Windows Speech Recognition profile-specific dictionary?

[Exergist]

Does anyone know if it is possible to have each WSR profile reference a different speech dictionary? Or is there only one speech dictionary and are separate profiles only useful for teaching WSR to recognize the voices of different people?

One example for why this would be useful is to better control how WSR references certain learned words depending on the application in use. For instance - generally speaking I wouldn't want Skyrim's dragon shouts to be available when I'm dictating to Word.

[SD48]

Also interested in this.