> Henry is of course right that this shift cannot be fully explained in
> a short E-mail. But it is interesting that this long A to O shift
> seems to occur rather regularly, including a second time in Hebrew.
> Thus we have the following changes in Hebrew:
>
> 2nd millennium BCE long A becomes (long) O (holem)
> 1st millennium BCE ? some short A's lengthened to long A
> 1st millennium CE ? new long A becomes O in Ashkenazi >
pronunciation
I seem to recall that in some Jewish pronunciation subgroups even
in modern times, the holem was at least occasionally pronounced
as a long a; an uncle of mine pronounced the holem in ywm (day)
as a long a.