| Budget Amount *help |
¥1,700,000 (Direct Cost: ¥1,700,000)
Fiscal Year 1990: ¥700,000 (Direct Cost: ¥700,000)
Fiscal Year 1989: ¥1,000,000 (Direct Cost: ¥1,000,000)
|
|---|
| Research Abstract |
When we attempt to clarify the underlying mechanism of natural languages, it is necessary to explicate the information structures of discourse and contextual information as well as grammatical structures of sentences.
Linguistic studies in the past have paid attention to the syntactic structures and formal rules of sentences. However, little attention has been paid to the interrelation between grammar and discourse. In the present study, an attempt was made to clarify the interrelation between grammatical knowledge and a variety of factors involved in discourse phenomena, comparing and examining those factors which characterize the information structures of Japanese and English.
In this study, especially the following problems were investigated. :
1) Pragmatic factors involved in the understanding of individual sentences which compose a given discourse (e. g., speech-act functions, elocutionary forces, perlocutionary forces, conventional implicatures, etc.),
2) Semantic and conceptual fact
… More
ors which characterize the propositional content of individual sentences in a given discourse (e. g., truth-conditions, idiomatic concepts of phrases and lexical meanings, logical connectives, etc.),
3) Those factors which are involved in the informational processing of pragmatic inferences (e. g., conversational implicatures, extra-linguistic knowledge, script and/or frame-based knowledge, etc.).
In the present study, an attempt was also made clarify the linguistic status of those factors which characterize the underlying structure of dialogs and those characterizing the covert contexts of discourses.
The results obtained through the present research are significant in the following respects. First, the pragmatic and semantic factors clarified in 1) and 2) above will contribute to the modelling of a discourse understanding system for natural languages. This kind of modelling is to explicate the internal information processing of contextbased understanding in our ordinary communication. Second, those results obtained in the study 3) above help us understand the concrete ways in which we gain pragmatic and extra-linguistic information from the context of a given discourse and from a variety of inferences based on that context. Less
|
