Hand forces and placement are modulated and covary during anticipatory control of bimanual manipulation.

Dexterous object manipulation relies on the feedforward and feedback control of kinetics (forces) and kinematics (hand shaping and digit placement). Lifting objects with an uneven mass distribution involves the generation of compensatory moments at object lift-off to counter object torques. This is accomplished through the modulation and covariation of digit forces and placement, which has been shown to be a general feature of unimanual manipulation. These feedforward anticipatory processes occur before performance-specific feedback. Whether this adaptation is a feature unique to unimanual dexterous manipulation or general across unimanual and bimanual manipulation is not known. We investigated the generation of compensatory moments through hand placement and force modulation during bimanual manipulation of an object with variable center of mass. Participants were instructed to prevent object roll during the lift. Similar to unimanual grasping, we found modulation and covariation of hand forces and placement for successful performance. Thus this motor adaptation of the anticipatory control of compensatory moment is a general feature across unimanual and bimanual effectors. Our results highlight the involvement of high-level representation of manipulation goals and underscore a sensorimotor circuitry for anticipatory control through a continuum of force and placement modulation of object manipulation across a range of effectors. This is the first study, to our knowledge, to show that successful bimanual manipulation of objects with asymmetrical centers of mass is performed through the modulation and covariation of hand forces and placements to generate compensatory moments. Digit force-to-placement modulation is thus a general phenomenon across multiple effectors, such as the fingers of one hand, and both hands. This adds to our understanding of integrating low-level internal representations of object properties into high-level task representations.