David Estlund holds that ultimate normative principles are insensitive to bad facts. This is a deliberately twisted appropriation of Jerry Cohen’s famous dictum that ultimate normative principles are fact-insensitive. In this paper, I will show why Estlund’s twist misses the point of Cohen’s argument. The fact-insensitivity claim is not a requirement to eliminate all facts from our normative theories because facts necessarily make these theories concessive. Instead, it may help us to locate the true origin of these concessions. In normative (...) theorizing, we have not explained why a fact supports a principle if we fail to articulate the higher-order principle in light of which that fact is normatively significant. (shrink)