Rig Veda, ta'ntra, nUl, and sUtra

[Dick Oliver]

>> Well, let me put it this way.  If someone shows me something and says that
>> it is one of a kind, and then I find something else that is of the *same*
>> kind and show it to her, I have not merely shown that her argument is
>> *improbable*.  I have shown that it is wrong.  No?

That would depend greatly on what you mean by "same kind." In linguistics,
one can generally only speak of similarities, not identities. Even with
exact cognates, it is always a possibility that they evolved independently
or through a natural association (e.g. "ma" as mother in various languages
may not indicate any linguistic sameness other than a natural association of
that sound with nursing). So I would question whether it is ever possible in
linguistics to "find something else that is of the *same* kind" in the
strictly logical way that you are indicating. We have to remember that
neither linguistics nor history is an experimental science, so we are really
always limited to inductive, not deductive logic.

I would still agree that arguments from uniqueness or universality are less
convincing than arguments of other kinds. But I would have a hard time
believing that any historical or linguistic phenomenon was truly and without
exception unique or universal anyway. When one argues that a word usage or
linguistic element is "unique" to a region or language, what one almost
always means is simply that it appears to occur with a much higher frequency
there. Such claims are always as rough and probabilistic as our knowledge of
history itself, and any refutation of such claims should also be weighed as
probabilistic rather than with strict "true or false" logic.

Going back to the question at hand, some interesting coindidences have been
pointed out. To interpret this as proving or disproving any major theory of
relationship between language groups would be premature to say the least. 


--Dick Oliver ----------
  Cedar Science Center
  Wolcott, Vermont USA