中国学习者二语花园幽径句解读研究

The Interpretation of L2 Garden Path Sentences by Chinese⁃speaking Learners

本研究通过操纵二语句子是否存在花园幽径现象、子句间语用合理性和歧义区域长度,采用阅读从句—主句复合句后回答一般疑问句的任务,考察中国学习者如何理解英语花园幽径句。结果发现,受试采用渐进方式阅读二语句子,遵守延迟关闭原则,花园幽径句理解的正确率显著地低于非花园幽径句;语用合理性不影响二语花园幽径句的理解;与长歧义区域条件中句子阅读相比,短歧义区域条件中句子阅读的正确率更高。这些结果不支持题元结构选择假设,但是支持重新分析模式:在二语句子理解中,句法解析器和语义处理器各司其职。研究结果对二语阅读教学有启发意义。

This paper reports an experimental study on the effects of L2 formal characteristics and pragmatic information on the syntactic analysis of temporally ambiguous sentences,specifically garden path sentences.The results indicate that the relative plausibility of two possible real world events did not influence the language processor's choice of ultimate syntactic analysis of an ambiguous string;namely,clear garden⁃path effects were observed in both relatively plausible and implausible reduced relative clauses;while no garden⁃path effect was observed for controlled sentences.Furthermore,the ambiguous region length played a significant role with regard to the ease with which Chinese⁃speaking L2 learners interpreted garden path sentences:the longer the L2 parser was committed to the ambiguous region,the more difficult it was for L2 learners to recover from the initial misanalysis.These re⁃sults do not endorse the thematic selection hypothesis,the essence of which is that pragmatic plausibility has an effect on the pro⁃cessing of sentences in such a manner as applying the initial syntactic parsing of an input constituent to identify major phrases,the language parser then makes use of real world knowledge and information about the current discourse to compare the pragmatic plausibility of whatever sets of relations between phrases are listed in the thematic structures associated with the head of a phrase;if the thematic selection process results in a set of relations which is incongruent with the selected syntactic parsing of the input,reanalysis will be attempted.However,the results support the selective reanalysis hypothesis,or the thematic reanalysis model,the essence of which is the proposal that the processor does not automatically proceed through the sentence forward or backward no matter what type of error is involved;rather,the processor makes use of whatever information indicates that the initial processing is not appropriate to attempt to diagnose the source of its error;if successful,this would permit it to selectively focus on just the fraction of the parsing which was responsible for the particular problem it encountered with the initial analysis.Furthermore,our results suggest that there exist distinct syntactic and semantic processors in the mechanism of L2 sentence comprehension.The re⁃search provides the pedagogical implications in L2 reading instruction.