Anscombe’s Fire, Hare’s Test: What Makes an Absolute Absolute?

Abstract: This essay is part of Real Morality’s larger project of recovering R. M. Hare’s universal prescriptivism from the charge that it reduces morality to calculation. Elizabeth Anscombe presents the strongest version of that charge: she argued that modern moral philosophy had lost the ability to say that some acts — murder, torture, rape, and condemning the innocent — must never be done. Anscombe was right to fear moral theories that can justify anything if the consequences look favorable. But absolute prohibitions cannot be made absolute merely by declaring them so. They require a criterion. Hare’s answer is universalization: a prohibition earns its authority when the contrary permission cannot be prescribed from all affected positions. Anscombe gives moral philosophy its fire; Hare gives that fire its test.

---

Elizabeth Anscombe was not a mild critic of modern moral philosophy. She did not think the reigning theories needed adjustment, balance, or a more humane vocabulary. She thought the enterprise had become corrupt at its root.

In “Modern Moral Philosophy,” Anscombe argued that philosophers should stop doing moral philosophy until they had first developed an adequate philosophy of psychology: an account of action, intention, wanting, pleasure, motive, and practical reason. She also argued that the modern language of “moral obligation,” “moral duty,” “moral right,” and “moral wrong” should be abandoned, because it had become a survival from a divine-law conception of ethics whose framework no longer generally survived. In Jennifer Frey’s reconstruction, Revisiting Modern Moral Philosophy, Anscombe was not asking philosophers to rearrange the furniture of moral theory more elegantly. She was asking them to “burn down the house.”  

That is why Anscombe deserves to be treated differently from many of Hare’s later critics. She was not trying to weaken moral philosophy in the name of nuance, sentiment, identity, or contextual discretion. She was trying to preserve something severe: the possibility of saying that some acts must never be done.

For Anscombe, the central scandal of modern moral philosophy was its inability to preserve absolute prohibitions. Murder, rape, torture, and the judicial condemnation of the innocent are not actions to be weighed, balanced, and perhaps justified if enough good consequences can be produced. They are unjust in themselves. They must not be chosen as means to any end. Frey argues that this concern unifies Anscombe’s three theses: modern moral philosophy had become unable to articulate why intrinsically unjust acts must never be done.  

That is a serious charge. It is also a serious moral insight.

A theory that leaves the murder of the innocent permanently open for recalculation has already lost something essential. If the right mixture of predicted benefits can justify condemning an innocent person, torturing a prisoner, murdering a child, or betraying a friend, then the theory has not merely become demanding. It has become dangerous. It has taught its adherents to redescribe corruption as rigor.

Anscombe saw that danger clearly. Her fire was real.

But fire is not enough.

The question she does not adequately answer is the one Hare can answer:

What entitles a prohibition to be absolute?

It cannot be enough simply to call an act intrinsically unjust. That may be true. But unless we can say why the prohibition has that authority, “absolute” becomes a moral intensifier attached to a prior conviction. It tells us how strongly the rule is held. It does not yet explain why the rule binds.

This is easy to see if we choose a ridiculous example.

Suppose someone declares an absolute prohibition on owning poodles. Or, more gravely but no more plausibly, an absolute prohibition on shearing poodles into ornate shapes. The prohibition is stated in absolutist form. No one may do it. No consequences can justify it. No exceptions are permitted.

But the prohibition fails immediately.

Why?

Not because we dislike absolutes. We accept absolutes all the time in ordinary moral life. Do not murder. Do not rape. Do not torture children. Do not frame the innocent. Do not betray those whose lives have been entrusted to you. These claims have force. They are not merely preferences. They do not feel like optional policies.

The poodle prohibition fails because it lacks moral intelligibility. It does not connect to justice, agency, suffering, trust, responsibility, fairness, human flourishing, animal welfare, or any other structure capable of justifying such a demand. It cannot survive being stated as a rule for everyone. It cannot explain why anyone should be bound by it.

The word “absolute” does no work by itself.

This is the problem with Anscombe’s absolutism. She is right to insist that some acts must be excluded from practical deliberation. But she does not escape the need to justify which acts they are. Her favored examples are compelling because they attach to patterns of human action that are already deeply recognizable as unjust. Murder, rape, torture, and judicial condemnation of the innocent are not arbitrary entries on a list. They involve the destruction, violation, domination, or instrumental use of persons in ways that no one could accept as a general rule from all positions.

That is the point. Their authority does not come from being called absolute. They can be called absolute because their contrary permissions cannot be universalized.

Here Hare’s universal prescriptivism supplies what Anscombe needs but refuses to use.

A moral judgment is not merely an expression of feeling. It is not a statement of taste. It is not a tribal signal. It is a prescription. And to be moral, a prescription must be universalizable. If I say that an action ought to be done, I must be willing to prescribe the same action in relevantly similar circumstances, regardless of whether I am the agent, the victim, the beneficiary, the official, the stranger, the parent, the child, the powerful, or the exposed.

That test is not a decorative procedure. It is the grammar of moral seriousness.

If I say it is permissible to condemn an innocent person whenever doing so prevents disorder, I must be prepared to accept that rule when I am the innocent person condemned. Not merely say the words, “Yes, I accept it,” while remaining imaginatively untouched. I must understand what the rule authorizes from inside the position it sacrifices. I must accept a world in which innocence gives me no protection once others calculate that my destruction is useful.

No one can rationally universalize that rule while continuing to rely on the moral meaning of justice, office, evidence, trial, innocence, or law. The permission destroys the institution it claims to serve.

This is why the prohibition against condemning the innocent is not arbitrary. It is not an intuition floating beyond reason. It is a rule without which the practice of judgment collapses.

The same is true of murder. If I permit myself to intentionally kill the innocent whenever doing so produces enough benefit, I must accept a world in which my own innocent life may be taken whenever others judge the balance favorable. The rule does not merely threaten me. It destroys the meaning of innocent protection. It converts every person into available material for calculation.

That is not a difficult moral theorem. It is a basic fact about what the rule authorizes.

Anscombe sees the horror of the authorization. Hare explains why it cannot be universalized.

There remains, however, a deeper Anscombean challenge. Anscombe’s complaint about the modern moral “ought” was not merely historical. It was semantic. She thought modern philosophers had retained the language of moral obligation after abandoning the divine-law framework that once gave such language sense. Without a lawgiver, the language of being morally bound looked to her like displaced theology: a psychological residue, still forceful, but no longer intelligible. Frey emphasizes this point: for Anscombe, modern uses of “moral” retained the suggestion of force while no longer signifying a real concept.  

Hare’s answer is that Anscombe mistook one historical source of obligation-language for its only possible logic.

The authority of the moral “ought” need not come from a divine commander. It can come from the structure of prescription itself. To say that someone ought to act is not merely to express pressure, feeling, or preference. It is to prescribe conduct. To prescribe morally is to accept that the prescription applies across relevantly similar cases, including cases in which one’s own position changes.

The binding force is not theological. It is structural.

This also clarifies the relation between Anscombe’s absolutism and Hare’s two-level theory. Anscombe thinks some prohibitions must be absolute in principle. Hare may allow, at the critical level, that moral reasoning asks what one would prescribe under full information, full impartiality, and full imaginative identification with every affected position. That makes Hare look, to Anscombe, as if he has left the door open to monstrous exceptions.

But that is too quick.

Hare’s two-level theory distinguishes between critical moral thinking and intuitive moral practice. Critical thinking is idealized. It asks what would be prescribed under conditions human beings almost never actually satisfy. Intuitive moral practice is how finite, biased, frightened, self-interested, socially pressured creatures must live most of the time. At that level, we need stable rules, trained dispositions, and prohibitions strong enough to survive panic, ambition, loyalty, fear, and rationalization.

For actual human beings, the practical difference between Anscombe and Hare may be smaller than it first appears.

Anscombe says: never entertain murder, torture, or condemning the innocent as options.

Hare can say: human beings should treat these prohibitions as effectively absolute because they are almost never competent to calculate past them. The moment a person under pressure says, “Perhaps this innocent person must be sacrificed for the greater good,” we have already entered the danger zone Hare’s two-level theory was designed to explain. The agent is not an archangel. The agent is a frightened, partial, self-protective creature with limited information and a spectacular talent for redescribing advantage as necessity.

Anscombe may reply that “effectively absolute for creatures like us” is not absolute enough. But that objection assumes that only a metaphysical prohibition can block monstrous calculation. Hare’s point is different: a prohibition can be rationally grounded and still be unavailable for ordinary human recalculation. The imagined Archangel is not a moral permission slip for frightened, partial, self-excusing agents.

So Hare does not need to mock absolute prohibitions. He can explain them.

They are not mysterious primitives. They are indispensable intuitive rules for beings like us. They preserve the moral architecture that calculation would otherwise corrode.

This is a better answer than Anscombe’s because it explains both why the prohibitions matter and why they must not be casually reopened. It also explains why ridiculous absolute prohibitions fail. “Never murder the innocent” survives universalization under the full description of what the contrary permission would mean. “Never own poodles” does not. The difference is not that one has been declared absolute more solemnly. The difference is that one is structurally necessary to moral life and the other is not.

Anscombe may object that this still makes the prohibition depend on something outside itself. It does. But that is not a defect. It is an advantage. If a prohibition cannot explain its authority, it should not govern us absolutely.

The deeper value in Anscombe lies elsewhere. She insists that moral theory cannot begin with calculation alone because calculation depends on description. Before we ask whether an act can be justified, we must ask what act is being done. Is it killing, murder, allowing death, self-defense, execution, sacrifice, betrayal, refusal, negligence, or rescue? Under what description is the agent acting? What is intended, and what is merely foreseen? What is the act, and what is the further purpose for which it is done?

Frey emphasizes this point in her reconstruction. Anscombe’s project requires a philosophy of action and intention because without one we cannot distinguish the act itself from the motive, the intended means from the foreseen side effect, or the action-description that makes the deed what it is.  

That is a genuine contribution. It should not be dismissed.

Indeed, it may be exactly what a Harean theory needs in order not to become crude. Universalization cannot operate on a falsely described act. If the agent describes judicial murder as “restoring public order,” torture as “information gathering,” betrayal as “strategic adjustment,” or coerced obedience as “alignment,” the universalization test has already been corrupted before it begins.

So Anscombe’s question must come first:

What are you doing, under what description?

Then Hare’s question follows:

Can you universalize the prescription that permits it?

Together, they form a powerful moral discipline.

Moral evasion begins in two places: by misdescribing the act, or by exempting oneself from the rule. Anscombe is strongest on the first. Hare is strongest on the second.

This is why Anscombe should not be treated as merely another anti-Hare critic. She is not offering the familiar academic escape from moral structure into “complexity,” “integrity,” “thick concepts,” or “the particulars.” She is trying to save moral refusal from the machinery of justification. She fears, rightly, that human beings are far too eager to call an injustice necessary when something they value is at stake.

But her solution is incomplete. Absolute prohibition is not self-justifying. It must still answer the Harean question: can the rule be prescribed universally? If not, it is not morality; it is command, taboo, inherited conviction, or metaphysical assertion.

The great irony is that Hare may protect Anscombe’s fire better than Anscombe does.

Anscombe protects absolutes by trying to remove them from calculation. Hare protects them by explaining why finite human beings are not entitled to calculate past them. Anscombe says the unjust act must never be done. Hare asks what rule would authorize it, who would have to live under that rule, and whether the rule could be accepted by anyone who fully understood what it licensed.

In the strongest cases, the result is the same.

Do not murder the innocent. Do not torture. Do not condemn the innocent. Do not rape. Do not betray those entrusted to your care.

But in Hare’s account, these prohibitions do not depend on a mysterious faculty for detecting intrinsic wickedness. They arise from the impossibility of universalizing the contrary permissions without destroying the very moral relations those permissions exploit.

That matters because moral philosophy cannot merely possess the right prohibitions. It must know why they are right. Otherwise it cannot distinguish the genuine absolute from the fanatic’s taboo, the sadist’s discipline, the zealot’s purity rule, or the absurd prohibition against poodle ownership.

Anscombe’s fire burns away the corruptions of calculation. Hare’s structure tells us where the fire may rightly be set.

That is why Anscombe is a serious antagonist but not a refutation. She forces Hareans to become more careful about action, intention, and description. She reminds us that some moral lines must be strong enough to withstand the pressure of consequence. But she does not show that the moral ought is unintelligible. She shows only that it becomes dangerous when severed from truthful description and disciplined refusal.

Hare’s answer is not to abandon the moral ought. It is to make it answerable.

Anscombe gave moral philosophy its fire: the insistence that some acts must be refused. Hare gave it the test that fire still needs: the rule must be one we can universalize. Without Anscombe, moral philosophy risks forgetting that some acts must not be calculated into permissibility. Without Hare, it cannot explain why the prohibition is more than inherited conviction, taboo, or command.

The act must be truthfully described.The rule must be stated.The prescription must be universalized.And the human agent must remember that not every calculation they can perform is one they are morally entitled to trust.

That is not moral weakness. It is the beginning of moral sanity.

---

Afterword: Why Anscombe's Absolute Prohibitions Matter for Hare

This essay belongs to the larger Real Morality project of recovering R. M. Hare’s universal prescriptivism from the objections that have obscured it. Among the most common is the fanatic objection: if a sadist or zealot is willing to universalize cruelty, does Hare’s theory have any answer?

It does. The objection mistakes verbal assent for successful universalization. A fanatic does not pass the test by saying, “I accept this even if I am the victim.” To universalize adequately is to occupy every affected position with full imaginative and empirical seriousness. A prescription that survives only by flattening the victim’s reality or misdescribing the act has not been universalized. It has been protected from universalization.

Anscombe’s challenge matters because it arises from the same fear: that moral calculation can be used to justify atrocity. She was right to fear that. But the answer is not to make absolutes self-certifying. The answer is to explain why some prohibitions survive universalization and others do not. Hare gives us that explanation.

Anscombe reminds us that some acts must be refused. Hare explains how to tell which refusals can claim moral authority. That is why her fire and his test belong together.