We highlight two aspects of research into social learning that have been neglected in existing developmental research, namely:(1) The role of social learning in learning to use tools, and (2) Whether children's social learning involves copying the actions themselves ('blind' imitation or mimicry), or alternatively, the effects of those actions (emulation). In Part I of the paper we argue that the failure to distinguish between these different mechanisms is closely related to the lack of research on the social transmission of tool use and that both omissions limit our understanding of early social learning. We conclude Part I by outlining the requirements for an adequate study of these two issues. In Part II, we use this analysis to critically examine data from existing developmental research with children. We also assess the data currently available in the comparative literature which address these issues more directly. We conclude that children learn only what actions to perform via observation ('blind' imitation or mimicry), and not why those actions are effective (emulation). We close by identifying important potential pitfalls and unresolved questions for the future study of the social learning of actions on objects.