Latin pinso etc. (was: *(H-)p/bh[-r/l-] again again)

From: Richard Wordingham
Message: 29552
Date: 2004-01-14

--- In cybalist@yahoogroups.com, "alex" <alxmoeller@...> wrote:
> Richard Wordingham wrote:
> > --- In cybalist@yahoogroups.com, "Marco Moretti"
> > <marcomoretti69@...> wrote:
> >
> >> Now I make some search on Pokorny IE Wo"rterbuch, and I found
*peit-
> >> "fat" > "nourishment" (German Nahrung).
> >> I was definitely wrong in quoting Greek /pi:^sos/ as "pea".
> >> It is /pi/sos/ with short /o/, while /pi:^sos/ has the meaning
> >> of "swamp, bog". I beg your pardon. But in my Latin vocabulary I
> >> found /pi:sum/ with long /i:/. Probably it is a printing mistake.
> >
> > If so, it is an authoritative mistake. Charlton and Lewis (as
> > quoted at http://www.perseus.tufts.edu/cgi-bin/ptext?doc=Perseus%
> > 3Atext%3A1999.04.0059%3Aentry%3D%2336467 ) gives it with a long
> > vowel. French _pois_ does argue for a short vowel.
> >
> > Richard.
>
>
> it seems there are some another root here. If one thinks at the
meaning
> "peas", "fat" one will think at Rom."pãsat" as well.
> pasat= food of corns of millet, cf DEX < Latin "Pinsatum"

This is a past participle of one of the verbs in the family of _pi:ns-
o:, -ere_ 'grind'. Pokorny interprets this as part of a large family
of forms from PIE *pis 'grind' - _pi:(n)s-o:, ere_ (form with nasal
infix), _pis-o:, -a:re_ (zero grade - cf. _-dica:re_ and
_di:cere_ 'say', and _-iuga:re_ and _iungere_ 'join'), _pins-o:, -
a:re_ (intensive form, from the past participle _pinsus_) and
_pinsio_, which I believe is evidenced only by an irregular imperfect.

It is possible that Pokorny has overestimated the number of forms
actually used. Certainly it seems a longer list than I could find at
the Perseus on-line dictionary. There is an attested supine
<pisatum> (or equivalent), presumably with short /i/ and long /a:/.
It is not the usual supine - commoner forms are _pinsum_ and
_pistum_. This would give Romanian _pãsat_. A Classical form
<pinsatum> (whose existence I am not sure of) would yield Romanian
*pisat.

What other evidence is there for <pisatum> having short /i/?

Richard.