GLOW 31 Colloquium, Newcastle University, Newcastle upon Tyne, UK
The file is open to annotations via Hypothes.is. Comments are always welcome!
In this talk I argue that purely phonological (e.g. autosegmental) approaches to initial consonant mutations in languages such as Fula and Irish are insufficient to account for the facts fully. In many ways this does not represent my current thinking on these matters, as can be established by looking, for instance, at this paper.