The use
of a programmable virtual
environment combined with a
motorized manual interface allows a unique opportunity to study the
nature of human
motor adaptation. We theorize that
humans make use of percepts of force and motion in order to accomplish
goal
directed action. In virtual
environments, novel feedback conditions and dynamic interactions may be
devised
that can work with or against the expectations of the human operator. By controlling these conditions in human
subject experiments, we would like to discover what features of a
mechanical
interaction influences the performance and learning of manual skills.
In the
manual control of an object, interaction forces are not random but
directly relate to the kinematics and object properties. In such
situations with predictable behavior, the human motor system may be
able to form a task appropriate strategy with a simple computational
structure. In the current study,
we examine
differences
between human motor adaptation responses to changes in movement
specification versus object parameters. For this purpose, we chose a
motor task where interaction with an external inertia is the cause of
forces between the arm and an environment. The task is relevant to the
study of human movement in that it requires adaptation to force
perturbations, and yet the kinds of interactions are familiar to
humans. We hypothesize that the human motor system controls the motion
of external objects using an internal representation of the object that
is distinct from the representation of the motion plan.
|