1. The title of this essay is indebted to Newlands (1997), ‘The Metamorphosis of Ovid's Medea’, an article which has also influenced my impressions of Ovid's Medea in the Metamorphoses. This essay contains truncated sections of my dissertation, ‘Seneca's Medea and the Tragic Self’, completed at the University of Southern California in May 2011. I would like to thank A.J. Boyle for including my contribution in this volume and for all his helpful advice and corrections along the way. All inadequacies and errors are entirely my own.

2. Without making assumptions about Seneca's skill for invention (or lack thereof), we cannot know the full extent of Seneca's relationship to Ovid's tragedy, though some have tried—see, e.g., Cleasby (1907).

3. Newlands (1997), 192, 178.

4. See Met. 7.192-219 (Medea's incantation for the rejuvenation of Aeson); 7.309-11, 332-38 (to the daughters of Pelias); and 7.350-424 (escaping the daughters of Pelias, the Corinthian episode, and Medea in Athens).

5. As Newlands (1997), 187f., also observes.

6. See Walsh (2011), 173-86. Schiesaro (2003) and Guastella (2001) both explore this aspect of the Medea character.

7. All Latin and Greek texts are from the most recent editions of the Oxford Classical Texts series, except for Ovid's Metamorphoses and Heroides 12, for which I use Anderson's Teubner and Goold's Loeb editions respectively. I have noted where I disagree with Zwierlein's text of Seneca's Medea. All translations are my own, unless otherwise specified. In translating Seneca's text, I have focused my energy on conveying Seneca's word order, even at the expense of a smoother English reading; I have found that Seneca's placement of words is often particular and laden with meaning.

8. I am not the first scholar to note the increased isolation given to Medea's character in Seneca's tragedy. See also Fyfe (1983); Davis (1993); Benton (2003), 274.

9. She also maintains her social distinction as mother with Aegeus, since (in addition to any children she might bear for him), she becomes the functional mother of Theseus (cf. Ov. Met. 7.402-24, Sen. Phaed. 697).

10. A common enough practice, if we are to imagine the application of Roman divorce law to Medea's situation: see Abrahamsen (1999).

11. Others have also observed Medea's isolation from the chorus. See, e.g., Davis (1993), 37: ‘the constant absence of the chorus contrasts with Medea's constant presence…and serves to emphasize Medea's social and cultural isolation’; Fyfe (1983); and Benton (2003), 272-77, who, like me, includes Medea's gender as a factor in her isolated state.

12. See in Euripides, e.g., Medea's scenes with Creon (261-356), with Jason (446-622, 866-975, 1317-1404) and with Aegeus (663-758).

13. Eur. Med. 214.

14. Seneca's break from Greek tragedy in his use of the choral body is well-documented (see, e.g., Davis [1993] and Boyle [2006] for general changes from the Greek tradition). While Senecan choruses generally lack the level of participation present in Greek tragic choruses, the particular absence of choral interaction with characters in the Medea is striking.

15. Costa (1973) ad 56-115: ‘note that the chorus throughout is hostile to M. and friendly to Jason (102ff., 362, 596); in Eur. the chorus is friendly to M. (e.g. 267-8)’.

16. One might contrast the choral activities in Seneca's Oedipus, Troades and Hercules Furens, where the choruses converse with several characters and represent clearly defined communities of citizens. I suggest that the chorus of the Medea is also clearly defined (as Corinthian citizens, made apparent in the first choral ode) despite their general distance from conversation.

17. For a fuller discussion of these passages as they relate to Seneca's Medea, see Walsh (2011), 134-38.

18. See Davis (1993), 37, who believes that ‘the chorus is absent from the stage in every act of Medea except during Act 5, scene 1 (879-90) when they are required as recipients for the messenger's news’.

19. Leaving aside any thematic similarities (of which there are many), the use of imperatives from Nurse to female protagonist is clearly part of the schema. Compare, e.g., Med. 174 with Phaed. 165 (compesce); Med. 157 (siste furialem impetum) with Phaed. 248, 263, and Ag. 203 (siste furorem—siste furibundum impetum—siste impetus); Med. 381 with Ag. 224 (comprime); Med. 175 (animosque minue) with Phaed. 256 (animos coerce). Contrast Atreus who in his conversation with the Satelles (Thy. 204-335), despite its length, rarely uses the imperative.

20. That is, unlike in Euripides, there is no opportunity for the chorus to warn/counsel Medea against taking certain actions, a phenomenon which is also noted by Fyfe (1983), 79-81.

21. . Showing again that the final monologue of Seneca's play is not the sine qua non of meaning, as some critics tend to treat it, though Gill himself does admit that the difference between these two Medeas probably reflects ‘larger differences between Euripidean and Senecan dramaturgy and psychology’ (Gill [1987], 25, from which the passage quoted is also taken).

22. Several critics observe that Medea openly models herself on her past: Guastella (2001), 203-17, Schiesaro (2003), Star (2006), Gill (1987), 31-36, Boyle (1997), 58f., etc. Schiesaro argues for Medea's ‘desire to push her life backwards, to deny the future any real possibility of unfolding and deviating from the past’ (208). He also highlights the fact that Seneca invents this desire: ‘Euripides' heroine makes no attempt to win back the object of her passion’ (209). See generally 208-14. The following analysis aims to unpack Medea's reliance on the past as a means of self-creation. On the premise that Medea is anxious about being forgotten (including a comparison to Ovid's Medea of the Heroides and Ariadne of Catullus 64), see Walsh (2011), 173-86.

23. Here I agree with Fitch & McElduff (2002), who argue that the unfortunate result of Medea's self-assertion is that she ‘misidentifies’ her self with one set of characteristics; she ‘associat[es herself] too completely with a particular role’ (30). Fitch & McElduff, however, fail to problematise the mechanism of Medea's self-fashioning, and address only the outward consequences. I also find their opposition between Medea-as-wife and Medea-as-mother to be a too simplistic reading of her character. Benton (2003) argues (also too simplistically, in my opinion) that Medea chooses to become the character of the ‘Other’, a xenophobic stereotype which the Corinthians have created for her.

24. Much has been said about the paradoxical connection between Medea's motherhood and the necessary severity of her crimes. See, e.g., Guastella (2001), Nussbaum (1996), McAuley (2008) for a range of better interpretations (though none of the above, in my opinion, interprets the problem to a satisfying end).

25. See Walsh (2011), 198-200, for a more detailed assessment of this passage.

26. For the presence and absence of the chorus in Seneca's tragedies, see Davis (1993).

27. For Medea adjusting her speech to fit the expectations of others, see Walsh (2011), 23-44.

28. Though not connected directly with her return home, Medea uses ‘re-’ words also to describe the return of the Argonauts to Greece (235, 238), and in other contexts (recedens, 282; recidas, 296). The ‘re-’ vocabulary also appears in the Nurse's address to Medea at 380-430 (381, 383, 385, 389, 425). Perhaps of interest is Medea's language in Ovid's Heroides 12, where she tells Jason redde and refer.

29. See, e.g., Med. 727.

30. In this, Medea is similar to Ariadne (who envisions herself going back home at Cat. 64.177-87), but she is unlike Vergil's Dido, who imagines herself travelling onwards with Aeneas to Italy, or waging war against the Trojans (Aen. 4.537-46).

31. De Prouidentia 2.10-12, which is Star's primary comparison: Star (2006), 218-21, 232-40.

32. This is usually the point in the monologue where critics identify Medea as having been overcome by madness or passion (or, as Gill might argue, an ‘akratic self-succumbing to passion’). See Gill (1987), 25f., Costa (1973) and Hine (2000) ad loc.

33. This is a bit paradoxical—a teleology that is also a turning backwards, but Medea's own speech is paradoxical: for example, she calls herself coniunx (1021 ) and mater (1012) after she has decided that both children must die, and after she has declared that her virginity has been restored (984). Her own logic is not internally consistent.

34. See Section II above.

35. A J. Boyle reminds me that this expression (the place being open for a wound) is reminiscent also of gladiatorial combat.

36. Fitch & McElduff (2002), 38.

37. For the connection between female and penetrability, see Walters (1998), esp. 41. For a lengthier discussion of Medea's impression of feminine penetrability and the adoption of a masculine persona as a means of securing her boundaries, see Walsh (2011), 67-74. Segal (1983b), 178f., and Nussbaum (1996), 453-57, also discuss Medea's quest to undo Jason's penetration, but they ignore the biological and gendered proclivity for penetration to be a problem (i.e., her female-ness), as well as the connection between Medea's penetrability and her adoption of masculinity throughout the play.

38. As I argue below, Medea's use of such verbs serves to reinforce her connection to the divine Juno in Seneca's Hercules Furens.

39. For further comparison of some of these passages with Medea's statement, see Walsh (2011), 83-86.

40. For more on the exigent circumstances of Jason's marriage to Creusa (and Creon's motivation for taking Jason in), see Fyfe (1983), 81, and Lawall (1979), 420f., 423f.

41. conflige is Avantius' emendation of the MSS confligere, which Zwierlein retains and obelises.

42. See, e.g., Boyle (1994), ad 614.

43. Her foreignness might also contribute to this perspective; though the social structure is familiar to her (having come from a land with a king and royal family), she is not socially or biologically Greek. For more on this topic, see Walsh (2011), 132-72, and Fyfe (1983) more generally on Medea's exclusion from the Corinthians.

44. Many scholars have also observed the similarities between Atreus and Medea, and there seems to be a consensus that Atreus' character is modelled to some extent on Medea's (or at least shares many of the same attributes, including a choice to see himself as a deity). See, e.g., Fitch & McElduff (2002), 36, Littlewood (2003), 38, 47, 97, 148-53, 174; Segal (1983a), 237-44.

45. Though it is outside the scope of this essay, one must reckon with the magic scene (lines 670-848), which comprises almost a fifth of the tragedy and is, perhaps, the clearest display of Medea's actual power. In the monologue itself (including the Nurse's descriptions and reported speech), Medea prays to Hecate to help her create the fire which, housed in the robe given to Creusa, will burn against the laws of nature (see the Messenger's report at Med. 888-90) and destroy the city. In the course of observing Medea's magical ability, Medea professes a power over nature—defined broadly: flora (707-30), fauna (675-704), landscape (752-59), seasons (e.g., 759-61), planetary bodies (768f.), and a power over the underworld (740-51).

46. A.J. Boyle reminds me that the theme of fulfilment of prayers is important in Senecan tragedy.

47. Nussbaum (1996), 463, mentions Phaethon alongside Icarus as proper warnings against ambition. While she argues extensively about the connection between Medea and Plato's Phaedrus, she does not examine in detail the resonance of the Phaethon myth.

48. For the ‘daring’ of the Argonauts, see Walsh (2011), 51-61; for the resonance of immemor, see 180-86.

49. For detailed verbal allusions, see Jakobi (1988) ad Med. 32, 33, 33f., 37f., 601f., 745, 758f.

50. The possible exception to this is the Fury at the beginning of the Thyestes, who could also make an interesting comparison to Medea.

51. For a fuller discussion of Medea's similarity to Juno, as well as the echoing themes of boundary violation and broken fides contained in both tragedies, see Walsh (2011), 100-16.

52. For Medea as dramaturge, see, e.g., Boyle (2006), 208-10, 215-18, Schiesaro (2003), 16-19 et passim; Littlewood (2004), 45-51, 57-60.

53. It is outside the scope of this essay to describe how Medea herself resembles the Fury specifically invoked by Juno, but it is significant in showing how Medea embodies her own agent of revenge: see Walsh (2011), 106. For Medea as a Fury in general, see Littlewood (2004), 37, 148-71 et passim.

54. HF 99 occurs with noster, a phrase used by Medea more than once (Med. 554, 907f., 1011), and a phrase that does not occur in Senecan tragedy outside these two plays. Though it is outside the current scope of my project, T. Habinek points me towards the likely intertext for an angry and mournful Juno: Vergil's Aeneid (and especially the opening: 1.4, 9, 11, 25, 130, etc.). See also Lawall (1983), who argues that Hercules' pietas and fortitude foil Juno's plans to destroy his reputation as hero.

55. According to Denooz (1980), of the 64 instances of dolor in Senecan tragedy, 14 occur in Medea (over 20%). Dolor occurs in Tro., Phaed., and Thy. 11, 10 and 10 times, respectively. Of the 70 instances of ira in Senecan tragedy, 22 occur in Med. (almost a third); Thy. has the next place with 13 occurrences.

56. Note that this line has two first person future verbs, one of which is extraham; Juno also uses extraham as the second verb of a two-verb sentence at HF 95.

57. For a discussion of Medea's agency and its relation to her use of first-person verb forms, see Walsh (2011), 194-200.

58. For a different interpretation of this part of the tragedy, see Fyfe (1983), 84f.

59. I am surprised that most critics, on the nature of Medea's love in this passage, tend to assume that Medea's love is love for Jason (thus she has conflicted feelings about him—anger and love). I think, however, that Medea's love here is directed at her children, since they are the most pressing cause of her body's rebellion. It makes more sense to see her conflicting emotions as (1) anger at Jason (the jilted wife who seeks revenge, which aligns her with Juno as well), and (2) love for her children (which would cause her to balk at murdering them). Structurally, the thing which is amor and pitted against ira seems to become expressed as pietas at Med. 943f.

60. See, e.g., Star (2006), Guastella (2001), Fitch & McElduff (2002).

61. Gill (2009) suggests a Stoic reading of Medea's final monologue, in terms of Medea's accordance with her ‘nature’.

62. In this, Medea also embodies the fulfilment of the chorus' wishes: to return to a pre-Argonautic past (with firm boundaries between lands) and to restore their relationship with the divine (which was lost when the Argonauts set sail). See, e.g., Walsh (2011), 50-67.

63. Newlands (1997), 192.

64. Newlands (1997), 207.

65. See, e.g., Med. 114, 150, 174, 189, 281, 381, 390, 538, 856.