Explanations in the natural historical sciences often take the form of stories. This paper examines two accounts of the sources of narrative’s explanatory power: Beatty’s suggestion that narrative explanation is closely connected to historical contingency, and that narratives explain by contrasting what happened with what might have happened; and Ereshefsky and Turner’s view that narratives explain by organizing events around a central subject with a distinctive direction of historical development. In both accounts, it turns out that non-epistemic values typically play (...) a role in the construction of narrative, and hence contribute to narrative’s explanatory force. Two case studies from historical science – the plate tectonic story of the Avalonian terrane, and the story of the evolution of trichromatic vision in (some) mammals – help to motivate and illustrate this argument. (shrink)