Why may the state punish? Retributive theories say punishment is deserved payback for wrongdoing. Consequentialist theories justify it by beneficial effects: deterring crime, protecting society, rehabilitating offenders. These justifications conflict: retribution asks 'what's deserved?' while consequentialism asks 'what works?' Most systems blend theories, but tensions remain over sentencing and proportionality.
Punishment is one of the most dramatic things the state does — it deliberately inflicts suffering on a person. Your prerequisite work on moral responsibility established that punishment seems to require that the person being punished is genuinely responsible for an offense: luck-induced behavior, coerced behavior, or behavior caused by severe mental illness complicates the moral case for punishment. But once responsibility is established, the deeper question remains: *why* may the state respond with punishment at all? This is where the major theories diverge.
Retributivism gives a backward-looking answer: punishment is deserved because the offender wronged others or violated the moral order, and justice demands that wrongdoing be met with proportionate payback. The offender's crime created an unjust advantage — they took more than their fair share by transgressing the constraints everyone else respects — and punishment restores a fair balance. Kant expressed this view in its starkest form: even if civil society were dissolved tomorrow, the last convicted murderer in prison should be executed, because justice demands it regardless of future consequences. The retributivist appeal is to moral intuition: some wrongs simply call for punishment, and failing to punish is itself a failure of justice. The crucial constraint this generates is that you may never punish an innocent person, regardless of the social benefits, because the innocent person does not deserve punishment.
Consequentialist theories give forward-looking answers: punishment is justified only by its beneficial effects. Deterrence (the threat of punishment prevents crimes), incapacitation (removing offenders from society prevents further crimes they would commit), and rehabilitation (transforming offenders into non-criminals reduces future crime) are all consequentialist goals. The standard objection is that pure consequentialism has no principled barrier against punishing innocents: if framing an innocent person would prevent a riot that kills dozens, the calculation might favor it. Classical retributivists see this as a reductio — the fact that consequentialism can license punishing innocents reveals that it has missed something essential about punishment's moral structure.
Mixed theories try to capture what is right in each. John Rawls's influential account distinguishes two levels of justification: we justify the *institution* of punishment by its social benefits (consequentialist reasoning), but within that institution, we justify *particular acts* of punishment by retributive principles (you may only punish the guilty, proportionately). This division preserves the retributive barrier against punishing innocents while allowing consequentialist reasoning to determine what the punishment system should look like. Most legal systems operate with some version of this blend — proportionality in sentencing reflects retributive constraints, while parole, rehabilitation programs, and sentencing guidelines reflect consequentialist concerns.
The practical stakes of this philosophical debate are high. If retribution is the primary justification, then mandatory minimum sentences and "three strikes" laws may be required by justice even when they produce worse social outcomes. If rehabilitation is the goal, then incarceration for non-violent offenses may be counterproductive — it produces worse outcomes for the person and society both. If deterrence drives the system, then we should be troubled by evidence that the death penalty does not deter murder more effectively than life imprisonment. Every criminal justice reform debate ultimately involves, at its core, disagreement about which of these justifications is primary.
Topics in reflective domains aren't scored by quiz answers. Read, reflect, and mark when you've thought it through.
No topics depend on this one yet.