Language redundancy effects on the prosodic word boundary strength in Standard German

Abstract

The Smooth Signal Redundancy hypothesis proposes a complementary relation between language redundancy and acoustic redundancy mediated via prosodic prominence and boundary structure. To test this hypothesis, the current study investigates the effects of lexical frequency on boundary-related segmental (+pause) duration patterns at prosodic word boundaries in Standard German. Results are consistent with predictions made by the Smooth Signal Redundancy hypothesis, showing an inverse correlation between lexical frequency and duration: Word boundary-related target intervals for frequent words were more than 10% shorter than corresponding intervals for infrequent words, and effects on non-boundary related intervals were not significantly different for frequent vs. infrequent words. These effects suggest a preference for producing stronger prosodic boundaries in case of low language redundancy. The effects appear to be stable across varying speech tempo by different speakers and targets, even when the factor of speech tempo is controlled for. This is consistent with the view that speech tempo, as a global factor that modulates the overall utterance, does not interfere with the localized acoustic redundancy.