Dear Milton, Thanks for the reply. This conversation has been very useful to me, and I hope to you too. The core of your view, I think, is a sort of ICANN exceptionalism: ICANN is, by your account, "the only truly global institution with globally binding authority over aspects of the Internet," and that makes the stakes high. But regardless of its global charter, I think ICANN's potential range of action is, as a political and operational/technical matter, very narrow. Let's run through some examples that David Post provided in another forum:

ICANN adopts a rule: no registrations of names to any person found (in a UDRP-like proceeding??) to have infringed copyright more than X times in the last Y years. It flows the requirement through to registries, registrars, and everyone further down the domain name chain. Or, how about a UDRP for "pornography"?

First, I'm curious how possible you think either of these policies is. My own sense is that, under ICANN's current structure, they are non-starters. ICANN would not promulgate such policies, not least because most of the interests in the room -- even after subtracting the public interest advocates that you think are not sufficiently staffing the barricades -- are indifferent or opposed to both of these ideas. The first policy is an administrative nightmare, of course -- I really can't imagine even the publishers wanting to go for it, especially because most of the infringement they care about is taking place through P2P nets, which rely on domain names not at all. There'd have to be theories of contributory infringement -- a la Youtube -- and once in that zone, there are very powerful interests who would stop it before it started. This is not an argument from theory, of course; new facts could undermine my belief. So suppose ICANN did pass these policies. First, there exist, and will continue to exist indefinitely, ccTLDs that have not agreed to flowed-down requirements from ICANN like the UDRP. People would migrate their domain name business there. (Reason enough for registries like .com to resist any new requirements.) Those individual TLD operators -- often themselves independent sovereigns -- would have to be forced into acceding to a scheme that they haven't even subscribed to for simple cybersquatting. Second, try to force those TLDs into line through ICANN's own power -- whether or not the TLD is part of the root -- and indeed there would be a precipitating event that would cause major ISPs and others who make use of DNS information to get their information elsewhere. I don't know if it would truly and fully split the root, but they would simply manually cache the information about where, say, .tv could be found, and only subscribe to the ICANN root for data about everything else. The inertia of the existing system is strong. The same could be said for IANA and the regional registries in numbering: imagine a policy that somehow came out that said "people who do X will lose their IP addresses." IPs cannot be yanked; they can merely be allocated in the sense of broadcasting to the world that conflicts will arise if anyone other than the designated party tries to use them. If the allocator tries to work too much policy into maintaining blocs (much less individual IPs!) that have already been assigned, the ISPs who control the routing tables will give a shrug. A map might indicate a road going into the ocean, but the driver will break rather than follow the map. So, we return to the question of how much to try to involve the public at large on this. I certainly don't begrudge you the time you've committed to watching ICANN and participating in its processes. But from the point of view of legal academia, I see no shortage of attention and articles about ICANN. I think ICANN's exceptional position as a global body concerned with the Internet is indeed what draws people to those meetings. They provide valuable networking opportunities and perhaps test runs of how to collaborate on policy issues that defy the borders of a single sovereign. But I do not see the link between this and some special lever that realistically can be deployed to effect major censorship. Our starting point was content-based decisions about new TLDs. That has ICANN's power at its height, because there's not the inertia of an existing TLD. Add none, add five, or add a hundred -- I don't see much difference, except perhaps in the price of a name. Best, JZ