Abstract:

Abstract: Automatic perspective taking refers to an individual tracking other people's opinions or being affected by others' perspectives, even when he or she is not explicitly required to adopt such perspectives. This phenomenon is revealed by the consistency effect, which suggests that a worsening perspective taking performance is triggered when others' perspectives and one's own perspective are different. There are two main views regarding automatic perspective taking: the implicit mentalizing view holds that automatic perspective taking is a domain-specific process that spontaneously selects and processes others' perspectives, and the submentalizing view proposes a domain-general process, such as reflective attentional orientation or the spatial coding of a location, that simulates the role of the mentalizing process in thesocial environment.

The common tasks in automatic perspective taking include the dot perspective taking task, the "social" Simon effect, the ambiguous number task, and the anticipatory looking paradigm. The consistency effect found in each task can be explained by both the implicit mentalizing view and the submentalizing view. Among the abovementioned tasks, the dot perspective taking task is the most typical type of automatic perspective taking task due to its utilization of the least cognitive resources compared to other automatic perspective taking tasks. Most behavioural experiments involving the dot perspective taking task pay great attention to the implicit mentalizing factors (including the social relevance of cues, the visual attribution state of cues, and the acquisition of a social perspective), but the results of the same factors support different views. In addition, cognitive neuroscience studies mainly compare the distinct neural mechanisms of one's own perspective with those of others' perspectives, but the consistency effect from a single perspective is less explored.Therefore, the interpretation of automatic perspective taking is still debatable.

The evidence of the two views in automatic perspective taking shows that one view is unable to completely negate the other one, so it is possible to synthesize the two views into a framework, i.e., the synergistic model of implicit mentalizingand submentalizing processes.The consistency effect in automatic perspective taking can be achieved in three ways:(1) after the presentation of a visual stimulus that is first perceptually processed, if only directional cues but no social cues are found, then the consistency effect will be triggered by a directional process;(2) if social cues are present but do not reach the threshold of social perception activation, then they are instead processed as directional cues, and thus, automatic perspective taking will be activated by the submentalizing process; and (3) if social cues are present and reach the threshold of social perception activation, then automatic perspective taking is activated by the implicit mentalizing process. If social cues are present and reach the threshold of social perceptual activation and directionality is enhanced, then the implicit mentalizing and submentalizing processes will be performed together. The processing threshold of activating social perception has not been clearly defined, so further evidence is required to determine this threshold and to establish the occurrence context of minimum automatic perspective taking.

Previous studies adopting the dot perspective taking task have focused more on the implicit mentalizing factors with different task details, in contrast to the submentalizing factors, so it is difficult to directly compare the results. Second, the small sample sizes of the experiments may reduce the reproducibility of the results. In addition, special individuals, such as infants, groups with autism spectrum disorder, and deaf children, may have different developmental performance levels in terms of perspective taking. Therefore, manipulating the implicit mentalizing and submentalizing factors of special individuals, as well as utilizing eye-tracking and/or cognitive neuroscience technologies, can provide further evidence of the underlying mechanisms of automatic perspective taking.

Key words: automatic perspective taking, implicitmentalizing, submentalizing