I’m flagging individual comments that clearly violate norms with brief mod-notes, which link back to this comment.
Before I dive into them, I wanted to summarize my understanding of Benquo’s position, based on our conversation yesterday.
Ben’s epistemic state is that the Punch Bug article is advocating for a change with a decent chance of escalating into violence with serious consequences.
Most of his concern is not about what Duncan said or intended to say, but in what the likely consequences of it getting promoted are. One important consideration here is downstream effects as the message gets propagated and simplified by the sorts of people who don’t make serious efforts to interpret 14,000 word texts and retain their nuance.
Independent of whether Ben was right in his assessment of the chance of serious escalation, he considered important that, at least in principle, there could be posts on LessWrong advocating for things with dangerous downstream consequences, and we should be cognizant of how our discourse norms will affect our ability to talk about that.
Ben’s model of why and how the post is likely to lead to bad outcomes are informed by the mechanics of how anti-Jew pogroms (and Nazis in particular) have functioned. This is not a claim about the magnitude of how strong the concern is, but it is a claim about the nuances of the dynamics involved. (He specifically noted: it very much would not have made sense to compare it to other forms of escalating violence, genocide, etc, because the dynamics there are different).
Ben experienced a lot of frustration that it is difficult to communicate about this sort of claim without people assuming the claim is about magnitude. In particular, he wanted it recognized that if our discourse norms include “Nazis are special, and either can’t be used as examples or comparisons, at least without a lot of additional interpretive effort on the part of the commenter”, this is creating a world where specific classes of behavioral patterns are harder to talk about.
That all said, by the end of the conversation, Ben and I seemed mostly in agreement on most of the specific areas where I considered his points clearly over the line. (see next section).
Bright Lines
There are a few specific bright lines that Benquo crossed. I don’t think avoiding these specific issues would have been sufficient for this discussion to have gone well, but it’s easiest to start with the clearest issues:
Irrelevant, inflammatory points – In Ben’s original comment, while I was generally worried about casually invoking Nazis and pogroms for rhetorical emphasis, the “german car” quip seemed particularly unjustified. Maybe it was an important part of Ben’s thought process that helped him notice something subtle, but as an argument it was clearly irrelevant to the broader point – I doubt Ben’s concerns depend on whether kids had invented Punch Bug or Punch Ford.
Attributing words to Duncan that he didn’t say – It’s not okay to substitute different words with different connotations and then ascribe them to someone as if they’d made a different claim. It’s doubly not okay when the connotations are as intensely bad as “ghetto.”
LessWrong does need to be the sort of place where, if you think someone is advocating for something with bad consequences, you are able to describe those consequences. But it is not okay to do that without clearly indicating that this is an inference you are making.
I think this issue was exacerbated by an issue Duncan noted, of Ben pointing towards norms that are indeed good and valuable, without acknowledging that Ben had already violated them.
Ascribing “Violent Sadist” as a motivation – This comment had a number of things going wrong with it. First, it was worded in an inflammatory way that made the discussion more adversarial. Ben reworded the comment in an attempt to alleviate that. I don’t think he went far enough in the rewording. But mostly I think the comment has a deeper, underlying issue – a failure of some combination of charity, imagination, empathy, or curiosity.
Said recently made some comments about penalizing “inner workings” rather than outward actions. I think there is something important to that – at the very least I think LW users (and mods) have a responsibility to be very careful with that sort of reasoning.
In this case, I can lay down some concrete rules/policies – if you find yourself making guesses about the motivations of someone you’re debating, and your explanation is “violent sadism”, you have a responsibility to at least come up with one alternate explanation. A serious explanation, not “house elves are stealing our magic”. (You should probably generally do this whenever someone advocates for something you find baffling)
But this rule doesn’t really feel sufficient to me. It accounts for this particular failure mode, but doesn’t quite carve reality at the joints and prevent a broad class of related issues. And here we get into a bit less “bright line” territory.
I have guesses about what inner-generator Ben was running here. In our in-person conversation we chatted a bit about it. While wearing my official moderator hat, what I feel comfortable saying is:
There will be times when we see a user making a string of bad comments. We need to be able to both say “this is bad for this specific reasons”, and also “something about your inner generator that is producing these comments is reliably degrading LessWrong discourse, and it is your responsibility to figure out what, and do something differently.” It cannot always be the full responsibility of the LessWrong team to tell all commenters precisely what needs to change.
Importantly, I think that even if Ben had avoided the error modes listed above, there’s a good chance the conversation still might have deteriorated (perhaps creating an even harder moderating challenge, since instead of clear over-the-line comments to critique, there would have been a bunch of juuuuust under-the-line comments requiring difficult judgment calls).
I don’t have anything specific to say here, other than “in a thread like this (where the primary discussion is about shifting social consensus), I think it’s important for everyone to spend more than what-feels-like-their-fair-share of charity and interpretive effort, because otherwise things tend to slowly escalate.”
Towards Clearer Policies
We had considered straight-up deleting some of Benquo’s comments a week ago. We didn’t end up doing so. I don’t think we had principled reasoning one way or the other, but my own post-facto introspection is that, at least for myself, it felt like most ways we could intervene had a strong chance of making things worse. It felt important to talk to Ben in person, and in general take the time to think and gather information before taking much explicit action.
There are certain moderation tools that would have made this easier to handle – I think we probably should have locked the entire thread after our initial warning and Ben’s subsequent reply, until we’d had a chance to talk in person.
I am embarrassed to note that I explicitly considered and discarded the idea that we could manually lock replies on each individual comment (since we do have the tool to do that), which would have taken maybe 10 minutes of effort, and my brain just straight up slided off that because it felt silly to spend 10 minutes doing a thing that I could envision how to automate.
There were a few other actions we could have taken with minimal effort that would have made this experience much less stressful for everyone involved without making major tradeoffs, and all I have to say is that none of us thought of them. In some cases I notice specific biases that led to us not thinking of it, in other cases it was just a matter of “this is a high dimensional problem space that we were not very experienced in – we had not yet formed good models of it to make thinking about it easier.”
Over the past week we’ve collectively spent a couple dozen hours thinking about both this specific discussion, and moderation in general. I don’t think our next major moderation crisis will be anywhere near perfect, but I do expect us to be better at noticing quick, low-cost actions we can take to communicate better and keep the situation from escalating.
I also expect that at least for situation similar to those we’ve faced before, we’ll be able to act faster.
Explicit Warnings
With all that in mind, specific mod notices for Ben.
For blatant violations of the type described above, we’ll be aiming to quickly flag comments that are crossing the line. Depending on circumstances, we may delete comments that cross the line, with an explanation of why (note: our delete button sends the user a PM containing the text of the comment, so the content isn’t lost)
In the future, in discussions where you notice yourself tempted to make comments similar the ones we’re flagging here, I ask that you put a lot more effort into ensuring the comments are high quality. This is important both for avoiding the explicit bright lines as well as grey-area-just-under-the-line comments.
Yes, this is a cost that makes certain kinds of conversations harder. And we should be cognizant of that. But my current take is that this is just a necessary cost we have to pay.
I’m fairly confident, based on extended conversations I’ve had with Ben over the years as well as the one last night, that the situation won’t escalate past the point of periodic warnings. I disagree with Ben on some important object level (and meta level) things, but I trust Ben’s underlying goals, principles and error-correction algorithms to be aligned with LessWrong’s overall mission.
But because we do need legibly clear rules, it’s important to note that if this sort of discussion became a pattern, we would follow a policy of “warnings of increasing severity and publicity, followed by escalating temp bans.”
I also want to note what I think is the biggest, most obvious mistake I made here. When I wrote the initial comment criticizing Benquo, I went through several drafts. Some of them focused more on specific examples of why I thought the comment was unacceptable and where it crossed the line. Other drafts focused on an overall lack-of-charity, in a discussion where charity was crucial.
Those drafts felt long/meandering/hard-to-parse, and after a couple hours I said “gah, fuck it, ship it”, and then while in fuck-it-ship-it mode, deleted a couple paragraphs in an attempt for brevity. In the process, I gutted the actual core point, leaving a pretty superficial criticism about “don’t literally invoke Nazism please.”
I think this was the point where making slightly different choices with a clearer head could have had a huge impact on the overall frame of the conversation.
Then I didn’t really re-examine the comment until a few days later. And at that point, it had seemed to me that it made more sense to wait till I’d talked to Ben, and then write a more thorough comment that fully addressed everything. I still wasn’t modeling costs to Duncan, and didn’t fully understand that until the point where Duncan called them out explicitly.
I agree the “german car” bit was wrong, and I’m sorry I included it.
On the second and third points, I can see reasons to discourage how I said what I said regardless of what I meant by it, but dispute some of the characterization of my point of view in those comments.
The connotations in my head when I think “ghetto” are pretty mild. Maybe you’re thinking specifically about the way they were used during WWII as staging areas for genocide, but they’d been around for hundreds of years before that & weren’t really considered atrocities on their own, just … places people lived, that helped keep a minority group visibly distinct. Ghettos aren’t some sort of unspeakable horror, just an obviously dangerous precedent.
“Violent sadists” wasn’t a good way to say it, but I still worry that you’re not parsing the literal content of my words there. In the specific context of that comment, the hypothesis that my interlocutor thought everyone would derive some benefit from getting to punch others with impunity sometimes was a reasonable hypothesis to generate. I can see how the emotional valence of the term I used would make this hard to see, though.
On the other hand, the bullet points directly trying to describe my overall point of view seem like a basically fair summary. Thanks for doing that work.
Ghettos aren’t some sort of unspeakable horror, just an obviously dangerous precedent.
I think this is a plausible interpretation of ghetto in-a-vacuum, but the associations with are strong enough that you basically can’t bring up the world ghetto without going out of your way to disclaim that.
And if you’re bringing it up as a description of someone else’s words, I think you actually need to substitute the definition for word. For a more extreme but hopefully clear example: If you say about someone “I’m not saying they’re going to put people in a literal nazi ghetto, I’m just saying <whatever you’re actually saying>”, one of the primary actions you’re doing is associating them with the phrase “literal nazi ghetto”, even if you specifically said you were not doing that, because that’s how language and brains work.
In this particular case, I’m honestly a bit skeptical that you think the connotations of ghetto are mild – maybe they’re mild in isolation, but your whole point was that this wouldn’t remain in isolation (which is the exact reason why ghettos game to have their associations in the first place)
In general, I think a policy of “if you’re going to bring up things associated with Nazi Germany, you are responsible for putting a much larger share of interpretive labor in, moreso than I usually think is necessary for an overton-window-fight.
This reads to me like a demand that I try to cooperate with attempts to prevent common knowledge of the likely bad consequences of policy proposals for me and people like me. Maybe that’s not what you mean and you’ve thought through the relevant balance. If so, I’d like an account of that, or a clear description of what aspects of this you still consider unsolved problems / asymmetric burdens. The latter could be a comparatively easy way to remove some time pressure for the former.
This does not seem to follow from what I’m saying at all. (in particular, it seems extremely reasonable to me to not attribute words to other peoplethat have different/stronger/worse connotations)
I’m asking you to cooperate with the general principle of building a conversational structure where we can actually talk about anything important, because if we do the thing you’re doing here you don’t get to have that.
I’m asking you, if you enforce a norm that there should be a very high burden of interpretive labor for using terms like “ghetto” in that way, to equally enforce a norm that there should be a very high burden for promoting proposals to create things that at all resemble ghettos, since the former norm makes those proposals substantially more difficult to criticize within the bounds of accepted discourse. Insofar as you’re asking me not just to abstain from literally saying that Duncan’s a Nazi or planning a genocide (I didn’t, and as far as I can tell he’s not), but from referring to structural similarities between his comparatively mild proposal and some things widely acknowledged to be bad, it seems like there should be a similarly sized avoidance zone around proposals like his, and unambiguously approving linkposts to such proposals.
I have two responses, one very strong/confident, one less confident.
1) Strong/Confident – It is definitely not okay to attribute words to people they didn’t say, in the first place. This is a fundamental cornerstone of How to Have Nice Things. The fact that I’m further increasing the burden for highly-charged rewordings doesn’t seem like it should be coming into play here.
You literally said Duncan said people who didn’t like punchbug could have a ghetto. He didn’t say that. That isn’t okay.
If someone is advocating for something that will lead to ghettos, the way to say that is “All the interpretations I can think of how to implement this sound like a ghetto.” This does not seem at all like an unreasonable burden.
2) Weaker claim that I’m less certain of:
I think there is stronger pressure (on LW and elsewhere) not to say things that pattern-match to anything resembling abuse, than there is to not talk about Nazis. (Moreover, outside of LW, there is the reverse pattern – comparing people to Nazis is incentivizedso badly that for LW to be different than mediocre internet discourse, it has to exert a strong counterpressure to be more cautious with it).
Meanwhile, the last two times Duncan attempted to say anything nuanced that vaguely pattern-matched to “thing that might possibly be dangerous”, he got hounded with hateful internet rhetoric. Most of the the responses I’ve seen to Punch Bug have ranged from “this makes good points but I don’t like Punch Bug” to “this seems abusive and bad.” There might be possible worlds where this is risking escalating off a cliff, but this world doesn’t seem to be one of them.
Yes, it is important that we build a place with rules that protect everyone. I think they are in fact protecting you. But people are always incentivized to notice the times when the system is imposing costs on them.
You literally said Duncan said people who didn’t like punchbug could have a ghetto. He didn’t say that. That isn’t okay.
Thanks for making this clear and direct.
I characterized what I thought Duncan’s words meant, in my own words.
The first time I did this I quoted the passage I thought implied this in the same comment. I don’t think it’s a plausible construal of this at all that Duncan had literally written the words not explicitly quoted, which were directly pointing to a following, overtly blockquoted, verbatim quote.
The other times I did this, on the same page (but sometimes not in the same thread), it’s more plausible that a reader who hadn’t carefully read the whole page could get that impression. But this is not the same thing as a literal assertion that Duncan had said that thing in those words.
Usually when person X writes “person Y said that Z”, there is some ambiguity about whether Z is a quotation (exact or approximate) or a characterization in Y’s own voice. I meant to say the latter, but I see how I was unclear in a way that let people wrongly infer the former. If anyone was genuinely misled by this, I’m sorry for my negligence.
Your specific proposal seems like a substantial improvement over the way I worded things, and I hope to meet that standard in the future. I think that I made an inadequate effort to distinguish between those two possible construals. But it’s hard for me to engage with criticism that pretends that a case with genuine ambiguity is unambiguous.
This seems like a fair response in isolation; if the only thing were paraphrasing Duncan’s suggestion as forming a ghetto, that seems plausibly acceptable. In particular, my understanding is that some school cafeterias are entirely peanut-free (including barring children from bringing products containing peanuts from home) in order to make the environment more safe for children with peanut allergies; a proposal to move from that environment to one in which there is simply a peanut-free section of the cafeteria could be said to be isolating those students with allergies. Calling that ghettoizing those children might be extreme (because of the difference in scale), but would be correct about the direction of the change (especially since it is ‘for their protection,’ which was originally true of many of the Jewish ghettos in Europe).
It seems to me like the primary question then is something like: “is enough of the original claim maintained that this summary seems fair, at least to an outside observer?” (The original author agreeing is probably too high a standard, as a summary that’s harsh but fair is unlikely to be acknowledged as fair.) A paraphrase that loses too much in the way of nuance, or which ends up being too hostile or uncharitable, seems like it should call for some sort of criticism or correction. It’s easiest to settle claims about whether or not direct quotes are accurate, but it’s also important to be tracking whether summaries or interpretations are fair and sensible.
Furthermore, moderation often involves noticing patterns of behavior and responding to them as a whole, using particular pieces of behavior as examples and evidence rather than independent cases, so that a long sequences of small pieces of slack doesn’t result in large amounts of unaccountability. In the context of the whole conversation, it seems difficult to separate the question of whether or not this paraphrase is fair or accurate from the question of whether or not a picture is being painted of Duncan as proto-Nazi in ways that are epistemically unsound. (That is, the Litany of Tarski applies here; if Duncan is actually a proto-Nazi, I desire to believe it, and if not, I want to not believe it, and that relies on having epistemic machinery that reliably points towards the truth.)
I was specifically objecting to the details of Raemon’s framing. I agree that in the context of the other things I said I should have been much more careful about the term “ghetto.” I strongly agree that moderators should attend to things like this. Your explanation basically makes sense to me.
I’m very unhappy that this ended up in a place where “is Duncan a proto-Nazi?” is a plausible framing of the question. I’m going to put quite a bit of thought to whether there was a way to object to this clearly and vividly, in a way that didn’t obscure the underlying threat model, without accidentally making it about Duncan personally.
One important thing I notice, that I hadn’t said, and I think would have been really helpful to figure-out-and-then-say-much-earlier:
If you find yourself wanting to communicate a concept that feels likely to get misinterpreted or escalate tensions in the conversation, I’d be very enthusiastic, both as a mod and as a person who knows Ben and thinks he’s trying to do and say important, nuanced things, to try to help do that. (esp if we hash out some thoughts privately before you post publicly)
This is easier the earlier it is in the conversation. I think it’d have been better if we reached out in that frame about your initial comment here, and it’d have been better still if you’d noticed as you were writing the comment that it was the sort of thing likely to get misinterpreted, and (if we had an easy way to do so), reach out to us about how to approach the conversation.
i.e. I think this sort of conversation requires a lot more interpretive labor than normal, but I’m happy to help that. (With limited bandwidth I can’t do this all the time, but it seems like a good approach insofar as we can, and worth prioritizing)
Yeah, I think there was something off about my framing, especially for a comment I was tagging as “high confidence.” I basically endorse Vaniver’s take here, and I think a more accurate and fair thing I would have said was:
1) I have strong sense, that I am quite confident is pointing at a real, important thing, that the ghetto comment was the sort of thing that reliably leads to a deteriorating discussion (and that this should be relatively common knowledge)
2) I had a moderate confidence that the shape-of-the-reason was close to the reason I gave (but, it did depend on context with other things), and I should have been some combination of “express less explicit confidence that I’m getting the details right” and “maybe orient more curiously about what some details/nuances were.”
On point 2, I think this is one of those cases where depending on your metric, there’s a strong power gradient in one or the other direction.
On one hand, I agree that Duncan’s faced a lot of often quite nasty criticism on the internet. While I haven’t looked at the stuff on Tumblr, I’m ready to believe that a lot of it was terrible and unfair. There is also a lot of criticism in the comments to this linkpost, and not a lot of praise or agreement by volume.
On the other, Duncan’s employed as an instructor at CFAR—an official representative of the community’s rationality standards, insofar as there’s any official apparatus at all—and the post was approvingly linked by another CFAR staff member.
Also, at the time of my initial comment, one of the top comments—maybe the most-upvoted at the time, not sure—was thanking Duncan for his bravery in writing this, with a response by a LessWrong moderator endorsing this. Other fairly central members of the Bay Area Rationalist and LessWrong communities participated in Dragon Army.
I agree that individuals’ perceptions of costs are likely to be biased in the way you say. I’m not asking you to privilege my perspective here.
the hypothesis that my interlocutor thought everyone would derive some benefit from getting to punch others with impunity sometimes was a reasonable hypothesis to generate.
So I think, from your epistemic state, that this is a plausible interpretation of what’s going on (esp. because neither Duncan nor anyone else has actually given much account here of why Punch Bug might be fun). But yes, I’m pretty confident I understand what you’re saying here, and think you’re missing important stuff. (At the very least,enjoying violence isn’t the only thing going on)
Some things that may or may not be going on with Punch Bug (I haven’t actually formed a clear opinion on this and it is a a bit confusing so I’m sympathetic here) include:
competition – sports are often filled with pain, but the enjoyment is largely about physicality, competition, overcoming difficult obstacles. Something like boxing is explicitly about punching people, and I’m sure “enjoying punching” is part of that, but I bet the sort of person who likes that cares more about proving themselves against a live opponent.
part of the whole point here is that pain can be calibrated, and if you’re calibrated to find a moderate punch to be something you can shrug off, that changes the nature of how you perceive it. Compare with verbal fencing – some people enjoy trading barbs/quips/trashtalk. And one thing that can happen with a clever insult is that people feel genuinely hurt/ashamed by it, but another thing that happen is that everyone involved is thinking something more like “ah, what clever competitive art we are engaging with.”
I think part of the thing with roughhousing in general is just getting to touch other humans. (and if you’re calibrated to the pain of a punch a la above, then you’re mostly getting a particular flavor of the positive-aspects of that, rather than the negative.)
I’m not just worried about the emotional valence of violent sadist, but that those words, together, convey a clear(ish) connotation that goes beyond “person that enjoys inflicting pain.”
These all seem like things that would strongly favor games like “sometimes roughhouse with your friends with punchback allowed and sensitivity over repeated interactions to how people are feeling about the whole thing” over games like “punchbug”. I’m justified in deprecating hypotheses somewhat that would predict different things than the ones that actually happened.
I’m definitely not claiming to not be missing important stuff. A more clearly grounded query would have been way better, of course, but as it turned out, generating more hypotheses about the appeal of punchbug would have been entirely beside the point, as the problem turned out to be that SilentCal thought I meant something much narrower by “asymmetric,” which excluded formally symmetrical rules with disparate impact.
You’re saying that I generated not only a wrong but an unreasonably wrong hypothesis, implying that it’s on the level of “house elves are stealing our magic.” I’d like you to say exactly what you think that hypothesis was, because it’s entirely unclear to me.
Ah, sorry – I was saying something slightly different:
I’m not sure if I think “people enjoy X because violent sadism” is a reasonable hypothesis or not. I brought up house-elves-steal-magic” to describe, not that hypothesis, but whatever second hypothesis you might generate.
I’m flagging individual comments that clearly violate norms with brief mod-notes, which link back to this comment.
Before I dive into them, I wanted to summarize my understanding of Benquo’s position, based on our conversation yesterday.
Ben’s epistemic state is that the Punch Bug article is advocating for a change with a decent chance of escalating into violence with serious consequences.
Most of his concern is not about what Duncan said or intended to say, but in what the likely consequences of it getting promoted are. One important consideration here is downstream effects as the message gets propagated and simplified by the sorts of people who don’t make serious efforts to interpret 14,000 word texts and retain their nuance.
Independent of whether Ben was right in his assessment of the chance of serious escalation, he considered important that, at least in principle, there could be posts on LessWrong advocating for things with dangerous downstream consequences, and we should be cognizant of how our discourse norms will affect our ability to talk about that.
Ben’s model of why and how the post is likely to lead to bad outcomes are informed by the mechanics of how anti-Jew pogroms (and Nazis in particular) have functioned. This is not a claim about the magnitude of how strong the concern is, but it is a claim about the nuances of the dynamics involved. (He specifically noted: it very much would not have made sense to compare it to other forms of escalating violence, genocide, etc, because the dynamics there are different).
Ben experienced a lot of frustration that it is difficult to communicate about this sort of claim without people assuming the claim is about magnitude. In particular, he wanted it recognized that if our discourse norms include “Nazis are special, and either can’t be used as examples or comparisons, at least without a lot of additional interpretive effort on the part of the commenter”, this is creating a world where specific classes of behavioral patterns are harder to talk about.
That all said, by the end of the conversation, Ben and I seemed mostly in agreement on most of the specific areas where I considered his points clearly over the line. (see next section).
Bright Lines
There are a few specific bright lines that Benquo crossed. I don’t think avoiding these specific issues would have been sufficient for this discussion to have gone well, but it’s easiest to start with the clearest issues:
Irrelevant, inflammatory points – In Ben’s original comment, while I was generally worried about casually invoking Nazis and pogroms for rhetorical emphasis, the “german car” quip seemed particularly unjustified. Maybe it was an important part of Ben’s thought process that helped him notice something subtle, but as an argument it was clearly irrelevant to the broader point – I doubt Ben’s concerns depend on whether kids had invented Punch Bug or Punch Ford.
Attributing words to Duncan that he didn’t say – It’s not okay to substitute different words with different connotations and then ascribe them to someone as if they’d made a different claim. It’s doubly not okay when the connotations are as intensely bad as “ghetto.”
LessWrong does need to be the sort of place where, if you think someone is advocating for something with bad consequences, you are able to describe those consequences. But it is not okay to do that without clearly indicating that this is an inference you are making.
I think this issue was exacerbated by an issue Duncan noted, of Ben pointing towards norms that are indeed good and valuable, without acknowledging that Ben had already violated them.
Ascribing “Violent Sadist” as a motivation – This comment had a number of things going wrong with it. First, it was worded in an inflammatory way that made the discussion more adversarial. Ben reworded the comment in an attempt to alleviate that. I don’t think he went far enough in the rewording. But mostly I think the comment has a deeper, underlying issue – a failure of some combination of charity, imagination, empathy, or curiosity.
Said recently made some comments about penalizing “inner workings” rather than outward actions. I think there is something important to that – at the very least I think LW users (and mods) have a responsibility to be very careful with that sort of reasoning.
In this case, I can lay down some concrete rules/policies – if you find yourself making guesses about the motivations of someone you’re debating, and your explanation is “violent sadism”, you have a responsibility to at least come up with one alternate explanation. A serious explanation, not “house elves are stealing our magic”. (You should probably generally do this whenever someone advocates for something you find baffling)
But this rule doesn’t really feel sufficient to me. It accounts for this particular failure mode, but doesn’t quite carve reality at the joints and prevent a broad class of related issues. And here we get into a bit less “bright line” territory.
I have guesses about what inner-generator Ben was running here. In our in-person conversation we chatted a bit about it. While wearing my official moderator hat, what I feel comfortable saying is:
There will be times when we see a user making a string of bad comments. We need to be able to both say “this is bad for this specific reasons”, and also “something about your inner generator that is producing these comments is reliably degrading LessWrong discourse, and it is your responsibility to figure out what, and do something differently.” It cannot always be the full responsibility of the LessWrong team to tell all commenters precisely what needs to change.
Importantly, I think that even if Ben had avoided the error modes listed above, there’s a good chance the conversation still might have deteriorated (perhaps creating an even harder moderating challenge, since instead of clear over-the-line comments to critique, there would have been a bunch of juuuuust under-the-line comments requiring difficult judgment calls).
I don’t have anything specific to say here, other than “in a thread like this (where the primary discussion is about shifting social consensus), I think it’s important for everyone to spend more than what-feels-like-their-fair-share of charity and interpretive effort, because otherwise things tend to slowly escalate.”
Towards Clearer Policies
We had considered straight-up deleting some of Benquo’s comments a week ago. We didn’t end up doing so. I don’t think we had principled reasoning one way or the other, but my own post-facto introspection is that, at least for myself, it felt like most ways we could intervene had a strong chance of making things worse. It felt important to talk to Ben in person, and in general take the time to think and gather information before taking much explicit action.
There are certain moderation tools that would have made this easier to handle – I think we probably should have locked the entire thread after our initial warning and Ben’s subsequent reply, until we’d had a chance to talk in person.
I am embarrassed to note that I explicitly considered and discarded the idea that we could manually lock replies on each individual comment (since we do have the tool to do that), which would have taken maybe 10 minutes of effort, and my brain just straight up slided off that because it felt silly to spend 10 minutes doing a thing that I could envision how to automate.
There were a few other actions we could have taken with minimal effort that would have made this experience much less stressful for everyone involved without making major tradeoffs, and all I have to say is that none of us thought of them. In some cases I notice specific biases that led to us not thinking of it, in other cases it was just a matter of “this is a high dimensional problem space that we were not very experienced in – we had not yet formed good models of it to make thinking about it easier.”
Over the past week we’ve collectively spent a couple dozen hours thinking about both this specific discussion, and moderation in general. I don’t think our next major moderation crisis will be anywhere near perfect, but I do expect us to be better at noticing quick, low-cost actions we can take to communicate better and keep the situation from escalating.
I also expect that at least for situation similar to those we’ve faced before, we’ll be able to act faster.
Explicit Warnings
With all that in mind, specific mod notices for Ben.
For blatant violations of the type described above, we’ll be aiming to quickly flag comments that are crossing the line. Depending on circumstances, we may delete comments that cross the line, with an explanation of why (note: our delete button sends the user a PM containing the text of the comment, so the content isn’t lost)
In the future, in discussions where you notice yourself tempted to make comments similar the ones we’re flagging here, I ask that you put a lot more effort into ensuring the comments are high quality. This is important both for avoiding the explicit bright lines as well as grey-area-just-under-the-line comments.
Yes, this is a cost that makes certain kinds of conversations harder. And we should be cognizant of that. But my current take is that this is just a necessary cost we have to pay.
I’m fairly confident, based on extended conversations I’ve had with Ben over the years as well as the one last night, that the situation won’t escalate past the point of periodic warnings. I disagree with Ben on some important object level (and meta level) things, but I trust Ben’s underlying goals, principles and error-correction algorithms to be aligned with LessWrong’s overall mission.
But because we do need legibly clear rules, it’s important to note that if this sort of discussion became a pattern, we would follow a policy of “warnings of increasing severity and publicity, followed by escalating temp bans.”
I also want to note what I think is the biggest, most obvious mistake I made here. When I wrote the initial comment criticizing Benquo, I went through several drafts. Some of them focused more on specific examples of why I thought the comment was unacceptable and where it crossed the line. Other drafts focused on an overall lack-of-charity, in a discussion where charity was crucial.
Those drafts felt long/meandering/hard-to-parse, and after a couple hours I said “gah, fuck it, ship it”, and then while in fuck-it-ship-it mode, deleted a couple paragraphs in an attempt for brevity. In the process, I gutted the actual core point, leaving a pretty superficial criticism about “don’t literally invoke Nazism please.”
I think this was the point where making slightly different choices with a clearer head could have had a huge impact on the overall frame of the conversation.
Then I didn’t really re-examine the comment until a few days later. And at that point, it had seemed to me that it made more sense to wait till I’d talked to Ben, and then write a more thorough comment that fully addressed everything. I still wasn’t modeling costs to Duncan, and didn’t fully understand that until the point where Duncan called them out explicitly.
I agree the “german car” bit was wrong, and I’m sorry I included it.
On the second and third points, I can see reasons to discourage how I said what I said regardless of what I meant by it, but dispute some of the characterization of my point of view in those comments.
The connotations in my head when I think “ghetto” are pretty mild. Maybe you’re thinking specifically about the way they were used during WWII as staging areas for genocide, but they’d been around for hundreds of years before that & weren’t really considered atrocities on their own, just … places people lived, that helped keep a minority group visibly distinct. Ghettos aren’t some sort of unspeakable horror, just an obviously dangerous precedent.
“Violent sadists” wasn’t a good way to say it, but I still worry that you’re not parsing the literal content of my words there. In the specific context of that comment, the hypothesis that my interlocutor thought everyone would derive some benefit from getting to punch others with impunity sometimes was a reasonable hypothesis to generate. I can see how the emotional valence of the term I used would make this hard to see, though.
On the other hand, the bullet points directly trying to describe my overall point of view seem like a basically fair summary. Thanks for doing that work.
I think this is a plausible interpretation of ghetto in-a-vacuum, but the associations with are strong enough that you basically can’t bring up the world ghetto without going out of your way to disclaim that.
And if you’re bringing it up as a description of someone else’s words, I think you actually need to substitute the definition for word. For a more extreme but hopefully clear example: If you say about someone “I’m not saying they’re going to put people in a literal nazi ghetto, I’m just saying <whatever you’re actually saying>”, one of the primary actions you’re doing is associating them with the phrase “literal nazi ghetto”, even if you specifically said you were not doing that, because that’s how language and brains work.
In this particular case, I’m honestly a bit skeptical that you think the connotations of ghetto are mild – maybe they’re mild in isolation, but your whole point was that this wouldn’t remain in isolation (which is the exact reason why ghettos game to have their associations in the first place)
In general, I think a policy of “if you’re going to bring up things associated with Nazi Germany, you are responsible for putting a much larger share of interpretive labor in, moreso than I usually think is necessary for an overton-window-fight.
This reads to me like a demand that I try to cooperate with attempts to prevent common knowledge of the likely bad consequences of policy proposals for me and people like me. Maybe that’s not what you mean and you’ve thought through the relevant balance. If so, I’d like an account of that, or a clear description of what aspects of this you still consider unsolved problems / asymmetric burdens. The latter could be a comparatively easy way to remove some time pressure for the former.
This does not seem to follow from what I’m saying at all. (in particular, it seems extremely reasonable to me to not attribute words to other people that have different/stronger/worse connotations)
I’m asking you to cooperate with the general principle of building a conversational structure where we can actually talk about anything important, because if we do the thing you’re doing here you don’t get to have that.
I’m asking you, if you enforce a norm that there should be a very high burden of interpretive labor for using terms like “ghetto” in that way, to equally enforce a norm that there should be a very high burden for promoting proposals to create things that at all resemble ghettos, since the former norm makes those proposals substantially more difficult to criticize within the bounds of accepted discourse. Insofar as you’re asking me not just to abstain from literally saying that Duncan’s a Nazi or planning a genocide (I didn’t, and as far as I can tell he’s not), but from referring to structural similarities between his comparatively mild proposal and some things widely acknowledged to be bad, it seems like there should be a similarly sized avoidance zone around proposals like his, and unambiguously approving linkposts to such proposals.
I have two responses, one very strong/confident, one less confident.
1) Strong/Confident – It is definitely not okay to attribute words to people they didn’t say, in the first place. This is a fundamental cornerstone of How to Have Nice Things. The fact that I’m further increasing the burden for highly-charged rewordings doesn’t seem like it should be coming into play here.
You literally said Duncan said people who didn’t like punchbug could have a ghetto. He didn’t say that. That isn’t okay.
If someone is advocating for something that will lead to ghettos, the way to say that is “All the interpretations I can think of how to implement this sound like a ghetto.” This does not seem at all like an unreasonable burden.
2) Weaker claim that I’m less certain of:
I think there is stronger pressure (on LW and elsewhere) not to say things that pattern-match to anything resembling abuse, than there is to not talk about Nazis. (Moreover, outside of LW, there is the reverse pattern – comparing people to Nazis is incentivized so badly that for LW to be different than mediocre internet discourse, it has to exert a strong counterpressure to be more cautious with it).
Meanwhile, the last two times Duncan attempted to say anything nuanced that vaguely pattern-matched to “thing that might possibly be dangerous”, he got hounded with hateful internet rhetoric. Most of the the responses I’ve seen to Punch Bug have ranged from “this makes good points but I don’t like Punch Bug” to “this seems abusive and bad.” There might be possible worlds where this is risking escalating off a cliff, but this world doesn’t seem to be one of them.
Yes, it is important that we build a place with rules that protect everyone. I think they are in fact protecting you. But people are always incentivized to notice the times when the system is imposing costs on them.
Thanks for making this clear and direct.
I characterized what I thought Duncan’s words meant, in my own words.
The first time I did this I quoted the passage I thought implied this in the same comment. I don’t think it’s a plausible construal of this at all that Duncan had literally written the words not explicitly quoted, which were directly pointing to a following, overtly blockquoted, verbatim quote.
The other times I did this, on the same page (but sometimes not in the same thread), it’s more plausible that a reader who hadn’t carefully read the whole page could get that impression. But this is not the same thing as a literal assertion that Duncan had said that thing in those words.
Usually when person X writes “person Y said that Z”, there is some ambiguity about whether Z is a quotation (exact or approximate) or a characterization in Y’s own voice. I meant to say the latter, but I see how I was unclear in a way that let people wrongly infer the former. If anyone was genuinely misled by this, I’m sorry for my negligence.
Your specific proposal seems like a substantial improvement over the way I worded things, and I hope to meet that standard in the future. I think that I made an inadequate effort to distinguish between those two possible construals. But it’s hard for me to engage with criticism that pretends that a case with genuine ambiguity is unambiguous.
This seems like a fair response in isolation; if the only thing were paraphrasing Duncan’s suggestion as forming a ghetto, that seems plausibly acceptable. In particular, my understanding is that some school cafeterias are entirely peanut-free (including barring children from bringing products containing peanuts from home) in order to make the environment more safe for children with peanut allergies; a proposal to move from that environment to one in which there is simply a peanut-free section of the cafeteria could be said to be isolating those students with allergies. Calling that ghettoizing those children might be extreme (because of the difference in scale), but would be correct about the direction of the change (especially since it is ‘for their protection,’ which was originally true of many of the Jewish ghettos in Europe).
It seems to me like the primary question then is something like: “is enough of the original claim maintained that this summary seems fair, at least to an outside observer?” (The original author agreeing is probably too high a standard, as a summary that’s harsh but fair is unlikely to be acknowledged as fair.) A paraphrase that loses too much in the way of nuance, or which ends up being too hostile or uncharitable, seems like it should call for some sort of criticism or correction. It’s easiest to settle claims about whether or not direct quotes are accurate, but it’s also important to be tracking whether summaries or interpretations are fair and sensible.
Furthermore, moderation often involves noticing patterns of behavior and responding to them as a whole, using particular pieces of behavior as examples and evidence rather than independent cases, so that a long sequences of small pieces of slack doesn’t result in large amounts of unaccountability. In the context of the whole conversation, it seems difficult to separate the question of whether or not this paraphrase is fair or accurate from the question of whether or not a picture is being painted of Duncan as proto-Nazi in ways that are epistemically unsound. (That is, the Litany of Tarski applies here; if Duncan is actually a proto-Nazi, I desire to believe it, and if not, I want to not believe it, and that relies on having epistemic machinery that reliably points towards the truth.)
I was specifically objecting to the details of Raemon’s framing. I agree that in the context of the other things I said I should have been much more careful about the term “ghetto.” I strongly agree that moderators should attend to things like this. Your explanation basically makes sense to me.
I’m very unhappy that this ended up in a place where “is Duncan a proto-Nazi?” is a plausible framing of the question. I’m going to put quite a bit of thought to whether there was a way to object to this clearly and vividly, in a way that didn’t obscure the underlying threat model, without accidentally making it about Duncan personally.
[note: Somewhat off the cuff]
One important thing I notice, that I hadn’t said, and I think would have been really helpful to figure-out-and-then-say-much-earlier:
If you find yourself wanting to communicate a concept that feels likely to get misinterpreted or escalate tensions in the conversation, I’d be very enthusiastic, both as a mod and as a person who knows Ben and thinks he’s trying to do and say important, nuanced things, to try to help do that. (esp if we hash out some thoughts privately before you post publicly)
This is easier the earlier it is in the conversation. I think it’d have been better if we reached out in that frame about your initial comment here, and it’d have been better still if you’d noticed as you were writing the comment that it was the sort of thing likely to get misinterpreted, and (if we had an easy way to do so), reach out to us about how to approach the conversation.
i.e. I think this sort of conversation requires a lot more interpretive labor than normal, but I’m happy to help that. (With limited bandwidth I can’t do this all the time, but it seems like a good approach insofar as we can, and worth prioritizing)
Thanks, good to have this offer :)
Yeah, I think there was something off about my framing, especially for a comment I was tagging as “high confidence.” I basically endorse Vaniver’s take here, and I think a more accurate and fair thing I would have said was:
1) I have strong sense, that I am quite confident is pointing at a real, important thing, that the ghetto comment was the sort of thing that reliably leads to a deteriorating discussion (and that this should be relatively common knowledge)
2) I had a moderate confidence that the shape-of-the-reason was close to the reason I gave (but, it did depend on context with other things), and I should have been some combination of “express less explicit confidence that I’m getting the details right” and “maybe orient more curiously about what some details/nuances were.”
(Note: I am running low on bandwidth to talk about this particular issue. I’ll respond up to 2 more times to clarify any remaining points)
On point 2, I think this is one of those cases where depending on your metric, there’s a strong power gradient in one or the other direction.
On one hand, I agree that Duncan’s faced a lot of often quite nasty criticism on the internet. While I haven’t looked at the stuff on Tumblr, I’m ready to believe that a lot of it was terrible and unfair. There is also a lot of criticism in the comments to this linkpost, and not a lot of praise or agreement by volume.
On the other, Duncan’s employed as an instructor at CFAR—an official representative of the community’s rationality standards, insofar as there’s any official apparatus at all—and the post was approvingly linked by another CFAR staff member.
Also, at the time of my initial comment, one of the top comments—maybe the most-upvoted at the time, not sure—was thanking Duncan for his bravery in writing this, with a response by a LessWrong moderator endorsing this. Other fairly central members of the Bay Area Rationalist and LessWrong communities participated in Dragon Army.
I agree that individuals’ perceptions of costs are likely to be biased in the way you say. I’m not asking you to privilege my perspective here.
So I think, from your epistemic state, that this is a plausible interpretation of what’s going on (esp. because neither Duncan nor anyone else has actually given much account here of why Punch Bug might be fun). But yes, I’m pretty confident I understand what you’re saying here, and think you’re missing important stuff. (At the very least, enjoying violence isn’t the only thing going on)
Some things that may or may not be going on with Punch Bug (I haven’t actually formed a clear opinion on this and it is a a bit confusing so I’m sympathetic here) include:
competition – sports are often filled with pain, but the enjoyment is largely about physicality, competition, overcoming difficult obstacles. Something like boxing is explicitly about punching people, and I’m sure “enjoying punching” is part of that, but I bet the sort of person who likes that cares more about proving themselves against a live opponent.
part of the whole point here is that pain can be calibrated, and if you’re calibrated to find a moderate punch to be something you can shrug off, that changes the nature of how you perceive it. Compare with verbal fencing – some people enjoy trading barbs/quips/trashtalk. And one thing that can happen with a clever insult is that people feel genuinely hurt/ashamed by it, but another thing that happen is that everyone involved is thinking something more like “ah, what clever competitive art we are engaging with.”
I think part of the thing with roughhousing in general is just getting to touch other humans. (and if you’re calibrated to the pain of a punch a la above, then you’re mostly getting a particular flavor of the positive-aspects of that, rather than the negative.)
I’m not just worried about the emotional valence of violent sadist, but that those words, together, convey a clear(ish) connotation that goes beyond “person that enjoys inflicting pain.”
These all seem like things that would strongly favor games like “sometimes roughhouse with your friends with punchback allowed and sensitivity over repeated interactions to how people are feeling about the whole thing” over games like “punchbug”. I’m justified in deprecating hypotheses somewhat that would predict different things than the ones that actually happened.
I’m definitely not claiming to not be missing important stuff. A more clearly grounded query would have been way better, of course, but as it turned out, generating more hypotheses about the appeal of punchbug would have been entirely beside the point, as the problem turned out to be that SilentCal thought I meant something much narrower by “asymmetric,” which excluded formally symmetrical rules with disparate impact.
You’re saying that I generated not only a wrong but an unreasonably wrong hypothesis, implying that it’s on the level of “house elves are stealing our magic.” I’d like you to say exactly what you think that hypothesis was, because it’s entirely unclear to me.
Ah, sorry – I was saying something slightly different:
I’m not sure if I think “people enjoy X because violent sadism” is a reasonable hypothesis or not. I brought up house-elves-steal-magic” to describe, not that hypothesis, but whatever second hypothesis you might generate.
FYI, I don’t think this feature actually works.
I see two PMs of the “Comment deleted…” type in my inbox. When I click on one, here is what I see:
https://dl.dropbox.com/s/0irgz8k7vgnu515/Screenshot%202018-06-02%2016.08.28.png?dl=0
(On GreaterWrong, I see something different; I have no idea whether these things are related.)
Hrmm. Well that definitely shouldn’t happen. Browser and other specs?
Same thing happens with Firefox 60.0.1 (same machine), and with Opera 49.0.2725.64 (also same machine).