Abstract
Considerable debate has focused on whether adults possess an implicit system for representing others' mental states. Some argue that people automatically represent the perspective of others using evidence from altercentric interference-cases in which another agent's perspective affects the speed with which one can report one's own perspective. Others have argued that altercentric interference is not always specific to social stimuli and thus may represent a simpler process such as submentalizing. To distinguish between these positions, Study 1 developed a novel measure of altercentric interference-a "sandbox" measure-that allowed us to more sensitively assess altercentric interference across social and non-social conditions. We replicated previous findings showing that participants experience both egocentric and altercentric interference, but we found that these effects emerge equally in social and non-social conditions. To further test whether altercentric interference emerges in social perspective-taking situations, Study 2 conducted a conceptual replication of a study which used a novel "goggle" paradigm to assess whether individuals implicitly represent others' perspectives. Although we failed to find evidence of altercentric interference in response times, participants' accuracy reflected the possibility of interference from others' perspectives. We argue that these findings provide support for the idea that altercentric interference in response to social stimuli (an avatar) is driven by perspective-taking mechanisms, while such interference in response to non-social stimuli (an arrow) is driven by attention-cuing mechanisms.