[Hans Blom, 931104]
(Rick Marken (931103.1200))
I'm sure Bill Powers will "do his dooty" and point out (for the
400,000th time) all the problems with feedforward models of
behavior.
Rick, didn't I point out the disadvantages of feedforward control
sufficiently? Do you think that I was unfair or unrealistic in positing
that there are circumstances where they DO work? I know that you are CSG's
most orthodox feedback afficionado, but does that close your eyes com-
pletely for other approaches that might also (sometimes) work?
Do you keep making the
same walking movements (with legs in the air, face on the ground)
after you trip over your wife's shoes which have been left
unexpectedly in your path?
Now I understand why you permanently have to pay attention to that
unreliable, capricious world around you. You have a WIFE!
Stimulus-response behavior, motor or mental, often takes over
after learning is complete
How long does this takeover last after a person has completely
learned to toss a ball at a target in a highly reliable
environment? How come when Gary Cziko applies a disturbance
to the throwing arm immediately after this learning has taken
place the S-R behavior seems to relinquish its "take over" (and
the throwing arm resists the disturbance)?
How about this addition: feedback controlled behavior, motor or mental,
again takes over as soon as we discover that the stimuli that our senses
receive do not correspond anymore to well-known patterns that we have
developed adequate reaction patterns to. This presupposes a "sensor" or
"recognizer", at some level, that discriminates familiar from unfamiliar,
or maybe danger from safety or some such. There is strong evidence that
such a "device" exists and is well developed already in small babies.
In this sense, you might say that feedforward is for experts and
that feedback is for beginners
Yes, a wind up doll is a real expert, all right. And it can walk
in the dark, too! Imagine that
Aren't we all wind up dolls in a great many respects?
(Bill Powers (931103.1340 MST))
An interesting and thorough discussion of feedforward. You make a
number of points with which nobody can argue
Thank you. That was all I wanted to do: to explain the concept of feed-
forward and some of the differences between feedforward and feedback.
4. Under HPCT, it is not necessary that a perceptual signal be
available to consciousness or awareness in order for it to play a
part in feedback control.
Sure. I do not assume consciousness or awareness in feedforward systems
either. In many respects, feedforward responses can be compared to uncon-
ditionally evoked reflexes that run off once started. This typifies the
behavior of most of us under most circumstances, doesn't it? But note the
advantage: processing resources can be freed for tasks that have NOT been
well-trained (high-level feedBACK processes?).
5. A purely feedforward system can't work, precisely because the
"tuning" of which you speak is necessary.
This is a non sequitur. Once trained, they CAN work. I thought you agreed
to that.
after some period of operation it
is necessary to recalibrate the feedforward system because both
it and its environment will suffer secular drifts in sensitivity
to inputs. So every successful feedforward system must initially
and periodically be part of some larger feedback loop
In-born feedforward systems could have a fixed tuning, dependent upon the
species. Otherwise, retuning or recalibration is equal to learning, which
may or may not be a random process. It is remarkable, by the way, how
robust feedforward may be in the face of parameter changes. A rock thrown
into the air will return to earth -- or some other chunk of matter --
under a great variety of gravities.
6. The range of situations under which feedforward control can
perform usefully is limited.
Yes, I thought I mentioned that.
7. The kinds of variables amenable to feedforward control are
those that naturally change slowly, over seconds or longer, even
in the presence of normal disturbances.
Not necessarily. We want predictability, not constancy. After you have
thrown a football, your idea of where it might land might be sufficiently
accurate to continue the play of the game with other concerns in mind. Of
course, the ball might hit an overflying bird, be picked up by a UFO, or
hit a shoe thrown by a female fan. But those are things that you usually
can disregard.
9. An adaptive compensatory system can't learn to control
"better" unless its outer loop is a feedback control system.
Right. But once it has learned, it might not need feedback anymore, or
only occasionally, or only under special circumstances.
I knew a doctor of the sort you mention. He was knowm to his less
flambouyant colleagues mainly for the inaccuracy of his
diagnoses, resulting primarily from his insistence that his first
impression was nearly always right. He also refused to consider
subsequent evidence against him (including death of the patient
from some undiagnosed cause). Such outcomes were never his fault.
Was this doctor any better when he used feedback control, i.e. in his more
difficult cases?
While, as you say, apparent cases of feedforward abound, a great
many of them are based on superstition and after-the-fact
"prediction."
Skinner talked about "superstitious learning". Now I could easily show you
that in any model of learning that I am aware of, ALL learning is necessa-
rily superstitious in the sense that it takes into account only the evid-
ence at hand -- a more or less reliable correlation between action and
outcome. If pecking at that bar is effective to keep the grain coming, do
it -- even though, unknown to you, the grain comes at random intervals
that cannot be influenced by pecking. It IS a solution, although a sub-
optimal one. But how would YOU act if in all previous experiments your
only method of control was pecking at bars? Finding _optimal_ solutions is
difficult in any walk of life, even for a pigeon.
The kind of control achieved by feedforward is
necessarily very poor except under specially controlled
circumstances
Is this the same as the more positive "The kind of control achieved by
feedforward can be good under some carefully delimited circumstances"?
More than such an admission I do not ask from you.
I can agree with almost everything you say about the nature and
characteristics of feedforward control. I just can't agree that
it accounts for very much of human behavior.
OK, thank you. We both think that feedforward / stimulus-response control
might work. We just differ in our estimate of how much it accounts for
human behavior.
(Bill Powers (931103.1530 MST))
Your use of the term "reference signal" in a completely open-loop
system is somewhat strange to me. Why not just call it the
"input?"
I don't care much for names. "Input" is fine with me, if you insist. I
just wanted to stress the similarity with a reference signal. You might
think of a "virtual" feedback signal provided by a well-tuned model, such
as in the following diagram:
ref -------- action
------>|+ |------> world
--->|- | |
> -------- |
> -------- |
----|model |<---
ยทยทยท
--------
where the model is purely internal; it only "knows" our actions / output,
not their effects on the outside world. In fact, the model PREDICTS those
effects given the actions and thus does something like your "imagination
mode". Am I clear enough or are all those transformations between methods
of drawing the control diagram too confusing? Mathematically / functional-
ly, they can be shown to be equivalent.
So Black considered feedforward first, and found it wanting.
His compensation method DID work, however. His scheme worked because,
although his amplifiers were nonlinear, amplification of a SMALL signal
(the difference) introduces significantly less distortion than amplifi-
cation of a large signal. Black's method is still used once in a while in
cases where the phase characteristics of the system are so funny that no
stable feedback system can be designed.
In no place, Bill, have I wanted to establish the impression that, given
some circumstances, feedback cannot be better than feedforward. What I
wanted to say was just that both feedback and feedforward have their place
in control systems design -- so probably in biological control systems,
organisms, as well.
(Tom Bourbon [931103.1641])
You say that your actions during your close encounter with the metal
obstruction exemplify feedforward and did not entail feedback. Did you
close your eyes (to eliminate visual perceptions), release the wheel and
remove your feet from the pedals and the floor (to eliminate kinesthetic,
proprioceptive, and tactual perceptions), and cover your ears (to
eliminate auditory perceptions)? Do you do similar things to eliminate
all other perceptions when you walk through the darkened room at night?
Tom, you must discriminate those perceptions that are used to provide the
feedback information from all the other old perceptions that we have but
that are not relevant to the task at hand. When I navigate my way through
a darkened room, I have a great many perceptions, but if the floor is even
-- no female shoes lying around -- I have no indication as to where I am,
except maybe from some things other than my immediate perceptions, such as
an internal model.
Greetings,
Hans