On Wed, 16 Feb 2005 17:18:03 +0100, Miguel Carrasquer
<
mcv@...> wrote:
>I have no clear picture
>in my mind at present as to how and why a.p. a u-stems
>became mobile in Slavic, but I'll have another think about
>it.
One tuna steak with mashed potatoes and guacamole and a good
think later...
First off, there _must_ have been other a.p. a u-stems,
certainly a.p. I (pre-Dybo barytone), but also post-Dybo
(I'm thinking of e.g. stanU and darU, which should have been
just as much prey for Hirt's law as synU).
The only reasonable scenario I can come up with is the
following:
PIE: Pedersen: Hirt:
N *suh1nús > *súh1nus
A *suh1núm > *súh1num
V *suh1neu
G *suh1néus > *súh1neus
D *suh1néwei > *súh1newei
L *suh1né:u > *súh1ne:u
I *suh1núh1 // *suh1numí
N *suh1néwes > *súh1newes
A *suh1núns > *súh1nuns
G *suh1néwom > *suh1newóm
D *suh1númos > *suh1numós
L *suh1núsu > *suh1nusú
I *suh1númi:s > *suh1numí:s
NA *suh1núh1 > *súh1nuh1
GL *suh1néwous > *suh1newóus
DI *suh1númoh3 > *suh1numóh3
Now, despite what happened to z^I"rny, z^I"rnUv-, it's
possible that the above paradigm, which is neither barytone
nor mobile, was maintained as an exceptional case in a
high-frequency word like synU. We can allow for bizarre
exceptions in Proto-Slavic as well.
It's even possible that after Dybo's law, this special
accentuation spread to the select circle of other a.p. a
u-stems as well. We would have had the following three
u-stem paradigms:
ap "a" ap b ap c
sy"nU domU' vIr^xU
sy"nu domu' vIrxu' (vIr^xu)
sy"novi domo'vi vIr^xovi
sy"nu domu' vIrxu'
sy"nUmI domU'mI vIrxUmI'
sy"nove domo've vIr^xove
sy"ny domy' vIr^xy
synovU' domo'vU vIrxovU'
synUmU' domU'mU vIrxUmU'
synUxU' domU'xU vIrxUxU'
synUmi' domU'mi vIrxUmi'
sy"ny domy' vIr^xy
synovu' domo'vu vIrxovu'
synUma' domo'va vIrxUma'
This may have worked fine until the breakdown of the
separate identity of u-stems. When the progressive merger
with the o-stems asserted itself, the krivaja of synU and
that of the mobile o-stems fell together, ictus-wise and in
isolated word forms, if not intonation-wise: barytone
singular + NA pl/du vs. oxytone pl/du oblique. Apparently,
it was enough to cause transferral of a.p. a u-stems to the
mobile class.
I suppose this explanation is basically what Kortlandt had
in mind, except that I don't have to account for Kortlandt's
yer-yer law, which would have retracted the accent to the
root in the Isg. (where analogical retraction is likely in
any case), but more importantly, in the D and L pl.
=======================
Miguel Carrasquer Vidal
mcv@...