Of course the brain encodes sensory input, but encoding is too fine a
word for the process that takes place in the nervous system. Encoding
implies the hiding of the meaning expressed in a natural language, for
later decoding and recovery of the meaning. Sorting should possibly be
a better word. The sensory input is sorted through a structure
established by the genome. It is sorted, but we are stuck with the
word "encoding".
The encoding is not in English, and it is not to produce a picture in
Descartes' Theater. The input (olfactory, visual, auditory, tactile,
etc.) is all encoded for only one purpose, to trigger a motor program
generator and thus produce a motor program (and eventually a motor
act) appropriate to the situation. Appropriateness is defined by the
genome as evolved in the last four billion years of cellular life.
Evolution is the random mutation of nucleotides in the DNA molecule,
nothing more, nothing less.
"In this new century, understanding the workings of the genome is of
paramount importance in almost any biological discussion." We must
live in this new century; let us take this to heart.
There is no question that visual input is encoded. The first step in
the encoding was worked out in the beautiful work of Hubel and Wiesel.
We are able to understand this first encoding of line segments with
given position and orientation. We are not able to understand the
subsequent encodings, but there is no reason why we should. Men are
deluded by the notion that the encodings should all come together like
a moving picture displayed on a wall. Actually, the cortex is designed
by the genome not to display a picture, but to trigger an appropriate
motor program generator.
Is a neural ensemble deterministic? We answer yes! But this does not
mean that a particular sensory input leads to the same neural pattern,
for all neural experience alters synaptic parameters. You cannot step
into the same river twice. Nevertheless, the action is deterministic.
The stunning results of Hubel and Wiesel showed that when the output
of the retina hits the striate cortex, the pattern is predictable.
Area seventeen is constructed by the genome to respond to line
segments. Line segments with a given position and orientation. Line
segments moving in an orthogonal direction. Line segments closely, but
not identically positioned, in both eyes. We have no reason to believe
that this structure is not repeated throughout the brain, always
selecting for patterns as assembled by the genome until a motor
program generator is triggered.
Can a given pattern between sensory input and motor program be
recorded and examined? The answer should be yes, since we have proven
capable of decrypting enciphered text, but it would be hellishly
difficult, and certainly not worth the effort
Perhaps, I should insist on "sorting", after all.
Ray