Seeing emotions in the eyes – inverse priming effects induced by eyes expressing mental states

OBJECTIVE Automatic emotional processing of faces and facial expressions gain more and more of relevance in terms of social communication. Among a variety of different primes, targets and tasks, whole face images and facial expressions have been used to affectively prime emotional responses. This study investigates whether emotional information provided solely in eye regions that display mental states can also trigger affective priming. METHODS Sixteen subjects answered a lexical decision task (LDT) coupled with an affective priming paradigm. Emotion-associated eye regions were extracted from photographs of faces and acted as primes, whereas targets were either words or pseudo-words. Participants had to decide whether the targets were real German words or generated pseudo-words. Primes and targets belonged to the emotional categories "fear," "disgust," "happiness," and "neutral." RESULTS A general valence effect for positive words was observed: responses in the LDT were faster for target words of the emotional category happiness when compared to other categories. Importantly, pictures of emotional eye regions preceding the target words affected their subsequent classification. While we show a classical priming effect for neutral target words - with shorter RT for congruent compared to incongruent prime-target pairs- , we observed an inverse priming effect for fearful and happy target words - with shorter RT for incongruent compared to congruent prime-target pairs. These inverse priming effects were driven exclusively by specific prime-target pairs. CONCLUSION Reduced facial emotional information is sufficient to induce automatic implicit emotional processing. The emotional-associated eye regions were processed with respect to their emotional valence and affected the performance on the LDT.