Proceedings of the Annual Meeting of the Cognitive Science Society

Poetry and lyrics across cultures, from Sonnets to Rap, demon-strate an obvious human cognitive capacity for the perceptionand production of various multi-syllable sound patterns. Herewe use entropy to measure discrete serialized representationsof phones and to explore the complexity of these sound struc-tures across genres of creative language arts. The present ex-ploratory analysis has two main objectives. First, our aim isto broaden the scope of cognitive processes and data that areconsidered in statistical learning approaches to phonologicallearning and language acquisition. Second, we hope to to pro-vide a basis for more targeted computational and phonologicalinvestigations of these patterns. We compare the conditionalentropy of sequences of phonological patterns in lyrics and findthat, in general, Battle Rap and Sonnets maintain noticeablylower entropy than other genres across sequence sizes, whilelyrics from Electronic music and Hip-Hop display relativelyhigh entropy.