rcb1 at LEX.LCCC.EDU (Ron Blue) wrote:
>Assume for that your brain is like a radio. If you remain on the
>same station learning would be restrained. If you change frequency
>channels learning would be enchanced. If you remember which channel
>you were one when you learned something new you could store the information
>under that reference frequency and modulation timing.
>Therefore, to function effectively you would need variation, and precise
>timing. Ron Blue
This is a good metaphor.
I guess I've been a bit too harsh on the _precise timing_ thing. I
should have been more clear. Indeed, precise timing may be a real
basis for coding info in the brain. Its just that I have found that
precisely timed patterns do not show up consistently over trials in a
behavioral task. Nevertheless, this is really not evidence against
precise timing in neurotransmission. Rather, the complexity of inputs
to a neuron may give rise to many precisely timed patterns and these
patterns may be the basis of the "richness" that allows for
flexibility in the CNS.
Moreover, I found that, for neurons in the rat basal ganglia and
cerebral cortex, there are trials, in a reaction-time task, on which
the neurons fail to fire at all. This occurs for cells that fire
phasically even at high rates (from less than 5 to nearly 100 Hz)!
These same sorts of failures occur if I 'build' spike trains from
Poisson processes. Maybe these failures are a basic limitation for
information transmission in the CNS. Could it be that info trans is
limited by the poissonian nature of the generators underlying spike
activity?
(I doubt that I am the first to say of such a thing. Does anyone know
of refs for previous statements like mine above from the
neurophysiological or computational literature?)
-----------------------------------
Mark Laubach
Dept. of Physiology & Pharmacology
Bowman Gray School of Medicine
Wake Forest University
Winston-Salem, NC 27157
laubach at biogfx.neuro.wfu.edu
-----------------------------------
