optimality party line

[Suzanne E Kemmer]

Often, issues relevant to how functionalists might view
Optimality Theory get discussed on the Optimality listserver
based at UCSD.

For example, the question of how faithfulness constraints might be
initially set in UG (low or high) brought some responses from several
people who clearly saw the issue in terms of processing, learning,
etc.  The person who brought up the issue said he was shocked to
receive such responses, which presented a view of language "so foreign
to those which I've come to view as standard in theoretical
linguistics."

It looks like processing and learning have no place in optimality
theory (or at least 'standard' OT), since this person went on to
describe his approach to the study of OT acquisition, which "takes the
grammar as a well-defined computational component of the human
cognitive system, independent of the production system and
environmental factors."

This stance is of course very familiar (and very alien) to
functionalists, as is the apparent stance against gradient
phonological phenomena.  (Whatever looks gradient is either just
phonetics, or sublinguistic, or something--contra Pierrehumbert
and others' cogent arguments to the contrary)

Of course, it may be (in fact probably is, in light of the 'shocking'
responses) that the theory is being used by many who take no such
positions, and may prove useful to non-believers in the basic party
line on particular issues.

Finally, I collected a very fascinating correspondence from an
optimality doubter, detailing exactly what properties of OT he
objected to and what phenomena it apparently has a great deal of
trouble with (in a nutshell, lexical). I found it an extremely useful
refresher of various phonological issues and how natural phonologists
etc. dealt with them, and very useful in understanding OT in its
historical context (which tends to get blurred very quickly when
people jump to a new theory with its own favored problems to work on
without worrying about what other problems the old theories did or
didn't solve.) I'd be happy to send it to anyone who asks.

--Suzanne

P.S. In case you want to subscribe to the optimality list:

mail to listserv at ucsd.edu, with the body of the message:

  subscribe <yourloginaddress>  optimal

Similar for unsubscribing:

  unsubscribe <yourloginaddress> optimal

Send actual submissions to optimal at ucsd.edu