Re: Levels of selection (Was: Can cognition override natural

WM:-
I have to start out by thanking Guy for a nice explanation of why meiosis 
negates the gene as a unit of selection 

>snip<

JE:-
Once real epistasis (non additive 
genomic gene associations) and not 
just phoney epistasis (additive
genomic gene associations) which 
appear to typify every single genomic 
gene _fitness_, enters selective
reasoning, then genes cannot be 
independently selected no matter 
if Hamilton's rule suggests that 
they can. If a gun fails to shoot
then it is always more parsimonious
to suggest the gun is faulty rather
than the laws of nature have suddenly
changed...

> GH:-
> OK.  Here is my quick answer to why it does not make sense.  I think
> the most concise way to explain it has been frequently argued by John
> Edser. [Note that I have always agreed with John on this aspect of his
> arguments, while disagreeing with other aspects.]  Individuals are not
> even close to merely the additive effects of genes. 

JE:-
All real gene fitness is epistatic.
Thus real gene fitness is entirely 
dissolved at meiosis, leaving only 
the organism level of fitness for
selection to act on. Thus all independent
gene fitnesses that are supposed to be
independently selectable at Hamilton's
proposed gene level of selection are
fiction, leaving only one level of selection
within _testable_ evolutionary theory. 
A simplified model is simply, not a theory 
and is misused if it is allowed to compete 
and win against any theory it was simplified
from (as if such a thing was not 100% obvious).


 GH:-
> ..it is still rather inconceivable to me how natural selection
> can operate when the reproducing agents exist at a scale of
> organization different from that where fitness is manifested (others
> might prefer to word "evaluated" here), I hope that you (or anyone)
> can explain it to me.  I am not aware of any logical explanation being
> provided by Dawkins, and IMHO there is no such logic.

JE:-
All you seem to be asking is how can a 
multilevel have different levels
such that one level "acts" while another
produces all the "fitness". This is possible
when a lower level is only selectable
at a higher level, as is the case with
gene fitnesses. Dawkins just has everything
in REVERSE. Since no independent gene level
exists then genes are actors for individual
fitnesses that incorporate 100% of gene
fitness epistasis. Amazing as it may seem,
Darwin's suggestion that Darwinian fitness
is heritable is more accurate than all the
algebra used by population geneticists
re: fitness at the gene level because
Darwin's remark includes all epistatic 
gene fitness.

The challenge is to reformulate genetics
and population genetics so that non additive 
gene relationships become "heritable" and not 
just "inherited". Fisher's model is so entrenched 
nobody is even willing to think outside his 
(tiny) square.


Regards,

John Edser
Independent Researcher

PO Box 266
Church Pt
NSW 2105
Australia

[EMAIL PROTECTED]