This paper deals with the phenomenon of Ecuadorian /s/-voicing in intervocalic environments, drawing its data from Quito radio speech. While this sibilant voicing in Quito is often considered an exceptionless case of complementary distribution, with [z] occurring intervocalically before a word boundary and [s] appearing in all other environments, the author finds that intervocalic /s/ voicing does not categorically occur at word boundaries in the Quito dialect, and the voiced variant appears in word-medial position as well. A Varbrul analysis shows that, while not categorical, word boundary is the most influential factor in /s/-voicing in Quito, but word-medial voicings seem to be constrained by the entirely different factor groups of following vowel and word frequency. Based on these findings, this paper highlights the inadequacy of any complementary distribution analysis espousing categorical application of the /s/-voicing rule in Quiteño Spanish.
Selected Proceedings of the 5th Workshop on Spanish Sociolinguistics
edited by Jim Michnowicz and Robin Dodsworth
Table of contents