Adaptation to sensory-motor temporal misalignment: Instrumental or perceptual learning?

Abstract

Sensory-motor delays vary over the course of development and under different environmental conditions. Previous research has shown that humans can compensate for the resulting temporal misalignment while performing sensory-motor tasks (e.g., Cunningham, Billock, & Tsou, 2001a), but remains silent on the question of whether perceptual learning—similar to that involved in adaptation to spatial misalignment (e.g., Redding & Wallace, 1993) and in adaptation to purely intersensory misalignment (e.g., Fujisaki, Shimojo, Kashino, & Nishida, 2004)—is also involved in this adaptive response. Following an attempted replication of Cunningham et al.'s (2001a) study in a preliminary experiment, we present in this paper two experiments that demonstrate that after-effects of adaptation to temporal misalignment do not spontaneously decay. The literature on adaptation to spatial misalignment suggests that, while instrumental learning spontaneously decays in the absence of reinforcement, perceptual learning persists. Therefore our results are consistent with adaptation being effected through perceptual learning.

Keywords:

• Temporal adaptation
• Sensory-motor co-ordination
• Decay of after-effects
• Perceptual learning
• World learning

This manuscript is based on the first-year PhD research of the first author, under the supervision of the second and third authors. We would like to thank Aline Bompas for reading and commenting on a draft of this paper and Lorraine Woods for producing line drawings of the apparatus.

Notes

¹ The difference in after-effect between the control and the delay condition was not significant, t(18) = 0.8, p = .2. However, the relationship between after-effects and test speed showed a strikingly similar pattern of results to that in Cunningham et al.'s (2001a) data. Since pretest and posttest speeds were identical for each participant, in the absence of true after-effects speed should have no effect on the after-effect measure. Conversely, when there is an after-effect, higher speeds would translate into larger errors on the screen, so one should see a correlation between speed and after-effect. Indeed, in both Cunningham et al.'s (2001a) data and in ours, there was a moderate negative correlation in the delay condition (r = –.7, p = .01 in their data, r = –.6, p = .04 in ours) and not in the control condition (r = .01, p > .99 in their data, r = .04, p > .99 in ours). Thus, although we did not replicate Cunningham et al.'s (2001a) direct evidence of an after-effect, there is an indication that an after-effect was present that was too small to be reliably detected by a t test. Further information on this experiment may be obtained from the authors.