Abstract:  Increasing evidence has indicated that emotional information, and particularly threatening visual input, elicits faster behavioral responses than non-threatening stimuli. This superior processing of threatening information is also found under conditions where consciousness is absent. However, recent studies found that faster unconscious detection of emotion-associated stimuli than neutral stimuli may be due to their unmatched physical characteristics, rather than by their emotional content. Thus, it is necessary to test whether emotional stimuli still have the processing advantage over neutral ones in unconscious conditions when low-level visual properties are matched. In order to investigate whether unconsciously prioritized processing still occurs with emotion-associated stimuli which are physically identical, we used the conditioning paradigm to manipulate the affective significance of Gabor patches. Participants performed two challenging visual detection tasks under the breaking Continuous Flash Suppression (b-CFS) paradigm. In experiment 1, differently oriented Gabor patches (45° and 135°) were used as material. During an initial learning phase, one oriented Gabor patch (e.g., 45°) was paired with an alarm sound (CS+), whereas the other was never paired with the alarm sound (CS–). The emotional rating indicated that negative emotion could be elicited by the alarm sound in the participants. The orientation of CS+ Gabor patches was counterbalanced across participants. In the subsequent testing phase, participants were required to discriminate the location of the Gabor patch relative to the central fixation as quickly and accurately as possible. In this phase, Gabor patches were suppressed by dynamic noise using b-CFS. The procedure in experiment 2 was the same with that in experiment 1, except that the color of the Gabor patches was also varied, between red and green. In experiment 1, there was no difference in the accuracy rates between CS+ stimuli and CS– stimuli (99% vs. 99%). Suppression time results showed that CS+ stimuli emerged from suppression faster than CS– ones. In experiment 2, there was no difference in the accuracy rates for different learning condition. For the analysis of suppression time, the “learning effect” was computed to represent difference between experimental conditions and control condition. Integrated learning showed a significant learning effect, while there was no remarkable learning effect in orientation learning or in color learning condition. These findings revealed an unconscious processing advantage for aversive conditioned stimuli. Furthermore, the learning effect was specific to the conditioned stimuli and could not generalize to other similar objects. Taken together, this study provided further evidence for the optimized processing of affectively significant visual stimuli in unconscious conditions.

Key words: affective learning, breaking Continuous Flash Suppression, unconscious processing