My 2 cents' worth

[wbaxter]

Hello, all--

I've been a fan (not really a practioner) of HPSG and related tendencies
(GPSG, Montague Semantics, Situation Semantics, categorial grammar a la
Carpenter, CCG...) for over 20 years now.  This status of
non-participant observer perhaps has at least the virtue of offering a
different perspective than that of those more closely involved.  A few
thoughts, as I've watched this discussion:

1.  I agree with the general strategy of trying to offer insightful
analyses of real problems, and with the rhetorical technique of
presenting them in a non-technical way first, followed by a fully
explicit version.  That is, it's good to do good linguistic work that
provides insight into particular problems.

2.  It is good for at least some people to deal with really big-picture
questions--like, what characterizes human languages?  What is it that
we  can do that other species can't? Realistically, how might this
ability or these abilities have evolved?  Does it make sense to classify
human languages into types (e.g. "configurational" or
"nonconfigurational"), or is there just one type of language, namely
human language?  In what ways to human languages differ from each
other?  Part of Chomsky's appeal is that he does have claim to have
something to say about big-picture questions (though not necessarily
these particular ones).

3.  Moments that stand out in my mind are (1) the suggestion (at least
as early as Pollard & Sag's first HPSG book) that languages can be seen
as populations (or lattices, or, I guess, semi-lattices, actually) of
linguistic signs, the distinction between "rule of grammar" and "lexical
item" being seen as ultimately artificial; (2) Mark Steedman suggesting
that what humans are especially good at is the operation of
function-composition; (3) Donohue & Sag, "Domains in Warlpiri", which
seemed to bring Warlpiri out of the nonconfigurational wilderness; (4)
"French clitic movement without clitics or movement"; (5) the analysis
of the English auxiliary system, showing what 'affix-hopping' was really
about, and how the differences among closely-related varieties of
English can be represented; (6) the literature on German word-order,
which seems to show that--never mind Warlpiri--even in such a well-known
language as German, accounting in a precise way for word order is a
nontrivial problem.

4.  Whenever anybody uses the word "standard", I reach for my wallet (to
make sure it is still there).  In this day and age, what linguistic
analysis or approach could possibly deserve to be called this?  BTW, can
anyone verify my suspicion that this expression was borrowed from
quantum physics (where there really is, I believe, a relatively stable
"standard" theory)?

5.  It may seem discouraging that GB/MP (for want of a better term) has
a greater influence than many of us feel it deserves, but one aspect of
this is encouraging, in a way:  like it or not, a lot of bright people
are attracted to the framework which appears to be dominant (from a
sociological point of view); once they get there, at least some of them
will not be satisfied with it in its current state and will gnaw at it
from within to try to improve it.  This means that at least some of the
GB/MP will turn out to be worth reading.  When HPSG-ers compare their
framework with GB/MP, they are most impressive when they demonstrate
that they have read it closely and understood it, and can run rings
around it.

6.  Not all the important theoretical work will seem sexy to the general
linguistic public; I suspect it is equally important for those who
understand foundational issues to hash them out among themselves, even
if the average linguist would yawn with boredom (or tremble at the
difficult math).

7.  One possible consequence of point 6 is that small conferences are
not necessarily bad.  It is true that "conferences" are an important
genre of academic ceremony and display, but you may actually get more
important work done at small workshops.  I attended a meeting at
Carnegie-Mellon a few years ago which was not terribly big, but (from
what I could tell) seemed useful; it appeared that real issues were
being discussed.

8.  One of the functions of conferences is to help students get jobs.  I
can't presume to give authoritative advice on this, but if I were on a
committee looking to hire a syntactician, I would hope to find someone
who (1) can communicate, politely, with the general linguistic public
(maybe even a bigger public than that!), on what they are doing and why
it is worthwhile; (2) is familiar enough with frameworks outside their
own to be able to be an impressive participant in inter-framework
debates; (3) has a broad enough perspective to be able to see where the
possible weak points are in their own framework; (4) isn't afraid of math.

Bill Baxter
University of Michigan