THEORY: What OT needs [was Re: Moraic codas]

On Fri, 13 Jul 2001, SMITH,MARCUS ANTHONY wrote:

> I agree with you completely. OT is in need of many things, and a good
> theory of constraint ranking is one of them. Another thing, closely
> related, is a theory of what a constraint can be, and how they are
> learned.
Jason Eisner at UPenn has proposed an extremely restrictive
theory of what OT constraints can be. He allows for two kinds of
constraints: the implication family and the clash family. The
first family requires linguistic elements to cooccur, while the
second family prohibits the cooccurrence of linguistic elements.
He claims to be able to derive the effects of all of the
constraints posited in the literature by using only these two
constraint mechanisms. Bruce Nevin used this restrictive version
of constraints in his dissertation on Pit River Phonology. And
to good effect, I might add.

Both Eisner's and Nevin's work can be found at the Rutgers
Optimality Archive. If you're of a formal bent, Eisner's papers
are good reading. Nevin is great description of the phonology of
a California Indian language; he's a field linguist who's paid
his dues.

> As long as people continue to create "arbitrary" constraints
> based on a description of the data, rather than use some kind of
> theoretical guidelines, Optimality Theorists are likely to get
> many explanatory results.
I agree that people have been playing fast and loose with OT. I
try to be more constrained in my use of it in my work, but since
my aims are primarily descriptive rather than theoretical, it
isn't my primary concern. That is, I'm more likely to fret about
the proper characterization of voiceless vowels in Shoshoni than
I am about implications of the factorial typology of a set of
constraints.

Dirk

--
Dirk Elzinga                          dirk.elzinga@m.cc.utah.edu

"The strong craving for a simple formula
has been the undoing of linguists."               - Edward Sapir