the liver and the brain
rsn_ at _comcast.net
Sat Sep 4 07:38:51 EST 2004
On 4 Sep 2004 03:31:22 -0700, rscanlon at nycap.rr.com (ray scanlon)
>In my humble opinion, when Hubel and Wiesel first stumbled on the
>truth that the cortical cell they had probed responded to a line
>segment rather than a spot, a great divide in human thought occurred.
>(Actually, that first cell was a complex cell that responded to a
>moving line segment.) This was not shared by all. There is a letter in
>which Marr referred to H&W's "silly little spots of light". Marr
>wanted the cell to respond to a computation of the entire visual
>scene, or at least a good portion of the scene.
I believe the real "eye-opener" (irony intensional) was the earlier
work of Lettvin, Maturana, McCullough and Pitts in "What the Frog's
Eye Tells the Frog's Brain" (Proc. of the I. R. E. 47 (11): 1940-
1951, 1959. ). That, plus the paper in Sensory Communication (Two
remarks on the visual system of the frog. Sensory Communication. Ed.
W. A. Rosemblith. pp. 757-776, 1961. ) demonstrated for the first time
(I believe) that quite peripheral sensory cells can do a lot of
processing and that it was no longer sufficient to stimulate with
simple spots of light turning on and off. The description of the
cells in the retina and, especially in the tectum, were done in pure
behavioral and teleological terms. There really are very specific
"bug detectors" in the frog CNS!
There were precursors from invertebrate systems, like the lateral
inhibition in Limulus eye (Hartline, 1956). But these were
invertebrates, not "real" animals. The Hubel and Wiesel discovery of
complex and hypercomplex cells extended that type of information to
mammals. No longer could we view the visual system as simple
television cameras, projecting an accurate picture of the world onto
the visual cortex for the :"little man in the brain" to view.
You are absolutely right. That decade, mid 50's to mid 60's
represented a major shift in thinking about sensory information
processing. It was not initially shared by all, as is always the case
in "paradigm shifts". But it didn't take long for the wave of change
to take over. Your quote from Marr represents the absolute opposite
extreme. H&W shows that "silly spots of light were, in fact,
inadequate stimuli. Marr's notion was similar to those who suggested
that, to study the auditory system, instead of playing clicks and
tones to the subject, we had to play a Beethoven symphony. This led
to the short-lived idea of the "grandmother cell", the supposed cell
somewhere deep in the visual processing region of the brain that
supposedly fires only when you see your grandmother which had some
play (though mostly to deride the notion) in the late 60's.
Incidentally, these directionally sensitive detectors truly
demonstrate the nature/nurture interaction. In cats, at least, the
ability to make form the appropriate circuits is clearly part of the
genetic program. However the actual construction of the circuits
requires there to be the appropriate stimulus. You say "the genetic
program constructs a brain that is plastic". That is the essential
point. In that sense, there is no fight between genetic and learned
(at least in humans) -- both parts are essential and interdependent as
part of the whole organism.
More information about the Neur-sci