Re: [tied] The disappearance of *-s -- The saga continues

--- In cybalist@yahoogroups.com, "Mate Kapovic" <mkapovic@...>
wrote:

> But in Slavic monosyllables need not be circumflexed. In Croatian

there is

> aorist bi^ from bi``ti "to be" < *byti (PIE *bhuH-), but aorist

from bi``ti

> "to beat" < *biti (PIE *bhiH-) is bi`` (ra`zbi, u`bi etc.).

I know that, these things got lexicalized, and levelling analogy hit
biti 'beat' and many other, but not some very prominent and
important verbs like biti 'be', dati 'give' and a number of others.
Phonetic rules hhave to made on the basis of the examples that
cannot be analogical, and forms that differ from each other are not
analogically connected.

> So we have
> circumflex in a. p. c where 2./3. sg. of aorist has initial

stress, but we

> have acute in a. p. a where 2/3. sg. of aorist has root stress

same as

> anywhere. So you can't prove anything with this. In Slavic there

is no

> problems with non-circumflexed monosyllables.

No, the rule applies to an older stage.

> P. S. what about Ab. of o-stems in BSl? Why do we have circumflex

here? Why

> doesn't PIE *-o:d yield acute here (yes, I *am* aware of the

difficulties

> with Slavic -a : Lith. -o~).

That ending was apparently disyllabic too. It probably also
contained an a-vowel, given the Baltic timbre; Stang wrote this.

Jens