Furthermore, we conducted a grasping with pulling task (modified action task) to further test Ganel
and Goodale’s (2003) results. This task was necessary because the demands of the grasping task are
problematic; specifically, an obstacle avoidance mechanism might affect the grasp data. For example,
while reaching for the width of an object, a portion of the object extends into the participant’s hand;
the longer the object, the more it extends into the palm. Consequently, the actor might open the hand
wider for long objects than for short objects to avoid colliding with the object. This avoidance mechanism
may counteract the assumed effect of the task-irrelevant dimension (length) because short objects
would be grasped wider than long objects and vice versa. To rule out the possibility that a taskirrelevant
dimension effect on grasping is confounded by an obstacle avoidance mechanism, another
sample of adults grasped the objects by the right edge and pulled them a few centimeters across a table
(grasping with pulling task). When an object is grasped by its right edge, there is no need to avoid a
portion of the object because the object always extents into the palm by the same degree. If there were
no task-irrelevant dimension effect on grasping in the grasping with pulling task, this result would
confirm that of the original action task and provide a base from which to test children. If there were
a task-irrelevant dimension effect on grasping, this result would challenge that of the original action
task.