Is perceptual learning unimodal?

Date of Completion

January 2004


Psychology, Experimental|Psychology, Cognitive




Two experiments assessed whether experience in a supplemental perceptual modality facilitates perceptual learning in a focal perceptual modality. Experiment 1 evaluated change in the perception of an environmental property (object length) in each of three focal modalities (vision, audition, and haptics) when perceivers were provided with the opportunity to experience the same environmental fact by means of a supplemental modality (for example, haptics supplementing vision, vision supplementing audition, or auditory supplementing haptics). Experiment 2 did so over a time frame that allowed for analysis of long-term structure in learning dynamics. In both experiments, perceptual learning was assessed by changes in attunement—correlation of perceptual reports with a task-specific energy pattern—and calibration—the accuracy and consistency of perceptual reports with respect to their actual metrical values. In Experiment 2, long-term structure was considered at the level of the learning curve and at the level of the trial-to-trial variability. Both experiments revealed an increased correlation with I 1 a task-specific inertial property, for perceived length by audition. Perceived length in all modalities increased in precision regardless of whether training included experience in a supplemental perceptual modality. Haptics and audition additionally showed increased accuracy, but only when training included experience in a supplemental perceptual modality. Experiment 2 showed that learning curves in each of three perceptual modalities could be accommodated by a single learning curve in which auditory perception over the short term was shown to scale to haptic perception over the mid-term and both were shown to scale to visual perception over the long term. Analysis of trial-to-trial variability revealed patterns of long-term correlations in all perceptual modalities regardless of training. Although this pattern is somewhat at odds with the analysis of learning curves, it is consistent with recent research on long-term correlations in cognitive and motor tasks. ^