June 1, 2001
To: Stuart Lynn, ICANN
From: David Post
Re: Your Discussion
Draft (28 May 2001) "A Unique Authoritative Root for the DNS"
You have asked for
comments concerning the ideas that you have put forward in this draft. There
are any number of points that you make that are worthy of serious discussion in the
Internet community; I will confine myself to a small number of comments about what
I regard as the most serious and far-reaching of them.
1. From the Abstract:
"Although the Internet allows a high degree of decentralized activities, coordination
of the assignment function by a single authority is necessary where unique parameter
values are technically required. Because of the uniqueness requirement, the
content and operation of the DNS root must be coordinated by a central entity."
is, simply and most profoundly, incorrect. The analogy - and it is as precise
an analogy as one can ask for - is to language systems generally, or any naming system
(e.g., the system of 'biological classification' under which species are given their
proper 'names'). [If you are interested in a more comprehensive development
of this argument, please see http://www.temple.edu/lawschool/dpost/TheFreeUseofOurFacultiesFinal.PDF]
The rules of the English language give names to objects - table, chair, domain, computer
... There is a "uniqueness requirement" if we are all to understand one another with
precision; if you call the object on which I am writing this note a "tomato,"
and I call it a "computer," we will not be communicating very efficiently.
Confusion and chaos will reign.
But somehow we muddle through without a
central coordinating entity. It's amazing, no? You and I can somehow
communicate with one another - not perfectly, not without ambiguity, but reasonably
effectively. All without an authoritative source of definitions! Sure, there
is a great deal of confusion in this system; there are innumerable private languages,
dialects, jargons, etc. out there, and navigating through them all is a complex task.
But that is the price we pay (and we must pay) for the evolution of the languages
we speak. Not only do we take for granted that we do not have a central coordinating
entity - some ICANN-like agency that tells us which words we need and which we don't,
or how conflicting uses are to be resolved - we take for granted that such an entity
would be a catastrophe for the continued richness of the language itself.
I want to set up an "alternative naming system," I do not need Stuart Lynn's permission
or anyone else's; I am entirely free to do so - subject, of course, to the constraint
(an only to the constraint) that my new system will only be effective to the extent
I can convince other English-speakers to use it.
If you think that you
have a better plan for English, I respectfully suggest you're wrong. Many people
throughout history have thought they had better plans, by the way. ICANN
has its predecessors. The 18th century saw a vigorous debate about the need
for a central coordinating entity for natural languages - and ICANN lost that debate.
The pronouncements of the Academie Francaise about "authoritative" French usage had,
let us not forget, the force of law at one point in time; and well-meaning and deeply
intelligent people thought we needed such a "central coordinating entity" for English
as well. (John Adams among them; Adams proposed an equivalent Academy of English,
to check the "natural tendency" of language to "degenerate," to the Continental Congress).
History has shown the misguided nature of these efforts, and it will show the misguided
nature of ICANN's efforts as well.
This is not to say that there is no room for
entities who play a coordinating role; far from it. Dictionaries, learned societies,
usage pundits, and all the rest have an equal claim to providing an "authoritative"
source of names - and we, the users of the English language, get to decide which
(if any) of them we want to obey. It's called unspoken consensus, and it works
- indeed, no other system can work as well. ICANN can earn that consensus position
by virtue of its stewardship of the DNS; it cannot, and it must not, simply declare
itself to be playing that role and demand that we all follow its dictates.
From the Summary: "Put simply, deploying multiple public DNS roots would raise
a very strong possibility that users of different ISPs who click on the same link
on a web page could end up at different destinations, against the will of the web
"Against the will of the web page designers"? Since when
is that the criterion of effective DNS performance? Let me repeat: I
can call the machine I'm working on a tomato. It is my will that everyone on
the planet do so as well. It would be nice, I suppose, if I could impose my
will on others, but I can't, and I shouldn't be able to.
3. From Section
5 ("Experimentation"): "DNS experiments should be encouraged. Experiments,
however, almost by definition have certain characteristics to avoid harm: (a) they
are clearly labeled as experiments, (b) it is well understood that these experiments
may end without establishing any prior claims on future directions, (c) they are
appropriately coordinated within a community-based framework (such as the IETF),
and (d) the experimenters commit to adapt to consensus-based standards when they
emerge through the ICANN and other community-based processes."
is a nice word. The notion that ICANN's process are consensus-based is a bit
much, don't you think? By all means, DNS "experiments" should be subject to
consensus-based processes - as they will be if ICANN leaves them alone, for no DNS
experiment can "succeed" unless somehow the hundreds of thousands of Internet Service
Providers out there agree to resolve names in the manner proposed. As
we used to say in Brooklyn, that's the beauty part; it is inherent in the nature
of language that it requires consensus to be effective.