On 2006-04-30 10:48, Benct Philip Jonsson wrote:
> I found
>
> # *anti̯ó- (germ. *andja-) in got. andeis,
> # anord. endir, as. endi, ags. ende m.,
> # ahd.anti, enti m. und n., nhd. Ende;
>
> Which answers the question.
> Sorry to have bothered you all!
Well, it doesn't answer the question. It's mostly a matter of taste
whether one transcribes the PGmc. word as *anDija- or *andija- (sic,
with *-ija- by Siever's Law!), since there was no phonemic contrast
between [D] and [d] at that stage. Actually, the phoneme /d/ in Gothic
stood for phonetic [D] in most cases, and when devoiced word-finally, it
was realised as [T] and spelt <T> (though admittedly _not_ after nasals,
which shows that Gothic had the allophone [d] post-nasally). The fact
that all historically attested Gmc. languages seem to show [nd] doesn't
necessarily mean that this pronunciation is of PGmc. date, since
parallel development in different subbranches is quite thinkable,
especially as the change is of a very natural kind (cf. the distribution
and historical development of voiced stops and fricatives in languages
such as Spanish or Modern Greek).
Piotr