At several points in the book so far we have run into the problem of moral luck, and now it is time to face it squarely. The term itself was invented by Bernard Williams in a classic 1975 paper, “Moral Luck”, to which Thomas Nagel responded in a paper by the same name. (I will be citing from the revised Williams paper in his 1981 anthology Moral Luck (1981a) and from Nagel's revised version in his 1979 anthology Mortal Questions.)

Both papers begin with the widespread assumption among moral philosophers, lawyers and the general public that morality is – and should be – immune to luck. This perceived immunity has deep Christian and Platonic roots, but received its most systematic philosophical elaboration in Kant's moral theory. According to this “Kantian conception” (as Williams calls it), if the agent's intentions are good, or at least not bad, and she has taken all the care and precautions required by the situation (that is, she is not reckless or negligent), then she cannot be held morally responsible for any bad effects that arise from bad luck alone. Indeed, it works both ways: she cannot claim responsibility for positive outcomes arising from good luck. To keep the discussion tidy, I am going to stick to bad luck, as did Williams and Nagel.