As usual, I recommend not reading this if allergic to Abd Thought. Some of you are. Consult your physician.

At 08:37 AM 6/3/2010, Daniel R. Tobias wrote:

On Tue, 01 Jun 2010 10:18:03 -0400, Abd wrote:

...

Durova's history is a classic example. She was hounded by a screaming mob when she made a mistake, even though she recognized the error and undid it within an hour.

I might well be counted as part of that "screaming mob" since I was one of the critics at the time, but my intended target was never Durova personally (who is now a Facebook friend of mine), but the entire system and its associated mindsets, in which a group of "insiders", with closed mailing lists of their own, takes on a "circling the wagons" mentality against "trolls and harassers", leading to snap judgments that can get people blocked or banned for saying politically incorrect things.

Yes. Durova became a stand-in, a poster girl, for that situation. But she was not the typical stubborn and abusive admin. She was not a knee-jerk "our faction right or wrong" enforcer. She made a mistake, admitted it immediately, and took responsibility for it. She has been very helpful in confronting admin abuse.

JzG, on the other hand, never admitted error, disappeared when it got hot at RfAr/Abd and JzG, resigned his bit a few months later, and, as I recall, complained that I was the reason -- even though I was site-banned at the time and all I'd done was to point out his use of tools while involved with Cold fusion -- and then, later, asked for the bit back, and since he'd only been admonished and not actually desysopped, it was routine.

And then went after his old nemesis, Pcarbonn, who had quietly returned to editing by making suggestions on the Cold fusion Talk page. JzG claimed that Pcarbonn was pushing the same POV that had gotten him banned, and the cabal jumped in to chant "yeah!" Of course, Pcarbonn had not been banned for his POV, he'd been banned for allegedly treating Wikipedia like a battleground, and JzG had successfully framed the issues that way a year before. In fact ....

Pcarbonn's ban was renewed, and GoRight got slapped for pointing out that there wasn't any evidence of misbehavior, I was blocked for allegedly violating my MYOB ban because I voted in a related poll (I was allowed to vote in polls!) and then commenting on the situation (it had become about me!) on Talk:GoRight; I was blocked for disagreeing with the administrator who blocked me. If I'd cared enough, that would have gone before ArbComm. Might still, I suppose, but ... I do have a real life.

And so it goes, on and on. I really don't care any more, I just have a habit of saying what I've seen, from time to time. My story is far from unique, it has been repeated over and over, and until it's realized that the lack of sane decision-making structure that would restrain the nutty unpredictability of how Wikipedia operates is the core problem, and it's addressed, Wikipedia will continue to foul its nest, building up reservoirs of people who have been burned.

Hipocrite, who stirred up the shit that led to RfAr/Abd-William M. Connolley, practically wears a blinking neon sign, "I'm a troll." He's been completely outrageous. But he's not the one who usually gets blocked. It's his targets, and how in the world does this happen? It's easy. There are admins who don't like his targets. He gives them cover to act.

In a sane structure, this kind of behavior would be spotted and interdicted quickly.

I had some comments on that situation in my essay I wrote as a rebuttal to one of JzG's essays:

http://dan.tobias.name/controversies/cyber/wiki2.html

Good essay, Dan. In my view, part of the solution to the Wikipedia problem would be off-site structure, for if Wikipedia is to fulfill its mission, it must have the faculty of *independent* judgment. One of the classic ways that organizations, once an oligarchy develops, suppress this, is through central control of communication. It's an error to pin this problem on the "bad guys." Rather, it's a phenomenon that naturally develops as part of the Iron Law of Oligarchy. The solution is to decentralize communication, such that, while there remains central communication, it cannot be dominant unless it sits reasonably with the consensus of what is decentralized. So off-site structure, because it cannot be controlled, bypasses the central. Only if a significant number of central participants, though, connect with the off-site structures -- and it's obvious that there must be many of them, not just one! -- can this become an effective restraint. Wikipedia Review, however, is already functioning as a bit of an ombudsman. When really outrageous behavior is noticed there, there is a tendency for some correction to happen.

But it's not reliable enough. I see the mailing list as the device that, being push, is most likely to be functional. And, yes, the cabal used and uses mailing lists. That wasn't the problem, as I think you realized, Dan. The problem was lack of balance.

The hysteria around that "secret" mailing list became, more or less, the official position of ArbComm with the EEML case. There, the very existence of an off-wiki list, private, with some apparent bias and some coordination of on-wiki activity (Canvassing! An article was semiprotected by an admin when notified by the list of IP revert warring! Some people commented in AfDs!) led ArbComm to completely set aside privacy concerns, using list archive material apparently obtained and published illegally, to throw the book at list participants, sanctioning far in excess of what could be justified by on-wiki behavior. And that should have been the standard, not the existence or non-existence of "collusion," which, after all, is just another name for cooperation. If it's okay for the "administrative cabal," as Jimbo called it, to cooperate, why not any other faction?

I joined the EEML when I saw the flap; I was under my site ban at the time, and, given my theories about the Wikipedia solution, I wanted to see, and, as well, to encourage them to work effectively and *properly*. They welcomed me, and they also started inviting others to join. The problem had been the secrecy, not what they actually did, for the most part. One editor had posted his password and invited others to use it while he was away to, say, effectively sock. There is no sign that any of the list members actually did this, but there the offer was, in the list archive. He was properly trout-slapped for that. My own suspicion was that this was trolling for others to violate policy, and that he may have been the one to leak the archive, but list members seemed to think that he'd merely been foolish. Arbitrators, on the other hand, clearly were attempting to sanction off-wiki communication with exemplary punishment, as a message to others to not do the same thing, and that was explicit in the deliberations.

Of course, nothing was accomplished but to drive away a loyal WP administrator, Piotrus, an academic, a sociologist, with peer-reviewed publications about Wikipedia, quite naive, actually. He believed in the system, which slapped him down when that was found convenient. And the editors with the most knowledge and, in fact, the most belief in neutrality, were also banned. (Some were hotheads, to be sure, there are those in every faction, and this is visible in list traffic.)

The system eats those who support it. It either literally rejects them or it burns them out, by failing to build stable structure and thus stable articles and predictable process. Flagged revisions could be a step toward fixing this, if accompanied by a review structure that's designed for reliability. But that is being vigorously resisted.

The paradox is that the crowd-sourced, ad-hoc structure of Wikipedia is an excellent and efficient device for originally building articles, if hybridized with a dispute resolution structure that would be easy to access, facilitated, and reliable. Many early decisions -- and non-decisions --, however, became locked in place. It is not that they were necessarily wrong, it is that mechanisms were never developed for efficiently finding wide consensus; the ad-hoc system that worked for rough article development is lousy for building stable policy and behavioral management when the scale becomes large, as it rapidly did.

There are lots of available solutions, but ... those who know and understand them and try to implement them are often blocked and banned. Within months of becoming active on Wikipedia, in 2007, I saw it happen. While it wasn't surprising -- this is how organizations work, that had been my study -- it was still disappointing, because Wikipedia had excellent policies and guidelines, generally, and I had some hope that they reflected the actual operating consensus. They didn't and don't. See User:Abd/Rule 0. It's not that actual practice is the actual consensus, it isn't. It's that the actual consensus is mostly asleep, and only comes up when there are extraordinarily striking problems and discussions that bring in massive attention.

Example: Jimbo intervened in Wikiversity, deleting some pages and blocking a user who, he declared, was globally banned. It caused a huge flap on Wikiversity, and Jimbo engaged in a bit of bluster, unfortunately, being technically correct (as a defacto representative of the WMF Board, he had the right to take action to protect the entire family of WMF projects) but not necessarily taking the wisest action; since WMF projects depend on massive volunteer labor, pushing the fact of central authority in the face of the volunteers that their consensus is not in charge isn't, shall we say, productive. It's legally correct, but in organizations like those of the projects, the legally controlling Board, and thus its representative or representatives, will normally align itself with the volunteer consensus, unless it finds it legally necessary to do soemthing else.

So elements within the WV community, and probably representing the majority opinion there, filed an RfC on meta to remove the founder bit from Jimbo, that had allowed him to take these actions. It was running 2:1 opposed. Now, that much anti-Jimbo opinion should have been a big red flag to the WMF (and to Jimbo). But elements within the community representing the "cabal," i.e., the circle-the-wagons-against-trolls faction, were, I'm sure, smugly confident that the trolls were being crushed, and they thought of the 1/3 as being simply malcontents, and the RfC ridiculous.

And then Jimbo deleted porn on Commons, bypassing community process. To repeat my summary of this sequence: "You can take away our academic freedom, but don't touch our porn!"

That, of course, is an oversimplification. The real issue was exactly the same, but Commons is much more central and much more noticeable. It got a lot of attention. And what happened then at the RfC on the Founder Flag probably represents a truer consensus, and this affair shows exactly how participation bias can create a false impression. I've seen plenty of examples where particpation bias produced an impression of a "consensus" in one direction, whereas an aroused community went overwhelminingly in the other direction. The !vote, last time I looked, was about 3:1 in favor of removing the flag. Jimbo caved and gave up the intrusive privileges, they were removed from the flag set.

The point is not "Jimbo screwed up!" The point is that ad-hoc dispute resolution doesn't work, except at the lowest levels. I.e., if it works ad-hoc, if disputes are actually resolved, fine. That's efficient. But you have never resolved a dispute when you have blocked a disputant. You have, generally, taken a side in the dispute, which isn't resolution, it is awarding victory to one side. Which then encourages that side to again push the same position with the next victim or opponent. While it's obvious that sometimes blocks and bans are necessary, they, far too easily, become part of the problem instead being genuine solutions. The Scibaby ban is still creating enormous disruption, with ban enforcement having created far more disruption than the damage an individual POV-pusher could have done. If actions had been taken to integrate Scibaby into a cooperating community, he'd probably have gone away, or would have settled down into occasional contribution. And the community has never actually gone back and looked at what happened with Scibaby, only a few have bothered, instead, even though Scibaby was never banned per WP:BAN, by a consensus of *uninvolved editors*, the community simply assumes that the ban was legitimate; after all, hasn't it been enforced for years? Scibaby was screwed over, in short. And human beings tend to resent that. Don't we?

So then we have two juvenile positions, locking horns: Scibaby: You can't stop me! The cabal: Yes we can! We're in charge! However, what Scibaby has successfully shown is: "You can't stop me without harming yourself more than I'm harmed. I'm having fun! Are you?"

You cannot have a neutral project if one faction is in charge, excluding others. Obviously, there must be some central authority of some kind, or a project could not attain the uniformity of expectation that's needed for reasonable reliability. But it must be one which simply facilitates genuine consensus. It's been done, I've pointed to examples, ad nauseum. Until the Wikipedia community wakes up enough to see that it desperately needs this, it will careen from crisis to crisis, rolling the boulder up the hill and watching it roll down again, until the boulder-rollers give up and do something else, being replaced by new generations of boulder-rollers, not yet burned out, all the while wondering why it's so difficult.

Actually, my opinion, it's not necessary for "the community" to wake up, not directly. It would only take a small number of editors who recognize the nature of the problem and who start to cooperate toward a solution.

It is more than any one person can accomplish, but I don't know what the necessary number is. In many contexts, I've been told that it might be as low as two or three. Now, that might seem easy, but try to find two! What happens is that groups of two form, but can't accomplish enough to start the larger ball rolling. So, probably, three. Good luck. I'll help, if asked. I generally understand the necessary structures. But I never wanted to *control* it. I leave that desire to those who imagine they know better than an informed community consensus, or who imagine that they represent the consensus, which is the general position of the default oligarchy in most organizations, whether it's true or not. Usually, it is somewhat true, but that may be because most who differed were shut out or left in frustration or didn't join in the first place.