THEORY: vowel harmony [was CHAT: Another NatLang i like]

On Thu, 24 Jun 1999, Nik Taylor wrote:

> Boudewijn Rempt wrote:
> > Ob conlangs: I've once tried to make a conlang with vowel harmony,
> > but it failed quite quickly because I couldn't prevent wrong
> > words from entering the lexicon...
>
> I did that too, altho I had a slight twist, instead of front-back
> harmony, as exists in Turkish, I had a height-based harmony, thus all
> vowels being high, middle, or low (thus, kitusi, ketose, and ka"tosa" [I
> don't remember exactly what the orthography was] would all be possible,
> but *kitesi wouldn't).  Unfortunately, I don't have my notes on it
> anymore.  :-(  I know of no natlang with that kind of harmony, but I'd
> be very surprised if there *weren't*.
I wouldn't put any money on it if I were you. It has already been noted
in recent discussions that the most basic contrast in vowels is one of
height (thus some have argued that Kabardian really has only two vowels:
<schwa> and [a]; I personally find the arguments compelling). Height
harmony would neutralize this very basic kind of contrast by requiring
all vowels within a harmonic domain to be of the same height. As far as
I know, this situation does not exist in natural languages. Features
which do trigger harmony include: fronting/backing, rounding, or tongue
root advancement/retraction. Height is sometimes involved in harmony;
there are systems where vowels harmonize for some feature (usually
rounding) iff they are of like height--Yawelmani/Yowlumni is like
this--but height contrasts are never neutralized in such a system.

One of the most interesting harmony systems is found in Nez Perce, an
American Indian language spoken in the Northwest US. There are five
vowels in Nez Perce:

        i       u
                o
        ae      a

These vowels belong to one of two sets: R={i, ae, u} and D={i, a, o}
such that if a morpheme contains any vowel of set D, then all vowels
within the harmonic domain (=word) are of set D. Otherwise, all vowels
are of set R. (Note that the intersection of the vowel sets is not
empty since [i] is found in both of them.) What is interesting about
this harmony pattern is that there is no obvious phonetic correlate to
harmony; that is, there doesn't seem to be any one feature which
triggers harmony (well, there really is, but it's a strange, twisted
tale, and I fear I'm testing your patience already :-).

Chukchi, a language of Siberia, has a harmony system very similar to Nez
Perce, with the difference that instead of [i] in set D, Chukchi has
<schwa>, making a total of six vowels (and, BTW, providing an important
clue as to the identity of the harmonizing feature).

Dirk

--
Dirk Elzinga
dirk.elzinga@m.cc.utah.edu                   "All grammars leak."
http://www.u.arizona.edu/~elzinga/                 -Edward Sapir