Rationality and self-interest are routinely attributed an explanatory priority as an inherently understandable basis - as an ideal of natural order - for all social scientific explanation. We argue that this is not consistent with a causal-mechanistic understanding of science and that using self-interest and rationality heuristically as a default baseline biases social scientific research. From a naturalist perspective, both rationality and self-interest are empirical objects of explanation. We discuss one such explanatory hypothesis, according to which consistent self-interested behavior is sustained by a social norm.