Andrew Botterell

This article concerns the problem of factual uncertainty in negligence law. We argue that negligence law’s insistence that fair terms of interaction be maintained between individuals—a requirement that typically manifests itself in the need for the plaintiff to prove factual or “but-for” causation—sometimes allows for the imposition of liability in the absence of such proof. In particular, we argue that the but-for requirement can be abandoned in certain situations where multiple defendants have imposed the same unreasonable risk on a plaintiff, where the plaintiff suffers the very sort of harm that rendered the risk unreasonable, and where the plaintiff cannot prove which of the defendants was the but-for cause of her loss. This approach provides one way to understand the Supreme Court of Canada’s recent decision in Resurfice Corp. v. Hanke. We find support for our approach in various concepts that underlie negligence liability quite generally. These underlying concepts are normative in nature, and manifest core notions of justice and fairness. We argue that approaches to the problem of factual uncertainty that appeal to such normative principles to make sense of atypical cases of causation are in no way inconsistent with the nature and structure of negligence law. Rather, the opposite is true: in taking negligence law seriously as law, such approaches are instead reflective and supportive of it.

Imagine the following. You have been asked to critically evaluate the criminal process in your home jurisdiction. In particular, you have been asked to determine whether the criminal process currently in place appropriately balances the need to maximize the chances of getting things right—of acquitting the innocent and convicting the guilty—with the need to minimize the chances of getting things wrong—of acquitting the guilty and convicting the innocent. How would you proceed? What rules of evidence and procedure would you put in place? Would you exclude germane inculpatory evidence that has been obtained in violation of the accused’s constitutional rights? Would you permit spouses to testify against each other, or allow the jury to draw adverse inferences from an accused’s failure to testify on his or her behalf? These are the sorts of epistemological issues addressed by Larry Laudan in his superb Truth, Error, and Criminal Law. The purpose of legal epistemology is to identify legal rules in order to assess them rationally, and, if necessary, to modify or replace them. The more and more widely Truth, Error, and Criminal Law is read, the more likely it is that legal epistemology will attract the attention of lawyers, legal academics, and philosophers, attention that can only contribute in a positive way to rethinking criminal law.

According to a familiar and influential view, rights are not absolute. To the contrary, they can sometimes be permissibly interfered with. I find such a view of rights attractive. John Oberdiek thinks otherwise. In a recent paper in this journal, Oberdiek has argued that any account of rights that incorporates a distinction between infringing and violating a right is indefensible. My aim in this paper is to argue that Oberdiek's worries are misplaced. The paper proceeds as follows. After some terminological stage-setting I present a familiar puzzle about rights and compensation and argue that the proper response to the puzzle is to distinguish between various ways in which rights can be interfered with. I then turn to a discussion of the general theoretical picture on which this account of rights rests, and I present some reasons for thinking that Oberdiek's criticisms of that picture are not successful. My conclusion is that the distinction between infringing and violating a right is a plausible one, and that an account of rights that rests on it is not for that reason problematic.