Relation between geometry and kinematics of articulatory trajectory associated with emotional speech production

We investigate whether articulatory movement trajectories follow the nonlinear invariant relationship between the tangential velocity and the curvature, i.e. the one-third power law. The power law holds for articulatory trajectories of phonetic rendering, although the exponent is in the range 0.35–0.42 when averaged across speakers and emotions but is relatively invariant under speaker/emotion differences. However, the velocity gain factor, a proportional constant in the power law, is sensitive to such variations. While the power law reflects some common articulatory movement characteristics for phonetic rendering across speakers and emotions, the prosodic variation/speaker idiosyncrasy are reflected in the velocity gain factor.