Re[2]: [tied] OE "afor"

At 6:16:45 PM on Thursday, December 4, 2003, Piotr
Gasiorowski wrote:

> 04-12-03 21:37, Brian M. Scott wrote:

>> At 2:41:19 PM on Thursday, December 4, 2003, alex wrote:
>>> It seems the *abhros would fit as well the both Germanic
>>> forms; more, it will fit the Rom. form too.

>> It wouldn't fit the Gmc. forms at all. If I'm not mistaken
>> it would yield Gmc. */aBraz/ and OE *afr-, *afor, not
>> <a:for>.

> Or rather OE *<æfer> [æv&r], since Anglo-Frisian
> brightening normally applies in such words (e.g. æcer
> 'acre' < PGmc. *akraz < PIE *h2ag^ros).

Thanks; I'd got the <o> of <a:for> stuck in my head and
was thinking that the back vowel would block AFB, but of
course it wasn't there. But while I'm thinking about it,
why <o> in <a:for> and <e> in <æcer>? Front-back
assimilation? Or is the representation of the unstressed
vowel simply a bit unpredictable?

Brian