Social Structure Limits Language Complexity

Social networks enables linguistic behaviours


Jun 4, 2018 | News

Linguists came together for the third talk in the Department of Linguistics and Multilingual Studies’ seminar series to hear about Assistant Professor Luca Onnis and Ph.D. candidate Matthew Lou-Magnuson’s fascinating research into the relationship between the social structure of a community and its long-term impact on grammatical change over generations of speakers. They contrasted intimate, village-like communities with hierarchical, city-like ones, and demonstrated how physical properties affect the development of agents within them.

The methodological focus of the talk was on how agent-based, computational models, while being abstract descriptions of a phenomenon, allow researchers to examine language development at a scale which otherwise cannot be evidenced. The speakers also argued that this approach ties together the traditional inductive and deductive modes of scientific inquiry, and is an example of the so-called third way to do science, namely:  generative scientific research.

Asst Prof. Onnis and Lou-Magnuson explained that they chose to implement a network linguistics approach to tackle this problem. In doing so, they revealed how complex-network methods, computational modelling and agent-based simulation can be combined to form a powerful research tool; this tool can be readily used to investigate the interconnected changes across the lifespans of individuals, communities, and generations.

With computer models and complex-network methods, linguists are able to simulate a number of factors, such as the size of language communities, different degrees and types of language contact, social (in)stability, the density of social networks, the sharing of information, vocabulary size, and the ratio of children to adult learners of a language, just to name a few. By exploring in a controlled manner the processes that circumscribe linguistic behaviour, scientists are able to gain fresh insight into how these variables interact to weave the tapestry of language.