Re: Evidence and proof; was More on Gosse's OMPHALOS

From: george murphy (gmurphy@raex.com)
Date: Fri Feb 16 2001 - 10:02:21 EST

  • Next message: george murphy: "Re: [METAVIEWS] 008: The Big Tent and the Camel's Nose,by Eugenie Scott"

    Iain Strachan wrote:

    > George wrote:
    > > 1) The "nobody was around to see what happened" argument is
    > valueless.
    > > We receive signals from the past - light, radio &c for astronomers,
    > fossils for
    > > paleontologists, &c. Some theorizing is needed in order to get
    > information from
    > > these signals about their sources - as is the case with _all_
    > > observations. Of course we have fewer signals from events 10^9 years ago
    > than
    > > for comparable events today & the required inferences are more complex,
    > but
    > > that's a matter of degree.
    >
    > It could equally be said, rather than "inferences are more complex", that
    > the degree of speculation and extrapolation required is more extreme. In my
    > (admittedly highly specialised) area of data fitting and neural networks,
    > the less data you have, the less complex model you can get away with, and
    > hence the weaker the predictions you can make with any certainty.
    >
    > ... and before anyone is quick to point it out, I also think that at least
    > some of the arguments advanced by YEC's also involve a ludicrous amount of
    > extrapolation; for example the highly debatable experimental 1% drop in the
    > measured speed of light over the last 300 years leading to the theory that
    > it could have been much faster in the past, compressing the timescale from
    > billions to 6000 years. I think the problem of distant starlight is clearly
    > the most difficult challenge for the YEC position; I know of work that is
    > going on into trying to get a cosmology that accomodates this, but the
    > problem is far from solved.

           1) My basic point, which I will restate as bluntly as possible,
    remains. The "nobody was around to see what happened" (or "we can't see the
    past") argument is utterly inept and cannot be used by anyone who has any
    knowledge of the way the world works. How speculative theories may need to be
    to interpert the signals we receive, and in particular to try to determine how
    far in the past they originated, is another matter.
           2) Many "creationists", & not only those of the YEC variety, don't like
    theories, labelling them speculation, against common sense, &c. But if they try
    to deal with observational data at all they MUST theorize. The problem is that
    they don't know how to do it! A common approach which shows up in many YEC
    claims (e.g., changing speed of light, shrinking sun, decay of earth's magnetic
    field) is to take some data gathered in the past couple of centuries &
    extrapolate that data into the distant past. If you're not careful that may
    mean you're just extrapolating observational uncertainties or errors, which can
    just produce nonsense. (Though sometimes one just gets lucky, as with Lowell's
    prediction of Planet X.)
            The more fundamental error, however, is the failure to realize that what
    they're doing IS theorizing. In order to extrapolate, you have to make a more
    or less intelligent guess about the form
    of the curve you're going to use - linear, exponential, or whatever. & of
    course that's a theory. But just drawing a curve through points on a graph
    without having any ideas about the underlying physical processes isn't likely to
    give any real insight.
            This doesn't mean that extrapolation is always worthless, but you need
    to have some idea how & why to do it. E.g., extrapolation of changes in the
    earth's magnetic field as simple exponential decay would make sense IF the field
    were simply frozen in to a fluid conducting core. In fact, you don't even need
    any data to determine that with such a model the half-life for decay of the
    field would be a few thousand years. But there are good reasons for rejecting
    such a model.
            In other words, if you're going to do theoretical physics, learn to do
    it right. & you can't do that if you're contemptuous of the whole concept of
    scientific theory.

    Shalom,

    George

    George L. Murphy
    http://web.raex.com/~gmurphy/
    "The Science-Theology Interface"



    This archive was generated by hypermail 2b29 : Fri Feb 16 2001 - 09:59:26 EST