Changing sensitivity to duration in human scalar timing: An experiment, a review, and some possible explanations

Evidence from a number of studies of human timing, using temporal generalization and bisection tasks, suggests more sensitive behavioural adjustment to presented durations under conditions in which the timing task demands discriminations between more closely spaced stimuli. An experiment using temporal generalization demonstrated this effect, as discrim-ination between a 600-msec standard duration and non-standard stimuli both shorter and longer than 600 msec was better when non-standard stimuli were more closely spaced around 600 msec. A review showed similar effects in other temporal generalization tasks and in a number of bisection studies, where time discrimination improved as the ratio of the long and short standards on the bisection task decreased. A standard model of human temporal gen-eralization explained the experimental data in terms of a decrease in the response threshold under more dif ® cult conditions, rather than changes in the representation of the standard duration. On the other hand, data from bisection could be modelled by assuming the con-trary; that representations of the short and long standards of the task were more precise under the more dif ® cult conditions. Explanations of some of these effects in terms of atten-tion to duration and/or arousal-induced changes in the speed of an internal clock were discussed. The theory of sca lar timing, originally proposed by Gibbon (1977) and developed later with other colleagues (Gibbon, Church, & Meck, 1984) must surely rank as one of the most successful recent imports from the animal to the human psychological laboratories.