User:Jayron32/On positions of authority at Wikipedia

Statement of the problem
Regarding all of the recent controversies at Wikipedia, there seems to be some general feeling that the people who are placed into positions of authority at Wikipedia, such as Admins, Bureaucrats, ArbCom, Checkusers, etc, etc. are somehow untrustworthy, and need to be watched closely or else chaos will ensue. Consider all of the following pages, instituted in the past year or so:


 * User:Tony1/AdminWatch
 * Review Board
 * User:Thatcher/Quis custodiet ipsos custodes
 * Requests for de-adminship
 * Removing administrator rights/Proposal
 * Admin Accountability Alliance
 * Community enforced administrator recall
 * Administrators open to recall
 * And of course the rather acrimonious 2008 ArbCom elections

Here is the basic situation. For one reason or another, people elected/appointed/whatever to these positions of trust in the community may misbehave. And based on that misbehavior, some method needs to exist to remove their authority should that become necessary. Now, nearly EVERY proposed method of dealing with this perceived problem results in one of two basic types of processes:


 * 1) Community-based sanctions
 * 2) Appoint a board of trusted members to provide oversight

Both of these situations have their problems. Lets see why...

Community based sanctions
There already exists, at Wikipedia, means for discussing and implementing community-based sanctions. Some of these include:
 * Request for Comment: User Conduct
 * Administrators Noticeboard
 * Administrators Noticeboard/Incidents

So, here's the idea. We need a place to hold centralized discussions whereby the community can look at the evidence and decide to remove the trusted status from an Admin/ArbCom/Checkuser/whatever. Seems like a good idea, right? I mean, Wikipedia is built on consensus, its built on discussing matters then deciding on what the community wants? So why is this a bad idea? Here's why:


 * 1) These process all work the same way. Someone raises a complaint.  Someone proposes a sanction.  People weigh in on what, if any sanction should or should not occur.  All of this already happens.  Why are we creating a NEW process, just to do exactly the same thing that the OLD processes do?  There isn't any other way to do this sort of thing.  It can only work one way, and so if the OLD processes are inadequate, the NEW processes will be inadequate for exactly the same reasons.  Its the system of "putting it before the community" that carries its own problems, and any new system that does this will suffer from the same inadequacies people complain about with the old system.  Consider the failed experiment of Community Sanctions Noticeboard.  Why didn't it work?  It was brought down by the same inadequacies that led people to think it was needed in the first place.
 * 2) Who shows up to these community based discussions? Two groups of people: A) People who are interested in policy and procedure and the behind the scenes work at Wikipedia, and these people are almost exclusively admins or future admins.  B) People who think they have been wronged.  The problem with these sorts of processes is that no one impartial to the situation voluntarily shows up to read and discuss.  You can't force them.  So what you get is either an organized group of editors with a vendetta against a specific editor, or you get no one at all commenting on the situation.  Its inherant in ANY community-based system of sanctioning editors.  This will always be a problem, and creating new community-based systems of sanctions won't solve anything.

People knew about these problems 200 years ago. Read Federalist No. 10. James Madison knew that direct democracy had these problems in 1787. Direct democracy STILL has these problems. Don't get me wrong, I do think that ANI and RFC/U are quite useful, but insofar as they DON'T work, creating a new place to have these sorts of discussions won't work for the same reasons.

A board or committee of some sort
So, there are problems inherant with direct democracy. The alternative is to set aside a trusted group of editors who the community trusts to oversee these questions. So, we have admins. Then, we need a group of MORE trusted users to have access to the technical information that Admins don't. SO we have Checkusers. And we need a group of EVEN MORE trusted editors to deal with all the really hard matters that admins and checkusers and the community are not equiped to deal with, largely for the exact reasons spelled out in the section above. So we have ArbCom; Wikipedia's court of last resort. ArbCom is supposed to handle all of the sorts of cases where the community has tried or failed, or where the community could not be reasonably be expected to deal with it. ArbCom is small (currently 18 members), and elected by the community on expectations of the level of impartiality, wisdom, and trustworthiness they will bring to the job.

So, then we get to the ultimate question: Who watches the watchmen? Or, in Latin, Quis custodiet ipsos custodes. Thatcher has written a fantastic essay about this problem at Wikipedia. See User:Thatcher/Quis custodiet ipsos custodes. I recommend that everyone read it. However, it suffers from one major flaw. The question is unanswerable. Eventually, there needs to be unwatched watchmen.

At some level, there needs to be a group of people who are so trusted by the community that they have the ultimate say. At Wikipedia, that group is ArbCom. It's 18 members, hardly an unweildy number. If the ArbCom has lost that trust, it's the people who hold the position that are the problem, not the organization itself. Are the people in ArbCom untrustworthy? Then we need to get trustworthy people into ArbCom. The solution is not to create a new committee or review board or commission or whatever. Because guess what? That body will be made of people. Who need to be trustworthy. If we just create another board which is doing what ArbCom is supposed to be doing, that is making decisions about the really hard stuff that the community as a whole cannot work out, then its just ArbCom under another name. And it will be just as vulnerable to the same sorts of weaknesses that ArbCom is; that is it will be made up of people who may someday lose the community trust.

So we create a Review Board which will review the actions of Checkusers and Admins and Arbitrators and decide, if the Checkusers and Admins and Arbitrators have stepped out of line, and if they have then they have the power to remove those people from power. And then what if the Review Board steps out of line?!? Do we create an Oversight Committee with the power to remove Review Board members? Then what happens if the Oversight Committee falls out of trust? Do you see where this is going? Eventually, there needs to be an ultimate end group, which is small enough to be effective, with enough power to be useful, and which is still open to periodic re-elections to be responsive to the community. We have that group. Its called ArbCom. Creating new ArbComs does not remove the problems with the old ArbCom, because the new ArbCom, whatever you call it, is still made of people, and those people may or may not be trustworthy.

So what do we do?
Nothing.

Seriously. We have existing systems that need to be expected to work. No, we need to demand that they be made to work. Why? Because any new systems we create will only suffer from the same inadequacies of the old systems. Creating new systems is not a way to escape these problems or prevent them from happening. Accept that they will happen. But demand that the systems we have in place be made to work as intended. If you don't, and just think you can create new systems which will somehow be immune from these problems, you will be sorely disappointed.