Re: [tied] Enclosed Places (was: The unexplained link between Greek

12-07-03 22:57, alex wrote:

> I guess here for this word is not important the metathesis in Slavic.
> Here is something else important Piotr. The /a/>/o/ in Slavic.Neither
> Albanian nor Romanian has the /o/ there but simply /a/. If this should
> be a Slavic loan, then you have to sustain the change /a/>/o/ in Slavic
> is a very recent thing, meaning somewhere in the X century. Is this
> change /a/>/o/ so late or is this soundshift somehow early to be dated?

In earliest Slavic the pronunciation of what we _spell_ *o was close to
[a] (this was in fact the original Balto-Slavic value of the vowel). I
wrote about that at length explaining liquid metathesis and the
etymology of <daltë> (check the archive if you don't remember the
details). The change of [a] > [o] is not as recent as you suggest, but
it probably swept through Slavic about AD 800 (only a very rough dating
can be suggested) as part of a more general qualitative "vowel shift".
If *[dalta] was borrowed as <daltë>/<daltã>, there's nothing irregular
about *[gardU] --> <gardh>/<gard>.

Piotr