How mechanical context and feedback jointly determine the use of mechanical variables in length perception by dynamic touch.

Earlier studies have revealed that both mechanical context and feedback determine what mechanical invariant is used to perceive length by dynamic touch. In the present article, the authors examined how these two factors jointly constrain the informational variable that is relied upon. Participants were to judge length while wielding a rod or while holding it stationary. In two experiments, it was tested whether perceptual learning effects in the wielding condition transferred to the holding condition and vice versa. There was an asymmetry in transfer across mechanical conditions: Improvements in the holding context transferred to the wielding condition, but not vice versa. Examining the individuals' exploitation of mechanical variables, we found that, after feedback, participants changed in information usage in both mechanical conditions. For many participants, these changes were not confined to the mechanical context in which the feedback was provided. Indeed, feedback in one mechanical context brought about changes in information usage that often manifested themselves in the other mechanical condition. The authors explore the implications of these findings for research on perceptual learning in dynamic touch and the salience hypothesis.