Gaze-induced joint attention persists under high perceptual load and does not depend on awareness

The automaticity of gaze-induced joint attention is well known in relatively easy cognitive tasks; but its role in harder tasks had never been examined. This encouraged us to study automaticity in hard tasks, tasks presenting the subjects with high perceptual loads. The Rapid Serial Visual Presentation (RSVP) paradigm was used to present participants with two streams of bilaterally displayed digit-flows while they fixated at the center of a synthetic representation of a human face. The face was presented both above (Experiments 1 and 2) and below (Experiment 3) the face's visual threshold (henceforth called "supraliminal" and "subliminal", respectively). Interocular suppression was used to make the face stimulus invisible (subliminal). In the critical trials of all three experiments, the gaze direction shown on the face was randomly diverted to either the left or to the right. This directed the participant's gaze either towards or away from the location of a target in the RSVP. The perceptual load was always relatively high. It was either set (Experiments 1 and 3) or manipulated (Experiment 2) during the experiment. In all three experiments, an appreciably higher and significant level of target detection was found when an uninformative gaze-cue was congruent with the location of the target. This result, which had only been reported with relatively easy tasks previously, is called the "gaze-cueing effect". Our novel findings include showing that: (i) the attentional effect of gaze persists under high perceptual loads, and (ii) awareness of the gaze stimuli is not required to obtain the gaze-cueing effect. They also serve to validate prior support for an important role of automaticity in gaze-induced joint attention.