Kantian Ethics - Duty and the categorical imperative

Good will and duty

Kant asks what makes a will morally good even when its action fails to produce the hoped-for result.

A good will chooses from respect for moral law. Intelligence, courage and wealth can help a harmful person; their value depends on how they are used. A good will has unconditional moral value because of its principle of choice. This is not the claim that a careless good intention excuses ignoring available means. Kant distinguishes a genuinely committed will using its resources from merely wishing for a good outcome.

Duty is the necessity of acting from respect for moral law. An action can be in accordance with duty without being from duty. A seller might charge a fair price solely to protect their reputation. The action conforms to duty, but the determining reason is self-interest. Another seller acts fairly because deceiving customers is wrong, even when cheating would remain secret. Kant locates moral worth in that respect for duty. The action does not need to be unpleasant: enjoying fairness is compatible with acting from duty if respect for law determines the choice.

This is a deontological approach because moral obligation is not determined by maximising desirable consequences. It has an absolutist character: moral law binds rational agents universally rather than varying with personal goals. Neither word means blind obedience to whatever law a government passes. Kantian moral authority is rational and autonomous; an unjust command can fail the moral test.

The distinction also protects agents from moral luck. Two equally careful rescuers may have different outcomes because one faces an unexpected obstacle. The bad outcome does not automatically show a bad will. Nevertheless, the judgement of the will differs from deciding whether an act fulfilled its duty: motive, action and result are related questions, not interchangeable measures.

Hypothetical and categorical commands

An imperative is a command of reason, but only a categorical imperative binds independently of an agent's desired outcome.

A hypothetical imperative has conditional force: if you want a particular end, take the necessary means. If you want to pass an examination, revise effectively. The instruction can be rational advice while the goal remains optional. It can even serve a wrongful goal: a fraudster can reason efficiently about how to deceive.

A categorical imperative commands unconditionally. It is not “be honest if you want a good reputation”, but a requirement on choice whatever the agent wants. Moral obligation cannot depend on whether the agent happens to desire its result, because an unwilling agent would then escape it simply by changing their preference.

This distinction is about the ground of obligation, not a ban on thinking about consequences at all. A duty to help people requires knowledge of which help would be effective. What consequences cannot do is become the sole foundation of the duty. Likewise, a sentence that lacks the word “if” may still express a hypothetical command: “Train daily” can mean only “train if you want to win”.

The categorical imperative has several formulations expressing the same underlying moral law from different perspectives. OCR requires the law of nature, end in itself and kingdom of ends. They examine universality, rational dignity and common lawgiving; they are not three optional rules from which we choose whichever permits our action.

The law of nature formulation

Universalisation tests whether the principle of an action could coherently govern every rational agent in relevantly similar circumstances.

A maxim is the agent's principle of action, including relevant circumstances, the proposed act and its purpose. It is more precise than a one-word action such as “lie”. Paraphrasing Kant's law of nature formulation: act only on a maxim you could will as a universal law of nature. Imagine the maxim operating reliably for everyone, not merely ask whether you would enjoy its consequences.

Take the maxim “When I need money, I will make a promise I intend to break in order to obtain a loan.” If this became universal, assurances would no longer create trust, so promising could not operate as the maxim requires. It defeats the practice it uses. This is a contradiction in conception, supporting a strict or perfect duty against false promising. The objection is not merely that many lies would make people unhappy.

Kant also considers maxims that can be conceived universally but cannot rationally be willed universally. “I will never help anyone in need” does not destroy the concept of help. Yet a finite rational agent has ends for which others' assistance can be necessary. Willing universal refusal conflicts with willing the support necessary for those ends. This contradiction in will supports an imperfect duty of beneficence: an obligation to adopt helping others as an end, with latitude concerning particular opportunities.

Maxim description matters. “I may lie whenever it benefits me” and “I will refuse to provide private information to someone threatening violence” are different principles. An agent must avoid tailoring a maxim with arbitrary personal exceptions. Relevant context can belong in it; “only when I am the person doing it” cannot make a selfish exception universal.

Ends in themselves and common lawgiving

Kant treats rational persons as having dignity, so they cannot be valued merely as instruments for someone else's plan.

The end in itself formulation requires treating humanity, in oneself and others, always also as an end and never merely as a means. Ordinary cooperation uses people as means in a permissible sense: a paid driver helps a passenger reach a destination. The moral problem is merely using them, as when deception or coercion prevents their rational participation. Consent is important evidence, though obtaining superficial consent does not automatically justify exploitation.

False promising fails here too. The lender cannot choose whether to lend on the real facts because the borrower conceals their intention. The borrower recruits the lender's agency while denying it an informed role. Respect for persons also has a positive side: their legitimate ends matter, giving a reason to support others' capacities and welfare rather than simply refrain from assault.

The kingdom of ends formulation asks us to act as rational members and legislators of a possible community governed by common moral laws. Each person is both subject to those laws and able rationally to endorse them. It is an ideal of reciprocal respect, not a prediction of a perfect state or permission for each individual to invent a private law.

Imagine an employer concealing dangerous working conditions to save money. Universalisation asks whether a deception-based principle can be common law. The end-in-itself formulation asks whether workers can make an informed choice and whether their dignity is respected. The kingdom of ends asks whether the employer can defend the principle to other rational members as an equal lawgiver. These perspectives explain why profitability alone cannot settle the action's moral status.

The theory offers strong protection against convenient exceptions. Its difficulty is that identifying relevant maxims and duties can require substantial judgement. Knowing the formulations supplies a method of scrutiny, not an automatic device giving an uncontested answer to every practical problem.