RE: [humanmarkup-comment] Referents

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

Righto.  I have a lot of work to do on 
the sign experiment.  I have a good handle 
on the pieces, but need to fit themtogether.

One issue is that in the Gudwin model, 
that the reflexive processor has 

Y=f(x)

and that is grossly simplified given there 
is/are feedback loops.  His layering looks 
right though.  Or maybe he just intends that 
function for the reflexive layer.  As 
additional layers are added, it get more 
complicated and becomes more like objects 
routing messages (and VRML likes that) 
and consulting its database.

The sign experiment is a way to bottom 
up toward the primary top down.  We 
get a useful secondary for beating on 
the primary.  When Manos and Sean said 
the primary is a kind of loose ontology, 
they hit the nail on the head.   
Our commitment to RDF stripe the secondary 
should also drive out aspects of the RDF 
representation of the primary.  Don't be 
surprised if we discover that we need more 
RDF up there or that we end up being 
more comfortable with the RDF representation 
because it proves to be easier to extend.

But I need Manos to keep the RDF straight. 
It isn't my bag.  It may be before we are 
done.   I think it more important now than 
I did before to ensure what we build is 
accessible to and from the Semantic Web 
because that is one source of a semiote's 
data.  A semiote is an agent; it has been 
modeled on our concepts of how humans 
communicate.  That is it's primary 
differentiator.

len

From: Rex Brooks [mailto:rexb@starbourne.com]

That's sorta what I thought.

However, I did want it in the list for looking at again when we 
finish the first pass through the existing schema.