1. On the question of unity see most recently Michelini, A.N., Euripides and the Tragic Tradition (Madison 1987), 133f.Google Scholar, including relevant bibliography; Heath, M., ‘“Jure Principem Locum Tenet”: Euripides’ Hecuba’, BICS 34 (1987), 43ff.Google Scholar; and Segal, C. P., “Violence and the Other: Greek, Female and Barbarian in Euripides’ Hecuba’, TAPA 120 (1990), 109Google Scholar. As Michelini observes (133), ‘complaints about the play’s disunity are usually based on presuppositions that derive from the work of Sophokles.’ Two recent excellent studies of this play appeared too late for me to consider fully for my main argument, but I do acknowledge them in my notes. Zeitlin, F.I., ‘Euripides’ Hekabe and the Somatics of Dionysiac Drama’, Ramus 20 (1991), 53–94CrossRefGoogle Scholar; Thalmann, W.G., ‘Euripides and Aeschylus: The Case of the Hekabe’, CA 12 (1993), 126–59Google Scholar.

2. See Zeitlin, F.I., ‘The Corruption of Sacrifice in the Oresteia’, TAPA 96 (1965), 463–508Google Scholar; Vidal-Naquet, P., ‘Hunting and Sacrifice in Aeschylus’ Oresteia’, in Myth and Tragedy in Ancient Greece (New York 1990), 141–60Google Scholar. On revenge tragedy in general see Burnett, A.P., ‘Medea and the Tragedy of Revenge’, CP 68 (1973), 1–24Google Scholar.

3. In particular, see Burkert, W., Homo Necans, tr. P. Bing (Berkeley and Los Angeles 1983; orig. publ. 1972)Google Scholar; Girard, R., Violence and the Sacred, tr. P. Gregory (Baltimore 1977; orig. publ. 1972 as La violence et le sacre)Google Scholar. The latter author’s applicability to classical studies has been hotly debated. On the uses and abuses of Girard for Greek tragedy see now Mitchell, R.N., ‘Miasma, Mimesis and Scapegoating in Euripides’ Hippolytus’, CA 10 (1991), 94–118Google Scholar; and Drew Griffith, R., ‘Oedipus Pharmakos? Alleged Scapegoating in Sophocles’ Oedipus the King’, Phoenix 47 (1993), 95–114CrossRefGoogle Scholar.

4. Foley, H., Ritual Irony: Poetry and Sacrifice in Euripides (Ithaca 1984), 89Google Scholar. Foley’s introductory chapter on the applicability of anthropological theory for Greek tragedy is quite useful.

5. Michelini (n.1 above), 170. On retributive and mimetic revenge see also Nussbaum, M.C., The Fragility of Goodness: Luck and Ethics in Greek Tragedy and Philosophy (Cambridge 1986), 410–15Google Scholar.

6. Adkins, A.W.H., ‘Basic Greek Values in Euripides’ Hecuba and Hercules Furens’, CQ 16 (1966), 192–219CrossRefGoogle Scholar. See also Michelini (n.1 above), 154f.

7. Michelini (n.1 above), 149. Kovacs, D., The Heroic Muse: Studies in the Hippolytus and the Hecuba of Euripides (Baltimore 1987), 105Google Scholar, 110f. and 118, views the absence of the gods positively. For a different reading see Segal, C.P., “The Problem of the Gods in Euripides’ Hecuba’, MD 22 (1989), 9–21Google Scholar.

8. For a general, although differently focused, study of the role of Achilles in this play see King, K.C., ‘The Politics of Imitation: Euripides’ Hekabe and the Homeric Achilles’, Arethusa 18(1985), 47–66Google Scholar.

9. However, Segal (n.1 above), 129, argues that the blinded Polymestor’s new prophetic powers turn him into a perverted deus ex machina.

10. King (n.8 above), 50, observes that Euripides transformed the tradition ‘in order that the original martial heroism of Achilles, distilled into savage egoism, might be condemned for its effect on human morality.’

11. Loraux, N., Tragic Ways of Killing a Woman, tr. A. Forster (Cambridge MA 1987; orig. publ. 1985 as Façons tragiques de tuer une femme), 37Google Scholar: ‘The virgins cannot fight alongside the men, but, in times of peril, their blood flows so that the community of andres may live…. It must be virgin blood, or as the sacrificers proclaim as they go about their work, “pure blood”.’

12. See Lloyd-Jones, H., ‘The Guilt of Agamemnon’, CQ 12 (1962), 198fCrossRefGoogle Scholar.

13. King (n.8 above), 54, notes aptly: ‘One might expect the debate to resolve around anagkē, the necessity to choose between two evils: either to commit human sacrifice or to return home…. But anagkē is never mentioned in this context.’

14. As Seaford, R., ‘Homeric and Tragic Sacrifice’, TAPA 119 (1989), 89Google Scholar, has most recently argued, ‘the aggression required for the group to kill an outsider (the enemy, or the prey) is created, sustained, or coordinated by the killing of an insider, a female member of the group. The sacrifice serves a state of mind’ [my emphasis]. Since Polyxena is outside the group, her death cannot serve the intended psychological purpose.

15. Since all of these concerns figure in both this drama and Aeschylus’ treatment of Iphigenia, Richard Seaford’s recent thoughts (n.14 above, 90) about the negative role of sacrifice in tragedy support my point here: ‘In tragedy too the uncontrolled violence characteristic of hunting and warfare is in tension with (ritual) cultural control (sacrifice, burial, etc.), but the tension tends to result in the failure or subversion of the rituals…. This perverted or anti-sacrificial quality of the “sacrifice” of the hare helps to explain why it gives rise specifically to the sacrifice of Iphigenia.’ Thus the antisacrificial quality of the sacrifice of Polyxena, I shall argue later, results in the sacrificial revenge against Polymestor.

16. I stress that Iphigenia contrasts, not compares, with Polyxena. We remember Agamemnon’s daughter here only to reinforce how utterly different this situation is. Here I part company with Segal (n.1 above, 114), who argues Polyxena is, like Iphigenia, ‘the ideal victim’ since she submits passively to the warriors. Ideal victims serve their purpose.

17. On the contradictions and dissonances created by Polyxena’s adoption of Achillean language and values, see Rosivach, V., ‘The First Stasimon of the Hecuba 444ff.’, AJP 96 (1975), 359f.Google Scholar, and Thalmann (n.1 above), 147f.

18. Loraux (n.11 above), 56–60, grants the erotic impact of the breasts but shifts the focus to Polyxena’s intentions, which are, of course, not sexual. However, given the setting and the sheer number and identity of the oglers, her intention ultimately does not matter much beyond the immediate heroic moment. Segal (n.1 above, 111f.) also sees the power of this scene’s eroticism, although he does not relate it to the sacrifice. Segal has an excellent summary and criticism of the relevant bibliography. Most recently, see Thalmann (n.1 above), 142–45.

19. On the need to keep sexuality and sacrifice separate see Burkert (n.3 above), 58–72, who further argues that the sacrifice of the maiden was part of the renunciation of male sexuality to arouse aggression needed for warfare. At the end of a war, however, violence needs to be cleansed, so the sacrifice is utterly misdirected. On pure and impure blood and the necessity of avoiding contact with the miasma see Vernant, J.-P., ‘The Pure and the Impure’, in Myth and Society in Ancient Greece, tr. J. Lloyd (New York 1990), 121–42Google Scholar.

20. For a recent, succinct and eloquent analysis of this paradigm, see Loraux (n.11 above), esp. 37–42.

21. There may be a ‘textual memory’ at work here also. Burkert (n.3 above), 67, speculates that Polyxena’s name may point to a practice of maiden sacrifice on the Volga where the victim had to give herself to all participants in the funeral; hence the phrase poluxenai neanides (Pindar fr. 122.1).

22. The name Neoptolemus does work metrically, and is used in the Andromache (e.g. 14), so one would presume its absence in the Hecuba is motivated. Segal (n.1 above), 127 n.53, also notes the exclusively patronymic references, but draws no conclusions from it.

23. Schlesier, R., ‘Die Bakchen des Hades: Dionysische Aspekte von Euripides’ Hekabe’ M’tis 3 (1988)Google Scholar, esp.117f.: ‘Der zerteilte Körper des Polydoros beweist der Mutter klar: Der Knabe war das Opfer eines Ritualmordes.’ Schlesier, however, does not tie that corruption of ritual to the failure of the sacrifice in the play. Zeitlin (n.1 above), 61, criticises Schlesier’s conclusions. The presence of the murdered corpse would present a great danger of contagious miasma; see Vernant (n.19 above), esp. 134.

24. Especially featured are cognates of sphazein (22, 41), tumbos (37, 41, 47, 50), taphos (30, 47), phainomai (37, 41) and phantasma (53) which is actually picked up by Hecuba’s description of her dreams immediately afterwards (70).

25. If Delebecque, É., Euripide et la guerre du Péloponnèse (Paris 1951), 154–59Google Scholar, argues correctly that Euripides patterned Polymestor after Seuthes, the contemporary king of Thrace, the untrustworthy, murderous, greedy ally of the Athenians, and the audience saw the connection, then Polymestor would really be a scapegoat for the Athenian audience as well.

26. On the implications and effects of shifting ethical affiliations in drama see Berger, H., Jr., ‘King Lear: The Family Romance’, The Centennial Review 23 (1979), 348–76Google Scholar; id., ‘Text against Performance: The Gloucester Family Romance’, in R. Erickson and C. Kahn (eds.), Shakespeare’s Rough Magic: Renaissance Essays in Honor of C.L. Barber (Newark, Del., 1985), 210–29Google Scholar; id., Imaginary Audition: Shakespeare on Stage and Page (Berkeley and Los Angeles, 1989)Google Scholar; Burke, K., ‘Coriolanus and the Delights of Faction’, in Language as Symbolic Action (Berkeley and Los Angeles 1966)Google Scholar. On the applicability of Berger’s ideas to Greek tragedy see Mitchell (n.3 above).

27. Obviously, my use of the term ‘catharsis’ is not rigorously Aristotelian, and thus needs (at least) some brief explanation. These thoughts in a sense actually derive from Plato’s fears of the effect of ignoble literary characters on the stability of the soul. Euripides compels his audience to invest their emotions in the suffering Hecuba, and then proceeds to redirect this investment towards a different project: the ethical complexities of revenge and its representation. No one can observe Hecuba’s transformation without discomfort. How then, does Euripides direct these emotions? See now also the similar speculations of Zeitlin (n.1 above), especially 56f. and 85f.

28. xenou (7), xenoi (19), xenos (26), and then another fifteen times afterwards: 82, 86, 710, 715, 774, 781, 790, 794, 852, 890, 1047, 1216, 1235, 1244, 1247. On xenia in the Hecuba, see Nussbaum (n.5 above), 406–10.

29. Schlesier (n.23 above), 111 n.2, notes that Euripides draws attention to the meaning of Polymestor’s name—‘who knows or gives much advice’—in his function as a prophet at the end. He also sees Polydorus’ name—‘the much-giving one’—in the context of the xenia (113 n.8).

30. Zeitlin (n.1 above), 53, observes references to epithets of Hades in their names, which is possible, but there are nineteen occurrences of forms of xenos in the play, in addition to the eight instances of Polyxena’s name, so the operating structure of thought, in my view, is the xenia.

31. Zeitlin (n.1 above), 69, also sees the revenge as compensating for both deaths.

32. A.P. Burnett (n.2 above), 2.

33. Girard, R., ‘Hamlet’s Dull Revenge’, Stanford Literature Review 1 (1984), 163fGoogle Scholar.

34. When Hecuba uses rhetoric to denounce rhetoricians like Odysseus, it is clear, as Michelini points out (n.1 above), 156, that ‘the speaker herself has become adept in the technique of saying bad things well. Her denunciation of those professionals exposes the self-destructive principles of contemporary rhetoric, while it reveals Hekabe’s own moral vulnerability.’

35. Segal (n.1 above), 128, observes how Polymestor ‘keeps probing for the places where he can hurt Hecuba until he finally hits the prophecy about Cassandra’s murder, the detail that will give her the most pain.’

36. Comparison with Phaedra’s similar struggle and consequent defeat is apt; see R. Mitchell (n.3 above).

37. Note Agamemnon’s language with respect to Polymestor: philos (858), koinon (860).

38. Conacher, D.J., Euripidean Drama: Myth, Theme and Structure (Toronto 1967), 156Google Scholar.

39. Steidle, W., ‘Zur Hekabe des Euripides’, WS 79 (1966), 136Google Scholar, further observes that the placement of this wish at the end of the episode would underscore its significance.

40. Hecuba further takes the heroic language used by Odysseus (as reported by Polydorus’ ghost, 134) to describe Achilles, ton ariston Danaōn panton (‘the best of all the Danaans’), and transfers it to her description of how she killed Polymestor’s children sun tais aristais Trōasin (1052), ‘with the best Trojan women’. Polyxena, donning the heroic mantle of Achilles, becomes psukhēn aristē (‘best in soul’, 580). Her daughter’s preference of a glorious death over ignoble life imitated the heroic Achilles of Homer, but Hecuba’s vengefulness imitates the bloody wraith inhabiting Euripides’ tragedy.

41. Segal (n.1 above), 126, is fairly close to my point here, although he restricts ‘the increasing collapse of differences’ primarily to just Hecuba and Polymestor.

42. Schlesier (n.23 above), 129 n.50, drawing on parallels between Polymestor and myths about Lycurgus, argues that the wolf symbolises Polymestor, and thus the deer is Polydorus. Hecuba’s language in the speech about the source of her dream oscillates back and forth between male and female, so the dream would seem to prophesy the death of both children.

43. This also essentially joins the two halves of the play psychologically, since they are thus part of the same experience and meaning.

44. A glance in the LSJ at okhlos and words cognate with it yields a great number of negative associations. Plēthos, on the other hand, seems relatively neutral. Hence, coming from the mouth of the royal (in spirit still, at least) Hecuba, it is a term of utmost reproach.

45. Schlesier (n.23 above), 131, observes the significance of knives in the play and links it (n.57) to the use of sharp weapons in Dionysian cult.

46. However, in King Lear as in the Hecuba, reason seems inadequate to the demands of the play’s world.

47. Schlesier (n.23 above), 111 n.2, observes the importance of Euripides’ invention of Polymestor’s name for this play and sees its meaning (‘he who gives much advice or knows much’) as underscoring his function as a prophet at the play’s end. Neither Schlesier nor Zeitlin have noticed the Odyssean overtones of the name, however.

48. For the similarities between Polymestor and Polyphemus see Wilhelm Schmid in Schmid, W. and Stählin, O., Geschichte der Griechische Literatur 1.3 (Munich 1940), 466Google Scholar; more recently Segal (n.1 above), 123–27, and ‘Golden Armor and Servile Robes: Heroism and Metamorphosis in the Hecuba of Euripides’, AJP 111 (1990), 309fGoogle Scholar.

49. See, most recently, Seaford (n.14 above), 91.

50. Segal (n.1 above), 121.

51. The sacrificial quality is underscored by the next word, sphakta, used to describe Polymestor’s children. This verbal repetition is important since Polymestor is using exactly the same language to describe Hecuba’s actions as she had for his.

52. See the listing for phasganon in Chantraine, P., Dictionnaire étymologique de la langue grecque, histoire des mots (Paris 1980)Google Scholar. Chantraine also notes that the post-Homeric use of phasganon is an archaism and poetic. Euripides may be calling attention to the etymology by pointedly using the same word for both sacrificial knife and the dagger of the revenge. This argument may be supported by a similar usage in the Ajax. The sword of Ajax is designated by a variety of terms—for example, egkos (558)—but, suddenly, as Ajax goes to kill himself he calls it a ‘sacrificer’ (sphageus, 815), and then a phasganon (834) with which he will cleave his side in two, just as he divided up the mistaken object of his wrath. Tecmessa then also calls the sword a phasganon (899).

53. Polyxena (561): katheisa pros gaian gonu; Polymestor (1150): kampsas gonu.

54. For a more sceptical view of the role of the gods here see Segal (n.1 above), 129–31, who finds that Agamemnon’s inability to see any larger question of a moral undermines any reading of the end as displaying the presence of the gods. However, Agamemnon’s silence here probably speaks more about his shallowness and perhaps his unease at what Polymestor has just said about him. On the return of the winds I am in agreement (although from a completely different perspective) with D. Kovacs (n.7 above), 110.

55. sphagēs (1037), sphakta (1078).

56. Griffiths, F.T., ‘Girard on the Greeks/The Greeks on Girard’, Berkshire Review 14 (1979), 27Google Scholar.

57. On the Hecuba’s complicated relationship with the Oresteia, see now Thalmann (n.1 above).

58. This process would complete the undermining of contemporary martial ethics in the midst of the Peloponnesian War which was begun with the Euripidean Achilles and makes ‘Achilles’ martial heroism and Athenian war politics comment on each other to the detriment of both’ (K. King [n.8 above], 48). With respect to the sons of Theseus, Segal (n.1 above), 111, wonders ‘if Euripides is reflecting pessimistically on his own city’.

59. We might recall here that in the Iphigenia at Aulis Euripides similarly places the Iliad in jeopardy by presenting an Achilles who almost runs off with Agamemnon’s daughter.

60. The ideas in this article were first presented in 1989 before a hospitable audience at Bryn Mawr College. For their comments on various written versions, I would like to thank Albert Cook, David Konstan, Daniel Tompkins, and, especially, Greg Thalmann. All remaining mistakes, eccentricities and inanities are due to my own foolishness.