more on OBJ(theta)

[Joan Bresnan]

The meaning of the "restrictedness" property of secondary or
restricted objects (OBJ_theta) has always evoked questions when I have
presented it, and the several recent messages on OBJ(theta) by Helge
Lodrup, Alex Alsina, and Lori Levin have been very helpful and
clarificatory to me.  I would like to add a couple of observations to
the discussion.

There is a *syntactic* distinction between primary and secondary
objects that has typological validity across many languages; it
differs from the indirect/direct object contrast (see, e.g. Matthew
Dryer's 1986 Lg article, "Primary Objects, Secondary Objects, and
Anti-Dative").  As Alex Alsina notes, this syntactic difference was
designated by OBJ and OBJ2 for primary and secondary object in early
LFG.  Secondary objects share properties with objects, as Lori Levin
notes: e.g. they are syntactically encoded without semantically
restricted adpositions or case markers; they are generally excluded as
the complements of Adjectives and Nouns; thay may be the subjects of
(i.e. functionally control) secondary predicates, at least arguably in
English depictives ("he served them the meat raw")--Norwegian differs
in this last respect, according to Helge Lodrup.  But secondary
objects also differ from primary objects in a number of respects that
need to be explained.

In her 1986 dissertation Lori Levin first proposed eliminating the
primitive syntactic distinction between primary and secondary objects
by reducing it to an independently motivated property of semantic
restrictedness.  A bonus is that the property of semantic
restrictedness would immediately allow multiple secondary objects
(since restricted objects would now be indexed by role), and this
certainly occurs in many languages.  The property of semantic
restrictedness had been used in my 1982 "Control..." paper for the
somewhat different purpose of distinguishing direct functions
(subjects and objects) from obliques.  It was Levin's original
contribution to reinterpet this feature as part of a system for
functional underspecification and linking.  The lexical mapping theory
grew out of this innovation.

In her recent msg to this list, Lori Levin notes that "the property
which [OBJ(theta)] shares with OBL(theta) is its restrictions in
linking only to specific semantic roles" and observes that this
characterization need not prevent functional control by OBJ(theta).
Her proposal can be interpreted in two different ways: (i)
restrictedness means having a specified or fixed semantic role; (ii)
restrictedness means having *some* semantic role (i.e. not being
athematic, as are expletives, raised subjects, etc.).  Given her
proposal under either of these intepretations, the failure of obliques
to be functional controllers would still have to be attributed to
something else--perhaps the role restrictions of their adpositional or
case markers, she suggests.

Alex Alsina refutes Lori Levin's proposal by mentioning the evidence
from the Bantu languages Gitonga (Firmino) and Kitharaka (Harford)
showing that both OBJ and OBJ(theta) can be goals or patients.  This
is an excellent point, showing that it would be incorrect to interpret
Levin's proposal as in (i).  But the interpretation in (ii) remains
untouched by this evidence: secondary objects and obliques still
differ from subjects and primary objects in not occurring athematically
(ie. as expletives, etc).  

Alsina also argues that by admitting restricted objects into the
theory we are parochially assuming that what appears in English is
universal, and are leaking illicit semantic information into the
syntax in the form of thematic role labels.  In view of the
typological generality of the concepts of primary and secondary
object, I don't think we need worry about the first point.  Concerning
the second point, we shouldn't be mislead by the *name* for the
primitive feature "restricted", whose history I outlined above.  Like
any primitive, its actual content is given by its various roles within
the axioms of the theory, e.g. lexical mapping theory, the theory of
functions, etc.  Moreover, "restricted" can be defined without
any reference to thematic roles:  for example,  in the interpretation
suggested above, it simply means that the function must be indexed to
a lexical role within the argument list brackets.  This of course has
empirical content when coupled with the semantics, as in the recent
work of Dalrymple, Lamping, Pereira, and Saraswat.  

Joan Bresnan


-----------------------------------------        ______     __o       __o
Joan Bresnan	bresnan at csli.stanford.edu      ______     _`\<,_    _`\<,_
-----------------------------------------       ______   (*)/ (*)  (*)/ (*)