漢語動詞及物性轉換的二語習得研究─以英語和日語為母語之學習者為研究對象

Abstract

本論文旨於藉由論元結構之習得，探究第二語言習得之本質，了解母語系統、普遍語法及語言輸入之作用，承繼前人研究如Yuan（1999）、Montrul（2000）以及Hirakawa（2001）而來，然而這些研究尚有不足之處；此外，根據Ionin, Zubizarreta& Maldonado（2008）的冠詞習得研究，母語系統、普遍語法以及語言輸入對於第二語言習得皆有重要影響，因而開啟本論文的研究契機，將以論元結構為素材，討論以上三者對於二語習得的影響。 本論文的研究問題有三點：第一、從句法表徵的角度探究普遍語法或母語系統在習得論元結構的作用為何？第二、從構詞的角度探究母語遷移作用，當母語有構詞線索標記動詞論元轉換時，是否有助於學習者習得論元轉換？第三、探究學習者的熟練度與習得歷程之關聯性為何？ 受試對象為以英語、日語為母語之學習者，分別招募初級、中級與高級程度受試者，實驗工具為文法判斷測驗，以了解學習者是否具備漢語轉換動詞、及物動詞、非對格動詞以及非作格動詞的論元結構知識。 實驗結果顯示，在習得論元結構方面，主要為母語遷移的作用；但某些句式則需要依靠熟練度的提升，這是由於學習者需要更多的語言輸入幫助他們建立該動詞完整的語言表徵。此外，學習者在論元轉換的判斷，亦發現母語遷移作用，但母語有構詞線索並非絕對有利，而是同時具有正遷移與負遷移的現象。最後，本論文亦發現語言輸入的可獲取性對習得成果有關鍵影響，匱乏輸入遠比可得輸入難以習得，若匱乏輸入為跨語言相同之表徵，則可仰賴母語系統的支持而習得，但若匱乏輸入為跨語言相異之表徵，則習得表現將始終與母語者有差距。 在教學應用方面，選用認知理論與認知碼教學法作為教學應用的出發點，在課堂教學應注意動詞論元結構與語義的說明、加強語言正例的質量、運用語言反例彌補跨語言差異與語言輸入之不足。This study on Mandarin Chinese as a second language investigates the logical problem of second language language acquisition by focusing on the the acquisition of argument structure, following the studies of Yuan (1999), Montrul (2000) and Hirakawa (2001). Ionin, Zubizarreta& Maldonado (2008) found the L1 transfer, Universal Grammar and L2 input all play an important role in L2A. This study investigates the sources of L2 knowledge above using the L2 acqusition of argument structure alternation. The three main research questions are as follows: (1) to investigate how UG or L1 work on the acquisition of the interface of synax; (2) to investigate whether L2 learners learn better on the transitivity alternation when their L1 posesses the morphological cues; (3) to investigate how proficiency effect interacts with the developmental trajectory of argument structure and transitivity alternation. Participants were divided into experimental and control groups. The experimental groups consisted of Englisn and Japanese native speakers of Chinese learners. The experimental group was divied into basic, intermediate and advanced levels according to their Chinese proficiency. The control group consisted of Chinese native speakers. Two experiments, namely Grammaticality Judgment Task and Vocabulary Translation Task were administered to undertand whether participants have acquired the grammatical knowledge of the argument stuctures of inchoative-causative, transitive, unaccusative, and unergative verbs or not. Results of the Grammaticality Judgment Task show that L2 learners rely on L1 system to learn argument structures, but the linguistic representation is not fully constructed until the latter stage. Meanwhile, there is no evidence to show that L2 learners have a better judgment on the transitivity alternation when their L1 posesses the morphological cues, but both the positive and negative transfers exist. Also, the accessibility of inputs exerts great influence on the acquition. The negative evidence is far difficult than the positive evidence. Furthermore, if there exists a crosslinguistic difference of representation, the learners’ performance may maintain a gap from native speakers. Contrarily if there is no crosslinguistic difference of representation, learners can finaly perform native like by their L1 support. As for teaching suggestions, this study suggests that instructors should explain the argument structures and semantic features of verbs explicitly, intensify the quantity and quality of positive evidence and use the negative evidence to bridge the gap of crosslinguitic differences.