Wikipedia talk:Community assent

[edit] Changes to policy

From the earlier Wikipedia:Village pump (policy)#HMAINS comments thread about the date-linking change to the MoS. I call it out here because it's caused me to consider the differing expectations people have about how guidelines (and to a lesser extent all policies) can change over time:

Over the past three years observing Wikipedia policy, I think I've observed two major but very divergent points of view with regards to policy and guideline changes, which I'd summarize as:

1. Be bold in making policies better. Editors who think a point of policy is incorrect, or see a way to make a guideline better, first simply edit the policy page to change it. They assume that those who care will be watching the policy page, and will revert it, dispute it, or use the other mechanisms we're all accustomed to in the article space. If the change remains undisputed for some period of time, then the changed text is every bit as actionable as any text previously in the guideline.
2. Get consensus first, then edit. Editors who have a problem with language in a guideline should first comment on the policy's talk page asking for consensus to change it. If no one responds, some editors will go ahead and make the change, but this does not appear to be universal. Any change made without following this process does not have consensus and is hence inactionable. A policy page that has been riddled with unreverted edits made without prior consensus can be considered "rotten" and the whole may become less actionable because of the bad edits.

Wikipedia:Policies and guidelines is silent on which of these views is correct, except for noting, "Amendments to a guideline should be discussed on its talk page, not on a new page - although it's generally acceptable to edit a guideline to improve it." "Generally acceptable" is, as someone said recently, a hole large enough to fly a large cargo plane through. Wikipedia:No binding decisions reaffirms that policies are subject to change by consensus, but does not prescribe a mechanism for doing so—though it does seem to affirm the idea that voting or a formalized amendment process is not necessary for change.

The problem with the lack of clarity on which change philosophy is right should be clear. The consequences of this lack of clarity show up on this page nearly every day.

Those who adhere to the "be bold" philosophy will participate in consensus discussions started by those adhering to the "get consensus first" philosophy, though they may complain about the overuse of process ("{{sofixit}}" or the like) or boldly incorporate the text suggested while people are still discussing it. They will generally perceive any text that has been stable in a guideline as actionable, and will behave accordingly, even to creating bots or running high-velocity, semi-automated edits to enforce the guidance. The idea that a statement that has persisted in a policy for a long time may not actually represent current policy is incomprehensible to them. They are liable to treat those in the other camp as whiners ("{{sofixit}}"), opportunists ("you're only complaining because you don't like the guidance"), or worse.

Those who follow the "get consensus first" philosophy are generally (it appears to me) less attentive to changes in policy until one surprises them. They point out that they're busy writing an encyclopedia, and the velocity of changes across policy pages is too great for them to audit each one to see if they disagree with it. They will assume that any important change would be brought to their attention one way or another (either at the village pump, or via a new section on a policy talk page). Any change that occurs without being brought to their attention is illegitimate, and a policy that has been extensively edited without prior consensus to do so is, at best, suspect ("a big ball of mud") and at worst has become completely inactionable from the weight of illegitimate edits.

Both sides accuse the other of wikilawyering from time to time. And in fact, the disputes that erupt tend to generate much heat but little light, with both sides at least insinuating that the others are not acting in good faith. Sometimes the result is a movement to ratify or reject the disputed policy change via a new consensus discussion or vote—which the "be bold" adherents will usually submit to, with some grumbling, because in their world view, the discussion should have already taken place following a quick revert of the original change. The new discussion does not address the legitimacy of the original edit, and so the question remains unanswered. Other times, the "get consensus first" adherents will simply walk away, dismissing the policy as no longer actionable. By their very (in)action, they reduce the value of the policy as a dispute resolution aid.

Not everyone falls into these two camps, naturally. My intuition is that it's something of a bimodal distribution. Even the boldest policy editor would think twice before making a major change to WP:NPOV without discussing it first, and even the strongest advocate of getting prior consensus would likely think nothing of updating a minor guideline in a straightforward way to account for a new software feature.

This suggests an avenue of compromise, one that most editors would probably already agree with me on: that sweeping changes that affect a large number of pages or editors or should be discussed prior to editing, and that minor changes that affect very little can be made without prior discussion. But what about all the other cases? When should prior consensus be expected? And is a change made without prior consensus illegitimate? If so, shouldn't all such changes be reverted as soon as they are discovered?

I think a policy change policy is called for here. I'm willing to write up a proposal, but I'd like to get some feedback first, in case I've misjudged the issues. (Please don't take that as evidence that I'm hostile to the "be bold" camp. ;-) --TreyHarris 09:40, 24 February 2006 (UTC)

A case in point: 3RR and in whole or in part

The characterization of those less attentive to changes in policy until one surprises them[, as] they're busy writing an encyclopedia, and the velocity of changes across policy pages is too great for them to audit each one to see if they disagree with it, certainly applies to me and my concern over in whole or in part. The challenges I has encountered in raising the matter are a case in point as to how policy changes or is preserved.

When I saw the change, I asked myself how recent it was and weighed that against that so broad a change went in with no apparent discussion. This in my mind overrode the sense that by sitting unchallenged for a few months, it had gained legitimacy. For raising the concern, I found myself immediately the brunt of cries of wikilawyering and gaming the system, which tended to obscur my actual concern: the dynamics of a policy change which undermined core foundation principles of neutral point of view and anyone can edit by rendering a broad spectrum of edits, especially edits in compromise, 3RR-countable.

This took what was to be an electric fence, a bright line not to be crossed, and rendered it a minefield, where one might not even realize one had violated the rule, spirit or letter. One could find oneself in a harmonious editing session with a collaborator, exchanging edits in compromise and tacitly accepting corrections of fact, only to find edits in that session counted towards 3RR when reverting an outsider once.

The deeper and more general paradox is that policy changes affect most disproportionately those editors who are unlikely to read policy pages.

How does one raise such a concern? First I reverted the changes (once), explaining the challenge and informing interested parties. A call went out to admins to prevent an edit war on the 3RR page— one can see here the genesis of what followed. Although I have never in the several years at Wikipedia been blocked for 3RR, I was immediately accused of edit warring, gaming the system, etc. An element to any policy discussion is the deeply biased outlook admins have about non-admins, particularly those concerned about 3RR issues. Towards anyone actually stigmatized by a 3RR block, the bias is considerably worse.

Although the matter is clearly in dispute, the {{activediscussion}} flag was repeated removed from WP:3RR, with admin threats of blocking and disruption. This brings us to another point: admins watch policy pages (good), but often act like they own policy pages (bad). Removing the flag would appear to be an attempt to starve the discussion of new participants. It is certainly against policy to forcibly remove the flag, but threats of blocking for disruption keep it off.

In principle, WP:CIVIL, WP:OWN, etc, should suffice to allow a reasoned consideration of the matter. Clearly, policy has to resist random attacks by the merely disgruntled in order to maintain coherence. However, with the manner of unsanctioned force and incivility now propping the policy up, how can it claim to be the result of legitimate consensus?

StrangerInParadise 16:54, 24 February 2006 (UTC)