Newsgroups: comp.ai.alife
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!pipex!sunic!news.chalmers.se!news.gu.se!gd-news!d6173.shv.hb.se!sa209
From: sa209@utb.shv.hb.se (Claes Andersson)
Subject: Re: Lamarckian Evolution
Message-ID: <sa209.106@utb.shv.hb.se>
Sender: usenet@gdunix.gd.chalmers.se (USENET News System)
Nntp-Posting-Host: d6173.shv.hb.se
Organization: Department of Scocial Science
References: <3ec0a0$7es@gap.cco.caltech.edu> <3eg6an$k8m@laplace.ee.latrobe.edu.au> <sa209.102@utb.shv.hb.se> <3et14l$1hl@laplace.ee.latrobe.edu.au>
Date: Wed, 11 Jan 1995 11:05:08 GMT
Lines: 60

In article <3et14l$1hl@laplace.ee.latrobe.edu.au> khorsell@ee.latrobe.edu.au (Kym Horsell) writes:

>In article <sa209.102@utb.shv.hb.se> sa209@utb.shv.hb.se (Claes Andersson) writes:
>> Still, what's most comforting with the Darwinian evolution theory is that 
>>it is very logic. Natural selection can bring order to chaos just because of 
>>it very logical ability to do so, nothing strange with it. What, on the 
>>other hand, would be strange AND hardly logical is a step between where the 
>>gametes in a magic way "knows" in advance what will be good for the 
>>offspring. What someone who advocates any sort of Lamarckian evolution 
>>implies is that something picks the right genes by hand, or how would the 
>>environment which evidently acts on the phenotype be able to alter the 
>>GENOTYPE is a meaningful way? I think it falls on the fact that it is 
>>impossible.. unless ordinary Darwinian evolution has evolved some sort of 
>>gene-alterating device with an ability to look into the future. Genetic 
>>psychohistory! :->

>I don't see that Larmarck is very different from the "logical Darwinian"
>theory in this respect. The genes don't "know" anything other than
>the parent SURVIVED TO THE POINT necessary to produce offspring. The
>information carried to the offspring can only include (under Larmarck)
>what the parent experienced (and developed biological countermeasures to).
>Darwinian theory syas gametes only arry what parent inherited at birth.

 That is exactly what I'm saying isn't it. The problem is that eblish, or 
any other language as well, don't have a better word than "know" to use, it  
is always  followed by a little explaination that it doesn't mean "know" in 
the usual way. Read what I've written once again and you will see.

>It should be pretty clear to establish this one way or the other
>using alife techniques. At what speed does a population develop
>immunity to disease? For diseases that do not have very high
>mortality we might expect Lamarckian evolution to develop quick
>solutions: Darwinian evolution might be slower.

 Once again: Natural selection selects entire phenotypes, due to crossover 
individual genes are selected for or agains in the long run. How would the 
laws of nature know how to alter a genotype so that it generates the proper 
phenotypic results? It just can't happen! I'm a programmer.. assume that I'
ve made a program which is tested by end-user with no knowledge whatsoever 
of computer programming. It's nothing strange happening as long as they 
point out the bugs and I correct them but when they start debugging it 
themselves without being able to program: it is ODD. The probability for a 
non-programmer to debug a complex software with success is VERY slight and 
one can certainly not rely on it.

>As far as I know it is suspected the sites of "crossover points" are
>themselves encoded genetically. It might therefore be the case that such 
>a mechanism is used to select between alternatives using environmental factors
>as offspring are produced and thereby affect a Lamarckian effect.

 But that is not Lamarckism... There  are many examples of genes affecting 
the genome.. once again: it isn't magic! Since the genome is in the absolute 
vicinity of the genome, why wouldn't the genes "try" to affect it? Isn't 
restriction enzymes etc. examples of this as well? Genes trying to put the 
crossover point before or after themselves and not in the middle does not 
represent an example of Lamarckian evolution in any way. Lamarckian 
evolution is that the environment actively alters the phenotype in a way 
that is possible to inherit.

Claes Andersson. University of Bors. Sweden
