"centum"/"satem" "exceptions" [was Re: Northwest IE attributes]

[Miguel Carrasquer Vidal]

Stanley Friesen <sarima at friesen.net> wrote:

[**i and **u]
>I agree that some loss is likely.  I was question *total* loss.  At least
>that was how I interpreted your original model, since you discussed loss or
>change to e/o in all environments.

I don't really have a model (yet).  I was merely playing with the
Slavic parallel.

I forgot to mention Class. Armenian, where *i and *u (but not *a,
*e and *o) were dropped in unstressed (pretonic) position, at
least orthographically (in some cases they were still pronounced
as [@]).

=======================
Miguel Carrasquer Vidal
mcv at wxs.nl