Meta: LW Policy: When to prohibit Alice from replying to Bob’s arguments?
In light of recent (and potential) events, I wanted to start a discussion here about a certain method of handling conflicts on this site’s discussion threads, and hopefully form a consensus on when to use the measure described in the title. Even if the discussion has no impact on site policy (“executive veto”), I hope administrators will at least clarify when such a measure will be used, and for what reason.
I also don’t want to taint or “anchor” the discussion by offering hypothetical situations or arguments for one position or another. Rather, I simply want to ask: Under what conditions should a specific poster, “Alice” be prohibited from replying directly to the arguments in a post/comment made by another poster, “Bob”? (Note: this is referring specifically to replies to ideas and arguments Bob has advanced, not general comments about Bob the person, which should probably go under much closer scrutiny because of the risk of incivility.)
Please offer your ideas and thoughts here on when this measure should be used.
I downvoted this post (with the implied “despite prominently opposing Alicorn’s behaviors when they occured” intended to lend at least some degree of emphasis). We didn’t need it. Meta crap is tiresome, doesn’t need a new post ever time and that particular subject had already played itself out in the immediate context of the conflict in question. (I downvote most other meta threads too, so this isn’t specific.)
I also don’t think making this post was the politically optimal move for Silas to make. I tend to downvote people who don’t seem to be acting effectively toward their perceived objectives.
Oh, and what the hell, I may as well give an actual answer while I’m here:
Never. Impement a jolly ignore feature already. It’ll make everyone better off. Until that time use the “ignore feature” that everybody already has built in and just don’t pay any attention or make any response to people you don’t want to listen to.
The above said, I still would mostly avoid replying to Alicorn—or Bob or Alice or whatever the euphemism is supposed to be—if she happened to personally ask me to. Not out of obligation but because I do care to some degree about the preferences of others and also am aware of the cost associated with engaging in public conflict without good reason. (Sometimes I choose to pay that cost, but not always.) I would still reply when there was a position I particularly desired to oppose and when I did so my reply would clearly be addressing the audience and not personally with Ali-Bob-Corn. But if there wasn’t a strong need to comment directly on a particular statement I would make my contributions to the conversation via particularly carefully constructed sibling comments and to whatever extent possible draw the attention of the conversation away from the Ali-Bob-Corn social rival and in a direction more useful for whatever my goals happened to be at the time.
Agreed. Especially when you don’t know the context of the discussion and it gives the feeling of LW being about a small clique of people and their issues, rather than actually talking about interesting things.
In general the site runs really well, all the stuff recently about not feeding trolls or filtering out newer people seems unnecessary and counterproductive.
I would think that being anchored by a strong (possibly uncharacteristic) recent event would make one’s judgment worse, which is (part of!) why I avoided giving that context, and said as much.
I consider meta discussion in general tiresome, and meta discussion that is being used as a code for personal disputes particularly tiresome.
Its not a matter of judgement as a matter of what I am interesting in discussing and what I am in favour of seeing on LessWrong. For example, I could engage in a discussion of literary theory without bias, but don’t particularly want to do so, and don’t think LessWrong would be the place to do so. I would be especially averse to such a discussion if it was being used as some sort of proxy for a dispute between two people. There are other places to discuss literary theory, and other better ways to resolve personal disputes.
If I see non-standard, extreme methods of community forum administration, and I wish to stop it or make it consistent, I have two options:
1) I can discuss the merit of the policy in the abstract, divorced from any particular instance.
2) I can discuss it with specific reference to the most recent events, thus rebooting that discussion and escalating it to a flamewar (or a worse flamewar if it’s already one)
No matter which way I go, you can come up with a reason why I did the stupidest/most inflammatory method. So, your comment doesn’t tell me a lot about what I should do instead—unless your position really is, “That’s a great policy, don’t bother even talking about it.”
Or perhaps that was the game—if I argue the abstract, you accuse me of being passive-aggressive about the particular; and if I argue the particular, you accuse me of rekindling and widening the existing drama. Either way, potentially abusive moderation gets a free pass.
The hard part: tell me what I should have done, that would met with your non-disapproval.
I sympathise with the problem as you state it, but don’t know enough about the particular circumstances to know if that is a fair summary or what you would be best doing about it.
Personally I would have preferred if you had mentioned the context in the original post in something like the format of: This thing happened [link and explanation], are we ok with this form of moderation being the norm on here?
Describing the issue in a very abstracted way gives an impression of subterfuge, and makes people feel excluded from the discussion.
I don’t disagree with any of that; I just don’t know if I’d get as much criticism had I done it that way, or if I’d just be told, “HOW DARE YOU SPREAD THAT CONFLICT TO THE REST OF THE SITE YOU F***ING TERRORIST!”
(My estimate is that you were wise to refrain from providing explicit references.)
Really? Why?
Suppose I have the goal of convincing LW to stop thinking about FAI. So I write a brilliant article on the dangers of uFAI. The utter disconnect between the goal and the article means you downvote?
That’s a different position that downvoting because you don’t approve the political move.
Oh, wow. I’d totally forgotten that whole thing.
Oops. I feel bad about fanning the flames, now.
Fanning flames is generally a bad policy for paper machines...
I’m not that kind of paper machine. What do you think I am, Clippy’s older brother? Jeesh.
There’s not (last I checked) a community consensus on the issue, and I’d rather isolate the meta-discussion to its own thread, rather than splattered all over anywhere Alicorn posts a comment.
An ignore feature would indeed be a significant improvement. I’m not convinced that it’s strictly necessary or sufficient, but I do think that it would be better to do than not.
Never. Either warn Alice, ban Alice, or leave it alone. Attempting to give Alice an Internet restraining order is only going to cause her to find more roundabout ways of expressing her views (e.g. dragging the conflict into other threads.) Just keep it simple.
Even more simple: If Bob doesn’t like Alice’s reply, then Bob should downvote Alice’s reply.
Schelling was in favor of a graduated scale of punishments in The Strategy of Conflict; either by smaller degrees of punishment, or a smaller possibility of a large punishment.
That seems to be most lacking part of his analysis of bargaining tactics, historically speaking. Many historians think that strategy made the Vietnam War worse than it had to be.
But is there any formal principle behind purposefully allowing yourself to carry out only a single type of punishment?
I saw something for the first time today. I replied to a comment that had been down-voted, and the site asked me,
So, if one person dislikes a comment, it shouldn’t be responded to? I disagree strongly. This makes the site enforce a tyranny of the majority. It punishes resistance to groupthink.
I don’t think Alice should be prohibited from responding to Bob, ever. If two users create drama with back-and-forth responses, they have both chosen to do so.
I think it requires at least 3 downvotes for the penalty to apply.
There have already been several threads related to this change. Opinion is divided; nearly everyone opposes it, but Eliezer supports it.
I’m not sure which way this bears on the “Lw is just a cult of personality around Eliezer” hypothesis. On the one hand, lots of people opposed Eliezer on something. On the other, we let him get away with this shit and don’t just leave.
Perhaps LW was originally a personality cult around Eliezer and now it just has Eliezer around as historic legacy that is too hard to get rid of for logistical reasons (like namespace ownership). Kind of like the UK still has a Queen as a head of state.
I think it says more about Trike Apps and SI kowtowing to Eliezer than anything.
On reflection, this has made me less inclined to donate to SI than I was before. (I had already been thinking recently that FHI was probably a better organization to donate to for a variety of reasons, ever since I discovered that donations from US citizens to FHI can be tax deductible.)
How would “just leaving” solve anything?
Well, there’s the obvious answer: this forum exists because people participate in it. If people didn’t participate in it, it would cease to exist. If I value its continued existence (either for its own sake, or for the sake of greater rationality in the world, or for the sake of greater fundraising opportunities, or for whatever reason), then a credible threat of non-participation by a significant fraction of active users threatens something I value, and I might therefore be motivated to change my behavior due to such a threat. One way to make such threats more credible is by demonstrating that users will in fact stop participating over behaviors in a specific class.
In addition to what TheOtherDave said, forums are cheap, and we could just as well be on a site without a dictator if we felt like it.
I missed some of the earlier threads and didn’t want to reignite them. I feel more comfortable replying to PhilGoetz’s comment since it’s only from two days ago.
One problem that I didn’t see anyone discuss is that this feature is likely to drive away new users. This policy discourages interaction with new users because unpopular comments overlap significantly with comments from new users. By discouraging commenters from responding to the low quality posts of new users, we disincentivize the picking of low hanging fruit, which is the opposite of what we should be doing. In addition, by doling out karma penalties at a set level rather than as a fraction of total accumulated karma, new users face much heavier fees than regular users, which will also result in increased insularity.
You’ve exaggerated in a few places, as follows.
It’s mild evidence for the statement that it shouldn’t be responded to.
It’s not clear that there is a “tyranny” with typical connotations, so the word shouldn’t be used without clarification.
Not “ever”. The choices that affect many other people negatively should be discouraged, or their effect neutralized in some way, if possible.
Hm, how about “philosopher-king of the majority”?
None, but a “block list” or similar filter should be implemented, allowing posters to screen out all posts/comments from specific other posters. Further, posters engaging in deliberate harassment should have direct action taken against them.
To clarify, would you count a reply like, “You make a valid point, but possibly overstate the strength of your evidence—the examples given are extremely exceptional, and far from the general case we should be preparing for” to be “harassment”?
No. I would count replies like “You suck, stop posting” as harassment.
I’d rather there wasn’t any official policy on this, and it was just solved informally among the concerned parties.
Either Alice can abstain from replying to Bob (answering in sister comments instead, if she really needs to), or Alice can reply to Bob at the risk of looking like a douche. It doesn’t seem like a huge sacrifice for either party, and not one worth escalating over.
A more general policy that would be worth having is that if some users start creating too much drama and bickering between themselves, the moderators should feel free to start nuking posts.
When a human moderator makes a judgment call.
What if Bob is a human moderator?
For obvious reasons, participants in a personal feud should not have moderator powers.
I pretended not to, until I convinced another moderator to approve a certain algorithm. (Surely bothering Eliezer on a per-case basis would not be the best choice, if there is such an algorithm and he in fact approves it.)
Yes: generally, one important rationalist skill is recognizing when your bias is rendering your judgment unusable while there are others you can turn to—for must the same reason that a rationalist should be able to understand “Do not murder for the good of the tribe, even for the good of the tribe” without their head exploding.
(For a while—though I seem to have been near alone in this—I avoided voting on subthreads in which I was an active, arguing participant, recognizing that this would compromise my judgment.)
I still feel a bit uncomfortable if it’s Bob’s friend who makes the judgement call. Wouldn’t a “some stated policy + judgement calls per basis” system be preferable to a “judgement calls per basis” system in terms of preventing abuse?
What if someone gets into a feud with Eliezer? Does he have to step down? This doesn’t seem like a very practical rule.
I took it to mean, “should not have moderator powers with respect to their feud”. If Eliezer and some other guy are fighting, Eliezer’s not allowed to ban that guy not matter how badly he behaves—he has to prevail on e.g. lukeprog to do it.
Additionally, he should not pick moderators over which he has unusually high influence—for an extreme example, if he’s the dominatrix (atror?) for that moderator. (I don’t know if that applies to lukeprog, but I’m pretty sure it doesn’t...)
I will note that the rule I intended was not “no Less Wrong moderator shall ever be involved in a conflict with another human being”. Crucial details of the present situation are that (1) SilasBarta is himself a prominent and respectable member of the LW community, with exactly 7700 karma at the time of writing; and (2) Alicorn’s issue with him originated on LW and predates Alicorn’s being made a moderator.
So we’re not talking about a situation where e.g. some troll with a longstanding feud with EY from outside LW comes onto the site to make trouble.
Per fezzwig, it just means, “Hey, other moderator, I’m not close enough to impartial here, so look at this thread and make the judgment on it.”
That would mean that if moderator A nobly “avoided” getting another moderator, but then picked a moderator B, over which A has unusual influence, and did it precisely “when it actually mattered [to A]”, that would also not be kosher.
Bob asks Casey (another moderator) to make a judgment call.
… directed by M. Night Shyamalan.
Define human, moderator, judgement call, makes, and “when”.
The parent is bad, but someone banned it (I unbanned it for now), and I’m not aware of a policy that permits banning in such cases. Please clarify.
Incorrect is a suspected Will Newsome sockpuppet and I’ve been told to—er—fire at will.
It was supposed to be a sarcastic response about being too strict with definitions but obviously didn’t end up being funny.
I am not a Will Newsome sockpuppet. I’ll refrain from making the lower quality subset of my comments henceforth.
I agree about Will Newsome, who recently fell under the “Put on probation if reaches minus hundreds of 30-day Karma” (his current 30-day Karma is minus 334), but “suspected Will Newsome sockpuppet” doesn’t seem reliable enough to me. (I wasn’t suspecting that, for example, did I miss something obvious?)
Began suspecting in this thread. Agree/disagree?
Plausible, not conclusive. It’s an old enough user account to be more conservative on the odd chance that there is an alternative explanation. (Unbanned Incorrect in that thread so that others can judge, it’s not that toxic on its own. As a result, there are currently no (recent) comments by Incorrect that are banned.)
I admire your restrain. I’m also really glad to see Alicorn taking the hands on approach to Will’s disruptions. As a curious observer, and for whatever it is worth I’m approximately neutral with respect to the “Ban Incorrect” decision. I naturally agree that banning on suspicion is a potentially abhorrent heuristic. Justice systems are so much better when “beyond reasonable doubt” is applied. However in this case the mistake isn’t banning someone who is innocent. It is banning someone who is acting like a willfully obnoxious fool who delights in baiting people for responses by saying stupid things in a way that looks like it could be Will Newsome. If the last part happens to be false then hey, at least Alicorn still got rid of comments by someone who was deliberately being an asshat.
Even when it results in false positives (people behaving like Will at his worst are banned as Will sockpuppets) or false negatives (Will acts like someone who isn’t being a dick on one of his sockpuppets so isn’t banned) the policy still results in Mission. Fucking. Acomplished.
Purely out of interest Alicorn, do you just ban individual comments by the supsected sockpuppets when they also happen to be silly comments? Or do you just ban all of them? The former option seems like more work but more power to you if you are willing to put in the effort. The latter option would be ok, but then it seems like you just need a ‘ban account’ feature instead!
The former. We don’t have a “ban account” feature, and I read-or-at-least-skim all comments on the entire site anyway.
Well, looking at the banned comments from Will’s user page it certainly seems that you’re not limiting yourself to silly comments.
Yes, I actually noticed that my favourite comment on this entire thread seems to have been banned (for being made by Will). This is disappointing but something I can accept if it is part of the price of also getting rid of the actually disruptive anti-social contributions.
I didn’t personally ban them all.
How? Through the RSS feed?
Through the Recent Comments pages.
The ‘Recent Comments’ text is a link. Wow. Never noticed that. Thanks.
I had to go back and reread this to really understand the implications. Suddenly it’s much clearer to me why you and other moderators care so much about keeping Recent Comments clean and want features like the troll feeding fee. Low quality comments, no matter in what thread, impact your experience much more than mine.
(Not to imply that reading many threads, or Recent Comments, is something only done by moderators.)
Account age isn’t a guarantee—I know Will has older accounts than his current main, although I haven’t seen them in circulation recently—but I’ll abide by your conservatism here.
I’m not saying that the age of an account (strongly) argues against the account being Newsome’s, I’m saying it (significantly) increases the disutility of wrongful banning. Which in turn argues for requiring a higher level of certainty for banning an older account. Hence, the distinction between “plausible” and “conclusive” becomes relevant, where it wouldn’t be so for a new account that only had a few very-bad-and-highly-suspicious comments.
Love it.
What is the current LW moderation policy? I did a search but couldn’t find it. (I have the power to ban individual comments and posts but have never used it except on obvious spam.)
The policy that I follow and that others don’t seem to be violating is that apart from spam and very rare unusual cases (like retroactive edits by sockpuppets), the only comments that can be banned are bad comments by users who managed to accumulate minus hundreds of 30-day Karma (or applied a lot of concentrated and highly downvoted effort more quickly). Bad comments are not banned if they are rare or in form of occasional bursts, for example. If I’m not forgetting someone, of the active users, currently only Will_Newsome and sam0345 have this distinction. (Edit: and now also eridu.)
At least as of today this also applies to Eridu.
Right, didn’t see that yet. Minus 169 30-day Karma at this moment.
Basing any action on 30-day karma seems unfair in this case. There was a relatively short window during which the comments were being downvoted, and now they’re banned; typically for a controversial topic, those would have been reversed by upvotes over the next week. I don’t know if that would have happened in this case, but now we’ll never know.
Systematically downvoted comments indicate some sort of failure mode, even when it’s not the failure mode of being wrong, or when it’s a failure mode on the part of downvoters. It’s usually possible to reframe the debate in a more constructive manner, long before you get remotely close to start getting banned. I’m not aware of any cases where persisting in highly downvoted behavior on LW achieved anything, apart from some unpleasantness (more unpleasantness the more the behavior persists).
I might agree with you about this in most cases, but that does not really ameliorate the problem of unfairness I referenced. If those comments were not banned and you showed up tomorrow, for example, eridu might not have made the list.
Perhaps notably, if I were aware that Eridu’s downvotes would result in banning of his comments, then I might have correctively upvoted his comments so that they would not be banned, since I do believe there were some valuable pieces in that discussion. Though I don’t know if we want to encourage my sort of behavior.
There is none. Moderation power is exercised arbitrarily at the whims of the enforcers.
(Arbitrarily rather than at random, to be precise.)
Thanks, fixed
What is the distinction that you wish to draw?
Not sure why the parent was downvoted.
“Arbitrary” and “random” tend to be used in different senses. “Random” connotes unpredictability, while “arbitrary” connotes subjectivity to individual judgement.
I was not intending to claim that the moderators’ actions are based on rolling dice, for example.
(“Meaning” and “sense” have very similar meaning and sense to me.)
And so we see that I am a terrible communicator. I would blame the study of philosophy for introducing me to wonderful distinctions that no one else uses, though I’m sure a student of philosophy would tell me that “sense” and “reference” are the relevant subcategories of “meaning”, and both of what I referred to above fall under “sense”. Is Frege or C.S.Peirce in the house?
But seriously, fixed (I hope).
This is false, see above (or refer to counterexamples).
(Edit: I shouldn’t have made this comment, it doesn’t usefully move the discussion.)
No, neither of those is false. There is no stated LW moderation policy, which as far as I’m concerned is equivalent to having no moderation policy. And given the lack of policy, moderation power is necessarily exercised arbitrarily. This does not imply that your judgement is bad, nor does it imply that other moderators’ judgement is good.
If there is some official LW moderation or comment policy, I’d appreciate being pointed to it. But again, I’ve been active here since the beginning and I’m not aware of one, so it might as well not exist.
As I understand it, the word “arbitrary” refers to lack of relevant or systematic explanation or reason for something. I’m not sure what meaning you intend, the word is confusing the way you use it in this context. (Suppose hypothetically that the policy I stated above was more prominently stated previously.)
Ah, glancing at a dictionary, I had intended “Based on or subject to individual judgement or preference” with a splash of “despotic”.
As I understand it, moderators are expected to use their judgement, and do not have any firm guidelines on where to apply it. Alicorn recently commented that the best she got for guidance was “ban shoe ads” (quoted from memory).
Yes, if the putative policy you stated above was more prominently stated before, that would help. Notably, if it were on record somewhere, endorsed by those who run the site, in a place where we could cite it and more importantly argue that it’s inappropriate and should be changed.
Convince me this isn’t an attempt to obfuscate the discussion with unnecessarily legalistic definitions. Otherwise I’m chalking it up to logical rudeness.
I’m guessing the “human” bit was to make sure Clippy (sorry, “User:”Clippy) doesn’t get control of the process :-P
I saw a BAN link on a list of comments, wasn’t even sure if it was attached to the draft of a post above or the comment below. I hovered over it hoping for a hint as to what it does. Then I did something that often helps you figure out what something does: I clicked it.
It changed from a link saying BAN to a non-link saying BANNED. I still have no idea what it does.
1) What does it do? 2) How in general should I get answers to questions about what various features of the site do?