Siftin' Sand

[Abbott_Bruce_Indiana]

[From Bruce Abbott (941221.1030 EST)]

Rick Marken (941220.1745)

I still have no Turbo for the PC

Does this mean you now HAVE a PC? (:->

I agree that the evolutionary model you posted (assuming that it
works; I still have no Turbo for the PC) operates on the basis of
"selection BY consequences". I would imagine that there will be a
bunch of bugs with "good parameters" running around after a few
selection episodes. But is this a model of purposeful behavior?
The behavior of the bugs themselves is not selected by consequences.
What is selected by consequences?

No, the behavior of the bugs themselves is a product of their two-level
control systems interacting with the nutrient environment. What is being
selected is a control system that works, as opposed to one that fails.

I'm really glad to hear you say that you believe in "selection BY
consequences," because the principle at work here (Darwinian variation and
natural selection) is, according to Bill Powers, the model he used for his
hypothetical reorganization process. Parameters of lower-level systems
(including their interconnections) are varied until a perceptual variable is
brought under control. What is selected is not a behavior but a control
system. Thus I am puzzled by your assertion that

"Selection BY consequences" exists BUT IT DOES NOT PRODUCE CONTROL.
And, to the extent that consequences select the outputs (or the behavior)
of a system that is already organized as a control system (as was your
E. coli model) they INTERFERE with the system's ability to control.

Is this statement not incompatible with Bill's conception that effective
control can be established through such means? In reorganization, do not the
consequences of a set of system parameters (control or failure to control)
determine whether those parameters will be preserved in the system or
abandoned in favor of yet other variations?

Are you saying that this new demo shows that control (purposeful behavior)
can result from selection by consequences?

Yes, of course!

If so, what is the controlled variable?

Stored nutrient level ("fuel").

How do you know that it is under control?

It stays near its reference level despite disturbances (fuel consumption and
variable nutrient density).

I think that throughout this debate, you have been repeatedly describing to me
the operation of intact control systems and I have been repeatedly trying to
get through to you that I am NOT discussing how intact control systems behave
but rather, how such systems are created as a result of experience with the
consequences of a given system's attempts to control some perception. Until
you are able to grasp the difference, we will continue to talk past one
another.

A system that learns from its mistakes (and successes) must include negative
feedback (by definition), but it is not necessarily a perceptual control
system within the definition given by PCT, for it may lack a gain greater than
1.0 and thus fail to resist continuous disturbance, as in the pendulum example
of your previous post. The process of evolution provides one example, with
the environmental realities an organism faces playing the role of gravity.

Behaviorists claim that purposeful behavior (like that seen in operant
conditioning experiments) can be explained as "selection BY consequences".

I agree with you that purposeful behavior is produced by a set of perceptual
control systems rather than by a simple "selection by consequences"
reinforcement mechanism.

The E. coli demos show that this is wrong; flat out, incontrovertibly
wrong.

The e. coli demos show that purposive behavior is best understood in terms of
perceptual control systems. My ECOLI4a demo showed that the parameters of a
successful control system can be established through a process in which the
organism varies its control parameters based on the immediate consequences of
its own behavior. Your demonstration that such learning could easily be
disrupted by changing the behavioral consequences has the same force as
"proving" that PCT doesn't work by demonstrating that human participants fail
to control their e. coli spots when wearing blindfolds.

PCT shows that purposeful behavior MUST be viewed as selection OF
consequences. Control theory shows how organisms must be organized in
order to be able to select and produce the selected consequences of
their actions.

Yes, and I agree. But you are talking about execution, not acquisition.

Why is this so hard?

I dunno--you tell me. Sand in your gears? (:->

Regards,

Bruce

[Tom_BOURBON3]

Tom Bourbon [941222.0919]

[From Bruce Abbott (941221.1030 EST)]

Rick Marken (941220.1745)

Bruce:

I'm really glad to hear you say that you believe in "selection BY
consequences," because the principle at work here (Darwinian variation and
natural selection) is, according to Bill Powers, the model he used for his
hypothetical reorganization process. Parameters of lower-level systems
(including their interconnections) are varied until a perceptual variable is
brought under control. What is selected is not a behavior but a control
system. Thus I am puzzled by your assertion that

"Selection BY consequences" exists BUT IT DOES NOT PRODUCE CONTROL.
And, to the extent that consequences select the outputs (or the behavior)
of a system that is already organized as a control system (as was your
E. coli model) they INTERFERE with the system's ability to control.

Is this statement not incompatible with Bill's conception that effective
control can be established through such means? In reorganization, do not the
consequences of a set of system parameters (control or failure to control)
determine whether those parameters will be preserved in the system or
abandoned in favor of yet other variations?

Rick [From Rick Marken (941221.2300)] has already replied to this, Bruce. I
agree with his reminder to you that the *consequences* do not "select" or
determine anything. The only consequence that matters in reorganization is,

     intrinsic error = intrinsic r - p = 0.

When that condition is not met, reorganization (small random variations in
system parameters) begins. When the condition is met, reorganization
ceases. If the condition is not re-established, the creature dies. (It is
not "selected against" by anything; it simply dies.) Other than the fact
that continued survival depends on the organism "selecting" a satisfactory
configuration for the system parameter(s), there is no "selection" going on
here.

When Bill says he used something like the idea of Darwinian variation and
natural selection as his early model for a reorganization process, there is
more to the story. For one thing, I believe he often says that in order to
continue a dialog in which someone suggests the similarity. But he has
often gone on to say that there is no real "selection" going on in or by the
environment.

The reading of evolution that I have seen Bill use most often when he talks
about PCT and reorganization is desceibed briefly in one of his previously
unpublished papers that is reprinted in _Living Control Systems II_. The
paper, written in 1983, is, "Learning and evolution." In it Bill recounts
how his ideas about reorganization resemble, but go beyond, Ross Ashby's
concept of the ultrastable, or superstable, system. Ashby wrote about a
double negative feedback system, in which the first system interacts
directly with the environment, while the second alters parameters in the
first whenever the relationship between the first system and the
environment goes unstable. Ashby's account has much in common with our
adaptive PCT models, which we have tested as models for control systems that
reorganize. One of the major differences is that, where Ashby discussed the
actions of the second system (the adaptor, the reorganizer) as step
functions, we have modeled its actions as continuously variable.

In "Learning and evolution," Bill once again recounted the similarity
between his own ideas about reorganization, and Donald Campbell's principle
of "blind variation and selective retention" in evolution. Blind variation
and selective retention. Blind variation occurs in the organism and affects
its interaction with the environmant. Selective retention occurs in the
organism: whichever state it is in when its interaction with the
environment becomes "OK" (as determined in the organism) sticks -- it is
retained -- variation stops. To me, "blind variation and selective
retention" has a cleaner feel to it than, "natural selection (by the
environment)." For one thing, to many people, "natural selection" has come
to imply that consequences do the selecting, whether it is in evolution, or
in learning, or in the reorganization of living control systems.

···

==========

The holidays are near. Almost time to go down to the beach, sift a little
Texas sand, and enjoy some broiled flounder while watching the surf.

Later,

Tom