Doctoral Dissertation Research: Inferring the linguistic past of an indigenous language family of North America

语言的进化方式与生物物种相似。最近的语言学研究使用了生物科学的计算系统发育方法来解决有关语言学过去的问题：法语与意大利语或西班牙语的关系更密切？诸如德语和英语等日尔曼语系是什么时候开始从它们共同的祖先中分化出来的？前者是推断语言谱系树内遗传关系的一个例子，后者是估计遗传相关群体分化日期的一个例子。关于这些方法在语言学中的应用，以及对它们产生的结果有多大的信心，还有很多东西要学。为此，本研究采用计算系统发育方法来推断北美土著语言家族的遗传关系和分歧日期，并检查结果与以前的研究有何不同。这项工作也试图了解为什么会出现这些差异，以及这些结果对语言家族的语言历史有什么新的启示。语言系统发育研究中用于解决类似研究问题的数据在相当大的程度上有所不同。因此，通常不清楚使用哪种数据类型或数据应如何编码，以及这些决策如何影响结果。此外，不同类型的语言数据可能表现出不同的进化动力学，并且使用相同的进化模型可能不适合所有类型的数据。因此，本研究探讨如何不同的选择（a）数据类型，如词汇和语法数据，（B）编码方案，（c）进化模型的影响结果。为了确保数据编码的准确性和一致性，这项研究涉及语言学，档案和实地研究。由于有些语言已不再使用，并包含“缺失”的数据，本研究还采用了相同的方法来推断语言的过去，以预测和填补数据中的缺失值，从而有助于语言振兴工作。该奖项反映了NSF的法定使命，并被认为值得通过使用基金会的智力价值和更广泛的影响审查标准进行评估来支持。

Languages evolve in similar ways to biological species. Recent linguistic work has used computational phylogenetic methods from the biological sciences to address questions about the linguistic past: Is French more closely related to Italian or to Spanish? When did the Germanic languages, such as German and English, begin to diverge from their common ancestor? The former is an example of inferring genetic relationships within a language family tree and the latter is an example of estimating the dates when genetically related groups diverged. There is still much to learn about the use of these methods in linguistics and how much confidence can be placed on the results they produce. To this end, this study employs computational phylogenetic methods to infer the genetic relationships and divergence dates of an Indigenous language family in North America and examines how the results differ from those of previous studies. This work also attempts to understand why these differences occur and what new light the results shed on the linguistic history of the language family.Data used in linguistic phylogenetic studies to address similar research questions have varied to a considerable degree. Therefore, it is often unclear which data type to use or how the data should be coded, and how these decisions affect the results. Additionally, different types of linguistic data may exhibit different evolutionary dynamics and using the same evolutionary model may not be appropriate for all kinds of data. Therefore, this study examines how different choices of (a) data type, such as lexical and grammatical data, (b) coding scheme, and (c) evolutionary model influence the results. To ensure that the data is coded accurately and consistently, this study involves philological, archival, and field research. As some languages are no longer spoken and contain ‘missing’ data, this study also employs the same methods for inferring the linguistic past to predict and fill in missing values in the data which can then be helpful for language revitalization efforts.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.