On Sun, 01 Feb 2004 15:42:59 -0500 (EST), "Brent J. Ermlick"
<
brent3@...> wrote:
>Yes, but is the new evidence that Jens mentioned incompatible with any
>of the various proposals? I've seen, for instance, a suggestion that
>the traditional *d, etc, were lax stops. I don't see how combining
>the traditional *t (whatever its assumed realization in any
>particular proposal) with a laryngeal would produce a lax stop.
The glottalic theory does open the possibility of *h3 having a voicing
effect without being voiced itself. If, for instance, it was /?W/
(labialized glottal stop), then *pip?Weti could give *pip'oti, which in
turn would have given (according to the glottalic theory) *pibeti. An
analysis of *h3 as /?W/ has the added advantage of explaining why *h3
doesn't aspirate or pre-aspirate, unlike *h1 (/h/?) and *h2 (/x/?) which
do. If we analyze *h3 as /GW/ (a labialized voiced velar fricative), then
that explains the voicing effect, but I frankly don't understand how
/pipGWeti/ could have given /pibeti/ instead of */pibheti/. It doesn't
make any phonetic sense. At the very least, *h3 must have been a voiced
_uvular_ fricative (/RW/) to explain the lack of aspiration.
=======================
Miguel Carrasquer Vidal
mcv@...