Abstract
In this thesis I propose that the distribution of /s/ in Blackfoot can be explained by positing that /s/ is inherently moraic in Blackfoot, and explore this hypothesis via two proposals about moraic /s/. The first is that /s/ in complex onsets, e.g., stsiki, ‘another’, is extrasyllabic, and that a moraic /s/ reduces the markedness of these extrasyllabic segments. The second is that because /s/ is moraic, it can act as a syllable nucleus, which explains why the distribution of geminate /ss/ is more similar to long vowels than to geminate consonants. In Blackfoot, clusters of more than two consonants occur only with /s/, and clusters of more than three consonants occur only with geminate /ss/. The Blackfoot syllable seems to be overwhelmingly simple, with /ss/ clusters being the only outliers. While all other geminates occur between vowels, geminate /ss/ often occurs before, after, or between other consonants. This thesis aims to make three specific contributions: (i) to describe the distribution of /s/ in Blackfoot, (ii) to propose that a non-vocoid may be inherently moraic, and (iii) to introduce the PROSODICSEQUENCING constraint, which explains the tendency for onsets to be non-moraic, and predicts that light CVC syllables will be less marked than heavy CVC syllables.
|
