Re: Addressing Scientific Reductionism


"Wirt Atmar" <atmar@xxxxxxxxxxxxxxxxx> wrote in message news:e09bu0$2gpu$1@xxxxxxxxxxxxxxxxxxxxxx
Robert Kolker wrote:

dkomo wrote:

Moreover, in a book I've just recently slogged through, _The
Plausibility of Life_, such modular organization has been strongly
selected by evolution because it greatly facilitates the process of
viable phenotypic variation. Organisms are modular because they can
evolve more easily if they are. A small amount of genetic change can
result in big changes in organism characteristics. No need for
Darwinian gradualism.

Now that is quite fascinating. That could very well account for
punctuated equilbrium. It is interesting to note there is much more to
(possible) evolutionary mechanisms than the sculptoring and culling done
by natural selection.

I am beginning to understand why thinkers like Dennett consider the
theory of evolution the greatest thing since sliced whitebread.

If you think about it for a minute, you will quickly come to the
conclusion that such a systems design is inevitable. The notion that
"one gene encodes one trait" was completely dispensed with by 1945
(although the idea remains implicit in mathematical genetics). Rather,
by the 1930's, it was becoming obvious that any complex character in the
phenotype was the result of the expression of a significant portion of
the genome.

If that's so, the question then becomes: "how do you evolve such a
structure?" If the interaction matrix between gene products and their
affected traits is such that everything affects everything, evolution
becomes impossible. The entire design becomes a house of cards. You
can't change any bit of code without affecting the entire organism.

The evolution of modular design ("organ-ization") is the only way out,
and it's advantages have been discovered independently by human
engineers and nature, albeit several billion years displaced in time.

Pleiotropy is the inevitable consequence of highly interactive systems.
Pleiotropic effects ? in any complex system ? are intrinsically
minimized by modular partitioning of the subprocesses. And evolution, as
an adaptive process, has just as clearly discovered the values attendant
to modular code and partitioned behaviors. Indeed, such structural
partitioning is the mechanism that allows the common phenomenon of
"mosaic evolution," the evolutionary process where one part of the body
(such as mandibular morphologies or gut chemistries) are allowed to
rapidly change in response to new evolutionary pressures while other
organs (such as the eyes or feet) remain relatively unmodified.

Engineers who work on very complex systems tend to define two qualities
as attributes of complexity: "coupling" and "cohesion." Cohesion is a
measure (somehow calculated) of how closely the internal parts that make
up a module belong together when seen from different perspectives.

Coupling, in contrast, measures how much the submodules interact. Good
structuring and clever systems design works to partition complex systems
into minimally interacting modules by reducing the implicit complexity
by minimizing pleiotropic interactions between the subparts.

[snip relevant advise from Microsoft regarding coupling and cohesion.]

I don't deny that examples of strong intra-module cohesion and weak
intermodule coupling can be found in organisms. One good example is
the prokaryote 'operon' as an organizing principle for the genetic
material. It is very interesting that there is growing evidence that
much of 'horizontal gene transfer' actually happens as 'horizontal
operon transfer'. That is, microbes typically don't swap single genes.
They swap entire metabolic pathways consisting of multiple protein-coding
genes plus a lot of regulatory information. See, for example, the paper
"The Natural History of Nitrogen Fixation" available through this
Wiki article:
http://wiki.cotch.net/index.php/Nitrogen_fixation

But there are also examples in which Nature apparently did not
follow Microsoft's good advice and came up with a solution I would
consider very non-modular. (Or perhaps, my sensibilities as to
what a module should look like are different than Nature's.) My
favorite example is attenuation in the tryptophan operon. Another
is the weird way in which the seleno-cysteine codon is recognized.

One of the reasons I resist the ideology that life is inherently
modular is that it sometimes closes minds in advance of the evidence.
If we already 'know' that life is modular, then we will not find
the exceptions to the rule. Lipmann, for example, provided a nice
modular model for the production and use of cellular energy. But
it actually made it more difficult for Mitchell to convince people
that the situation is more complicated than that.

There is also the population genetics argument. Wirt writes:

If the interaction matrix between gene products and their
affected traits is such that everything affects everything, evolution
becomes impossible. The entire design becomes a house of cards. You
can't change any bit of code without affecting the entire organism.

Ok, but that is not an argument using natural selection at the organism
level for maximum advantage here and now. It is an argument appealing
to retaining the potential for future improvements. NS can take this
kind of criterion into account only if selection takes place at a
higher level - the species level, for example. And, population genetics
tells us that selection at that level must be fairly weak.

Hmmm. I wonder if what is 'wrong' with the pop gen argument is the
assumption that mutation is random and undirected? If a significant
amount of lateral gene transfer takes place, and the likelihood of
a transfer being accepted is related to the cohesion (and lack of
coupling) of the genomic information transferred, then perhaps even
believers in organism-level selection can see the inevitability of
modular organization.


.

Relevant Pages

  • Re: Addressing Scientific Reductionism
    ... such modular organization has been strongly ... theory of evolution the greatest thing since sliced whitebread. ... can't change any bit of code without affecting the entire organism. ... but that is not an argument using natural selection at the organism ...
    (sci.bio.evolution)
  • Re: Addressing Scientific Reductionism
    ... such modular organization has been strongly ... theory of evolution the greatest thing since sliced whitebread. ... can't change any bit of code without affecting the entire organism. ... but that is not an argument using natural selection at the organism ...
    (sci.bio.evolution)
  • Re: James assuary and Archeopteryx
    ... >>> evolution predicts that any particular organism. ... It's the fact that evolution doesn't predict the further mutation ... That's far beyond random chance. ... > However, Evolution includes not only mutation, but selection. ...
    (talk.origins)
  • Re: Human design and natural "design"
    ... >>> Nope, it is called selection. ... Evolution all the way down. ... >>environment to the organism to tell it how to adapt. ... The environment never determines which choice to generate. ...
    (talk.origins)
  • Re: Addressing Scientific Reductionism
    ... such modular organization has been strongly ... theory of evolution the greatest thing since sliced whitebread. ... can't change any bit of code without affecting the entire organism. ... Coupling, in contrast, measures how much the submodules interact. ...
    (sci.bio.evolution)