--- In
cybalist@yahoogroups.com, Kim Bastin <kimb@...> wrote:
> On Mon, 20 Sep 2004 11:37:20 -0000, you wrote:
>
> >
> >What rules out the present participle?
> >
> >Richard.
>
> Plural forms? Even if ama:bam, -a:s, -at somehow continues
> *ama:nts+ba:m etc., it's very hard to get ama:ba:mus -a:tis -ant
from
> *ama:nte:s+ba:mos etc.
ama:ntsbHa:m > ama:nsbHa:m > ama:sbHa:m > ama:ba:m > ama:bam
regentsbHa:m > regensbHa:m > rege:sbHa:m > rege:ba:m > rege:bam
An alternative route might be
regentsbha:m > regensbha:m > regesbha:m > rege:ba:m > rege:bam
In either case we get a long vowel matching the reduced (nasalised)
nominative singular of the participle.
The plural forms are not a problem. Analogy would very rapidly have
reduced ama:ba:t : amante:ba:nt to ama:ba:t : ama:ba:nt.
The problem that I do see is that, although the sound changes I
propose did occur in Latin, I am by no means confident that they
occurred in the right order. On the other hand, given reductions
like *aksla: > a:la 'wing', it seems entirely plausible. But did
post-consonantal IE *bH yield /f/ or /b/ in Latin. Also, were there
not Oscan congates to the Latin imperfect? If there were, we would
have to explain -sbH- > Latin -b- and Oscan -f-. -sn- was certainly
possible in Oscan - Oscan _fíísnam_ and Umbrian _fesnafe_ from *dHes
and related to Latin _fa:num_ and Greek _theos_, unless the Latin
word be derived from Etruscan.
Richard.