Using Cross Language Analysis to Investigate Factors for Differential Marking

这个项目使用跨语言分析来调查“差异标记”的因素。在语言学中，“差异标记”指的是名词（如主语）以特殊编码出现的形态模式。通过这种编码，说话者可以传达关于该名称的非语法信息，例如意外、不可预测性或事件中实体的意外参与。说话者不自觉地使用差别标记来包装信息。相反，似乎是语法、话语、语义和语用因素的复杂组合，预测了不同的标记，包括名词的固有属性（例如，人、animacy或计数与质量），谓语的属性（及物性、完成的动作），或名词在较长连接的言语中的位置（例如，在对话或故事中第一次提到）。该项目将包括对学生进行编码和语法分析方面的培训。语言数据、编码协议和基于python的工具将在北德克萨斯大学数字图书馆和/或通过GitHub存储库进行存档和免费访问。本项目采用一种创新的文献方法来收集决定差异标记的因素的信息。首先，被调查语言的母语语言学家将对关联语篇中的名词进行编码，以寻找与差异标记相关的因素。当非说话者分析数据进行差异标记时，微妙的含义可能会在翻译中丢失。由训练有素的母语人士编码将更准确地捕捉到说话者想要表达的意思。该项目将开发一个编码手册来标准化编码。其次，与非语言使用者小组讨论数据将用于改进编码。关于语法的小组讨论往往会唤起用法和解释的场景，而这些场景是独立研究数据的调查人员无法回忆起来的。第三，该项目将开发基于python的工具来比较不同数据集之间的差异标记因素，无论是在一种语言内还是在不同语言之间，以找到因素之间统计上显著的对应关系。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This project uses cross language analysis to investigate factors for 'differential marking.' In linguistics, 'differential marking' refers to morphological patterning where a nominal, such as a subject, occurs with special encoding. With this encoding, a speaker can communicate non-grammatical information about that nominal, such as surprise, unpredictability, or unexpectedness of the involvement of an entity in an event. Speakers do not consciously use differential marking to package information. Rather, there appear to be a complex combination of grammatical, discourse, semantic, and pragmatic factors that predict differential marking, including inherent properties of the noun (e.g., person, animacy, or count versus mass), properties of the predicate (transitivity, completed action), or the position of a nominal in longer connected speech (e.g., mentioned for the first time in a conversation or story). The project will include the training of students in coding and grammatical analysis. Language data, the coding protocol, and Python-based tools will be archived and freely accessible at University of North Texas Digital Library and/or through a GitHub repository.This project uses an innovative documentary method to gather information on factors determining differential marking. First, native speaking linguists of the investigated languages will code nominals in connected discourse for factors associated with differential marking. When non-speakers analyze data for differential marking, nuanced meanings can be lost in translation. Coding by trained native speakers will more accurately capture the meanings intended by the speaker. The project will develop a coding manual to standardize coding. Second, discussions about the data with groups of non-linguist speakers will be used to refine coding. Group discussions on grammar tend to evoke scenarios of usage and interpretation that are not recalled by investigators working on data on their own. Third, the project will develop Python-based tools to compare factors for differential marking across various data sets, both within one language and across different languages, to find statistically salient correspondences between factors.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.