The geologic timescale is a relative timescale based on physical and biological sequences and discontinuities in the rock record thought to record events in the history of the earth and of life. In some sense going beyond relative dating, absolute dating of rocks, fossils, and evolutionary events (such as branching events on the tree of life) can be accomplished through the use of radiometric dating, chronological signals extractable from fossil growth patterns, astrochronology, and the “molecular clock.” Sometimes these different methods, which start from largely independent assumptions and evidentiary bases, converge in their temporal estimates, resulting in a consilience of inductions. At other times they fail to agree, either because fossils and molecules are giving temporal information about different aspects of nature and should not be expected to agree, or because of flawed assumptions that give rise to an inaccurate estimate. The dispute over which kinds of evidence are more reliable is sometimes called the “rocks vs. clocks debate.” I argue that, despite the fact that it can be difficult to integrate disparate kinds of evidence, the principle of total evidence should be applied to the dating of evolutionary events. One complication, however, is that various putatively independent lines of evidence for the dating of past events have had their independence compromised in the way that dating methods are calibrated, resulting in pseudorobustness. I propose a return to the robustness analysis of William Wimsatt which emphasizes learning from failures of robustness in the ongoing effort to improve method.