Intensional Predicates without Biclausal Structures * "exceptional Accusative" Predicates 2.1. Data

ing away from the technical details, what the data in (14)-(15) suggests is that the Case marker on a nominal, be it nominative or accusative, can be omitted as long as the nominal is adjacent to the predicate. With this observation in mind, let us examine the data in (16) with the exceptional accusative predicate. (16) Case-marker drop and the exceptional accusative predicate: a. Taroo-wa tyuugokugo-ga/-o suki-da yo. Taro-Top Chinese-Nom/-Acc like-Cop.Pres SFP 'Taro likes Chinese.' b. (?)Taroo-wa tyuugokugo-∅ suki-da yo. Taro-Top Chinese like-Cop.Pres SFP 'Taro likes Chinese.' The Case-marker on the theme NP of the exceptional accusative predicate can be as freely omitted as the object of other kinds of predicates. This suggests that the theme NP of the exceptional accusative predicate is postulated in the position adjacent to the predicate. 3.3. Quantifier Interpretation As argued in Kuroda (1970) and Hoji (1985), among others, a sentence such as (17) whose subject and object are quantified expressions, is ambiguous for a subset of Japanese speakers, with either of the quantifer taking scope over the other. (17) Dareka-ga daremo-o aisi-te i-ru. someone-Nom everyone-Acc love-Ger be-Pres 'Someone loves everyone.' someone > everyone, ??-?*everyone > someone 8 It has been argued that sentences with two quantifiers are unambigous in Japanese unless scrambling is involved (Kuroda 1970, Hoji 1985). However, as Kuno (1976) notes, some native speakers indeed find examples like (17) ambiguous. As extensively discussed in Ueyama 1998, the possibility of having ambiguous interpretation of a sentence with two quantifiers may depend on the choice of quantifiers used; in particular, when the quantifier in the object position is of the kind which easily induces a specific group reading, then the ambiguous interpretation is easier to obtain.