The speech-recognition sharing object supports the IEnumSRShare interface, which enumerates, selects, or creates a new shared engine/audio-source pair.