Newsgroups: comp.ai.alife,comp.ai.philosophy,comp.ai,alt.consciousness
From: ohgs@chatham.demon.co.uk (Oliver Sparrow)
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!pipex!bt!btnet!peernews.demon.co.uk!chatham.demon.co.uk!ohgs
Subject: Re: Thought Question
References: <3eh97p$8oh@agate.berkeley.edu> <3fmba3$nmf@vixen.cso.uiuc.edu> <3fmn70$cpa@pentagon.io.com> <3fnjie$449@lyra.csx.cam.ac.uk> <3g0nri$6vu@pentagon.io.com>
Organization: Royal Institute of International Affairs
Reply-To: ohgs@chatham.demon.co.uk
X-Newsreader: Demon Internet Simple News v1.27
Lines: 32
X-Posting-Host: chatham.demon.co.uk
Date: Thu, 26 Jan 1995 08:52:29 +0000
Message-ID: <791110349snz@chatham.demon.co.uk>
Sender: usenet@demon.co.uk
Xref: glinda.oz.cs.cmu.edu comp.ai.alife:1971 comp.ai.philosophy:24967 comp.ai:26799

The discussion of whether a neuron can or cannot be simulated is not, I suggest 
a very helpful one. I am aware of nothing that neoorons do which is not 
understandable in terms of what we know about how they do it; but the 
properties of neurons, astrocytes and diffusable substances on each other and 
acting in ensembles completely transcend these terms of reference. A model 
which captured all of what a neuron can do - its transfer function emsemble, if 
you like - will be a part of the model which you need in order to understand 
the properties of the ensemble. As you go up the organising tree, the data 
structures which are captured in the ensembles become the driving element of 
what is going on: you still need the neuron model - as you still need chemistry 
- but these have become backdrops to the principle components of variation.

This is something which engineers - who, not unnaturally tend to live in a 
world of unambiguous components which have a defined role in the scheme of 
things - but is a way of looking at the world which biologists find natural.
A set - what it is useful to call a "thing" - a cell, a flower - is defined by 
the roles which it plays. In acting in the scheme and having a set of 
consistent properties which govern how it acts, it becomes a component in the 
engineering sense; but in a self-assembling structure. You can only understand 
these properties if you understand the system in which they are displayed. 
 
They cannot be transferred down the hierarchy and deduced from the properties 
of the components. One can create objects which have properties which are 
possessed by cells, flowers and the other actors in the ensemble and let them 
bang against each other in a computer and the results are often quite 
realistic, but - and it is a *huge* but - you have to know what these 
properties are at the level of the system before you put them in the computer. 

 _________________________________________________

  Oliver Sparrow
  ohgs@chatham.demon.co.uk
