Groundwork of the Metaphysics of Morals

Overview

Ninety pages, three sections and a preface, and Kant is explicit that it is groundwork and not a building. He is not trying to tell you what to do. He is trying to isolate the principle that ordinary moral judgement already uses, state it exactly, and then ask what has to be true about us for it to apply. The ethical system itself came later, in the Metaphysics of Morals of 1797.

The method is worth naming because it explains the odd shape. Sections 1 and 2 are analytic: they start from something granted, common moral opinion in the first case and the idea of a rational agent in the second, and work back to the principle that must underlie it. Section 3 is synthetic: it starts from the principle and tries to show it genuinely binds us. Kant says openly that the third section is the hard one, and it is where most readers, and most philosophers since, get off the bus.

The prose is terrible in a specific way. Kant piles subordinate clauses and then introduces a distinction in a parenthesis. The sentences are not confused, only overloaded, and they unpack cleanly if you find the main verb first.

The argument

Section 1 begins with common moral knowledge. Talents, temperament and good fortune can all be used badly, so none of them is unconditionally good. Only a good will is, and its goodness does not depend on what it achieves. Kant then separates acting in accordance with duty from acting from duty, using the shopkeeper who is honest because it pays and the person who preserves his life because he enjoys it. To see duty clearly you need a case where inclination pulls the other way: the man who has lost all taste for living and preserves his life anyway.

If moral worth cannot lie in the purpose of an action, it must lie in the principle on which the agent acts. And if you strip a principle of every particular purpose, what remains is only lawfulness as such. That yields the first version of the imperative, and Kant claims this is what ordinary people use already when they ask whether they would be happy for everyone to behave this way.

Section 2 argues that you cannot get morality from examples or from human nature, because no observed case is ever certainly moral and a principle drawn from human psychology would bind only humans. So the law has to be derived from the concept of a rational will. Imperatives divide into hypothetical, which tell you what to do if you want something, and categorical, which command absolutely. Only one imperative can be categorical, and Kant gives it in several formulas he treats as equivalent: act on universalisable maxims; act as if your maxim were to become a law of nature; treat humanity always as an end; act as a legislating member of a kingdom of ends. Four examples run through the formulas: suicide from despair, borrowing on a promise you cannot keep, letting your talents rot, and refusing all help to others. The first two, Kant says, cannot even be thought as universal laws. The second two could be, but nobody could rationally will them.

The section’s key concept is autonomy. A will bound by something outside itself, by God, by happiness, by feeling, is heteronomous and therefore not free, and Kant thinks every previous moral philosophy made this mistake.

Section 3 asks what entitles us to think the law applies. A free will and a will under moral law turn out to be the same idea, which threatens a circle: we call ourselves free because we are under the law, and under the law because we are free. Kant breaks it by distinguishing two standpoints. As appearances we belong to the world of sense and are determined; as intelligences we belong to a world of understanding and are free. He ends by conceding that he cannot explain how a categorical imperative is possible, only why no explanation is available, and he calls that the comprehension of the law’s incomprehensibility.

How to read it

Read the preface, then Section 1 slowly, then Section 2 twice, then Section 3 once without panicking. A first reading that ends at the kingdom of ends has still got most of the value.

Do the four examples yourself before reading Kant’s treatment of them. Write out the maxim in the form he uses, an act plus a purpose plus a circumstance, and see where the contradiction falls. Most objections to the universal law formula come from stating the maxim carelessly.

Note that Kant’s target is not people who do wrong. It is people who do right for the wrong reason, and philosophers who ground morality in anything empirical. Hume and the British sentimentalists are offstage throughout.

The common misreading is that Kant condemns feeling and recommends coldness. He says the opposite elsewhere, that cultivating sympathy is itself a duty. His point in Section 1 is diagnostic: he needs a case with no sympathy in it to see what duty looks like on its own.

Questions it raises

• Can any maxim be formulated precisely enough that the universal law test gives a determinate answer?
• If moral worth requires acting from duty alone, is a person who enjoys being kind worse off morally than one who does not?
• Does the formula of humanity actually follow from the formula of universal law, or are they separate principles Kant asserts to be one?
• What is left of the self that legislates, once you have removed everything empirical about it?
• Is the two-standpoints solution a real answer to the circle, or a restatement of the problem in new vocabulary?
• Kant says we can never know whether any action was truly done from duty, including our own. What does that do to moral self-knowledge?