Re: [tied] *kuningaz (again)

-----Original Message-----
From: george knysh [mailto:gknysh@...]
Sent: Wednesday, January 30, 2002 4:11 PM
To: cybalist@yahoogroups.com
Subject: Re: [tied] *kuningaz (again)

*****GK: Thank you Piotr (and thanks to Miguel for his
parallel contribution). How does one explain the sound
shift away from "g" in Slavic? Esp. East Slavic. In
the latter, one has both "Varyagi" and "Variazi"
developing from "Vaeringar" (with the nasal loss), but
I have never seen "knyagi" only "knyazi". And the "g"
to "z" may even have antedated the loss of nasals
here, since there is an Arabic text of the 8th c.
which explains that "knez" is the Saqaliba term for
"malik".******

 

 

Strictly speaking there have been two different (though probably ontogenetically related) phonetic processes in Late Proto-Slavic, called the second and the third palatalization of velars, both of which could change */g/ to */dz/.

 

The second palatalization is a change */k/>/c/([ts]), */g/>/dz/, */x/>/s/ or /s^/(West Slavic) before */e^/ ([æ:]~[e.:]) and */i/ if those */e^/ and */i/ are from earlier *[ai]. The /v/ between */k/,*/g/,*/x/ and */e^/,*/i/ sometimes blocked, but sometimes didn't block the second palatalization.

This palatalization succeded in all Slavic dialects but _completely_ failed in Krivichian, which simply retained old */k/,*/g/ and */x/.

Typical examples would be:

*ke^lU 'whole' > OCS,ORuss ce^lU, Krivichian ke^lU

*ge^l- 'strong' > OCS dze^lo, ORuss ze^lo

*xe^rU 'gray' > OCS,ORuss se^rU, Proto-Polish *s^e^rU, Krivichian xe^rU.

*vUlxvi 'sourcerers' > OCS vlUsvi

*gve^zda 'start' > OCS dze^zda, but Proto-Polish *gve^zda

*kve^t- 'colour, flower' > OCS cve^t-, but Ukr. kvit-.

 

The third palatalization is the same change */k/>/c/([ts]), */g/>/dz/, */x/>/s/ or /s^/ but _after_ */i/,*/I/,*/Ir/ or  */e,/ if that */e,/ is from earlier *[in] or *[In]. It has sometimes succeded, sometimes failed, and there are no strict rules to predict that succes or failure. It's often stated, though, that it was blocked by a consonant or a middle or back vowel (first of all, */y/,*/U/, also sometimes probably */u/,*/o/) following */k/,*/g/ or */x/. There are very few exmples of the third palatalization in Krivichian (Standard Old Russian influence?).

 

Typical examples of success would be:

*vUlc^ika 'she-wolf' > vUlc^ica

*liko 'face' > OCS lice

*zIrkalo 'mirror' > OCS zIrcalo

*vIxU 'all'> OCS,ORuss vIsI

 

Failures can be observed in the same positions and even lexemes:

 

*zemjanika 'strawberry' > Russ zeml'anika

*likU 'face' > OCS likU

*zIrkalo 'mirror' > ORuss zIrkalo

*vIxU 'all'> Krivichian vIxU

 

The relative chronology of these processes is unknown, though the Krivichian evidence might be interpreted as pointing to the earlier date for the _third_ palatalization.

 

Now, if you are still following, back to the prince-word.

 

Nom. sing. *kUne,dzI (> Standard Old Russian *kUn'azI)is a (nearly) regular result of the second palatalization applied to earlier *kUnIngU (*/U/ would be expected to block, but probably gave up under the analogical influence of oblique cases, like G. *kUne,dza etc). Even quirky Krivichian, if I recall (I'll check it up in some hours), has kUn'azI.

 

Nom. _plural_ *kUne,dzi could get its */dz/ both from either the second or the third palatalizations depending on which operated earlier.

 

Standard Old Russian Nom. sing. var'agU 'viking' is most likely a Krivichianism (which seemes to hate both palatalizations), or just has been borrowed after the period of the third palatalization's phonological activity, or is just another example of an unconditional failure of the third palatalization.

Standard Old Russian seems not to have Nom. _plural_ **var'agi (only var'azi, its /z/ being a regular result of the second palatalization, i.e., *g before *i<[ai]) -- this form is most likely a scribal artefact of the later times, when analogical levelling (under the influence of, eg, Acc. pl. var'agy) restored etymological *g in that form.

 

In short, both attested ORuss var'agU and unattested **var'azI are OK as to the diachronical phonetical rules operated in Slavic and relate to each other as, eg, Russian zeml'anika and unattested (in Standard Russian), but possible **zeml'anica (cf. devica 'girl').

 

Sergei