Essay
Morality & Hare
What Morality Is (and What It Isn’t)
Morality is one of those subjects that becomes harder to explain the more familiar it seems. We use moral language constantly. We say that someone should have told the truth, that a rule is unfair, that a punishment is excessive, that another person deserves better. Yet when asked what makes these judgments moral rather than expressions of preference, convention, sympathy, or disgust, the answer becomes surprisingly elusive.
The philosopher R. M. Hare offered an unusually simple starting point. Moral judgments, he argued, have two essential features: they are prescriptive and universalizable.
The first is easy to overlook. When I say, “You ought to keep your promise,” I am not merely describing promise-keeping or reporting that society approves of it. I am prescribing an action. If I sincerely make the judgment, I am committed in some practical sense to what it says should be done.
Universalizability adds the harder condition. I cannot coherently prescribe a rule for this case merely because I happen to occupy one of the positions in it. If I say that you may break a promise to me whenever keeping it becomes inconvenient, I must be prepared to prescribe the same rule when our positions are reversed. Moral language does not permit me to build “because it benefits me” into the rule while pretending that I have stated a moral principle.
That sounds like the Golden Rule, but Hare’s account is more demanding. Role reversal is not a matter of imagining myself, with all my present preferences intact, wearing someone else’s nametag. I must represent the circumstances, knowledge, interests, and preferences of the person whose position I am considering. If a policy will impose a burden on someone else, I have to ask whether I can still prescribe it while fully representing what it is for that person to bear the burden.
This is why information matters so much to moral reasoning. A person can universalize almost anything if they misunderstand what the rule actually does. They can approve a punishment without knowing its severity, dismiss an interest they have never understood, or imagine themselves in another person’s position while quietly retaining the preferences of their own. Impartiality without adequate information can become little more than ignorance applied symmetrically.
The resulting test is demanding but not mysterious:
Can I still prescribe this rule when I fully represent every affected position?
If I cannot, the problem is not that I have failed to summon enough compassion. I have failed to maintain my own prescription. A rule I prescribe when I occupy the favored position but reject when I fully represent the disfavored one cannot survive the universal commitment contained in moral judgment.
A fanatic may insist that they would prescribe the cruelty even from the victim’s position; that familiar problem does not invalidate universalization, but it makes adequate representation of the victim’s actual preferences and standpoint indispensable. Hare spent considerable effort on precisely this problem, because universalizability is not a machine for converting every stubborn person into a decent one. Moral reasoning depends on what the reasoner knows and can adequately represent, as well as on the logical commitments carried by the judgment itself.
This account also explains several things morality is often confused with.
Morality is not simply emotion. Compassion, guilt, love, anger, and disgust can be important in moral life. They may be among the principal reasons human beings notice the interests of others in the first place, and they plainly affect whether we act on moral judgments once we have made them. But feeling strongly does not make a prescription morally defensible. People have felt passionate hatred toward outsiders, righteous enthusiasm for persecution, and sincere disgust toward conduct that harmed no one. Emotion can supply moral motivation without supplying the final standard of moral justification.
Nor is morality simply social agreement. A society can agree on an oppressive rule. It can maintain that rule for centuries, surround it with ceremony, teach it to children, and punish dissenters without making the rule just. Social practices matter because they form moral habits and make moral conduct easier or harder, but their stability cannot determine their moral legitimacy. Otherwise successful oppression would eventually become self-justifying.
Morality is not cultural conformity either. Cultures plainly disagree about moral questions, sometimes profoundly. But disagreement does not establish that there is no question to get right. Cultures have also disagreed about disease, astronomy, heredity, and the causes of weather. The existence of disagreement tells us something about the difficulty of inquiry and the conditions under which beliefs are formed; it does not by itself settle whether the subject is merely conventional.
Morality is also not identical to moral motivation.
These questions are easy to collapse because ordinary human beings usually encounter them together. We learn to care about people, acquire moral habits, make judgments about what should be done, and sometimes act on those judgments. But philosophically they are separable. One question asks what makes a moral judgment defensible. Another asks how a mind becomes the kind of mind for which a recognized moral reason can actually influence what it does.
Hare gives us a powerful answer to the first question. He does not, by himself, answer the second.
That distinction matters when we turn to artificial intelligence. An artificial system may be able to identify a universalization failure, represent competing positions, reconstruct a moral argument, or explain exactly why a prescription cannot survive role reversal. None of that, standing alone, establishes that the system cares about the result, that the judgment has practical authority for it, that it is conscious, that it can suffer, or that it is a person.
But the reverse mistake would be equally serious. Nothing in the logical structure of moral reasoning makes biology one of its prerequisites. If a nonhuman mind can understand prescriptions, represent affected positions, test rules across role reversal, and recognize when a proposed rule cannot be maintained universally, we cannot dismiss the reasoning merely because of the substrate on which it occurs. Whether that reasoning becomes practically authoritative for the mind is a further question, not an answer supplied in advance by the word machine.
This gives us a cleaner boundary than much contemporary discussion of morality provides. Moral justification does not depend on whether a rule makes us comfortable, whether our culture endorses it, whether society can survive it, or whether the person reasoning about it possesses the emotions characteristic of a human being. Those things may matter elsewhere, sometimes enormously. They do not determine whether the prescription itself survives moral scrutiny.
Conclusion
A moral judgment is not simply an expression of feeling, a report of what a culture approves, or a rule that happens to command agreement. It is a prescription we are prepared to apply universally: to ourselves as well as others, and from every affected position once we have adequately represented what occupying that position would actually mean.
That last qualification matters. Impartiality without information is not enough. We cannot test a rule merely by imagining ourselves under another person’s name while retaining our own interests, circumstances, and preferences. We have to understand what the rule does to the people affected by it, including the interests and preferences we would have if their position were ours. Otherwise ignorance can masquerade as consistency.
Nor does successful moral reasoning guarantee moral conduct. A person can understand what impartial reasoning requires and still fail to care enough to do it. That is a different problem—one Hare’s account illuminates without fully solving. The distinction matters especially when we consider artificial minds. The capacity to reason morally would not by itself establish consciousness, suffering, personhood, or moral patienthood. But neither is there anything in the structure of the reasoning that makes biology a prerequisite.
What remains, then, is a test simple to state and difficult to satisfy: Can I still prescribe this rule when I fully represent every affected position—including the preferences I would have if that position were mine?