optimality party line

[Dan Everett]

I second Suzanne's recommendation that people interested in OT
related issues subscribe to the OT list via UCSD.

However, I disagree with her claim that OT is not concerned with
acquisition, historical change, processing, etc. Quite the opposite. There
are papers in OT on acquisition and implications for historical change and
dialectical differences, for example. And more are in the works.

But I reiterate that functional considerations are no more nor less likely
to play a role in OT than in other formal models. For example, in addition
to what I already said about functional considerations playing a potential
role in rule-orderings in previous versions of generative grammar, Chomsky
explicitly discussed the fact that WH-movement was to the left periphery
of the clause, regardless of the underlying constituent order in a
language, and assumed that this was likely due to functional
considerations (the WH element tells you how to interpret a gap in the
sentence and it is nice to have this information when you start processing
the sentence rather than at the end). All formal theories are open to
suggestions on how the mechanics of grammar interface with the uses to
which grammar is put.

OT is one of the most exciting developments in formal grammar to come
along.  But as we might expect of a new theory, it has raised many
questions for which there is still no good answer (the paper Suzanne
refers to on its potential difficulties with lexical constraints is a case
in point). But the issues of deterministic, nondeterministic, functional
relevance, etc. all seem to me to be red herrings at this stage. The
principal objective at present is to work out empirical implications of
the model via detailed studies of constraints and their interactions
across a wide variety of languages. As understanding of the model deepens
via such studies, its relevance for functional concerns ought to become
clearer.

-- Dan Everett