Deriving Consonant Cluster Phonotactics: Evidence from Singapore English

Abstract

Consonant clusters are often targeted by phonological processes that typically make them simpler. This brings up two questions of theoretical interest. First, why do cluster processes occur? One possible answer is that they occur to improve syllable structure (e.g. Steriade 1982, Borowsky 1986, Ito 1988, Blevins 1995); another is that they improve the perceptibility of stop place cues (Labov 1997, Hume 1998, Cote 2000, Steriade 2001). Second, how does a language choose among possible cluster processes? In Optimality Theory (Prince and Smolensky 1993), the choice is uniquely determined by the language specific constraint ranking. A potential problem arises in languages with multiple cluster processes that sometimes interact, sometimes do not, and sometimes occur in free variation. We will consider these issues in the light of evidence from Singapore English (Tay 1982, Mohanan 1992, Bao 1998, Poedjosoedarmo 2000). This variety of English exhibits several consonant cluster processes that fall into three types: epenthesis, reordering (Metathesis, Copy) and lenition (Fricativization, Place Assimilation, Deletion), illustrated in (1).