Re: vowel harmony

At 10:51 PM -0600 21/11/05, Herman Miller wrote:
>That's a funny coincidence, I was just listening to the Routledge
>Mongolian tapes earlier today, and thought that the sound spelled
>with the Cyrillic letter that looks like "o" with a line through it
>sounded quite a bit like [8]. Considering that the IPA symbol for
>[8] looks like this Cyrillic letter, I wonder if there's some
>relation (one borrowed from the other). I do recall that the vowels
>they were calling "front" vowels didn't sound a bit like front
>vowels, and that the Cyrillic letter that looks like a capital Y
>moved downward was pronounced pretty much like /u/ in American
>English (not a fully back [u], but not far enough forward to be
>anywhere near [u\], either). I don't recall if there was any
>difference between short [8] and the long equivalent, but if
>anything, the long equivalent of [U] sounded a bit like [o:] (or
>perhaps [7:]).
Yeah, you're right. (I think... not up to speed on the current ASCII
IPA flavor!). In crude terms, you could even say that it's a
rounded/unrounded distinction, not a front/back one. That's certainly
what it sounded like to me when I was learning and working with it.

The description of vocalism in the Mongolian languages has been
driven, until very recently, by 19th century Turkological models. (In
general, the state of linguistics of Mongolian languages is amazingly
primitive; the historical-comparative tail has been wagging the
linguistic dog for generations -- but that's another rant.) Hence the
definition of certain vowels as "front", and the tradition of
transliterating and transcribing them as such. I think it's pretty
safe to say that in reality, there is not a single Mongolian language
with the vowels /y/ or /2/ -- in fact, I'm prepared to go further and
claim that there *never has been* one, either. (G'wan, someone, prove
me wrong...) It's just an artifact of 'prescriptive descriptivism' by
early Western students of those languages, and a reliance on
'argument from authority' by their successors.

Of course, the fact that the traditional Mongolian script is borrowed
from Uighur, and reflects Uighur vocalism, further muddies the
picture. I never got around to it, but at one point I had the idea of
looking at Middle Mongolian texts in non-Turkic-derived scripts
(Chinese and hphags-pa) with a more critical eye to the presentation
of the vowels and vowel harmony. I'd gotten the impression that
modern scholars of those texts had kind of shoehorned that data into
the front/back distinction they presumed 'must' be there -- but I
could well be totally off the mark there.

As far as vowel harmony in Mongolian goes, it definitely still is
alive and well in at least the 'southern' dialects of modern
Mongolian -- Chahar and so forth. I'm not so familiar with Khalkha,
but I doubt it's much different. It *can* be difficult to spot it
immediately, given the tendency of any short unstressed vowel to turn
into some kind of schwa-like microvowel, but it's definitely there,
and works just as described in the literature. (Notwithstanding
quibbles about what those vowels *are*; the pattern is correct
enough.)

A couple years ago Bert Vaux (sp.?) at Harvard published a paper
suggesting that vowel harmony in the Tungusic languages is actually
+/-ATR harmony -- at least in origin, even if it's not intact or
clearly preserved in all members of the family today. I'd already
seen some part of that argument in the older Russian literature on
Even ("Lamut"), but Vaux's paper sold me on the bigger idea -- enough
to build it into my own faux-Tungusic conlang from the ground up.
Being pretty much totally out of academe these days, I don't know if
anyone has looked into whether ATR could play a role in the harmony
systems of any of the modern Mongolian languages, but it seems
plausible to me on the surface. Certainly "front/back" is quite
inaccurate for Mongolian, even if it is still appropriate for at
least some of the Turkic languages.

Oooo-kay... that was a longer rant than I intended. Sorry. But it's
pretty exciting to see multiple people on the list talking about
Mongolian -- and listening to it! Pity all that gets any coverage
these days is Khalkha, it's got to be the ugliest-sounding of all the
dialects... ;)

Kim
--

CONLANG Code: G/S v1.1 !lh@ cN:L:S:G a+ x0 n3d:1d !B A--- E--- L-- N0
Iv/m<s k-- ia--@ p- s+@ m- o P--- S++