Support from a TMS/MEP study for a direct link between positive/negative stimuli and approach/avoidance tendencies.

Previous behavioral studies using stimulus-response compatibility tasks have shown that people are faster to carry out instructed approach/avoidance responses to positive/negative stimuli. This result has been taken as evidence that positive/negative stimulus valence directly activates a tendency to approach/avoid, which in turn, facilitates execution of instructed approach/avoidance behavior. In these studies, however, it cannot be excluded that the results reflect a purely semantic link between stimulus valence and instructed responses. According to this alternative interpretation, positive/negative stimuli do not elicit an approach/avoidance tendency, but instead they interact with the positive/negative valence of the instructed responses, and in this way, produce the observed compatibility effect. To circumvent this possible disadvantage of compatibility tasks, we used a novel method for the measurement of early action tendencies: TMS induced MEPs. In two experiments, participants were first trained to abduct the index finger to approach and the thumb to avoid. Then, they observed a series of positive and negative stimuli. Each stimulus was followed by a TMS pulse (at 400 ms post-stimulus onset) and MEPs were measured continuously on the muscles of both fingers. These observation trials were randomly intermixed with response trials, in which neutral stimuli were presented and participants were instructed to approach/avoid the stimuli. In Experiment 1, participants received clear visual feedback on the outcome of their response in the response trials. In Experiment 2, we omitted this feedback to test whether it was necessary for the effect to occur. The results indicated higher MEPs for the approach/avoidance finger after positive/negative stimuli in Experiment 1 but not in Experiment 2. Analyses on the data aggregated over both experiments suggest that the visual feedback was necessary for stimulus valence to elicit action tendencies. Taken together, the results are in line with the results of behavioral studies with compatibility tasks, suggesting that stimulus valence directly elicits specific action tendencies already at 400 ms but they indicate that clear visual feedback is necessary for this effect to occur.