Glance, glimpse or stare? The discrimination of gaze duration

Abstract

While a considerable amount of work has focused on the perception of gaze direction, there has been little consideration of gaze duration. This is surprising insofar as social cues mediated by gaze are contingent on interactions between perceived gaze direction and duration. Participants were presented with two computer-generated faces one on either side of fixation. In the mutual gaze condition, participants were asked to judge which was the longer of two intervals defined by a shift in gaze from one side of the participant's head to mutual gaze and back. In the averted gaze condition, intervals were defined by a switch in gaze from one side of the participant's head to the opposite side and back. The standard gaze duration presented on one side of fixation lasted 625 ms. The comparison interval presented on the other side ranged from 100 to 1600 ms to define a psychometric function. The order of presentation was randomised over trials. The discrimination threshold was defined as the slope of the psychometric function. For upright faces we found that participants' discrimination of mutual gaze duration was better than discrimination of intervals of averted gaze. In contrast, there was no difference between mutual and averted discrimination for inverted faces. In a second experiment, using similar methods, we show that participants' discrimination of eye-colour did not vary as a function of gaze direction, when changes in iris colour coincided with changes in gaze. This argues against selective enhancement of upright mutual gaze duration discrimination mediated by greater attentional capture or elevated arousal, as these accounts predict improved discrimination across a range of psychophysical tasks. We propose that the detection of upright mutual gaze recruits a gated domain specific timing mechanism, with a high temporal resolution, allowing the precise representation necessary to interpret the social cues afforded by mutual gaze.