Spontaneous visual perspective-taking: level 2 representations of another’s perspective are not related to what they actually see

Abstract

The ability to understand what another person can see is a fundamental feature of human interaction, commonly referred to as ‘visual perspective-taking’ (VPT). Traditionally, this ability has been conceptualised along two ‘levels’, in line with developmental observations (Flavell, Everett, Croft & Flavell, 1981). ‘Level-1 VPT’ is a simple, automatic, and rapid process enabling understanding of what another individual can see. In contrast, ‘Level-2 VPT’ is a more complex process requiring conscious control, which enables understanding of how features of the environment look from another’s location. This has led to a widely accepted hypothesis that Level-1 VPT is necessary in the formation of Level-2 understanding. Recent research however is beginning to challenge these previous notions. Using a new paradigm which incorporates mental rotation (Shepard and Metzler, 1971) and perceptual interference (Riecke, Cunningham & Bülthoff, 2007) phenomena, Ward, Ganis and Bach (2019) demonstrated that Level-2 representations emerge rapidly, spontaneously and can facilitate (as well as inhibit) one’s own perceptual judgements. Building upon these findings, the current experiment implemented barriers into Ward et al’s paradigm. By blocking the other person’s line of sight to the target stimuli, this manipulation was able to assess how Level-1 judgements affect Level-2 understanding. Surprisingly, findings indicated that Level-2 representations of how the other person saw the target stimuli persisted, even when they could not see the stimuli in reality, and in contrast to participants' own verbal assessments. Implications for the relationship between levels and conceptualization of Level-2 VPT are discussed along with suggestions for further research.