Philosophers and scientists often engage in historical reconstruction, outlining sweeping narratives on topics such as the evolution of cognition, language, morality, and so on. These narratives are closely related to what is known as "how possibly explanation" but little work has been done connecting the literature on these two concepts of historical reconstruction. In this paper, I synthesize these concepts and give my own account of narrative explanation. I then move on to show how work on how possibly explanation gives us a framework for understanding the multiple ways in which sweeping narratives are anchored. Patrick Forber's notions of global and local constraints are a good start, but some streams of evidence, such as experiments in highly artificial laboratory settings, do not fit neatly into either category. I attempt to develop Forber's framework to better accommodate such evidence streams.