One way to understand tort law is as a functional response to the social problem of accidental personal injury. That puts the negligence action at the center, and emphasizes the doctrinal choice between negligence and strict liability, while downplaying the intentional torts and the torts that do not involve physical injury. It also foregrounds the policy choice between tort and other means of dealing with accidents. This functional treatment is not uncontroversial today, but it is certainly orthodox.1
Here I propose to bring back into view some neglected aspects of the intellectual origins of the accident-centered approach to tort law. When torts was emerging as an important doctrinal category in the common-law world during the late nineteenth century, the early commentator who did the most to organize it around the problem of accidental injury was the young Oliver Wendell Holmes, Jr. The influential slant he gave to the subject turns out to have resulted from his struggle with doubts, surprising and possibly instructive to us, about whether torts was a viable legal category at all.
Neither Holmes' doubts about torts nor the theory with which he resolved them had much to do with his views about proper social policy toward industrial accidents. He was mainly responding to the inner dynamics of a juristic debate about the taxonomic arrangement of the substantive law, a debate that had been triggered by the legislative abolition of the common-law forms of action. Jurists drawing on conceptual traditions inherited from Roman law favored adopting tort as a basic category, while those influenced by the analytical jurisprudence of Bentham and Austin pressed the other way. After first taking the Bentham-Austin side, Holmes discovered that centering tort law around the problem of accidents could justify its recognition as an important subject after all.
Coincidentally, the burst of personal injury litigation that accompanied the growth of railroads and factories in the late nineteenth century made Holmes' accident-centered formulation of tort law especially salient in practical terms, and his theory went on to gain the dominant position it holds today, at least in the United States. It thus turned out that in resolving an abstruse theoretical puzzle about the arrangement of the law in the way he did, Holmes was helping to construct an understanding of torts that is still dominant a century later, when its origins have largely been forgotten. As a final twist, we now have our own quite different doubts about torts, based more on concerns about accident policy than on views about conceptual arrangement--and the accident--centered conception that Holmes devised to justify the subject in the first place turns out to leave it especially vulnerable to these doubts.
I'll start by sketching a version of the accident-centered conception of torts as it exists today--the one I teach in my own introductory course on the subject. Then I'll show why torts was by no means certain to become a fundamental category, a subject taught in every law school, at the time our basic legal taxonomy emerged in the late nineteenth century. Next I'll trace the steps through which Holmes moved from his early rejection of torts as a category to justifying it as a body of law organized around negligence and accidental injury. Finally, I'll note the difficulties that the accident-- centered conception poses for the continued survival of torts as a primary division of our substantive law today.
I. WHAT IS A TORT?
Students come to law school with ideas about contracts, property, crime, and constitutions, but "tort" is a purely legal term corresponding to none of their ordinary notions or experiences. Telling them that it is a French word for "wrong" doesn't help much; a breach of contract is a civil wrong too, and they will learn that not all tort liability involves wrongdoing.
So I begin my introductory course in torts not with a definition, but with a sketch of its relation to the rest of the introductory curriculum, which I map using the distinctions between private and public, civil, and criminal, and substantive and procedural law. …