A Nearby Span of Time

A study of rhythmic applause in a theater places two recording positions on the same timeline. In the microphone near one audience member, adjacent claps were initially separated by roughly four or five tenths of a second. About ten seconds after the recording begins, this interval lengthens to around eight tenths of a second and stays for a while at the longer position. During a nearby span of time, the shared accents in the full-hall recording gradually become clear; the average noise intensity, calculated with a three-second window, falls. The observation in the first half thus moves from the nearby position toward the whole hall: first the interval between one audience member’s adjacent claps, then the accents that only appear after many claps have been superposed.

The shared accents were clear for a while in the full-hall recording.

After about twenty seconds, the shared accents in the full-hall recording gradually become unstable. The three-second average noise intensity begins to rise again. In the same stretch of recording, that audience member’s clapping intervals are also shortening overall, while still showing marked fluctuations along the way.

It is not hard to tell these changes as a neat round trip: clapping slows, shared accents appear; clapping speeds up, shared accents disperse. But a sentence like this closes sooner than the sound does. When the article returns from the whole hall to the nearby position, it does not find a person restored to the original speed, only a sequence of intervals that are shortening overall while adjacent values remain uneven. The local clapping and the hall’s accents turn during nearby periods, but they do not combine into a body that chooses fast or slow. Several changes lie close together, and the story acquires a unified actor; yet that audience member’s intervals in the latter half refuse to supply its ending. Overall, he clapped more densely. Between two adjacent claps, the interval still kept lengthening and shortening by turns.