Humans are social beings, wired to connect with and also to get to know and understand each other. For this we are well equip, from early in life, with a generally applicable causal theory of agents and actions. This "theory of mind" helps us understand that mental states - intentions, goals, desires, beliefs, knowledge, and emotions - drive behavior, and that these states of mind are linked with, but importantly separate from, the states of the world outside the mind. Much evidence suggests that, despite differences in language, executive functioning, and socialization practices, the time course of theory of mind development is largely stable across cultures. But there are well-documented cultural differences beyond early childhood in a host of social-cognitive understandings which relate to our early theory of mind. Do we need a separate framework for accounting for these later developments? In this talk, I will argue that we do not. The argument rests on a fundamental property of causation and causal inference, which is that causal and counterfactual reasoning are intimately connected (Lewis, 1986; also Pearl, 2000; Woodward & Hitchcock, 2003). This connection leads to a set of theoretical claims about our earliest causal representations of human action (that they support counterfactual inference, even in infancy) and can help clarify where cultural information fits into these representations. For the remainder of the talk, I will focus on the methodological possibilities suggested by this theoretical claim: namely, that asking children about alternative possible actions reveals a broad explanatory framework, beyond belief-desire psychology, in which cultural knowledge plays an important role. This framework has potential for unifying several areas of research related to social cognition, self-awareness, and self-regulation.