Dutch Reduction Domains: Between Syllables and Feet

Abstract

This paper proposes that Dutch has a separate prosodic domain that encodes the difference between reduced and unreduced vowels: the Reduction Domain (see also van der Hulst & Moortgat 1980 for a similar proposal). This proposal is supported by evidence from morphotactics and vowel reduction. I argue that Reduction Domains allow for a superior account of the morphotactics of underived verb and adjective stems in Dutch (Trommelen 1989). I also show how the Reduction Domain yields an alternative account of restrictions on (optional) vowel reduction in Dutch (Kager 1989; van Oostendorp 1995). While this approach performs equally well on the basic facts of Dutch vowel reduction as Martínez-Paricio (2013), the Reduction Domain approach makes testable predictions both for Dutch and for phonological typology, distinct from those made by Martínez-Paricio (2013). Thus, the Reduction Domain forms an interesting case for potential addition to the Prosodic Hierarchy, whether as a language-universal representational unit, or as a representational unit specific to certain languages.