Mittwoch, 14.30 Uhr
Complex predicates present grammatical frameworks with interesting interface problems. In this paper, I will be mainly concerned with some of the issues that complex predicate formation raises for the 'syntax-phonology' interface. This is an aspect, which hasn't received much attention in the literature; the only work I am aware of being that of Hayes and Lahiri (1991) which addresses the issue of complex predicates in Bengali.
I will take Romance complex predicates into consideration and in particular restructuring (and causative) verbs in Italian. The issue of their appropriate prosodic representation will be addressed and two alternatives will be considered: either the restructuring verb(s) and the verbal complement merge into a prosodic word (PW) or they form a Phonological Phrase (PPh).
The behavior of restructuring verbs with respect to stress and to certain phonological rules, such as Intervocalic s-Voicing, will be considered (cf. Nespor and Vogel 1986). Evidence will be thus provided against the possibility that the verbal complex forms a prosodic word. On the other hand, the way the two verbs behave with respect to phonological rules such as Raddoppiamento Sintattico, Stress Retraction and Final Lengthening (which are rules that have the PPh as domain of application) seems to suggest that the restructuring verb(s) and the verbal complement merge into a Phonological Phrase.
Following Nespor and Vogel (1986), I will suggest that there is no clear isomorphism between syntactic structure and prosodic structure. This is particularly evident in the case of restructuring verbs which have two different syntactic structures associated with them, depending on whether they trigger clitic climbing or not (cf. also Rizzi 1982). However, I will show that this distinction is not reflected at the prosodic level where there will be the same structure in both cases. In addition, I will question the appropriateness of the algorithm introduced by Nespor and Vogel to derive Phonological Phrase constituency from syntactic constituency and I will make suggestions on how it should be extended.