Anti-realists like Paul Roth conceive of historical narratives as having no genuine explanatory power, because historical events are not ready-made and reveal themselves only to the retrospective gaze of the historian. For that reason, the categories with the help of which historians identify historical events do not map onto categories of general theories of the world required for a genuine explanation of them. While I agree with Paul Roth that the significance of a historical event is revealed only retrospectively, I argue that this does not imply that historical narratives do not provide genuine explanations. In this context, it is however important to distinguish between the description used by historians to identify the event as being part of the narrative and the description under which the occurrence of the very same event could be causally explained within the narrative. Both types of descriptions possess a certain degree of conceptual independence from each other. I argue that historical narratives incorporate both dimensions: what I also call the view from and with the view from below. Historical narratives do explain, even though they differ from scientific theories.