Re: [wg-c] Re: nine principles for domain names

["Mark C. Langston" <skritch@home.com>]

On Tue, Feb 22, 2000 at 09:16:29AM -0800, Bret A. Fausett wrote:
> > As to "memorability" vs. "meaning"...  This attempts to draw a stronger
> > distinction than the context allows, especially since this context is not
> > one limited to psychologists and/or linguists.  The distinction also might
> > be wrong.
> 
> IANAL (I am not a linguist), but I simply wanted to point out what I thought
> was a missing element of the proposed principle. A good definition will
> encompass both meanings, and I think Philip's reformulation and his
> statements on this list satisfy that concern.
> 

IAACAYGTAL (I'm as close as you've got to a linguist -- a.b.d in
experimental cognitive psychology, focusing on human memory,
comprehension, questioning, and inference generation, with heavy
linguistic crossover -- references available on request, most current
work published in "Understanding Language Understanding: Computational
Models of Reading", Ashwin Ram & Kenneth Moorman (eds.), 1999, MIT
Press., Chapter 5.)

I'd just like to point out that if you're going to start paying
attention to research in these areas w.r.t. gTLDs, I'm going to
strongly recommend people start seriously considering my pleas to
consider the past 30-40 years' research on categorization,
memorability, and generalization over in WG-B.

What lawyers would like to consider memorable and what psychologists'
hard data point to as memorable do not overlap as often as many here
would like to believe.

Same goes for meaning.  But that's another argument, dealing with
contextual dependency.  Suffice it to say that most of the discussion
regarding memorability and meaning has been missing the mark.  The
answer, which many of you may not like, is that there is no meaning
without context.  And in this arena, context comes from two places:
the TLD, and the current services provided under the IP pointed to by
the FQDNs under the given SLD -- which, I'll remind everyone again, is
more than just web pages.

These two are nonseperable; the TLD itself only has meaning in the
context of the FQDNs under it, which themselves derive meaning as
above.

A string of letters in isolation has no meaning, except that which
prior and current context brings to the table.  That meaning is fluid
and dynamic, and wholly dependent upon context.  That context is fluid
and dynamic, and wholly dependent upon the actors who work to create
said context.

In short, a TLD has no intrinsic meaning.  An SLD has no intrinsic
meaning.  A FQDN has no intrinsic meaning.  None of these have any
deep memorability, except as provided by meaning, and dictated by how
humans categorize.  It's nonsensical to talk about the memorability of
a TLD in isolation, as it's never used in that way.  At best, you can
facilitate or hinder the human's ability to 'chunk' the TLD.  Same
goes for the SLD, and the SLD and TLD in combination.  This all falls
under shallow memory effects, like the "7 +/-2 rule".  Similarly, it
goes out the window once there is a deeper meaning associated with the
string of letters and numbers that make up a FQDN.  Then, it's all
context, and all categorization.


-- 
Mark C. Langston
mark@bitshift.org
Systems & Network Admin
San Jose, CA