Simon and _The sciences of the artificial_

[From Bruce Abbott (960906.2045 EST)]

Rick Marken (960906.1540) --

In an earlier post you said [Bruce Abbott (960906.1255 EST)]:

What he [Simon] says above not only sounds good, it IS good: it's based NOT
on a version of the old input-output view of behavior, but on the view that
the ant is a servomechanism.

Now one could get the impression from your statement here that you are saying
that Simon does not view behavior in terms of an input-output model but,
rather, in terms of a servomechanism model that (one might imagine) is more
consistent with PCT. I think the point Francisco was making with his quotes
from Vera & Simon (1993) is that, if that _were_ what you were saying, then
you _would_ have been speaking hogwash;-)

If someone _were_ to draw that inference, they'd be jumping to conclusions.

What I illustrated in my quoted material was that Simon, writing in 1969,
clearly recognized that the behavior of the ant was purposeful, goal
directed, could be explained via a simple mechanism (which turns out to be a
servomechanism, although he does not state that in the portion of his
writing I quoted), and that the apparent complexity of its behavior was a
consequence, not of an internal complexity in the ant, but of the
environment in which it found itself.

Whatever else Simon may have said before or since does not change that fact,
leading me to wonder what relevance Francisco's quoted material has to the
information I presented. I don't think it has any whatsoever. I never
claimed that Simon fully understood the ramifications of his insight (e.g.,
behavior as the control of perception, etc.), nor did I claim that he
pursues this view today. But at the time Simon wrote the material I quoted,
he was on the right track and was able to see one of the important
implications: that such complex behavior is a consequence of a complex
environment, and that the mechanism beneath that complex behavior may be
wonderously simple despite all appearances. Moreover, the mechanism he
suggested might be at work in the ant was the good ol' closed-loop negative
feedback control system. Not a bad guess, I'd say.

Bill Powers (960906.1500 MDT) --

Interesting discussion of Herbert Simon. If you read his description, he's
talking about feedback control, but if you read his explanation, it's
stimulus-response. The critical part comes when he says that this is a
goal-directed process -- and then explains the goal as being a motor program
produced out of perceptions of the environment.

What is the intellectual barrier that prevents so many scientists from
taking that last step?

Good question. Reading Simon's _The sciences of the artificial_, I keep
wanting to say "and? and?". Simon is so close you want to lead him by the
hand; you keep looking for the punch-line, but it never comes. It's the
same feeling I get when reading certain passages from Aristotle.

Sill, in this wonderful little (118 page) book (based on his Karl Taylor
Compton lectures at MIT in the spring of 1968), Simon offers a number of
ideas which are highly congenial to the PCTers way of thinking (and some, it
must be noted, which are not, but let's focus on the positive rather than on
the Markenian negative). In many ways reading this book prepared me for
what Bill Powers had to say when I finally encountered B:CP, and I would
recommend it even today, especially the first 25 pages or so and some stuff
about evolution later on.

If anyone is curious about what these congenial ideas are, I'd be happy to
describe them (and provide some quotes from Simon), but at the moment I've
got supper waiting at home.