Allow the user to correct text-to-speech mispronunciations and/or retrieve pronunciation corrections from a central source