Re: [tied] Re: PIE laringeals

On Sat, 03 Mar 2007 04:57:06 -0000, "etherman23"
<etherman23@...> wrote:

>There are lots of theories and no one really knows. My own views are
>in constant flux. There does seem to be a widespread assumption that
>h3 has a labial element but as you rightly point out labials and
>labio-velars don't cause rounding. I think this very argument is
>evidence against the existence of a labial element. All that is
>necessary is that h3 cause the vowel to become more back. Unmarked
>back vowels are rounded. Backing is therefore likely to be accompanied
>by rounding. Also widespread is the assumption that h3 is voiced. I
>don't think that is necessary for assimilatory voicing to be induced.
>It should be noted that only the stops have phonemic voicing. If the
>Glottalic Hypothesis is correct then the traditional unaspirated
>voiced stops are really just voiceless glottalic stops. This suggests
>that h3 need not be voiced but does cause glottalization. One
>candidate for glottalization and backing of vowels is the glottal
>stop.

I don't think the glottal stop as such causes backing of
vowels.

>So far as I know cuniform does not have a symbol for the glottal
>stop so the symbol for /x/ would have been the closest thing, which is
>why Hittite would have used it.

Assyro-Babylonian cuneiform did introduce special symbols
for /?/ plus vowel. Sometimes the signs corespond to
Sumerian vowel signs (e.g. Sum. ú = Akk. ?u), sometimes to
Sumerian signs with /x/ (e.g. Sum. ah/uh = Akk ?a/?e/?i/?u).

The correspondences of Hittite syllabograms including /x/
with Akkadograms and Sumerograms are:

Hitt Akk Sum
ha ?a4 ha
he hi
hé hí hé
hu [mus^en ("bird")]
Vh ?V ah, uh
hal hal
hap hab
hVr har, àr, hur, ur5
has^
hVt hat.
hul hul
hup hub
luh làh làh, luh
mah mah
s^ah s^ih s^ah
tVh tah

>For h2 I favor /h/ because it induces aspiration in stops in
>Indo-Iranian.

That could also be due to /x/ or /X/. In fact, /kh2/
doesn't give aspirated /k`/ in Armenian, as we would expect
for [kh], but it gives /x/ (as in Balto-Slavic), as we would
expect for [kx] ~ [kX].

>I'm not convinced that h1 even exists. Putting that aside, my best
>guess is /S/ but the evidence for that is mighty weak.

What is that evidence?


=======================
Miguel Carrasquer Vidal
miguelc@...