Abstract: The question of the evolution of language continues to excite many scientific disciplines. Clearly there is a biological component to the answer because language requires extraordinary brain capacities. There is also a social component because language only makes sense in communities where members have to cooperate intensely to survive and raise children. Here I will focus on the third, equally important, cultural component: what are the interaction patterns and cognitive functions by which a community can self-organize a language system with similar properties as we find in human languages. I have been tackling this approach with my group using experiments with artificial agents (possibly embedded in humanoid robots) that play language games. This experimental approach allows us to pose very precise questions, such as how and why does phrase structure grammar arise or how and why can a case system emerge and collapse, and to test out theories to answer these questions. The talk will not only discuss concrete examples of this approach but also extract the general theory of the evolution of languages that has come out of this work. References : Luc Steels, Agent-based models for the emergence and evolution of grammar, Phil. Trans. R. Soc. B 371:20150447 http://dx.doi.org/10.1098/rstb.2015.0447 Luc Steels, Do languages evolve or merely change? Preprint submitted to Journal of Neurolinguistcs, 10 pp.
origin, evolution, language