Abstract
Winkielman, Berridge, & Wilbarger (in press, PSPB) report that unseen happy and angry faces influence consumptive behaviors such as drinking and the rated value of a beverage. Central to their claim is the assumption that the emotionally expressive faces were not consciously perceived. Participants in their study were shown emotional faces for 16 ms, which were replaced by neutral faces for 400 ms. These display sequences influenced beverage consumption, even though participants were at chance in their attempts to identify the emotion in the prime faces. In the present work, we asked whether perception of either the prime or the mask face was influenced by interactions between features of both faces. In phase 1, participants made speeded classifications of happy and angry masks (450 ms) that were preceded by angry, happy, or neutral primes (22 ms) at one of three intervals (22 ms, 45 ms, or 67 ms). In phase 2, participants were instructed to classify these same prime faces as either happy or angry. Prime-mask congruence had different influences in the two tasks. In the mask classification task, emotionally congruent primes led to faster responses than incongruent primes and this effect increased with prime-mask interval. Featural similarity in the faces played no role in priming. In the prime classification task, the effect of emotional congruence interacted with featural similarity of the faces. For emotionally congruent faces, increased feature similarity improved accuracy whereas for incongruent faces it impaired accuracy. This underscores the importance of examining prime-mask interactions on unconscious influences on consumptive behavior. A second study examined the issue of task relevance in the masking of faces. We discuss the implications of these findings for unconscious action priming and conscious object recognition in the realm of rapid emotional processing.
Natural Sciences and Engineering Research Council of Canada, Izaak Killam Memorial Trust