定式表現研究の動向と今後の課題

―日本語の定式表現研究の発展に向けて―

抄録

This paper reviews previous research on formulaic language in English and Japanese and discusses results and issues based on the adopted approaches to investigate the development possibility in Japanese research. According to the results of the previous studies, the research in Japanese is carried out by methods of descriptive research and corpus-based approach. The researcher's intuitive judgment plays a vital role in the former, while in the latter not only the intuitive judgment but also statistical methods based on frequency is adopted, so is more objective since it is based on analysis of a large amount of data. In the recent English research, on the other hand, the triangulation approach is adopted, where corpus-driven approach and psycholinguistic approach merges gradually, and the research tends to diversify. In this approach, the advantages and disadvantages of the corpus and the experiment become complementary, and it is possible to study formulaic language from multiple viewpoints. However, there are very few triangulation approach studies in Japanese. Therefore, it is necessary to investigate the formulaic language frequently used by native speakers using the corpus-driven approach to analyze grammatical and pragmatic characteristics. Also, studies on formulaic language using psycholinguistic approach are needed to clarify the mental lexicon of native speakers and learners.