Re: Various loose thoughts

From: willemvermeer
Message: 35998
Date: 2005-01-18

--- In cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:

> > I also don't really understand Kortlandt's argument about R.
> > détjam, détjax, ljúdjam, ljúdjax. I gather the idea is that
> > the accent was retracted in i-stem ljudImÚ, ljudIxÚ from the
> > final yer all the way to ljúdImU, ljúdIxU, skipping the
> > middle yer, but how does that invalidate the possibility of
> > ljudÍmU, ljudÍxU > ljúdImU, ljúdIxU? What am I missing?

Then "Thomas Olander" <olander@...> wrote:

> I missed that point, too. I consulted Kortlandt's "Slavic
accentuation",
> p. 15 (which he refers to in his paper in Baltu filolog'ija 2004,
p. 27),
> but it didn't really help.
> Maybe someone from Leiden could tell us?


I'll try, but I know no more than you do, having only FK's
publications to go on. Unfortunately I have to start with a few
generalities to avoid misunderstandings.

The most intricate and innovative part of FK's accentological theory
is his relative chronology of late Slavic innovations (i.e. the
second and third chapters of his book). By far the most important
purpose of that chronology is to generate a system of vowel
quantities from which the attested distributions (which are more
intricate than is generally realized) can be derived.

Historically, Slavic vowel quantity has rarely been treated as a
problem in its own right. The traditional terminological system that
uses words like "acute" and "neoacute" (etcetera) to refer to
prosodic phenomena talks only about what happens in stressed
syllables and combines quantity with tone into undigestable units.
Interest in vowel length in unstressed syllables has been extremely
limited.

There have of course been exceptions, notably in the early decades of
the twentieth century. For us the one that is most relevant is Stang,
who describes, or at least mentions, a whole range of phenomena
involving vowel quantity that a good theory of Slavic accentuation
should account for. FK's theory is an attempt to do just that. (FK's
theory is an attempt to solve Stang's problems in the light of
Ebeling's chronology and the early publications of the Moscow
School.)

In the post-Stang era, however, the influence of the Moscow School
has caused interest to be focused rather onesidedly first on problems
involving the place of the stress and second to problems common to
Slavic and Baltic. The present wave of attempts to reformulate Hirt's
and Winter's laws illustrates that onesidedness as well as anything.

Many features of FK's theory that have struck outsiders as odd or
worse have the purpose of generating the correct Slavic vowel
quantities in the correct positions. That includes the oft-deplored
decision to postpone the elimination of the reflex of the product of
the merger of the laryngeals to the Slavic period and spread it out
over three stages. And a host of other features that have struck
people as objectionable, or have remained unnoticed and uncommented
on.

Now back to the specifics of the theory.

One of the features of FK's theory that have often been overlooked is
the way the loss of the stressability of final jers interacts with
Dybo's law: first *final* jers (non-final jers are another story)
lose the ability of carrying the stress and only then Dybo's law
takes place. As a consequence, such forms as pre-Dybo *st'olU retain
the stress, which can no longer move to the next syllable, which
consists of a final jer only, contrast Gsg *st'ola > *stol'a.

This chronology is motivated entirely by the quantitative phenomena
it is designed to explain (see Slavic Accentuation pp. 15-17).
Whoever is not aware of that fact, will tend to reject it out of hand
as perverse, if it is noticed at all.

A change like ljudÍmU, ljudÍxU > ljúdImU, ljúdIxU is impossible
because non-final jers are not held to have lost the stress at the
stage involved and because those particular jers ultimately became
strong, so reconstructed *ljudÍmU and *ljudÍxU would end up with
stress on the ending in the attested material, just like *sestr'amU
*sestr'axU or *voz'omU *voz'e^xU or what have you, leaving R. ljúdjam
ljúdjax (and other phenomena) unaccounted for.

I don't want to suggest that FK's reconstruction is without
complications, far from it. In order to account for ljudjam/ljudjax
and related phenomena he has to assume that the loss of the
stressability of word-final jers took place before the classical
distinction between strong and weak jers arose. Not everybody will
like that.

Does this help?


Willem