Finite Small Clauses in Japanese and Their Theoretical Implications

Finite Small Clauses in Japanese and Their Theoretical Implications Masakazu Kuno University of Tokyo Japan masakazu-kuno@nifty.com Abstract This paper investigates the internal structure of finite small clauses (FSC). I will propose that a FSC is base-generated at Spec-CP and a null operator is involved to check the formal features of the embedded T and turn a sentence into a predicate. 1. Introduction Japanese has small clauses as shown in (1a). 1 In this example, the small clause subject has its Case feature checked against the matrix verb, resulting in an accusative Case-marking. This is indeed the only way for the small clause subject to check its Case feature. Since there is no finite T in the small clause, as is indicated by the non-inflected form of the predicate, the small clause subject cannot appear in the nominative as in (lb): (1) a. John-ga [sc Mary-0 [PRED kasikoku]] omotta� (koto) nom� -acc� intelligent considered (fact) "John considered Mary intelligent." b. *John-ga [Sc Mary-ga [PRED kasikoku]] omotta� (koto) nom� -nom� intelligent considered (fact) However, it is often said that Japanese allows another type of small clause, in which a small clause predicate shows up in a finite form as in (2a). I will refer to this type of small clause as (mite small clause (FSC). That there exists a finite T in the FSC can be made clear by the availability of the nominative Case marking on the embedded subject as in (2b): (2) a. John-ga Mary-o kasikoi to omotta� (koto) nom� -acc intelligent C considered (fact) b. John-ga Mary-ga kasikoi to omotta� (koto) -nom� -nom intelligent C considered (fact) 1 Koto "the fact" is added to the end of some of the Japanese examples to avoid the unnaturalness resulting from the lack topic in a matrix sentence. I ignore koto in the translation of those examples. 237 The string (2b) can be analyzed as a complex sentence that embeds a finite clause, in which the subject Mary-ga "M-nom" is moved from the predicate-internal subject position to Spec-TP as is shown in (3): (3) John-go [cp [Tp Mary-go; [PRED kasikoi]] to] omotta� (koto) -nom�-nom� intelligent C considered (fact) "John considered that Mary is intelligent." Then, what is the internal structure of FSCs like (2a)? I will answer this quetion by considering the two questions (4) and (5): (4) Where is the FSC subject base-generated? a. At predicate-internal subject position b. At Spec-CP of the embedded clause (5) Where is the FSC subject moved to before Spell-Out? a. Spec-TP of the embedded clause b. Outer Spec-vP of the matrix clause c. Spec-CP of the embedded clause There are two conceivable answers for the former question as in (4a-b) and three answers for the latter as in (5a-c). Therefore, there are logically six combinations. However, the combination of (4b) and (5a) can be precluded in the first place because it requires lowering operation in overt syntax, which runs counter to the extension condition in the sense of Chomsky (1993). The other five possible combinations …

Content truncated to 500 words