Re: long, flat, full

From: tgpedersen
Message: 60598
Date: 2008-10-06

--- In, "Arnaud Fournet" <fournet.arnaud@...>
> ----- Original Message -----
> From: "tgpedersen" <tgpedersen@...>
> > ==========
> > My point of view about PIE *l is that this "surface"
> > correspondence covers more than one proto-phoneme.
> > When PIE *l corresponds with PAA *l as in *pel = full = Arabic
> > Hafil = Touareg balal,
> > the expected correspondence in ST should be yod.
> Why?
> > > For that matter, these ST roots are highly dubious.
> > In what way are they dubious?
> ============
> A ST root cognate to PAA *p_l = PIE *p_l should be *p_y
> otherwise, I consider it does not work.

Should I infer from that that if a reconstructed ST root doesn't
follow the rule you have set up, it is highly dubious?

> To say nothing about the unpleasant voiced / voiceless alternation,
> which is an open-door to create phonetic fancies.
> Arnaud
> =========
Edwin G. Pulleyblank
Outline of Classical Chinese Grammar, pp 10-11
'As our understanding of the phonology of Old Chinese improves, it is
becoming possible to explain some of this morphology in terms of
affixes of various kinds. The following are some of the most important
(b) Alternation between Middle Chinese voiceless and voiced initials
is often found in verbs with transitive and intransitive or neuter
meaning respectively, e.g., jiàn .. (EMC kenh) 'see,' also read xiàn
(EMC Genh < *g-) 'appear' (now written .. in this meaning); zhu ..,
.. (EMC tc,uawk) 'to attach, enjoin,' shu .. (EMC dz,uawk) 'be
attached, belong.' This probably reflects a prefix *a-, cognate to
Tibetan ha-c^hun^ and Burmese a(?) (Pulleyblank 1973a, 1989)
(c) ... '

> > There are at least two possibilities :
> > 1. all ST words are loanwords from IE languages,
> > 2. ST words may also be borrowings from SouthEast Asian languages
> > that keep *l unyodized.
> Explain.
> ========
> It's possible SE Asian languages keep *l (=PPA *l = PIE *l)
> unchanged.
> I have not checked this.
> I consider all Northern Asia languages should have yod for (=PPA *l
> = PIE *l)
> Uralic, Turcic, Yukaghir, Chinese have yod.
> Languages that keep *l as l or r should preferably be removed from
> ST.
> Arnaud
> ===========
Or you should change your rule.

> > In any case, you have to be aware that Matisoff is not a very
> > reliable ST expert.
> Because?
> Besides, it doesn't matter much, since the root reconstructions are
> Benedict's.
> Torsten
> ==========
> Well, then the master was not very good and the disciple does not
> look like an improvement.
> Is there anything in ST studies we should remember Benedict for ?
From Matisoff's Preface in his
Handbook of Proto-Tibeto-Burman
'My involvement in Tibeto-Burman (TB) and Sino-Tibetan (ST)
comparative reconstruction dates from my first fieldwork on Jingpho,
Burmese, and Lahu in the 1960's, and especially from my intense
contact with Paul K. Benedict when I was teaching at Columbia
University (1966-69). The manuscript version of Benedict's
Sino-Tibetan: a Conspectus (STC) had been lying around unpublished
since its composition around 1940; it was exciting for me to
contribute to its eventual publication in 1972. With its nearly 700 TB
cognate sets, and over 300 TB/Chinese comparisons, the Conspectus
ushered in the current renaissance of TB and ST comparative
linguistics. Its rigor and precision, as well as the breadth of its
vision, have made it the indispensable point of departure for
subsequent work in the field.
While there is certainly room for tinkering with a few details of
Benedict's reconstructive scheme for Proto-Tibeto-Burman (PTB), the
major features of the system itself remain basically unassailable. The
real progress that has been made in the past 30 years lies elsewhere.
An avalanche of new data from recent fieldwork has strengthened the
support for previously reconstructed etyma and has permitted the
reconstruction of hundreds of new roots at all taxonomic levels of TB,
though many more undoubtedly remain to be discovered. The harnessing
of the computer for etymological research has speeded up the
identification of new cognates and provided a powerful tool for
testing the validity of proposed reconstructions. A better
understanding of the variational processes at work in TB and ST
word-families has enabled us to decide more accurately whether sets of
forms that bear partial phonosemantic resemblances to each other are
really variants of the same etymon or etymologically independent. On
the Chinese side, the successors to Karlgren have made profound
changes in the reconstructive scheme for Old Chinese, and it is no
exaggeration to say that the field of historical Sinology is now going
through a period of ferment. Still, almost all of STC's suggested
Chinese comparanda for PTB etyma have gone unchallenged.'