When repetitively lifting an object with randomly varying mechanical properties, the fingertip forces reflect the previous lift. We examined the specificity of this "sensorimotor memory" by observing the effects of an isolated pinch on the subsequent lift of a known object. In this case, the pinch force was unrelated to the fingertip forces necessary to grip the object efficiently. The peak grip force used to lift the test object (4 N weight) depended on the preceding task. Compared with repetitively lifting the 4 N test object, the peak grip force was 2 N greater when a lift of the same object was preceded by a lift in which a hidden mass was attached to the object to increase the weight to 8 N. This 2 N increase in grip force also occurred when subjects lifted the 4 N test object after pinching a force transducer with a force of 8 N. Thus, similar grip forces were stored in sensorimotor memory for both tasks, and reflected subjects' use of 7.9 +/- 1.1 N to lift the 8 N object. Similar effects occurred when the preceding pinch or lift was performed with the opposite hand. The peak lift force was unaffected by the isolated pinch, suggesting that a generalized increase in fingertip and limb forces did not occur. We conclude that the sensorimotor memory is not specific for lifting an object. It is doubtful that this particular memory stores the physical properties of objects or reflects a forward internal model for predictively controlling fingertip forces.
Quaney, B. M., Rotella, D. L., Peterson, C., & Cole, K. J. (2003). Sensorimotor memory for fingertip forces: evidence for a task-independent motor memory. The Journal of Neuroscience : The Official Journal of the Society for Neuroscience, 23(5), 1981–6. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12629204