Friday, January 30, 2009

Variable syllabic boundaries, 9 contd, unreleased stops and glottals

The rule that deletes interconsonantal //t θ d// under the appropriate conditions is part of a large and very complex rule that stipulates where stops are unreleased.

Lack of release triggers glottalization of the heavy (i.e., underlying voiceless) stops (//t// becomes /ʔ/ or /tˀ/, depending on the context).

Glottalization is thus complementary to release in such stops.

The vastly misunderstood rule of glottalization has been alleged to provide evidence for syllabization. That the relationship is not simple can be seen by examining the tempo-variable behavior of /p k c/ before //s//.

In slow tempos, one hears [cˀ] in axe and [pˀ] in cops;
In faster tempos, glottalization is replaced by release.
Note, however, that where interconsonantal //t// is replaced by length in
acts ['æcˀːs] and Copts ['kʰapˀːs] can sound like axe and cops (although in yet faster tempos acts and copts can sound like axe and cops).

A failure to understand this gradient variability is responsible for the lack of sucesses of the static analyses

No comments: