Abstract
Research on tool integration has focused on whether structural representations of tools are assimilated into the "body schema", our continually updated unconscious representation of our body's posture and extension. However, little attention has been paid as to whether tool assimilation occurs in how we plan for action. In particular, does the forward model for action assimilate a structural representation of grasped tools? Smooth pursuit (SP) is a result of a predictive forward model that anticipates the direction and the velocity of visual objects. However, SP also occurs when visually tracking one's own hand, even in total darkness. When SP occurs for the hand in the absence of visual input, this is due to a prediction of the direction and velocity of the target from the forward model. Thus, if subjects show greater SP for an occluded tool than an imagined tool, or no stimulus, this would suggest that the forward model also contains a prediction of tool direction and velocity. To test this possibility, we recorded SP eye movements (N = 9) under eight conditions. In four "visible" conditions subjects tracked: (1) a moving dot, (2) their finger, (3) the tip of a grasped rod, and (4) an imaginary rod tip, while grasping a rod handle. In the "occluded" conditions subjects made movements behind a screen displaying dynamic noise. In these conditions they tracked: (5) no movement, (6) their finger, (7) a rod tip, and (8) an imaginary rod tip while grasping a rod handle. As predicted, we found greater SP for the occluded tool condition than in either the occluded tool handle condition (p <.021) or the no stimulus condition (p <.047). This supports our hypothesis that the forward model assimilates a structural representation of grasped tools into its model of action outcomes.
Meeting abstract presented at VSS 2012