chouse at uoguelph.ca (Chris J House) writes:
> Counting
>: synapses is relatively easy -- there are about 10^14 of them in the
>: -- Bill Skaggs
> I'd hesitate that counting all of the synapses in the human brain is a
>relatively or otherwise easy task. The numbers I always see quoted in
>the literature as to the number of synapses in the human brain is
>performed through averaging. I am sure you are not implying that someone
>somewhere has sat down at their microscopes with a billion and more brain
>slices counting synapses.
I thought Bill was being facetious there... :)
> After realizing that these averages are taken from different locations
>in the brain (Cortex vs cerebellum, limbic vs medullary.. etc), we have
>to also take into account that a good deal of our observed memory stems
>from learned tasks and preprogramed motor skills (eg. typing my name on
>the keyboard). Do these skills actual require memory?
Why yes, they do. Not the same kind of memory as is used when you are
given a list of words and have to recall them 5 minutes later, but a form
of memory ("procedural" is one term) nonetheless. Even a lowly reflex
action is a form of memory.
> And do all the
>multitude of synapses that fire in the process of creating this task
>constitute cerebral, cerebellar and spinal synapses within your count.
>Oh, wait.. now, I'm confused. The count always seems to be cortical (or
>at least synapses within the cranium) as an estimate. What about patern
>generators in the spinal cord? What about the multitude of reflexes that
>never make it to the brain? I do not think that an estimate of the
>number of synapses is in any way relative to a level of storage.
I've wondered about these cell and synapse estimates too -- they always
seem to be for cortex, not including the rest of the brain.
In fact, I'd say this is discrimination against the subcortical areas of
the nervous system. Subcortical areas of the brain, unite! (sorry, now
I'm being facetious)
> Finally, the processes of the brain are not just storage and memory per se.
No, they're not, but in neural networks, processing and storage occur in
the same computational elements.
Kevin
-----------------------------------------------------------
Kevin Spencer
Cognitive Psychophysiology Laboratory and Beckman Institute
University of Illinois at Urbana-Champaign
kspencer at s.psych.uiuc.edu
-----------------------------------------------------------
