Published online by Cambridge University Press: 28 February 2013
Thucydides' account of faction in the Peloponnesian War (3.82–83) is perhaps more often admired than appreciated; and indeed its structure and wording have not always satisfied commentators. Two kinds of question raised by the passage may be identified:
(1) What does it say and what is it about? We need to examine it as a coherent train of thought, and at the same time – for it is plainly not a mere digression or show-piece – as an essential part of Thucydides' whole work. Both the historian's practice and his programme suggest that the reader should do this. For in putting his thoughts in so highly generalized a form, he invites us to apply them widely. Even when they occur in speeches, where a fallible or even dishonest orator is trying to persuade an audience, general statements may be thus widely applicable; how much more when they come directly from the historian. We should also be ready to consider whether what are presented as remarks on stasis may not apply beyond situations of stasis. For war and faction as twin causes of violent death are closely connected in the proem to Thucydides' whole work (1.23.2); and 3.82.1–2, which echo the proem (1.22.4), bring out that the growth of stasis was a natural consequence of war and that in stasis, as in war, human nature is revealed.
1. With most editors, I take 3.84 to be an interpolation; for literature on this point, see Dionysius of Halicarnassus: On Thucydides, tr. and comm. Pritchett, W.K. (1975) 117Google Scholar.
2. Cf. Meister, C., Die Gnomik im Geschichtswerk des Thucydides (Diss. Zürich 1955) 82Google Scholar. For a valuable account of the wider relevance of Diodotus, ' sententiae in 3.45Google Scholar, see Müri, W., ‘Beitrag zum Verständnis des Thucydides’, MH 4 (1947) 251–75Google Scholar = Wege der Forschung 98 (1968) 135–70Google Scholar.
3. in 1.22.4 is a larger concept than in 3.82.2. But ‘human affairs’ must include ‘human nature’.
4. Cf., in general terms, Parry, A., ‘Thucydides' use of abstract language’, Yale French Studies 45 (1970) 3–20CrossRefGoogle Scholar. The most helpful detailed studies so far are Landmann, G.P., Die Friedensmahnung des Hermokrates (Diss. Basel 1932)Google Scholar; Kakridis, J. Th., Der thukydideische Epitaphios (1961)Google Scholar; Kohl, W., Die Redetrias vor der sizilischen Expedition (1977)Google Scholar.
5. The text of this sentence is sound. On the omission of after , see Poppo–Stahl ad loc.; Wackernagel, J., Vorlesungen über Syntax I (1926) 294Google Scholar. The missing main verb in the first limb of the sentence () is to be supplied by the reader. The ellipse is analogous to one familiar in double conditionals: i.e. in a sentence of the form ‘if A, then B; but if not / if C, then D’, the second element (‘then B’) may be omitted if the context makes it too trivial or too obvious to need stating (sometimes with powerful rhetorical effect, as e.g. in Iliad 1.135–9): cf. Thuc. 3.3.3 with Poppo–Stahl ad loc.; Kühner–Gerth II §577.4d. So here , makes it obvious what the consequence of the genitive absolute is, and what happened in peace – anyway, in effect nothing – is not the historian's concern. For similar ellipses in Thucydides see 1.35.4 (after ); 4.63.2 (after ); 6.11.2 (after ).
6. On this and contrasting ideas, see Dover, K.J., Greek popular morality (1974) 109–12Google Scholar, and add e.g. Eur., Ion. 1045–7Google Scholar; Isoc. 4.174, 7.4; Arist., Pol. 1265b12, 1334a25–8Google Scholar.
7. Cf. Thuc. 4.126.1; Plat., Gorg. 453 DGoogle Scholar; Ep. 7.332A; also and esp., Thuc. 5.86 (though there the word has overtones of compulsion). I take it that there is again a hint of oxymoron when in Democritus (B 76) , or in Diodorus Siculus (1.8.9) , is called . There is a similar oxymoron below in the word (3.82.7), which denotes what is in effect aggression, not self-defence.
8. See Xen., Anab. 2.6.12Google Scholar. I owe this reference to Dr. A.C. Cassio.
9. Cf. 1.78.2, 82.6, 84.3, 140.1; 2.11.4; 4.18.4. This is not to say that in Thucydides man has to be the plaything of circumstances; see Edmunds, L., Chance and intelligence in Thucydides (1975)CrossRefGoogle Scholar; Herter, H., ‘Freiheit und Gebundenheit des Staatsmannes bei Thucydides’, RM 93 (1950) 133–53Google Scholar = Wege der Forschung 98 (1968) 260–81Google Scholar; idem, ‘Thucydides und Demokrit über die Tyche’, WS N.F. 10 (1976) 106–28. The locus classicus is 2.60–65 (cf. 1.140.1).
10. See L. Edmunds (n.9) 1–2. Further, Plat., Gorg. 448CGoogle Scholar and Dodds ad loc; Legg. 889 B–D.
11. Cf. 1.41.3, where the Corinthians' sententia is true, but also reveals that the argument it is meant to support is doomed to fail.
12. See Dodds on Plato., Gorg. 490E 9–11Google Scholar; also Isoc. 13.12; cf. Ar., Vesp. 1044Google Scholar; Nub. 547.
13. Cf. Radermacher, L., Artium Scriptores C 72Google Scholar; [Cic., ] Ad Her. 3.3.6Google Scholar; Quintil. 3.7.25. It may well be that it was partly rhetorical theory and practice which suggested such a formulation to Thucydides, as to Plato, (Rep. 560 DGoogle Scholar; Gorg. 491 E 2 and Dodds ad loc), Isocrates (7.20; 12.131; 15.284) and the Anonymous Iamblichi (6.1). In general on 3.82.4–5, see Müri, W., ‘Politische Metonomasie’, MH 26 (1969) 65–79Google Scholar.
14. See Radford, R.S., Personification and the use of abstract subjects in the Attic orators and Thucydides (1901) 31–2Google Scholar.
15. For war personified with this implication, see Hdt. 7.144; Isoc. 8.19; Dem. 18.89; Aeschin. 3.148 (here as ‘teacher’). Note also the bon mot of Hyperides quoted and discussed by [Longinus, ] De subl. 15.10Google Scholar. Cf. Radford (n.14) 36; Denniston, J.D., Greek prose style (1952) 32Google Scholar. Particularly striking is Plat., Legg. 709AGoogle Scholar.
16. Cf. de Romilly, J., ‘Thucydide et l'idée du progrès’, ASNP 35 (1966) 158–67Google Scholar. On the limits of Thucydides' idea of progress, see den Boer, W., ‘Progress in the Greece of Thucydides’, Med. der kon. Ned. Akad. van Wet., afd. Lett., 40.2 (1978)Google Scholar.
17. Thucydides watchfully observes and carefully describes the innovations that the Peloponnesian war made in military and naval skill or technology: see e.g. 2.75–6; 4.100; 7.36; in general, cf. 1.71.3, 6.18.6, and see further de Romilly (n.16), 167–70. In general for comments in contemporary literature on how have developed, see Dodds, E.R., The ancient concept of progress (1913), 11–12Google Scholar; also Isoc. 9.7; Dem. 9.47.
18. The motives of , and (1.75.3, 76.2) and the characteristic Athenian (1.70) are given powerful expression, with terrible consequences, in 6.17–18; see further on that speech as an interpretation of imperialism, ‘Rhetoric and history: Thuc. 6.16–18’, Quaderni di Storia 2 (1975) 54–62Google Scholar. In general, cf. de Romilly, J., The rise and fall of states according to Greek authors (1977) chs. 1–3, esp. p. 18Google Scholar.
19. On this idea, see further de Romilly (n.16), 154–8, with ample documentation. So even if Diodorus' immediate sources are not pre-Hellenistic, clearly ‘necessity the mother of invention’ is familiar to Thucydides; see also Burton, A., Diodorus Siculus Book I: a commentary (1972) 49CrossRefGoogle Scholar.
20. Cf. Aesch., P.V. 436–506Google Scholar; Soph., Ant. 332–63Google Scholar; Hippocr., Vet. Med. 3Google Scholar.
21. Cf. [Critias] D.-K. B 25, a fragment now rightly ascribed to Euripides by Dihle, A., ‘Das Satyrspiel “Sisyphos”’, Hermes 105 (1977) 28–42Google Scholar; Moschion frag. 6.27, 30 N.; Anon. Iambl. 3.6.
22. See also Arist., Pol. 1252b 29–30Google Scholar: On sophistic as ‘life-saving’ see Heinimann, F., ‘Eine vorplatonische Theorie der ’, MH 18 (1961) 118–9Google Scholar.
23. Twelve words for ‘die / death / kill’ in seventeen lines of O.C.T. For such massive repetition, cf. above all 2.64.3: amidst and against which is set the sobering
24. Also Aesch., P.V. 443–4Google Scholar; Soph., Ant. 354–5Google Scholar; Anaxag. B 21b; Xen., Mem. 4.3.11Google Scholar. But Theseus in Eur. Supp. and the Chorus in Soph. Ant. go on to envisage, like Thucydides, the errors of .
25. Also Soph., Ant. 354Google Scholar; Lys. 2.19; Plat., Protag. 322AGoogle Scholar; Xen., Mem. 4.3.12Google Scholar; Isoc. 4.47–9.
26. By Edmunds, L., ‘Thucydides' ethics as reflected in the description of stasis’, HSCPh 79 (1975) 73–92Google Scholar; for contrasts between Hesiod and Thucydides, see there pp.86, 91.
27. For analogies in Solon, see Ryffel, H., Metabole politeion (1949) 16–22Google Scholar. For a contemporary account of which has much in common with Thucydides, see Anon. Iambl. 7.1–9.
28. This word either means ‘amplius deliberare sive consultare’ (Poppo–Stahl) or is equivalent to the simple (Steup, , Anhang, p.274Google Scholar). It is used in this unique way to reinforce the paronomasia, which also makes the use comprehensible; any attempt to emend it is therefore misguided.
For the word-order of , see, besides Gomme ad loc, 6.77.2 with Classen-Steup's and Dover's notes. For the sense of (‘with a view to safety’, ‘without taking risks’) Stahl compares 3.56.5, where the word is opposite to note also 1.69.5, where , = ‘safety-minded’, ‘prudent’.
29. For similar plays on words, see e.g. Il. 21.523–4; Od. 8.523–4; Thuc. 2.62.3; 3.39.2; 7.15.1 Further, [Cic., ] Ad Her. 4.30Google Scholar and Caplan ad loc.
30. On this kind of word-play, see ‘Reason and necessity: Thuc. 3.9–14, 37–48’, JHS 98 (1978) 66 n.8Google Scholar, and add Aesch., Ag. 1630–2Google Scholar; Hdt. 1.11.2; Gorg., Hel. 14Google Scholar; Thuc. 1.118.2; Plat., Lach. 187E 9–10Google Scholar; Tim. 55D 1; Dem. 9.17–18.
31. In general on the word , and cognates in Thucydides, see L. Edmunds (n.9), 9–10; Huart, P., Le vocabulaire de l'analyse psychologique dans l'oeuvre de Thucydide (1968) 279–90Google Scholar.
32. Admirable on these is Schneider, C., Information und Absicht bei Thukydides (1974)CrossRefGoogle Scholar.
33. Cf. Plat., Lach. 197 A–BGoogle Scholar; Prot. 360 D. A Socratic influence on the historian, as in the Melian Dialogue (cf. ‘Form and meaning in the Melian Dialogue’, Historia 23 [1974] 389–91Google Scholar; GRBS 18 [1977] 233 n.12Google Scholar)?
34. Darius' speech begins by deriding speech: there is an irony here – the ‘man of action’ is, and needs to be, a master of words (both explanations to his comrades and deceptions of his enemies); cf. Thuc. 4.10 and Historia 23 (1974) 394Google Scholar. Sophistic influence is also strong. The attempt to prove that lying and truth-telling are in effect the same recalls the pattern of argument in the Dissoi Logoi (D.-K. 90) 1–4. For the principle that gain is the motive of all human action, see Radermacher, L., Artium Scriptores C 46Google Scholar; for the aim as the differentia of actions, see Isoc. 4.130. Note also Dihle, A., Die goldene Regel (1962), 98–9Google Scholar; ‘Herodot und die Sophistik’, Philol. 106 (1962) 217Google Scholar. But though the sophistic influence stands out, as does an interest in ethical theory, Herodotus also has a good historical reason for inventing such a speech: the Persians had special objections to lying (cf. 1.136.2) which could be thought to need overcoming; and thus Darius displays the typical and of the stasiotes (cf. Thuc. 3.82.6, 8). Similarly, Darius' speech in 3.82, another piece influenced by Greek sophistic, concludes with an argument tailored to a Persian audience: that monarchy is best as the country's traditional and well-proven form of government (cf., for this as part of the apparatus of Greek rhetoric, Thuc. 6.18.7; Plat., Gorg. 513 B–CGoogle Scholar; Arist., Rhet. 1365b 21–1366a16Google Scholar; and for the use of constitutional ideology in gaining power, Thuc. 3.82.8). So sophistic or rhetorical art converges with historical imagination in Herodotus' speeches, as in Thucydides'.
35. On the prevalence of and other themes in 82–83 in connection with Euripides' Orestes, see the thoughtful and balanced remarks of Longo, O., ‘Proposte di lettura per l'Oreste di Euripide’, Maia 27 (1975) 265–87Google Scholar.
36. Cf. 1.22.4; 3.38.2–4 Gorg., Hel. 13Google Scholar.
37. The author of ch. 84 clearly noticed this (though he was also inspired by 3.45).
38. The arrangement of the words favours this translation: we should expect the order of as adjectival clause and as predicate to correspond to the order of chiasmus could be a wilful obfuscation here. (The sentence is worded deliberately in such a way that the reader has to tease out its meaning; the more important, then, that there be no arbitrary difficulties in the puzzle.) Gomme objects, with Dionysius: ‘Why should the stupid be thought good?’ But Thucydides says this is less common, implying downright uncommon (it is characteristic of him to say 'x rather than 'y, when y is in effect negated; cf. 1.34.2, 140.2; 3.37.2; 64.3, 65.2); and that was surely as true in his time as it is in ours.
39. Cf. Lysias frag. 75 Thalheim ; Plat., Theaet. 176DGoogle Scholar (sc. ). Also Hume, A treatise of human nature, Book III, Part iii, Section iv: ‘There are few, who are not as jealous of their character, with regard to sense and knowledge, as to honour and courage; and much more than with regard to emperance and sobriety. Men are even afraid of passing for good-natur'd; lest that shou'd be taken for want of understanding.’
40. For these notions in this kind of context, cf. Anon. Iambl. 6.1; Eur., Phoen. 509–10, 531–2Google Scholar: on that play in connection with Thuc. 3.82–83, see de Romilly, J., ‘Les Phéniciennes d'Euripide ou l'actualité dans la tragédie grecque’, RPh 39 (1965) 28–47Google Scholar. , coming after , is a kind of paronomasia (for a more explicit play on the two senses of , see Plat., Polit. 305C2Google Scholar; Isoc. 4.119), which brings out how basic a motive is. and its motives, which include and (cf. e.g. 1.75.3, 76.2; 5.97), are of course a central theme of Thucydides' whole work.
41. See further Ryffel (n.27).
42. See e.g. 2.39, 65; 3.37–38, 42–43; 6.16–24; 7.55.2.
43. Cf. Hdt. 3.82.3; Isoc. 3.18; Xen., Mem. 2.6.20Google Scholar; Arist., Pol. 1306a 12–19, 1306b 22–7Google Scholar.
44. Classen–Steup's rather vaguely formulated objections (p. 173) to the ‘repetitions’ in 83.1 and elsewhere in the passage are thus misguided.
45. On the exploitation of oaths, legal processes and constitutional values, cf. GRBS 18 (1977) 232–3Google Scholar, and passim. In the narrative about Corcyra, see 3.70.5–6, 81.4.
46. Cf. the specious use of the catchword ‘freedom’: see Diller, H., ‘Freiheit bei Thukydides als Schlagwort und als Wirklichkeit’, Gymnasium 69 (1962) 189–204Google Scholar = Wege der Forschung 98 (1968) 639–60Google Scholar.
47. Cf. Arist., Rhet. 1358b 13–17Google Scholar. For the near-identification of justice and the public interest, cf. Arist., E.N. 1160a 11–14Google Scholar: .
48. On this notion, see esp. Plat., Legg. 679 B-EGoogle Scholar, and cf. Thuc. 2.37.3 on ‘fear … of the unwritten laws’. Note that for Thucydides it is an essential part of civilized co-existence, whereas for Plato it is typical of a time before came into being. For opposed to , cf. 2.40.5.
49. See further Huart (n.31), 336–44.
50. Cf. Ep. ad Amm. 3, where ; is an example of a ‘poetic’ word.
51. Cf. Ep. ad Amm. 17.
52. See esp. Rep. 348 C11-D6; Gorg. 491 E2 and Dodds ad loc.
53. It does not follow from his paraphrase that Dionysius read in his text of Thucydides, as Gomme claims. He seems rather to have rightly taken as subjective genitive with (‘benefit caused by the established laws’) and quietly smoothed out the asymmetry of with Cf. Ros, J., Die Metabole als Stilprinzip des Thukydides (1938) 178Google Scholar; his alternative rendering (‘zum Nutzen der Gesetze’) is perhaps possible, though I cannot parallel the notion of ‘benefiting’, as opposed to ‘defending’ or ‘respecting’, the laws.
54. Note esp. Polyb. 13.3.2–3, a passage adduced but not fully exploited by L. Edmunds (n.26) 87 n.40. Cf. e.g. Thuc. 3.43.3.
55. See Ricks, C., Milton's grand style (1963)Google Scholar.
56. Berni, F., ‘Capitolo a fra Bastiano del Piombo’, line 31, in Rime (1969) 179Google Scholar. ‘Voi’ are the imitators of Petrarch.
I owe valuable criticisms of an earlier draft to Drs. D.C. Innes, D.M. Lewis and C.B.R. Pelling.