From: Sergejus Tarasovas
Message: 7799
Date: 2001-07-04
> How about [stressed a] > o, [unstressed a] > [schwa]/[e]? One couldalso imagine intermediate stages such as:
>Looks neatly (*& ~ *e distribution is still to be explained somehow).
> **รก > **a: > *o
> **a > **a > *& ~ *e (also subject to colouring by *h2 and *h3)