出版社：Japan Science and Technology Information Aggregator, Electronic
摘要：“Modifier Noun no Noun” constructions in Japanese ( e. g. chiisana kodomo no te ) are structurally ambiguous in that the Modifier can modify either the first Noun or the “Noun no Noun” as a whole. But in actual use they are not usually felt to be ambiguous, because we can understand the unambiguous messages underlying them by making use of world knowledge and contextual information which are stored in our memory. The purpose of this paper is to construct a model on the process of disambiguating these constructions. The model proposed in this paper is based on the analysis of many examples found in literary works etc., and invokes the integrated processing hypothesis proposed by Schank & Birnbaum (1984) and the hierarchical memory model proposed by Tanaka (1976). In section 1 are discussed those cases which can be disambiguated out of context, whereas in section 2 are discussed those cases the complete disambiguation of which requires contextual information.
