[comp.ai.neural-nets] subnet specialization

silber@sbphy.ucsb.edu (03/04/89)

A recent contributor suggests that the human brain is a homogeneous
net and that if any region with a specialization is damaged, the
function can be recovered elsewhere.  I am no neurobiologist, but
it seems to me that i've encountered repeated references to the fact
that some brain regions' functions cannot simply be assumed by other
subnets after damage

mmm@cup.portal.com (Mark Robert Thorson) (03/07/89)

The human brain is not at all a homogeneous organ.  It is only homogeneous
at the local level.  The brain is like hundreds or perhaps thousands of
separate organs.

The largest homogeneous structure is the cerebellum, where a small number
of neurons form a group which is repeated across the whole organ.

The transition between one organ of the cerebral cortex and another is
easily visible in a fresh brain.  The number of layers of cells, and the
thickness of the layers, changes from one organ to another.

A microphotograph of the transition between area 17 (visual area 1, at the 
extreme rear of the brain) and area 18 (visual 2, just forward of visual 1)
can be seen on p. 214 of MODELS OF THE NERVOUS SYSTEM by Sid Deutsch.
This book is out of print, it's from 1967.  Nice book, nevertheless.