|Abstract: ||近二十年來，英語作為共通語(ELF)在學術場所中的研究已如雨後春筍般的發展，同時，相關學者也提出許多對於傳統英語教學的挑戰，包括對於將語言學習者視為合理的語言使用者或者將學習者錯誤視為變異等看法，雖然如此，英語作為共通語與學術寫作及其教學法尚未受到重視，有鑑於此，本研究旨在探討(1) 英語作為共通語情境之下兩個句法規則(冠詞與過去式標記)的變異情形以及(2)教師對於這兩種變異用法在學術寫作中的看法。針對第一個研究目標，研究者從一門大學通識課蒐集188篇的反思短文以及從一門研究所研究與寫作的課程蒐集84篇研究草稿，利用VARBRRUL的統計方法進行迴歸分析。在冠詞變異使用方面，結果發現聽者的背景知識、名詞的可數性和在子句中的位置對於冠詞使用的變異有影響，進而歸納非標準冠詞使用可主要歸因於母語遷移或語用及篇章效應。此外，就過去式標記變異方面，結果顯示當動詞類型為BE動詞或規則動詞時，ELF 使用者傾向使用現在式而非過去式，代表過去式與現在式的語音差異程度對過去式標記有顯著影響，另外，當動詞時貌為狀態動詞時，使用者偏好使用現在式而非過去式，代表動詞本身內涵靜態性、持續性與無終點性影響使用者過去式標記的情形。了解語言內部對於變異用法的影響後，研究者進而訪問十位在台灣曾經利用英語作為共通語授課的大學教授來了解他們的看法，透過內容分析，結果發現所有的受訪者對於文本中出現這兩種變異用法的情形均不會造成理解上的問題，但對於過去式的變異用法稍有負面觀感，此外，受訪者也透露幾個風格上的因素會影響是否接受變異型態句法的使用，這些因素包含課程本質、言談社群、寫作結果風險及班級人數，如果課程為非研究導向、言談社群相對不開放、寫作風險低與班級人數多的情形，英語作為共通語的變異特色可以較為接受，反之，如果課程為研究導向、言談社群為開放、寫作風險高與班級人數少，這些因素則影響受訪者不能接受英語作為共通語的變異特色。另外，就審查投稿期刊而言，大部分受訪者認為統一標準還是需要存在，這些看法顯示在討論英語作為共通語及學術寫作時，言談社群的概念必須納入考量，具備使用符合特定言談社群的語言使用能力才能被視為圈內人。總結來說，變異用法並非混亂的錯誤而是具有系統性、有規則的自然用法，因此這些變異用法應被視為合理的變異情形。採取這樣包容觀點並結合質化訪談中所獲得的結果，研究者提出教師與學生首先需要對英語作為共通語有態度上的轉變，透過語料庫、接觸具有共通語變異特色的文章或者同儕評閱等教學活動，學生與老師雙方皆能對英語作為共通語的現象有正面的概念。另外，老師也需要提醒學生學術寫作中風格因素的限制，在哪些情境之下，變異型態句法能夠被接受，而在哪些情境則無法，在評量方面，研究者建議教師需要先區分正式/非正式的用法，再區分合文法/不合文法的用法，最後區分標準用法/變異用法。|
The past two decades have seen a tremendous amount of research on ELF spoken communication in academic settings. Substantial reconceptualizations have also been proposed to challenge the traditional concepts in English teaching, including rethinking language learners as legitimate language users and learners’ errors as different uses. However, studies on ELF academic writing and its pedagogy are still sparse. This research, therefore, aimed to investigate (1) the variation of two ELF morphosyntactic features, variant use of articles and nonstandard past-tense marking, and (2) how teachers perceived the two ELF features in academic writing. In response to the first goal, the researcher collected 188 short essays of class reflection from one general education course at undergraduate level and 84 draft research papers from one research and writing class at graduate level. The written texts were analyzed by a variable rule analysis (VARBRUL) to show systematic patterns of the two ELF features. In terms of the variation of the variant use of articles, the results showed that NPs assumed to be known to the hearer (+HK) and countable plural nouns favored zero article when the was supposed to be used. +HK and clause initial position favored zero when a was supposed to be used. +HK and clause initial position favored the when zero was the obligatory article. In general, the findings suggested that L1 transfer effect and pragmatic and discourse effects contributed to the variant use of articles. With respect to the variant use of past-tense marking, the results showed that when verbs were copula and regular verbs, they favored present tense instead of past-tense marking. This indicated that phonetic saliency had impacts on the use of past-tense. In addition, when verbs were state verbs, they also favored present tense rather than past-tense marking, suggesting lexical aspects had effects on the variant use of past-tense marking. To address the second question on teachers’ perceptions of the two linguistic features in writing, ten professors who had experience in using ELF for teaching were recruited for interviews. By using content analysis, the findings revealed that all of the interviewees showed no incomprehension when variant use of articles and past-tense marking were present in the texts, but more negative attitude was revealed toward the past-tense marking variants. Moreover, several style-related factors that could influence the acceptance of ELF were identified, including the nature of the course, discourse communities, stakes of writing activities, and class size. It is found that when courses are not research-oriented, have local discourse communities, low-stakes writing activities and large student number, the use of ELF features are acceptable. However, when courses are research-oriented, have global discourse communities, high-stakes writing activities and small student number, standard English is more preferred. In addition, in contexts such as journal publication where ELF is the common situation, standard forms are still the norm in that it is the practice of a discourse community. Drawing insights from the quantitative and qualitative analyses, the researcher proposed some implications for ELF writing pedagogy. First, it is argued that the variant usages should not be seen as chaotic errors but natural uses. Accordingly, in line with the major claims in ELF paradigm, variant usages in the present study should be seen as variants but not simply errors because they have systematic and rule-governed patterns. To incorporate this idea into ELF academic writing, it is suggested that using corpus, texts with ELF features, and peer-reviews are useful resources that should be introduced to teachers and students for an attitude shift. In addition, students should be informed to be sensitive to the stylistic factors and the existing standard so that they could select which code is more appropriate. Finally, with respect to assessment, it is suggested that teachers need to distinguish between formal and informal styles at discourse level, followed by grammatical/ungrammatical uses and standard/variant forms.