[WikiEN-l] Re: Arbitrator recusal

[Michael Snow]

James D. Forrester wrote:

> geni wrote:
>
>> On 11/12/05, Delirium <delirium at hackish.org> wrote:
>> > There's an informal (formal?) rule that if a majority of arbitrators
>> > ever recuse on a case, they all are automatically unrecused.  The
>> > justification is that, in all probability, a situation that 
>> resulted in
>> > a user having personal conflicts with nearly every single 
>> arbitrator is
>> > more likely to be the fault of the user than the fault individually of
>> > every arbitrator.  A more practical justification is that it serves 
>> as a
>> > deterrent to trying to "win" a case by forcing everyone to recuse.
>>
>> No. Last time we disscussed this the general conclusion was that we
>> would call up past arbcom members and people from other languages if
>> posible. The solution you list runs into the problem that if you have
>> multiple sibjects to abitration it is posible for none of them to have
>> gone overboard in disputes.
>
> Erm. What on Earth do you mean, "no"? Mark was entirely correct - that 
> is exactly the policy we came up with back in January 2004, when we 
> wrote the policies in the first place.

That was the policy as it existed in its original draft form, but not 
the policy that was actually adopted. The provision for "unrecusing" 
already recused arbitrators was dropped from the policy based on my 
objections, see [[Wikipedia talk:Arbitration policy comments]]. I refuse 
to accept that this could be policy, formal or informal, since it's 
never been done in practice and was specifically objected to at the time.

The better approach is simply for arbitrators not to recuse themselves 
over trivial issues, or when they have merely been targeted by the 
editor in question. For these reasons, I support the refusal of Kelly 
Martin to recuse herself in the Silverback case. The conduct of the 
arbitrator, not the person whose conduct is being arbitrated, is what 
determines the necessity of recusal.

Geni is correct that in March there was some discussion of how to deal 
with arbitrator recusals. See this message I posted about the issue: 
http://mail.wikipedia.org/pipermail/wikien-l/2005-March/020510.html
My proposal received some acclaim, but I wouldn't say it was formally 
adopted, and in practice we've never had occasion to need it. I still 
think the "all arbitrators have to recuse themselves" scenario is pretty 
chimerical.

However, I maintain that we could benefit greatly from developing a 
group of temporary arbitrators, or magistrates, for the purpose of 
reducing the workload, filling in for recusals, and allowing cases to be 
handled by smaller panels. It would also get us out of the dilemma we 
currently have, between elections where we as a community make guesses 
about who's qualified and will actually do the work, or having Jimbo 
appoint everyone based on his own guesses and the advice of those he trusts.

--Michael Snow