The Costs of Non-Simultaneity of Action

Raibert (1978) has pointed out that our limbs are useful tools only if they will do our bidding. While it is abundantly clear that, for the most part, our limbs do indeed do as bidded, what is not clear is how the language of our wishes is translated into a language which motoneurons and muscles understand. A simple example should suffice: “close your eyes and move your hand to grasp the dial and then twist counterclockwise”. Even though this specification gives no explicit information about the requisite muscle forces, the action is carried out as intended. This, in spite of the fact that answers to the question only recently phrased by Stein (1982, in press) as “What muscle variable(s) does the nervous system control in limb movements?” still eludes us. The behavioural approach adopted in the present set of experiments aims to shed some light on the above issue by examining timing constraints. A pervasive finding in movement control and coordination is that the relative or proportional timing of an act remains invariant over a wide range of kinematic changes in the act (e.g. amplitude and force). In previous work (Kelso, Southard and Goodman, 1979) it was shown that even under differing task demands in a two-handed coordination task, the two hands exhibit simultaneity of response. In the present study movement initiation and execution are examined when the timing demands of the task are altered for each hand. Reaction times in the first two experiments were inflated relative to control conditions which required simultaneity of movement. This was taken as preliminary indication of the costs involved in preparing for non-simultaneous action. The third experiment further examined the issue in a choice reaction time paradigm. Either simultaneous or non-simultaneous two handed movements were required. Twelve subjects performed 4 blocks of 48 trials in which they responded to one of two equi-probable stimuli by performing a two-hand response. The costs of non-simultaneity of action are revealed by examining the characteristics of the speed-accuracy relationship. Reaction times were inflated on those trials requiring a non-simultaneous response. However, movement times for the hand striking first (in the non-simultaneous trials) were not significantly different from the movement times of those trials in which a simultaneous response was required. Evidence is provided that timing is intrinsic to the organization (as distinct from the execution) of an act. This regulation of timing and costs incurred by disruption are further discussed from a dynamics perspective.