Timing

Investigations of Timing

Finger-tapping tasks are a primary means of investigating timing in the laboratory. Precise measurements of a series of self-paced pulses can sometimes reveal subtle deviations from strict timing, even in highly skilled musicians. Some observations first made over 125 years ago include the tendency to alternate between pulses of longer and shorter duration, and slower drifts in tempo. Varying the conditions under which individuals perform the task can also systematically affect timing. Researchers make inferences about the nature of a timing mechanism that could give rise to their observations and, in turn, make new predictions from these functional models. Increasingly, researchers use neuroimaging techniques to attempt to associate functional models with specific structures in the brain.

The two-level model proposes an internal timekeeper that measures out the duration of each pulse and triggers the onset of tap execution. Under the assumption that timekeeper and execution processes independently operate, longer-shorter pulse alternation is thought to arise from a random delay in tap execution, which should both lengthen the pulse that it completes and shorten the one that it initiates. At the heart of the model is a mathematical procedure applied to timing deviations in a series of self-paced pulses that isolates timekeeper error (e.g., random mistimings of pulses) from execution error (e.g., random delays in the onsets of tap movements). Some striking dissociations support the two-level model.

First, timekeeper and execution error have are selectively increased in special populations, such as stroke patients. Second, whereas timekeeper error increases at slower tempi, execution error remains constant across tempi. These findings support the model's prediction that execution errors are the result of a process peripheral to timing. Moreover, to the extent that time perception is also less accurate at longer intervals, these findings imply that the perception and production of timed intervals share a common timekeeping process. Music-relevant extensions to the two-level model have been proposed. Some account for synchronization to an external pulse, proposing an additional error correction process that adjusts for timing gaps between produced tap and perceived pulse. Others account for the production of musical rhythms; one proposal is that hierarchical rhythmic structure is reproduced by a hierarchy of timekeepers capable of simultaneously measuring out different durations.

...