主題-評論結構(topic-comment)在主題顯著(topic prominent)的語言(如中文)扮演十分重要的角色,尤其是在描述第二語言學習者的中介語時特別重要．但是有關研究,對於中國學生在學習英文時產生的中介語,並沒有對主題-評論結構提供全面且確切的描述．本研究旨在探討中國學生學習英文時所產生的中介語之主題-評論結構, 運用三種測驗來引出中介語—文法判定測驗, 引導寫作測驗, 及翻譯測驗．受試者包含台灣區域位於台北的三所國中挑選的94位國二生,以國一期末考英文成績分為三組．此外,並於政大語視中心挑選14位以英語為母語的外籍生作為本研究的控制組．本研究探討的四種主題-評論結構是由中英文主題結構對比中所產生的, 分別為主題省略(topic drop), 主題移前(topicalization), 主題置左(left-dislocation), 及雙主詞結構(double-subject construction)．本研究所運用的中介語之理論架構是參考三種第二外語習得理論所發展出來的, 分別為第一語言影響(L1 influence), 普遍語法(universal grammar)的存在, 以及附屬集合理論(the Subset Principle)． 研究結果顯示, 第一語言轉換(L1 transfer)在所有程度的學生的中介語當中都十分普遍, 但是普遍語法只有在程度最高的學生的中介語較為明顯．因此本研究推論中介語系統內部運作過程包含三階段, 開始為第一語言轉換, 接著是第一語言轉換及普遍語法並存, 最後階段為普遍語法． / The role that the topic-comment structure plays in a topic prominent language such as Chinese is very important when it comes to SLA learners’ interlanguage. However, the related research has not provided a comprehensive framework in describing topic structures of Chinese EFL learners’ interlanguage. The purpose of this study is to analyze the topic structures in Chinese EFL learners’ interlanguage elicited by three tasks—the grammaticality judgment task, the guided writing task, and the translation task. The subjects include 94 junior high school students chosen from three separate schools in Taipei, Taiwan, and they were divided into three proficiency levels. In addition, 14 native speakers from the language center of NCCU served as the control group. This study investigated four topic structures—topic drop, topicalization, left-dislocation, and double-subject construction—which were determined by a comparative study of Chinese and English topic structures. The current theoretical framework of interlanguage operation is developed from three SLA perspectives—L1 influence, the existence of universal grammar (UG), and the subset principle. Results indicated that L1 transfer was prevalent in all proficiency levels but UG was only obvious in the most advanced level. It was thus inferred that the internal operation of interlanguage system should include three stages—first L1 transfer, then L1 transfer and UG, and finally UG.

「原住民族語言能力認證」是我國為了推動原住民族的民族語言教育，在行政院原住民族委員會的策劃下，針對全國原住民各族族語能力所進行的一項資格檢定工作。本論文從語言規劃的角度評估族語能力認證制度的規劃與執行，一方面透過詞彙分析，評估認證考試題庫的詞彙設計；另一方面運用問卷調查方法，探究規劃單位、執行單位、認證委員、考生、以及非考生等五類受試者對於「族語能力認證制度」與首屆「族語能力認證考試」的各項態度。最後，對未來認證制度的實施，以及族語復振政策的方向，提供建議。 本論文第一章說明研究緣起與相關問題點。第二章描述族語能力認證實施過程的背景。第三章探討與本研究相關的理論與文獻，以利於理解族語能力認證在語言規劃和語言政策範疇裡的意義。第四章為研究設計，描述內容包括對首屆認證考試題庫中的詞彙進行分析的方法，以及對首屆認證考試的內容進行量化研究的問卷設計與資料分析方法。第五章為研究結果，提出各族認證題庫的詞彙、各族語言能力、各族語言使用情形、以及各族對族語認證制度的態度等四個層面的調查結果，並加以分析與討論。第六章為結論與建議。研究結果發現，首屆族語能力認證的擬定與實施，大體獲得各類受試者的肯定，為往後認證制度的規劃、執行與評鑑奠立了基礎。整體受試者並認為族語認證工作最重要的意義在於挽救族語流失，本研究因而建議往後的認證制度可朝此方向推展，以帶動更廣泛的族語學習與族語復振。本論文的主要貢獻在於為族語能力認證及相關語言規劃的評鑑步驟，建立初步的模式。 / This thesis aims to evaluate the planning and the first implementation of the Accreditation of the Aboriginal Languages Proficiency (AALP), which is directed by Council of Indigenous Peoples, Executive Yuan, with the view of promoting the education of aboriginal languages in Taiwan. There are six chapters in this thesis. Chapter One describes the purpose of this study and presents some related problems. Chapter Two presents the backgrounds of AALP. In Chapter Three, the related theories and studies are reviewed and discussed. Chapter Four depicts the research designs, including the methodology to evaluate the vocabulary test in the first AALP, questionnaire design to elicit subjects' opinions about the policy of AALP and its first implementation, and methods for data analysis. Chapter Five reports the results in terms of vocabulary analysis, language proficiency, language use, and language attitudes toward AALP. The last Chapter offers conclusions and suggestions. One of the major findings lies in that AALP is widely recognized, and the results of its first implementation may serve as a guideline for its future implementation. It is also believed that AALP may help to revitalize aboriginal languages in Taiwan and thus reverse the language shift related. Therefore, it is suggested that AALP should be continued by following the current model.