¹ Homer on Life and Death (Oxford, 1980), p. xivGoogle Scholar.

² CQ n.s. 26 (1976), 161–87CrossRefGoogle Scholar, esp. pp. 161 and 183. Griffin also examined the scholiasts' use of the cognate adjectives and adverbs. This discussion seems to bear out Griffin's apparent assumption that πάθος means ‘pathos’, but one would have liked to see this assumption defended, or at least explained. A revised form of the article appeared in Griffin's book as ch. 4, ‘Death, Pathos and Objectivity’. From here on in this article the terms pathos and ēthos will normally appear in transliterated form.

³ cf. Rutherford, William G., A Chapter in the History of Annotation = Scholia Aristophanica ii (London, 1905), pp. 138–46Google Scholar, who discusses the rhetorical background to the commentators' use of ēthos and pathos, andPohlenz, Max, πρέπο 16 (1933), 53–92, esp. pp. 67–70Google Scholar.

⁴ I am concerned here only with cases where the two terms are contrasted. They are often coupled, especially in the form ἤθη καì πάθη cf. Pohlenz, op. cit. p. 67 andKeuls, Eva, Plato and Greek Painting (Leiden, 1978), pp. 98 ff.Google Scholar; and their meanings seem sometimes to merge in the scholia, cf.Kroll, W., ἤθεɩ, Philologus 75 (1918), 68–76CrossRefGoogle Scholar. I do not propose to explore those developments here.

⁵ I shall refer simply to Longinus, without inverted commas or brackets, denoting the author of the treatise On the Sublime. As for his date, ‘the received opinion today is that we have to do with a book of the first century A.D.’, Russell, in Russell, D. A. and Winterbottom, M., ed., Ancient Literary Criticism: The Principal Texts in New Translations (Oxford, 1972), p. 461Google Scholar.

⁶ cf. Lucas, D. W., ed., Aristotle: Poetics (Oxford, 1968)Google Scholar, adloc, who clarifies this confusing passage. On tragic ēthos, see Poet. 2, 1448a l–18, 13 passim, esp. 1453a 16–17, 15, 1454a l6 ff., 1454 b8 ff. Cf.Schütrumpf, Eckart, Die Bedeutung des Wortes ēthos in der Poetik des Aristoteles, Zetemata Monographs 49 (Munich, 1970), esp. pp. 52–63Google Scholar, who effectively defends the view that, when Aristotle in the Poetics describes people as ‘good’ (e.g. σπουδαȋοɩ, πɩεɩκεȋς) with respect to ēthos, he is using these terms in the same sense as in the ethical writings. See also n. 12 below. For the connection of ēthos with προαίρεσɩς, and with speeches denotative of this, see Poet. 6, 1450b8–ll, 15, 1454a 17–19.

⁷ See 18, 1455b34–5, 6, 1449b24–8, 13, 1452b28 ff.

⁸ 12, 1452b 11–13; cf.Rees, B. R., ‘Pathos in the Poetics of Aristotle’, G & R 19 (1972), 1–11Google Scholar, who discusses the puzzling phrase τ ϕανερ θάνατοɩ and suggests that it probably means deaths described by messengers as well as those exhibited on the ekkuklēma or shown on stage.

⁹ See 14, 1453b 11 ff. The OT owes much of its emotional impact to the violent pathos the play uncovers, Oedipus' killing of his father (1453b29–32), and the violent acts this discovery provokes in Jocasta and Oedipus (perhaps alluded to at 14, 1453b3–7). But the OT is a classic example of a ‘complex’ tragedy rather than a pathetikē one, because these elements are integrated in a plot which hinges on a striking combination of peripeteia and anagnōrisis (11, 1452 a 29–33).

¹⁰ This is not explicit, but may be implied. Cf. Lucas on 60a 12, 62b 13.

¹¹ For Aristotle's view of Achilles' ēthos see 15, 1454b8–15, discussed fully inElse, Gerald F., Aristotle's Poetics: The Argument (Cambridge, Mass., 1957)CrossRefGoogle Scholar, ad he. All subsequent references in this article to ‘Else’ denote this work, unless otherwise specified.

¹² See Poet. 2, 1448a 1–5, 11–12, 3, 1448a25–9, 4, 1448b24 ff. This contrast between the Iliad and Odyssey I propose to explore elsewhere; but an obvious example of what I have in mind is that the dialogue and narrative of the Odyssey is permeated with adjectives and adverbs which divide people into two ‘camps’ : those who are sensible, wise, thoughtful, and those who behave with recklessness, thoughtlessness, lawless arrogance, whereas this is not so obviously true of the Iliad.

¹³ 13, 1453a30–6. In 14, 1454a4–9, notoriously, he takes a different view; cf. Moles, John, ‘Notes on Aristotle, Poetics 13 and 14’ , CQ n.s. 29 (1979), 77–94Google Scholar, esp. pp. 82 ff.

¹⁴ 13, 1453a 1–7; cf. Else, pp. 531 ff.

¹⁵ Lamentation for Hector, It. 24. 704 ff.; cf. the scene of mutual lamentation between Achilles and Priam, 507 ff. The latter scene contains the speech of Achilles on the arbitrariness of divine punishment and reward that Plato found so offensive (527 ff.; cf. Resp. 379d). Aristotle's theory of tragedy, by contrast, accepts undeserved suffering as a prime tragic theme, provoking the ‘proper pleasure’ of pity and fear; cf. Poet. 13, passim, andGlanville, I. M., ‘Tragic error’, CQ 43 (1949), 47–56, esp. pp. 52–6CrossRefGoogle Scholar.

¹⁶ In this respect, Aristotle's thinking seems to anticipate the later link between ēthos and comedy; though, of course, he knows only old and middle comedy and not the Menandrian comedy later critics associated with ēthos. See furtherPost, L. A., ‘Menander in current criticism’, TAPA 65 (1934), 13–35, esp. pp. 14 ff.Google Scholar, ‘Aristotle and Menander’, TAPA 69 (1938), 1–42Google Scholar; and n. 57 below for the later sense of ēthos.

¹⁷ Stinton, T. C. W., ‘Hamartia in Aristotle and Greek Tragedy’, CQ n.s. 25 (1975), 221–54, esp. pp. 228–35CrossRefGoogle Scholar. One way of describing some at least of the cases of Aristotelian μαρτία Stinton thinks appropriate to tragedy would be to say that the central figure acts in the grip of pathos instead of expressing his, generally good, ēthos; see Stinton's cases (2), (3), (4) on 232. In the Poetics it is only in 15, 1454b 11–15 (Achilles et al. ρϒίλους but πιεκεȋς) that Aristotle comes close to considering the interplay of character and emotion.

¹⁸ e.g. 6, 1449b24–8, 9, 1452a 1 ff., 13–14, passim.

¹⁹ 14, 1453b 18 and context; cf. 14, 1453b3–7, 13, 1453al7–22.

²⁰ 17, 1455a22 ff. Lucas, ad loc, thinks Aristotle has in mind both ‘the original situation, as it existed at Thebes during the plague for instance, or the situation as it was to be represented in the theatre of Dionysus’, but it is certainly the latter Aristotle has in mind in a26–9. In Quint. 6. 2. 32 and De Sublim. 15. 1 ff., which seem, like Hor. Ars P. 101 ff., to be developments of this passage (cf. Lucas on 55 a 32), the question of άρϒεια is extended to the point of view of the fictional character. The poet imagines the scene as the character sees it and is thus enabled to convey the character's emotional response in words and so make the audience see the scene (and feel the corresponding emotions) from the character's viewpoint.

²¹ 1455a29–32. My translation follows the interpretation of Lucas; cf.Bywater, Ingram, Aristotle On the Art of Poetry (Oxford, 1909)Google Scholar, ad loc. Lucas finds Aristotle's identification of the poet with the performer ‘somewhat bizarre’, and Else finds it inconceivable and reinterprets the whole passage accordingly (taking σχήμασιν to mean ‘figures of speech’ and referring τοȋς to the — unmentioned — dramatic characters). ButJones, John, On Aristotle and Greek Tragedy (Oxford, 1962), pp. 34–5Google Scholar, is surely right to insist that we retain, and accept, the plain meaning of the text. As Else, in spite of his interpretation, points out, Aristotle's model of the poet goes back to Plato; at least, ‘This idea, that the poet lurks behind and in his characters to establish an emotional sway over his audience…, is in fact Plato's assumption in the third book of the Republic’ (pp. 491, cf. 494, Resp. 392d ff, cf. Ion 553d–e and Resp. 605d, κροώμενοι Όμήρου ἢ ἄλλου τινòς τς τραϒωδοποι μιμουένου τιν τν ρώων πένθει ὂντα… πόμεθα συμπάσχοντες). Aristotle sėems to see the poet as a quasi-orator (as well as rhapsodeactor); see πιθανώταοι in a 30. Hence the apparent attraction of the passage for orators; see refs. in n. 20 and Cic. De Or. 2. 188–97. Lucas' comparison to Ibsen, ‘I have to visualize his whole exterior too, down to the last button…’, i.e. visualization of the appearance of the character (note on 55 a 32), is not very appropriate.

²² 1455a32–4. The passage either adopts (Lucas, Bywater) or modifies (Else) the Platonic idea of the ‘madness’ of the poet.

²³ Rh. 1. 2, 1356a 1 ff., 1. 9, 1366b23 ff., on προαίρεσις and πρξις 1. 9, 1367b22 ff.; cf. 2.21, 1395bll ff., 3. 16, 1417a 16 ff.

²⁴ 1.2, 1356a 14–16, 2. 1, 1377b20ff.

²⁵ There are at least two qualifications to be made to this claim (cf. also n. 32 below). One is that the analysis of the pathē in Rh. 2. 2 ff. is clearly meant to provide the speaker with guidelines about which emotions he can plausibly express, e.g. anger or indignation, as regards his treatment at the hands of others. Nevertheless, it is the arousal of such emotions in the audience that Aristotle stresses, e.g. 2. 2, 1380al–5,2. 3, 1380b 31–4 and the conclusion in 2–11, 1388b29–30. It is true, also, that the section 2. 12–17 characterizes the ēthē of different sections of the potential audience (young, old, rich, poor). But the point of considering the typical attitudes and emotional inclinations (κατ τ πάθη 2. 12, 1388 b 31) of these groups is to enable the speaker to assimilate his attitudes to theirs and so more effectively play on their emotions. Their ēthē are not being considered from the same standpoint as the (individual, moral) ēthos of the speaker: cf. Cope, E. M., An Introduction to Aristotle's Rhetoric (London, 1867), pp. 108–18Google Scholar; Süss, Wilhelm, Ethos, Studien zur Älteren Griechischen Rhetorik (Leipzig and Berlin, 1910), pp. 152 ffGoogle Scholar; Kennedy, George, The Art of Persuasion in Greece (Princeton, 1963), pp. 91–4Google Scholar; Solmsen, Friedrich, ‘Aristotle and Cicero on the orator's playing upon the feelings’, CP 33 (1938), 390 ff.Google Scholar; Schütrumpf, op. cit. pp. 28 ff.

²⁶ 1. 2, 1356a4–13, 2. 1, 1377b20–1378a5.

²⁷ For exceptions, see n. 32 below.

²⁸ Attic law distinguished, at least in the case of homicide, between acts committed κ προνοίας, from forethought, and those not so committed; cf.Irwin, T. H., ‘Reason and responsibility in Aristotle’, in Aristotle's Ethics, ed. Rorty, Amélie Oksenberg (Berkeley and Los Angeles, 1980), pp. 119–20Google Scholar and refs. Therefore, one might have expected Aristotle to discuss the defence that one acted, in such cases, from pathos and not from a πρόνοια that expressed one'S Ḕthos, e.g. in his discussion of the motives of wrongdoing,Rh. 1. 10–12. But there seems not to be any such discussion, despite his interest in such topics in the ethical writings; cf. Irwin, 120 ff.

²⁹ For Plato's views see Resp. 603c ff.; an d for Aristotle's response see refs. in n. 6 an d n. 17 above. The clearest account of this aspect of the debate between Plato and Aristotle is given by Else in The Structure and Date of Book 10 of Plato's Republic, AHA W (1972) esp. pp. 44–5. (Else's account is not vitiated by his controversial view that the relevant section of Book 10 of theRepublic is a separate, late composition, written in response to an Aristotelian proto-Poetics, pp. 55–7). Aristotle's determination to avoid the Platonic position on this question is probably his strongest motive for avoiding discussion of pathos in tragic figures, a point I pursue further in an article forthcoming in Poetics Today, ‘The question of character and personality in Greek Tragedy’.

³⁰ cf.Poet. 19, 1456a37–b2, and Rh. 1. 2, 1356a 1 ff. Aristotle's failure, in the Poetics passage, to mention the ethical proof probably reflects his sense that he has dealt with Ḕthos elsewhere (e.g. in ch. 15) and is concerned only with dianoia here; cf. Else, p. 563. There are, in any case, clear parallels between the treatment of Ḕthos in the two treatises; cf. e.g.Poet. 6, 1450b8–12, 15, 1454a 16-19, with Rh. 1,2, 1356a 1–13, 2. 21, 1395b 11–19, 3. 16, 1417a 16 ff. (including the famous discussion of Antigone). See further Dale, A. M., ‘Ethos and Dianoia: “character” and “thought” in Aristotle's Poetics’, AUMLA 11 (1959), 3 ff.Google Scholar, esp. pp. 11–16, who argues strongly for rhetorical influence on Aristotle's dramatic theory; and Else, pp. 565–6, who maintains that Aristotle saw the parallels between the two areas but also advocated that the differences be recognized as well.

³¹ Here I am trying to paraphrase the problematic sentence in 1456 b 2–7, drawing on the exegesis of Bywater, Else and Lucas, ad loc; cf. Hubbard's, M. E. note on her translation of the Poetics, in Ancient Literary Criticism, ed. Russell, and Winterbottom, , pp. 116–17, n. 5Google Scholar. In the Greek of 1456a37–b7, Aristotle does not in fact distinguish as clearly as I have done between the emotional effect the characters have on each other (or the audience) and the emotional effect the dramatist has on the audience; cf. the tendency to conflate these two effects in ch. 17, 1455 a22 ff., discussed above.

³² 3. 21, 1395a23–4, ἔστι δ παӨητικ μν οιοεἴ τισ ργιξΌμενο7sigma; κ.τ.λ 3. 7, 1408a 16, παӨητικ δ ν μν ᾖ ὕβρισ, ργιξομνο λγισ κ.τ.λ., 1408a23–4, σνομοπαӨεî κοων εì τω παӨητικωσ λγο7nu;τι, κἄν μηӨν λγ In 3. 16, 1417a36–7, τν παӨη7tau;ικν has a rather different sense, ‘acts expressive of emotion’, where the emotions are those of the speaker's opponent. Cf. Cope-Sandys, , The Rhetoric of Aristotle (Cambridge, 1877),Google Scholarad loc.

³³ See, e.g., 2. 21, 1395a20–34, 3. 16, 1417a 16–b 10, 3. 17, 1418a 12–21. Cf. alsoḔthikos logos (1. 8, 1366a8–16, 2. 18, 1391b25–7), i.e. logos designed to show that the Ḕthos of the speaker is sympathetic to that of the audience in political attitudes (cf. n. 25 above).

³⁴ 3.7,1408al0–b20,esp.a25–32andbll–20.For this kind of Ḕethos, cf.Poet 15,1459al9–24. The passage on ‘ethical’ style is sometimes taken to refer to the speech-maker's attempt to make the speech verbally suitable for the client in question; see, e.g., Kennedy, , The Art of Persuasion in Greece, pp. 91–2Google Scholar. However, as Elaine Fantham points out, ‘women and foreigners’ (Aristotle's examples here) ‘could not plead their own cases, so Aristotle cannot be discussing how to characterize a client, but rather how to include narrative quotations in the style of those quoted’, ‘Ciceronian Conciliare and Aristotelian Ethos', Phoenix 27 (1973), 262–75Google Scholar, quotation from p. 272.

³⁵ Eloc. 28. ‘Oμοιοτλετα are not useful for pathos and Ḕthos, for both of these require simplicity and naturalness. The meaning of Ḕthos is not explained, but pathos is clearly associated with the excitation of emotions such as grief. For authorship and date of this treatise, see Ancient Literary Criticism, ed. Russell, and Winterbottom, , p. 172Google Scholar.

³⁶ See 69: quot officia oratoris tot sunt genera dicendi… modicum in delectando (cf. iucundum, 128); (‘grand’) vehemens in flectendo; in quo uno vis omnis oratoris est (cf.in quo uno regnal oratio, 128). See further 97–9: (‘grand’) quae cursu magno sonituque ferretur, etc. (cf. rapide fertur, 128),tractare animos…permovere (cf.quo perturbantur animi et concitantur, 128), ardens…inflammare (cf. incensum, 128); medius suavis (cf. 91–2 and iucundum, 128). On Cicero and the three styles, see Douglas, A. E., ‘A Ciceronian contribution to rhetorical theory’, Eranos 55 (1957), 18–26Google Scholar.

³⁷ cf. Fantham, op. cit. p. 262. Both the speaker-related and the audience-related functions of Ḕthikon have Aristotelian parallels (cf. nn. 23, 25 and 33 above). The link between the two treatises is particularly clear at De Or. 2. 213 (cf. Oral. 128), esp. hoc [genus], quod ad vitam et mores accommodatur. In De Or. it is clear (from 2. 182–4) that, in this phrase, Cicero primarily has in mind the favourable portrayal of the client (and advocate), though he also thinks this will guarantee a favourable response from the audience; cf. n. 41 below.

³⁸ See De Or. 2. 115, 121, and Arist. Rh. 1. 2, 1356al ff.; the Ḕthos-passage is De Or. 182–4; the pathos-passage is 2. 185–215. Cf. Fantham, op. cit. pp. 262 ff.; Solmsen, op. cit. pp. 396–402; Kennedy, , The Art of Rhetoric in the Roman World (Princeton, 1972), pp. 222–3Google Scholar.

³⁹ cf. lenitas vocis, vultus pudoris significatio, verborum comitas (2. 182); tantum autem efficitur sensu quodam ac ratione dicendi, ut quasi mores oratoris effingat oratio (2. 184). On the link presupposed here between style, esp. verbal style, and socially approved qualities of character, see Fantham, op. cit. p. 263, on lenitas, and Zucker, Friedrich, ‘AνηӨο7pi;οíητο’, in Semantka, Rhetorica, Ethica. Deutsche Akademie der Wissenschaften zu Berlin, Schriften der Sektion für Altertumswissenschaft (Berlin, 1963), 33 ffGoogle Scholar.

⁴⁰ See, e.g.,De Or. 2. 192, 200–1, and cf. Kennedy, , ‘The rhetoric of advocacy in Greece and Rome’ AJP 89 (1968), 419–36, esp. 428–35Google Scholar. Fantham, however, sees confusion of thought in Cicero, resulting from an uncritical reproduction of Aristotle's Ḕthos-theory, in which the advocate-client distinction played no part; op. cit. pp. 264–6, 271–3.

⁴¹ De Or. 2. 182, cf. 2. 114–15 and Arist. refs. in n. 26 above. In Cicero's later rhetorical works, the function of conciliare tends to be replaced by delectare (see Or. 69, and cf. Fantham, op. cit. pp. 273–5). Fantham points out that in the later stages of Cicero's discussion (De Or. 2. 212, 216), the distinction between conciliare and movere (and the distinction between the correlated styles) tends to merge: ‘Conciliatio has here become merely a label for one section of movere’ (Fantham, op. cit. p. 267). This development may help to explain why Quintilian presents Ḕthos and pathos as alternative types of emotional modes for the orator to use, indeed as two different types or levels of emotion (see discussion below).

⁴² DeOr.2. 183; cf. 211–12. In 2. 128–9, three stylistic qualities are linked with three oratorical functions (acumen with docere, lenitas with conciliare, vis with concitare). This is not quite identical with the more fully elaborated system of three styles and functions given in Oral. 69 ff.; Cicero there seems to have in mind stylistic attractiveness (suavitas), aimed at pleasing (delectare), rather than gentleness of tone(lenitas), aimed at winning over(commendare); cf. nn. 36 and 41 above. However, the distinction is a fine one; character should be presented suaviter as well as actione leni in De Or. 2. 184, and in Orat. 128 the ‘ethical’ style should be pleasing (iucundum) as well as winning (come…adbenevolentiam conciliandam paratum). In general, it seems fair to say that Cicero associates character-presentation with some version of the middle style, and associates emotion, consistently, with the grand and intense style, both in De Or. and in Orat.

⁴³ 2. 189 ff. The passage is closer to Arist. Poet. 17 (note the reference to the Platonic ‘mad’ or ‘inspired’ poet in 2. 194); cf. Poet. 1454a32–4, and Lucas, note on 55a32. Cf. Hor. Ars P. 102–3, si vis meflere, dolendum est /primum ipsi tibi, and on the question of the orator's sincerity in these emotional states see Brink, C. O., Horace on Poetry (Cambridge, 1981), p. 186Google Scholar.

⁴⁴ For the Aristotelian usages cf. nn. 33–4 above. For Ḕthos in Dionysius, see Lockwood, J. F., ‘HOIKH AEξI’, CQ 23 (1929), 181–5Google Scholar.

⁴⁵ D.H., Opuscula, ed. Usener, H. and Radermacher, L. (Leipzig, 1899), 1Google Scholar. 143, Dem. 8; Opuscula, vols. 1–2 (ed. Usener-Radermacher) = D.H. Teubner vols. 5–6. Cf. D.H.,Comp. 11. 37, for a similar list of contrasted qualities, discussed by Russell, D. A., Criticism in Antiquity (London, 1981), pp. 133–5Google Scholar.

⁴⁶ Dem. 2, Usener-Radermacher, 1. 131; cf. Dem. 43, Usener-Radermacher, 1. 224. Cf. Lockwood's comment on Dem. 22 (see n. 47 below), on Ḕthos: ‘the reference is to the effect on the hearer's mental outlook. Isocrates is like a military band; he makes you stick out your chest and think yourself “no end of a fellow”’ (op. cit. p. 182 n. 6).

⁴⁷ Dem. 22, Usener-Radermacher, 1. 176. For Ḕethos/pathos used in a quite Aristotelian way, see Dem. 18, Usener-Radermacher, 1. 166.

⁴⁸ cf. Solmsen, op. cit. pp. 395–6; Zucker, op. cit. p. 43; Keuls, op. cit. p. 99 n. 37.

⁴⁹ e.g. under Ḕthos we find the long-standing association with good character and ethics (6. 2.8,11,13,18); with depiction of types, such as rusticos, 6.2.17, cf. 1.9.3, though not re-creation of their style, as in Arist. Rh. 3, 7, 1408a25 ff., cf. n. 34 above. We also find the association of Ḕthos with a ‘calm’ style, 6. 2. 13, identified with the ‘middle’ style, 6. 2. 19, as in Cicero; cf. n. 42 above. For his treatmentof pathos, cf. nn. 52-3 below. For the theme of the speaker's emotional self-involvement, cf. 6. 2. 25 ff., including the reference to actors, 35, and Cic. De Or.2. 189 ff.; for the link between emotional self-involvement and visualization, cf. 6. 2. 29 ff., Arist. Poet. 17, 1455a22–34, and n. 20 above.

⁵⁰ In De Or. 2. 182–4, the portrayal of good character is linked with a calm emotional mode (and with certain stylistic features, 2. 212–14), but these connections are argued for. In Quintilian, 6. 2. 10–19, the connections are assumed, and the various themes run into each other without clear demarcation.

⁵¹ De Or. 2. 178, 185, 189 ff., 206 ff.

⁵² The discussion of Ḕthos and pathos begins with this concern (6. 2. 1–7); but in his detailed discussion of pathos as well as Ḕthos his attention is on the orator, and his selection of the appropriate mode, rather than the impact on the audience (cf. refs. in n. 53 below; exceptions include comments in 6. 2. 9, adperturbalionem…praevalere, and 6. 2. 20–1, a brief attempt to distinguish the emotions we feel ourselves and those we induce in others). The topic of the emotional appeal to the audience is resumed in 6. 2. 23 ff., in discussing δεíνωσισ and visiones.

⁵³ 6. 2. 9 (concitatos…vehementer commotos, mitesque aique compositos); 6. 2. 12 (amor and caritas; cf. 6. 2. 17); stylistic indicators in 6. 2. 19–24. In Cic. Oral. 128, Ḕthikon and pathetikon could also be interpreted (and perhaps were interpreted by Quint.) as oratorical-emotional modes of this kind (cf. n. 41 above, on De Or. 2. 212, 216).

⁵⁴ 6. 2. 9, more and less violent emotions; 6. 2. 12, Ḕthos and pathos differ only in degree, cf. 6. 2. 17. Conceivably, the quidam who maintained that Ḕthos was perpetuum, while pathos was temporale, had psychological theory in mind; cf. Arist. EN 2. 5, EE 2. 2. But for Quintilian the concern is a practical one, relating to the emotional level of the speech; cf. Kennedy, The Art of Rhetoric in the Roman World, pp. 505–6, and for the underlying rhetorical issue Solmsen, op, cit. esp. pp. 400–1.

⁵⁵ cf. Hor. Ars P. 89 ff. for a similar association of intense emotion and high style with tragedy, rather than comedy (despite exceptions). Cf. Brink, op. cit. ii pp. 174 ff. and refs., esp. note on 1. 89, and Donatus' comment that Terence et morem ( =Ḕthos) retinuit, ut comoediam scriberet, et temperavit affection ( =pathos), ne in tragoediam transiliret.

⁵⁶ At least, the Odyssey is associated both with comedy (13, 1453a32–9) and with Ḕthos (24, 1459b 13–16); cf. n. 16 above.

⁵⁷ See Quint. 6. 2. 8–19; compare the Ḕthe, or type-portraits of rusticos, superstitiosos, avaros, timidos (6. 2. 17), with the portrayal of personae, such as rusticorum, irascentium, deprecantium, mitium, asperorum, ascribed to Menander in 10. 1. 71. For moral themes, see e.g. Men. Dys., Ter. Ad. On Menander's style, see Quint. 10. 1. 69; cf. Plu. Mor. 853 d-e.

⁵⁸ In Quint. 10. 1. 68, Euripides is described as in adfectibus mirus, esp. miseratione. In 6. 2. 35, he speaks of actors as inflaming our emotions falsis adfectibus (and as themselves fentes after a performance), but he mentions comoedos here as well as (tragic) histriones. Cf. n. 55 above.

⁵⁹ cf. ‘Longinus’, On the Sublime, ed. Russell, D. A. (Oxford, 1964), xxxvi–iiGoogle Scholar. (See further Russell', s ‘Longinus Revisited’, Mnemosyne 34 (1981), 72–86, on the place of pathos in the overall strategy of the treatise.) All subsequent refs. to ‘Russell’ denote his edn of ‘Longinus’, unless otherwise specifiedCrossRefGoogle Scholar.

⁶⁰ 1. 4. Cf. the contrast of ethical/emotional styles in Cic. Oral. 128, D.H., Dem. 8, Usener-Radermacher, 1. 143.

⁶¹ In Aristotle, as we saw, pathos was primarily associated, in the Rhetoric, with the impact on the audience, except in discussion of the emotional style; cf. n. 32 above. In Poet. 17, Aristotle develops the theme of emotional self-involvement in the writer that becomes a dominant theme in the orators; cf. nn. 20–1 above, and Russell, , Criticism in Antiquity, pp. 81–2Google Scholar.

⁶² See esp. 9. 10 ff. (Homer), 10 (Sappho), 15 (Euripides). Cf. discussion below.

⁶³ 13. 2–3; cf. Russell, notes ad loc. The process is sometimes described more prosaically: it resembles ‘the reproduction of good character (ἠӨν) in statues and works of art’, 13. 4; cf. Russell,ad loc.

⁶⁴ 15; cf. Russell, notes ad loc, Quint. 6. 2. 25–36, nn. 20, 21, 47 above.

⁶⁵ 15. 1–8. In 15. 2 (as in Arist. Poet. 17), the process is depicted rather as a relationship between writer and audience, though at 15. 8 the role (and the imagined viewpoint) of the dramatic character is brought in (Orestes πατξεται ταȗӨ’ οτι μαíνεται); cf. Quint. 6. 2. 32 and n. 20 above. Euripides is typically seen as the master of pathos; cf. Quint. 10. 1. 68, in adfectibus mirus. The theme goes back to Arist. Poet. 13, 1453 a 23–30; whereas Sophocles is the master of ethos, 25, 1460b33–5, 3, 1448a25–8.

⁶⁶ 15. 2, 15. 9, 15. 11 (κπληκτικΌν in rhetorical visualization); cf. Russell, note on 15. 2.

⁶⁷ See 18–22, 25–7; cf. 10. 5–6 and 39. 3.

⁶⁸ 17. 2; cf. 32.4: τ…σποδρ πάӨη καì γεαîο Ǔψοσ…constitute τια λεξιπρμακτῷ ῥοӨíω τσ πορσand 38. 5. Cf. William, K. Wimsatt, and Brooks, Cleanth, Literary Criticism: a short History (New York, 1957), pp. 100–1, 103–5Google Scholar.

⁶⁹ 29. 2; cf. Russell, adloc. and pp. xxx ff. In the last clause, ὕψοσ and δον seem to be the desired effects, pathos and ethos the contributory means. Ḕthos is hard to render properly here; Russell suggests ‘characterization’ (note ad loc), but this may be too specific. Longinus may have in mind the various elements, attractive character, calm emotional tenor, sweet style, that make the ‘ethical’ style a pleasant one.

⁷⁰ 34. 2. Again τῸ ἠӨι7kappa;Ὸν is difficult. Longinus seems to have in mind the self-presentation of the speaker as a good person, charming, amusing, etc., rather than the impersonation of different types of speaker.

⁷¹ 34. 3. νηӨοποΊητο‘without sense of character’, but the sense ‘unattractive’, which he also notes, seems the more relevant; cf. Zucker, op. cit. passim, esp. pp. 37–8. Demosthenes' style lacks Hypereides' ‘charm’ but instead it has intense emotional power.

⁷² 34.4. The emotion in Demosthenes gives power to the emotional effects (pathḔ), constitutes an antidote to any artificiality in his expression (cf. nn. 67–8), and so takes over the minds and emotions of the audience. Cf. 12. 3–5; Demosthenes, compared to Cicero, has more passion in him and so is more emotionally effective (ἅτε παӨητικώτεροκ.τ.λ.).

⁷³ 9.10. One might have expected Longinus to say the lines convey Ajax's Ḕthos (cf. the scholiast cited by Russell, adloc, ӨαμαστῸν τῸ ἤӨοσ). But it is the heroic pathos Longinus is interested in, the courageous impulse that demotes death below an honourable fight and puts Zeus in his place. Cf. the famous comment in 9. 7, “Oμηροσ γρ μοι δοκεî…νӨρώποσ ὅσο πí τῇ δνμ7epsiv;ι Өενὺσπεποιηκνατοὺσ δ νӨρώποσ.

⁷⁴ 9. 11, “Oμηροσ…οὔριοσ σνεμνεἶ τοἶσ γ7sigma;ι τ.λ.; cf. LSJ, οὔριοσ III, and Demosthenes’ ӨεΌπεμ7pi;τα δωρματα (34. 4). The idea that Homer is Өεἶοσ is, of course, traditional; see e.g. Horn. Il. 2. 484–92, Ar. Ran. 1033, PI. Ion 536a ff. For the idea that authors, including Homer, can be Өεἶοι, ἥρωεσ or ῚσΌӨεοι cf. 4. 4, 4. 6, 35. 2, 36. 2. In the passage quoted, Il. 15. 605–8, the raging Hector is acting under the direct encouragement, and with the help, of Zeus, 596–614.

⁷⁵ 9. 13; the term seems to pick u p the ‘wind ’ motif in οριοσ…σνεἶ (cf. previous note), and to carry also the suggestion of divine inspiration (cf. 13. 2, many writers λλοτρíω Өεοποροȗνται πνματι 33.5, κκεíνησ τσ κβολσ τοȗ αιμονíο πνεματοσ; and Russell, note on 9. 13).

⁷⁶ Similar sorts of qualities are ascribed to the ‘emotional’ Demosthenes' oratory; see 18. 1, μπρακτΌτερα (cf. πρακτικοȗ, 9. 14) καì σοαρώτερα σντεíνει τ λεγΌμενα…τΌ ἔνӨον καì ξρροτον κ.τ.λ. 20. 2, τῇ παλλλω…πορ, cf. the technique ascribed to him in 22. 3–4. Cf. also at 22. 1 the natural expression of those in the grip of pathos: τῇδε κκεîσε γχιστρΌπωσ ντισπώμενοι κ.τ.λ., and Russell,adloc.

⁷⁷ cf. 15. 1, “Oγκο καì μεγαληγορíασ καì γνοσ…αΊ παντασíαι παρασκεαστικώταται; 15.9. παντασíα…τοîσ λΌγοισ ναώνια καì μπαӨ προσεισπρειν…οὐ πεíӨει τῸν κροατν μΌνον, λλ καì δουλοȗτα. I take it Longinus has in mind the Homeric similes that are so prominent a feature of the Iliad; but perhaps he has in mind the various ways, in narrative and dialogue, Homer makes the action vividly real to us. Cf. the use of παντασíα in rhetoric, κλλιστον εì τῸ ἔμπρακτον καì νληӨε's preference, on ethical grounds, of narrative over dialogue (Resp. 396c and context). Plato, of course, distrusts the emotional involvement of the listener in expressed or represented emotions which Longinus values (Resp. 605 c ff.).

⁷⁹ For the connection between βíοσ, Ḕthos, and (Menandrian) comedy, see Russell, notes on9. 15, and nn. 55–7 above. In the later books of the Odyssey, in Ithaca, there is more scope for human interchange on a mundane, unimpassioned level than we find in the Iliad, of the kind often associated with the ‘ethical style’; cf. nn. 39, 49–50, 69–71 above. For subtleties in the psychological portrayal in the Odyssey, esp. in connection with women, see Griffin, , Homer on Life and Death, pp. 56–65Google Scholar. The Odyssey is also more ‘ethical’ in demarcating the good and bad Ḕthos of its figures (cf. n. 12 above), and thus shares the ethical concerns of some comedies (cf. n. 57 above); but it is not clear that Longinus has this in mind: see also discussion below.

⁸⁰ For pathos as a quality, or power, within the writer, cf. the images of the setting sun, and the Ocean ‘withdrawing into itself’ (9.13). Ḕthos also, in 9. 15, seems partly to denote a lower emotional level in the author.

⁸¹ cf. discussion above, and refs. in nn. 65 ff.

⁸² Poet. 24, 1459b13–16; cf. discussion above.

⁸³ Poet. 12, 1452b11–13, cf. n. 8 above;De Sublim. 9. 13, cf. the heroic violence of the pathetikon Iliad, 9. 10–12.

⁸⁴ cf. refs. in nn. 69–71 above. The ‘charm’ comes perhaps from the fabulous and incredible (9. 13–14), as well as the domestic and ‘comic’ (9. 15), neither of which makes intense, passionate demands on the listener.

⁸⁶ cf. n. 79 above, and contrast Aristotle's view, discussed above.

⁸⁶ Poet. 13, 1453a30–39, De Sublim. 9. 15.

⁸⁷ cf. discussion and nn. 14–16 above.

⁸⁸ cf. refs. in nn. 6, 12, 23 above.

⁸⁹ 44. 6–12; cf. 6–7, and Russell, pp. 185–6, xxii, xlii. See further Charles Segal, P., ‘“YψOσ and the problem of cultural decline in the De Sublimitate’, HSCP 64 (1959), 121–46Google Scholar.

⁹⁰ 8, 9, 10, 15, 35; hence, in part, Longinus' appeal in pre-Romantic modern Europe, cf. Russell, pp. xlv–xlvii.

⁹¹ cf. Segal, op. cit., esp. pp. 131 ff.

⁹² 9. 7, 10, 11, 13.

⁹³ Late nineteenth-century and Edwardian critics seem inclined to interpret questions of moral ambiguity in great art in Longinian terms, at least as far as tragedy goes. ‘We could wish that Aristotle had gone farther and said explicitly that in power, even more than in virtue, the tragic hero must be raised above the common level; that he must possess a deeper vein of feeling or heightened powers of intellect or will; that the morally trivial, rather than the morally bad, is fatal to tragic effect’, Butcher, S. H., Aristotle's Theory of Poetry and Fine Art ⁴ (London, 1904, repr. New York, 1951), p. 317Google Scholar; cf. Bradley, A. C., Oxford Lectures on Poetry (Oxford, 1909, repr. 1965), pp. 87–8Google Scholar.

⁹⁴ Rutherford, op. cit. 138–46; Lockwood, op. cit.; Zucker, op. cit.; Russell, , Criticism in Antiquity, pp. 154, 201Google Scholar.

⁹⁵ cf. nn. 33–4 above.

⁹⁶ In Dionysius of Halicarnassus, there is a further development: Ḕthos signifies the state of mind of the person responding to the ethical style; cf. n. 46 above.

⁹⁷ cf. nn. 32–3 above.

⁹⁸ The idea that the author will move the audience if he is emotionally moved himself goes back a long way; cf. nn. 21, 41, 49, 80 above.

⁹⁹ In rhetorical theory, the production of pathos is sometimes thought to depend on the speaker's successful presentation of his Ḕthos; cf. nn. 26, 41 above. Yet orators also admit, with a greater or lesser degree of frankness, that the emotional style can effectively transform an audience's view of the Ḕthos of the person and the rights and wrongs of the situation (see Arist. Rh. 3. 7. 4, Cic. De Or. 2. 194 ff., esp. 200–1, magis affectis animis iudicum quam doctis, Quint. 6. 2. 4–7, 23–4, De Sublim. 18. 2, cf. 22. 4).