Innovated innovation: New approaches to computational historical linguistics in the Asian context

Quantitative methods have been applied to historical linguistics since the early 20th century. However, it is only in recent times that multiple computational methods have flourished in the exploration of the histories of East Asian languages. This workshop showcases diverse computational approaches to understanding linguistic evolution. Zhang Menghan (Fudan University, China), Lai Yunfan (Trinity College Dublin) and Nathan Hill (Trinity College Dublin) will present from distinct angles: first, the inference of prehistoric trajectories within Sino-Tibetan and Kra-Dai populations; second, the application of information theory to explore language subgrouping via sound correspondences; and third, the automated phonological reconstruction of proto-languages employing finite state transducers. Supported by the FAHSS Events Fund. 2-2:40pm - Zhang Menghan (Fudan University), "Phylolinguistic insights into the prehistories of Sino-Tibetan and Kra-Dai languages" The study of language origin and divergence is important for understanding the history of human populations and their cultures. As the most linguistically diverse region on earth, East and Southeast Asia have witnessed extensive sociocultural and ethnic contacts among different language communities. Especially, there are several language families such as the Sino-Tibetan languages, which is the second largest one in the world, and the Kra-Dai languages that exhibit tremendous socio-cultural importance in East and Southeast Asia. How to reconstruct the evolutionary processes of these different language families remains a great challenge in traditional historical linguistics. The major reason is the lack of sufficient and continuous historical accounts of their prehistory. Fortunately, applying Bayesian phylogenetic frameworks and triangulating empirical evidence from linguistics, genetics, and archaeology can provide new insights into studying the complex prehistory of languages in the world. In this presentation, we are showing our recent advances in the reconstruction of the prehistory of Sino-Tibetan and Kra-Dai languages in East and Southeast Asia. 2:40-3:20pm - Lai Yunfan (Trinity College Dublin), "Mutual predictiveness of sound correspondences for reconstruction and language subgrouping" This paper introduces a novel method for evaluating proto-language reconstructions in historical linguistics, using Gyalrongic languages—a conservative subset of Sino-Tibetan—as a case study. Mutual Implicative Entropy (MIE) is applied to assess the genetic distance between related languages and to construct Neighbornet networks for visualizing the subgrouping of Gyalrongic languages. The resulting networks align with qualitative historical linguistic analyses, facilitating adjustments to previous subgroupings derived from Bayesian phylogenetic inference. This method offers a means of detecting nuances in lower sub-branches that might be overlooked by lexicon-based methods, making MIE a quick and efficient tool for validating reconstructions and refining the preliminary structure of language subgroupings in historical linguistics. 3:20-4pm - Nathan Hill (Trinity College Dublin), "Enhancing Proto-Language Reconstruction with CAPR: Advancements in Rule-Based Phonological Analysis" This paper presents CAPR (Computer-Assisted Phonological Reconstruction), a software tool that employs finite state transducers to assist in the reconstruction of proto-languages, with a focus on Proto-Burmish. CAPR automates aspects of phonological analysis, including the detection of loanwords, and supports the detailed examination of sound changes within daughter languages. By utilizing both forward and backward reconstruction methods, CAPR enhances the efficiency of hypothesis testing in historical phonology. This tool is particularly useful for linguistic communities with limited research resources. The paper outlines the technical framework of CAPR, its application in studying the Bola language, and its impact on streamlining phonological research in computational historical linguistics.