Trying to work on the mammoth problem of rhythm control with a teenager it becomes clear that one cannot arrive at rhythmic control without considering tonal action. In a Beethoven sonata, for example, the many layers of tonal activity so directly affect awareness of the beat that the two are inextricably intertwined. Don't even try to treat one apart from the other: death will surely ensue--either the player's, from boredom, or Beethoven's from assault.

I liken the problem to the heart monitor at the hospital bedside, happily purring away at a steady rate until a bell rings. The thing goes wild. Are you going to believe that all that activity occurs within the same time unit as the quiet, predictable purring?