Clarification

[Dan Everett]

I agree with Bill Turkel, namely, that the report of Guy's talk on
"deterministic" vs. "nondeterministic" is unclear. These are terms which
linguists often use in nonstandard (for CSrs, which I am not) ways, or so
my colleagues in CS tell me.

As for the compatibility of OT with functional models, I do not see how it
would be any more or less compatible than other formal models. It is possible
that functional motivations might affect constraint rankings, but in a
1965 Aspects or SPE model functional considerations could just as easily
affect rule-orderings.

-- Dan Everett