[From Dag Forssell (940502 1000)]
I just got this from a NewsNet reader. Nice to hear that someone
reads CSGnet that way, and that Bill Silvert's FTP site is being
used. Mikael, I hope you don't mind my answering on the net. That
way I can share my reply with the net at minimum cost, and you can
get other input.
···
Date: Mon May 02, 1994 5:14 am PST
From: Mikael Nyberg MBX: mikaeln@ifi.uio.no
Subject: PCT-demos
I downloaded some of the files, and after looking at them and
thinking a bit I would like to see more. But I can't remember
where I got them. Is there a single PCT ftp-site with all the
stuff available?
How is PCT-research funded, by the way? Are there grants
administred from a group or is the research physically bound to
a specific place?
I am a graduate student of computer science, and my interest in
PCT is due to the seeming lack of "metamodel"-research, which to
my opinion results in narrow models and weak expressibility in
science. My idea is some sort of modelling theory which is
polyscopic, that is it must allow arbitrarily many models to
coexist on different degrees of "detail", the only real
constraint beeing consistency. And there must be ways of viewing
"from different perspectives" in the global model, and also
creating unions, intersections, closures etc. I feel that my
motivation resembles the motivation I could interpret from
reading the documents. I will for instance have to consider
problems like the quantitative-> qualitative axis, and it seems
that this is a topic in PCT too. I would appreciate any comments.
I try to read the news-group. Thanks,
Mikael Nyberg, University of Oslo, Norway.
First: Yes there is a single FTP site. Look for it in the
CSGINTRO document which Gary Cziko will post momentarily (beginning
of every month).
Second: Some PCT'ers wish we were funded. PCT shows that
psychological academia is barking up the wrong tree, and is
generally useless. PCT is not popular, understood or welcome to
join the funded establishment. All PCT research is a labor of love
and personal conviction.
Third: The Control Systems Group has some tolerance for
...arbitrarily many models to coexist on different degrees of
"detail", the only real constraint beeing consistency.
but PCT is specifically about modeling human beings -- ourselves.
PCT research is seeking the ONE model which explains the design of
the human nervous system. We argue about which design that is, but
are not interested in "artificial intelligence" in general, just
because a design happens to be feasible on a digital computer. We
are interested in the design that explains humans as they can be
shown to function.
Those who have studied PCT in depth (as it is presently laid out)
believe that it has plenty to say to those who seek workable models
for "artificial intelligence" and will welcome focused questions on
PCT. We do appreciate that you become familiar with PCT. The
CSGINTRO document gives all the basic references. It has been
updated this month with addresses and order forms for most of the
literature.
Hope this answers your question.
Best, Dag