Re: [tied] PIE laryngeals

On 2007-03-02 18:58, C. Darwin Goranson wrote:

> Yay necromancy! But that aside, I have been modifying some of my
> ideas on laryngeal identity, especially after reading some Winifred
> Lehmann. I think the result is part Lehmann, part Rasmussen in its
> nature, with a little bit of Goranson thrown in. ^.^
>
> h1 = either a glottal stop or a glottal fricative. These may have
> been allophonic.
> The voiced version of h1 is a schwa.
>
> h2 = voiceless velar or uvular fricative. I prefer the latter myself,
> but I could understand the possibility of it being velar before high
> vowels, i.e. that h2 is allophonically either a voiceless velar or
> uvular fricative.
> The voiced version of h2 is a lowered schwa, represented as an upside-
> down a.
> Adjacent to h2, *e eventually merged with natural PIE *a, but when
> this happened (Early-, Mid-, Late-, Post-PIE?) is uncertain. I think
> that before the merger, *e around h2 may have become a low front or
> central vowel, and since that wasn't too far different from *a, the
> two became pronounced the same. This would have DEFINITELY happened
> once or before the loss of the laryngeals.
> Perhaps the intervocalic voicing of the h2 in Hittite is present in
> PIE, perhaps not - it's something to seriously look into.

Does anything depend on it?

> h3 = (rounded) voiced velar fricative. It seems likely to have been
> rounded, based on examples of laryngeal hardening in Germanic such as
> *kwikwaz from *gwih3os (lively).

Lat. vi:vus, Skt. ji:vá-, OCS z^ivU, Lith. gy'vas, Gk. zo:ós all require
*gWih3wó-. In an article to be published later this year I derive that
ultimately from reduplicated *gWi-g(W)w-ó- with the dissimilation of *g

> *[G] = *h3. Germanic may have preserved the original formation with a

velar stop. Jens proposed the reverse change of *h3 > pre-Grimm *g > *k
in a 1989 article. In either case the articulation of *h3 and *g must
have been similar, but rounding needn't be assumed.

> The voiced version was a backed rounded version of schwa, represented
> as an o with a central bar.

The voiced version of a voiced sound? (see below).

> Adjacent to h3, *e eventually merged with natural PIE *o, but again,
> when this happened is uncertain. I think that before the merger, *e
> around h3 may have become a mid front rounded vowel; this isn't too
> far from *o, and eventually the two merged. The exact phonetic
> position of this one, admittedly, is probably the weakest of the
> bunch.
>
> Does this sound plausible?

Quite plausible, except that you seem to be using "voiced" with the
meaning normally given to "syllabic" or "vocalised".

piotr