Wikipedia and bullying
Jun. 6th, 2007 04:34 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
garranSince reading cola's recent post I've been thinking again about how a lot of people become disillusioned with or (as she says) prejudiced against wikipedia. The usual reason, and one of the most compelling reasons, seems to be that wikipedia is an environment where, if you're determined enough, bullying, and being so obnoxious that others give in so that they don't have to deal with you anymore, are pretty effective tactics for getting your way.
So, mostly what I want to talk about (especially since I'm not hugely knowledgable about wikipedia as a specific community; most of my information on the subject comes from watching Rachel) is this tactic, and how, unfortunately, it's actually pretty difficult to set up a discussion environment where it isn't effective. From my days at the Megatokyo Story Discussions forum, my default preference for the government of a message board is something sort of laissez-faire, letting people post as they please and the most interesting, enlightening and on-topic threads naturally grow and come to more general attention (and in the meantime, haiku and spontaneous roleplaying blossom in the crannies). This was a pretty privileged point of view for me to be able to hold, on account of how there were so few trolls around SD, although there were a constant stream of confused new people; but as well as lacking a mechanism for dealing with trolls (aside from the old 'ignore them as best you can' routine), it also lacks one for dealing with other people who don't hold to it themselves -- people who are sometimes-constructive regulars but who want to force everyone else to conform to a standard of discussion, or an agreement on some still more trivial point, that they are in the minority in preferring, and are willing to take disruptive actions such as responding to anything which doesn't meet their criteria with loud attempts to shift the discussion to one of whether such conformity ought to take place.
One way of dealing with both these types of disruptive posters is Hobbes' solution to the state of nature -- you have some general authority figure fear of whom keeps folks generally in line, and if someone is behaving unreasonably enough then you can appeal to them and they'll come and punch the offender in the face or take their vowels away. Teresa Nielsen Hayden has made the argument that you literally "can't have a good online discussion without moderation" -- without, in other words, somebody in power who is willing and able to ruthlessly censure people whose habits of discussion become abusive. It probably isn't what she meant, because wikipedia doesn't to my knowledge have any system of ultimate authority like this even for its founder, but when Rachel in cola's comments remarked parenthetically that she was "glad to hear Jimbo is on the case", I had a brief vision of something similar: a 'Jimbo Wales ex machina', where he would descend on a talk page and sort everybody out with rewards and punishments.
There are reasons that essentially autocratic strategies like that employed by Making Light aren't really feasible for something like wikipedia, though. One is the fact that it's an institution that one might hope and expect to outlast its founders and original leaders; it is of course not certain that one's authority figures in the first place will be fair, decent and wise enough to be an acceptable solution to the problems of people unfairly pushing others around instead of just a massive contribution to it, but even if they are it seems to become less and less likely that the person who replaces them, and the person after that, and so on, will all have as much integrity or ability. The community is in the position of Barrayar, which functions well because Emperor Gregor is canny and benevolent, but which will fall into bloody chaos the next time the bloodline produces a Serg or a Yuri if he can't find a way to build checks into the power of the position. Communities like wikipedia are also much too large for one person to pay attention to and moderate, and so what's necessary is some sort of moderating caste -- so that now you've spread out the opportunities for someone abusive to find themselves in such a role not only over time, but over (notional) space.
So then we are faced with the question of how to deal with the community's systems becoming corrupted -- with, that is, the sort of disruptive poster I described before actually being someone in a position of authority or clout. This happened in the Megatokyan case I was alluding to before: an individual whose name I sha'n't invoke, who was very angry and insistent that the discussions people were having outside of his threads were out of order (and why were we oppressing him so?, and so on), eventually became a moderator and was able to start wiping out those errant threads directly, which he did. (It's worth noting, though it's probably obvious, that I was very much of the opposite point of view to this fellow; I don't think, even given that bias, that my characterizations are too far off the mark, but the possibility should be acknowledged.) This guy wasn't always crazy or hysterical -- he had done some worthwhile things in the community, and I don't think that whomever gave him moderating power was being completely irrational -- but it's nonetheless the case that he used all the power in his possession to implement a policy of which, in the conversations we'd had about it beforehand, he was one of the only proponents in a sea of disagreement. In the wikipedian case that Hannah talks about, the guy is an admin; he didn't use his admin powers specifically to carry out his systematic removal of links to a site whose maintainer he had quarreled with*, but the fact that he had them did mean that he couldn't be dealt with by the standard wikipedia authority system of getting another admin to warn him off. In cases like these, an appeal might be made to some still higher authority, or to someone of similar rank but more widely respected in the community, but this is a stopgap, and it's unreliable; the system of keeping bullies in line through appeal to authority has essentially let us down.
(* I know that regarding edits to wikipedia we're supposed to assume good faith, but this is pretty obviously what was going on.)
So, if we don't have an authority figure, then we're likely to get a Hobbesian state of nature where every conversation is at the mercy of its most unscrupulous participant; if we do have one, the position is open to abuse, and also there's the small matter not previously mentioned that, ideologically speaking, projects like wikipedia are interested in giving as much power to as many people as possible (which I'm in sympathy with, as you might gather from my earlier description of how I prefer my message boards to be run). A sort of mixed system like wikipedia's of many admins chosen by general acclaim seems to have some of the weaknesses of both.
This would be a great place for me to describe some comprehensive solution, but I don't really have one; as I mentioned at the beginning, this is difficult. Mostly I'm just poking at the problem. The nearest thing I have, I guess, is the attempt to make some teleological alteration to the culture of the community; if you can make bullying conversational tactics things that are Not Done, and which a whole lot of people will speak out in shocked disapproval of if they show up. This is something you can actually do formally and overtly in a community the size of Windsor House -- you shut down the school for a while and talk it out and get enough people to agree to consciously change their cultural assumptions, and then when you start up again, suddenly in the general mindset "you don't get caught bringing illegal drugs to school" has been replaced by "you don't bring illegal drugs to school". In cases like that, it's an enormously potent tool. If a group is much larger, it's correspondingly harder, because there's simply no way to change everyone's minds at once; but it's still necessary, if you want to bring about some real and comprehensive change, because simply having a law against something that's generally culturally accepted is not going to stop people doing it -- see prohibition, or marijuana in Vancouver, for example.
The difficulty of trying to do this in this particular case, of course, is that attempts to get other people to do and believe what we want them to are endemic to human conversation, and many of them are constructive and necessary; a comprehensive and compelling theory of which of the others are unacceptable bullying and should be rejected is beyond me at least today. (Edit: But see comment.) There's also the difficulty, once a coherent idea of that sort is arrived at, of arguing it so persuasively that it becomes generally accepted, a meme strong enough to discourage many people internally before they even begin and to get people like the admin bully met with a uniform wall of serious disapproval. (And that not everybody is dissuaded by uniform disapproval is still another problem for the efficacy of this method, though it would at least cut down.) Accomplishing this in the general community of all wikipedia editors seems basically impossible, considering how numerous, diffuse, only sporadically involved, and endlessly multiplying most of that number are (it would be easier on a message board); doing so only among some subset, such as the admins and the people who care enough to be involved in choosing the admins, seems more feasible, but still dauntingly unlikely.
So in fact I don't have an adequate solution; sorry, internet. Now I'm going to go home and eat.
So, mostly what I want to talk about (especially since I'm not hugely knowledgable about wikipedia as a specific community; most of my information on the subject comes from watching Rachel) is this tactic, and how, unfortunately, it's actually pretty difficult to set up a discussion environment where it isn't effective. From my days at the Megatokyo Story Discussions forum, my default preference for the government of a message board is something sort of laissez-faire, letting people post as they please and the most interesting, enlightening and on-topic threads naturally grow and come to more general attention (and in the meantime, haiku and spontaneous roleplaying blossom in the crannies). This was a pretty privileged point of view for me to be able to hold, on account of how there were so few trolls around SD, although there were a constant stream of confused new people; but as well as lacking a mechanism for dealing with trolls (aside from the old 'ignore them as best you can' routine), it also lacks one for dealing with other people who don't hold to it themselves -- people who are sometimes-constructive regulars but who want to force everyone else to conform to a standard of discussion, or an agreement on some still more trivial point, that they are in the minority in preferring, and are willing to take disruptive actions such as responding to anything which doesn't meet their criteria with loud attempts to shift the discussion to one of whether such conformity ought to take place.
One way of dealing with both these types of disruptive posters is Hobbes' solution to the state of nature -- you have some general authority figure fear of whom keeps folks generally in line, and if someone is behaving unreasonably enough then you can appeal to them and they'll come and punch the offender in the face or take their vowels away. Teresa Nielsen Hayden has made the argument that you literally "can't have a good online discussion without moderation" -- without, in other words, somebody in power who is willing and able to ruthlessly censure people whose habits of discussion become abusive. It probably isn't what she meant, because wikipedia doesn't to my knowledge have any system of ultimate authority like this even for its founder, but when Rachel in cola's comments remarked parenthetically that she was "glad to hear Jimbo is on the case", I had a brief vision of something similar: a 'Jimbo Wales ex machina', where he would descend on a talk page and sort everybody out with rewards and punishments.
There are reasons that essentially autocratic strategies like that employed by Making Light aren't really feasible for something like wikipedia, though. One is the fact that it's an institution that one might hope and expect to outlast its founders and original leaders; it is of course not certain that one's authority figures in the first place will be fair, decent and wise enough to be an acceptable solution to the problems of people unfairly pushing others around instead of just a massive contribution to it, but even if they are it seems to become less and less likely that the person who replaces them, and the person after that, and so on, will all have as much integrity or ability. The community is in the position of Barrayar, which functions well because Emperor Gregor is canny and benevolent, but which will fall into bloody chaos the next time the bloodline produces a Serg or a Yuri if he can't find a way to build checks into the power of the position. Communities like wikipedia are also much too large for one person to pay attention to and moderate, and so what's necessary is some sort of moderating caste -- so that now you've spread out the opportunities for someone abusive to find themselves in such a role not only over time, but over (notional) space.
So then we are faced with the question of how to deal with the community's systems becoming corrupted -- with, that is, the sort of disruptive poster I described before actually being someone in a position of authority or clout. This happened in the Megatokyan case I was alluding to before: an individual whose name I sha'n't invoke, who was very angry and insistent that the discussions people were having outside of his threads were out of order (and why were we oppressing him so?, and so on), eventually became a moderator and was able to start wiping out those errant threads directly, which he did. (It's worth noting, though it's probably obvious, that I was very much of the opposite point of view to this fellow; I don't think, even given that bias, that my characterizations are too far off the mark, but the possibility should be acknowledged.) This guy wasn't always crazy or hysterical -- he had done some worthwhile things in the community, and I don't think that whomever gave him moderating power was being completely irrational -- but it's nonetheless the case that he used all the power in his possession to implement a policy of which, in the conversations we'd had about it beforehand, he was one of the only proponents in a sea of disagreement. In the wikipedian case that Hannah talks about, the guy is an admin; he didn't use his admin powers specifically to carry out his systematic removal of links to a site whose maintainer he had quarreled with*, but the fact that he had them did mean that he couldn't be dealt with by the standard wikipedia authority system of getting another admin to warn him off. In cases like these, an appeal might be made to some still higher authority, or to someone of similar rank but more widely respected in the community, but this is a stopgap, and it's unreliable; the system of keeping bullies in line through appeal to authority has essentially let us down.
(* I know that regarding edits to wikipedia we're supposed to assume good faith, but this is pretty obviously what was going on.)
So, if we don't have an authority figure, then we're likely to get a Hobbesian state of nature where every conversation is at the mercy of its most unscrupulous participant; if we do have one, the position is open to abuse, and also there's the small matter not previously mentioned that, ideologically speaking, projects like wikipedia are interested in giving as much power to as many people as possible (which I'm in sympathy with, as you might gather from my earlier description of how I prefer my message boards to be run). A sort of mixed system like wikipedia's of many admins chosen by general acclaim seems to have some of the weaknesses of both.
This would be a great place for me to describe some comprehensive solution, but I don't really have one; as I mentioned at the beginning, this is difficult. Mostly I'm just poking at the problem. The nearest thing I have, I guess, is the attempt to make some teleological alteration to the culture of the community; if you can make bullying conversational tactics things that are Not Done, and which a whole lot of people will speak out in shocked disapproval of if they show up. This is something you can actually do formally and overtly in a community the size of Windsor House -- you shut down the school for a while and talk it out and get enough people to agree to consciously change their cultural assumptions, and then when you start up again, suddenly in the general mindset "you don't get caught bringing illegal drugs to school" has been replaced by "you don't bring illegal drugs to school". In cases like that, it's an enormously potent tool. If a group is much larger, it's correspondingly harder, because there's simply no way to change everyone's minds at once; but it's still necessary, if you want to bring about some real and comprehensive change, because simply having a law against something that's generally culturally accepted is not going to stop people doing it -- see prohibition, or marijuana in Vancouver, for example.
The difficulty of trying to do this in this particular case, of course, is that attempts to get other people to do and believe what we want them to are endemic to human conversation, and many of them are constructive and necessary; a comprehensive and compelling theory of which of the others are unacceptable bullying and should be rejected is beyond me at least today. (Edit: But see comment.) There's also the difficulty, once a coherent idea of that sort is arrived at, of arguing it so persuasively that it becomes generally accepted, a meme strong enough to discourage many people internally before they even begin and to get people like the admin bully met with a uniform wall of serious disapproval. (And that not everybody is dissuaded by uniform disapproval is still another problem for the efficacy of this method, though it would at least cut down.) Accomplishing this in the general community of all wikipedia editors seems basically impossible, considering how numerous, diffuse, only sporadically involved, and endlessly multiplying most of that number are (it would be easier on a message board); doing so only among some subset, such as the admins and the people who care enough to be involved in choosing the admins, seems more feasible, but still dauntingly unlikely.
So in fact I don't have an adequate solution; sorry, internet. Now I'm going to go home and eat.
