著者

小亀 拓也 Kogame Takuya コガメ タクヤ

出版者

大阪大学日本語日本文化教育センター

雑誌

間谷論集 (ISSN:18821111)

巻号頁・発行日

no.15, pp.45-66, 2021-03-31

---

研究論文This paper aims to examine whether Mihara's (1995) claims are supported from a quantitative viewpoint and to study why it is difficult for forms such as -kamoshirenai, nichigainai, -rashii and -darou to occur within adnominal form clauses. Mihara (1995) indicated that, even among the forms conventionally said to occur easily in adnominal form clauses, some of them result in unnatural expressions when used in the clauses. Using the BCCWJ, this research endeavors to assess quantitatively whether his analysis is valid. I believe that Mihara's(1995) results are partially unsupported, from the following results of the investigation: (i)the low frequency of the use of -kamoshirenai and -nichigainai —the forms that according to Mihara occur without issue in attributive clauses—observed in actual usage and (ii)the lower asage frequency of the attributive usage of -kamoshirenai and -nichigainai compared with that of -rashii. From the investigation results above, the research paper presents the following hypothesis: "More grammaticalized predicate forms which express conjectural judgment exhibit higher difficulty of use in attributive clauses." With this hypothesis as the basis, I explain that -kamoshirenai, -nichigainai, -rashii and -darou hardly appear in attributive clauses because these forms have been highly grammaticalized as expressing a speech act's meaning of "conjecture". On the other hand, examples of this usage do exist—albeit in small numbers—because these forms still retain its property as a content word.