Newsgroups: comp.ai.alife
Path: cantaloupe.srv.cs.cmu.edu!nntp.club.cc.cmu.edu!godot.cc.duq.edu!ddsw1!news.kei.com!hookup!swrinde!pipex!lyra.csx.cam.ac.uk!sunsite.doc.ic.ac.uk!uknet!bcc.ac.uk!uu-igor-mac.open.ac.uk!user
From: A.J.Hirst@uk.ac.open (Tony Hirst)
Subject: Re: Lamarckian Evolution
Sender: news@ucl.ac.uk (Usenet News System)
Message-ID: <A.J.Hirst-060195122247@uu-igor-mac.open.ac.uk>
Date: Fri, 6 Jan 1995 12:22:46 GMT
References: <D1ty93.9xJ@lincoln.gpsemi.com> <D1un1B.Er2@watdragon.uwaterloo.ca>
Organization: HCRL, The Open University, UK
Followup-To: comp.ai.alife
Lines: 47

In article <D1un1B.Er2@watdragon.uwaterloo.ca>, bpvanstr@yoho.uwaterloo.ca
(Brian Van Straalen) wrote:

> In article <D1ty93.9xJ@lincoln.gpsemi.com>,
> David Whipp <whipp@roborough.gpsemi.com> wrote:
> >The Lamarckian view of evolution is that the experience of an
> >individual may be passed to the next generation when it reproduces
> >(i.e. in the genes).  Although largely discredited in wet-life

so called lamarckian effects have been suggested as arising (especially in
single celled organisms) from adaptive (rather than random) mutation and
epigenetic inheritance (ie inheritance of the functional state of DNA eg as
occurs during differentiation in multicellular organisms).

> >evolution, there is no reason why it can't be attempted in a-life.
> >

I completely agree - alife provides us with the opportunity to play around
with various theoretical biologies that suit the computational medium;
alife is grounded in a logical rather than a chemical-structural physics,
so there's no reason why we shouldn't experiment with logically plausible
inheritance regimes even if we find no corollary in natural biology. 


> 
> Lamarkian mechanisms could be the needed computational shortcut we need
> to make truly complicated (large genomed) alifeforms possible.
> 
>   (unfortunately, using this as a biology tool wouldn't be a good idea,
>   since nature seems quite content to wait for darwinian processes)
> 

it may not be a good idea in alife either (depending on the application) -
matural evolution may be slow, but it does give rise to robust agents; if
we allow for lamarckian inheritance, then we may find that we are evolving
increasingly brittle children that can survive only in a very restricted
range of environments; this is ok if you're tackling an optimisation
problem with a single solution, but if you're interests lie, for example,
in a robot control system that is likely to encounter many unexpected
situations, then i suspect there'll be a high mortality/extinction rate....


--------------------------------------------------------------------------
 All opinions etc etc...
--------------------------------------------------------------------------
      | Tony Hirst ("Monty")          | e-mail:  A.J.Hirst@open.ac.uk
       -------------------------------------------------------------------
