Wikipedia talk:Governance reform/Archive 2

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

Proposed additions

I believe something along the lines of the following could be reasonably added.

"Wikipedia has always been an encyclopedia which anyone can edit. The freedom to edit expressed in that statement will remain a cornerstone of wikipedia. 'That government is best which governs least' is a position which will be taken if reform would take place. There would be no intention to create any more policies or guidelines than are absolutely required by circumstances. However, there are now, and have been in the past, situations when policies or guidelines needed to be enacted, and with some speed. It is for these instances, and these instances alone, which the policy making body is being considered for creation. Other events, recently and in the more distant past, have also indicated that it would be useful for other, more formal procedures and practices to be established in some instances, to prevent the possibility of abuse of the system. The fundamental goals and ideals of wikipedia would remain the same should this proposal be enacted. However, it is clear that some changes are clearly needed, and it is only to address those required changes that this proposal is being made."

OK, I know I repeat myself more than a little, and that I tend to use too many words as often as not, but I hope at least the idea is clear enough. John Carter (talk) 23:46, 24 April 2008 (UTC)

Um, more formal procedures and practices typically allow easier abuse of a system. This seems to be counter-intuitive to some people, but that doesn't make it less true ^^;; --Kim Bruning (talk) 02:31, 25 April 2008 (UTC)

more formal procedures and practices typically allow easier abuse of a system - I disagree. In the real world, the rule of law is prized for a reason; having formal rules, which must be obeyed by everyone (even those in power), ensures that those in power are held to account and cannot act arbitrarily. If punishment is handed out without reference to the rules, then people cannot reasonably conform their conduct to the rules and thereby escape sanction. (See the work of Joseph Raz for more on this topic.)

In the specific context of Wikipedia, it is a lot easier for a new user if they can just look at a page of rules and find out how things work. If the rules are unwritten, or made up by the admins as they go along, then new users can find themselves blocked without realising that they did anything wrong. (This actually happened to me on Wiktionary, where I created an account some time ago as wikt:User:Eric the Gnome in order to see how newbies were treated. But that's another story for another time.) Policies and guidelines are therefore inherently a good thing, though they should also be kept in check. Walton^One 08:22, 25 April 2008 (UTC)

Documentation of best practices is always a good thing. I agree so far. The part I disagree with is that having strict, prescriptive policies is a good thing. A good example is declaring things "non-negotiable". While several policies were marked "non-negotiable", people started "non-negotiating" their way out of following NPOV, for instance. Oops ;-) --Kim Bruning (talk) 08:25, 25 April 2008 (UTC)

I don't know quite what you mean there; NPOV is indeed a non-negotiable policy (and rightly so, since it defines the character of the encyclopedia). Obviously, in practice, the way NPOV should be applied to some contentious topics is debatable and negotiable (the evolution-creation controversy, for instance), but this doesn't mean people are rejecting the policy. I've never, so far, seen anyone justify their changes to an article by asserting that we don't need to follow NPOV (rather, they tend to assert that their opponent is not following NPOV).

The point of having strict, prescriptive policies is that everyone is bound by them, from the most experienced admin to the newest editor; and if an admin violates policy, they can be held to account. It's exactly the same principle as having strict, prescriptive laws in real life. Do you think the United States Constitution should be downgraded to a "documentation of best practice"? Power, wherever it is found, has to be held in check. Walton^One 11:24, 25 April 2008 (UTC)

Kim seems to be me to be maybe arguing a case which no one else is arguing against. There is no intention that I know of for this to become a bureaucracy. However, that is not the same as saying that there is not a need for more policies than we currently have. One of the current, unfortunate, cases before ArbCom is about a case where an admin, probably with no intended unacceptable motives, acted in a way which just about everybody else agrees was not in the best interests of wikipedia. In such a case, a policy or guideline (I personally question how much functional difference there is) relevant to an admin acting in a situation where he could be perceived as having a COI would be one that just about everyone could agree with. A few policies/guidelines have been suggested to deal with such situations, because there is no reason to think they would not recur if such policies/guidelines do not exist to mitigate them.

If ArbCom, Jimbo, or, potentially, a clear consensus on one or more of the Admin noticeboards indicate that there is just cause for a policy/guideline to be created, then I can't see any reason why one cannot be proposed. If through examination that policy or guideline should be found to be either inherently unenforcable or otherwise problematic, then clearly that policy or guideline will be rejected. Also, if someone were to violate one of these polices or guidelines, with extant mitigating circumstances, there wouldn't necessarily be serious consequences, although that possibility would clearly exist.

And I don't know of anyone suggesting that anything but the most core policies are necessarily "non-negotiable", because any policy can be and generally is changed. In fact, creation of such a policy board would probably help ensure that the extant policies and guidelines are more likely to be kept "up to date". With a sufficiently broad base, it could also help ensure that these new policies/guidelines don't create more harm in some area than good in general.

I don't imagine anyone is thinking of this as being a major editorial responsibility, to the degree that being a member of ArbCom is. With sufficient members to divide the responsibilities, it would leave most free to do other activities most of the time. But it would also permit us to not have to repeatedly invoke policies or guidelines which may not clearly be appropriate in the situation, like invoking IAR in cases of "emergencies" or problematic situations which could be and are anticipated, and should be addressed, but haven't been yet because there hasn't been sufficient focused attention to the matter for a policy or guideline to become enacted. John Carter (talk) 14:12, 25 April 2008 (UTC)

"Kim seems to be me to be maybe arguing a case which no one else is arguing against.". Tango argues that we should abolish consensus. Walton argues here that we should adopt hard rules. Together, this would be the start of some of the most sweeping changes since the Nupedia project set up a wiki. (And quite frankly, it would be a reversal). --Kim Bruning (talk) 15:11, 25 April 2008 (UTC)

Then your comments should probably be directed clearly toward them, rather than posted as a response in a thread which kind of indicates the exact contrary. And I would expect, in the initial stages of any discussion, there would be any number of ideas. That is to be expected, and probably even encouraged, as part of an open process. But to write a response to one proposal, which seems to be expressing concerns regarding the possible consequences of other proposals, strikes me as, well, odd. And I read WaltonOne's statement above, which dealt exclusively with NPOV, which basically already is a non-negotiable policy, although the exact phrasing of it is and always should be open to question and amended, as circumstances require. And to put together just those two ideas, out of all that have yet been proposed, and say that, on the basis of a minority of comments on a new proposal, the proposal should be rejected, is at best dubious thinking. John Carter (talk) 15:45, 25 April 2008 (UTC)

Another approach

I think a key approach is missing from the page, which is the actual current consensus process. Might want to add that. :-) --Kim Bruning (talk) 05:03, 25 April 2008 (UTC)

Well, it hardly needs to be proposed again, does it? ;-)

More to the point, Wikipedia's consensus model has certain failure scenarios that the community has been unable to address by mere incremental modification. The consensus-building process inherently relies on the BRD model—in other words, gradual change is achieved by a series of changes which are contested and individually resolved. The key here is the "Revert" portion; if it is replaced with something other than merely a reversal of the original change—such as, say, a block or desysopping—the model begins to break down quite rapidly. (Or, alternatively: the model works only insofar as the initial bold change can be performed without costing the user making it their ability to participate in the discussion and/or the project.)

This is a very significant practical issue with policies that govern (or describe, if you prefer) the use of certain privileged access rights. Ordinary policies—which is to say, those dealing with editor actions—are fairly amenable to such individual-driven change. Policies dealing with administrative tools are less so; people being unduly bold with them tend to wind up desysopped. At the far end of the scale, the policies dealing with things like oversight and checkuser access are effectively set in stone, since any deviation from them invites all manner of highly unpleasant consequences.

Consider, for example, the current debate on whether inactive admins should have their access to the tools removed. Such a policy (not policy document, but rather actual in-practice policy) cannot feasibly develop through the BRD approach; no steward would "boldly" desysop a few users merely because they thought it a good idea. And so any attempt at change devolves to establishing consensus a priori—which, for a change of this magnitude, draws in more people than can actually come to a reasonable consensus through ordinary discussion, leading to the predictable no-consensus referendum outcome.

(More generally, one can argue about whether certain policy should lean towards the prescriptive rather than the descriptive side of the spectrum—conduct policy, if made truly descriptive, would lean rather too much towards an unpleasant lowest common denominator—but that's really secondary to the practical issues with the current model.) Kirill 08:10, 25 April 2008 (UTC)

I agree wholeheartedly with Kirill. The wiki-"consensus" model is fine for editing articles, and it works very well on a small scale. But when we have to co-ordinate policy changes across a vast community with thousands of members, "consensus" doesn't work, and we need a more formal and regulated system. Walton^One 08:17, 25 April 2008 (UTC)

At 2,346,917 articles; 12,895,573 pages; 218,874,408 edits; 782,443 media files (excluding commons); 6,970,823 registered users; of which 1,538 have administrative tools; over a period of 8 years numbers will have changed by the time I hit submit, I kind of like your concept of small scale. O:-) And that's just what has worked up till now. You're arguing it won't continue to work in future? Heh. :-) --Kim Bruning (talk) 15:19, 25 April 2008 (UTC) The same argument was made around the time when we had just 1M articles. I guess we managed the next million without too much trouble. I figure the next million after this will work out too, we're already partway there. "Wikipedia can never work"; "Wikipedia is doomed". Yeah right! :-P

The small scale here is the 4 people discussing some article, as opposed to the large scale discussions of policy. There is an emergent effect that allows those small interactions to scale to something extraordinary of course, unfortunately for policy no similar emergent effect exists. (although I am slightly interested in the old Wikipedia:Delegable proxy scenarios) - cohesion 22:33, 25 April 2008 (UTC)

Please, no more "Wikipedia is not a democracy" BS

I hate it when someone says it. Originally, this phrase was meaningful, and the meaning was, "ok, we cannot really tell if people vote twice, so we have to be careful about that, so we cannot decide things by simply voting". But now it's being used as a "victory" argument, when some minority wins, regardless if majority had a different opinion, especially in more subjective matters (and there is lot of them). Also, as in this proposal, it is used as an "obvious statement" which protects the proposal from outcry of admins, who feel democracy is threatening their position (and if you don't believe many many admins feel that way, see informal vote about admin recall of RfA page - large majority of normal users wanted such feature, but large majority of admins were against it).

Wikipedia needs more democracy, that's it. Democracy is a natural extension of consensus model, which works fine for small communities. Some will say that democracy means rule of the majority. Yes, but the only other possibility is rule of the minority. Which one do you prefer? I prefer rule of the majority to rule of the minority, because majority is more probably right and more probably, I will be a member of it. To prefer minority rule over majority rule is very elitist position, contrarian to the spirit of good faith in people on Wikipedia. I know that losing a vote hurts, but if you are reasonable, you will win most of the voting because of the majority rule.

So, I wouldn't bother building a legislative body - they tend to fail in the real world, and will probably fail here too (there is a big difference if you vote for a person and if you vote for an issue, because in the former, you have to implicitly trust that someone, in the latter, you don't have to trust, which is a good thing).

Instead, I would like to see 2 things:

1. Establish very precise rules for decision-making by voting when consensus cannot be reached. This means rules for who exactly is eligible to vote (based on number of edits, say 200, or some other metric - it doesn't have to be perfect, but it does have to be exact), how to make proposals for voting, how the votings have to be announced (so people would know about them), how long they will take and how to penalize the sockpuppets. ArbComm should probably decide contentious votings, and that should be it.

1. Unrelated thing, but also good. Further divide the powers of administrators, so that banning and article deletions/protections has to be done by different people.

Samohyl Jan (talk) 06:26, 25 April 2008 (UTC)

So you'd like to replace the consensus by a majority? As a member of a consensus group, you have the ability to influence every decision in detail. As a member of a democratic minority, you might not be able to influence a decision at all. So you're essentially proposing that we effectively disenfranchise you in particular situations, with no appreciable gain for yourself. ;-)

Are you sure that that's what you really want? --Kim Bruning (talk) 07:55, 25 April 2008 (UTC)

Kim, that's a false opposition. Consensus and democracy are not opposites. When you have a small organization, where people can take the time to consider the issues and negotiate and resolve disagreements, consensus process can be exhilarating. But as years pass and precedents pile up, not to mention decisions that are made and considered the "status quo," true consensus rules can become highly conservative and oppressive even in small organizations. Wikipedia doesn't generally seek true consensus, but this happens even with the rough consensus that we use. Basically, when the status quo favors an active group, that group can block any changes. Essentially, what started as consensus rule becomes minority rule, where a minority, in the name of consensus, can continue a practice only supported, in fact, by a minority. The conclusion I came to, years ago, was that consensus was *important,* that the degree of consensus found is a measure of the success of the organizational structure, but that, ultimately, the decision of what decisions to make by majority rule and what decisions to make by other processes must bge up to ... the majority. If a majority are foolish, they will impose their ill-considered opinions on the minority. But if they majority is wise, it will respect consensus and seek it. The trick is: how?

In order to function efficiently without becoming a tight oligarchy, routine decision-making must be broadly distributed. This is how Wikipedia works. Decisions are made by rough consensus of those sufficiently interested to participate. It works well. Sometimes. Sometimes not. It is when it does not work well that something else is needed. we have escalating circles of dispute resolution, but they can become mob scenes, deliberation becomes practically impossible, and there is little way to determine if the results actually enjoy consensus. I.e., broad consensus. What happens instead is that increasing numbers of editors become burned out and disgusted with the process, which can take extraordinary energy to make even small decisions. The ideas that my friend introduced here were designed to deal with exactly this problem. It allows ad-hoc estimation of large-scale consensus based on small numbers of participants; it would allow what is basically the same structure as we have to continue to function when it works, with larger-scale involvement only as necessary. It does not introduce any bureaucracy, no elected offices, but, quite possibly, certain new practices and procedures for advising community servants. And for advising the Foundation, when necessary. This is WP:PRX, if the experiment is tried and applied. It is not about making decisions by voting. It's about estimating the consensus that any proposition might enjoy, if everyone had the time to actually become involved and make an informed decision, and it does this without elections or other contested process. Gee, I think it's a great idea. Wonder why so many wanted, immediately, to totally crush it? --Abd (talk) 16:13, 25 April 2008 (UTC)

No, I think Samohyl Jan is essentially right. The traditional nebulous concept of wiki-"consensus" works for a small community where everyone knows each other. It does not work for a project of this size and scale. What ends up happening is that those with power - usually the admins - make the decisions. And as regards policy changes, nothing ever gets done; it's impossible to build a "consensus" from the thousands of active members of the community. If we ratified policy changes by a majority vote, there would at least be some policy changes, and we could start fixing some of our broken processes.

I have been arguing for a long time that we need to end "Wikipedia is not a democracy", in its present form. It should probably be replaced with "Wikipedia is not a nation-state", emphasising the point that we are not primarily a political community, and that the encyclopedia is more important than developing rules of governance for their own sake. But democracy, in the form of majority voting on contentious decisions, is the only reasonable direction that this community can now take. Walton^One 08:15, 25 April 2008 (UTC)

Size arguments thoroughly debunked on this page already :-) --Kim Bruning (talk) 08:21, 25 April 2008 (UTC)

No, what's up on the page is a bunch of mildly interesting but not very meaningful statistics. It doesn't really make sense to simultaneously argue that editing a policy page is not prescriptive or significant to the policy itself, but also that the number of people editing that page has any meaningful correlation to the number of people involved in discussion of said policy.

Consider, for example, the blocking policy. It intimately affects hundreds, even thousands of editors every day; it is regularly debated on dozens of pages—user talk pages, noticeboards, arbitration cases; it lies at the root of dramatic conflicts among groups of editors; and yet most of this occurs away from the policy page itself! Certainly, neither the actual blockers nor the actual blockees are likely to be editing the policy (the former because they've presumably found what they needed in it, the latter because they can't). Kirill 08:58, 25 April 2008 (UTC)

Kim, I think I haven't been clear enough in explaining what the "size argument" is. The data shows there are somewhere between 10k and 30k active editors on the site, depending on how you measure "active", as well as hundreds of thousands of occasional editors. The "size argument" is that the community has reached an overall size where some sort of representative system is the only responsible and effective way to make difficult decisions. Certainly, if everyone stays away from policy pages then things will look nice, but what that means is that instead of having a responsible, deliberative, representative group writing policy we have a self-selected, and often biased, group doing so. On the other hand, any time even 250 people decide to participate in a single discussion, no consensus is likely to be found.

The number of editors per article isn't relevant to that argument. You continue to claim the size argument is debunked, but I don't think you have actually addressed the argument that is being made. — Carl (CBM · talk) 12:22, 25 April 2008 (UTC)

In the response to Samohyl Jan, Kim is right that, by advocating a representative system, individual editors would have less "power" in policy discussions than they do now. But I don't think that corresponds to being "disenfranchised". Any system we set up will need to have very strong provisions to protect significant minorities, and I would be happy if the deliberative group used a more formal consensus process, in which any one member could hold a proposal for discussion indefinitely. But the main concern that I have is the opposite of Kim's. Kim is concerned that a representative group would use policy making as a way to gain power. I am concerned that there are groups of individual editors who already do this, and would prefer to see them replaced with a more responsible and accountable group. — Carl (CBM · talk) 12:29, 25 April 2008 (UTC)

That would be if it would be a solely representative system. I think this one is being proposed as being one in which representatives ultimately decide, but also one in which the broader community has a period in which it can discuss and express its concerns, much like the current ArbCom setup. That would allow all parties, representatives and otherwise, to have a very clear and noticable say. Regarding the "single-veto" proposal above, I might not go that far. If, however, the "veto" were based on the proposals flawed approach to type of subject or material, I have no doubt that the proposal would be altered to take into account the variations depending on subject, etc. And I agree that, right now, there do seem to be a small number of editors involved in writing policy, partially because the policy proposals aren't as obvious as they could be. By in effect making such policy writing better known, through elections, and probably more active discssions, the policy proposals would probably be better known and, very possibly, get even more input than they get today. But any step to bring it more transparent is probably a good thing, and this would probably help serve that goal. John Carter (talk) 15:39, 25 April 2008 (UTC)

So you would wish to discard consensus can change too? Ouch.

Also, note that advertising things too widely in a short timespan will basically kill any process (see discussions wrt dunbar's number earlier on this page). Don't Do That.

So if you first drag in more people than you can handle at once, and then try to fix that by tacking on a system to handle them, and then tangle with the problem that that system disenfranchises them to an extent... this is getting rather breathtaking. I think I need to sit down. --Kim Bruning (talk) 15:46, 25 April 2008 (UTC)

In response to Kim: As have been already noted above, democracy and consensus are not opposites, so it's not replacing consensus. Consensus, by narrow definition, is agreement of all parties, so if there was consensus and we would then vote about the proposal, everybody would vote the same. It's hence a special case of majority support, in this case 100% majority. Now in broader definition, it also means accompanied discussion and decision based on merits of the arguments. I don't intend to replace the discussion by voting, in fact, I think the rules should state clearly how long the discussion should take place before voting (but also, after how long discussion there should be voting, if there is disagreement). However, there is a problem - there has to be a body, which ultimately decides which arguments have good merit and which have bad merit. Obviously, if there is no such body, anyone can come along and say that his argument has merit, so there is just chaos. So you need such a body, and it also have to be completely neutral, with no interest in any issue. So the simplest such body is just to take fair aggregation of view of all interested people, ie. decide by voting. Any representation just murkies the idea who actually decides, so I oppose it. If the representatives should always follow the will of those who voted them in, then they are useless - we can as well decide directly, without need to trust them. If not, then the system is not fair, because then they don't represent the interests of whomever voted them in. So I want a direct democratic system because it is simple, fair and transparent.

Also, I would like to point out again as an example the proposal of admin recall. If there would be voting, it would be already decided in favour of recall. But there isn't such process, so anyone can come along and claim this or that proposal doesn't have consensus, because arguments on that side are wrong. In the meantime, we have a system that majority (including me) opposes. It's nice that you talk I am part of the consensus group, when I am not, because there is no consensus, just status quo. I don't see how this would be different (from my perspective) if I were overruled by majority, except that in the latter case, more people would be happy. Samohyl Jan (talk) 17:50, 25 April 2008 (UTC)

In addition, I would like to note that the democratic decision is not an end of the world. Fundamental property of democracy is reversibility - every law can be later redone (except laws for democratic process, obviously), if the people wish so. So if some minority would be overruled by majority, there would be a specified period (say, half a year), after which we could vote again about the issue (if, say, 5% people voting previously would want to), and this time, the arguments from minority could convince the majority. So this rule would be in sharp contrast to current practice which says "consensus can change", but doesn't say how to objectively recognize it has changed (for example - if there was 100 people coming, one each week, and saying, I don't like it, and each week, the same group of 50 people says, we like it in the response, is that a change in consensus?). That's precisely why such rules are needed, and majority voting itself is just one of them. Samohyl Jan (talk) 18:04, 25 April 2008 (UTC)

The BLP claim is not true

The BLP claim is not true. There was a community consensus for it. See User talk:WAS 4.250/Archive 05#Talk:WP:BLP for details regarding its progress from creation to being policy. While it is true that the timing of its becoming policy was due to a push from Jimbo, none-the-less there was a community consensus for it and in time it would have been standard best practice and then elevated from guideline to policy as a descriptive policy. Jimbo pushing it to move from guideline to policy when he did sped up the process and thus made it a prescriptive policy. But there was community consensus for it and Jimbo's involvement merely sped up the process. "Forced on us from above" is nonsense. I created the initial proposal and I've never been an admin (and don't want to be). WAS 4.250 (talk) 12:39, 25 April 2008 (UTC)

Your archived post there ends with "and its a policy because Jimbo wanted it to be a policy". Conversely, my impression is that WP:ATT is not policy only because Jimbo decided to yank it; if he had supported the policy tag on it, the tag would have remained. But we don't have any system other than Jimbo to promote a disputed proposal to policy; because as long as it's disputed, it won't become policy by consensus. In these situations our "consensus" system comes down to waiting to see which side of the discussion grows tired first, which I find very irresponsible. — Carl (CBM · talk) 12:50, 25 April 2008 (UTC)

The context of that archived post was me describing what happened when and why to someone interested in that. Above I flesh out "and its a policy because Jimbo wanted it to be a policy" in the context of a claim that it was forced on us explaining 1)it did have consensus 2}we normally promote a guideline after it becomes standard best practice rather than before 3)I was referring to when not that it was made policy so "and its a policy because Jimbo wanted it to be a policy" should be understood to mean "and it became a policy when it did because Jimbo wanted it to be a policy". WAS 4.250 (talk) 13:33, 25 April 2008 (UTC)

Sure, it wasn't forced on WP out of the blue. Jimbo's seal is more useful for taking something that has significant agreement but not consensus, and convincing those who disagree to back down. I think that happened with the BLP policy but not with ATT. — Carl (CBM · talk) 13:38, 25 April 2008 (UTC)

Yes, well said. Except on BLP rather than getting anyone to back down, his suggestion to make it policy was useful in preventing the creation of an opposition movement in the first place. There never existed significant opposition to BLP which is as David likes to say "a hard assed implementation of the other policies" and as he likes to leave out an insistence on treating living people like living people and not like a building or some other subject of an article (which was Daniel's original complaint - people told him they could edit his article anyway they liked so long as it met wikipedia policies and any harm it caused him was none of their concern). WAS 4.250 (talk) 13:53, 25 April 2008 (UTC)

It is actually quite possible to resolve such disputes rather more rapidly, by defining particular forms of blocking discussion as disruptive and thus blocking those editors (see Wikipedia_talk:Disruptive_editing#Blocking_consensus , User:Dmcdevit/On_edit_warring#Usefulness_of_repeat_protections). This is not the the only way to resolve such situations, just the most expeditious.

Actually, commonly there is merely a dearth of good faith, and resolving that will tend to solve the problem to quite a degree. (The list of documented procedures and manned systems to apply to that problem is rather long, so please forgive me for not linking here. Suffice to say that a little skill at mediation can go a long way. :-)

This way of approaching the problem can be and is already being used, and is also happens to be rather less invasive than turning the entire wiki-model on its head.

Which policy pages do you currently have the problem on?

--Kim Bruning (talk) 13:16, 25 April 2008 (UTC) I'm not saying it's the case here, but I have often noticed that the people complaining loudest about consensus being a waiting game, are actually involved in causing the problem in the first place. (This is statistically intuitive: the person who is {unintentionally} responsible for certain kinds of situation will tend to encounter it rather more often than others. ;-P )

What looks like 'blocking consensus' to one person may look like a good-faith objection to another. I don't see it as feasible to start blocking large swaths of editors in order to get some change I prefer moved into policy, especially sice I can be wrong sometimes.

One example at the moment, on which I am not involved, is the possible change to permit BLP articles to be deleted if there is no consensus to keep at an AFD. There are strong arguments on both sides there, so neither side is acting in bad faith. What we want to happen there is for a group of editors to carefully weigh our encyclopedic mission against the BLP issues and come to a difficult decision. I don't know which way that decision would go, it's genuinely hard. But the system we have doesn't encourage that sort of deliberative, responsible process for policy making. — Carl (CBM · talk) 13:24, 25 April 2008 (UTC)

Good faith objectors shouldn't be reverting every day, for instance, I would wager. (not even to The Consensus Version, not even to The Wrong Version). Nor should they be requesting page protection, I should think. And they shouldn't be wikilawyering about which policy xyz applies, or at least not longer than one or two posts, after which it should become clear to them that mere robotic behavior won't work (we have pywikipedia bot for that), and that they should start displaying some strong ai now.

Good faith people should be trying to work towards Consensus (which none of the above things are).

I do realize that some people might find that a novel concept.

--Kim Bruning (talk) 13:48, 25 April 2008 (UTC) I'm sure that xkcd would argue that import strong_ai should do the trick... but then why do we have humans here? Also, I'm skeptical about xkcd. I got an error when I tried to import antigravity . Very marginally on-topic, I had no luck with import soul either.

Ok, and the more direct problem at hand is BLP? I see. I'm starting to get the idea that BLP might be better off if we spun that off to wikinews entirely. But that's just my impression right now.

I'm willing to look into those discussions as well, if you like. Have you asked editor assistance or the mediation cabal for help already? (so that I don't accidentally cross someone else's work?) --Kim Bruning (talk) 13:56, 25 April 2008 (UTC)

Short brroad-brush history of policy making at the English language Wikipedia

First we had no rules. Then some basic rules defining the project are make (the owner of the hardware, Jimbo, declares what that hardware will be allowed to be used for). Then rules get added based on existing best practice (we know the rules have consensus because these new rules are descriptive, not prescriptive). Then we get lots of new editors who take the rules to be rigid laws and apply them rigidly. Now changes to old rules can no longer occur though people gradually changing best practice because of rigid enforcement of an all encompassing set of rules. So now changes must be established as consensus by voting. But voting does not work when there is no way to establish who is an eligible voter. So we are increasingly stuck without a workable mechanism for changing existing rules. But that's not a crisis because the rules are quite good as they are. But for anyone who thinks they have a great new rule, they are frustrated by a lack of an easy was to get the proposed rule made into policy. Maybe that's for the best. Maybe it should be that hard to change our policies. WAS 4.250 (talk) 13:51, 25 April 2008 (UTC)

Actually you can still document existing best practice using just normal wiki-rules for editing. Most of our current policy is still written that way, the process has never actually changed. At last count over 90% of policy and guidelines had been made and maintained using the normal wiki method.

Some people get a different impression, either because they've been pushing prescriptive practices, or because they have seen others do so (where those others have obviously been causing more heat than light)

--Kim Bruning (talk) 14:00, 25 April 2008 (UTC)

Yes, and that's where the true problem lies. Perhaps my Welcome Template is a start towards fixing the problem where it began. Kurt Weber (Go Colts!) 14:04, 25 April 2008 (UTC)

Kim, I agree. But I'm also raising the issue that a major change from existing policy can not be done though a gradual change in behavior anymore. WAS 4.250 (talk) 14:09, 25 April 2008 (UTC)

Jimbo by panel

The Delegates as proposed are basically the new Jimmy. As he's basically off doing WMF, Wikia, and scuba diving with Tony Blair on Richard Branson's island, the proposal as formulated basically looks like a realottment of the authority he once enjoyed to the community in an official capacity. The community elects trusted delegates to x length terms; the delegates basically handle the dirty work of policy clarification, change, and implementation, and if it's a major change can draft a WP:3RR style vote for the community to decide. Behavioral stuff stays with the Arbcom. The Policy Delegates will not answer to Arbcom or to Jimmy, but if there's a legal question (GFDL, etc.) they can ping Mike Godwin. Grab your 15 to 50 most trusted, chuck 'em on 1-year terms, and let no one serve more than 2 consecutive terms to avoid stagnation, and I think you've got a winner. This is a great proposal. It's the epitome of community since it empowers the community to take supreme control of about everything in a fashion that will scale over a time frame of years. Lawrence Cohen § t/e 14:11, 25 April 2008 (UTC)

That sounds like what we wanted to happen in the first place. Someone should write a page for the proposal on exactly what would happen. STORMTRACKER 94 ^{Go Irish!} 14:49, 25 April 2008 (UTC)

<Eyes the salesman on the doorstep> If I actually wanted a new Jimmy, that'd be totally true. But like, I don't want any Jimmys anywhere near policy anymore, that was totally a stopgap. Why would I want a new one? <looks rather suspicious> --Kim Bruning (talk) 15:05, 25 April 2008 (UTC) (I don't think JWales minds one bit that the community can take care of itself now :-) )

Haven't your heard those Hendrix lyrics, "Move over, Rover, and let Jimmy take over"? Oh wait, maybe that was "Jimi." Chin Chill-A Eat Mor Rodents (talk) 17:58, 27 April 2008 (UTC)

Calling it "Jimmy" was just a descriptive, really. Lawrence Cohen § t/e 15:35, 25 April 2008 (UTC)

Why can't I just edit policy myself? Strange things, these wikis are. They work just fine for such mere trifles as multi-million page encyclopedias, but fail utterly for such important and complex things as describing how a single website works. --Kim Bruning (talk) 15:49, 25 April 2008 (UTC) Sarcasm? What sarcasm? O:-)

Because CCC in the classic sense and the old ways will be dead when we have 10 times as many people as a viable solution. Jimmy's hippy ways aren't going to scale up unfortunately. I think the point of this as I understand is to make the site viable in this regards no matter how big it gets. The bigger it gets, the more structured some things would I think need to be. Lawrence Cohen § t/e 16:18, 25 April 2008 (UTC)

I completely agree with Lawrence Cohen above. Walton^One 21:48, 25 April 2008 (UTC)

Bzzt

Lawrence cohen said:

Because CCC in the classic sense and the old ways will be dead when we have 10 times as many people as a viable solution.

Which possibly isn't going to happen for at least another 10 years or so, looking at statistics over time. Wikis tend to scale very well. (I'm not kidding. 10 years is a conservative estimate, in case we happen to be at the edge of some curve we can't quite see yet.)

Jimmy's hippy ways aren't going to scale up unfortunately.

Jimmy's hippy ways? Don't you mean Ward Cunningham et-al's rather well thought out patterns? wiki:

I think the point of this as I understand is to make the site viable in this regards no matter how big it gets. The bigger it gets, the more structured some things would I think need to be.

So why do you want to stomp all over the existing structure like an elephant, without even looking at it? :-)

--Kim Bruning (talk) 17:46, 25 April 2008 (UTC)

Oh, I don't. Just weighing in, is all. But I fully support anything that would take policy-power out of the hands of the few and legitimately push it back into the hands of the many. The problems are that 1) a small group of wonks can stop policy talk; 2) the more people that join a discussion the more likely, as Kirill observed, that it will be an automatic no-consensus. Lawrence Cohen § t/e 17:56, 25 April 2008 (UTC)

Oh, I thought you were taking it out of the hands of the many, and putting it in the hands of the few. I'm totally opposed to that, of course. --Kim Bruning (talk) 18:50, 25 April 2008 (UTC)

My understanding of this proposal as Matthew drafted it is the exact opposite. This would be the exact opposite of disenfranchising the few; it would disenfranchise policy trolls, policy wonks, and people that loiter in Wikipedia-space rather than building an encyclopedia to give the power to the masses to make policy decisions. Lawrence Cohen § t/e 18:52, 25 April 2008 (UTC)

But the "masses" (geesh) ... the wikipedia editors already have that power, they may collectively form consensus on any matter and carry it out. This proposal pulls that power out of their hands, gives it to some committee, and then (at the end of the day) it's thecommittee tells the editors what they're supposed to do, basically. Oh and in between that there's lots of voting yelling, and basic way-past-dunbars-number-human-condition going on too... erm... I'll shut up now. It's just I need this broswer window to test... and the temptation... <reaches for logout button> --Kim Bruning (talk) 20:02, 25 April 2008 (UTC)

Graphical workflow of what the ideas "look like"

This is based on what all the ideas here kind of mesh together as:

How close is this to what everyone is envisioning? Lawrence Cohen § t/e 16:12, 25 April 2008 (UTC)

An observation on this: no one will be disenfranchised by such a system in any bad-faith sense. Since "one" voice or a small minority of voices shouldn't have more weigh or authority than any other comparable group of the same size or large on the wiki, the only people who would be disenfranchised by such a system is the group of policy-wonks that tend to currently dominate policy changes and discussions--the policy "regulars" would be made partially redundant and kicked down to the level of everyone else. This is not a bad thing, in the grand scheme. Lawrence Cohen § t/e 16:27, 25 April 2008 (UTC)

I'm sorry, but I would strongly oppose the creation of a power structure such as that.(Though I have to say that without a touch of "this or that", I might support it.) My main concern is its role in the creation of policy. I still would prefer that this group just be a "review board", and possibly also as a braintrust that other groups (such as arbcomm) might solicit opinion from. - jc37 16:39, 25 April 2008 (UTC)

I oppose that and any attempt to create a Parliament or any other sort of decision making body along these lines. I'd like to see something in which we could get the whole community to engage with policy creation, but not a parliament. The wiki process, and it pains me to admit this, is the best we've got at the moment. Although we need a better method for tracking all the changes. Wikipedia has certainly outgrown the watchlist model. Hiding T 16:48, 25 April 2008 (UTC)

Well, this map is sort of exactly what we have now. Policy proposal > Discussion > Determine consensus > policy change goes live. This is sort of just a stewarding of it, but on a scale that may theoretically scale up to many x times the number of users we have, and seems (based on what everyone has said here) to be an attempt to strip away some of the anarchy in the process, to prevent any small group or lone users from stopping policy change. Stopping lone wolves or small interest groups on-wiki from blockading policy change alone would make something like this worthwhile--we don't need policy wonks. What don't you like about this? I'm torn on the overall idea myself, even though I'm more in favor in general of structure to things rather that chaos. Lawrence Cohen § t/e 16:49, 25 April 2008 (UTC)

If step 3 were also sent down to a vote/poll to determine acceptance, so that this group were more a bunch of Policy Stewards, or simply trackers/gatekeepers to major policy change, would that be less offensive? A Policy Steward body would be more appealing to me personally. Lawrence Cohen § t/e 16:50, 25 April 2008 (UTC)

A modified workflow based on this feedback:

Lawrence Cohen § t/e 17:05, 25 April 2008 (UTC)

(multiple ec's later)I share some of the concerns of JC, Hiding, and maybe others above, but probably not to the degree that the editors themselves necessarily do. In terms of creation of new policy, I think that, if the model were enacted, the initial "drafts" of new policies would almost certainly be in most cases a simple statement of a behavior or other standard, and possible exceptions to that standard which would emiliorate the situataion. But having a body which could draft policies for discussion, and at the same time hopefully ensure that the drafts don't have huge defects in them which have to be altered later, would probably be a good thing.

I do think that if there were a specifically designated "review period" for editors in general, most of the concerns they express would almost certainly be considered, up to and including rejection of the current draft. I think that it should be made a bit clearer that the community input section would be ongoing throughout the period of a policy being considered, up to and including the !voting itself. But I agree that this would open up the discussions more. There are other ways to do that, including a new WikiProject noticeboard, which can and should be added to any number of other pages, and the like. But that noticeboard, when placed everywhere it should be, and the Signpost, which could probably run a regular bit on the policy proposals, developments, etc., should both increase the level of involvement and participation of editors, while at the same time ensuring that the current stagnation ceases. I have some concerns regarding some of the ancillary details, like size, whether ArbCom members and bcrats would get in automatically, possible development of other governmental entities, and the like, but most of those can be dealt with later. The flow chart looks like a good starting point, even if some of the details clearly need to be worked on, like maybe the additional step Lawrence proposed above, which I don't have any objections to adding, with one proviso. In the event the process were started by a request/demand for a policy from Jimbo, ArbCom, the legal office, or some similar entity, then the Stewards might be allowed to vote for the proposal at the end, if the statement from Jimbo or ArbCom were to somehow indicate that such a policy, even a very basic, undeveloped one, were somehow necessary.

And, to specificy what I meant by ancillaries above, as I expect there will be questions, I'm thinking of group makeup, numbers, removal of members of the body for misconduct, and the possibility of members recusing themselves from certain discussions. Also, it might help if we had a group of specifically designated Arbitration and policy enforcers, and standards for enforcement, like ensuring parties with apparent, if not necessarily real, possible conflicts don't try to to enforce arbitration decisions regarding parties or issues with which they have conflicts. Doing so would help ensure that the appearance of COI doesn't happen very often, if at all. The latter development would probably be "liberating" in a sense as well, as someone being considered for AE would be able to be a bit surer that there would be no "action by enemies" involved. I think that any real proposals which really do limit freedoms in a unnecessary, potentially damaging way, which I don't think these do, would probably be dismissed unless there were obvious need for it for whatever reason. But, in terms of the negative effects of creating new policy/guidelines, having a clearer idea of what they are in advance, and what the consequences of acting contrary to them are, is generally much more liberating than finding out after the fact that you've done something wrong which you can be penalized for. John Carter (talk) 17:10, 25 April 2008 (UTC)

(Looking over the new chart) - For one thing, I would split the election of the committee members and the tasks of the committee members to two separate charts (they're separate concepts to discuss)

So for the latter topic:

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

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

Review = reviewing a page to see whether it meets/falls within a.) Foundation principles b.) current general usage c.) the current policy framework d.) current community consensus, etc.

If I'm missing something, please ask for clarification : ) - jc37 17:59, 25 April 2008 (UTC)

FA doesn't scale at all, it's a one man show (two man now). Arbcom does not scale and has needed replacing for quite some time. (We've shored it up by reducing the workload with layers of mediation, some of which does scale, but that's really not the way to go in the long run). If you are hell-bent on making a new system, please at least base it on something that isn't about to fall apart. :-P --Kim Bruning (talk) 18:12, 25 April 2008 (UTC)

What happened to the Kim Bruning who was suggesting that we look to the policy and process already in place? Or are we cherry-picking what "we" prefer? : P - jc37 18:29, 25 April 2008 (UTC)

LOL no! You totally have the right idea going there. And I guess we totally should discuss arbcom, FA, and also *FD... all of those have (had) scaling issues. And that's a totally interesting fact, I suppose. --Kim Bruning (talk) 18:44, 25 April 2008 (UTC)

We can, and I'd probably enjoy the discussion : )

But talking about "scale", consider what we're discussing: Policy pages.

Now compare the number of policy pages to the number of editors who get into disputes, or the number of pages up for deletion, or even the number of FC up for review.

Thousands or millions compared to what, several dozen?

Honestly, my main concern is that the members (if they only focused on their committee work) would be twiddling their collective thumbs. But with the addition of being a support body to arbcomm (and others), I think they'd be kept fairly busy : ) - jc37 18:55, 25 April 2008 (UTC)

The one advantage, of course, to not keeping them busy on committee work is that they would be less likely to suffer burnout from doing too much committee work. I personally think that the proposal would be best if it resembled what the US Congress started as, a group of "part-timers" who give up some of their time to write the bloody policies and guidelines, but get to spend much/most of their time doing something else, which they're probably more interested in anyway. John Carter (talk) 18:59, 25 April 2008 (UTC)

current method

The current method to maintain policy pages is documented at Wikipedia:Consensus, and that page (and many pages like it) are being maintained by that very process right now. :-) (It'd better be, if we don't even trust our own dogfood, we're lost.)

The advantage of that method is that it works, and that it works now.

This method is characterized by quick turnaround, the ability for anyone to contribute at any time, the ability to teach people about the policy by allowing them to directly modify it, and many more besides. This method has been used to create 90% of our policy pages or more, as well as over two million encyclopedia pages. It is proven, fast, well tested, and resilient even under continuous pressure by vandals. It also scales extemely well, because it lacks central choke points or particular time requirements.

There are also several extensions and improvements that have been devised over a period of 8 years that are built upon this system. Several processes and systems have been designed around it or built upon it.

Before we abandon current best practice, perhaps it should be studied more closely. Just because you are trying things at random and failing doesn't mean that an existing process isn't already documented. :-)

--Kim Bruning (talk) 18:08, 25 April 2008 (UTC)

You know I love you, Kim, but this is a classic example of what I meant by one person burying a proposal or change with "machine gunned criticism". :) No one is as heavily commenting on this as you, and it's beginning to drown out the page. ;) The tenacious should not win simply for being tenacious. :) Lawrence Cohen § t/e 18:35, 25 April 2008 (UTC)

Oh my. I was actually considering scaling back on my replies, I guess you noticed too. I have some coding to do too, so possibly I might go off and do some of that. There's just so many issues with this particular proposal, I don't really know where to start. :-/ Anyway, this particular section wasn't too critical, was it? --Kim Bruning (talk) 18:46, 25 April 2008 (UTC)

No, not too critical. I'm more in favor of decentralized but structured control, you're more in favor of just decentralized control. Apples and oranges, both good for you. Just a question of what vitamins are best for our job. :) Lawrence Cohen § t/e 18:54, 25 April 2008 (UTC)

Majority = consensus

I think that's a big part of the disconnect. Which is more important to us on discussions, and policy changes? Getting your opponents on board, or going by what the majority wants? Tricky. If the majority want a given change, but the minority don't, what should matter more? What the majority want, or the concerns of the minority? Should a minority be able to quash the majority, as commonly happens now? This is one of the reasons I tend to not weigh in much on policy stuff after some aborted attempts. I've seen too many discussions where 15, 20 people endorse something, and some loudmouth 1-5 people come in with a machine gun of complaints and bury the changes. It ends up being a situation where the most dedicated, overwrought (I'm no less guilty on that mark, unfortunately, cf my RFA), or persistent people win, even if their wishes don't align with the majority.

It has the net effect of the users who spend the vast bulk of their time on policy pages running the show. This is not a good thing. Everyone knows the folks I'm talking about, as well: look at their contributions, and of their last 1000, 90%+ will be to Wikipedia/Wikipedia talk pages. Our current model rewards the tenacity of the vocal minority rather than the will of the majority. Lawrence Cohen § t/e 18:02, 25 April 2008 (UTC)

Example. If I edit a policy page so that it now reads, in full: "poop". Does the policy change? --Kim Bruning (talk) 18:19, 25 April 2008 (UTC)

With the current system, no, certainly not - it would be instantly reverted, and the page has to stabilise before the changes are meaningful, whatever meaning you see them as having. With the proposed system (at least, by my understanding, which differs from some), if the assembly voted to change the policy page to "poop", then that would be the new policy (for as long as it took for people to IAR and get the hell rid of the insane assembly). --Tango (talk) 18:27, 25 April 2008 (UTC)

Actually the policy just stays the same old, no matter what change you make to the page. People will do whatever they like, and what they do is what's called policy. The page just documents what they do. Like any other wiki page, it can fall behind, be vandalized, etc etc... --Kim Bruning (talk) 18:34, 25 April 2008 (UTC)

Yes, but: I think lately from things I've seen is that there is more of a shift that policy as written, in an accepted form, to be enforceable. This sort of proposal would be that sort of thing. It's really more of a cultural proposal change than a silly new process. "If the majority says policy is this, you need to do this, and thats that." Lawrence Cohen § t/e 18:37, 25 April 2008 (UTC)

Right. But that's not really a good idea, is it? That basically disenfranchises up to 49% of all Wikipedians. (And that's the optimistic version :-P ) --Kim Bruning (talk) 18:41, 25 April 2008 (UTC)

Well, 51% isn't a clear majority, though. If you applied RFA-type pass/fail standards, it would be a whole other game. If you had 100 people all vote on a change to a given policy that is not a WMF/legal matter, and 80 to 90 of them endorse the change, that's it, game over: change should be done. Lawrence Cohen § t/e 18:46, 25 April 2008 (UTC)

Sure, people can do whatever they like, but they'll end up getting blocked for it. --Tango (talk) 18:46, 25 April 2008 (UTC)

Well, majority certainly doesn't equal consensus - that much is easy. If there is a consensus on a matter, then the way forward is clear. If there isn't, then the first step is to try and form one (ie. discuss, convince and compromise until everyone is a least willing to go along with it). The problems arise when we are unable to do that and it's then that we have to decide between doing nothing or using something other than consensus for decision making. At the moment, we generally fall back on polls, but they rarely go well. This proposal is an alternative method for dealing with situations without a consensus. --Tango (talk) 18:27, 25 April 2008 (UTC)

If there's no consensus, we can wait. m:Eventualism :-) --Kim Bruning (talk) 18:35, 25 April 2008 (UTC)

In some cases, sure, but not in all. Sometimes the status quo is the worst of all the options. --Tango (talk) 18:46, 25 April 2008 (UTC)

Also, Kim is taking it for granted, possibly fallaciously, that, while waiting for "eventualism" to triumph, we don't have the ship sinking as a result of a problem which has arisen while we're waiting for eventualism to kick in. Realistically, I think we have to admit that just putting everything off till later will almost certainly have the consequence of creating additional problems we don't need, and could have resolved earlier, had we not been almost religiously relying on eventualism to ultimately save everything. This is particularly important if what we're relying on eventualism to solve is a legal matter which could potentially destroy the project immediately. John Carter (talk) 18:49, 25 April 2008 (UTC)

Curiously, that ATT poll had what, 67% support? What if that had been 90%? If Jimbo had come along to quash the ATT move then, what would have happened? He would have been out of bounds and the community would have been within their power to quash him in return. Majority certainly can equal consensus, and should, but our systems and discussion methods are setup in a way that encourages watering down of strong ideas by discussion for (often) pointless discussions' sake, and with our harebrained "voting is evil" methodology, it often discourages people from expressing if they simply do support something. Not all decisions require 100kb of discussion per person; not all changes require 10 pages of archives. That discourages many people from joining in. If someone does 800 edits a month, 600 to articles, 100 to various user/article talks, and at best 100 to Wikipedia spaces, why shouldn't his opinions on a policy matter carry the same weight as say Kim, or Until 1=2, or Ned Scott, who tend to spend as much time on Wikipedia internal matters as our example editor spends on articles? We are all supposed to be exactly equal here. Lawrence Cohen § t/e 18:42, 25 April 2008 (UTC)

(In reply to Lawrence) This somewhat simplifies the issue with changing policy. It is rarely a case of X% support and Y% oppose, it is usually in multiple shades of gray. A% support the whole thing, B% oppose the whole thing, C% support parts 1 and 5, D% supports parts 2 and 3, E% supports parts 2 and is strongly opposed to part 1, etc. You could rewrite the proposal to exclude the parts that got the least support, but then the people who supported the parts that were removed may no longer support it or a completely different group of people might give their opinions. There's also the issue of "what is consensus?" This was one of the major problems with the rollback discussion. Unless you do it as a straight Yes or No vote like ArbCom elections, the more people that get involved, the bigger variety of opinions you get and the less likely you are to get consensus for any one idea. This is also why WP:RFC/U is the worst discussion format we have, it basically encourages this. Mr.Z-man 18:30, 25 April 2008 (UTC)

Was the 3RR vote a good thing or a bad thing? Lawrence Cohen § t/e 18:37, 25 April 2008 (UTC)

That's a meaningless question - there is no such thing as "good" and "bad", just "better" and "worse". You need something to compare it with. --Tango (talk) 18:46, 25 April 2008 (UTC)

Fine, compare the 3RR vote to the ATT vote. The 3RR vote appears successful, passed, and from poking around old stuff, didn't appear that controversial. See also the vote for the Main Page redesign. Compare to ATT, where everyone fought like mad, mainly over not giving up community control to a vote: reading that was painful. It seemed like the anti-vote was as much "voting is evil" dogma as it was some people refusing to yield policy control to the unwashed masses. Lawrence Cohen § t/e 18:50, 25 April 2008 (UTC)

(ec) If your change has 100% support, it might still not be policy. (a textbook example happened recently with the recent NFCC policy page, where 100% of those present supported an alteration, but the alteration got reverted anyway. ) --Kim Bruning (talk) 18:39, 25 April 2008 (UTC)

Diff? NFCC is a legal/WMF matter, however. Those are immune to consensus. If you had 40 people all endorse a change to WP:NPA with 100% support, and one person reverted it out to stonewall, that one person is as they say here in the States, "Shit out of luck." Lawrence Cohen § t/e 18:44, 25 April 2008 (UTC)

What? No fillibuster? (sorry, had to : ) - jc37 18:47, 25 April 2008 (UTC)

Yeah, yeah. Filibustering is not a productive thing on here. :) Lawrence Cohen § t/e 18:48, 25 April 2008 (UTC)

I wasn't involved in that particular case, but surely the person that reverted didn't support it, so there wasn't 100% support... --Tango (talk) 18:46, 25 April 2008 (UTC)

Well, lets say 25 people endorse a change to BLP and I say, "Hell no." and try to stonewall. How far should I be able to take that? Lawrence Cohen § t/e 18:48, 25 April 2008 (UTC)

Potentially? all the way through the DR process, to arbcomm, if necessary. Here's the thing, though. Arbcomm will tell you that they don't decide which version is "right". So as you said above, eventually, you're SOL. Better go make some friends (grin). - jc37 18:58, 25 April 2008 (UTC)

Exactly, bingo. And this sort of process would make it so that no one is ever SOL, in that way, since any good-faith policy change could get a shot at being heard before the whole community, as I read it. It'd make policy edit warring pointless; it'd make sure that crappy changes or changes that some people push through to benefit their own Wiki-interests get quashed; and it would make sure that only broadly-supported changes go through, and none by anyone's fiat. Lawrence Cohen § t/e 19:00, 25 April 2008 (UTC)

Exactly! FloNight♥♥♥ 19:02, 25 April 2008 (UTC)

Then that's "policy review", not "policy creation". Since that'd be reviewing the edits made. Or am I missing something? - jc37 19:15, 25 April 2008 (UTC)

Policy is created either by someone posting a proposal, and the majority of users endorsing it; or someone drafting a policy that reflects practice, and the majority of users endorsing it. It doesn't really matter if one person drafts it, or two, ten working in a team, or 50 working in a panel, though, does it? If the policy itself or a change to policy has legs, regardless of how it originated, it will either fly or it won't. Let's say some guy that has a Great New Idea (to him) for policy, but not a lot of great knowledge to socialize and implement this under the current, present day model. He can come to these Stewards, or Gatekeepers, or Delegates, or whatever we call them, and put it up for review: "This is my great idea!" If it's sound, the clueful people we elect after consideration can put it forth in a well-worded manner to the masses: "Do you guys want this? Does this reflect the Wikipedia you live in, or want to live in?" Then we all, the body members included, decide.

Another thing that this reform proposal strikes me as: rather than small policy changes languishing forever in back-corners of Wikipedia, this will throw them out to the public broadly. If there are five policy changes in a month on a "docket", for lack of a better term, this will let the interested public possibly look them over in one centralized place.

At least that's how I read it. Taking away the backroom dealing, taking policy control out of the hands of the backroom, out of the very few and giving it up to all. Lawrence Cohen § t/e 19:55, 25 April 2008 (UTC)

But anything on a policy page isn't binding because it's on a policy page, it's "binding" because that's what everyone does. So no matter how much backrooming goes on (which I assure you isn't as much as you'd think :-P ) , it wouldn't make a whit of a difference. That's a misunderstanding a couple of people have. Oh... sorry.. I was supposed to be coding... getting back to that. :-) --Kim Bruning (talk) 19:58, 25 April 2008 (UTC)

You know, I've heard a lot of people say that there isn't that much being done in secretive, furtive, concealed locations as a lot of people think. And they're probably right. And if all those freaking idiots out there who seek this silly thing called "transparency" could just accept that they don't need to know anything that goes on in the backrooms, we'd all be happy. Why, oh, why, can't they realize that they're better off being kept in the dark about these matters? ;) John Carter (talk) 20:02, 25 April 2008 (UTC)

The secret is that there are no backrooms for this kind of thing. By assuming that there are, and acting on it, you're actually shooting yourself in the foot. (oops) --Kim Bruning (talk) 20:04, 25 April 2008 (UTC)

The key problem with the proposed model

Draft mockup interpretation

An elected body of representatives? Seems to blankly contradict the fact that Wikipedia is not a democracy. Anyone who wants can participate in policy reform and creation, and that's way it should be. Policy comes organically from within the community, and to do otherwise violates the basic principle of operation that has made this project successful. Laying the power to alter/create policy in the hands of delegates alone is not something I would ever accept, even if that process includes a ratifying vote. VanTucky 20:20, 25 April 2008 (UTC)

Every one can add content, yes. That everyone can be directly involved in all the other process has not scaled as the Project grew. FloNight♥♥♥ 20:48, 25 April 2008 (UTC)

But handing over the responsibilities that ArbCom and the sysops each respectively have has disenfranchised no one's ability to have their direct say in the formation of the policy that governs Wikipedia, it simply handed additional powers to a few to enforce that policy. ArbCom, admins and others only act on the basis of policy that is open to reform by everyone. Taking away the rights of everyone to reform and create policy is cutting the legs off of what makes those systems equitable. VanTucky 20:57, 25 April 2008 (UTC)

Where in Kiril's proposal (as mocked up in my silly little graphic as close I could) does it indicate anyone loses control? If anything, this gives MORE control to MORE people over policy content. The only people who will lose theoretically are the policy wonks that sit on the policy pages all day, as they'll no longer be primarily in control of and stewarding the pages. Lawrence Cohen § t/e 21:00, 25 April 2008 (UTC)

I read the draft graphic before. Yes, I lose the ability to just hit the edit button and be bold in updating this content. A proposal which takes the ability of all to edit policy, and adds a bureaucracy that will stop users and say, "no, you have to jump through these legislative hoops to get policy changed" is something I oppose 100%. Policy on Wikipedia is not legislation, and I don't want it turned to that either. I don't care how you frame it, but any forced process whereby the power over policy is in the hands of delegates is bad. The point is: the privilege to edit policy directly is power over that policy. I'm not handing that power directly in to the hands of a select few, be they elected or no, without a fight. There's simply no pressing reason why I should. VanTucky 21:20, 25 April 2008 (UTC)

How do we deal then where you have one person or some ultra-minority that doggedly wear out and nuke any proposed policy change until everyone gives up on it, or the ones where some out of the way group works to ram through their own policy shifts and then goes to war until their other opponents give up? Should victory go to the policy wonks, policy trolls, and the people who are willing to sit and fight for months on end? Or should ultimate authority over policy reside with the collective masses, and to have shifts in policy presented in a centralized manner for them to decide? Lawrence Cohen § t/e 21:23, 25 April 2008 (UTC)

It's not having a centralized location to discuss policy changes that I oppose. It's creating a specified step-by-step bureaucratic process (which will inevitably require much instruction creep) to do so, and more importantly, having to protect that bureaucracy by handing over the power to edit policy to a select group of delegates. I would much rather see the way policy is handled stay the way it is (by those interested enough to devote time and energy) than see another bureaucratic process and a hierarchical position of power created. Again, the way policy is handled on Wikipedia is just another thing that shouldn't work (at least when you describe in the terms you just did), but in my opinion, it does. VanTucky 21:28, 25 April 2008 (UTC)

You may feel that policy governed by the dedicated people that loiter on policy pages and are willing to get into months-long battles over the odd sentences works, but consensus apparently (so far) on this page seems to be counter to that. Why should policy be governed by people who are willing to babysit policy pages to deflect any and all comers? It engenders policy ownership, and falsely, inappropriately, and unethically empowers long-time users or the simply persistent who fight on pages until everyone gives up. That's the very definition of a problem, because it de-powers anyone from affecting policy that isn't willing to fight through weeks or months of trolling and conflict by vocal minorities. Look at the massive outcry from a very small number of users that comes up whenever people try to strengthen WP:BLP as a good example. You may not think BLP needs tightening, and they may not think it, but any system that allows a handful of people to derail a change from a much, much larger group is flawed. We need a way to smoke out the policy rats that will cause this ship to sink when we add more and more people; some sort of system that makes it so established every user has an exactly equal voice in the policy making process--without them having to spend 12 weeks doing battle with a couple of users. Lawrence Cohen § t/e 22:19, 25 April 2008 (UTC)

Again, I'm of the opinion that the proposed alternative is much, much worse than the present way of doing things. I don't want an elected bureaucrat holding court over the policy that I am obliged to follow and enforce. I want to be able to participate directly in shaping it, and I don't mind that occasionally letting all have access to it creates a problem. It's still a net positive. Again, the sausage making metaphor is applicable here. If the way policy is edited, created, and approved is so awful, then why is our policy so good? And yes, I do think our policies are, as a whole, very very good. That's part of why I wanted to become an admin. VanTucky 23:04, 25 April 2008 (UTC)

I hate adding extra bureaucracy. I'll oppose it almost every chance I get. However, from my recent experiences, getting policy changed in any significant way is virtually impossible. Right now the people who are willing to fight the longest win and in the majority of cases those are the people fighting to maintain the status quo. We are a dynamic content website with fairly static policies. That isn't good IMO. Mr.Z-man 23:34, 25 April 2008 (UTC)

If it's not a democracy, what is it then? You think people come here and invest their time so that some other people would command them? If so, it is at least immoral, and I find it offensive that you would agree with such practice. Also, you're misquoting - the original statement was "Wikipedia is not an experiment in democracy", which has completely different meaning. Samohyl Jan (talk) 09:09, 26 April 2008 (UTC)

Democracy is not the only system of community participation, though it is one of the oldest. On wikipedia, we make decisions based on consensus of editors, rather than by democratic majority. We already knew that consensus was the best way to determine facts and NPOV, and we started to use it elsewhere mainly to just to remain consistent. But over time this has turned out to be a blessing in disguise, as consensus has turned out to actually be quite useful for purposes of governance.

One advantage of consensus over democracy is that in a democracy, the minority is just simply out of luck. In a consensus system, a minority still might not always get everything they want, but they certainly still get a say in what happens, because everyone gets a say.

As you might surmise, I've become rather enamored with it.:-) --Kim Bruning (talk) 20:08, 27 April 2008 (UTC) So be careful before you clamor for a downgrade to democracy, because you will lose rights and become partially disenfranchised.

Power

Just curious. Is this proposal an attempt to expand consensus, and fill a current possible vaccuum in the process, or is it entirely about who should have the "power" over policy?

The more I read the discussions, it's starting to appear to be a case of "I want somebody to be in charge, so that (the enemy) can't stop me from doing what I wanna do."

Is this really what's being proposed?

A caveat: I've faced similar situations, and have been rather disgusted that a single POV-pushing editor can stop clarity or growth.

But to deal with this concern we're going to set up a student council, or possibly an elective dictatorship?

Would one of the original proposers please clarify this for me? - jc37 21:14, 25 April 2008 (UTC)

I think you've slightly misinterpreted it. We all agree that the wiki-consensus model works fine for the vast majority of decisions taken on the site - content decisions, etc. But for major policy changes, or the adoption of new policies, it is incredibly difficult to build consensus when you have hundreds or even thousands of participants in a discussion. There's no formal process for adopting policy; strawpolls are almost always inconclusive, and a vocal minority can block changes desired by the majority. In other words, what Wikipedia lacks is an effective legislative process. We're simply proposing an elected council to fill the legislative gap in Wikipedia. Obviously, everyone would have input into policy-building, just as they do now; but it would be the council who would vote on it and make the final decision. Walton^One 21:54, 25 April 2008 (UTC)

Misinterpreting? I look at my post above, and my comment about a "student council", and I look below and see what? A student Wikipedia council. I know you're a strong proponent for Democracy on Wikipedia, you've made no bones about it. But that's not the "wiki-way", which is, by the way, a foundation issue. I strongly doubt that any sort of legislative body is going to fly. - jc37 10:36, 26 April 2008 (UTC)

When the heck have we ever had hundreds or thousands of participants descend on a page without first having to be prompted by drastic measures (sitenotice for instance) ?

If inviting hundreds or thousands of people all at once b0rks your precious proposal... then... Don't Do That Then! Invite people a couple at a time and discuss properly with all of them. :-)

You're basically first causing your own problem, and then proposing to turn wikipedia upside down to solve your own problem that you caused for yourself in the first place ;-)

--Kim Bruning (talk) 22:06, 25 April 2008 (UTC)

Discuss and get buy-in from all or most participants? Never. The problem is the community should never on any discussion or process shift be held hostage by a vocal minority. If a policy, decision, or action is sound for the whole, showing it in plain language to the whole is perfectly fine. If the majority/community wants it, they'll take it. But we certainly will never need or desire detailed discussion-based buy-in from hundreds of users, the idea is absurd, Kim. We're not here to be social, and sway each others' views to our own, we're here to make an encyclopedia. If 90 people buy into something, 5 kind of do, 2 sort of dislike, and 3 vehemently are against it, too bad for the 3. Lawrence Cohen § t/e 22:12, 25 April 2008 (UTC)

Doesn't that contradict your "SOL" comments? - jc37 10:36, 26 April 2008 (UTC)

Beware of the man who would disenfranchise you. :-) It totally depends on the arguments put forward. Ever watch 12 angry men? It's a good thing the minority wasn't overruled by the majority, or an innocent man would have gotten the chair. On the other hand, if a minority is being totally unreasonable, we already use rough consensus, and unreasonable people get overruled and even sanctioned.--Kim Bruning (talk) 20:14, 27 April 2008 (UTC)

Vocal minorities have been very effective in preventing the implementation of the three revert rule, rollback, biographies of living persons policy, and even the arbitration committee. Oh, wait, all of those things exist. Are we just not comfortable with the idea that new proposals need community support? – Luna Santin (talk) 22:59, 26 April 2008 (UTC)

A detailed proposal

Feel free to pull this apart/debate the details, but here is a general proposal based on the discussion above.

• The Wikipedia Council (sorry for the pretentious name; if anyone has any better ideas, feel free to suggest them) will be a body of 40 members, elected annually. (40 should be enough to represent a diverse range of views and avoid cliquery, while being small enough to be manageable.)
• The voting system used to elect Council members will be approval voting, the same as that currently used for ArbCom. Any registered user in good standing can vote and can stand for election; candidates do not have to be administrators.
• Major changes to a policy, or adoption of a new policy, will require a vote by the Council. The change will require a 75% majority (30 out of 40 votes) in order to pass. That last part is just a suggestion; I don't know whether we really need to require a supermajority or not. Please discuss.
• The Council will not vote on a policy change until the community has had an opportunity to discuss the proposal and refine it to its final form. Addition: However, the Council will, of course, have the power to amend the proposal before voting on it, or to pass an amended version. But they should take account of community opinion as far as possible. Walton^One 22:24, 25 April 2008 (UTC)
• A more controversial part: The Council will also have the authority to order the desysopping of an administrator, by a 75% majority vote. I realise many people will think this point is a bad idea, so please discuss.

This may sound both pretentious and bureaucratic, but I believe it's what we need, considering the failings of the current wiki-legislative process. And from the discussion above, it looks like there's a rough consensus for it. Walton^One 22:02, 25 April 2008 (UTC)

The desysopping should stay with arbcom, I think. The legislative body should be for making policy. FloNight♥♥♥ 22:16, 25 April 2008 (UTC)

On reflection I think you're right; ArbCom is the "judicial branch", such as it is, and so it's probably better-placed to deal with such issues (which may require complex investigation). Walton^One 22:19, 25 April 2008 (UTC)

We don't need ArbCom in the first place. If we do decide to go ahead with this "judicial branch" idea, then arbcom needs to be re-elected and the terms shortened. Monobi (talk) 02:45, 26 April 2008 (UTC)

We don't have a wiki-legislative process in the first place. From the sound of it, having one fail sounds like a good idea. ;-) --Kim Bruning (talk) 22:08, 25 April 2008 (UTC)

Ah, this stems from your repeated insistence that policy is not in fact a prescriptive set of rules (a view you share, ironically enough, with Kurt Weber). However, as I stated earlier, I am firmly of the opinion that the rule of law is a desirable attribute in any community; rules must be clear, static and binding on everyone, including those in power, in order to prevent abuses of power and to allow users to know where they stand.

No one should ever be blocked, for instance, unless the blocking admin can point to a specific rule the blockee has violated; otherwise, how are new and inexperienced users supposed to know what is and isn't a blockable offence? A bad block can drive away a valuable contributor, and remonstrating with the blocking admin after the fact is often too late. So while some (particularly those who have been editing here since the early years, such as yourself) cling to the traditional model of informality, consensus and collegiality, it simply cannot work for a community of this size. (Or, more accurately, it can work, but it will lead to many good editors being driven away by abuses of power, and to problems not being fixed because of the community's inability to change things.)

Anyway, I don't want to re-hash the same discussion that you've had with about 20 different people in the threads above. I'm hoping for some broader input on this specific proposal. Walton^One 22:16, 25 April 2008 (UTC)

Policy is not prescriptive; consensus is. Rule of law does lead to stability, but it also leads to a system which can be easily gamed, wikilawyered, corruption and abuse of power by those in a position to dictate the rules, a strictly hierarchical community structure which we've been trying to avoid for years, stagnation and the death of "be bold" and the consensus model, and of course it would remove the ability of regular users to influence policy with the support of the community. A lot of people may trust "some committee," but when it comes time to choose the members of that committee, who do we trust with that sort of power? – Luna Santin (talk) 23:08, 26 April 2008 (UTC)

Kim, I don't want to sound snippy, but do you have any criticism of this proposal beyond "We've never done it this way"? :( Lawrence Cohen § t/e 22:13, 25 April 2008 (UTC)

Oh no! no no! If we had never done it this way before, I'd have told you to totally go for it. (provided you don't force others to use your way too)

But we have done it this way before, and it has never worked before because you're starting out with a bottleneck (commissions, central pages (potentially), votes), making it very hard to get things to scale very well at all. Get rid of bottlenecks and other enemies of scalability, and see where you end up. If the argument is that wikipedia is getting bigger, then logically you must introduce scalability, not remove it, else there's a bit of a fib hiding somewhere in your story, eh?

The other problem is that you're being all prescriptive and building rather large multi-tier castles, rather than getting some basic tenets down and then putting the show on the road and seeing where the issues are, live and direct on-wiki.

Finally, it sounds like you want to funnel all policy work through just this one commission, amplifying all of the above many times.

If you can resolve those issues, you'll have a winner. :-) But watch out, with the current starting point, that's going to be really hard, so you may need to start thinking from scratch. There's an entire profession dedicated to systems design, but remember: first put down the requirements, then start thinking of how to meet them. If you do that, you won't go too far astray.

In the mean time, we already have a method that works and which many folks are using, so my two major worries there are that you draw too many people away to an unproven method (not too bad), or worse, that the untested method immediately becomes the only way for wikipedia to maintain project space pages. FYI the latter a pretty good definition of "bureaucratic screwup", so don't let that happen!

I hope you take these points in the constructive spirit in which they are offered.

--Kim Bruning (talk) 16:39, 26 April 2008 (UTC)

I don't get this proposal. The community drafts the proposal and then they don't get to decide if it passes? This seems counter productive and counter intuitive. A two tier system won't work. You're looking to fix a problem that doesn't exist. The major problem on Wikipedia is one of factionalism. You won't solve that with any form of council or Parliament, you'll entrench it further. Hiding T 22:15, 25 April 2008 (UTC)

I agree that factionalism is a big problem, but at the moment vocal minorities can prevent any change whatsoever. We are never going to eliminate factionalism, and we are never going to all agree on how to change policy. But policy does need changing; and going with the majority view is better than no change whatsoever. What this proposal will do is give us an effective method for changing policy, and prevent vocal opponents of the change from derailing it. Walton^One 22:18, 25 April 2008 (UTC)

Hiding, that's not how I read the proposal, but I already asked Kiril who wrote it to weigh in. The way I read it is that the "Council" or whatever they're called in the end are the policy stewards. Anyone can propose a policy change, but it would (seem) to go in one central forum, rather than all over creation like it is now. Everyone weighs in that watches the page, and if it's some valid request/change/thing then it goes to the wider community to decide, like how we voted to adopt WP:3RR, and it goes "on the books". The method would make all policy-changes super-visible, and would make edit-warring over policy (as often happens today to try to quash change or enforce change) pointless. It basically will make all the Wikipedia Policy Wonks that guard their pet policy pages unemployed, and give full control to the masses. That's how I read it, anyway. Lawrence Cohen § t/e 22:24, 25 April 2008 (UTC)

Some of the drafting of policy would be done by the Wikipedia Council, I think. But the Community would still have a large part in it as well. Certainly will have a place for the Community to make suggestions for changes or express disapproval. But in the end, if the Wikipedia Council approves it, the the new policy or re-writes would be made. Or course, further review and changes are possible after a trial period. FloNight♥♥♥ 22:28, 25 April 2008 (UTC)

For me the detailed proposal sounds quite reasonable, and not in the least bureaucratic. There's one difficult point however: that the community refines a proposal to its "final form" - this will probably run into old problems (there will be no consensus as to what is the "final form"). It's a good idea to have as much community input on proposals as possible, maybe even leading to alternate versions of a proposal for discussion - but the "Wikipedia Council" probably needs the option to modify proposals (or select from different variants) as needed for compromise. --B. Wolterding (talk) 22:21, 25 April 2008 (UTC)

You're right, I've added wording to that effect. Walton^One 22:24, 25 April 2008 (UTC)