An Application of the Bilingual Asymmetry Model to Japanese: A Dissociation Between Kanji-to-English and Kana-to-English Translation

機関リポジトリ Web Site

書誌事項

タイトル別名
  • バイリンガル非対称モデルの日本語へ適用:漢字から英語への翻訳と仮名から英語への翻訳との間の乖離
  • バイリンガル ヒタイショウ モデル ノ ニホンゴ エ テキヨウ カンジ カラ エイゴ エノ ホンヤク ト カナ カラ エイゴ エノ ホンヤク トノ アイダ ノ カイリ

この論文をさがす

抄録

論文(Article)

This study addresses two questions: the first question is about whether backward translation (i.e. the second language to the first language or L2-to-L1) exploits a different mental process from forward translation (i.e. the first language to the second language or L1-to-L2). A qualitatively difference has been proposed by Kroll and Stewart's (1994) asymmetry model of bilingual memory representation: forward translation is largely conceptually mediated whereas backward translation is largely lexically mediated. Thus, the model predicts that backward translation will be faster than forward translation and also both L1 and L2 picture naming. We have examined this hypothesis using Japanese learners of English, who have rarely been taken up so far in the research of the bilingual memory. Our results of the backward translation task by Japanese-English bilinguals show that backward translation is conceptually mediated like forward translation, a contradicting result to Kroll and Stewart's model. The second question is how kanji (Japanese logographic script) will be processed in the forward translation, i.e., whether kanji script accesses L1 lexicon (the set of L1 phonological labels) through or without phonological mediation. Our experiments show that kanji in the first encounter is most likely processed via phonological mediation. Thus we have obtained decisive evidence regarding the phonological-mediation controversy in the area of kanji recognition.　Keywords： asymmetry model, forward translation, backward translation

収録刊行物

キーワード

詳細情報 詳細情報について

問題の指摘

ページトップへ