FIGURE 9 .1 Time interval production with break in Fo rtin and Massé's (2000) experiments.

b a c break, and to resume timing at the break offset until they judged that the total tone duration corresponded to the target interval. For example, in experiment 2 of Fortin and Massé (2000) the target duration was 2.5 sec. The prebreak duration (a in Figure 9.1) could be 800, 1300, or 1800 msec, and break duration could be 2, 3, or 4 sec. In a trial, the participant started the temporal production by pressing a 0 key on a keyboard, which triggered a tone lasting for 800, 1300, or 1800 msec. The tone was then interrupted for 2, 3, or 4 sec, and then resumed until the participant ended the temporal production by pressing 0. The participant was instructed to execute this second key press when the total tone duration was equal to the practiced target duration. So in principle, if prebreak duration was 800 msec, the participant had to wait 1700 msec after the break in tone presentation to produce a perfect 2.5-sec target interval.

Location (which corresponds to prebreak duration) of breaks and their duration were varied in these experiments, as illustrated in Figure 9.1. The procedure is similar to that in animal break or gap experiments (e.g., Meck et al., 1984; Roberts and Church, 1978; Roberts, 1981). In those experiments, normal rats seem to be able to time a stimulus before and after a gap in stimulus presentation, while interrupting timing during the gap. This defines a stop mode where the switch controlling pulse accumulation from the pacemaker to the accumulator is opened during the break, thus stopping the accumulation process for this period. In human experiments, a gap in stimulus presentation was used similarly to induce a break in time estimation. Produced intervals were defined as the sum of the prebreak and postbreak durations (a + c). In this task, a stop mode would result in produced intervals corresponding to the practiced target duration (e.g., 2.5 sec) and should not be affected by break location or duration.

The mean produced intervals in Fortin and Massé's (2000) experiment 2, averaged over the three break durations, are shown in Figure 9.2 as a function of break location. Perfect production in a stop mode should result in 2500-msec intervals and should not vary with break location. Actually, the main finding was that mean produced intervals lengthened proportionally with increasing value of location, that is, of prebreak duration. In fact, even though mean produced intervals were generally close to the 2500-msec target duration, they were clearly positively related to pre-break duration. This lengthening is called the break location effect.

0 0

Post a comment