Hamilton meets causal decision theory.

In this paper, I contrast two mathematically equivalent ways of modeling the evolution of altruism, namely the classical inclusive fitness approach and a more recent, "direct fitness" approach. Though both are usually considered by evolutionists as mere different ways of representing the same causal process (i.e. that of kin selection), I argue that this consensus is misleading, for there is a fundamental ambiguity concerning the causal interpretation of the DF approach. Drawing on an analogy between the structure of inclusive fitness theory and that of causal decision theory (Stalnaker, 1972), I show that only the inclusive fitness framework can provide us with a proper, and unambiguous causal partition of the relevant variables involved in the evolution of altruism.