Quintilian Institutio Oratoria 5.2

[1] As regards decisions in previous courts, these fall under three heads. First, we have matters on which judgment has been given at some time or other in cases of a similar nature: these are, however, more correctly termed precedents, as for instance where a father's will has been annulled or confirmed in opposition to his sons. Secondly, there are judgments concerned with the case itself; it is from these that the name praeiudicium is derived: as examples I may cite those passed against Oppianicus ¹ or by the senate against Milo. ² Thirdly, there are judgments passed on the actual case, as for example in cases where the accused has been deported, ³ or where renewed application is made for the recognition of an individual as a free man, ⁴ or in portions of cases tried in the centumviral court which come before two different panels of judges. ⁵

[2] Such previous decisions are as a rule confirmed in two ways: by the authority of those who gave the decision and by the likeness between the two cases. As for their reversal, this can rarely be obtained by denouncing the judges, unless they have been guilty of obvious error. For each of those who are trying the case wishes the decision given by another to stand, since he too has to give judgment and is reluctant to create a precedent that may recoil upon himself.

[3] Consequently, as regards the first two classes, we must, if possible, take refuge in some dissimilarity between the two cases, and two cases are scarcely ever alike in all their details. If, however, such a course is impossible and the case is the same as that on which the previous decision was given, we must complain of the negligence shown in the conduct of the previous case or of the weakness of the parties condemned, or of undue influence employed to corrupt the witnesses, or again of popular prejudice or ignorance which reacted unfavourably against our client; or else we must consider what has occurred since to alter the aspect of the case.

[4] If none of these courses can be adopted, it will still be possible to point out that the peculiar circumstances of many trials have led to unjust decisions; hence condemnations such as that of Rutilius ⁶ and acquittals such as those of Clodius and Catiline. We must also ask the judges to consider the facts of the case on their merits rather than make their verdict the inevitable consequence of a verdict given by others.

[5] When, however, we are confronted by decrees of the senate, or ordinances of emperors or magistrates, there is no remedy, unless we can make out that there is some difference, however small, between the cases, or that the same persons or persons holding the same powers have made some subsequent enactment reversing the former decision. Failing this, there will be no case for judgment.