Quintilian, Institutio Oratoria, Book 5
Harold Edgeworth Butler, Ed.

1. 1. To begin with it may be noted that the division laid down by Aristotle¹ has met with almost universal approval. It is to the effect that there are some proofs adopted by the orator which lie outside the art of speaking, and others which he himself deduces or, if I may use the term, begets out of his case. The former therefore have been styled ἄτεχνοι or inartificial proofs, the latter ἔντεχνοι or artificial. [2] To the first class belong decisions of previous courts, rumours, evidence extracted by torture, documents, oaths, and witnesses, for it is with these that the majority of forensic arguments are concerned. But though in themselves they involve no art, all the powers of eloquence are as a rule required to disparage or refute them. Consequently in my opinion those who would eliminate the whole of this class of proof from their rules of oratory, deserve the strongest condemnation. [3] It is not, however, my intention to embrace all that can be said for or against these views. I do not for instance propose to lay down rules for commonplaces, a task requiring infinite detail, but merely to sketch out the general lines and method [p. 159] to be followed by the orator. The method once indicated, it is for the individual orator not merely to employ his powers on its application, but on the invention of similar methods as the circumstances of the case may demand. For it is impossible to deal with every kind of case, even if we confine ourselves to those which have actually occurred in the past without considering those which may occur in the future.