Abstract
Automatically recognising threat-related expressions is critical for humans’ survival as it allows for rapid evaluation of a threatening stimulus that subsequently elicits an immediate behavioural response. Research has consistently found that angry expressions are automatically processed and recognised. However, research investigating automatic recognition of fearful expressions has provided conflicting results. One possible explanation for why evidence has not always been found for automatic processing of fearful expressions is that eye-gaze direction was not considered as a factor modulating emotion recognition. Typically, only direct gaze stimuli have been used in experimental studies. However, the shared signal hypothesis and appraisal theory both propose that greater recognition of angry expressions is associated with a direct gaze, whereas averted gaze enhances the recognition of fearful expressions. Accordingly, eye-gaze direction was incorporated into the current experimental design. The present study investigated whether the automatic recognition of task-irrelevant, threat-related expressions is modulated by eye-gaze direction. In three experiments, angry and fearful emotion words were superimposed upon faces expressing angry, fearful or neutral expressions to create congruent, neutral and incongruent trials. The faces also depicted a direct or an averted eye-gaze direction. Participants were instructed to ignore the face and categorise the emotion word into the correct emotion category as quickly and accurately as possible. In Experiment 1 the stimuli depicted natural portrayals of gaze aversion, whereas in Experiments 2 and 3 gaze aversion was digitally manipulated. Additionally, in Experiment 3 emotional intensity was manipulated to depict a mild or intermediate version in order to reduce the discriminability of angry and fearful expressions and enhance the significance of gaze cues. The results across the three experiments showed significant congruency effects for angry and fearful expressions regardless of gaze direction, providing evidence that both threat-related expressions are automatically recognised. Additionally, the results from Experiment 3 showed weaker congruency effects in the mild intensity condition compared to the intermediate condition, confirming that emotional expressions were less discriminable in the mild intensity condition. However, reducing emotion discriminability had no effect on the significance of gaze cues as gaze direction did not impact emotion recognition in either intensity condition. In conclusion, this study indicates that the automatic recognition of threat-related expressions is not modulated by eye-gaze direction in a selective attention task.