From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wupost!ukma!nsisrv!mimsy!harwood Thu Dec 26 23:57:36 EST 1991
Article 2317 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!wupost!ukma!nsisrv!mimsy!harwood
>From: harwood@umiacs.umd.edu (David Harwood)
Newsgroups: comp.ai.philosophy
Subject: Re: Scaled up slug brains
Message-ID: <45281@mimsy.umd.edu>
Date: 20 Dec 91 18:36:09 GMT
References: <1991Dec18.071959.4921@daisy.ee.und.ac.za> <45103@mimsy.umd.edu> <1991Dec20.164840.4658@aifh.ed.ac.uk>
Sender: news@mimsy.umd.edu
Organization: UMIACS, University of Maryland, College Park, MD 20742
Lines: 50

In article <1991Dec20.164840.4658@aifh.ed.ac.uk> bhw@aifh.ed.ac.uk (Barbara H. Webb) writes:

>>	Eg - some animals have color vision. There is no evolutionary
>>"scaling-up" of nervous systems with eyes having 1 pair of cones, 2 pair,
>>3 pair, ..., N pair. No- they have genetically coded "color vision"
>>cells and "circuitry", or they don't.
>
>This doesn't seem like a good example: something like 1% of human males
>have only two cones (not the usual three) due to a genetic difference
>(have you ever heard of colour blindness?).
>There are people who have only one cone (I have seen experimental reports on
>such a subject). Turtles (and possibly some other animals) have four
>cones. Colour vision is certainly not "all or nothing" in the natural
>world.
\\\\\\\\\\\\\\
	No - I know this. I did not make myself clear (as I was aware
when I hurriedly replied). I have done research on detection of eye
disease, and work on computer vision; yes I am familiar with these 
things.)
	I meant animal brains show very major "qualitative" differences,
due to even small genetic variation (even such as you describe above).
(There are very many specialized neurons and circuits in the brain, due
to these genetic specifications. For example, as I mentioned elsewhere,
it is not just a matter of retinal photoreceptors (that have types as
you agree with me- a qualitative difference), but there is the lateral
geniculate nucleus of the brain which transforms this output of these
rods and types of cones, specifically for color vision.
	You thought I was incrementing number of types of cone cells,
but I was not. I was trying to say that a brain with color vision
cannot be obtained by increasing simply by "scaling up" slug brains
by incrementing their cells, because they don't even have appropriate
types of neurons (and genetic programs). There are radical innovations in 
brain structure, which are not due to incrementing numbers of cells or
simply rewiring circuitry.
	I was giving an illustration, supposing for simplicity that there
were two populations (types) of retinal transducers. (I did not want to
get involved in rods and various numbers of cone-types, so I chose two
types.) And I was trying to illustrate that no matter how many cells
of these two types there were - there was no "scaling up" (in numbers of
tokens) that would change the type or (quality) of color vision.
	A slug brain without cones cannot "scale-up" to mammalian color
vision, no matter how many cells of the same kind you add or rewire.
(Of course you can add a different type of photoreceptor-neuron, but that
is not simply "scaling-up" - that is a new type of neuron; and even so
you may also have to add new types of cells to the LGN.)
	Sorry for the confusion. Actually, you are elaborating on my
same point by talking about these genetic defects and species-specific
variations.




