Monkey Do, Monkey See? The Effect of Imitation Strategies on Visuospatial Perspective-Taking and Self-Reported Social Cognitive Skills.

Classical social cognitive conceptions suppose that the existence of common representations between agents constitutes the basis that represents the world from others' perspectives. Alternatively, recent contributions support that the ability to distinguish self- from other's representation would rather be at the origins of social inferences abilities. In the present study we compared the effects of two types of imitation training: mirror imitation (for which gesture could be represented in common referential) and anatomically congruent imitation (which requires not only a representation of the gesture of the model but also distinguishing between one's own and others' representations). We observed that a 4 min training of anatomically congruent imitation, but not of mirror imitation, improved performance on a visual perspective-taking test. This short training did not significantly impact self-reported measures of social cognitive skills. These results suggest that a unique transversal cognitive mechanism of co-representing and switching between self-related and other-related representations could be involved at both the motor and the mental-state levels. Opportunities for innovative social cognitive interventions at the motor level are discussed.