There are text animations, one of them is Typing, but they are only functional for SWF export, not recommended. Synchronizing with audio would be tough.
You could use a shape with the color of the background on top of the text, and apply a motion path to that shape so that it clears the text gradually. Length of the motion path can easily be edited.
You could create a multistate object, having one word more in each state. Have it then appear in steps with a While loop, Delay command followed by Change to next state.
Maybe could offer more possibilities, but want to end with a critical note: do you really think this will add to the learning efficiency? When I need to use bulleted text – which I avoid whenever possible – I use the Multistate approach and attach audio to the bullet lines. But for words or characters in a sentence?