way you want it to. His point is that rationalists should win. This does not mean that rationalists will necessarily win every time. Rather, rationalists will win more often than non-rationalists. In Newcomb’s Problem Omega has a historical record of being wrong 1 times out of 100, so even if all players have so far been making the correct, rational choice (which is the one-box choice), then 1 out of 100 of them have still been losing. This does not demonstrate that those who picked the one-box and still lost were not, in fact, being rational on those occasions. That they were being rational is demonstrated by the fact that the ones who adopted their exact same strategy won 99 times out of 100.
In Newcomb’s Problem Omega has a historical record of being wrong 1 times out of 100, so even if all players have so far been making the correct, rational choice (which is the one-box choice), then 1 out of 100 of them have still been losing. This does not demonstrate that those who picked the one-box and still lost were not, in fact, being rational on those occasions. That they were being rational is demonstrated by the fact that the ones who adopted their exact same strategy won 99 times out of 100.
There is a level of rationality where when you know what happens 99 times out of 100, you can win 99 times out of 100. And then there is a higher level, where you figure out how to predict that 1 in 100 deviation, and win all the time. The protestations that you followed the rules when you were wrong are excuses not to pursue the higher level.
Seconded. This whole conversation appears to be the result of someone stepping on a social land mine, by using an incorrect pronoun. We’ve got people arguing he should have known it was there and detoured around it (presupposing gender is bad); people arguing that he acted correctly because the land mine was on the shortest path to his destination (Spivak pronouns are awkward); people arguing that it ought not to be a social land mine in the first place (the offense taken was disproportionate).
And now, it seems, we’ve gone meta and somehow produced an analogy to Newcomb’s problem. I still don’t understand that one.
I’m embarrassed to note that I misread “presupposing gender is bad” as (presupposing (gender is bad)) rather than ((presupposing gender) is bad), and was halfway through a comment pointing out that nobody was presupposing any such thing before I realized I was being an idiot.
The post I replied to was so ridiculous that I was forced to be the unreasonable one in order to fully communicate my distaste to JGWeissman.
As it turned out, it appears I misjudged—my comment was not as far out of line as I had thought it would be. If I could do this again, I would not have put the parenthetical disclaimer in.
All in the service of bizarrely attempting to refute my point, which is an elementary point about word meanings that any English speaker should be be aware of, that the word “correct” has more than one meaning, and that “factually true” is only one, and clearly not the one meant.
Did Eliezer write any post about the abuse of cleverness to promote stupidity? Seeing as citing Eliezer posts seems to be a shortcut to karma fortune.
You’ve retreated to the argument that there is no excuse for failing to be omniscient.
I’m sorry, but this simply does not fly in context. If you apply this in context, then the complaint that was raised against dfranke becomes a complaint that he failed to be omniscient. None of us are omniscient. It is arbitrary to single out and selectively attack dfranke for his failure to be omniscient.
I am not saying that dfranke should be forever banished to Bayesian Hell for his mistake, but he did make a mistake. Like I said:
It is true that people make mistakes, and we should be able to react by improving ourselves and moving on, but the first step in this process is to stop making excuses and admit the mistake.
I actually agree with you that dfranke may have made a mistake, but I disagree about the identity of the mistake. The possible mistake would be the inverse of what you have been arguing. You cited Eliezer posts to the effect that obeying social rules is no excuse for being irrational. But the purported problem here surely is that dfranke broke certain social rules—the purported rule to make no assumptions about a poster’s gender when referring to them. It is the breaking of social rules, not irrationality per se, that typically causes offense. And offense is what was caused here.
No, mere offense is not the problem. AstroCJ reports:
When someone calls me “he”, the strong and immediate association in my mind is that they are about to verbally abuse or assault me. “He” as a default might be lovely and convenient for cisgendered men, but it’s not polite to women, and it stumbles across very, very negative and visceral associations to transwomen.
It happens that transwomen who physically look like men get physically asaulted, by people who identify the transwomen as a (defective) man, and make a point emphasizing this identification during the assault. So when someone else identifies her as a man, she anticpates (through the representitive heurestic) that she is about to be assaulted. This anticipation, though irrational and inaccurate and possibly even contradicted by more accurate explicit beliefs, is highly stressful. This stress is a real consequence of the misidentification, and I think we should be able to recognize this consequence as a bad thing independantly of social rules.
If I understand correctly, you agree that dfranke made an actual mistake about what decision to make to get good consequences, rather than merely violating a social rule.
Given that context, is there anything I said in the previous discussion that you were previously confused about that you understand now, or any assertions you may have confidently made that you should now reconsider?
I am only agreeing to the specified point, which is that the stress caused to AstroCJ is a bad thing independently of social rules.
It does not follow that dfranke necessarily made a mistake of rationality, given what dfranke knew at the time, and even given what dfranke was responsible for knowing at the time (to take the criterion of responsibility up a notch).
Would it be a normal psychological reaction for dfranke now to feel guilt and apologize for the stress caused, even if dfranke has genuinely done nothing wrong? Maybe. Recall this post. Quoting:
Fourth, guilt sometimes occurs even when a person has done nothing wrong.
As a matter of fact—and here I’m re-introducing the idea of the social norm—it may be a social norm for dfranke now to apologize even if dfranke has done nothing wrong. Such a social norm could be built on top of the psychological regularity that Yvain pointed out.
Ok, so you previously said that you agree “that dfranke may have made a mistake”, and you now agree that this mistake was not a violation of social rules. You still assert that it was not a “mistake of rationality”.
Would you agree that it was a mistake that dfranke, and others who behave the same way, should take note of and avoid repeating in the future? Ultimately, my point is that whatever rules were correctly or incorrectly followed to lead to this bad outcome, the bad outcome should be a red flag that says we should try to understand what happened, and fix the rules or follow the rules better or whatever will work to not repeat the mistake.
The general problem with arguing that bad outcomes were not caused by a mistake is that whatever denotations you use to make it technically correct, it is bringing in connotations that there is nothing to fix, which is flat out false.
Ok, so you previously said that you agree “that dfranke may have made a mistake”, and you now agree that this mistake was not a violation of social rules.
No, I retract entirely the claim that he may have made a social mistake. I do not substitute for it any other claim.
You thought that dfrank had made a mistake of violating a social rule.
I argued that the mistake was not merely violation of a social rule.
You accepted my argument, thus modifying your belief to: there was no mistake?
Tabooing “mistake”, would you agree that a bad outcome occured, and that in the future we should make better decisions so that similar bad outcomes do not occur?
No, I did not accept your argument. I accepted a particular point that you raised. I quoted the point that I accepted
Fine, consider line 3 to be modified to say “You accepted a point in my argument” instead of “You accepted my argument”. I still want to know: is the result that you now believe there was no mistake? If so, how did that happen? If not, what was the result?
(Was that distinction really so important that it had to take the place of responding to my questions?)
And the really important part of that comment was:
Tabooing “mistake”, would you agree that a bad outcome occured, and that in the future we should make better decisions so that similar bad outcomes do not occur?
Tabooing “mistake”, would you agree that a bad outcome occured, and that in the future we should make better decisions so that similar bad outcomes do not occur?
I think that even ideal decisionmaking will, in the face of uncertainty, occasionally produce bad outcomes. Therefore the occurrence of a single bad outcome is not proof, and may not even be strong evidence, that a bad decision was made.
I think moreover that thinking about modifying the rules in the immediate wake of a specific bad outcome can be a dangerous thing to do, because the recency of the particular event will tend to bias the result toward avoiding that class of event, at the disproportionate expense of those who are inconvenienced or bothered by the imposition of the rule. I’m pretty sure that this class of bias has been named here before, though I don’t recall the name.
I think that even ideal decisionmaking will, in the face of uncertainty, occasionally produce bad outcomes.
The problem here is that what we are using is not even close to ideal. Yes, you should consider the reasons of why you made the decision the way you did, and how modifying it prevent the recent bad outcome may make you vulnerable to other bad outcomes. But that concern doesn’t mean that you should avoid even considering how to improve. It may also be that after looking for ways to improve you can’t figure anything out with acceptable tradeoffs. But you still should take note there is something you are dissatisfied with and would like third alternatives for.
the recency of the particular event will tend to bias the result toward avoiding that class of event, at the disproportionate expense of those who are inconvenienced or bothered by the imposition of the rule.
In this case, if it were an available action to make everyone feel more welcome in communities where they are not the dominant gender, at the expense of making everyone accept the inconvenience of learning new pronouns, taking that action would be a no brainer. The tradeoff is clear even before looking at the visceral physical fear our ignorance can cause in victims of those who are actively for less tolerant than ourselves.
In this case, if it were an available action to make everyone feel more welcome in communities where they are not the dominant gender, at the expense of making everyone accept the inconvenience of learning new pronouns, taking that action would be a no brainer. The tradeoff is clear...
Not without numbers. Would you prefer that one person be made to feel horribly unwelcome in an online community, or that 3^^^3 members of the community go to the trouble of using new pronouns?
Not without numbers. Would you prefer that one person be made to feel horribly unwelcome in an online community, or that 3^^^3 members of the community go to the trouble of using new pronouns?
Ok, in the real world where we are making the decisions I am talking about, there are not 3^^^3 people at all. Yes in that world I would say fine, the trivial convenience of those 3^^^3 trumps the inclusiveness for 1 person. But in the real world, there are about 7 billion people, and a substantial fraction of them are subject to the problems of exclusiveness.
Dfranke apologising would be faux pas. Or at least it would be a strategically poor social move.
Really? If I unintentionally do something to offend someone, I apologize. If that holds for unintentionally bumping into someone, or spilling coffee on their shoe, then as a logical extension it holds true for things I say, whatever medium I use to say them. The relevant aspect in this case isn’t what I say, it’s what effect that has. If I said (or wrote) something that seemed reasonable at the time, but offended someone or hurt their feelings, then I’m sorry to have hurt their feelings. I won’t necessarily censor myself forever after, or even change the things I say, but I will apologize because it’s a social ritual that hopefully makes me feel less guilty and the hurt/offended party feel less offended or hurt.
If that holds for unintentionally bumping into someone, or spilling coffee on their shoe, then as a logical extension it holds true for things I say, whatever medium I use to say them.
I would apologise for spilling coffee on someone but not in this situation. The analogy is not a good one and definitely not one of logical deduction! Some relevant factors:
Astro was being obnoxious and disrespectful. (Barring a couple of exceptions that would not apply in this case) apologising to people when they are being obnoxious and disrespectful legitimises people behaving that way to you.
This isn’t direct personal interaction going on in good faith. It’s an absurd public spectacle. It’s an entirely different situation and one in which people’s judgement changes drastically, losing perspective. An apology here wouldn’t just be
Give an inch and they’ll take a mile. See JGWeissman’s behaviour here with Constant for an illustration. An apology would be twisted into a confession of guilt. As though Dfranke actually did something wrong. (Apart from spam the forum with Qualia nonsense—I’d appreciate an apology for that!)
Dfranke didn’t call Astro a dude—it was a guess that it was even one distinct individual and picking an arbitrary gender for the hypothesised individual isn’t saying anything about Astro at all. In fact the unknown downvoter could just as easily have been me. My voting patterns (everything by Dfranke in this thread down whenever I noticed it) match exactly what he described.
Dfranke apologising would be a (minor) slight to all those who have defended him from perceived unjust accusations. The clear consensus (by voting pattern) is that Astro was behaving inappropriately and there was a solid base of support for Dfranke at least as far as pronoun use goes. You don’t undermine that without good reason.
Dfranke basically isn’t involved in this discussion. That’s a good way to be. Some people have taken it as an excuse to push their spivak related political agenda but he has chosen not to try to desperately justify himself. Staying uninvolved is a wise move and if he did choose to make a statement it would be significant primarily as a political feature, not an instrument of furthering interpersonal harmony.
If Dfranke did feel guilt (or, more realistically given that it would be a response to public criticism, shame) then that is a problem of miscalibrated emotions and not something to submit to. Guilt would not be serving him in this instance and he has the opportunity to release that feeling and move the stimulus response pattern (disapproval → shame → supplication) one step closer to extinction.
Even if an apology is met with approval in the moment it is not necessarily producing an overall good outcome for you. It may get an apparently encouraging response from a minority but would not lead to being treated with respect in the future either by those people doing the encouraging or by others. You apologise when you have actually done something wrong, not because someone else tries to emotionally bully you.
See JGWeissman’s behaviour here with Costanza for an illustration.
I may have missed something, but I think the bulk of the interaction was with me, though Costanza added a comment at the end. The username similarity is pure coincidence.
I may have missed something, but I think the bulk of the interaction was with me, though Costanza added a comment at the end. The username similarity is pure coincidence.
Astro was being obnoxious and disrespectful. This isn’t direct personal interaction going on in good faith. It’s an absurd public spectacle. It’s an entirely different situation and one in which people’s judgement changes drastically, losing perspective.
I guess maybe I did not read the entire comment string, since I didn’t notice any ‘obnoxious’ comments from Astro, or much of an ‘absurd public spectacle’. You may be right about that.
Dfranke basically isn’t involved in this discussion. That’s a good way to be. Some people have taken it as an excuse to push their spivak related political agenda but he has (wisely) chosen not to try to desperately justify himself.
Agreed!
Guilt would not be serving him in this instance and he has the opportunity to release that feeling and move the stimulus response pattern (disapproval → shame → supplication) one step closer to extinction.
I would still apologize. That is the person I’ve chosen to be (and by extension, the person I’ve chosen to represent myself as). It may not produce an overall ‘good’ outcome, but I’m not sure what you define as ‘good’. I’ve never been treated with disrespect by people I’ve apologized too.
It was the correct assumption to make, in precisely the same way that 2 for 1 odds on a coin flip is the correct bet to take. That is why I included the context of “given his knowledge or priors” directly after the part you quoted.
In any event, dfranke failed to multiply this small prior probability by the huge negative utility of bringing up associations in a transwoman of being cruelly treated as a defective male instead of the female she sees herself as. The art did not fail him in assigning low probability to the truth, he failed the art in not considering the potential consequences of low probability possibilities.
It is true that people make mistakes, and we should be able to react by improving ourselves and moving on, but the first step in this process is to stop making excuses and admit the mistake.
Ever played poker? You can tell if a player’s going to improve a lot or only a little by looking at whether they reward themselves for making the right play win or lose, or for winning the hand right play or no. Analogously, dfranke made the right play and got unlucky.
I can invoke selection effects and the dust specks vs torture post and especially a failure to multiply to explain why the disutility of accidentally insulting a transgendered person appears to outweigh the disutility of adopting a different communication style but does not, but you should be doing that for yourself.
It was a reasonable assumption, but not a “completely correct” one. Certainty, for example, wouldn’t be justified (but it wasn’t expressed either, this sub-discussion rather refers to shokwave’s “completely correct” characterization).
Interesting. I got completely stuck on shuffling around the existing words instead of looking for a substitute.
“Reasonable” may suffer from the same problem (immediately I can imagine “reasonable people don’t go around Xing all the Ys”) as correct, but to a lesser extent. At the very least, thanks for opening up my thought process on the matter.
Making assumptions usually trades off correctness for simplicity (which is often a good idea), raising merely likely to the status of certain. By its nature, making of assumptions won’t be characterized by “complete correctness”.
What I am aiming for is to be able to examine the process a person used in producing their assumption, compare it to a prototypical process that always produces the best possible assumption from all given knowledge, background knowledge, and prior distributions, and then be able to say “this person made the best possible assumption they could have possibly made under the circumstances”.
Something similar to how you can look at a person making a bet and say whether they have made that bet correctly or not—before they win or lose.
It might be that ‘correct’ is simply contraindicated with ‘assumption’ and I have to find another way to express this.
(I appreciate that you are taking the time to engage with me politely, especially after I have previously been (rightly or wrongly) impolite due to anger.)
dfranke didn’t make a “correct” assumption, they[1] made an “unnecessary” assumption. I find it really quite surprising and disheartening that the Less Wrong community doesn’t have an interest in making a habit of avoiding these—yes, even to the point of thinking for a tenth of a second longer when using vernacular speech. Good habits, people.
There are numerous other problems here; if the community assumes that everyone in the community is male, then the community is more likely to lose female (or third-gender members) - witness both Alicorn’s and my strong irritation at being misgendered. You might chose to ignore third-gender folk, since they’re not numerous, but ignoring the [potential] presence of the entire female gender is not healthy for the individual or for the community.
If I were strictly third gender and I had complained about someone referring to me as “he/she” or similar, then I think your point here would stand; the commenter would have signalled clearly that they had made no assumptions about my gender, even if they had also signalled at the same time that they had made assumptions about gender in general. I would then be being unreasonable.
Finally, “indignation is not the correct response” because “it’s not a community norm”. Since a good number of people are avoiding gendered assumptions whilst posting here, I think indignation might well be the only way to point out to some people just how rude they are being.
[1] Edited after Perplexed pointed out that dfranke had not explicitly identified as male.
dfranke didn’t make a “correct” assumption, he made an “unnecessary” assumption.
Excuse me, I know you are not the first person to use the pronoun ‘he’ regarding dfranke, but are you certain it is appropriate? (Incidentally, I did notice that you avoided making that assumption in your initial complaint about being labeled a ‘guy’. Has dfranke self-identified as male somewhere since then?)
I’m doing it egregiously and on purpose (if you doubt this read the first paragraph of this comment :D) to satisfy my sense of irony, to (perhaps unethically) see if I could trick other commentors into using the pronoun too, and because there is no possible way in which dfranke could hold me accountable for misidentifying his or her gender, given the debate that has sprung up.
You’re quite right; by paraphrasing shokwave in my rebuttal, I picked up a male pronoun. I’ve now edited the relevant comment to remove this. Thank you, on two levels.
EDIT: I didn’t actually consciously avoid it in my first post.
dfranke didn’t make a “correct” assumption, he made an “unnecessary” assumption.
I should have included “if he wished to gender his pronouns”. I meant to communicate that the assumption he made was the correct one given his information and priors at the time; I grant that it spilled over into saying that gendering his speech was a correct choice and I did not intend that.
I find it really quite surprising and disheartening that the Less Wrong community doesn’t have an interest in making a habit of avoiding these
Actually we do—as I said in the previous comment we are partial to this practice, but it is not (yet) a community norm the way that, say, having read the Sequences, or arguing in good faith allowing for the possibility of changing your mind is. I fully expect it will soon become a norm.
A note on indignation: although it’s a greasy social psychology point, indignation isn’t the correct response unless it is a community norm. Reacting indignantly to something which is normally reacted to neutrally or ignored marks you as the unreasonable one, instead of the person that casually insulted you. Of course, this is only where “correct response” means “response that achieves the goal you want”. (There’s another interpretation of “correct response” that would say that indignation is a correct response, and that it fails to achieve the goal you want is a fact about the environment, not about the response).
if the community assumes that everyone in the community is male, then the community is more likely to lose female or third-gender members
Given the concern that LessWrong already suffers from style and interest deficiencies in such respects, this is a crucial matter. I don’t know how to address it other than to increase my efforts to avoid gendered speech and more often point it out to others.
A note on indignation: although it’s a greasy social psychology point, indignation isn’t the correct response unless it is a community norm.
Leaving aside semantics around “correct,” I agree that getting indignant over X when most people around me think X is unobjectionable often has results I don’t want.
That said, sometimes things become community norms as a consequence of the expressed indignation of individuals and the community’s willingness to align with those individuals.
Predicting when that second result is likely is easy to get wrong. Sometimes it’s worthwhile just to try and see.
Yes. I feel that is an extension on my parenthetical about the other interpretation of correct response—that it could lead to changing the environment.
Predicting when that second result is likely is easy to get wrong.
I’d like to put it down in writing somewhere that I predict a community norm of using nongendered speech, at least on the level of the norm of “read the Sequences”, to be fully formed and applied by six months from now.
By nongendered to you mean ve, ver, vis? Conditional prediction: If there is a move away from “he,she,etc.” I predict “they/them/their” will dominate.
By nongendered speech, I mean speech that does not indicate male or female gender. So they/them/their, ve/ver/vis, ey, or any other gender-neutral pronouns. It also includes my preferred way of avoiding gendered speech—you, the poster, and using the poster’s name. Yes, it’s fairly broad :P
I totally endorse that, and I’m pretty good about it myself (at least, I think I am), but I’d be very surprised if it ever became a reliable LW community norm.
dfranke didn’t make a “correct” assumption, he made an “unnecessary” assumption.
It’s not completely unnecessary, it’s grammatically more convenient to use a specific gender. It’s a question of priorities in deciding what to say, not of factual knowledge. You would be incorrect to argue that no a priori knowledge about your gender exists, or that it doesn’t say “probably male”.
If someone wants to avoid specifying gender, ey have options.
Spivak pronouns look weird and are hard to read for most people who aren’t used to using them. Just use the singular they. Much simpler and has been used colloquially for centuries.
Edit: This seems to be just way too much drama. Can we all just agree that English is a sucky language and that no matter what we do we’re going to be using some kludge and just get along?
I got used to Spivak pronouns in less than a day. People generally are capable of learning new vocabulary, if we don’t indulge their excuse that they aren’t used to it.
[blanked]
I have no desire to continue this upsetting conversation.
But, dfranke didn’t do these things. He made a completely correct assumption based on his knowledge of LessWrong’s population, or on his prior for a random sample of the population. He didn’t know anything about the downvoter except that they downvoted—he had to guess at the rest of their characteristics, he chose (again, possibly not entirely consciously) to guess at their gender in order to express himself the way he wished to.
If he had known he was speaking of a transgendered downvoter, you would be justified in being angry. As he did not, you should not be angry. Note that in the past, commenters have been corrected on their usage of male gendered pronouns when explicitly referring to other posters who do not appreciate that practice, and these corrections have been upvoted—as I believe Alicorn may have mentioned.
If you wish to criticize the practice of using gendered pronouns in common communication, you may do so; LessWrong is already partial to this argument, but it’s not a community norm, so indignation is not the correct response.
You are declaring “completely correct” an assumption that turned out to be wrong.
“Correct” clearly is not, in shokwave’s statement, intended to refer to its truth value. There are other forms of correctness, such as obeying the rules.
There is no rule that excuses you for being wrong because you followed the rules.
“If you fail to achieve a correct answer, it is futile to protest that you acted with propriety.”
I’m slightly saddened that you were so rapidly upvoted for lazily mis-applying Eliezer cites as a substitute for doing your own thinking.
I’ll take the cites one by one.
The first cite (which you badly paraphrased) concerns people who fail to follow Bayesian rationality and who attempt to excuse this by saying that they followed some social-custom-based rationality. His point in this article is that rationality, true rationality, is not a matter of social rules.
You have mis-applied this. True rationality does not preclude having false factual beliefs on occasion. It is possible, for example, for the evidence to temporarily mislead an ideally rational person.
The second cite is an improvement on the first since you are providing a quote but still does not apply in the way you want it to. His point is that rationalists should win. This does not mean that rationalists will necessarily win every time. Rather, rationalists will win more often than non-rationalists. In Newcomb’s Problem Omega has a historical record of being wrong 1 times out of 100, so even if all players have so far been making the correct, rational choice (which is the one-box choice), then 1 out of 100 of them have still been losing. This does not demonstrate that those who picked the one-box and still lost were not, in fact, being rational on those occasions. That they were being rational is demonstrated by the fact that the ones who adopted their exact same strategy won 99 times out of 100.
There is a level of rationality where when you know what happens 99 times out of 100, you can win 99 times out of 100. And then there is a higher level, where you figure out how to predict that 1 in 100 deviation, and win all the time. The protestations that you followed the rules when you were wrong are excuses not to pursue the higher level.
(I’ll gladly take the downvotes for this.)
what the hell is this crap
Seconded. This whole conversation appears to be the result of someone stepping on a social land mine, by using an incorrect pronoun. We’ve got people arguing he should have known it was there and detoured around it (presupposing gender is bad); people arguing that he acted correctly because the land mine was on the shortest path to his destination (Spivak pronouns are awkward); people arguing that it ought not to be a social land mine in the first place (the offense taken was disproportionate).
And now, it seems, we’ve gone meta and somehow produced an analogy to Newcomb’s problem. I still don’t understand that one.
I’m embarrassed to note that I misread “presupposing gender is bad” as (presupposing (gender is bad)) rather than ((presupposing gender) is bad), and was halfway through a comment pointing out that nobody was presupposing any such thing before I realized I was being an idiot.
I feel oddly compelled to confess to this.
I parsed it the same way, and did not even catch the mistake.
Where does this post fit into your ideas about community norms and indignation?
The post I replied to was so ridiculous that I was forced to be the unreasonable one in order to fully communicate my distaste to JGWeissman.
As it turned out, it appears I misjudged—my comment was not as far out of line as I had thought it would be. If I could do this again, I would not have put the parenthetical disclaimer in.
It would not surprise me if, had you posted it without the disclaimer, it would have been downvoted. Of course, I have no data to back that up.
I expect so too, but I would have gladly taken the downvotes for it, disclaimer or no.
All in the service of bizarrely attempting to refute my point, which is an elementary point about word meanings that any English speaker should be be aware of, that the word “correct” has more than one meaning, and that “factually true” is only one, and clearly not the one meant.
Did Eliezer write any post about the abuse of cleverness to promote stupidity? Seeing as citing Eliezer posts seems to be a shortcut to karma fortune.
Knowing about biases can hurt people is the first thing that comes to mind.
You’ve retreated to the argument that there is no excuse for failing to be omniscient.
I’m sorry, but this simply does not fly in context. If you apply this in context, then the complaint that was raised against dfranke becomes a complaint that he failed to be omniscient. None of us are omniscient. It is arbitrary to single out and selectively attack dfranke for his failure to be omniscient.
I am not saying that dfranke should be forever banished to Bayesian Hell for his mistake, but he did make a mistake. Like I said:
I actually agree with you that dfranke may have made a mistake, but I disagree about the identity of the mistake. The possible mistake would be the inverse of what you have been arguing. You cited Eliezer posts to the effect that obeying social rules is no excuse for being irrational. But the purported problem here surely is that dfranke broke certain social rules—the purported rule to make no assumptions about a poster’s gender when referring to them. It is the breaking of social rules, not irrationality per se, that typically causes offense. And offense is what was caused here.
No, mere offense is not the problem. AstroCJ reports:
It happens that transwomen who physically look like men get physically asaulted, by people who identify the transwomen as a (defective) man, and make a point emphasizing this identification during the assault. So when someone else identifies her as a man, she anticpates (through the representitive heurestic) that she is about to be assaulted. This anticipation, though irrational and inaccurate and possibly even contradicted by more accurate explicit beliefs, is highly stressful. This stress is a real consequence of the misidentification, and I think we should be able to recognize this consequence as a bad thing independantly of social rules.
All right, point taken.
If I understand correctly, you agree that dfranke made an actual mistake about what decision to make to get good consequences, rather than merely violating a social rule.
Given that context, is there anything I said in the previous discussion that you were previously confused about that you understand now, or any assertions you may have confidently made that you should now reconsider?
I am only agreeing to the specified point, which is that the stress caused to AstroCJ is a bad thing independently of social rules.
It does not follow that dfranke necessarily made a mistake of rationality, given what dfranke knew at the time, and even given what dfranke was responsible for knowing at the time (to take the criterion of responsibility up a notch).
Would it be a normal psychological reaction for dfranke now to feel guilt and apologize for the stress caused, even if dfranke has genuinely done nothing wrong? Maybe. Recall this post. Quoting:
As a matter of fact—and here I’m re-introducing the idea of the social norm—it may be a social norm for dfranke now to apologize even if dfranke has done nothing wrong. Such a social norm could be built on top of the psychological regularity that Yvain pointed out.
Ok, so you previously said that you agree “that dfranke may have made a mistake”, and you now agree that this mistake was not a violation of social rules. You still assert that it was not a “mistake of rationality”.
Would you agree that it was a mistake that dfranke, and others who behave the same way, should take note of and avoid repeating in the future? Ultimately, my point is that whatever rules were correctly or incorrectly followed to lead to this bad outcome, the bad outcome should be a red flag that says we should try to understand what happened, and fix the rules or follow the rules better or whatever will work to not repeat the mistake.
The general problem with arguing that bad outcomes were not caused by a mistake is that whatever denotations you use to make it technically correct, it is bringing in connotations that there is nothing to fix, which is flat out false.
No, I retract entirely the claim that he may have made a social mistake. I do not substitute for it any other claim.
Let me get this straight:
You thought that dfrank had made a mistake of violating a social rule.
I argued that the mistake was not merely violation of a social rule.
You accepted my argument, thus modifying your belief to: there was no mistake?
Tabooing “mistake”, would you agree that a bad outcome occured, and that in the future we should make better decisions so that similar bad outcomes do not occur?
No, I did not accept your argument. I accepted a particular point that you raised. I quoted the point that I accepted.
Fine, consider line 3 to be modified to say “You accepted a point in my argument” instead of “You accepted my argument”. I still want to know: is the result that you now believe there was no mistake? If so, how did that happen? If not, what was the result?
(Was that distinction really so important that it had to take the place of responding to my questions?)
And the really important part of that comment was:
I think that even ideal decisionmaking will, in the face of uncertainty, occasionally produce bad outcomes. Therefore the occurrence of a single bad outcome is not proof, and may not even be strong evidence, that a bad decision was made.
I think moreover that thinking about modifying the rules in the immediate wake of a specific bad outcome can be a dangerous thing to do, because the recency of the particular event will tend to bias the result toward avoiding that class of event, at the disproportionate expense of those who are inconvenienced or bothered by the imposition of the rule. I’m pretty sure that this class of bias has been named here before, though I don’t recall the name.
The problem here is that what we are using is not even close to ideal. Yes, you should consider the reasons of why you made the decision the way you did, and how modifying it prevent the recent bad outcome may make you vulnerable to other bad outcomes. But that concern doesn’t mean that you should avoid even considering how to improve. It may also be that after looking for ways to improve you can’t figure anything out with acceptable tradeoffs. But you still should take note there is something you are dissatisfied with and would like third alternatives for.
In this case, if it were an available action to make everyone feel more welcome in communities where they are not the dominant gender, at the expense of making everyone accept the inconvenience of learning new pronouns, taking that action would be a no brainer. The tradeoff is clear even before looking at the visceral physical fear our ignorance can cause in victims of those who are actively for less tolerant than ourselves.
Not without numbers. Would you prefer that one person be made to feel horribly unwelcome in an online community, or that 3^^^3 members of the community go to the trouble of using new pronouns?
Ok, in the real world where we are making the decisions I am talking about, there are not 3^^^3 people at all. Yes in that world I would say fine, the trivial convenience of those 3^^^3 trumps the inclusiveness for 1 person. But in the real world, there are about 7 billion people, and a substantial fraction of them are subject to the problems of exclusiveness.
I don’t see how this profits anyone. Constant has been precise enough already.
Dfranke apologising would be faux pas. Or at least it would be a strategically poor social move.
Really? If I unintentionally do something to offend someone, I apologize. If that holds for unintentionally bumping into someone, or spilling coffee on their shoe, then as a logical extension it holds true for things I say, whatever medium I use to say them. The relevant aspect in this case isn’t what I say, it’s what effect that has. If I said (or wrote) something that seemed reasonable at the time, but offended someone or hurt their feelings, then I’m sorry to have hurt their feelings. I won’t necessarily censor myself forever after, or even change the things I say, but I will apologize because it’s a social ritual that hopefully makes me feel less guilty and the hurt/offended party feel less offended or hurt.
(For the sake of abstract curiosity:)
I would apologise for spilling coffee on someone but not in this situation. The analogy is not a good one and definitely not one of logical deduction! Some relevant factors:
Astro was being obnoxious and disrespectful. (Barring a couple of exceptions that would not apply in this case) apologising to people when they are being obnoxious and disrespectful legitimises people behaving that way to you.
This isn’t direct personal interaction going on in good faith. It’s an absurd public spectacle. It’s an entirely different situation and one in which people’s judgement changes drastically, losing perspective. An apology here wouldn’t just be
Give an inch and they’ll take a mile. See JGWeissman’s behaviour here with Constant for an illustration. An apology would be twisted into a confession of guilt. As though Dfranke actually did something wrong. (Apart from spam the forum with Qualia nonsense—I’d appreciate an apology for that!)
Dfranke didn’t call Astro a dude—it was a guess that it was even one distinct individual and picking an arbitrary gender for the hypothesised individual isn’t saying anything about Astro at all. In fact the unknown downvoter could just as easily have been me. My voting patterns (everything by Dfranke in this thread down whenever I noticed it) match exactly what he described.
Dfranke apologising would be a (minor) slight to all those who have defended him from perceived unjust accusations. The clear consensus (by voting pattern) is that Astro was behaving inappropriately and there was a solid base of support for Dfranke at least as far as pronoun use goes. You don’t undermine that without good reason.
Dfranke basically isn’t involved in this discussion. That’s a good way to be. Some people have taken it as an excuse to push their spivak related political agenda but he has chosen not to try to desperately justify himself. Staying uninvolved is a wise move and if he did choose to make a statement it would be significant primarily as a political feature, not an instrument of furthering interpersonal harmony.
If Dfranke did feel guilt (or, more realistically given that it would be a response to public criticism, shame) then that is a problem of miscalibrated emotions and not something to submit to. Guilt would not be serving him in this instance and he has the opportunity to release that feeling and move the stimulus response pattern (disapproval → shame → supplication) one step closer to extinction.
Even if an apology is met with approval in the moment it is not necessarily producing an overall good outcome for you. It may get an apparently encouraging response from a minority but would not lead to being treated with respect in the future either by those people doing the encouraging or by others. You apologise when you have actually done something wrong, not because someone else tries to emotionally bully you.
I may have missed something, but I think the bulk of the interaction was with me, though Costanza added a comment at the end. The username similarity is pure coincidence.
That’s the one! Fixed.
I guess maybe I did not read the entire comment string, since I didn’t notice any ‘obnoxious’ comments from Astro, or much of an ‘absurd public spectacle’. You may be right about that.
Agreed!
I would still apologize. That is the person I’ve chosen to be (and by extension, the person I’ve chosen to represent myself as). It may not produce an overall ‘good’ outcome, but I’m not sure what you define as ‘good’. I’ve never been treated with disrespect by people I’ve apologized too.
It was the correct assumption to make, in precisely the same way that 2 for 1 odds on a coin flip is the correct bet to take.
It was the correct assumption to make, in precisely the same way that 2 for 1 odds on a coin flip is the correct bet to take. That is why I included the context of “given his knowledge or priors” directly after the part you quoted.
So what?
In any event, dfranke failed to multiply this small prior probability by the huge negative utility of bringing up associations in a transwoman of being cruelly treated as a defective male instead of the female she sees herself as. The art did not fail him in assigning low probability to the truth, he failed the art in not considering the potential consequences of low probability possibilities.
It is true that people make mistakes, and we should be able to react by improving ourselves and moving on, but the first step in this process is to stop making excuses and admit the mistake.
Ever played poker? You can tell if a player’s going to improve a lot or only a little by looking at whether they reward themselves for making the right play win or lose, or for winning the hand right play or no. Analogously, dfranke made the right play and got unlucky.
I can invoke selection effects and the dust specks vs torture post and especially a failure to multiply to explain why the disutility of accidentally insulting a transgendered person appears to outweigh the disutility of adopting a different communication style but does not, but you should be doing that for yourself.
It was a reasonable assumption, but not a “completely correct” one. Certainty, for example, wouldn’t be justified (but it wasn’t expressed either, this sub-discussion rather refers to shokwave’s “completely correct” characterization).
He completely correctly made the assumption that....
Would this phrasing illustrate the nuance I was aiming for better?
“Made the reasonable assumption” strikes me as most appropriate.
Interesting. I got completely stuck on shuffling around the existing words instead of looking for a substitute.
“Reasonable” may suffer from the same problem (immediately I can imagine “reasonable people don’t go around Xing all the Ys”) as correct, but to a lesser extent. At the very least, thanks for opening up my thought process on the matter.
Making assumptions usually trades off correctness for simplicity (which is often a good idea), raising merely likely to the status of certain. By its nature, making of assumptions won’t be characterized by “complete correctness”.
What I am aiming for is to be able to examine the process a person used in producing their assumption, compare it to a prototypical process that always produces the best possible assumption from all given knowledge, background knowledge, and prior distributions, and then be able to say “this person made the best possible assumption they could have possibly made under the circumstances”.
Something similar to how you can look at a person making a bet and say whether they have made that bet correctly or not—before they win or lose.
It might be that ‘correct’ is simply contraindicated with ‘assumption’ and I have to find another way to express this.
A 50% chance of each outcome of a coin toss would not count as an assumption about the outcome in my sense.
Works for me.
(I appreciate that you are taking the time to engage with me politely, especially after I have previously been (rightly or wrongly) impolite due to anger.)
dfranke didn’t make a “correct” assumption, they[1] made an “unnecessary” assumption. I find it really quite surprising and disheartening that the Less Wrong community doesn’t have an interest in making a habit of avoiding these—yes, even to the point of thinking for a tenth of a second longer when using vernacular speech. Good habits, people.
There are numerous other problems here; if the community assumes that everyone in the community is male, then the community is more likely to lose female (or third-gender members) - witness both Alicorn’s and my strong irritation at being misgendered. You might chose to ignore third-gender folk, since they’re not numerous, but ignoring the [potential] presence of the entire female gender is not healthy for the individual or for the community.
If I were strictly third gender and I had complained about someone referring to me as “he/she” or similar, then I think your point here would stand; the commenter would have signalled clearly that they had made no assumptions about my gender, even if they had also signalled at the same time that they had made assumptions about gender in general. I would then be being unreasonable.
Finally, “indignation is not the correct response” because “it’s not a community norm”. Since a good number of people are avoiding gendered assumptions whilst posting here, I think indignation might well be the only way to point out to some people just how rude they are being.
[1] Edited after Perplexed pointed out that dfranke had not explicitly identified as male.
Excuse me, I know you are not the first person to use the pronoun ‘he’ regarding dfranke, but are you certain it is appropriate? (Incidentally, I did notice that you avoided making that assumption in your initial complaint about being labeled a ‘guy’. Has dfranke self-identified as male somewhere since then?)
I’m doing it egregiously and on purpose (if you doubt this read the first paragraph of this comment :D) to satisfy my sense of irony, to (perhaps unethically) see if I could trick other commentors into using the pronoun too, and because there is no possible way in which dfranke could hold me accountable for misidentifying his or her gender, given the debate that has sprung up.
You’re quite right; by paraphrasing shokwave in my rebuttal, I picked up a male pronoun. I’ve now edited the relevant comment to remove this. Thank you, on two levels.
EDIT: I didn’t actually consciously avoid it in my first post.
I should have included “if he wished to gender his pronouns”. I meant to communicate that the assumption he made was the correct one given his information and priors at the time; I grant that it spilled over into saying that gendering his speech was a correct choice and I did not intend that.
Actually we do—as I said in the previous comment we are partial to this practice, but it is not (yet) a community norm the way that, say, having read the Sequences, or arguing in good faith allowing for the possibility of changing your mind is. I fully expect it will soon become a norm.
A note on indignation: although it’s a greasy social psychology point, indignation isn’t the correct response unless it is a community norm. Reacting indignantly to something which is normally reacted to neutrally or ignored marks you as the unreasonable one, instead of the person that casually insulted you. Of course, this is only where “correct response” means “response that achieves the goal you want”. (There’s another interpretation of “correct response” that would say that indignation is a correct response, and that it fails to achieve the goal you want is a fact about the environment, not about the response).
Given the concern that LessWrong already suffers from style and interest deficiencies in such respects, this is a crucial matter. I don’t know how to address it other than to increase my efforts to avoid gendered speech and more often point it out to others.
Leaving aside semantics around “correct,” I agree that getting indignant over X when most people around me think X is unobjectionable often has results I don’t want.
That said, sometimes things become community norms as a consequence of the expressed indignation of individuals and the community’s willingness to align with those individuals.
Predicting when that second result is likely is easy to get wrong. Sometimes it’s worthwhile just to try and see.
Yes. I feel that is an extension on my parenthetical about the other interpretation of correct response—that it could lead to changing the environment.
I’d like to put it down in writing somewhere that I predict a community norm of using nongendered speech, at least on the level of the norm of “read the Sequences”, to be fully formed and applied by six months from now.
By nongendered to you mean ve, ver, vis? Conditional prediction: If there is a move away from “he,she,etc.” I predict “they/them/their” will dominate.
By nongendered speech, I mean speech that does not indicate male or female gender. So they/them/their, ve/ver/vis, ey, or any other gender-neutral pronouns. It also includes my preferred way of avoiding gendered speech—you, the poster, and using the poster’s name. Yes, it’s fairly broad :P
Huh. Confidence interval?
I totally endorse that, and I’m pretty good about it myself (at least, I think I am), but I’d be very surprised if it ever became a reliable LW community norm.
It’s pretty uncalibrated but let’s say 90% confidence interval of 2 months to 2 years.
Upvoted for being willing to put numbers to it. I’ll never remember to come back and check, though.
It’s not completely unnecessary, it’s grammatically more convenient to use a specific gender. It’s a question of priorities in deciding what to say, not of factual knowledge. You would be incorrect to argue that no a priori knowledge about your gender exists, or that it doesn’t say “probably male”.
If someone wants to avoid specifying gender, ey has options.
Spivak pronouns look weird and are hard to read for most people who aren’t used to using them. Just use the singular they. Much simpler and has been used colloquially for centuries.
Edit: This seems to be just way too much drama. Can we all just agree that English is a sucky language and that no matter what we do we’re going to be using some kludge and just get along?
I got used to Spivak pronouns in less than a day. People generally are capable of learning new vocabulary, if we don’t indulge their excuse that they aren’t used to it.
You can learn to use new vocabulary, but this is not the same thing as adjusting aesthetic perception of its use.
It’s “ey has”.
Conjugation fixed.
I can personally attest that dfranke is male.