Deodands put a price on objects that caused death, a concept rooted in the Latin “deo dandum”—“that which must be given to God.” Under English common law before 1846, any moveable chattel—a cart wheel, a barrel, a ladder—that directly killed an adult became forfeit to the Crown. The coroner would assess the object’s value, theoretically for charitable use, but in practice the money often went to the bereaved family. A 1829 floor collapse at the Norfolk Arms Hotel illustrates the scale: a jury valued the splintered timbers at five shillings to divide among thirty victims’ families, a paltry sum even then. The law was messy and inconsistent—church bells were excluded as real property, while a whole cart with horses could be seized if a man fell on its pole—yet it provided an accessible, if imperfect, mechanism for compensation when no human perpetrator could be blamed.
The railway broke this logic. When a train struck someone, the moving object was undeniably responsible, but its value—an entire locomotive—was orders of magnitude greater than a barrel or cart. Juries responded by assessing deodands at sharply higher amounts during the Industrial Revolution, arguably as punishment for perceived industrial negligence or because companies could afford it. In some cases jurors explicitly suggested they would drop the deodand claim if the railway made provision for the family. The disparity between a local merchant paying shillings and a railway company facing the cost of a locomotive made the doctrine politically untenable.
Parliament abolished deodands in 1846, a move framed as progress. Yet as legal historian Harry Smith argues, the immediate effect was to protect railway companies: families of passengers, trespassers, and railway employees killed on the job lost even the minimal compensation deodands had provided. The repeal created a gap that modern liability law would eventually fill, but not before decades of industrial injuries went uncompensated. The episode reveals a deeper tension: a medieval law built around discrete, human-scale objects could not scale to complex, high-value systems where the “object” causing harm is a corporate asset rather than a personal chattel—a problem that resurfaces today with autonomous vehicles and AI systems where the moving, decision-making entity is neither a simple tool nor a human actor.