Re: [tied] Verner's Law (Germanic)

On 2006-05-03 18:40, Andrew Jarrette wrote:

> Why then doesn't OE have -/s/ for the nominative singular of
> masculine a-stems, -(/e/)/s/ for the nominative singular of
> masculine i-stems, and -(/u/)/s/ for the nominative singular of
> masculine u-stems? Or is it that there had to originally be a vowel
> following the *-z, which was lost causing new final *-z, then
> devoicing to /s/? (I.e. distinction between old *-z and new *-z
> from *-zV(z))

WGmc. final *z was lost early (except in monosyllabic words, where it
dialectally survived). I propose the following relative chronology of
changes:

(1) Loss of word-final *z
(2) Loss of final vowels in polysyllabic words
(3) Devoicing of word-final fricatives
(4) Rhotacism

OE -s is preserved in the nom.pl. of thematic masculines (which was
originally *-o:s-es > PGmc. *-o:ziz) and the thematic gen.sg.
(originally *-oso > PGmc. *-aza), but not in the reflexes of *-os, *-is,
*-us, etc., where it became PGmc. -z and was dropped at stage (1).

Piotr