RE: [humanmarkup-comment] RE: [topicmaps-comment] multilingual thesaurus - language, scope, and topic naming constraint

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

AFs could conceivably be created dynamically.  They 
are a syntax for formally stating relationships among 
architectures.   Steve is the expert, but dynamism 
in such linking (sync engines, really) is why I 
became part of the HyTime effort very early on. 

See http://www.infoloom.com/gcaconfs/WEB/seattle96/prog.HTM

So, as to establishing creds, can we drop that part 
of the discussion?  You are a consultant I guess, 
need to establish a consortium or some such, and the 
more ideas you can put your name on, the better for funding.
But life is too short and the work I did in this 
domain, I did almost thirteen years ago before there 
was a WWW.  We just wanted to streamline and hook 
businesses together digitally, so the communications 
process became a vital issue.  I am not trying to 
be insulting; just having to write fast in the middle 
of a work day.  I have a music video to film this weekend 
so time is a limiter for me.

Steve wrote that bio, and I'll live with it.  At that 
time, (late eighties) I was working for GE and asked 
to consider where we would be relative to global enterprises 
in ten years.  So I coined a term for it:  enterprise 
engineering and said that we couldn't solve the problems 
of the human (the human WANTS to miscommunicate) but 
that the use of markup applied to very large distributed 
hypermedia systems could help detect noise in the limited 
domains needed.  We could engineer out the worst parts 
and human intelligence would do the rest.  As Shannon 
said, don't worry about the semantics until you can 
ensure the reproducibility of the messages at the 
endpoints.

I think history proves that one to be spot on.  

However, and as pointed out in those 
papers, the closer a system attempts to measure 
and interpret real time events, the more the latency 
(both semantic and physical) begins to mire the system 
in itself.  That is why the notions of stable cooperating 
systems were pointed out in those papers and a good 
deal of effort was spend discussing chaotic systems 
(a relatively new concept at that time).  I pointed 
out that performing musicians deal with this problem 
every day and solve it to a satisfying degree.   They 
do use a notation of time that is elastic and that 
enables them to cope with some of the communication 
event problems.  They also use multiple systems 
of gesture and signal to keep the communications 
coherent across the extent of the performance even 
with mistakes in some local parts (redundancy).  
Ringo Starr was asked how he kept time with 
thousands of women screaming in the days of no 
foldback systems.  He said simply, "I watched 
John and Paul's butts."  Not that difficult if 
you can find a clear channel. 

But NOTHING solves the human in the loop problem. A guy who 
deliberately takes a solo over the bridge while 
the singer is peaking, is just an asshole and has 
to be fired.  That is the reality of any negotiated 
process of communication; it requires the right 
for any participant to walk out or be blackballed. 
Sad but so.

I don't think we are in conflict here.  I consider it 
an implementation issue because any dynamic system 
that attempts to deal with real-time association 
and interpretation has to live with the overhead of 
the system itself and the natures of the thing 
being systematized.  Otherwise, there is nothing 
particularly daunting about the tasks except the 
politics of getting people to agree to use the 
system or abide by any results it produces.   

The application of computer systems to human 
systems analysis is self-limiting.  The tool 
limits, the network limits, the language model 
limits.  I suspect the third of these is what 
is of the biggest concern in this thread.  Yet, 
the real limit is the will of the humans.  
Radar guns breed radar detectors.  Processes 
are not emergent; processes engage and 
controls emerge.

But at some point, pick a system and run 
with it, then measure results and try to 
do better.  It is usually better to start 
with a simple system. In markup, gencoding wins 
everytime.  AFs are mechanism to cope with 
that reality.

len

-----Original Message-----
From: psp [mailto:beadmaster@ontologystream.com]
Sent: Friday, February 01, 2002 12:14 PM
To: Bullard, Claude L (Len); Topicmaps-Comment; Thomas B. Passin
Cc: Douglas Weidner; Tim Barber; Dorothy Denning; Doug Dearie; Dr.
Robert Brammer; Rita Colwell; James L. Olds; eventChemistry;
Humanmarkup-Comment; Katarina Auer; Paul Zavidniak; William Sander;
Dennis Wisnosky; Albright; Ivan Prueitt; Pharris(Contr-Ito); George
Lakoff
Subject: RE: [humanmarkup-comment] RE: [topicmaps-comment] multilingual
thesaurus - language, scope, and topic naming constraint


Len said:

"The interpreter rules.
HyTime AFs serve as a means to make an explicit map among
enabling architectures.

These are not heady concepts.  They are implementation
constraints."

***

<Paul Prueitt>

<header> Sorry for the cc list participation in this discussion, feel free
to use the delete button, but Len's view point is the point that I am trying
to make.  The discussion will die down in a day or so...  So just hit
delete.  Live with it.. *s

Perhaps it is important for a policy discussion to occur on this.
Particularly those in the government in this cc list, I feel that you have
the responsibility to engage in this discussion.  You may disagree, but you
are the civil servant, not I. </header>

***

<Paul Prueitt>

With all respects, the notion that situated scope IS an implementation
constraint, that can be engineered in advance, is simply the current
paradigm that blocks the world from making progress towards true knowledge
technologies.

This is my claim.

Len's comment MIGHT be an example of how a paradigmatic block is
instantiated ?  Perhaps I misunderstand, Len?