Re: The evolution of "evolution"

From: Chris Taylor (christ@ebi.ac.uk)
Date: Sat 15 Oct 2005 - 02:00:22 GMT

  • Next message: Robin Faichney: "Re[2]: [2] The evolution of "evolution""

    > Thus it's really a question of intelligent versus
    > random adaptation. As I've said, intelligence will re-enter biology one way
    > or another, either via the mind of God or the minds of organisms ranging
    > from bacteria to hominids.

    God? Intelligence? What's all this about then?

    And on the non-genetic inheritance thing; there is little in the way of generational encoding of experience in genes in some Lamarckian sense, but there surely is non-genetic inheritance in everything from egg structure and constitution through to dietary effects passed down through social class (alphas breed alphas). These run in parallel (more or less tied to the gene line). Just throwing that in to confuse the issue really :)

    > If developing organisms merely mimic the actions of their predecessors,
    > design is unnecessary. Neither genes nor the migrating cells that contain
    > them need know what sort of body they're building-- so long as they hew to
    > species tradition. The notion that an organism must follow a set of
    > specifications is a clear-cut example of anthropomorphosis. We are
    > projecting our own mechanistic ways onto the natural world. It's no
    > accident that the other design-theory arises from theology, the ultimate
    > exemplar of anthropomorphosis.

    When you say the 'other' design theory (ID) what is doing your designing? Evolution by natural selection is like carving out a shape from phenotype space. Kind of like hand development (loss of webbing) you see what sculpture is in the rock by chipping away the bits you don't want, rather than putting in the bits that you do want. In this case (fin to hand) I think you can kind of assert a recapitulation argument about the degree of refinements in sculpting organisms (reminds me of 'I went to the shop and I bought...'); some are a bit easier to make, some trickier and the trickier ones usually have embryos that go through that much less exquisite stage.

    > Nobody has ever calculated the correct combination of genes needed for the
    > timing of penicillin production in the haploid mold Aspergilla. That's
    > because the number of possible combinations is 2 to the 1000th power (or 10
    > to the 300th power), way beyond the realm of calculability. This is to say
    > nothing of the production of multicellular organs out of diploid genomes.
    > See *Reflections on a Theory of Organisms,* by Walter Elsasser, for which
    > Rubin contributes an introduction.

    This is just the same old reductio Im afraid. 'It's hard' is not a proper objection as this is a miscasting of the problem. Consider glycoproteins etc. The number of combinations of those things should be beyond comprehension but because of genetic history/ancestry only certain reactions can be catalysed and so the number of 'families' of sugar structures and indeed the actual total number seen are vanishingly small as a percentage of the theoretically available space of possible sugar decoration.

    The point being that genes were never in blueprint form. They are the brain of the cell, not the bauplan. They are an intimate part of an interrelationship of parts that predates time almost. This is also the fallacy of the 'RNA world'; why do we need a
    (probably white, anglo, male, muscled, blond) molecule to step up to the plate and sort the chemical chaos into a biosphere? It was always more the hypercycle thing and then one sort of compound increasingly becomes a sort of canalising information exchange point and switching network, just like the nervous system evolved out of the workaday cells.

    To answer the penicillin point above; the possible combinations may be astronomical but the point is that that space was never thoroughly explored by the organism. It went where its history
    (both long-term genetic and short-term metabolic) _let_ it, in little steps, and had a hard copy of exactly where it was
    (genetically and metabolically) at all times. There is nothing awe-inspring about this; there are an almost infinite number of paths I could walk across a football field, or even a cargo net, but to say that retrospectively, the chance of me taking any one path is astronomically small misrepresents the scenario. I was always going to find a way that worked, or I couldn't have been said to have crossed (analogue: proliferated as a species) but how densely packed with good points (strings/ridges) is the space of possible solutions, and how multidimensional?

    I still want to know more about methylation. Obviously the patterns are copied, but it seems like an excellent way to play with the gene network, looking for new and interesting patterns; toggle-able, short-term mutations effectively. Is it also stochastically-governed? Could stress affect methylation patterns, or some other environmental?

    One last thing about explanations. Obviously there is no way to prove a negative, or in this case that what seems perfectly explicable given the causes to hand may actually be inextricably affected by another unseen effect, but when modelling one should attempt to use a measure of explanatory power and these always penalise extra terms (when, for example modelling some polynomial with more polys than you need). I fail to see what we cannot currently explain, or at least where we find something to be beyond the scope of explanation being conceivable within our current frame of reference.

    Cheers, Chris.

    ~~~~~~~~~~~~~~~~~~~~~~~~
      chris.taylor@ebi.ac.uk
      http://psidev.sf.net/
    ~~~~~~~~~~~~~~~~~~~~~~~~

    =============================================================== This was distributed via the memetics list associated with the Journal of Memetics - Evolutionary Models of Information Transmission For information about the journal and the list (e.g. unsubscribing) see: http://www.cpm.mmu.ac.uk/jom-emit



    This archive was generated by hypermail 2.1.5 : Sat 15 Oct 2005 - 02:17:35 GMT