Many activities require the ability to estimate intervals of time in an accurate and flexible manner. A traditional and popular account suggests that humans possess a kind of internal stopwatch that can be started, paused and stopped at will. Here we test this idea by measuring variable performance errors in three experiments. Participants had to compare the total time accumulated during one to three short target intervals with a single standard interval. With two or more target intervals, participants had to pause, but not reset, their putative internal stopwatches. By establishing baseline performance at two different standard durations and extrapolating based on Weber’s law, we were able to estimate how much performance should have deteriorated when target segments contained breaks. The decrement in performance we observed far exceeded the stopwatch prediction, and also exceeded the simulated predictions of a modified stopwatch with a slowing pacemaker. The data thus favour either a counter that cannot be paused during sub-second durations or alternative models of sub-second interval duration discrimination which do not posit a count-based metric for time. We discuss several possible strategies which participants might have implemented in order to apply such clocks in the split-interval task.
展开▼
