Coronal underspecification and Optimality Theory

In this paper I address the issue of the role of underspecification of input representations in Optimality Theory (Prince and Smolensky (1993), McCarthy and Prince (1993a). The arguments against the need for underspecification in OT are presented and shown to be unconvincing in the face of typical patterns of place assimilation in consonants. The only full-specification OT accounts put forth appeal to a stipulated preferential treatment of marked elements under Faithfulness. I present an alternate analysis of place assimilation assuming underspecified representations and demonstrate how this analysis allows Faithfulness and constraints on markedness to apply without stipulations regarding constraint characterization and ranking.