Research suggests that public figures can play an influential role in forming public opinion; yet, little research has experimentally tested the efficacy of public figures on the cognitive formation of boundaries. Using an experiment embedded within a nationally representative survey, we examine how two science popularizers, Francis Collins and Richard Dawkins, influence perceptions regarding the boundaries between religion and science. We find that learning of Dawkins does not influence people’s perceptions of the religion–science relationship, while learning of Collins shifts respondents toward a collaborative view of religion and science. Findings suggest that figures with unexpected views might be more effective in changing conceptual boundaries.