Why are there so many different languages across the world? How come these languages are so vastly different, while at the same time yielding recurrent features? How can we learn, use and process language?

We tackle these questions by simulating the emergence and evolution of language in agent-based models in which language evolves spontaneously as the side-effect of how communicative agents locally interact with each other, thereby displaying many processes that are also observed in nature but on a cultural level.

Language dynamics is a rapidly growing field that focuses on all processes related to the evolution, emergence, change, competition and extinction of languages. One of the major outcomes from this field of research is that language can be viewed as a complex adaptive dynamical system that evolves through the process of self-organization and self-regulation. According to this viewpoint, a community of language users can be seen as a dynamical system that collectively solves the problem of developing a shared communication framework through the back-and-forth signalling between the participating individuals.

Sony CSL – Paris has been pioneering the activities in this area, with a special focus on how a population develops a shared set of names (Naming Game), categories (Category Game), or rules (Rule Dynamics) from scratch. The basic framework is grounded in theories of Language Games, but frequently borrows concepts and methods from the areas of statistical physics and complex systems science. These areas have proven to be extremely powerful in providing various quantitative insights into language emergence and evolution.