Re: [tied] Germanic decads

On Tue, 27 Mar 2001 07:47:56 +0200, "Piotr Gasiorowski"
<gpiotr@...> wrote:

>These forms are rare and indeed found only in the earliest texts. The idea that <sibunzo> is elliptic for unattested *sibunzo hund = ten heptads

Hmmm, I was wrong about there being no collectives in *-d-...

>and that this reconstructed phrase is parallel to <sibuntehund> goes back to Osthoff and Brugmann;

That's nice to know.

><sibunzo> has also been construed as an ordinal. Either solution implies that this sporadically attested form of the upper decads reflects something unusual and unique to OHG. I prefer the much simpler explanation of A.S.C. Ross: -zo is a reflex of West Germanic *-tox '-ty' ~ *-tux < *-téxu-, the old barytone counterpart of *-tog, with the final *-x exceptionally lost in unstressed position. This provides an exact parallel for my conjectural *sibuntexu.

Your solution is much more palatable than Szemerényi's. But I still
have the feeling that there's a Gpl. hiding in the -e: ~ -o. Besides
the heptads and the decads (*-d), another possible argument is East
Slavic <devjanosto>, if that represents somehow the same construction
as the Germanic forms (*devendU sUto).


=======================
Miguel Carrasquer Vidal
mcv@...