Query: Dendritic Networks
paul at phy.ucsf.edu
Thu Feb 8 17:32:13 EST 1996
In article <JAN.96Feb5142724 at cava.neuroinformatik.ruhr-uni-bochum.de>, jan at neuroinformatik.ruhr-uni-bochum.de (Jan Vorbrueggen) writes:
|> In article <mike.lowndes-0102961103290001 at mac3.anat.ox.ac.uk>
|> mike.lowndes at anat.ox.ac.uk (Mike Lowndes) writes:
|> You can bet yr bottom pound sterling that anything written about it by a
|> physicist/mathematition/ modeller is almost bound to turn out WRONG. The
|> only way to get it right is to understand the physiology as a first step
|> and we as neurscientists don't even understand that.
Your first sentence seems to contradict your last sentence (below). But in
general I agree with your argument.
|> Wrong it what sense? I think your statement epitomizes what's wrong with
|> (neuro)biology: the general unwillingness to take a more or less simplified,
|> substitute system to try to understand the general features of the system
|> under study (here: the brain), and to hell with the details. Of course the
The problem with this argument is that such models have been constructed, a large
number of them. None of them has led to any great insight, I think because they
sacrifice too much in order to produce a tractable analysis. That and the fact
that many of this type of modeller pay only cursory lip servive to actual brain
data. Of course, none of the biophysically-based neural models have produced a
major breakthrough yet either, at least not since Hodgkin and Huxley/Rall 30 or
40 years ago. But I live in (confidant) hope.
|> brain is a much more complicated case than, say, all those semiconductors,
|> where this programme has worked very well; but to say that any model must
|> explain all known data (including the bad and the contradictory, which nobody
|> is willing to weed out by writing a well-reasoned and opinionated review) is
This job (writing a comprehensive, objective review) would be a fulltime
occupation for each subfield in Neuroscience. But it certainly would be helpful.
I have often thought that Neuroscience would benefit more from hiring people to
do this job than hiring a few more experimentalists.
|> On the other hand there is nothing wrong with trying to /mimic/ the brains
|> processes- which is what modellers are doing.
|> Duh. What else than mimic could we possibly do? Or, to phrase the question
|> differently: where do you draw the line between "mimic" and "understand"?
Good question. Difficult to answer, but I bet that if you could build a model of
the brain that duplicated all its functions then you would have a pretty good
understanding of how it works.
More information about the Neur-sci