Fixed Action Patterns vs. Fixed Motor Patterns

[From Bill Powers (970101.0500 MST)]

Gary Cziko 970101.0530 GMT --

Here's an interesting passage from a book I've been looking through.

==================================================

"Neuroscientists distinguish between a FAP (fixed action pattern), the
stereotypical production of some particular motor _result_--a limb
movement, say--and an FMP (fixed motor pattern), the electrical activity
in the motor neurons that brings about that result, FMPs are comparatively
rare.

This sort of thing has been noted throughout this century. It was either
Charles Sherrington or Hughlings Jackson who said that the cortex "thinks in
movements, not muscles." In my (very old) Ranson and Clark neuroanatomy
text, there is this:

"Stimulation of a given point of the motor zone [of the cortex] controlling
foreleg movements, for example, caused the contralateral foreleg to assume a
'final position,' making whatever preliminary adjustments that were
necessary to attain it. Depending on the position prior to stimulation, the
limb might begin its approach to the final position with protraction or
retraction, movements exactly opposite in direction." (p. 304)

Sherrington called this common effect "the instability of the motor point."
I take it as a possible indication that the stimulation was exciting a nerve
cell that normally carried a position or joint-angle reference signal.

So even a "fixed action pattern" is not fixed with relation to stimulation
of the brain. It all depends, of course, on what level in the hierarchy the
investigations happened to have lighted upon. Not knowing about levels of
organization, and not thinking in terms of perceptual signals, error
signals, and output signals of a control system, these early researchers
(and their descendants, apparently) didn't know what they were looking at.

Effects like these keep being discovered, forgotten, rediscovered, and
forgotten again, because there has been no underlying model to make sense of
them. In the first years of my work on control theory I enthusiastically
planned to inhabit the library of the hospital where I worked, going through
all the neurology literature there and finding out how the brain is
organized, so I could look for control systems. The reality was a bitter
disappointment. There were lots of fragments like the above, but they were
scattered and disorganized; nothing was followed through. The general
impression was sort of like reading one of those Ripley's Believe It Or Not
books -- a collection of fascinating vignettes having nothing to do with
each other.

Cricket songs come close; the neural activity involved in the
production of these songs is tightly controlled in regard to the number >and
the {132.1} timing of motor impulses (Ewing and Hoyle 1965). Hoyle (1984,
p. 405) describes an even more impressive case: The courtship behavior of
a tiny male grasshopper produces stereotypic motor output accurate to
within a millisecond of the firing of individual nerve impulses.

Isn't this what you'd expect when looking at the output function of a
low-level, perhaps first-order, control system in a simple CNS? The real
question isn't the relation of the observed result to the driving signals,
but where the driving signals come from. Are they error signals? Outputs of
higher systems? Spontaneous? Research done without any theoretical questions
in mind is essentially random -- useless. Just when you get to something
interesting, the researcher gets bored and goes on to some other topic. I
call this "gee-whiz" research. "Gee whiz," says the researcher, "will you
look at that! I wonder what else I can find that's even more fascinating."
Its sort of like channel-surfing on satellite TV. I found a lot of that in
neurology texts.

ยทยทยท

-----------------------------------
Gary Cziko (970101.0313) --

RE: variable mutation rates:

Dawkins suggests that this could be the case for asexually reproducing
organisms but not for sexually reproducing ones. He argues that for
sexual organisms specific genes could not "learn" to increase their
mutation rate in response to certain environmental conditions since sexual
reshuffling would separate the mutator gene from the structural one. But
I don't think his argument holds against a general increase in mutation
rate throughout the genome of the type I think Bill is arguing for. See
Dawkins pp. 82 ff.

Why should there be a "mutator gene?" If you think of all the genes as a
_system_, it would be the organization of the whole system that produces
these and other effects, just as it is the organization of the whole brain,
not a single neuron, that produces behavioral effects. I think that attempts
to find one-to-one correspondences between genes and large organizational
effects are terribly naive. It's like removing a transistor from a boom-box,
and when the music stops announcing that you've found the music transistor.

After all, we know that genes are just an organism's way of making another
organism, don't we?

Best,

Bill P.