Ruaridh Purse will be defending his dissertation proposal, tentatively entitled: Categoricity and Gradience in Coronal Stop Deletion, on Thursday the 9th of January at 3pm in the Linguistics department Seminar Room. All are welcome to attend.

Please find an electronic copy of the proposal document here, and the corresponding abstract below.


Title: Categoricity and Gradience in Coronal Stop Deletion

Supervisor: Meredith Tamminga

Proposal Committee: Mark Liberman (chair), Jianjing Kuang, David Embick

Time: 3:00pm - 4:30pm

Place: Seminar Room, 3401C Walnut St. Suite 300


Coronal Stop Deletion (CSD) in English has long enjoyed pride of place as the test case for approaches to variable phonology. However, CSD's realisation as a categorical alternation has always been assumed, and not demonstrated. This dissertation seeks to provide a unied analysis of CSD, exploring the details of its production and perception for clues about its representation.

In this proposal, I present pilot EMA data from five speakers, in which it seems that complete deletion with no tongue tip raising is rare but does exist. Moreover, some individuals exhibit articulatory profiles that could be consistent with a structure non preserving variable allophony analysis. Simultaneously, effects of morphology, task, and speech rate on a gradient measure of tongue tip raising suggest current phonological analyses of CSD still do not capture the whole story.

The core of the dissertation will involve gathering a large-scale Electromagnetic Articulography (EMA) corpus of semi-spontaneous speech. This data will be crucial in addressing questions of categoricity and gradience in the implementation of CSD, as well as individual differences along these parameters. The articulatory core of the dissertation will additionally be supplemented with a perception study and a laboratory acoustic production study. The perception study should conrm whether listeners can tell the difference between coronal stops that would traditionally be classified as deleted but differ in terms of tongue tip raising. The acoustic production study will investigate an alternative representation for CSD that allows for some level of gradience in tongue movements without challenging the modularity of the grammar; namely that words may differ in their prosodic structure, giving rise to different rates of lingual undershoot.