On Mon, 17 Jan 2005 14:14:52 +0000, Thomas Olander
<
olander@...> wrote:
>--- In cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:
>> But obviously Kortlandt doesn't think it didn't work,
>> otherwise he'd just stated that, and nothing more needs to
>> be said about Olander's theory.
>
>Sorry, just a few words more...
>Kortlandt's article in Baltu Filolog'ija touches upon a relevant problem for the idea that the
>seemingly final (not thematic) stress of the ending-stressed word-forms of the Lithuanian
>and Slavic mobile paradigms is due to Saussure's and Dybo's laws. Jens also mentioned it
>when he read my article in his festschrift. This is the mobility of Pre-BSl. *suhnús in
>Lithuanian and Slavic. If this paradigm was thematically stressed when Hirt's law worked,
>we would expect all forms of the word to have root-stress.
Which they have, modulo secondary mobilization.
My point against Kortlandt was that if we assume original
_final_ stress in the mobile paradigms (the "traditional"
view), we'd _also_ get root-stress in words affected by
Hirt's law, if not directly, then by analogy.
What Kortlandt seems to be implying is that there was (until
when?) a kind of "mixed" accent paradigm (in the i- and
u-stems at least) with barytonic singular (except the
ins.sg.) and oxytonic plural oblique, distinct from both
a.p. a (< original PIE u-stem barytones) and a.p. c (< PIE
u-stem oxytones unaffected by Hirt's law). I don't think
there ever was such a paradigm. Perhaps a case could be
made that the expected forms after strict application of
Hirt's law (basically: a.p. a singular, a.p. c plural) led
to a situation where *su:nus could be either classified as
(a) or (c), but given that _all_ a.p. (a) u-stems have
become mobile in Slavic, not just the ones with Hirt, I
doubt it.
>I explain the mobility in this word in the same way as Mate does (# 35909) - Dauks^a and
>other Old Lithuanians (plus Illich-Svitych 1963: 75-76, p. 59 in the translation) actually do
>vote for original root-stress in this word. The mobility in Modern Lithuanian might not be
>original; spread of mobility is a phenomenon we've heard of before in this language. For
>the Slavic mobility, as Mate said, at some point u-stems seem to have become mobile.
In sum, I'd say that *su:nus cannot be used to falsify or
prove the "thematic stress" theory.
In fact, I can't of anything right now that *can* falsify
it. As you say:
>It's just funny that our well-known
>accent laws would yield the same result and, as a bonus, they solve the riddle of the
>thematic stress of the dative plural.
Leaving the question of the dat.pl. aside for now (I just
paid E 3,= to ILL for Kazlauskas' article), it *is* funny
that the same results can be reached in multiple ways:
Pedersen-Hirt works, but so does Olander-Dybo. I would
consider the second route to be less robust (it crucially
depends on Meillet's law to create the enclimomena before
Dybo's law can act upon them), but I cannot falsify it.
However, having now read "The Ending-Stressed Word-Forms of
the Baltic and Slavic Mobile Paradigms", there's another
point I'd like to make, concerning the motivation for
"theme-stress" and its PIE roots.
An important line of argumentation in the paper is that
assuming stress on the theme vowel (before Dybo's law in
Slavic, and Saussure's law in Lithuanian, of course) renders
the Balto-Slavic forms more in line with the original PIE
accentuation of vowel stems (e.g. Skt. pl. obl. priyá:n.a:m,
priyébhyas, priyés.u, priyáis (priyébhis)). That seems to
be accurate (up to a point, see below), but I have two
objections:
(1) the Balto-Slavic innovation is precisely that columnar
stress in the vowel stems was replaced by mobility. The BS
ictus *should* and *does* differ from the columnar PIE
pattern in vowel stems. If the oxytone vowel stems mimicked
the athematic mobile retraction of the accent in the
accusative singular (as well as in the dative/locative),
something that remains undisputed, then what we would expect
in the plural/dual oblique is also an accent shift, this
time forward, towards the stressed endings of mobile
athematic paradigms: *-ó(:)m, *-mós, *-sú, *-mí:s, dual
*-óu(s), *-móh3.
(2) And we would expect the same forward shift in the verb,
analogically after the athematic 1/2pl. endings *-mé(s),
*-té(s) [or whatever they were in PBS, I don't want to go
into that now]. In fact, in the verb, that is the *only*
shift that occurred, at first. And this is where I think the
argument breaks down that assuming theme-stress in the
(Balto-)Slavic mobile paradigms allows us to maintain the
PIE place of the ictus. The thematic verbs that become
mobile in (Balto-)Slavic are the _barytones_, with PIE
accent on the root: *bhéro:, *bhéresi, *bhéreti, *bhéromes,
*bhéretes, *bhéronti. Nothing is to be gained,
continuity-wise, by assuming PBS theme-stress here. There
never was a paradigm +bheró:, +bherési, +bheréti, +bherómes,
+bherétes, +bherónti, which is neither PIE, nor PBS. The
PBS paradigm was initially *bhéro:, *bhéresi, *bhéreti,
*bheromés, *bheretés, *bherónti, with mobility applied to
the _plural_.
As I've said here before, I certainly believe PBS *had*
theme-stressed paradigms, corresponding to PIE theme-stress.
In Slavic these are basically the verbs in -né-, -jé-, -í:-
(classes II, III, and IV), albeit with numerous posterior
rearrangements.
Now in the 3pl. of mobile verbs, I can certainly see how
this novel idea of applying Dybo's law to mobile paradigms
may open a promising pathway. The accentuation -oNtÍ makes
no sense from a PIE point of view. Thematic 3pl. *bhéronti
would have given *bherónti (after athematic *h1sénti), but
never *bherontí. A shift *berónti > *berontí caused by
Dybo's law makes perfect sense (and the same might then also
be acceptable, in principle, in the nominal paradigms).
With 2/3sg. beres^í, beretÍ (and athematic 1sg. esmÍ, but
*not* thematic béroN), things are more complicated. One
would like to know where this bizarre accentuation came
from, but applying Dybo's law doesn't help and is not an
option.
=======================
Miguel Carrasquer Vidal
mcv@...