This work is a contribution to studies on anticipation in speech, in the visual perception domain. Anticipatory gestures produced by two French speakers are analysed using kinematic data for VunroundedCnVrounded sequences, in order to examine the timing of labial anticipation and its expansion. Our results show that anticipatory gestures can be initiated within configurations related to the unrounded vowel. Moreover, our data do not confirm the predictions of dominant models for anticipatory production, except for those of the Movement Expansion Model. The perceptual efficiency of these gestures in the visual modality was tested. Results confirm that visual perception of anticipatory gestures follow their expansion. In the visual modality, a few conditions have to be fulfilled in order to allow recognition of the upcoming rounded vowel: A high velocity; At least 60 ms from the beginning of the anticipatory movement; In order to enhance recognition, anticipatory movements should accelerate, then decelerate near the final target. In other words, when movements are performed with several minor velocity peaks, the rounded is recognised later. Thus, in dynamic conditions, perceptual efficiency of anticipatory gesture is independent of the duration preceding the onset of the rounded vowel, static configuration or absolute kinematic event.
Cite as: Roy, J.-P. (2005) Visual perception of anticipatory rounding gestures in French. Proc. Interspeech 2005, 2949-2952, doi: 10.21437/Interspeech.2005-778
@inproceedings{roy05_interspeech,
author={Johanna-Pascale Roy},
title={{Visual perception of anticipatory rounding gestures in French}},
year=2005,
booktitle={Proc. Interspeech 2005},
pages={2949--2952},
doi={10.21437/Interspeech.2005-778}
}