In the course of development organisms learn to associate their actions with the effects these actions have in the environment. Recent studies have shown that perceiving or anticipating action-effects automatically activates actions, which formerly have been experienced to cause these effects (Elsner & Hommel, 2001). Using subliminal priming paradigms and electrophysiological measures I investigated whether subliminally (i.e., not consciously perceivable) presented action-effects can automatically activate associated actions and if so, whether this response priming by action-effects can bias free-choice actions. Secondly I investigated whether action-effects with different emotional valences influence response selection differently. To address the first question three experiments were performed. Each experiment consisted of two experimental phases. The first phase, the acquisition-phase, was a learning phase were simple key-press actions were associated with simple visual stimuli (i.e., action-effects; diamond or square) that were contingent on the actions. Immediately after the acquisition-phase the test-phase followed, in which participants performed free-choice actions after the presentation of a Go-signal. In Experiments 2 and 3 a NoGo-signal indicating that responses had to be withheld could appear with the same likelihood as the Go-signal. Unknown to the participants, one of the former action-effects (diamond or square) was presented subliminally prior to each Go- and NoGo-signal to investigate the influence of unconscious action-effects on response selection. Taken together, the results of the test-phases provided strong evidence that even subliminally presented (i.e., unconscious) action-effects can automatically activate associated responses. The response priming by action-effects became evident in the lateralized readiness potential (LRP), an electrophysiological indicator of specific response activation processes. Under certain circumstanced this automatic response activation can bias free-choice actions although participants experienced the actions as freely chosen. In the test-phase of the first experiment more acquisition-phase-consistent than –inconsistent responses were chosen. If, for instance, a left key-press had been associated with a square during the acquisition-phase, the left key was chosen significantly more often after the subliminal presentation of a square in the test-phase. At least three factors seemed to influence which responses were chosen and executed: The strength of the priming effect, the complexity of the task (i.e., pure Go-blocks or intermixed Go/NoGo-blocks), and the elapsed time between the prime stimulus and the Go-signal. To address the second question simple key-press actions were linked to action-effects with different emotional valences (positive vs. negative pictures accompanied by high or low tones) during the acquisition-phase. In the subsequent test-phase, the effects-tones that had been associated with negative or positive pictures were presented and followed by a Go-signal, after which participants had to freely choose to press one of the two response keys. Results indicated that the anticipation of the emotional valence of an action-effect influenced free-choice action. Whereas the effect-tones induced a clear response bias (i.e., more acquisition-consistent than –inconsistent key-choices) if they had been associated with a positive emotional valence, this response bias was not reliable for action-effects associated with negative emotional features. In summary, the present results provide further proof for ideomotor theories of action control (James, 1890; Elsner & Hommel, 2001) which state that actions are automatically activated by anticipating their consequences.

In the course of development organisms learn to associate their actions with the effects these actions have in the environment. Recent studies have shown that perceiving or anticipating action-effects automatically activates actions, which formerly have been experienced to cause these effects (Elsner & Hommel, 2001). Using subliminal priming paradigms and electrophysiological measures I investigated whether subliminally (i.e., not consciously perceivable) presented action-effects can automatically activate associated actions and if so, whether this response priming by action-effects can bias free-choice actions. Secondly I investigated whether action-effects with different emotional valences influence response selection differently. To address the first question three experiments were performed. Each experiment consisted of two experimental phases. The first phase, the acquisition-phase, was a learning phase were simple key-press actions were associated with simple visual stimuli (i.e., action-effects; diamond or square) that were contingent on the actions. Immediately after the acquisition-phase the test-phase followed, in which participants performed free-choice actions after the presentation of a Go-signal. In Experiments 2 and 3 a NoGo-signal indicating that responses had to be withheld could appear with the same likelihood as the Go-signal. Unknown to the participants, one of the former action-effects (diamond or square) was presented subliminally prior to each Go- and NoGo-signal to investigate the influence of unconscious action-effects on response selection. Taken together, the results of the test-phases provided strong evidence that even subliminally presented (i.e., unconscious) action-effects can automatically activate associated responses. The response priming by action-effects became evident in the lateralized readiness potential (LRP), an electrophysiological indicator of specific response activation processes. Under certain circumstanced this automatic response activation can bias free-choice actions although participants experienced the actions as freely chosen. In the test-phase of the first experiment more acquisition-phase-consistent than –inconsistent responses were chosen. If, for instance, a left key-press had been associated with a square during the acquisition-phase, the left key was chosen significantly more often after the subliminal presentation of a square in the test-phase. At least three factors seemed to influence which responses were chosen and executed: The strength of the priming effect, the complexity of the task (i.e., pure Go-blocks or intermixed Go/NoGo-blocks), and the elapsed time between the prime stimulus and the Go-signal. To address the second question simple key-press actions were linked to action-effects with different emotional valences (positive vs. negative pictures accompanied by high or low tones) during the acquisition-phase. In the subsequent test-phase, the effects-tones that had been associated with negative or positive pictures were presented and followed by a Go-signal, after which participants had to freely choose to press one of the two response keys. Results indicated that the anticipation of the emotional valence of an action-effect influenced free-choice action. Whereas the effect-tones induced a clear response bias (i.e., more acquisition-consistent than –inconsistent key-choices) if they had been associated with a positive emotional valence, this response bias was not reliable for action-effects associated with negative emotional features. In summary, the present results provide further proof for ideomotor theories of action control (James, 1890; Elsner & Hommel, 2001) which state that actions are automatically activated by anticipating their consequences.