On 2006-04-30 10:11, Benct Philip Jonsson wrote:
> can anyone tell me what happens with a voiced
> fricative resulting from Verner's Law when it
> occurs after a nasal? My hunch is that
> [mB nD NG] were not tolerated, but became
> [mb nd Ng] with voiced stops. I'd appreciate
> if anyone could provide examples, whichever
> way it goes.
Eventually, PGmc *B, *D, *G became stops after nasals, but it's hard to
date the change, as the difference was merely allophonic in early
Germanic. My personal theory is that the change is post-PGmc. (and even
post-PNWGmc. in the lineage of English). I intend to publish the
argument soon, so please don't quote me yet, but the general idea is
that the final syllable in polysyllabic words was apocopated in early
WGmc., and if a final voiced fricative was exposed in word-final
position as a result, it became devoiced and remained devoiced (unless
*-D > -d was restored analogically). The result of devoicing in the case
of *D is *T, not *t, which indicates a fricative pronunciation at least
for this particular point of articulation, e.g. *bHéronti > *beranDi
(cf. Goth. bairand) > *beranD > *beranT > *bera~:T > OE beraT (likewise,
*-z > *-s before rhotacism).
Piotr