On the expressiveness, validity and reproducibility of models of language evolution Comment on 'Modelling language evolution: Examples and predictions' by Tao Gong, Shuai Lan, and Menghan Zhang

摘要

Simulation models of language evolution (SMoLE) as reviewed by Gong et al. [1] span a very promising field of research. They contribute substantially to a paradigmatic shift from synchronic to diachronic linguistics. Computer simulations also give access to systems whose dynamics is otherwise non-accessible or for which empirical data is out of reach. Simulation modelling may further help to bridge the gap between yet unrelated, though equally promising fields of research such as experimental semiotics [2], robotic experiments [3] and theories of grounding [4]. All of this is excellently reviewed and exemplified by Gong et al. [1]. In spite of this and related success stories, simulation models raise questions about their expressiveness, validity and reproducibility whose answers may foster future research in this area.