2/2 THE AMERICAN JOURNAL OF SOCIOLOGY
These conclusions will appear to many persons so paradoxical that it may be worth while to enforce them by a brief reference to still more patent phenomena in a closely allied department of morals. The form of justice we have thus far been dealing with is called distributive. It assumes the existence of a given amount of wealth, and asks by what principles this shall be apportioned among the members of society. Our first impulse, as has been pointed out, is to give the reins into the hands of gratitude, and attempt to make reward commensurate with desert. But grati- tude, and still more frequently its correlative resentment, may be called in to solve another problem, that of retribution. A crime has been committed, and it becomes necessary to determine the severity of the penalty. Common sense would again reply, let the punishment be proportioned to the desert. But here a diffi- culty at once arises over and above those which this conception has already been seen to involve. The degree of guilt is meas- ured by the amount of effort that would have been requisite to overcome the temptation. Assuming for the moment that this amount could be determined and expressed in "units of effort" a supposition which we have seen to be absurd how determine the unit of punishment which shall be equivalent to a unit of demerit? Shall it be a day in the penitentiary at hard labor, or a week, or a month? If a man, under the stress of a given temptation, has stolen ten dollars from his employer's till and thereby contracted five "units of guilt" shall he be imprisoned five weeks or five months? No one can really answer, and the only principle which is capable of a consistent application in such cases is, by an exploration of the records of past experience to determine what penalties are most effective in preventing crime and what methods of punishment promise to do most for the reformation of the criminal. These are the principles adopted by modern penology, but they involve the application of the criterion of well-being and not of desert. Our conclusion with regard to the standard of distributive justice thus seems to be confirmed by a study of retributive justice. Our denial of the possibility of apportioning reward as well as penalty to
