Submentalizing or Mentalizing in a Level 1 Perspective-Taking Task: A Cloak and Goggles Test

It has been proposed that humans possess an automatic system to represent mental states ('implicit mentalizing'). The existence of an implicit mentalizing system has generated considerable debate however, centered on the ability of various experimental paradigms to demonstrate unambiguously such mentalizing. Evidence for implicit mentalizing has previously been provided by the 'dot perspective task,' where participants are slower to verify the number of dots they can see when an avatar can see a different number of dots. However, recent evidence challenged a mentalizing interpretation of this effect by showing it was unaltered when the avatar was replaced with an inanimate arrow stimulus. Here we present an extension of the dot perspective task using an invisibility cloaking device to render the dots invisible on certain trials. This paradigm is capable of providing unambiguous evidence of automatic mentalizing, but no such evidence was found. Two further well-powered experiments used opaque and transparent goggles to manipulate visibility but found no evidence of automatic mentalizing, nor of individual differences in empathy or perspective-taking predicting performance, contradicting previous studies using the same design. The results cast doubt on the existence of an implicit mentalizing system, suggesting that previous effects were due to domain-general processes. Public Significance Statement The ability to represent in one's own mind what other people see, think, or believe is important for social interactions and relationships. There is wide agreement that this 'mentalizing' ability depends on a late developing, slow and effortful system, but much debate on whether humans also possess a fast and automatic mentalizing system. The present studies tested whether participants automatically represented what an onscreen human avatar could see. Objects' visibility was manipulated by using either a set of telescopes or goggles. One of each set allowed objects to be seen, and the other did not. Participant response times were predicted to be faster when what they saw corresponded to what the avatar saw, and slower when there was a difference. However, this did not occur, providing no evidence for an automatic mentalizing system, suggesting rather that representing others' mental states is effortful not automatic.