What Can Mathematical, Computational, and Robotic Models Tell Us about the Origins of Syntax?

This working group staked out the landscape of mathematical and computational modeling approaches to language evolution. The authors attempt both to provide a survey of existing research in this field and to suggest promising novel directions for investigations. The focus was set on cultural evolution; biological evolution is covered in other chapters of this volume. The field is young, manifold, and highly productive, with contributions not only from theoretical linguists but also from theoretical physics, robotics, computational neuroscience, and machine learning. This chapter provides a systematic presentation and comparison of different modeling paradigms, which range from a rigorous mathematical analysis of macroscopic language transmission dynamics to computational (even robotic) studies of complex systems populated by learning agents. Original contributions of modeling research to the understanding of cultural language evolution are highlighted, shedding light on the magnification of learning bias through cultural transmission, restricting the space of possible grammars, coevolution of categories and names, and the emergence of linguistic ontologies.