previous next

Of the two divisions of this third book, proposed at the conclusion of Bk. II, and the opening of Bk. III, περὶ λέξεως καὶ τάξεως, πῶς χρὴ τάξαι τὰ μέρη τοῦ λόγου, the first having been dispatched in the preceding chapters 2—12, we now proceed to the second, on the arrangement of the parts of the speech: this will include a criticism of the anterior, and the current, divisions, with a new classification in c. 13: and an explanation and discussion of the proper contents of each. A full account of the various divisions which prevailed before and after Aristotle has been already given in the Introd. p. 331, 332, and the notes, and need not be here repeated. It will be sufficient to say that Aristotle in this chapter takes the fourfold division, adopted by Isocrates, and accepted by his followers, as the author of the Rhet. ad Alex., viz. προοίμιον, διηγήσεις, πίστεις, ἐπίλογος, criticizes it, and reduces it to two, πρόθεσις and πίστεις, as the only two parts necessary to the speech; adding notices of some superfluous distinctions introduced by Theodorus (of Byzantium) and Licymnius. [See Rössler's pamphlet, Rhetorum antiquorum de dispositione doctrina, pp. 30, Budissin, 1866; and Volkmann, die Rhetorik der Griechen und Römer, § 38.]

On the importance attached to the arrangement of the topics of these divisions, and especially to the order of the proofs, Whately has some good remarks, Rhet. c. 1 (Encycl. Metrop. p. 256). This is illustrated by the contest between Demosthenes and Aeschines. “Aeschines strongly urged the judges (in the celebrated contest for the Crown) to confine his adversary to the same order in his reply to the charges brought which he himself had observed in bringing them forward. Demosthenes however was far too skilful to be thus entrapped; and so much importance does he attach to the point, that he opens his speech with a most solemn appeal to the judges for an impartial hearing; which implies, he says, not only a rejection of prejudice, but no less also a permission for each speaker to adopt whatever arrangement he should think fit. And accordingly he proceeds to adopt one very different from that which his antagonist had laid down; for he was no less sensible than his rival that the same arrangement which is the most favourable to one side, is likely to be least favourable to the other.”


‘Of the speech there are (only) two parts: for it is only necessary first to state the subject, and then to prove (your side of) it. It follows from this necessary relation between them (διό), that it is impossible (if the speech is to be complete) either to state your case without going on to prove it, or to prove it without having first stated it’, (the impossibility lies in the absurdity of the supposition: it is a moral impossibility): ‘for proving implies something to prove, and a preliminary statement is made in order to be proved’. All this implies that the speaker has some object in view, some case to make out. It would not apply to all declamations; though it is true that, as a general rule, even they try to prove something, however absurd it may be.


‘Of these the one is the statement of the case (the setting forth of all its circumstances, as a foundation for judgment and argument), the other the (rhetorical) arguments in support of it, just as if the division were (the dialectical one) the problem (alternative question proposed or stated) and its demonstration’. πρόθεσις, propositio: Rhet. ad Al. c. 29 (30). 2, προεκτιθέναι τὸ πρᾶγμα. Ib. § 21, τὴν πρόθεσιν ἐν ἀρχῇ ἐκθήσομεν. c. 35 (36). I, φροιμιαστέον...πρῶτον προθεμένους τὰς προθέσεις: πίστις confirmatio.πρόβλημα διαλεκτικὸν θεώρημα, Top. A 11, 104 b 1, quod in disputando quaestione bipartita efferri solebat, ex. gr. voluptas estne expetenda, annon? mundus estne aeternus, annon?” Trendelenburg, El. Log. Ar. § 42, p. 118.


‘The present’ (current, Isocrates') ‘division is absurd; for surely narrative (διήγησις narratio, the detailed description of the circumstances of the case) belongs only to the forensic speech, but in a demonstrative or public speech how can there be a narrative such as they describe, or a reply to the opponent; or an epilogue (peroration) in argumentative or demonstrative speeches?’ On this Quint. says, III 9. 5, Tamen nec iis assentior qui detrahunt refutationem (sc. τὰ προς τὸν ἀντίδικον) tanquam probationi subiectam, ut Aristoteles; haec enim est quae constituat, illa quae destruat. Hoc quoque idem aliquatenus novat, quod prooemio non narrationem subiungit, sed propositionem. (This is one of Quintilian's ordinary misrepresentations of writers whom he quotes. Ar. says nothing here of the prooemium, theoretically disallowing it: though in compliance with the received custom he afterwards gives an account of it and its contents). Verum id facit quia propositio genus, narratio species videtur: et hac non semper, illa semper et ubique credit opus esse. The last clause very well explains Ar.'s substitution of πρόθεσις for (προοίμιον and) διήγησις.

In Introd. p. 333, I have given at length from Cic. de Inv. I 19. 27, the distinction of διήγησις in its ordinary sense and πρόθεσις. It is here said that the narrative or statement of the case, strictly speaking, belongs (he means necessarily belongs) only to the forensic branch of Rhetoric: there there is always a case to state: in the declamatory, panegyrical branch, not a regular systematic narrative or detailed statement as of a case; in this the διήγησις is dispersed over the whole speech, infra 16. 1: and, in δημηγορία equally, there is not universally or necessarily, as in the law-speech, a διήγησις, because its time is the future, and a narrative of things future is impossible: when it is used, it is to recall the memory of past facts for the purpose of comparison—which is a very different thing from the forensic διήγησις. Comp. c. 16. 11. The author of the Rhet. ad Alex. c. 30 (31) includes διήγησις in the deliberative branch, δημηγορικὸν γένος; no doubt following Isocrates. On διήγησις see Dionysius Hal., Ars Rhet. c. x § 14.

The same argument applies to the refutatio, τὰ πρὸς τὸν ἀντίδικον, and with more force than to the preceding, for in the epideictic branch there is no adversary, and therefore can be no refutation of his arguments, at least such as those who lay down this division intend: though it is true that a panegyrist may have to meet adverse statements or imputations on the object of his panegyric, real or supposed. In fact, it is only in the forensic branch that there is necessarily an opponent. On this division, see III 17. 14, 15.

ἐπίλογον τῶν ἀποδεικτικῶν] This is understood by Victorius, Majoragius, and Schrader of the demonstrativum genus, ἀποδ. being supposed to be put here for ἐπιδεικτικῶν. This in Aristotle I hold to be impossible. Nor have I found any example of it elsewhere, though Victorius says that Isocrates uses ἀποδεικνύναι for ἐπιδεικνύναι more than once in the Panath. speech. I have supposed (in note on p. 335 Introd.) that his text of Isocrates may have exhibited this interchange from the uncorrected carelessness of transcribers. What is true is, that Isocrates, twice in the Paneg. §§ 18 and 65, does use ἐπιδεικνύναι in a sense nearly approaching, if not absolutely identical with, that of ἀποδεικνύναι. The words can only mean, as I have translated them, that there may be some speeches which consist entirely of proof or arguments, and that a summary of these would not correspond to the ἐπίλογος in its ordinary sense—described c. 19. 1—of which only a small part is a recapitulation.

‘And again προοίμιον (preface, opening or introduction), and comparison (setting over against one another side by side) of opposing (views, statements, arguments), and review, are found in public speeches then only when there is a dispute (between two opponents)’: as in Demosthenes' Speech for the Crown, of which the προοίμιον has been before referred to. ἐπάνοδος, ‘a going over again’=ἀνακεφαλαίωσις, summary recapitulation of the foregoing topics of the speech, appears also in Plato Phaedr. 267 D, τὸ δὲ δὴ τέλος τῶν λόγων κοινῇ πᾶσιν ἔοικε συνδεδογμένον εἶναι, τινὲς μὲν ἐπάνοδον, ἄλλοι δὲ ἄλλο τίθενται ὄνομα. The ἄλλο ὄνομα may be ἐπίλογος or ἀνακεφαλαίωσις, or παλιλλογία (Rhet. ad Alex. c. 20 (21). 1). It is properly a subdivision of the ἐπίλογος, and as such is here condemned as superfluous.

ὅταν ἀντιλογία ] “The object of the prooemium is to conciliate the audience, and invite their attention, and briefly intimate the subject of the ensuing speech. In recommending this or that measure to the assembly, unless there is an adversary who has poisoned the hearers' minds against it and its author, or some other special reason, there is no occasion for this: and also, the audience is usually well acquainted with the subject. See further on this, c. 14. 11. Comparison of argument, and review, can only be required when there is an opposition.” Introd. pp. 335, 6. The Rhet. ad Alex. expressly tells us, c. 28 (29) ult., that the προοίμιον is “common to all the seven species, and will be appropriate to every kind of (rhetorical) business.”

The following argument καὶ γὰρπολλάκις is a reductio ad absurdum of the preceding. You say that προοίμιον, ἀντιπαραβολή and ἐπάνοδος are essential parts of the public speech—‘Why at that rate (is the reply) so are accusation and defence, for they are frequently there’—this involves the absurdity of introducing the whole contents of the forensic genus into the δημηγορικὸν γένος as a mere part of the latter—‘but not qua deliberation’: not in the sense or character of deliberation, which is essential to the deliberative branch, but as mere accidents.

There can be no question that we should read for συμβουλή. So Victorius, Schrader, Buhle, Spengel. Bekker alone retains . The following clause requires an alteration of punctuation to make it intelligible; suggested long ago by Victorius, Majoragius, Vater, and adopted by Spengel; not so by Bekker. Spengel also rejects ἔτι [delendum aut in ἐστὶν mutandum]. With the altered reading, ἀλλ᾽ ἐπίλογος ἔτι οὐδὲ δικανικοῦ κ.τ.λ. it is certainly out of place. I am by no means persuaded of the certainty of this alteration—perhaps Bekker had the same reason for withholding his consent to the two alterations—I think it quite as likely that a word or two has dropt out after ἐπίλογος.

‘But further’ (if ἔτι be retained) ‘neither does the peroration belong to every forensic speech; as for instance if it be short, or the matter of it easy to recollect; for what happens (in an ordinary epilogue) is a subtraction from the length’—not the brevity, of a speech: i.e. an epilogue is appropriate to a long speech, not a short one. This is Victorius' explanation, and no doubt right (that which I gave in the Introd. is wrong, and also not Victorius', as stated in the note).

‘Consequently the (only) necessary parts are the statement of the case, and the proof’.


‘Now these two are peculiar to, and characteristic of, speeches in general’.

It is possible that ἴδιον here may be the proprium of logic, one of the predicables: that which characterizes a thing, without being absolutely essential to it, as the genus and differentia are. The proprium is a necessary accident or property, though it is not of the essence itself: “but flowing from, or a consequence of, the essence, is inseparably attached to the species” (J. S. Mill, Logic, I p. 148). All this would apply very well to these two parts. They are not of the essence of the speech, and do not enter into the definition: the speech could exist without them. At the same time they are immediate consequences of that essence, and inseparably attached to all species of speeches, according to the view put forward here.

We might therefore be satisfied with these. ‘If we add more’ (following the authorities on the subject), ‘they must be at the most, preface, statement of case, confirmatory arguments, conclusion: for the refutation of the adversary belongs to the proofs’ (Quint. u.s. III 9. 5, Tamen nec iis assentior qui detrahunt refutationem, tanquam probationi subiectam, ut Aristoteles; haec enim est quae constituat, illa quae destruat), ‘and counter-comparison, (a comparative statement of your own views and arguments placed in juxtaposition with them to bring them into contrast,) which, being as it is a magnifying (making the most) of one's own case, must be a part of the confirmatory arguments, or general proof: for one who does this proves something: but not so the prologue; nor the epilogue, which merely recalls to mind’.


‘Such divisions, if any one choose to make them, will be pretty much the same as the inventions of Theodorus and his school, that is, to distinguish narration from after-narration and fore-narration, and refutation and per-re-refutation’. In this compound word ἐπί ‘in addition’ is represented by re, and ἐξ, ‘out and out’, ‘outright’, ‘thoroughly’, ‘completely’ by per. διά and per in composition are the more usual and direct exponents of ‘thoroughness’ or ‘complete carrying through’, of a thing. On ἐπιδιήγησις, repetita narratio, see Quint. IV 2. 128, res declamatoria magis quam forensis. He accepts it as a division, but thinks it should be rarely used. Plato, Phaedr. 266 D seq., in speaking of these same superfluous divisions of Theodorus, leaves out ἐπι- and προ-διήγησις, and introduces πίστωσιν καὶ ἐπιπίστωσιν in their place. These plainly correspond to the other pair ἔλεγχος and ἐπεξέλεγχος, the one being confirmatory, the other refutatory arguments. See Camb. Journ. of Cl. and Sacred Phil. No. IX. Vol. III p. 285, and Thompson's notes on the Phaedrus.

The general drift of the last clause is this; if you introduce such divisions at all, you may go on dividing and subdividing for ever, as Theodorus does in his τέχνη. This is followed by the statement of the true principle of division: the foundation of my own twofold division, hints Ar. ‘But a name (like one of these, the class-name, or, as here, the name of a division) should be given to mark a kind and a specific difference’. It is the genus plus the specific (εἰδοποιός, species-making) difference that constitutes the distinct species or kind. Now these names, though supposed to mark distinct kinds, have no specific differences which thus distinguish them. A special name demands a real distinction of kinds. Waitz ad Categ. 1 b 17. Trendelenburg, El. Log. Ar. § 59.

‘Otherwise they become empty and frivolous, such as Licymnius' inventions in his art, the names which he coins, ἐπούρωσις, ἀποπλάνησις and ὄζοι’. On Licymnius and his productions, see Heindorf ad Phaedr. u. s. p. 242, and Camb. Journ. of Cl.and Sacred Phil. No. IX. Vol. III pp. 255—7; where an attempt is made to explain these three obscure names. Licymnius was a dithyrambic poet, supra III 12. 2, as well as a rhetorician, and his prose style seems to have participated in the dithyrambic character. ἐπούρωσις I take to be a word coined by Licymnius for his own purposes: it is a ἅπαξ λεγόμενον. It seems to be formed from ἐπουροῦν, a synonym of ἐπουρίζειν, ‘to speed onward by a fair gale’, also συνεπουρίζειν, Hist. Anim. VIII 13. 9, de Caelo, III 2. 17: Polybius has ἐπουροῦν II 10. 6, and κατουροῦν, I 44. 3, 61. 7, both as neut. The Schol. quoted by Spengel, Artium Scriptores p. 89, defines ἐπόρουσις (ἐπούρωσις) τὰ συνευπορίζοντα καὶ βοηθοῦντα τοῖς ἐνθυμήμασι, καὶ ἁπλῶς ὅσα λέγονται βοηθοῦντα τῇ ἀποδείξει. All which seems to favour the notion that the figurative rhetorician represented ‘subsidiary’ or ‘confirmatory arguments’, Theodorus' πίστωσις and ἐπιπίστωσις, under the image of ‘a fair wind astern’. ἀποπλάνησις is no doubt, as in Plato Polit. 263 C, ‘a digression’, wandering off from the main subject, Schol. τὰ ἔξω τοῦ πράγματος; and ὄζοι, ‘branches’, most likely means places in which the discourse ‘branches off’ in different directions, ‘ramifications’: unless the same Scholiast's explanation be preferred, τὰ ἄκρα, ἤτοι τὰ προοίμια καὶ τοὺς ἐπιλόγους. This would mean the ‘branches’ opposed to the stock or trunk, as something extraneous, or at all events non-essential. (I think this is preferable.)

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 United States License.

An XML version of this text is available for download, with the additional restriction that you offer Perseus any modifications you make. Perseus provides credit for all accepted changes, storing new additions in a versioning system.

hide Display Preferences
Greek Display:
Arabic Display:
View by Default:
Browse Bar: