If it’s worth applying moderation action and developing novel moderation technology to (among other things, sure) prevent one user from potentially sometimes misleading users into falsely believing X, then it must surely be worthwhile to simply outright tell users ¬X?
Adding a UI element, visible to every user, on every new comment they write, on every post they will ever interface with, because one specific user tends to have a confusing communication style seems unlikely to be the right choice. You are a UI designer and you are well-aware of the limits of UI complexity, so I am pretty surprised you are suggesting this as a real solution.
But even assuming we did add such a message, there are many other problems:
Posting such a message would communicate a level of importance of this specific norm, which does not actually come up very frequently in conversations that don’t involve you and a small number of other users, that is not commensurate with its actual importance. We have the standard frontpage commenting guidelines, and they cover what I consider the actually most important things to communicate, and they are approximately the maximum length I expect new users to read. Adding this warning would have to displace one of the existing guidelines, which seems very unlikely to be worth it.
Banner blindness is real, and if you put the same block of text anywhere, people will quickly learn to ignore them. This has already happened with the existing moderation guidelines and frontpage guidelines.
If you have a sign in a space that says “don’t scream at people” but then lots of people do actually scream at you in that room, this doesn’t actually really help very much, and more likely just reduces trust in your ability to set any kind of norm in your space. I’ve really done a lot of user interviews and talked to lots of authors about this pattern, and interfacing with you and a few other users definitely gets confidently interpreted as making a claim that authors and other commenters have an obligation to respond or otherwise face humiliation in front of the LessWrong audience. The correct response by users to your comments, in the presence of the box with the guideline, would be “There is a very prominent rule that says I am not obligated to respond, so why aren’t you deleting or moderating the people who sure seem to be creating a strong obligation for me to respond?”, which then would just bring us back to square one.
My guess is you will respond to this with some statement of the form “but I have said many times that I do not think the norms are such that you have an obligation to respond”, but man, subtext and text do just differ frequently in communication, and the subtext of your comments does really just tend to communicate the opposite. A way out of this situation might be that you just include a disclaimer in the first comment on every post, but I can also imagine that not working for a bunch of messy reasons.
I can also imagine you responding to this with “but I can’t possible create an obligation to respond, the only people who can do that are the moderators”, which seems to be a stance implied by some other comments you wrote recently. This stance seems to me to fail to model how actual social obligations develop and how people build knowledge about social norms in a space. The moderators only set a small fraction of the norms and culture of the site, and of course individual users can create an obligation for someone to respond.
I am not super interested in going into depth here, but felt somewhat obligated to reply since your suggested had some number of upvotes.
First, concerning the first half of your comment (re: importance of this information, best way of communicating it):
I mean, look, either this is an important thing for users to know or it isn’t. If it’s important for users to know, then it just seems bizarre to go about ensuring that they know it in this extremely reactive way, where you make no real attempt to communicate it, but then when a single user very occasionally says something that sometimes gets interpreted by some people as implying the opposite of the thing, you ban that user. You’re saying “Said, stop telling people X!” And quite aside from “But I haven’t actually done that”, my response, simply from a UX design perspective, is “Sure, but have you actually tried just telling people ¬X?”
Have you checked that users understand that they don’t have an obligation to respond to comments?
If they don’t, then it sure seems like some effort should be spent on conveying this. Right? (If not, then what’s the point of all of this?)
Second, concerning the second half of your comment:
Frankly, this whole perspective you describe just seems bizarre.
Of course I can’t possibly create a formal obligation to respond to comments. Of course only the moderators can do that. I can’t even create social norms that responses are expected, if the moderators don’t support me in this (and especially if they actively oppose me). I’ve never said that such a formal obligation or social norm exists; and if I ever did say that, all it would take is a moderator posting a comment saying “no, actually” to unambiguously controvert the claim.
But on the other hand, I can’t create an epistemic obligation to respond, either—because it already either exists or already doesn’t exist, regardless of what I think or do.
So, you say:
I’ve really done a lot of user interviews and talked to lots of authors about this pattern, and interfacing with you and a few other users definitely gets confidently interpreted as making a claim that authors and other commenters have an obligation to respond or otherwise face humiliation in front of the LessWrong audience.
If someone writes a post and someone else (regardless of who it is!) writes a comment that says “what are some examples?”, then whether the post author “faces humiliation” (hardly the wording I’d choose, but let’s go with it) in front of the Less Wrong audience if they don’t respond is… not something that I can meaningfully affect. That judgment is in the minds of the aforesaid audience. I can’t make people judge thus, nor can I stop them from doing so. To ascribe this effect to me, or to any specific commenter, seems like willful denial of reality.
The correct response by users to your comments, in the presence of the box with the guideline, would be “There is a very prominent rule that says I am not obligated to respond, so why aren’t you deleting or moderating the people who sure seem to be creating a strong obligation for me to respond?”, which then would just bring us back to square one.
This would be a highly unreasonable response. And the correct counter-response by moderators, to such a question, would be:
“Because users can’t ‘create a strong obligation for you to respond’. We’ve made it clear that you have no such obligation. (And the commenters certainly aren’t claiming otherwise, as you can see.) It would be utterly absurd for us to moderate or delete these comments, just because you don’t want to respond to them. If you feel that you must respond, respond; if you don’t want to, don’t. You’re an adult and this is your decision to make.”
(You might also add that the downvote button exists for a reason. You might point out, additionally, that low-karma comments are hidden by default. And if the comments in question are actually highly upvoted, well, that suggests something, doesn’t it?)
(I am not planning to engage further at this point.
My guess is you can figure out what I mean by various things I have said by asking other LessWrong users, since I don’t think I am saying particularly complicated things, and I think I’ve communicated enough of my generators so that most people reading this can understand what the rules are that we are setting without having to be worried that they will somehow accidentally violate them.
My guess is we also both agree that it is not necessary for moderators and users to come to consensus in cases like this. The moderation call is made, it might or might not improve things, and you are either capable of understanding what we are aiming for, or we’ll continue to take some moderator actions until things look better by our models. I think we’ve both gone far beyond our duty of effort to explain where we are coming from and what our models are.)
In the first part of the grandparent comment, I asked a couple of questions. I can’t possibly “figure out what you mean” in those cases, since they were questions about what you’ve done or haven’t done, and about what you think of something I asked.
In the second part of the grandparent comment, I gave arguments for why some things you said seem wrong or incoherent. There, too, “figuring out what you mean” seems like an inapplicable concept.
I think we’ve both gone far beyond our duty of effort to explain where we are coming from and what our models are.
You and the other moderators have certainly written many words. But only the last few comments on this topic have contained even an attempted explanation of what problem you’re trying to solve (this “enforcement of norms” thing), and there, you’ve not only not “gone far beyond your duty” to explain—you’ve explicitly disclaimed any attempt at explanation. You’ve outright said that you won’t explain and won’t try!
Adding a UI element, visible to every user, on every new comment they write, on every post they will ever interface with, because one specific user tends to have a confusing communication style seems unlikely to be the right choice. You are a UI designer and you are well-aware of the limits of UI complexity, so I am pretty surprised you are suggesting this as a real solution.
But even assuming we did add such a message, there are many other problems:
Posting such a message would communicate a level of importance of this specific norm, which does not actually come up very frequently in conversations that don’t involve you and a small number of other users, that is not commensurate with its actual importance. We have the standard frontpage commenting guidelines, and they cover what I consider the actually most important things to communicate, and they are approximately the maximum length I expect new users to read. Adding this warning would have to displace one of the existing guidelines, which seems very unlikely to be worth it.
Banner blindness is real, and if you put the same block of text anywhere, people will quickly learn to ignore them. This has already happened with the existing moderation guidelines and frontpage guidelines.
If you have a sign in a space that says “don’t scream at people” but then lots of people do actually scream at you in that room, this doesn’t actually really help very much, and more likely just reduces trust in your ability to set any kind of norm in your space. I’ve really done a lot of user interviews and talked to lots of authors about this pattern, and interfacing with you and a few other users definitely gets confidently interpreted as making a claim that authors and other commenters have an obligation to respond or otherwise face humiliation in front of the LessWrong audience. The correct response by users to your comments, in the presence of the box with the guideline, would be “There is a very prominent rule that says I am not obligated to respond, so why aren’t you deleting or moderating the people who sure seem to be creating a strong obligation for me to respond?”, which then would just bring us back to square one.
My guess is you will respond to this with some statement of the form “but I have said many times that I do not think the norms are such that you have an obligation to respond”, but man, subtext and text do just differ frequently in communication, and the subtext of your comments does really just tend to communicate the opposite. A way out of this situation might be that you just include a disclaimer in the first comment on every post, but I can also imagine that not working for a bunch of messy reasons.
I can also imagine you responding to this with “but I can’t possible create an obligation to respond, the only people who can do that are the moderators”, which seems to be a stance implied by some other comments you wrote recently. This stance seems to me to fail to model how actual social obligations develop and how people build knowledge about social norms in a space. The moderators only set a small fraction of the norms and culture of the site, and of course individual users can create an obligation for someone to respond.
I am not super interested in going into depth here, but felt somewhat obligated to reply since your suggested had some number of upvotes.
First, concerning the first half of your comment (re: importance of this information, best way of communicating it):
I mean, look, either this is an important thing for users to know or it isn’t. If it’s important for users to know, then it just seems bizarre to go about ensuring that they know it in this extremely reactive way, where you make no real attempt to communicate it, but then when a single user very occasionally says something that sometimes gets interpreted by some people as implying the opposite of the thing, you ban that user. You’re saying “Said, stop telling people X!” And quite aside from “But I haven’t actually done that”, my response, simply from a UX design perspective, is “Sure, but have you actually tried just telling people ¬X?”
Have you checked that users understand that they don’t have an obligation to respond to comments?
If they don’t, then it sure seems like some effort should be spent on conveying this. Right? (If not, then what’s the point of all of this?)
Second, concerning the second half of your comment:
Frankly, this whole perspective you describe just seems bizarre.
Of course I can’t possibly create a formal obligation to respond to comments. Of course only the moderators can do that. I can’t even create social norms that responses are expected, if the moderators don’t support me in this (and especially if they actively oppose me). I’ve never said that such a formal obligation or social norm exists; and if I ever did say that, all it would take is a moderator posting a comment saying “no, actually” to unambiguously controvert the claim.
But on the other hand, I can’t create an epistemic obligation to respond, either—because it already either exists or already doesn’t exist, regardless of what I think or do.
So, you say:
If someone writes a post and someone else (regardless of who it is!) writes a comment that says “what are some examples?”, then whether the post author “faces humiliation” (hardly the wording I’d choose, but let’s go with it) in front of the Less Wrong audience if they don’t respond is… not something that I can meaningfully affect. That judgment is in the minds of the aforesaid audience. I can’t make people judge thus, nor can I stop them from doing so. To ascribe this effect to me, or to any specific commenter, seems like willful denial of reality.
This would be a highly unreasonable response. And the correct counter-response by moderators, to such a question, would be:
“Because users can’t ‘create a strong obligation for you to respond’. We’ve made it clear that you have no such obligation. (And the commenters certainly aren’t claiming otherwise, as you can see.) It would be utterly absurd for us to moderate or delete these comments, just because you don’t want to respond to them. If you feel that you must respond, respond; if you don’t want to, don’t. You’re an adult and this is your decision to make.”
(You might also add that the downvote button exists for a reason. You might point out, additionally, that low-karma comments are hidden by default. And if the comments in question are actually highly upvoted, well, that suggests something, doesn’t it?)
(I am not planning to engage further at this point.
My guess is you can figure out what I mean by various things I have said by asking other LessWrong users, since I don’t think I am saying particularly complicated things, and I think I’ve communicated enough of my generators so that most people reading this can understand what the rules are that we are setting without having to be worried that they will somehow accidentally violate them.
My guess is we also both agree that it is not necessary for moderators and users to come to consensus in cases like this. The moderation call is made, it might or might not improve things, and you are either capable of understanding what we are aiming for, or we’ll continue to take some moderator actions until things look better by our models. I think we’ve both gone far beyond our duty of effort to explain where we are coming from and what our models are.)
This seems like an odd response.
In the first part of the grandparent comment, I asked a couple of questions. I can’t possibly “figure out what you mean” in those cases, since they were questions about what you’ve done or haven’t done, and about what you think of something I asked.
In the second part of the grandparent comment, I gave arguments for why some things you said seem wrong or incoherent. There, too, “figuring out what you mean” seems like an inapplicable concept.
You and the other moderators have certainly written many words. But only the last few comments on this topic have contained even an attempted explanation of what problem you’re trying to solve (this “enforcement of norms” thing), and there, you’ve not only not “gone far beyond your duty” to explain—you’ve explicitly disclaimed any attempt at explanation. You’ve outright said that you won’t explain and won’t try!
It’s important for users to know when it comes up. It doesn’t come up much except with you.