From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!samsung!uunet!tdatirv!sarima Tue May 12 15:49:38 EDT 1992
Article 5477 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!samsung!uunet!tdatirv!sarima
>From: sarima@tdatirv.UUCP (Stanley Friesen)
Newsgroups: comp.ai.philosophy
Subject: Re: syntax and semantics
Message-ID: <16@tdatirv.UUCP>
Date: 7 May 92 21:52:11 GMT
References: <1992Apr8.215800.18021@mp.cs.niu.edu> <92099.194744JPE1@psuvm.psu.edu> <1992Apr9.174840.3407@organpipe.uug.arizona.edu> <5674@mtecv2.mty.itesm.mx> <2@tdatirv.UUCP> <571@trwacs.fp.trw.com>
Reply-To: sarima@tdatirv.UUCP (Stanley Friesen)
Organization: Teradata Corp., Irvine
Lines: 61

In article <571@trwacs.fp.trw.com> erwin@trwacs.fp.trw.com (Harry Erwin) writes:
|sarima@tdatirv.UUCP (Stanley Friesen) writes:
|
|>WRONG me bucko!   We *do* know, in large part, how a neuron works.
|...
|Which model of neuron functioning are you refering to? The computer
|science variant (highly oversimplified) or the neuroscience variant (which
|frankly admits its ignorance).

The neuroscience variant, which admits its ignorance *past* *a* *certain*
*point*.  The basic functionality,in terms of chemistry and physiology is
quite well understood.  It is only certain details that are not known.

As I stated, this is true for *all* sciences - there are always aspects
of any phenomonon that are mysterious, no matter how well studied.

As an example, in this week's Science News there is a report on a several
experiments designed to measure the mass of a neutrino.  It seems they
consistantly get unexpected results, and unless this is due to some sort
of interference, this means there is something about neutrino production
that is unknown.

|Dowling and Boycott pointed out in 1965
|that few neurons in the eye prior to the ganglion cell layer generate
|nerve impulses, and all interactions are via graded polarization.

Hmm, that is certainly interesting, but it hardly suggests a major
revision of neural theory.  After all, it is only the ganglion layer
cells of the eye that are anything like normal neurons in any way.
So, it is hardly surprising that such unusual cells as the bipolar cells
and the amacrine cells have unusual modes of action.
[Actually the amacrine cells correspond to a rare cell type of the cerebral
cortex, found occasionally in the outermost layer, so it may have some
relevance to computational neural network theory].

|In this context: has anyone looked at the evolution of neural network
|components? The neuron, although primitive, is a complex beastie. What is
|a realistic evolutionary sequence from ectoderm cell to basic neuron?

The Hydra, as I remember, has epithelial cells (or is it muscle cells)
that are intermediate in form.

|Just for a laugh: Christen, Ratto, Baroin, Perasso, Grell, and Adoutte
|(1991) claim to have shown that plants, diploblasts (cnidaria, ctenophora,
|and porifera), and triploblasts (the remaining metazoan phyla) are derived
|independently from a protist assemblage, with plants and diploblasts being
|more closely related than either are to triploblasts. They come to this
|conclusion from RNA sequencing data. One implication of this result is
|that the neuron predates multicellular life, since it is found in both
|diploblast and triploblast phyla.

Not necessarily, convergence is something that must always be considered
in this type of case.  Given the functional constraints, and the prior
existance of intercellular signalling based on hormones, I suspect that
essentially identical neurons could evolve more than once.

[note that most neurotransmitters are actually hormones, and the others act
like hormones with regard to the nerve cell].

Finally, the results mentioned are still unconfirmed by other researchers,
and so are still in limbo, scientifically speaking.


