On 2007-12-12 00:37, stlatos wrote:
> The timing is important. It's not very likely that the borrowing
> occurred before s>s.>x after u/etc. A long o: or O: could easily be
> borrowed as u: after original o: > a:. It could be almost any time
> before U/I > 0 (no in at end of syl., so in > eN).
Proto-Slavic had *au from the merger of *au and *ou. It was later
monophthongised to *u and is spelt as such in the traditional system of
PSl. reconstruction (which is somewhat artificial and anachronistic).
Anyway, any early loan with foreign *au would have Slavic *u. Long
Germanic *o: also ends up as Slavic *u, but -o:- and -o:n- stem
feminines often joined the *u:-declension in Slavic (nom.sg. in -y,
other cases with -Uv- before the declensional ending); a good example of
both treatments in one and the same word is Goth. bo:ka 'letter', pl.
bo:ko:s 'book, scripture' --> Slavic *buky, gen. bukUve.
Piotr