iterative processing

[James M. Bower]

Read carefully, Richard's response to my previous email indicates pretty
clearly the degree to which prior thinking about how the olfactory system
works influenced the development of the olfactory model.  This is the sole
point I have been trying to make.

In particular he states that:

"The 1986 monograph was an early and fruitful step in the field of modeling
of real biological systems."

Of course there was no model in the monograph -- the discussion involved
speculations based on biological data and certain assumptions concerning
olfactory processing (e.g. iterative refinement of olfactory response).
Richard's long history makes pretty clear that it was those assumptions
that drove the subsequent modeling.  This, I presume, is what Richard meant
by a "fruitful step".  The "fruit" in this case, being the model.

Richard's recounting of the history also makes it clear that the original
claim that the model was based on a wide range of biological data,
including data published in 1990 and after is somewhat difficult to
reconcile with the statement that:

"two papers (were written) in 1988 and four in '89, all attempting to
identify various aspects of function from the structure and operation of
the system."


Finally, there are many technical and biological issues that could be
raised concerning the assumptions and conclusions of this particular
modeling effort (The location of LTP, or evidence that rats can apparently
recognize odors on a single sniff (theta cycle), or the issue of whether
olfactory perception space is really heirarchically clustered), however, it
was never my intent to argue about the model itself.  Instead, I was trying
to make the point that one has to be very careful to distinguish between
models that assume a particular function, and then try to identify a
biologically plausible structure that might provide it, and models that
start with structure, and try to infer function.   It may be that this
distinction is blurry to some -- however, if one has done the later, the
distinction is obvious and important.

Jim Bower


================ Message 3 of 4 ================