HomeResearchOverview Skill Acquisition Multi-tasking Time perception Learning from instructionsACT-R/LISA PublicationsPersonal

Time perception

Although most current cognitive architectures have modules for visual and auditory perception, they generally do not incorporate the perception of time. Existing models of time perception focus just on time perception, for example the following model:

Although specific models have been successful in explaining various aspects of time perception, they often fail when they involve tasks in which time perception is just a part of a larger cognitive context. In the diagram above, the "rest of cognition" is represented by "Memory" and "Comparison". Even worse, general aspects of cognition like Attention are modeled using specific elements in the time estimation model, for example the "Attention" and "Gate components in the diagram. It is therefore a much better idea to view time perception as a component of a larger cognitive architecture, and let the architecture handle the more cognition-general aspects like attention and learning.

Together with Hedderik van Rijn I have developed a time perception module that is capable of modeling many phenomena documented in the time perception literature (Taatgen, van Rijn & Anderson, 2007). Instead of a specific attention component, we use ACT-R's serial bottleneck theory to model attention, and instead of a specific learning component, we use ACT-R's instance learning to model gradual improvements in accuracy. Although we haven't yet tested it empirically, ACT-R's production compilation should be able to account for automatization of time estimation.

On the basis of new experiments, we have shown that our model provides better predictions of the impact of attention on time perception than the currently dominant Attentional Gate Theory.

Out latest works has focussed on the question whether and how people are able to perceive multiple intervals in parallel. We have found evidence for the fact that people are reasoanbly accurate at this, even though they only use a single timing mechanism (van Rijn & Taatgen, submitted).

Funding

This project is funded by the Office of Naval Research

Key references

Taatgen, N. A., Rijn, H. v., & Anderson, J. R. (2007). An Integrated Theory of Prospective Time Interval Estimation: The Role of Cognition, Attention and Learning. Psychological Review, 114(3), 577-598. (pdf and models)

van Rijn, H. & Taatgen, N.A. (2008). Timing of multiple overlapping time intervals: How many clocks do we have? Acta Psychologica, 129(3), 365-375. (pdf)