Next: Syntactic Tests for Argument
Up: Prepositional Phrases and Verb
Previous: Introduction
The traditional treatments of dative PPs force the semantic
contribution of these PPs to be specified explicitly in the lexical
entry for each verb which can appear with a dative PP, essentially
treating the preposition as a marker for a particular verbal relation
and a particular semantic role, with no independent contribution.
These approaches seem to demand the syntactic subcategorisation of
dative PPs, even in cases in which they are syntactically optional
such as exp2, since they model the dative alternation as a
permutation of a verb's subcategorisation list. This ignores
generalisations over the contribution of the PPs, in that the PPs seem
to add similar information regardless of which specific verb they
appear with. Thus in each of exp2a and exp3a, the PP
for Mary specifies who benefits from (some aspect
of)
the event described in the
remainder of the sentence.
John made a drawing for Mary.
John made Mary a drawing.
Likewise, the contribution of the inner NP in the alternate in
exp2b of exp2a and the alternate in exp3b of
exp3a can be identified as specifying who receives benefit from
the outer NP (a cake and a drawing, respectively).
A general question arises from the observation of such
generalisations -- what is the status of these PP elements? Should they
be treated as subcategorised-for complements or as adjuncts which make
an independent, identifiable, semantic contribution across verbal
heads? This will be investigated below through a series of standard
syntactic and semantic tests for argument structure. The two types of
dative PPs, -PPs and -PPs, will be contrasted in this
investigation. The analysis will show that -dative PPs and
certain -dative PPs behave as complements syntactically while
behaving as adjuncts semantically. The remaining -dative PPs
behave as complements both syntactically and semantically.
Next: Syntactic Tests for Argument
Up: Prepositional Phrases and Verb
Previous: Introduction