Doctoral Dissertation Research: Testing the Limits of Adult SLA: A Psycholinguistic Study of Near-Native Sentence Processing

成人学习者能否通过第二语言获得类似母语的知识和能力一直是第二语言习得领域的一个重要问题。以前对这一问题的研究大多采用传统的纸笔方法，沿袭了正式的语言学传统，并得出了相互矛盾的结论。很少有研究使用现代心理语言学技术来探索如何将接近母语的语法知识应用于在线句子处理。另一方面，迅速发展的第二语言句子处理的新领域主要是遵循现有的第一语言研究，并刚刚开始回答第二语言习得特有的问题。在美国国家科学基金会的支持下，在Bill VanPatten博士的指导下，Jill Jegerski女士将进行一项学位论文研究，以回答以下研究问题：成年第二语言学习者能否获得类似母语的句法加工行为？该项目将借鉴二语习得和心理语言学的研究成果，从句子处理的角度考察语言学习的上限。具体地说，在四个自定节奏的阅读实验中，近母语人士的西班牙语表现将与母语人士的表现进行比较，这些人在墨西哥进行了类似母语的熟练程度和长期沉浸的筛选，并独立测试了离线语法能力。每一项都将测试浅层结构假说的不同预测。不同类型的句子加工将在单一的被试内设计中进行测量，阅读时间将被报告并对接近母语的个体和群体进行统计分析--这些都是方法论上的创新，因为之前的句子加工研究都没有采用这两种方法。因此，这项调查的设计和结果将促进现有知识的发展，并促进未来关于语言习得、语言处理和总体人类语言能力的实证研究。

The question of whether adult learners can acquire native-like knowledge and ability with a second language has long been of fundamental importance in the field of Second Language Acquisition (SLA). Much previous investigation of this issue has employed traditional pen-and-paper measures, following the formal linguistics tradition, and has yielded conflicting findings. Few studies have used modern psycholinguistic techniques to explore how near-native knowledge of grammar can be applied in online sentence processing. The rapidly growing new area of second language sentence (L2) processing, on the other hand, has mostly followed existing first language research and is just now beginning to answer questions unique to SLA. With support from the National Science Foundation and under the direction of Dr. Bill VanPatten, Ms. Jill Jegerski will conduct a dissertation study that addresses the following research question: Can adult second language learners acquire native-like syntactic processing behavior? The proposed project will draw on previous research from both SLA and psycholinguistics, examining the upper limits of language learning from a sentence processing perspective. Specifically, the performance of near-native speakers of Spanish, screened for native-like proficiency and long-term immersion in Mexico and independently tested for offline grammatical competence, will be compared to that of native speakers in four self-paced reading experiments. Each will test a different prediction of the shallow structure hypothesis. The different types of sentence processing will be measured in a single, within-subjects design and reading times will be reported and analyzed statistically for individual near-native participants as well as for the group--these are methodological innovations as neither has been employed in previous study of sentence processing. Thus, the design and outcome of this investigation will both advance current knowledge and facilitate future empirical research on language acquisition, language processing, and the human language capacity in general.