status of words; HPSG and CG

[kay at cogsci.berkeley.edu]

On Thu, 16 Aug 2001 dmellow at sfu.ca wrote:

> Hello,
>
> I have a question about the status of words and the different feature
> complexes in HPSG and Construction Grammar.
>
> As I understand it, HPSG assumes strict lexicalism: word structure and
> phrase structure are governed by independent principles (a quote taken right
> from the Stanford web page).  CG makes a different assumption (Goldberg,
> 1995, p.7):
>
> “In Construction Grammar, no strict division is assumed between the lexicon
> and syntax.  Lexical constructions and syntactic constructions differ in
> internal complexity, and also in the extent to which phonological form is
> specified, but both lexical and syntactic constructions are essentially the
> same type of declaratively represented data structure: both pair form with
> meaning.  It is not the case, however, that in rejecting a strict division,
> Construction Grammar denies the existence of any distinctly morphological or
> syntactic constraints (or constructions).  Rather, it is claimed that there
> are basic commonalities between the two types of constructions, and
> moreover, that there are cases, such as verb-particle combinations, that
> blur the boundary.

Although what Andreas Kathol writes about differences between Berkeley and
Illinois dialects of CG is correct, in recent conversations with Adele
Goldberg I find we have more in common than I thought, particularly in the
use of monotonic multiple inheritance - something she doesn't make a big
fuss about in her book but relies on more or less implicitly I have
discovered. Anyway, I think the above paragraph is fair to my own views
(whatever that's worth) with one important exception. It is not the case
in Berkeley CG that *all* constructions pair form with meaning. Most of
the constructions in Berkeley CG that correspond more or less to the
phrase structure schemata of "classic" HPSG (Non-inverted finite clauses
["Subject-Predicate"], SAI, VP, and so on) do not themselves pair sound
with meaning but are (often) inherited by construcitons that do.  I'm not
sure what Goldberg's position on this would be. If she would (still? Is
that fair?) maintain that all constructions carry meaning, that would
indeed be an important difference between Illinois and Berkeley CG.
Adele, are you following this thread?

On the whole, I agree with Andreas that it's not easy to find non-cosmetic
differences between Berkeley CG and the flavor of HPSG (I hope it's okay
to mix a metphor if it's tasty) represented by Sag on Relative Clauses (Jl
of Ling 1997) and the recent Ginsburg and Sag opus.  Differences are at
the level of CG not caring about subcategorizational locality and so
allowing valence constructions to reach inside of complements in order to
capture the very rare cases when this happens (causing us to bite the
bullet of having nothing to say about the rarity of the phenomenon). Of
course, the architectures are such that CG could institute something like
HPSG's subcategorizing only for sysnsem objects, not signs, if it wanted
to and HPSG could allow subcategorizing for signs if *it* wanted to.  I
think the differences about whether words are completely different from
phrases are of the same order.  Either view coudl easlily be expressed in
either approach. It's sometimes useful to distinguish between substantive
hunches or esthetic preferences that practitioners of a particular
framework have favored and decisions that are enforced by the formal
machinery of a framework. An important mistake not to make, I think, is a
a view that Chomsky advocated long ago (and I don't know whether he has
since or would now reject): that the *notation* should embody the
empirical claims of the theory.

BTW, Dean (and others) there's a new CG mailing list.  So far as I know,
nothing has been posted to it yet, so you could be the first.  To sign up,
go to http://mailman.icsi.berkeley.edu/mailman/listinfo/cxg.

Paul
__________________________________________________________
 Paul Kay                      Department of Linguistics
 kay at cogsci.berkeley.edu       University of California
 www.icsi.berkeley.edu/~kay    Berkeley, CA 94720, USA