I'd say *mori too, if I had to pick one. Since PIE */a/
where not conditioned by */h2/, */k/, */g/ or */gh/, often
seems to be the result of nasalization of */o/, to me this
looks like a case of sporadic nasalization *ma:ri ~ *mã:ri
(> *mori ~ *mari) already in PIE.
--------------------
You say PIE /a/ seems to result from nasalization of /o/, then go on to say that /ma:ri/ was nasalized which could produce either /mori/ or /mari/. So is it /o/ or /a:/ that is nasalized after *m, and
which are you saying was the original vowel in *mori/mari?
>> >L pars < *porti-s
>>
>> From *pr.Htí-, cf. Skt. pu:rtí-. *port- gives port-
>> as in
>> portare, porta, portus, etc.
>
> I'd say that *pr.tí-s > *porti-s > pars. If
>*pr.Htí- then *pra:ti- in L.
Yes. The form is irregular in any case (perhaps *prh1tí- >
*pr.thí-?). I just don't see how it supports a development
por- > par- in the face of clear and regular forms as
portare, port, portus, etc.
------------------
Couldn't [portare], [porta], [portus] come from *pr.ta:-, *pr.ta:, *pr.tus? Compare [portus] with Germanic *furduz from *pr.tus.
Andrew
.
