Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!mcnc!rti!dg-rtp!ellerbe!poirier
From: poirier@ellerbe.rtp.dg.com (Charles Poirier)
Newsgroups: comp.ai
Subject: Re: evolution is not a scale
Message-ID: <1990Dec27.192128.22054@dg-rtp.dg.com>
Date: 27 Dec 90 19:21:28 GMT
References: <70996@bu.edu.bu.edu> <1870014@hpwrce.HP.COM>
Sender: usenet@dg-rtp.dg.com (Usenet Administration)
Organization: Data General Corporation. RTP, NC.
Lines: 43

In article <1870014@hpwrce.HP.COM> kingsley@hpwrce.HP.COM (Kingsley Morse) writes:
>/ hpwrce:comp.ai / colby@bu-bio.bu.edu (Chris Colby) /  7:43 pm  Dec 18, 1990 /
>
>>The idea that evolution is a linear scale with humans at the pinnacle
>>is simply not true.
>
>I'm interested in evolution, and applying it to AI. I've been assuming that 
>evolution is goal directed, in a sense. I'm interested in feedback on my 
>assumption. 
>
>Do you think it's possible that Darwin's theory of natural selection uses
>"fitness" as a goal? In other words, Darwin proposed that "survival of the
>fittest" causes succeeding generations to be ever better adapted to
>the environment. Could successively fitter generations be considered a "scale",
>in that any given branch of evolution has "fitter" genes at it's leaves than
>at the root?

"Fitness" can not be defined in isolation; one is only "fit" with respect to
a given environment.  The poster's point of view would be fine, were it not
for the fact that the environment itself is continually changing.  Bear in
mind that the "environment" includes other living, and thus evolving,
creatures.  Climate and terrain are other factors that vary over time.  I
expect that if today's species could be somehow transported to a past era,
that some (perhaps many) would be less fit *for that environment* than their
forebears who evolved for fitness in that world.  If so, we can't claim
consistent progress toward an idealized goal of "fitness".   We lack a good
crystal ball to tell us the shape of some future environment; so we can't
say whether or not the current environment directs current natural selection
toward future fitness.  In short, evolution is not goal-directed because
the goal does not exist.

Of course, there is continuity as well as change.  One could attempt to
abstract those aspects of the environment that remain relatively constant
over a given span of space and time, and claim that natural selection
is directed towards the goal of fitness to survive and reproduce in that
(abstract) environment.  But this is a slippery concept.  One might counter
that the local environment of every individual is unique in some way,
implying a multitude of distinct goals: one for each creature.  One is then
stuck with the unsatisfying image of each and every living thing as the
pinnacle of a chain of length one.

	Cheers,
	Charles Poirier   poirier@dg-rtp.dg.com