Abstract:

During social communication, eyes gaze indicates the focus of people's attention. There are two types of gaze cue: direct gaze (straight) and averted gaze (left or right), which have different social functions. Previous researchers assume that gaze perception involves two different processing stages. The first stage refers to earlier visual feature analysis of eyes such as luminance contrast and geometry of white sclera and iris. The second stage involves the perceptual integration of eyes’ features and the extraction of gaze direction. However, this assumption ignores the differences in processing mechanism between direct gaze and averted gaze. Other researchers hypothesize that perception of direct gaze is mediated by a fast-track pathway via subcortical face processing route. Direct gaze perception has characteristics of automatic process, such as capturing attention and being processed unconsciously. Its automaticity shows that direct gaze has a processing advantage over averted gaze. We inferred that direct gaze perception was independent of attentional resources, while averted gaze perception required more attentional resources. The present study aimed at investigating how attentional resources affect perception of averted and direct gaze.

We combined Lavie’s perceptual load task and a short-term gaze adaptation paradigm to investigate how attentional resources influence the gaze adaptation aftereffect of direct gaze and averted gaze. Participants attended to a letter string superimposed on gaze adaptor (200 ms), and identified the target letter (X or N) embedded in the letter string of either 6 identical letters (low load) or 6 different letters (high load). Subsequently, a gaze probe was presented for 200 ms. Participants had to make judgements concerning the direction of probe (i.e. leftward, straight, or rightward). Sixteen na?ve participants (8 females) were tested in Experiment 1, which measured the modulation of averted gaze adaptation effect by perceptual loads. And twenty-two na?ve observers (11 females) participated in Experiment 2, which explored the effects of perceptual loads on direct gaze adaptation effect.

The results of Experiment 1 revealed that a significant aftereffect of averted gaze directed toward the adapted side in the low load condition. However, such gaze adaptation effect was eliminated under high load condition. Crucially, the gaze adaptation effect in low load condition was significantly greater than that in high load condition. It was suggested that the extraction of averted gaze direction (leftward or rightward) was modulated by attentional resources (perceptual load). When a demanding competing task has exhausted all available processing capacity, the direction of averted gaze could not be extracted. While the load was low, “spill over” capacity was available to the processing of task-irrelevant gaze direction. Results of Experiment 2 showed a significant direct gaze adaptation effect both under low and high load conditions, which were not significantly different from each other. These results indicated that the perceptual adaptation of direct gaze was not influenced by attentional resources. Although limited processing capacity was fully consumed by a high load task, direct gaze could be processed as well. Compared with averted gaze, processing of direct gaze required fewer attentional resources. Therefore, requirements to attentional resources were different between processing of averted and direct gaze.

In conclusion, the current study demonstrates that the processing of averted gaze direction is not mandatory automatic and it requires some degree of attentional resources. But gaze perception does not require attentional resources for direct gaze. Our findings indicated that processing of averted and direct gaze may be mediated by different underlying mechanisms, although distinct gaze directions are coded by a similar multichannel system with separate channels coding left, direct, and right gaze. Our results replicate and extend recent findings of different processing mechanisms between averted and direct gaze. Moreover, it may promote integration between different gaze perception models, such as multichannel system and fast-track modulator model.

Key words: direct gaze, averted gaze, gaze perception, gaze perceptual adaptation aftereffect, perceptual load