--- In
cybalist@yahoogroups.com, "Torsten" <tgpedersen@...> wrote:
> --- In cybalist@yahoogroups.com, "Torsten" <tgpedersen@> wrote:
> > --- In cybalist@yahoogroups.com, "Richard Wordingham" <richard.wordingham@> wrote:
> > >
> > > --- In cybalist@yahoogroups.com, "Torsten" <tgpedersen@> wrote:
> > > > One of those oldest loans listed by Kästner is the Christian
> > > > term msza "mess", which is Pan-Slavic.
> > >
> > > > <snip>
> > >
> > > > The thing that puzzles me is: how would Slavs get the idea to
> > > > end-stress a word borrowed from a language with generalized
> > > > initial stress?
> > >
> > > Wouldn't the yer have been lost regardless of its original
> > > stress?
> > >
> >
> > According to
> > http://en.wikipedia.org/wiki/Havl%C3%ADk%27s_law
> > yes. Of course you can't stress a non-existent vowel, but I
> > haven't seen stress-shift as a consequence of HavlÃÂk described
> > anywhere. Does anybody know?
> >
> I tried to figure out what Miguel thought about it
> On *-a: :
> http://tech.groups.yahoo.com/group/cybalist/message/47648
> 'It was stressed when it was stressed, and unstressed when it
> was unstressed'
Well, obviously we have a problem here, because in neither
OHG missa, mëssa nor
Latin missa
is the final -a stressed. The stress has been moved in the loan process.
> Also:
> http://tech.groups.yahoo.com/group/cybalist/message/47648
> 'In Balto-Slavic, two things happened that caused nom. and
> acc. stress to diverge. In the first place, a posttonic long
> vowel (lengthened grade, not two contracted vowels or a
> vowel + laryngeal) attracted the stress. Secondly, syllabic
> resonants developed into diphthongs R. > iR (uR).'
>
> which doesn't apply here, since it's *-a: < *-ax (*-ah2)
>
I had an idea.
I earlier came up with the idea that there once was a suffix *-aN which, because the instability of the nasal vowel, had alloforms *-a, *-am/-aN, *-o:, *-en/-in-, and that these alloforms were organised by orderly-minded native grammarians into two paradigms
1) the thematic feminine: nom. *-a, acc. *-am/-aN (by imitation of the thematic masculine), later extended with more cases
2) the n-stems: nom. *-o:, oblique stem *-en-/-in- (by imitation of the athematic stems)
This new idea says:
a) The suffix was originally an independent word, hence the resulting unverbation with it could be stressed either on that suffix or on the preceding word
b) alloform condition: when the suffix was stressed, it lost its nasalization.
http://tech.groups.yahoo.com/group/cybalist/message/47648
So
*ak^mó:, but *ák^men- "stone"
and
*galwá:, but *gálwaN "head"
with subsequent stress leveling, as the case may be.
At least one subdialect of Slavic had the confusion between nom. and acc. f. which I assume as the starting point:
Paul Wexler
Explorations in Judeo-Slavic Linguistics
p. 189
'Until detailed descriptions of Jewish-Slavic speech are produced for all the Slavic lands (see the definition of type 4 in section 2 above),201 it is safest to attribute grammatical errors in the Slavic component of these folklore materials to Yiddish substratal influence; an example would be the loss of case in the Slavic component (there is almost no case marking in Yiddish), as in Jew-Uk til`ko pravda skazaty 'only speak the truth' (for stUk pravdu acc sg),...'
Msza "mass" would then have been placed in that f. paradigm like this:
nom. mÑšá
acc. méšaN
of which the latter violates HavlÃk's rule (by which jers to be dropped are counted from the last one), which I therefore assume to be the result of filtering (the 'wrong' forms out), so it was paradigm-regularized out.
Torsten