'Tis a gift to be simple

[From Rick Marken (930116.1000)]

Avery Andrews (920116.1233)

One aspect of feedback & sequential order that strikes me is this: there
seems to be a bit of a tendency to construe the issue in either-or terms
(central pattern generation or feedback-driven response chaining), but
this seems to me to be completely unmotivated.

It is motivated (in PCT) by the requirement that there be something
IN THE MODEL that can "switch" from central pattern generation to
feedback control and back. Parsimony seems to demand that the same
model handle all phenomena -- until observation forces a change in
the model. But PCT can handle the phenomena that seem to require a
central pattern generator -- whether or not PCT handles it correctly
must be determined by research.

In the case of a "fast trill", PCT would say that the pianist is
setting a reference for perception of trill at a level which
demands very fast changes in the lower level perceptions (of finger
position) that produce this perception. What is happening is that the
references for the lower level systems are changing faster than
the rate at which those systems can control the perceptions demanded
by these references; this causes no problems as long as there are
no disturbances to the lower order perceptions; but, if you disturb the
lower level perceptions (like by making the force required to press a key
suddenly change) the lower level systems cannot operate fast enough
to compensate for the disturbance; so it looks like the lower level
system is going "open loop" because it is not "controlling"
position. But there is still feedback -- closed loops --you just have a
higher level system that is setting a reference for the lower level
(finger movement) system that forces that lower level system to
operate beyond its ability to control. This, by the way, is really what
happens in the hierarchical control model. You can test whether
this is what is going on in the subject by introducing disturbances
of different frequencies to the the lower level perceptions (assuming
that you know what they are). You should see evidence of SOME degree of
control -- especially with the lower frequency distburnaces.

One respect in which the arm demo is an unfair critique of inverse
kinematics is that real arms have a lot more degrees of freedom that
the current version of the simulated arm, which makes it a lot less
obvious that simple-minded feedback schemes will work.

The only part of this I agree with is that we (PCTers) will never
be able to build ANYTHING that would convince anyone of the virtues
of PCT who isn't willing to be convinced. In fact, I think the
arm demo (and my spreadsheet model, which simultaneously controls
even more degrees of freedom of more different types than does the
arm -- though in an enormously simpler environment) makes it clear that
the only hope for building robots that control many degrees of freedom
in complex environments is a "simple minded" feedback scheme.

Actually, that reference to "simple minded" reminds me of one of
the main reasons that PCT is not taken seriously in many of the
communities where it could be so helpful -- it is (at heart) very
simple (making it all the more interesting that so few people
understand it). I think there is an assumption in robotics, artificial
intelligence, motor control,and many other fields that if an idea is
not blindingly complex -- at the very edge of comprehension by the
layman -- it couldn't possibly by useful. My feelings are just the
opposite -- if an idea is not simple, it can't be very useful.

I think the goal of science should be to find the simplicity at the heart
of all the complexity we see. Of course, the heart of all the complexity
might itself be complex -- but I have yet to be convinced that that is
the case. Newton made a big impression on me.