In this course we will examine how representational explanations in phonology work, when we should consider using them in preference to others, and what kind of predictions representational theories can and cannot make. Focusing on phenomena in segmental phonology, we will consider how (whether?) representational devices such as underspecification, privativity, the contrastive hierarchy, and geometric structure can constrain our expectations regarding what a possible phonological process is, as well as how they relate to each other. Our main focus will be on what kind of data can motivate us to adopt a representational solution. We will consider both mainstream substance-based views of representation in the SPE tradition and various substance-free approaches to feature theory.
The course will be quite practical, with a focus on hands-on discussion of curated data sets. This course is most suitable for attendees who already have a grasp of how to construct a basic phonological analysis. As such, there is no set reading, but a familiarity with analysis techniques will be assumed that can be gained by at least a cursory examination of a textbook in generative phonological analysis, insofar as it covers underlying and surface representations, rules and rule ordering, and a basic introduction to phonological features.