[From Rick Marken (921111.2100)]
In the interests of trying to satisfy Chuck Tucker's desire
to see the feedforward issue resolved, let me propose that
we try to agree to the following Nominal Agreement on
Feedforward Through Argument (NAFTA). NAFTA is my approach
to describing what I think we should now be able to agree on
regarding feedforward. These resolutions are based on my
memory of the feedforward dicussion. I have no interest in
trying to quote chapter and verse. I just want to understand
the possible place of feedforward in our understanding of
human nature.
1. Feedforward refers to efferent connections from brain to output.
In a control loop, feedforward is a direct connection from a reference
signal to an output function, bypassing the comparator. In an open loop
system (one in which there is no effect of output on input) all efferent
connections are feedforward.
2. Bill Powers has posted two mathematical descriptions of the dubious
benefits of adding a feedforward connection to an "ordinary" control
loop. I used the spreadsheet simulation to compare the behavior of the
standard control model to a control model with feedforward; the model
with feedforward did a teensey bit better than the feedback-only model
when there were no disturbances. Otherwise, it was a wash. Feedforward
might improve the performance of a standard control loop -- but not by
much, and probably not at all.
3. A feedforward connection is certainly a possible augmentation of the
pure feedback model. But at the present time, given the data that is
available, feedforward mechanisms do not improve the ability of the
feedback model to act like a living control system. So feedforward
does not improve our ability to simulate control behavior.
4. Loss of sensory input during behavior places the system in an open
loop situation with respect to the sensed variable (if what is being
controlled includes that variable). This fact, in and of itself, does
NOT require the addition of a feedforward component to the feedback model.
Continued behavior during loss of some subset of sensory inputs is
handled in several ways by the existing HPCT model:
a) existence of other controlled lower level perceptions assures
some continuity over brief loss of other variables.
b) reorganization to use remaining sensory variables to control
higher level variables
c) no significant deterioration of control expected for a long
time if sensory loss affects only higher level variable.
d) output integration maintains variables near last level prior
to sensory loss.
Again, with no data regarding behavior during intermittant sensory loss
it is hard to know which solution will handle the data. But, since the
feedback control model already behaves nearly exactly like a living system
when there is no sensory loss, it seems like the feedback-only options
should be the ones to try first when we want to account for what happens
when sensory loses occur (this was Tom Bourbon's point in his last post).
4. There is absolutely no evidence at all that the mechanism underlying
control of a sensed variable (like the visual distance between
hand and sheets) changes when the variable is no longer sensed
(the lights go out). Control of that variable is simply lost. If
that variable is consistently brought to a reference state even
though it is no longer sensed, then this MIGHT be the result of
a feedforward process BUT it might also be a SIDE EFFECT of
controlling other, sensed variables.
5. In order to prove that an unsensed variable is being brought to
a reference state by feedforward processes, it must be shown that
a) the variable is NOT controlled
b) the unsensed variable is NOT a side effect of another control
processes.
If BOTH of these conditions are met, then the unsensed variable's value
is probably being produced by feedforward processes.
6. PCT maintains a strong SHOW ME attitude regarding the possible
existence of behaviors that are generated by feedforward processes
ONLY. PCT does not rule out the possibility that feedforward might
be ADDED to feedback to improve performance but the bias would be
to have the feedforward be there ALL THE TIME. But mathematical
analysis (see 2) suggests that the benefits of feedforward to a
control system are small, if they are beneficial at all.
7. Conventional models of living systems view virtually ALL behavior
as SOLELY a result of feedforward processes. PCT shows that this estimate
of the probable prevalence of feedforward based behavior has been
GREATLY EXAGGERTAED. In fact, there is no evidence that ANY behavior
is generated by feedforward processes. Thus, PCT people get a little
touchy when people go WAY out of their way to see the merits of
feedforward models of behavior. The feedforward model (where feedforward
in one form or the other is exclusively responsible for behavior) is
precisely what PCT says is the WRONG model of behavior. Some feedforward
might tune things up or help out with the details but it is unquestionably
NOT of much significance (according to PCT) in terms of understanding
what organisms do.
Now, let's see if this will pass CSGNet.
Best
Rick