FROM STATEMENTS MADE BY JUDGES AND MAGISTRATES, THE PRINCIPAL AIMS OF SENTENCES ARE DETERRENCE, REFORMATION, INCAPACITATION, RETRIBUTION, DENUNCIATION, AND REPARATION. THE AIM OF RETRIBUTION USUALLY IS DEFINED IN TERMS OF PUNISHMENT OR SUFFERING IN RETURN FOR CRIME COMMITTED. SOME IDEA OF PROPORTIONALITY OFTEN IS INCLUDED IN THE CONCEPT, SINCE IT IS THOUGHT THAT THE AMOUNT OF RETRIBUTION SHOULD INCREASE WITH THE GRAVITY OF THE CRIME. REPARATION INVOLVES MAKING AMENDS FOR A CRIME, EITHER TO THE VICTIM DIRECTLY OR TO SOCIETY IN GENERAL. ITS SUBJECTIVE NATURE, EXCEPT FOR DIRECT RESTITUTION OF STOLEN GOODS, IS SIMILAR TO THAT OF RETRIBUTION. DENUNCIATION INVOLVES THE USE OF SENTENCES TO DEMONSTRATE SOCIETY'S ABHORRENCE OF CRIME, OR AT LEAST A SENTENCER'S ABHORRENCE OF CRIME. INCAPACITATION INVOLVES ELIMINATING THE POSSIBILITY FOR AN OFFENDER TO COMMIT CRIMES IN THE COMMUNITY FOR A SPECIFIED PERIOD. THIS IS PROBABLY THE EASIEST AIM TO DEFINE OBJECTIVELY AND INVESTIGATE EMPIRICALLY, AT LEAST IN PRINCIPLE. MUCH RESEARCH HAS BEEN CONDUCTED TO DETERMINE THE EFFECTS OF SENTENCING ON THE AIMS OF DETERRENCE AND REFORMATION. DETERRENCE INVOLVES MAKING AN OFFENDER OR POTENTIAL OFFENDER LESS LIKELY TO COMMIT CRIME BECAUSE OF HIS OR HER FEAR OF THE LEGAL CONSEQUENCES. REFORMATION REFERS TO IMPROVING AN OFFENDER BY REDUCING HIS OR HER CRIMINAL INTENT OR DESIRE TO COMMIT CRIME. THE MAJOR PROBLEM IN CONDUCTING RESEARCH ON THE GENERAL DETERRENT EFFECT OF SENTENCES IS THAT IT IS DIFFICULT TO INSURE THAT ANY DECREASE THAT MAY OCCUR AFTER A CHANGE IN LEGAL PENALTIES OR EXEMPLARY SENTENCES IS CAUSED BY THE SENTENCE CHANGE RATHER THAN BY ONE OF MANY OTHER CONCURRENT SOCIAL CHANGES. WITHOUT SURVEYING POTENTIAL OFFENDERS, ONE DOES NOT KNOW IF ANY CRIME REDUCTION CAN BE ASCRIBED TO DETERRENCE. IN PRINCIPLE, INDIVIDUAL DETERRENCE AND REFORMATION ARE EASIER TO INVESTIGATE THAN GENERAL DETERRENCE BECAUSE IT IS MORE FEASIBLE TO CARRY OUT RESEARCH IN WHICH EXTRANEOUS INFLUENCES ON OFFENDERS ARE CONTROLLED. IF OFFENDERS ARE NOT ALLOCATED RANDOMLY WHEN EVALUATING DIFFERENCES IN CRIMINAL CAREERS BETWEEN OFFENDERS WITH DIFFERENT TYPES OF SENTENCES, IT IS ALWAYS POSSIBLE THAT VARIATIONS REFLECT PREEXISTING DIFFERENCES RATHER THAN EFFECTS OF SENTENCES. VERY FEW STUDIES HAVE COMPARED BEFORE-AND-AFTER MEASURES ON OFFENDERS OR A SENTENCED GROUP WITH AN UNSENTENCED GROUP. MUCH INFORMATION IS AVAILABLE ON RECONVICTION RATES FOR OFFENDERS RECEIVING DIFFERENT SENTENCES. THE GENERAL POINT OF THE ARTICLE IS THAT THE LITERATURE DOES NOT INDICATE HOW EFFECTIVE ANY GIVEN SENTENCE IS IN ACHIEVING AIMS RELATED TO OFFENSES AND OFFENDERS. (DEP)