Taking That Last Step

[From Fred Nickols (960907.1030 EST)]

Replying to [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? >>

I'll take a shot at answering that . . . but I'll probably come at it in
a roundabout way . . .

As a fire control technician in the Navy (gunfire, missiles, etc.), I
learned a lot about servomechanisms, amplifiers, positive and
negative feedback, and related concepts early on. Those were
"hard" systems. (And that was a long, long time ago.)

Later on in my Navy career, I was trained as a trainer and then
as an organization development (OD) specialist. Again, similar
concepts showed up, this time in the context of open or "soft"
systems. (Frankly, the first thing I noticed about "soft" systems
was the lack of good schematics).

The big difference, for my money, isn't between "hard" or "soft"
systems, but is instead a distinction between "natural" and "
contrived" systems. A human being (or Simon's ant) is a natural
system. An organization, a gunfire control system (including all
its amplifiers and servomechanisms), or a modern data processing
system are all examples of contrived systems. Their reference
conditions have been placed or programmed into them. In slightly
different terms, these are not systems that establish their own
reference conditions.

Herbert Simon is a computer whiz. He is probably a genius, too.
He has certainly provided me with a lot of good reading over the
years. I recently learned that his father was a servomechanisms
engineer. I haven't read all of Simon's work, but I've read a lot of it,
and I don't think Herbert Simon's inability to take that last step, as
Bill put it, is intellectual -- I think it's philosophical -- and political.

To take that last step is to recognize that human beings are
autonomous entities, that their reference conditions are self-
generated, not imposed from without. This, in turn, means that,
in the last analysis, the concept of social control reduces to what
it has always rested on: force, coercion, manipulation, misinformation,
propaganda, and the like. For the scientific world to adopt, en masse,
an autonomous, self-governing view of human beings and their behavior
would put them at odds with the politicians and the generals. I don't
think scientists in general are ready to pick that fight.

Herbert Simon has spent a good portion of his lifetime wrestling
with the problem of control in organizations and they, of course,
are an interesting, dynamic mix of natural and contrived systems
(e.g., people, and processes). However, for the most part, people
have always been components in this framework, not actors. If
Herbert Simon were to adopt the PCT view, he'd have to say to his
organizational and governmental clientele, "Well, you can control this,
and that, but you can't control the people." That would put him right
in sync with Peter Drucker who has been saying for the last 25 years
that we have to focus on controlling the work, not the worker.

Aw, the heck with it. This is getting way too long . . . I don't
think they're going to take it on their own; they'll have to be
nudged.

Regards,

Fred Nickols
nickols@aol.com

[From Bruce Abbott (960907.1145 EST)]

Fred Nickols (960907.1030 EST) --

Thanks for the interesting presentation, both of your own background and of
your thoughts on Herbert Simon.

The big difference, for my money, isn't between "hard" or "soft"
systems, but is instead a distinction between "natural" and "
contrived" systems. A human being (or Simon's ant) is a natural
system. An organization, a gunfire control system (including all
its amplifiers and servomechanisms), or a modern data processing
system are all examples of contrived systems. Their reference
conditions have been placed or programmed into them. In slightly
different terms, these are not systems that establish their own
reference conditions.

Interesting. In _The sciences of the artificial_, Simon makes a case for a
different, though similar-sounding cut: between the natural and the
artificial. Simon defines these terms a little differently than commonly
understood. An artifact is something whose properties were selected so as
to serve some purpose; in this sense an old-growth forrest is natural but a
tree-farm is not, although both are composed of trees. For artificial
systems, it makes sense, then, to ask what is the purpose of the system.
Biological systems whose characteristics have been selected over the course
of evolution to provide specific functions are artificial in this sense,
whereas a stone found lying on a stream bed is not. Thus from Simon's point
of view a human being, the ant, an organization, or a modern data processing
system are all examples of artifacts.

Such systems have an internal organization suited to carrying out their
functions and an outer "environment" in which those functions must be
carried out. The normal behavior of the system tells us more about the
environment than about the inner system whose mechanism has been adapted so
as to carry out its function(s) in that environment. Indeed it is a
characteristic of systems so defined (by their functions) that a variety of
inner mechanisms can be designed that carry out the same functions.

Regards,

Bruce