Networks in aid of historical linguistics and linguistic typology: Uncovering hidden patterns in language evolution and language diversity

titleNetworks in aid of historical linguistics and linguistic typology: Uncovering hidden patterns in language evolution and language diversity
start_date2023/06/13
schedule10h30-12h
onlineno
location_infosalle 512 bis & en ligne
summaryIn recent years, an increasing number of studies have utilized networks to represent typological universals of linguistic co-expression, and reveal associations between the meanings that are co-expressed (e.g., Jackson et al., 2019; Youn et al., 2016). These studies have shown that networks are not only effective for visualizing data, but they are also a valuable tool for uncovering details that would otherwise remain hidden (see also List et al., 2013; List 游函, 2016). This talk aligns with this trend and aims to demonstrate how graph-theoretic techniques and networks can enhance our understanding of fundamental questions in Historical Linguistics and Linguistic Typology. Specifically, this talk will highlight how semantic maps and colexification networks, two popular graph structure representations employed in typology (Croft, 2022; Georgako­poulos, 2019), can be used to uncover pathways of meaning evolution, and capture meaning­ful generalizations about the organization of the lexicon and its reshaping over time. In addition, it will present a protocol for integrating the diachronic dimension into lexical semantic maps (François, 2022; Georgakopoulos & Polis, 2021). In doing so, we can delve deeper into the potential of colexification networks, and reveal interrelations between 1/ association patterns found across multiple languages and 2/ the frequency of occurrence of specific words within individual languages. This parallel investigation will shed new light on the evolution of polysemy patterns over time, and showcase the potential of using cross-linguistic data to investigate language-specific questions.
responsiblesSarda