phonetic resemblances

[H. Mark Hubey]

----------------------------Original message----------------------------
Max W Wheeler wrote:
>
> ----------------------------Original message----------------------------
> On Thu, 28 Jan 1999, H. Mark Hubey wrote:
>
> >
> > Indeed there is no proof outside of mathematics. Even physics cannot
> > "prove" that the sun will rise tomorrow. It's just that its
> > probability is so high that we accept it. The same goes with other
> > "laws" of physics like F=man, pV=nRT, etc. What separates physical
> > sciences from social sciences is the great uncertainty and great
> > complexity of the social sciences. The uncertainty comes from the
> > complexity. Since we don't know the laws, we cannot predict anything
> > and it looks extremely difficult, and it is. Some accept this to
> > mean that it can't be done, but I don't.
>
> Indeed, and for that very reason I carefully avoided the words "prove"
> and "proof" in my posting. I spoke of "confirming" or "refuting" a
> hypothesis. You may say I went too far in saying "refute"; "fails to
> confirm" might be more accurate. But at the point where one's doing that
> sort of hypothesis testing, phonetic resemblance is neither here nor
> there. Systematic correspondence, above the chance level, is what we're
> after.

Actually you brought up something even more interesting. This problem
of what is science has been and still is being debated by many thinkers,
philosophers, epistemologists, Ai researchers, etc. of the past and
present. From what I know of works by people like Chalmers, Feyerabend,
Popper (the few that stand out), the historical development went
something like this:

1. Baconian: the usual data,hypothesis,theory,experiment, proof cycle.
2. Debate over induction (abduction in some cases) and why there can be
no proof. This led to "verificationism" i.e. experiment verifies the
theory.
3. Debated over verificationism leads to confirmationism because
verification
still is based on finite number of data and is still induction.
4. More debates: an experiment cannot confirm either, because it still
leads to
the same problem of induction.
5. Falsificationism: this is the last stop championed by Popper. There
can be
nothing else except falsification of general statements. No theory can
be proven, verified, confirmed, etc. It can only be falsified. So a
theory with lots of good
predictions (confirmation really informally) is assumed to be true until
proven
false.

So, as far as historical linguistics goes, there is no proof of
geneticity, only
degrees of belief. The systematic correspondance of course is another
way of
saying that the odds of it being due to chance is virtually nil. But the
data that is being interpreted and evaluated is based on nothing more
than "phonetic distance" (whose inverse is informally called resemblance
or similarity).


> Max Wheeler
>
> ___________________________________________________________________________
>
> Max W. Wheeler <maxw at cogs.susx.ac.uk>
> School of Cognitive and Computing Sciences
> University of Sussex, Falmer, Brighton BN1 9QH, UK
> Tel: +44 (0)1273 678975; fax: +44 (0)1273 671320
> ___________________________________________________________________________

--
Best Regards,
Mark
-==-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
hubeyh at montclair.edu =-=-=-= http://www.csam.montclair.edu/~hubey
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=