How does phonology make reference to morphology? We will survey a range of answers to this question, focusing our discussion on how we might be able to provide our own new, and hopefully improved answers by extending our research on Harmonic Serialism (HS) and Harmonic Grammar (HG) into this domain.

HS suggests new ways of formalizing the interleaving of morphology and phonology (as in cyclic and lexical phonology). This has already been explored extensively by Matt Wolf in OT-CC, a variant of HS that McCarthy (2007) proposed to handle phonological opacity. We’ll thus start by reading Wolf’s dissertation, which has the added attraction of providing comprehensive and lucid discussion of earlier alternatives. We’ll then go on to explore other recent proposals about morphology-phonology interaction, including stratal OT, indexed constraints, and Optimal Paradigm theory (others?).

HG has the potential of contributing new ways of formalizing gradiently productive morphophonology, and accompanying learning algorithms. We’ll start this discussion by going back to the empirical and theoretical roots of all of this (including OT itself): Rumelhart and McClelland’s connectionist model of the English past tense (and various responses…)