From:    John Cowan <[log in to unmask]>
> I think the problem is that BE can have "the jury is agreed" and
> "the jury are disagreed", so unless you want to say that "jury" is two
> different lexical items in such a case, you have to allow for the effects
> of semantics.  As And says, plural verb agreement appears when the noun
> is felt *in that particular context* to be plural, regardless of its form.

Yeah, I think this is a problem.  Just the other day, after I sent
off that response to your email, I was doing some casual reading
in Dandamaev's massive _Slavery in Babylonia_, wherein I noticed
the following sentence:

  "The overwhelming majority of slaves was of local origin, i.e.,
  Babylonian" (103).

I realized when reading this that my dialect would greatly prefer
plural agreement with majority:  "the majority of slaves were",
where singular agreement sounds rather stilted and formal.  Thus,
the LFG notion of agreement, without a particular controller, would
have problems with this unless you read "majority of" as a lexicalized
quantifier of some kind. Attraction phenomena like this (or case in
Greek) would still be awkward.

> You can always solve grammatical problems by introducing more and more
> homonymous lexical items, but eventually old William will start looking
> at you, er, sharply.

Yeah.  Although I am increasingly skeptical about the mechanisms
behind optimality theory, perhaps some kind of stochastic OT (or
something like it) would work best.  This would shift some of the
weight back onto grammar.

Thomas Wier	       "I find it useful to meet my subjects personally,
Dept. of Linguistics    because our secret police don't get it right
University of Chicago   half the time." -- octogenarian Sheikh Zayed of
1010 E. 59th Street     Abu Dhabi, to a French reporter.
Chicago, IL 60637