Confronting Normative Uncertainty: Deciding Cases When You Don’t Know How to Decide

Abstract

The intellectually honest and humble judge faces a very serious problem about which little has been said. It is this: how to decide what she ought to do when, despite knowing all the relevant facts, laws, and theories of adjudication about how she ought to decide, she remains uncertain. Drawing from emerging debates in moral theory, I call this problem the problem of “normative uncertainty.” In this Article, I explain the problem, distinguish it from related debates in the legal literature, and show that the obvious solution—engage in some legal reasoning to figure it out—is a nonstarter. Judges ought not simply muddle through, following whichever theory of adjudication they find most plausible. Sometimes, like in Brown v. Board of Education or Google v. Oracle, they ought not follow their preferred theory but do what another theory suggests instead. Developing a full solution will be difficult. To show why, I sketch a solution and apply it to Brown I. I do not defend it. Instead, I use it to illustrate why developing a solution to normative uncertainty is considerably more difficult than developing solutions to other kinds of uncertainty. I then apply it to Google to show how, even without a solution, recognizing the problem changes how we evaluate judging. Specifically, we can consistently maintain both that a court erred legally, and that a court acted rationally given its uncertainty. Importantly, these two critiques are against different benchmarks—jurisprudence and practical rationality—and so going forward, we need to be clear which critique we are making.