Newsgroups: comp.ai.alife
Path: cantaloupe.srv.cs.cmu.edu!das-news.harvard.edu!news2.near.net!MathWorks.Com!yeshua.marcam.com!usc!crash!snodgras
From: snodgras@crash.cts.com (John Snodgrass)
Subject: Re: Defining "Life"
Organization: CTS Network Services (CTSNET), San Diego, CA
Date: Sun, 25 Sep 1994 19:56:55 GMT
Message-ID: <snodgras.780523015@crash.cts.com>
References: <35acgm$232@network.ucsd.edu> <snodgras.780094130@crash.cts.com> <35o0pv$ah9@masala.cc.uh.edu>
Sender: news@crash.cts.com (news subsystem)
Nntp-Posting-Host: crash.cts.com
Lines: 53

In <35o0pv$ah9@masala.cc.uh.edu> clifton@rodin.cs.uh.edu ( Clifton B. Davis ) writes:

>In article <snodgras.780094130@crash.cts.com>,
>John Snodgrass <snodgras@crash.cts.com> wrote:
>>In <35acgm$232@network.ucsd.edu> reality@ucsd.edu writes:
>>
>>

[del]

>>in an obvious way with his professed view.  Are machines self-organizing?
>>No? Well, let's leave the definition of life open, shall we? 
>>     JES

>   Well, yes, there can be self organizing machines.
>   A brilliant and wide-ranging mind describes Minsky.
>   No one is stopping you from discussing the definition of life to your
>       hearts content.
>   Chill.
>   What, in your opinion, is the evidence for vitalism?  Is it simply
>a failure on your part to see how mechanism could be an adequite
>explanation, or do you have a stronger reason?

     Vitalism and mechanism are two modeling strategies. They are equal
and opposite in their power. Mechanism provides the ability to model things
about which little is known of their internal workings, either because they
are too big or too small. Vitalism makes much more sense when looking at
macroscopically self-organizing systems, because these systems are purposefully
functioning to maintain their own stability. This is why physics is not
applied easily to biology, and vice-versa. But we shouldn't want to take
this specialization of modeling based on our relationship with our world
to heart, insist that one is "better" than the other (as people like Minsky
do). Is matter at bottom self-organizing or externally controlled by physical
law? It is impossible to show, since one or the other must be initially
assumed as the basis for the modeling strategy. The more interesting
question is: how do we seamlessly integrate the two strategies? How do we
build a new science which recognizes the power of both? 

     If I sometimes sound like I'm saying matter is self-organizing -- it
should be taken to mean that this is a functional modeling strategy, on a
par with the mechanistic approach. 

     This is related to AI. People are commonly heard to suggest that
lifeforms are "just a different kind of machine" (e.g. Data on STTNG is
just a different kind of person). This is the belief that underlies AI.
But machines are not self-organizing, cannot be made to be so, because they
are designed (design and self-organization are opposites). With lifeforms
there is ambiguity as to what they ultimately are, but not with machines.
Therefore the comparison is unbalanced. The basis for the notion of AI is 
an oversight. 

     JES
 
