Re: testability

[Bill Adlam <bill.adlam@st-peters.oxford.ac.uk>]

On Tue, 28 Feb 1995, Lars Bergquist wrote:

>                     ...Galileo was the first to actually *use* the
> hypothetic-deductive method systematically, but the prelates
> insisted that as he could not logically prove the exclusive truth
> of his own (Copernican) position, he had no right to deny the
> truth of the Roman church's (Ptolemaic) world picture. Both
> "saved the appearances", i.e. the old view was not "falsified".

I agree.  Then, and probably until the mid-twentieth century, both 
Ptolemaic and Copernican schemes were entirely consistent with all the
observations.  Galileo preferred the Copernican system because it was 
more parsimonious: it required only a few parameters rather than the
many epicycles needed to make geocentric views fit more and more 
precise measurements.  To put it another way, the Copernican system is 
intrinsically more liable to falsification and so the fact that it 
survived was more significant.

Much the same happens today when reconstructing phylogenies.  All our 
evidence is consistent with the theory (hypothesis if you prefer) that
birds are a subclade of the theropods.  However, it could be that birds
are really the sister group of the mammals, or are polyphyletic - all
we can say is that if they are, many more changes must have occurred.  
Popper implies that any appeal to parsimony is bad science.  I think it
would be fairer to say that it not distinctive of science in the way
that the hypothetico-deductive method is, but nevertheless it is vital to
it (just like logic, maths, clear comunication, etc).

I am reluctant to call Popper an epistemologist because this means he 
was wrong.  Believing everything that could be disproved but hasn't 
been is not a good way of acquiring knowledge.  Describing him as a
philosopher of science is also a more appropriate way of depicting his 
influence on 20th century philosophy.

BTW, I think abolishing the philosophical theory-hypothesis distinction
(although it's artificial) would be a bad idea.  Where would 
macroevolution be if it didn't distinguish patterns from processes?
Maybe I was too hard on the scientists, though, it might have been the 
philosophers who changed the meanings of "theory" and "hypothesis".

                                                        Bill Adlam