Re: Finnish KASKA

On 2008-02-22 07:43, Miguel Carrasquer Vidal wrote:

> Yes. What I have trouble with is seeing this as the result
> of a phonetic development (soundlaw). The thematic vowel is
> /e/ before unvoiced (including final) and /o/ before voiced,
> independent of the stress. We have o-stems stressed on the
> root as well as on the thematic vowel, both in the nominal
> and verbal systems (*bhóros, *bhorós; *bhére-, tudé-). If
> the thematic vowel is e/o when unstressed (as in *bhór-o-s,
> *bhér-e-ti), what room is there for unstressed -i-? There is
> only the possibility that unstressed thematic vowels which
> appear as e/o were originally stressed, and had vr.ddhi of
> the root vowel (which caused the accent to shift back): a
> form like *bhér-e- would then come from original **bha:r-á-.
> I believe that is indeed the explanation for barytone
> e-grade thematic forms. But if we apply this rule
> consistently, it seems as if the whole basis for unstressed
> thematic vowels falls away: at that pre-stage of PIE, the
> definition of the thematic vowel would have included
> carrying the stress.

I'd prefer a solution which would also account for the distribution of
*e and *i in the reduplication syllable. We have *sí-sd-e/o-,
*g^í-g^n(h1)-e/o-, etc., with and *i that _must_ be somehow explained
phonetically. It's quite obvious that the stress was originally on the
thematic vowel (LIV and many authors even reconstruct the PIE stem as
*si-sd-é/ó-, against whatever evidence there is), and that there was a
time when the reduplicated stem was something like **s&-s&d-é-, with the
first reduced vowel two syllables away from the location of stress. I
imagine the reduction of *e/o to *i in compounds to have been something
similar: at some "early PIE" stage predating the retraction of the
accent to pretonic full vowels, stress was already initial in some major
types of compounds, so the second element in the *kóm-moini- type became
destressed, which resulted in the phonetic weakening of the thematic
vowel. The change of **bHe(:)r-é- to *bHér-e- is more recent (as is
**si-sd-é- > *sí-sd-e-), which is why the thematic vowel retains its
full quality.

Piotr