Mathematical modeling of grammatical diversity supports the historical reality of formal syntax

Recent studies have taken advantage of computational techniques to investigate the evolution of Indo-European languages [1-3]. However, these methods are not able to overcome the time constraints on lexical evolution, which limit a broader application of the Classical Comparative Method, and therefore cannot be used above the family level. For this reason, evidence from cross-family relationships must come from other domains (e.g. phonetics, [4, 5]). Reference [6] shows that another domain, syntax, is a potential source for cross-family comparison. In this paper, we evaluate the method proposed in [6], the PCM, and argue through a random generation of possible grammars that syntactic distances can be useful to detect signals of historical relatedness above the Indo-European level, within some confidence probabilistic intervals.