Re: [tied] -st

On 04-11-03 11:43, tgpedersen wrote:

> An n-/gn- alternation together with the *s-d- root, aha.

This Slavic <gn-> is completely isolated, and it doesn't alternate with
<n-> within Slavic itself. Coupled with the other Slavic irregularity,
*e^ instead of *I, it strongly suggests some kind of folk-etymological
contamination (in Slavic too initial <gn-> may carry the phonaesthetic
suggestion of pressing or crushing).

> So maybe the
> two phenomena belong to the same donor language, which would be a
> Nordwestblock one (given the place names in -st).

Certainly. A Nordwestblock loan into Armenian and Sanskrit.

> Or a North European loan getting a double whammy of Grimm. Are there
> cognates outside Germanic?

Yes, this time in Celtic, Latin (<turdus>) and Balto-Slavic (OPr.
tresde, Lith. strazdas, Slavic *drozdU. A northern word, definitely, but
then the song thrush is mostly a north European bird.

Piotr