New Results - Explicit Modeling of Speech Production and Perception