04-11-03 17:20, alexandru_mg3 wrote:
> As an example : somebody wrote here that alb. "bese:" is a
> loanword in Albanian and we cannot take it into account based on
> its "s", or based on the Pokorny's list etc.....however Hamp takes
> seriously the Albanian "s" problem when he writes regarding "besë" :
[snip]
> So please show me the proof ...I will read, if needed, additional
> stuf on Albanian historical linguistics, before to come back...and if
> your explanation is a clear one I will be the first to change my
> mind...and happy to learn something new...
OK. Alb. besë is not a loan. It's an inherited word. The problem is that
<besë> is a _modern_ word, and Albanian has changed a lot (and sometimes
in strange ways) since the time Thracian was spoken (i.e.,
approximately, from the Roman period). For example, what is now Albanian
<s> used to be an affricate, just as Hamp says -- the kind of affricate
that had resulted from the palatalisation of *kW before front vowels (an
inherited *s would have become Modern Alb. <sh>). Its treatment in
"substratal" words in Romanian and the treatment of some early Slavic
loans in Albanian demonstrate that it was an alveopalatal consonant
(*c^) and that the change *c^ > s occurred later than the first contacts
between Slavic and Albanian (by which time Thracian was no longer spoken
anywhere).
To sum up, <besë> comes from earlier *bec^ë. Both the *e and the *c^ can
come from more than one source, but it's generally accepted that *besë
is an abstract noun derived from the well-known root *bHendH- 'bind'
(metaphorically, , and indeed it can be derived without the slightest
difficulty from *bHéndH-tah2 (*-twah2 would also work; both suffixes
form nomina actionis) > *bént(s)ta: > *bêc^ë > besë (*ê was a nasal
vowel, which developed into <e> before a following *c^ as in *penkWe +
-a: > *pêc^ë > pesë '5'). In other words, the reconstructed ancestor of
<besë> was pronounced more or less like "bencha" (*bêc^ë) in Roman
times. If that's similar to Bessi, I'm similar to Gandhi.
It's hard to exclude that Thracian Bess- also comes from *bHendH- (we
know so little about Thracian that it's hard to exclude anything). But
even if it were, it wouldn't be directly connected with <besë>; it would
only represent an independent development of a Thracian derivative of
the same IE root. Reflexes of *bHendH- are attested widely (cf. Eng.
bind, Skt. bandH-) and the fact that two branches have them doesn't
suggest any special relationship between them. Indeed, if Bess- and
<besë> should be reflexes of the same word (for the sake of the
argument), it would only mean that Thracian was quite different from
Proto-Albanian!
Piotr