From Greg Williams (920326)
Rick, in your reply to Martin which he posted, your logic is faultless. No
doubt, if someone thinks that "open-loop action is control," then they are
SIMPLY WRONG, PROVIDED that they accept your DEFINITIONS that "closed-loop
action (only) equals behavior" and "behavior equals control," which reduce to
"closed-loop action (only) equals control."
But science is supposed to be more than a matter of definitions. I thought the
argument was about whether some organisms use open-loop, pre-calibrated
behavioral organizations to achieve what look to observers as adaptations.
Contra Bill's latest post (I think), I think the argument was about how
organisms are organized INTERNALLY. For example, does Horst Mittelstaedt's
preying mantis pre-calibrate its trajectory before lunging and then follow
that "aiming" during the lunge, or does it adjust its trajectory as it lunges
on the basis of ongoing perceptual inputs? I'm a bit lost by the newest turns
(?) in the debate, seemingly toward tautological exclusion of some positions
and/or bringing in EXTERNAL (environmental) organization. Maybe I'm just slow
to catch on, but I'm still wondering why a person couldn't, when reaching, use
BOTH pre-calibration (to get started) AND end-point-perceptual-control (near
the goal). After all, when the lights go out, pre-calibration begins to seem
pretty important!
Greg
[From Rick Marken (920326 12:00)
Greg Williams (920326) says:
Contra Bill's latest post (I think), I think the argument was about how
organisms are organized INTERNALLY. For example, does Horst Mittelstaedt's
preying mantis pre-calibrate its trajectory before lunging and then follow
that "aiming" during the lunge, or does it adjust its trajectory as it lunges
on the basis of ongoing perceptual inputs?
What this discussion has made me realize is that the argument is really
NOT about how organism's are organized internally. And I think that this
is an important point -- and a new realization for me. The negative feedback
situation defines the organism as a control system, regarless of how the
organism is internally organized. The internal organization of the organism
just has to transform sensory inputs into outputs that effect the input
in a negative feedback sense. There are probably many organizations that
will do this -- and one of these has closed loop systems manipulating
the reference for the inputs controlled by other closed loop systems
-- that's the HPCT model of the organism, but it is not the only
one (as Bill would readily admit).
There is nothing tautological about the fact that negative feedback
means that the organism is a control system. The existence of negative
feedback connections can be determined by inspection and test. And the
fact that the system is controlling input variables can also be determined
by inspection. There is no tautology because the inspection could reveal
that there is no negative feedback (or controlled input). If there is
negative feeback (because of the closed loop SITUATION) then the
behavior of the organism cannot be correctly described by an equation of
the form o = f(i). It must be described by two simultaneous equations which,
when solved for output (o), show that o = -g(d) -- outputs depends on
disturbance, NOT INPUT. That is not a tautology, it is a fact. And it is
true of ANY organism (no matter what its internal organization) when it
is in a high gain, negative feedback relationship with its environment.
I am arguing that, if you don't know that this is the case when you are
dealing with an organism in this situation (negative feedback) then you
cannot possibly come to the correct conclusion about what its internal
functional organization might be based on observation of its behavior.
Regards
Rick
···
**************************************************************
Richard S. Marken USMail: 10459 Holman Ave
The Aerospace Corporation Los Angeles, CA 90024
Internet:marken@aerospace.aero.org
(310) 336-6214 (day)
(310) 474-0313 (evening)