Doctoral Dissertation Research: Documenting Tense, Aspect, Mood and Polarity in a Language with a Complex Verbal System

时间是一个重要的概念，渗透到日常生活的许多方面。时间是如此重要，以至于所有被证明的人类语言都至少有一种语法策略来编码时间关系。然而，大多数语言都有相对复杂的时间编码系统(通常称为时态/体系统)，它们与语法的几个特征相互作用。个体的时体系统，除了其内部的复杂性外，在不同的语言中也有很大的差异。由于已经观察到的跨语言差异，任何给定的语言都有可能对可能的时间编码策略的程度以及时态和体可以承载的可能意义的程度提供新的见解。该项目的学习语言是Iyasa(Yko)，是喀麦隆西南部沿海地区的一种班图语，约有2000人使用。班图语的时体研究尤其重要，因为这些语言的系统被认为是世界上最复杂的系统之一。因此，该项目将利用语言田野调查和语言分析的方法，对Iyasa的时体系统进行全面的文件和描述性记录，以揭示该系统的细微差别。更广泛的影响包括博士生的专业发展，公开可访问的数据，以及一名美国语言学家在喀麦隆的存在，以帮助维持和改善两国之间的学术关系。此外，该研究项目及其产品可以将语言意识和注意力带到目前科学文献中代表性不足的语言、地区和说话者社区。濒危语言的时体系统的记录在质量上差异很大，从每一个时体的单纯标签(让读者来判断每个标签应该有什么确切的含义)到几乎每一个与时体相关的细节都包括在内的描述。为了对Iyasa时/体系统进行充分详细的描述，研究人员将使用多种方法。这些方法包括(但不限于)管理本项目研究人员已经开发和新开发的时态/体问卷，分析自然叙事和会话话语中的时态/体，定向语义诱导(研究人员提供了一系列仅在有意义的方面略有不同的语境)，以及通过视频剪辑或故事板诱导的定向叙事。这个项目的另一个重要方面是，它将纳入广泛理论的见解，包括时体认知理论，这些理论摆脱了时间作为单一线性时间线的概念。该项目除了为Iyasa社区提供重要的文献资料和语言培训机会外，还将为如何改进自然语言处理的时体计算模型提供重要的见解，并有助于加强我们对人类如何能够并确实用语言编码时间的总体理解。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Time is an important notion that pervades many aspects of daily life. Time is of such importance that all attested human languages have at least one grammatical strategy to encode temporal relationships. Most languages, however, have relatively complex systems of encoding time (which are generally referred to as tense/aspect systems) that interact with several features of the grammar. Individual tense/aspect systems, in addition to their internal complexity, also vary significantly from language to language. Because of the cross-linguistic variation that has been observed, any given language has the potential to offer new insights into the extent of possible temporal encoding strategies, as well as the extent of possible meanings that tense and aspect can carry. The language of study for this project, Iyasa (yko), is a Bantu language spoken in the Littoral Region of southwestern Cameroon by approximately 2,000 people. Tense/aspect studies in Bantu are especially critical as the systems in these languages are noted as being among the most complex in the world. This project will, thus, provide a thorough documentary and descriptive record of Iyasa's system of tense and aspect, utilizing methods in both linguistic fieldwork and linguistic analysis in order to uncover the nuances of the system. Broader impacts include the professional development of a doctoral student, the publicly accessible data, and the presence of an American linguist in Cameroon to help maintain and improve the academic relationship between the two nations. Additionally, the research project and its products can bring linguistic awareness and attention to a language, region, and community of speakers currently under-represented in scientific literature. Records of tense/aspect systems of endangered language vary greatly in quality from mere labels of each tense and aspect (leaving it to the reader to assume what precise meanings should be associated with each label) to descriptions where nearly every detail related to tense and aspect is included. In order to achieve an adequately detailed description of the Iyasa tense/aspect system, researchers will make use of a wide variety of methodologies. These methodologies include (but are not limited to) administering tense/aspect questionnaires both already developed and newly developed by researchers on this project, analyzing of tense/aspect in natural narrative and conversational discourse, directed semantic elicitation (where a series of contexts that differ only slightly in meaningful ways are provided by the researcher), and directed narratives that are elicited by means of video clips or storyboards. Another important aspect of this project is that it will incorporate insights from a wide range of theories including cognitive theories of tense and aspect that move away from the notion of time as a single linear timeline. This project, in addition to providing important documentary materials and linguistic training opportunities to the Iyasa community, will provide important insights on how to improve computational models of tense and aspect for natural language processing and help strengthen our general understanding of how humans can and do encode time in language.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.