Adjudication is a conspicuous example of decision making under uncertainty, with “burdens of proof” employed to structure the decision. At the center of the following discourse is a set of ideas that collectively constitute the burdens of proof. These ideas inhabit the intersection of diverse disciplines, including law (its principal subject), epistemology, and decision theory. My goal is to elucidate this fascinating interplay of ideas. This work's focus is not heavily doctrinal or historical, although doctrine plays an important part, as does the history of ideas. It is, ultimately, an elaboration of a philosophy of factual adjudication.

In evidence law scholarship, much more attention has been paid to the rules of admissibility over the last two centuries, perhaps because rules regarding the “sufficiency” of evidence have languished: many that had once existed have disappeared, and the few that remain can be learned easily, at least at a superficial level. But among those who are concerned with the process and logic of proof, in the context of evidence that has passed the hurdles of admissibility, the rules that structure the decision take on great significance. And despite a recent surge of scholarly interest in the subject of legal proof, it remains decidedly undertheorized.

It will be widely agreed that adjudication of disputed facts is properly based on the weight of evidence. But “weight” can mean many things. For most lawyers, the weight of evidence is understood as the degree to which the evidence favors one side of the dispute over the other, what I will call its “discriminatory power.” In modern legal scholarship, however, there has been considerable interest in the weight of evidence in a sense developed, most prominently, by John Maynard Keynes – which refers to the total amount of relevant evidence considered, regardless of which side is favored thereby. The interest in this subject has been limited to a relatively small number of theorists, to judge by the impact that it has had on mainstream evidence scholarship. Perhaps this is because there remains significant disagreement about what lessons to take from the conversation. This is an unfortunate state of affairs, I believe, because these lessons are of considerable theoretical and practical import.