Information about relative phase in bimanual coordination is modality specific (not amodal), but kinesthesis and vision can teach one another.

How is information from different sensory modalities coordinated when learning an action? We tested two hypotheses. The first is that the information is amodal. The second is that the information is modality specific and one modality is used in first learning the action and then is used to teach the other modality. We investigated these hypotheses using a rhythmic coordination task. One group of participants learned to perform bimanual coordination at a relative phase of 90° using kinesthesis. A second group used vision to learn unimanual 90° coordination. After training, performance using the alternate modality was tested in each case. Snapp-Childs, Wilson, and Bingham (2015) had found transfer of 50% of learned performance of 90° coordination between the unimanual and bimanual tasks when each had included use of vision. Now, we found essentially no transfer (≈5%) indicating that the information was modality specific. Next, post-training trials performed using the untrained modality were alternated with trials in which kinesthesis and vision were used. The result was that performance using the untrained modality progressively improved. We concluded that trained modality was used to teach the untrained modality and that this likely represents the way information from different sensory modalities is coordinated in performance of actions.