John Cage used to demonstrate time brackets to skeptical audiences by conducting one unit of time with his outstretched arms, first one, then another, imitating the motion of a clockhand — the same conducting techique used in his Concert for Piano and Orchestra and Atlas Eclipticalis — during which time period a skeptical audience member was to make a sound, any single sound of their own choice.
This apparently very simple exercise contained all the ingredients necessary to create a musical event which was itself anything but simple; in fact, all of the elements of musical time were in play for the composer, the skeptical performer from the audience, and the rest of the audience. First, the duration of the time unit described by the clock-like motion and the speed of the clock -- as it needn't go 'round the clock face at a constant rate -- were variables. The beginning and completion of the cycle were markers within a larger context. Is the time "outside" of the cycle zero time? Does the completion of a cycle suggest imply its repetition? And the performance of the skeptic noisemaker introduced articulation that was inevitably musical: Did it coincide precisely with either the beginning or end of the cycle? Did it divide the cycle into some recognizable fraction? Did it occur slightly before or after either the termina or some rational division of the cycle? What about the duration of the sound? Did that duration relate — was it proportional — to the duration of the cycle as a whole? Did the sound contain internal subdivisions or subarticulations, further marking the cycle? Did the interaction of cycle and sound articulate or suggest a tempo or metre, perhaps -- when the sound even itself has subarticulations -- a polyphony of tempi or metre? Or even movement or rubato within that tempo or metre? And so on.
The most striking thing to me about this exercise — which I witnessed Cage carry out three or four times with rather different audiences in rather different settings — was how musically robust it was, or rather, how temporally robust music might be. Even if the skeptic wanted to go beyond skepticism and sabotage the exercise in some way (which usually takes the form of a very loud sound coinciding precisely with either the beginning or end of the cycle), it was nearly impossible to ruin the exercise as there was, as far as I could or can imagine, no relationship between the cycle and the sound which was not — or at least potentially not — musically useful.
Other composer and musicians look at this condition, this peculiar condition of musical time, as if peering into the abyss and ask How could music be any more arbitrary that this? And they draw the conclusion that music, if so arbitrary in possibilities, must be easy. I believe that this is a false conclusion, and that music is in fact extremely difficult precisely because of this huge vector space of possible states, for as potentially useful and rich in implications as each might be, this was an exercise, not yet composed and not yet music. In order to go beyond the arbitrary temporal world of the exercise, the event music be placed in a larger context, and that one supposes, is composition.