Strip punishment of every possible benefit. It protects nobody, repairs nothing, deters no future harm, and teaches no one. If inflicting pain still feels morally necessary, the remaining claim is that suffering is good because the wrongdoer deserves it.
I do not buy that claim.
Wrongdoing calls for condemnation. A person who hurts someone may owe an apology, restitution, changed conduct, or the loss of freedom needed to protect others. None of that requires believing pain has moral value by itself. Justice should answer harm with safety and repair. Suffering that accomplishes neither is revenge with a theory attached.
Desert asks for an impossible author
Philosophers use “basic desert” for a particular idea. The Stanford Encyclopedia of Philosophy defines it as deserving blame or praise simply because an act was knowingly wrong or exemplary, apart from any good that blaming or praising might produce. In criminal punishment, positive retributivism treats desert itself as a reason to punish. The pain is supposed to fit the past wrong.
That picture demands a kind of authorship no person has. No one chooses the past that exists before a first choice or the capacities used to make it. Later choices can shape character, but those choices come from the character and capacities already present at the time. Trace the chain backward and there is no moment when a person creates a self from nothing.
Randomness cannot rescue the idea. As the moral responsibility skepticism entry explains, indeterminism does not supply the control that basic desert requires. A dice roll inside the mind supplies only chance. Authorship remains absent.
The free will debate extends much further. Compatibilists argue that meaningful freedom can exist even in a causally determined world, and the morally useful question can be whether a person acted through a capacity that recognizes and responds to reasons. That standard preserves the differences among a deliberating adult, someone acting under coercion, a young child, and a person unable to understand what an action means.
The capacity to recognize and respond to reasons can justify censure and demands for repair. Purposeless suffering still lacks a justification.
Accountability survives desert
A person can cause harm without being the ultimate author of every trait that led to it. The act still belongs to that person in morally important ways. It can express values, intentions, and judgments. Other people can ask for reasons and reject the answer.
The philosophical literature calls this answerability. As the Stanford entry on skepticism about moral responsibility explains, rejecting basic desert can leave us able to demand justification, apology, reform, and reconciliation. A victim can say, “You did this. It was wrong. You need to repair what you can.” That address treats the wrongdoer as someone capable of understanding a moral demand.
A sentence can make somebody miserable without requiring honesty, restitution, or change. Accountability asks the person who caused harm to face what happened and participate in making the future safer.
Moral judgment survives too. Calling cruelty wrong does not depend on proving that its author created a character without causes. The victim still suffers, and someone else’s freedom was violated. We can condemn the choice because of what it did to another person and because it expressed a disregard that the actor could be asked to examine.
Severity is a poor measure of safety
Evidence on sentencing exposes the weakness of severity as a safety policy. The National Institute of Justice’s review of deterrence research says the certainty of being caught matters far more than punishment severity. It says increasing already long sentences produces, at best, a modest deterrent effect. Incapacitation is real when confinement stops someone who presents a continuing danger. The claim that harsher sentences deter other people is different and far less supported.
A 2014 National Research Council consensus report reached a similar conclusion. It found little marginal crime reduction from long sentences through either deterrence or incapacitation and recommended reducing incarceration while rebuilding rehabilitation. The report also acknowledged that prison remains appropriate in specific cases. Protecting people sometimes requires restraint. The question is what level and form of restraint the evidence can justify.
Retribution answers that question with a feeling of moral proportion. More serious wrong means more suffering. When extra years do little for safety, the additional pain expresses vengeance and produces little public benefit.
Useful outcomes need hard limits
Retributivists have a serious objection. A justice system aimed only at useful outcomes could punish an innocent person if officials believed doing so would reduce crime. It could impose a grotesque sentence on one guilty person to frighten everyone else. The classic objection to consequentialist punishment is that it can turn human beings into tools for producing social benefits.
Crude outcome maximization fails this test. Its failure cannot grant moral value to purposeless suffering.
Dignity, equal citizenship, and limits on state power can ground absolute rules against punishing the innocent. Proof and culpability remain necessary because coercing someone who did not commit the harmful act violates that person’s rights. Proportionality remains necessary because the state must use the least restrictive response adequate to protect others and repair the harm. Regular review matters because risk changes, and the state should have to prove that continued restraint remains necessary.
Public safety language creates a different danger when it becomes an excuse for indefinite confinement based on predictions instead of acts. A humane system must identify proven harm, distinguish intent from accident and coercion, and justify every burden it imposes. Abandoning retribution should reduce state cruelty. A predictive bureaucracy that confines people indefinitely would betray that aim.
Repair asks more than revenge
Restorative justice offers one imperfect example of accountability built around harm. A Campbell Collaboration review of ten randomized studies found a lower frequency of subsequent crimes on average than standard justice alone. Participating victims reported greater satisfaction, more sincere apologies, less desire for revenge, and fewer symptoms of posttraumatic stress.
Those studies covered willing victims and offenders, and the strongest effects appeared when conferences supplemented conventional proceedings. The model will not fit every case. These meetings ask more of a wrongdoer than passive suffering does. The person has to hear how the victim was harmed, answer for it, and confront the work of repair.
Some people remain dangerous. Some harms cannot be repaired. Victims owe nobody forgiveness or contact. A justice system still needs boundaries, restraint, and condemnation. It should impose burdens because those burdens protect people or make repair possible, never because pain itself is a moral achievement.
We are shaped by causes we did not choose, and we remain capable of answering reasons, changing conduct, and taking responsibility for what comes next. Justice should address that capacity wherever it exists.
Pain that helps no one has no claim on us.