Re: Slavic endings / transferred from "contintental Celtic List"

--- In cybalist@yahoogroups.com, "hwhatting" <hwhatting@...> wrote:

>
> --- In continentalceltic@yahoogroups.com, "mcvwxsnl" <mcv@> wrote:
> > For the development of the Slavic endings from PIE, my model

works

> > with just six "Auslautgesetze", in chronological order:
> >
> > 1) raising of circumflex vowels in final syllables (e~ > i:, a~ >
> o:,
> > o~ > u:)
> > 2) raising of back vowels before final nasal (-aN > -uN, -a:N > -
> > o:N, -o:N > -u:N)
> > 3) raising of back vowels before final -h < -s/-s^ (-ah > -uh, -
> a:h
> > > -o:h, -o:h > -u:h)
> > 4) shortening of final diphthongs (-V:R > -VR)
> > 5) lengthening before -Rh in the acc.pl. and o-stem ins.pl.
> > 6) j-umlaut (ja > je, ju > ji, etc.)
> >
> > Since (1), (2) and (3) do not occur in Baltic, and they must

occur

> > before (4) and (5), it follows that the shortening of long
> diphthongs
> > occurred independently in Slavic and Baltic. Since the soundlaw
> > itself is trivial enough (reduction of 3-moraic sequences is

quite

> a
> > natural thing to happen), I don't think that poses much of a
> > problem. It's much more curious that, quite independently from
> > Slavic of course, Sanskrit also has soundlaw (5) [lengthening in
> the
> > acc.pl. and o-stem ins.pl.].
>
> How does the APl. of the stems in -yo- and -ya:- develop under your
> model?

We have *-ju:Nh (c.q. -ju:N, a long nasal vowel) in both case forms
after the first 5 laws. The j-umlaut produces *-jü:N, which undergoes
(probably by lowering to *-jö:N) a special development to *-je:N.
This develops to -jê in North Slavic, -jeN in South Slavic.
There are other curiosities about the j-umlaut.
ja > je, but not jau > ju or jaN > joN
ja:, jo:, je: > je: > ja: (or something like that)
ju(:) > jü(:) > ji(:), but jü:N > jö:N > je:N > jê/jeN

> > > What is your position on the Slavic Nsg. and ASg. of the o-

stems?

> >
> > My position is that -U is the regular outcome of both *-os (*-as
> (3)>
> > *-uh > -U) and *-om (*-am (2)> -uN > -U) [Although the
> accentuation
> > of mobile o-stems suggests that the nominative was in fact
> replaced
> > by the accusative: we should have e.g. N. *snê'gU with neo-acute

vs.

> > A. snê~gU with circumflex]. The o-stem neuter ending -o (Lith. -a
> > ["neuter adjective"]) comes from the pronouns (*-od), via the
> > adjectives (regular -U < *-om is seen in the barytone neuters

such

> as
> > dvorU < *dhw󲯭). The s-stem NAsg. ending -o must also be

analogical.

> >
>
> The neuter s-stems have always been the point that convinced me

that

> PIE *-os > Slavic -U must be analogical. Is there naything speaking
> against the assumption that this development was not a sound law,
> but an analogical development (Nsg / Asg. u-stems -uh / -uN , i-
> stems -ih/-iN, therefore a-stems -a(s/h) / -uN > -uh/-uN) that
> spread from the o-stems to other cases of *-os? Because at some
> point, both *-us and *-as (or, *-uh and *ah, or whatever would be
> the expected form at that time) were possible in a frequently used
> form, and from there this parallelism could spresd to other forms,
> e.g. to the ending of the 1st Pl. In that case, the variation -mU
> vs. -mo in Slavic might represent not two different PIE endings

(say

> *-mos vs. *-mo), but reflect this parallel usage.

But why should such an analogy affect the Dpl. *-mos > -mU?

In any case, I think there is quite a lot of evidence that the
raising of back vowels (a, a:, o:) before -s (> -h) is phonetically
regular. We not only have the cases of *-os > -U (o-stem Nsg., verbal
1pl., general D.pl.), but we also have the a:-stem Apl. *-a:ns > -y,
which requires one extra round of raising compared with a:-stem Asg.
*-a:m > -oN, and the raising in o-stem Ipl. *-ojs > *-u:jh > -y.
Finally, and decisively, we have the pronominal forms *no:s > ny and
*wo:s > vy (cf. dual *noh3 > *no: > na).

I see no other solution than that *-os > -U is regular. That leaves
only s-stem *-os > -o unexplained. There are two ways to explain this
form: (1) analogically, after all the other neuters in -o, or (2)
phonetically. That is in fact Piotr's suggestion, but I think it goes
something like this: if we consider that the raising was produced by
final -h, and that in the NAsg. of the s-stems, belonging to a
paradigm where /s/ was retained in the oblique cases (*slovos,
slovese), final -s may have failed to become -h, that would also
explain the lack of raising. There was a parallel situation in the
phonetic history of English, but I don't recall what it was.