Re: MD On Faith - Improbability ?

From: Jim Ledbury (
Date: Sat Oct 16 2004 - 02:12:53 BST

  • Next message: Scott Roberts: "Re: MD On Faith - Improbability ?"

    Mark Steven Heyman wrote:

    >When I say Darwinism, I have been explicitly using it in the sense of
    >evolution solely through chance and natural selection. I don't see
    >how the evidence you describe indicates that evolution comes about
    >through "solely through chance" and natural selection, rather than in
    >some other way.
    >msh says:
    >But isn't the fossil record loaded with examples of biological false
    >starts, goofy non-viable mutations, and dead ends? (I bet Jim can
    >provide plenty of examples.)

    Only after having read about them :-) I try not to get too bogged down
    in details, although I can quote you the structure of all 20 coded amino
    acids and recite the periodic table. My quotes were basic "but
    evolution is only a theory" put-downs without relying on the crudity of
    "but creationism is only a fairy-story" - *not* that I am suggesting
    anyone here is a creationist, but having deducted creation from the
    realm of reasonable explanations you are left with evolution. Quite how
    MoQ meshes with this is also alas completely 'unscientific', made purely
    on aesthetic grounds and also completely retrodictive.

    >If this is so, then wouldn't this be
    >evidence (retrodictive tests, to use Jim's phrase) supporting the
    >idea of random mutation rather than design, which (design) is what
    >I'm assuming you mean by "some other way."
    >As for testability, since when must ALL elements of a scientific
    >theory be immediately testable? General Relativity wasn't completely
    >testable for, what, 10 years before science was able to measure the
    >warp of starlight passing through the Sun's gravitational field.
    >Does this mean the theory wasn't scientific?

    If you want to go further, general relativity (GR) made very few
    testable predictions and almost all its independent retrodictions were
    known about. There are numerous other explanations of the phenomena
    observed (many of which are crank in that they reject the idea of curved
    space-time as an article of faith, the only one which grabs me in anyway
    is the explanation of gravity in terms of interaction with zero point
    radiation and that because of its generality), few of which attain the
    elegance of curved geometry (think ZMM when Phaedrus has his first
    breakdown because he can pose more questions than are testable). It's
    a nice theory, but it is one which is accepted almost purely on
    aesthetic grounds rather than data-specific ones. This is mainly due to
    the fact that gravity is actually very difficult to experiment with due
    to the weakness of this force compared to other so-called fundamental
    forces. Even only now is one of of the sine qua nons of GR starting to
    be tested in terms of frame dragging - the following of non-accelerated
    frames around a rotating body - by the Gravity Probe B
    ( An "experiment" (astrophysics like
    natural history has to rely on what nature has set up for us to see) a
    couple of years ago to try and prove that gravity travelled at the speed
    of light (another believed but untested assumption of GR) was fraught
    with interpretation problems - many scientists concluded that the
    experiment had only uncovered a fancy and inaccurate way of measuring
    the speed of light (sorry, can't pull an url from the top of my
    bookmarks - all I know is it involved how Jupiter's gravity affected the
    observation of certain binary stars). Fact: GR is selected because of
    mathematical elegance not because it ranks high in and passes
    testability. But then all GR's competitors suffer from the same
    problems of testability. The rather large amounts of human labour that
    have to go into their testing (reflected in the price tag) detract
    somewhat from the scientific ideal. My position with particle physics
    is similar: very few theories actually get to go at the bucks of CERN,
    BNL or Fermilab.

    >Besides, what sort of test would one conduct to show that life and
    >consciousness can arise randomly? I suppose we could cook up a
    >primordial soup and bombard it with random flashes of lightning for a
    >million years, then wait around for a few billion years for self-
    >aware life to start tapping on the glass. But even if such an
    >experiment were possible, there would be no scientific defense
    >against the claim that we had in no way excluded the possibility of
    >divine intervention. It's a no win situation for science because,
    >ultimately, the question is not a scientific one.

    :-) Yup. If we had knowledge of say even 10 or even 2 or 3 ecosystems,
    then biology might be said to be testable rather than phenomenological.
    Sadly these seem to occur rarely (or at least observably rarely: I am
    tempted to suspect that there are ecosystems deep within the gas giants
    and possibly even in the interior of the sun, but these involve pure
    speculation as to possible 'chemistry' and inheritance mechanisms and
    are unobservable to current human science and to any science probably
    within my lifetime - I am intrigued what if anything such ecologies if
    technological would or could say about the possibility of life on
    Earth). Nothing about Mars grabs you and says "this is biological"
    (detection of what should be short lived gases like methane and ammonia
    in the Martian atmosphere can be explained away by clathrates -
    unambiguous detection or elimination of biological processes is several
    years and quite a few billions of euros away). Roll on the new year and
    data from the Huygens probe.

    However I think we can possibly do better with the possibility of
    testing that some form of awareness can arise from particular
    configurations of matter in the field of artificial inteligence. If we
    can show that certain forms of hardware demonstrate insights that say
    purely algorithmic approaches do not, then we are a long way forward in
    answering that particular question. This has the virtue of being
    testable and repeatable (subject to the constraints of the global
    budget). The test would be about "non-algorithmic perception", the sort
    of thing that Roger Penrose goes on about - although whether he has hit
    the nail on the correct test or not is a different matter, and I am far
    from agreeing with him about certain of his assertions (I don't think
    the secret for consciousness lies in the solution to quantum gravity for


    MOQ.ORG -
    Mail Archives:
    Aug '98 - Oct '02 -
    Nov '02 Onward -
    MD Queries -

    To unsubscribe from moq_discuss follow the instructions at:

    This archive was generated by hypermail 2.1.5 : Sat Oct 16 2004 - 02:15:19 BST