RE: markup race

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

So one could try to document the theories, not 
assert truth with regards to individuals past 
some shallow observables.

The use of physical characteristics for identification 
based on observables is purely process-oriented but 
points out that surface level characteristics are 
limited for inferencing.  My guess is the 
bigger problems are in the RDF applications where 
inferencing is part of the system and even though 
that is a GIGO dilemma, something that has been 
pointed out multiple times about the SW project 
and the necessity of labeling opinion and theory 
as exactly that.

Warning labels on Web sites. PICS rulz.

Len 
http://www.mp3.com/LenBullard

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h


-----Original Message-----
From: Kurt Cagle [mailto:kurt@kurtcagle.net]

Here's where things get complicated, however. One way of looking at HumanML
from a physical description standpoint is that it is a measure of
phenotype -- blonde hair, blue eyes, that sort of thing. We know today that
phenotype is an expressive rather than an intrinsic set of qualities that
are based upon genotype; this is one of the reasons why phenotypic
classification is so difficult ... the finer the granularity of the
classification, the more obvious it becomes that phenotype is a poor
classification mechanism, yet at the same time I think it is likely that no
one here wants to tackle the issues of creating a genetic map of human
beings.