著書

フィルモアの提唱したフレーム意味論の立場から、動詞と関わる意味論的諸問題に多面的に取り組む日本で初めての書物である。言語表現が喚起する背景知識の総体である「フレーム」の概念を用いて、11名の研究者が、語、構文・イディオム、語形成などに関する意味論の課題に新しい答えを見出している。

構文理論の立場から、日本語のいわゆる主要部内在型関係節と逆接文を分析し、両者の統語論的、意味論的、語用論的比較を行った上で、これらの構文間の関係を論じた。

構文理論の枠組で、日本語のいわゆる主要部内在型関係節と、この構文に統語論的、意味論的、語用論的に類似する他の節接続構文とを比較分析し、構文理論が日本語の分析に有効であることを示した。

日英両言語の、数量や度合の比較に関する構文を比較対照し、フレーム意味論・構文文法の枠組みで分析した。フレームネットの新たな方向性の一つであるConstructiconという概念を用いて分析を行った。

講座日本語コーパス『6. コーパスと日本語学』監修者：前川喜久雄、編集者：田野村忠温．

In constructionist theory, a constructicon is an inventory of constructions making up the full set of linguistic units in a language. In applied practice, it is a set of construction descriptions – a “dictionary of constructions”. The development of constructicons in the latter sense typically means combining principles of both construction grammar and lexicography, and is probably best characterized as a blend between the two traditions. We call this blend constructicography. The present volume is a comprehensive introduction to the emerging field of constructicography. After a general introduction follow six chapters presenting constructicon projects for English, German, Japanese, Brazilian Portuguese, Russian, and Swedish, respectively, often in relation to a framenet of the language. In addition, there is a chapter addressing the interplay between linguistics and language technology in constructicon development, and a final chapter exploring the prospects for interlingual constructicography.

Eve Sweetser (1996)の論文 “Changes in figures and changes in grounds: A note on change predicates, mental spaces and scalar norms”の翻訳。

This chapter discusses relations between frames and constructions, based on the constructicon-building project within the Japanese FrameNet (JFN) project. The aims are: to clarify distinctions between a framenet lexicon and a constructicon; and to contribute to the on-going discussion on whether all constructions are “meaning-bearing.” I will argue that a framenet analysis involves annotating frame-based syntactic/semantic structures of words (simple words and multiwords), while a constructicon annotation pertains to describing the internal and external syntax/semantics of linguistic objects that have complex structures. While maintaining that all constructions are meaning-bearing, I will point out that meaning structures of some constructions may not involve frames and propose a frame-based classification of constructions. Finally, I will suggest that a constructicon annotation needs both semantic frames and interactional frames.