输出修正策略、任务类型与语言学习

Output Modlfication,Task Type and Language Learning

文章以Long的"交互假说(Interaction Hypothesis)"、维果茨基的"社会互动论"和认知心理学的注意理论为基础,考察了中国英语学习者所使用的输出修正策略种类以及任务类型对修正策略和语言的正确使用的影响。第一个观察研究的结果表明,外语学习者根据言语表达时的反馈信息共采取了10种输出修正策略:确定性检查、澄清请求、理解检查、自我纠错、他人纠错、重复、支架式补充、元语言式讨论、请求帮助、主动提供解释等。这些策略能促成和改善言语表达,且有助于学习者中介语系统的重构。第二个实验研究探讨了任务类型对单向和多向支架式补充策略的使用量和语言的正确使用量的影响。研究发现:(1)从总数上看,语言信息必换的封闭性单一目标任务引发受试采用了更多的支持性修正策略;(2)封闭性单一目标任务也促使学习者使用了更多的双向支持性修正,使学习者更多地把所修正过的话语正确地整合到其语言使用中;(3)语言信息不必互换的、开放性多重目标任务却使受试采用了更多的单向支持性修正,同时学习者在延时测试中能正确使用话语的比例相对较少。

On the basis of Long＇s Interaction Hypothesis Vygotsky＇s Social Interactionism and the theory of noticing from cognitive psychology the research explored types of output modification moves used by Chinese-speaking learners of English and the impact of task type on modification moves and correct language production in L2 learning.The results from the observational study showed that the Chinese-speaking learners of English used 10 types of modification moves in their interaction with peers including confirmation checks clarification request comprehension checks self-correction other-correction repetition scaffolding metalinguistic talk solicitation of assistance and unsolicitated explanation.We believe that these modification moves may push＇learners to improve language production and may help them to reconstruct their interlanguage system.The second study examined the effect of task type on the amount of mutual scaffolding and one-way scaffolding moves used and correct use of language which produced these finding a in contrast to the open optional information exchange task with divergent goalsthe close required information exchange task with convergent goal pushed the learners to use greater amount of scaffolding move b in addition the close required information exchange task also elicited more mutual scaffolding move which helped the learners to integrate greater amount of language into their interlanguage system and c the open optional information exchange task brought about more one-way scaffolding move and lower percentage of correct language integration in the delayed test.