Speech Prosody 2004
Our production experiment in Scottish English revealed that the duration of a rhyme immediately followed by a Level II suffix such as -s (the 1st person singular/plural/possessive suffix) and -t (the past tense suffix) was significantly longer than that of a monomorphemic counterpart. Such a durational difference between suffixed forms and monomorphemic forms was absent when the Level II suffix was -er (the agentive suffix) or -ing (the progressive suffix). Those results may indicate that morphological boundaries (i.e. stem-suffix boundaries) are not directly influencing acoustic duration adjustment and support a prosody-phonetics interface hypothesis that the phonetic component is only accessible to prosodic structure (but not to morpho-syntactic structure). We also found that the rhyme duration of suffixed forms was shorter than the duration of a comparable rhyme in two-word sequence forms. This result, however, does not necessarily refute a hypothesis that there is a lexical/prosodic word boundary at the stem-suffix boundary because the presence of the word boundary at the stem-suffix boundary still allows "polysegmental" shortening to be applied to a higher word that dominates both the stem and the suffix.
Bibliographic reference. Sugahara, Mariko / Turk, Alice (2004): "Phonetic reflexes of morphological boundaries at a normal speech rate", In SP-2004, 353-356.