Segment-to-syllable alignment and vocalization in Chilean Spanish

This paper analyzes the vocalization of stop consonants that occurs in Chilean Spanish as a strategy to avoid the emergence of consonantal segments that may not be aligned with the left edge of a syllable. Vocalization is the best alternative to grammars that are pressed to parse every consonantal segment in syllable initial position but do not want to do so at the expense of losing or gaining segments. By providing an unfaithful correspondent that is non-consonantal, output forms are able to avoid the emergence of misaligned consonants and still hold the position of the offending segment. Consonant-to-syllable alignment constraints also force the simplification of complex onsets when the first member of the cluster is a voiced stop whose output correspondent may not surface as such due to an independent process of spirantization. But even in this case, vocalization also serves to avoid the parsing of a consonant in a position that is not leftmost within the syllable since the second member of the cluster becomes syllable initial. Despite ridding the output form of misaligned consonants, vocalization cannot completely erase but simply shift markedness from the coda to the nucleus because it requires the formation of a diphthong.