JOHN HARRIS
Abstract
By attributing stand-alone phonetic interpretability to each melodic prime, we rid phonological output of segmental redundancy without compromising the ability of representations to be mapped onto articulation and auditory perception. Under this approach, the repertoire of constraints on segmental output is automatically restricted to those that refer to lexically distinctive primes. One specific consequence is that there is no need for major-class features. An analysis of glide hardening in Cypriot Greek demonstrates how [consonantal] can be jettisoned without jeopardizing the phonetic interpretability of phonological representations.
[PDF file]