with commentary by Paul Guyer
It is one of the most demanding ideas in the history of ethics, and one of the most misquoted. Immanuel Kant's categorical imperative is often reduced to a bumper sticker: do what you'd want everyone to do. The categorical imperative is Kant's single supreme principle of morality, laid down in the Groundwork of the Metaphysics of Morals in 1785: act only on a maxim you could at the same time will to become a universal law. It commands unconditionally, whatever you happen to want, and in its second formulation it forbids treating any person's humanity as a mere means. Paul Guyer, who has spent a lifetime translating and interpreting Kant, shows that the real idea is stranger, tighter, and more uncompromising than the slogan, and that Kant himself did not always make it work.
The Good Will and the One Supreme Principle
Kant starts with a single startling claim: only a good will is good without qualification. In the Groundwork, he writes that "nothing can possibly be conceived in the world, or even out of it, which can be called good without qualification, except a Good Will." Intelligence, courage, wealth, even happiness can all be turned to evil ends. Only the good will is good no matter what. And by good will, Guyer explains, Kant means something specific and unglamorous: "the intention to do the right thing." Not the outcome, which luck can spoil, but the intention. A rescuer who dies failing to save a drowning child still acted rightly; a murderer whose bad aim happens to kill someone genuinely wicked still acted wrongly. "The basis of moral evaluation," Guyer says, "is the will, or the intention."
From that foundation Kant derives his single supreme principle. In the Groundwork's Second Section he states it: "There is therefore but one categorical imperative, namely this: Act only on that maxim whereby thou canst at the same time will that it should become a universal law." Guyer unpacks the word categorical first. A categorical imperative, he explains, "is one that applies without restriction whether you like it or not." It differs from hypothetical imperatives, which are conditional on what you want: if you want to reach the museum, turn left on Madison; if you want to poison someone, use this. The moral law is not conditional. It commands whether or not it suits you, which is exactly why it arrives as an imperative, a "you must," rather than a suggestion.
A Maxim Is a Policy, Not a Loophole
The heart of the test is the word maxim, and this is where most popular versions of Kant go wrong. A maxim, Guyer stresses, is not a narrow description of a single act but a general policy. "When you make a choice, you're implementing a policy," he says. "Roughly speaking, that's what a maxim is." This matters because clever readers try to defeat Kant with hyper-specific maxims: surely "anybody with red hair named Hilly Flitcraft could rob banks on a northeast corner" could be universalized without collapsing the banking system. Guyer dismisses this as a misunderstanding. "These sort of highly customized maxims that could be universalized without contradiction, that's not what Kant means by a maxim." Kant's own worked example is not bank robbery but the false promise: making it your policy to escape debt by "promising to pay back loans that I have no intention of doing." The test is to imagine that policy made universal. In a world where everyone made lying promises, the very institution of promising would dissolve, and the maxim would destroy the thing it depends on. It contradicts itself.
Guyer locates the philosophical engine underneath all this in the principle of non-contradiction, which he argues Kant treats as the first principle not only of logic but of morality. When you treat someone immorally, Guyer explains, "you're treating them as if they did not have wills of their own, but you know perfectly well that they do. So there's a contradiction involved." Immorality, on this reading, is a kind of practical self-contradiction: acting on a policy you cannot coherently will everyone to share.
Treat Humanity as an End, Never Merely as a Means
This is also why Kant offers a second, famous formulation, which many readers find more humane than the first. In the Groundwork he writes: "So act as to treat humanity, whether in thine own person or in that of any other, in every case as an end withal, never as means only." Guyer draws out what Kant means by humanity here, and it is precise. Humanity, in Kant's late definition, "is simply the capacity to set oneself ends." Plug that in, and the formula becomes concrete: always treat the end-setting capacity, in yourself and everyone else, as something of value in itself, never as a mere tool. To use a person merely as a means is to override the very capacity that makes them a person.
Where Kant's Own Arguments Fall Short
What makes Guyer an unusually trustworthy guide is that he will not pretend Kant's arguments all succeed. He is candid that the first formulation's four illustrations do not all hold, and singles out the suicide case as the weakest. Kant argues that a man wearied of life cannot will self-destruction as a universal law, because a nature designed to preserve life cannot coherently also destroy it. Guyer has "always regarded this argument as very weak." The flaw, he explains, is the assumption that a single natural tendency cannot produce different outcomes in different circumstances. That is simply false, and so the argument does not carry the weight Kant puts on it. A lesser commentator might smooth this over. Guyer names it.
So what is the categorical imperative, really? It is not the golden rule and not a calculation of consequences. It is a test of the policies behind your actions: can you coherently will that everyone act as you propose to act, and does your policy treat every person's end-setting capacity as an end rather than a mere means? Behind both formulas lies a single demand, that morality be grounded in reason alone, binding on every rational being whether they like it or not. Guyer shows both the grandeur of that ambition and the places where Kant's own reasoning falls short of it. The imperative is harder than the slogan, and, in Guyer's honest reading, more human for having its weaknesses named.