--- In
cybalist@yahoogroups.com, "Piotr Gasiorowski" wrote:
>> Is there anything to support this assumption [*-c^ --> -s^]?
>> I never met that in historical Romanian phonetics.
>
> I don't know. One would have to examine the behaviour of
> sufficiently early loans with final /-c^/. The
> sixty-four-thousand-dollar question is: is there
> any independent material to examine? Perhaps some
> really old loans from Slavic would do, but I can't think
> of any obvious example right now. Maybe later.
In the list of the first Slavic loanwords in Romanian (those
shared with similar phonetism by Daco-Romanian and Aromanian)
there is only one with final /-c^/, "kolac^I" which got early
regularised by alternance analogy both in DR and AR dialects,
reconstructing a final /-k/ for the singular: "colac" (DR),
"culac" (AR), same phenomenon which occured to substratal "copac"
for an older "copaci(u)" (but not to monosyllabic "baci(u)").
On another hand, there is to wonder why the ancient form of
modern "atunci" < Lat. "ad tuncce" or "aici (aice)" < "ad hicce"
did not underwent this transformation if it really happened.
Note that in Balkan Romance the phonemes /c^/ and /s^/ made
their "entry" quite simultaneously, so the hypothesis that
some final /-c^/ from a supposed loanword would have been
reduced to /-s^/ rather than being preserved as such doesn't
look very convincing from this point of view.
Regards,
Marius Iacomi