Abstract: Observing another person’s averted eye gaze leads to automatic shift of attention to the same object and facilitates subsequent early visual processing. This phenomenon is termed ‘Joint attention’. Joint attention processing proceeds through two main stages: gaze perception and gaze following. Gaze perception refers to analysis of the perceptual features of a gaze cue. By contrast, gaze following refers to the tendency of observers to shift attention to locations looked at by others, which is indicated by gaze cueing effect (GCE). Most of researchers maintained that the latter process relied on the former (serial model). They hold an idea that mechanisms involved in gaze perception precede those involved in attentional cueing from gaze. However, other results suggested that gaze perception and gaze following might involve dissociable mechanisms (parallel model). As Doherty et al. (2009) reported in young children, it was possible for gaze following to occur in the absence of precise gaze perception. Thus, it remains unclear what role the gaze perception plays in gaze following behavior. In the current study, a gaze adaptation technique and a gaze cueing paradigm were combined to investigate the effects of gaze perceptual adaptation on gaze cueing effect. We hope to figure out the relationship between gaze perception and gaze following.
Three experiments were performed in this study with the gaze adaptation technique and the gaze cueing paradigm. The gaze direction aftereffect was re-examined using perceptual adaptation technique in experiment 1. The gaze cueing effects in different gaze angles (5 degree and 10 degree) were examined in experiment 2. Crucially, the influence of gaze perceptual adaptation on the shift of attention induced by averted gaze was investigated in experiment 3. Forty two subjects participated in the three experiments (12 subjects in experiment 1, 12 subjects in experiment 2, and 18 subjects in experiment 3).
In experiment 1, we repeated a strong gaze direction aftereffect, finding that perception of gaze in the adapted direction was significantly impaired, which was consistent with previous studies. It provided a clear evidence for distinct neuronal channels tuned to leftward- and rightward-oriented gaze in humans. In experiment 2, the GCE induced by 10° gaze was significantly stronger than 5° condition, which suggested that gaze following was influenced by stimulus salience through a bottom-up mechanism. Crucially, the shift of attention induced by averted gaze was found significantly decreased after gaze direction adaptation in experiment 3, which indicated that the extraction of gaze direction modulated by perceptual adaptation could affect the gaze following behavior.
In conclusion, the present study provided critical behavioral evidence that there existed a direct link between gaze perception and gaze following under conscious condition. The spatial encoding and attention shift happen only if gaze direction has been extracted correctly. Based on its neural underpinnings of adaptation, “the psychologist’s microelectrode’’, we inferred that a cortical visual pathway flowing from superior temporal sulcus (STS) to intraparietal sulcus (IPS) could relate observed gaze to visual attention. Moreover, gaze following was not purely reflexive. It was also sensitive to top-down modulation triggered by perceptual experience.

Key words: joint attention, gaze perception, gaze following, perceptual adaptation, gaze cueing effect