On occasion, I’ve found myself wanting to downvote a comment for being unreasonably mean to someone. There really isn’t a category for that – “incorrect” comes closest, but I feel like taxonomically that’s for “the statement is incorrect” rather than “the rhetorical style is incorrect”. So I reached out to jcs, and he asked me to create a meta post, and here we are.
Troll seems like it fits with the idea of a post being inflammatory in tone.
There’s a subtle difference between being mean and being a troll, I’d suggest–I’ve got a little experience with both, many many years ago, in my misspent youth; I’d certainly never engage in either today! ;)
If a comment or post is “trollish”, then its primary purpose is to antagonize or bait other posters. As we all recognize this, it also tends to mean that a post flagged as “trollish” will be dismissed out of hand. We all assume that even if good points are being made that is incidental to the purpose of the post and hence they can be ignored. Typically, ignoring troll posts is the best thing to do, because if the troll isn’t provoking a response they lose interest and go elsewhere, or decide to engage in useful conversation.
If a comment or post is “mean”, though, it may be posted and argued quite honestly and genuinely, but be disagreeable in tone. It can be considered a valid point, it can be considered well-argued, but it may ruffle feathers. The difference between it and a troll comment is that we can expect to engage the author in a reasonable discussion, perhaps even suggesting a different way of wording things.
Again, a subtle difference, but it means all the world for discourse.
I think there’s an enormous difference. Trolling is posting something in bad faith in order to garner responses; it can be inflammatory, for example using racial or sexual slurs to attract condemnation, but it doesn’t have to be. You can troll by posting a deliberately incorrect answer to a technical question in the hopes that people will rush to correct you, for example. (Trolling in this way is often a more successful tactic than simply asking the question, although I don’t do it because I feel it’s unethical.) Meanwhile, you can be inflammatory in tone without trolling — the easiest way is simply to just say inflammatory things you sincerely mean, but even if you’re just looking for an excuse to bully someone, it isn’t trolling if you aren’t trying to get a response out of them. Many bullies, I know from sad experience, are just seeking to silence their victims, not watch their “amusing” reactions.
With that, I will return to my silence here.
Should we just rename Troll to be more generic?
Flame maybe? That’s already mostly what I used it for, though fortunately not too often. “Troll” is a notoriously vague term, but we don’t really have trolling in the classic sense of deliberately bad-faith posting. So in this context I already think of it as meaning something like jerkish/flamey.
I like unproductive.
I’d vote for a more generic “breech of site etiquette / code of conduct”, if it would come with a similarly-named section on the about page :-)
I trust our judgement when moderating more than any code.
The written code isn’t so much for the voter as for the receiver of the votes. It’s particularly useful for new community members so they can figure out what is acceptable behaviour here.
Then you’ll get people complaining that they didn’t violate the code. And like any documentation it will get out of date. Better to encourage people to lurk for a while before posting.
Actually, yes, if that’s on the table I’ve wished for a while that it were something I felt was applicable to posts like that. I know it’s intentionally chosen to not be a catch-all, but I don’t believe I’ve ever used it at all. “Personal attack”?
Big Brother is watching. Always watching.
For my part, I’d like to see a community where we spend a lot more time and energy up-voting things that matter / we agree with, rather than down-voting things that don’t. Whilst I don’t like to see rudeness or ad hominem attacks any more than the next guy/gal, I also don’t like the negativist mindset that constant clobbering / down-voting / de-karma'ing / closing (SO anyone?!) seemingly creates. Let’s focus on the positive and let the negative take care of itself. Then again maybe I’m just showing my age.
Lobsters is an excellent, highly civil community, where I feel the position you’re espousing is practical and a good idea on at least 95% of threads. :) I really like that a downvote is several clicks, here, and requires some thought about what justified it - I vaguely think jcs has talked about that explicitly being intended to discourage downvotes, and if he hasn’t, it has that effect anyway and it’s a good thing.
Just a point of reflection:
“the rhetorical style is incorrect” is something that can be addressed in a reply to the comment, and presumably over time the poster fixes their style. If they continue doing so, they’re a troll, and we’ve already got a flag for that.
I’m a bit skeptical about having a “this is mean” downvote button, because it tends to conflate “this post is incorrect but fine” with “this post is correct but hurts somebody’s feelings” with “this post is crafted malicious garbage”.
I think it might be nice to have a way of flagging mean without causing the actual number of votes to decrease. Thus, we can decouple the idea of “this post is poor in rhetorical style” from “this post is poor in quality”.
As a proof: a mean post by, say, Torvalds or Drepper or de Raadt may be downvoted to oblivion because it’s “mean”, but still have the best technical content in the thread!
There is an unstated assumption here: that it’s more important to preserve 100% of posts with high-quality technical content than to maintain a friendly environment. I disagree with this assumption. I am happy to miss the occasional informative-but-jerky comment if it means that the lobste.rs community remains an inviting place for people to talk in good faith. That seems a very small price to pay to me. I already have a reading list a mile long. I will never be able to read everything interesting that I want to read before I die. So who cares if I miss a lobste.rs comment that might have had some value hidden beneath the bile?
Thank you, and likewise.
Agreed 100%, and this is a point that could stand to be added to every single discussion of discussion in a technical context.
We’ll have to agree to disagree. :)
I tend to be concerned with what I’ve observed to be the endgame of these noble intentions.
I agree – “mean” can be objectively deserved in some cases; I’m thinking more about the case of “unreasonably mean”, or “excessively mean”. All of that’s in the eye of the beholder obviously. What I was thinking about was a kind of safety valve – giving people the ability to downvote a jerkish comment without starting a flamewar and clogging threads with commentary about acceptability of tone every time it happens.
I’m a fan of adding “excessively mean” as a downvote option. It captures the concept of “unsportsmanlike conduct” while being appropriate for a website.
I think the safety valve idea has merit. People WANT to give feedback that they felt someone was mean – but currently don’t have a place to do so. That said, I agree with angersock that it doesn’t feel like something that should go into the score. Leave the scores as is, I honestly WANT to see brilliant mean comments. Mean is sometimes an appropriate and reasonable way to communicate.
It almost feels like this would be a new feature if implemented: tone feedback. Something that the author of the comment (even if highly upvoted) could always see, and it could go well beyond “mean” – because different people dislike different things “arrogant”, “bullying”, “condescending”, “demeaning”, … I can’t think of one with E, so I will stop there. Not sure I want such a feature, but is interesting to think about if it would change poster (and therefore community) behavior at all – if on an upvoted post (positive feedback) they got a bunch of “arrogant” tone feedback, would that possibly cause a behavior change?
I definitely don’t want to see such comments, and I don’t think it’s an appropriate way to communicate, at least on lobste.rs. I’m happy if someone corrects me if I posted something incorrect. Why should they yell at me, call me names, tell me I’m so stupid I should commit suicide, or otherwise try to do their best imitation of a 1980s Usenet discussion? How does that improve anything? I don’t want those kinds of comments here, and I think they should be marked “flame” and the score downranked accordingly.
Do you have examples of comments on lobste.rs that were egregiously mean to the person they were replying to, and yet “brilliant” and a good example of the kinds of comment we want in the community?
While certainly one can write blatantly uncivil comments (“you should commit suicide”), there are a lot of ways of phrasing something civilly and “meanly” that are both entertaining and informative to read…and crucially, which can contain the additional emotional content that would be lost in a more formal expression.
The sort of practical issue that often occurs is that the threshold for what’s mean or not invariably gets stricter and stricter. Humor is usually the first casualty, because most human humor is rooted in the misfortune of others.
This can still be a tolerable state of affairs, but as we loosen our standards and let in non-pure-tech articles (say, news or politics or whatever), the conversations start to become less and less fact-based and more and more rooted in feelings. Eventually, there is neither the technical content or the entertainment value, and the site loses its initial members.
I agree there’s a risk of broadening the focus so much that it becomes “geeks chat about random stuff”, which tends to not be always very high-quality discussion (arguably a popular web forum run by a certain startup incubator fills that role). I don’t see that as that strongly tied to the civility thing, though. In principle, it could be related: non-mainstream community norms can help keep a community from being diluted, by serving as a kind of in-group mechanism. But it’s really tricky to get right, and you can easily drive away a lot of interesting technical people and end up with less good discussions, because they also don’t feel part of the in-group (or just don’t like it). It can become really tedious when you have a number of people who are very abrasive and always flaming, and people can lose patience for it and go elsewhere. The old comp.lang.lisp was in that category for me, bad enough that I think it served as an active negative for the overall CL community.
I have never seen anything approaching what you just mentioned on lobste.rs – but maybe I just missed it or it was already downvoted to oblivion for being troll (telling people to commit suicide, WTF, is this League of Legends?). Are people seriously telling each other to commit suicide on here?
I have seen people be “mean” which is generally saying stuff like “I find you arrogant and condescending” – which is a bit “mean” – but is also completely sincere. There often is a nice way to say something, but sometimes their isn’t, and “mean” is appropriate. When I think of “mean” in terms of lobste.rs – I think of Linus posting biting comments on LKML.
Ah sorry, I’m getting current-situation and hypothetical-norms parts of the discussion kind of muddled in my comment. You’re right, lobste.rs doesn’t have anything like that, which is also why I don’t feel the need to use the “troll” downvote much here. The “commit suicide” example was a kind of oblique reference to some Linus Torvalds mailing-list posts, which some people seem to think is an ok style of discussion, but which I would rather not see here.
If a Linus post got downvoted because of its antisocial style, that would probably be valuable feedback to the author on how to be taken more seriously in the future.
If they’re truly a Linus, we have empirical proof that their methods work and work well enough. Unless you have some genuine concern that Linus Torvalds isn’t being taken seriously?
We also have empirical proof that it drives people away.
Discussing the symptoms is not very interesting, “what would linux be with a more approachable maintainer” is a more interesting one.
It might be more interesting, but it’s also quite hypothetical.
We can ask the question “Can a massively influential and beneficial software project be run by a cranky git and with a culture of meanness and still be successful?”, and see (because history) that the answer is yes.
We can’t ask the stronger question “Is Linux successful because of it’s abrasive leadership?” because of any numerous factors…but we similarly can’t assume that the nicer approach is automatically valid.
We can’t claim that the project is worse off (or better off) for the people it drove away.
Sure, it is. And that makes it the more interesting one.
There’s a an ample amount of projects that live well with very nice leadership, so we do have material to talk about.
Eh, that seems like way too many distinctions to me, especially in a smallish community that I’d like to enjoy participating in. A good post participates in the community constructively, a bad one doesn’t. There are a number of ways of being bad, and being a huge jerk who’s needlessly insulting your fellow community members ranks pretty highly there for me, maybe at the very top.
Do we want to downvote, and therefore push to remove from the conversation, imflammatory content based solely on their rude tone?
If a comment is unreasonably mean, there is a good chance the post was written poorly to begin with, therefore making it elligible for other downvote categories. This is an assumption I’m making based on my time in online communities.
If an author continually makes mean remarks, they’ll gain a reputation for having a poor attitude. If they make a one off mean comment, but with good quality content, it is best to ignore it in the grand scheme of things.
Should consider having a politeness score as well as overall score?
I don’t want each user’s profile to look like an RPG stats listing, but it’s an interesting thought.
I’m old-school, so I don’t believe in feeding the trolls, even if they’re unintentional trolls. Ignoring them is much better. Lavishing attention should be done in positive swaths to those deserving. If a downvote is actually necessary, I’d rather it be for some calculable objective error, rather than my view that their tone doesn’t jive with my personality (or worse, my perception that it might not jive with somebody else’s).
Be the change you want to see in the world. That doesn’t mean stamp out the changes you don’t want to see. All that does is highlight the things you should be drawing eyes away from.
Basically, if you don’t like what someone is saying, give people something better to see. There’s no more sure victory over a troll or a bully than to see their efforts ignored. Negativity just begets more negativity.
I normally do decide not to respond at all to comments I just don’t like because they’re attention-seeking. Personal attacks are different, and need to be handled differently - it’s important for the targets to know they’re not alone and undefended, and it’s important for bystanders to see that those things aren’t allowed to pass.
Even knowing that it’s important to speak up, there’s still a decision that has to be weighed, because the benefit is still balanced by the negative effect of giving the attacker attention. I see downvotes for those situations as falling in-between saying nothing and saying something about how the attack is inappropriate and unwelcome.
I realize that there are people (not on Lobsters, as far as I’ve seen, but in general) who view a negative score as something to aspire to, but the score isn’t for or about them to begin with, it’s to get their remarks moved to the bottom of the page so they don’t waste as much of everyone else’s emotional energy. And it rewards them less than a response in words does, while still reinforcing the community’s standards with reasonable efficacy.
I think the handling of abberations like direct attacks is for moderators to weigh in on. Community moderation through the use of flags (to call mod attention to a thing) is the closest to effective I’ve ever seen, and still results in mods having to sometimes say “no, lynch mob, they’re not actually doing anything wrong.” Downvotes invite more of that abuse in my experience.
If we want to foster a community where such solidarity is a thing, I’m all for it, but the downvote mechanism is a bad choice for it. A mod should negate the posts scorability altogether (pulling it to the bottom instantly), and replace the text with “moderated for X content” in the worst case (depending on well defined rules about censoring). With a note of why (maybe to just the poster, depending on context?). This would also mean no more replying to that subtree.
Culture is hard. Especially as you get larger.
I think downvotes would work fine. They work for everything else.
I’m not against that approach. I’ll let moderators weigh in as to the burden of it.
[Comment removed by author]
Very eloquently put! I’m really in two minds. One one hand I think that one can simply add a reply to churlish comments letting people know the error of their ways, but you lose anonymity that way. This community is gentler than lkml, so telling someone they’re rude here shouldn’t require a Sharpish backbone, but I can still imagine situations where anonymity is desired.
Isn’t that what we have now, but with 5 predefined choices? ;)
[Comment removed by author]
I wrote a longer comment below about how part of the value of the downvote is that it provides an anonymous option which simply registers disapproval. Making it not-anonymous with freeform text is the same as just replying, except for the numerical implication, which to me has always felt redundant when I’m responding in words also. This would move downvotes out of the middle-ground position, and that position is useful; I elaborated in the other comment.
I think there could be a sensible discussion about whether the person whose comment it is should get to see the reasons. I also am not inherently opposed to freeform downvote tags, just not in conjunction with removing anonymity. There’s a UX case to be made against freeform entry, but I doubt that’s going to be relevant. Dwc, do you consider the non-anonymity to be the main point of this proposal? Do you feel the freeform entry serves a purpose even if it’s anonymous?
[Comment removed by author]
Thanks; that makes sense.
Then, I’ve given my position on anonymity. I don’t have a strong opinion on the nature of the tags. Has the tag data been useful for anything in the aggregate, or is its only significant purpose as feedback to the author? Are the current choices all roughly evenly used?
Sorry for the off topic, how downvotes work with tags? Is it for admins to see? I am curious to know what purposes it serves for end users and also for admins.
The per-reason downvotes are shown to moderators (always) and the author of the comment (when the score is less than 0).
I don’t like “breech of site etiquette / code of conduct” as new downvote category. We want people to stick to certain behaviors because they’re reasonable, not because they are ~the rules~. And “nonproductive” implies productivity as criterion for noteworthiness which we would however only apply when disagreeing with someone. It’s pseudo objective.
Trolling is about impression, being mean is about expression. They read differently, are differently long (troll optimizing for maximum impact with few lines), and differently motivated.
I once added a short comment to a story which you could call provocative. It was in the form of “So it’s basically like <oversimplification of proposed technique>?”. Not to provoke people but to provoke answers as to how it differs from previous approaches. I got a “troll” downvote for this but I was just trying to be honest and save people the time of reading a long paragraph. Give me a comment next time.