Re: [tied] Definite adjectives: correction

On Tue, 16 Dec 2003 06:55:44 -0500 (EST), "Brent J. Ermlick"
<brent3@...> wrote:

>In article <local.pie/tovttv4ieq84c06es79vjjib8q7rm81ao0@4ax.com> Miguel
>Carrasquer <mcv@...> wrote:
>. . .
>> initially (anti-hiatus *h-) the normal result is -0-, but medially
>> (anti-hiatus *-oho-, *-eho-) we find -g- (> -h- in Upper Sorbian, Czech,
>> Slovak, Belorussian, Ukrainian, Southern Russian) and -v- (Northern
>> Russian).
>
>But in Upper Sorbian, Czech, etc, all /g/ -> /h/.

As you can see above, I wasn't denying that. However, I do think some of
the evidence for initial h- before o-/oN- (sometimes a-, e-, etc) could
indicate that h- may have survived just long enough to make it to the g > h
development, and therefore survives mainly in languages that have such a
new /h/ (Ukrainian, Belorussian, Sorbian [although h- is most popular in
Lower Sorbian, where g = g]).

>I can't get to
>my Czech historical grammar right now, but I seem to remember Czech
>words occurring in early Latin prose (10th & 11th century) in Bohemia
>with "g" rather than "h". This would imply (but not require) the
>change of *g -> *h in the historic period.
>
>I also don't recall any other instances where PIE or pre-Slavic
>medial *s develops into Slavic _g_.

No, medial -s- remains as -s-. The Slavic (as well as Lithuanian) genitive
[of the o-stems and pronouns, of course] derives from the IE ablative. The
only exception is <c^eso> < *kWesyo, a good example of medial /s/ remaining
as such.

My point was that the ablative (pronominal *-eo, *-oo; o-stem *-oo) was
still uncontracted in Balto-Slavic, which explains why Lith. has -o~
(contracted *-o: would have given *-uo~) and Slavic inserted a filler /h/
to break the hiatus.

>I see what you're thinking, but I have reservations about a phonetic
>change that only occurs in one unique grammatical environment.

Sometimes that can't be helped, take for instance the unique development in
the o-stem dat.sg. (*-o:i > Lith. -ui, Slavic -u).

=======================
Miguel Carrasquer Vidal
mcv@...