Kant's Ethics: The Categorical Imperative, Duty, and the Autonomy of the Will

In a world where good actions produce disastrous results and base actions yield prosperity, what can still be called good without qualification? Immanuel Kant answered that question with a clarity that has unsettled moral philosophy for more than two centuries: only the good will deserves to be called an unconditioned good. Not talent, not courage, not even happiness — for all of these can be turned toward wicked ends. Only the will that acts out of pure respect for the moral law, independently of its consequences, possesses moral worth in itself. From this inaugural insight, Kant constructed, in the Groundwork of the Metaphysics of Morals (1785) and the Critique of Practical Reason (1788), the most ambitious and most contested ethical architecture of modernity.

1. The Good Will and Acting from Duty

The opening of the Groundwork contains one of the most quoted sentences in philosophy: “It is impossible to think of anything at all in the world, or indeed beyond it, that could be considered good without limitation except a good will.” What makes a will a good will is not what it achieves, nor what it produces — it is the principle according to which it acts.

Kant distinguishes three relations between action and duty:

1. Contrary to duty: the shopkeeper who cheats customers. No moral worth whatsoever.
2. In conformity with duty, but from inclination: the philanthropist who genuinely loves helping others. His action has moral propriety, but not moral worth in the strict sense, because inclination does the motivating work.
3. From duty: the person who, without any inclination toward helping others — oppressed by grief, cold to the suffering around him — nevertheless acts beneficently because he recognizes it as his duty. Here, and only here, does the action possess genuine moral worth.

This distinction is deliberately provocative. Kant is not saying that feelings like compassion and generosity are bad. He is saying that, from the moral point of view, what confers worth on an action is not the emotion accompanying it but the maxim — the subjective principle — on which the agent acts. Action that is both in conformity with duty and done from duty has moral worth; action in conformity with duty done merely from inclination may be praiseworthy, but its goodness is contingent: if the inclination changes, the action ceases.

2. Hypothetical Imperatives and the Categorical Imperative

Every practical law is expressed through imperatives, commands of reason addressed to beings who can act against it (for a perfectly rational being would follow the law spontaneously, without needing to be commanded). Kant distinguishes two fundamental types:

Hypothetical imperatives take the form “If you want X, you must do Y.” Their obligatory character is conditional: they bind only those who happen to have the relevant end. “If you want to pass the exam, study.” “If you want to lose weight, exercise.” These are rules of skill (for possible ends) or counsels of prudence (for happiness as a general aim), but they are not moral laws.

The categorical imperative, by contrast, commands unconditionally: “Act thus,” with no conditional clause. Its binding force does not depend on desires, inclinations, or outcomes. It holds for every rational being simply in virtue of its rationality. This unconditional character is precisely what makes it a moral law: a law that binds us only when convenient is not a moral law — it is a prudential strategy.

The general formula of the categorical imperative can be stated as: act only according to the maxim by which you can at the same time will that it should become a universal law. The test is not empirical (“what would happen if everyone did this?”) but rational: can I, without contradiction, will that the maxim governing my action be a law for all rational beings?

3. The Three Canonical Formulas

Kant presents the categorical imperative in several formulations that he takes to be equivalent in substance, though emphasizing different aspects of the same moral law.

3.1. The Formula of Universal Law

“Act only according to the maxim by which you can at the same time will that it should become a universal law.”

A variant, the Formula of the Law of Nature, reads: “Act as if the maxim of your action were to become through your will a universal law of nature.”

The universalizability test functions as a rational sieve: if the maxim guiding my action cannot be universalized without contradiction, the action is morally impermissible. The contradiction may be of two kinds: (a) a contradiction in conception — the practice universalized destroys the very practice that makes it possible; (b) a contradiction in the will — the maxim can be conceived as a universal law, but cannot be willed as such by a rational being.

3.2. The Formula of Humanity

“Act so that you treat humanity, whether in your own person or in that of another, always as an end and never merely as a means.”

This is arguably the most influential formulation in contemporary applied ethics. It encodes dignity (Würde) as the absolute value of every rational being: rational beings possess a worth that admits no equivalent and cannot be exchanged for any price. To instrumentalize a rational being — to treat it solely as a tool for ends not its own — is to violate its dignity.

The word “merely” is crucial: Kant does not forbid all use of persons as means (which would make any social relationship impossible), but forbids treating them only as means, disregarding that they are also ends in themselves. A hired worker is used as a means, but if her employer respects her rights, her autonomy, and her person, he does not treat her merely as a means.

3.3. The Formula of Autonomy and the Kingdom of Ends

The third formulation unites the previous two in an ideal: the autonomy of the will and the kingdom of ends (Reich der Zwecke), an ideal community of all rational beings in which each is simultaneously legislator and subject. In the kingdom of ends, each member acts according to maxims it could have legislated for all others; each treats the others as ends in themselves; and all share an equal and inalienable dignity.

This ideal is not utopian in a naive sense: it is the regulative standard by which the morality of concrete actions is assessed. The moral world Kant envisions is one in which the practical reason of each agent resonates with the practical reason of all others — a system of freedoms that mutually sustain one another.

4. The Four Examples and the Distinction Between Perfect and Imperfect Duties

To illustrate how the formulas work, Kant presents four examples in the Groundwork that have become canonical. They are organized along two axes: duties toward oneself or toward others; perfect duties (which admit no exception) or imperfect duties (which leave room for the agent’s discretion).

1. Suicide from despair (perfect duty toward oneself): The maxim “to shorten my suffering, I will end my life” fails the universal law test: a law of nature commanding the destruction of life whenever it caused suffering would contradict the very function of life as a force of self-preservation. The maxim cannot be universalized without a contradiction in conception.

2. False promise (perfect duty toward others): The maxim “when I need money, I will promise to repay it knowing I cannot” cannot be universalized: if everyone adopted it, the institution of promising would collapse, and the maxim would destroy its own presupposition. Contradiction in conception.

3. Cultivating one’s talents (imperfect duty toward oneself): The maxim “I will let my talents rust” can be conceived as a universal law without immediate contradiction, but a rational being cannot will it as universal — for she would simultaneously want to rely on her own talents whenever she needed them. Contradiction in the will.

4. Beneficence (imperfect duty toward others): The maxim “I will help no one in need” can be conceived as a universal law, but it cannot be willed as such by a rational being who might at any time need the help of others. Contradiction in the will.

The perfect/imperfect distinction has practical consequences: perfect duties (do not lie, do not kill) prohibit specific actions without exception; imperfect duties (cultivate talents, be benevolent) impose ends without specifying the exact action, leaving agents latitude of choice regarding the how and when.

5. Autonomy and Heteronomy

The concept of autonomy of the will (Autonomie des Willens) is the heart of Kantian ethics. An autonomous will is one that gives itself the moral law — whose determination comes from no source external to practical reason itself. By contrast, heteronomy occurs when the will is determined by something external: the desire for pleasure, the fear of punishment, the authority of a god or sovereign, social pressure.

For Kant, all previous ethics was heteronomous — including Aristotle’s eudaimonist ethics (whose end is happiness, empirically determined), theological ethics (whose foundation is divine will), and the moral sentimentalism of Hutcheson and Hume (whose criterion is a feeling of approval). None of these can provide a moral law that is necessary and universal, because all depend on contingent factors: happiness varies from person to person; God may or may not exist; feelings vary culturally and individually.

Only pure practical reason, legislating to itself, can ground a moral law that holds for all rational beings at all times. Autonomy is therefore the foundation of dignity: precisely because we are capable of giving ourselves our own law, we are ends in ourselves and not mere means.

The article dedicated to the concept of autonomy across the history of philosophy traces its trajectory from Greek thought to contemporary bioethics; here the focus is the internal architecture of the Kantian ethical system.

6. Transcendental Freedom and the “Fact of Reason”

Kantian ethics presupposes freedom: if the moral law orders me to act in a certain way, I must be capable of acting otherwise — I must be free. But transcendental freedom — the capacity to initiate a causal series from reason, independently of natural determinations — cannot be demonstrated theoretically. Nature, as studied by science, is a necessary chain of causes. How can freedom exist in such a world?

Kant answers by distinguishing two standpoints on the same being: as phenomenon (member of the natural world), the human being is subject to deterministic causality; as noumenon (member of the intelligible world), he is free. Freedom is postulated, not demonstrated.

In the Critique of Practical Reason, Kant goes further: the moral law is itself the only undeniable datum — the “fact of reason” (Faktum der Vernunft). We do not need to prove that we are free before knowing that we ought to act morally; it is the consciousness of the moral law that reveals, from within, that we are free. The moral law and freedom mutually imply each other.

7. The Highest Good and the Three Postulates of Practical Reason

Kantian ethics does not ignore happiness — but it subordinates it. The highest good (höchstes Gut) is neither pure happiness nor pure virtue, but the proportioning of virtue and happiness: the morally worthy agent ought also to be happy. This proportion is not guaranteed by the natural order, which is indifferent to morality. Practical reason therefore postulates conditions that make the highest good possible:

1. Freedom — the agent must be capable of acting morally; already linked to the moral law as its presupposition.
2. Immortality of the soul — perfect virtue (the holiness of the will) requires an infinite progress that presupposes the indefinite continuation of existence.
3. The existence of God — only a being endowed with infinite knowledge and power can guarantee, in the end, the proportion between virtue and happiness.

These postulates are neither articles of faith nor theoretical demonstrations: they are demands of practical reason, necessary conditions for the moral effort to make sense. Kant reintroduces through the servants’ entrance of practical reason what theoretical reason expelled through the front door: God and immortality.

8. Classical Objections and Responses

8.1. The “Empty Formalism” (Hegel)

Hegel argued, especially in the Philosophy of Right (1820), that the categorical imperative is an empty formalism: in testing whether a maxim can be universalized, we introduce no real moral content — we merely check the logical consistency of an already accepted practice. The test prohibits the false promise, but only by tacitly presupposing that the institution of promising ought to exist; if one held a maxim abolishing promises altogether, the categorical test could not condemn it as immoral without assuming what it sets out to prove.

The objection is powerful, but Kantians respond that the distinction between contradiction in conception and contradiction in the will provides sufficient normative content — and that the Formula of Humanity, which Hegel tends to neglect, anchors morality in the mutual recognition of dignity.

8.2. The Case of Benjamin Constant and the “Lie of Beneficence”

The Swiss-French liberal politician and philosopher Benjamin Constant (1767–1830) challenged Kant with a concrete case: if a murderer knocks at my door and asks where my friend is — intending to kill him — may I lie? Constant argued yes: the duty of truthfulness extends only to those who have a right to the truth, and a murderer has forfeited that right.

Kant replied in the pamphlet On a Supposed Right to Lie from Philanthropy (1797) with a position that shocked many readers: no, even in this case, lying is wrong. Kant’s argument is that truthfulness is the foundation of all communication and all right; whoever lies assumes responsibility for the unforeseeable consequences of the lie (and if, by chance, the friend has already left and the murderer finds him by another route, the liar has become co-responsible). Only sincerity is entirely within the agent’s control; consequences are not.

This position generated fierce controversy. Many contemporary Kantians — including Christine Korsgaard and Onora O’Neill — argue that Kant’s own framework permits a different conclusion: the murderer has violated the conditions of a rational social contract by using the question as a trap, and therefore lacks the standing of an honest interlocutor.

8.3. Schiller and the Tension Between Duty and Inclination

Friedrich Schiller, in his essay On Grace and Dignity (1793), expressed discomfort with Kantian ethics: the Kantian ideal seemed to render goodness of heart morally suspect — as though an action had worth only when inclination struggled against duty. Schiller proposed the ideal of the beautiful soul (schöne Seele): one in whom duty and inclination coincide, who does good joyfully and without tension. The critique was direct enough that Kant replied to it in a footnote added to the second edition of Religion within the Boundaries of Mere Reason.

Kant’s response was that he does not despise beauty of character — it is certainly admirable. But from a strictly moral standpoint, what counts is the principle of the action, not the emotional state accompanying it.

8.4. The Material Ethics of Values (Scheler)

Max Scheler, in Formalism in Ethics and Non-Formal Ethics of Values (1913–1916), argued that Kantian ethics is excessively formal: by excluding all empirical content from moral principles, Kant renders morality too abstract to orient concrete choices. Scheler proposed a material ethics that recognizes objective hierarchies of values (vital, sensible, spiritual, sacred), apprehensible through emotional intuition (Fühlen) irreducible to subjective inclination.

8.5. Williams and Foot: Kantian Coldness and the Problem of Reasons

Bernard Williams argued, in Ethics and the Limits of Philosophy (1985), that Kantian “morality” — centred on obligation, guilt, and impartiality — is a peculiar and in many respects impoverishing system. By requiring the agent to justify saving his own wife by invoking a universal moral maxim, Kantianism introduces “one thought too many”: the personal relationship should itself be sufficient reason. Radical impartiality threatens what Williams called ground projects — the commitments that constitute a person’s identity.

Philippa Foot, in “Morality as a System of Hypothetical Imperatives” (1972), questioned whether moral imperatives really possess the categorical force Kant attributes to them, or whether they ultimately depend on some form of the agent’s interest. Although Foot later revised part of this critique when developing the ethical naturalism of Natural Goodness (2001), her intervention showed that the unconditional normativity of Kantian morality is not self-evident and requires grounding.

9. Contemporary Reception: Kantian Constructivism

In recent decades, Kantian ethics has found a path of renewal in Kantian constructivism, associated with John Rawls, Onora O’Neill, and Christine Korsgaard.

John Rawls, whose theory of justice has deep Kantian roots, proposed a reading of Kantian ethics that dispenses with its more controversial metaphysical postulates (immortality, God) and retains the procedural core: moral principles are those that rational agents would choose under conditions of impartiality. The “veil of ignorance” of the original position is a Rawlsian translation of the standpoint of the categorical imperative.

Christine Korsgaard, in The Sources of Normativity (1996), reinterprets autonomy as self-constitution: by acting on principles she herself reflectively endorses, the will constitutes itself as a unified agent. The foundation of normativity is not an external moral law nor a brute fact about the world, but the reflective structure of practical self-consciousness.

Onora O’Neill, in Constructions of Reason (1989) and Towards Justice and Virtue (1996), argues that the categorical imperative, properly understood, is not empty: it requires that principles be adoptable by all affected agents — which has concrete implications for global justice, bioethics, and power relations.

10. Balance: The Persistence of a Question

Kantian ethics survives the avalanche of objections not because it is immune to them, but because it touches a question that cannot be evaded: what does it mean to act morally for reasons? Every critique of Kantianism presupposes that there is something morality requires — whether happiness (against formalism), personal integrity (Williams), or natural flourishing (Foot). But by articulating with precision what it means to act on principles any rational being could endorse, Kant gave ethics a language that still structures debate.

The concept of the a priori — central to Kantian epistemology — reappears here: the moral law is not learned from experience, but discovered (or rather, constituted) by pure practical reason. The distinction between consequences and intentions, between duty and inclination, between autonomy and heteronomy — these distinctions have not disappeared from the ethical vocabulary, even for those who reject them.

To discuss the trolley problem and the tension between consequentialism and deontology, to understand the debate between utilitarianism and duty-based ethics, or to situate Kant within the broader landscape of normative ethics, one must have this architecture clearly in view.

Essential Readings

• KANT, Immanuel. Groundwork of the Metaphysics of Morals. Translated and edited by Mary Gregor. Cambridge: Cambridge University Press, 1997. (Original: Grundlegung zur Metaphysik der Sitten, 1785.)
• KANT, Immanuel. Critique of Practical Reason. Translated and edited by Mary Gregor. Cambridge: Cambridge University Press, 1997. (Original: Kritik der praktischen Vernunft, 1788.)
• KANT, Immanuel. On a Supposed Right to Lie from Philanthropy (1797). In: Practical Philosophy. Translated and edited by Mary Gregor. Cambridge: Cambridge University Press, 1996.
• PATON, H. J. The Categorical Imperative: A Study in Kant’s Moral Philosophy. London: Hutchinson, 1947.
• KORSGAARD, Christine M. Creating the Kingdom of Ends. Cambridge: Cambridge University Press, 1996.
• O’NEILL, Onora. Constructions of Reason: Explorations of Kant’s Practical Philosophy. Cambridge: Cambridge University Press, 1989.
• RAWLS, John. Kantian Constructivism in Moral Theory. In: Journal of Philosophy, 77(9), 1980, pp. 515–572.
• WILLIAMS, Bernard. Ethics and the Limits of Philosophy. London: Fontana Press, 1985.
• WOOD, Allen W. Kant’s Ethical Thought. Cambridge: Cambridge University Press, 1999.

← Articles