--- In
cybalist@yahoogroups.com, "P&G" <G&P@...> wrote:
> That flies in the face of everything we know about laryngeals,
doesn't it?
> *romh1C > romaC (or romiC in I-I) - short vowel
> *romh1V > romV regularly
> Only after syllabic consonants (R.H) can we get a long vowel as
the outcome
> (Skt gu:rta etc)
The reconstruction *bhér-o-m&1no-s for Gk. pherómenos and Avestan
bar&mno: /baramna-/ via Proto-Indo-Iranian *bháramïna-s (using
double-dotted i for the vowel arising from schwa in Indo-Iranian) is
regular under Brugmann's original formulation of the law if the
interconsonantal laryngeal is realized phonetically as something
like [h&], i.e. a modulation beginning as a fricative and ending as
a vowel, perhaps a gradual lessening of the frictional noise in the
course of the articulation. That apparently made the first part of
the "schwa" consonantal enough to act as a second consonant causing
[mh] to close the preceding syllable and prevent lengthening of the
[o].
Jens