This talk will review recent work on the status of the geodesic principle in general relativity and the geometrized formulation of Newtonian gravitation (Newton-Cartan theory). These results suggest that, despite various claims to the contrary, the status of the geodesic principle is strikingly similar in these two theories. That said, in neither case is the situation as clean as one might have expected: in both theories, to prove the geodesic principle as a theorem, one needs to make assumptions that are not clearly more fundamental than the geodesic principle itself. I will argue that these complications do not represent explanatory flaws in either theory. Instead, the foundations of a physical theory are best conceived as a network of interrelated, mutually interdependent principles. On this view, the interesting project is not to identify the fundamental axioms or postulates of a theory, but rather to exhibit the details of the myriad interdependencies that obtain between its various parts.