【作者】 王宇红；

【导师】 戴炜栋；

【作者基本信息】 上海外国语大学 ， 英语语言文学， 2009， 博士

【摘要】 本研究以互动假说(Interaction Hypothesis)为理论框架,结合学习者注意(noticing)因素,研究不同类型的纠正性反馈(corrective feedback)对二语发展的作用。互动假说强调互动对二语习得的重要的作用,认为二语学习者和本族语者或水平较高的对话者之间的对话互动(conversational interaction),使二语习得过程中的输入(input)、反馈(feedback)和输出(output)与学习者内在能力(learner internal capacity)有机结合,从而有效促进二语习得(Long,1996)。互动论认为对话者的反馈使二语学习者有可能注意到自己的二语输出形式和目的语形式之间的差异,从而促进二语习得。反馈研究是互动型二语习得(interaction-driven SLA)研究的重要组成部分,以实证研究为主。研究背景包括自然课堂互动(natural classroom interaction)和实验室一对一互动(laboratory dyadic interaction),或描述反馈类型的分布和频度,或探索反馈的习得价值及其制约因素。涉及的理论包括输入假说(Krashen, 1982, 1985)、输出假说(Swain, 1985,1995)和注意理论(Schmidt, 1995; Robinson, 1995a, 2003)。反馈习得价值的研究旨在揭示一种或几种类型的反馈对中介语发展(IL development, IL)的作用。此类研究多为实证性,操作方法各不相同,关于反馈习得效果的结论也不尽一致,不具有普遍性。近年来,研究者们开始关注各种内在和外在因素(learner- internal and external factors)对反馈习得效果的制约作用,研究也趋向细化,但是仍存在若干不足。反馈的明显性和隐含性(explicitness/implicitness)是受到较多关注的问题之一。依此特点反馈可分为显性反馈和隐性反馈。显性反馈中研究较多的是元语言反馈(metalinguistic feedback),隐性反馈中对重述(recasts)的研究力度最大。但是由于各种原因,这些研究的结论并不具有普遍性。有关显性纠错(explicit correction)和形式磋商(negotiation of form)的研究还远远不够。对反馈这一复杂的互动行为的研究角度也较为单一。此外目前有关反馈注意(noticing of feedback)的研究还很贫乏,较少采用直接手段测量学习者对反馈的注意并对其进行定性研究。反馈习得效果仅限于学习者的简单语言项的习得和发展。鉴于以上问题,本论文从多个角度,结合学习者注意因素,研究和比较显性纠错、重述和澄清要求对二语发展的作用,以期丰富反馈研究,为此研究领域的发展提供一定支持。显性纠错、重述和澄清要求三种反馈之间既有共性又有差异。首先,从明显性/隐含性角度,重述和澄清要求都属于隐性反馈,显性纠错属于显性反馈。其次,从输入/输出角度,显性纠错和重述给学习者提供目的语输入,因此修正输出(modified output)的可能性较小,而澄清要求产生修正输出的可能性较大。Lyster & Ranta (1997)就是从这一角度,将澄清要求归为形式磋商(negotiation of form)或提示(prompts),即学习者和对话者就语言形式方面的问题进行的磋商。Lyster (1998b)认为,显性纠错显性程度最高,其次是形式磋商,最后才是重述。但是,Lyster (1998b)的研究仅仅描述了对学习者对三种反馈的领会(uptake),对它们的习得效果并未作探讨。因此本研究将三种反馈的特点因素析离出来,并探讨它们的习得效果。根据上述三种反馈的共性和差异的比较,显性纠错属于显性反馈,重述和澄清则属隐性纠错。两种隐性反馈中,重述界定为提供输入型反馈(input-providing feedback),澄清要求界定为促使输出型反馈(output-prompting feedback)。通过研究这两种不同性质特点的隐性反馈,为互动型二语习得中的输入和输出作用的研究提供一定的补充。本研究选择了结构复杂的过去时虚拟条件句为目的项,旨在揭示反馈对复杂语法项目的习得作用。本论文的研究问题如下:1.显性纠错、重述和澄清假设三种反馈对二语发展的相对效果如何?2.在对话互动过程中,学习者是否注意重述的纠错意图?学习者是否能注意澄清要求的形式磋商意图?3.重述和澄清要求这两种隐性反馈中,哪种反馈更能引起学习者对语言形式的注意?4.学习者对两种隐性反馈的注意和二语发展是否有关?5.在学习者注意到反馈意图的情况下,提供输入型的反馈和促使输出型的反馈对二语发展的效果如何?就以上问题,本论文的假设如下:假设1:显性纠错对二语发展的效果优于重述和澄清要求的效果。假设1.2:澄清要求对二语发展的效果优于重述的效果。假设2:在对话互动过程中,有些学习者会注意到重述的纠错意图,而有些学习者则不会;同样有些学习者会注意到澄清要求的形式磋商意图,而有些学习者则不会。假设3:学习者对澄清要求的注意高于对重述的注意。假设4:学习者对隐性反馈的注意和其二语发展之间存在联系。当他们准确注意到反馈意图时取得的二语发展大于没准确注意到反馈时的二语发展。针对第5个研究问题,因为当这两种反馈的意图都被准确注意的情况下,二语发展效果就取决于其中各自的输入和输出因素。因此,本研究并未就此问题提出假设,旨在根据实验结果来进行分析。本研究采用前测-后测的实验设计,反馈处理通过2次任务型的对话互动进行。每次互动包含2个任务,共用时60分钟左右。注意数据的收集使用的是刺激性回忆(stimulated recall, Gass & Mackey, 2000)。实验参与者为郑州大学73名大学二年级的非英语专业EFL学习者,语言水平为中级。受试选自4个英语自然班,处理手段在四个班中任意分配,成为显性纠错组、重述组、澄清要求组和控制组,每组人数大致20人左右。每个处理组内以小组讨论的形式完成对话任务,研究者本人不仅参与小组讨论,也提供反馈。本研究中只针对语言形式提供反馈。实验过程包括三次测试:前测、即时后测和延时后测。三次测试均包括了提示性口语产出任务(cued oral production task)和提示性笔语产出任务(cued written production task)。口试中目的项句子有8个,笔试中目的项句子有5个。记分采用百分比形式。实验程序如下:前测(第1天)→处理1(第2天)→处理2(第3天)→即时后测(第3天)→刺激回忆报告(第3天/即时后测之后的48小时内)→延时后测(第4天,处理之后的第2周)→问卷调查(第4天)。分别对三次测试的口语任务进行了录音,对两次处理进行了录像,用作刺激回忆报告的提示材料;另外对受试者的刺激性回忆报告进行了录音和定性分析,随后进行编码(coding)。最后对受试者三次测试和回忆报告的成绩进行统计分析。本研究发现,(1)显性纠错对学习者过去时虚拟条件句各个部分的发展效果明显优于重述和澄清要求两种隐性反馈。在两种隐性反馈中,重述对于过去时虚拟条件句各部分的发展具有明显的短期效果,而澄清要求的效果则表现在延时后测的过去时虚拟条件句的某一部分的笔语产出;(2)纠正性反馈能够刺激学习者对目的项的注意,在重述组和澄清要求组内都有学习者报告对反馈意图的注意;(3)澄清要求促发的学习者的注意程度高于重述,但两者的差别并不显著;(4)一方面,学习者对隐性反馈的准确注意对于过去时虚拟条件句的二语发展整体上呈现出促进效果,另一方面,隐性反馈的效果随反馈类型、测试时间和产出形式而呈现出一定的差别。注意到重述意图的学习者短期内过去时虚拟条件句得到显著提高,但是长期效果却未达到显著;对澄清要求的准确注意虽然短时期内对于过去时虚拟条件句的发展没有显著效果,但是对此结构中的if从句的笔语产出却呈现出显著的延时效果;(5)在学习者准确注意的情况下,提供输入型(recasts)的隐性反馈对过去时虚拟条件句的发展具有短期效果,促使输出型的隐性反馈(clarification requests)则呈现出部分的延时效用。本研究中显性反馈的显著效果,说明纠正性反馈自身的显性程度(explicitness/salience)对于过去时虚拟条件句这样复杂的语法结构的习得有着显著的促进作用。显性反馈较隐性反馈更能引起学习者对目的项形式方面的注意,从而促进习得。而本研究中的显性纠错除了自身的显性特点以外,还将目的项的错误形式和正确形式并置在一个话步中,这种并置使得学习者很容易地注意到正确形式和错误形式之间的差别(noticing the gap),并对两者进行认知比较(cognitive comparisons),从而促进对二语发展。隐性反馈对于复杂语法项目的习得效果较多地取决于学习者注意这一因素,而对隐性反馈的注意又受到诸如反馈类型、互动环境以及反馈的操作方式等因素影响。隐性反馈中的输出因素似乎较输入因素更能够引起学习者对反馈意图的注意,这也一定程度上印证了输出假说所提出的输出的注意功能。重述被学习者准确注意时,对目的项发展只有显著的短期效果,这说明重述注意虽然有可能引起学习者的认知对比,但是这种提供输入型的纠正性反馈所激发的认知加工深度较小,这种对目的项的浅加工可能只使其保留在短期记忆中,还不足以使目的项的正确形式保存在学习者的长期记忆中;与之相比,当澄清要求被准确注意时,这种促使输出型的反馈激发的对目的项的加工深度较大,对二语发展有长期的效果。而另一方面,本研究中澄清要求的即时效果的缺失又说明,输出型反馈对二语发展的的作用有可能受到学习者有关目的项的现有知识、目的项复杂度以及任务等因素的制约。本研究的理论意义在于,它丰富了现有的反馈研究的体系,通过结合学习者注意因素对三种不同类型的反馈进行研究,加深了我们对于反馈、注意和二语习得三者之间的关系以及期间若干制约因素的认识。尤其是将两种隐性反馈中的输入和输出因素析离出来并探究其对二语发展的作用以及其它制约因素,为输入和输出理论在反馈研究中的应用提供了新的视角,并且丰富了二语习得互动理论。本研究的发现对于互动型的二语教学也有着指导意义,强调了互动型的课堂教学中教师应结合语言项目的特点以及学习者认知能力和水平适当选择有效的反馈类型,并且以有效的方式提供反馈,最大程度的发挥纠正性反馈对二语发展的促进作用。更多还原

【Abstract】 The present study investigates the effects of different types of corrective feedback on second language acquisition in relation to learner noticing. Taking the Interaction Hypothesis (Long, 1996) as its theoretical framework, which claims that it is during interaction that input, feedback, output and learner noticing combine to drive second language acquisition forward, this study addresses the following issues: (1) the overall comparative effects of explicit correction, recasts and clarification requests on second language development, (2) the relationship between feedback types and learners’noticing, (3) the effect of learners’nociting of corrective feedback on their subsequent second language development, and (4) the role of the componential elements of recasts and clarification requests in second language development.While a large volume of studies have been done on corrective feedback, there are still gaps in this research scope. First, feedback types involved are categorically inconsistent and limited in number. Explicit correction and prompts are relatively underexplored. Second, the effects of corrective feedback are not differentiated from other instructional strategies due to combined provision of the two. Third, the componential features of feedback have not been fully explored in the account of the source of feedback effect on L2 learning. Fourth, the effects of corrective feedback on L2 development of more complex morphosyntactic items are yet to be investigated, as the linguistic items included in the studies to date have been mostly simple items. Lastly, studies on the relationship between learners’noticing of corrective feedback and L2 development using more direct measures to collect qualitative noticing data are relatively insufficient. Besides, learners’noticing of clarification requests and subquent L2 development suffers a severe lack of investigation, warranting more studies for a better understanding of the link between feedback, noticing and L2 development.The research questions to be investigated in the present study are as follows:RQ1: What is the relative effect of explicit correction, recasts and clarification requests in general?RQ2: Do learners report noticing of the intent of recasts provided during conversational interaction? Do learners report noticing of the intent of clarification requests provided during conversational interaction?RQ3: Which of the two types of implicit feedback, namely, recasts and clarification requests leads to more noticing?RQ4: Is there a relationship between noticing of feedback and L2 development?RQ5: What is the relative effect of input-providing feedback (recasts) and output-prompting feedback (clarification requests) on L2 development in the case of noticing?By addressing the research questions, the present study endeavors to fill the aforementioned gaps. Comparisons of the effect of the three feedback types are made in terms of explicitness/implicitness distinction. The effect of the two types of implicit feedback (recasts and clarification requests) are then explored in relation to learner noticing by a qualitative measure of noticing, the stimulated recall protocal (Gass & Mackey, 2000). The present study also tackles the multi-faceted features of corrective feedback as a complex phenomenon by defining recasts as input-providing feedback and clarification requests as output-prompting feedback. The developmental value of input element in recasts and output elements in clarification requests are then further explored in the case where both types of feedback are noticed by the learners. In assessing the learners’L2 development the present study chooses as its target linguistic item past hypothetical conditionals in English, a cognitively and syntactically complex structure, also deemed as the difficult structure for Chinese EFL learners.The present study adopted a pretest-posttest design. 73 college EFL learners at the intermediate level were randomly assigned to four experimental conditions (explicit correction, recasts, clarification requests and control) and carried out the task-based conversational interactions with the researcher in group work, during which corrective feedback was provided to the learners in the treatment groups upon their non-targetlike production of the target structure. L2 development was operationalized as increased targetlike production of the target structure and assessed by cued oral and written production in the immediate and delayed posttests. The learners’noticing of the two types of implicit feedback was operationalized as verbal stimulated recall report indicating their awareness of being corrected on the target structure.The present study found a significantly greater effect for explicit correction over recasts and clarification requests on short-term and long-term development of all clause categories of the target item. Recasts are effective in short term, whereas clarification requests have delayed effect. A similar pattern was also found for the effect of noticing of recasts and of clarification requests on L2 development. The study also found that clarification requests led to a higher rate of learner noticing than did recasts, although the difference between the two were not statistically significant. In the case where learners’noticing of the two implicit feedback occurred, the input-providing feedback (recasts) was found to be more effective than the output-prompting feedback (clarification requests) on the immediate incoporation of the targetlike form, whereas the effect for the output-prompting feedback took time to emerge on just one part of the target structure.The findings of the present study point to the significant role of explicitness of feedback in facilitating L2 development on complex structure. It is also suggested that the effectiveness of implicit corrective feedback is mediated by learners’noticing, which in turn is constrained by a number of learner- internal and external factors (e.g., cognitive capacity, learning context, type of feedback, operationalization of feedback, and complexity of the target form). The present study also lends empirical support for the Output Hypothesis, particularly for the noticing and metalinguistic functions of output. The differential effectiveness of input element and output element of the two implicit feedback in the case of learner noticing can be attributed to the modulation by learners’cognitive process, depth of processing, and the learners’prior L2 knowledge.The present study provides a fine-grained picture of corrective feedback within the theoretical framework of Interaction Hypothesis. It investigates the effects of different types of corrective feedback in a multifaceted fashion, taking into account the multiple features of feedback and the constraining factors. In doing so, it not only extends our knowledge of corrective feedback but also enriches the relevant theories, hence providing pedagogical implications that will hopefully feed into interaction-driven second language learning and teaching.更多还原