Re: Chance and Selection

From: Chris Cogan (ccogan@telepath.com)
Date: Fri Dec 01 2000 - 16:49:02 EST

  • Next message: Bertvan@aol.com: "Chance and Selection"

    ><snip>
    >Bertvan:
    >I find it hard to believe any random mutation might improve any delicate
    >piece of machinery, much less a living organism, which is many times more
    >complex than any humanly conceived machine. You wouldn't open up your
    >computer and invite a monkey to rearrange the pieces, expecting a beneficial
    >mutation. You suggest that "given enough time" one might occur. We aren't
    >speaking of infinite time for the evolution of the biosphere. I believe
    >Dembski's augment is that it is mathematically impossible for such complexity
    >to have occurred by chance in some four billion years.

    Chris
    Bad analogy. Here's a better one: You have a billion computers, almost all
    alike, and all designed with a lot more flexibility than today's computers,
    so that, for example, a significantly lowered voltage in one part will not
    normally disable the computer, and a malfunctioning subroutine in the
    software will not usually cause the computer to quit functioning. Further,
    the things that this computer are to do are not precisely defined, as they
    are in current computers. There are no accounting programs that have to be
    right to a penny on billions of dollars. No, in this computer, everything
    is approximate, or is even a kind of computerized "guess." We might imagine
    a game-playing computer, for example that responded in certain general
    ways, by moving an image in certain ways, but, from the untrained user's
    perspective, it would seem to be playing the same way even if it was in
    fact playing somewhat differently.

    Now, you'd like a better design. You get a billion monkeys to sit at a
    keyboard and enter codes that make small changes in parts of the design.
    They can only make relatively small changes, or somewhat larger changes
    *if* these larger changes use components that are already present (but
    perhaps not currently used). Further, the software that processes the
    design changes can only handle changes that fall within certain parameters.
    For example, it could not process a redesign that would, if it worked,
    produce a computer with a fundamentally different instruction-set. But, it
    could handle minor changes, both in the design of the software, and in
    itself. The monkeys can make both kinds of changes, and the changes they
    make may be more or less probable depending partly on the degree of change,
    so that the smallest changes are the most probable, while the larger
    changes are the least probable and happen less often. This can be arranged
    easily by a preprocessor that causes many of the monkey-actions to trigger
    small changes, and relatively and progressively fewer to generate larger
    changes.

    Now, the "language" of the computer and the "design" is such that a great
    many variations on the design are such that at least the specified computer
    can be produced, even if it won't be functional.

    So, we let the monkeys do their thing on a billion computers, and a billion
    new computers are produced based on their (mostly slight) changes. In some
    cases, the change might be only an increase in power supply wattage. In
    other cases, it might be an increase in the amount of memory. In other
    cases, it might be a difference in some software subroutine.

    Now: Is there a reasonable chance that some of these changes might be for
    the better (by *any* standard of "better" that might be reasonably applied
    in such a case)? I think there is.

    What's wrong with your analogy?

    1. It assumes a single, isolated computer and a single monkey. You wouldn't
    want that monkey fiddling with your computer because it's the only computer
    you have, and it is *very* likely that the monkey would fuck it up somehow.
    It is also not being reproduced, so you would lose the only computer you
    have and not get a replacement.

    2. In the real world, nearly all the modifications do not apply to the
    *parent* organism, but to some few or many of the *offspring* organisms. If
    your computer was about to undergo some sort of reproduction, you might
    *well* let the monkey have at it, in the hope that it just *might* do
    something like increase the output wattage of the power supply. Why?
    Because you'd still have the computer you *already* have, *unchanged*.

    3. In the real world, there *are* (typically) billions (or at least *many*
    copies of the organism whose offspring are to have modifications. If you
    had billions of computers, all of them with what you would consider
    weaknesses, you might well, again, let the monkeys have at their designs,
    in hope that at least a few of the resulting changes would be for the better.

    4. In the real, biological world, *many* changes that might be made to an
    organism would be inconsequential. Some would not be, and would be fatal. A
    smaller number would be beneficial. But, *because* computers are designed,
    we design them to fairly rigid specifications because that makes the design
    process a *lot* easier. If we were designing with monkeys, we'd want a
    mechanism that was much more "forgiving" to the "randomness" of the
    monkeys' behavior. We could design such a mechanism, and people are even
    using such design principles to design some things that have to work under
    a wide range of conditions, and with parts missing, etc. But, *mostly*, we
    don't do this because it is easier to ensure that the conditions under
    which things are going to be used are limited (voltage is within a certain
    narrow range, for example), as is temperature, etc.

    5. Chemistry provides *just* such a means of establishing a certain
    *flexibility* with respect to changes. In particular, it is not usually the
    case that a single nucleotide change will prevent the resulting organism
    from forming or from being viable once it does form. This type of mechanism
    has a kind of inherent tolerance for sloppiness, and the DNA mechanism
    allows for redundancy and error correction in cases where it is really
    necessary that something not change to a large degree or frequently.

    There may be other respects in which your analogy misrepresents the true
    situation in genetics, biology generally, and naturalistic evolutionary
    theory generally, but, I trust that these five items are enough to make the
    point that if your argument depends on the kind of thinking represented by
    the analogy/metaphor of the computer and the monkey, it is *way* off the
    mark. It is not even *relevant* because it does not translate even *the*
    main features of the idea of random changes producing beneficial results
    *sometimes*. In short, you have specified an analogy that grossly
    misrepresents the theory you saying you have such a problem with.

    If your analogy *did* truly represent any idea that was central to
    naturalistic evolutionary theory (NET), I'd reject it myself. But the claim
    you are making is one that, as far as I know, no serious NET supporter has
    *ever* claimed as part of his theory. You might "pull a Jones," of course,
    and find some remark made by Margulis or Gould (etc.) and take it out of
    context and make it *appear* to people that they were in fact making such a
    claim, but I challenge you to show both that some serious evolutionist
    *has* made such a claim, and, even more importantly, that such a claim is
    *required* by NET (even if some evolutionist were intemperate enough to
    actually say such a thing, it doesn't refute NET unless you can show that
    it is a crucial part of NET -- which, obviously, it is not). Good luck in
    your search, in both cases.

    You have commented recently that you don't like it when people get angry
    with you because you *disagree* with them. I have yet to see *any* case in
    the past two years where any NET supporter has done so. What some of us
    (including me) *have* gotten angry about are instances like the above
    analogy, in which you misrepresent NET in *fundamental* ways, refute or
    reject that pseudo-evolutionary theory, and then pretend that NET in
    general is like the theory you have shown or claimed to be unacceptable.

    I hope you do not find it odd of me to regard this practice of grossly
    misrepresenting something and then criticizing the theory that you yourself
    have created and doing so *as if* it were in fact NET in general is a
    dishonest practice, especially when it has been pointed out to you
    literally dozens of times that theory you are rejecting is something
    *OTHER* than anything that any serious evolutionist would even *want* to
    support. Perhaps I should not have been so hard on you in the past, but I
    have assumed that you could not possibly be *so* extremely forgetful that
    you would *never* remember the corrections offered to your statements that
    are supposed to represent Darwinism. Another factor that has provoked anger
    is that you make such grossly false statements about Darwinism and NET
    generally with the apparent expectation that your *readers* will be stupid
    enough to accept them mindlessly. Frankly, for someone who has frequently
    complained that people are calling you stupid, this seems hypocritical,
    because presenting such a theory as if it were what NET supporters actually
    *do* support implies that *we* are stupid, stupid enough to accept such a
    stupid theory, and stupid enough to think that what you are treating as NET
    is really essential to NET, when in fact virtually *no* supporters of NET
    support such claims as you attribute to NET generally. Why do you find it
    objectionable that people might become angry at such insults to their
    intelligence and to the millions of hours of hard research and thought that
    has gone into developing NET?

    But, now I must ask you:

    Is there even a *slight* chance that you will remember and at least reduce
    the *frequency* with which you misrepresent Darwinism in the future? Will
    you at least make an *attempt* to represent it honestly and *fairly* before
    you say things like, "You wouldn't open up your computer and invite a
    monkey to rearrange the pieces, expecting a beneficial mutation," as if
    *that* had something to do with evolution? (If you did, you would not make
    such remarks to begin with, because you'd instantly realize that they are
    not representative of *any* main variant of NET).

    Please think about it.

    I sincerely *hope* you are not the sort of person who gets so much fun out
    of misrepresenting your opponents' views and watching their reactions that
    you feel compelled to keep on doing it.

    --Chris



    This archive was generated by hypermail 2b29 : Fri Dec 01 2000 - 17:50:58 EST