14.04.2026 Remo Nitschke

In this study we present a methodology that allows us to infer whether a communication system is more likely generated by a regular grammar or a supra-regular grammar. We demonstrate this on artificially generated data and human generated data to validate the methodology and present mathematical reasoning for why our approach works. Finally, we show that some datasets sampled from various primates do not pattern like pure regular grammars or pure supra-regular grammars but show signals of both.