20-10-03 23:33, alex wrote:
> Piotr, can it be that this Slavic *kUne~dzI is the one who is the basis
> of later "knezU"> Rom. "cneaz"?
Of course.
> But if this is a Germanic loan then the "dz" is just an ordinary
> palatalisation in Slavic, isn't it?
It's the so-called second palatalisation of velars (more precissely, its
"progressive" part), a rather late phenomenon in Common Slavic. The word
was borrowed into Proto-Slavic before the progressive palatalisation, so
it behaves just like any inherited item. By the same token, *korljI was
borrowed before the subbranch-specific metatheses and pleophonies, so we
get <król>, <král>, <kralj>, <korol>, as if the word dated back to
Proto-Slavic.
> I wonder in fact about how stared some other words remained. For
> instance, Marius mentionated in a previous mail SerboCRoatian "ploska"
> which is a loan from Germanic *flaska ; but what a loan !!! in the time
> as the "p" was stil "p" in Germanic and not "f". The question should be,
> where have the Serbians lived togethere with Germans in that time?
No, it isn't a pre-Grimm loan. Early Slavic languages had no /f/, so
they substituted /p/ for it in loans from German, Latin, etc. We have
the same phenomenon in ecclesiastic loans and borrowed Christian names
in earliest Polish: <lucyper> for <lucifer>, <Pabian> for <Fabianus>,
<Szczepan> for <Stephanus>, <Pabir> for <Faber>, etc. These Polish words
can't be earlier than the 10th c. The substitution of Slavic /o/ for
Germanic short /a/ is also normal in early loans, cf. West Slavic
*kostelU 'strong-walled building' (hence Cz. kostel, Polish kos'ciól/
'church') <-- OHG kastel <-- Lat. castellum.
> BTW serbs.. is this a curiousity:
> serb vs. Latin servus
> slav vs Grk/Lt sklavus
<sclavus> is just a Latin adaptation of the Slavic ethnonym (Latin had
no /sl-/, so /-k-/ was automatically inserted to make the word
pronouncible in terms of Latin phonotactics). <servus> is much too old
in Latin to be a loan from Slavic; in fact, it's an inherited IE item.
*sIrbU, on the other hand, can't reflect Lat. servus (if borrowed early,
it would have given South Slavic *sre^b-). The similarity is coincidental.
Piotr