At 08:31 AM 6/18/2008, Carl Beckhorn wrote:
On Tue, Jun 17, 2008 at 11:47:40AM -0400, David Goodman wrote:
This is a proposal that will encourage administrators to not act responsibly, by destroying the principle that an administrative action can be overturned by another administrator.
Independent of BLP issues, that principle has always been a problem. Permitting any admin to unilaterally reverse any other admin's action harms the collegiality of the admin corps and undermines the individual responsibility of administrators. When it is clear that the original admin would not agree to having their actions reversed, discussion is in order, not unilateral reversal.
Actually, no. How the Wikipedia system works has often been missed. Any admin can reverse any admin's decision, and this is a protection. Essentially, any admin may unilaterally take an action, thus allowing many actions to be taken with no discussion. If there is no opposition, then it's done, which is highly efficient. This, by the way, is not as far from standard democratic process in direct democracies as we might think.
But if there is an objection, properly, the original action shouldn't stand, until and unless it is confirmed. Procedurally, courtesy would suggest discussion before reversing what an admin has done, but that discussion, for efficiency, could be confined to the original admin and one who disagrees. It should not require appeal to a broader circle to reverse an opposed action, for it is quite possible that the original aministrator will decide that it's not worth the fuss.
But once an admin action has been reversed once, both the original admin and the reversing admin shouldn't touch it again, any further action should require yet another admin, and broadening circles of discussion. A single reversal of an admin action isn't wheel-warring. Repeated reversals, by the same admin, are. Wikipedia, when it works, gathers together, ad-hoc, sufficient numbers of editors that a community consensus becomes apparent.
BLP policy may set certain biases in this process, there can be a bias toward removal of controversial information, pending appeal. ArbComm, however, seems to be setting too high a standard, one of "clear consensus." That's dangerous, and is well-known, in consensus organizations, to lead to a kind of dictatorship of the minority. Rather, that's what we have ArbComm for, to make decisions when consensus isn't clear. It would seem that ArbComm is attempting to replace its function with rigid rules, which is why I noted that it seems they've lost their collective mind, and they seem to be moving beyond their franchise.
ArbComm can issue temporary injunctions; indeed, a procedure could be that any ArbComm member can issue such an injunction, which can be cancelled by another, leading to a broader decision if necessary. So if ArbComm wants to set up some operating rules to govern BLP issues, it can easily do it. It is setting up rigid rules that change how the community traditionally operates, without finding consensus for that in the community, that's a problem.
Here is the collision that could take place, if the guideline is accepted: a majority of editors believe that some information is properly sourced and balanced and belongs in an article. (and I mean by "majority," a majority of those who are informed on the issue). Similarly, there is a majority of administrators. But this could be short of "clear consensus." An admin takes the information out and protects the article. After discussion -- which wouldn't even be necessary with a non-BLP issue, for protecting a favored version is generally a big no-no -- another admin sees support for putting it back in, and does it. Is ArbComm going to allow some kind of automatic blocking or desysopping for a good-faith action by an administrator taken with majority support? Is it going to reverse all prior practice and desysop that admin itself, based on a violation of the new rules it set?
What I'd see as proper and not conflicting with the legitimate goals of the proposal is that any admin who has not previously touched the article would, once, take the allegedly offending information out. Given the conditions (the existence of controversy, and mere majority support for it being in), it shouldn't be difficult to find such an admin, and quickly. The proper BLP bias would at this point suggest much broader discussion before putting it back in; and, if controversy continued, ArbComm could step in with an injunction, quickly, and then consider the case in more deliberative fashion.
I think we need to understand and respect how Wikipedia works, first, before fixing it. What I've described as proper already exists or could be made to be so by unilateral action by administrators or members of ArbComm. I'd consider it proper for any member of ArbComm to issue a preliminary injunction that would stand,pending a decision, unless opposed by a majority of arbitrators (it could thus shift back and forth a little, if the matter is truly controversial within ArbComm, as Arbitrators show up and weigh in on it.)