Most of the standard arguments against the retributive theory of punishment are hardly new. That the retributive view of punishment is but a rationalization of a primitive urge for revenge; that the retributivists, instead of providing an answer to the question about the source of our moral right to add a new evil (punishment) to an already perpetrated one (the crime), simply assert dogmatically that punishment is an intrinsic good, i.e. something that needs no further moral justification; that it is impossible to apply the lex talionis in practice; that the retributivist thesis that the criminal has a right to punishment is absurd, because the criminal himself would be the first to deny that he has any such right; that the retributive theory is incompatible with the claims of forgiveness and mercy; that the practical consequences of the theory are conservative, the theory itself being in fact an apology for the existing laws and the existing social order; that the Hegelian idea of punishment as a ‘negation’ or ‘annulment’ of crime is either unintelligible or essentially utilitarian in character–most of this had been said already in the last century, or even earlier. All these arguments are still in use.1 But in recent literature–in a number of papers published in the last twenty years or so–we find a new argument against the retributive theory. It is only natural that this particular argument should not have been used earlier: methodologically it is typical of a philosophical orientation which emerged only in our century–analytical philosophy.