Acoustic characteristics of prolongations in spontaneous Italian speech

Abstract

This study concerns the analysis of segmental prolongations from an acoustic and functional perspective. Since the lengthening of sounds is a common feature of speech, recognizing disfluent-related occurrences is a complex operation and the main method commonly used is a “pragmatic approach” based on subjective perceptive judgment. However, there is a lack of investigations concerning acoustic characteristics that, in addition to duration, contribute to the perception of markedness and distinguish the portion of the segment identifiable as a prolongation from that identifiable in context as an unmarked part of the lexical element. From the results, it emerges that prolongation instances exhibit longer duration, reduced intensity, lower jitter and shimmer values, and higher Harmonics-to-Noise Ratio (HNR) measurements. Furthermore, their functional differences have been observed to correlate with different acoustic realizations.