We argue that theory-of-mind (TOM) approaches, such as “theory theory” and “simulation theory”, are both problematic and not needed. They account for neither our primary and pervasive way of engaging with others nor the true basis of our folk psychological understanding, even when narrowly construed. Developmental evidence shows that young infants are capable of grasping the purposeful intentions of others through the perception of bodily movements, gestures, facial expressions, etc. Trevarthen’s notion of primary intersubjectivity can provide a theoretical framework for understanding these capabilities. His notion of secondary intersubjectivity shows the importance of pragmatic contexts for infants starting around one year of age. The recent neuroscience of resonance systems (mirror neurons, shared representations) also supports this view. These ideas are worked out in the context of an embodied “Interaction Theory” of social cognition. Still, for more sophisticated intersubjective interactions in older children and adults, one might argue that some form of TOM is required. This thought is defused by appeal to narrative competency and the Narrative Practice Hypothesis (or NPH). We propose that repeated encounters with narratives of a distinctive kind is the normal route through which children acquire an understanding of the forms and norms that enable them to make sense of actions in terms of reasons. A potential objection to this hypothesis is that it presupposes TOM abilities. Interaction Theory is deployed once again to answer this by providing an alternative approach to understanding basic narrative competency and its development.