國家教育研究院

　　本研究以英美地區的知名英語辭典為考察對象，旨在探究其辭典編輯上的方法和原則，聚焦於意義的設立、例句的收錄、搭配詞、句型等，實際的考察對象以主要以辭典中的典範《牛津辭典》為主，輔以其它辭典，如柯林斯辭典、劍橋辭典、韋氏辭典、朗文辭典、麥克米蘭辭典等，本研究發現，受惠於語料庫語言學的日益發展，上述辭典的編輯都有賴於大型平衡語料庫的建立，語料庫所提供的大量真實例句讓辭典編輯變得更加省力，但同時卻能更精準的捕捉某個字或詞的正確用法，而語料庫的建立更能也能與時俱進的提供某個字或詞的演變歷程，此外，上述辭典都將母語人士使用的辭典和英語學習者使用的辭典分開處理，以滿足各自不同的需求，而不論是哪一種類型的辭典，他們都很注重使用者友善，因此他們開發出很多使用者友善的查詢和使用介面，讓使用者能更輕易上手，此外為了配合時代的脈動，這些辭典也都允許查詢結果和知名的社交網站(如facebook, twitter 等)連結，藉由分享，可以更快速的拓展該辭典的知名度與影響力，這些都可作為本國語文辭典日後編修時的參考方向和依據。語文教育是一切學習的基礎，其中最重要的基石當屬辭典，這對於奠定一個國家的語文教育起了莫大的功能，是成功的語文教育中不可缺少的一個重要環節。若能考察並借鏡其它地區的本國語文辭典編務，並從中吸取編輯的原則、做法和實務經驗，對於本國語文辭典的編輯將有莫大的助益。而辭典編輯立即可行的建議包含：(一)著手開始建立平衡語料庫、(二)著手培訓辭典編輯人員、(三)將母語人士使用的單語辭典和華語文學習者的辭典分開，此外，關於辭典編輯的中長期建議擇包括：(一)平衡語料庫必須持續更新、(二)設立專區廣納各方意見、(三)需要維持編輯人員的穩定性。

　　By examining monolingual dictionaries in the user’s native language in the UK and the US, this study aims to delve into the methods and principles for dictionary compilation. Special attention will be paid to the postulation of sememes, the collection of data, collocation, and sentence patterns. We specifically focus on Oxford Dictionary and, when necessary, compare it with other dictionaries such as Collins Cobuild Dictionary, Cambridge Dictionary, Merriam-Webster Dictionary, Longman Dictionary, and Macmillan Dictionary. Owing to the recent development in corpus linguistics, the dictionaries mentioned rely increasingly on the corpus data and are able to provide readers with more authentic examples, making the compilation of dictionaries less painstaking. It is hence easier for learners to have a clearer understanding of the usage of a given word/phrase and to sense the new evolution(s). Therefore, the importance of using corpora for the compilation of dictionaries cannot be neglected. In addition, it has now become accepted practice to distinguish between native language dictionaries and learner’s dictionaries. Both types of dictionaries place great emphasis on the principles of user-friendliness and, therefore, put a lot effort in designing websites that are easy to use, understand, and operate. Further, these dictionaries allow users to share their search results with others via popular online social networking websites such as facebook, twitter, etc. Recommendations for policy for the present include (a) the building of a balanced corpus, (b) the training of dictionary compilers, and (c) the distinction between monolingual dictionaries in the user’s native language and learner’s dictionaries. Long-term suggestions include (a) the continual renew of the balanced corpus, (b) constant interaction between dictionary compilers and users, and (c) the stability of staff.