Is automatic imitation based on goal coding or movement coding? A comparison of goal-directed and goal-less actions

A key issue for research on automatic imitation is whether it occurs primarily at the level of movements, that is, by automatically activating a representation of the movement/effector involved in the execution of the observed action, or at the level of goals, that is, by triggering a representation of the action goal, irrespective of how the motor act is physically instantiated. The present study presents two experiments aimed at investigating the contribution of movement coding and goal coding to automatic imitation, by assessing participants' performance in a spatial compatibility task where the observed stimuli were goal-directed and goal-less actions, which have been demonstrated to elicit, respectively, goal and movement coding. We found a significant automatic imitation effect both when the stimuli were goal-less actions and when they were actions directed toward a goal. However, the effect was stronger for the goal-less actions, even after controlling for saliency effects. These results suggest that goal coding contributes to automatic imitation, but to a lesser degree compared to movement coding. The implications of these results for theory and research on automatic imitation are discussed.