There is a set of predicates, i. e.‘epistemic’ predicates in Japanese, which express the speaker's epistemic attitudes toward the propositional content, such as daro, rasii, kamosirenai, etc. In spite of abundant literature on Japanese auxiliaries, the syntactic and semantic structure of ‘epistemic’constructions has been only partially clarified in the history of Japanese grammar. In this paper I am mainly concerned with formulating the derivational process of them in terms of rules such as Subject Raising to Subject Position (SRS hereafter), Predicate Raising (PR hereafter), Restructuring.
I argue for the following:
(a) SRS enables us to formulate Japanese ‘epistemic’ constructions as well as English ones in the common framework of syntax, though its existence is denied in Kuno (1976).
(b)‘Restructuring rules’(in the sense of Akmajian, Steele, and Wasow 1979) are applied in the derivation of ‘formal nouns+epistemic predicates’constructions. For instance, in the case of [[S]-no]_NP+[daro]_Pred, the formal noun no is ‘restructured’ to the epistemic predicate daro to derive the complex epistemic predicate no-daro.
(c) The relevant rules are applied in the following order: Restructuring Rules>SRS>PR
(d) In ‘epistemic’ constructions, formal nouns such as no or koto represent the speaker's proposional attitudes. no refers to the ‘situational’ entity around the speaker and implies that the judgement is based on facts, while koto objectifies the propositional content, and implies that there is a greater psychological distance between the speaker and the content.