Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site telesoft.UUCP
Path: utzoo!linus!philabs!cmcl2!seismo!harvard!talcott!panda!genrad!decvax!ittatc!dcdwest!sdcsvax!telesoft!kst
From: kst@telesoft.UUCP (Keith Thompson @stroke)
Newsgroups: net.origins
Subject: Re: No Cigarettes, Please ("fitness")
Message-ID: <170@telesoft.UUCP>
Date: Fri, 2-Aug-85 03:53:22 EDT
Article-I.D.: telesoft.170
Posted: Fri Aug  2 03:53:22 1985
Date-Received: Mon, 5-Aug-85 00:15:07 EDT
References: <1318@uwmacc.UUCP>
Reply-To: kst@telesoft.UUCP (Keith Thompson @stroke)
Distribution: net
Organization: TeleSoft, SanDiego CA
Lines: 58
Keywords: fitness, natural selection

In article <1318@uwmacc.UUCP> dubois@uwmacc.UUCP (Paul DuBois) writes:
> ...
> Organisms are or are not fit.  Fitness is the thing we want to measure,
> but how do we do it?  We can say that fitness is some abstract
> quantity, but which is a covert variable, not directly observable.  So
> we need an overt variable that we can hang a number on.  The overt
> variable needs to be in some sort of relationship to fitness, and
> reproductive success is thought to have that characteristic.  If the
> relationship is strong, then we have in differential success a good
> indicator of the fitness of the population.  My question is:  how do we
> know that differential reproductive success is a valid indicator?  My
> answer is that we don't.  It is thought to be so, but I'm skeptical.
> Is an increased number of offspring a necessary consequence of
> fitness?  I don't think so.
> ...

Differential reproductive success is indeed the key issue here.  
This is, by definition, the *only* indicator of the fitness of a
population.  Specifically, "fitness" is the set of characteristics
that tend to result in reproductive success.  (This is my personal
definition; I don't claim that biologists define the term the same
way.)  What these characteristics are depends on what kind of
environment the organism lives in.  Longevity is certainly one
characteristic that contributes to fitness (the longer an organism
lives, the more time it has to reproduce), but by no means the only
one.  I would say that the common housefly is considerably more
successful as a species than, say, the Galapagos tortoise.

Note that what I have stated so far is a definition, not a (testable)
theory.  What the theory of evolution by natural selection states, as
I understand it, is that there are indeed such characteristics, and
that they can be passed on to an organism's descendants.  In other
words: 
(1) Reproductive success is (at least partially) a result of certain
    characteristics inherent to an organism; it is not purely a result
    of random chance.
(2) These characteristics can be passed on; successful organisms tend
    to have successful offspring (and lots of them).

Note that there is a distinction between fitness and success.  Fitness
is a set of characteristics inherent to an individual or species.
Success is an observable *result* of fitness.  Note that success can
be a result of things other than fitness.  For example, blind luck can
play a significant role.  The more recent "post-Darwinian" theories of
evolution focus on factors other than natural selection.  However,
natural selection still plays a major part (I think).

Given that mutations occur, creating a large pool of different
characteristics in a population, and that some small fraction of these
mutations result in greater "fitness", and thus in differential
reproductive success, how can evolution *not* occur?

Comments appreciated; flames to /dev/null.

		-- The_Other_Keith
--
"Inevitably, their affair ended: Howard worried excessively about what
the pack would think, and Agnes simply ate the flowers."