Brain

Terence P. Ma tpm at anat.UMSMED.EDU
Wed Mar 11 16:22:34 EST 1992


Well, Tom, you have asked some really difficult questions.
Unfortunately, we really don't know the answers to most, if not all,
of them.  I will give you a brief answer to some of them ... but
bear in mind that they are incomplete, and subject to discussion and
debate.

In article <4379 at cluster.cs.su.oz.au> tom at minnie.cs.su.OZ.AU (Thomas James Jones) writes:
>	1) Is 'knowledge' thought to be distributed or locally stored.
>		(I would expect distributed, with some local elements)

I assume by "knowledge" you mean memory.  There are at least two
types of memory, short-term memory (which seems to be primarily
mediated through the hippocampus) and long-term memory (for which
there seems to be no specific site).  There are hypotheses that
suggest that some cells in some areas may modify the proteins within
the neuron and they represent storage mechanisms.  Others suggest
that the change in the connections between cells and represent
storage mechanisms.  And there are many more hypotheses.

>	2) How much/What sections of the brain are activated for various tasks.
>	 	(the two 'main' distinct classes of tasks I would describe 
>		as Perception and Reasoning).

Classically, the brain is divided into "sensory" and "motor"
pathways where sensory usually refers to incoming information and
motor to outgoing information.  However, there are so many different
types of combinations between the two that it is difficult, for many
areas of the brain, to rationalize which is which.  Perception
often refers to a specific function within the sensory modality and
I don't know enough about it.  Reasoning, I think, is anybody's
guess at this time (although recently Dr. Bill Newsome at Stanford
has demonstrated that individual cortical cells are able to respond
for presumably cognitive processes).

>	3) For those with some understanding of Artificial Neural Networks (ANNs)
>		- to what extent does that technology mimic _potential_ reality,
>		and for all of you, how would you suggest that our brain be
>		simulated (by a computer).

Computer simulation of brain function is really difficult.  On the
motor side (in which I work), we simulate brain function using two
basic techniques.  First, we develop bioengineering models (based on
systems engineering) and do signal analysis.  Second, we do 3 layer
neural networks, train the networks, and then analyze the hidden
layer.  However, there is no proof, nor any reasonable demonstration
that what happens in the hidden layer actually happens in the
brain.  There are certainly some similarities, but I think the jury
is still out as to what it all means.  Personally, I like to think
of what happens in the hidden layers as a possible mechanism of
function, I am in the process of developing a series of studies
where I take the data about the hidden units, develop mathematical
constructs from them, and then apply them to the classical "boxes"
in systems models in order to do signal analysis.

>	4) What evidence of RECURRENCY (feedback) is there in the
>		activity of the brain with respect to a particular
>		task. Personal view: 
>		tasks of perception are 'feed-forward' , do not require feedback
>		tasks of reasoning do require feedback.
>	[I have expressed this in terms of ANN terminology, I can clarify
>	  if required]

It seems that the brain is highly redundant.  Certainly, the old
anatomical maxim that "everything is connected to everything" is
awful close to the truth.  I don't know a single function of the
brain that does not depend on feedback mechanisms of some sort.  For
example, perception of visual stimuli requires very sophisticated
feedback mechanisms at every level.  Similarly, on the motor side,
lack of feedback leads to run-away systems or dynamically unstable
systems.  My terminology is based on systems modeling and
neuroscience, I never am certain whether they are equivalent to the
mathematician's!

>vague questions:
>	4) What is the _basic_ macroscopic & microscopic view on the brains
>		activity?

Ummm ... macroscopic and microscopic refer to anatomy.  Activity
usually refers to physiology.  So, are you asking "how does the
brain function?" [;-)].  It's a tough one and difficult to answer
without a better understanding of what you want to know.

>	5) What is perception? [in terms of the brain]
>	6) What is reasoning?

I'll leave that to the cognitive and psychophysics types around.  As
a neuroanatomist and neurophysiologist of the motor system, I don't
know as much about perception and cognition as I'd like.

Regards!
-- 
Terence P. Ma, Ph.D.		     VOICE:	601-984-1654
Department of Anatomy		     FAX:	601-984-1655
University of Mississippi Med. Ctr.  INTERNET:	tpm at anat.UMSMED.EDU
Jackson, MS 39216-4505		     UUCP:	... uunet!tpm-sprl!tpm



More information about the Neur-sci mailing list