Consonant durations in clusters

The durations of initial and final consonant clusters in monosyllabic and bisyllabic words within a frame sentence were studied from spectrograms of readings by 3 speakers. The durations of consonants within a cluster varied with the features of the consonant and its phonetic environment, such as voicing and point and manner of articulation. A durational model was proposed based on two mechanisms. An articulatory mechanism was attributed to effects involving coarticulation and restrictions in the motion of the tongue and lips during a cluster. Shortening of consonants in clusters seemed to arise from the shorter distances that the articulators travel in clusters. Lengthening of consonants before fricatives and voiced consonants and aspiration effects were noted. The other factor was a phonological mechanism, related to the use of duration as an acoustic cue in consonant perception. Single consonants varying only in the voicing characteristic had substantial durational differences which could aid in distinguishing them. However, phonological restrictions, such as common voicing among stops and fricatives, arise in the clusters, and the redundancies allow the durational differences to become less.