利用語境消除英語多義片語動詞(polysemous phrasal verbs)歧義性的能力，在語言學習中通常被視為重要的一種能力。然而，儘管語境在消除岐義性扮演重要角色，在英語作為外語(EFL)的臺灣英語課堂中，將多義英語單字或詞語等視為可獨立於語境且在教學中直接提供其中文翻譯，仍是普遍的做法。雖然此種教與學的方法可歸於英語教師在有限教學時間下所產生的權衡，這種直接提供中文翻譯的做法引發了一個疑問: 直接提供多義片語動詞中文翻譯的教與學，可能會如何阻礙學習者利用語境消除多義片語動詞歧義性的能力。本研究利用數據驅動學習的詞意區別任務並比較兩組學習者在詞意區別任務的表現，來回答上述的疑問。研究採用準實驗設計，實驗參與者為70名來自桃園某公立高中的十年級學生。這些學生由兩個不同班級所組成，兩班學生隨機被分配到對照組以及實驗組。這兩組學生均被邀請完成詞意區別任務。在為期三週的任務中，對照組學生沒有多義片語動詞的中文意思資源可以參考，僅能透過語境嘗試消除歧義性。而實驗組學生在前兩週則有多義片語動詞的中文意思資源可以運用；在第三週，中文意思資源則不再提供給實驗組學生。此實驗設計之目的在於探討實驗組學生在習慣有中文翻譯的學習情境下，當他們沒有中文翻譯的輔助後，其利用語境消除多義片語動詞歧義性的能力是否會受到阻礙。本研究計算實驗參與者在詞意區別任務中，正確區分多義片語動詞之不同意思的數量，並利用獨立樣本t檢定分析實驗組與對照組在任務中的表現。 研究結果顯示，在第一及第二週的詞意區別任務中，實驗組與對照組達到統計上之顯著差異。實驗組學生正確區分多義片語動詞之不同意思的數量顯著多於對照組。然而，在第三週當實驗組學生沒有中文翻譯可以參照時，他們利用語境消除多義片語動詞歧義性的表現與對照組學生並無統計上之顯著差異。 綜上所述，直接提供中文翻譯的教與學，對於是否真正提升英語作為外語學習者利用語境消除多義片語動詞歧義性的能力，長遠看來，似乎仍待商榷。實驗組在第三週的下降表現恰巧反映出學習者的好表現並不總是一定是學習的可信指標。畢竟，正是中文翻譯的輔助使實驗組學生在前兩週的詞意區別任務有較好的表現。當中文翻譯不再提供給實驗組學習者，也就是如同第三週的情境，習慣有中文翻譯的實驗組以及僅能透過語境嘗試消除歧義性的對照組之間並無顯著差異。本研究之發現與實驗後訪談學生之結果，提供了培養學習者利用語境消除英語多義性面向能力之可能方向。;The ability to use context to disambiguate polysemous phrasal verbs is generally recognized as important to the language learning process. However, even in light of the essential status of context noted in polysemy resolution, the item-oriented direct instruction such as providing Chinese translation is still often emphasized in English as a Foreign Language (EFL) classrooms in Taiwan. While this approach to teaching and learning seems to be a trade-off in view of the limited classroom time teachers are given, this typical item-oriented direct instruction practice raises a question concerning how this approach by providing the intended meaning of a polysemous phrasal verb may potentially hinder learners’ subsequent use of context to disambiguate polysemous expressions. This study approaches this question using a data-driven approach through a task called the sense discrimination task and by basically comparing two groups of learners: the item-oriented group versus context-only group. A quasi-experimental design was used to investigate and a total of seventy 10th grade students from a municipal high school in Taoyuan, Taiwan, were randomly assigned to the context-only group and item-oriented group. Both groups were asked to complete the sense discrimination task. There were three experimental sessions, each session once a week. Students in the context-only group were not provided with Chinese translations during all the three-week sessions, while students in the item-oriented group were given Chinese translations in the first and second sessions, but not in the third. The purpose of such design is to see if learners in the item-oriented group who had the item-like translation information would be hindered in their ability to disambiguate polysemous expressions when no translations are available in subsequent task sessions. To see whether there is such hindrance, the performance in the sense discrimination task by the item-oriented group is compared with the context-only group. An Independent sample t-test was used to analyze their performances on the sense discrimination task and their correct sense distinctions were calculated and analyzed. The results indicate that item-oriented group significantly outperformed students in the context-only group in the first and second experimental sessions. However, in the third session when the participants in the item-oriented group were not given Chinese translations, no significant difference between the two groups was found. This raises the question whether providing Chinese translation directly really helps to cultivate learners’ uses of context to disambiguate polysemous expressions, especially in the long run. The decline in the final session of the item-oriented group suggests that good performance may not always be a reliable indication of learning. After all, it was the item-based support which accounted for better performance of the item-oriented group but only when receiving direct translations each time. When the item-like information was withheld, no significant difference was found between the item-oriented group receiving the translations and the context-only group working with context only. The findings of the present study and the interview responses provide a possible direction for developing of learners’ capacity to use context to address polysemy.
