Abstract

This paper presents explication on how paleontologists reconstruct the past using fossils when _good_ modern analogues are not available. I call these _pastist_ methods to differentiate them from presentist methods in which such analogues are available. I do so by presenting two fossil cases: the problematica and graphoglyptids. I describe a forgotten heuristic, “analogue chaining,” that involves jumping from fossil anomaly to fossil anomaly using one to make sense of the other in successive fashion, using the relations _between fossils_ to guide reconstruction. I relate this to the philosophy of historical sciences in four ways. First, that methods like analogue chaining have a “linearity” meaning that there are limited ways in which to learn about specimens using analogues. Second, that they are intrinsically difficult to notice, i.e. invisible. Third, that linearity and invisibility put pressure on some accounts of optimism about historical sciences. Fourth, our cases provide novel forms of optimism based on epistemic enhancement: the phenomena that some questions regarding an event are _better_ answered millions of years after its occurrence.