Melodic heads, saliency, and strength in voicing and nasality

Abstract

Melodic Heads in Element Theory (Kaye et al. 1985; Harris & Lindsey 1995; Backley 2011) have long been associated with higher acoustic saliency of the headed prime’s properties (Lindsey & Harris 1990; Backley 1995; Harris & Lindsey 1995; et alia) and with the relative strength (e.g. alignment of melodic heads with strong positions and robustness of headed expressions against lenition) of a melodic head compared to a dependent (e.g. Backley & Nasukawa 2009). Following substantial work on the interaction of voicing and nasality (Nasukawa 1997, 2005; Ploch 1999; Botma 2004) it is commonly assumed that voicing and nasality are both represented by the same prime |L|, with dependent |L| encoding nasality and headed |L| encoding voicing. In this paper I counter some of the arguments for the universality of this implementation, and develop an alternative view of a unified voicing–nasality prime, in which voicing is encoded by dependent |L| and nasality by headed |L|. I show how this analysis is more consistent with both the saliency and strength arguments by considering arguments based on the represented acoustic patterns, positional strength, nasal sharing (nasal harmony within onset–nucleus pairs), and cross-linguistic biases against loss of nasality. Finally, I show how this account is compatible with a more restrictive, recursive view of the phonological interpretation component following the set theoretic model of Element Theory in Breit (2013). Based on these arguments I conclude that we have good reason to doubt the universality of Nasukawa (1997, 2005) and Ploch’s (1999) implementation; instead we must give serious consideration to the reverse option with headed |L| for nasality and dependent |L| for voicing. I suggest that there are two possible responses to this situation: we can either make the attempt to radically adopt the alternative, or we can adopt a more relativistic position (in the sense of Cyran 2011, 2014) which allows a choice between both options. This article is part of the Special Collection: Headedness in Phonology