On 2007-09-10 03:24, stlatos wrote:
>> Slavic *strumenI (a late transformation of a *-men- stem)
>
> After RUKI (including H) *-mYn.- > *-mn.y- before a V. This happens
> after y>iy after heavy syl., etc., so the cluster is broken by e.
The original Slavic declension was nom.sg. *-my reflecting *-mo:n (OCS
kamy 'stone', plamy 'flame', Polab. komói, Kash. kamë, plamë, widespread
diminutives in -y-kU), acc.sg. *-menI < *-men-m., gen.sg. = nom.pl.
*-mene < *-men-es. The variant *-men-, which appears instead of expected
*-mon- in the acc.sg. and nom.pl. and instead of *-mn- in the weak
cases, came from masculines in *-mé:n/*-m(e)n-, whose declension merged
with that of *-mo:n (Lith. piemuo~/gen. piemen~s and akmuo~/akmen~s).
Nearly all the modern Slavic languages have replaced the exceptional
nom. sg. with the acc.sg. in -menI. The Church Slavic writers of the
11th c. must already have felt the form <kamy> to be archaic/literary
and often misused it as a hypercorrect acc.sg.
The development of *-e:n/*-o:n and *-te:r/*-to:r was similar, resulting
in nom.sg. *-y (from *-o:n) and *-ti (from *-te:r) alternating with
*-en- and *-ter- (elsewhere).
Piotr