Principles of economy in nonfinite complementation

Date of Completion

January 1995


Language, Linguistics




This thesis investigates the role of Economy Principles in infinitival and participial complementation.^ In chapter 2, I examine English ECM and control infinitivals. I provide evidence against c-selection for various types of infinitival clauses and the standard binding-theoretic account of the distribution of PRO. I argue that both ECM and control infinitivals are IPs, the IP status being forced on them by Economy of Representation. I give an account of infinitival complementation based on the Case-theoretic account of the distribution of PRO and the s-selectional properties of higher predicates.^ In chapter 3, I provide an Economy account of Pesetsky's (1992) observation that agentive verbs cannot ECM lexical NPs based on Hale and Keyser's (1993) proposal that agentive constructions contain a null agentive verb. I also examine properties of French believe-class infinitivals, which, in contrast to their English counterparts, license PRO. I show that French believe-class infinitivals differ from English believe-class infinitivals in that they are specified as (+ Tense). I provide a uniform account of the different behavior of English and French believe-class infinitivals with respect to their Tense specification and licensing of PRO based on the Case-theoretic account of the distribution of PRO.^ Chapter 4 investigates the relevance of infinitival complementation for existential constructions and expletive replacement. Based on the tacts concerning infinitival complementation, I argue that the associate of there is Case-marked by be and that Chomsky's (1993) Greed holds. To reconcile the conclusions with the expletive replacement hypothesis, I adopt Chomsky's/Lasnik's proposal that there is an LF affix and propose that to overcome its morphological inadequacy, there undergoes affix hopping in LF. Empirical evidence is provided for the affix hopping analysis.^ Chapter 5 examines participle (P)-movement in Serbo-Croatian. I provide evidence against the theoretically anomalous long-head movement to C analysis of Serbo-Croatian P-movement (Rivero 1991, Roberts 1994). Instead, I argue that Serbo-Croatian participles adjoin to the element located under Aux$\sp0.$ The analysis is extended to P-movement in Dutch and shown to have important consequences for second position cliticization, excorporation, and the direction of adjunction operations. ^