SKM 2023 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
SOE: Fachverband Physik sozio-ökonomischer Systeme
SOE 8: Semantic Networks, Language and Culture
SOE 8.3: Talk
Tuesday, March 28, 2023, 14:30–14:45, ZEU 260
Dynamics of language features on complex networks: statistical physics analysis, and the introduction of linguistic temperature — •Christopher Kitching1, Jordan Abbott1, Tobias Galla1, 2, Henri Kauhanen3, Deepti Gopal4, and Ricardo Bermúdez-Otero1 — 1The University of Manchester, UK — 2Instituto de Fisica Interdisciplinar y Sistemas Complejos (CSIC-UIB) — 3University of Konstanz, Germany — 4University of Cambridge, UK
Statistical physics has made important contributions to the study of language dynamics. Here, we focus on a stochastic individual-based model of the spatio-temporal evolution of language features. Features evolve through a combination of descent across generations, and the influence of geographically neighbouring languages.
The dynamics are a variation of the two-state voter model, with spontaneous changes of state and transmission errors in the imitation process. Previous work is restricted to lattices, here we focus on complex networks. We establish a parabolic relation between the feature frequency and the density of active interfaces, modulated by the network, and `linguistic temperature' - a measure of a feature's stability.
Results are obtained by combining the voter model, pair approximation and a new approach to networks, based on random walks. While the theoretical work has applications in other areas (e.g. opinion dynamics), our main focus is in linguistics: the frequency and isogloss density of real-world language features can be obtained from WALS and from this we calculate their temperatures.