to appear). Agent-based modeling of the evolution of vowel harmony

A central tenet of the study of microparametric variation (MPV) is that closely related languages will reveal which observable parameters of languages are correlated; in particular, this correlation can lead to the discovery of more abstract parameters underlying the behavior of surface patterns. The study of MPV is a way of approximating diachronic experimentation on a language, where the linguist (if it were possible) would alter a parameter of a language and see what occurred. In this paper, we propose agent-based simulation as a complementary methodology for examining the feasibility of abstracting parameters through analysis of related languages. Agent-based simulations are a way to model the sort of complex system entailed by the evolution of grammars within a speech community. The focus of our research is vowel harmony in Turkic (Altaic) languages. These systems exhibit a great deal of change—instability, even—over the more than one millennium during which these languages have been recorded in writing. Most Turkic languages have two autonomous harmony systems, one based on tongue backness and one on lip rounding. The systems range from robust, nearly exceptionless harmony to highly variable or restricted harmony to no harmony at all. They thus provide rich material for constructing typologies of harmony and models of language change. With our model, we attempt to show how changes at the level of the individual aggregate to give language-wide evolution in backness harmony systems. Agent-based simulation, where a community of computational agents interacts over time, is appropriate here, as language changes takes place in a social context; moreover, the changes occur over multiple generations of individuals. 1 The authors wish to thank Mark Liberman, Donca Steriade, Kie Zuraw, Andy Wedel and two anonymous reviewers for their helpful insights. We also gratefully acknowledge the expertise of native speaker consultants of Tuvan, Uzbek, and Turkish