> At 04:44 PM 12/13/97 -0500, Lloyd Eby wrote:
>
> >I'm quite familiar with Popper's views on falsifiability. (My Ph.D.
> >dissertation was entitled: "Objective Knowledge and the Knowing Subject:
> >The Popper-Kuhn Debate.") I think the situation with theory-proposal and
> >theory-choice is somewhat diØfferent from what Popper thought. The famous
> >experiment, where light was shown to be bent when it passes a massive
> >object (e.g. the sun), in some sense "proved" Einstein's theory just as
> >much as it falsified Newton's.
>
> Foo on that statement!
>
> PhD in Philosophy or not, I cannot buy into that that one. The fact that
> the outcome of an experiment is correctly predicted can hardly be considered
> to be a "proof". If it were widely accepted as such by scientists in
> general, then the scientific community would call the General Theory of
> Relativity a "fact" or a "law", which they do not. You appear to have
> constructed your own personal set of logic rules and definitions.
Well, for me not much hangs on this particular question. My point was that
I'm willing to expand the notion of "proof" to cover some (a very few?)
cases of what we might call "positive proof for a scientific theory," as
opposed to just the usual notion of proof within a formal system (logic
and some parts of mathematics are formal systems). I'm not especially
interested in defending that view, however, because I don't see much, if
anything, coming out of it -- it would just be another (different)
meaning of "proof." If no one else finds that interesting, I'm perfectly
happy to drop it.
Therre is a far more important objection to Popperian falsificationism,
however, that was raised by Imre Lakatos (and others). That is that all
scientific theories have falsifying evidence, but working scientists
ignore or set aside that falsifying evidence if they think the theory is
true or if it gives them a fruitful research program. Thus, Lakatos moved
to speaking of scientific research programs and trying to distinguish
between good and degenerating ones. That seems to me to be the right way
to go.
Lloyd Eby