Consequences, Rules, and Character

A friend is hiding in your house. Someone comes to the door intending to kill them and asks whether they are inside. The case is Kant's, and he used it to defend an answer most readers find appalling: that you may not lie even here.

Whatever you think of his verdict, the case does what a good case does. Three families of theory give three different accounts of why the answer is the answer, and the accounts diverge before the verdicts do.

Look at what happens

The reasoning at the door is short. Telling the truth produces a death; lying produces an inconvenienced murderer. Lie.

The appeal of the family is that it explains why morality matters at all. Suffering is bad, and a theory that tracks suffering is tracking something that anybody can see is worth tracking. It is also impartial in a way that is hard to argue against: your friend's life counts, and it counts the same as a stranger's, because the badness of a death does not depend on whose it is.

What has to be added is an account of which outcomes are better, and the versions of consequentialism differ mainly in what they put there. Classical utilitarianism, from Bentham and Mill, puts happiness or pleasure. Other versions substitute preference satisfaction, or a list of goods, or wellbeing understood some other way. The intermediate tier takes those apart.

The standing pressure on the family comes from cases where the numbers point somewhere ugly. If the only thing that matters is the outcome, then framing an innocent person is required when it prevents a riot, and a doctor who could save five patients by harvesting one is not merely permitted but obliged. Consequentialists have replies, and the replies are the substance of the intermediate article. The point here is that the objections all have the same shape: the theory has no place for a consideration that resists being weighed.

Look at what you are doing

The reasoning at the door is different in kind. Lying is a way of using another person's rational agency against them, and doing that is not made acceptable by what follows. Kant held that the duty is unconditional, and that a person who lies becomes answerable for whatever the lie leads to, while one who tells the truth has done no wrong even if the worst happens.

The appeal here is that it captures something consequentialism keeps losing. There is a difference between what happens to a person and what you do to a person, and ordinary moral thought treats that difference as fundamental. Rights, promises, and the demand that people not be used as instruments are all badly served by a theory that converts them into quantities.

The standing pressure is the mirror image of the one above. A constraint that holds regardless of consequences will sometimes require an outcome nobody could defend, and the theory must either accept that or start admitting exceptions and explaining what governs them. There is also a structural puzzle, sometimes called the paradox of deontology: if lying is wrong because of what lying is, it is hard to see why you may not tell one lie to prevent five other lies. A theory that forbids you to minimise the very thing it treats as forbidden owes an account of what constraints are.

Look at who you are becoming

The third approach begins by rejecting the question. Asking for a rule that settles the case treats moral competence as the ability to apply a formula, and Aristotle's view was that it is nothing like that. It is a trained capacity to perceive what a situation calls for, developed by habituation, and possessed by people who have become a certain way rather than by people who have learned something.

At the door, the honest person is not consulting a principle about lying. Honesty is not the disposition to produce true sentences on demand; it is a concern with truth that a person of practical wisdom exercises with judgement, and every reading of Aristotle has the virtuous agent deceiving the murderer without difficulty.

The appeal is its realism about moral life. It has room for the fact that people are better or worse rather than merely acting rightly or wrongly, for moral education, for emotions as part of virtue rather than obstacles to it, and for the observation that knowing the rule and being decent are different accomplishments.

The standing pressure is that it seems not to answer the question that was asked. Told that the right action is what a virtuous person would do, the person at the door still does not know what to say. Defenders respond that this is a feature: the demand for a formula was the mistake, and pointing at exemplars is what moral guidance actually looks like. Critics respond that the account is then either empty or covertly borrowing its content from one of the other two.

Where the disagreement actually lies

The families are often introduced as three answers to one question, which understates the disagreement. They differ about what the question is.

Consequentialism asks which state of affairs to bring about, and treats action as the means. Deontology asks what this agent may do, and treats the relation between agent and act as basic. Virtue ethics asks what sort of person to be, and treats particular actions as expressions of that.

This is why the debate does not resolve by finding a case that settles it. A case that embarrasses one theory is usually a case its own framing tells it to handle differently, and each family has a story about why the others go wrong that its rivals do not accept.

Next

Kant's case is vivid and it is not very diagnostic, because almost everyone reaches the same verdict and disagrees about why. The next article takes a pair of cases designed so that the verdicts come apart, and asks what the gap between them shows.