real proto-lang

[Ante Aikio]

On Fri, 20 Jul 2001, petegray wrote:

> I would make the abstractionist position even thinner.   It is not the final
> result of application of the method, but merely a description of the
> attested state of affairs.   I think it is application of the method which
> moves beyond mere description to posit an original sound or root.

> Thus saying that the symbol *bh is a code for the correspondences in
> attested IE languages is mere description plus label;   whereas positing an
> actual sound *bh is application of the method.

I don't think I can fully agree with you here: it seems that the
abstractionist position cannot be merely "a description of the attested
state of affairs", because it involves etymological claims (X and Y are
cognate etc.). No sound correspondences can be established even in the
abstractionist framework through mere description; this is, of course,
what the method is for.

> Interpretation comes in
> when we then try to make sense of that sound, and say either that PIE needed
> to have *ph as well, or that the sound we write *bh was breathy voice, or
> allophonically aspirated, or a fricative or whatever.

> There are thus (for me) three stages, and the refusal to move from the first
> to the second seems an unnecessary limitation on what we can claim to  know.

Of course, I agree. I don't think any serious comparativist would really
refuse to move beyond the first stage; but the question is rather what the
comparative method does. It seems to me that the method, in the strict
sense, is only applicable in the first stage and everything beyond that
has actually nothing to do with the method itself but merely with the
result.

-----

A concrete example also comes into mind here. In 1981, several problematic
correspondences (concerning e.g. the origin of long vowels in Finnic, and
certain words showing a lengthened vowel before intervocalic *k in Samic)
between the various branches of Uralic were solved by Juha Janhunen. The
exact details are not relevant here, but the solution involved positing a
new Proto-Uralic consonant, which has not been preserved unchanged in any
branch. The consonant was symbolized with *x by Janhunen, as its exact
phonetic nature is unknown. (The paper appeared in Journal de la Société
Finno-Ougrienne 77, but it is in Finnish.)

I'm sure you wouldn't consider this explanation merely a description of the
attested state of affairs, because no phonetic value was suggested for *x?

regards,
Ante Aikio