2. Though Jamie Masters might contest this for Lucan: see Poetry and Civil War in Lucan's ‘Bellum Civile’ (Cambridge, 1992), 216–59Google Scholar.

3. See Hardie, P., ‘Closure in Latin epic’, in Dunn, F. M., Fowler, D. P., Roberts, D. H. (eds.) Classical Closure (Princeton, forthcoming)Google Scholar. I am most grateful for early sight of this paper.

4. See Vessey, D. W. T., ‘Pierius menti calor incidit: Statius' epic style’, ANRW 2.32.5 (1986) 2965–3019Google Scholar, on ‘The Aeneid enclosed’ (2966–7), e.g. ‘The Aeneid is written into the Thebaid.’

5. Cf. Aen. 8.494 omnis furiis surrexit Etruria iustis, of justified anger directed against a tyrant.

6. A point made by Hardie, P., The Epic Successors of Virgil (Cambridge, 1993), 46Google Scholar.

7. Of course, Elena Theodorakopoulos is probably right to say that it may be impossible to be a purely passive reader. She observes that Theseus' promise of burial to Creon could be seen as a resolution of the ambiguity at the end of the Aeneid about whether or not Aeneas accedes to Turnus' request that he be returned to his father for burial, et me seu corpus spoliatum lumine mauis | redde meis (935–6).

8. Cf. Hardie, ‘Closure in Latin epic’.

9. The word ‘supplement’ is not without problems, of course, in that it might suggest something additional and not essential.

10. Thanks to John Henderson for this observation (and many others).

11. manus miscere is usually ‘to join battle’: OLD misceo 3b.

12. Perhaps there is a play here with the etymolgy of triumphus from thriambos: see Maltby, R., A Lexicon of Ancient Latin Etymologies (Leeds, 1991)Google Scholar; Versnel, H. S., Triumphus: an enquiry into the origin, development and meaning of the Roman triumph (Leiden, 1970) 11–15Google Scholar.

13. Hardie, , Epic Successors 48Google Scholar.

14. The depth of the women's grief may help account for the potentially alarming note in the Bacchant simile at 792–3 where the women are described as amentes, magnum quas poscere credas ∣ aut fecisse nefas (‘crazed, maybe bent on or fresh from an atrocity’): thus Statius indicates the unrestrained emotions that cannot be relieved.

The close of the Iliad is, of course, significantly different. In Iliad 24 the three women's laments (723–75) represent a reprise of major themes of the poem, with Andromache emphasising past and future suffering for the Trojans, with Hecuba both expressing grief and celebrating the favour shown by the gods and with Helen praising Hector's kindness to her. Yet the close of the Iliad is not closural, because of the emphasis on the temporary nature of the twelve-day truce for burial (778–81) and the Trojans' fear that the Greeks will attack them (799–800), which looks ahead to the renewal of the fighting (and ultimately to Achilles' death). See Macleod, C. W.'s commentary on Homer Iliad 24 (Cambridge, 1982) on 24.723–76Google Scholar.

15. On nautical imagery see Curtius, E. R., European Literature and the Latin Middle Ages (New York, 1953), 128–30Google Scholar; on the open sea as a metaphor for epic poetry see Wimmel, W., Kallimachos in Rom. Die Nachfolge seines apologetischen Dichtens in der Augusteerzeit (Hermes Einzelschriften 16, Wiesbaden 1960), 227–30Google Scholar and e.g. Virg. Georg. 2.39–45 with Thomas's notes (Thomas, R., Virgil: Georgics I–II (Cambridge, 1988)Google Scholar. For a nexus of ideas and texts linking completion of a sea voyage with completion of a poem, see Hardie, A., Statius and the Silvae: poets, patrons and epideixis in the Graeco-Roman world (Liverpool, 1983), 163Google Scholar with n. 25 on Stat. Silv. 3.2 and Hor. Od. 1.3: ‘Statius may have thought Horace was talking about Virgil's composition of the Aeneid, and that safe arrival in Greece was the allegorical equivalent of completion of the epic poem.’

16. On the unusual nature of this sermo addressed by poet to his creation see Vessey, , ‘Statius' epic style’, 2975Google Scholar.

17. Cf. Vessey, , ‘Statius' epic style’, 2976Google Scholar: ‘it is not the divinity but the vestigiality of the Aeneid which is enshrined and embedded in the textual universe of its adoring worshipper’.

18. For details of the intertextuality see Dominik, W. J., The Mythic Voice of Statius: power and politics in the Thebaid (Leiden, 1994), 174Google Scholar. On Horatian intertexts for the coda to Ovid's Metamorphoses see Henderson, J., ‘Statius' Thebaid/Form premade’, PCPS 37 (1991) 30–80, p. 30Google Scholar. Hardie, , ‘Closure in Latin epic’. suggests Odes 3.30Google Scholar may in turn be a reworking of or allusion to the first close of Ennius' Annales (Book 15, before he extended it by another three books).

19. Cf. Putnam, M. C. J.'s concluding remarks in ‘Wrathful Aeneas and the tactics of pietas in Virgil, Ovid, and Lucan’, in Virgil's Aeneid: Interpretation and Influence (Chapel Hill and London, 1995)Google Scholar, where he briefly depicts the Thebaid as providing a kind of'sealing’ to classical Latin epic (240): ‘Only with Statius’ Thebaid, in the next generation, do we get at least some forms of conciliation, in the plot itself and in the poet's concluding sphragis that looks back on his Virgilian model with a public modesty that scarcely hides the rivalry lurking just below the surface. Clemency, in the figure of Theseus, does at last seem to work, and acts of mourning bring a closure that deliberately leaps back beyond Virgil to recall the Iliadic past common to both poets.’

20. Adrastus, accompanied by the Argive women (the chorus), asks for Theseus' help in burying their dead; after putting the matter to a democratic vote, Theseus agrees and proceeds to argue fiercely with the Theban herald who reports Creon's edict forbidding anyone to assist Adrastus and the Argive women; Theseus then marches on Thebes; a messenger reports his victory and the burial of the dead; the bodies of the sons of the Argive women process, accompanied with lamentations; then Evadne kills herself by throwing herself from a height onto her husband Capaneus' pyre; and an alliance is formed between Argos and Athens.

21. Melville, A. D.'s translation (Statius Thebaid, Oxford, 1992)Google Scholar of clementia as ‘compassion’ is too Christian.

22. Philip Hardie suggests that this phrase also makes a metapoetical point: these words, which amount to a definition of martial epic, will give way in the face of clementia; that is, clementia will bring to an end the unending cycle of violent epics.

23. Rieks, R., Homo, Humanus, Humanitas: Zur Humanität in der lateinischen Literatur des ersten nachchristlichen Jahrhunderts (Munich, 1967), 221–2Google Scholar. Creon's behaviour finds its first-century CE analogue in the heaps of victims executed by Tiberius: iacuit inmensa strages … dispersi aut aggerati. neque propinquis aut amicis adsistere, inlacrimare, ne uisere quidem diutius dabatur, sed circumiecti custodes et in maerorem cuiusque intenti corpora putrefacta adsectabantur dum in Tiberim traherentur, ubi fluitantia aut ripis adpulsa non cremare quisquam, non contingere. interciderat sortis humanae commercium ui metus, quantumque saeuitia glisceret miseratio arcebatur (Tac. Ann. 6.19).

24. See Weinstock, S., Divus Julius (Oxford, 1971), 233–43Google Scholar and Burgess, J. F., ‘Statius' Altar of Mercy’, CQ 22 (1972) 339–49, at 339–42CrossRefGoogle Scholar.

25. See Weinstock, , Divus Julius, 241–3Google Scholar on the temple; Weinstock sees Cicero as the initiator of clementia as a political term (238) where I would prefer to see Cicero as one of several organs of Caesar's political self-representation, which can be seen in part at least as an attempt at self-definition through antithesis with the cruelty of Sulla.

26. For these and other references, see Weinstock and Burgess in n. 24 above.

27. See Coleman, K. M.'s commentary on Statius Silvae 4 (Oxford, 1988) on 4.3.13–15Google Scholar. Cf. Laguna, G.'s note (commentary on Estacio, Silvas III, Madrid, 1992) on 3.3.167b–71Google Scholar on Domitian's clementia, in which he refers to Ovid and the subsequent tradition of the basilikos logos.

28. See Howell, P., A Commentary on Book One of the Epigrams of Martial (London, 1980)Google Scholar, ad loc.

29. Tacitus' cynicism is illustrated by the story of the accusation of treason against Antistius (Ann. 14.48–9). Tacitus asserts that Nero intended to use his veto to save the man from death and thus demonstrate his clementia, but was thwarted when the Senate preempted what he regarded as his imperial prerogative by supporting Thrasea's proposal of exile and confiscation of property instead of the death penalty. The phrase publicae clementiae at the end of Thrasea's reported speech (48) is very pointed.

30. Weinstock, Divus Julius, plate 19.3; the Hadrianic coin reproduced by Weinstock plate 19.4 shows her standing.

31. And that, if anything, is the point of the juxtaposition of these two groups of women in the poem. Hence I cannot agree with those who see the juxtaposition of the suppliant Argive women and the captured Amazons as a problem, e.g. Ahl, F. M., ‘Statius' Thebaid: a reconsideration’, ANRW 2.32.5 (1986), 2892–8Google Scholar, who believes that Theseus is humiliating the Amazons he has conquered by parading them and threatening them in a scene designed to contrast with the other group of women which has just arrived in Athens, the Argive women at the altar of Clementia.

32. In an unpublished paper on the relationship between fifth-century Athenian and Augustan images of the barbarian Other; I am most grateful to Philip Hardie for making this material available to me.

33. E.g. La Rocca, E., Amazzonomachia: le sculture frontali del tempio di Apollo Sosiano (Rome, 1985), e.g. 83–102Google Scholar ‘Il programma augusteo nei Circo Flaminio: la decorazione del tempio di Apollo Sosiano’ on Augustus' appropriation of Greek symbolism for his own political purposes; Fink, J., ‘Amazonenkämpfe auf einer Reliefbasis in Nikopolis’, OJh 47 (1964–1965), 70–92Google Scholar, who shows that parallels between Augustus and Theseus include the imagery of Amazonomachy (80); Kähler, H., ‘Die Ara Pacis und die augusteische Friedensidee’, JDI 69 (1954), 67–100, at 84–5Google Scholar; Simon, E., ‘Zur Bedeutung des Greifen in der Kunst der Kaiserzeit’, Latomus 21 (1962), 749–80, esp. 777–8Google Scholar on the Amazons as identified with the superbi of Aeneid 6, the enemies of Roman imperium.

34. For links between Theseus and the Flavian emperors, Vespasian and Domitian, see Vessey, D., Statius and the Thebaid (Cambridge, 1973), 314–15Google Scholar and Snijder, H., Statius Thebaid: a commentary on book III (Amsterdam, 1968) 17–21Google Scholar, citing passages from Silv. 4.1, Statius' panegyric of Domitian.

35. Reworking Turnus' words at Aen. 12.11: nulla mora in Turno. For most of the poem, the Thebaid has been characterised by mora, like Lucan's poem where delay (mora) is the central feature of the narrative (or the resistance to narrative). On the lengthy morae earlier in the Thebaid, e.g. the Hypsipyle episode, see e.g. Vessey, , ‘Statius' epic style’, 2988–93Google Scholar. On mora in Lucan see Bramble, J. C., ‘Lucan’, in Cambridge History of Classical Literature, vol. 2: Latin Literature, edd. Kenney, E. J. & Clausen, W. V. (Cambridge, 1982), 540Google Scholar (Lucan ‘refuses to narrate’); cf. Henderson, J., ‘Lucan/The word at war’, in Boyle, A. J. (ed.), The Imperial Muse: Ramus essays on Roman literature of the empire (Berwick, 1988), 122–64Google Scholar, on the text's aposiopesis (132), and esp. Masters, , Poetry and Civil War, e.g. 3–10Google Scholar, index s.v. ‘mora’.

36. Cf. the closural capacity of Jupiter in the Aeneid according to Feeney, D. C., The Gods in Epic (Oxford, 1991), 137–8Google Scholar; ‘From the beginning Jupiter is associated with the end.’

37. See Feeney, , Gods, 337–91Google Scholar for an extensive discussion of the role of the gods in the poem, particularly in terms of personification and allegory. As will become clear, I do not agree with Feeney's view (385) that the duel ‘prepares for the disappearance of the allegorical narrative mode’; rather, the abstract figure of Clementia maintains the allegorical dimension of the narrative and invites an interpretation of Theseus as the embodiment of clementia.

38. Feeney, , Gods, 356Google Scholar, cf. 376.

39. See OLD adsum 13.

40. Similarly there are Jovian aspects to Aeneas at the end of the Aeneid: see Hardie, P., Virgil's Aeneid: cosmos and imperium (Oxford, 1986), 147–50Google Scholar, which makes Statius' scene in this respect a replay of the final scene of Aeneid.

41. I find it hard to accept Hershkowitz's interpretation of this simile as ‘a picture of the development of madness’ (Hershkowitz, D., ‘Patterns of madness in Statius’ Thebaid, JRS 85 (1995), 52–64, at 64Google Scholar), although her view of the close of the poem is nearer the mark: that Theseus' ‘frenzy-fuelled excessiveness cannot be sustained but instead diminishes and finally disappears’. But she does not allow an adequate role for Theseus in the expulsion of madness.

42. Cf. Hor. Od. 3.5.1–4, the classic locus for the complementarity of Jupiter and emperor, reiterated by Statius at Theb. 1.29–31.

43. E.g. Ahl, ‘Statius’ Thebaid, 2803–912, who focuses upon boundaries and definitions (2898–902).

44. Vessey's triumphalism: Statius and the Thebaid, 316. Unease for some critics might arise from phrases like insana fulmina (655), wielded by Jupiter in the simile; the possibility that Theseus is breathing in and being inspired by hellish vapours (712–13, perhaps alluding to Aen. 12.945–6 postquam saeui monumenta doloris ∣ exuviasque hausit, furiis accensus); the indiscriminate slaughter of Thebans by Theseus' army (738); Theseus' taunt to Creon (768), which is left unglossed by Statius and which could be seen as an intimidatory conundrum toying with the victim's fate; and the possibly hybristic connotation of horridus Aegides and ore superbo (769–70), which might interact with Aen. 12.930, ille humilis supplex … Interpretation of all these phrases will, as usual, depend on emphasis and contextualisation. Here I shall do no more than respond to the phrase ore superbo (770), used of Theseus immediately prior to killing Creon, that superbus in Virgil can bear a range of meanings from hybristic pride (e.g. of pre-fall Troy (3.2), Brutus (6.817), Mezentius (8.481) and Turnus (10.514 and 12.326)) to patriotic commendation (e.g. of Agrippa's naval crown (8.683) and the doors of Augustus' temple (8.721)). And in the words nec fallunt iussa superba magnanimi Iouis (Aen. 12.877–8) whether you perceive a tension or a convergence between superba and magnanimi will depend upon other elements of your reading of the poem.

45. Johnson, W. R., Darkness Visible (Berkeley and Los Angeles, 1976), 8–16CrossRefGoogle Scholar.

46. The most recent rounds are: Galinsky, K., ‘The anger of Aeneas’, AJPh 109 (1988), 321–48Google Scholar; Putnam, M. C. J., ‘Anger, blindness and insight in Virgil's Aeneid’, in Nussbaum, M. C. (ed.) The Poetics of Therapy = Apeiron 23 (1990), 7–40Google Scholar; Galinksy, , ‘How to be philosophical about the end of the Aeneid’, ICS 19 (1994). 191–201Google Scholar; Putnam, , ‘Wrathful Aeneas and the tactics of pietas’, 201–45Google Scholar. I suspect these two combatants will never agree. For some discussion of the philosophical elements here see Braund, S., ‘Virgil and the Cosmos: religious and philosophical ideas’, in The Cambridge Companion to Virgil (Cambridge, forthcoming)Google Scholar.

47. D. P. Fowler, ‘Epicurean anger’, M. R. Wright, ‘Ferox uirtus: anger in Virgil's Aeneid’ and Gill, C., ‘Passion as madness in Roman poetry’, in The Passions in Roman Thought and Literature, edd. Braund, S. M. & Gill, C. (Cambridge, 1997)Google Scholar.

48. Cairns, F., Virgil's Augustan Epic (Cambridge, 1989)CrossRefGoogle Scholar; quotation from p. 78, Turnus as tyrannos p. 67. It is salutary to remind ourselves that there is no recorded criticism of Aeneas for his killing of Turnus until Christian writers such as Lactantius and St Augustine and that Servius, the fourth-century commentator on Virgil, found the close of the Aeneid unproblematic: he says that it reflects glory upon Aeneas, because he considers sparing his enemy and avenges for Evander the death of Pallas by killing Turnus.

49. On the tensions see Hardie, , Cosmos and Imperium, esp. 153–4Google Scholar; for imperialist readings see Hardie, 's broadly favourable review of Cairns in JRS 80 (1990) 209–10Google Scholar.

50. On open-endedness see Hardie, , Epic Successors, 32–5Google Scholar, quotation from 35.

51. Proponents of this view include Lewis, C. S., The Allegory of Love (Oxford, 1936)Google Scholar and Vessey, Statius and the Thebaid, who even gives the ending Christian overtones: ‘an epic not of sin but of redemption, a chronicle not of evil but of triumphant good’ (316).

52. Vessey, , Statius and the Thebaid, 312Google Scholar.

53. This is well represented in the Ramus volume Imperial Roman Literature II: Flavian epicist to Claudian (Ramus 18 (1989)) edited by Boyle, TonyGoogle Scholar, devoted to Roman epic. In it, McGuire, D. T. Jr, ‘Textual strategies and political suicide in Flavian epic’, 21–45Google Scholar, reads Flavian epic as political commentary on civil war and tyranny and Dominik, W. J., ‘Monarchal power and imperial politics in Statius' Thebaid’, 74–97Google Scholar, articulates explicitly the veiled criticism of the regime he finds here, especially ‘the cost of institutional monarchy in individual loss and suffering’ (78). Dominik develops these ideas more fully in The Mythic Voice of Statius. Fantham, Elaine, ‘“Envy, and fear the begetter of hate”: Statius' Thebaid and the genesis of hatred’, in The Passions in Roman Thought and Literature, edd. Braund, S. M. & Gill, C., (Cambridge, 1997)Google Scholar, sees the Thebaid as a poem dominated by destructive forces.

54. For detail see Dominik, , The Mythic Voice of Statius, 92–8, 157–8Google Scholar. Theseus' first appearance in the poem is as a victor in a triumphal procession, humiliating the Amazons he has conquered by parading them and threatening them: Ahl, ‘Statius' Thebaid’; his violent despatch of Creon is seen as a criticism of the violent methods employed in the Flavian restoration: Dominik, , ‘Monarchal power’, 92Google Scholar.

55. Ahl, ‘Statius' Thebaid’, quotations from 2904 and 2898 respectively.

56. Feeney, , Gods, 362–3Google Scholar.

57. Henderson, ‘Form premade': the quotation from Barthes’ S/Z appears in footnote 2.

58. Hardie, , Epic Successors, quotations from 46 and 48Google Scholar.

59. Hershkowitz, , ‘Patterns of madness’, 63Google Scholar.

60. Some critics resist attempts to invoke (the ancient) contemporary settings to illuminate ancient texts. Of course, I concede that we can, ultimately, only respond to texts as products of our own (collective and individual) histories. Of course, I cannot deny the validity of readings which respond to ancient texts from the perspective of twentieth-century western democracies. In the case of the Thebaid, with our (awareness of recent) history, we can hardly avoid the reflexes of loathing autocracy and tyranny; I'm a member of Amnesty International too. But I think one of the central challenges and rewards of studying ancient texts is the invitation they offer to enter a different mind-set. That seems worth the effort. In this case, we are dealing with the mind-set of a first-century CE Roman court poet – and that is is not necessarily easy to understand. I hope that this paper has offered at least some useful and persuasive points of entry.