On Wed, 22 Nov 2000 23:19:40 +0100, "Piotr Gasiorowski"
<
gpiotr@...> wrote:
>Now the final problem: what evidence do we really have for the phonemic contrast between *h2 and *h3?
>
>How can we tell, for a given form, whether attested *o- or *o: stands for underlying *h3o-/*oh3 < **h3e-/**eh3 rather than *o accompanied by one of the other laryngeals?
I've looked a bit at "drink" and "live" and "know" and such, and on
the whole I'm still convinced that only *h3 can explain forms like
Greek zo:- etc. The thing is of course that most of the data is
irrelevant to the question anyway: the typical IE language does not
distinguish *a[:] from *o[:], and we're basically left with Greek and
Latin and a bit of Armenian. And then only if we can prove e-grade
(possibly zero grade if one accepts the three Greek schwas). And
given that nobody fully understands what o-grade is really all
about[1], that makes it really hard to convince the unfaithful.
I'll see if I can come up with a real good example (apart from
Kartvelian *os'txw-, that is), but I'm not holding my breath.
[1] If you ask me, o-grade is an amalgam of at least three phenomena:
1) sporadic rounding of *e after a labialized consonant, e.g. *h3,
*mw, etc.
2) a very ancient form of vrddhi, back when *e was **a. A lengthened
grade **a: (as in Baltic, Germanic, M.English or Amsterdam Dutch, but
unlike Slavic, Hungarian or Standard Dutch) was backed to /A:/, /O:/,
while short **a was fronted to /ae/, /E/. As per Brugmann's Law, *o
was still long in open syllables [... some other conditions ...] in
Proto-Indo-Iranian. Otherwise, the length was phonologically
redundant (there being no short **o), and the result is PIE *o.
[After palatals [.. some more stuff ...], as in the case of
*ye:kwr(t), **a: was not backed, and remained in opposition with short
*e.]
3) secondary lengthening of reduced **@ (presumed first stage of
Nullstufe) in lengthening environments like the masc. nominative sg.
This **@: merged with **A: and further developed to *o [cases of
poimé:n ~ dáimo:n, etc.]
=======================
Miguel Carrasquer Vidal
mcv@...