Book 3
13.3
But the division now generally made is absurd; for narrative only belongs in a manner to forensic speech, but in epideictic or deliberative speech how is it possible that there should be narrative as it is defined, or a refutation; or an epilogue in demonstrative speeches?The generally accepted divisions are: προοίμιον (exordium), διήγησις (narrative), πίστις (proof), ἐπίλογος (peroration). ( διήγησις is a species of πρόθεσις, which is used instead of it just before.) Aristotle objects that it is (as a rule) only the forensic speech which requires a regular διήγησις, a full and detailed statement of what has happened before. In epideictic and demonstrative (deliberative) speeches, the object of which is to prove something, there is no need of another existing division called the refutation of the adversary, and in the demonstrative there can be no room for an epilogue, which is not a summary of proofs and arguments. Thus the necessary divisions of a speech are really only two; πρόθεσις and πίστις, or at most four. In deliberative speeches, again, exordium, comparison, and recapitulation are only admissible when there is a conflict of opinion. For both accusation and defence are often found in deliberative, but not qua deliberative speech. And further, the epilogue does not even belong to every forensic speech, for instance, when it is short, or the matter is easy to recollect; for in the epilogue what happens is that there is a reduction of length.i.e. its use is to recall the main facts briefly (sect. 4 end), which in a short speech is needless.
