Syntactic and discourse-level constraints in native and non-native pronoun resolution

尽可能快速准确地建立代词和先行词之间的指称联系对于成功地理解句子和篇章至关重要。然而，所涉及的心理机制的本质，以及指导先行搜索的约束类型，仍然没有完全理解。无论是语言学模型还是心理语言学模型，都几乎完全是由成年母语人士的数据提供的，而且绝大多数在线加工研究都是针对英语的。目前关于代词解析的时间过程的假设在处理过程中句法和语篇层面的限制何时起作用以及它们如何随时间相互作用的问题上存在分歧。本项目将调查和比较德语母语和非母语代词解析过程中句法和话语层面约束的时间，目的是测试当前关于实时指代解析的理论模型和假设的跨语言和跨人群有效性。因此，该项目属于XPrag优先方案的核心领域“实用现象模式”，也符合该方案将对实用现象的调查扩大到非标准人口的目标。来自母语和非母语人士的数据将进行比较，以检验现有的回指解析模型的稳健性，并有助于改进。一种对时间过程敏感的实验方法，即阅读过程中的眼动监测，将用于检查在处理句法和/或篇章级信息时是否以及何时限制了对先行词的搜索。非母语人士在句法信息处理过程中往往表现出较低或延迟的敏感性，这被预测会导致非母语人士在处理回指时比母语人士更依赖语篇层面的信息。在不同人群中，句法和话语层面约束的使用或相对时间可能存在差异，这将对现有代词解析模型的普遍有效性提出质疑，同时为潜在的表征和/或计算系统的模块化提供证据。这个项目的结果将有助于促进我们对实时代词解析的心理结构、机制和语言表征的理解。该项目的研究结果将对代词解析的理论语言学方法和当前的处理模型产生影响，也将对非母语习得领域的研究人员产生兴趣。

Establishing referential links between pronouns and their antecedents as quickly and as accurately as possible is vital for successful sentence and discourse comprehension. Yet the nature of the mental mechanisms involved, and the types of constraint which guide the antecedent search, are still not fully understood. Both linguistic and psycholinguistic models of anaphor resolution have almost exclusively been informed by data from adult native speakers, and the vast majority of online processing studies have investigated English. Current hypotheses about the time-course of pronoun resolution differ with regard to the questions of when during processing syntactic and discourse-level constraints come into play, and how they interact over time.This project will investigate and compare the timing of syntactic and discourse-level constraints during native and non-native pronoun resolution in German, with the aim of testing the cross-linguistic and cross-population validity of current theoretical models and hypotheses about real-time anaphor resolution. The project thus falls within the XPrag priority programme's core area "models for pragmatic phenomena" and also meets the programme's aim of extending the investigation of pragmatic phenomena to non-standard populations. Data from native and non-native speakers will be compared in order to test the robustness of, and help improve, existing models of anaphor resolution.A time-course sensitive experimental method, eye-movement monitoring during reading, will be used to examine whether and when during processing syntactic and/or discourse-level information constrain the search for an antecedent. Non-native speakers have often shown reduced or delayed sensitivity to syntactic information during processing, which is predicted to lead to a greater reliance on discourse-level information during non-native compared to native anaphor resolution. Possible differences in the use or relative timing of syntactic and discourse-level constraints across different populations would call the universal validity of existing models of pronoun resolution into question, while at the same time providing evidence for the underlying representational and/or computational systems' modularity. The results from this project will help advance our understanding of the mental architecture, mechanisms and linguistic representations that underlie real-time pronoun resolution. Findings from the project will have implications both for theoretical linguistic approaches to pronoun resolution and for current processing models, and will also be of interest to researchers in the field of non-native language acquisition.