IUBio Biosequences .. Software .. Molbio soft .. Network News .. FTP

memory (and SOMs)

yan king yin (dont spam) y.k.y at lycos.com
Tue Mar 12 15:43:20 EST 2002


Sorry I disappeared for a week, was feeling very tired....

"Matt Jones" <jonesmat at physiology.wisc.edu>
> Well, we know a -little- about cortical circuitry (and by "we", i mean
> other people who actually know about cortex, not myself). And there
> are a number of theories about long-term memory in the cortex. Marr
> had the classical work in the early 70's which, although probably
> wrong in a number of specific details, laid the initial conceptual
> threads that still run through current cortical modeling. More
> recently Rolls has proposed
> a theory for the transition from short-term (hippocampus and EC) to
> long-term (cortex) storage.

I'll check out the works of Marr and Rolls, thanks.

I dont know of any studies done on the connectivity of the limbic system
surrounding the hippocampus, I guess it would be difficult to construct
a theory of memory transfer without that piece of information.

I've read somewhere that long-term memory is supposedly stored in
the temporal lobe. There could be specific circuits for that.

> > According to one study, there are about 32 visual areas in the macaque
> > visual cortex, connected by 305 connections. Now it seems that these
> > visual maps cannot be the result of self-organization, so they must be
> > genetically specified.
>
> Wait, why can't they arise through self-organization? Not to step
> right into the middle of the nature-vs-nurture mess again, but there
> are plenty of experimental demonstrations that cortical circuitry
> (e.g., receptive field maps) do reorganize in response to
> environmental changes (e.g., the "classical" but grisly experiments
> involving sewing kittens' eyes shut, and gluing monkeys' fingers
> together, then recording changes in cortical receptive fields). I
> think there's very clear evidence for an initial developmental program
> (i.e., nature) as well as self-organization (e.g., nurture).

Thanks for reminding me of plasticity. There's also the fMRI study that
shows that blind people uses their visual cortex to read Braile. This is
rather paradoxical in that the visual cortex appears to be wired up
specifically for vision. But I dont think they have figured out the details
of such reorganizations yet. Moreover, reorganization does not
neccessarily mean that cortical circuits are not fixed. The real distinction
between a fixed connection pattern and a self-organized pattern is
that the former does not show variations among individuals.

The examples that you quote are reorganizations of receptive fields
within a topographical map, such as V1 (for that cat) and the primary
somatosensory area (for that monkey) -- is that correct? I guess they
are not reorganizations of connections *between* areas, only those
*within* a map. This is an important distinction. Maybe the patterns
within maps are plastic, but the pathways connecting different maps
are not. In developmental neurobiology, one talks of axonal pathfinding
and target invasion as two distinct events governed by different molecular
mechanisms. Though this does not neccessarily mean that long-range
pathways must be "fixed", it does seem likely to be so.

And if there really exists fixed long-range connections in the cortex,
then understanding the brain would require knowing those connections.
I cant imagine solving the problem without that piece of the puzzle.

The long-range connections of V1-V4 and MT do not show variations
among individuals. Other long-range connections extending to the
parietal and temporal lobes have also been described and they also
seem to be constant.

In short, I still think that figuring out cortical connections should be a
high priority.

PS.  I have been searching for papers on cortical circuits but they are
really hard to find. Do you think some journals would have more of
these articles?






More information about the Neur-sci mailing list

Send comments to us at biosci-help [At] net.bio.net