Re: A bit of Proto-Slavic relative chronology

From: tarasovass
Message: 13710
Date: 2002-05-14

I repost, because all the c-cedillas (for palatal fricatives) were
"translated" to plain c-s comehow. If I fail this time, here's a
clue: [c] is intended to mean [c-cedilla], the passage:

"which would in turn point to the following processes: *tj > *[tJ] >
(*[c]) > [kJ], *dj > *[dJ] > (*[J-]) > [gJ]."

being the only exception.

> --- In cybalist@..., "Piotr Gasiorowski" <gpiotr@...> wrote:
>
> > Willem R. Vermeer ("The rise of the North Russian dialect of
Common
> Slavic", _Studies in Slavic and General Linguistics_ 8 [1986])
argues
> that the monophthongisation of the Proto-Slavic diphthongs had not
> diffused to (Proto-)Krivichian before the progressive/second
> regressive assimilation (regarded as one and the same change)
> expanded to the Novgorod/Pskov area. Since *ai was still a
diphthong
> with a non-front first element, there was no input at all to the
> "second regressive" part of the assimilation, hence its failure in
> Krivichian. The "progressive" part resulted in a series of new
> palatals that were restricted to stem-final positions and therefore
> marginal to the system (Vermeer treats the Krivichian "c'okanje" as
a
> later innovation, not an archaism). As such, they were vulnerable
to
> analogical elimination. The one that normally survived was *c',
which
> was well entrenched in productive suffixes.
> >
> > Vermeer's analysis is attractive. The only inconsistency I can
see
> in it is the non-occurrence of the progressive assimilation after
old
> *ai and *ei in Krivichian. If they had preserved their final glide,
> it ought to have triggered the palatalisation. One can propose (as
> Vermeer does) that the glide had been lowered and the Proto-
> Krivichian reflexes of the diphthongs at the time of the
> palatalisation were something like *ae, *ee., which may be a little
> ad hoc, though surely not implausible.
> >
> > I haven't got the original article to hand, but I have a PDF copy
> of a more recen article by Vermeer (2000, from _Russian
Linguistics_
> 24), which contains an exposition of his ideas. I am sending it to
> you off-list.
> >
> > Piotr
> >
>

Unfortunately I haven't got your e-mail. If you sent via Yahoo's
server, it probably just failed to deliver, as it usually does. In
case you would decide to retry, my e-mail is:

S.Tarasovas @ sl.lt

(please remove spaces before and after '@').

Vermeer's position, as far as I can evaluate it from your summary,
doesn't look very attractive to me. Eliminating one problem (why the
second regressive assimilation failed) it introduces another (why
the
Proto-Krivichian was late for the train as to the
monophthongization), and there are no extra evidence (apart from the
failed assimilation itself) for such a late monphthongization. While
the (Proto-)Slavic dialects were indeed a bit inconsistent as to the
output of the second wave of assimilations, I've never heard of any
observable inconsistency as to the monophthongization (with the only
exception of vowel+{*m,*n,*l,*r}-type diphthongoids, which are not
proper diphthongs), which would require rather early date for the
process, hardly after 500 AD (when North Krivichians reached the
Novgorod/Pskov area). Probably loans (both into and from Proto-
Slavic) could shed some light on that, but nothing relevant comes to
my mind at the moment.

And I see at least two more inconsistencies in his analysis:
1. North Krivichian _kIrky_ 'church' is left without explanation.
2. the process *ei > (lowering) *ee. > (raising) i(:) is not very
natural, and the process *ai > *ae > (in curcumflexed syllables) i
(:)
is just implausible.

I would dare offer an alternative scenario, but first I would like
to
summarize the results of the consonantal assimilations to the front
vowels and the palatal glide in Krivichian (E -- any front vowel
triggering the first regressive assimilation, I -- any (sequence of)
segment(s) triggering the progressive assimilation, [J] -- voiced
palatal fricative, [G] -- voiced velar fricative, [CJ] --
palatalized
[C]):

1.
*kE, *kj > [tç], *skE, *skj > [çkJ]
*gE, *gj > [J], *zgE, *zgj > [JgJ]
*xE, *xj > [ç]

2.
*tj > [kJ], *stj > [çkJ]
*dj > [gJ], *zdj > [JgJ]
*sj > [xJ] (didn't merge with [ç])
*zj > [GJ] (didn't merge with [J])

*sE > [ç]
*zE > [J]

3.
*Ik > [k]/[tç]
*Ig > [g]/[J]
*Ix > [xJ] (actually we have only one example -- the stem _vIx-_
'all')

It must be noted that I failed to find any evidence of anything
aberrant as to the progressive assimilation in general (except _vIx-
_) and *Ig-assimilation specifically both in Zaliznyak's thesaurus
to the birch bark inscriptions (plus other early documents from the
Novgorod/Pskov areas) and in Nikolaev's papers on contemporary
Russian dialects formed on the North Krivichian substrate, so I have
no idea what makes Vermeer think that "The one that normally
survived
was *c', which was well entrenched in productive suffixes.".

Now the scenario proper.

The progressive assimilation was a rather early process, and its
'stronger' phonetical environment points to that indirectly: it took
_close_ front vowel (*i or *I) to trigger it, (open-)mid *e was not
enough (and *i2 just didn't occur before velars).
Proto-Krivichian hadn't yet acquired [kJ] and [gJ] at that time.
Indeed, Finnish _kaatio_ 'thigh, trouser-leg' point to Proto-
Krivichian *[ga:tja:] (*tj -- still a cluster) or *[ga:tJa:], which
would in turn point to the following processes: *tj > *[tJ] > (*[c])
> [kJ], *dj > *[dJ] > (*[J-]) > [gJ].
Proto-Slavic clusters *skE, *skj (and *zgE, *zgj) probably first
assimilated their velars to the dental *s, yielding *[stJ] (and
*zdJ), before merging with the reflexes of *stj (and *zdj) in [çkJ]
(and [JgJ]) in North Krivichian.
Finding no support in existing phonems, phonetically palatalized *
[kJ] and *[gJ] either remain intact or jumped to the nearest
existing
phonemes -- *tç and *(d)J (rather than *tj or *tJ and *dj or *dJ),
while *[xj] remained intact, because a supporting phoneme -- *xJ (<
*sj) -- already existed.

Later, by the time of the second regressive assimilation, *[tJ] and *
[dJ] had already yielded *[kJ] and *[gJ] (and *[xJ] remained), thus
velars, phonetically palatalized before *e^ and *i2, were
phonologized by a merger to already existing phonemes, forming a
unique feature of the North Krivichian phonemic inventory -- the
/kJ/,/gJ/,/xJ/ series, that triggered some interesting
restuructrings
of the declensional system.

Sergei