From: Willem Vermeer
Message: 39477
Date: 2005-08-01
On 27 Jul 2005 at 15:36, elmeras2000 wrote:
[On the valencies of the Moscow School]
> I find it amazing how frequently the valencies are predictable. All acute vowels are heavy, [...].
That looks to me like a misunderstanding. Acute vs. non-acute and "H" vs. "L" are independent,
as shown by the Lithuanian paradigm (3), where the stem is acute but low. There are four
possibilities: (1) Acute+H; (2) Non-acute+H; (3) Acute+L; (4) Non-acute+L.
> I intuitively consider it close to cheating to just write a feature of valency into the segments, but I admit that
> by Balto-Slavic this system has been brought about, and it does seem to be adequate for the explanation of younger
> and productive forms.
I'm not sure it is cheating in a strict sense, but compared with more traditional approaches it
doubles the number of prosodic possibilities in unstressed position, thereby critically enhancing
the opportunities for fishing in muddy waters.
The Leiden people appear to regard the valency system as pre-PIE (where it has to contribute to
the rise of the distinction between voiceless and aspirated consonants, laut Lubotsky), perhaps
wholly or in part as PIE (considering the way elements of it are held to have survived into Greek
and Indo-Iranian, laut Kortlandt) and presumably assume that it collapsed (i.e. developed into a
system of contrastive stress) at an early Balto-Slavic stage (but I'm not sure they say so anywhere
explicitly).
I'd written:
> > (i) In Moscow, Sergej Nikolaev has personally seen to it that Dybo's law has disintegrated into a "rightward
> drift" comprising an entire series of changes which I for one don't understand at all, but which is held to have
> enormous consequences for PSl dialectology.
You wrote:
> Is that wise?
Obviously not, but Nikolaev's messy "rightward drift" (which is a succession of rightward shifts)
still appears to be the party line.
And:
> It seems that dialectology always wreaks havoc when specialists want all variants to be accounted for
> as if they were all equally old. [...]
I couldn't agree more.
I'd written:
> > And in that connection, de Saussure's law has been reformulated for the first time in almost a century.
And you reacted:
> Really? Saussure' law is as neat as anything we've got: A long vowel takes the accent from a preceding short vowel -
> the strong one wins. If the long vowel is given heavy valency in a younger system, the law can logically be
> abolished; is that what has been done?
This one really has to be seen to be believed. (Note again that valencies are independent of
length.) The stress is held to have moved to an internal syllable only if the latter is _of the same
valency_ as the originally stressed syllable. (For final syllables de Saussure's formulation still
holds, of course until further notice.) But don't take my word for it, ça vaut le détour: Dybo, V.A.
& Nikolaev, S.L., "Novye dannye i materialy po balto-slavjanskoj akcentologii". In: Problemy
slavjanskogo jazykoznanija: Tri doklada k XII Mez^dunarodnomu s"ezdu slavistov, M., 1998, 5-
70, p. 54. Same text (nearly): Dybo, V.A., "Iz balto-slavjanskoj akcentologii. Problema zakona
Fortunatova i popravka k zakonu F. de Sossjura", Balto-slavjanskie Issledovanija 1998-1999,
2000, 27-82, p. 75.
I'd written:
> > (iii) Kortlandt's law is by no means unopposed among the four (five?) specialists in Old Prussian. To put it
mildly.
And you reacted:
> How so? Kortlandt's observation that geminated spelling is regularly
> followed by accent mark on vowels is a rare case of empirical proof.
> I am desperately behind in my reading. Could you tell us some more?
It's a general impression ...
(A point observers tend not to like is the idea that geminated spelling points to stress on the
_following_ rather than the _preceding_ syllable, never mind the empirical proof.)
[...]
> ... The smaller and more specialized the field, the less room does it offer for acceptance of the opinions of others.
> That is just petty rivalry.
Well, yes, but earlier accentologists (up to the generation of Belic/, Lehr-Spl/awin/ski, van Wijk,
etc.) would assiduously review and discuss and dissect each other's work, simplifying the role of
the onlookers mightily.
[...]
I'd written:
> > Literally nobody outside Leiden accepts Kortlandt's glottalic interpretation of it, yet I'm not familiar with even
> a single systematic discussion. Given the virtual consensus one would have expected several.
And you wrote:
> But what would there be to discuss systematically?
The way it is embedded in a general conception of the development of the phonological history
of Balto-Slavic.
> If (apart from aRD > a:RD) Winter's law works immediately before the accent, then the Latvian tone would be the
> glottal one anyway. Shintani said that.
OK, but there are counterexamples, such as the 'nit' word.
On the documentation of accent types:
> > Another point is the rickety foundation of the attribution of accent types to individual items, particularly in the
> case of nouns. The word for 'hand' (Lith. _ranka_) has the reflex of fixed stem stress in Lithuanian (2) and is
> overwhelmingly mobile in Slavic. The case is typical (most items are attested with more than a single accent type)
> and the tradition is just to state apodictically whichever solution seems most convenient, often on the basis of data
> from unpublished and unanalysed manuscripts or dialects. (The massive use of factual material that cannot be publicly
> evaluated was initiated by Illich-Svitych, sorry to say.)
And you wrote:
> Sure, if ''hand' is 2 and c, don't use it; same with material that cannot be trusted.
Consistency in this respect would force one to throw away just about all the evidence.
> ... I for one find Il.-Sv.'s presentation ond the conclusions drawn from it absolutely compelling. I consider his
> oeuvre and his genius to be on the same level as Saussure's.
Ill.-Sv. had very little time: at the age Mozart died, he [Ill.-Sv.] had been dead for five years. And
he belonged to the generation that had to reinvent historical linguistics after it had been reduced
to the merest trickle as a consequence of the Russian revolution and its aftermath. He is a hero.
But nobody is perfect. De Saussure would never have dreamed of quoting factual material from
sources without first reading the introduction _and_ being on the look-out for signs that
everything is not well _and_ acquainting himself with the general context. Whereas Ill.-Sv.'s
recklessness in this respect really has to be seen to be believed. His observation that the
Common Slavic transition of msc o-stems from (b) to (c) failed to take place in the peripheral
north-west of C^akavian (the phenomenon later called "(d)") is exclusively built on palpably
unreliable dialect descriptions coupled with ignorance of the basics of SCr dialectology. And this
happens to be a point we _can_ check because those descriptions have at least been published
and subjected to a certain amount of public evaluation. So I tend to be a bit sceptical about those
cases where we can't judge for ourselves because we have no access to the primary facts. I'm
convinced that in Vedic or Greek or Old Germanic studies nobody would get away with this kind
of thing. Why should Balto-Slavic accentology be different?
[...]
Willem