--- In cybalist@..., Miguel Carrasquer Vidal <mcv@...> wrote:
> But I
> find it hard to imagine that a law that should have been regularly
> blocked in the Nom. and Acc. masc. (-U) would have come through the
> paradigmatic Ausgleich with the success it evidently did have
(outside
> Krivichian, that is).
>
1. Some clarification on Krivichian and
2. Some (rather vague) second thoughts on George's request.
1. I've just re-read the two papers on Krivichian phonology and
morphology I have and inventorized nearly all the examples relevant
to the regressive palatalization. While the (second) _progressive_
palatalization (before */e^/ and */i2/) seemes to fail completely
both in stems and inflexional endings*, the regressive palatalization
demonstrates the normal Late Common-Slavic chaos** observed in other
Slavic idioms, vIxI being the only non-standard feature -- so the
(third) regressive palatalization _hasn't failed_***.
2. Old Russian indeed shows only unpalatalized variants in such
Scandinavian borrowings as kUlb'agU 'viking' (cf. also toponym
Kolb'agi), *bur'agU 'viking?' (reconstructed from toponyms like
Bur'agi etc), s^el'agU and s^c^Il'agU 'a money unit' (cf. Old Church
Slavonic skUle,dzI and Russian Church Slavonic stIl'azI, s^c^l'azI --
South Slavic seemes to have palatalized). This could well be
explained without Krivichian (which has clear examples of succeded
regressive palatalization) -- eg, as a result of a rather late
borrowing. But, interestingly enough, all of the above words could
well be borrowed via Krivichian: toponyms Kolb'agi and Bur'agi are
registered on the former Krivichian territory only
(masc.N.pl.{kolob'agI} /kUlUb'age^/**** is attested in birch bark
inscriptions), and due to their geographical position Krivichians
were the first candidates to hear Scandinavian prototypes of
s^el'agU/s^c^Il'agU. All that provokes to explain East Slavic
(actually, Standard Old Russian + Krivichian) resistance to the
regressive palatalizations in late borrowings in the following way:
by the time of the borrowings the regressive palatalization ceased to
operate in Krivichian (but continued to opeate in other dialects)
because of lack of support across paradigm, triggered by failed
(second) progressive palatalization (so o-stems L.sg,
N.pl.,Acc.pl,Ins.pl left unpalatalized), and the borrowings in
question came to East Slavic via Krivichian. This could also explain
forms like N.pl. var'agi mentioned by George (though I haven't come
across them) as Krivichianisms.
---------------------------------
* thus N.pl. of Kriv. var'age 'viking' (-e in N.sg. of -o-stems!)
would be var'age^ (with -e^ rather than -i, which, IMO, could point
to the fact that in some Slavic dialects circumflexed *-oi in auslaut
merged to /e^/ rather than /i/.
** or indeed reflexes various now 'invisible' stress patterns plus
some analogical levelling
*** this could indicate that the regressive palatalization preceded
the second progressive one
**** sic! 1. the graphemes {I} and {e^}, as well as {U} and {o} are
interchangeable in birch bark inscriptions 2. *UrT yields *UrUT in
Krivichian