--- In
cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:
> Do you think that this mobile accent in the u-stems was secondary?
I would assume it had come about just like the accentual mobility
seen in other paradigms, including the verb. Wherever a stem was
followed by a flexive with an underlying vowel in it the accent
moved one syllable towards the end of the word. In the case of *-tu-
this works fine, for when added to a root-final consonant this gave
a strem ending in at least three consonants and so caused insertion
of a vowel -e- before the stem-final /w/. Before the working of the
ablaut, the acc. would be *pér-tew-m, while the genitive would be
*per-téw-os with accent advancement. After the ablaut this comes out
as *pér-tu-m, *pr-téw-s.
In the u-stem adjective, however, I really do not know what is going
on. I believe I have some intuitive understanding of ntr. *pélH1-
u 'viel' and of the type *H2ug-ró-s 'strong', but I find it very
difficult to accomodate *plH1-ú-s, gen. *plH1-éw-s in a sensible
system. One might like to regard *plH1-ú- as an end-stressed (i.e.
adjectival) variant of a substantival *pélH1-u-, but then of what is
*H2ug-ró- the end-stssed variant? And if it is not a variant, why
this difference?
Do we have to posit
*bhéng^h-u- => *bhng^h-ú-
*H2éwg-ri-s => *H2ug-ró- ?
And do the substantival forms proceed from older
*bhéng^h-ro- > *bhéng^h-o- > *bhéng^h-u-
*H2éwg-ro- > *H2éwg-ri- ??
The contrastive accent will appear to have worked in a setting:
*bhéng^h-u- => *bheng^h-ú- > *bhng^h-ú-
*H2éwg-ro- => *H2ewg-ró-,
i.e. before the substantival form reduced *-ro- to *-ri.
This could yield a principled account for the difference between -u-
and -ri- of the substantival forms: *-ro- loses its /r/ after roots
that have a sonant of their own; but the thematic vowel itself
undergoes a heavier reduction after a longer sequence, hence -ri-,
but -u; after a light root (without sonant) the reduction of *-ro-
was poserior to the creation of the adjectival counterpart in which
then the suffix *-ró- survives unreduced. This takes stock of the
facts we have, including the substantive type seen in OCS
dIbrI 'depth', Gk. ákris 'peak' which plays a key role in writings
on these matters.
It would mean that the suffix split into *-ro- and *-u- must have
been completed before the production of zero-grade. Therefore it was
not, after all, a heavy clustering like "*bhng^h-ro-", but already
its full-grade prestage *bhéng^h-ro- (or *-ru-?) that lost its -r-
by reduction of a cluster.
It looks strange perhaps that the -u- of the nom.-acc. acts like a
vowel when it receives the accent but as a consonant when the suffix
is expanded to /-ew-/ after a cluster in the weak cases where it
gets the accent as a consequence of accentual mobility. This may
hold the key: The status of the high vowels would seem to have been
unstable.
It would appear that the reduced form of the thematic vowel -u-
acted just like the later sonants: It stays a vowel before a
consonantal ending (-us, -um, -ubhis, -usu), but becomes a sonant -w-
when followed by a vowel. That produced pre-ablaut gen. *bhéng^h-ew-
os which with the second-syllable accent of adjcetives became
*bheng^h-éw-os and ended up IE *bhng^h-éw-s.
The lesson to be learned apears to be that adjectives made by
contrastive accent on the basis of substantives were not mobile, but
columnal, in their accentuation.
This may take some digestion. Where is it wrong? Where coould it be
different and still honour the facts?
Jens