Wikipedia talk:Governance reform/Archive 3

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

Election Committee

I think an election of this large a group will need a designated election committee. The Community can still assist but I think we need a group of users to be responsible for making sure the election is well run. Should the Election Committee be appointed by ArbCom after the users are self nom and discussed by the Community? How many? FloNight♥♥♥ 23:42, 25 April 2008 (UTC)

See what I mean about piling on the bureaucracy? Reason number three why I heartily oppose the proposal. VanTucky 23:57, 25 April 2008 (UTC)

VanTucky, this would decrease process wonk by removing power from the abusive ArbCom and giving it to the actual editors. Monobi (talk) 02:40, 26 April 2008 (UTC)

Again, maybe I'm in a crazy minority, but I don't think ArbCom is abusive. And besides, since when was the power to create and edit policy solely in the hands of ArbCom? Never. It's a moot point. VanTucky 21:19, 26 April 2008 (UTC)

Thankfully we already have a Election Committee. We call them Bureaucrats. Geoff Plourde (talk) 01:22, 26 April 2008 (UTC)

I was thinking more that, since ArbCom members can't be members of the Council as well (in order to ensure a separation of powers), one or two designated ArbCom members could supervise the elections. (Newyorkbrad might be willing to take on the job, I would expect). Whoever was designated as election clerk would, of course, have to be impartial and abstain from voting in the election. Walton^One 07:24, 26 April 2008 (UTC)

Why this is unnecessary

This proposal seems to completely misunderstand what Wikipedia's policy-making method is at the moment. Apparently it has broken down -- this is a myth! There never was a policy-making method! It is said that the last policy to be created was BLP. How many others can you name? In my memory, policies that have been approved in this way can be counted on one hand -- 3RR, the Arbitration Committee, BLP. But policy has changed since BLP was instituted. For instance, the (misguided, IMHO) policy of community sanctions exists, even though it has never been put to the vote. It is practised and it works to a certain extent; certainly it is Wikipedia's practice, and the way it has always worked is that practice is policy.

This proposed change seems to me to be highly unneeded. The whole point about Wikipedia is that our entire way of working can be summed up in three words "ignore all rules". This isn't just a policy, or a recommendation, or a current working practice, it is also the way other policies, recommendations and working practices are formed.

If people are trying to find a way of fixing the broken method of creating policy -- the one that involves straw polls -- maybe they should question whether that actually is the way of making policy.

Sam Korn ^(smoddy) 12:56, 26 April 2008 (UTC)

That may have been how it worked initially. Now, ignoring all rules is becoming increasingly frowned upon, as policy is increasingly being seen as prescriptive, so changing policy by adjusting practice is frowned upon (because it means violating current policy). Changing policy by explicitly proposing policy changes doesn't work because all it takes is one reasonably-sized chunk of opposition to scuttle any policy change. Besides that, there are some things - BLP protection being the one that I'm most fond of invoking - that are too important, and currently not done well enough, to be left to the consensus process. Sarcasticidealist (talk) 13:03, 26 April 2008 (UTC)

Can you name another? Four or five policies in seven years seems a bad reason to make this proposal.

The day that ignoring all rules ceases to be acceptable is the day Wikipedia dies. I have hardly ever -- in the best part of four years on the site -- read a policy page and I don't intend to start anytime soon.

Sam Korn ^(smoddy) 13:12, 26 April 2008 (UTC)

Two questions

There are two questions here, to my view. First, is the current way things get done working to handle the large and important issues confronting the project? .. and second, if there is a problem with that, is this proposal the way to solve the problem?

I think the answer to the first question is yes no, things are not working. I think we have a building crisis with BLPs that we are not effectively coming to grips with. I think Kim (who above has been repeatedly (!!) saying "there is no problem, everything is fine") is wrong. The many attempts to put forward solutions to the BLP problem by prescriptive policymaking have failed, and I think most of us recognize there IS a BLP problem.

But the second question? No way. I think this proposed government structure is unimplementable, and if it were actually implemented, I think while it might be able to tinker round the edges, it would fail to come to grips with the actual important problems. So I oppose this solution. Policy here remains descriptive, and as long as you have influential people such as Sam, above, proud that they do not read policy pages, it's not likely that will change. No, I feel the same way about this as I do about the Wikicouncil proposal... not a good idea. ++Lar: t/c 14:16, 26 April 2008 (UTC)

Do you mean that the answer to the first question is "no"? I'm having trouble reconciling a yes answer with the rest of your paragraph (or indeed with the views I've seen you express elsewhere). Sarcasticidealist (talk) 14:50, 26 April 2008 (UTC)

Yes, that was a typo. Struck and corrected, thanks. ++Lar: t/c 16:21, 26 April 2008 (UTC)

Indeed, that looks kind of like a typo. ;-)

More to the point, is there any feasible alternative (to the broad concept of a legislative body, not necessarily the specific details being discussed here) that would allow effective policy change to take place? Kirill 14:54, 26 April 2008 (UTC)

Yes, I think there is a feasible alternative. Embrace the wiki way. This initiative is trying to move policy to be prescriptive. I think that is a very hard move, and counter to the wiki way. I believe policy here should, and further that it will, remain descriptive. Therefore the way to effect policy change is to change what one does, and encourage others to change as well. The way to solve the BLP crisis is not to propose prescriptive policy after prescriptive policy, each to go down in flames (because 65% approval isn't consensus, or whatever) but instead, act on the principles behind the proposed policies... close marginal BLP AfDs as deletes. Lock contentious BLP articles after stubbing them out. Delete marginally non notable BLPS if the subject turns up, shows they aren't notable, and requests deletion. Give more credence to the arguments and assertions credentialed people than anonymous ones (absent any other reason to make a difference). Sign and adhere to the Responsible Editing pledge. And so on. Those are all ways to move the descriptivist policy in the right direction, and there are other ways that are BLP related as well. That can work for more than just BLP as well. ++Lar: t/c 16:57, 26 April 2008 (UTC)

PS, another thing I think would help a lot would be if ArbCom stepped up to the plate more. ArbCom's role is not to make policy, but when the trend is in the right direction and cases come up, it could be more forceful in its articulation of findings, especially principles related to policy, and recognise emergent policy earlier, and come down a lot harder on those who are standing in the way of emergent policy. What ArbCom did in the BadlyDrawnJeff case was helpful, rearticulation of the BLP principles was very good. Some of its more recent decisions have been far less helpful (the IRC case springs to mind for me but others may have other examples) ++Lar: t/c 17:03, 26 April 2008 (UTC)

I regularly find Lar's opinions quite insightful, and here I am again. Descriptive policy is the sum of countless individual consensus decisions across the wiki, based on community input and oversight, and that in many ways better expresses the will of the community than a small voting body ever could or would. All we need to do to change descriptive policy is change what we do, and to do so in a way that others agree with and support the changes -- consider it a viral thing, if you will. If ArbCom misses the ball, there's still a possibility of community resolution; if the proposed policy committee misses the ball, what's the next move? – Luna Santin (talk) 23:40, 26 April 2008 (UTC)

Actually, I really should add a couple of caveats to what I said above. I don't disregard policy -- it is vitally important. I don't, however, spend my time reading it. I know what it says, as much from just spending time here. You shouldn't need to read policy -- it should be totally obvious! The other caveat is that there was a time when I took great care to read policy -- when I was a member of the Arbitration Committee. Then it was important to understand why people took sometimes strange action claiming policy support -- and often in good faith: they misread the generalities of the policy on account of its specifics. The reason I don't think prescriptive policy works is that you end up being unable to see the purpose of a policy due to the specifics of it and you end up enforcing it in the wrong way and for the wrong reasons.

Furthermore, prescriptive policy doesn't work on a site like Wikipedia. It is a radically free and open project; prescriptive policy will not work under those conditions.

Sam Korn ^(smoddy) 15:43, 26 April 2008 (UTC)

In total, there have been maybe 10 prescriptive policies ever made. One of the last things I did in cooperation with Radiant while he was still visibly around was to put a huge amount of sweat into tearing down the old broken prescriptive policy system, because it was an utter waste of time. It never worked, and Radiant and I finished it off, mercifully! Wikipedia has been around for 8 years, and we arguably(!) have maybe at most 8 prescriptive policies.

But look at how many policies there are in Category:Wikipedia official policy, a lot more! And there's also plenty of guidelines in Category:Wikipedia guidelines. So, we actually have an existing method that is used to maintain Policy and guideline pages. It's called Consensus. I'll even admit to occasionally lording our superior consensus system over other wikis <very innocent look>.

So I'm not so pleased about people attempting to ressurect the prescriptive system. Some people even want to kill the consensus system alongside. Well, that might not be entirely cool :-) --Kim Bruning (talk) 16:56, 26 April 2008 (UTC) Remember that all project namespace pages are intended for projects or to document best practices. Radiant's original categorization scheme divided the documentation pages into 3 categories: Policy, Guideline, and Essay. I still think that this categorization scheme uses misleading names and is not fine grained enough for our current needs, but that's what we've got right now. People seem to have all kinds of grand ideas about what that scheme is meant to represent though. It's just a categorization scheme ;-)

I'm not entirely sure that the policies and guidelines being proposed fall in the "proscriptive system" being defined above. Personally, except in very obvious cases (WP:DONOTKILLPEOPLE, for instance), I think most policies and guidelines should be arranged for at least reasonable exceptions. But to oppose something on the basis of one's possibly incorrect views of what it might, potentially, at some point become, when it is being argued for the most part as being inherently limited, strikes me as being, well, odd. John Carter (talk) 17:04, 26 April 2008 (UTC)

Do you understand the difference between prescriptive and descriptive? I am opposed to "should" , and "arranged" and "reasonable exceptions" and all they entail. I'm not opposing what it might become one day. I'm pretty sure you're already there, because -while I totally respect you- you simply might not know any other way! --Kim Bruning (talk) 17:15, 26 April 2008 (UTC) I get the impression that I'm speaking an alien language to many people here... sounds like there's some acculturation problems. :-) If you're relatively new to wikipedia governance, you might think I'm totally bonkers or speaking in tongues or so. If so, I'll explain in more depth.

Not an alien tongue. The issue, I think, is that different people are defining what they feel policy "is", each in their own way. It may turn out that there will need to be a discussion on what policy is, before there can be a discussion concerning the governance of policy. (Probably not a bad idea, I suppose. Though I would have thought most of the peoiple in this discussion were on-the-same-page about this, as it were - and please pardon the pun : ) - jc37 17:22, 26 April 2008 (UTC)

Well, you could discuss it, but it'd be a redefinition, if you're not careful. To me {{policy}} it's a category (and template) from a categorization scheme for wikipedia project space documentation, which was set up by Stevertigo, Netoholic, and Radiant (&co.), whom I knew fairly well. --Kim Bruning (talk) 17:27, 26 April 2008 (UTC)

And it represents *what* to you? (I think I know, but I also think you're enjoying making it appear that you're talking in circles : ) - jc37 17:33, 26 April 2008 (UTC)

What the policy category represents? To me, nothing more than that. It's a semi-arbitrary category. Tags get shifted around all the time, and at times important pages are categorized "lower" than less important ones. Personally, I'd prefer a more accurate and fine grained system, for instance one where pages are scored by how many times they are linked to, or by averaging importance scores attributed by users ("How would you rate the usefulness of this page. Does that answer your question? Or did you mean to ask a different one? --Kim Bruning (talk) 17:52, 26 April 2008 (UTC)

Ever see Chicago? ("Now presenting: A tap dance..."  : ) - jc37 18:22, 26 April 2008 (UTC)

Nope, though I could watch it sometime if it's a nice movie. What's the relevance? --Kim Bruning (talk) 18:49, 26 April 2008 (UTC)

To me, what I think in most cases what would be done wouldn't apply as proscriptive or restrictive, but, like Kim suggested earlier, "best practices", with possibly at the end, "if you don't do this, you have a very real chance of getting in deep doodoo." Like is the case with most laws todays, I very sincerely doubt that there will be creation of many new policies, but rather, more often than not, just statements made about the application of existing policies in some more specific instances. I doubt very seriously that we would very often have creation of entirely new policies, regarding things that aren't yet covered, at least in a broad sense, by any extant policy, though. John Carter (talk) 19:07, 26 April 2008 (UTC)

Lar: I'm not quite saying that nothing is wrong. I recognize that some people have a problem. I also have a bit of a problem with those people, and possibly they with me. Not because we hate each other, but because we can't understand each other right now.

I guess the problem might be an area in which you haven't been looking yet? To wit: it's an acculturation problem. It becomes rather hard to solve problems together when you're not all on the same page eh? :-) But once people *are* on the same page, it's surprisingly easy to cooperate and work together, (even when updating and maintaining policy pages.) Folks who are on the arbcom mailing list could ask their colleagues (and ex-members) about acculturation issues too. You'll find interesting answers. :-)

I've been trying out different methods to try to solve acculturation issues. I think Wikipedia:Lectures might be the best approach so far. Drop by, listen in, and help out! :-)

--Kim Bruning (talk) 17:34, 26 April 2008 (UTC)

Not sure that lectures and acculturation can solve this. Starting to think that there is too much talk and not enough do. :) ++Lar: t/c 18:20, 26 April 2008 (UTC)

Won't solve the BLP issue by themselves. Might solve some of the too much talk issues (ironically :-P ) --Kim Bruning (talk) 18:51, 26 April 2008 (UTC)

Oh, also addressing the BLP problem. Yes we have one. A really big one. I don't think that wikipedia is structured in a way that can ever resolve the biographies of living persons. The requirements (clarity, accuracy, morality, all delivered on the first public edit) are impossible to attain with the 100% public wiki-model used by en.wikipedia, which takes time and many edits for an article to get into shape. If you drop the public part from the wiki-model, things become possible. This is similar to the model already used at wikinews: where they first work on articles, and don't publish until they are done.

As I don't want to sacrifice the current public wiki model, my current position on BLP's is that en.wikipedia is the wrong site to host them altogether. We should consider moving such content to (for instance) Wikinews instead.

--Kim Bruning (talk) 17:44, 26 April 2008 (UTC)

I'd be sad to see that (moving BLPs off en:wp entirely) as the solution, but if no other solution can work, then that may be what needs doing. It would require some assurance that the problem gets solved by the move, not just moved elsewhere. As well as analysis of the edge effects of removal of bios... what of the articles that remain that, while not bios, still refer to living people? I'd rather see every other option explored first. But ya. ++Lar: t/c 18:20, 26 April 2008 (UTC)

So would I. President of the United States would have to be moved (several are alive), all sports teams pages would have to be moved because of living players, recent movies, TV shows, even some biographies of historical personages who have had recent, notable biographies written about them, all would have to be removed if all BLP's were to be moved off. I doubt we'd have any more than 20% of the current articles remain if that were done. John Carter (talk) 19:57, 26 April 2008 (UTC)

At which point we should better just close down the wiki. Fortunately, I don't think it's quite that bad, and Lar is correct that a middle ground might be found. Many BLPs are not contentious at all. Possibly we might only transwiki contentious ones. :-) --Kim Bruning (talk) 20:21, 26 April 2008 (UTC)

Kim, that is easy to say unless you are the person that is being publicly talked about on one of the largest web sites in the world. For many people, a Google search brings back a Wikipedia articles as the first hit for their name. People genuinely get upset when the article misrepresents who they really are. FloNight♥♥♥ 23:27, 26 April 2008 (UTC)

If you read back up the thread a bit, you'll see that my position is that we may have to ban BLP entirely! The post you're replying to is a confirmation that I'm willing to compromise on that position somewhat, if someone figures out a different solution for our mutual concerns, or if it turns out that our/my position is untenable on practical grounds. --Kim Bruning (talk) 20:18, 27 April 2008 (UTC) ie. a tacit reaffirmation that I'm willing to participate in the consensus process

Proposals so far

So far as I can tell, these are the various proposals so far suggested:

• (1) A group of elected editors who in effect write policy and guidelines at the request of the community, the foundation, or other parties - PolCom, if you want
• (2) A group of elected editors who review extant policy and guidelines, adding and pruning as required or indicated
• (3) A group of elected editors who function much like a legislature, writing and amending policy and guidelines as required

Personally, I would favor something like the third alternative, as it seems to incorporate both of the first two. However, I also think it should be acknowledged that there should be very serious limitations on what the group would be permitted to do. In effect, a very detailed constitution would be indicated as well, specificying in detail under what particular circumstances this body would be enabled to act. Anyone who sees anything I missed, or wants to add other options to the list above, please feel free to do so. John Carter (talk) 14:26, 26 April 2008 (UTC)

Though there are a myriad details throughout this page, it looks like a decent enough summary : )

Since I proposed it, it's probably no surprise I support #2 (review). And strongly oppose the creation of any body to "make" policy. In addition, I'd probably oppose #3 on the grounds of WP:NOT#BURO, but really this whole page falls a bit befoul of that : ) - jc37 14:35, 26 April 2008 (UTC)

(2) and (3) are virtually identical from a practical standpoint; if the group can change existing policy, then it can make new policy so long as it finds an existing page to add it to. The only real distinction that I could see being drawn would be that (2) would be unable to create a policy which is entirely new (i.e. has no connection to any present policy); but that doesn't seem like much of a limitation.

(Personally, I'd prefer (3) or something close to it.) Kirill 14:40, 26 April 2008 (UTC)

Thank you, btw. You just answered one of my questions higher up this page. - jc37 14:43, 26 April 2008 (UTC)

Also, 2 and 3 are decidedly different. #2 reviews pages created by the community. #3 creates such pages. Therefore, #2 doesn't create policy, but rather is more "ratifying" a page which documents what already exists. #3 would actually create policy. Bad idea. - jc37 14:46, 26 April 2008 (UTC)

Wouldn't that be "ratifying as required" instead of "adding and pruning as required", then? If the group can add to a policy page once it's in their hands, then they can effectively create policy by doing so, no?

(More generally, even if all the group does is take ratifying votes on community-proposed amendments, any member of the group could—acting as a community member—propose an amendment themselves, so the group would wind up being able to change policy anyways.) Kirill 14:52, 26 April 2008 (UTC)

As long as we keep the "hats" straight (and separate). Any arbcom member could submit an RfAr, recusing themselves from their arb hat. I don't see this as any different. - jc37 14:55, 26 April 2008 (UTC)

John: you forget "(4) use consensus and wiki-editing process." :-) --Kim Bruning (talk) 16:59, 26 April 2008 (UTC)

I may be repeating myself, but if #2 means "ratifying a proposal which already exists", it does not mean a great advance over the current situation. For the community would first have to agree on a proposal which should be submitted for ratification. The community should have the possibility to comment, suggest changes, and prepare alternate versions, before the actual vote is cast; but preparing the actual "final version" for voting should rather be done by a small group. Even with 40 or 50 editors that will be difficult; with >1000 it's hardly possible. --B. Wolterding (talk) 21:39, 26 April 2008 (UTC)

The actual number is typically under 10, for any given day.--Kim Bruning (talk) 22:16, 26 April 2008 (UTC)

And that's the point - the final version should be determined by a well-described process in which everybody has a fair share, not by the 10 people who just happen to edit it on the last day. --B. Wolterding (talk) 22:24, 26 April 2008 (UTC)

There is no final version, and there is no last day. The process is described in great detail and mostly enforced by software. And that's what ensures that everybody who wants to can make any input they want, for as long as they want, to any level of detail that they want. And if they change their mind, they can come back and do it all over again. And again. And again. Whenever they feel it's necessary. --Kim Bruning (talk) 22:31, 26 April 2008 (UTC)

And this instability in our policies is one part of the problem. Most people do better if there is consistency in the world. When policies are variable and open to constant revamping it causes unnecessary conflict. I'm increasingly concerned that are processes are needless making people have unhappy days. :-( FloNight♥♥♥ 22:57, 26 April 2008 (UTC)

Interesting that this is being argued both ways: the committee is at once the silver bullet which manages to enable substantive changes to policy, even as it also prevents substantive changes to policy. – Luna Santin (talk) 23:11, 26 April 2008 (UTC)

I don't know, I'm kind of enjoying watching the great dance here. Kim does a remarkable job at getting everyone out of their seats and on the dance floor. I don't even mind the tune for the most part. (Though at times... : ) - jc37 23:18, 26 April 2008 (UTC)

@Flonight: This was discussed a long time ago. I think we agreed that the world around wikipedia changes, so we need to be able to change with it. That has upsides and downsides, of course. --Kim Bruning (talk) 01:15, 27 April 2008 (UTC)

@Flonight (late addition): And how could I forget. If policies make you nervous or depressed, you may of course simply ignore them. *gentle smile* --Kim Bruning (talk) 20:29, 27 April 2008 (UTC) this was one of the first rules of wikipedia. :-)

Kim, it amazes me that you continuously argue with "consensus". What is the "process" you're talking about, with respect to consensus? Who determines this consensus? Technically, everybody can edit the policies today; but few people do - not because they don't have an opinion or wouldn't support a change; but because they're tired of edit warring, or because they're too polite to override others. Currently, it doesn't seem to be a fair consensus that determines our policies or guidelines - rather it's a consensus of the most reckless. --B. Wolterding (talk) 23:01, 26 April 2008 (UTC)

Wikipedia:Consensus (and much of the project namespace in fact) doesn't document well enough? If not, we can recruit people to help out. :-) --Kim Bruning (talk) 01:15, 27 April 2008 (UTC)

Wikipedia:Lectures

Wikipedia currently has about 100 pages marked policy or guideline. Since I need a subject for Sunday's talk anyway, why don't I go ahead and explain the methods by which those have been created and maintained? Would people be interested?

Sign up at Wikipedia:Lectures, or just show up at 15:00 UTC (that's 11 am Eastern Standard Time, for you usa-ians )

--Kim Bruning (talk) 16:59, 26 April 2008 (UTC)

A different opinion: Wikipedia governance is thriving.

<Grin> Since two arbitrators started all this, I figured I should shoot back a bit, if good-naturedly (especially when the writers started taking shots at other dispute resolution processes) :-P

Arbcom a one stop shop? Why not compare the amount of cases handled by EA, 3O and Medcab together, versus number of cases handled by the Arbcom. You probably won't be surprised at all.

--Kim Bruning (talk) 20:18, 26 April 2008 (UTC) With apologies to those arbcom members who didn't start this page. Arbcom does a great job, and I don't want to turn this into mediation versus arbitration :-)

The lack of success in developing consensus is the reason that we have chronic problems. Dispute resolution of all types, including Arbitration, has been unsuccessful in ending these conflicts. Because the Arbitration Committee limits our self to addressing editor conduct issues instead of resolving content disputes, these conflicts feaster in the community. If the Arbitration Committee wanted to break with custom and write policy we could settle these disputes, but I do not think that is in the best interest of the Community. FloNight♥♥♥ 23:13, 26 April 2008 (UTC)

I find it disturbing that anyone on ArbCom seriously considers the idea that they have the authority to dictate policy. :) – Luna Santin (talk) 23:20, 26 April 2008 (UTC)

We don't, where did you get that idea? FloNight♥♥♥ 23:32, 26 April 2008 (UTC)

"If the Arbitration Committee wanted to break with custom and write policy we could settle these disputes..." makes it sound as if the only thing keeping the committee from doing so is its own opinion; if that view is accepted, it implies they have that authority and simply choose not to use it. – Luna Santin (talk) 23:42, 26 April 2008 (UTC)

I'm finding it hard to see how that statement does anything but acknowledge that the Arbitration Committee DOES NOT have the authority to write policy. FloNight♥♥♥ 23:47, 26 April 2008 (UTC)

If you can name specific areas where there is difficulty obtaining consensus, something can be done about that. :-) --Kim Bruning (talk) 00:06, 27 April 2008 (UTC)

No one is hiding these issues from the Community. Read the last two years Arbitration cases. Keep in mind that frequently the reason that they reach Arbitration is because the Community can not reach consensus and one or more users gets so frustrated that the editor(s) engage in misconduct. We see this happening over and over again. FloNight♥♥♥ 00:14, 27 April 2008 (UTC)

Isn't that what the arbcom is for? :-) Do you feel that the arbcom is able to handle all the substantial cases that come its way, or do you end up rejecting a lot of cases that you really would have wanted to take on? --Kim Bruning (talk) 00:45, 27 April 2008 (UTC)

Some of our users have unhappy days, maybe even weeks, from their volunteer work on Wikipedia. If I can find a way to prevent some of this unhappiness by preventing some of these major flair ups over policy related issues, I want to do it. And as soon as possible. Life is too short... FloNight♥♥♥ 00:58, 27 April 2008 (UTC)

Totally Agree with that much. That's been my entire aim for years. I also talked with DMCDevit today, but he won't have time 'till summer. We should put our heads together soon. --Kim Bruning (talk) 01:20, 27 April 2008 (UTC)

Thank you for clarifying your original comment. Seems we've had an unfortunate miscommunication. – Luna Santin (talk) 00:33, 27 April 2008 (UTC)

I'm sorry if my wording confused you. I meant to make it clear that the Arbitration does not write policy because we do not have the authority to write it. FloNight♥♥♥ 00:52, 27 April 2008 (UTC)

My head...

I have spend more than two hours reading this page (during which I should be normally sleeping); I honestly find it particularly difficult for anyone to follow this discussion, even if some people do have the time and energy to read the page. For the love of the Unicorn, 236 kilobytes in less than five days? It is ridiculous. I propose that we should try and rationalise this discussion. Can every participant please state their opinion, as clearly and succinctly as possible, about what changes exactly each desires to see take place (or not)? Then we could perhaps see some common themes and ideas and build on them. Waltham, The Duke of 00:50, 27 April 2008 (UTC)

Sure. (It's also higher up in the threads at #Graphical workflow of what the ideas "look like".) I support the idea of having a policy review committee. One which does not "make" policy, but which is made up of elected editors who would review pages submitted by the community.

Something like the following (copied from above):

• A. "Request for policy and guideline review" (PAGRV): A request for a guideline page to be reviewed to become policy OR concerns about an existing policy page come to light
• B. (possibly) a request for an essay to be reviewed to become a guideline, or a guideline review to become an essay. (I say possibly, because we may wish to leave essay/guideline discussions solely with the community, with the PAGRV committee merely being solicitied for an opinion on policy.) Acting as mediator/arbitrator concerning only policy/guideline, and not the actions of editors involved - which would instead be arbcomm's jurisdiction
• C. The PAGRV committee is requested by arbcomm, or some other group, to offer a "finding" and/or "principle" concerning policy and/or the interpretation of policy.
• D. WMF (or some such personage or body) has determined that a guideline (or some core concept) should be or become policy (I'm strongly tempted to remove this section, as I really don't like the idea of this review body even coming close to "creating" policy or even a policy page. Perhaps this could be downgraded to PAGRV supervising the organisation of a panel/project to create a policy page, upon request of WMF, or whomever, and then reviewing that panel's results, and presenting both the results and the review to the requesting body.)
• E. A - D lead here, to a "presentation" page (a combination of arbcomm's evidence and workshop) which anyone can edit.
• F. E leads here, to a discussion between the members, with community allowed discussion on a related talk page (comparable to "final decision" in arbcomm, or a bureaucrat chat)
• G. Resolution/Interpretation/Opinion presented.

This essentially is (roughly) how Featured article process works, Arbcomm works, and a host of other examples. It's inclusive of the community, while having a set of individuals acting as the "closers" to the discussion (rather than just a single individual, such as Jimbo Wales (in the past, and, though rarely, sometimes the recent present), or an admin (XfD), or a bureaucrat (RfA), or whatever.)

Review = reviewing a page to see whether it meets/falls within a.) Foundation principles b.) current general usage c.) the current policy framework d.) current community consensus, etc. - jc37 01:10, 27 April 2008 (UTC)

Very good, thank you. Now, next participant. No comments whatsoever; I want to see a succession of proposals. Discussion can continue afterwards, preferably after some reflection. Waltham, The Duke of 01:19, 27 April 2008 (UTC)

My very modest and novel proposal is to use the standard wiki-editing model and consensus, which has been very successful in bringing us up to this point. The wiki model allows everyone to participate, everyone can provide input, and they can do so at any time. When everyone assumes good faith, you can get huge amounts done in a very short period of time. This is also why (surprisingly to some) the wiki model can also be very good at handling rapid response in emergencies.

I understand people have particular issues, but they're very normal wiki-issues, and are no more difficult to solve than any issues that went before. People have predicted the downfall of wikipedia (by dinnertime at latest) every day since day one. It still hasn't happened (even though we've had to work hard at times). If there are open issues, let's solve those issues. We can solve them rapidly and efficiently.

However, If you do want to try out alternates, that's fine, start today! The wiki-model allows you to do it even. ;-) Just set up your system and get to it and have fun! As long as you don't interfere with other methods, you'll do fine. Besides, a policy patrol might do wonders at times. :-)

• Wiki-model works great most of the time.
• If it breaks, fix it. :-)
• If you want to do something else, you're allowed to start today even.
  • as long as you play nice with others, who might use a different system

--Kim Bruning (talk) 02:33, 27 April 2008 (UTC) How many folks here have read Excession? The Culture ships are organized by principles which are strikingly similar to wikipedia consensus. :)

I have two separate proposals.

• Radical proposal: As I suggested above, we could introduce a "Wikipedia Council" of 40 members, elected every year by approval voting (the same system as that used for ArbCom). They would approve policy changes by vote, acting effectively as a legislature.
  • Pros: Would ensure that our policies were fixed and definite, therefore avoiding confusion for new users. Would enable much-needed policy changes to be efficiently implemented, without getting bogged down in massive strawpolls and trying to build consensus from thousands of users.
  • Cons: Would require the creation of a new process, and would involve a selective departure from the traditional wiki-consensus model.
• Moderate proposal: Create Wikipedia:Requests for policy change, a centralised discussion forum for proposing new policy or changes to existing policy. Discussions could be structured rather like an RfA.
  • Pros: Would enable centralised discussion rather than spreading it over many talk pages, so community-wide consensus would be easier to determine. Would not move away from the traditional consensus method.
  • Cons: Who would close the discussions and determine consensus? Also, would not eliminate the problems we have with the consensus model.

Discuss. Walton^One 08:43, 27 April 2008 (UTC)

Discuss, yes; here and now, no. Let's gather a few more proposals first, shall we? So, anyone else? Waltham, The Duke of 12:22, 27 April 2008 (UTC)

Support the new governance proposal

As the drafter of this proposed policy points out, the current policy governance model is broken. I propose a Policy Committee (PolCom), with nine members elected to one year terms, who will govern our polices. The PolCom will make/approve policy changes, upgrade guidelines to policy when and if necessary, and advise on policy application and interpretation, such as currently occurs at the different noticeboards, like the Reliable Sources Noticeboard. Thus, the policy pages would be completely protected (locked-down), and only a PolCom member would be allowed to edit them. Editors who want to make a policy change would present their argument on the PolCom page, and the PolCom would debate it and do a straight-up vote. Cla68 (talk) 05:11, 27 April 2008 (UTC)

Hmm, PolCom may be a better name than "Wikipedia Council" (the name I proposed, which may sound rather pretentious). Perhaps we should take a strawpoll on what to call it (then we have, at least, hammered out one of the details of this proposal). But I disagree that it should be only 9 members; I proposed 40, because I think we need a broader range of viewpoints and community input. We don't want to develop a cabal, and I see no good reason to have so few members. Walton^One 08:37, 27 April 2008 (UTC)

Here's a simple reason why I will never support this (as I noted above, I am a proponent of direct democratic system): If the PolCom will makes few decisions per year, you can as well vote directly about what you want. On the other hand, if PolCom makes a lot of decisions per year, it will become difficult to keep track if the candidates are acting in your interest. Just think about it, and you will see that voting directly about the issues you care about is easier and more in your interest. Samohyl Jan (talk) 08:53, 27 April 2008 (UTC)

The number on the board doesn't matter, IMO. I chose nine just to choose a number. As far as for tracking how the candidates are voting...so be it if it's hard to track precisely. Look at a few decisions, read some outside comments on the how the PolCom is acting, such as will probably occur on WR, and recognize some general trends. That should be good enough for most people, it is for me. Cla68 (talk) 12:13, 27 April 2008 (UTC)

In general, I like your ideas. I think the number needs to be higher than nine, I'd say into the the twenties, at least. I want the people on the Committee to be well rounded people; continue as article content contributor and have a life outside of Wikipedia. The workload will be better distributed with 21-25 people, I think. I'm not sure about acting as advisers on noticeboards. I'll need to give that some more thought. I do think that our policy pages should be protected and new drafting of them worked in a work space. The name change to something more related to policy is fine with me also. It will keep the Committee focused on this work. FloNight♥♥♥ 16:15, 27 April 2008 (UTC)

Yes, I definitely think we need a larger group than nine; burnout will likely be as significant an issue here as it is on ArbCom, so we need enough leeway that a few people going on break won't cripple the committee. Kirill 17:22, 27 April 2008 (UTC)

Not a bad idea. The only reason I would object to Samohyl Jan's proposal is that, as I think we all know, the only people voting in his/her proposal would be those who are either policy wonks or among those who would likely be very committed to one side or the other of the proposed policy. We would certainly have a distorted view of the opinion of the community, one way or another, if that were to happen. And then there are the matters of creating socks for the purpose of piling up votes, and all the other problems of direct democracy in general, and internet democracy in particular. John Carter (talk) 17:53, 27 April 2008 (UTC)

You're saying "we all know", but this isn't obvious. I would specifically want the rules so there would be designated page about the upcoming votings. So anyone could watch just this one page and decide to vote on any issue he would like to. How you, on the other hand, guarantee that those on the committee won't be policy wonks? If people are not interested in the issue enough to show up for voting, they are probably not interested in the issue, that simple it is. Direct democracy may have it's problems, but I have yet to hear about a problem with direct democracy that doesn't manifest much more in the indirect democracy (and I would like to hear of real-world evidence, not a feeling - 8 years ago everybody got "feeling" that Wikipedia will be complete disaster; in fact you could use the same argument - how you guarantee that not only crazy wonks or those committed to one side will edit the articles?). Samohyl Jan (talk) 19:12, 27 April 2008 (UTC)

Actually, if you looked at the existing elections for ArbCom, RfAdmin, RfB, etc., I think that "we all know" people don't vote is already a fairly clearly indicated statement. Look how comparatively little attention any of them ever get. However, I suppose that you could argue that, as someone who has never expressed an opinion in any such matters, but are making comments here, that that might not hold in all cases. And it probably doesn't. But demographics I think tend to back up the statement that people who don't get involved tend to stay not invovled. Regarding ensuring that policy wonks don't get on the committee, I don't think that's the case at all. I think the committee probably would be composed of people who are concerned with policy. And, to an extent, I can't see a problem with some editors who are active elsewhere, to varying degrees, and are also probably generally knowledgable about policy, being the ones to take part in such discussions. Like Sonny Bono heard in response to a question he asked why so many congressmen were lawyers, it's because what they do in Congress is, well, write laws. As long as they welcome input from others regarding the subjects under discussion, like ArbCom, which I think all the proposals above do, no one would be disenfranchised, and we would tend to make sure that the single-purpose voters don't get overrepresented. John Carter (talk) 19:29, 27 April 2008 (UTC)

I don't care about most of the policies, only very few of them (I am a very casual user, and now even more so, since I have seen some horrible things related to forming of Wikipedia policies). I don't care about who is on ArbCom or who is Admin, because I don't know these people in most cases and it's hard to keep track of what they are doing, so it's hard to say if they should be elected or not (also, I would like to note there is a fundamental difference between voting for someone and voting for something, as in the first case, you have to trust the person). And precisely because I am casual user and I trust community in matters that are of no interest to me, I wish for direct democracy. I don't want to select someone who is deciding instead of me - I would have to check up on these people, if they are actually doing work representing me. I don't know if you actually want committee to always represent interests of whomever voted them in. If not, then it's elitist and I certainly disagree with that. If yes, then it's just more overhead; just imagine if only a selected few on Wikipedia could edit articles, and people would have to come to that someone and politely ask to make the change. It would be insane - and yet it is precisely the same thing you propose with this committee. Samohyl Jan (talk) 22:51, 27 April 2008 (UTC)

What if you only had to appoint one trusted person, and yet they (and their appointees, if any) did all the work of representing you? And you could fire your representative at any time, or overrule their decision on something, without needing to get other voters' consent? That is what liquid democracy is all about. Chin Chill-A Eat Mor Rodents (talk) 23:25, 27 April 2008 (UTC)

I think this is interesting idea in theory, but in practice I can't imagine that. What you are saying that decision could retroactively change if enough voters would disagree with their representative? And also, who would keep track on who's who representative and that this information is accurate (I ask because I have seen an administrator being a member of "admins for admin recall club", voting against admin recall in the poll, and when I pointed it out, he said, well, sorry, that was old information; so if people don't keep track of their own stuff, how do you expect them to keep track of other's stuff)? It seems overly complex system to me. I would prefer to have regular referenda about change in a given policy say each 6 months, during which any changes requested by other people could be voted about. So if you would care about particular policy, you would have to show up twice a year. That's no big hassle. Anyway, now that I am thinking about it - direct democracy seems to be a special case of liquid democracy (because, in direct democracy, there are no representatives). So if people would feel liquid democracy is too much of an overhead, they would failover to direct democracy, so it would be a good thing. I would perhaps need to see more concrete proposal to consider it further. Samohyl Jan (talk) 06:40, 28 April 2008 (UTC)

Usually what people do is create a table of who is who's representative. Every citizen has the right to participate personally in decisions, and if both the citizen and his representative vote, then the representative's vote counts only for himself (not for both him and the citizen). You're right that if no one appointed a representative, it would just revert to direct democracy. The corporate world has used systems like this for a long time and methods have been developed of checking credentials, etc. See proxy voting. Chin Chill-A Eat Mor Rodents (talk) 12:54, 28 April 2008 (UTC)

Walton, Wikipedia Council would be a bad idea, because it would create confusion with the WikiProject Council.

Samohyl, a basic element of the problem we are encountering is the sheer number of involved editors trying to be heard in the various policy-making discussions; even if the body in question makes relatively few decisions, they will still be made in a much more effective way than they are done now.

FloNight, I quite agree; if we are to place our trust in a certain group of people, and give them the power to handle our policies, we must ensure that they will be in constant touch with the community, and not adopt a more insular and isolated mentality (as Wikipedia editors in general have been said to have, by the way, but this is a completely different matter). Waltham, The Duke of 18:13, 27 April 2008 (UTC)

Waltham: Show me this sheer number of editors involved. And if they are trying to make prescriptive policy, perhaps we should dissuade them? --Kim Bruning (talk) 23:33, 27 April 2008 (UTC)

I support this proposal, because something must be done about the ownership of policy pages, the endless edit warring to force through special agendas, the resulting confusion and the time this drains from article writing. Nine is far too small, though, and members should be elected, and terms should be no longer than one year: these measures should help with the ownership issues and yield more stable and reasonable policy pages. Further, if this model is enacted and works well on policy pages, it might be extended as a method to solve the interminable disputes at MoS. SandyGeorgia (Talk) 18:27, 27 April 2008 (UTC)

I fear that this particular medicine is worse than the malady. Groups of users have picked up some bad habits, preventing people from normal editing of policy pages. Admins are actually (unintentionally) complicit in this, because they have a tendency to do page protections and thus hand their victory to the edit warriors and filibusters on a silver platter.

I think a better solution might be to somehow act directly against such users, and at the same time to better educate admins in how to deal with such situations.

--Kim Bruning (talk) 23:32, 27 April 2008 (UTC)

By having one year terms, it makes the memembers of the PolCom more accountable to the community for their decisions. I believe that the number of members should be an odd number to help break ties in those cases where all members cast a vote. How about 21 members? I believe the current system used to vote for ArbCom members would work here. Cla68 (talk) 01:21, 28 April 2008 (UTC)

Votes are a bottleneck. Can you make the system work without them? --Kim Bruning (talk) 01:32, 28 April 2008 (UTC)

To what are they a bottleneck? I see that they prevent a small group from making rapid changes, but some sort of formalized opinion-gathering process seems very responsible to me, as a way to tell whether the small gorup active on the page represents broader opinion. — Carl (CBM · talk) 01:37, 28 April 2008 (UTC)

They make a lot of wikipedia editors come to one single place, which is the leading cause of wikidrama. For the same reason, they also require a lot of extra layers of organization. All in all, they place a palpable strain on the community.

If you can avoid a vote at all, no matter how, do so!

Instead of trying to shoehorn in voting one way or the other, try one of many other methods that have already been pioneered on wiki, to get your people together. :-)

--Kim Bruning (talk) 01:46, 28 April 2008 (UTC)

Needless to say, placing complete authority over our policies in the hands of a small group is very serious. As you know, policies govern what can go into articles, how we edit articles, and how we administer the project. Thus, a process that invites full community participation in an orderly and fair process is very important. The ArbCom election process does this with one exception- giving Jimbo final appointment authority. I recommend that we use the ArbCom voting mechanism, but take Jimbo out of the process. Cla68 (talk) 02:01, 28 April 2008 (UTC)

Ah, and therein lies the misunderstanding. Our policies do not govern anything. Our policies do not govern. They document.

This is, in fact, policy!

I'm not sure that what I just said will convince you of anything yet, but maybe it'll given you an inkling that there is more between heaven and earth that you may wish to encompass in your philosophy? --Kim Bruning (talk) 02:12, 28 April 2008 (UTC)

Votes are not a bottleneck, and anything that brings more eyes to any page is not a bad thing. Anything that "breaks" authority so that it's in the hands of the many instead of the hands of the few in the end is inherently a good thing. If it "breaks" our traditional wiki-way of doing things or alters it in a paradigm shift over time, this is not an inherently bad thing either. It's evolution. Evolution is good, because to continue to survive, evolution is required. If this ends up making some of us not matter to the ruling and leadership of Wikipedia, well, the dinosaurs and hairy elephants went away... and the world kept on doing fine. Lawrence Cohen § t/e 13:27, 28 April 2008 (UTC)

Indeed. Our fundamental goal is the building of the encyclopedia; our policy system is merely a means to that end. Some people seem set on maintaining traditional approaches in perpetuity only because they're traditional, not because they're necessarily the best ones for the task. I think we need to avoid being dogmatic on such points; if changing the existing system will allow the project to function better, then we should not shy away from doing so. Kirill 13:44, 28 April 2008 (UTC)

Another proposal: Polling on binary issues

Here's another suggestion: We poll on issues, advertising through the watchlist. For example, the recent rollback drama might have gone easier if we had started from the position of a poll with everyone accepting a two-third majority and a minimum input of 100 editors. Note this wouldn't be the method for writing the policy or guidance, this would just be getting community assent to write it. For example:

Poll on whether spoiler tags should be used in the encyclopedia. If they shouldn't, nothig changes. If they should, interested parties work up the guidance per WP:CONSENSUS. And so on and so forth. Can also work backwards, allowing editors to work up a proposal and choose to have a poll on the acceptance of it. This shouldn't be obligatory. Hiding T 14:45, 27 April 2008 (UTC)

Personally, no objections. And, in fact, part of what is proposed above was already included in the proposed PolCom, based more or less on the existing ArbCom setup. Clearly, even in the more extreme setups, I can't see how, if a process were proposed and rejected by the majority of editors, it would stand a chance of passage. The one exception being if a policy dealing with a particular subject were "very much requested" by the legal office or some similar entity. John Carter (talk) 15:20, 27 April 2008 (UTC)

Policies, Guidelines, and Constitution

I found it remarkable that even Kim, in the statements above, wasn't apparently able to clearly differentiate between a policy and a guideline. If a person who did as much work on most of them as Kim did can't, we have a real problem here. I think that we might benefit from having, at least initially, some group of editors, either formally chosen or informal participants, going through the existing policies and guidelines, trying to come up with definitions of the terms as relevant here, and maybe at the same time creating some sort of structure to the policies that emerge. Functionally, this wouldn't be that much different from a written constitution. As a possible preamble, stating our most core policy, it might start something like this:

"Wikipedia is an organization dedicated to the dissemination of reliable, verifiable information, presented in a neutral way, on subjects of encyclopedic merit. All of the policies and guidelines this organization has exist to further this one, central, goal."

By doing something like this, we would indicate what the most core "policies" (if that's how the word will ultimately be defined) of the project are, and maybe go on from there into the comparative "details" of policy and guidelines. John Carter (talk) 15:20, 27 April 2008 (UTC)

"I think that we might benefit from having, at least initially, some group of editors, either formally chosen or informal participants...". Ahem. Quis custodiet ipsos custodes? If possible—and I have grave doubts that it is—let's pick noncontroversial souls, if anything similar is enacted. Ling.Nut (talk) 16:06, 27 April 2008 (UTC)

Valid point, said the Watchmen fan. But the initial proposal was only for the possible creation of some sort of definition and review process. Arrogant as I am, I don't want to try to define the details before anyone else has even agreed to the raw proposal yet. John Carter (talk) 16:17, 27 April 2008 (UTC)

I don't differentiate between P/G/E at all in fact. I was one of the holdouts on that particular consensus. I'm still holding out good-natured-ly. :-) Are you telling me that you actually might want to come (partway) my way on this? --Kim Bruning (talk) 18:24, 27 April 2008 (UTC) *hardly dares hope*

(edit conflict) The impression I have formed, Mr Carter, is that people here are not sure whether the proposal should apply on policies alone or on both policies and guidelines. However, the difference between policies and guidelines is one of applicability (fewer exceptions from policies are acceptable than from guidelines) and modification (changes can be made easier to guidelines than to policies). And, of course, policies are more widely accepted and constitute a more important part of the project's character. Apart from these... Well, what kind of differences are you expecting to find?

On another note, I hope your comment on arrogance is not some kind of provocation; I shouldn't like to engage in a contest now—although we could certainly schedule it for a later time... Waltham, The Duke of 18:32, 27 April 2008 (UTC)

In response to Kim, I'm not sure what s/he is talking about in terms of meeting partway. (sorry, but in the US, where I'm from, I've seen the name used for both genders). I never saw this proposal, and I doubt anyone else did either, as being a "regular job" requiring lots of time, just a way to fix some problems in a way that they will actually be addressed by some sort of policy or guideline, when to date we haven't been able to generate either in some cases where they seem to at least some, like the ArbCom, to be necessary.

To the Duke, unfortunately, I can't find that stupid white glove right now, but I will keep looking. ;) Regarding the difference between policys, guidelines, etc., your statement about the differentiation between the two, and about the question about whether this reform should address both, is probably accurate. I get the impression that this discussion is getting to the point where there should be additional pages detailing the specific proposals and variations being put forward. But, before that would make a lot of sense, it would probably help if we could all agree that there would be cause for the various proposals to be spelled out and discussed separately. So far as I can tell, there might be as many as three people who might be counted as completely opposed to these reforms, Kim Bruning, Kurt Weber, and Lar, and all the others agreeable to at least some possible variation on at least one of the proposals. Do any of the rest of you think that there is just cause to try to elucidate the various proposals at this point, or would that still be too premature? John Carter (talk) 19:00, 27 April 2008 (UTC)

Based on my continued learning/understanding of the actual proposal made on the main page, if I am understanding it, I think I'm also opposed to it as it stands.

And if I am still confused, to clarify then: I am opposed to the creation of a legislature of any kind. I support the idea of a policy review committee, simply because the concept, at least, would seem to mesh well with other Wikipedia processes of editor or page review. That said, I'm thinking now that perhaps I should present it as a separate proposal, as this page's intended proposal seems to be (as it's named, and I apparently wasn't paying enough attemtion) based on the idea of creating a body for Wikipedia governance. So at the moment, I guess you can add my name (shudder) to the opposers. - jc37 19:29, 27 April 2008 (UTC)

For what it's worth, I was considering your possible proposal as one of them. Whether its a "legislature" (which is a potentially loaded word) or just an ArbCom like body, which I think is what most people are agreeing to, might be a matter of semantics. But I was very much thinking that your potential proposal be one of those to be written up and considered. John Carter (talk) 19:33, 27 April 2008 (UTC)

Fair enough. As I said, I'm still not sure what the initial proposers (KL and CM) had in mind when proposing this. - jc37 19:39, 27 April 2008 (UTC)

For what it's worth, KL's early comments indicate he wasn't real sure himself. John Carter (talk) 20:07, 27 April 2008 (UTC)

Alternative voting systems for an assembly

Approval voting may not be the ideal voting system. Its main advantage is that it is cheap to implement, and marginally better than the bloc voting normally used in multi-candidate U.S. elections (where the ballot says "vote for not more than n"); but since this is an online community, the programming issues are not as complicated as they would be in, say, an electoral district having to implement new software on many precincts' voting machines.

Single transferable vote would allow voters to rank candidates in an order of preference, which could help the results more accurately reflect the will of the electorate. Interactive representation would allow each member of the assembly to cast a number of votes equal to the number of votes by which they were elected. So, say, if Mr. X barely gets elected to the assembly, squeaking by with only 50 votes, and Mr. Y gets elected by 150 votes, then Mr. Y's influence over decisions will be three times Mr. X's.

There are many other possible systems as well. As mentioned above, liquid democracy would allow users to have representatives, without the need for a formal assembly at all. Chin Chill-A Eat Mor Rodents (talk) 17:31, 27 April 2008 (UTC)

You're proposing we downgrade from consensus? (because it's too complicated or so?) --Kim Bruning (talk) 18:25, 27 April 2008 (UTC) wait, I know who you are.... :-)

How to create it (start with a shadow legislature)

I am thinking, any editor or group of editors could create a sort of advisory shadow legislature immediately, without going through the proposal process. Giving it actual powers could be done through a proposal, or the powers could evolve informally through a gradual building of consensus. Of course, if the legislature were created and started giving bad advice, then consensus presumably would not form to give it powers, and no harm would be done. It would be very similar to the proposed process for creation and evolution of a UN Parliamentary Assembly. Chin Chill-A Eat Mor Rodents (talk) 19:35, 27 April 2008 (UTC)

This is basically my idea. The point is to create a deliberative body that can function when the scale is large. It is an old problem, and most of the old solutions tend to break down in various ways. What has never been tried, though, after the scale has already become large, is to form a deliberative body deliberatively, without first setting up a representative system to reduce the scale. And most people -- including political scientists -- will, at first glance, think this impossible under the conditions we face. However, there are certain techniques that can be tried, and as CC notes, there is one which does no harm if it fails. The problem with traditional systems involving elected representation is that (1) unless STV or other proportional representation systems are used, the representation tends to get rather warped and (2) it involves bureaucracy, election systems, and complications. However, there are certain principles we can derive from traditional forms, while hewing more closely -- much more closely -- to the direct participation that built Wikipedia.

Suppose those who want to form an assembly, instead of trying to create a community consensus first, just form it. Any registered editor may join and, initially, all have identical rights. However, from tradition, any assembly may develop its own rules for participation. These rules are special to the assembly and don't affect anything outside. Again, by tradition, the majority may protect itself, just as it may consider it wise to likewise protect the minority; otherwise the consensus measured by votes in the assembly becomes warped, no longer truly representative of the membership.

As to how the assembly proceeds, I could -- and have -- come up with suggestions. Delegable proxy, and the related Asset Voting, are ideas (the latter is quite old, and delegable proxy is really only a formalization of what happens in many peer organizations informally) as to how to compose a truly representative assembly without contested elections and bureaucracy. But, ultimately, the assembly itself determines its own rules, and it is sovereign over them. My suggestions are just that -- it happens to be a problem that I've spent thirty years on, but that only generates advice, not authority.

And if all the assembly does is to generate deliberated advice, binding on no individual or other body, together with a measurement or estimation of the consensus the advice enjoys within the assembly, it cannot harm.

So what is the difference between this and what we currently have? Well, for starters, we don't have any pure deliberative environments set up. ArbComm is close to it, in some ways: imagine, actually gathering evidence and arguments before voting! In deliberative bodies there are no votes (on ordinary motions) until a whole process has been completed, including, as the last step, a decision that it is time to vote. At that point, generally, every reasonable argument, known to any participating member, has been presented. What we have now, instead, is that someone makes a proposal (such as an AfD), and, immediately -- even without a second -- voting begins, with debate mixed in. So early votes may not have seen the evidence, nor considered the arguments. It's no wonder we distrust voting. We are often looking at knee-jerk votes, made without consideration of the arguments.

This immediate assembly, formed in advance of any consensus to create it, can cut through the thicket of noise that envelops any serious proposal for change. It can create NPOV reports on issues (don't we already know how to do that?). (Note that a report of a vote, properly done, is NPOV. Votes in this kind of context are facts, not opinions, generally. Facts do not control outcome, rather, they may be used as a part of advice; hopefully, our decisions are advised by facts! But by judgment as well.)

And the first business of the assembly will be to determine its own process. That will take time. There is a tendency here to start something and if it doesn't work or complete a task within a short period, we mark it as failed and move on. However, there are centuries of experience about how to undertake this task; it merely requires adapting the experience and precedents to the special needs (and special opportunities) of on-line process. Real-time meetings are probably out of order at the beginning, for example. Large deliberative bodies deal with complex issues through referral to committee, so then a crucial issue comes to be how committees are formed.

What will be noticed is that committees, again, don't properly make decisions; rather they collect testimony and evidence and consider proposals and then report back to the full body. This is the classic solution to the problem of scale: reduce the scale for full deliberation to a representative body, in this case a committee. Delegable proxy can, in theory, form ad-hoc committees that are broadly representative.

But delegable proxy also allows constant maintenance of true representation, if proxy assignments can be changed at any time. Further, to the extent that proxy assignments are public rather than secret -- I vastly prefer open systems -- direct voting can remain possible in any process, for the problems of scale really only impact deliberation. Most members will sensibly not vote on an issue when they have a trusted representative participating who takes the time to research it; that representative can communicate directly with the member if the member doesn't understand the rep's position, and it is my expectation that usually representative votes will roughly represent the conclusions of those represented after communication over the issue has taken place. This is why I describe proxy expansions as estimating consensus. For advisory purposes it is not necessary to have an exact measurement. Pretty clearly, if, for example, a vote is close enough that a few votes this way or that changes the "result," we aren't seeing a consensus, but a situation where the community has not yet found agreement. Strictly, the "result" is the vote itself, not a decision. Decisions are made by those advised.

Maybe later, the community could assign decision-making power to the assembly. But I'd prefer, actually, the present model. The assembly would advise; it would advise the community of editors, first and foremost, through deliberative discovery of consensus, and it would advise the servants of the community, being the administrators, bureaucrats, founder, and board members of the Foundation, and none of these are bound, per se, by that advice. But if the consensus measured was strong and broadly representative, and the deliberative process thorough, any of those advised could see that and would neglect the consensus at their own peril. Yet if they think the process flawed, they retain the right of decision according to IAR; what they decide is how they use their own personal power; the Foundation board decides similarly, subject as well to the constraints of law, for they also represent the State of Florida.

"Legislature" is the wrong term. What would be started would simply be a meeting of the community, an "assembly," and it would have no legislative power, with only one exception: it can make its own rules for its own process. It would presumably make these rules, tentatively, by majority vote. And if it does so abusively, there is a huge protection. It should be clear at the outset that it is possible to have more than one assembly. In the end, though, the assembly that is most representative and that has the fairest rules will probably absorb the others. (But sub-assemblies may still exist, they are what I call "caucuses." They each have their own rules, and they can restrict membership in ways that a full assembly would not.) "Extra" assemblies won't be formed, I'd predict, unless intractable disagreements appear over process. If most participants understand that the goal is broad representation and the discovery of true consensus, I don't think that we will see more than a few split-offs attempts, if any. If the full assembly rules are good enough, there won't be a need. But the possibility could help keep the assembly honest.

And where should this assembly exist? Because it should be free to make its own rules, I highly recommend it be off-wiki; the simplest form is a mailing list or (as in on experiment I've been working on), a family of mailing lists, being a top-level list and then committee lists. I've used Yahoogroups for this in the past, because the tools are adequate, and it's free. The decision, really, belongs to the person bold enough to actually start the Assembly and solicit participation. I've done that in the past, many times. Here, I'd rather see it be started by someone else. The person starting the list is the list Owner, and is thus, properly, a trustee, and is advised by the list. When I've started such lists, I've pledged to follow the list consensus. That gives me a certain right to, particularly at the beginning, disregard an apparent majority if I feel that it isn't representative. And, of course, the list members can take their marbles elsewhere: if a delegable proxy system is set up, with a tradition that a proxy appointment and acceptance involves an exchange of direct contact information, the list members, if enough of them have named/accepted proxies, actually *can* move elsewhere, even if the gatekeeper, the Owner, doesn't like it and bans them.

(An Assembly off-wiki could still have on-wiki pages, and an on-wiki membership list, but those would be supplemental and not critical. The independence of the Assembly is crucial. It must be free to make its own rules, which rules may, in some cases, conflict with Wikipedia rules, and at least one example is obvious: if the Assembly becomes large, it may restrict who can post directly to the proceedings of the Assembly; if this restriction is impossible, the problem of scale will impact the Assembly just as it impacts process here.)

So who is going to start the first Assembly? Be aware, there are elements here which are quite hostile to the idea, and some of these editors have administrative tools and are not shy about using them, given an excuse. I don't recommend that a new editor try it. But, hey, I'll consider anything .... it's easy to merge Assemblies of the kind I've described, so what is really important is that the process start. Failure, in fact, of any given attempt, can be success in the long run, if the next attempt is informed by the first failure. The true failure is to assume that the whole thing is impossible. --Abd (talk) 16:58, 28 April 2008 (UTC)

If you create an assembly that fails and is MFD'ed, then it creates a precedent allowing speedy deletion under G4 of the next assembly someone creates. So, there isn't really an opportunity to incrementally improve the idea if that happpens. If we stuck to the wiki way that Kim suggests for handling wiki-organizations, I don't think that would be an issue. It basically sounds like freedom of association in that people can just start their little committee and do what they want, as long as there's manpower to keep it going. Chin Chill-A Eat Mor Rodents (talk) 17:09, 28 April 2008 (UTC)

Sure, about the MfD. That's one reason why you don't start on-wiki. If you have an off-wiki "Assembly" -- and note that a number already exist, though I'm not aware of any that are functioning as deliberative bodies, though I think there might be one that was set up by ArbComm fairly recently, restricted membership, appointed -- it can't be deleted or obstructed, except by joining it and trying to stop it that way, which isn't likely to succeed. (Unless truly cogent arguments are presented.) If it starts to get some participation, then it's possible to create on-wiki connecting structures that, for example, invite participation, describe what decisions have been made so far (and how to try to change them if you don't agree with them), etc. Deleting those on-wiki links would be less than effective and would probably accelerate the process. Esperanza and AMA, besides their own internal problems which led to sluggishness and lack of enthusiasm (which may well have been temporary), were totally vulnerable to MfD. Off-wiki, they would have faced only the natural enemy of apathy. Which can be quite bad enough. --Abd (talk) 20:17, 28 April 2008 (UTC)

Hey abd your sockpuppet demonstrated his hand too quickly this time:

my prediction: http://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Governance_reform&diff=208321611&oldid=208320414

What happens: http://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Governance_reform&diff=208573896&oldid=208573416

Who turns out to be: http://en.wikipedia.org/wiki/User:Chin_Chill-A_Eat_Mor_Rodents --87.112.64.32 (talk) 00:13, 29 April 2008 (UTC)

Proposing creation of draft pages for the various individual proposals

The discussion here, particularly the more repetitive parts, are getting in the way of the various ideas put forward. I suggest that we consider the possibility of reform in some way to have received sufficient impetus for the various proposals suggested to be written up individually as separate proposals. Jc37 has indicated a proposal which I believe has merit, and others have as well. I think it would make sense for the discussion on this page from this point forward to be limited to discussion of the idea of reform in general. Discussion of particular variations on reform, including any that haven't been proposed yet, could probably best be made on the various separate pages for each proposal. I am in the process of starting one such page myself, and will put in a link to it at the bottom of the current page. I think that would be the easiest way to ensure that the discussion regarding this idea remains comprehensible. John Carter (talk) 21:02, 27 April 2008 (UTC)

I think it might be a little too soon for that. I have started a thread above ("My head..."), where I have requested that the participants in this conversation should state their proposals. Perhaps this (expanded with further proposals) could do for the moment? It will be easier to try and find common ground between the various views this way; if some of the participants here could come, through discussion, to a single proposal instead of each of them posting their own, there would be less clutter and the proposals would have more support from the beginning. Consolidation, I think they call it. Waltham, The Duke of 21:20, 27 April 2008 (UTC)

Moving forward rapidly

While I'm convinced that the basic architecture is still sound, there are several things that can be done to make day-to-day work rather less frustrating for people. At the same time, I wouldn't want to actually block or compete with other maintenance mechanisms. The more the merrier :-) So here's my first unified proposal.

A group is formed.

(In whatever way we can sort out. Possibly "Wikiproject tidy project documentation"? :-) You could probably pick a less boring name. ;-P

The requirements for this group are at least

• scalability (a requirement for all new wiki-processes)
  • Monkeysphere. Be mindful of Dunbar's number
  • Choke points. Make sure that there is no central council or overused central page that can become a chokepoint. Make sure you never go over 75 people (dunbar/2, 50% redundancy) at any one location. If you can keep it under 10 or so, that's even better. If you can make it 1, perfect. If you can make it 0 (software task) we shall erect an altar and worship you. (see also: Burn out, Mediawiki)
  • Burn out. Keep workload for individuals low. Make it easy to split tasks up into small chunks or steps, that can be carried out by any volunteer at any time. Tasks that can be handled by software should be.
  • Simplicity. Try to keep tasks as simple and as rapid to carry out as possible.
  • simplicity 2. Try to keep the number of steps in the process as low as possible. Preferably very close to 1.
  • Mediawiki. Try to "go with the grain". Figure out tasks in such a way that mediawiki already partially supports what you want to do (dividing people into subgroups with pages, reporting on issues via templates and looking them up in categories, etc). Then the human side of the task becomes much easier.
  • Flexibility: The wiki environment is constantly changing. Keep structures in flux, simple, and easy to change, so that you can adjust your group to changes on the wiki or in real life. Don't become a dinosaur like AMA or Esperanza!
• transparency
  • People should be able to see what the group does at all times
  • it should be easy to join.
  • it should be easy to leave.
  • it should be easy to be an observer.
  • it should be easy to look up past actions
  • it should be easy to change the way the group works.
• speed
  • Decision cycle. Be mindful of the OODA loop
  • Keep it short. Make your decision cycles short. The perfect is the enemy of the good. It is more important to decide on time. Come back later to improve things further.
  • Take the initiative. act inside the decision cycle of others, take the initiative
  • Keep the initiative. act quickly enough that others can't get inside your decision cycle, or you lose the initiative
  • Morale is good. A solid tempo boosts morale. A slow tempo harms morale

Maintain the Consensus system properly.

At the same time, the normal consensus/wiki-editing process could also use a boost. Just like a democracy, consensus needs maintenance, else it falls apart just as badly. In fact, consensus probably needs a little more maintenance than a democratic system, since it requires slightly more intelligence and education to use. (Not so much that people aren't able to learn, but enough to be noticeable :-) )

• Find better methods to deal with the occasional special interest groups, edit warriors and filibusters. They're ruining it for the rest of us. Educate admins on these methods.
• The consensus system is tricky to learn at first, Make better documentation available (this might be a good first job for the doumentation wikiproject)
• Provide other (multimedia) course materials to bring people up to speed.
• Make projects, rewards and incentives for people who do some of the above
• We need to show a little pride in our systems, and maintain them accordingly. :-)

Resolve underlying issues.

I think there are some unspoken underlying issues as well.

• Too much pressure on arbcom
  • Explain consensus process and the requirement to assume good faith in more detail, so that people can learn to avoid disputes upfront.
  • Make the dispute resolution system more accessible.
    • People shouldn't wait to call in an editor assistant for advice. They should call them in early and often. Editor assitants are like the GPs in medicine, a gateway to further help.
    • Explain how dispute resolution works often!
  • Teach people how to mediate themselves. If everyone has some mediation skill, at the very least if a conflict breaks out, there will always be someone on hand to help deal with things
  • The mediation cabal is currently basically running a mediation school. Take your small conflicts to them, and/or join for a while, to give these people a chance to learn! If you are experienced, help out inexperienced people. If you are inexperienced, ask for help and learn!
• BLP and None-free content
  • You can't force things from above on a wiki. Understand and respect this
  • Policy talk pages act like a giant WP:BRD net. Have people "on staff" there at all times, and you can use those talk pages to educate new people on the importance of these policies
  • Hand off: In the worst case we can transwiki tricky cases
    • commons: is better equipped and trained to handle Free images. Worst case ban image uploads to en.wikipedia, and leave the job to commons.
    • wikinews: is better equipped and trained to deal with Biographies of living persons. Worst case ban biographies of living persons on wikipedia proper, and leave them to wikinews.
• It is impossible to develop new top down policy.
  • Good. That's because you're not supposed to ;-)
  • The current policy mechanism is structured for bottom up policy change, not top down.
  • If you want to change policy, change the way you behave, and convince others to do the same.
    • When this fails, it fails mostly due to people not understanding the system. Patiently explain how policy change starts withWP:IAR, WP:BOLD, and why you're allowed to as per WP:CCC etc... (yes, there was a reason we had those. ;-) )

--Kim Bruning (talk) 00:20, 28 April 2008 (UTC) FINE then. If you want to make a group, at least do it properly. None of this pussyfooting around! ;-)

Along the lines of making it "easy to change the way the group works" might we implement freedom of association? Let editors form whatever little mini-cabals they want, to experiment with different systems. Certain editors seem to have to tendency to want to charge through the crowd saying, "Break it up, break it up" (e.g. by MFD'ing a wiki-organization) which seems a bit contrary to wiki concepts of self-organization and self-regulation. This proposed legislature could very well fall victim to a similar attempt to crush it, or it might be destroyed by the community while still in an embryonic stage, before anyone has a chance to see how well (or badly) it would have worked. It shouldn't be that we can only form groups that the community has gone through a bureaucratic process of officially sanctioning. Chin Chill-A Eat Mor Rodents (talk) 00:39, 28 April 2008 (UTC)

Technically we already allow all that. Maybe 2 or more competing paradigms might work, but you do need enough people to man them all! --Kim Bruning (talk) 00:45, 28 April 2008 (UTC)

If the community (or a vocal subset thereof) thinks that your group's work is unhelpful to the encyclopedia, though, don't they typically try to crush and salt it? (E.g., WP:Esperanza, WP:AMA) Who's to say that if someone tried to start a legislature, that they wouldn't attack it (through MFD, or by placing a rejected tag on it) for the same reasons (i.e. "too much bureaucracy," etc.) Chin Chill-A Eat Mor Rodents (talk) 00:57, 28 April 2008 (UTC)

Yes, no, it depends. Design it right and they won't. --Kim Bruning (talk) 01:01, 28 April 2008 (UTC)

Sure. Be God, get it right the first time. So what's the problem? WP:PRX was a proposal to simply set up proxy files and see what happens. An experiment, no policy changes. There was no specific application that was actually part of the proposal, merely some suggestions of how some editors might use the information generated. What was wrong with it? Well, it was "REJECTED." By whom? Two editors, basically, Kim "God" Bruning, and Mangojuice. Kim's rejection was perhaps, prescient, as befits his middle name. Mangojuice, however, soon joined by some others, radically misrepresented the concept, and what was rejected was, largely, "voting." Which wasn't part of the proposal itself. So, sure, if we can somehow come up with the perfect proposal.... lotus born .... nah, I still think it would be rejected. I decided this about myself long ago: if God himself presented me with the perfect idea, I'd reject it. I'm sure I'd find something wrong with it. It takes time to understand new ideas, usually. If they were really so obvious that everyone would immediately accept them, first time they hear about it, we'd already be doing it. On the other hand, if we can just figure out the perfect presentation, and surely God would get it absolutely right, with every possible phony argument presented and refuted .... oops! nobody would read it. Too long. --Abd (talk) 17:10, 28 April 2008 (UTC)

So ask someone who knows what they're doing? :-) --Kim Bruning (talk) 19:42, 28 April 2008 (UTC)

(multiple ec's later) For what it's worth, I generally agree with Kim. There does seem to be a bit too much work thrown at ArbCom in particular, and having the other processes be more active would help a lot. By which, I mean quicker responses. Unfortunately, I for one ain't even close to being mediator material. It also might help if there were a perhaps more clearly "rigid" formalized process of progression there, so that people don't go to ArbCom as soon as they're annoyed. Personally, don't know much anything about wikinews, so can't say anything there one way or another. The only point I would likely dispute is the "top down don't work" point, only because I don't think that, if it was structured right, the group in question would ever do that sort of thing, except maybe by "persuasion" by Jimbo, the legal office, maybe ArbCom, situations like that. Regarding the group structure, I think a parallel to ArbCom, maybe with regular reports in the Signpost, would match the criteria Kim laid out. Basically, based on the other data, around 30-50 people, probably formally or informally broken into groups, should meet those criteria. There might be problems getting enough people for the "rapid-response" factor to work, but it should be doable. Regarding Chin's points, if a system works, and it works well, I don't think the special interest opponents would have enough support to win a delete. Check to see how many times the XfD pages themselves have been nominated for deletion. If the bureaucracy is minimal, and the results are good, the enemies of the system will almost certainly be outnumbered by its supporters. John Carter (talk) 01:03, 28 April 2008 (UTC)

I'd like to add "Anything relating to Wikipedia:Notability" to the list of underlying issues. It's probably the most controversial page with a checkmark at the top, and typically generates at least one arbcom case every few months, all the way back to 2005. Nifboy (talk) 19:15, 28 April 2008 (UTC)

Possibility of binding content mediation

Why do I have a feeling that this is going to get a lot of negative reaction? Anyway, as ArbCom has specifically said on several occasions, they do not deal with content questions, but behavior questions. That's fine, no reason to change that. But it does not address the fact that there often are questions regarding content which lead to articles being locked, editors being blocked or banned, and the like. Presumably, in most such cases, both sides are trying to present their positions accurately. The question, however, still remains. Maybe something like this could be proposed:

• (1) After a given discussion has been through RfCs, and other such attempts to resolve disputes, if the problem still remains, it is likely that the article will be at least somewhat locked for some time regarding the dispute. After or during one such lock, it might be possible and useful to propose that the disputing parties agree to binding mediation regarding the content in question. Such binding mediation would not necessarily be an "official" process. However, if parties on both sides agree that they will accept the results of the binding content mediation, then the mediated outcome could be seen as being binding, at least in regards to the evidence presented up to that time.
• (2) Under such a setup, the parties involved in the dispute would present their positions and evidence to support their positions to the mediation panel. We're probably thinking about a group of about the size of ArbCom. Individual members of such a panel who could reasonably be seen as having a potential conflict of interests for whatever reason could be asked to recuse themselves from certain discussions, but there would probably still be enough members to present a reasonable body to hear the case.
• (3) This body would review the evidence presented by those who have already been involved in the discussion, and any other parties who they might ask to present information to them, and, eventually, decide based on the evidence presented how the content should be adjusted, if at all. They could also indicate, if the case is simple enough, what kinds of evidence would be required for the content regarding a particular point of dispute to be reasonably changed. Should evidence that qualifies be produced, then there would be just cause to change. Also, of course, if the material related to the content of the article changes subsequently in a broader sense, by having new material relating to the subject come out, then the content in general would be of course subject to change.

By creating the group as an unofficial group whose decisions the disputants acknowledge to be binding, we might be able to dodge the "formal policy" problems, while still providing a way to resolve the matter at hand. And, considering that such a process would only be necessary in the case of a seemingly intractable dispute, I think the chances of it being used are ultimately rather good. Anyway, like I said in the beginning, I expect some criticism of the idea, because I know it is still only a basic one. Feel free to fire away. John Carter (talk) 14:09, 28 April 2008 (UTC)

The roads saga might be instructive in this area as a lesson/case study. (some may say a case study on how NOT to resolve content matters :) ... there was significant deviation from the consensus process, by consensus, or alternatively, by fiat, depending on who you talk to, in some parts of the resolution process) I'll try to dig up a few pointers for further reading if there's interest. ++Lar: t/c 14:40, 28 April 2008 (UTC)

The concept of binding mediation crops up in mediation circles from time to time. You could try canvas medcom and medcab and 3O about this. :-) They might bite. O:-) --Kim Bruning (talk) 17:19, 28 April 2008 (UTC)

I think binding content mediation would be good. Right now the best we can do is hope that a non-binding solution sticks or wait for people to start being disruptive so arbcom will handle it. Mr.Z-man 23:51, 28 April 2008 (UTC)

Policy patrol

Wikipedia:Policy_patrol is currently unused. It meets many of the above requirements, except there's no real drive behind it. Could we extend that concept, or get a couple more vigorous editors on board? --Kim Bruning (talk) 17:16, 28 April 2008 (UTC)

heh it has a talk with Kim first rule still in there? That's nice for an early start, but that might need some changing. :-) <scratches head> --Kim Bruning (talk) 17:21, 28 April 2008 (UTC)

Why we need this - a case study

An excellent example of community consensus based decision failing to scale is the "discussion" that, among other things, resulted in the "rollback" right being implemented.

What happenened?

Feel free to fill in any gaps in the timeline

• Wikipedia:Rollback for non-administrators - an RFC-type discussion with about 20 different views. 65KB of discussion on the main page, 23KB talk.
• Wikipedia:Rollback for non-administrators proposal - the first proposal. After 92KB of talk page discussion, didn't really get anywhere.
• Wikipedia:Non-administrator rollback - this is where most of the discussion took place, multiple straw polls and alternative proposals. 96KB on the main page, 353KB on the talk page (including archive)
• Wikipedia:Non-administrator rollback/Poll - the main straw poll, split from the main page after closing. 455 "voters" - 215KB
• Wikipedia:Non-administrator rollback/Archived proposals - 2 alternate proposals split from the main page. 47KB
• bugzilla:12534 - Bug request that resulted in it being turned on, 17 comments.
• Wikipedia:Administrators' noticeboard/Archive121#Rollbackery - The first AN thread, 11KB
• Wikipedia:Administrators' noticeboard/Archive121#Wikipedia:Requests for rollback - discussion began shortly after the right was added. Multiple comments by Jimbo and a statement by Anthere. 82KB
• Rollback consensus RFAR - 35 statements, 73KB
• Wikipedia talk:Requests for rollback/Archive 4 - the talk page of the current process page shortly after it was implemented, about 4 days in January. 212KB
• Wikipedia:Miscellany for deletion/Wikipedia:Requests for rollback - MFD of the process page, withdrawn after 9 minutes of discussion, 7KB
• Wikipedia:Requests for rollback/Draft poll - Plan for a second poll. 23KB on the page, 35KB talk.
• Wikipedia:Requests for rollback/Vote - another poll, this time a straight vote. 31KB and 153KB talk.
• Wikipedia:Rollback policy - failed proposal for a rollback policy, never got off the ground.

The results

The rollbacker right was implemented with admins able to grant and remove it and a stable process on WP:RFR was formed. After only 1.5MB of discussion, polling, voting, and arbcom - a new process (the policy proposal failed) was created. A slow motion edit war over whether Wikipedia:Requests for rollback was a disputed process carried on for a few weeks after the discussion calmed down and the Wikimedia shell users are unwilling to do things for enwiki after being threatened with ArbCom sanctions for turning on rollback.

Discussion

Now this may be the ideal way to add/change policies and processes, but reviewing it all, I don't think it is and it certainly isn't what's described at Wikipedia:Consensus. Mr.Z-man 20:40, 28 April 2008 (UTC)

I agree. Our current way of adding new policy is far less than ideal as this example shows. This is a good example of what I meant when I said that our current process of forming policy is making too many people have really bad days. :-( While any process is going to lead to some disagreement, I think the need for dramatic actions to make change happen brings about needless discourse. FloNight♥♥♥ 20:51, 28 April 2008 (UTC)

The vote on rollback I began is a great example of how people can kill things.[1] Shortly after it was clear that consensus was heading for rollback to go live, period, but with just a question of which form, the Vote was nuked from orbit. If the Vote had ended up being 300 for rollback with process, 200 for rollback with no process, and 50 against, it would have been clear that we had consensus for some form of rollback. We could have then done a simple final runoff vote between the two versions if needed, and done. But I'm convinced that some people are opposed to this sort of thing because it means that no one gets to personally be in charge of something, to decide consensus or own process or policy pages. The rollback saga is a fantastic example of the need for reform as outlined by Kirill. See also Wikipedia:Three revert rule enforcement for an example of how policy can come into being as the end-process of this idea. Lawrence Cohen § t/e 21:29, 28 April 2008 (UTC)

The rollback saga and ATT saga are examples of what happens when people act in utter ignorance and contempt of the existent system. These two cases went against all documentation.

Any new system you put in place will be treated in the same way. There's nothing wrong with having a system. The problem is in making sure everyone knows and respects and works with the system.

This is the major problem with wikipedia. We're the new usenet, and or eternal september is dawning. We need to push that back as hard as we can, by helping, teaching, mediating, and educating. We don't do that by sitting around making up ever more elaborate ivory towers. :-)

We can work on this project or not, but in the mean time, Who here is willing to help me? There's plenty of work to go around.

--Kim Bruning (talk) 22:13, 28 April 2008 (UTC)

Kim, we can certainly push to educate new people, but the problem is that many people seem to agree that the old way itself is what is broken in regards to policy matters. IAR applies to our traditional ways of thought, as well. If something is preventing us from improving the encyclopedia, we ignore it. The classic methodology appears to be preventing us from improving policy to benefit the encyclopedia. Rollback, if my vote had been allowed to run, would have been fine. It would have been a clearly binary view of what consensus really was from a high altitude, just as the 3RR vote worked fine. ATT would have ran fine for it's vote process, had Jimbo not decided to interfere and override the community (which was headed to the same conclusion apparently as him: no consensus).

Would you say Wikipedia:Three revert rule enforcement was a success? Why, or why not? If voting is so wrong, why do we vote for arbiters and the WMF board rather than discussing it? Why did we vote repeatedly on the main page redesign? Were those successful? Wikipedia seemed to have survived them fine... Lawrence Cohen § t/e 22:29, 28 April 2008 (UTC)

(edit conflict)

ATT was headed for no consensus because it was so abysmally handled. It was not handled according to known best practices. If something is handled in a known pessimal way and it breaks: Do we blame the manual that says not to do that? Or do we blame ourselves?

Sure votes like that occasionally work, but then again occasionally votes like that asplode. The safety record could use some work ;-)

(And no, I intensely dislike the 3RR and the 3RR vote. 3RR gets gamed more than practically anything else on the wiki. Rule: "Hard Rules" are for gaming and trolling.

Note that Arbcom and WMF board are outside the wiki system. This is a very very old story: Summary: Different system, different tools. Right tool for right job. Voting on wiki-encyclopedia: BAD PLAN (1+1=3 gets 200 votes in support!). Voting elsewhere: DEPENDS.

I can expand on either point if you like. )

--Kim Bruning (talk) 22:45, 28 April 2008 (UTC) Did I mention that such votes are NOT according to best practice? Don't blame the current practice. Blame people not reading it. :-)

Best practice is clearly broken and owned by a small subset of users. It's time to take the 2003-2004 best practice out behind the toolshed and shoot it in the head. Anything that allows a tiny subsection of users to control the system is broken. Lawrence Cohen § t/e 23:19, 28 April 2008 (UTC)

There were plenty of non-vote discussions in the rollback saga. The problem is twofold. 1) The sysadmins can't be bothered to read 10,000 words of threaded discussion to determine if there is a consensus. 2) Drive-by commentors who make a comment after seeing an RFC or a village pump notice then never return to discuss. The latter is more of a problem as far as the current system is concerned. Mr.Z-man 23:40, 28 April 2008 (UTC)

Re Lawrence Cohen's comment of 21:29. Some people who OWN various policies do so in good faith. Another downside of the consensus system for policies is that, if a change isn't discussed or removed, it is presumed to have consensus. This means that, when someone randomly edits a policy page in some way that doesn't have consensus, someone has the unpleasant task of undoing it at some point. But any significant change that is broadly advertised is unlikely to gain consensus in the first place.

An example of this is a recent change to NFCC#8; the person who changed the policy did so in good faith, and thought the change had consensus, but it didn't. It took a very long time before another user reverted the change. But really there hasn't been consensus for the present version of NFCC for some time.

Personally, I would be very happy to delegate my voice in shaping policy to a committee if I knew everyone else was doing so as well, and knew that I would have a voice in selecting the committee. — Carl (CBM · talk) 22:35, 28 April 2008 (UTC)

That would be another example of having policy gatekeepers in a format organized by the community. Just take all the present pages tagged policy, put them under their stewardship (which amounts to them playing gatekeeper for the community) and then our community-delegated group would mediate the changes out for us. Policies are supposed to be somewhat static over time, but not because tiny handfuls of individuals control them. This would have the great effect of taking policy control away from any one or five or ten individuals, and giving it wholly to the "people". Lawrence Cohen § t/e 22:40, 28 April 2008 (UTC)

In the case of NFCC, 7 editors (inadvertently) hogged the policy, and the people took it back none too gently. :-) --Kim Bruning (talk) 22:51, 28 April 2008 (UTC) I'm not sure how handing power to a small group == handing power to the people. Is this the new newspeak? :-)

Carl: You People failed to follow documented policy, and suffered the documented consequences. I can imagine you folks might be somewhat unhappy about that. --Kim Bruning (talk) 22:51, 28 April 2008 (UTC) I'm starting to get the feeling that this talk page is manned mostly by people who didn't follow or didn't understand documented process, and got burned somewhere along the way. Is this true?

I wasn't actually involved in the NFCC change, until after someone else had reverted it, at which point I felt obligated to advocate for the free-content side of things. So I wasn't personally upset by it, but I can see how the people who advocated the change would feel put out. Also, I have to disagree there is "documented policy" on how to make policy. There isn't even consensus whether policies are normative or not, much less on how to make them. — Carl (CBM · talk) 22:58, 28 April 2008 (UTC)

Ah sorry, did I mistake you for someone else? Apologies. (fixed).

These people felt put out because they didn't read what would happen if they worked that way. Who is to blame? I blame myself, for not being able to get that information to them on time. I only did a little bit of mediation post-hoc, which is technically already too late, of course.

There is no consensus on whether policies are normative or not? heh. Either they're normative, and consensus is a figment of our imagination, or they are not, and wikipedia is run by consensus. You decide. :-) Also, there's currently a document on how to create policy, and there are all the normal editing guidelines and essays, which also apply equally to policy. What kind of information are you missing? --Kim Bruning (talk) 23:07, 28 April 2008 (UTC) This is why I feel that we should be working on the education aspect of this so heavily. Every time and every where we see that people don't understand how to use the systems at their disposal. Things can certainly be made more more clear and more obvious. --Kim Bruning (talk) 23:07, 28 April 2008 (UTC)

Kim, we know how to use the systems at our disposal. I've personally suggested policy changes myself. The problem is that we (collectively) keep finding situations where a tiny group is self-empowered to stop any forward progress. Coupled with policy talk being limited to "insiders" who normally stay on the policy pages, it causes massive problems to ever do forward motion that has wide community support. You can't get 200 people to weigh in on WP:N's talk page. But if you had a central place... you could. For a true consensus that a minority of users couldn't game or stop. Lawrence Cohen § t/e 23:17, 28 April 2008 (UTC)

I don't understand. Can you show me this insidious minority someplace? :-)

In the mean time, 200 >> 150. Dunbar's number is for social groups what the critical mass is for uranium. You don't want to go over that value. :-)

--Kim Bruning (talk) 23:21, 28 April 2008 (UTC) And you can't legislate different values for e, π, or Dunbar's number

Wikipedia talk:Spoiler: A minority of users rammed through a change to eliminate spoilers. WP:ATT was drafted and forced through by a minority of users. Rollback was rammed through by a minority of users. Wikipedia:Private correspondence was nuked by a very small number of users. Any one of these could have turned out quite, quite differently had they received a wide communal review. Not an insidious minority, but no minority let alone one user (Jimmy included, who no longer owns this site, and matters less now) is entitled to the ability to stop a policy change. See my Example below. No one small tiny group could stop a policy change like that if the community endorsed it. Lawrence Cohen § t/e 23:31, 28 April 2008 (UTC)

what's stopping you from bringing those back up again? :-) --Kim Bruning (talk) 00:31, 29 April 2008 (UTC)

He probably would rather not be involved in another month's worth of discussion that will likely end up confirming the status quo, as happens to most significant policy discussions. Mr.Z-man 01:06, 29 April 2008 (UTC)

Exactly, and this is due to a new discussion being 99% of the same people. Its just like us trying to sway Kim here. He appears to be unwilling to be swayed, and if that is the case, if he said so, we would be spared a month of trying, and could move on to convincing other editors of our perceived need for this change. Would that be bad faith, to ignore Kim and move on to others, and hoping to build a consensus that bypasses him? This is why a system like this is needed. No games, no waste of time discussion for the sake of discussion. Lawrence Cohen § t/e 13:27, 29 April 2008 (UTC)

That's totally unfair. I've actually helped out by listing the basic requirements that you need to meet to have a viable wiki-group. How about instead of accusing me of stonewalling, we start thinking of ways to get all those to fit together, eh? :-) Now who is stonewalling whom? Start out with a small set of wiki-pages, and note how people behave on those. Notice how mediawiki already covers most of the requirements? Now without going against the grain too much, start blocking out your system. Do you need help with that? (or is this going to end up like my discussions with Esperanza about their need to reform?) --Kim Bruning (talk) 17:46, 29 April 2008 (UTC) And yes folks have made viable wiki-groups before in recent history, and yes they meet those requirements, and groups that didn't make the cut did get MFD'd... so no whining, more working! No one said it's easy, but you say you want to do it. So less whining, more working!  :-P )

Lets try the focused side of things. Take a look at the new section I posted below--post your own understanding/pro/con, so we can understand and address point by point your concerns. And do a response to my own section, please. :) Lawrence Cohen § t/e 17:55, 29 April 2008 (UTC)

The problem with CCC is that it is very badly defined (I have the same problem with what you call decision-making by consensus - there is no formal process, so everybody can claim they are right). Just imagine a following example: 50 people have some policy page in their watchlist, and are happy with it. Each 2 days, a different person comes over the policy page and says, I disagree with the policy and want such and such change. Each time, some of these 50 people will respond - there is consensus for that now. This happens for 200 days. So, on one side, each individual that wanted change was clearly overruled by 50 individuals watching the page. On the other hand, there was 100 of such individuals, so actually more people support the change than there are people happy with the current policy. Now tell me, which side is right. Has the consensus changed or not? Samohyl Jan (talk) 07:34, 29 April 2008 (UTC)

Consensus changed on day 1. There's two approaches. The first is that the person who goes away watchlists the page, and when a 2nd person comes along, they now can second that person... so the pair of them are less likely to be sent away.

The other approach is to directly WP:BRD, which allows you to talk with each of those 50 people, one by one, and convince each one that your method is better. (Though that might take up to 50 days. Fortunately, in real life you typically only need to convince a handful of people, so you might need less than a week of evenings, say).

Or you can combine the two approaches. And you can draw in other people who use the same method to come and discuss. And other people will slowly start to show up.

The methods you can use are not vague at all. Each of those pages has actual white-line points where you can know exactly what to do, and what will likely happen next.

Does that make sense so far? --Kim Bruning (talk) 17:57, 29 April 2008 (UTC)

Discussion section break 1

I think the rollback saga is actually a better argument for more formal polling procedures than the creation of a policy commission. As Lawrence says, had we had proper poll procedures, the mess largely wouldn't have happened. As to all the scrabblings after the devs implemented it, a lot of that was consensus editing. Some people felt there should be another poll, and through consensus editing the idea was considered and dismissed. Consensus, isn't pretty, but at times it gets where it is going. I believe people will more readily accept decisions made through this sort of manner than those handed down by some sort of committee. Remember the userbox saga? Consensus editing has pretty much solved that. I think the problem you are trying to solve is only a problem if you take the view that a decision has to be made on everything immediately. As to filibustering, there's been a lot of that at WP:FICT recently, but in the last couple of days I think people just realised they could simply edit the thing rather than continually answer and rebut objections, and it is now moving a lot faster. Eventually it will reflect consensus. I don't know when, but is there a deadline? Hiding T 12:05, 29 April 2008 (UTC)

I think that many people only dislike polls and voting even as a barometer of consensus is because it limits the ability to skew or manipulate the "perception" to whatever our desired ends are. Thats why we in AFD or RFA tend to attack responses they disagree with. Not to sway the poster (which almost rarely happens), but to make the poster's ideas look unappealing to other !voters. Conducting a poll or vote minimizes the power of each participant to being just one-of-many, and heavens forbid everyone be equal. That would be bad. ;) Lawrence Cohen § t/e 13:31, 29 April 2008 (UTC)

I agree completely. My experience in discussions with people about direct democracy shows that they have problem understanding democracy at all, and they often expect either that someone enlightened will decide in things in their favor, or that they are part of the enlightened few who have natural right to decide for others. Samohyl Jan (talk) 15:40, 29 April 2008 (UTC)

Um, you realise that you've listed proof of the exact opposite above, don't you? Two polls on rollback, both ending in 67% support, one declared no consensus and one declared consensus. Some of the best disputes on Wikipedia are over how to interpret poll results. Have you looked at the way the pass %'s have changed at RFA and RFB over the years? Dickering over poll results and trying to skew them is in the top ten of what Wikipedians do best. Hiding T 20:41, 29 April 2008 (UTC)

Oh, and polls don't make everyone equal on Wikipedia. Have you voted in every poll that is currently open? Polls make everyone aware of the vote and who vote equal. Consensus is the only method which makes everyone equal, because at any moment at any time I can change anything. With a poll, you miss the poll and it's done. Bye bye, tough luck, because we had a poll and our survey said, and that's what we're doing. You can't change that now, and no we're not redoing the poll because the result would be the same. Polls are evil. Hiding T 20:44, 29 April 2008 (UTC)

(Hmm, and here I thought you might be talking about XfD, as some people view it : ) - jc37 05:55, 30 April 2008 (UTC)