Re: Information request re: Dawkins' "weasel" algorithm

From: Wesley R. Elsberry (welsberr@inia.cls.org)
Date: Tue Oct 10 2000 - 12:35:52 EDT

  • Next message: billwald@juno.com: "Re: (non-flame post) good chess programs intelligent?"

    Richard Wein wrote:

    [...]

    RW>On re-reading, I see that Dembski's description of the weasel
    RW>model is less clear than I first thought. But it can just
    RW>about be reconciled with Dawkins' original.

    I doubt it.

    RW>Correctly described, each randomization of the remaining
    RW>unmatched characters is considered one step, and proceeds
    RW>whether or not any new match was achieved at the last
    RW>step. Dembski has the algorithm randomizing the remaining
    RW>unmatched characters when, and only when, a new match is made.
    RW>This creates following potential problems: (a) One has to
    RW>assume that each randomization (of all remaining unmatched
    RW>characters) is completed before proceeding to check whether
    RW>any new matches have occurred, but this is not clear from
    RW>Dembski's account. (b) Dembski's account implies that if, at
    RW>any stage, randomizing the remaining unmatched characters
    RW>fails to produce another match, then the algorithm ceases;
    RW>this is clearly wrong, but one can assume it's not what
    RW>Dembski intended. (c) It's not clear from Dembski's account
    RW>what constitutes a "step".

    Richard, you are "missing the elephant". Dembski says that
    "correct" letters are immune from mutation; where does Dawkins
    say that? Hint: It is not said anywhere in TBW that "correct"
    letters are immune from mutation, AFAICT.

    RW>I think the issue is one of poor writing on Dembski's part
    RW>rather than a difference in interpretation of Dawkins'
    RW>model. If one knows what Dawkins' really wrote and interprets
    RW>Dembski generously, then there should be no problem.

    There is rather a large conceptual difference between treating
    "correct" letters as immune to mutation and treating all
    letters as equally likely to mutate. I continue to see this
    as a problem.

    Wesley



    This archive was generated by hypermail 2b29 : Tue Oct 10 2000 - 11:49:08 EDT