Wikipedia talk:Governance reform/Policy and guideline review

Comments

Rather like the essay/guideline/policy hierarchy as defined here. If that were to be implemented, it would probably make the creation of completely new policy all but unnecessary, as there will almost certainly be at least one or two extant essays on virtually any subject. Personally I have some reservations about the comparatively small number of members, as it might not be enough to ensure that at least one individual in the group is necessarily familiar with all the variety of content we have, but I imagine the numbers are probably among the least "set in stone" of all the points in these proposals. John Carter (talk) 00:15, 5 May 2008 (UTC)

I don't see this as radically different from Kirill Lokshin's "Assembly", although the mood music is a bit different. To that extent I support it.

• Can you explain the difference between "current general usage" and "current community consensus"? To the extent that they are different, my problem with "community consensus" is that it often means whatever the person charged with determining "consensus" wants it to mean. That's why I'd just rather accept that the committee represents the community and let them decide without reference to any mythical consensus. (But probably that would be easier to swallow with a bigger committee, see below).
• If a criterion for acceptance is consistency with current usage, no "jump" policy change would ever be allowed (e.g. introduction of WP:FLAGGED).
• As written, the committee only has the option of rubber-stamping pre-written documents. What happens if the committee feels that a key policy needs to be changed? They demote the current version to an essay...and then we have no policy at all? I guess in practice someone writes a rival essay that is prompted instead, but this seems bizarrely procedural compared to just authorizing an edit to the policy. Maybe the last sentence provides a get-out on this.
• Do we need a separate discussion page as well as the policy talk page? We do this for FAC etc but (prospective) policy talk pages are largely about whether the policy should be "promoted" anyway.
• I'm not convinced about extending the remit of the committee from policy to cover guidelines and essays as well. This whole plan is (necessarily) very WP:CREEPy but this makes it quite a lot more so. Also, that enormously increases the workload.
• As Kirill pointed out, it is a bit futile to try to prevent such a committee from initiating policy, since members can always submit new "essays" to promote as simple community members. If you explicitly withdraw this right from committee members they will just get their friends/supporters to do it instead, and you have succeeded only in obfuscating what is going on.
• Agree with John Carter that 16 people is too small. As well as what he said, you want to avoid group think, i.e. the committee should contain more than one faction (ideally several).
• Who decides that candidates are "well versed in policies and guidelines"? Why not let the voters decide?

PaddyLeahy (talk) 00:54, 5 May 2008 (UTC)

Comments

My suggestions:

• Make it clear that the committee's decisions are binding.
• Allow the committee to review and approve/disapprove changes to the wording of policies/guidelines in addition to adding/deleting/promoting entire policies and guidelines.
• Make the member's terms 1 year instead of 2. This makes them more accountable to the community for their actions since their continued service is subject to a vote more often.
• Make it 25 members instead of 16 to ensure enough participation in the discussions by members. Cla68 (talk) 04:05, 5 May 2008 (UTC)

Some responses to the above

Since some similar things were brought up, it's probably just easier to answer in one location : )

(Note that some of the concerns above are actually already adressed in the text, such as the ability to review existing policy pages.)

"Binding" - I'm avoiding that word, since it's confusing to the intent. (Binding has a sense of finality which seems anti-wiki to me.) The determination of the committee would be "binding" in that this would be the only way a page could be marked as policy. Essentially by authorising this committee, the community would be relinquishing the ability to be bold in tagging a page as "policy". So in a sense, yes, it's binding, though only until the next time the page is reviewed.

Terms - I'm sorry, but I disagree. We actually shouldn't want the comittee members quite so politically answerable to the community. I would presume that we would want them to be able to make "the tough choices" necessary, without needing to be constantly "kissing babies". Also realise that there would be a "changing of the guard" of 4 members every 6 months. So yearly, half the committee changes. That should be enough, yet retain the institutional memory that should also be requisite with such a position.

And that goes along with the idea that the comittee needs to be some huge group of people. Honestly, 4 people could do this. (Imagine Kim Bruning, Kirill Lokshin, John Carter, and Hiding all at a table. Now imagine 4 more such people, and now add 4 more, and now 4 more. 16 is likely more than enough. - Personally, I actually think 9 would probably be more than enough...) After some discussion at the governance proposal talk page, 16/4 seemed to be the most likely to find agreement/consensus. Again, remember that 4 seats will be up for election every 6 months. Compare that to arbcomm: yearly elections, and terms of 3 years.

So if anything we should start small. If it turns out that more are needed, for whatever reason, we can always discuss increasing that amount.

I agree that sections of a policy page (not a guideline page, since those would still be boldly editable by the general community at any time) could be brought for review, I just am not sure how that could be phrased in order to weed out the potentially constant stream of frivolous requests. ("I don't like this specific sentence becuase of x, y, and z.")

As for the committee members being "well-versed..." - The whole point of the elections is to allow the community to decide. The presumption is that the election itself should act as that process.

And anyone should be able to suggest a page for review. Though presumably a member may recuse themselves from review if they feel they should.

• Can you explain the difference between "current general usage" and "current community consensus"? - Sure. they are as different as the community says or shows that they are : )

Consensus is what people agree upon. General usage is what people are doing. They are not always the same thing.

What I'm essentially suggesting is that the committee will weigh all those things and more, as appropriate, when making the review. As a review is a "case-by-case" basis sort of thing, we should leave that part as non-defined as possible, precisely in order to avoid WP:CREEP.

I hope this clarifies. Though I welcome further thoughts/etc. - jc37 06:03, 5 May 2008 (UTC)

So if general usage and community consensus are different, no policy would be approved! What I'm saying is that this is poor drafting. I think we both agree that the committee should be able to overrule existing practice and/or consensus if it believes it is necessary, i.e. ultimately it has a free hand. We both also agree that it should respect the existing system as much as possible, which in many ways has been incredibly successful. But I would prefer to see the free hand made explicit, rather that being authorized by a vague and contradictory remit which falsely suggests that this is primarily a judicial committee interpreting current practice and consensus, even though insiders will see that it is indeed a policy-making body. And given that such a committee is capable of re-writing its own rules, I don't see much point in trying to impose restrictions in the remit of the committee. Instead, again, let the voters decide. There is a strong conservative tendency within wikipedia which should ensure that the committee doesn't step out of line (and members who do will get thrown out at the next election).

On elections: with a 6-monthly term, we will be subject to a semi-permanent election campaign. I just don't see that as organizationally feasible. If we make the elections coincide with board elections we significantly reduce the overheads. On numbers, let's agree to differ. If this goes anywhere that will be hammered out later. PaddyLeahy (talk) 12:44, 5 May 2008 (UTC)

They're 2-year terms. Merely 4 different seats every 6 months. It's a way of staggering terms, rather than have the whole committee up for election at the same time.

• "So if general usage and community consensus are different, no policy would be approved!"

I don't see how a has anything to do with b.

I was listing some things which the committee should take into consideration. Not what the committee was "bound" to.

In general, those 4 things (at least) reflect what Wikipedia policy is. That doesn't mean that exceptions aren't possible. (And incidentally, personally I wouldn't "vote" for a member unless they had an absolutely firm grounding in actually understanding the appropriate application of IAR.)

And finally, I think perhaps you feel that it's semantics, but this isn't a legislative body. Note that the community is writing the pages, even in the cases where the committee is asked to supervise. (Compare to the task force requested by JW concerning irc.)

If I had to shorten this to its essentials (nutshell):

The committee is merely a replacement to an admin closing a "policy" discussion, and to the ability of any editor to be bold in tagging a page as "policy"; and provides a forum for discussion/review of "policy"; and provides a resource concerning "policy" for other bodies to draw upon.

The rest is merely the format/design/process in which those essentials are implemented. - jc37 14:35, 5 May 2008 (UTC)

I don't want to imagine the scenario like you presented above too vividly, as I have a rather pronounced aversion to violence in any form. My point in suggesting that we might benefit from having a larger number is primarily that I personally think that there are probably more than 16 "core areas" that wikipedia deals with, which would mean that there would probably be more than 16 people needed in such a group to ensure that there is sufficient knowledge and experience to be able to effective write any proposals which would include material relating to all those groups. If I am wrong in that, I would withdraw the complaint. But at this point I think that I may not be. And one of the disadvantages of a truly "democratic" selection process is that it would stand a very real chance of overrrepresenting one area of content, lets say television and film, pronouncedly over another field of content, which might be at least as important but perhaps with fewer involved editors, say mathematics or physics. I'm not saying that I necessarily have an answer to that problem myself, but do think that it probably should be addressed somehow. John Carter (talk) 15:01, 5 May 2008 (UTC)

While not speaking for other editors, as someone who can be fairly active in XfD, we don't presume that everyone has intimate knowledge, so much as "hope" that they've done their research. Same for this process. Also, by having the intermediate stage of the review process open to all Wikipedians, those more fluently "in-the-know" would hopefully enlighten and or share such information based upon their experiences. (Again, similar to XfD; and Arbcomm, for that matter.)

Incidentally, this is another reason to not have this be an enclosed legislative body. By leaving it "open", as merely a review board, we benefit from the knowledge/understanding/experience of all Wikipedians, not just those elected. - jc37 15:31, 5 May 2008 (UTC)

Have you thought of other classification schemes?

Policy/Guideline/Essay (henceforth P/G/E) is not entirely entrenched yet. Can you think of other classification schemes? --Kim Bruning (talk) 16:11, 7 May 2008 (UTC)

Sure. Probably an infinite number. But as the P/G/E classification seems to be current usage/convention, I was defaulting to that.

If I were to suggest what I would prefer, it would probably be usage of namespaces.

For example, I still strongly support an MoS namespace for all page "guidelines" such as MoS pages and naming conventions.

So if we had that and a "Policy" namespace, everything else would be "Stuff in Wikipedia space". (And don't kid yourself, there would still be a lot.)

That said, I don't foresee the idea of separate namespaces gaining concensus any time soon. So, I'm taking your advice, and attempting to work within the current system/conventions : ) - jc37 23:40, 7 May 2008 (UTC)

Well, I suppose you already know that I think that naming important documentation "policy" was a big mistake. ;-) (no matter how useful it seemed at the time) ^^;; --Kim Bruning (talk) 17:17, 8 May 2008 (UTC)

What other term would you suggest? - jc37 21:16, 8 May 2008 (UTC)

the wrong approach

I admit to a certain frustration about the process of changing policy, but I do not think it is a good idea to remove it from the community as a while. The decision about the status should not be delegated to a body who, it may turn out, will not take account of any views than their own. The use of established people here will have the effect of discouraging change, by discounting the influence of relative newcomers with fresher views. This is a proposal that will decrease flexibility and preserve stagnation. More specifically, this ignores the actual means of policy change--when a guideline no longer has consensus the guideline is very rarely rejected--it is usually changed until it does have consensus. I suppose this could operate asa way of forcing this sort of change just as Afd forces improvement--adjust it to meet our views or we'll try to get it rejected. But that would be looking at it backwards--the need is to find some other way to force improvements than to use such threats. DGG (talk) 00:04, 12 May 2008 (UTC)

I think there are several things you may be not understanding about this proposal. Not the least of which is community inclusion. Would you please describe what you think you see, and perhaps I can clarfy. - jc37 00:08, 12 May 2008 (UTC)

I tend to agree with DGG. All of a sudden you've created a new class of users, who have been granted power through elections. And they're allowed to pronounce "a clarification, an interpretation, a finding, or a principle (etc.) of Wikipedia policy" and have that pronouncement matter. Because, if it didn't, if it mattered as much or as little as anybody else's, we wouldn't have needed to elect them. Either you have empowered these people to be different from the rest of the community, or you haven't. If you haven't, there's no point electing them. If they can do something the community can't DGG has a valid point. If they can't, they aren't needed. Hiding T 13:43, 15 May 2008 (UTC)

Except it isn't that duality. There are two separate issues here: The process, and the participants in that process.

There would be an open process in which the community is welcome to participate.

The PAGR committee members would essentially be "empowered" by the community to "close" the policy discussion.

This is comparable to the recent irc request of arbcomm, by User:Jimbo Wales, as well as comparable to election committee members.

Essentially this is a case where it does and should take more than one person to "close" such a discussion. Arbcomm has said it would like to stay out of "making policy". So this is a way to have a committee which would be involved with that.

And it should deal with the "governance" concerns by others, without leaving the concensus model behind (as several seem to wish). - jc37 00:22, 16 May 2008 (UTC)

If these committee members are able to "close" a policy discussion, I take it this means they would have the ability to force a decision against the will of some participants in the debate? Hiding T 09:17, 16 May 2008 (UTC)

(Looking around, looking for the bear trap...) Um probably? Perhaps even "against the will" of some of the members. Compare directly to Arbcomm. - jc37 18:53, 16 May 2008 (UTC)

Then it removes power from normal editors, and as such I oppose it on the basis of what that entails. Comparing it to arbcomm is especially bad, because look at how far arbcomm has meandered from its stated purpose, and look how clandestine its work is now. The awful thing is, if arbcomm were better fulfilling their role, a polcomm would be unnecessary, because what is harming the consensus model is disruptive editors, and arbcomm is meant to be the body that deals with disruptive editors in a binding manner. Instead of instigating another committee, we should work out why arbcomm is failing and look to correct those faults. Otherwise we will potentially duplicate the issues we have, and possibly ramp them up exponentially. Hiding T 10:15, 19 May 2008 (UTC)

Not to get off on too much of a tangent, but arbcomm seems to me to be caught in somewhat of a catch-22 for exactly the reasons you state.

"Fixing" arbcomm would be simple if they were pushed a little in direction Y. But then group B gets upset, and says that they really should be pushed more in direction C. It's as if squabbling children were allowed to tell their parents how to adjudicate and discipline. That's the problem with arbcomm. However, because we're consensus-based, that's the trouble. It's a social dynamic that requires a "parent" to finally close a discussion. And what we're starting to find is that by saying adminship is "no big deal", we're running into problems with admins being closers of the "big stuff". (However each person wishes to define it.) And so they run to arbcomm to be the "ultimate closers". Well arbcomm won't close a policy discussion that way, and so now we have a vacuum. And as nature abhors a vacuum, we have problems on those front. Which is why, I think, that we see a few arbcomm members starting the governance discussion in the first place. I personally think the simplest solution is to create a sister committee to arbcomm to deal with policy. It keep the arbcomm's hands "clean" of policy discussions, and which allows them more "plausuble deniablility" (and "suspension of disbelief", I suppose). Which is what they would prefer, so that they can focus on behavior and not policy. If you have a better solution to deal with the vacuum, I'm all ears. (But, as you know, I'm strongly opposed to all these suggestions of democratic governance, such as creating a legislature.) So I'm not sure what else you might suggest. But it's clear from watching and reading, the "status-quo" is going to eventually sink Wikipedia, if not dealt with. (In which case, arbcomm will get pulled more and more into questions of deciding policy.) - jc37 21:20, 19 May 2008 (UTC)

Well, I have a couple of issues here. First, I've never understood why things need closing. Okay, I can understand WP:RFA having a deadline, but beyond that, I guess there's a point to XFD needing a deadline, but beyond that... enforcing a decision or closing a discussion appears to indicate a a decision is required and necessary, and contradicts WP:CONSENSUS. Consensus is not something which is declared by a bloke or a woman in fancy robes and a wig with a hammer, or whatever your particular fancy, it's decided by what the community will tolerate. And that's really wehere a lot of issues come in, when people attempt to enforce a deadline when one isn't really needed. Getting back to arbcomm, I disagree entirely. Arbcomms mandate is actually fairly simple; they simply have to look at an editors behaviour and work out if it is in keeping with WP:CIV, WP:AGF, WP:CONSENSUS and WP:EP, or whether they are disruptive. It's that simple. They don;t need to rule on policy or anything else. They simply need to work out whether any given editor is disruptive, and if they are, come up with a solution. The problem is, there's been politicking and slippage at arbcomm. They introduced the idea that the community didn't have to respect an arbcomm decision, which is understandable, but defeats the point of arbcomm being binding. At that point the contract was broken and arbcomm became somewhat redundant. The other mistake they made was condoning disruptive behaviour because they agreed with the goals. That lost arbcomm a lot of respect, since it became apparent they were biased. Now, given those issues, I cannot see how any other system can work. The only real method we can have is the one we've got now, which does tend to scale, as long as people who disrupt it are dealt with effectively. The problem is, it requires a lot of intelligence by those involved. They have to maintain and engage, and decide which editors are not interested in building a consensus and effectively sideline them through refusal to engage. With regards discussing, we perhaps need to look at some form of scaled polling, by which I mean getting a respondents actual thoughts on a scale of 1-10, or having sections along the lines of "Will not accept", "Would prefer it not to be implemented", "don't know", "Prefer amendments", and so on. How many proposals have failed even though consensus was behind the broad point but the dispute was over the details? That these fail indicate that there is disruption operating within consensus, possibly by those seeking to impose a deadline. There is no need to close a debate. Either there is consensus, or there isn't. It really is that simple. The problem is when people invest so much in their position that they have to see it implemented and refuse to consider the weight of those against them. Those were the very issues our behavioural policies were meant to guard against, which arbcom are supposed to settle. Arbcomm is the reason for this problem; the solution lies in creating an effective method for dealing with disruptive users. Unfortunately, it appears there isn't one. They all involve points the community is loathe to enact, either through individual responsibility or the removal of said. The interesting thing is how well we are establishing an elite within a project founded on the idea that there should be no elite. Through the refusal to countenance opinions that differ with our own, through the decisions to enforce outcomes when none truly are needed, we have driven contributors away, we have instigated hierarchies and we have decided that we are no longer the encyclopedia anyone can help to build. The funny thing is, the people who voice most in defending that we are are those who do most to drive them away. Arbcomm was a mistake. It should be abolished, but we should not enact any other body in its place. The community should instead of shirking its responsibilities, engage with them. It should appreciate why we have behavioural policies and it should make sure they are respected. Then, many issues would disappear. That's how I feel anyway, and why I simply cannot countenance establishing anything of the kind described here. Hiding T 13:49, 20 May 2008 (UTC)

I've read over your response several times. What keeps coming to mind is Animal Farm. IN the beginning, Beasts of England, and the communal discussions resolved (roughly) by consensus, as "all animals are equal" (among several other basic "policies"). In reading your post above, you almost directly refer to how the farm changed, including the development of an "elite". ("All animals are equal, but some are more equal than others.")

Back to Wikipedia... The problem, of course, is that people are involved. With all their own personal foibles and biases, and expectations. Everyone seems to have their own set of "IWANT"s. So the question is how to guide that energy into the positive construction of the encyclopedia. There simply need to be "conventions" so that everyone can work towards similar goals, rather than constantly having editors working towards divergent goals, leading to the work of one or the other (or both) being undone. That also, unfortunately, leads to voluminous amounts of guidelines, which most editors probably won't read prior to editing. There probably should be a BRE (Bold-revert-educate (point to guideline)). Teaching, or at least indicating the "why" of a certain guideline I think would be much better than the typical RBI responses. (Though of course, we would be presuming that editors are well-meaning, which often isn't the case, unfortunately.)

So that's essentially why we need "someone" to close discussions: People are involved. I just think that (as you note above) that closure requires a bit of intelligence, or rather a bit of Solomonic wisdom. Anyone can "count votes". The question is: can they read for content? And if adminship is "no big deal", and we don't require such wisdom when granting "the tools", aren't we being self-defeating? It seems to me that a crisis is waiting around the corner. Look at the current BLP controversy. Honestly, I think that we may have to eventually remove all BLP articles to either their own namespace, to wikinews, or to their own wiki. It's probably Wikipedia's number one problem right now, and it's ongoing.

And that's the obvious troubles with BLP. There are innumerable conventions on WIkipedia. And note the several factions who wish to do away with consensus for a more "democratic" system, such as "vote counting". The vast number of DRV noms which happen because someone didn't like the fact that the admin actually did read for content and didn't count votes is staggering. And if you wish to talk about an "elite", consider that the "DRV regulars" are the only ones who the community seems to accept to close "the big stuff". So in a sense, we already have such an elite. (Indeed, at one point I suggested that all of the regulars be nominated for bureaucrat. They might as well make it "official" in some way.) Now they don't intend it to be that way, but that's roughly been the way things are done. And isn't "common practice" one of the definitions of policy? : )

I realise that in an ideal world, we shouldn't have to have such measures. But this isn't an ideal world.

It's my "hope" that by having the "elections" be Wiki-wide, that at least, each member "might" be vetted to be a bit wiser than those who are currently pushed through RfA, as "needing the tools". And hopefully prevent an oligarchy-in-the-making. So far, the only other solution I've seen that's likely to "pass", is Arbcomm taking over being the "closers" to the "big stuff". And that will lead to greater troubles than this proposal. I tend to like the idea of interpretation of behavior and interpretation of policy being the demarcation line of "separation of powers".

All of that said, I'm really starting to doubt that any change such as this will be implemented, unless someone "important" (whatever that means) decides to be bold, and the implementation manages to survive the typical community upheaval. But then, after all, rollback went through, so I suppose anything's possible. - jc37 18:04, 21 May 2008 (UTC)

It wouldn't work otherwise. Even so, consensus also does require the ability to say to some people: your view is so far away from what the rest of us think, that we need to disregard it. A system which cannot do that in some manner becomes government by Liberum veto, unable to accomplish any positive change that might disturb anyone--well designed to maintain the position of whatever group has the power in the beginning. A consensus-based system needs to be designed to force compromise, but to still be able to proceed if compromise is impossible.

Thus the rule that arbcom refuses to act unless everything else has failed--their acceptance by the community depends upon that--upon the community's knowledge that the alternative is anarchy, and that a final decision of some sort is necessary by some means. I think most of the dissatisfaction with arbcom over the past year has been not because of a tendency to assume unwarranted power, but the disinclination to actually make such a final decision,and to return the case to the community without any real decision. DGG (talk) 16:09, 17 May 2008 (UTC)

That's part of what I was saying above. Some feel that they don't go far enough, and some feel that they go too far. It's all a question of implementation.

We ask these volunteers to make the tough decisions for us because, for whatever reasons, we're unable to come to agreement ourselves, and feel that some resolution is necessary.

And honestly, that's the same of what we're expecting of those who close discussions. The difference, of course, is that there is no "DRV" for arbcomm. That and while they tend to follow previous precedent, as Consensus can change, they are by definition not bound to precedent. (Which seems to drive some of the lawyers, and wannabe lawyers, in the communty more than a little nuts : )

I do find it interesting that discussing this has turned towards discussing the usefullness and efficacy of arbcomm. Which is probably a good thing, as the proposal compares somewhat to that committee. - jc37 18:18, 21 May 2008 (UTC)

The practical control of he community over arbcom is that one-=third of the members need to be elected each year, and that there is wide participation in the vote. DGG (talk) 05:51, 23 May 2008 (UTC)