"As shown, the user interface 380 may include graphical representations 382-386 of one or more items, such as a key, a doll and a bell in this example, that the user may select in response to the spoken word's separated sound units." . . . .