Stylistic Inversion and Information Structure in Subordinate Clauses of French

本研究の目的は，フランス語の書き言葉で頻繁に用いられる文体的倒置が従属節中に出現する場合において，様々な従属節における倒置の頻度を明らかにし，前提を表す従属節における倒置の要因が「情報の重要度」によって説明できることを示すことである. 従属節には断定を示すものと前提を示すものがある. Lahousse (2011)は理由や譲歩を表す副詞節と非制限的関係節を断定を表す従属節とし，その節内の倒置は主語の非トピック化によって起こると示した. この制約によって，断定を表す従属節中の倒置の頻度は前提を表す従属節よりも低いと指摘したが，詳細なデータは提示されず，前提を表す従属節については倒置の要因の十分な説明がなされなかった. そこで本研究ではコーパスFrantextを用いて，様々な従属節の接続詞ごとに倒置の頻度を明らかにした. その結果，断定を表す従属節のうち，理由を表す従属節と譲歩を表す従属節における倒置の頻度が最も低く，非制限的な関係節については，制限的関係節と比較すると倒置の生起頻度がより低いことを明らかにした. また，前提を表す従属節中の倒置は「情報の重要度」という概念を用いることで説明が可能となり，主語が担う情報の重要度が動詞の担う情報の重要度よりも相対的に高い場合に倒置が起こりやすいことを示した. 前提を表す従属節でも情報の重要度に沿った語順が選択されうることを明らかにした.This study aimed to investigate the frequency of stylistic inversion in various subordinate clauses in French and to analyze the cause of inversion in presupposed subordinate clauses. Lahousse (2011) identified causal or concessional clauses and non-restrictive relative clauses as assertive and found that inversion in assertive clauses is caused by non-topicalization of the subject. She pointed out that the frequency of inversion in assertive clauses is lower than that in presupposed clauses because of this constraint. However detailed data was not included and the cause of inversion in presupposed clauses was not fully explained. Therefore this research extracted data from the corpus Frantext and analyzed the frequency of inversion quantitatively for each conjunction of various subordinate clauses. As a result for assertive clauses the frequency of inversion was the lowest in causal and concessive clauses. In addition the frequency of inversion in non-restrictive relatives was lower than that in restrictive relatives. Furthermore this study revealed that inversion in presupposed subordinate clauses occurred more frequently when the importance of information conveyed by the subject was relatively higher than that carried by the verb. Accordingly inversion in presupposed subordinate clauses occurs depending on "the degree of importance of information.