Task demand modulates the effects of reward learning on emotional stimuli.

The current study used event-related potentials (ERPs) to examine the ability of task demand in modulating the effect of reward association on the processing of emotional faces. In the learning phase, a high or low reward probability was paired with male or female facial photos of angry, happy, or neutral expressions. Then, in the test phase, task demand was manipulated by asking participants to discriminate the emotionality or the gender of the pre-learned face with no reward at stake. The ERP results in the test phase revealed that the fronto-central N1 (60-100 ms) and the VPP (160-210 ms) components were sensitive to the interaction between reward and emotion, in that the differences between the mean amplitudes for high- and low-reward conditions were significantly larger in the neutral face and angry face conditions than in the happy face condition. Moreover, reward association and task demand showed a significant interaction over the right hemisphere for the N170 component (140-180 ms), with amplitude difference between high- and low-reward conditions being larger in the emotion task than that in the gender task. The later N2pc component exhibited an interaction between task demand and emotionality, in that happy faces elicited larger N2pc difference waves than angry and neutral faces did in the emotion task, but neutral faces elicited larger N2pc difference waves than angry faces did in the gender task. The N2pc effect aligned with behavioral performance. These results suggest that reward association acts as an 'emotional tagging' to imbue neutral or angry faces with motivational significance at early time windows. Task demand functions in a top-down way to modulate the deployment of attentional resources at the later attentional selection stage, but does not affect the early automatic processing of either emotion or reward association.