Wikipedia talk:Governance reform/Archive 4

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

Example, from how the ideas feel today

If I'm following what the supporters of reform generally are in agreement over...

I feel User-conduct RFCS should be certified by 3 or 4 involved people, rather than 2, to help cut down on frivolous and useless RFCs. I post my request and reasoning to someplace like WP:Proposed policy changes. A notification of my request goes to the User conduct RFC talk page linking back to the discussion I initiated. The Group, or Delegates, weigh in with their opinions but do not have the ability "nix" or automatically enforce my change. The general community then weighs in on the discussion. If there is clear public support for my change on the proposal page, after x amount of time, the discussion is tagged "success", and one of the Group makes the appropriate changes, and you now have a new standard for User conduct RFCs. If my idea fails, the policy and it's enforcement stays the same. Any users attempting to edit the policy page to change it arbitrarily would be reverted back--the policy page is thus "bound" strictly to wide consensus in a central forum. The Group are essentially selected for their policy experience to offer opinion and policy analysis but have no authority but to revert out and protect policy from forced attempts at change outside a wide community acceptance.

Am I getting this right? Lawrence Cohen § t/e 22:52, 28 April 2008 (UTC)

You would have to be very careful. There's a major flaw in the "discuss first" paradigm when applied to wikipedia. It provides a loophole for filibuster. People tried modeling a consensus process with enforced discuss first, and they failed to find one that did not have this flaw.

So if you want to enforce that, you will need to flowchart your process carefully to be sure that you don't get bitten by that same bug.

so ... I think your proposal as it stands basically hands the keys to the wiki to the filibusters... :-/ --Kim Bruning (talk) 00:25, 29 April 2008 (UTC) in fact, in large lines this process looks pretty much like one of the "enforced filibuster discussion" models we were working on. ^^;; I orignally thought those models would turn out to be a bit inefficient, but that that might be a survivable compromise. Instead, they turned out to be unworkable.

But that's the case with the present model as well; any time you want to change an enforced policy, you need to discuss the change beforehand.

The key, I think, is limiting the ability of small groups of obstinate editors to tire everyone out, driving them away from the discussion. We need a method by which we can readily cause the community to make a decision on a proposal, rather than dragging "discussion" on interminably. Kirill ^(P) 00:34, 29 April 2008 (UTC)

Basically I have found and documented that this is not true. You can typically break through the barriers that people erect. Once you have found the people blocking a situation and negotiated with them, you can then continue using more regular wiki-editing.

--Kim Bruning (talk) 00:55, 29 April 2008 (UTC) part of doing BRD right is to get people to stop talking about policy this and policy that, and start talking about what they actually think is the best outcome (you can practice this concept by mediating a lot of disputes). As they approach that point, consensus slowly begins to work normally again.

If your first bold move gets you desysopped or banned, that tends to make further discussion rather difficult, at best. Not everything is so lacking in prescriptive provisions as you would have people believe; there are definitely parts of policy that one chooses to ignore at one's own peril. Kirill ^(P) 00:58, 29 April 2008 (UTC)

(ec) I'll cover the ignore at one's peril a bit later, because that's very interesting. But before then, this bit has me concerned much more: Desysopping and banning are the responsibility of the arbitration committee. You yourself therefore already have a measure of control and responsibility over whether the process works or not. Did you realize the extent of that before starting this page? --Kim Bruning (talk) 01:21, 29 April 2008 (UTC)

Oh yes; why do you think I started it? For at least as long as I've been on the Committee, the policy formation process has been very broken; almost ubiquitously, attempts to introduce significant changes to core policy—policy dealing with the use of admin tools in particular, but that's by no means the only case—have resulted in wide-spread edit-warring, wheel-warring, and other forms of highly disruptive conflict, with the predictable bannings and desysoppings following as a result. Further, several key areas of policy remain undefined or contradicted by other policy, turning things like the community ban process into a game of Russian roulette for any admin that uses their tools in those contexts.

(The alternative to what I've proposed here, I suspect, would be increasing imposition of policy changes by way of ArbCom fiat, as we've already done in a number of cases. I think, personally, that a more democratic and less ad-hoc method would be better for the long-term health of the project.) Kirill ^(P) 01:36, 29 April 2008 (UTC)

I think we already covered that. Don't impose changes. The system is designed to resist that. (aka. "Descriptive not Prescriptive"), My mentor who taught me that still happens to be on the arbcom mailing list. Did you think to ask him for advice? :-) There is also documentation that states that doing this kind of thing is most unwise. Descriptive policy really is that, descriptive. It can be very powerful indeed at times. Even the arbcom can get into trouble if it doesn't follow it, let alone normal editors. And indeed, it sounds like that is exactly what happened. Flaunting the basic principles caused much drama. --Kim Bruning (talk) 01:55, 29 April 2008 (UTC) ignore policy at your peril indeed!

Amusing, but I don't really subscribe to the "Wikipedia policy is perfect and needs no changes" school of thought. Much of it is like a leaky old boat; it kind of works in the nominal case, but starts sinking in the slightest rough spots.

(In any case, ArbCom has had little trouble imposing changes; it's more that we don't really want to be in that position. It's other editors' attempts at causing change that tend to blow up in their faces.) Kirill ^(P) 02:07, 29 April 2008 (UTC)

Heh, I'll be the last person to ever say that policy is perfect and doesn't need changing. Wikipedia exists in a changing world, so policy needs to adapt with it. The interesting thing is that some people actually would like to keep policy static instead.

I hope the arbcom hasn't had too much recent experience with imposing changes. :-P And I agree that editors should learn how to alter things in a safe way. <scratches head> I think it's easy because I've been doing it for ages. You're saying it's too difficult for other folks? --Kim Bruning (talk) 02:57, 29 April 2008 (UTC)

When are the last three times you made a significant change of a policy on a non-technical matter, that would have had behavioral impacts on Wikipedia, and received push-back from other editors? How long did the push-back last? Lawrence Cohen § t/e 13:32, 29 April 2008 (UTC)

Hopefully never. You have it backwards. I don't want to impact behaviour on wikipedia, I want to document which behaviors work.

Other than that, define significant edits? Sometimes improved understanding can be expressed in a single line (like removing the "non-negotiable" text from all policies... people thought it meant they could "non negotiate" their way out of situations :-P You're probably laughing now, but boy was that annoying to mediate! Definitely significantly improves my blood-pressure, that.) , sometimes a page needs a cleanup and has been significantly rewritten (like Wikipedia:Consensus). And sometimes convincing people to change their behavior (such as allowing non-latin Wikipedia:Usernames) allows for improved cooperation between wiki's.

Remember, these things just document best practices. The intent is not to change people's behavior directly... (though if documenting best practices teaches people better ways to achieve their aims, then that's what it's for eh? :-) ) --Kim Bruning (talk) 14:46, 29 April 2008 (UTC)

Er... I think we're just having fundamental disconnects here as you tend to stay on the meta side of issues and a lot of us have fought over behavioral control policies--thats what they are--like WP:CIV, the banning and blocking policies, PRIVATE; various content (the most important policies of all as they effect our articles) issues like Notability, ATT, SPOILER, Fiction, and the grand poohbahs, NPOV and BLP, probably our most important policies (try implying that BLP is not non-negotiable, see how far that gets you). Have you ever tried to implement any changes to policy on any possibly controversial policies? Or just the fluffy ones? No offense. Lawrence Cohen § t/e 15:02, 29 April 2008 (UTC)

And, Kim, has anyone ever blockaded your policy changes, in the past? When? How did you resolve it? Provide example, please. Lawrence Cohen § t/e 15:16, 29 April 2008 (UTC)

Break 1

When was the last time BRD worked for any significant change to a policy? It tends to either be BRRRRRR or BRDRDRDRDRDRDR. Mr.Z-man 01:08, 29 April 2008 (UTC)

Today. :-) FT2 applied BRD, and I did major edits in return. (or you can look at it backwards, I did major edits, and FT2 reverted... the situation is more fluid with experienced editors.) I conceded mostly to FT2. Check our edit logs. --Kim Bruning (talk) 01:21, 29 April 2008 (UTC)

That was on an obscure technical description page, not a core policy. I rather doubt even someone of your persuasive talents could freely change something like the blocking policy in a significant way and not get caught up in either edit-warring or interminable debate; and the average editor has no chance at all. Kirill ^(P) 02:10, 29 April 2008 (UTC)

Well, this is basically a near-identical request to the rollback permission one. This time I figured I'd get in on it early and prevent a redo. ;-)

But alright. A core policy? I managed to make a change in the NPOV policy recently. Does that count?

But you're saying it's too hard for people? It seems to be easy enough for other people too, at least, whenever I'm around. <scratches head> Would my presence be influencing my observations? --Kim Bruning (talk) 02:57, 29 April 2008 (UTC)

Actually, I think your presence would influence things. You're profoundly on the "descriptive policy" side of things; when you change a policy page, it is almost certain to be a mere change of the text, not preparation for any sort of enforcement of the newly-written policy, and I expect it is regarded as such. To put it quite simply, I don't think you're seen as a threat.

Changes to things like the various administrative policies, on the other hand, usually come down to a question like "Do we block for X?" (regardless of how it's worded); and it's typically quite obvious that the question is not merely a theoretical one. So we immediately have at least two groups (the presumptive blockers and the presumptively blocked) which are on opposite sides of the matter, and which have a very direct and practical stake in the outcome. It's pretty much the same thing with the other major policies governing administrative actions (which, at this point, function essentially as normative ones, regardless of what the theoretical underpinnings might be), as well as some of the more-useful-as-hammers-in-disputes content policies (which are also becoming increasingly enforced as normative standards). Kirill ^(P) 03:15, 29 April 2008 (UTC)

I've been involved in WP:BRD at WP:CIV the last couple of days. I remember rewriting WP:SOCK a while back. Um... there's been some BRD at WP:DPR and WP:DGFA involving me recently. Back in March me and Jc37 managed to capture consensus with a change to WP:NOR such that sourcing is not always complicated. Are those helpful? Hiding T 11:54, 29 April 2008 (UTC)

Well, I am involved in changing policy though. Generally when doing one kind of dispute resolution or another, folks come across diverse issues where the policies are unclear or simply give bad advice. We then discuss better ways of doing things. Later when the dispute or problem is resolved, someone will tend to write down how to prevent similar problems in future. Sometimes that's me. And actually that's how I got into maintaining policy you see. Now I guess you're seeing the same things? But you're going about it with a rather grand gesture, I must say. :-) --Kim Bruning (talk) 15:08, 29 April 2008 (UTC) I've also been known to tell off admins who do something stupid :-P

(Outdent) Do you all notice the general deference that the community gives to the ArbCom? In my opinion, this is because the community recognizes the need for a certain few to have decisive authority in dealing with certain matters in order to maintain, for lack of a better term, law and order in the project. One reason that the community is willing to give deference to the ArbCom is because the members are elected in a fairly democratic process in which every vote, support or oppose, has equal weight. The same idea should work for a PolCom. I believe that a democratically elected committee comprised of a sufficient number of members will enjoy community support in managing the policies and considering community inputs for changes and modifications. Cla68 (talk) 13:42, 29 April 2008 (UTC)

I share this view, Cla68. Based on my interaction with the Community, I think the majority of the Community would welcome a more organized way to create and review Wikipedia policy. The Policy Committee's role would be to oversee Policy related matters. The Community should still play a large role in policy development and review but now an organized system would be in place to make policy changes. If done well this Committee will make routine policy changes discussions more effective and less contentious, and the result will be policy that the majority of the Community understands to be an accurate statement of the way that Wikipedia works. FloNight♥♥♥ 14:09, 29 April 2008 (UTC)

You have high hopes. I've seen committees operate in the wiki-world before. I'm not particularly impressed by them. Are you sure they won't make the situation worse, rather than better? The community gives deference to arbcom, because the arbcom follows the rules set by the community. If it goes too far outside that remit, things start to get shaky. (a similar story applies to medcom).

A fun fact is that you can insert just about any subsystem into the community in a similar way (so it's not impossible to add a new subsystem). Just if it's slow and ponderous and doesn't keep up, it can just as easily slide downhill (vide WP:AMA, and even at one point almost WP:MEDCOM, until they vigorously tidied up their act.) Committees don't have infinite credit, and I'm not sure how much of that credit comes from being elected. --Kim Bruning (talk) 15:52, 29 April 2008 (UTC) Why listen to me? Well, check out the AMA debates for instance. Basically the old medcab jumped in in style :-) . Nicholas Turnbull was trying to reform AMA, I was covering for Editor Assistance, a new scalable replacement. Either solution would be a win (in the end EA won out). Also note the closers in the MFD debates. ;-) We adhered strictly to policy, and were a neutral 3rd party, but obviously we were very involved.

Month long polls

Here's an idea. Any wikidrama which blows up over a binary issue should be put to a month long poll, poll closed by a bureaucrat and results chewed over by editors. A month seems long enough for people to work out a consensus on the issue through editing so much so that the results of the poll will be fairly obvious. Because really, want causes the problem in the first place is the drama, not the individual opinions. Everyone wants to have their input. Give them a poll and a month to express it, who knows what will happen. And it doesn't harm the consensus method. It actually strengthens it, because after the month there'll be an indication of the direction to take. We could have a little chart for how to interpret polls, like:

• 90% or above support -> This is the way to go.
• 80 - 90% support -> This is the way to go, slowly.
• 70-80% support -> This may be the way to go, with caveats.
• 60-70% support -> This is nearly the way we may be going, but people going the other way need to be compromised with. We're going off at a related tangent instead.
• 50-60% support ->Draw up a page expressing both sides of the situation, with the one with majority at the top, making it clear both views are equal and editors shouldn't edit war to get their preferred version.

Yes, it's half assed, but it's the germ of an idea. Hiding T 12:13, 29 April 2008 (UTC)

There is a role for polling in conjunction with a new Policy Committee. The Policy Committee would work with the Community to develop better ways to create and review policy. Polls can have a role in assisting with determining policy. But having stand alone polls be the final determinate is a problem. Polling as now done on Wikipedia is too often flawed to find consensus. Many issues that seem binary are not. For them to work there needs to be a long organized discussion period prior to the poll. Then a well thought out poll written. These efforts take someone knowledgeable about the specific issues, existing Wikipedia policy and guidelines. Polls can be a good tool if used properly. Having a Committee make use of them when appropriate would be a good idea. FloNight♥♥♥ 13:39, 29 April 2008 (UTC)

Decision cycles need to be fast. Can you think of methods to speed up your decision cycle considerably? If you're going to actually make the decision cycle take longer, then not only is there less heat, there's also less light (as well as more room for issues to sneak through the gaps). So that'd be somewhat unhelpful. Can you think of a way to get your own (proposed) decision cycles either a) faster, or b) get them to produce less "heat" (conflict) without slowing them down?

Also, would this method supplement in-the-field policy making, or would it supplant it, in your opinion? --Kim Bruning (talk) 16:04, 29 April 2008 (UTC) The question is really how quickly can we get feedback on actions back to the people taking them. And how do we ensure that feedback is summarized and documented. hmm..

Why do decision cycles need to be fast? Thats not the point of this, at least I've read it. An RFA-physical layout to the page, perhaps, with a week for feedback, and a week for polling, to toss out an example. That's about as long as any policy discussion I've seen is useful and consensus is almost always clear on a change in under a month. Lawrence Cohen § t/e 16:21, 29 April 2008 (UTC)

The reason why decision cycles need to be fast is the subject of stacks of literature rather too large to supply here. You can start at OODA loop and keep on reading from there. ;-) Obviously, if your decision cycle is too long, you'll tend to "miss the boat", make a decision after the problem is long gone. if your cycle is short enough, then you can rely on making small incremental improvements, rather than large world-shifting improvements all the time, That reduces stress. Finally in a real-world environment, the organization of system with the shortest decision cycle holds the initiative. The strategic importance of initiative is left for another day. :-) --Kim Bruning (talk) 16:42, 29 April 2008 (UTC)

With an organized review of policy the need for changes can be noted sooner and different approaches to addressing the problems can be initiated before a crisis develops. Currently, too often changes are made in the middle of difficult situations and do not stick because they are reactionary. FloNight♥♥♥ 16:52, 29 April 2008 (UTC)

Well, fair enough, I can live with more upfront review work. At the same time, inflexibility in the face of crisis can lead to an organization breaking, not bending. So we need to at least retain flexibility. Does that make sense? --Kim Bruning (talk) 17:11, 29 April 2008 (UTC)

Kim, are you serious? On one side, you argue that "there is no need for change current processes" and "policies are descriptive". On the other hand, you say "decision cycles need to be fast". What exactly do you want to accomplish by discussing here? Is it just me, or do also other people feel you are just trolling? Samohyl Jan (talk) 16:25, 29 April 2008 (UTC)

I have a feeling this is all stonewalling to prevent the old-timers from having less power on Wikipedia control matters. :( Lawrence Cohen § t/e 16:35, 29 April 2008 (UTC)

On the one hand, I am arguing to at least retain the current system alongside anything new we make, at least as a backup. On the other hand, if people positively must make a new system, then we had better darn well do the best job we can (see also#Moving_forward_rapidly), so I'm contributing to that side of the discussion too, rather than simply stonewalling it.

I believe this approach shows that I'm willing to cooperate and shows a will to compromise and find middle ground (and in fact I'm doing it that way quite deliberately, to make sure that I'm not stonewalling, because I hate it just as much when people do it to me :-P ). Would you be willing to withdraw your accusation, or do you wish to let it stand? --Kim Bruning (talk) 16:45, 29 April 2008 (UTC)

Thanks for explaining, Kim. Any system change designed to specifically take power away from the 'minority' in policy matters is key for long term growth here. I'm glad you agree we need this, and need to do it right. Lawrence Cohen § t/e 16:52, 29 April 2008 (UTC)

I think the act of taking power away from anyone (called disenfranchisement), is a bad idea, and should be vigorously opposed wherever it rears its ugly head. What I'm willing to contribute to is a system that potentially equally gives everyone more power and control over wikipedia. Fair enough? --Kim Bruning (talk) 16:57, 29 April 2008 (UTC) In short: less taking, more giving. :-)

But this is power they should not have to begin with, as it's contrary to our ways that anyone can bring forward change with support. NO minority is authorized or empowered under our policies to dictate how things are to happen on this sort of thing. Why would we endorse a system that allows a tiny group to hold sway over a larger body? Lawrence Cohen § t/e 17:17, 29 April 2008 (UTC)

Yeah, I was being (deliberately) harsh. Because frankly, you have more proposals and comments on this page than other people combined, and your position is still not quite clear to me; which is kinda sad considering that what I understand from comments of other people discussing here, they all agree with the following:

• Wikipedia decision making process is broken, and sometimes ends with status quo, where (powerful) minority wins over (scattered) majority.
• There is a need for more rigid rules of decision making, which should make gaming the system harder.
• Voting (or polling) is not a bad thing.

We may differ on details, such as if there should be an elected decision-making body, or how exact procedures should work, but I believe we all agree with the 3 above points (if anyone wants to correct me, go ahead). So, if you claim you want to find a "middle ground", this is a middle ground in the group discussing on this page. I don't know what group you want to find middle ground in you have in mind, but if you mean some larger group than that, you should say, what points above you disagree with and why. Samohyl Jan (talk) 17:15, 29 April 2008 (UTC)

I disagree with all those points. Rigid rules are game-able rules, voting has failed spectacularly on wikipedia, many times over. Decision making does not typically end in status quo, and there are known methods to prevent that. (as discussed elsewhere on this page)

But before I discuss that further with you, we have a more immediate problem. You have accused me of trolling. Are you willing to withdraw that accusation? --Kim Bruning (talk) 17:29, 29 April 2008 (UTC)

Kim, the Policy Committee will be tasked with improving the way that policy is created, reviewed, and changed. Like in most large organizations, there are different types of policy. The core policies, that affect everyone, need to have broad review prior to changing them. Wikipedia will benefit from getting more peoples input prior to changing core policy. Some of the issues involved are complex and need to have in depth discussion in order to decide which approach is the best policy. FloNight♥♥♥ 16:26, 29 April 2008 (UTC)

However, the committee will not be issued a magic wand, sadly. So we need to figure out real world ways in which the committee can actually achieve those goals, right? :-) --Kim Bruning (talk) 16:49, 29 April 2008 (UTC)

Wide, wide view of policy changes and proposals in a format that allows the most possible people to weigh in with support or dislike of the changes will be the Magic Wand in question. The PolCom or whatever its called is just what will organize that. Getting policy control out of the hands of small numbers of people is the Magic Wand. Lawrence Cohen § t/e 16:53, 29 April 2008 (UTC)

I'm still unsure why you believe that policy is controlled by small numbers of people. But see above. I have no problems ensuring that everyone has a say. What I want to prevent is the situation where we start taking away power from anyone. Along that path lies dictatorship, not democracy (or consensus) --Kim Bruning (talk) 16:59, 29 April 2008 (UTC)

The old model of Jimbo dictating change and/or any small minority being able to stonewall or filibuster any change is exactly what you describe, and exactly what this new model by Kirill will shoot dead. Whats a more fair way to decide a policy change? 5-20 people deciding on a policy talk page, and then a handful of people being able to stop their changes? Or a system where the policy change is broadcast in a very public place, everyone gets to say their peace, and then everyone collectively decides the outcome? How is that possibly dictatorial? I think you're grasping now, to try to poison the idea of something that will kill a traditional way of doing things for a new way. Tradition is useless if its not able to grow and evolve. Lawrence Cohen § t/e 17:20, 29 April 2008 (UTC)

Well, I think that 20 people trying to make a decision for everyone on a talk page is reprehensible. That is what I call prescriptive policy making, and what you happen to describe is the worst form of that.

Many want to have the situation where people get together and collectively decide things, but I'm not even quite happy with that. Though I can live with it, if that means people have at least thought things through upfront... see discussion with Flonight above)

My preference is for decisions to organically be made in the field, and people can then document how people in the field are deciding. That's what I've been doing. (see my discussion with Kirill about that. He thinks that that's entirely uncontroversial even... though that's not true unfortunately).

My problem with several proposals made here is that they haven't actually been thought through very well. I've seen a lot of systems and proposals come and go over time. So some of the proposals here are repeats or variants of things that are actually known to cause some of the ills that these "new" ideas purport to cure :-( . So I guess I'm saying that people should actually be systematic and think carefully, not just make things up willy-nilly --Kim Bruning (talk) 18:19, 29 April 2008 (UTC)

Basing it on just "field work" is flawed if based on one's own perception. Any one of us sees only a tiny portion of the behemoth that is Wikipedia, and nothing more. What 10, 20, or 30 people do in the field is not a baramoter of what the other 9,000 active users are doing or even want. Kiril was not wrong, he right: try doing that stuff to a meaty, important policy (banning/blocking, BLP, Notability, V, RS, or NPOV) and see how fast it gets revert warred back over time. Even if 40 users are doing something, that's a miniscule fraction of who we are. Not enough. Lawrence Cohen § t/e 18:25, 29 April 2008 (UTC)

Heh, I did actually edit NPOV recently, remember? --Kim Bruning (talk) 20:02, 29 April 2008 (UTC)

Hmm, on review, it looks like it's going to need some more work, as things are starting to slip back in again. <scratches head>. That is the one problem of maintaining policy. Non-best practices keep slipping back in as well, over time, as less experienced editors sometimes add information that was already fixed. That's very different from claims being made here though. --Kim Bruning (talk) 20:07, 29 April 2008 (UTC)

• Kim, yes, you're right, the decision model needs to be fast. That's why the poll should be a month long. So that it takes the drama out of actually solving the problem, and allows the problem to get solved rather swiftly before the poll ends. The poll isn't the solution to the problem. The poll is the solution to the problem which is stopping the problem getting solved, which is that everyone wants to own the solution. Sticking a month long poll in front of everyone makes it quite clear that no-one gets to own the solution. But perhaps that's too clever. Hiding T 19:18, 29 April 2008 (UTC)

  Oh. I see. That IS clever. <scratches head> First clever thing I've seen here , to be frank ^^;; Has some downsides too though, of course, but you've probably thought of that too. I wonder if there's a way to cover that? <scratches head some more> --Kim Bruning (talk) 20:01, 29 April 2008 (UTC)

  I haven't solved all the issues. Personally I think it is unworkable, but often if you mention polling long enough people work on a consensus method to avoid even discussing polling. Hiding T 20:32, 29 April 2008 (UTC)

  (I wonder if that's like putting a rejected tag on a discussion...) - jc37 05:55, 30 April 2008 (UTC)

AN clerks

Hmmm, continuing from the small note above, possibly we could use some WP:AN clerks to write down outcomes of AN debates in a useful summary fashion, and to categorize them in useful ways. This would end up being useful documentation of consensus all by itself, possibly more useful than the current policy system, in fact .... <scratches head>. Can anyone tell me why this wouldn't work? --Kim Bruning (talk) 16:09, 29 April 2008 (UTC)

For some topics a quick discussion works. But other topics need to have in depth review to learn whether the situation can be improved with changes and if so what type of changes. Anecdotal comments can trigger in depth review but alone should not be the basis for changing some types of policy. FloNight♥♥♥ 16:38, 29 April 2008 (UTC)

Right, but discussion alone doesn't work either. You need to discuss, test, discuss, test, and discuss again. A bit like the scientific method, you might say. And AN already provides that platform. So we've basically been sitting on a potential treasure trove that we haven't begun to adequately tap yet. :-) So how can we begin to use that? --Kim Bruning (talk) 16:54, 29 April 2008 (UTC)

Yes, having access to better organized discussions will definitely help. AN and AN/I are searchable. If we add keywords when we close discussions, maybe it will make the discussions more useful in a meta way. FloNight♥♥♥ 17:08, 29 April 2008 (UTC)

Understanding, Pros, and Cons to Kirill's model

Feel free to add to here, but please use the formatting listed and move threaded discussion to the discussion/reply section so this stays organized and useable.

Lawrence Cohen

Understanding of how it works

My understanding of this is that the PolCom, or The Group, whatever they will be called, will be simply stewards and advisors over policy matters, with no official power beyond making sure that no one tries to force through policy change without wide consensus. All new/major (i.e. non-trivial) changes to policy (elevating guidelines to policy, substantive changes, etc.) will be organized in a central place like WP:Proposed policy changes. People will post proposals in a simple format (maybe a physical "layout" like RFA, AFD, or DRV). The PolCom will weigh in with their views and opinions over perhaps a week, then the entire community gets to weigh in for 1-4 weeks (to be determined for length) in a poll/RFA/AFD/DRV style, with the usual threaded discussion below that. If the idea has support, policy is changed. If not, not. Try again later, since consensus can change. Policy talk pages will be still used to discuss policy in general, and to nurture ideas for change--but substantive change must be enacted in this manner for the widest review and to judge the true consensus and authority of the community. Lawrence Cohen § t/e 17:51, 29 April 2008 (UTC)

Pros

1. Easy tracking of policy change and evolution.
2. Provides a high-level view of what true consensus on a given policy proposal/change is.
3. With it's apparent 3-section system ("Policy Committee review/analysis", "Poll", "Discussion") provides a very clear 3-part easy to read and follow overview of where consensus on the matter is from veteran users (the PolCom), everyone else (Poll), and then provides for detailed discussion below (Discussion).
4. Centralizes all pending/proposed policy changes, so that none can ever catch any part of our userbase unaware. Everyone knows what is possibly changing and can speak up.
5. The clear "Polling", similar to RFA/DRV/AFD prevents any small groups from stonewalling, filibustering, or otherwise taking an inappropriate superior position over the majority of users.
6. Keeps our principles of discussion for change--but just reformats its layout to encourage participation from many more people, as it should be.
7. Makes all established users as close to equal as possible for policy decisions, as it should be.

Cons

1. Possibly delays obvious or "quick" changes to written policy, but such changes without demonstrated support may not have value anyway.
2. Slower, but more focused policy change and growth.

Support

• I support this. I would go for 4 week poll - change in policy isn't just an AfD, and more casual users may want to have their say too. Also, it should be noted what "has support" means, if it's just majority or 70% majority or what. Samohyl Jan (talk) 18:09, 29 April 2008 (UTC)

FloNight

Understanding of how it works

The model that Lawrence describes works for me as long as we include systematic review of current policy. Also much longer discussion that one week is needed for some topics so the Community can have a chance to collect information needed to make good choices.

Pros

• Flexibility
• Ongoing reviews

(more later)

Cons

(more later)

Kim Bruning

Understanding of how it works

Blocking of consensus can be at least partially explained a flowchart drawn by Kevin Murray, where he tried to include "discuss first" into the model.

In this variant of consensus process, an edit proposal can pathologically get caught in a loop with no change ever actually occurring. The loop runs from "should process continue"->"propose a compromise"->"is the result accepted" and back. As the "should process continue" box is not well defined, a skillful filibuster can easily get a "yes" each time

We found that that particular design lends itself especially well to filibuster and blocking of consensus. (see chart for details). Note that this is one of several charts documenting bad, good, and proposed simplifications of consensus process. Possibly we can make a page discussing all these at some point.

When looking at real life situations, it turns out that people have in fact already been applying variants of this method. What happens in reality is that a small group can hold out and WP:OWN a page for quite a while, until someone with WP:BRD skill comes along and breaks up the party. (I have done so occasionally over time). I'll admit that this situation has been getting somewhat worse over time, to the point where there has been a discussion about whether such behavior constitutes disruptive editing, see:Wikipedia_talk:DE#Blocking_consensus. Though I don't think it's disastrous yet, since so far we've been able to knock down those blockades, and get people editing according to the actual consensus process again.

Lawrence Cohen's understanding resembles this flowchart. The policy committee might seize control of the "should process continue" box, and perhaps improve things, but it's a very heavyweight approach to the problem.

Imho A better solution is to eliminate that step from the process. In current practice, this is done by enforcing some of the more optimal flows as documented at Wikipedia:Consensus (there's more documentation on talk page and in archives, for those who are interested.)

Pros

• It looks good on paper.
• If applied carefully, it can prevent people from blocking consensus, in a roundabout way.

Cons

• Single point of failure: If the wrong people get on the committee, they basically take power of the wiki.
• If the group is not careful, or if it is manned by people who don't understand the situation, it can very easily end up reinforcing existing deadlocks.
• reverting people or protecting policy pages is the best way to create deadlocks. This must either be prevented, or it must be compensated for elsewhere.
• Centralization on a large wiki leads to possibility of surpassing Dunbar's number. This must be eliminated, or the system will fail.
• Enforcing long procedures stretches the decision cycle (OODA loop) too long, and makes it impossible to come to timely resolutions. (the perfect is the enemy of the good). This must be prevented, or the community will out pace the new system, rendering it ineffective.
• Prescriptive policy is very hard to enforce on a large sprawling wiki. The group may want to stick to describing best practice, as is the current known best practice.
• Such commissions have been suggested before, once on enwiki (Wikipedia:Wikirules_proposal), and once on nlwiki as far as I can recall. In the case of the enwiki commission, the committee was strongly outpaced by the consensus system Wikipedia_talk:Simplified_ruleset#Historic_information. The final outcome of that situation can be found today at Wikipedia:Five pillars.

Hiding

Understanding of how it works

Basically it's a policy approval committee dressed up as a steering committee.

Pros

I can't actually think of any. May please policy wonks?

Cons

Secret mailing lists. Ulterior motives. Poor representation. The community becoming herded. Unnecessary. Anti-wiki. A shift in the power balance which currently leaves every editor equal. A new class of editor on Wikipedia. More process, when less is better. Every flaw inherent in every committee known. For example, over time they get dressed up in their own importance. After a while a seat on the committee becomes something to have just for having's sake, rather than for the actual purpose. It will eventually assume importance. The checks and balances are woefully inadequate and always will be. For instance, how long before only policy com is allowed to edit the policy com pages, and decide how the policy-com system works? And then we'll have policy clerks. It will institutionalise. There will be some form of systemic bias which will be hard to counter, since no committee can be as fully representative of the community as much as the community is of itself. It will factionalise debate. Probably many more. Look, I know why this being proposed, and I know people will dismiss this as worst possible case scenario, but have a look at how flawed other processes are criticised as being. Look at the problems IRC is causing. Look at the issues people have with arb-com. Look at the flaws in the RFA process. Look at the issues with AFD. They all exist, and no-one saw them coming or perhaps thought they were above them. If there really is a problem that has to be solved, then the board will solve it, because they're the ones who have to do stuff. The community can't be expected to solve everything, and it won't, because not everything can be solved. Look at all the articles on pederasty. WTF are we supposed to do about those? A lot of the issues we;re having aren't with consensus editing, they are with disruptive editors who are attempting to own issues, and who aren't being dealt with effectively by existing processes. It's not the fault of consensus, it's not a problem another committee will solve. It's a problem that dealing with disruptive editors in a better way will solve. Sorry, I haven't meant to stamp on people's feet too hard, but I really think this is a bad idea. We're all supposed to be equal. Hiding T 19:52, 29 April 2008 (UTC)

Discussion

Response to Kim Bruning

In regards to Kim's cons...

1. Single point of failure: If the wrong people get on the committee, they basically take power of the wiki.

   Actually, in my interpretation of Kiril's model, the "PolCom" can't take over the Wiki. They have no power but to basically interpret consensus, and make sure no one forces policy change without the public review Kiril's model pushes. They're advisors and gatekeepers, nothing else.

   Originally yes. But it's very easy for it to morph into something else. --Kim Bruning (talk)

   Easy enough to fix. We make it a mandate, like how "Arbcom doesn't do policy". Seperation of powers. Fixed. :) Lawrence Cohen § t/e 20:02, 29 April 2008 (UTC)

2. If the group is not careful, or if it is manned by people who don't understand the situation, it can very easily end up reinforcing existing deadlocks.

   They can't do this either, as they don't decide. The community still decides, just on a central/wider scale via a poll in support or opposition of the proposed change, plus the associated discussion. If the wider community doesn't want or wants the change, it happens. Thats how things should be.

   They guard the process, and they need to take special care, is what I mean --Kim Bruning (talk)

3. reverting people or protecting policy pages is the best way to create deadlocks. This must either be prevented, or it must be compensated for elsewhere.

   A wide review of changes as proposed by Kiril's model does exactly this. If there is a historic deadlock, exposing that deadlock to 100+ additional eyes will demonstrate the value of the various argued points quite quickly. Stupid ideas will die fast in this model, and good ideas will rise fast.

   100+ additional eyes in one place tends to slow things down to a crawl. --Kim Bruning (talk)

   I sincerely believe that this is the primary disconnect you're seeing with Kiril's proposal. Having the discusions structured in a manner like how RFA is will encourage more people to join in, and it will go swimmingly I think on this manner. If more people weigh in, great--the good ideas will shine, if the community really wants them, and the bad ones will sink like a rock. Lawrence Cohen § t/e 20:02, 29 April 2008 (UTC)

4. Centralization on a large wiki leads to possibility of surpassing dunbar's number. This must be eliminated, or the system will fail.

   AFD and DRV seem to be working just fine. Anything that limits visibility of discussion or that requires or encourages babysitting of pages to get your way (policy wonks) is evil, and must be stopped as it encourages a ruling policy class.

   DRV is so obviously broken that I don't know how to respond to that. AFD is surprisingly actually working a little better than it used to. Not that this helps much, its hiccups still occasionally makes the presses on a slow news day, IIRC. But they do split discussions up over multiple pages, and make good use of the wiki. I wouldn't recommend that model, it's an ugly hack on top of an ugly hack with multiple templates and transclusions and bots, but it does work. It's not true that it's simply a centralized discussion page. That it most DEFINITELY is not.  :-P

   Like I mentioned on my talk page, we'll see what? 5, 10 proposals a month at most? We don't even have that many 'pending' policy changes now. I think the volume and an RFA-type physical layout with a few transcluded pages will work fine. It won't be any kind of an overload. DRV's fault is that it's mandate is ill-defined. It can use more structure. Lawrence Cohen § t/e 20:02, 29 April 2008 (UTC)

5. Enforcing long procedures stretches the decision cycle OODA loop too long, and makes it impossible to come to timely resolutions. (the perfect is the enemy of the good). This must be prevented, or the community will out pace the new system, rendering it ineffective

   It's a darn good thing this process is pretty simple, then. Propose a change. The committee weighs in with views, discussion commences on the proposal page. A week or so later, we open up the poll/AFD/DRV/RFA etc. style part of the page. Consensus will be apparent readily then. The method is a formatted "Propose > Discuss > Decide" model.

   Okay, that's better, but now you have enforced time-limits. *sigh*. That CAN work, has worked in practice... but is not very nice. Nlwiki uses a system like that... a more factitious place you have never seen... I already don't edit on nlwiki, because their system really does suck that much. --Kim Bruning (talk) 19:44, 29 April 2008 (UTC)

   We're not here to be nice, to be all-inclusive in the sense of swaying people to consensus over time. That worked years ago, I'm sure, but it's not needed. If someone won't turn around, they won't turn around. We're not here to be masters of socially convincing others of ideas of merit, we're here to improve Wikipedia. Time limits are fine--we have them on AFD, DRV, and RFA. If a consensus isn't obvious by a reasonable time, then it's a safe bet the idea isn't worth implementing. Lawrence Cohen § t/e 20:02, 29 April 2008 (UTC)

6. Prescriptive policy is very hard to enforce on a large sprawling wiki. The group may want to stick to describing best practice, as is the current known best practice.

   Rigid adherence to historic practice is futile. Just as consensus can change, our methods to determine consensus can change too. Nothing is immutable on Wikipedia except legal concerns. We can dump anything and everything else if a better solution presents itself, from a line of article text, to policy determination, to the entire WMF board and WMF employees.

   Now that is irony. In effect you're basically saying here that "Rigid adherence to the historic practice of non-rigidity is futile, therefore we must introduce a new practice of absolute rigidity!". Maybe you'd like to reword? :-)

   Nope, no reword. Theres nothing rigid about Kiril's proposal. Is it rigid to ask people to comment within a couple of lines, rather than aimless, rambling sprawling archives of text that only the hard core policy wonks will follow? Lawrence Cohen § t/e 20:02, 29 April 2008 (UTC)

7. Such commissions have been suggested before, once on enwiki, and once on nlwiki as far as I can recall. Both times the commissions failed to work, the consensus system outpaced them rather strongly at the time.

   Because something may not work is not a reason to not try it. Wikipedia:Wikirules proposal looks totally different than Kiril's proposal; it would have been an appointed ruling class to decide rules. Kiril's proposal is to empower the community in a more organized fashion. Who would have thought an encyclopedia anyone can edit would work?

   Like I said, I love the principle. Empowering people is a fine concept. But at the same time, I think the proposals put forward so far look like they will do the opposite, some even resemble the mindset behind nupedia. That's why I posed some basic requirements. If they are met, then there's a good chance that the principle would actually end up being put into practice. :-) In the mean time, I want to continue some of my own activities which do tend to empower people over time, as I have observed.

   I experienced that people would never have thought wikipedia would work. But it did. It worked because people had thought very carefully about governance over time, and implemented many procedures to make it work. Unfortunately, over time, a lot of those people have left, and now we sometimes have people actually *on* wikipedia, who still don't believe that wikipedia can work ;-) --Kim Bruning (talk) 19:44, 29 April 2008 (UTC)One of my modest proposals is to teach them how to make it work for them. But that's a story for another day.

   It's actually the opposite of nupedia, because Kiril's model and proposal would make it so that the "policy wonks" are of no more value to the policy determination process than any other user, and because it encourages more people with a stake in Wikipedia to weigh in with their thoughts. If you ask a random good user with 300 article edits, 200 talk edits, and another 200 edits across various project pages to weigh in on a policy proposal, what if the guy doesn't have the time to sit and respond in depth for days, weeks, or months on the change? He has a stake in it--should his voice have less value in the ultimate decision because he doesn't have time to sit and work the policy talk page until everyone is swayed? Maybe he just really likes or dislikes the change, and says why. That one edit/post should count just as much as the lady with time to post 200 times to that policy page in a month and to reply to everything, because that lady spends 550 of the 700 posts she has time for in a month on policy pages. Lawrence Cohen § t/e 20:02, 29 April 2008 (UTC)

Responses in-line. Lawrence Cohen § t/e 19:19, 29 April 2008 (UTC)

Response to Hiding

How does putting all policy decision in one central place, with everything done in public, dragged out of the backrooms of policy talk pages with a method of:

1. I propose a change on the central page/cross notify to the policy talk page
2. The PolCom members weigh in with their take/views on it
3. The community weighs in with their take/views on it on the proposal page
4. A 1-4 week poll on the proposal commences
5. That poll decides the fate of the proposal
6. PolCom simply guards against random changes to policy that have no support and advises on policy interpretation

How does that exclude anyone... or make some of us less equal in voice? ALL our voices should be 1:1. None of us should be 2:1, or 3:1, or 5:1, as some policy wonks are now by ownership of their pet pages. Is your major problem the so-called steering committee? What if it was:

1. I propose a change on the central page/cross notify to the policy talk page
2. The community weighs in with their take/views on it on the proposal page
3. A 1-4 week poll on the proposal commences
4. That poll decides the fate of the proposal

Would this be less offensive if we left off the "PolCom" bit? What I want myself is to have everyone have an equal voice in policy decision, in a way that does not reward or empower those with time to spend hours and hours daily on guarding policy pages. Kim? Thoughts? Lawrence Cohen § t/e 20:11, 29 April 2008 (UTC)

• Um, okay. Let's look at your bottom suggestion. Let's try and take this for a test run. You propose that everyone start their comment on a talk page with hello. We then have to go through stages 2 3 and 4? If not, why not?
• Also, as it stands we already have 1 and 2, why do we need 3 and 4 mandated? They are there if we want to use them, and we have used them in the past.

• My major problem with your proposal is that it will go bad, because power corrupts. And I don't care how you dress it up, anything which involves electing people involves power. Now god knows I like polls, but I don't like polls being used to mean anything. I like polls being used to force people into doing anything but listening to the poll, because if the poll is going to return a unanimous decision we wouldn't need the poll in the first place because there wouldn't be any contention. So therefore we will only have polls when the community is divided, and they then only divide the community even more. See above for my thoughts on how to read the results of a poll. Some of the best disputes on Wikipedia are how to determine the results of a poll. Hiding T 20:28, 29 April 2008 (UTC)

OK, you're against the model with a committee. So lets cover the rest in the bottom version.

Kiril's model: You want to make a substantive change to WP:N. You post the change proposal and reasoning on something like WP:Proposed policy changes. Post a notification back to the PPC page on the WP:N talk page. Everyone weighs in for a week lets say, *ON* the proposal page in the central location. People just need to bookmark that one page then to see all pending proposals. After a solid week's discussion, we fire up a poll. A policy change shouldn't happen unless there is obvious significant support for it--thats the point of them being commonly accepted by the masses. Most policy change attempts fail. It will be the same under this model--but it will allow a lot more people to weigh in with their thoughts and views, without having to sit and debate for weeks on end (unless if they want to).

Let's take it the other way: today's model. You post your change to WP:N's talk page for discussion. People watching WP:N weigh in. Lets say I notice your proposal, and say "I don't like this for xyz, don't do it" or "I like this for abc, please do this". Let's say I never pop back into that page for 2-3 months. Should my statement and voice in that policy change have value and affect consensus, without my having to sit there and defend my views for days, weeks, or months?

On polls... In regards to steps 3/4, the poll, what is a fairer way to ensure that all users in good standing have an exactly equal voice, on a 1:1 basis, as they are entitled to in the operations of our website? Should some users have a 2:1 weight, or 3:1? Sure--based on their opinions, but just like on RFA, people's opinions sway each other. But we're a volunteer project; we spend time as we will. For my voice to be counted I shouldn't have to watchlist every comment I make to fend off all challenges. That's absurd. If I pop onto an AFD and say xyz, my statement has weight, and should be counted in consensus, regardless if I never return to it. This Kiril model defends the average user's voice against policy rats and policy wonks. It makes all users the same.

Kiril's model also, like AFD and DRV, would allow for returns to old discussions, to rerun them--just as AFD and DRV. Consensus can change over time, but like a real policy talk page, are you going to try to ram through a change monthly for a year till you get your way? Of course not. This focused method would make it easier to track perennial requests that have no merit. Polls can be rerun at any time. If you abuse that, you'll just be ignored, though, without presenting a compelling argument...

So... I have maybe 10-15 hours a week to spend here, for example. On a good week 10+ of that is article work. Research, writing. Two FAs and two GAs down, and another 2-3 FAs hopefully in the pipeline. Should I have to spend 10+ of my 15 hours per week on policy matters for my voice to have the same value as those who already spend that much time on policy pages? Which of those two models is more fair to people that want to build the encyclopedia, and not build policy? Lawrence Cohen § t/e 20:42, 29 April 2008 (UTC)

I think you missed a whole chunk of my post. You also completely misunderstand today's model. First off, I want to change policy, I just change it. Or, I only have ten to 15 hours to edit, so I don;t care what policy looks like. I'll just edit the way I always do and let the policy wonks edit it, because they're so busy doing that they'll never come into article space and actually find out it doesn't get implemented, we're too busy building the encyclopedia through consensus editing. The best model for creating the encyclopedia is to create it. And then describe how we do it. Why on earth would you want to do anything different? How do you know what works until you've tried it? Sorry Lawrence, I don't see the nut you're trying to crack. Maybe we need a specific example. Hiding T 20:50, 29 April 2008 (UTC)

A specific example would be Wikipedia_talk:Biographies_of_living_persons#Straw_poll, which has mega-impact. Something like that would be perfectly tailored to Kiril's model, as it would guarantee the widest exposure for something like this. Unless people happen to watch the BLP page, or caught it on the Village Pump (not enough watch this, and you have to find it before it scrolls), or BLPN, they may not know about it. Every active user that edits has a stake in that discussion, and all should have have the same change to weigh in without having to watch the few places it was hawked. Lawrence Cohen § t/e 21:00, 29 April 2008 (UTC)

Actually, that poll is a staggeringly bad example of the problem. Because I missed the poll and tried to amend policy, and my changes were reverted because the poll is ongoing and now you can't change the policy until the poll is done, which is stupid. (Curiously, I've been reverted by an arbitrator on those changes, which makes you wonder whether arb-com is divorced from policy making. Would you revert an arbitrator?) But the poll is now mentioned in the deletion process, guidance on deletion for admins and at WP:AN, so it broadens the scope further. And your assumption seems to be that the poll won't solve anything. Otherwise you wouldn't be using it as an example. And if all you are concerned about is coverage, stick a link on the top of the watchlist. The problem with the BLP solution is people keep having separate straw polls on all the separate solutions in different places on different days of the week. The consensus seems to be moving towards semi-protecting BLP's, but fairly slowly. All that said, if this really was a mega-issue, the board would already have mandated the answer. And we'd still be in the same logjam now as we would be if Kiril's proposal was implemented. How would that have solved this issue? Hiding T 21:18, 29 April 2008 (UTC)

I also have an example to share - WP:SPOILER. Normal people were busy writing encyclopedia, and some of them included spoiler warnings. The WP:SPOILER page contained description of the practice. But then policy wonks chimed in, and removed all the spoiler warnings (under false pretenses), then changed the WP:SPOILER to describe what they did - ie. Wikipedia doesn't contain spoiler warnings. They used circular argument - there are no spoiler warnings, so the policy documents the current practice. If someone tried to add SWs back, they were reverted or even banned on the grounds that they contradict to the policy. The spoiler warning template was later removed in TfD which ignored majority. Your view is unfortunately very idealistic, and we don't live in a vacuum. From time to time policy wonks notice the article space and start doing their thing (usually deleting stuff). You may consider WP:SPOILER as a minor thing, and I generally agree, but for me, it was an eye-opener regarding Wikipedia politics. Samohyl Jan (talk) 21:24, 29 April 2008 (UTC)

Spoilers are one of those issues no debate would solve. You either have them or you don't, it's binary and one side loses no matter what. It appears as though consensus is with removal though, since consensus can change and it appears it hasn't yet. Maybe you should restart debate at the village pump or amend WP:SPOILER, see if consensus has changed. Hiding T 21:57, 29 April 2008 (UTC)

Actually, it isn't so binary as it seems - I gave two compromise proposals which would be in between, but they were rejected by the anti-spoiler people (and since they had more power, there was no point fighting for that). It's nice that you say that consensus can change, the problem is, there was no consensus for their removal in the first place. I don't want to bitch about it here though, I am just giving an example where what you described above ("just edit the mainspace and then document the practice") didn't work. If you actually admit that there are issues in which one side loses no matter what, then voting would guarantee that the minority would lose in such cases, thus making less people unhappy. That's why I want voting to decide things (I wouldn't mind to lose in SW debate by voting, but I do really object to lose by underhanded tactics). Samohyl Jan (talk) 22:33, 29 April 2008 (UTC)

Okay. Let's walk through that and see where we end up. So for starters, what makes you say there was no consensus? Weren't the spoiler warnings reverted back in, for instance? *listening carefully* --Kim Bruning (talk) 23:20, 29 April 2008 (UTC)

I don't want to go through that, it's irrelevant to this discussion, and it's beating of a dead horse now. And I already explained above that those who tried put SWs back to the articles were reverted (and pointed to the recently changed policy) or even banned. Mostly these people were maintainers of the articles, and were unaware what's going on on WP:SPOILER. Just read the discussion from June and July 2007 on WP:SPOILER (the RfC, for example, where they closed the polls as soon as they started to be unfavorable to their cause), and you will understand that you cannot call that consensus. Samohyl Jan (talk) 05:19, 30 April 2008 (UTC)

Well, you brought it up now, and if you're right that something went off the rails there, it bears investigating, right? (if prefer to take it elsewhere, leaving a message on my talk page works fine too.) --Kim Bruning (talk) 18:03, 30 April 2008 (UTC)

• In reference to the KL's proposal as I "think" Hiding understands it, I pretty much agree with Hiding's assessment of potential shortcomings. - jc37 05:55, 30 April 2008 (UTC)

Response to User

Personally, I think it would be interesting to scrap all policy but the core 5 ones and see how that works now in large scale. IAR all the time, as it were. I bet things won't degrade too much. Too much energy is spent on policy squabbling anyway. JeanLatore (talk) 03:02, 30 April 2008 (UTC)

What do you consider to be the "core 5 ones"? - jc37 05:31, 30 April 2008 (UTC)

Probably means WP:5P. Carcharoth (talk) 06:05, 1 May 2008 (UTC)

Heh. Exactly what I was thinking. Only I would have called it a "Modest Proposal" or something like that...Ben Standeven (talk) 10:18, 1 May 2008 (UTC)

If you know the history of WP:5P, you might know that historically, this has had my support as well. A lot of policy that is actually important is enforced by the mediawiki engine, and/or could be programmed in.

One could alternately go further yet, or hold back just a little bit.

<scratches head> I realize we're several years onward now, but still, I wonder to what extent that's still a viable plan? I'm somewhat skeptical, but it would be cool if it was still possible, especially if we can get more buy-in than previously. But how to proceed on that?

--Kim Bruning (talk) 20:20, 1 May 2008 (UTC) I have the feeling that the current system might be subject to the Abilene paradox

The trifecta nearly works. Oddly, the simplified rule-set looks too complicated. The single rule-set is possible, of course: do your best to improve the encyclopedia. I'm not sure if we have passed that point already though. Hiding T 09:56, 2 May 2008 (UTC)

What was the catalyst for this?

What was the catlyst for this? Is it the current BLP "discussion"? - jc37 05:56, 30 April 2008 (UTC)

I suspect it's a variety of factors, one of which I see as ArbCom's frustation with cases where policies collide. The tension builds and conflicts between editors are almost never resolved satisfactorily. Nifboy (talk) 09:40, 30 April 2008 (UTC)

If arb-com actually looked at the behavioural patterns of editors involved in that dispute it might not have festered so long. There was one sock puppet involved in that case and some very tendentious editors. The flaws in that case are the flaws this will not solve, because the community as a whole is rather perfectly split between inclusionists and deletionists. The other flaw in that case is that there are editors who don't wish to treat problems case by case and there are editors who feel that no precedent is made after a large number of articles have been treated case by case. There is also no quarter given on either side, which would see us retain anything which resembles an encyclopedic article regardless of being sourced mainly from primary sources. But that's a separate issue. What I would like to know is how this proposal will solve issues where the community is finely balanced in two camps. Either policy-com is the final arbiter, in which case power has been removed from editors and invested in a committee, with all the flaws that entails, or we have the status quo, so the policy com is redundant. The problem isn't with the system, it's with the ability to sell changes to the community, and with our methods of dealing with editors who don't seek to build consensus. Arbitration on Wikipedia currently doesn't do what it is supposed to do. Hiding T 10:10, 30 April 2008 (UTC)

I think the collision of policies, including that decision (no offense to the ArbCom), is completely a false dilemma. Taking the example linked above, there is only a contradiction if each page is considered in a vacuum, which is a silly way to consider things. Verifiability, no original research, reliable sourcing, notability and "what Wikipedia is not" all contradict that version of the "Wiki is not paper" principle is a similarly superficial manner. Can we create topics with hundreds of subarticles and other related content? Sure. The wiki is not paper. Does that mean we disregard our basic content and inclusion principles? Of course not. There is no inherent contradiction except when considering the individual principles of themselves with a rather literalist reading. Vassyana (talk) 01:58, 8 May 2008 (UTC)

A related essay

The essay User:SteveBaker/Consensus Essay may be of interest in relation to this proposal. DuncanHill (talk) 23:29, 1 May 2008 (UTC)

Hm, interesting essay. He came to the conclusion that voting and some rules are necessary, even through the usual denial of Wikipedians that it is not. His proposal has problems, however - there is no time limit for the voting, meaning anyone can stop it at a convenient time, and also there is no way someone interest in the voted proposal could know about it (no central page); also, his supermajority requirement means that 1/3 minority can hold status quo against 2/3 majority, but that's a minor bug in my eyes. Anyway, I always wonder why so many Wikipedians believe voting and hard rules are such a bad thing. As the above essay have proven formally, there is no way around voting. And I have always interpreted "ignore all rules" not as cowboy-era "you can do whatever you please" but more like "if you are new here, you don't have to read 20 thick books on various policies, and in general, if you occasionally make a mistake, that's fine". Samohyl Jan (talk) 05:52, 2 May 2008 (UTC)

(I've always felt that the correct interpretation of "Ignore all rules" is "Recognise that there are exceptions to every rule - and that in the end, making a better encyclopedia is sufficient justification for claiming that exception"). SteveBaker (talk) 18:56, 2 May 2008 (UTC)

Thanks to Duncan for posting my essay - and for reminding me that I wrote it(!). The main thrust of my essay was to convince people that the system is broken. I agree with Samohyl Jan that my concluding proposals are probably not of great merit. What is important is that we can agree that there are situations where consensus breaks down. Having done that we will be better equipped to address a proper solution.

From what I've read above, quite a few people are still in denial and believe that consensus can work. I'd like to hear what they have to say about my essay - what DO you do if consensus cannot be reached on a matter that has no default position or in a situation where everyone agrees that action is essential but the precise form of action cannot be agreed upon?

I maintain that consensus still works for deciding what words go into an article - because we can always compromise ("Say both" or "Say neither"). It works moderately well for small groups of editors who are on friendly terms because people will go out of their way to back down when they see they are obstacles to consensus. It's also OK for things like WP:FAQ and WP:AfD where the 'default' action in the event of 'no consensus' is clear (don't promote the article, don't delete the article). Where consensus fails is when there are a large number of people involved - or where there is no good agreement on what "no consensus" means.

Sadly, I think this effort to introduce voting is doomed because we require consensus in order to be able to replace consensus - and this debate is rapidly becoming a poster-child for questions-that-consensus-will-never-answer.

SteveBaker (talk) 18:48, 2 May 2008 (UTC)

I agree. It seems to me, from the arguments against voting here, most Wikipedians who despise voting do so because they have bad experience with it. Unfortunately, this bad experience is due to lack of hard and respected rules. If people do not respect rules of voting and the result of voting, then the act of voting is completely meaningless. It is important to realize that voting and hard rules for voting go hand in hand. Anyway, I am still quite optimistic - there are many people who are starting to notice the problem, and by debating it, they can find a good solution. Samohyl Jan (talk) 21:04, 2 May 2008 (UTC)

Nah, it can work, the Weimar republic managed to abolish not just consensus, but even voting altogether (might be preferred by those who would prefer the thing run by arbcom fiat, for instance). You'll need a time machine to go live there though.

Of course, the Weimar republic was the extreme example of the case. The more general relevance is that a participatory system *can* be downgraded, if sufficient support can be gained from people who lack respect for the system or knowledge of how to use it. The lesson learned is that those familiar with the system should put up the most strenuous resistance in the short term, and work hard to educate people to learn how to use it in the long term.

This particular page should have been a wake up call, but too many people simply have never looked at it still, even today. That's a weak spot in participatory systems, people can get too complacent, if you're not careful. --Kim Bruning (talk) 10:30, 3 May 2008 (UTC)

I'm surprised by SteveBaker's conclusion that "many" people are in denial. As far as I can see, only Kim Bruning is defending the current system. Given the monumental filibuster he is pulling off here, he obviously finds it very congenial. Not surprising as he is demonstrably one of the few people with sufficient clout (i.e. admin buttons plus reputation) to be able to make changes to policy pages; edits to policy by non-admins are rapidly reverted and result in blocking if they try to press the point (yes, it happened to me).

FWIW, I strongly support the model suggested by Kirill Lokshin. Wikipedia changes, and new policy is required from time to time. Right now we have no way to produce it. Example: WP:Flagged revisions is coming. It has many switches that have to be set. The community will never decide on how to do this, but a formal policy assembly, making decisions by simple majority voting, could. Otherwise we effectively leave the decision to the developers, who certainly don't want to make them. PaddyLeahy (talk) 15:22, 3 May 2008 (UTC)

• Kim doesn't, I believe, have admin buttons. I do, and will throw my lot in with Kim's. I believe polls have a part to play, but only when consensus dictates. Consensus is the best decision making process we have. A lot of the issues people have with the consensus model is that they imagine the issue has to be settled within their preferred timescale. Anyone who gets blocked by an admin for editing policy needs to kick up a big stink. All the way to arbitration, if need be, and let arbitration do what it is supposed to, which is defend the decision making process and settle disputes unequivocally. Yes there are flaws in the process. Kim's doing a very good job at fixing those on the ground as and when he can. Further on, maybe the community needs to rethink its criteria for electing arbitrators. Too often they're picked on popularity rather than insight or suitability. Which is why I think pol-com is a staggeringly bad idea. Hiding T 19:26, 3 May 2008 (UTC)

Heh, I ain't got no admin flag, and I'm not sure how many people still know me at all. People come and go at a startling rate. I figure that's part of the problem. People don't ever get a chance to learn how to do this stuff properly.

You got blocked for attempting to edit a policy page? Okay, let's go take a look there. Can you link?

--Kim Bruning (talk) 21:33, 3 May 2008 (UTC)

No, let's not get diverted talking about ancient history. Suffice to say the block was technically legit but IMHO as a newbie (then) I'd have got off with a warning if it hadn't been a dispute over policy with a bunch of admins. In the context of the current debate, it's worth saying that in the wake of that dispute I was sufficiently impressed by the disconnect between the gobledegook about consensus and the actual way that policy is made and enforced on wikipedia that I drastically cut down on editing, while other editors quit completely. This is yet another reason the current system is "broken": the lack of process clarity greatly increases the stress about even relatively trivial disputes, leading to the losing side feeling they have been steamrollered by powerful cliques. If there was a pol-com, users could clearly see who was deciding what, accept the decision pro tem and if they really cared, vote accordingly (or stand themselves) at the next election.

It's encouraging to hear Kim is not an admin; as for the other thing, in this context the reputation that counts is with the admins who take an interest in policy, and I think you are pretty well known to and trusted by that group. PaddyLeahy (talk) 15:29, 4 May 2008 (UTC)

Agreed regarding Kim's reputation being sufficient to earn respect, admin or not. In response to Hiding, part of the problem might be those cases when people are wanting policy change "on their timescale" because of a perceived difficulty which could blow up in everyone's face if it isn't addressed comparatively rapidly. In all honesty, I have no clear idea how to address that in a way satisfactory to all. The best option I can personally think of, and it may not be a particularly good one, would be to maybe draft a guideline regarding the matter which would address the perceived current difficulty effectively but possibly only that perceived current difficulty, and see if it can get consensus as a guideline. If it can, then the proximate problem will likely be addressed, while at the same time hopefully not being an undue burden on freedom. Also, it would allow for further "refinement" of language, terms, and the like to remain ongoing.

As someone who supports the model proposed, my biggest personal reservation would be in regards to those individuals who wrongly expect the group to function in the way we are used to legislatures doing, like passing new laws every day or so. That would definitely not be what I and I think others are contemplating. I'm thinking less of a legislative model than a crisis management model, whether the term "crisis management" itself is applicable or not. The members of the committee should be selected on the basis of familiarity with extant policies and guidelines, a working knowledge of optimal phrasing for new proposals, and such so that, effectively, if it is perceived as being good or necessary to have a new policy or guideline proposed, the members of the committee can be those individuals who can produce a functional draft of the policy or guideline quickly and effectively, know between them enough about wikipedia to be able to draft something without glaring loopholes in it, and be able to adequately make any changes resulting from reasonable possible complications noted by others later. By settting it up in that way, it would allow the matter to come to broader community discussion more quickly and possibly lead to a quicker reasonable resolution of the proposal. If it is drafted and still remains only a proposal, with no changes to policy or guidelines resulting, fine. But we would want to make sure that no one is expecting the group to pass 100 new policies every term or anything similar to a legislature. John Carter (talk) 15:51, 4 May 2008 (UTC)

(outdent) I don't think a legislature is too bad a model, if not taken too literally. A pol-com would be fundamentally limited by the time volunteers can contribute, so 1 new policy per week (let alone day) seems highly unlikely. But once created, it would be up to pol-com to decide how to run itself, since it could presumably change any policies defining its role (unless WMF were to impose a constitution, which would be a Good Thing but seems unlikely). Pol-com would be one of wikipedia's most powerful bodies (in some ways more powerful than the WMF trustees, given that they can't or won't interfere on content). Hence pol-com members must be, and be seen to be, representative of the community. That implies

1. a group big enough to reflect the major strands of opinion in the community (minimum 20, 50 would be better but might be hard to recruit);
2. democratic elections with careful control of voter eligibility (like the WMF board) to prevent gaming the system;
3. moderately long terms (say 2 years) because such elections are expensive of volunteer time and effort.

I don't see drafting as a major problem. The problem is agreeing (i) that the process of drafting is finished and (ii) whether or not to implement the draft. Nothing stops pol-com from inviting non-members (e.g. with legal training) to draft policies, from voting on policies suggested by non-members (although it can also ignore them), or running votes or !votes within the general community (although it doesn't have to abide by the results). As a democratic body, members wanting re-election would have to listen to the community. All of these are typical of legislatures, and militate against pol-com becoming too autocratic (though of course the whole point here is to be a bit more top-down than the current system, so as to be able to make actual decisions). John Carter's carefully "selected" (by whom?) crisis-management committee sounds more like an executive than a legislature, and despite the best of intentions at the start that would likely develop rapidly into a very top-down system. I would agree that, unlike a legislature, pol-com would be wise to act only where a significant dispute blows up, and let the current system continue for minor changes (e.g. most MOS issues) where it generally works OK. PaddyLeahy (talk) 18:06, 4 May 2008 (UTC)

Actually, I never said "carefully", and I was more or less considering that they would be eligible for election like your proposal above, although possibly with some input either at the beginning or end from the Foundation, Jimbo, or whatever, if they would request that, much like ArbCom so far as I can tell. I can't help but thinking that the two ideas are basically similar, even if the terms to describe them aren't identical. I would agree that it would probably function best like the original US Congress was supposed to, passing new issues only when required. But, considering that the legislatures we now have have their candidates run on the basis of the bills they personally proposed or passed, I think that using the word "legislature" will carrying some meanings to at least several people which might confuse at least a few of the issues regarding the subject. John Carter (talk) 18:18, 4 May 2008 (UTC)

In fact I notice that I ended up just repeating the original proposal... I agree it is mainly a matter of emphasis. WMF vetting pre- or post-election might backfire though: it would be better for WMF to specify a mechanism for dispute resolution if WMF and pol-com end up arguing (e.g., WMF power to call fresh pol-com elections).

Presumably candidates only run on bills that were popular: is that a bad thing? Would it be bad if pol-com passed many policies that were popular with the community? More likely, editors will prefer to work with a fairly stable set of rules, but these things have their own dynamic, which is hard to predict or tie down at the start. PaddyLeahy (talk) 18:59, 4 May 2008 (UTC)

It's been said above that the current ArbCom members sometimes get elected for popularity, not necessarily clarity of thought. On that basis, making it easier for people to be swayed by "popularity" is probably not a particularly good thing. Also, I think the "vetting" process would involve letting someone know your real name, so they can check to see if you're a lunatic mass murderer who'll never leave jail for the rest of his life. In a lot of cases I don't think that's the sort of thing we spread around about ourselves too often. But the Foundation, who doesn't have to say what they know to us, or Jimbo, would be in a position to ensure that we don't have Hannibal Lecter determining content here. Somehow, I think his opinions on "food" content might be, well, problematic. John Carter (talk) 19:29, 4 May 2008 (UTC)

Given the influence of pol-com and the real-life monetary value of wikipedia, I would be in favour of pol-com candidates being required to reveal their real name and age to the electorate, and have this verified by the election board. Furthermore, off-wiki activities should be regarded as completely legitimate topics for discussion about candidates. Then all the vetting is done by the electorate, rather than by some council of mullahs. In democracies, candidates get elected because they are popular. One of the known drawbacks. But it's too late to go back to benevolent dictatorship (come to think of it, Jimbo was popular too...)

...and then I read your actual proposal at Wikipedia:Governance reform/Policy Committee version 1. This seems pretty much the opposite of what I thought Kirill was proposing. More comments here. PaddyLeahy (talk) 20:06, 4 May 2008 (UTC)

Decision methods

(If it's the latter, we'd be back to the same problem of having no usable method for causing the community to come to a decision.) Kirill ^(prof) 03:35, 5 May 2008 (UTC)

I personally would much rather see some form of direct democracy, or at least open door for direct democracy, so I wouldn't quite like if some committee (albeit elected) would ultimately decide everything (but it's a good start anyway). I think your argument is false - there needs to be method to elect the committee anyway, so there obviously is a method for community to come to a decision. I would like to ask you - if you oppose direct democracy (I am not sure) - would it be acceptable for you if the committee decides differently than what the majority of eligible voters (in the committee election) wishes? Samohyl Jan (talk) 04:38, 5 May 2008 (UTC)

I was contrasting the proposal not with a hypothetical system of direct democracy—I have no fundamental objections to that—but rather with the current system, in which nobody votes, and instead we sit debating things ad infinitum (or as long as the most persistent members of the opposition want to, in any case).

My main practical concern with fully direct voting is that we don't really have a closed voter set, since people can join or leave at any point. This means that while a referendum-style vote is good for settling a single question, it's difficult to use for formulating a "most popular" variant of a proposal, since there's no way of gauging voter feeling continuously. (Compare a committee version, which lends itself more to a classical consensus-like "work on it until we have a version which passes" approach.) This isn't to say that direct voting would be impossible; but I think we'd need to develop some system for it that would be easier to use. Perhaps some sort of dual approach, with a committee approving a preliminary policy that would then be ratified by direct vote (similar to how, say, U.S. constitutional amendments work)? Kirill ^(prof) 04:58, 5 May 2008 (UTC)

Thanks for answer, nice to hear that. I agree that almost any system is better than the current one. I would agree with a ratification system, and some other people proposed system like that in the above discussion too. As long as the committee would be willing to take input from the community, and community could overrule it (by sufficient enough votes), I will be fine with it. Samohyl Jan (talk) 05:24, 5 May 2008 (UTC)

That can be a telling point. Another possibility, which I have seen arise in other discussions, is that for whatever reason a person contests a given opinion, disappears for several weeks, during which everybody else tries to address that editor's concerns as well as they can, without benefit of that contesting person's input, effectively resolve the situation to there own consensus, only to have the original contender re-appear thereafter, say his concerns have not been met and there is no consensus, and start the discussion all over again. The process can repeat indefinitely. While I would acknowledge that there are specific situations when such would be unavoidable, and would hope that a model like those proposed would take them into account, such possible attempts to game the system would be particularly disruptive regarding policy debates. Creating some sort of system which would make such events less likely to occur would be to I think everybody's advantage. John Carter (talk) 15:07, 6 May 2008 (UTC)

So, why not also join in in discussions at Wikipedia talk:Consensus, where filibuster and prevention thereof is a common topic? (we're discussing it on the basis of flowcharts). --Kim Bruning (talk) 22:09, 16 May 2008 (UTC)

Foundation-l Posting by Eric Moller

A couple of weeks ago, Eric Moller (deputy director of the Wikimedia Foundation) made an interesting post to the Foundation-l mailing list, responding to the fuss about the recent board resolutions. It reads in part:

The Board has given the community a clear "go" signal to explore models of self-governance and decision making processes, be they councils, direct voting, committees, or other processes which work. This allows for the rapid, parallel evolution of mechanisms of self-governance and a "survival of the fittest" decision-making processes. That's a very real alternative to a top-down decision to explore one particular model (Volunteer Council) and, arguably, preferable.

Actually, this sounds to me like a definition of civil war but at any rate can be taken as support for the current discussion. With this in mind I've made Kirill Lokshin's original suggestion on the project page here (i.e. the "Wikipedia Assembly" model) a bit more specific. I find myself in total agreement with all the points Kirill has made on this and related pages, and unsympathetic with the various attempts to bend the idea away from being a legislative body. Whether Kirill (or anybody else) agrees with the specific changes I've made I don't know...any reactions? PaddyLeahy (talk) 13:04, 5 May 2008 (UTC)

I'm fine with them; the idea of allowing the current model to continue except where specified otherwise, in particular, seems like a very good approach, and would help avoid turning the body into a bureaucratic bottleneck for uncontroversial matters.

More broadly, I would tend to favor more traditionally legislative models over less legislative ones; but I view the establishment of any usable model of policy modification to be more important than the particular type, and so have no problems with some of the non-legislative approaches if the community is more comfortable with those. Kirill ^(prof) 13:31, 5 May 2008 (UTC)

As I think you probably know by now, I strongly oppose the creation of a "legislative assembly".

From everything I've read of your (KL's) concerns so far, all you seem to be after is "someone" to act in the position of an admin closing a discussion. As noted at consensus (and particularly at Consensus decision-making), a facilitator is needed to "close" as discussion.

I don't have large problems with that, presuming that such a person or persons is directly responsible to the community. (For example, the position shouldn't be "for life, and should be subject to some sort of review itself.)

So let's attempt to resolve your concerns without throwing the baby out with the bathwater : ) - jc37 14:44, 5 May 2008 (UTC)

Well, I don't think the two are all that distinct. Consider, for example, the practice of having multiple bureaucrats consult before closing a contentious RFA. I'd think a similar approach would be useful here, given the length and acrimony of some policy debates—we wouldn't really want to have things overly dependent on which individual editor closes the discussion—so we'd basically wind up with a group of people ("PolCom", for lack of a better term at the moment) that would get together and determine what the outcome of a policy debate was. But there is very little practical difference between this model (community debates, PolCom decides outcome) and the legislative one (community debates, Assembly decides outcome). PolCom would, in theory, only be enunciating existing consensus rather than "deciding" policy, true—but given a sufficiently large and equal-sided debate, it's quite possible to pull either outcome from it, so PolCom would in essence still be making the final decision.

So I don't think there's as much difference between the two approaches as might seem at first glance, particularly if PolCom makes its own internal determinations by some method resembling voting. In practical terms, either one is suitable, in my opinion, for making sure that policy debates actually have some conclusion, and are not controlled by the loudest and most obstinate editors; the differences between the two are mostly in the details of implementation, which are entirely open to debate in any case. Kirill ^(prof) 00:42, 6 May 2008 (UTC)

• Ah. That puts a whole new shape on it all. I'm actually getting rather fed up of the board's method of communicating with the individual wiki's, and their methods of getting what they want. POLITICS, n. A strife of interests masquerading as a contest of principles. ~ Ambrose Bierce in The Devil's Dictionary. Hiding T 13:35, 6 May 2008 (UTC)

Natural consensus, polls, and no committees, please

Discuss→viewpoints summary→viewpoints poll→discuss→consensus→modify.

I've tried writing several Council, Community Council, Community Committee, and Steering Committee proposals on my internal wiki, and for each one I've always felt a sense that they are not the right thing to do — they're either too bureaucratic, too complicated, too community-resource-intensive, or too-(something bad).

The core of the issue at hand seems to be that some editors are fed up with other users holding sole claim to certain policy pages. I ask: does this alone warrant the formation of a policy committee or policy review system, which would suck huge amounts of community resources? No. Take a look at this page, which is nearly 600 kilobytes long as I write; given 30 kilobyte articles, 20 (very substantial) articles could have been written with all of the discussion that's been generated here. Policy is apparently a "big issue", and I am not prepared to see it becoming the focus of the community.

Let's look at the big picture before painting even a small part of it.

• Wikipedia is an encyclopedia dedicated to bringing the sum of human knowledge to every person on the planet, free, in their own language.
• We, as a community, are here to further this goal, not obstruct it.
• As a community existing to develop an encyclopedia, we require certain rules to dictate what content is acceptable in the encyclopedia and what behaviour is appropriate in the community. Rules are needed in order to ensure quality of content and productivity of community; without clear rules, editors may be treated unfairly or inappropriately.
• If any of these rules prevent us from improving or maintaining Wikipedia, we are obliged to overlook them for this purpose and this purpose only.
• For the most part, these rules, hereafter referred to as policies, should document existing cultural practices of the project. However, some policies are not debatable:
  • content inserted in articles must be presented in a neutral and verifiable manner, and must be notable (in other words, not original research);
  • behaviour in the community must be collaborative and cooperative.
• Beyond these principles, requirements for encyclopedic content and community behaviour are open to constructive community discussion, but no more than discussion; in particular, policies, guidelines, procedures, systems, and groups may only be formed for the purpose of improving or maintaining the encyclopedia or the community (and I do not believe that either a policy committee or a policy review system would improve or maintain either the encyclopedia or the community).
• To extent, policies define (whether for good or for bad) practices that editors are expected to follow unless they have a very good reason to do otherwise. Thus, policy pages are important, and they must be controlled by the community, not by selfish individuals or cults.

Which, after a view of the big picture, brings us to the crux of the issue being discussed on this page: how do we, as a community, take control of policies (and guidelines, procedures, systems, and groups) out of the hands of selfish individuals and cults and into the hands of the community?

I would argue that we should do so in two ways:

• by blocking, aggressively, anybody who carries out modifications in a manner that deliberately goes against community consensus — in order to avoid accusations of administrator abuse, considerations of such blocks should be supported by a minimum of two (preferably more) neutral, uninvolved administrators;
• by encouraging — no, enforcing — a discuss→viewpoints summary→viewpoints poll→discuss→consensus→modify cycle for significant changes (where viewpoints summary→viewpoints poll→discuss is repeated as many times as necessary; it may not even be needed at all) (carrying out a regular, but non-binding, 24-hour poll ensures that no one can misinterpret consensus) (see the top of this post for a simple diagram demonstrating this process).

If a council, committee, or system was started sometime in the future, I would insist that it serve the purpose of encouraging, assisting, and enabling community members to contribute constructively to the project, either in the form of a pure Steering Committee or Content Committee which had a very minimal overhead and got on with building the encyclopedia after doing its job:

• a Steering Committee (my preferable option) would ultimately aim to steer the project toward its goal through resolving disputes (including policy disputes), advocating and representing between the community and the general public, actively developing the encyclopedia, and encouraging others to do all of these things;
• a Content Committee would act as a last resort, making binding resolutions to articles and other pages in accordance with acceptable, established community practices; this Committee could, in a binding fashion, resolve neutrality/verifiability/original-research issues in articles and consensus issues in other pages (such as policy pages, etc.).

Ideally, I would most definitely prefer that a new council, committee, or system was not started (although, of course, I reserve the right to change my position on this view); once we start thinking about the huge resource-sucking consequences, it becomes apparent (at least to me) that such a group would almost inevitably cause less editors to improve and maintain the encyclopedia and community. If some group or procedure was started, it would have to be very carefully planned and laid out to ensure maximum effectiveness and efficiency, minimal interference with other aspects of the community and encyclopedia, and recognisable bonus to the project.

Best and friendly regards — and make sure you remember what we're here for, folks. — Thomas H. Larsen 01:13, 6 May 2008 (UTC)

Just curious, how large was your wiki? There is a consensus among people who want this change that consensus system works very well for small communities, and only if people start to become largely unknown, it becomes a practical problem. Samohyl Jan (talk) 05:52, 6 May 2008 (UTC)

Sorry, I was not clear enough. I was referring to a single-user wiki installed on my GNU/Linux computing system — I use it for developing pages and playing around with. In other words, it's not an Internet wiki. — Thomas H. Larsen 00:38, 9 May 2008 (UTC)

Instead of polls, we've been using page editing as our primary method of testing consensus. In some ways that seems almost anarchistic, but at the end of the day when you examine it carefully, it actually scales better than polling, much of the time. (There are exceptional circumstances when you are forced into a poll).

FORCING the discuss then modify cycle has been very thoroughly checked out using flowcharts, and has also had some testing in the field. Unlike the Bold first model, it allows people to filibuster indefinitely. People who mistakenly (or deliberately but naively) tried to push this model have suffered a lot of embarrassment on en.wikipedia. If anything can help to prevent the ills people complain about, it is to ensure that people are permitted to be bold first. It is not surprising that that particular guideline has endured so long.

Samohyl Jan makes the traditional scaling argument, but I'd like to point out that we've already discussed numbers and scaling, and have found that the wiki itself scales very well, and thus that scaling issues have not been a major problem throughout most of the history of the wiki. There is a caveat there. A naive process designer is more likely to cause a scaling issue in his/her design today than they were 8 years ago, for 2 reasons. The first is that yes the wiki has grown, so you need to be extra careful. The second is that less people are aware of how the wiki works (so they are more likely to be naive). --Kim Bruning (talk) 13:37, 6 May 2008 (UTC)

I think you misunderstood my point. Polls can be used to determine consensus (only after discussion, of course), not facilitate filibusters. The "bold edit, revert, discuss" model, however, only works if both the editor and the reverter are both reasonable people; otherwise, that very same process can continue indefinitely. — Thomas H. Larsen 00:38, 9 May 2008 (UTC)

Mostly people are reasonable. Also, it's a tad dirty, but other reasonable people you meet via BRD can either pressure or even tag team out the rare unreasonable person. So so far I've gotten BRD to work fairly well for me, specifically as a tool to break filibusters and blocking-of-consensus. --Kim Bruning (talk) 22:13, 16 May 2008 (UTC)