The Trouble with Tracing

In this paper I argue that the widely relied upon idea of “tracing ” in the theoryof moral responsibility is considerably more problematic than has been previ-ously acknowledged. The difficulty I raise stems from requirements imposed by the knowledge condition on moral responsibility. Suppose you believed that being a responsible agent (at least paradigmati-cally) involves being suitably sensitive to situation-relevant moral concerns. On this view, agents that are not suitably sensitive to the relevant moral concerns (these may include non-human animals, young children, and presumably some adults with particular conditions) are not appropriate candidates for ascriptions of moral responsibility. Though this sort of theory may have the sheen of plausi-bility to it, it is not difficult to generate problem cases for the theory as stated. For instance, suppose that while he is intoxicated Luis decides to drive home while his intoxication has rendered him insensitive to the relevant moral concerns. When he runs his Buick over a young mother and two kids on the way home from the bar, we do not thereby conclude that because he was in a state of intoxication suffi-cient to render him insensitive to the relevant moral concerns he was not morally responsible.1 The standard response to problem cases of this sort is to introduce a notion of tracing. Tracing is the idea that responsibility for some outcome need not be anchored in the agent or agent’s action at the moment immediately prior to outcome, but rather at some suitable time prior to the moment of deliberation or action. So one thing that we might say is that Luis ’ responsibility for running over