【作者】 张凯；

【导师】 王同顺；

【作者基本信息】 上海交通大学 ， 外国语言学及应用语言学， 2016， 博士

【摘要】 纠错反馈是二语习得研究的重要议题。众多研究结果表明,无论口头反馈还是书面反馈均对部分形态语法形式习得有效。然而近年来,相关研究也发现,反馈并非是促进二语习得的唯一变量,其效果同样受到相关变量的交互影响(Ellis,2010;Bitchener&Ferris,2012),从而使学界围绕该议题的争论日渐增加。通过探讨相关变量对反馈效果的影响,已有研究证明不同反馈模式和类型会影响学习者对目标语法形式的注意程度,并继而导致不同的语言学习效果。但是,对其促学效果的研究却因反馈的不同模式和类型而结论迥异。此外,由于句法的复杂特性(Celce-Murcia&Larsen-Freeman,1999;Shintai et al.,2014),关注句法错误的反馈研究相对较少,针对句法内部特征(比如过去虚拟语气主从句)的研究则更为鲜见。为了更好地解释反馈的纠错效果,将句法结构和学习者语言水平纳入反馈研究的变量范畴已成学界共识。本研究基于Ellis(2010)纠错反馈研究框架,多维度探讨反馈模式(口头vs.书面)、反馈类型(输入型vs.输出型)以及学习者对纠错反馈的注意程度和其所持态度对纠错反馈效果的影响;并注重考察不同类型纠错反馈对不同语言水平的学习者过去虚拟语气主从句习得发展的影响。本研究旨在回答如下研究问题:1)口头纠错反馈和书面纠错反馈分别对中国英语学习者习得英语过去虚拟语气有何影响?2)不同反馈类型对纠错反馈的效果有何影响?3)书面反馈结合写作练习是否存在长时效果(特别是在后续写作中)?4)中国学习者的英语水平对他们习得过去虚拟语气主从句会产生怎样的影响?5)中国英语学习者对目标语的注意程度以及对教师实施的纠错反馈所持态度,对其习得过去虚拟语气有何影响?为了解决上述问题,本研究共设计了两项实验,采用前测和后测的实验设计以及语言重构任务作为实验干预。182名中国非英语专业大学生分别编为口头输入组、口头输出组、口头控制组、书面输入组、书面输出组和书面控制组,参与实验一。100名中国非英语专业大学生编为书面直接反馈组和书面间接反馈组,参与了实验二。随后,本研究将所收集到的材料进行口头转写,并对文本材料、语法测试、写作测试、语法判断测试成绩以及事后问卷结果进行分析。本研究主要发现如下:1)纠错反馈的效果与反馈模式和类型相关。实验一数据表明,书面输入纠错与口头输入纠错两组之间存在显著差异,这说明纠错模式(口头vs.书面)对纠错效果产生一定影响,这种差异主要是受到不同模式的显性特征以及加工负荷的影响。输出型反馈组在准确率上均优于输入型反馈组,这种优势在后测中更为显著。这也证明输出型纠错反馈在促进学习者自我修正和深层加工方面优于输入型纠错反馈。通过自我修正,学习者能够对现有的知识体系进行反思,并实现对其语言产出的有效监控。同时,通过加大输入频率和规则阐释,降低认知负荷等手段,输出型反馈可以激发学习者有意识的学习,促使其认知加工的深层发展,从而增强他们知识记忆的持久性。2)纠错反馈能够产生长时效果。除口头输入型反馈组,其他口头和书面反馈组受试在语法准确性方面都有显著提高,表明显性反馈有助于中介语的重构。同时,写作测试后测结果显示纠错反馈在后续写作中能够保持这种学习效果,这表明通过反复练习和知识提取,书面反馈具有内化和巩固陈述性知识的潜在作用。3)纠错反馈有利于学习者对句法结构的习得,但习得效果受到句法内部结构复杂性差异的影响。实验二结果显示,过去虚拟语气从句习得效果优于主句习得,证明句法内部结构存在不同的发展阶段和语法复杂度。这表明学习者习得语法形式需按照一定的顺序,只有在学习者中介语水平发展到一定阶段才能习得该语法形式,在未发展到适当阶段时则无法从反馈中获取相关信息。4)学习者英语水平对反馈效果产生影响。研究发现高水平学习者能够受益于两种不同的反馈形式,而间接反馈对于低水平学习者作用更加明显。这可能是因为间接反馈通过划出错误区域、指明错误类别以及提供规则阐释等反馈功能降低了任务难度,为学习者提供了可理解性的语言形式输入,并推动学习者自我修正,从而加深其对正确语言形式的记忆。而直接反馈的纠错性并未被低水平学习者领会,他们仅仅注意到教师的反馈,而不能理解反馈的内容。5)反馈效果不仅与注意多少有关,更与注意程度有关。本研究中所有的反馈类型都能够帮助学习者或多或少注意到目标语法形式,但其注意程度却因反馈类型不同而有所差异。与隐性和输入型反馈相比,显性反馈和输出型反馈有助于提升学习者对目标语的有效注意,使他们更好地注意到当前语言结构与目标语结构之间的差距,并最终影响到反馈的有效性。同时,被试对纠错反馈的态度没有显著差异,说明中国外语学习环境下的学习者对于纠错反馈普遍持积极态度。总体而言,纠错反馈不仅能促进英语句法结构的显性习得,其学习效果在后续写作中也能得到保持和加强,这一发现拓宽了纠错反馈研究范畴,并为纠错反馈的长时效果提供了有力证据。此外,本研究不仅证明了口头和书面反馈均受到反馈模式、反馈类型和对反馈的注意程度等相关变量的影响;同时,书面反馈还受到语言特征、学习水平的交互影响。本研究在纠错反馈促进句法结构习得以及相关调节变量上的发现对于未来的理论和实践研究都具有重要参考价值。更多还原

【Abstract】 Corrective feedback(CF)has become one of the foci of Second Language Acquisition(SLA)research.It is now widely recognized that either oral CF or written CF is beneficial to the acquisition of certain morphosyntactic forms.However,empirical studies have found that CF per se is not the only variable enhancing the effect of L2 learning,and some other variables are very likely to be involved(Ellis,2010;Bitchener & Ferris,2012).To examine the impact of relevant variables that may mitigate the efficacy of CF,a substantial number of studies have been conducted,demonstrating that different modes and types of CF may result in various degrees of noticeability of target features,thus possibly leading to differential learning effects.However,the extent to which the modes and types of CF can influence acquisition is still a controversial issue.Moreover,less attention has been paid to the effect of CF on syntactic features given their complexity and difficulty(Celce-Murcia & Larsen-Freeman,1999;Shintai et al.,2014).To date though,few empirical studies have been conducted to explore the relative effects of CF on Chinese learners of English in their learning the main and subordinate clauses of past hypothetical conditional.Finally,given the paramount importance of individual differences,learners‘ L2 proficiency also needs to be taken into account to investigate how the effects of CF types on L2 learning interact with those of L2 proficiency.Based on the componential framework proposed by Ellis(2010),this study attempts to explore the effect of CF on L2 learning from multiple perspectives,including the effects of CF modes(oral vs.written),CF types(input-providing vs.output-prompting),and the interactive effects of learners‘ proficiency,grammatical features as well as noticing of CF and attitude to CF on Chinese EFL learners‘ acquisition of past hypothetical conditional.Accordingly,five research questions are addressed:1)What are the effects of oral and written CF on Chinese EFL learners‘ acquisition of past hypothetical conditional?2)To what extent does the effectiveness of CF vary with CF types?3)Do the effects of written CF incorporating writing practice on Chinese EFL learners‘ acquisition of past hypothetical conditional endure over time,especially in the subsequent pieces of writing?4)What are the effects of written CF on high-and low-proficiency Chinese EFL learners‘ acquisition of the main and subordinate clauses of past hypothetical conditional?5)To what extent do Chinese EFL learners‘ noticing of target forms and their attitude to teacher‘s CF moves affect their acquisition of past hypothetical conditional?To address these research questions,the present study adopted a pretest-posttest research design and conducted two experiments involving the use of reconstruction tasks.One hundred and eighty-two adult EFL learners at the tertiary level participated in Experiment 1,who were assigned to the six treatment conditions,namely,oral input-providing,oral output-prompting,oral control,written input-providing,written output-prompting,and written control.One hundred adult EFL learners participated in Experiment 2,who were assigned to one written direct CF group and one written indirect CF group.The collected data included oral transcripts and written texts,scores of grammar tests,writing tests,and grammatical judgment tests as well as the data from exit questionnaires.This study has obtained the following findings.First,the effect of CF is found to be related to CF modes and types.Specifically,based on the results of Experiment 1,the written input-providing CF was more effective than the oral input-providing CF,suggesting that CF mode has an impact on the effects of CF.The effectiveness of the written CF may be affected by its more explicit corrective force and less cognitive load than the oral CF.Besides,the output-prompting CF groups outperformed the input-providing CF groups in terms of the improvement of accuracy rates.Meanwhile,the former groups also achieved more significant gains than the latter groups in the posttests.This suggests that the output-prompting CF type be more conducive to promoting self-repair and deep processing than the input-providing CF type.The self-repair induces learners to reflect on their existing knowledge and help them acknowledge the suggested solution.In addition,the output-prompting CF type also engages learners in noticing the gap and fosters deep processing through the intensified frequency and rule explanation,thus reinforcing the knowledge in long-term memory.Second,it has been evidenced that CF can facilitate EFL learners‘ long-term grammatical improvement.Both the oral and written CF types(except the oral input-providing CF)significantly improved grammatical accuracy,suggesting that the explicit CF is helpful for interlanguage restructuring.Meanwhile,the results of writing tests showed that the written CF could help L2 learners better use the target features in the new pieces of writing than the control group,thus validating the potential for learners to internalize and consolidate their declarative knowledge gained through written CF together with the practice and repeated retrieval.Third,CF is proved to be beneficial to the acquisition of syntactic features but such effects may be mediated by the inherent nature of the syntactic features.Results of Experiment 2 showed that the mean scores of the subordinate clause of the past hypothetical conditional were significantly higher than those of the main clause.This validates the claim that the two syntactic structures are in fact different in the acquisition order(Chou,2000).It demonstrates that written CF can only be effective if it is directed at grammatical features that the learners are ready to acquire.Fourth,the effect of written CF is mediated by the learners‘ language proficiency.It was found that the high-proficiency learners benefitted from both CF types in the acquisition of main and subordinate clauses of past hypothetical conditional.Meanwhile,the low-proficiency learners benefitted from the indirect CF most in learning the target features.The possible explanation is that the indirect CF helps simplify the task,reduce the cognitive load and provides the learners with comprehensible input by pointing out the locus and nature of the error and providing rule explanations.In addition,it induces the repair or output which assists learners in noticing the target features and noticing the gap.Such a sequence of repairs can be helpful for the storage of the target features in long-term memory.However,the learners in the direct-low group may simply notice the error correction but do not understand the corrective intention.Fifth,there is a relationship between the effectiveness of CF and the degree of noticing.CF involved in the present study could direct the learners‘ attention to the target features,but the degree of noticing varied with the type of CF.Compared with the less explicit CF and the CF which offers no opportunities for repair,the more explicit CF and the CF which offers more opportunities for self-repair tend to be more powerful in directing learners‘ attention to the discrepancies between the erroneous form and the target form.However,the investigation into learners‘ attitude to CF failed to find significant between-group differences with respect to the relevant variables,revealing that learners in the EFL context generally hold positive attitude to teachers‘ correction.In all,the findings of the current study have proved that CF helps to develop the explicit knowledge of syntactic features.Meanwhile,the learning effect can be sustained and consolidated in the subsequent writing,thus providing evidence for the extension of the CF research scope and the effects of CF on long-term improvement.Besides,this study has found that the effectiveness of CF varies with other variables including CF mode,CF type,grammatical feature,L2 proficiency level and the noticeability of CF.The findings of the present study has important implications for EFL theory and practice in the Chinese context.更多还原