Re: Mutationism Redux

Arlin wrote:

The infinitesimalist argument (specifically the
empirical side of the argument) has two parts: 1) show that
differences between closely related species are composed entirely of
infinitesimal changes and 2) extrapolate from this to all evolved
differences. The problem is that *the facts never allowed the first
step*. Orr and Coyne, 1992, systematically review the results of
quantitative analyses on this very question, and they write "We
conclude-- unexpectedly-- that there is little evidence for the neo-
Darwinian view: its theoretical foundations and the experimental
evidence supporting it are weak, and there is no doubt that mutations
of large effect are sometimes important in evolution." Furthermore,
they conclude that the MS architects *given the evidence available to
them at the time* were not justified in their conclusions either.

I hate to badger you for more references, but:

Who ever held the idea that "differences between
closely related species are composed entirely of
infinitesimal changes" in the first place?

My understanding of gradualism was always that each of
the individual changes at the /genome/ level was small -
not that they had small *phenotypic* effects.

Of course, jumps across phenotype space are much
more allowable. Alter a regulatory gene that
has effects early on and you may well wind up
with a dwarf organism, half the size of the
original - who *might* nontheless be viable.

I note a genetic interpretation given on:

http://en.wikipedia.org/wiki/Modern_evolutionary_synthesis#Tenets_of_the_modern_synthesis

``2. Evolution is gradual: small genetic changes, recombination
ordered by natural selection.''

Note that it is the *genetic* changes which are small,
not their *effects*.

IMO, "small" probably /ought/ to be interpreted in
information-theoretic terms, as the information needed
to specify the change - though I'm sure this was /not/
what was historically intended.

I.e. duplicating a chromosome makes a lot of data -
but it can be specified with only a little more
information. Endosymbiosis could be interpreted
similarly: the information that one organisms now
lives inside another one is only a few bits.
Large deletions might be large genetic changes,
but they too can be specified with just a few bits.
This would then be a statement of the principle
that there are no large *constructive* jumps
across gene space.

Anyway, the idea that mutations with large *effects*
violates darwinian gradualism seems like a rather
unsympathetic reading to me.
--
__________
|im |yler http://timtyler.org/ tim@xxxxxxxxxxx Remove lock to reply.

.