RE: loose ends

From: Glenn Morton (glennmorton@entouch.net)
Date: Sun Aug 03 2003 - 10:40:43 EDT

  • Next message: Glenn Morton: "RE: The Diatribe"

    Hi Michael,
    >-----Original Message-----
    >From: asa-owner@lists.calvin.edu [mailto:asa-owner@lists.calvin.edu]On
    Behalf Of Michael
    >Spence
    >Sent: Sunday, August 03, 2003 3:33 AM
    >To: asa@calvin.edu
    >
    >
    >Subject: Re: loose ends I hate to sound smug, and I'm not trying to be, but
    doesn't "whipping
    >up another test" suggest design? When you type something that is
    gibberish, haven't you
    >designed it? It doesn't mean anything, but haven't you designed it, as in
    intentionally
    >brought it to be?

    There are a couple of points here to comment on, and you bring up a good
    point. Of course I have designed the test, and of course I have designed
    the meaningful sentences. but what I haven't claimed is that I can tell
    which sequences are designed and which aren't. That is what Dembski et al
    are really claiming. They assert that they can look at DNA organization,
    whose origin is at question, and tell whether or not it is designed. They
    claim they have a scientific and mathematical technique for indicating
    design. Fine, I am merely offering a test for their methodology. If you did
    the same to me, designed such a test, I would be utterly incapbable of
    telling which sequences were designed and which one weren't. If faced with
    such a test, I would acquiesce and acknowledge that I have no ability to
    detect design. They won't either acknowledge their inability nor will they
    take the test and prove, experimentally, that they can detect design. If
    they can do it with even 95% accuracy, I will become the biggest proponent
    of their case. But they aren't interested in proving their point. They are
    interested in selling books.

    Indeed, Michael,there are in reality fundamental limitations upon our
    ability to detect design in an unknown sequence. There are coding systems
    which have the ex post facto ability to turn any sequence you present me
    with into a fully meaningful English sentence. If you present me with a
    sequence, generated by a random number sequence, and I 'decode' your
    sequence with one of those systems, it would suddenly appear that the random
    number generator produced a semantically meaningful, but coded sentence. I
    did that on this list a couple of years ago. Even Dembski et al would not
    acknowledge that a random number generator can produce meaningful sentences,
    but if I claim that the sequence with an unknown origin is really a code,
    then I can decode it into any meaningful sequence I wish.

    >Even though I understand what you're saying regarding ID, I don't grasp
    >what the test's purpose is, what exactly it really tests, and what it
    reveals besides some
    >interesting linguistic lessons.

    I would suspect that you haven't been a student of information theory. If
    that assumption is correct, I would ask you a question, 'How do you know
    that what the ID group tells you about it is actually correct?' You may
    wish that they be correct, but wishing doesn't make them so.
    The ID group specializes in writing books for people who don't really
    understand the science of information.

    As to the test, when psychics claim to be able to read minds, or hidden
    cards, etc, it is standard operating procedure in science to set up tests.
    They can be double blind or single blind, to test the ability of the psychic
    to accompllish what he claims. What I am trying to do is get one of the ID
    folks to submit to standard scientific methods for claims testing.

    About a century ago, a guy had a horse, Clever Hans, who could do
    mathematics. The guy made lots of money with this horse. People would pay
    to give him math problems and Hans would stamp out the correct answer. Lots
    of people believed in the smart horse. But what was happening, was that the
    trainer knew the answer and was subtly signalling Hans when to stop tapping.
    This is what Dembski is doing, except that the trainer and Hans are rolled
    up into one. Dembski uses previous knowledge to signal himself that this
    sequence is designed. When the trainer is removed, Hans couldn't cipher
    worth a spit. Can Dembski's methodology show us anything about sequences
    whose origin are unknown to him? I say no. The unwillingness to take the
    test clearly shows that they don't really believe their own hype.



    This archive was generated by hypermail 2.1.4 : Sun Aug 03 2003 - 10:41:09 EDT