In the response I would have wanted to see, Duncan would have clearly and correctly pointed to that difference. He is in favor of people asking for examples [combined with other efforts to cross the gap], does it himself, gives examples himself, and so on. The unsaid [without anything else] part is load-bearing and thus inappropriate to leave out or merely hint at. [Or, alternatively, using “ask people for examples” to refer to comments that do only that, as opposed to the conversational move which can be included or not in a comment with other moves.]
I agree that the hypothetical comment you describe as better is in fact better. I think something like … twenty-or-so exchanges with Said ago, I would have written that comment? I don’t quite know how to weigh up [the comment I actually wrote is worse on these axes of prosocial cooperation and revealing cruxes and productively clarifying disagreement and so forth] with [having a justified true belief that putting forth that effort with Said in particular is just rewarded with more branches being created].
(e.g. there was that one time recently where Said said I’d blocked people due to disagreeing with me/criticizing me, and I said no, I haven’t blocked anybody for disagreeing/criticizing, and he responded “I didn’t say anything about ‘blocked for disagreeing [or criticizing]’. (Go ahead, check!)” and the actual thing he’d said was that they’d been blocked due to disagreeing/criticizing; that’s the level of … gumming up the works? gish-gallop? … that I’ve viscerally come to expect.)
Like, I think there’s plausibly a CEV-ish code of conduct in which I “should”, at that point, still have put forth the effort, but I think it’s also plausible that the correct code of conduct is one in which doing so is a genuine mistake and … noticing that there’s a hypothetical “better” comment is not the same as there being an implication that I should’ve written it?
Something something, how many turns of the cheek are actually correct, especially given that, the week prior, multiple commenters had been unable, with evidence+argument+personal testimony, to shift Said away from a strikingly uncharitable prior.
Mine either, to be clear; I felt by that point that Said had willingly put himself outside of the set of [signatories to the peace treaty], turning down many successive opportunities to remain in compliance with it. I was treating his statements closer to the way I think it is correct to treat the statements of the literal Donald Trump than the way I think it is correct to treat the statements of an undistinguished random Republican.
(I can go into the reasoning for that in more detail, but it seems sort of conflicty to do so unprompted.)
If Said ate breakfasts of only cereal, and Duncan said that was unhealthy and he shouldn’t do it, it is not quite right to say Duncan ‘thinks you shouldn’t eat cereal’, as he might be in favor of cereal as part of a balanced breakfast; but also it is not quite right for Duncan to ignore Said’s point that one of the main issues under contention is whether Said can eat cereal by itself (i.e. asking for examples without putting in interpretative labor).
I’m a little lost in this analogy; this is sort of where the privileging-the-hypothesis complaint comes in.
The conversation had, in other places, centered on the question of whether Said can eat cereal by itself; Logan for instance highlighted Said’s claim in a reply on FB:
Furthermore, you have mentioned the “inferential gap” several times, and suggested that it is the criticizer’s job, at least in part, to bridge it. I disagree.
There, the larger question of “can you eat only cereal, or must you eat other things in balance?” is front-and-center.
But at that point in the subthread, it was not front-and-center; yes, it was relevant context, but the specific claim being made by Said was clear, and discrete, and not at all dependent-on or changed-by that context.
The history of that chain:
Said includes, in a long comment “In summary, I think that what’s been described as ‘aiming for convergence on truth’ is some mixture of” … “contentless” … “good but basically unrelated to the rest of it” … “bad (various proposed norms of interaction such as ‘don’t ask people for examples of their claims’ and so on)”
gjm, in another long comment, includes “I don’t know where you get ‘don’t ask people for examples of their claims’ from and it sounds like a straw man” and goes on to elaborate “I think the things Duncan has actually said are more like ‘Said engages in unproductive modes of discussion where he is constantly demanding more and more rigour and detail from his interlocutors while not providing it himself’, and wherever that lands on a scale from ’100% truth’ to ‘100% bullshit’ it is not helpful to pretend that he said ‘it is bad to ask people for examples of their claims’.
There’s a bunch of other stuff going on in their back and forth, but that particular thread has been isolated and directly addressed, in other words. gjm specifically noted the separation between the major issue of whether balance is required, and this other, narrower claim.
Said replied:
If “asking people for examples of their claims” doesn’t fit Duncan’s stated criteria for what constitutes acceptable engagement/criticism, then it is not pretending, but in fact accurate, to describe Duncan as advocating for a norm of “don’t ask people for examples of their claims”.
Which, yes, I straightforwardly agree with the if-then statement; if “asking people for examples of their claims” didn’t fit my stated criteria for what constitutes acceptable engagement or criticism, then it would be correct to describe me as advocating for a norm of “don’t ask people for examples of their claims.”
But like. The if does not hold. It really clearly doesn’t hold. It was enough of an out-of-nowhere strawman/non-sequitur that gjm specifically called it out as ”???” at which point Said doubled down, saying the above and also
Duncan has, I think, made it very clear that that a comment that just says “what are some examples of this claim?” is, in his view, unacceptable. That’s what I was talking about. I really do not think it’s controversial at all to ascribe this opinion to Duncan.
It seems like, in your interpretation, I “should” (in some sense) be extending a hand of charity and understanding and, I dunno, helping Said to coax out his broader, potentially more valid point—helping him to get past his own strawman and on to something more steel, or at least flesh. Like, if I am reading you correctly above, you’re saying that, by focusing in on the narrow point that had been challenged by gjm and specifically reaffirmed by Said, I myself was making some sort of faux pas.
(Am I in fact reading you correctly?)
I do not think so. I think that, twenty exchanges prior, I perhaps owed Said something like that degree of care and charity and helping him avoid tying his own shoelaces together. I certainly feel I would owe it to, I dunno, Eric Rogstad or Julia Galef, and would not be the slightest bit loath to provide it.
But here, Said had just spent several thousand words the week prior, refusing to be budged from a weirdly uncharitable belief about the internals of my mind, despite that belief being incoherent with observable evidence and challenged by multiple non-me people. I don’t think it’s wise-in-the-sense-of-wisdom to a) engage with substantial charity in that situation, or b) expect someone else to engage with substantial charity in that situation.
(You can tell that my stated criteria do not rule out asking people for examples of their claims in part because I’ve written really quite a lot about what I think constitutes acceptable engagement or criticism, and I’ve just never come anywhere close to a criterion like that, nor have I ever complained about someone asking for examples unless it was after a long, long string of what felt like them repeatedly not sharing in the labor of truthseeking. Like, the closest I can think of is this thread with tailcalled, in which (I think/I hope) it’s pretty clear that what’s going on is that I was trying to cap the total attention paid to the essay and its discussion, and thus was loath to enter into something like an exchange of examples—not that it was bad in any fundamental sense for someone to want some. I did in fact provide some, a few comments deeper in the thread, though I headlined that I hadn’t spent much time on them.)
So in other words: I don’t think it was wrong to focus on the literal, actual claim that Said had made (since he made it, basically, twice in a row, affirming “no, I really mean this” after gjm’s objection and even saying that he thinks it is so obvious as to not be controversial. I don’t think I “ought” to have had a broader focus, under the circumstances—Said was making a specific, concrete, and false claim, and his examples utterly fail to back up that specific, concrete, and false claim (though I do agree with you that they back up something like his conception of our broader disagreement).
I dunno, I’m feeling kind of autistic, here, but I feel like if, on Less Wrong dot com, somebody makes a specific, concrete claim about my beliefs or policies, clarifies that yes, they really meant that claim, and furthermore says that such-and-such links are “citations for [me] expressing the sentiment [they’ve] ascribed to [me]” when they simply are not—
It feels likeemphatic and unapologetic rejection should be 100% okay, and not looked at askance. The fact that they are citations supporting a different claim is (or at least, I claim, should be) immaterial; it’s not my job to steelman somebody who spent hours and hours negatively psychologizing me in public (while claiming to have no particular animus, which, boy, a carbon copy of Said sure would have had Words about).
I think there’s a thing here of standards unevenly applied; surely whatever standard would’ve had me address Said’s “real” concern would’ve also had Said behave much differently at many steps prior, possibly never strawmanning me so hard in the first place?
I think this situation is, on some level, pretty symmetric.
I think the features of Said’s commenting style that people (not just Duncan!) find annoying are things that Said is deliberately optimizing for or the results of principled commitments he’s made, so it’s not just a simple bug that can be fixed.
I think the features of Duncan’s conflict resolution methods that people find offputting are similarly things that Duncan is deliberately optimizing for or the results of principled commitments he’s made, so it’s not just a simple bug that can be fixed.
I think the asymmetry breaks in that, like, a bunch of people have asked Said to stop and he won’t; I’m quite eager to stop doing the conflict resolution that people don’t like, if there can pretty please be some kind of system in place that obviates it. I much prefer the world where there are competent police to the world where I have to fight off muggers in the alley—that’s why I’m trying so hard to get there to be some kind of actually legible standards rather than there always being some plausible reason why maybe we shouldn’t just say “no” to the bullshit that Zack or Said or anonymouswhoever is pulling.
Right now, though, it feels like we’ve gone from “Ben Hoffman will claim Duncan wants to ghettoize people and it’ll be left upvoted for nine days with no mod action” to “Ray will expound on why he thinks it’s kinda off for Said to be doing what he’s doing but there won’t be anything to stop Said from doing it” and I take Oli’s point about this stuff being hard and there being other priorities but like, it’s been years. And I get a stance of, like, “well, Duncan, you’re asking for a lot,” but I’m trying pretty hard to earn it, and to … pave the way? Help make the ask smaller? … with things like the old Moderating LessWrong post and the Concentration of Force post and the more recent Basics post. Like, I can’t think of much more that someone with zero authority and zero mantle can do. My problem is that abuse and strawmanning of me gets hosted on LW and upvoted on LW and people are like, well, maybe if you patiently engaged with and overturned the abuse and strawmanning in detail instead of fighting back—
I dunno. If mods would show up and be like “false” and “cut it out” I would pretty happily never get into a scrap on LW ever again.
At the risk of guessing wrong, and perhaps typical-mind-fallacying, I imagining that you’re [rightly?] feeling a lot frustration, exasperation, and even despair about moderation on LessWrong. You’ve spend dozens (more?) and tens of thousands of words trying to make LessWrong the garden you think it ought to be (and to protect yourself here against attackers) and just to try to uphold, indeed basic standards for truthseeking discourse. You’ve written that some small validation goes a long way, so this is me trying to say that I think your feelings have a helluva lot of validity.
I don’t think that you and I share exactly the same ideals for LessWrong. PerfectLessWrong!Ruby and PerfectLessWrong!Duncan would be different (or heck, even just VeryGoodLessWrongs), though I also am pretty sure that you’d be much happier with my ideal, you’d think it was pretty good if not perfect. Respectable, maybe adequate. A garden.
And I’m really sad that the current LessWrong feels really really far short of my own ideals (and Ray of his ideals, and Oli of his ideals), etc. And not just short of a super-amazing-lofty-ideal, also short of a “this place is really under control” kind of ideal. I take responsibility for it not being so, and I’m sorry. I wouldn’t blame you for saying this isn’t good enough and wanting to leave[1], there are some pretty bad flaws.
But sir, you impugn my and my site’s honor. This is not a perfect garden, it also not a jungle. And there is an awful lot of gardening going on. I take it very seriously that LessWrong is not just any place, and it takes ongoing work to keep it so. This is approx my full-time job (and that of others too), and while I don’t work 80-hour weeks, I feel like I put a tonne of my soul into this site.
Over the last year, I’ve been particularly focused on what I suspect are existential threats to LessWrong (not even the ideal, just the decently-valuable thing we have now). I think this very much counts as gardening. The major one over last year is how to both have all the AI content (and I do think AI is the most important topic right now) and not have it eat LessWrong and turn it into the AI-website rather than the truth-seeking/effectiveness/rationality website which is actually what I believe is its true spirit[2]. So far, I feel like we’re still failing at this. On many days, the Frontpage is 90+% AI posts. It’s not been a trivial problem for many problems.
The other existential problem, beyond the topic, that I’ve been anticipating for a long time and is now heating up is the deluge of new users flowing to the site because of the rising prominence of AI. Moderation is currently our top focus, but even before that, every day – the first thing we do when the team gets in the morning – is review every new post, all first time submissions from users, and the activity of users who are getting a lot of downvotes. It’s not exactly fun, but we do it basically everyday[3]. In the interests of greater transparency and accountability, we will soon build a Rejected Content section of the site where you’ll be able to view the content we didn’t go live, and I predict that will demonstrate just how much this garden is getting tended, and that counterfactually the quality would be a lot lot worse. You can see here a recent internal document that describes my sense of priorities for the team.
I think the discourse norms and bad behavior (and I’m willing to say now in advance of my more detailed thoughts that there’s a lot of badness to how Said behaves) are also serious threats to the site, and we do give those attention too. They haven’t felt like the most pressing threats (or for that matter, opportunities, recently), and I could be making a mistake there, but we do take them seriously. Our focus (which I think has a high opportunity cost) has been turned to the exchanges between you and Said this week, plausibly you’ve done us a service to draw our attention to behavior we should be deeming intolerable, and it’s easily 50-100 hours of team attention.
It is plausible the LessWrong team has made a mistake in not prioritizing this stuff more highly over the years (it has been years – though Said and Zack and others have in fact received hundreds of hours of attention), and there are definitely particular projects that I think turned out to be misguided and less valuable than marginal moderation would have been, but I’ll claim that it was definitely not an obvious mistake that we haven’t addressed the problems you’re most focused on.
It is actually on my radar and I’ve been actively wanted for a while a system that reliably gets the mod team to show up and say “cut it out” sometimes. I suspect that’s what should have happened a lot earlier on in your recent exchanges with Said. I might have liked to say “Duncan, we the mods certify that if you disengage, it is no mark against you” or something. I’m not sure. Ray mentioned the concept of “Maslow’s Hierarchy of Moderation” and I like that idea, and would like to get soon to the higher level where we’re actively intervening in this cases. I regret that I in particular on the team am not great at dropping what I’m doing to pivot when these threads come up, perhaps I should work on that.
I think a claim you could make is the LessWrong team should have hired more people so they could cover more of this. Arguing why we haven’t (or why Lightcone as a whole didn’t keep more team members on LessWrong team) is a bigger deal. I think things would be worse if LessWrong had been bigger most of the time, and barring unusually good candidate, it’d be bad to hire right now.
All this to say, this garden has a lot of shortcomings, but the team works quite hard to keep it at least as good as it is and try to make it better. Fair enough if it doesn’t meet your standards or not how you’d do it, perhaps we’re not all that competent, fair enough.
(And also you’ve had a positive influence on us, so your efforts are not completely in vain. We do refer to your moderation post/philosophy even if we haven’t adopted it wholesale, and make use of many of the concepts you’ve crystallized. For that I am grateful. Those are contributions I’d be sad to lose, but I don’t want to push you to offer to them to us if doing so is too costly for you.)
I will also claim though that a better version of Duncan would be better able to tolerate the shortcomings of LessWrong and improve it too; that even if your efforts to change LW aren’t working enough, there are efforts on yourself that would make you better, and better able to benefit from the LessWrong that is.
This is fair, and I apologize; in that line I was speaking from despair and not particularly tracking Truth.
A [less straightforwardly wrong and unfair] phrasing would have been something like “this is not a Japanese tea garden; it is a British cottage garden.”
I probably rushed this comment out the door in a “defend my honor, set the record straight” instinct that I don’t think reliably leads to good discourse and is not what I should be modeling on LessWrong.
I didn’t make it to every point, but hopefully you find this more of the substantive engagement you were hoping for.
I did, thanks.
gjm specifically noted the separation between the major issue of whether balance is required, and this other, narrower claim.
I think gjm’s comment was missing the observation that “comment that just ask for examples” are themselves an example of “unproductive modes of discussion where he is constantly demanding more and more rigour and detail from his interlocutors while not providing it himself”, and so it wasn’t cleanly about “balance: required or not?”. I think a reasonable reader could come away from that comment of gjm’s uncertain whether or not Said simply saying “examples?” would count as an example.
My interpretation of this section is basically the double crux dots arguing over the labels they should have, with Said disagreeing strenuously with calling his mode “unproductive” (and elsewhere over whether labor is good or bad, or how best to minimize it) and moving from the concrete examples to an abstract pattern (I suspect because he thinks the former is easier to defend than the latter).
I should also note here that I don’t think you have explicitly staked out that you think Said just saying “examples?” is bad (like, you didn’t here, which was the obvious place to), I am inferring that from various things you’ve written (and, tho this source is more suspect and so has less influence, ways other people have reacted to Said before).
Said to coax out his broader, potentially more valid point
Importantly, I think Said’s more valid point was narrower, not broader, and the breadth was the ‘strawmanning’ part of it. (If you mean to refer to the point dealing with the broader context, I agree with that.) The invalid “Duncan’s rule against horses” turning into the valid “Duncan’s rule against white horses”. If you don’t have other rules against horses—you’re fine with brown ones and black one and chestnut ones and so on—I think that points towards your rule against white horses pretty clearly. [My model of you thinks that language is for compiling into concepts instead of pointing at concepts and so “Duncan’s rule against horses” compiles into “Duncan thinks horses should be banned” which is both incorrect and wildly inconsistent with the evidence. I think language is for both, and when one gives you a nonsense result, you should check the other.]
I will note a way here in which it is not quite fair that I am saying “I think you didn’t do a reasonable level of interpretive labor when reading Said”, in the broader context of your complaint that Said doesn’t do much interpretive labor (deliberately!). I think it is justified by the difference in how the two of you respond to the failure of that labor.
(Am I in fact reading you correctly?)
I am trying to place the faux pas not in that you “reacted at all to that prompt” but “how you reacted to the prompt”. More in the next section.
clarifies that yes, they really meant that claim,
I think this point is our core disagreement. I see the second comment saying “yeah, Duncan’s rule against horses, the thing where he dislikes white ones”, and you proceeding as if he just said “Duncan’s rule against horses.” I think there was a illusion of transparency behind “specifically reaffirmed by Said”.
Like, I think if you had said “STRAWMAN!” and tried to get us to put a scarlet S in Said’s username, this would have been a defensible accusation, and the punishment unusual but worth considering. Instead I think you said “LIAR!” and that just doesn’t line up with my reading of the thread (tho I acknowledge disagreement about the boundary between ‘lying’ and ‘strawmanning’) or my sense of how to disagree properly. In my favorite world, you call it a mislabeling and identify why you think the label fails to match (again, noting that gjm attempted to do so, tho I think not in a way that bridged the gap).
I think there’s a thing here of standards unevenly applied; surely whatever standard would’ve had me address Said’s “real” concern would’ve also had Said behave much differently at many steps prior, possibly never strawmanning me so hard in the first place?
I mean, for sure I wish Said had done things differently! I described them in some detail, and not strawmanning you so hard in the first place was IMO the core one.
When I say “locally”, I am starting the clock at Killing Socrates, which was perhaps unclear.
if there can pretty please be some kind of system in place that obviates it.
Do you think Said would not also stop if, for every post he read on LW, he found that someone else had already made the comment he would have liked to have made?
(I do see a difference where the outcomes you seek to achieve are more easily obtained with mod powers backing them up, but I don’t think that affects the primary point.)
If mods would show up and be like “false” and “cut it out” I would pretty happily never get into a scrap on LW ever again.
So, over here Elizabeth ‘summarizes’ Said in an unflattering way, and Said objects. I don’t think I will reliably see such comments before those mentioned in them do (there were only 23 minutes before Said objected) and it is not obvious to me that LW would be improved by me also objecting now.
But perhaps our disagreement is that, on seeing Elizabeth’s comment, I didn’t have a strong impulse to ‘set the record straight’; I attribute that mostly to not seeing Elizabeth’s comment as “the record,” tho I’m open to arguments that I should.
I think a reasonable reader could come away from that comment of gjm’s uncertain whether or not Said simply saying “examples?” would count as an example.
...
I should also note here that I don’t think you have explicitly staked out that you think Said just saying “examples?” is bad (like, you didn’t here, which was the obvious place to), I am inferring that from various things you’ve written (and, tho this source is more suspect and so has less influence, ways other people have reacted to Said before).
To clarify:
If one starts out looking to collect and categorize evidence of their conversational partner not doing their fair share of the labor, then a bunch of comments that just say “Examples?” would go into the pile. But just encountering a handful of comments that just say “Examples?” would not be enough to send a reasonable person toward the hypothesis that their conversational partner reliably doesn’t do their fair share of the labor.
“Do you have examples?” is one of the core, common, prosocial moves, and correctly so. It is a bid for the other person to put in extra work, but the scales of “are we both contributing?” don’t need to be balanced every three seconds, or even every conversation. Sometimes I’m the asker/learner and you’re the teacher/expounder, and other times the roles are reversed, and other times we go back and forth.
The problem is not in asking someone to do a little labor on your behalf. It’s having 85+% of your engagement be asking other people to do labor on your behalf, and never reciprocating, and when people are like, hey, could you not, or even just a little less? being supercilious about it.
Said simply saying “examples?” is an example, then, but only because of the strong prior from his accumulated behavior; if the rule is something like “doing this <100x/wk is fine, doing it >100x/wk is less fine,” then the question of whether a given instance “is an example” is slightly tricky.
I think this point is our core disagreement. I see the second comment saying “yeah, Duncan’s rule against horses, the thing where he dislikes white ones”, and you proceeding as if he just said “Duncan’s rule against horses.” I think there was a illusion of transparency behind “specifically reaffirmed by Said”.
Yeah, you may have pinned it down (the disagreement). I definitely don’t (currently) think it’s sensible to read the second comment that way, and certainly not sensible enough to mentally dock someone for not reading it that way even if that reading is technically available (which I agree it is).
Like, I think if you had said “STRAWMAN!” and tried to get us to put a scarlet S in Said’s username, this would have been a defensible accusation
I perhaps have some learned helplessness around what I can, in fact, expect from the mod team; I claim that if I had believed that this would be received as defensible I would’ve done that instead. At the time, I felt helpless and alone*/had no expectation of mod support for reasons I think are reasonable, and so was not proceeding as if there was any kind of request I could make, and so was not brainstorming requests.
*alone vis-a-vis moderators, not alone vis-a-vis other commenters like gjm
I do think that you should put a scarlet P in Said’s username, since he’s been doing it for a couple weeks now and is still doing it (c.f. “I have yet to see any compelling reason to conclude that this [extremely unlikely on its face hypothesis] is false.”).
In my favorite world, you call it a mislabeling and identify why you think the label fails to match (again, noting that gjm attempted to do so, tho I think not in a way that bridged the gap).
I again agree that this is clearly a better set of moves in some sense, but I’m thinking in a fabricated options frame and being, like, is that really actually a possible world, in that the whole problem is Said’s utterly exhausting and unrewarding mode of engagement. Like, I wonder if I might convince you that your favorite world is incoherent and impossible, because it’s one in which people are engaging in the colloquial definition of insanity and never updating their heuristics based on feedback. Or maybe you’re saying “do it for the audience and for site norms, then,” which feels less like throwing good money after bad.
But like. I think I’m getting dinged for impatience when I did not, previously, get headpats for patience? The wanted behavior feels unincentivized relative to the unwanted behavior.
When I say “locally”, I am starting the clock at Killing Socrates, which was perhaps unclear.
No, that was pretty clear, and that’s what generated the :((((((((. The choice to start the clock there feels unfair-to-Neville, like if I were a teacher I would glance at that and say “okay, obviously this is not the local beginning” and look further.
Do you think Said would not also stop if, for every post he read on LW, he found that someone else had already made the comment he would have liked to have made?
I am wary of irresponsibly theorizing about the contents of someone else’s mind. I do think that, if one looks over the explosive proliferation of his threads once he starts a back-and-forth, it’s unlikely that there’s some state in which Said is like “ah, people are already saying all the things!” I suspect that Said (like others, to be clear; this is not precisely a criticism) has an infinite priority list, and if all the things of top priority are handled by other commenters, he’ll move down to lower ones.
I do think that if you took all of Said’s comments, and distributed 8% of them each into the corpus of comments of Julia Galef, Anna Salamon, Rob Bensinger, Scott Garrabrant, you, Eliezer Yudkowsky, Logan Brienne Strohl, Oliver Habryka, Kelsey Piper, Nate Soares, Eric Rogstad, Spencer Greenberg, and Dan Keys this would be much better. Part of the problem is the sheer concentration of princely entitlement and speaking-as-if-it-is-the-author’s-job-to-convince-Said-particularly-regardless-of-whether-Said’s-skepticism-is-a-signal-of-any-real-problem-with-the-claims.
If Kelsey Piper locally is like, buddy, you need to give me more examples, or if Spencer Greenberg locally is like, but what the heck do you even mean by “annoying,” there’s zero sense (on my part, at least) that here we go again, more taking-without-contributing. Instead, with Kelsey and Spencer it feels like a series of escalating favors and a tightening of the web of mutual obligation in which everybody is grateful to everybody else for having put in so many little bits of work here and there, of course I want to spill some words to help connect the dots for Kelsey and Spencer, they’ve spilled so many words helping me.
The pattern of “give, then take, then give, then take, then take, then take, then give, then give” is a healthy one to model, and is patriotically Athenian in the frame of my recent essay, and is not one which, if a thousand newbies were to start emulating, would cause a problem.
But perhaps our disagreement is that, on seeing Elizabeth’s comment, I didn’t have a strong impulse to ‘set the record straight’
I don’t think that mods should be chiming in and setting the record straight on every little thing. But when, like, Said spends multiple thousands of words in a literally irrational (in the sense of not having cruxes and not being open to update and being directly contradicted by evidence) screed strawmanning me and claiming that I block people for disagreeing with my claims or criticizing my arguments—
—and furthermore when I ask for mod help—
—then I do think that a LessWrong where a mod shows up to say “false” and “actually cut it out for real” is meaningfully different and meaningfully better than the current Wild West feel where Said doesn’t get in trouble but I do.
The problem is not in asking someone to do a little labor on your behalf. It’s having 85+% of your engagement be asking other people to do labor on your behalf, and never reciprocating, and when people are like, hey, could you not, or even just a little less? being supercilious about it.
But why should this be a problem?
Why should people say “hey, could you not, or even just a little less”? If you do something that isn’t bad, that isn’t not a problem, why should people ask you to stop? If it’s a good thing to do, why wouldn’t they instead ask you to do it more?
And why, indeed, are you still speaking in this transactional way?
If you write a post about some abstract concept, without any examples of it, and I write a post that says “What are some examples?”, I am not asking you to do labor on my behalf, I am not asking for a favor (which must be justified by some “favor credit”, some positive account of favors in the bank of Duncan). Quite frankly, I find that claim ridiculous to the point of offensiveness. What I am doing, in that scenario, is making a positive contribution to the discussion, both for your benefit and (even more importantly) for the benefit of other readers and commenters.
There is no good reason why you should resent responding to a request like “what are some examples”. There is no good reason why you should view it as an unjustified and entitled demand for a favor. There is definitely no good reason why you should view acceding to that request as being “for my benefit” (instead of, say, for your benefit, and for the benefit of readers).
(And the gall of saying “never reciprocating”, to me! When I write a post, I include examples pre-emptively, because I know that I should be asked to do so otherwise. Not “will be asked”, of course—but “should”. And when I write a post without enough examples, and someone asks for examples, I respond in great detail. Note that my responses in that thread are much, much longer than the comment which asked for examples. Of course they are! Because the question doesn’t need to be longer—but the answers do!)
(And you might say: “but Said, you barely write any posts—like one a year, at best!”. Indeed. Indeed.)
There is no good reason why you should resent responding to a request like “what are some examples”.
Maybe “resent” is doing most work here, but an excellent reason to not respond is that it takes work. To the extent that there are norms in place that urge response, they create motivation to suppress criticism that would urge response. An expectation that it’s normal for criticism to be a request for response that should normally be granted is pressure to do the work of responding, which is costly, which motivates defensive action in the form of suppressing criticism.
A culture could make it costless (all else equal) to ignore the event of a criticism having been made. This is an inessential reason for suppressing criticism that can be removed, and therefore should, to make criticism cheaper and more abundant.
The content of criticism may of course motivate the author of a criticized text to make further statements, but the fact of criticism’s posting by itself should not. The fact of not responding to criticism is some sort of noisy evidence of not having a good response that is feasible or hedonic to make, but that’s Law, not something that can change for the sake of mechanism design.
It’s certainly doing a decent amount of work, I agree.
Anyhow, your overall point is taken—although I have to point out that that your last sentence seems like a rebuttal of your next-to-last sentence.
That having been said, of course the content of criticism matters. A piece of criticism could simply be bad, and clearly wrong; and then it’s good and proper to just ignore it (perhaps after having made sure that an interested party could, if they so wished, easily see or learn why that criticism is bad). I do not, and would not, advocate for a norm that all comments, all critical questions, etc., regardless of their content, must always be responded to. That is unreasonable.
I also want to note—as I’ve said several times in this discussion, but it bears repeating—there is nothing problematic or blameworthy about someone other than the author of a post responding to questions, criticism, requests for examples, etc. That is fine. Collaborative development of ideas is a perfectly normal and good thing.
What that adds up to, I think, is a set of requirements for a set of social norms which is quite compatible with your suggestion of making it “costless (all else equal) to ignore the event of a criticism having been made”.
The content of criticism may of course motivate the author of a criticized text to make further statements, but the fact of criticism’s posting by itself should not. The fact of not responding to criticism is some sort of noisy evidence of not having a good response that is feasible or hedonic to make, but that’s Law, not something that can change for the sake of mechanism design.
I have to point out that that your last sentence seems like a rebuttal of your next-to-last sentence
They are in opposition, but the point is that they are about different kinds of things, and one of them can’t respond to policy decisions. It’s useful to have a norm that lessens the burden of addressing criticism. It’s Law of reasoning that this burden can nonetheless materialize. The Law is implacable but importantly asymmetric, it only holds when it does, not when the court of public opinion says it should. While the norms are the other way around, and their pressure is somewhat insensitive to facts of a particular situation, so it’s worth pointing them in a generally useful direction, with no hope for their nuanced or at all sane response to details.
Perhaps the presence of Law justifies norms that are over-the-top forgiving to ignoring criticism, or find ignoring criticism a bit praiseworthy when it would be at all unpleasant not to ignore it, to oppose the average valence of Law, while of course attempting to preserve its asymmetry. So I’d say my last sentence in that comment argues that the next-to-last sentence should be stronger. Which I’m not sure I agree with, but here’s the argument.
Said, above, is saying a bunch of things, many of which I agree with, as if they are contra my position or my previous claims.
He can’t pass my ITT (not that I’ve asked him to), which means that he doesn’t understand the thing he’s trying to disagree with, which means that his disagreement is not actually pointing at my position; the things he finds ridiculous and offensive are cardboard cutouts of his own construction. More detail on that over here.
BTW I was surprised earlier to see you agree with the ‘relational’ piece of this comment because Duncan’s grandparent comment seems like it’s a pretty central example of that. (I view you as having more of a “visitor-commons” orientation towards LW, and Duncan has more of an orientation where this is a place where people inhabit their pairwise relationships, as well as more one-to-many relationships.)
Sorry, I’m not quite sure I follow the references here. You’re saying that… this comment… is a central example of… what, exactly?
(I view you as having more of a “visitor-commons” orientation towards LW, and Duncan has more of an orientation where this is a place where people inhabit their pairwise relationships, as well as more one-to-many relationships.)
That… seems like it’s probably accurate… I think? I think I’d have to more clearly understand what you’re getting at in your comment, in order to judge whether this part makes sense to me.
Sorry, my previous comment wasn’t very clear. Earlier I said:
Duncan is trying to suggest what is permissible or impermissible is more relational and deals with people’s attitudes towards each other (as suggested by gjm here).
and you responded with:
I also—and, perhaps, more importantly—think that the interactions in question are not only fine, but good, in a “relational” sense.
(and a few related comments) which made me think “hmm, I don’t think we mean the same thing by ‘relational’. Then Duncan’s comment had a frame that I would have described as ‘relational’—as in focusing on the relationships between the people saying and hearing the words—which you then described as transactional.
I think that the sense in which I would characterize Duncan’s description as “transactional” is… mostly orthogonal to the question of “is this a relational frame”. I don’t think that this has much to do with the “‘visitor commons’ vs. ‘pairwise relationships’” distinction, either (although that distinction is an interesting and possibly important one in its own right, and you’re certainly more right than wrong about where my preferences lie in that regard).
(There’s more that I could say about this, but I don’t know whether anything of importance hinges on this point. It seems like it mostly shouldn’t, but perhaps you are a better judge of that…)
“—then I do think that a LessWrong where a mod shows up to say “false” and “actually cut it out for real” is meaningfully different and meaningfully better than the current Wild West feel where Said doesn’t get in trouble but I do.”
I think Vaniver right now is focusing on resolving the point “is Said a liar?”, but not resolving the “who did most wrong?” question. (I’m not actually 100% sure on Vaniver’s goals/takes at the moment). I agree this is an important subquestion but it’s not the primary question I’m interested in.
I’m somewhat worried about this thread taking in more energy that it quite warrants, and making Duncan feel more persecuted than really makes sense here.
I roughly agree with Vaniver than “Liar!” isn’t the right accusation to have levied, but also don’t judge you harshly for having made it.
I think this comment of mine summarizes my relevant opinions here.
(tagging @Vaniver to make sure he’s at least tracking this comment)
I note (while acknowledging that this is a small and subtle distinction, but claiming that it is an important one nonetheless) that I said that I now categorize Said as a liar, which is an importantly and intentionally weaker claim than Said is a liar, i.e. “everyone should be able to see that he’s a liar” or “if you don’t think he’s a liar you are definitely wrong.”
(This is me in the past behaving in line with the points I just made under Said’s comment, about not confusing [how things seem to me] with [how they are] or [how they do or should seem to others].)
This is much much closer to saying “Liar!” than it is to not saying “Liar!” … if one is to round me off, that’s the correct place to round me off to. But it is still a rounding.
I agree that the hypothetical comment you describe as better is in fact better. I think something like … twenty-or-so exchanges with Said ago, I would have written that comment? I don’t quite know how to weigh up [the comment I actually wrote is worse on these axes of prosocial cooperation and revealing cruxes and productively clarifying disagreement and so forth] with [having a justified true belief that putting forth that effort with Said in particular is just rewarded with more branches being created].
(e.g. there was that one time recently where Said said I’d blocked people due to disagreeing with me/criticizing me, and I said no, I haven’t blocked anybody for disagreeing/criticizing, and he responded “I didn’t say anything about ‘blocked for disagreeing [or criticizing]’. (Go ahead, check!)” and the actual thing he’d said was that they’d been blocked due to disagreeing/criticizing; that’s the level of … gumming up the works? gish-gallop? … that I’ve viscerally come to expect.)
Like, I think there’s plausibly a CEV-ish code of conduct in which I “should”, at that point, still have put forth the effort, but I think it’s also plausible that the correct code of conduct is one in which doing so is a genuine mistake and … noticing that there’s a hypothetical “better” comment is not the same as there being an implication that I should’ve written it?
Something something, how many turns of the cheek are actually correct, especially given that, the week prior, multiple commenters had been unable, with evidence+argument+personal testimony, to shift Said away from a strikingly uncharitable prior.
Mine either, to be clear; I felt by that point that Said had willingly put himself outside of the set of [signatories to the peace treaty], turning down many successive opportunities to remain in compliance with it. I was treating his statements closer to the way I think it is correct to treat the statements of the literal Donald Trump than the way I think it is correct to treat the statements of an undistinguished random Republican.
(I can go into the reasoning for that in more detail, but it seems sort of conflicty to do so unprompted.)
I’m a little lost in this analogy; this is sort of where the privileging-the-hypothesis complaint comes in.
The conversation had, in other places, centered on the question of whether Said can eat cereal by itself; Logan for instance highlighted Said’s claim in a reply on FB:
There, the larger question of “can you eat only cereal, or must you eat other things in balance?” is front-and-center.
But at that point in the subthread, it was not front-and-center; yes, it was relevant context, but the specific claim being made by Said was clear, and discrete, and not at all dependent-on or changed-by that context.
The history of that chain:
Said includes, in a long comment “In summary, I think that what’s been described as ‘aiming for convergence on truth’ is some mixture of” … “contentless” … “good but basically unrelated to the rest of it” … “bad (various proposed norms of interaction such as ‘don’t ask people for examples of their claims’ and so on)”
gjm, in another long comment, includes “I don’t know where you get ‘don’t ask people for examples of their claims’ from and it sounds like a straw man” and goes on to elaborate “I think the things Duncan has actually said are more like ‘Said engages in unproductive modes of discussion where he is constantly demanding more and more rigour and detail from his interlocutors while not providing it himself’, and wherever that lands on a scale from ’100% truth’ to ‘100% bullshit’ it is not helpful to pretend that he said ‘it is bad to ask people for examples of their claims’.
There’s a bunch of other stuff going on in their back and forth, but that particular thread has been isolated and directly addressed, in other words. gjm specifically noted the separation between the major issue of whether balance is required, and this other, narrower claim.
Said replied:
Which, yes, I straightforwardly agree with the if-then statement; if “asking people for examples of their claims” didn’t fit my stated criteria for what constitutes acceptable engagement or criticism, then it would be correct to describe me as advocating for a norm of “don’t ask people for examples of their claims.”
But like. The if does not hold. It really clearly doesn’t hold. It was enough of an out-of-nowhere strawman/non-sequitur that gjm specifically called it out as ”???” at which point Said doubled down, saying the above and also
It seems like, in your interpretation, I “should” (in some sense) be extending a hand of charity and understanding and, I dunno, helping Said to coax out his broader, potentially more valid point—helping him to get past his own strawman and on to something more steel, or at least flesh. Like, if I am reading you correctly above, you’re saying that, by focusing in on the narrow point that had been challenged by gjm and specifically reaffirmed by Said, I myself was making some sort of faux pas.
(Am I in fact reading you correctly?)
I do not think so. I think that, twenty exchanges prior, I perhaps owed Said something like that degree of care and charity and helping him avoid tying his own shoelaces together. I certainly feel I would owe it to, I dunno, Eric Rogstad or Julia Galef, and would not be the slightest bit loath to provide it.
But here, Said had just spent several thousand words the week prior, refusing to be budged from a weirdly uncharitable belief about the internals of my mind, despite that belief being incoherent with observable evidence and challenged by multiple non-me people. I don’t think it’s wise-in-the-sense-of-wisdom to a) engage with substantial charity in that situation, or b) expect someone else to engage with substantial charity in that situation.
(You can tell that my stated criteria do not rule out asking people for examples of their claims in part because I’ve written really quite a lot about what I think constitutes acceptable engagement or criticism, and I’ve just never come anywhere close to a criterion like that, nor have I ever complained about someone asking for examples unless it was after a long, long string of what felt like them repeatedly not sharing in the labor of truthseeking. Like, the closest I can think of is this thread with tailcalled, in which (I think/I hope) it’s pretty clear that what’s going on is that I was trying to cap the total attention paid to the essay and its discussion, and thus was loath to enter into something like an exchange of examples—not that it was bad in any fundamental sense for someone to want some. I did in fact provide some, a few comments deeper in the thread, though I headlined that I hadn’t spent much time on them.)
So in other words: I don’t think it was wrong to focus on the literal, actual claim that Said had made (since he made it, basically, twice in a row, affirming “no, I really mean this” after gjm’s objection and even saying that he thinks it is so obvious as to not be controversial. I don’t think I “ought” to have had a broader focus, under the circumstances—Said was making a specific, concrete, and false claim, and his examples utterly fail to back up that specific, concrete, and false claim (though I do agree with you that they back up something like his conception of our broader disagreement).
I dunno, I’m feeling kind of autistic, here, but I feel like if, on Less Wrong dot com, somebody makes a specific, concrete claim about my beliefs or policies, clarifies that yes, they really meant that claim, and furthermore says that such-and-such links are “citations for [me] expressing the sentiment [they’ve] ascribed to [me]” when they simply are not—
It feels like emphatic and unapologetic rejection should be 100% okay, and not looked at askance. The fact that they are citations supporting a different claim is (or at least, I claim, should be) immaterial; it’s not my job to steelman somebody who spent hours and hours negatively psychologizing me in public (while claiming to have no particular animus, which, boy, a carbon copy of Said sure would have had Words about).
I think there’s a thing here of standards unevenly applied; surely whatever standard would’ve had me address Said’s “real” concern would’ve also had Said behave much differently at many steps prior, possibly never strawmanning me so hard in the first place?
I think the asymmetry breaks in that, like, a bunch of people have asked Said to stop and he won’t; I’m quite eager to stop doing the conflict resolution that people don’t like, if there can pretty please be some kind of system in place that obviates it. I much prefer the world where there are competent police to the world where I have to fight off muggers in the alley—that’s why I’m trying so hard to get there to be some kind of actually legible standards rather than there always being some plausible reason why maybe we shouldn’t just say “no” to the bullshit that Zack or Said or anonymouswhoever is pulling.
Right now, though, it feels like we’ve gone from “Ben Hoffman will claim Duncan wants to ghettoize people and it’ll be left upvoted for nine days with no mod action” to “Ray will expound on why he thinks it’s kinda off for Said to be doing what he’s doing but there won’t be anything to stop Said from doing it” and I take Oli’s point about this stuff being hard and there being other priorities but like, it’s been years. And I get a stance of, like, “well, Duncan, you’re asking for a lot,” but I’m trying pretty hard to earn it, and to … pave the way? Help make the ask smaller? … with things like the old Moderating LessWrong post and the Concentration of Force post and the more recent Basics post. Like, I can’t think of much more that someone with zero authority and zero mantle can do. My problem is that abuse and strawmanning of me gets hosted on LW and upvoted on LW and people are like, well, maybe if you patiently engaged with and overturned the abuse and strawmanning in detail instead of fighting back—
I dunno. If mods would show up and be like “false” and “cut it out” I would pretty happily never get into a scrap on LW ever again.
:(((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((
This, more than anything else, is like “just give up and leave, this is definitely not a garden.”
I didn’t make it to every point, but hopefully you find this more of the substantive engagement you were hoping for.
At the risk of guessing wrong, and perhaps typical-mind-fallacying, I imagining that you’re [rightly?] feeling a lot frustration, exasperation, and even despair about moderation on LessWrong. You’ve spend dozens (more?) and tens of thousands of words trying to make LessWrong the garden you think it ought to be (and to protect yourself here against attackers) and just to try to uphold, indeed basic standards for truthseeking discourse. You’ve written that some small validation goes a long way, so this is me trying to say that I think your feelings have a helluva lot of validity.
I don’t think that you and I share exactly the same ideals for LessWrong. PerfectLessWrong!Ruby and PerfectLessWrong!Duncan would be different (or heck, even just VeryGoodLessWrongs), though I also am pretty sure that you’d be much happier with my ideal, you’d think it was pretty good if not perfect. Respectable, maybe adequate. A garden.
And I’m really sad that the current LessWrong feels really really far short of my own ideals (and Ray of his ideals, and Oli of his ideals), etc. And not just short of a super-amazing-lofty-ideal, also short of a “this place is really under control” kind of ideal. I take responsibility for it not being so, and I’m sorry. I wouldn’t blame you for saying this isn’t good enough and wanting to leave[1], there are some pretty bad flaws.
But sir, you impugn my and my site’s honor. This is not a perfect garden, it also not a jungle. And there is an awful lot of gardening going on. I take it very seriously that LessWrong is not just any place, and it takes ongoing work to keep it so. This is approx my full-time job (and that of others too), and while I don’t work 80-hour weeks, I feel like I put a tonne of my soul into this site.
Over the last year, I’ve been particularly focused on what I suspect are existential threats to LessWrong (not even the ideal, just the decently-valuable thing we have now). I think this very much counts as gardening. The major one over last year is how to both have all the AI content (and I do think AI is the most important topic right now) and not have it eat LessWrong and turn it into the AI-website rather than the truth-seeking/effectiveness/rationality website which is actually what I believe is its true spirit[2]. So far, I feel like we’re still failing at this. On many days, the Frontpage is 90+% AI posts. It’s not been a trivial problem for many problems.
The other existential problem, beyond the topic, that I’ve been anticipating for a long time and is now heating up is the deluge of new users flowing to the site because of the rising prominence of AI. Moderation is currently our top focus, but even before that, every day – the first thing we do when the team gets in the morning – is review every new post, all first time submissions from users, and the activity of users who are getting a lot of downvotes. It’s not exactly fun, but we do it basically everyday[3]. In the interests of greater transparency and accountability, we will soon build a Rejected Content section of the site where you’ll be able to view the content we didn’t go live, and I predict that will demonstrate just how much this garden is getting tended, and that counterfactually the quality would be a lot lot worse. You can see here a recent internal document that describes my sense of priorities for the team.
I think the discourse norms and bad behavior (and I’m willing to say now in advance of my more detailed thoughts that there’s a lot of badness to how Said behaves) are also serious threats to the site, and we do give those attention too. They haven’t felt like the most pressing threats (or for that matter, opportunities, recently), and I could be making a mistake there, but we do take them seriously. Our focus (which I think has a high opportunity cost) has been turned to the exchanges between you and Said this week, plausibly you’ve done us a service to draw our attention to behavior we should be deeming intolerable, and it’s easily 50-100 hours of team attention.
It is plausible the LessWrong team has made a mistake in not prioritizing this stuff more highly over the years (it has been years – though Said and Zack and others have in fact received hundreds of hours of attention), and there are definitely particular projects that I think turned out to be misguided and less valuable than marginal moderation would have been, but I’ll claim that it was definitely not an obvious mistake that we haven’t addressed the problems you’re most focused on.
It is actually on my radar and I’ve been actively wanted for a while a system that reliably gets the mod team to show up and say “cut it out” sometimes. I suspect that’s what should have happened a lot earlier on in your recent exchanges with Said. I might have liked to say “Duncan, we the mods certify that if you disengage, it is no mark against you” or something. I’m not sure. Ray mentioned the concept of “Maslow’s Hierarchy of Moderation” and I like that idea, and would like to get soon to the higher level where we’re actively intervening in this cases. I regret that I in particular on the team am not great at dropping what I’m doing to pivot when these threads come up, perhaps I should work on that.
I think a claim you could make is the LessWrong team should have hired more people so they could cover more of this. Arguing why we haven’t (or why Lightcone as a whole didn’t keep more team members on LessWrong team) is a bigger deal. I think things would be worse if LessWrong had been bigger most of the time, and barring unusually good candidate, it’d be bad to hire right now.
All this to say, this garden has a lot of shortcomings, but the team works quite hard to keep it at least as good as it is and try to make it better. Fair enough if it doesn’t meet your standards or not how you’d do it, perhaps we’re not all that competent, fair enough.
(And also you’ve had a positive influence on us, so your efforts are not completely in vain. We do refer to your moderation post/philosophy even if we haven’t adopted it wholesale, and make use of many of the concepts you’ve crystallized. For that I am grateful. Those are contributions I’d be sad to lose, but I don’t want to push you to offer to them to us if doing so is too costly for you.)
I will also claim though that a better version of Duncan would be better able to tolerate the shortcomings of LessWrong and improve it too; that even if your efforts to change LW aren’t working enough, there are efforts on yourself that would make you better, and better able to benefit from the LessWrong that is.
Something like the core identity of LessWrong is rationality. In alternate worlds, that is the same, but the major topic could be something else.
Over the weekend, some parts of the reviewing get deferred till the work week.
This is fair, and I apologize; in that line I was speaking from despair and not particularly tracking Truth.
A [less straightforwardly wrong and unfair] phrasing would have been something like “this is not a Japanese tea garden; it is a British cottage garden.”
I have been to the Japanese tea garden in Portland, and found it exquisite, so I think get your referent there.
Aye, indeed it is not that.
I probably rushed this comment out the door in a “defend my honor, set the record straight” instinct that I don’t think reliably leads to good discourse and is not what I should be modeling on LessWrong.
I did, thanks.
I think gjm’s comment was missing the observation that “comment that just ask for examples” are themselves an example of “unproductive modes of discussion where he is constantly demanding more and more rigour and detail from his interlocutors while not providing it himself”, and so it wasn’t cleanly about “balance: required or not?”. I think a reasonable reader could come away from that comment of gjm’s uncertain whether or not Said simply saying “examples?” would count as an example.
My interpretation of this section is basically the double crux dots arguing over the labels they should have, with Said disagreeing strenuously with calling his mode “unproductive” (and elsewhere over whether labor is good or bad, or how best to minimize it) and moving from the concrete examples to an abstract pattern (I suspect because he thinks the former is easier to defend than the latter).
I should also note here that I don’t think you have explicitly staked out that you think Said just saying “examples?” is bad (like, you didn’t here, which was the obvious place to), I am inferring that from various things you’ve written (and, tho this source is more suspect and so has less influence, ways other people have reacted to Said before).
Importantly, I think Said’s more valid point was narrower, not broader, and the breadth was the ‘strawmanning’ part of it. (If you mean to refer to the point dealing with the broader context, I agree with that.) The invalid “Duncan’s rule against horses” turning into the valid “Duncan’s rule against white horses”. If you don’t have other rules against horses—you’re fine with brown ones and black one and chestnut ones and so on—I think that points towards your rule against white horses pretty clearly. [My model of you thinks that language is for compiling into concepts instead of pointing at concepts and so “Duncan’s rule against horses” compiles into “Duncan thinks horses should be banned” which is both incorrect and wildly inconsistent with the evidence. I think language is for both, and when one gives you a nonsense result, you should check the other.]
I will note a way here in which it is not quite fair that I am saying “I think you didn’t do a reasonable level of interpretive labor when reading Said”, in the broader context of your complaint that Said doesn’t do much interpretive labor (deliberately!). I think it is justified by the difference in how the two of you respond to the failure of that labor.
I am trying to place the faux pas not in that you “reacted at all to that prompt” but “how you reacted to the prompt”. More in the next section.
I think this point is our core disagreement. I see the second comment saying “yeah, Duncan’s rule against horses, the thing where he dislikes white ones”, and you proceeding as if he just said “Duncan’s rule against horses.” I think there was a illusion of transparency behind “specifically reaffirmed by Said”.
Like, I think if you had said “STRAWMAN!” and tried to get us to put a scarlet S in Said’s username, this would have been a defensible accusation, and the punishment unusual but worth considering. Instead I think you said “LIAR!” and that just doesn’t line up with my reading of the thread (tho I acknowledge disagreement about the boundary between ‘lying’ and ‘strawmanning’) or my sense of how to disagree properly. In my favorite world, you call it a mislabeling and identify why you think the label fails to match (again, noting that gjm attempted to do so, tho I think not in a way that bridged the gap).
I mean, for sure I wish Said had done things differently! I described them in some detail, and not strawmanning you so hard in the first place was IMO the core one.
When I say “locally”, I am starting the clock at Killing Socrates, which was perhaps unclear.
Do you think Said would not also stop if, for every post he read on LW, he found that someone else had already made the comment he would have liked to have made?
(I do see a difference where the outcomes you seek to achieve are more easily obtained with mod powers backing them up, but I don’t think that affects the primary point.)
So, over here Elizabeth ‘summarizes’ Said in an unflattering way, and Said objects. I don’t think I will reliably see such comments before those mentioned in them do (there were only 23 minutes before Said objected) and it is not obvious to me that LW would be improved by me also objecting now.
But perhaps our disagreement is that, on seeing Elizabeth’s comment, I didn’t have a strong impulse to ‘set the record straight’; I attribute that mostly to not seeing Elizabeth’s comment as “the record,” tho I’m open to arguments that I should.
To clarify:
If one starts out looking to collect and categorize evidence of their conversational partner not doing their fair share of the labor, then a bunch of comments that just say “Examples?” would go into the pile. But just encountering a handful of comments that just say “Examples?” would not be enough to send a reasonable person toward the hypothesis that their conversational partner reliably doesn’t do their fair share of the labor.
“Do you have examples?” is one of the core, common, prosocial moves, and correctly so. It is a bid for the other person to put in extra work, but the scales of “are we both contributing?” don’t need to be balanced every three seconds, or even every conversation. Sometimes I’m the asker/learner and you’re the teacher/expounder, and other times the roles are reversed, and other times we go back and forth.
The problem is not in asking someone to do a little labor on your behalf. It’s having 85+% of your engagement be asking other people to do labor on your behalf, and never reciprocating, and when people are like, hey, could you not, or even just a little less? being supercilious about it.
Said simply saying “examples?” is an example, then, but only because of the strong prior from his accumulated behavior; if the rule is something like “doing this <100x/wk is fine, doing it >100x/wk is less fine,” then the question of whether a given instance “is an example” is slightly tricky.
Yeah, you may have pinned it down (the disagreement). I definitely don’t (currently) think it’s sensible to read the second comment that way, and certainly not sensible enough to mentally dock someone for not reading it that way even if that reading is technically available (which I agree it is).
I perhaps have some learned helplessness around what I can, in fact, expect from the mod team; I claim that if I had believed that this would be received as defensible I would’ve done that instead. At the time, I felt helpless and alone*/had no expectation of mod support for reasons I think are reasonable, and so was not proceeding as if there was any kind of request I could make, and so was not brainstorming requests.
*alone vis-a-vis moderators, not alone vis-a-vis other commenters like gjm
I do think that you should put a scarlet P in Said’s username, since he’s been doing it for a couple weeks now and is still doing it (c.f. “I have yet to see any compelling reason to conclude that this [extremely unlikely on its face hypothesis] is false.”).
I again agree that this is clearly a better set of moves in some sense, but I’m thinking in a fabricated options frame and being, like, is that really actually a possible world, in that the whole problem is Said’s utterly exhausting and unrewarding mode of engagement. Like, I wonder if I might convince you that your favorite world is incoherent and impossible, because it’s one in which people are engaging in the colloquial definition of insanity and never updating their heuristics based on feedback. Or maybe you’re saying “do it for the audience and for site norms, then,” which feels less like throwing good money after bad.
But like. I think I’m getting dinged for impatience when I did not, previously, get headpats for patience? The wanted behavior feels unincentivized relative to the unwanted behavior.
No, that was pretty clear, and that’s what generated the :((((((((. The choice to start the clock there feels unfair-to-Neville, like if I were a teacher I would glance at that and say “okay, obviously this is not the local beginning” and look further.
I am wary of irresponsibly theorizing about the contents of someone else’s mind. I do think that, if one looks over the explosive proliferation of his threads once he starts a back-and-forth, it’s unlikely that there’s some state in which Said is like “ah, people are already saying all the things!” I suspect that Said (like others, to be clear; this is not precisely a criticism) has an infinite priority list, and if all the things of top priority are handled by other commenters, he’ll move down to lower ones.
I do think that if you took all of Said’s comments, and distributed 8% of them each into the corpus of comments of Julia Galef, Anna Salamon, Rob Bensinger, Scott Garrabrant, you, Eliezer Yudkowsky, Logan Brienne Strohl, Oliver Habryka, Kelsey Piper, Nate Soares, Eric Rogstad, Spencer Greenberg, and Dan Keys this would be much better. Part of the problem is the sheer concentration of princely entitlement and speaking-as-if-it-is-the-author’s-job-to-convince-Said-particularly-regardless-of-whether-Said’s-skepticism-is-a-signal-of-any-real-problem-with-the-claims.
If Kelsey Piper locally is like, buddy, you need to give me more examples, or if Spencer Greenberg locally is like, but what the heck do you even mean by “annoying,” there’s zero sense (on my part, at least) that here we go again, more taking-without-contributing. Instead, with Kelsey and Spencer it feels like a series of escalating favors and a tightening of the web of mutual obligation in which everybody is grateful to everybody else for having put in so many little bits of work here and there, of course I want to spill some words to help connect the dots for Kelsey and Spencer, they’ve spilled so many words helping me.
The pattern of “give, then take, then give, then take, then take, then take, then give, then give” is a healthy one to model, and is patriotically Athenian in the frame of my recent essay, and is not one which, if a thousand newbies were to start emulating, would cause a problem.
I don’t think that mods should be chiming in and setting the record straight on every little thing. But when, like, Said spends multiple thousands of words in a literally irrational (in the sense of not having cruxes and not being open to update and being directly contradicted by evidence) screed strawmanning me and claiming that I block people for disagreeing with my claims or criticizing my arguments—
—and furthermore when I ask for mod help—
—then I do think that a LessWrong where a mod shows up to say “false” and “actually cut it out for real” is meaningfully different and meaningfully better than the current Wild West feel where Said doesn’t get in trouble but I do.
But why should this be a problem?
Why should people say “hey, could you not, or even just a little less”? If you do something that isn’t bad, that isn’t not a problem, why should people ask you to stop? If it’s a good thing to do, why wouldn’t they instead ask you to do it more?
And why, indeed, are you still speaking in this transactional way?
If you write a post about some abstract concept, without any examples of it, and I write a post that says “What are some examples?”, I am not asking you to do labor on my behalf, I am not asking for a favor (which must be justified by some “favor credit”, some positive account of favors in the bank of Duncan). Quite frankly, I find that claim ridiculous to the point of offensiveness. What I am doing, in that scenario, is making a positive contribution to the discussion, both for your benefit and (even more importantly) for the benefit of other readers and commenters.
There is no good reason why you should resent responding to a request like “what are some examples”. There is no good reason why you should view it as an unjustified and entitled demand for a favor. There is definitely no good reason why you should view acceding to that request as being “for my benefit” (instead of, say, for your benefit, and for the benefit of readers).
(And the gall of saying “never reciprocating”, to me! When I write a post, I include examples pre-emptively, because I know that I should be asked to do so otherwise. Not “will be asked”, of course—but “should”. And when I write a post without enough examples, and someone asks for examples, I respond in great detail. Note that my responses in that thread are much, much longer than the comment which asked for examples. Of course they are! Because the question doesn’t need to be longer—but the answers do!)
(And you might say: “but Said, you barely write any posts—like one a year, at best!”. Indeed. Indeed.)
Maybe “resent” is doing most work here, but an excellent reason to not respond is that it takes work. To the extent that there are norms in place that urge response, they create motivation to suppress criticism that would urge response. An expectation that it’s normal for criticism to be a request for response that should normally be granted is pressure to do the work of responding, which is costly, which motivates defensive action in the form of suppressing criticism.
A culture could make it costless (all else equal) to ignore the event of a criticism having been made. This is an inessential reason for suppressing criticism that can be removed, and therefore should, to make criticism cheaper and more abundant.
The content of criticism may of course motivate the author of a criticized text to make further statements, but the fact of criticism’s posting by itself should not. The fact of not responding to criticism is some sort of noisy evidence of not having a good response that is feasible or hedonic to make, but that’s Law, not something that can change for the sake of mechanism design.
It’s certainly doing a decent amount of work, I agree.
Anyhow, your overall point is taken—although I have to point out that that your last sentence seems like a rebuttal of your next-to-last sentence.
That having been said, of course the content of criticism matters. A piece of criticism could simply be bad, and clearly wrong; and then it’s good and proper to just ignore it (perhaps after having made sure that an interested party could, if they so wished, easily see or learn why that criticism is bad). I do not, and would not, advocate for a norm that all comments, all critical questions, etc., regardless of their content, must always be responded to. That is unreasonable.
I also want to note—as I’ve said several times in this discussion, but it bears repeating—there is nothing problematic or blameworthy about someone other than the author of a post responding to questions, criticism, requests for examples, etc. That is fine. Collaborative development of ideas is a perfectly normal and good thing.
What that adds up to, I think, is a set of requirements for a set of social norms which is quite compatible with your suggestion of making it “costless (all else equal) to ignore the event of a criticism having been made”.
They are in opposition, but the point is that they are about different kinds of things, and one of them can’t respond to policy decisions. It’s useful to have a norm that lessens the burden of addressing criticism. It’s Law of reasoning that this burden can nonetheless materialize. The Law is implacable but importantly asymmetric, it only holds when it does, not when the court of public opinion says it should. While the norms are the other way around, and their pressure is somewhat insensitive to facts of a particular situation, so it’s worth pointing them in a generally useful direction, with no hope for their nuanced or at all sane response to details.
Perhaps the presence of Law justifies norms that are over-the-top forgiving to ignoring criticism, or find ignoring criticism a bit praiseworthy when it would be at all unpleasant not to ignore it, to oppose the average valence of Law, while of course attempting to preserve its asymmetry. So I’d say my last sentence in that comment argues that the next-to-last sentence should be stronger. Which I’m not sure I agree with, but here’s the argument.
Said, above, is saying a bunch of things, many of which I agree with, as if they are contra my position or my previous claims.
He can’t pass my ITT (not that I’ve asked him to), which means that he doesn’t understand the thing he’s trying to disagree with, which means that his disagreement is not actually pointing at my position; the things he finds ridiculous and offensive are cardboard cutouts of his own construction. More detail on that over here.
This response is manifestly untenable, given the comment of yours that I was responding to.
BTW I was surprised earlier to see you agree with the ‘relational’ piece of this comment because Duncan’s grandparent comment seems like it’s a pretty central example of that. (I view you as having more of a “visitor-commons” orientation towards LW, and Duncan has more of an orientation where this is a place where people inhabit their pairwise relationships, as well as more one-to-many relationships.)
Sorry, I’m not quite sure I follow the references here. You’re saying that… this comment… is a central example of… what, exactly?
That… seems like it’s probably accurate… I think? I think I’d have to more clearly understand what you’re getting at in your comment, in order to judge whether this part makes sense to me.
Sorry, my previous comment wasn’t very clear. Earlier I said:
and you responded with:
(and a few related comments) which made me think “hmm, I don’t think we mean the same thing by ‘relational’. Then Duncan’s comment had a frame that I would have described as ‘relational’—as in focusing on the relationships between the people saying and hearing the words—which you then described as transactional.
Ah, I see.
I think that the sense in which I would characterize Duncan’s description as “transactional” is… mostly orthogonal to the question of “is this a relational frame”. I don’t think that this has much to do with the “‘visitor commons’ vs. ‘pairwise relationships’” distinction, either (although that distinction is an interesting and possibly important one in its own right, and you’re certainly more right than wrong about where my preferences lie in that regard).
(There’s more that I could say about this, but I don’t know whether anything of importance hinges on this point. It seems like it mostly shouldn’t, but perhaps you are a better judge of that…)
A couple quick notes for now:
I agree with Duncan here it’s kinda silly to start the clock at “Killing Socrates”. Insofar as there’s a current live fight that is worth tracking separately from overall history, I think it probably starts in the comments of LW Team is adjusting moderation policy, and I think the recent-ish back and forth on Basics of Rationalist Discourse and “Rationalist Discourse” Is Like “Physicist Motors” is recent enough to be relevant (hence me including the in the OP)
I think Vaniver right now is focusing on resolving the point “is Said a liar?”, but not resolving the “who did most wrong?” question. (I’m not actually 100% sure on Vaniver’s goals/takes at the moment). I agree this is an important subquestion but it’s not the primary question I’m interested in.
I’m somewhat worried about this thread taking in more energy that it quite warrants, and making Duncan feel more persecuted than really makes sense here.
I roughly agree with Vaniver than “Liar!” isn’t the right accusation to have levied, but also don’t judge you harshly for having made it.
I think this comment of mine summarizes my relevant opinions here.
(tagging @Vaniver to make sure he’s at least tracking this comment)
Thanks.
I note (while acknowledging that this is a small and subtle distinction, but claiming that it is an important one nonetheless) that I said that I now categorize Said as a liar, which is an importantly and intentionally weaker claim than Said is a liar, i.e. “everyone should be able to see that he’s a liar” or “if you don’t think he’s a liar you are definitely wrong.”
(This is me in the past behaving in line with the points I just made under Said’s comment, about not confusing [how things seem to me] with [how they are] or [how they do or should seem to others].)
This is much much closer to saying “Liar!” than it is to not saying “Liar!” … if one is to round me off, that’s the correct place to round me off to. But it is still a rounding.
Nod, seems fair to note.