> Not if we assume that the effect of /pr-/ is similar to that of /p-/,
> which retracts an unstressed *e (as well as *ae, *i, e:, *oe) to ã, cf.
> <pãcat>. Cf. also Rom. prãdá < praeda:re.
>
> Piotr
In modern Romanian, in many cases the plural constructions
oscillate. E.g., in modern Romanian, another possible plural variant
for pradã would be *prade. Yet, at the level of subdialects
(or "low classes"), a -ã-i plural is perceived as more traditional
than the -a-e one (and the latter has been reinforced by the
standardization of language in the last 150 years or so).
It is a popular tendency in all Romanian subdialects to generate -ã-i
plurals instead of -a-e (and -e-e) ones. (Cf. below cãSi, cãmãSi)
A few examples:
stradã, strãzi, never *strade (although pretty young in the
vocabulary, as
compared with many other words of this category)
cadã, cãzi, never *cade
ladã, lãzi, never *lade
vamã, vãmi, never *vame
canã, cãni, not *cane
mânã, mâni, not *mâne
sãptãmânã, sãptãmâni, not *sãptãmâne
vargã, 1. vergi, 2. vãrgi, never *varge
baltã, bãltzi, never *balte
barcã, bãrci, never *barce, but: barjã, barje, never *bãrji
barbã, bãrbi, never *barbe
casã, cã$i (regional & colloquial)
casã, case (standard)
cãmaSã, cãmãSi, no *cãmaSe
(regional: cãmaSe, cãmãSi / chimeSe & câmeSe, chimeSi & câmeSi)
cioarã, ciori (occasionlly, colloquially or jokingly cioare)
vioarã, viori
roatã, rotzi & (older) roate
Scoalã, Scoli & (older) Scoale
but: mamã, mame, not *mãmi; mumã, mume, not *mumi (but: maicã, maici,
and also occasionally maice; ditto with u, but no *muice); babã, babe,
not *bãbi (although, if insisted upon, one would get accustomed to this
one in a couple of years :-)); lamã, lame, not *lãmi; damã, dame, not
*dãmi; vadrã, vedre, not *vadre or *vãdri; scamã, scame, not *scãmi;
bamã, bame, not *bãmi; vacã, vaci, not *vãci or *vace; labã, labe (lãbi
might be there as regional variant, I ain't sure); palmã, palme (pãlmi
as regional var., I'm sure)
• gãluScã, gãluSte, but regional & popular also gãluSti [g@-'luSt‘]
• cireaSã, cireSe, but also cireSi, although this popular plural also
means the cherry *tree*
• cireadã, cirezi, no *cirede, *cirãzi
• gireadã, girezi, no *girede, *girãzi
• muscã, muSte, but (reg. & pop.) also muSti (In regions of
Transylvania, this plural without the fem. article has the same
pronunciation as the word "muscles" and the one meaning "you bite;
you're biting; you should bite" -- namely with the t uttered with a
palatal [k‘].)
#
treabã, trebi & treburi: there are subdialectal regions (esp. in
Transylvania) where the verb a trebui in the indicative sg. 3rd pers.
treabã is pronounced trabã or trãbã; yet never the noun treabã (some
people in other regions, incl. Bucharest assume that this one is also
de-diphtongized). Moreover, the colloquial and regional tendency to
diphtongize drac ("devil") esp. in Bucharest and surrounding counties
is also and to the same extent there esp. in West-Transylvanian
subdialects. This illustrates that one can never 100% rely on the
e<->ea<->a theories either. (Transylvania, subdialectally, is still the
most active converter of e and ea into a -- also because the
subdialects there have an additional "open" [e], which is unknown
beyond the Carpathian range in the South and East, and which, of
course, is closer to [a] -- in a similar way as it is the "a" in has,
have, hat, at, as; also similar to the [a] pronounced in
Hamburg-Deutsch.)
George
PS: the word entered the Hung. vocabulary as préda ['pre:d@] (I don't
know when and whether from late Latin or so)