This chapter defends a non-consequentialist approach to justifying harming innocent nonthreatening people. The discussion centers on variations of cases where either a trolley driver or a bystander may redirect a trolley so that it kills one person, rather than five people towards whom it is headed. The article begins by critically examining other non-consequentialist proposals for why turning the trolley may be justified in both types of cases. It then defends seven principles, which together comprise the “Doctrine of Productive Purity” (DPP). The DPP implies that interpersonal harms may be justified by what produces them as opposed to by the goods they produce as in consequentialist views. The article contrasts the DPP with threshold deontology and considers how those whom one might permissibly harm may still permissibly resist such harm. The chapter concludes by considering limits of the DPP, the moral significance of the DPP for relations between people, and a new trolley problem that has implications for the killing/letting die distinction.

Philippa Foot offered her original Trolley Case, where a driver will unavoidably kill either five or one, neither to present a moral quandary nor to pose any special problem for moral theory. Its purpose was simply to illustrate how her theory of duties – which she deployed to defend non-consequentialist intuitions about various other cases of harmful agency – could also handle cases like this one where consequentialist verdicts seem apt. The Trolley Problem arose when Judith Jarvis Thomson modified the case, using the bystander variation to raise difficulties for Foot. Intuitively, it seems that the bystander may equally turn the trolley, but that would amount to killing one to save five, which was ruled out by Foot’s theory (even though it allowed a driver to kill one to avoid killing five). That was the Trolley Problem, which amounted to a challenge to explain why the bystander trolley case is an exception to the usual prohibition against killing some to save others for the greater good. For decades Thomson saw it as a legitimate challenge and offered plausible ways of meeting it. Later, however, she changed her mind, concluding that it is not permissible for the bystander to turn the trolley after all. In this chapter, I begin by examining the relevant background, refining the understanding of the Trolley Problem, and considering basic methodological issues in this area of normative ethics. I then develop a solution to the Trolley Problem along with a diagnosis of what went wrong in Thomson’s later attempt to show that there was never really a Trolley Problem to begin with. In the end I hope to have shown why there was indeed an important theoretical challenge posed by the Trolley Problem and how it can be resolved in a principled way with minor revisions to Foot’s original theory.

Since the publication of Judith Thomson’s 1976 paper, solving the Trolley Problem has been a favorite preoccupation of utilitarians and deontologists. Why is it permissible to divert a runaway trolley, thereby killing one person to save five others, but impermissible to push a big man onto a track to save five others.To date, virtue ethicists have not shown any interest in the debate.An obvious reason for this lack of interest is that virtue ethicists reject the very idea that there are universal moral rules and principles, according to which actions can be evaluated as permissible or impermissible. It is possible to frame the Trolley Problem in terms of what a virtuous person would and wouldn’t do, but then a further problem emerges, namely, that trolley experiments are not good tests of character. They rule out many of the ways that virtuous people can distinguish themselves from the non-virtuous.I discuss some of these problems in the first part of the chapter. In the second part, and with a few reservations and qualifications in mind, I argue that a virtue ethicist can support our commonsense intuitions in two central cases – Bystander and Footbridge – while also offering a response to Thomson’s Loop challenge.