17.01.2014 Views

Predicate-Argument Structure

Predicate-Argument Structure

Predicate-Argument Structure

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Chapter 6. The <strong>Predicate</strong>-<strong>Argument</strong> <strong>Structure</strong> 255<br />

In an HPSG like framework this amounts to developing the content part of the sign in a way that<br />

yields characterisations of the constructions being analysed that are deeper (i.e., more abstract)<br />

than those that are obtained in the simple syntactic structure. To a great extent then the purpose of<br />

this chapter is to provide a description of the predicative relation that can be fitted into the content<br />

part of the predicative signs.<br />

We are interested in providing a description to the predicative relation that is:<br />

<br />

<br />

broad enough to cover most phenomena related to predication in the major western languages<br />

deep enough to overcome some of the differences arising from the surface realisation in<br />

different languages<br />

<br />

specific enough to maintain in a strict way the link to the surface relation<br />

Let us consider these aspects in turn. Firstly, the intended coverage should be broad enough.<br />

In particular we aim at characterising the relation of predication for all major categories in a<br />

consistent and coherent way. This means that all complementation patterns for verbs, nouns,<br />

adjectives and prepositions/adverbs are in the intended coverage of this section. In addition, it is<br />

intended that the treatment for all categories is consistent with one another; in a way that what is<br />

said about adjective complements is consistent with what is said about verb complements or that<br />

the treatment proposed for noun complements is not problematic for verb complements. To be<br />

more specific about these two examples, it is intended that the treatment of adjectives as predicates<br />

is compatible with what is said of raising and control verbs (and particularly of the verb to be),<br />

since in most occasions the predicative realisation of adjectives occurs when they are attributes<br />

(either of the subject or of the object); and what is said about the predicative use of adjectives<br />

should be compatible with what is said about their use as noun modifiers. With respect to the<br />

second example, it is clear that the discussion and proposal for the treatment of complements to<br />

nouns has consequences for the treatment of the complements of all categories; in this particular<br />

case, the difficulties in representing the arguments to nouns have been one of the basic reasons<br />

for adopting a particular basic structure of predicative signs for all categories.<br />

Secondly, the coverage of the specifications should be deep enough. Thus, we aim at generality<br />

by intending to treat the basic phenomena for a wide range of European languages. Since there are<br />

obvious differences in the complementation patterns of these languages, this amounts to trying to<br />

provide descriptions of the predicative relation that are “neutral” with respect to these differences:<br />

thus whether the relation a complement has to its head is expressed by means of case, of a<br />

preposition or configurationally should not imply a different representation. Similarly the proposed<br />

system should propose similar representations for language specific different constructions (a<br />

complement to a noun should obtain the same basic representation whether it is in the genitive<br />

case or a prepositional phrase).<br />

And finally, the description we are aiming at ought not to be as deep as possible, since depth<br />

in analysis alone does not provide consistency and coherence in the description. When trying to<br />

overcome surface differences one is tempted to go always deeper in the analysis; this however, if<br />

unconstrained, is a bad move since the deeper one goes the more difficult is to maintain a levelness<br />

and coherence in the obtained descriptions. At the same time, the link between the abstract and<br />

the surface representations is to be maintained, if we do not want to introduce a large amount of<br />

indeterminacy in our descriptions. These are then the limits to the purported depth of description.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!