Laryngeals

[Anthony Appleyard]

  Glen Gordon <glengordon01 at hotmail.com> wrote:-
>... A labial quality must be added to explain this - thus *H3 = /h.<w>/
>and it is a labial *H2. ...

  The moderator commented:-
>The evidence adduced in favour of voicing of *H3 is Skt. pibati, Latin bibit
>"he drinks", a reduplicated present from a root reconstructed as *peH3-, in
>combination with Sturtevant's Rule.  A tad thin, I admit. --rma

  Anthony Appleyard replied:-
>In my mouth at least, I find that the /e/ in /h.a/ tries to acquire a distinct
>flavour of /a/, and the /e/ in /3e/ (where /3/ is ayin) tries to acquire a
>distinct flavour of /o/. ...

  The moderator commented:-:
>There exist in the phoneme inventories of several NW Caucasian languages both
>voiceless and voiced labialized pharyngals, which cause rounding in adjacent
>vowels; thus, "o-coloring" need not imply voicing in *H3. --rma

If H3 was /h.w/ or similar, with a labial component, surely in at least one of
PIE's many descendant languages we would find H3 presenting as /w/?

[ Moderator's comment:
  Why?  What would privilege any such language to be preserved?  (Remember that
  the evidence we have for any reconstruction of any protolanguage results from
  the historical accident of preservation.)  We might indeed find such evidence
  at a future date--but we cannot demand it _a priori_.
  --rma ]