'Spontaneous' visual perspective-taking mediated by attention orienting that is voluntary and not reflexive.

Experiments revealing 'spontaneous' visual perspective-taking are conventionally interpreted as demonstrating that adults have the capacity to track simple mental states in a fast and efficient manner ('implicit mentalising'). A rival account suggests that these experiments can be explained by the general purpose mechanisms responsible for reflexive attentional orienting. Here, we report two experiments designed to distinguish between these competing accounts. In Experiment 1, we assessed whether reflexive attention orienting was sufficient to yield findings interpreted as spontaneous perspective-taking in the 'avatar task' when the protocol was adapted so that participants were unaware that they were taking part in a perspective-taking experiment. Results revealed no evidence for perspective-taking. In Experiment 2, we employed a Posner paradigm to investigate the attentional orienting properties of the avatar stimuli. This revealed cue-validity effects only for longer stimulus onset asynchronies, which indicates a voluntary rather than reflexive shift in spatial attention. Taken together, these findings suggest that attentional orienting does indeed contribute to performance in the Samson et al. avatar task. However, attention orienting appears to be voluntary rather than reflexive, indicating that the perspective-taking phenomenon measured may be less spontaneous than first reported.