On 04-12-14 13:41, g wrote:
> If such an assumed substratum language (akin to PAlb or in relationship
> with it as a, say, dialect) went through the same sound changes, then
> no wonder that what has resulted in the Romanian vocabulary doesn't
> reflect earlier stadia than the counterpart pieces of vocabulary in
> Albanian. (I haven't studied these correspondences, plus sound changes,
> but I'm skeptical of such words like balaur and mazãre as being simple
> reflexes made out of Alb. bollë and modhullë. If they are though to be
> seen as such, I'd expect them to be extremely old, since I don't think
> that these Alb -o-'s can get so easily -a-'s in Romanian (I'd rather
> expect forms such as *muzure/muzãre/muzrã (or even forms with -d-
> instead of -z-, say *mudurã) and *bulã or *burã).)
The only short vowels in Proto-Albanian were *i, *u, *a (with the
umlauted variant *ä) and the rising diphthong *iä (from "broken" short
*e). To explain the observed patterns of phonemic substitution between
early Albanian and Balkan Romance (both ways) it's enough to assume that
the rounding of stressed *a: had produced Proto-Albanian *O: (a round
low back vowel). The quality of Latin /o, o:/ was intermediate between
*u and *O:, which is why either of them could be substituted for BRom.
*o (with positionally governed preferences). On the other hand, the *O:
of Albanoid loanwords was sufficiently open to fall together with Latin
/a, a:/ rather than /o, o:/ in Proto-Romanian, hence *ra:tja: > *rO:tja
--> rat,ã etc. Perhaps the dialect from which the loans were taken had
delabialised its *O:, just like those American English accents in which
<caught> rhymes with <lot>.
As an alternative, one could entertain the possibility that Romanian
took the loans from a dialect similar to Albanian but without the
labialisation of inherited *a: (cf. the development of Old English /a:/,
which remained unchanged in Northern Middle English but changed into
/O:/ in the Southern and Midland dialects). This, however, is less
likely, since traces of original labiality can be seen e.g. in vatrã
(which reflects *wO:tra with a prothetic glide, presupposing earlier
*O:tra, pace Alexander; the Tosk unrounding of /vo-/ is a late process).
There remains the case of <cioarã> : <sorrë>, which either shows that
the substitution was variable (BRom. *a or *o for PAlb. *O:), or perhaps
owes its rounded vowel to the labialising influence of PAlb. *c^[W] <
*c'W. Cf. Romanian -oara = Hung. -vár.
Piotr