[Comp-neuro] Re: Reproducability, Funding, and a note on hippocampus

A. David Redish adr at adrlab.ahc.umn.edu
Wed Aug 20 18:53:13 CEST 2008

Re: unifying models

The real issue, I think, is that we do not yet agree on the correct
language with which to describe the brain.  Is it molecules, channels,
compartmental models, abstract networks... ?  My opinion is that the
correct descriptive language depends on the questions.  Since we don't
have the correct language yet, I don't think we would want to limit
ourselves at this point.  Imagine if physics limited it's models to
what was available conceptually before Newton or before Planck or
before Einstein...  I think that this sort of debate between multiple
potential explanatory languages is very healthy and should be
encouraged rather than stifled.

Re: Hippocampus

In terms of hippocampus, the models actually fall into identifiable
families (attractor networks, oscillator-models, compartmental
models).  These models do provide bases and predictions that are very
replicable between labs.  As one example, we can compare the attractor
network family, typified by work that Kechen Zhang did, work that
Alexei Samsonovich did with Bruce McNaughton, and work that I did with
Dave Touretzky, all in the 1990s, all of which came to similar,
highly-replicable conclusions.  As another, we can compare the recent
work done in the O'Keefe, Burgess, and Hasselmo labs on
dual-oscillator models, which, again, have made surprising predictions
that have been borne out.  (See, for example, the wonderful paper by
Giocomo/Hasselmo [Science 2007] testing predictions of multiple
dual-oscillator models from different labs.)

I don't think it would be good to have a single model base
(Neuron/Genesis/Neural-Nets) with which to create models because I
think that would be stifling and limiting.  What is critical (I think)
is to have the models available so that they can be replicated.

Re: Lone-wolfism

Jim Bower wrote:
> How do you know a real lone wolf from one who simply has rabies?

Isn't that the million-dollar question? [Literally.  That's the
funding question, right?]

Also, isn't that the fundamental question that we're all trying to ask
ourselves?  How do we know when our ideas are just crazy? :)

A. David Redish         redish at umn.edu
Associate Professor     http://umn.edu/~redish/ 
Department of Neuroscience, University of Minnesota

More information about the Comp-neuro mailing list