This paper addresses the question “what connection is there between our answer to the question of what we are, and the question, what our actions are?” Suppose that actions are reflexive changes of agents. On that supposition, there would be a direct connection between the answers to those two questions. An action of mine will be a reflexive change of me, and what I am will fix the nature of those changes. I hold that supposition to be true and consider (...) reasons in favor of believing it. However, the paper is not primarily aimed at defense of that thesis. It rather concerned with exploring what consequences accepting it has for the competing notions of what we are, given what we ordinarily think actions are, and bringing to light a tension between thinking of actions as reflexive changes of agents in this way, and a kind of causal understanding of actions that is prevalent. What emerges is that we should shift where we start our theorizing: we cannot assume that action theory primarily involves the task of characterizing the relation between an agent and changes caused, rather than a characterization of a particular kind of relation between the agent and herself. (shrink)