[discuss] Who-and-where vs. what

Andrew Sullivan ajs at anvilwalrusden.com
Wed Sep 3 13:02:20 UTC 2014


On Wed, Sep 03, 2014 at 03:29:51PM +0300, Jordan Carter wrote:
> sending it, Andrew. I agree with you largely. We do not want a system of
> Internet Governance that requires particular people or entities to be
> trusted: we need a responsive system where if bad actors get into a
> position to shape things, or an institution goes off the rails, the
> community can act together to correct things.

Well, what I want is _both_.  That is, I want the normal functioning
of the system to depend partly on trust and relying on people's
judgement, because I believe that any attempt to write perfect
procedures that require neither judgement nor trust will fail.  (It
will fail in two ways: the procedures and rules will be so complicated
that none of us could understand it, and anyway such rules do not
guarantee good outcomes.  GM was, as I understand it, fully ISO 9001
certified when it shipped all those faulty ignition switches.)  But I
want a mechanism to redress cases where that trust and judgement
fails.  If it is constructed correctly, it should be just painful
enough to use that it's not invoked all the time, and it should be
just easy enough to use that the corrupt aren't tempted to ignore it
because it can't be used effectively.

> A system of redress and change such as you suggest can achieve that.

I want to emphasise my suggestion was on purpose too low a bar.  I
note that the IETF has such a system, which is part of why I've been
convinced that the IETF's policy oversight of the protocol parameters
registries is just fine.  We have dispute-resolution mechanisms (which
we call "appeals procedure", but it's not like an appeal to an outside
body), and in the worst case we have a recall procedure.  They're hard
to use, and the bar for their use is intentionally high.  But that's
sort of the point: instead of recourse to "who's in charge", we have
an arrangement that forces people to make arguments and work out
disputes, _or else_ a much less happy outcome results.

Best regards,

A

-- 
Andrew Sullivan
ajs at anvilwalrusden.com



More information about the discuss mailing list