Re: (non-flame) Dawkin's METHINKITISLIKEAWEASEL model

From: Chris Cogan (ccogan@telepath.com)
Date: Sat Oct 14 2000 - 02:58:07 EDT

  • Next message: Chris Cogan: "Euphemisms and More (was What really killed T. Rex?)"

    Chris
    Thanks, Brian, for the Dawkins quote (below). I was reasonably sure that
    Dawkins was being misrepresented (*again*), but I did not get around to
    looking for the passage in TBW.

    Interestingly, it is not hard to modify the model to make it produce
    interesting results *without* any specific target. *Exact* physical targets
    can be awkward for DNA-based evolution, even though *functional* targets
    (such as resistance to heat and dryness as in the case of camels and
    several other organisms) can often be achieved fairly closely. Indeed, they
    can often be achieved in several *different* ways.

    With some modifications, it also becomes an example of the richness
    available to repeated small-scale variation (as well as a model of how a
    complex result can be achieved cumulatively). Why? Because it is clear that
    *every* string of letters of the same length can be generated by this
    method. Indeed, if every trial is saved and used as the basis for further
    trials (with most letters in each string-so-far being saved into its
    "offspring," with only one randomly chosen letter-position being allowed
    to vary randomly per trial, etc.), it is clear that the *only* ultimate
    limitation is the number of *possible* different strings.

    Further, this is true even if the strings are allowed to be a few billion
    letters long.

    However, without some kind of selection, the experiment will soon come to a
    halt because, even if it is run on a large computer system, there won't be
    enough disk space for all the variations.

    This is where it matters what standard (actual or implicit) is applied to
    selection. However, interesting results can be achieved even if selection
    is totally *random* on each generation.

    But, more interesting results can be achieved if there is some kind of
    selection "standard" such as is applied by Nature when an organism that is
    not suited to extreme hot climates is born into one, etc. With text, we can
    do this sort of thing by giving letters and letter combinations arbitrary
    phenotypic "meanings," so that each string must "survive" in an
    "environment" which may or may not be compatible with such phenotypic
    characteristics. In this way, we can study evolution without ourselves
    applying any "meaning" to either the phenotypic "meanings" of sub-strings
    or to environmental factors. The whole thing can be set up by a
    randomization process. In fact, we might even set it up so that we could
    not ourselves determine the nature of the environment *except* from the
    results of selection, thus eliminating the possibility for the intrusion of
    "intelligence" beyond the "intelligence" of a natural mechanistic
    unintelligent physical environment.

    As I've pointed out before, the "intelligence" in such a system can easily
    be limited to that which is computationally *necessary* to represent even
    *unintelligent* structures and systems in a computer. In particular, it can
    be excluded from both the variation mechanism and the selection mechanism
    in all ways that would reflect poorly on the results. Since the laws of
    physics are not generally claimed to be identical to the supposed
    "intelligent designer," if we make a selective environment that is no more
    intelligent than rocks and water, it should be acceptable even to ID
    theorists as evidence of the potential power of evolutionary processes.

    >>[...]
    >>
    >>Oh dear, oh dear. When will creationists/IDers stop beating up on this poor
    >>old straw man?
    >>
    >>Dawkins' METHINKITISLIKEAWEASEL model was only intended to demonstrate the
    >>power of cumulative selection vis-a-vis single-step selection. Dawkins made
    >>this perfectly clear at the time (in The Blind Watchmaker), and specifically
    >>pointed out the limitations of the model. (Sorry, I can't quote as I only
    >>had the book on loan from the library.)
    >
    >Since this occurs so often, I thought I would post the appropriate paragraph
    >below. Apologies if someone has already done so.
    >
    >But first a few comments. What is, to me, equally disturbing is that one
    >hardly
    >ever sees an acknowledgement of this error. I hate to brag :), but the only
    >creationist I've ever seen admit to this mistake is me :). I did talk.o
    >several
    >years ago when I was still a creationist. One time I posted this type of
    >criticism of the monkey. Someone, actually rather politely considering it
    >was t.o :), pointed the paragraph out to me and I must say I was tremendously
    >embarrassed. The possibility of just keeping quiet never occurred to me.
    >A public blunder of this magnitude requires (morally, IMHO) a public apology.
    >
    >Here's the quote:
    >
    >====begin quote======================
    >Although the monkey/Shakespeare model
    >is useful for explaining the distinction
    >between single-step selection and cumulative
    >selection, it is misleading in important
    >ways. One of these is that, in each generation
    >of selective 'breeding', the mutant 'progeny'
    >phrases were judged according to the criterion
    >of resemblance to a distant ideal target, the
    >phrase METHINKS IT IS LIKE A WEASEL. Life isn't
    >like that. Evolution has no long-term goal.
    >There is no long-distance target, no final
    >perfection to serve as a criterion for selection,
    >although human vanity cherishes the absurd notion
    >that our species is the final goal of evolution.
    >In real life, the criterion for selection is
    >always short-term, either simple survival or,
    >more generally, reproductive success. If, after
    >the aeons, what looks like progress towards
    >some distant goal seems, with hindsight, to
    >have been achieved, this is always an incidental
    >consequence of many generations of short-term
    >selection. The 'watchmaker' that is cumulative
    >natural selection is blind to the future and has
    >no long-term goal. -- Dawkins TBW
    >=====end quote=========================
    >
    >
    >
    >
    >Brian Harper
    >Associate Professor
    >Mechanical Engineering
    >The Ohio State University
    >"One never knows, do one?"
    >-- Fats Waller



    This archive was generated by hypermail 2b29 : Sat Oct 14 2000 - 03:10:05 EDT