A system for a multimodal unification of articulation includes a voice signal modality to receive a voice signal, and a control signal modality which receives an input from a user and generates a control signal from the input which is selected from predetermined inputs directly corresponding to the phonetic information. The interactive voice based phonetic input system also includes a multimodal integration system to receive and integrates the voice signal and the control signal. The multimodal integration system delimits a context of a spoken utterance of the voice signal by using the control signal to preprocess and discretize into phonetic frames. A voice recognizer analyzing the voice signal integrated with the control signal to output a voice recognition result. This new paradigm helps overcome constraints found in interfacing mobile devices. Context information facilitates the handling of the commands in the application environment.