【作者】 王奕凯；

【导师】 刘兵；

【作者基本信息】 太原理工大学 ， 外国语言学及应用语言学， 2014， 硕士

【摘要】 本研究基于收集研究生英语写作课的作业而建成的语料库,通过对比中介语分析方法(1)比较自建语料库和母语参照语料库,考察学习者在写作中的英语使用特征,并将对比结果应用于写作反馈；(2)通过历时比较,监控观察学习者语言发展过程,考察基于语料库的反馈在学术英语写作教学中的应用。本文报告了在学术写作教学中,根据学生作业建成自建语料库,通过语料库分析并应用于课堂反馈的实验研究。研究基于的写作课程每隔两周一次,每次2学时。受试者(N=205)来自两个学术英语写作课堂,学生是一年级研究生,来自不同专业。本研究采用混合研究方法。定量分析基于自建的语料库,由5个子库构成,每个子库包括一次学生作业。因为学生专业方向不同,所以作业依然采用比较普及的主题,但要求采用问题-解决型模式的写作。比较分析发现：发现：(1)中国学生和英国学生语言使用存在差异,具体在冠词、人称代词、情态动词、形容词等方面都有体现;(2)中国学生已经基本掌握问题-解决型标识词的使用,但在种类和具体用法上都比较有限；(3)比较5次学生的作业,呈现一个有趣的趋势。如果研究者发现学生有一个词使用频次显著少于母语使用者,在课堂反馈时,就把这一现象以图表的形式呈现给学生。在下一次作业,就会看到这个词的使用频次急剧上升,但是在接下来的几次作业中,又会出现频次慢慢下降的现象。同样,一些过度使用的词的变化正好相反。定量数据主要是基于第一次和最后一次写作课的开放性问题的回答。整体上说这次试验成积极效果。大部分学生认为基于语料库的反馈新奇有效,注意到了自己和母语作者的作文在词汇方面存在的差异,认为基于语料库分析的反馈帮助他／她们改进了写作并愿意将来继续学习使用。通过就一些词在语料库之间的表现以及学生的态度变化,本研究证明在高水平语言使用者的写作课堂引入基于语料库的反馈教学具有一定效果。研究结果有助于了解在写作中学生词汇发展特征,改进了写作教学方法。但限于中介语分析方法的本质决定,其只能发现“有”的特征。因此在本研究中表现出的学习者语言特征可能只是一小部分。此外由于教学课时和学生不同专业的限制导致研究样本偏小,主题单一,未能深入研究课堂中语料库的其他应用。希望未来研究可以突破这些限制,丰富语料库在写作课堂中的应用。更多还原

【Abstract】 This study investigates corpus-based feedback in teaching of academic English writing for Chinese English learners by Contrastive Interlanguage Analysis on the basis of a self-compiled written corpus consisting of assignments of Chinese postgraduate students from non-English majors with two reference corpora:the BNC and the BAWE. The study consists of two parts:(1) investigation of English usage of learners in writing;(2) application of corpus-based feedback in academic writing teaching and monitoring the language development of the learners.The paper reports the analysis and effect of an experiment to teaching academic writing in which students got feedback which was based on the data and results from the corpus analysis of their own assignments. The course was made up of6weekly2-hour sessions. The participants (N=205) came from2multidisciplinary classes of first-year postgraduate students. Mixed methods were used in the study. Quantitative analysis was concentrated on the self-compiled learner writing corpus consisting of5-session assignments with Problem-Solution Pattern. On the contrast analysis, the results showed:(1) There are differences between Chinese students and British students in terms of use of articles, personal pronouns, modal verbs, adjectives, nouns, etc.(2) Students already mastered some signals for Problem-Solution Pattern writing, but the types and flexibility were restricted.(3) On the analysis through5-time homework, an interesting tendency occurred. When the researcher found an underuse word and showed it to the students on the feedback period in the classroom, a dramatic increase of the use of the word could be found in the next assignment. But in the following assignments, we could watch a slow opposite tendency of the data. The overuse situation was the opposite.Qualitative data, collected from two survey questionnaires in the first and last session, showed generally positive results. A majority of students found it new and effective to use corpus-based feedback in the writing course. Most students noticed the differences between their own writing with that of native speakers and were enthusiastic about working with the corpus and agreed that the corpus-based analysis helped them improve their writing and intended to learn it more in future.The study demonstrated the effect of the corpus-based feedback in the English writing course in terms of specific words and students’ attitude to the new way of feedback and writing teaching with corpus tools. The results have some implications for understanding lexical development in L2academic writing as well as improving writing pedagogy. Due to the nature of Contrastive Interlanguage Analysis which could only find out existing problems, the distinctive features found in this study might be only part of the real situation. In addition, the short teaching period made it impossible to further explore the effect of the corpus and other corpus usages in classroom. However, the study practiced one corpus use in classroom, which could enrich the writing teaching with corpus.更多还原