7th Celtic Linguistics Conference, University of Rennes 2, Rennes, France
The file is open to annotations via Hypothes.is. Comments are always welcome!
I offer an analysis of gliding in Bothoa Breton, as presented in the thesis. I argue that high vowels (especially [i]) either become glides or coalesce with a preceding consonant, in order to avoid hiatus. However, this pattern is disrupted if the high vowel becomes a nucleus at an earlier stage in the cyclic derivation. This provides an argument for stratal models of phonology-morphology interactions, and a strong argument against Output-Output Correspondence, because some of the exceptional non-gliding cases cannot be explained by reference to a free-standing form with transparent non-gliding.