Expecting a break in time estimation shortens perceived duration. In time production, the effect is directly proportional to the duration for which the break is expected. Moreover, perceived duration is shortened further by increasing certainty about the break occurrence: higher levels of expectancy would increase the time allocated to monitoring the source from which the break signal will come, leading to greater loss in accumulation of temporal information. Taken together, the data presented here confirm the generality of the phenomenon. They also support its interpretation in terms of expectancy because it appears to be modulated by certainty, as are other expectancy-related effects in psychology.

The empirical work reviewed in this chapter is interpreted within an internal clock framework, where accumulation of temporal information is under attentional control (Gibbon et al., 1984; Meck, 1984; Zakay and Block, 1996). Expecting an interruption in accumulation perturbs this process, possibly because of attentional shifts from accumulation to monitor for the break signal. Linear functions in our time production experiments with breaks suggest that attentional shifts are regularly distributed during the expectancy period, as illustrated in Figure 9.8. The production data reported in the current chapter suggest that the total duration of shifts seems longer when certainty about the break is higher, which would result from more frequent or longer shifts from accumulation.

In time discrimination, the effect seems functionally similar, although a detailed analysis of performance suggested that the response is determined by a decisional criterion in that paradigm.

The role of attention in time estimation is a fundamental issue that may be analyzed with a variety of experimental strategies. Expectancy effects in temporally structured sequences of events have been thoroughly analyzed (e.g., Jones and Boltz, 1989; Jones et al., 1993). Dual tasks, where participants execute some nontemporal task and estimate its duration simultaneously, have been widely used to study attention in human timing research (e.g., Brown, 1985, 1997; Hicks et al., 1976; Macar, 1996; McClain, 1983; Zakay et al., 1983). Interpolating nontemporal tasks in time production (e.g., Burle and Casini, 2001; Fortin and Rousseau, 1987; Rousseau et al., 1984; for a review of studies using this paradigm, see also Fortin, 1999) or in time discrimination (e.g., Casini and Macar, 1997) has also been used in the same purpose. However, even though disruption or interruption in accumulation of temporal information was often inferred to interpret results from dual-task studies, none of them addressed the problem of interruption in timing itself. In contrast, effects of breaks in stimuli to be timed were systematically investigated using the peak-interval procedure in animal timing research (e.g., Meck et al., 1984; Roberts, 1981; Roberts and Church, 1978), which provided reference data and a useful theoretical framework to interpret interruption-related effects. In this chapter we reviewed recent experiments with human participants inspired from these break or gap experiments. The results show strong attentional effects related to expecting a break in human timing, in the absence of any concurrent nontemporal task during the interruption. We showed that these effects are reliable and may be generalized to diverse timing conditions. An internal clock framework — where brief closures of a flickering switch under attentional control during pulse accumulation are induced by break expectation — may account for these data. One major point made by these experiments is that attentional effects related to expecting an interruption in timing must be distinguished and considered independently from any additional cost of concurrency between temporal and nontemporal processes usually assumed to interrupt or disrupt timing.

Although results with human participants are often similar to those in animal experiments, they differ in some respects. For example, effects of break location, consistently observed with humans, as reviewed in this chapter, seem weaker and are not always reliable with rats (see, for example, Roberts, 1981, experiment 2; Meck et al., 1984). In gap experiments, rats and pigeons seem to use different processing rules when stimulus conditions are changed (Buhusi, this volume; Buhusi and Meck, 2000, 2002; Buhusi et al., 2002), which is not the case with humans, as shown in the two experiments reported here. Some discrepancies in results may be explained by differences in procedures, for example, by training with no-break trials, which is much more intensive in animal experiments than in human ones. Data from human and animal experiments with breaks have still to be compared systematically, however. In the future, such comparison should provide invaluable information on the role of attention in timing and, more generally, on rules common to animal and human timing.

0 0

Post a comment