Processing Mechanisms Of Eye-Head Cues And Eye-Finger-Pointing Cues In The Dot-Perspective Task

Calculating others’ visual perspective automatically is a pivotal ability in human’s social communications. In the dot-perspective task, the ability is shown as a consistency effect: adults respond more slowly to judge the number of discs that they can see when a computer-generated avatar sees fewer discs. The implicit mentalising account attributes the effect to relatively automatic tracking of others’ visual perspectives. However, the submentalising account attributes the effect to domain-general attentional orienting. Accordingly, three studies were conducted to elucidate the ongoing implicit mentalising vs. submentalising debate.

Study 1 (comprising Experiments 1 and 2) replicated the consistency effect either when real-human-face or spatial layout of discs was considered. Study 2 (comprising of Experiments 3 and 4) dissociated two accounts by manipulating real human’s facial cues. In Experiment 3, using a new visual access manipulation (i.e., a black rectangle placed on an agent’s eyes for rendering an invisible condition), a consistency effect was induced for eyes-opened but not eyes-covered faces with head direction, suggesting implicit mentalising. Experiment 4 firstly compared implicit mentalising (via consistency effect in the dot-perspective task) with attentional orienting (via a cue-validity effect in Posner task) when manipulating eye-head cues (head-front-gaze-averted versus head-turned-gaze-maintained). Neither effect was modulated by eye-head-related directional cue, but the cue-validity effect’s elicitation seemed to be related to the directional cue’s dynamic property. Overall, implicit mentalising as revealed in consistency effect cannot be purely reduced to attentional-orienting-related submentalising processes.

Study 3 (comprising of Experiments 5 to 7) further clarified the debate by considering the agent’s different body cues. Experiment 5 extended the findings of Experiment 4 by generating a new eye-head-cue comparison (head-front-gaze-averted vs. head-turned-gaze-averted). Directional cue modulated cue-validity effect but not consistency effect, favouring Study 2’s conclusion. Experiment 6 adopted a new body-cue-manipulation (gaze-averted vs. finger-pointing). Both cue-validity and consistency effects were elicited for finger-pointing but not gaze-averted agents, supporting submentalising. Experiment 7 combined finger-pointing with visual access’s manipulation (eyes-opened vs. eyes-covered) on the dot-perspective task. Visual access did not modulate the consistency effect when finger-pointing was simultaneously displayed, supporting submentalising. Altogether, gaze aversion cues appear to play a dominant role in moderating implicit mentalising on the dot-perspective task, but the process may be interfered by the easily-discriminable finger-pointing cues via an attentional orienting mechanism.