[From Rick Marken (950522.0820)]
Hans Blom (950522) --
Bill, we made a serious error in the interpretation of my demo.
Unmodeled dynamics ARE controlled away by a model-based con-
troller! I have to set some things straight.
Hans Blom (950522b) --
My demo includes a noise term (vt) that corrupts the perception only, NOT
what happens in the world (that is modelled by the term ft). Set vt to a non-
zero value, and see what is controlled: the world (xt), NOT the perception
(y).
These are VERY important comments from two VERY important posts from Hans.
I'll have to take a close look at them before doing anything else with PCT
(like reviewing "Action Science") because,if Hans is correct about what he
claims to have demonstrated about control, then PCT is wrong because one of
its two fundamental tenets is wrong.
The two fundamental tenets of PCT are:
1) purposeful behavior IS control and
2) control is the maintenance of perceptual variables in reference states.
Both tenets are neatly combined in the title of the basic text on PCT:
"Behavior: The control of perception".
Hans claims to have designed a control system that controls a physical
variable (xt), NOT the perceptual representation of that variable (y). This
demonstation, if valid, would show that the second tenet of PCT (that
perception is controlled) is not true. Hans' demo purports to show that
control can be the maintenance of _physical_ (not perceptual) variables in
reference states.
If Hans' demo is valid, then the PCT model of purposeful behavior is false;
behavior may be a process of control, but it is not control of perception.
The basic book on PCT (according to Hans) should be called "Behavior: The
control of external reality" and PCT should stand for Physical Control
Theory.
I'll have to wait until this evening to try out Hans' demos; but if Hans is
right, I'm going to have to go through a MAJOR system level reorganization --
OUCH!!
Those were GREAT posts, Hans. They really get to the nitty-gritty of the
difference between PCT and all other views of behavior. If your demos are
valid, you will have done what I have found it impossible to do for 15 years;
disprove PCT. It's gonna hurt if you are right -- but facts is facts.
Good luck
Rick