Miguel Carrasquer wrote:
>>Second, all the old derivatives of *korljI in Slavic have the extension
>>*-ev- (*korljev-IskU 'royal', *korljev-itjI 'prince', etc.), which might
>>be take as indicating an original stem in *-ju- rather than *-jo-. I'm
>>not sure how to explain it.
>
>
> That's an interesting observation. Are there any other
> words which have a compulsory -ev- or -ov- in composition? I
> can't think of any.
Of course old u-stems behave like that:
*medU 'honey, mead' --> *medvInU, *medva, *medv-e^dI;
*domU 'house, home' --> *domovU, *domovInU, *domovIcI, domovIskU,
*domovi 'homewards', *domovIje;
*synU 'son' --> *synovIskU, *synovIcI, *-synoviti.
The variant *-ev- is expected after a palatal, but unfortunately it's
hard to find an example of such a stem, since there was a strong
tendency in early Slavic to eliminate the u-stem nouns as a separate
class, and analogy played havoc with their derivational and inflectional
morphology. The only independent example of a stem in *-ju- that I can
think of is the 'rain' word if one accepts, as I do, Trubachev's
etymology (*dus-dju- > *dUzdjI). Among its archaic derivatives
*dUzdjevInU 'rainy' (cf. Pol. dz.dz.ownica 'earthworm') may indicate an
original u-stem.
> But if *karla- was influenced by *erla-, and the two words
> were in the same semantic sphere, does that not imply that
> we perhaps don't need Charlemagne after all?
There is, apparently, some rather puzzling onomastic evidence of Slavic
*korljI in Greek placenames datable to a period before Charlemagne.
Perhaps some derivative of *karla- was borrowed already into
Proto-Slavic with the meaning 'alderman, village headman' and changed
its meaning to 'king' ca. AD 800 in the wake of encounters with the
Carolingian empire. That would explain the regular metathesis or
pleophony throughout Slavic despite the relatively late historical
connection with Charlemagne.
This still leaves me in the dark as regards the hypothetical preform
*karl-ju-.
Piotr