Skip to main content Accessibility help

The social function of Attic tragedy: a Response to Jasper Griffin

  • Richard Seaford (a1)


Jasper Griffin's polemic, in this journal, against what he calls the ‘collectivist school’ of interpretation of Athenian tragedy is welcome, as it encourages clarification of fundamental differences. I do not have the space here to tackle him wherever I think he is wrong, still less construct an argument to the effect that Athenian tragedy was a ‘collective’ phenomenon. Rather I want to do two things. Firstly, the casual reader may have formed the impression that whereas the ‘collectivists’ operate with vague and unsubstantiated notions, Griffin's view has the advantage of being firmly grounded in the ancient texts. This impression I intend to dispel. In doing so I will confine myself to some of G.'s general remarks and to his attack on my own views, as a sample of the quality of his argument. Secondly, I also adduce new material in the hope of advancing the debate on this important issue.



Hide All

1 Griffin, J., CQ 48 (1998), 3961.

2 This extends even to the supposed political basis of the ‘collectivists’. First (40) we have the cliché that Marxism lingers on only among a few benighted intellectuals, but later (61) are warned about thinking that an ancient work chimes with our liberal ideas about the state.

3 We may even see the caricature in the process of unfolding. Quotes, G. from my Reciprocity and Ritual: Homer and Tragedy in the Developing City-State (Oxford, 1994), 153–4 the following ‘If that is so, then among the qualities that ensured the predominance of the Iliad and Odyssey over other possible versions of themselves and over the epic cycle was their exceptional embodiment of the aspirations of the early polis’, and on the very same page (49) paraphrases it as follows: ‘we recall that for Seaford the ascendancy of the Homeric poems was because of their “exceptional embodiment of the aspirations of the early polis”’. The disappearance of the words ‘among the qualities…’ creates an absurdly simplistic notion that is easy to dismiss.

4 This is welcome evidence that G. is prepared (contrary to the impression he sometimes gives: see below) to go beyond the Greeks’ own terms in interpreting their literature.

5 Svenbro, J., La Parole el le Marbre (Lund, 1976), esp. 31–5.

6 On this scene Dover, K. J. writes of ‘Aristophanes’ readiness to caricature both sides in a debate’ (Aristophanes Frogs [Oxford, 1993], 17) and that ‘comparison with Clouds indicates that Aristophanes has assimilated the contrast between Aeschylus and Euripides to the generalized contrast between old and new’ (22).

7 See esp. 1008–10, 1419, 1561–2 ‘save the polis’.

8 The theatre as reconstructed by Lykourgos is thought to hold between 14,000 and 17,000. Plato's 30,000 (Symp. 175e) for the earlier theatre may be an exaggeration. Even allowing for the presence of foreigners, metics, women, and slaves, it seems very likely that the audience contained many thousands of politically active male citizens. In the fourth century the number of adult male citizens is estimated at 20,000 or 30,000. For the early fifth century Hdt. (5.97.2) suggests (perhaps as a conventional number) 30,000. A high estimate for the year 432/1 is 60,000: M. Hansen, Three Studies in Athenian Demography. Historisk-filosoflske Meddelelser 56 (Copenhagen, 1988).

9 G. may have felt forced to make this concession by Pl. Ion 535–6, in which the rhapsode sees his vast audience weeping, amazed, and held (as it were) by a magnetic power that originates from the god and passes to them through poet and rhapsode.

10 Has this to do with the isolation of enjoying Greek poetry in the late twentieth century? G. seems to feel that the particularity of (his?) pleasure in poetry needs to be defended against the collective, the historical, and even the explicable.

11 Gorgias says, in G.'s own translation, ‘those who listen to poetry experience the shudders of fear, etc’ (Hel. 9). Plato says of Homer and the tragedians that ‘their audiences “take delight in their displays of pathos…”’ (Rep. 605cd); here ‘their audiences’ is G.'s paraphrase (though Plato actually says ‘the best of us…’—which in the context means even the best of us). Of Aristotle, G. refers to passages from the Poetics that refer to tragedy as producing pity and alarm and pleasure (with no mention of the audience).

12 For this anthropologically based argument see n. 62 below.

13 Aristotle regards the performance of tragedy as relatively unimportant (Poetics, ch. 6). For a historical explanation of the absence of the polis from Aristotle's account of tragedy, see Hall, Edith in Silk, M. S., Tragedy and the Tragic (Oxford, 1996), 295309.Salkever, S. G., on the other hand, argues that ‘Aristotle’s discussion presupposes the thought that tragedy is not only a political institution, but more particularly a democratic one’ (in Euben, J. P. [ed.], Greek Tragedy and Political Theory [Berkeley, 1986], 274303).

14 See J. Henderson in K. Morgan (ed.), Popular Tyranny (Texas University Press, forthcoming).

15 Note e. g. the ‘tyrannical’ behaviour of ‘the thirty’: Xen. Hell. 2.3.16, 4.2.1.

16 On this point, and in general, see Taylor, M. W., The Tyrant Slayers. The Heroic Image in Fifth Century BC. Athenian Art and Politics (Salem, NH, 1991).

17 Eating at public expense in the Prytaneum: see e.g. W. E. Thompson in AJP 92 (1971), 226–37; Rhodes on [Arist.] Ath.Pol. 24.3.

18 Something similar was probably also part of the oath taken by members of the Boule, by the mid-fifth century if not the late sixth, according to Ostwald, M. in TAPA 86 (1955), 103–28.

19 Perhaps the one cited at Ath.Pol. 16.10: see Rhodes ad loc; Lavelle, B. in Class, et Med. 39 (1988), 35–7.

20 Meiggs, R. and Lewis, D., A Selection of Greek Historical Inscriptions to the End of the Fifth Century B.C. (Oxford, 1969), no. 40, lines 32–4.

21 [Aristot.]Ath.Pol. 22. 3.

2 Aristophanes (Birds 1074–5) parodies the anachronism of this practice with the words ‘anyone who kills any of the dead tyrants’. The decree would have been appropriate early in the century.

23 In which I am heavily indebted to K. Rauflaub ‘Stick and glue: the function of tyranny in fifth-century Athenian democracy’, in Morgan (n. 14).

24 This cannot be attributed merely to the greater metrical convenience of τúραυυoς.

25 I set these respects out in ‘Tragic tyranny’ in Morgan (n. 14).

26 That he really does believe this is confirmed by his seriously telling us that even Euripides’ Suppliant Women, which he admits may have some sort of political motive, also ‘contains things of a very different tendency’. Well, of course it does—even though the category of ‘almost pure pathos’ that he applies to the suttee of Evadne and the grief of her father does not exactly shed a flood of light on the passage (except perhaps for the clinically insensitive), and may mislead those unaware of the complex way in which the scene relates to Greek ideas of the wedding (which are quite different from our own).

27 The evidence is too plentiful even to summarize here. How G. can claim that the ‘great tragic theme’ of leaving the dead unburied was ‘not political but religious in character’ (61) is beyond me (it is of course both), especially as he also writes that tragedy ‘loves scenes in which tyrannical power forbids the burial of the dead and is defied’ (56).

28 Though oddly the only evidence that he gives for it is from Herodotus, who was closely associated with democratic Athens.

29 G. does not consider my answer, but he does list striking events, some of them of the kind represented in tragedy, in Herodotus, Thucydides, and Xenophon. ‘What I am concerned with’, he writes (57), ‘is the general character of the political world in the fifth century, full of such episodes, and (at least equally important) populated by people who saw history in terms of such patterns and such dramatic confrontations.’ But the political world contains more than just ‘episodes’. For example, G. notes, appositely, that at Corcyra ‘father slew’ son (Thuc. 3.81), but does not observe that such (non-Homeric) horrors occurred, according to Thucydides, in conflict for control of the polis. Similarly, G. claims that the Athenians saw their history as determined by great personalities such as Pericles (58). But actually Pericles himself stresses (to the Athenians) the importance of various factors, notably money—as do others in Thuc. (2.13; cf. e.g. 1.80.3–4, 1.83.2, 1.121.2, 1.141.3, 6.34.2). Cf. e.g. A. Ag. 1638–9 (unthinkable in Homer). Understanding tragedy requires us to understand the polis, for which events described by Thucydides, Herodotus, and Xenophon are very limited evidence. Nor should we disdain to inform ourselves of the general understanding of pre-modern societies that has made some advance in the last two hundred years or so.

30 Against the link between democracy an d tragedy G. advances the possibility that tragedy was first performed under tyranny (47) and the fact that it was imitated by the Romans in their ‘eminently undemocratic city’ (60). But this argument works only if we assume that the political complexion of tragedy was always the same. In fact the politics of the only extant Roman tragedy (Seneca) are (precisely as those attacked by G. would predict) quite different from Athenian tragedy.

31 See esp. Rauflaub (n. 23).

32 Rauflaub cites Alcibiades’ justification to the Spartans of his aristocratic family's record of democratic leadership: ‘My family has always been opposed to tyranny. The people (δῆμoς) is the name given to any force that opposes δυυαστεíα (despotic power)’ (Thuc. 6.89).

33 Guépin, J. P., The Tragic Paradox (Amsterdam, 1968), 8990; Vernant, J. -P. in Vernant, J. -P. and Vidal-Naquet, P., Myth and Tragedy in Ancient Greece (Sussex, 1981), 87119.

34 It is perhaps curiously necessary to point out that this does not imply the absence of human consciousness.

35 ‘The tomb of Aias and the prospect of hero cult in Sophokles’, CA 12 (1993), 165–80. The argument is in some respects anticipated by Burian, P.Supplication and hero cult in SophoclesAjax’, GRBS 13 (1972), 151–6.

36 Here G. misleads. Of Aeschylus the only final plays of trilogies we have are Sept. and Eum.: at the end of Sept. hero-cult is indicated (1002–3; cf. Pausan. 9.18.3), and in Eum. cult for Orestes is predicted and cult for the Furies is founded. See also Soph. O.C.; E. Ale (445–52), Med. (1381–3), Held. (1031–6), Hipp. (1423–30), Suppl. (1196–1212), H.F. (1328–33), I.T. (1456–67), Hel. (1666–9), Pho. (1703–7). G. claims that no cult is predicted at the end of Eur. Ale., El., Andr., and Hec But in El. he has failed to notice 1270–2; in Ale the cult of Alcestis is predicted not at the end of the (prosatyric) play but before her return from the dead; and both Andr. and Hec. end with the establishments of tombs, at which in the former case we know there was an important hero-cult (see below). In Ba. the cult of Dionysos must have been established by the god in the lost part of his final speech. The ending of I.A. is spurious, and of Tro. exceptional as involving the destruction of the whole (exceptionally, non-Greek) community, which cannot therefore perform cult. Of Euripides’ extant tragedies that leaves only Ion and Or., each of which ends with the aetiology not of hero-cult but of the naming (of the Ionians, of the the town Oresteiori) after the hero. About his lost plays we can say little, except for the founding of hero-cult in the Erechtheus.

37 In Sophocles: Second Thoughts (Göttingen, 1997), 137, H. Lloyd-Jones and N. Wilson attempt to defend (against my argument, n. 3, p. 135) the conjecture υὑξ (Martin) by citing A. Cho. 65 τoùς ἂκρατoς (Schütz; ἂκραυτoς M) ᾤΧει υúξ (in a context of punishment), which however (on either reading) has the opposite sense to that given by Martin's conjecture to the OC passage.

38 Perhaps G. thinks that the presence of Oedipus’ body (without cult) will be enough. But it is through cult that benefit from the heroic dead (e.g. Theseus in the Theseion) is maintained: Burkert, W., Greek Religion (Oxford, 1985), 203–8. And indeed cult for Oedipus is clearly indicated at 1526–34 (cf. 624, 1642, 1644 τά δρώμευα with Kamerbeek ad loc), though its content must remain secret. The point of secrecy (cf. e.g. Eur. Held. 1041–2) was presumably to deny to enemies the chance to obtain the goodwill of the hero through offerings at his tomb, as is made explicit at E. Erechtheus 65.87–9: Kearns, E., The Heroes of Attica. BICS Suppl. 57 (London, 1989), 51–2. Hero-cult for Oedipus at Colonus is mentioned at Pausan. 1.30.4.

39 Although (against G.) I agree with Harrison (JHS 109 [1989], 173–5) that in Phil, it is hinted at at 1418–22: see further Seaford (n. 3), 138.

40 See e.g. Easterling, P. E., Sophocles Trachiniae (Cambridge, 1982), 910.

41 Kron, U., Die Zehn Attischen Phylenheroen. AM Beiheft 5 (Berlin, 1976), 172–6; Kearns (n. 38), 82, 141; Shapiro, H. A., Art and Cult under the Tyrants in Athens (Mainz am Rhein, 1989), 154–7.

42 Eur. Ale, which refers to both Sparta and Athens, I count under the latter.

43 At Trozen this was the cult of Hippolytus, who also, however, had an Athenian hero-cult alluded to at Hipp. 29–33. The cult foretold in Hel. is not specific to any particular place.

44 C. Carey in his A Commentary on Five Odes of Pindar (Arno Press, 1981), comparing our passage of Pindar with the tradition that the Delphians were at this time hostile to Neoptolemus, is puzzled, because unaware that to give hero-cult (in appeasement) to the enemy you have slain is standard practice: see e.g. M. Visser ‘Worshipping your enemy: aspects of the cult of heroes in ancient Greece’, HTR 75 (1982), 203–28. The Athenians, to judge from Eur. Andr., may have isolated this element of the cult as a reproach to the Delphians.

45 Seaford (n. 3), 93, 130–1, 311–8.

46 Ibid., ch. 8; id., Euripides Bacchae (Warminster, 1996), 44–52.

47 Hutchinson in his commentary deletes these lines. But cf. Seaford (n. 3), 347, n. 57.

48 The order in which these lines are spoken is a textual problem that does not concern us here.

49 See e.g. 764–5 (quoted below), 923, and the plot as a whole.

50 But the very same metaphor is in fact used by the messenger (795–6) to say that (after the battle) the city now is in clear water!

51 689–91, 720, 801, 813, 828 (retain ảτέκυoυς see n. 56 below), 877, 880–2, 954–5.

52 Along with most critics, I accept the overwhelming evidence that Sept. 1005ff. are interpolated (see Hutchinson's commentary ad loc). My remark (n. 3, 347) that ‘it is of course well known that the family and its sufferings were not yet at an end’ implied too great a concession.

53 As Hutchinson notes ad loc., there are verbal echoes here of the earlier description of the oracle.

54 In the lines quoted above. I believe that 820–1 should be printed after 801–2, but this problem makes no real difference to my argument.

55 On safety for the group by relinquishing a member as a basic pattern of Greek and other religions, see Burkert, W., The Creation of the Sacred (Harvard, 1996).

56 At 828 Hutchinson argues correctly that ảτέκυouς cannot mean ‘unfortunate in their birth’, but says that the notion of childlessness appears ‘much too abruptly here’. But the chorus in this stanza are responding to what the messenger has just said, which includes the fact that the clan has ‘indeed’ been destroyed!

57 The aorist διήρκεσαυ (842) implies that the chorus are (not unnaturally) thinking of the death of the brothers (the present ảμβλúυεται [844] then expresses a general principle).

58 Septem I have dealt with above. In O. T. the destruction of Oedipus’ household is powerful and complete enough to marginalize any sense of continuity in the succession of Creon, who as Jocasta's brother is from a different household.

59 Suppl. 439; cf. e.g. Pho. 560 πýτερα τυραυυεîυ 祰 πýλιυ σŵσαι θ⋯λεις;

60 see Zeitlin, F., ‘Thebes: theater of self and society in Athenian drama’, in Winkler, J. and Zeitlin, F. (eds), Nothing to Do with Dionysos? Athenian Drama in its Social Context (Princeton, 1990), 130–67.

61 For the audience, see esp. Pl. Rep. 606a.

62 This effect is ‘collective’ also in the sense that it is antithetical to the potentially anti-social effect of the same shared emotion when confined to the bereaved family. See my anthropologically grounded arguments (n. 3, 86–92, etc., ignored by G.) for the unifying effect of collective lamentation.

63 Visser (n. 44); Seaford (n. 3), ch. 4.

64 See n. 33 above, and J. Bremmer, ‘Scapegoat rituals in ancient Greece’, HSCP 87 (1983), 299–320. The ritual was performed in Athens.

65 For ambivalence towards the tragic hero, see e. g., in Aesch. Ag., the chorus’ clearly implied disapproval of the king as bad for the community (62, 447–9, 456–62, 763–84) with their subsequent grief at his death (1489–96,1538–50). On the ambivalence of the Oresteia towards its great individuals, see further Griffith, M., ‘Bright dynasts: power and politics in the Oresteia’, Classical Antiquity 14 (1995), 62129.

66 On the Athenian tyrants’ benefits for the polis, see e. g. Hdt. 1.59.6, Thuc. 6.54.5.

Recommend this journal

Email your librarian or administrator to recommend adding this journal to your organisation's collection.

The Classical Quarterly
  • ISSN: 0009-8388
  • EISSN: 1471-6844
  • URL: /core/journals/classical-quarterly
Please enter your name
Please enter a valid email address
Who would you like to send this to? *


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed