Wikipedia talk:Dispute resolution/Mediation and Arbitration discussion

Thanks :-)

It should be noted that this is a depressing failure of the goals and policies of Wikipedia. To wit, this proposal marks the death of this passage:


 * How are these policies enforced?


 * You are a Wikipedia editor. Wikipedia lacks an editor-in-chief or a central, top-down mechanism whereby the day-to-day progress on the encyclopedia is monitored and approved. Instead, active participants monitor recent changes and make copyedits and corrections to the content and format problems they see. So the participants are both writers and editors.

--The Cunctator 07:58, 12 Dec 2003 (UTC)


 * It was a bummer about the Easter Bunny too, and Richard Stallman used to make his password public, at least until his system got trashed by intruders. That WP passage was written without very much thought taken about how things were going to scale; did anybody really think that somehow two fanatics at each other's throats were not going to keep on reverting each other's changes every 30 seconds, day in and day out? I'm not at all depressed about mediation, because the unrestrained fanatics were defeating my goal, which was to add to the encyclopedia. Stan 08:47, 12 Dec 2003 (UTC)


 * No, mediation is just a system for setting people up to talk informally and work things out. Binding arbitration, on the other hand, would actually enforce things and that would be a failure of the goals and policies of Wikipedia. LDan

Actually, it was written with a lot of thought about how things would scale. Centralized systems scale more poorly than distributed ones. Yes, people thought that somehow two fanatics at each other's throats were going to keep on reverting each other's changes every 30 seconds. We expected it to happen. I'm confused how that would prevent you from adding to the encyclopedia. --The Cunctator 10:18, 12 Dec 2003 (UTC)


 * If there was a lot of thought that went into it, then I'm not impressed by the quality of the thought process. Edit wars hurt everybody's productivity because lots of people go to look at the flood of changes, they waste time trying to help when they don't realize that the warriors are not rational, and the ensuing meta-discussion is also a huge timesink, again because well-meaning people work hard to rationalize irrational behavior. The line that "centralized systems scale more poorly than distributed ones" is a misapplication of a CS precept; successful distributed systems in CS work because all the systems adhere to protocols which are very tightly controlled. Requiring rational behavior by editors is the equivalent of requiring the use of TCP/IP in a distributed computing system. Stan 18:14, 12 Dec 2003 (UTC)

Cunctator, neither mediators nor arbitrators will monitor or approve activities on a day to day basis. Most problems will still be resolved through bottom-up methods, such as discussion, informal mediation, and better articles. Arbitrators will not be editors-in-chief, and will not be GodKings.

Stan, you unfairly criticise those who founded Wikipedia for their inability to forsee the events of three years time, on a wiki with a thousand times as many pages, and almost as popular as slashdot. Consider how well fundamental policies like the "neutral point of view" have stood the test of time. Consider that when everyone else was saying that Wikipedia would never work a all, those early visionaries correctly saw that it would work, and came together to make it work. Maybe some of their views no longer seem applicable to you, but I suggest that our comments today may seem equally strange to those who read us in three years time. Martin 18:28, 12 Dec 2003 (UTC)


 * Note that I used the subjunctive in reaction to Cunc's assertion, because the evidence I've seen in old articles and talk pages suggests that people simply didn't think about the intensity of nationalistic passions. (I thought I knew my world cultures for instance, but never would have guessed at an Oder/Odra naming controversy!) That's OK; I've worked on GNU since 1989, when few people besides RMS imagined world domination was possible, and have spent part of that time fixing limitations that RMS put in because he didn't completely anticipate what world domination would look like. It doesn't mean he's not a visionary and a great programmer. In the WP case, I see Cunc and others grousing about WP going downhill rather than celebrating its success, just because the community is growing in ways that don't validate their social theories. GNU is run quite differently today than it was 14 years ago, and I expect WP will also evolve in ways that we simply can't foresee today. For all we know, a year from now the whole mediation/arbitration idea will be the butt of jokes that oldtimers tell to the newbies - but we can't get there without actually trying it. Stan 22:19, 12 Dec 2003 (UTC)

I'm not grousing about WP going downhill. I think it's great. Don't get me wrong. In fact, I strongly suspect that if someone did a study of contribution distribution, we'd find that although there are some bottlenecks which engage high amounts of energy for a few people, there's a vast sea of work being done that noone notices because it's utterly uncontroversial and isn't being done by participants who jabber on and on.

It's not really appropriate to make comparisons to GNU--Richard Stallman is much smarter than me, but he has a profound disability in understanding how other people and society works. He's learned a lot over the years but part of the reason for the serious growing pains of GNU rest in his personality.

A more appropriate comparison might be Perl, which was designed as a project by someone who understands the way people work. And though it's certainly developed in ways that Wall never imagined, it's done a very good job of sticking with the philosophical approach that Wall began with.

It's funny you should use TCP-IP as an example. It accepts an arbitrary number of dropped packets on the principle that the important thing is to work very well most of the time and to eventually succeed the rest of the time.

Edit wars *are* a basic problem. But that problem is grossly magnified when people think that they're a big deal. And from the perspective of the overall project, they're not. They're just dropped packets.

Getting back to the grousing hypothesis: my concern is not that Wikipedia is changing in ways that don't validate my social theories; my concern is that Wikipedia is validating my social theories. See the Wiki Life Cycle. --The Cunctator 00:43, 13 Dec 2003 (UTC)


 * "depressing failure" is not grousing? OK... Anyway, you're missing my points. First, I was talking about RMS' own personal coding, not his project leadership.


 * If I'd thought about the dropped packets aspect of TCP, I would have used a different protocol as example - but let's go with it. The ability to handle dropped packets is not an accidental property of TCP, it's a required feature of the protocol and a required part of implementations. A computer not capable of handling dropped packets will be unable to function in the distributed system, and the computers it communicates with will be unable to function effectively as well. In the WP context, those "computers" are people, and it turns out you're blandly condoning the mistreatment of people by others as "dropped packets". Wasn't it you that was unhappy because other editors wanted to delete all the 9/11 tributes? If you're not in an edit war, sure it seems as trivial as a dropped packet, but if you're the one having your hard work trashed by some ignorant person, who deletes over and over and won't listen to any argument, it can change your whole feeling about the project, and we know there are a bunch of knowledgeable contributors who have given up and gone away because no one would stand up to the irrational. If the goal is to build an encyclopedia, letting the nut cases chase away the knowledgeable people is not a good strategy to get there. Stan 05:42, 14 Dec 2003 (UTC)

We're both talking at cross-purposes here, with our insistence on using extended metaphors. I didn't make my TCP-IP metaphor explicit; suffice it to say your interpretation was interesting but not the one I was aiming at. Hopefully the below will make my intent more clear.

With respect to my happiness: I was and am unhappy that other editors want to delete the entries on victims of 9/11.

However, I don't think my happiness is crucial to the future of Wikipedia. I'm not arrogant enough to believe that Wikipedia would fall apart without my presence. The only reason for concern for Wikipedia would be if the events which are causing my unhappiness are indicative of a broader pattern of events which are detrimental to the future development of Wikipedia.

With respect to edit wars: There are four basic choices when faced with a "nutcase" insistent on a particular skewed viewpoint:
 * 1) figure out how to work with the nutcase to incorporate the information or viewpoint they consider important.
 * 2) edit war--that is, reversion;
 * 3) (attempt to) prevent the nutcase from editing
 * 4) let the edits stand, perhaps commenting on why they're wrong in the talk pages.

The first option is clearly the best from the viewpoint of Wikipedia's content and community, but requires the most effort on the part of the editors. The last is the worst for the content, but requires the least affort on the part of the community. The middle two are resorts to removal of freedoms for the nutcase (essentially violent responses) to protect Wikipedia's content and community.

I was using the "dropped packets" metaphor particularly to refer not to people dropping out due to upset--even though I am simply not convinced that happiness of long-time contributors is dreadfully important--but to content which is imperfect due to not being improved, since I believe that over time the likelihood that the imperfect content will disappear nears certainty.

That is, I think option four is a much more reasonable option than many people do. I know that most people have a lower emotional tolerance for this than I do, but I really think a lot of time a little avoidance of troll-feeding goes a long way, and that the troll-feeders are often at least partly to blame.

Edit wars with nutcases are rare--edit wars between functional adults with strongly held beliefs which involve a degree of non-rational thought are common. --The Cunctator 06:53, 14 Dec 2003 (UTC)


 * The first option is extremely time-consuming, and at least from what I've seen in the past ten months, rarely works anyway; by definition, a person who is irrational on some subject is not going to be persuaded by rational arguments, and will simply take advantage of the situation. Letting bad edits stand might have worked when they were fewer editors, but with thousands of contributors, it's now almost certain that there will be someone who will be unwilling to leave it alone - after all, we do encourage everybody to "be bold", right? But if all the knowledgeable contributors are chased away, WP will simply sink into a morass of advocates of conspiracy theory A edit-warring with advocates of theory B - some would say it's already happened :-). Your faith that everything will magically get better is touching, but real-life distributed systems only work because of the dedication of people committed to maintaining them - and defending them from assaults by the destructive. Stan 18:06, 14 Dec 2003 (UTC)

Please don't use denigrating language such as "Your faith that everything will magically get better is touching" if you wish for me to engage with you seriously.

With respect to option one, I had a feeling you'd say that. The reason I presented option one is that I've been always able to make it work. There's never been a nutcase whose content I haven't been able to incorporate through various methods of avoiding direct conflict. As I said, it does take a lot of work.

I have a strong bias against arguments which use language such as "destructive", as that is a word which specifically refers to non-reversible processes. At the individual level, no edit is destructive. The possibility of statistically destructive processes exists, but without evidence of such a trend, my bias is not to accept that violent measures need to be taken to prevent it.

It is true that if all the good contributors on Wikipedia are chased away Wikipedia would sink into a morass of uselessness. As I said, I don't believe a statistical analysis of the contributions to Wikipedia would show that is the trend. Maybe you believe that it is the trend. The only thing that should be clear to both of us is that that information is not something that can simply grokked by an individual.

Rather, the destructive trend for which I see copious evidence is that of agglomeration--the trending of entries toward unworkably long texts with poorly defined boundaries: see New York, New York and Roman Catholic Church sex abuse scandal for just two examples. The longer the text and the more poorly defined its boundaries, the more difficult the text becomes to edit. Also, the more poorly defined the coverage under each entry title, the less useful the interlinks between entries becomes.

When I've been making attempts to break out the content into parsable chunks of specific information, the work has been immediately reverted by people "protecting" the entries from "mutilations". There is a clear trend towards increased article length, which has been abetted by technological changes to the software which Eloquence made. In a remarkable coincidence, as an editor he is a strong advocate of long entries which cover multiple related topics. This trend is, I argue, antithetical to the things that make the wiki process work in creating a knowledge reference. --The Cunctator 18:35, 14 Dec 2003 (UTC)


 * How is a discussion of the benefits of agglomeration relevant to the Mediation and Arbitration proposal? Angela. 18:39, 14 Dec 2003 (UTC)


 * Incorporating nutcase content would work if the nutcase had left - do you have an example where you were able to engage the nutcase and convince them to stop?


 * Also, reversibility of edits doesn't guarantee that WP can't be destroyed. At the point of 100 irrational edits for every rational edit, the community of encyclopedia-builders will be gone. Stan 19:08, 14 Dec 2003 (UTC)

Angela, understanding the causes of conflict between editors is crucial to successful mediation and arbitration. --The Cunctator 19:27, 14 Dec 2003 (UTC)


 * I don't disagree, but I don't think having those conflicts on this page is a good idea. Angela. 19:33, 14 Dec 2003 (UTC)

Stan: I have successfully engaged nutcases, yes. Heimatvertriebene is one example.

I agreed that reversibility of edits doesn't guarantee that WP can't be destroyed; as I said, "It is true that if all the good contributors on Wikipedia are chased away Wikipedia would sink into a morass of uselessness....The possibility of statistically destructive processes exists, but without evidence of such a trend, my bias is not to accept that violent measures need to be taken to prevent it." --The Cunctator 19:27, 14 Dec 2003 (UTC)


 * What did you do on Heimatvertriebene? It's not obvious from history - I see only a single undated comment with your signature.


 * I think a number of editors have the feeling that they're struggling to hold back a flood of junk; I personally spend as much as 1/3 of my daily WP time just reviewing and fixing other people's bad edits - and that's with me steering generally clear of edit wars and not doing nearly as much cleanup as much some other editors do. I think that's an underlying cause of simmering anger over VfD for instance. We may not have statistics proving that editors are spending more and more time dealing with bad edits, but I don't think anybody has tried to measure it either. We do know that there are a bunch of respected ex-contributors who've quit because they got tired of fighting with nutcases that no one was willing to do anything about. Stan 01:20, 15 Dec 2003 (UTC)

It's not going to be clear what I did with Heimatvertriebene since it looks like the edit history doesn't reach back that far back. But it involved massively rewriting that entry and the entry on ethnicity and creating the nationalism, ethnic nationalism and diaspora entries, and possibly others. The method was to identify what the distinct concepts were within the controversial mess and to extract all of the content related to those distinct concepts. So when X said "heimatvertriebene was a diaspora of Germans" instead of arguing, as some did "it wasn't a diaspora" and "they weren't Germans", and instead of trying to define "diaspora" and "Germans" in the heimatvertriebene entry, I used the power of Wikipedia to bring clarity and specificity to the language. I also did some quick research into what heimatvertriebene was. I used as a basic assumption that there was a valid point that H. Jonat was trying to make, and that her one-issue attitude and her limited English was muddying that point. What needed to be done was to polish and shape her text to reveal the valid information hidden within. And she appreciated that I was working to do so.

Yeah, the feeling of the impossibility of holding back the flood is universal, and you may be right that the dam will break. But all I know is that feeling has been the same from the very start, and the nutcase issue doesn't seem to have gotten worse or better. However, other stuff has changed. It's massively more difficult to be bold in editing, as people feel attached to particular versions of entries and expect that all significant edits to be approved by them beforehand. As you can see from my contributions I've tried the experiment of editing in a significant amount again over this weekend after several months away and the results are in--insistent rollbacks in nearly every single case, from Philadelphia, Pennsylvania, to Sport, to Rick Santorum, to Roman Catholic sex abuse scandal. At least with sport the person who's taking it personally is willing to discuss the edits with me. It may be time to rehabilitate my public image. But I can't say I'm desperate to act kindly and avuncular just to build up a posse. On the other hand, a quick check on "Random Page" makes me feel very positively about Wikipedia--it's turned up a steady stream of nice entries. You should try it whenever you feel bad about the future of Wikipedia.

Stan, do you think I'm on to something with the identification of the deletionism/agglomeration trend? --The Cunctator 04:18, 15 Dec 2003 (UTC)


 * Your experience with surly reversion is not uncommon these days. Edit wars are tiring and irritating; a tired and irritated editor will be tempted to just revert a change instead of starting up yet another discussion that might become long and irritating. When I get tired of it all myself, I'll pick an unnoticed page and work it over, plus fill in some of the free links with new articles, nothing to fight about because they're new. :-) I've commented on deletionism and agglomeration on your respective pages. Stan 06:36, 15 Dec 2003 (UTC)

I'm curious why this page is listed as (proposal) when the committees have already been selected. Is there still something to be done before this goes live? - Hephaestos 08:18, 12 Jan 2004 (UTC)

This page was originally created by me as a place to document the proposal that was presented on the mailing list in succinct form. Anthere requested that it be posted in the Wikipedia space so it could be looked over and discussed. At that time the Role acount User:Mediator was created and then quashed. Some discussion continued about the process and then Jimbo started naming members of the committee.

This post would be more appropriate at Talk:Arbitration committee, but the article for that talk page has not been created yet, and I want someone to actually see this post. I seem to recall that Jimbo chose not just mediators, but also arbitrators. If my memory is not faulty, could someone please post the names of the arbitration committee? I've looked, but I can't seem to find them. Cheers, Cyan 19:02, 14 Jan 2004 (UTC)

From http://mail.wikipedia.org/pipermail/wikien-l/2003-December/008779.html

See arbitrators, mediators

--

Arbitration
Mediation has sometimes been utilized to good effect when coupled with arbitration, particularly binding arbitration, in a process called, appropriately enough, "mediation/arbitration." In this process, if the parties are unable to reach resolution through mediation, the mediator becomes an arbitrator, shifts the mediation process into an arbitral one, seeks any additional evidence needed (particularly witnesses, if any, since witnesses would not normally be called in a mediation), and renders an arbitral decision.