From: George Murphy (gmurphy@raex.com)
Date: Wed Oct 15 2003 - 08:06:00 EDT
Walter Hicks wrote:
..................
> If you can tell me a textbook that offers a theory that it clearly falsifiable,
> I will go buy it. And maybe tell me (and others ) what tests could falsify
> those current theories...............................
I'm jumping into the middle here & belatedly at that, but there's a basic issue
that hasn't been addressed. Falsifiability (F) is a useful working criterion for
scientific theories but in the end it isn't adequate.
The idea of F (Popper) is an advance on the naive idea that theories can be
tested simply b "verifying" them. F rules out theories which claim that phenomena are
caused by agents whose _only_ effects are the phenomena in question ("invisible demons"
e.g.) But it runs into problems with more complex theories. The basic difficulty is
that you can always avoid having your theory falsified by introducing additional
elements into the theory.
Case in point. You can maintain (against general relativity) that the geometry
of the world is always strictly Euclidean. That requires additional hypotheses about
effects of gravitation on rulers and clocks - & perhaps changes in these hypotheses as
new data are gathered. But no one can absolutely falsify the claim that the sum of
angles of a triangle is exactly 180 degrees.
A better way of understanding how science works, IMO, is that of Lakatos, which
has been developed by Nancey Murphy (especially in _Theology in the Age of Scientific
Reasoning_) in application to theology. Essentially the idea is this.
You have a "hard core", a theoretical claim at the center which you're going to
try to maintain in the course of investigation. & then there are surrounding theories
forming a "protective belt" which can be modified to protect the hard core. In the
above example, the hard core is strict Euclidean geometry (as a claim about the physical
world, not just math) & the protective belt is made up of ideas about the influence of
masses on rulers & clocks.
So how do you ever decide if a theory is good or bad? It's not a matter of
evaluating a static theory but of an ongoing research program - new observations &
experiments & theoretical development. If your theory can continue to predict "novel
facts" (i.e., those not used in construction of the theory) with no - or slight -
modification of the protective belt then it's a progressive research program. If you
can't predict novel facts & continually have to be changing the protective belt to
shield your hard core from the implications of new data then your research program is
degenerating. & if it keeps on degenerating, eventually most scientists will abandon
it.
One good example is steady state cosmology. In the 40s & 50s it had the merit
of being "more falsifiable" than BB theory because it made unique predictions about
things like counts of radio sources. But those predictions turned out to be wrong, & BB
theory predicted the MWB while classical steady state theory didn't. However, you can
introduce auxiliary hypotheses (our universe is just a fluctuation in the overall state
state, appropritae scattering of starlight by dust to make the MWB) to protect the hard
core of the state state theory - which is that the universe is, on the average,
unchanging - & there are still a few diehards who hold that theory. (Hoyle did up to
his death.) But most cosmologists think that the SS program is degenerating & the BB
progressive.
I've got to get to a meeting & don't have time now to apply this to the
Darwin-Wallace theory of evolution (which, in any case, ought to be done by a
specialist in that area). It seems to me though that it's certainly _more_ progressive
than special creationism or ID.
Shalom,
George
George L. Murphy
gmurphy@raex.com
http://web.raex.com/~gmurphy/
This archive was generated by hypermail 2.1.4 : Wed Oct 15 2003 - 08:07:24 EDT