Phonological complexity

The structures and patterns found in human languages can be measured according to the computational expressivity required to describe them. A range of claims have been made about the expressivity required for human language sound patterns, and this research aims to contribute to our understanding of this range in two ways. The first is to accurately define the apparent upper bound of expressivity that appears to be required, based on an analysis of a particular set of cases. The second is to explain the uneven distribution of sound patterns across the range of expressivity in terms of the relative learnability of different computational classes.

(joint work with Adam G. McCollum, Anna Mai, and Eric Meinhardt,

and in collaboration with Leon Bergen, Nadia Polikarpova, Shraddha Barke, and Rose Kunkel)

Manuscripts

Presentations