🚀 go-pugleaf

In article <30jqdl$fl@Jester.CC.MsState.Edu>,
Frank Peters <fwp@CC.MsState.Edu> wrote:
>Marc A. Belanger <3914MABE@vm1.ulaval.ca> says:
>>I stand corrected. But would you still say that a science has to make
>>predictions in order to be considered scientific?
>
>Pretty much.
>
>In order to be scientific a theory must be falsifiable.  That is, there
>must be some way of demonstrating that the theory is incorrect (it
>being practically impossible to demonstrate that a theory is correct).
>The only way that I can think of to do this is to make predictions of
>the nature of unobserved phenomena based upon the theory.  The more
>unique to your theory the predictions, the better (its not useful, for
>example, to say that your new theory of cosmology predicts that the
>western hemisphere of earth will turn to face the sun at HH:MM GMT
>tomorrow).  The prediction can then be tested.  If it is incorrect then
>your theory is wrong (or at least flawed).  If it is correct then your
>theory is more strongly supported.  Note that the predicitions don't
>have to be about new things, just newly observed things.
>
>If you can think of a way of testing the correctness of a theory which
>does not involve predicitions then I suppose that would serve.
>
>Notice that even wildly incorrect theories are entirely scientific if
>they can be tested.  Is just unscientific to stick to a theory after it
>has been proven incorrect.
>
>Also notice that a theory may be entirely correct but, until it makes
>testable predictions, it isn't scientific.
>

As pointed out by Sean Redmond, de Saussure's "coefficients" were
actually discovered in Hittite.  This is a good example of 
a "testable prediction" in historical linguistics.
But this is very rare indeed.  If the Hittites had never written
anything down, or the if the Hattussas archives had been destroyed,
de Saussure's theory would never have been "tested".  Historical
linguistics can't do "experiments", because it studies a unique
series of events, that has already taken place.  A "re-run" of
the experiment, a re-run of history, would most likely give a
completely different outcome (Stephen Jay Gould has written many
pages about these two "modes" of science, and he can explain it
much better than I can; I refer you to any of his books).

Synchronic linguistics is a different matter, I guess, and I
shall not speak of it.

-- 
Miguel Carrasquer         ____________________  ~~~
Amsterdam                [                  ||]~  
mcv@inter.NL.net         ce .sig n'est pas une .cig