In the last two decades, cognitive and evolutionary approaches have appeared as new and invigorating attempts to explain what religion is: how religious phenomena emerged, why they persist, and why we find recurring patterns across cultural and historical borders. When addressing such question from perspectives informed by evolutionary biology and cognitive science, a pertinent question arises: How do we reconcile these new theories, and more experimentally inclined approaches, with a more traditional historical and/or sociological study of religion? What can cognitive and evolutionary approaches teach a general science of religion? In this paper I argue that historiography must indeed take theoretical and explanatory models arising from cognitive and evolutionary approaches seriously, but that we need to conceptualize not only the relation between distinct explanatory levels, but also the constraints imposed by the scope of particular scholarly endeavors.