Re: Information theory, entropy, and evo

Danny Yee (danny@STAFF.CS.SU.OZ.AU)
Thu, 23 Feb 1995 10:32:18 +1000

Dwight Read writes:
> Neutral mutations and drift required a modification of this framework by
> virtue of the possibility of allele frequencies changing stochastically, even
> allowing for replacement of an existing allele by a new, mutant allele in the
> absence of any selective effects. Neutral mutations do not invalidate the
> general selectionist framework, but did require certain modifications; e.g.,
> stasis is not the null condition in the absence of selective effects.

Yeah, yeah, I've read Kiumura. Neutral mutations certainly don't require
the introduction of teleology, which is what I was really complaining
of in Steve's message.

> Recent work on self-organizing systems suggests that structures can
> arise prior to selection (the self-organizing structure) and so selection
> need not be invoked to account for the existence of these structures.

Protein self-assembly is just this. It has been known about for decades,
it is not a big new discovery. (Steve always seems to take things like
that and talk about them as if they were revolutionary; one of the reasons
I keep disagreeing with him!)

> Selection might favor the conditions which give rise to the self organizing
> structure, so that there could be selection at more of a macro, rather than
> a micro level. Again, this does not negate the selectionist framework, only
> that certain modifications will be needed such as macro structures possibly
> may be introduced as a unity rather than being built up piece by piece, as it
> were--i.e., macro evolution may be more than micro evolution writ large.

Hmmm... shifting balance style changes are one thing; saltationism is
another. No one has ever come up with good evidence for macromutations
playing an important role in evolution. Species level selection/evolution
is another kettle of fish, of course.

> I take it that Campbell is suggesting yet another way in which what selection
> acts on may be more complex than single alleles, namely that there may be
> some kind of internal organization or structure to DNA that affects what
> kinds of mutations are possible (my reading of what has been said here about
> Campbell's argument is that this is not meiotic drive). Again, this will not
> negate the selectionist framework. Rather, that it will (if the idea pans
> out) require further modifications. (I am intrigued by Campbell's argument as
> I have long thought that the phenomena which are produced via DNA has a level
> of complexity that is hard to accomdate within DNA which is "merely" a
> sequence of alleles each acting as an independent unit.)

But who forgot the proteins? The immediate gap between the complex
phenomena produced via DNA and the simple DNA structure is protein
self-assembly. One of the things I liked about _Chance and Necessity_
was that it put much more stress on proteins and protein self-assembly
than most.

> So I suspect that what we will be seeing is more evidence of the way in which
> the model of individual alleles independently susceptible to mutation and
> selection will need modification is too simplistic and that more and more
> structure (IMHO) will be identified that comes about not via selection
> (though selection will act on what is produced), but via the whole nature of
> chemical and biological processes involved.

Most structure "comes about" because of history -- Gould and Lewontin's
famous article looks at macroscopic features, but the same thing
holds at the molecular level too. There's a lot of debate about how
significant epistasis is, and whether there's anything in "shifting
balance theories", but it's not something new, and it certainly
doesn't provide the least bit of evidence for teleology in evolution.
(The main objection I had to Steve's summary was that the claim
about teleology.)

Conclusion:

Recent work on self-organising systems at a molecular/genetic level is
tremendously exciting (though it's not yet clear how significant the
phenomena are) but it *doesn't* require fundamental epistemological
or ontological changes in the way we approach evolution.

Danny Yee.