Wikipedia talk:Changing policies and guidelines

This proposal came from a discussion on WP:VPP where I noted the confusion developing between people who felt they should use the WP:BRD cycle for policy and those who felt that any policy changes that weren't widely discussed first were not actionable. I have attempted to explicitly document the current practice that has led to least discord, rather than to establish any new protocols or procedures. (For an example of a proposal that would change current practice, see Wikipedia:Community assent.) --TreyHarris 09:08, 2 March 2006 (UTC)[reply]

It is great that TreyHarris has proposed this policy. There is a lot of hard work put into the effort. We share a common gist, and I hope that our work is less competitive in the end.

Wikipedia:Community assent would not change practice in a worldly sense, so it is not that radical. The changes only reflect upon what has been done here on Wikipedia, which has only been live within the last decade. Parliamentrary procedure is common and has been around longer than the last decade. Most parliamentary procedures have been based on minute-by-minute, face-to-face debate motions, but this is online text, so the motions have to change a bit to suit the media. It is not easy to simply incorporate parliamentary procedure "as is." Further, the Wikipedia:Stable versions or certified versions seems like a new radical idea, but it is actually a proposal that has been on the table for awhile. Most source repositories have a stable branch of code, and that is the basic idea. Wikipedia:Community assent does not invent anything new, but it tries to incorporate a couple of common external-to-wiki practices into Wikipedia.

Some notes on the text so far:

"Policy pages allow administrators to "invoke a higher authority" in warning disruptive users to stop, and to block those users if they continue violating the policy pages' guidance. In this sense, policy pages help not just with article editing disputes, but with disputes between users as well."

The use of admin's ability to block for such reasons is not by any recommendation. There are specific cases where we can find an assertion of its use. However, there are many more general cases where such assertions do not apply. The basic recommendation is for admins to intervene or protect the article page. Such intervention usually means an active role in dispute resolution. Please, do not try to justify here a practice that has resulted in a user being blocked where there was a lack of proper intervention such as page protection or dispute resolution.

"It is always allowable to get consensus first."

There is no grounds for what is or is not always allowable. It is not always allowable to get consensus first because we should use the talk page conservatively. I was about to be bold myself and make the change myself, but I decided to allow this proposal some time to develop and just drop a note for future reference.

On "Two ways to change policy," since we have both used the same idea (and I used TreyHarris's format), we should make this section have its own guideline page. This section is clearly distinct as common pratice from either proposal.

— Dzonatas 11:10, 2 March 2006 (UTC)[reply]

The use of admin's ability to block for such reasons is not by any recommendation. I'm not sure what you mean, but I'll assume you mean "there is no justification for admins blocking for such reasons". As far as I know, every block should be because the user has repeatedly violated policy. If you're reading this sentence as saying that any admin can and/or should block any user who repeatedly violated any policy, without ArbComm intervention, then I'll see how to rewrite it so it can't be read that way—I would think that a definitional portion of a policy on changing policies would not be misread as giving administrators some sweeping new power they didn't have before and having nothing to do with changing policies! But I'll come up with some new wording to make that explicit.

There is no grounds for what is or is not always allowable. It is not always allowable to get consensus first because we should use the talk page conservatively. I find this a remarkable statement. Can you give me an example of an edit that we would disallow a user from discussing first, that they would be required to make on the page first? --TreyHarris 11:26, 2 March 2006 (UTC)[reply]

About the block, the words used gave the impression that the block is the only option if a user does not follow policy or guidelines. There are general recommendations to intervene for dispute resolution by the admin, by a mediator, or by anybody else and not just ArbCom. There was a vote to block if any user reverts four times in less then a day, but that was based on identical reverts. There is a widely accepted notion of the use of a block to stop continuous vandalism, or there is no substantial argument to allow vandalism outside of the sandbox page. There is also a note at at Wikipedia:Office Actions. [[1] None of these practices supports a recommendation to generally block as the first option.

Once a policy becomes official, any portion of its text may be used as "official," the 3RR page dispute is a good example where minor changes after such "wide acceptable" have been stated continually as "widely acceptable" as to justify further action. I'm sure you did not intend for it to be used in such a manner. Perhaps, official policy pages should have a section that points out exactly the official text and other sections for further descriptions and references.

For the example about being conservative on the talk page: I did not state that we disallow or require one or another method to edit a page, so there is not an example I can give to your question. — Dzonatas 13:39, 2 March 2006 (UTC)[reply]

I've removed the reference to administrators and blocking; given your comment, I can see that it confuses more than clarifies.

As for the other issue, I'm still confused. The sentence currently reads, "It is always allowable to get consensus first. It is not always allowable to make your change before getting consensus." If you disagree, then you must believe there is a case when it is not allowable to get consensus first, correct? What would that case be? I don't know what "being conservative on the talk page" means. --TreyHarris 00:52, 3 March 2006 (UTC)[reply]

In general, no. Otherwise specifically, attempts to get consensus to overide policy are not always allowable, like to get a consensus on a outnumbered view to move away from a nuetral view. — Dzonatas 01:54, 3 March 2006 (UTC)[reply]

Guidelines on wikipedia (I stress, GUIDELINES) , are current best practices on which there consists some level of consensus that that is indeed the best practice. Everything else is nomic. Ugh. This mess is propagated by trashy "policy" pages which are nothing of the sort :-( Kim Bruning 13:29, 3 March 2006 (UTC)[reply]

Wow. From your edit to VPP, I guess this really upset you. I apologize, but I'm confused. I was observing all the same policy disputes you were, and noted that the one thing that seemed to be setting most off was lack of agreement on exactly how policies can and cannot get changed. You refer to "nomic", which is a game I hadn't been aware of before. I can see where any talk about the rules, let alone changing the rules, could remind you of that. But I'm not playing "nomic" here; I wasn't trying to advance my own position in a dispute. If you doubt, look at my contributions of late and see if you can find a dispute that this proposal would help me in. I saw a problem, the problem seemed to be diverting a lot of energy from the encyclopedia, and was trying to be helpful in finding a way to resolve it. What's wrong with the proposal? How can it be fixed to address your concerns? --TreyHarris 16:49, 3 March 2006 (UTC)[reply]

The thing is that our original method to write down guidelinews was so obvious that no one really thought about it. Of course, now we end up stuck with folks who never heard of the original method :-P

It's really very simple. When you go out and do stuff, see if there's some guidelines to help you. If none really seem to apply or work, and you get depressed, well, just talk with folks and get some consensus to do it some more useful way for your situation. That's where the whole ignore all rules is coming from you see. Now to prevent the whole thing from happening again, it's probably a good idea to write down what you learned someplace, so make a new page in the wikipedia namespace, and maybe mark it as advice or "guideline", if you happen to be a tagophile.

Told you it was too logical to write down! ;-)

And yes this method still works, even on en.wikipedia. See WP:WOTTA and Wikipedia:Adminitis for some recent light-hearted examples. (By the way, writing down things in a funny way really helps keep people happy! :-) ) Kim Bruning 17:19, 3 March 2006 (UTC)[reply]

Note that Bold Revert Discuss is an emergency measure only intended to be used when the above fails utterly. Once conversation is moving, you can typically switch back to the logical method :-) Kim Bruning 17:28, 3 March 2006 (UTC)[reply]

Right. I don't think the proposal here seeks to change any of what you just said, and I don't read it as doing so. I'm not so concerned about additions that come about as you say; most of them will never result in dispute in the first place. What I'm concerned with and why I drafted this proposal was the cases where dozens of editors were getting sucked into protracted meta-disputes over whether a change to policy was or was not legitimate. There are lots of cases in article space where there are kneejerk editing reactions (usually reverts)—simple vandalism, spamlinking, nonsense. These kneejerk reactions are good things—they streamline the editing process and don't force us to stop and enter dispute resolution every time. All I'm doing with this proposal is validating a kneejerk revert on policy pages when the change causes a change in dispute resolution.

Look at any policy page history, and you'll see cases of bold, revert... re-revert. All the major disputes on VPP in the past few weeks about changes getting "slipped in" to policy were symptomatic of this. All I'm trying to do is clarify those cases, where people try to, as you say, play "nomic" with Wikipedia. This policy, in a sense, outlaws "nomic"—you don't get to change the rules if it will change the result of the "game". What I'm really trying to do is reduce the chances to "game the system" so that maybe people will stop treating policy pages as a big game. If people monitor policy pages and revert any change on sight that changes dispute resolution results, then we'll have a chance to go through the process you outline above, and we won't end up in these battles over whether the current revision of a policy is actionable or not. --TreyHarris 20:20, 3 March 2006 (UTC)[reply]