Re: Architecture For Inquiry

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

Topic :> Architecture For Inquiry :> Discussion 1

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

Jack,

Enough about football ...

And please excuse a few subject line experiments,
as I seem to have exhausted my supply of acronyms,
and find myself forced to restructure my thoughts
according to -- what shall I call them -- topics?

Also, I am just coming off a bad cold, so I hope
what I say here isn't just the Tylenol^TM talking.

But you have gone and reminded me of some more stuff that
I was thinking a lot about in the early 80's and then again
in the early 90's, so maybe it's one of those sunspot things.

I was thinking of Grossberg's competition/cooperation models.
It's a very cold memory, so I will leave it at that for now,
and hope that I remember more tomorrow.

I will say that I have always been wildly ambivalent about
so-called "connectionist" models, partly because the term
itself is so ambiguous.  On the one hand I like the idea
that "it's in the connections".  On the other hand, the
full strength of that idea is hardly ever explored in
the models that one commonly sees, which are limited
mostly to linear transforms, algebraically speaking,
and 2-adic junctures, graph-theoretically speaking,
and you know that just won't cut it, or splice it,
from a Peircean point of view.

Enough for now.  Hope my head is clearer tomorrow.

Jon Awbrey

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

> I am reposting here a stream of messages circulated between Jon Awbrey 
> and myself. Jon thinks, and I agree, that a larger audience may look at 
> what we are looking at and see things we haven't seen. That would be a 
> useful outcome. Herewith, reorganized in sequence, are some bits of the 
> conversation. In theory, our intent is to continue the discussion in 
> this forum, unless, of course, we get booed off stage.
> 
> 2004 is shaping up quite nicely!
> Jack
> 
> <jp>
> There are folks creating a "language" for the description of topic maps.
> I think they are on a track that could represent a contribution to
> architectures of inquiry and have been dogging it, them, and
> implementations of prototypes for quite a while now.
> Their latest effort is found at
> 
> http://www.isotopicmaps.org/tmrm/TMRM-3.10/TMRM-3.10.html
> 
> One of the designers, Patrick Durusau, is now suggesting that Z:
>
> http://spivey.oriel.ox.ac.uk/~mike/zrm/zrm.pdf
>
> would be a good description language in which to do that work.
> </jp>
>
> <ja>
> generally speaking, there are getting to be so many new languages
> every week that i don't think there's much hope keeping up with
> them unless we start finding better ways of teaching them to
> each other, and i am not seeing that happen, even on the
> far horizon.  you find me in the middle of my return to
> english, which has been a struggle for a dozen years
> now, after a decade of thinking in pascal, that is
> not spoke all that much anymore.  as a practical
> matter, it does not look like i will get either
> the time or energy to learn another artifact
> of that sort.  i will probably stick to the
> basics of logic and math language, though
> of a peircean bent, from here on out.
> of course, there are few innovations
> of syntax that would facilitate
> matters greatly, but we can't
> let the course of inquiry be
> impeded by the lack of
> a perfect language.
> </ja>
>
> <jp>
> I hear you.
> Nevertheless, I am profoundly locked in a dance with destiny which is 
> guided by an underlying thesis that there must be a way to program my 
> computer such that it will augment my own intellect. To do that, I am 
> stuck examining all the artifacts I can find, looking, always, for the 
> "perfect" match to the task I have posed. Along the way, I have discovered
>  inference engines which reason both forwards and backwards
>  Eurisko-like evolutionary programming methodologies
>  qualitative reasoning
>  dialog mapping and IBIS structures
>  neural net (analog and digital) representations
>  semantic net representations
>  category theory, Rosen, Raschevsky, etc
>  Peirce
>
> Somewhere in there lies the seeds of an architectural implementation, 
> one most likely guided by the diagrams you are now drawing, but 
> somewhere in there, motivated by desires sparked more by Rosen than by 
> anyone else; I fully subscribe to Raschevsky's notion that we have yet 
> to invent a representation scheme which allows us to discover and 
> understand that which constitutes life iself.  Along the way, I have 
> discovered quantum entanglement, and I have posited that *it* is the 
> *thing* we disentangle when we tease open living cells, whereupon they 
> die and from which we are completely unable to reconstruct them.
> 
> I am to understand that there is a group of people, perhaps driven by 
> Chris Langton (not sure on that) who think they are on the verge of 
> actually constructing a living cell from scratch. If they succeed, then, 
> so much for my quantum entanglement thesis. I'm personally betting that 
> they will not succeed, simply on the strength of words by Raschevsky and 
> Rosen.
> 
> At the same time, I am most interested in teasing structures of 
> understanding, grist for your architecture, out of the spoken and 
> written word.  My friend Patrick Durusau sent this URL along this morning:
>
> http://www.palgrave-journals.com/cgi-taf/DynaPage.taf?file=/ivs/journal/v2/n1/full/9500037a.html 
> 
> Food for thought in there.
> 
> Sure, we can analyze hell out of every word down to Peirces triads. Sure 
> we can construct commutative graphs ala Rosen. But, in the end, where do 
> you go with those?  One of the grand problems I had with all the 
> discussions going on in the Mikulecky letters [ 
> http://www.vcu.edu/complex/ ] was that nobody was able to take Rosen's 
> 10C-6 and "fill in the blanks."
> [10C-6 is the commutative diagram that is Rosen's "canonical organism", 
> a metabolism-repair diagram which, when drawn, elicits reproduction for 
> free, found in his book _Life Itself_] Nobody seems to know what to do 
> with a basic MR diagram. Does anybody know what to do with Peirce's triads?
> 
> I think we are a loooooong way from wherever it is we want to be.
> </jp>

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o
http://www.cs.bsu.edu/homepages/mighty/history.html
o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o