RE: [huml-comment] FW: [KMForum-KScience]"Motifs distinguish net works"

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

Wiring is communication channels, gazintas and gazoutas. 
I definitely think the pattern work will apply as you say.  
Hmm... I wonder if the same routing patterns show up in 
VRML ROUTEs.

The thing to look for of course, is how the HumanML 
category values would affect the choices of route 
patterns and do the route patterns affect the selectors 
for sign semantics (choosing the interpretation).  
Intuitively, I would think so because of such things 
as inheriting a pattern from the culture or the group. 
For example, my father was a believer in and taught 
me to do business with locals as much as possible 
over going to another county to buy a car to save 
a few hundred bucks.   He said the costs were 
insignificant next to the value of building reliable 
local relationships.  Sort of downhome keiretsu but 
proximity oriented.   A question then is, how do 
the choices of network patterns build or negate 
proxemic values and then, the effect overall of 
shifts in proxemics.  For example, choosing non-near 
nodes means learning non-local signs.  What are 
the payoffs and downsides?  Greater scope of command 
and control through broader if shallower communications?

Unfortunately, we have the news article but not 
the actual technical report.

len


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

If the motifs are recurring and show definable patterns, the recent 
work from Wolfram might apply. The advantage to that is that 
eventually these and other patterns  when recognized can be analyzed 
and manipulated mathematically. Or we might be able to provide new 
computational operators. Imagine being able to change initial 
conditions from one kind of pattern to another to see differing 
outcomes or choose between such outcomes for the appropriate initial 
governing conditions.

I do find the wiring metaphor a bit cumbersome.