I agree with you: deprivation of a person's liberty is, in itself, a harsh punishment, even without taking into account the details of their incarceration. How

I agree with you: deprivation of a person's liberty is, in itself, a harsh punishment, even without taking into account the details of their incarceration. However, reasons matter. If policy dictates the purpose is to punish, then that justifies longer sentences "to make sure they are sufficiently punished" and to be ill-treated because "they are here to be punished". If policy dictates the purpose is to act as a deterrent to other crimes, then sentences are not justified as any longer than required to serve that purpose. That can be empirically observed, and I believe it has been studied and shown there is a strong scope insensitivity to the magnitude of punishments. This is even more pronounced in criminals, who are disproportionately more likely to have lower-than-average general intelligence or impulse control issues. If policy dictates the purpose is to keep the criminal out of society until they are ready to rejoin it, then incarceration should be optimised for preparing them to return to society, and they should not be kept imprisoned any longer when they are ready. How to judge whether this has been accomplished is difficult, but the effectiveness of the methods can be empirically observed, principally by recidivism rates. Note that the purpose of "punishment" is the only reason of these that cannot be empirically observed. It can justify just about any mistreatment, and doesn't, as far as I can tell, have any firm ethical justification except for the religious or pseudo-religious. So even if it seems like a minor question, deciding the purpose of the criminal justice system is of the utmost importance and has wide-ranging policy implications.