Abstract
Duration discrimination thresholds in the seconds range are approximately proportional to the standard duration (Weber's Law). So are those for spatial length discrimination. Why? For space, a multi-resolution ‘rubber ruler’ model has been proposed. We suggest a similar model for time, consisting of a clock that has an adjustable tick rate, but a finite size accumulator. Because of the latter, the tick rate, and thus the resolution, must be approximately proportional to the duration being timed. We predict that observers will be impaired if they do not know at the start of each trial whether they are going to have to time a short (∼1 sec) or a long (∼ 4 sec) interval; and we shall report that this is the case.