Typological tendencies in verse and their cognitive grounding
This dissertation is about verse, some of its recurrent features, and cognitive aspects which can explain their prevalence. Verse includes a range of verbal phenomena, most typically songs and poems, but also nursery rhymes, religious chants or demonstration slogans. Compared to everyday speech, all these forms show additional layers of structure, like a regular alternation of accented syllables, a fixed melody, or a systematic number of syllables per utterance. Every linguistic community in the world engages in verse, but certain features seem suspiciously widespread.
On the one hand, I have developed computational tools in order to assess systematically how widespread individual verse features are. On the other hand, I have conducted behavioural experiments to investigate to which extent these widespread features may stem from properties of human cognition. Using these two approaches, the thesis examines three aspects of verse. The first part deals with constituent structure in verse and how it can emerge in the process of iterative learning. The second part measures final strictness in several languages, and proposes that it is a consequence of reduced attention at the beginning of lines. The last part develops a method to describe how linguistic and musical features are aligned in songs, and how to test the intuitions of native speakers experimentally. Although verse constitutes a prototypically creative activity subject to extensive cultural variability, it is nonetheless bound and shaped by our cognitive system.