The role of prosody and hand gestures in the perception of boundaries in speech

This paper investigates the use of prosodic, gestural, and syntactic information in the perception of boundaries in extracts of spontaneous speech in British English. Experiment 1 aimed at investigating the effect of prosody on naive participants' perception of boundary strength. 13 naive listeners had to rate boundary strength for 64 extracts on a 5-point scale. The stimuli all contained three tone-units, the second being a syntactic subordinate construction, which was established as a variable. The prosodic cues at the boundary between the tone-units were also established as variables, and were subject to manipulation (addition of a single cue associated with the perception of a prosodic boundary). Experiment 2 aimed at assessing the effect of gesture on naive participants' perception of boundary strength. In Experiment 2, 24 naive listeners had to measure boundary strength for 24 extracts on a 5-point scale. The stimuli all contained three tone-units, the second being a syntactic subordinate construction, which was established as a variable. The hand gestures produced in co-occurrence with the tone-units were established as variables, and were subject to manipulation. Results show that prosody modulates perceived boundary strength, but not gesture, based on the variables we included. Silent pauses have the strongest effect on perceived boundary strength, but final syllabic lengthening and pitch reset also have separate effects as single predictors. Our data also shows a trend concerning the production of two identical hand gestures in terms of configuration and trajectory.