subcategorization

[Dick Hudson]

A footnote to Joan Bresnan's interesting survey of problems for defining
complement-types in terms of phrasal categories rather than functions: 
Then there's always the problem of those complements which can't be defined
at all in syntactic terms, because they range over a variety of syntactic types.
a. The second complement after verbs like PUT and TREAT or WORD, whose only
restriction (apart from being obligatory) is semantic (respectively, they
must define a place and a manner). In syntactic terms they can be PP or NP.
b. Predicative complements, i.e. the one after verbs like BE and SEEM, which
have to define a `predicate' which may be expressed syntactically as AdjP,
PP, NP or various kinds of clause.
c. Equative complements after the equative BE, where virtually anything is
allowed (especially in cleft sentences), provided its referent can be
identified with that of the subject.
Not to mention the problem of the complement of GO which cannot even be
verbal (e.g. "The train went [whistling sound]" but, for older people, not
%"She went `I feel sick.'"). 

In function-based theories like LFG and Word Grammar, none of these
restrictions is particularly problematic because the complement concerned
can be identified functionally, independently of whatever restrictions have
to be imposed. If complements can only be identified (i.e. picked out from
other complements) in terms of phrasal categories, it is logically
impossible even to identify a complement which is *not* restricted in terms
of phrasal categories.
Richard (=Dick) Hudson
Department of Phonetics and Linguistics,
University College London,
Gower Street,
London WC1E 6BT
work phone: +171 419 3152; work fax: +171 383 4108
email: dick at ling.ucl.ac.uk
web-sites:
  home page = http://www.phon.ucl.ac.uk/home/dick/home.htm
  unpublished papers available by ftp = ....uk/home/dick/papers.htm