Binding and retrieval have been portrayed as key mechanisms in human action control. This theoretical framework posits that features of perception and action (e.g., object color, effector, etc.) are bound to each other. Afterwards, re-encountering one feature can retrieve bound features, thereby facilitating action repetition. Previous research focused on bindings with discrete response features (e.g., responding with a left or right button). This focus neglects the rich information required to control body movements. Here, we investigated continuous response features of extended movements related to both action goals and the implementation of these goals. Participants swiped their fingers on a tablet toward a small, briefly presented target circle that was accompanied by a sound. We analyzed response errors and similarity between successive response positions, depending on the sequence of (1) swiped hands, (2) irrelevant sounds, and (3) target positions. Sound repetitions did not bias responses toward preceding target or response positions whereas hand repetitions biased responding toward both positions, especially when target positions repeated. On the one hand, stimulus-response binding and retrieval seem to be remarkably focused on discrete response features, thus challenging the role of this mechanism in controlling overt body movements. On the other hand, fine-grained features of a response can be bound and retrieved from other response features, providing short-cuts to both the agents' previous action goals and the implemented deviation from these goals.Public significance statementHumans do not explicitly think about every action. Instead, they automatically connect what they perceive with how they act, a process that facilitates repetition of actions. Yet, which aspects of perception and action can become connected remains a topic of debate. Our study shows that these connections may extend beyond categorical action features (e.g., pressing a certain button on a phone) to metric action features (e.g., touching a specific position on a screen).
Foerster et al. (Wed,) studied this question.