Abstract

In morphology, the restriction of phonological processes to a subword domain is often seen to have one of two structural sources: either the process applies only within an independently motivated morphological constituent, e.g. due to cyclicity; or the process applies only within a certain prosodic constituent. Meanwhile, in the recent syntax-phonology interface literature, theories of phonological domain formation such as Selkirk’s (2009) Match Theory and Cheng & Downing’s (2012) edge-based mapping theory incorporate competing morphosyntactic and phonological constraints that give rise to a greater variety of domain types. Based on data from Huave, a language isolate of southern Mexico, I propose that these theories should be extended to sub-word phenomena, and I present some arguments in favor of edge-based mapping.