1. See chapter 1, n. 2 for the conjectured dates of Ovid’s poetry. Suetonius’s Lives of the Caesars was published between 119 and 122 CE. See Power 2021, 17. For Suet. Aug. 99.1 quoted above, the second hand in MS P and Philippus Beroaldus has corrected minimum to mimum, which makes better sense in the context. Regardless of whether we accept mimum or minimum, the metaphor of dramatic playacting is made clear by both πέπαισται and κρότον. See Myerowitz 1985, 187; Beacham 1999, 151; Boyle 2006, 160; and Power 2021, 146–50 for discussions of this passage.

2. Boyle 2006, 7. See further, e.g., Dupont 1985; Bettini 1991; Bartsch 1994; Flower 1995, 1996, 2004; Beacham 1999, 35–44 and passim; Boyle 2006, 3–7; and Gildenhard and Revermann 2010, 18–19 and n. 78 on institutional and social performance, including drama, in ancient Rome. See Wiseman 1998, 75–120 for the promotion of the gentes with spectacles, including the Roman dramatic genre of the fabula praetexta.

3. See Beard 2007 for a discussion of the Roman triumph.

4. See Flower 1996, 91–127, on imagines in Roman funerals; and 12–15 on the emotion of shame and the function of the imagines as an audience judging younger members of a family.

5. For an overview of Roman tragedy and its history, see, e.g., Beare 1965; Tarrant 1978; Currie 1981; Beacham 1991, 117–26; Fantham 2005; Paratore 2005; Schiesaro 2005; Boyle 2006, esp. 3–23; Gildenhard 2010; Manuwald 2010, 1–41; and Feeney 2016. For tragedy in the late republic and imperial period, see, e.g., Slater 1996 and Beacham 1999. Gildenhard (2010) and Feeney (2016) offer recent theories on the Roman motivation and interest in Greek drama. For material evidence of the Roman theater, see, e.g., Savarese 2007 and Borriello et al. 2010.

6. See Currie 1981, 2704 and passim; Fantham 2005, 116–17 and passim; Schiesaro 2005, 269–71 and passim; and Feeney 2016, 1–16 and passim for the influence of Attic tragedy on Roman drama.

7. Cicero (Brut. 72) calls it a fabula. See Boyle 2006, 28, 246 n. 7, and 8 for evidence supporting fabula to mean “tragedy,” and further ancient sources. See Boyle 2006, 27–36, and Feeney 2016, 45–64, for a biography of Livius Andronicus and summary of his work.

8. Feeney 2016, 17–44, 131–32, and passim.

9. Feeney 2016, 140–41, citing McElduff 2013, 78.

10. See, in particular, Feeney 2016, 62–64 and 116–36. He further applies Taussig’s idea of “mimesis and alterity”: “The competing cultures, then, oscillate between concentrating on otherness, by focusing on what is different about their rivals, and concentrating on similarity, by the imitative process that best enables them to define and master what makes up that otherness,” 133.

11. See further Feeney 2016, 80–81, citing Wardy 2000, 87, 140–51, 181–85; and Feeney 2016, 179–98 and passim. In a similar vein, Gildenhard (2010, 179) has argued that these shifts “introduce inevitable disjunctions between discourse and practice (especially if the texts have their origins in a foreign culture) and thereby expand the scope of what becomes conceivable in a given society.”

12. See Butler 1993, 12–16, for a clear but concise summary of her theory of the “assumption” of sexual identity and her debt to Lacan. See also Butler 2008, 183–93, for an earlier formulation of gender identification and performativity.

13. Butler 1993, 12–16, 93–119, and 244–46 n. 7–8; 2008, 185. Butler borrows the idea of the “citation” from Derrida in order to expand Lacan’s idea of the “assumption” of sex. She argues that an individual’s assumption of “sex” must be understood as constrained by ideology already at work in the historical and cultural context into which an individual is born. Derridean citationality introduces the idea of the iterability and performativity of an individual’s assumption of a “sex” which is “always derivative” (1993, 13), i.e., a choice of sexual identity already authorized and acknowledged as a legitimate sexual identity.

14. “It is thus not lack of cleanliness or health that causes abjection but what disturbs identity, system, order. What does not respect borders, positions, rules,” Kristeva 1982, 4. See further Kristeva 1982, 1–31 and passim, for her formulation of “abject.”

15. See also Butler 1993, 3–16 and 93–119, for further explanation of her understanding of “abject.”

16. See Curley 2013, 14–17 and 179–85, for a full explanation of code-modeling and his application of Conte’s concept. The “codes” of Phaedra and Medea encompass not only their characteristic dilemmas but also pointed allusions to their Euripidean instantiations and the entirety of their mythological “lifetimes,” or as Curley (2013, 180) terms it, the “synchrony” of their myths. Curley (2013, 177–85) follows, among others, Larmour 1990 and Newlands 1997 in noting how Ovid cites a heroine’s code through intratextual and Alexandrian footnotes, signaling that a reader should recognize, for example, Hypsipyle’s threat to play Medea in her letter to Jason (Her. 6).

17. Curley (2013, 4) also limits his use of “tragic” to its generic sense.

18. See Curley 2013, 14–18.

19. See Conte 1986; Keith 2000; and Panoussi 2009 for the tragic figure in Vergil’s Aeneid.

20. See also earlier studies by Bartsch 1997; Atherton 1998; and Murgatroyd 2007.

21. The first Hippolytus (Hipp. I) featured a Phaedra whose representation was by all accounts more aggressive and threatening. She probably approached Hippolytus herself. Hipp. I was sometimes referred to as Hippolytus Kalyptomenos because the character covered himself in shame at her proposal. See Barrett 1964, 11. Perhaps mocking this play in the contest between Aeschylus and Euripides of Aristophanes’s Frogs, Aeschylus states that he never made Phaedras or Stheneboias into prostitutes (ἀλλ᾽ οὐ μὰ Δί᾽ οὐ Φαίδραc ἐποίουν πόρναc οὐδὲ Ϲθενεβοίαc, 1043). For a discussion of Euripides’s two Hippolyti and Sophocles’s Phaedra, see Barrett 1964, 10–15; Snell 1964, 23–69; and Webster 1967, 64–76. See Herter 1940 for the mythological tradition and its Greek and Roman reception.

22. Barrett (1964, 6–10) conjectures that Phaedra’s part in the Hippolytus legend goes back to the sixth century BCE when Theseus was adopted as an Athenian hero. She is mentioned in Hom. Od. 11.321–25 in connection with Procris, Ariadne, and Theseus. Barrett suspects this passage to be an Attic insertion also dated to the sixth century. See the fourth-century Asclepiades’s Tragodoumena (FGrH 12 fr. 28), Paus. 1.22.2, and Apollod. Epit. 1.18–19 for variants on the myth not represented in Euripides’s extant play. See further Barrett 1964, 26–27, and Coffey and Mayer 1990, 9–10 and n. 30. The Alexandrian playwright Lycophron wrote a Hippolytus, but there is no evidence of Roman tragic treatments before Seneca. Pfeiffer 1968, 119–20, and Coffey and Mayer 1990, 10 and n. 31.

23. See Boyle 2012, 4–5, for a summary of Euripides’s treatments of Medea with bibliography. Boyle draws attention to the many mythological elements to which Euripides’s play Medea is our earliest testament, including the adultery, poisoned robes, Aegeus’s visit to Corinth, her escape in a flying chariot, and, especially, the intentional infanticide by Medea. For variants on the children’s death, see Boyle 2012, 3–4.

24. Foley (2001, 243–71) interprets Medea’s famous struggle at Med. 1021–80 not as a struggle between her reason and her passion but as her “force (e.g., courage) that directs the self to action,” guiding her “plans” (θυμὸς δὲ κρείσσων τῶν ἐμῶν βουλευμάτων, 1079; Foley 2001, 253). Foley argues that Medea’s plan for revenge follows the archaic model of “helping friends and harming enemies” familiar from Homer’s Achilles.

25. Hyg. Fab. 26; Apollod. Bibl. 1.9.28, Epit.1.4–6; Paus. 2.3.7–8; Plut. Thes. 12; Diod. Sic. 4.55–56; Ov. Met. 7.402–24. Sfyroeras (1994–95) has argued that Euripides’s Medea also looks forward to her Athenian career by shifting her relationship to her sons from mother to stepmother. Euripides’s Aegeus may have been set in Athens. See Webster 1967, 77–80, for a conjectured outline of the play, the evidence, fragments, and dating. Sophocles’s Aegeus also told the story of Theseus in Attica and may have included Medea. See Sutton 1984, 5–6, who further cites Pearson 1917, 15–21; Welcker 1938, I.393; Bates 1961, 166; and TrGF 4.19–25.

26. Sophocles’s Colchides may have also represented a young Medea in love as well as the murder of Apsyrtus in Aeetes’s palace. Scholiasts on Apollonius’s Argonautica cite Sophocles’s play, suggesting that it was a source for the Hellenistic epic. Pearson 1917, 15–23; Welcker 1938, 1.333; Bates 1961, 189; TrGF 4.336–49; Sutton 1984, 32–33; and Boyle 2012, 4.

27. These episodes were also featured on the tragic stage. See Boyle 2012, 3–5, and Gildenhard and Zissos 2013 for a summary of fifth-century BCE Greek treatments of Medea with bibliography. A fourth-century tragedian, Carcinus, also wrote a Medea (Aris. Rh. 2.23.28, 1400 b8 = Carcinus TrGF 1.70, 1e). On the fragment, see further West 2007 and Hall 2010, 18 and 24 n. 20. Aeschylus’s Trophoi may have dramatized her rejuvenation of Dionysus’s nurses; Sophocles Rhizotomoi may have featured a magical Medea; Sophocles’s Scythae may have followed the Argo home. Euripides’s Peliades described Pelias’s death.

28. See, e.g., Clauss and Johnston 1997, 6–7, and Bartel and Simon 2010, 1–2, for the complexity of Medea in myth.

29. See Polt 2013, 207 n. 14, with further bibliography. Boyle (2012, 30–31 n. 59) summarizes the account of Medea’s myth given by the first-century BCE historian Diodorus Siculus (4.45–56), which diverges in small ways from her more familiar variants. One detail significant to Ovid is the practice of sacrificing strangers who arrive in Colchis, which resembles the story of Iphigenia in Tauris (Euripides), put to use by Ovid in his exile poems.

30. See further Citroni 2003, 171–80; Farrell 2005, 427; and Feeney 2016, 152–78.

31. Feeney (2016, 169) cites Cowan 2013, 330.

32. See further Feeney (2016, 53–55, 62–64, 268–89 nn. 51–59), who cites Hinds 1998, 60–61.

33. Cowan 2013, 170–71.

34. Fitch (1987, 50–53) has settled on an early date for Phaedra of 41–54 CE, based on stylistic analysis.

35. Feeney 2016, 167, 305 n. 66.

36. See chapters 1 and 2 for evidence of Phaedra-like characters in Roman drama.

37. See Feeney 2016, 122–27, for the importance of drama in Sicily at this time.

38. See, e.g., Hinds 1993; Heinze 1997, 9–10; Newlands 1997; Fantham 2005, 119–20; Boyle 2006, 71–78, 92, 94, and 115–17; 2012, 6–16; Cowan 2010; and Curley 2013, 19–58 for Medea in Roman tragedy.

39. Medea Exul /Medea, ca. 203–169 BCE (Joc. frr. 103–16).

40. Medus, ca. late third century–ca. 140 BCE (Warmington fr. 231–65).

41. Medea sive The Argonautae, second half of the second century BCE (Dangel frr. 467–99).

42. See Jocelyn 1967, 342–47, for evidence and further bibliography on these plays. Jocelyn (1967, 122, 378–79) identifies fr. 112 as belonging to a second play by Ennius which takes place after the events of Euripides’s extant Medea in Athens.

43. See Vogt-Spira 2000, 273; Boyle 2006, 71–78; 2012, 6–19; Cowan 2010; and Feeney 2016, 143.

44. See Beacham 1991, 140–49; 1999, 140–47; Boyle 2006, 171–72; Garelli 2007; Lada-Richards 2007, 2013; Hall and Wyles 2008; Webb 2008; and Zanobi 2010 on the Roman pantomime.

45. Pantomime as a genre was very likely developing over the course of several decades. See Lada-Richards 2013, 110–13, for a discussion of its history. See also Beacham 1999, 142–43, and Boyle 2006, 171–72.

46. See Lada-Richards 2013, 111–12 and nn. 26–27 for pantomime’s long-standing connection to tragedy.

47. See Lada-Richards 2013, 113 and n. 31, for evidence and further references regarding the staging of Vergil (Eclogues) and Ovid (Amores and Heroides) to music. She demonstrates the similarities between the dance of this genre and Ovid’s detailed, body-focused descriptions of metamorphoses. She argues persuasively that readers of his Metamorphoses, familiar with the performance genre, would read Ovid’s verse from this perspective (2013, 115 and passim; 2016; 2018). See further Currie (1981, 2703 n. 4), who cites Ovid at Tr. 2.519–20 (“My poems have often been danced for the people, often they even have held your own eyes,” et mea sunt populo saltata poemata saepe,/ saepe oculos etiam detinuere tuos) and 5.7.25–27.

48. “I am carried here and there, as if full of the god” (feror huc illuc, ut plena deo, Sen. Suas. 3.7); “I was able to save: are you asking whether I can destroy?” (seruare potui: perdere an possim rogas, Quint. Inst. 8.5.6). For a discussion of Ovid’s lost tragedy, see Curley 2013, 19–58, esp. 37–49. For further bibliography on reconstructions of the play, Curley (2013, 21 n. 7) cites Arcellaschi 1990, 231–67, and Heinze 1997, 221–52.

49. See Curley 2013, 102, for a demonstration of this process in Ovid’s “translation” of Euripides’s Hecuba.

50. For the dialectic between elegy and epic in Ovid, see, e.g., Heinze 1919; Otis 1970, 1–44 and passim; Hinds 1987, 1992a and b, and 2000; Keith 2002; and Farrell 2009, 270–80.

51. See, e.g., Sullivan 1972. Farrell (2009, 370) calculates that 65 percent of Ovid’s extant poetry is in elegiac couplets, the hexameter poem Metamorphoses making up the remainder. He notes, however, that Ovid’s lost Aratea was likely composed in hexameter, while his Medea would have been composed in tragic meters.

52. Hinds 1987, 1992a and b. In his famous and influential interpretation of the epic and elegiac genres in Ovid, Heinze (1919) compares the treatment of the Persephone myth as it is represented in the elegiac Fasti and the epic Metamorphoses and concludes that the respective treatments are generically appropriate in theme and language. See also Farrell 2009, 373–76, for a summary of Heinze’s 1919 argument and scholarship responding to his conclusions including Hinds. Barchiesi (2001) and Farrell (2009, 374) note the influence of Kroll’s 1924 monograph which argues that generic transgression is a characteristic of Hellenistic and Roman verse.

53. Farrell 2009, 378.

54. On this poem and further bibliography, see De Caro 2003, 140–42; Curley 2013, 38–49; Westerhold 2013; and Filippi 2015, 198–99. Keith (1994), Wyke (2002, 1989) and Perkins (2011, 313–33) employ the personified Elegy as an interpretive tool for understanding the metapoetic function of the elegiac puella in Roman erotic elegy. 3.1 is commonly cited by those who wish to reconstruct Ovid’s short-lived career as a tragic poet. See Hollis 1977; Heinze 1997, 223; McKeown 1987–1998, 1:86–89, and 3:394.

55. See Curley 2013, 5–6, for a thorough review of the scholarship on this topic.

56. See, e.g., Jannacone 1953; D’Anna 1959; Bömer 1969–1986; Otis 1970, 376; Lafaye 1971, 141–159; Anderson 1972, 1997; Currie 1981; Larmour 1990; Ciappi 1998; Curley 1999, 2003, 2013; Gildenhard and Zissos 1999, 2007, 2013; Keith 2002, 2010b; Feldherr 2004, 2008, 2010; Williams 2012; and Westerhold 2013, 2014.

57. See, e.g., Jacobson 1974; Currie 1981; Barchiesi 1993; Hinds 1993; Bessone 1997, 11–41; Kennedy 2002; Davis 2012; and Curley 2013. See Filippi 2015 on archaic Roman tragedy in Ovid’s elegy.

58. See, e.g., Davisson 1984; Ingleheart 2010; and Boyle 2012, 21.

59. See Curley 1999, 2003, 2013; Gildenhard and Zissos 1999, 2007; and Keith 2002. Similar work has been done for Vergil’s Aeneid. See, e.g., Hardie 1997; Keith 2000, 2020, 91–117; and Panoussi 2009.

60. Goldberg 1996, 571; 2000, 52 and passim; 2005; Boyle 2006, 154; 2012, 16–18; Gildenhard 2010, 179–80; and Feeney 2016, 86–88, 131–36. See Bourdieu’s “Cultural Reproduction and Social Reproduction,” in Knowledge, Education, and Cultural Change, ed. Richard K. Brown, 71–112. London: Tavistock, 1973, for an explanation of the term “cultural capital,” which he uses to describe knowledge that identifies membership in superior social groups. In her treatment of Woman in Propertius, Fantham (2006, 189) notes that Propertius relies largely on mythological exempla drawn from Homer, tragedy, and Apollonius, for “these would be what his public had read or heard or seen, so that a phrase or even a mere epithet could evoke remembered texts or images without need for narrative.”

61. Cicero cites the opening lines in seven of the thirteen instances; Jocelyn 1967, 52–53. Canter (1936, 40–41) identifies the myth of Jason and the Argo as a frequent exemplum in Cicero’s works. See further Zillinger 1911; Jocelyn 1973; Shackleton Bailey 1983; Goldberg 2000; Gildenhard 2007; Zetzel 2007; and Boyle 2012, 16–19, on Cicero’s use of tragedy. Shackleton Bailey (1983, 244–45) provides statistics which he confirms against Zillinger’s (1911, 244 n. 33) data; Zetzel 2007, 2. See Zillinger 1911, 50–68, for citations of poets by genre in Cicero, and 97–124 for citations of Ennius.

62. Segal (1986, 52) defines this “megatext” as “not merely the totality of themes or songs that the poets of an oral culture would have had available in their repertoires but also the network of more or less subconscious patterns, or deep structures, or undisplaced forms, which tales of a given type share with one another.”

63. Her. 4 may also allude to moments from Euripides’s Hipp. I and Sophocles’s Phaedra which we do not recognize. For a careful comparison of Her. 4 to the extant Euripidean play as well as to the fragments and evidence for the two lost plays of Euripides and Sophocles, see, e.g., Jacobson 1974, 142–45, and Casali 1995.

64. Larmour 1990, 137–41. “The ‘pudor-amor’ conflict, the nurse’s disastrous attempts to help and the overwhelming power of the pathological libido—are incorporated, again through a sort of contaminatio, into other episodes: the Myrrha, Byblis and Scylla” (137).

65. Larmour 1990, 137, and Curley 2013, 91–92. Curley (1999, 197–98) notes that, in Scylla’s tale, Minos also fills this role to a certain degree, pointing to Minos’s astonished reaction (turbatus, 8.96) to Scylla’s gift of her father’s lock, which resembles Caunus’s reaction to Byblis’s letter (attonitus, 9.574).

66. Barrett 1964, 13.

67. Myrrha decides to hang herself because she is unable to suppress her incestuous desire (Met. 10.378–79). Her nurse discovers her and begs to intervene (382–430), at one point offering spells and herbal remedies (carmine sanet et herbis/ … magico lustrabere ritu, 397–98). It is the nurse who approaches Cinyras and arranges for the meeting. Phaedra in Hipp. II eventually hangs herself after writing her letter of accusation against Hippolytus (800–2), but at the beginning of the play, Phaedra has resolved to die (400–401). Her nurse intervenes (284–310, 433–81), offering spells and enchanting words as a cure (ἐπωιδαὶ … λόγοι θελκτήριοι/ … τι τῆcδε φάρμακον νόcου, Hipp. II, 478–79), and approaches Hippolytus in an attempt to seduce him on Phaedra’s behalf (565–615). See Anderson 1972, 501–17 ad 298–502; esp. 508–9, ad 371–76, 377–79; Larmour 1990, 138; and Bruzzone 2012.

68. For a comprehensive treatment of Euripides’s Medea as source text for Ovid’s own Medeas, see Curley 2013, 121–33, 141–53, and passim.

69. For a discussion of Ovid’s lost tragedy, see Curley 2013, 19–58, esp. 37–49. For further bibliography on reconstructions of the play, Curley (2013, 21 n. 7) cites Arcellaschi 1990, 231–67, and Heinze 1997, 221–52.

70. See further Jacobson 1974, 103, 120–23; Knox 1986, 1995, 196–201; Hinds 1993; Lindheim 2003, 125, 226 n. 104–6; Michalopoulos 2004; and Curley 2013, 177–216.

71. Curley (2013, 43) cites Currie 1981, 2704, who argues that what the poet omitted in the Met. he treated in his tragedy. See Williams 2012; Curley 2013, 43 n. 119; and Gildenhard and Zissos 2013 for further bibliography.

72. Newlands 1997, 192–95; Curley 2003, 186; and Williams 2012. See further Curley 2013, 177–216 and passim, for a broader treatment of Medeas in Ovid’s oeuvre.

73. “One, in fact, one woman of those before, I hear, who threw her hand on her beloved children: Ino driven mad by the gods” (μίαν δὴ κλύω μίαν τῶν πάρος/ γυναῖκ᾽ ἐν φίλοις χέρα βαλεῖν τέκνοις·/ Ἰνὼ μανεῖσαν ἐκ θεῶν).

74. Aristophanes also revised a play, Clouds, after a poor reception at the City Dionysia in 424–423 BCE, although it is unclear whether the revised version was produced or merely circulated as a text (Dover 1968, lxxxi). The extant Clouds, like Euripides’s Hipp. II, reflects the revised play. See Dover 1968, lxxx–xcviii for a discussion of the evidence.

75. Barrett 1964, 13.

1. Medea: Ars am. 1. 1.283–340, 2.99–104, 2.373–408, 3.32–40; Rem. am. 41–68, 261–88. Varro Atacinus’s Argonautae appears on a recommended reading list at Ars am. 3.329–48. Phaedra: Ars am. 1.283–340, 1.509–12, 1.741–46; Rem. am. 41–68, 741–49.

2. The dating for Ovid’s poetry is uncertain. Heinze (1997, 21) dates the first two books of Ars am. to the years 6–1 BCE, situating their publication after the first edition of Am., Medea, and the single epistles. He dates the third book of Ars am. to the years 1 BCE–1 CE, just after the second edition of Am. and Rem. am. and before his longer poems Fast. and Met. McKeown’s (1987–98, vol 1, 77) conjecture is 2 BCE–1 CE; Hollis (1977, xiii), before spring of 2 CE. Holzberg (2006, 40–42) follows scholars who argue that all three books were written together. He provides a summary of the arguments on both sides with bibliography. See further, e.g., Hollis 1977, xi–xiii; McKeown 1987–98, vol 1, 74–89, vol. 3, 384–86; Heinze 1997, 21–24; Gibson 2003, 37–43; Keith 2010a; and Curley 2013, 38–40.

3. See further in the introduction “Ovid, in Theory” and Kristeva 1982, 1–31.

4. See further in the introduction “Ovid, in Theory” and Butler 1993, 3–16, 93–119.

5. Segal’s (1986, 57) understanding of myth as a semiotic system derives from other theorists, including and especially Barthes (Mythologies, 1972) and Lévi-Strauss.

6. See also the introduction, “Ovid on the Intersection.”

7. “Therefore, however many understand the writings of the ancients and themselves are always in the company of the Muses know that Zeus once desired marriage with Semele, and they know that once beautifully shining Dawn seized Cephalus and carried him off to the gods because of desire: nevertheless, they dwell in the sky and they do not flee away from the gods.” (ὅϲοι μὲν οὖν γραφάϲ τε τῶν παλαιτέρων/ ἔχουϲιν αὐτοί τ᾽ εἰϲὶν ἐν μούϲαιϲ ἀεὶ/ ἴϲαϲι μὲν Ζεὺϲ ὥϲ ποτ᾽ ἠράϲθη γάμων/ Ϲεμέληϲ, ἴϲαϲι δ᾽ ὡϲ ἀνήρπαϲέν ποτε/ ἡ καλλιφεγγὴϲ Κέφαλον ἐϲ θεοὺϲ Ἕωϲ/ ἔρωτοϲ οὕνεκ᾽· ἀλλ᾽ ὅμωϲ ἐν οὐρανῶι/ ναίουϲι κοὐ φεύγουϲιν ἐκποδὼν θεούϲ, Eur. Hipp. II, 451–57). “Cypris gave in a deadly marriage to Alcmene’s son the Oechalian filly, unyoked to a marriage bed, formerly unhusbanded, unbetrothed, and joined her in wedlock away from the house of Eurytus, like a wildly roaming naiad and a Bacchante with blood, with smoke: O wretched in marriage. O holy wall of Thebes, O font of Dirce, you may confirm how Cypris comes: for, after wedding the mother of twice-born Bacchus to the flaming thunder, she put her to bed with a deadly destiny. For she inspires all terribly, like some bee flying here and there.” (τὰν μὲν Οἰχαλίαι/ πῶλον ἄζυγα λέκτρων,/ ἄνανδρον τὸ πρὶν καὶ ἄνυμφον, οἴκων/ ζεύξαϲ᾽ ἀπ᾽ Εὐρυτίων/ δρομάδα ναΐδ᾽ ὅπωϲ τε βάκ-/ χαν ϲὺν αἵματι, ϲὺν καπνῶι,/ φονίοιϲι νυμφείοιϲ/ Ἀλκμήναϲ τόκωι Κύπριϲ ἐξέδωκεν·/ ὦ τλάμων ὑμεναίων./ ὦ Θήβαϲ ἱερὸν/ τεῖχοϲ, ὦ ϲτόμα Δίρκας,/ ϲυνείποιτ᾽ ἂν ἁ Κύπριϲ οἷον ἕρπει·/ βροντᾶι γὰρ ἀμφιπύρωι/ τοκάδα τὰν διγόνοιο βάκ-/ χου νυμφευϲαμένα πότμωι/ φονίωι κατηύναϲεν./ δεινὰ γὰρ τὰ πάντ᾽ ἐπιπνεῖ, μέλιϲϲα δ᾽ οἵ-/ α τιϲ πεπόταται, Eur. Hipp. II, 545–64).

8. See Treggiari 1991, 3–36, for a detailed discussion of marriage customs in Rome, and 107–19 for the relationship of affinitas created by marriages.

9. McGinn 1998, 70, and Treggiari 1991, 277.

10. McGinn 1998, 92–93, 143, 156–71.

11. See McGinn 1998, 140–215, for a careful analysis of the lex Julia de adulteriis coercendis and its social and economic implications. See Edmondson 2008, 37–38, on the importance of the toga and stola as a costume in “stagecraft” (tragoedia; citing Polyb. 6.53–54, 56.9), defining the Roman community of men and women.

12. McGinn 1998, 171–94.

13. Segal (1986, 53, 57), citing earlier research (53, n. 13), traces literary paradigmatic analogies to Homer, which he conjectures predate the poet: “Homer is probably developing a systematic coherence already present in the mythic material.”

14. I am following the dating for Propertius’s third book of Goold 1990, 2; Hollis 1977, xiii; and McKeown 1987–98, vol 1, 75. Propertius, book 3, ca. 23–21 BCE; Ovid, Ars 1, 2, sometime between 20 and 2 BCE, but most likely closer to 2 BCE.

15. Bergmann (1996) has applied the term “pregnant moment” to these frescoes. For a careful discussion of this house and its frescoes, see Bergmann 1996 and Valladares 2021, 158–73.

16. Bergmann (2017, 218) describes the frescoes as depicting the “perfect moment” of each myth. The villa was built ca. 127 CE. The paintings may be as late as the third century CE. See, most recently, Newby 2016 and Bergmann 2017. The original placement of these paintings in the room is not known. See Bergmann 2017, 211–12, for conjectures.

17. Bergmann (2017, 209–23) argues that “galleries of heroines” can be traced to the fifth century BCE in Greece, and, further, that these galleries were influenced by and influenced similar catalogues in literature and performances on the stage. See Fredrick 1995, 274–75 and passim, Bergmann 2017, and Valladares 2021 for the associations suggested by painting groups, and Bergmann 2017 in relation to the Tor Marancia grouping in particular.

18. See Bergmann 2017, 213–15, and Valladares 2021, 158–73, for a discussion of Ovid’s literary catalogues and his Her. in relation to these groups.

19. MSS R, Y, P_a, and the second hand of MS H read agat. MS O alone reads cogat. MS A and all other manuscripts read aget. See Kenney 1959, 246–47, for arguments in support of reading aget.

20. I would suggest that in line 331, the participle furata which modifies filia hints at but suppresses the Megarian Scylla’s desire for Minos through its acoustic similarity to furor.

21. See Leach 1964 for the representation of women’s desire as excessive in the Ars am.

22. OLD, agere, 25.

23. Ovid’s own tragedy, Medea, may have been finished by the time of the Ars am. See McKeown 1987–98, vol 1, 77, 87.

24. Pasiphae: Euripides, Cretans; Accius’s Minos/Minotaurus. Aerope: Euripides, Cretan Women, according to the scholion on Aristophanes’s Frogs (Σ Batr. 849), and perhaps Sophocles’s Atreus. See Owen 1924, 216 ad 2.391 for a complete list of evidence of tragedies on the subject, including Eur., Cretan Women, and Webster 37–39 for a summary of evidence on the subject matter of Cretan Women. Scylla: according to Ovid himself, the subject of a play by an unknown tragedian (cf. Tristia 2.393–94: inpia nec tragicos tetigisset Scylla coturnos, / ni patrium crinem desecuisset amor); Hollis 1977, 91 ad 331–32; Owen 1924, 217 ad 2.393. Clytemnestra: Aeschylus, Oresteia; Sophocles, Electra; Euripides, Electra; Ennius, Eumenides, Iphigenia at Aulis; Accius, Clytemnestra or Aegisthus? Medea: Aeschylus, Trophoi; Sophocles, Aegeus?, Colchides, Rhizotomoi, Scythae; Euripides, Aegeus, Medea, Peliades; Carcinus, Medea; Ennius, Medea Exul / Medea; Pacuvius, Medus; Accius, Medea sive Argonautae; Ovid, Medea. Phthia: Euripides, Phoenix; Ennius, Phoenix (cf. Webster, 84–85). Phaedra: Euripides, Hipp. I and II; Sophocles, Phaedra. And Phineus’s wife: perhaps the subject of Aeschylus, Phineus, and Sophocles, Tympanistae and Phineus A and B.

25. These “tragic” representations would have been very familiar to Ovid’s audience. In addition to the plays featuring Medea (note 24 above), Romans would have seen Medea represented in the dramatic and poetic genres of pantomime, epic, lyric, and elegy; in the art of private homes and public spaces; and in the speeches of Roman orators. See in the introduction “Phaedra and Medea in Rome” for a summary of these depictions in the Roman republic and early principate with further bibliography. In her discussion of tenderness in Roman art, Valladares 2021, 130 and passim, argues that wall paintings engaged a viewer in meaning-making and narrative construction by recalling images from art and literature but not reproducing one in particular.

26. Loraux (1987, 7–30) argues the most dishonorable suicide, and the one associated with women, particularly in tragedy, was hanging.

27. Ovid’s treatment of Pasiphae reworks Vergil’s Pasiphae in Ecl. 6. See Armstrong 2006, 170–77, 180–81; Hollis 1977, 93 ad 289–326; and Pietropaolo 2020, 178–98, for a detailed discussion of this passage and comparisons of the two Pasiphaes.

28. Compare Vergil’s Pasiphae, who wanders through the mountains (Ecl. 6.52), reflecting the wandering of her senses and equating her with both the masculine pastoral lover and the wandering of the followers of Dionysus. Armstrong (2006, 174–77) has mapped out the instances of pastoral elements in the passage, drawing special attention to a resemblance to Corydon of Ecl. 2 and Gallus of Ecl. 6. See also Hubbard 1975, 61–62, for a similar argument.

29. In the Met., Medea’s struggle over her passion is described as a struggle between ratio and furor (7.10); Scylla is called furibunda (8.107); Byblis calls love furor (9.512, 541, 602), and the narrator refers to her love as a furor (9.583) and her as furibunda (9.637); Myrrha is also called furibunda (10.410).

30. Vergil likens the lovesick Dido to a bacchant (Aen. 4.301–3). In Ov. Met., Byblis is compared to a maenad (9.641–43). The connection between the bacchant and the sexually transgressive woman can be traced back to Greek representations. On this topic in Greek myth and literature, Seaford (1990) argues exposure to the wild was part of the initiation process which resulted in a son’s return to the natal household and a daughter’s transfer to a husband’s (161). “Maenadism represents in an extreme form the loss of control by the male of the female, and was indeed imagined as involving the danger of illicit sex” (163, emphasis added). Seaford offers tragic examples of out-of-control women associated with Bacchic worship (1990, 163–64); see also Pease 1967, 278–83; Warden 1978, 181–82; Keuls 1985, 349–79; Miller 1995; Janka 1997, 296–98; Janan 2001, 76–78; Gibson 2003, 367–68; Armstrong 2006, 97–98; Gardner 2013, 176, and n. 64; Spentzou 2018, 264; and Panoussi 2019, 117–67, for an analysis of this topic and a comprehensive list of comparanda from Greek and Latin literature.

31. Cf., e.g., Eur. Bacch., where Pentheus charges the maenads with drunkenness and that “each one in different directions slinks off to a secret place and serves the beds of men” (ἄλλην δ᾽ ἄλλοϲ᾽ εἰϲ ἐρημίαν/ πτώϲϲουϲαν εὐναῖϲ ἀρϲένων ὑπηρετεῖν, 222–223); or the Roman charges in the Bacchanalia, which included “the promiscuous debauchery of free men and women,” stupra permiscua ingenuorum feminarumque, Livy 39.8.

32. See Conte 1986, 159–84; Hardie 1997; and Panoussi 2009 on tragic contamination in Vergil’s Aeneid; for Ovid, see in the introduction “Generic Performances.”

33. Zeitlin (1996, 343–44) has argued that the feminine aspects of Dionysus reflected and associated the god with the intrinsically feminine genre of tragedy.

34. See Hollis 1977, 93 ad 289–326 for references; see Webster 1967, 87–92, and Collard and Cropp 2008, 516–19, for evidence and a summary of the play.

35. See Seaford 2006, 26–38, for an interesting discussion of the (mythical) kin-killing violence of the thiasos and its potential for community-building. Seaford argues that the dissolution of boundaries the individual participants experienced and metaphorical violence done to the private family created a strong cohesion in the community which was then harnessed through state-controlled festivals. Moreover, as the group most closely associated with the private sphere, the house, women figured prominently in myths of resistance to Dionysus and were the best representatives for the ritual dissolution of the private sphere in the actual performance of his rites (34). See also Panoussi 2019, 117–19, on this symbolism in Latin authors.

36. Livy (39.8–19) provides a description of the suppression and the events leading up to the senatorial decree; for the inscription preserving the senatus consultum see CIL I.196 and ILS 18. See Seaford 2006, 58–61, for a discussion of the social and political significance of the suppression.

37. See also Takács 2000 and Panoussi 2019, 120–39, for a discussion of the suppression and the mechanisms of the religious and ideological control used by the Roman elite. While Takács argues that the suppression of the worship of Bacchus was merely an excuse for exercising such control and curtailing Hellenistic influence, Livy’s account (39.8–19), probably published during Ovid’s early writing career, provides a different perspective. Anxiety over gender reversal and sexual deviance is evident in Livy’s narrative of the event: “In Livy’s narrative, the cult of Bacchus represents disorder and madness while the state represented by the (all male) Senate stands for order and sanity” (Takács 2000, 310), attesting to the reception of the event at the dawn of the Augustan era. For the dating of Livy’s Ab Urbe Condita, see Syme 1959; Walsh 1961, 1–19; and Ogilvie 1965, 1–2.

38. The suppression included over 6,000 executions. See Livy 39.17–19 for the sentencing and punishment of the conspirators named. Bauman (1992, 35–37) argues that the Senate’s suppression was “anti-feminist” and motivated by a fear of a primarily female cult plotting against the state. Most interesting is his discussion of the use of coniuratio by Cato (De Coniuratio; Livy 34.2.3) and Terence (Hec. 198; Hyg. Fab. 15:507) in reference to this and other women’s “movements” (36). See also Gruen 1990, 34–78.

39. Deiphobus describes Helen simulating Bacchic worship while aiding the Greeks hiding in the Trojan horse (Verg. Aen. 6.517–19). Amata, inspired (inspirans, 7.351) by a snake from Allecto’s hair, feigns Bacchic worship for the sake of hiding Lavinia and preventing her marriage to Aeneas, a marriage contract struck between the Trojan leader and her husband, Evander (7.385–88); the description of the feigned rites continues through line 405. Allecto finds Amata already distressed over the marriage, a distress qualified as womanly (“[Amata], whom a womanly anxiety and anger was tormenting over the arrival of the Teucrians and Turnus’ marriage,” quam super aduentu Teucrum Turnique hymenaeis / femineae ardentem curaeque iraeque coquebant, 7.344–45). In addition to the fire imagery, which we see in the elegiac descriptions of desire, Amata’s resulting furor (“without delay she raves out of her mind through the city,” sine more furit lymphata per urbem, 7.377) is inspired explicitly to disrupt the household: “so that, maddened by the monster [snake], she might throw the entire house into confusion,” (quo furibunda domum monstro permisceat omnem, 7.348). See Panoussi 2009, 124–33, on this passage.

40. See Miller 2004, 166–67, for a detailed discussion of intertextual relations between Antony, Amor, and Bacchus in Roman elegy (Am. 1.2.43–48, 51–52, and Prop. 2.16.41–42). See further Miller 2009, 18 and nn. 15, 17, 26–28, 176–77, 180–81, for Antony’s self-identification with Bacchus/Liber/Dionysus.

41. This is in contrast to Pasiphae’s speech in Euripides’s Cretans (TrGF 5.1.472e, quoted below in note 51) where Pasiphae points to the ridiculousness of this possibility as proof of her innocence.

42. See OLD, vir, 1., “an adult male;” 2a., “a husband”; see also Williams 1968, 528–29, 539, 542; Hollis 1977, 95 ad 310; Davis 1999, 445–46; and Miller 2004, 170–74 for the ambiguity of this term, particularly in elegy.

43. Oedipus famously embodied the tragic lack of self-knowledge and his anagnorisis (“recognition”) articulates an important part of the resolution of the tragic plot (cf. Arist. Poet. 11.1452a–b). Pasiphae’s daughter Phaedra compares time to a mirror in Eur, Hipp, II, 428–30: “Time, holding up a mirror as though to a young maiden, sometimes reveals the base among mortals, among whom may I never be seen” (κακοὺϲ δὲ θνητῶν ἐξέφην᾽, ὅταν τύχηι,/ προθεὶϲ κάτοπτρον ὥστε παρθένωι νέαι,/ χρόνοϲ· παρ᾽ οἷϲι μήποτ᾽ ὀφθείην ἐγω). Goff (1990, 23–24) argues that this passage is a case of misidentification, that it reveals her subjectivity caught between anticipation of her transgression and a nostalgia for her lost innocence.

44. See Pietropaolo 2020, 194, for the grotesque nature of this passage.

45. I am grateful to Cornell University Press’s anonymous reader for drawing my attention to Pasiphae’s abject offspring, the Minotaur. At Her. 4.57–58, Phaedra calls her mother’s pregnancy a crimen in a list of Cretan abject desire: Pasiphae mater, decepto subdita tauro, / Enixa est utero crimen onusque suo. Pietropaolo (2020, 197–98) points to a later mention of the minotaur at Ars am. 2.24 in connection to Daedalus’s escape.

46. If we assume that Ovid intentionally combined the two mythological characters, the monster of 332 could be read as another example of active female desire, whose abject “natural” form is unnatural and monstrous. There is no metamorphosis. The Scylla of line 331 is already the monster of 332. Ovid repeats this “mistake” on several occasions, as does Prop. 4.4.39–40, but not 3.19.21–28; and Verg. Ecl. 6.74–76. For a full list of hybrid Scyllas and references, see Hinds 1993, 15, n. 14, and Casali 2007, 181, n. 1. See Hopman 2012 and Lowe 2015, 73–84, for Scylla’s representation as monster in Greek and Roman art and literature.

47. It seems not even Creusa can escape the discourse of female desire, for her death by fire, flamma, recalls the use of flamma to describe masculine desire in line 282. In addition, so far as we know, the adjective Ephyraea used to describe Creusa occurs only once prior to this in Latin literature, in Prop. 2.6.1 to modify the name of a Greek prostitute from Corinth in a poem averring that Cynthia is more promiscuous than infamous prostitutes (“Not in the same way were they filling the house of Ephyraean Lais,” non ita complebant Ephyraeae Laidos aedis).

48. See Hollis 1977, 98 ad loc. for a summary of the story and Webster 1967, 84–85, for a summary of Eur. Phoenix.

49. See Hollis 1977, 98 ad loc. for references.

50. Cf. Prop. 3.19.11–22, where only the names of the first two examples, Pasiphae and Tyro, are suppressed.

51. While it is hard to tell from the extant fragments, Pasiphae’s speech in Euripides’s Cretans suggests that she is either responding to specific accusations of this kind or she is anticipating these charges. She denies any voluntary, and therefore active, desire, placing the blame upon the gods (ἐκ θεοῦ, TrGF 5.1.472e.9), exonerating her from actively playing the part of her own exchanger by choosing (another) man for a husband (οὐ μὴν δέμαϲ γ’ εὖ.[ca. 8 ll. ν]υμφίου˙ / τοιῶνδε λέκτρω[ν οὕνεκ᾽ εἰϲ] πεδοϲτιβῆ / ῥινὸν καθιϲ.[ca. 15 ll.]ται / ἀλλ᾽ οὐδὲ παίδων. [ca. 9 ll.] πόϲιν / θέϲθαι, TrGF 5.1.472e.16–20). Of note is that her argument relies on what is likely (ἔχει γὰρ οὐδὲν εἰκόϲ, TrGF 5.472e.11)—in other words, what is within the bounds of intelligible and reasonable human behavior. In addition, she argues that she would be an adulteress (μάχ[λο]ϲ) if she threw her body at a man (μὲν ἀνδρί): ἐγ[ώ] γὰρ εἰ μὲν ἀνδρί προὔβαλον δέμαϲ / τοὐμόν λαθραίαν ἐμπολωμένη Κύπριν, / ὀρθῶϲ ἄν ἤδη μάχ[λο]ϲ οὖϲ᾽ ἐφαινόμην, TrGF 5.1.472e.6–8). The implication is that, because her lover is not a man but a bull, the current situation (νῦν δ’, 9) is not recognizable as adultery. Ovid’s poet-praeceptor expresses a similar sentiment: siue uirum mauis fallere, falle uiro, Ars am. 1.310. The poet-praeceptor, perhaps echoing Pasiphae herself, urges like-for-like trade, with a play on the meanings of uir. See Hollis 1977, 95 ad 310; and Leach 1964, 143 n. 1a. In both of these passages, adultery with a man is recognizable and nameable (μάχ[λο]ϲ, fallere), although punishable. The madness, which in Euripides can only be explained as divinely inspired (ἐκ θεοῦ γὰρ προϲβολῆϲ, TrGF 5.1.472e.9), is, in Ovid, proof and punishment of Pasiphae’s abject position.

52. According to Butler (1993, 14–15), “the symbolic ought to be rethought as a series of normativizing injunctions that secure the borders of sex through the threat of psychosis, abjection, psychic unlivability …”

53. See above for a discussion of Dionysiac imagery as a generic marker in this passage.

54. Euripides and Sophocles both wrote tragedies of this myth. In Rome, Ennius, Accius, and Varius Rufus wrote plays depicting the two brothers. Varius’s play Thyestes was commissioned and produced for Augustus’s triple triumph in 29 BCE. For a full discussion of the tragic treatments of this myth, see in chapter 3 “Procne.”

55. See above n. 24 for the tragic plays depicting these myths.

56. See above n. 2 for conjectured chronologies of Ovid’s works with bibliography.

57. See further Leach 1964, 145, on the imagery of woman’s rage in this passage.

58. In book 1 the poet-praeceptor characterizes women as master dissimulators of erotic passion. Cf. uir male dissimulat, tectius illa cupit, 1.276, discussed above. He will later warn women to avoid rage because this emotion disfigures the face of women (3.501–8).

59. Compare Byblis’s entry at 1.284: arsit et est laqueo fortiter ulta nefas, with ulta in the same sedes as Medea’s above. In this line we may find the collocation of erotic fire (arsit) and revenge (ulta). Moreover, although Byblis’s revenge targets only herself, the outcome disrupts by ending a family line. She may not be used to create an affinity with another man, nor will she produce heirs to continue the family name.

60. See Feldherr 2008, 2010, 199–239, for Ovid’s intertextual development and crystallization of Procne as sign in the Met. 6 and Fasti 2, as he does here (signatum, 384). Feldherr argues that this is an objectifying process which allows the reader to define themselves in opposition to the other in Ovid’s verse (or on the tragic stage).

61. For Medea, see n. 24 above. Procne: Sophocles, Tereus; Livius Andronicus, Tereus; Accius, Tereus. See chapter 3 for further discussion of the tragic tradition of both heroines.

62. Procris is similarly described as “a Bacchant driven by the thyrsus” (ut thyrso concita Baccha, Ars am. 3.710; discussed below), and Phyllis, “as if renewing the triennial festival for Edonian Bacchus” (ut Edono referens trieterica Baccho, Rem. am. 593). See also n. 30 for comparanda outside Ovid’s erotodidactic poems and further bibliography.

63. The rumor, moreover, may further allude to Clytemnestra’s speech in Aesch. Ag., in which she complains of the rumors which visit a woman waiting for her husband to return from war (861–65).

64. Cf. “the shameful part is hidden under concealing clothing” (parsque sub iniecta ueste pudenda latet, Ars am. 2.618), describing the famous gesture of Praxiteles’s Cnidian Venus. In Remedia, the poet-praeceptor suggests the lover let himself see “obscene parts” (obscenas partes, 429) or “shameful marks on the dirty couch” (in immundo signa pudenda toro, 432) in order to get over a girlfriend.

65. Cf. “and through such approaches they seek shameful profit” (perque aditus talis lucra pudenda petant, Ars am. 3.442). The lying would-be lovers, who seek shameful (sexual) gains, are characterized as effeminate by the standards of the poet-praeceptor’s lessons, for they are more well dressed and coifed than the women they seduce (“when the man is smoother than the woman herself,” cum sit uir leuior ipsa, 437; see further 433–34, 443–46). Later the adjective describes the threat of rape for women who pass out at the banquet table after too much Lyaeo (Bacchus/wine, 765): “Many shameful things are accustomed to happen during sleep” (per somnos fieri multa pudenda solent, 3.768).

66. Cf. Ars am. 1. 327, where Aerope’s desire is called Thyestean: “If the Cretan woman had abstained from love for Thyestes” (Cressa Thyesteo si se abstinuisset amore). See nn. 24 and 54 above for a list of Greek and Roman tragedies depicting the house of Atreus.

67. See in the introduction “Ovid on the Intersection,” and above for Segal’s formulation of the megatext of myth.

68. See, e.g., Skinner 1993, 111–15, and Gleason 1995, xx–xxvi, 62–67, and passim, on self-control as a defining feature of masculinity, which must be constantly maintained in order to avoid slipping into a feminine position: “Ancient masculinity is thus intrinsically unstable and always at risk, but never so much as in the presence of the sexually experienced female, whose erotic energies are presumed to be boundless and whose erotic demands are correspondingly insatiable” (Skinner 1993, 111). See in the introduction “Ovid, in Theory” and Butler 1993, 12–16, and 2006, 183–93, for Butler’s formulation of gender identification and performativity.

69. See also Feldherr 2010, discussed in chapter 3, for the same reader responses to Ovid’s epic poem, Met.

70. Miller 2004, 160–83, has recently argued that Ovid’s parody of Augustan discourse, while relying on the assumed hegemony of such discourse, unintentionally exposes its contradictions and limits. Gibson (2003, 32–35, 2006, 2007, 112–14), citing 3.305 (sed sit, ut in multis, modus hic quoque) argues that the Ars teaches “an ethic of the middle way” (2007, 3). He maintains this ethics of moderation not only reflects the generic middle position of the Ars, situated between elegy and epic, but is also anti-Augustan, or at least subversive, in that it leads to a confusion of the sexual categories the lex Julia de adulteriis coercendis sought to clearly define.

1. See in the introduction “Ovid on the Intersection” and chapter 1 for an explanation of Segal’s (1986) “megatext” of myth.

2. See in the introduction “Ovid, in Theory” and chapter 1 for a full explanation of this term.

3. Heinze (1997) dates the single epistles to between 16 and 5 BCE, Ars am. 1 and 2 from 6 to 1 BCE, and Metamorphoses between 1 and 8 CE. See further chapter 1, n. 2 and, below, n. 79.

4. See Conte 1994a, 35–66. See also n. 32 below on the elegiac reinterpretation of amicitia, iniuria, and the client-patron relationship.

5. See also Barchiesi 1993, 346: “Ovid’s heroines … are conditioned by an intertextuality which is not simply mythological, but is specifically literary … the reader hesitates about their place in the literary tradition; they are characters who attempt to rewrite their story in terms of a new and different code.… Different genres diffract our perceptions and multiply differences.”

6. French feminist theories, especially the ideas of Cixous and Irigaray, which have not been considered in my reading, offer multiple approaches to women’s writing which could be productively applied to Phaedra and Byblis as female writing subjects. See Gold 1993 for a brief overview of feminist approaches, including theories of women’s writing developed by the French feminists, Cixous (écriture feminine), Kristeva and Irigaray (parler femme), and their utility for reading ancient male-authored texts. In her article, Gold applies Alice Jardine’s idea of “gynesis” to Propertius.

7. See Kristeva 1984, 21–106, especially 25–37, for an explanation of the semiotic chora: “The semiotic can thus be understood as pre-thetic, preceding the positing of the subject. Previous to the ego thinking within a proposition, no Meaning exists, but there do exist articulations heterogeneous to signification and the sign: the semiotic chora” (36). Spentzou (2003, 85–122) characterizes the creative potential of the female authors in the Heroides as a sort of Kristevan chora.

8. “All enunciation, whether of a word or of a sentence, is thetic. It requires an identification; in other words, the subject must separate from and through his image, from and through his objects. This image and objects must first be posited in a space that becomes symbolic because it connects the two separated positions, recording them or redistributing them in an open combinatorial system” (Kristeva 1984, 43).

9. See Kristeva 1984, 43–45, for an explanation of this process for a speaking subject, and 68–71 for the thetic in artistic practices.

10. Janan (1991) identifies the theme of repetition and difference, which surfaces in the Byblis passage as it does throughout the Metamorphoses, as a symptom of a uniquely Roman anxiety over belatedness in the literary tradition.

11. While literature, especially poetry, complicates the thesis of a subject and the symbolic system which this thesis makes possible, a literary work is still located within the logic of such symbolic systems. Therefore, the producer of a text and the textual product rely on the rules governing these systems: “Without the completion of the thetic phase … no signifying practice is possible” (Kristeva, 1984, 63). See Kristeva 1984, 62–67, for poetry’s reliance on symbolic systems and its simultaneous transgression of such systems. Roudiez (Kristeva 1984, 7–8), in his introduction to Kristeva, states: “Consciousness is far from dominating the process and … the writing subject is a complex, heterogeneous force.… The writing subject, then, includes not only the consciousness of the writer but also his or her unconscious.… The subject of writing also includes the non-conscious … an area covered by the notion of dominant ideology” (8). The writing subject’s positionality, i.e., status within a society, gendered, racial, class, and so on, also determines the strength and focus of the repressive function governing their desires and the literary representations they create.

12. Harvey’s (1992) formulation assumes that a man writing as a woman will always be different from a woman writing as a woman. Harvey advocates “a tactical essentialism, the belief that even while we recognize the constructed nature of gender, we can still adhere to a conviction that women and men (and their respective voices) are not politically interchangeable” (13).

13. Harvey’s (1992) critical analysis of the “double-voice” of transvestite ventriloquism focuses on “the gap between the male voice and the female voice it takes on” as a site for revealing gender construction: “In male appropriations of feminine voices we can see what is most desired and most feared about women and why male authors might have wished to occupy that cultural space, however contingently and provisionally” (32). See also Kennedy 2002, 226–31, and Lindheim 2003, passim, especially 136–76, for the double-voice in Ovid’s Heroides. Applying Harvey’s theory, Lindheim argues that the female writers of the Heroides construct themselves in order to sustain or renew their status as the addressee’s object of desire, cheating themselves of any true agency.

14. See Westerhold 2018 on Byblis’s feminine writing voice. See Rosenmeyer 2001 for an analysis of embedded letters in ancient Greek literature.

15. Conte (1994a, 35–66) argues elegy is the most reductive genre because it reinterprets all discourse to suit an all-encompassing erotic ideology. In her discussion of the Fasti, Newlands (1995, 108–9, 140–45) has noted a similar “failure” of elegy to accommodate grander, Augustan material into its code in the tale of Vesta and Priapus.

16. See in the introduction “Phaedra and Medea in Greece” for a full discussion and further references about these plays.

17. Barchiesi (1993) points out that in works such as Apollonius’s Argo. and Ovid’s Her., the allusion looks forward temporally from the point of view of a younger hero/heroine whose future has already been written by other writers (334). This presents the opportunity for dramatic irony, while a shared knowledge of the predetermined outcome creates “a sort of complicity between [the author and the reader] against the characters,” (334) and “forces the reader to enter into discussion with the text he is reading” (335); cf. Kennedy 2002, 224–25.

18. Phaedra announces to the chorus and her nurse that her suicide is meant to preserve her reputation, which would allow her male marital kin to live in Athens as free men enjoying freedom of speech (Hipp. II, 420–23). See Goff 1990, passim, and McClure 1999, 112–57, on the connection of silence, speech, and reputation in Hipp. II.

19. See Jacobson 1974, 142–45, for specific correspondences with Sophocles’s Phaedra and Euripides’s first Hippolytus. According to Jacobson, Her. 4 must follow Hipp. I (like Ov. Met. and Sen. Phae.): “The fact that in all Ovid’s poetry Phaedra is condemned outright, never extenuated, suits best the first Phaedra Euripides portrayed” (144). See also Davis 1995. Barchiesi (1993) argues that allusions point not only to single models but to participation in a broader literary tradition. Adducing Her. 12 (Medea’s letter) and Dido from Verg. Aen., Barchiesi suggests, “Augustan poets are particularly interested in this potential because they write in a tradition which has rich articulations and separate canons” (352). Pace Rosati 1985, who traces Euripidean influences as they are filtered through elegiac mediating texts and codes; see also Hinds 1993 on Met. 12. In addition, Hinds (1993) and Fulkerson (2005) demonstrate an intratextuality between the epistles themselves. See also Kauffman 1986, 42–43.

20. This may be because Hipp. II preserved certain elements from the earlier plays or because the Ovidian Phaedra is an amalgam of all of her tragic instantiations. Jacobson (1974) catalogues intertexts between Ovid’s depictions of Phaedra, Euripides’s Hipp. I, Hipp P. II, and Sophocles’s Phaedra (142–45), where he finds a number of correspondences with Hipp. II along with Hipp. I: “Ovid, we must apparently conclude, knew and was using both of Euripides’ plays (and, just possibly, Sophocles’ too)” (145). See also Casali 1995, passim, for allusions and intertexts with Hipp. II.

21. One could imagine that the Ovidian Phaedra (from Hipp. I) has seen the second play and is trying to present herself and the circumstances accordingly in her letter.

22. Byblis’s story was famous and told to Latin readers by Parthenius (11). Ovid follows this version, in which Byblis hangs herself after her rejection by Caunus, in Ars am. 1.283–84: “Why should I mention Byblis, who burned with a forbidden love for her brother and bravely avenged her unspeakable crime with a noose?” (Byblida quid referam, uetito quae fratris amore / arsit et est laqueo fortiter ulta nefas?). Greek versions included Apollonius of Rhodes (according to Parth. 11 in The Foundation of Caunus), Aristocritus (also according to Parth. 11, in On Miletus), Conon (preserved by Photius of Byzantium), Nicaenetus (quoted by Parth. 11), Antoninus Liberalis (30), Nicander (II), and Hyginus (243.6). See Anderson 1972, 449–50 ad 450–665; White 1982; Celoria 1992, 193–95; Lightfoot 1999, 433–36; and Jenkins 2000, 440–41, for the literary tradition. On Ovid, see also Bömer 1969–1986, 414–15 ad 450–471; Otis 1970, 386–88, 415–17; and Hollis 1977, 92 ad 283–84.

23. See in the introduction “Ovid on the Intersection” for evidence and references connecting Phaedra to these maidens: Scylla, Byblis, and Myrrha.

24. See further Larmour 1990 and Curley 2013, 59–94.

25. See OLD, fero, 38, “to get (from a source), derive”; see Hinds 1998, 2, on feruntur’s use as an Alexandrian footnote.

26. In some cases, love for a husband (e.g., Eur. Alcestis, Helen) or a son (Eur. Ion) has a happy ending.

27. Ehwald prints “quam” at Her. 4.1.

28. For Phaedra’s elegiac self-presentation, see Jacobson 1974, 147; Casali 1995, 3; Davis 1995; and Fulkerson 2005, 126.

29. Jacobson (1974, 147) points to the “juxtaposition of ‘geographical’ adjectives” which emphasize geographic and national distance. “The current relationship, wife and son of Theseus, virtually mother and child, is completely obscured,” and puella viro stands in as a new relation, “sharp with erotic-elegiac overtones.”

30. See James 2003, 47–52, on the relationship of the puella and ego to the traditional figure of vir as husband, and 71–107 on the avida puella.

31. Her argument resembles the argument Plutarch attributes to Phaedra in (presumably) Hipp. I: “that he made a Phaedra who accuses Theseus that she lusted after Hippolytus due to his transgressions” (ὅτι τήν τε Φαίδραν καὶ προσεγκαλοῦσαν τῷ Θησεῖ πεποίηκεν ὡς διὰ τὰς ἐκείνου παρανομίας ἐρασθεῖσαν τοῦ Ἱππολύτου, Mor. 27f–28a; Barrett 1964, 18). In Sophocles’s and Seneca’s Phaedra, Theseus has been absent for so long, he is believed to be dead (TrGF 4.686; Barrett 1964, 12).

32. See Rosati 1985, 119, on Phaedra’s construction of the two as equally victimized by Theseus. See Lyne 1980, 39–40, for Catullus’s use of iniuria as infidelity, especially in poem 72. He argues that, because it engages the language of amicitia with which Catullus describes the sacred foedus between himself and Lesbia, “Catullus means that Lesbia has not just committed wrongs against him (i.e., acts of infidelity), she has committed them with such wilful and inimical intent that the wholeness of Catullan love (exalted amicitia) is now impossible” (40). See also Oliensis 1997a, who charts how the client-patron relationship of amicitia and its vocabulary is translated into the elegiac code in the relationship between lover and beloved in Tibullus, Propertius, and Horace.

33. Jacobson (1974, 156) contrasts the Ovidian Phaedra’s renunciation of her maternal role to the Euripidean in Hipp. II, who maintains her loyalty to her husband and children to the end. Gordon (1997, 283) associates the Ovidian Phaedra’s attitude toward her children with her masculine representation in the poem.

34. These lines in effect answer the question Hippolytus poses to the nurse in Hipp. II, 651–55: “Even so, O base human, you at least have come to make an agreement with us for my father’s untouchable bed. I will cleanse with streams of water, pouring them into my ears. In fact, how could I be a base man, I who think I am impure because I have heard such things” (ὡϲ καὶ ϲύ γ᾽ ἡμῖν πατρόϲ, ὦ κακὸν κάρα,/ λέκτρων ἀθίκτων ἦλθεϲ ἐϲ ϲυναλλαγάϲ·/ ἁγὼ ῥυτοῖϲ ναϲμοῖϲιν ἐξομόρξομαι/ ἐϲ ὦτα κλύζων. πῶϲ ἂν οὖν εἴην κακόϲ,/ ὃϲ οὐδ᾽ ἀκούϲαϲ τοιάδ᾽ ἁγνεύειν δοκῶ). Barrett (1964, 18) and Webster (1967, 67), based on fragments from Hipp. I, argue that in this play Phaedra attempts to persuade Hippolytus to seize rule from his father (TrGF 5.1.432–34). This is perhaps Ovid’s source for her argument in this passage.

35. In her self-presentation, Phaedra makes some attempt to justify her desire and the renegotiation of kinship roles it requires by calling the institution which governs them, pietas (131), out of date (rustica, 132), and by appealing to divine examples and their absolute power over the mos maiorum (133–36).

36. Cf., e.g., Prop. 1.13.29. Addressing Gallus, Propertius calls his own girlfriend most like Leda who was worthy of Jupiter (Ioue dignae Ledae). See also Am. 1.3.20; Tib. 2.6.43; Sulpicia [Tib.] 3.1.8, 3.13.10.

37. Jacobson 1974, 150. The use of nequitia in an elegiac context to mean “depravity,” or “wantonness” [OLD, nequitia, 1,3], can be traced back to Gallus. Hollis 2007, fr. 145. At the start of Am. 2, Ovid figures his elegiac poetry as tales of his nequitia: “I have written this, too … I, that Naso, poet of my own wantonness” (Hoc quoque composui … / ille ego nequitiae Naso poeta meae, Am. 2.1.1–2). When Phaedra claims she does not commit adultery because of nequitia, she is, perhaps, implying that it was not Ovid’s poetry which gave her the idea. For uses of the word in elegy, see also, e.g., Prop. 1.6.26, 1.15.38, 2.5.1–2, 2.24.6, 3.10.24, 3.19.10; Ovid Am. 1.13.32, 3.4.10, 3.1.17, 3.11.37, 3.14.17.

38. In Hipp. II, by contrast, Phaedra’s attempts to conceal her abject desire are foiled by a meddlesome and rhetorically persuasive nurse. Rosati (1985, 122–28) outlines traces of the argument given by the Euripidean nurse, filtered by the Roman comic and elegiac lena, who serves a similar function, in Phaedra’s own arguments.

39. Cf. Prop. 1.1.1–8; Ovid, Am. 1.1.1–4, 21–30, especially Cupid’s direct address to the poet: “quod” que “canas, uates, accipe” dixit “opus!”, 24; to Her. 4.13–14. Cf. Am. 2.1.3: hoc quoque iussit Amor, to Her. 4.11. See Jacobson, 149.

40. See Luck 1969, 61, 121–22, for this theme in Roman erotic elegy. The poet-praeceptor of Ov. Ars am. is one example of a male elegiac subject who claims dominance. He offers as analogy, Achilles’s submission to his own praeceptor, Chiron: uerberibus iussas praebuit ille manus (Ars am. 1.16), an analogy which echoes Phaedra’s reported promise of Cupid that Hippolytus will offer his conquered hands to her: dabit victas ferreus ille manus (Her. 4.14). The praeceptor boasts, however, that his captive will be Cupid himself (Ars am. 1.21–24), from which he will exact revenge for the submissive elegiac position the praeceptor held in the past. Here the exception proves the rule that the submissive position, although feminized, belongs programmatically to the male elegiac amator, a gendered position his exemplary masculine analog demonstrates.

41. See Davis 1995, 44, on Phaedra taking the masculine role of amator.

42. Spentzou (2003, 43–84) has noted a tension in the Heroides between a longing for innocence and the ars associated with love. For Phaedra’s letter as deceptive, see Farrell 1998 and Lindheim 2003, 27–28 and passim. Such an interpretation is in line with the general precepts of cultus as a means of deceptive seduction (Ars am. 1.611; 3.101–28, 155, 210); on this topic, see Kauffman 1986, 52 and Conte 1994, 54a.

43. See Palmer 1898, 310 ad. Her. 4.86. for a note on militia applied to hunting; see also OLD, 1d., for figurative use with other “occupations or services.” Some MSS read materia for militia, strengthening the metapoetic resonance. We see an example of militia amoris in Am. 1.9.1: “Every lover is a soldier, and Cupid has his own camp” (Militat omnis amans, et habet sua castra Cupido). See Lyne 1980, 71–78, for this theme in Roman erotic elegy.

44. See Palmer 1898, 310 ad loc. and OLD, numerus, 12 b., “the successive movements performed in an exercise,” which cites this line in its examples. Ovid is fond of puns on the elegiac meter, the alternus versus; cf., e.g., Am. 1.1.3–4: “the following verse was equal; Cupid is said to have laughed and stolen one foot” (par erat inferior uersus; risisse Cupido / dicitur atque unum surripuisse pedem); 17–18; cf. Am. 3.1.7–10. See Sharrock 1990 for a discussion of this use of alternus in Roman erotic elegy.

45. The argument for alternating love with hunting also echoes the argument of Hippolytus’s Therapon in Hipp. II (88–120) and in Hipp. I (TrGF 5.1.428), addressed to Hippolytus by an unidentified speaker.

46. Phaedra does refer to Aphrodite’s curse of the Cretan women (53–66), but only in order to justify her desire as an unavoidable inherited trait. See below for further discussion.

47. Butler 1993, 12–14 and passim.

48. See Gill 1990; Goff 1990; and McClure 1999, 112–57. Casali (1995, 11) points out the irony of the Ovidian Phaedra’s “optimistic unconcern about being seen together with Hippolytus,” for she herself has identified the source of Venus’s anger against her family as the Sun, who saw Venus and Mars together.

49. The tone of this couplet recalls Ars am. 3.113, where the poet-praeceptor announces to his female students: “previously there was an uncultivated simplicity; but now Rome is golden” (simplicitas rudis ante fuit; nunc aurea Roma est). This is perhaps another clue that the two projects were simultaneous and/or we are meant to read Phaedra’s epistle as if she has read the Ars. This last suggestion does not exclude the possibility that Ovid has written books of the Ars as a prequel, so to speak, of his earlier amatory poetry.

50. In addition to Phaedra’s depiction as a stepmother in love with her stepson in Euripides and Sophocles, the fear of the “amorous stepmother” also appears in Latin literature and declamation. In Rome, a man would likely remarry a woman of childbearing age. In some cases the wife may be the same age or younger than her husband’s adult children. Such circumstances have been cited as possible causes for this fear. Gray-Fow 1988, 748–49; Treggiari 1991, 401; and Watson 1995, 136–39.

51. See Gray-Fow 1988; Treggiari 1991, 391–92; and Watson 1995, 92–206, for the tradition of the saeva noverca in Latin literature and its basis in Roman life and law. See Watson 1995, 109–13, for the portrayal of Phaedra as a saeva noverca in Latin literature.

52. The authenticity of the female voice in Heroides is much contested. Many follow Harvey’s (1992) conclusions regarding transvestite ventriloquism, described above, nn. 12 and 13. See, e.g., Gordon 1997 and Lindheim 2003. Fulkerson (2005, 5) argues that the gendered voice is entirely a construction, and not affected by biology.

53. As I argue elsewhere, Sulpicia’s elegiac speaking voice is qualitatively different; Westerhold 2018. See also Fulkerson 2017, 46–53, for evidence of women’s literacy and a discussion of “feminine” Latin.

54. Casali (1995, 1–3) notes that the intrusion of a letter into a preexisting story, in this instance, complicates the central problem in the story, i.e., the revelation of Phaedra’s desire. In addition, he notes the act of writing “is already a forewarning of Phaedra’s (tragic) letter to Theseus” (1).

55. “One could not ask for a better example of Jost’s category of kinetic letters, as Phaedra’s letter instigates violent action and reaction, links suicide with homicide, and requires divine intervention to ‘rewrite’ its contents” (88). See Rosenmeyer 2001, 65–66 and 65 nn. 12 and 13, for an explanation of the difference between “communicative” and “kinetic” letters as formulated by Jost.

56. As Rosenmeyer (2001, 90) points out, Phaedra’s letter “speaks” to Theseus: “it shouts, it shouts, the inconsolable tablet” (βοᾶι βοᾶι δέλτοϲ ἄλαϲτα, Eur., Hipp. II, 877). See further Rosenmeyer 2001, 61–97.

57. In the House of Jason, for example, Phaedra’s nurse carries wax tablets. See chapter 1 and Valladares 2021, 158–73, for discussion of this fresco. See further Jacobson 1974, 146, n.11; Casali 1995, 3, 13 n.12; Bergmann 1996; and Jolivet 2001, 249–50, n. 78, for further references. See also Leo 1878, 178–79, who argues for a seduction letter in Hipp. I.

58. As noted earlier, Webster (1967) argues that in Hipp. I Hippolytus veiled himself and probably left the stage in reaction to Phaedra’s approach (65, 67); see Barchiesi 1993, 337, who connects Phaedra’s invitation to read with Theseus’s unsuspecting opening of her letter in Hipp. II.

59. Watson (1995, 137) draws this distinction between the Greek and Roman legal status of stepmother. In Greece the union would only be considered adultery. See Watson 1995, 137 n. 11, for further bibliography on incest between stepparent and stepchild. At the end of the first century CE, after incest prohibitions had been relaxed, to the point that even a paternal uncle was allowed to marry his niece, a stepmother was one of four affines still prohibited to a Roman man as a marriage partner (Corbier 1991, 177).

60. Bassi (1998, 42–98) argues that all mediated speech in Greek literature and drama, such as a letter, is marked as deceptive in opposition to unmediated dialogue.

61. See Goff 1990, 1–26 and passim, and Zeitlin 1996, 219–84, on this theme in Euripides’s Hipp. II.

62. The Euripidean Phaedra manages to suppress Hippolytus’s name entirely. It is the nurse who names him, in connection with the issue of inheritance should Phaedra die (310), and after Phaedra has revealed her desire (352). His name, spoken aloud, is the trigger for Phaedra’s speech and revelation of her desire (310–53).

63. ὅστιϲ ποθ᾽ οὗτόϲ ἐσθ᾽, ὁ τῆϲ Ἀμαζόνοϲ, Eur., Hipp. II, 351. When the nurse names Hippolytus a second time, identifying him as Phaedra’s beloved (“Do you say Hippolytus?” Ἱππόλυτον αὐδᾷϲ; 352), Phaedra refuses any responsibility for the revelation, saying “you are hearing your own words, not mine” (ϲοῦ τάδ᾽, οὐκ ἐμοῦ κλύειϲ, 352). See Casali 1995, 2–3.

64. ὦ τάλαινα παῖ Κρηϲία, Eur., Hipp. II, 372. Casali (1995, 2–3) adds to the allusion the stereotype of the Cretan liar, pointing to Phaedra’s lying letter in Hipp. II, the current letter, and her elegiac pose. Fulkerson (2005, 129), in contrast, reads Phaedra’s self-identification as an allusion to and an identification with her sister, Ariadne.

65. Cf., e.g., Eur., Hipp. II, 331: “out of shame we contrive good” (ἐκ τῶν γὰρ αἰσχρῶν ἐσθλὰ μηχανώμεθα).

66. See Cairns 1993, 10–11; Kaster 1997, 3; and Scheff 1997, 209, for the similarities between αἰδώς and pudor.

67. See Kaster 1997 on the threat of judgment connoted by the word pudor.

68. Kaster 1997, 12–14. Kaster does note, however, that “pudor can also denote … a desire to avoid behavior that causes [the feeling of shame]” (4).

69. See Newlands 1995, 146–74, esp. 168, for the association of female speech, sexuality, and shame in the tales of Lucretia (Fasti 2) and Myrrha (Met. 10).

70. Aphrodite announces that she has inspired Phaedra with a love for Hippolytus and intends to reveal her desire to Theseus as a means to punish Hippolytus (Eur., Hipp. II, 21–50).

71. “I have a teacher of daring and courage … Eros, the most unconquerable god of all” (ἔχω δὲ τόλμηϲ καὶ θράϲουϲ διδάϲκαλον … Ἔρωτα, πάντων δυϲμαχώτατον θεόν (Hipp. I, TrGF 5.1.430; Webster 1967, 67). In Sophocles’s Phaedra, unattributed dialogue expresses the power of Eros (TrGF 4.684; Barrett 1964, 23). In Hipp. II, the power of Aphrodite and Eros is described, e.g., by the nurse (433–81) and the chorus (525–64, 1268–82).

72. Jacobson (1974, 148) reads the repetition of ter as a parody of an epic motif.

73. Compare the precepts of Ovid’s poet-praeceptor in his Ars am. (1.505–24). In fact, his exemplars include Phaedra and Hippolytus: “Phaedra loved Hippolytus, and he was not very dressed up” (Hippolytum Phaedra, nec erat bene cultus, amauit, 1.511; in addition to Ariadne and Theseus: “Theseus carried off Minos’s daughter, although he had not adorned his temples with a hair pin,” Minoida Theseus / abstulit, a nulla tempora comptus acu, 1.509–10). Jacobson (1974, 150) reads Phaedra as a poet-praeceptor in her own right. See further Fulkerson 2005, 1–22.

74. See Armstrong 2006, 1–28, 109–66, on Cretan women and sexuality, and 12–16 for a summary of Roman literary receptions of the myths of Cretan women.

75. The chorus of Hipp. II also provides examples of victims of Aphrodite and Eros, Iole and Semele (545–64). It is noteworthy that Iole is likened to a bacchant (ὅπωϲ τε βάκ-/χαν, 550–51), which is echoed in the antistrophe in the same metrical position by the genitive of Bacchus (τοκάδα τὰν διγόνοιο Βάκ-/χου, 560–61) in a phrase which identifies Semele. The echo reinforces the similarity of the two women while drawing attention to the similarity of women under the influence of Aphrodite and Bacchus.

76. Phaedra offers a second Alexandrian catalogue of actively desiring women and their beloved hunters—Cephalus and Aurora, Cinyras and Venus, and Meleager and Atalanta (93–104). The list provides examples which seem to be straightforward analogs to Phaedra and Hippolytus. Again, as Jacobson (1974, 153) notes, the list alludes to Hipp. II by repeating examples used by the nurse (Cephalus and Eos, Zeus and Semele, 454–56) in order to persuade Phaedra not to commit suicide. Her list appears to be an attempt to construct a more inclusive list of the desires that count. Phaedra’s examples are presented as egalitarian love affairs where lover and beloved vacillate from female to male, but the denouement of each mythological love contains the seeds of tragedies. See further Jacobson 1974, 153–54; Casali 1995, 6–7; and Davis 1995, 53–55.

77. Jacobson (1974) argues that the theme of deceit functions “as if to underscore the presence of deceit in her own attempt to win Hippolytus” (156).

78. Hipp. II, 1198–248. Hippolytus actually dies in the company of Theseus, but his accident is ultimately the cause of his death. Seneca later uses the same verb to describe Hippolytus’s dismemberment in his Phaedra, when his horses tear his body from the tree trunk where he is impaled: “and they end the delay and tear apart their master at the same time” (et pariter moram / dominumque rumpunt, 1101–2).

79. McKeown (1998) comments, in connection with Am. 2.18, on the scholarly consensus that 2.18.19–20 refers to the Ars, indicating the work was recently begun at the time of the second edition of Am., while Her. would have just been finished, and he suggests that the two projects may well have overlapped. McKeown 1989, vol 1, 74–89, vol. 3, 384–86: “The present tense of profitemur (19) and scribimus (22) seems to suggest that Ovid was actively engaged on more than one of his amatory works when 2.18 was written, perhaps not only the Amores and Heroides, but also, if line 19 refers to the Ars … that poem as well” (1, 87).

80. See in the introduction “Generic Performances” for a more detailed discussion of Am. 3.1.

81. In Medea’s letter to Jason, Medea announces that her mind is conceiving of something greater (Nescio quid certe mens mea maius agit, Her. 12.212), which Hinds (1993, 39–40) argues is a reference to her tragic tradition.

82. See Ferrari 2002, 54–56, 72–86, for a discussion of αἰδώς in Greek art and literature along with further references. See Kaster 1997, 3 n. 5, for the metaphor of pudor as a garment with references and comparanda. Ovid’s text echoes Vergil’s Dido, who also acknowledges both her “crime” and the need to “cover” it. There, too, pudor motivates this performance, for she has sworn an oath to pudor (4.27), and, although she later disregards her reputation and openly expresses her desire, she does so under the name of “wife,” again a performance: “and, in fact, Dido is not moved by appearance or reputation nor is she now contemplating a secret affair: she calls it a marriage, she hides the crime with this name” (neque enim specie famaue mouetur / nec iam furtiuum Dido meditatur amorem: / coniugium uocat, hoc praetexit nomine culpam, Aen. 4.170–73). See further Gildenhard and Zissos 2013, 96 and 125 n. 35. The female poet Sulpicia, by contrast, explicitly denies pudor power over her behavior, a poetic move which expresses both a sexual desire and her renunciation of normative feminine behavior. See Keith 1997 and Merriam 2005 on Dido as a possible literary paradigm for Sulpicia. See Fulkerson 2017 for a very recent introduction and commentary on Sulpicia’s poems. Fabre-Serris (2009) argues that Her. 4 (as well as 15) alludes directly to the Sulpician corpus. See further Westerhold 2018 on Sulpicia and Ovid with further bibliography.

83. Casali (1995) argues that the elegiac genre cannot change the tragic story: “When the expressive code is changed the story does not change: Phaedra will die, Hippolytus will die.… Behind the distorting filter of elegiac language the reader recognizes the prefiguration of the ‘tragic’ end of the story” (5). See also Verducci 1985, 19–20, who notes that allusions in the epistles work against the heroines: “The fictive speaker thus becomes, through her own words, the involuntary and unconscious victim of the poet’s authorial and often allusively literary parody” (19–20).

84. See chapter 1 for a discussion of these lists.

85. φεῦ / εἴθ᾿ ἦν ἐμαυτὸν προϲβλέπειν ἐναντίον / ϲτάνθ᾽, ὡϲ ἐδάκρυϲ᾽ οἷα πάϲχομεν κακά.

86. Fulkerson (2005, 146) argues that “the Heroides provoke us to ask to what extent a heroine has the textual authority to change her story but, more importantly, to what extent our own literary experience will permit her to effect this change” (146).

87. Contra Verducci 1985, 31–32 and passim, who, while also connecting the issue of allusive tension to the question of poetic authority, argues that the epistolary writers seem “real” because they successfully defy expectations set up by their source texts.

88. Quoting Erasmus, Harvey continues her argument that, although women are associated with disguise and deception, they remain “undisguisable (‘a woman is still a woman … no matter what role she may try to play’)” (62).

89. See Rabinowitz 1991 for the female body as the object of the male pornographic gaze.

90. This is the temporal aspect of Curley’s (2013, 95–133) argument that Ovid’s reinterpretation of Greek tragedy in the Metamorphoses effects changes to the story allowed by the generic transition from drama, which limits spatial and temporal aspects, “here-and-now,” to epic, which affords seemingly unlimited narrative possibilities in the “there-and-then.”

91. See Ahl 1985, 211–12; Janan 1991, 240; Raval 2001, 295–96 and passim; and Rosenmeyer 2001, 19. In addition to the pun on her name, Ahl draws attention to Byblis’s connection to water, in which papyrus grows. She is turned into a stream which bears her name. Her mother, Cyane, is a water nymph and her grandfather, the river Maeander, a river, Janan (1991, 243) notes, which acts as an “artistic paradigm” in the Met. See also Boyd 2006 on the Maeander river as a model and map, so to speak, for the narrative structure of the Met. Ahl (1985, 211–13) also points out Byblis’s connection to Io, another “writer,” who was reported by Apollodorus and Plutarch to have wandered to the town Byblos when searching for her lost son; to Daedalus through their shared reliance on wax; and to the Sibyl, acoustically. See Ahl 1985, 144–50, for his discussion of the Io episode.

92. Lindheim (2003, 22–23) notes that the letter writer (of the Heroides) is free of “the impediment of an ordering, external narrating voice that might accord to her tale secondary status, curbing its length and de-accentuating its importance” (23).

93. Kristeva (1984, 66–67) emphasizes the need for a semiotic “code” to be shared in order to function as a means of communication: “[A human being’s] semiotic ‘code’ is cut off from any possible identification unless it is assumed by the other (first the mother, then the symbolic and/or the social group)” (66). Looking to social anthropology, she further explores the shared dependence of the social and the symbolic on the thetic (72–85). Responding to Lévi-Strauss (1969, 249, n. 90), Kristeva remarks, “All things stemming from social symbolism, hence kinship structures and myth itself, are symbolic devices, made possible by the thetic, which has taken on social symbolism as such” (74, emphasis added).

94. Anderson (1972, 451–52 ad 468–71) notes that “Ovid’s dramatic sensitivity here anticipates Freud.” He likens Byblis’s erotic dream to the “subconscious” (451 ad 464–65). Jocasta, another tragic heroine, but one whose incest offered a model for Freud, also speaks of dreams of incest in Sophocles’s play Oedipus Tyrannos: “do not fear marriage to your mother: for many mortals already have lain with their mother in dreams” (cὺ δ᾿ ἐc τὰ μητρὸc μὴ φοβοῦ νυμφεύματα· / πολλοὶ γὰρ ἤδη κἀν ὀνείραcιν βροτῶν / μητρὶ ξυνηυνάcθηcαν, 980–82). In her account, however, the dreams are merely wish fulfillment. See Bassi 1998, 31–33, on Freud and this play. See Scioli 2015 for dreams in Latin elegy and 212–14 for a discussion of Byblis’s dream. For a summary of Freudian and Jungian dream theories, see, e.g., Freud 1915; Jung 1969; Segal 1984; and Csapo 2005, 86–110.

95. See in the introduction “Ovid on the Intersection” for a detailed discussion of the correspondences between these texts and further references. See Janan 1991, 246, n. 18; Raval 2001, 289, 300; and Curley 2013, 84–94, 183–84, for correspondences with Phaedra’s epistle (Her. 4); and Raval 2001 for correspondences with Her. in general. See Nagle 1983; Ahl 1985, 214; Rosati 1985, 121–22; Verducci 1985, 190–97; Larmour 1990; and Tissol 1997, 36–52, for a comparison of Phaedra’s epistle (Her. 4) and the Byblis and Myrrha episodes in Ovid’s Met.

96. Contra Anderson 1972, 451–52 ad 468–71, who argues that her suppression of spes obscenas (468) while awake “implies that Byblis has become aware of her desire.”

97. OLD, obsc(a)enus, 2–4; “Obscenum is said from scena; this word Accius writes, as the Greeks [do], scena … for that reason a shameful thing is obscaenum because it should not be spoken publicly except on the stage (scaena)” (obscenum dictum ab scena; eam, ut Graeci, aut Accius scribit scenam … quare turpe ideo obscaenum quod nisi in scaena palam dici non debet, Varro, Ling. 7.96); Miller (2004, 34), citing Barton (1993, 142 n.173) citing Varro (above), uses the English word obscene to describe Catull. 64.62–67, where the description of Ariadne violates her subjectivity in a way that is pornographic, obscene: “The alienation of either the flesh or the experience from the self, its specularization ob scaenum (for the stage) is the essence of the obscene.” Ernout and Meillet (1967 ad loc.) report that the etymology of obscenus, -a, -um is unknown, but also conjecture a connection between obscenus and sc(a)ena from the Greek word σκηνή (“stage”) through an Etruscan intermediary.

98. OLD, visum, -i, n [pple. of VIDEO], a. “a vision (usu. as presented in a dream).”

99. “We see the condition of the subject of signifiance as a heterogeneous contradiction between two irreconcilable elements—separate but inseparable from the process in which they assume asymmetrical functions” (Kristeva, 1984, 82). Kristeva’s (1984, 17) term, “signifiance,” refers to the process of signification associated with literary practice and the text, regulated by social, cultural, and legal systems but transforming these systems if it is introduced into their discourse. See Kristeva 1984, 81–82, 202–7, for a full description of this relationship in the process of signification.

100. See also Ahl 1985, 213–14, in his discussion of this episode. He identifies the “assault upon fixed forms of words” as a larger theme of social constraint and self-expression in the Met.

101. The opening of Myrrha’s tale makes this point even more explicit: choose one man from your many suitors, Myrrha, but don’t let it be that one, i.e., the one excluded by the incest taboo, your father (“From everywhere the finest princes desire you; from the entire east young men are here to compete for your hand; out of them all choose one for your husband, Myrrha, only let one not be among them all,” undique lecti / te cupiunt proceres, totoque oriente iuuentus / ad thalami certamen adest: ex omnibus unum / elige, Myrrha, uirum—dum ne sit in omnibus unus, Met. 10.315–18).

102. Kristeva (1984) argues that poetic mimesis “dissolves” the process of creating meaning, because its mimetic denotation points to a solely discursive construction (there is not a real object as a referent), and its mimetic enunciator is likewise solely discursive. “Mimesis, in our view, is a transgression of the thetic when truth is no longer a reference to an object that is identifiable outside of language,” and for this reason, “mimesis and the poetic language … prevent the imposition of the thetic from hiding the semiotic process that produces it” (58). This puts the subject of enunciation “in process/on trial.” See 57–59 for further discussion of mimesis and meaning. See esp. 16–17, 62–71, and 99–106 for an explanation of the “text” and its signifying practices.

103. Her geographical location doubles as a genealogy, since the town is named after her father: “and in the Asian land you constructed walls holding the name of the builder” (et in Aside terra / moenia constituis positoris habentia nomen, 448–49).

104. Perhaps another aspect of Byblis’s corporeal representation is her letter. Tissol (1997, 44–46) argues that Byblis’s letter foregrounds the physicality of the written word, especially when she erases soror and when her writing spills into the margins of the page.

105. Barchiesi 1993.

106. Kristeva (1986, 192 and passim) connects the diachronic characteristic of syntax in language to the idea of linear time. As a consequence, the temporality of grammar informs the “thesis” of the subject in relation to the predicate in a sentence (1984, 54–55), and the positing of a subject in culture (1984, 66–67, 72) includes a temporal delay until her symbolic role is recognized by other members of the community. When one considers, as Butler does, the positing of the subject as a performativity requiring repetition (1993, 12–14 and passim; 2000; 2008), the symbolic is limited by time (for each citation is only temporary, never permanent), but simultaneously relies on time (as the continual citation of the symbolic role or law over time creates the illusion of permanence).

107. Lindheim (2003, 13–77 and passim) argues that the epistolary form of the Heroides gives the author “the authority to arrange the narrative as she sees fit, in agreement with her own perspective on events” (22–23).

108. See Jacobson 1974, 147, n. 13, who likens Byblis’s deletion to Phaedra’s avoidance of kinship names in Her. 4. See above for a discussion of Phaedra’s strategies of renaming. Farrell (1998, 320–21) argues that the deletion of soror is the act which allows Byblis’s passion to be expressed sincerely. Kennedy’s (2002, 224–26) idea of “temporality” in the Heroides suggests the heroines have the power to reimagine their myths in the epistles: “The heroine’s stories, when we come to read their letters, are, in this sense, already written.… The epistolary form freezes them at a moment within the story, foreseeing or desiring a particular ‘end’ to their stories, which may or may not approximate to the ‘end,’ the outcome or consequences, with which the external reader is familiar” (225). Contra Lindheim (2003), who argues that the heroines do not rewrite their stories: “Rather than struggling against the tellings that the prior texts have provided in an effort to rewrite their stories, the heroines choose to inhabit their traditional, recognizable selves and stories” (34).

109. In all of the earlier versions of the Byblis story she is sister to Caunus.

110. Byblis’s epistolary greeting may also signal another Euripidean tragic influence, Aeolus (TGF², 14–41), for her greeting may recall that of Canace in Heroides 11 (“The daughter of Aeolis sends to the son of Aeolis a wish for good health, which she herself does not have,” Aeolis Aeolidae quam non habet ipsa salutem / mittit, Her. 11.1–2). Palmer (1898) does not include this line, which the MS S does. See Palmer 1898, xli–xlii, for a very brief discussion of these lines in the manuscript tradition. Canace and Macareus, children of Aeolus, fell in love and conceived a child. See further Curley 2013, 70 and n. 30. Raval (2001, 289–90) also points to Met. 9.506–8 as an allusion to Her. 11 (289–90), which characterizes Byblis as a reader of the epistles. See Paratore 1970; Verducci 1985, 191–97; and Raval 2001, 298–304, for further discussion of Byblis’s reading of Heroides and correspondences with Heroides 4 and 11 in particular. See also Fulkerson 2005, 1–22 and passim, who argues that the heroines of the Heroides demonstrate their familiarity with the literary tradition, including the other letters in the collection, in their own epistolary poetry. Consider also the earlier allusion, “Her right hand holds a pen, an empty wax tablet the other holds” (dextra tenet ferrum, uacuam tenet altera ceram, Met. 9.522), to “My right hand holds the pen, a drawn sword the other holds” (Dextra tenet calamum, strictum tenet altera ferrum, Her. 11.3), noted by Anderson (1972, 454–55 ad 520–22) and Raval (2001, 299–300).

111. And perhaps Ovid’s readers are reminded of Minos’s sexuality. According to a variant reported by Apollodorus (3.1.2) and Antoninus Liberalis (30), Miletus was fleeing the sexual advances of Minos. See Anderson 1972, 447–48 ad. 418–49 for a summary of this tradition.

112. Kaster 1997, 9. See above nn. 66–68 for a discussion of pudor and its association with feminine pudicitia.

113. So, too, the poet-praeceptor of the Ars uses the same verb when he claims women cover their desire better (tectius, 1.276).

114. See Raval 2001 for elegiac elements in Byblis’s letter, and 295–96 for Byblis’s status as writer and subject of her text, with references to woman as tablet in Greek literature and woman as text (scripta puella) in Roman erotic elegy. See Gold 1993; Keith 1994; and Wyke 1987, 1989, and 2002 for a more general discussion of the scripta puella.

115. See, e.g., McCarthy 1998 on the importance of wax tablets to the elegiac poet.

116. See in chapter 1 “Ars Amatoria 1.283–340” and n. 132 below for Pasiphae in Ars am. See Keith 2008b for the dress of the elegiac puella.

117. Ars am. 3.127–280; See also Rem. Am. 341–44, and Medic. For a recent discussion of the connection of cultus to femininity, see Shumka 2008, 173–78.

118. “Ah unhappy maiden, you eat bitter herbs” (a uirgo infelix, herbis pasceris amaris; Calvus, Courtney, fr. 9).

119. “Let us live, my Lesbia, and let us love, and let us assess all the rumors of the rather serious old men as worth one penny” (Viuamus mea Lesbia, atque amemus, / rumoresque senum seueriorum / omnes unius aestimemus assis, 5.1–3).

120. Another Catullan allusion may be perceived in Byblis’s comment that Caunus’s mother is neither a tiger nor made of rock or iron, nor did he nurse at the teat of a lion. Cf. Met. 9.613–15 with Catull. 64.154–57: “What lion bore you under a lonely rock, what sea conceived you and spat you out from its foaming waves, what Syrtis, what greedy Scylla, what vast Charybdis, you who give rewards such as these in return for your sweet life?” (quaenam te genuit sola sub rupe leaena, / quod mare conceptum spumantibus exspuit undis, quae Syrtis, quae Scylla rapax, quae uasta Carybdis, / talia qui reddis pro dulci praemia uita?); and Raval 2001, 304–5.

121. Raval 2001, 294. See Luck 1969, 37, 61, 121–22; Lyne 1980; Conte 1994b, 322–23; Greene 1998; McCarthy 1998; Fitzgerald 2000, 72–77; and Raval 2001, 294, n. 25, for the traditional elegiac theme of servitium amoris, the marginal position assumed by the elegiac subject in relation to his female beloved (reversing gender hierarchies) and in relation to the free-born Roman male citizen (as slave). Cf. Prop. 1.5.18–19: “Poor man, you will not be able to know who or where you are! At that time you will be forced to know a painful slavery to our girlfriend” (nec poteris, qui sis aut ubi, nosse miser! / tum graue seruitium nostrae cogere puellae / discere). For Byblis to take this position as slave to love is to assume the (emasculated but still) masculine guise of the elegiac amator. As Gold (1993, 89–90) has argued, however, the puella’s role is entirely relational. She therefore oscillates between dominator and dominated in opposition to the amator. See also Greene 1998 and Wyke 2002, 159–62, 172–73, on the puella’s inconsistent depiction as dominator/dominated.

122. In contrast, Phaedra correctly self-identifies with the puella-domina (Her. 4.14) in her attempt to construct herself as a feminine elegiac subject; however, her gendered position is blurred by her simultaneous identification with the amator who is compelled to write by love (Her. 4.13–18) and is being punished by Venus (Her. 4.53–54).

123. Byblis follows the example of the majority of the female writers of the Heroides, who draw attention to their own bodies. “Emblematic of their sexual passivity is their habit of dwelling on their own physical appearance while ignoring the man’s” (Gordon 1997, 280). Phaedra and Sappho are exceptional in this respect. In her epistle Phaedra describes her own acquisitive, objectifying gaze on Hippolytus’s exercising body (Her. 4.70–84). Ovid’s representation of her gaze of Hippolytus’s body mimics the male gaze in its eroticization of his movements; however, her scopic pleasure is not the result of his corporeal suffering but his demonstration of healthy strength. See Gordon 1997, 280–83, for a comparison of the feminine and masculine gaze of the Heroides. See Mulvey [1989] 2009, 19–22 and passim, for her famous formulation of the gaze in film theory.

124. Dura modifies, perhaps, arma (543), or perhaps stands as a substantive: “I bore adversities.” Metrically, however, the short a makes ego an alternative modified noun, adding another feminine elegiac persona (the puella dura) to the costume of puella Byblis wears for her letter.

125. These lines recall simultaneously the defenseless victim of divine desire Callisto (“She indeed fights back as much as a woman can [if only you were watching, Saturnian: you would have been gentler], she indeed fights back; but what man could a girl overcome, or who could conquer Jupiter?” illa quidem contra, quantum modo femina posset / (adspiceres utinam, Saturnia: mitior esses), / illa quidem pugnat; sed quem superare puella, /quisue Iouem poterat? Met., 2.434–36) and the superfeminine courage of Polyxena (“brave and unhappy and a maiden who is more than woman,” fortis et infelix et plus quam femina uirgo, Met., 13.451). While both characters are gender benders to some degree (Callisto refuses suitors and chooses the company of Diana; Polyxena offers her chest to the sword, a masculine death), both are maidens whose stories, like Byblis’s, are poised on the threshold between virginity and marriage.

126. Byblis’s story demonstrates that a writing subject is mute (muta, 655) and unable to be effective without a reader. We learn later that Byblis, after announcing her desire to everyone (“she admits her hope for forbidden desire,” inconcessaeque fatetur / spem Veneris, 638–39) and pursuing her brother through Caria (“she follows her fugitive brother’s footsteps,” profugi sequitur uestigia fratris, 640), finally collapses in silence (Bybli, iaces, 651; muta iacet, 655). See further Natoli 2017, 33–79, on speech loss in the Metamorphoses accompanying a character’s exclusion from community.

127. See Tissol 1997, 46–47, and Raval 2001, 304, for further discussion of this passage. “Speech requires the speaker’s presence, from which flow all the rhetorical advantages of gesture, of facial and bodily expression, and most importantly, of adjustment to the perceived response of the audience” (Tissol 1997, 47; emphasis added). See also Rosenmeyer 1996, 20 and passim, for the unpredictability of epistolary reception.

128. See also Jenkins 2000 for a discussion of this anxiety and its connection with anxieties over literary reception. Byblis’s desire remains in the realm of fantasy until it is received and becomes a speech act (439).

129. Cf. Am. 1.12.3–4. The amator has entrusted his tablets containing a letter to his beloved to the servant Nape. When the reply is negative, the amator blames an omen—Nape tripping on her way out the door: “There is something in omens; just now when Nape wished to leave, she hit the threshold with her toes and stopped” (omina sunt aliquid: modo cum discedere uellet, / ad limen digitos restitit icta Nape).

130. Am. 1.11 and 1.12, in their narration of the amator’s sending and receiving of love notes, share some of the same anxieties as Byblis’s tale. Both poems stress the power of the messenger and the reader in the success of the amator’s communication. Byblis’s messenger finds a fitting time to deliver the tablets (“The servant, finding a fitting time, went,” apta minister / tempora nactus adit, Met. 9.572–73), but when Caunus’s reaction is negative, Byblis suspects that the messenger did not find a fitting time, choosing one in which Caunus was distracted by other matters (“And perhaps the servant whom I sent made some mistake; he did not approach him suitably; and he did not choose a fitting time, I suppose, and he did not seek both an unoccupied hour and mind,” forsitan et missi sit quaedam culpa ministri; / non adiit apte, nec legit idonea, credo, / tempora, nec petiit horamque animumque uacantem, 9.610–12). Likewise, the amator enjoins Nape to find a fitting time when his beloved’s mind is free (“Give her the tablets while she is happily free,” uacuae bene redde tabellas, 1.11.15). The stress on the reader’s leisure indicates a desire for intellectual receptivity to the message.

131. I owe this insight to an anonymous reviewer for Cornell University Press. Cf., e.g., Catull. 68.139; Ov. Met. 2.37; Verg. Aen. 4.170–73. See n. 82 above on Dido’s covering of her culpa.

132. Byblis’s cultus recalls that of Pasiphae (Ars am. 1.303–7), whose fine clothes and mirror, the praeceptor tells her, has no effect on her bull. Byblis’s toilette is meant for the gaze of an inappropriate love interest, and, although we are not told it is to no avail, a literary reader can guess that Caunus is not charmed. Furthermore, Byblis’s jealousy over Caunus’s interest in other women (“and if any other is more beautiful to him, she envies her,” et si qua est illic formosior, inuidet illi, Met. 9.463) is reminiscent of Pasiphae’s more absurd, but equally horrifying, Junonian jealousy of her “rivals,” cows (Ars am. 1.313–22). See chapter 1, “Ars Amatoria 1.283–340,” for a discussion of this passage.

133. Zeitlin (1996) has identified the passivity effected by the peripeteia of Greek Attic tragedy, which leads to suffering and pathos, as a feminine experience which tragedy offers to the masculine Greek subject in the audience of Attic tragedy. She argues that the peripeteia results in “a shift … from active to passive, from mastery over the self and others to surrender and grief. Sometimes there is madness, always suffering and pathos” (363). See 349–52 for the body as a feminine theme in Attic Greek tragedy. See esp. 247–48 and 351 for Hippolytus’s feminine suffering which reflects Phaedra’s.

134. 9.439–665. I count Byblis’s letter as direct speech because it is quoted verbatim. Holzberg’s (2002, 134–35) count is 122. Holzberg, moreover, includes only 212 lines in the Byblis episode. I start at 9.439 because I consider the paratextual transition part of the frame and significant to the understanding of Ovid’s tale. See Keith 2002, 265, for direct speech as an indication of tragic influence in the Pentheus episode of Met. 3, which contains 167 out of 223 lines.

135. As noted in chapter 1, similes of bacchantes combine the metaphor of feminine sexuality as a wild, uncontrollable ecstasis with the performance site of tragedy through the divine figure of Dionysus.

136. See chapter 1 for a full discussion of this theme with comparanda and references.

137. See further Curley 2013, 69–94, 141–59, and passim, for Ovid’s representation of this struggle between pudor and amor, which he traces back to tragedy and locates in both Her. 4 and Met. 9.

138. Anderson (1972, 454 ad loc.) translates uiderit as “let Caunus see to his own actions,” citing other instances of this use at 10.624 and Ars am. 3.671. He argues the phrase “limit[s] the responsibilities of the speaker and place[s] the onus on the other person, the subject of uiderit.” My translation is intentionally ambiguous, allowing the multiple meanings of uideo to interact with other uses of the verb in this passage, creating a semantic network suggesting simultaneously sight, spectacle, and appearance.

139. See n. 56 above.

140. Raval 2001, 302. Raval (2001) also argues that Byblis is responding, as a reader of Ovidian elegy, to the poet-lover’s request that his beloved’s written response be effusive (2001, 301–2; Am. 1.11.19–22), but, as Raval points out in her discussion of Byblis’s reading of the Ars (297–98; Ars am. 1.455–58, 3.469–78), hers is a “misreading”: the poet-lover is asking for an effusive reply. Caunus has shown no erotic interest in Byblis, let alone sent her a love letter to which she responds. Farrell (1998, 319–20) argues that Byblis’s decision to write at all is a “misreading” of the Ars and a symptom of her gender transgression, since the poet-praeceptor enjoins his male pupils to use writing as the first tool of seduction (Ars am. 1.437–86).

141. See Janan 1991, 241–42 n. 11, for a summary of opinions on Ovid’s own expansiveness.

142. Janan 1991, 249. Like Byblis’s father, whose symbolic position defines her love for Caunus as incest, Ovid’s literary “fathers” define the appropriate material for poetry (Janan 1991, 252–53 and passim): “The poem replicates itself (and, through a complex pattern of allusion, its author Ovid’s poetic career), limiting creative options to the already known—the literary realization of incest” (242).

143. See Keith 2000, 36–64, for this process in Latin epic, and especially 50–52 in the Met.

144. Janan 1991, 253, n. 36.

145. Kristeva 1982, 4, 67–68, and passim; Butler 1993, 3.

146. In her analysis of Antigone, Judith Butler argues, as she also does for the categories of gender and sex (1993, 2008), that the ideal performance of a family role—she offers the father—is temporal and open to transformation through each of its iterations and instantiations. “Kinship is not simply a situation she [Antigone] is in but a set of practices that she also performs, relations that are reinstituted in time precisely through the practice of their repetition” (Butler 2000, 57–58). The symbolic position of a father, a sister, or a (step)mother is thus an “ideality” which is never fully realized in any individual who assumes one of these roles. Byblis’s and Phaedra’s bad performances of their kinship roles represent a renegotiation of the relations of prohibition in their families and a potential transformation; Butler 2000, 20–21.

147. The poetry of Sulpicia may, in fact, present an historic female elegiac amator, but her self-presentation differs in meaningful ways from the male amatores whom Phaedra and Byblis imitate. See further Westerhold 2018; Fulkerson 2017, 46–53, 221–94; and nn. 53 and 82 above.

148. Janan 1991, 245. Janan argues that Ovid’s text exposes the trap, the author’s complicity in its construction, and the necessity for readers to read beyond such a determined frame of reference.

149. Newlands 1995, 146–74, notes that, in Ovid, “the vocal woman is often punished for transgressing this [gender and kinship] cultural norm, and typically the punishment takes the form of the silence the woman has refused to keep” (166–67). She cites, in addition to Lara and Lucretia in the Fasti, Philomela, Byblis, and Myrrha in the Met.

150. In her study of Ovid’s Met., Enterline (2000, 1–90) notes a similar pattern of displacement in the rape narratives of the Met. She demonstrates a pattern of displacing anxieties about poetic authority onto the resisting, silent female body who is most often the object and subject of the poetry (e.g., Daphne, Syrinx, Philomela). The narrator embodies the potential for any poetic speaker to be alienated from his/her own tongue in stories about the difference between “his” voice and “her” resistance” (71). In the stories Enterline considers, “her” resisting body is changed and adopted as a tool for poetry (e.g., Syrinx becomes the pipes of Pan).

151. Farrell (1999, 127–41) argues along similar lines that throughout the Met. the fragility and temporality of the physical poetic corpus is juxtaposed against the immortal poetic voice. “At all points the liability of the text to change is tied to its material, bookish form. Opposed to this is the status of the poem as song, which being immaterial is not liable to those forces that threaten the stability of the material text” (141).

1. See Curley 2013, 68–74, on the connection between the elegiac “pathos of love” and tragic “pathos,” to which Ovid draws attention at Tr. 2.381–408.

2. Curley (2013, 180 and passim) refers to this model as a “master Medea code,” which is created through a network of “footnotes” to Ovid’s own poems as well as those of others. In addition to the catalogues and citations deployed as examples in his erotic elegy, didactic poetry, and exilic poetry, an example of which we consider in chapter 1, and the two poetic passages under consideration in this chapter, the Heroides includes a letter from Medea herself (12), the Met. features vengeful acts by Medea (7.1–424), Althaea (8.445–525), and Hecuba (13.399–575), and Ovid’s lone tragedy is Medea.

3. See the introduction “Ovid on the Intersection” for an explanation of Segal’s term “megatext.”

4. See especially the introduction “Ovid, in Theory” for a full definition of Kristeva’s and Butler’s theories of the abject.

5. Freud (1918) famously locates the origin of sacrifice and taboo in the first murder of the father. His murder is both an act of violence in retaliation for a perceived wrong (i.e., revenge) and the foundation of social taboos (i.e., laws and civil justice). See Girard 1977 and 1987 and his theory of the “scapegoat” for connections between revenge and justice. Girard theorizes that ritualized violence, committed by the community and justified by divine or civil law, absorbs the retributive cycle of vendetta culture. See further Csapo 2005, 116–20. Kristeva (1984, 75–83) proposes, as a counterpart to the thetic, sacrifice, and the ritual that accompanies the originary sacrifice, i.e., the dance enacting the violence leading up to first murder. From this dance, notes Kristeva, comes poetry (including and most significantly, tragedy). As a mimesis of what precedes the sacrifice which ultimately forecloses all other violence and makes social order possible, this ritual straddles the threshold of the thetic and the semiotic chora. “From its roots in ritual, poetry retains the expenditure of the thetic, its opening on to semiotic vehemence and its capacity for letting jouissance come through” (80). See Csapo 2005, 145–61, for dance, ritual, and sacrifice as theorized by Jane Harrison and the Cambridge Ritualists. Joplin (1984) applies Lévi-Strauss and Girard to Ovid’s Procne, along with other myths. She focuses on the violent institution and maintenance of difference through the sacrifice of female discourse.

6. See Clauss and Johnston 1997 and Heike and Simon 2010 for recent collections of essays which consider a variety of generic Medeas; and Ramus 2012, 41, “Roman Medea,” for essays exploring Medea in Latin literature. See the introduction of this book for a summary of Medea’s literary treatments.

7. Sophocles may have produced a second edition of Lemniae as Aristophanes did for Clouds. See Sutton 1984, xii, citing Stephanus of Byzantium (fr. 380).

8. See Jacobson 1974, 94–97 and 94 n. 1; Knox 1995, 170–71; and Fulkerson 2005, 41–43, for a list of possible sources and further bibliography. See further Jacobson 1974, 94–97, for a brief comparison of Her. 6 with Ap. Rhod. and other sources. There have also been many comparative readings of Her. 6 and 12 (the letters of Hypsipyle and Medea). See, e.g., Bloch 2000; and Fulkerson 2005, 40–66 and 40, n. 2 for further bibliography. See further Jacobson 1974; Verducci 1985, 33–85; Hinds 1993, 27–34; Heinze 1997, 38–39; Lindheim 2003, 114–33; Michalopoulos 2004; Curley 2013, 181–82; Vaiopoulos 2013; and Filippi 2015.

9. See, e.g., Burkert 1970 and Verducci 1985, 59–62, for the Hypsipyle myth and extant sources.

10. Fulkerson (2005) also reads the Heroides as a meditation on poetic authority and reception in the Augustan era. “For me, the Heroides provoke the metapoetic question of whether we might be willing to take Briseis’ word (Ovid’s word) over Homer’s, Ariadne’s (Ovid’s) over Catullus’. How wedded are we to the stories we know; how unwilling to view them differently?” 17–18. See chapter 2 for Phaedra’s attempts to change her own story in Her. 4.

11. Michalopoulos (2004, 113 and passim) also argues that the meaning of Hypsipyle’s letter requires a knowledge of her literary tradition and Medea’s but attributes the intertextual play to Ovid alone.

12. Lindheim (2003, 13–77 and passim) applies Lacan’s formulation of desire to interpret the epistolary strategies of the heroines in this collection, wherein some heroines reflect their beloved’s narcissistic fantasies, others, helpless women. See further 114–33 for Lindheim’s reading of Her. 6 and 12.

13. See chapter 1 on the threat abject female desire poses for normative masculine subjects in Ars am. 2.

14. Fulkerson (2005, 2 and passim) calls the writers of Heroides “excessively literate” and argues that we are to imagine them responding and alluding to each other’s epistles. As Verducci notes, the mimetic aim of epistolary fiction guides readers to attribute literary allusions to earlier work and their ironic effects to the heroines writing the letters (25–26); contra Vaiopoulos (2013, 125 and passim), who argues that Ovid puts the letters in dialogue while the heroines remain oblivious.

15. See further, e.g., Jacobson 1974, 94–108; Verducci 1985, 58–59; and Lindheim 2003, 118, for Hypsipyle’s familiarity with Euripides’s play.

16. “The barbarian of Phasis avenged her spouse’s crime and the violated oaths of marriage by means of her own children” (coniugis admissum uiolataque iura marita est/ barbara per natos Phasias ulta suos, 381–82). See chapter 1 for a discussion of this passage and the use of mythological paradigms as a language. Cf. Medea in Ov. Met.: “the mother wickedly took revenge and fled Jason’s weapons” (ultaque se male mater Iasonis effugit arma, 7.397). Medea’s exemplary status in Her. 6 is also noted by Michalopoulos (2004, 112).

17. Line 127 may even look “forward” to tragedies which recounted Medea’s dangerous role as stepmother to Theseus, when she (almost) murders the Athenian hero. Euripides’s Aegeus, Ennius’s Medea Exul, and Accius’s Medea may have all been set in Athens. See Jocelyn 1967, 342–47, for evidence and further bibliography on these plays. Jocelyn (1967, 122, 378–79) identifies fr. 112 as belonging to a second play by Ennius which takes place after the events of Euripides’s extant Medea in Athens, also told in Ov. Met. 7.398–424.

18. Along with allusions to other literary works. See Bloch 2000, 202.

19. Jacobson (1974, 102 and passim), who does not believe Ovid exploits fully the potential for irony in the Heroides, remarks, “This letter, however, is one of the major exceptions. Irony is so pervasive, so informing a factor that one is almost inclined to suggest that the poem exists for the irony in it,” 102.

20. See Leigh 1997, 607, on the metapoetic significance of fama in this epistle. See further Feeney 1991, 184–87, 247–49, and Hardie 2012 on fama as literary tradition and poetic innovation in Verg. Aen. and Ov. Met. (12.53–58; 15.871–79).

21. Jacobson (1974, 98 and passim) connects the theme of “telling, saying” to Hypsipyle’s opinion that Jason has insulted her and failed to recognize her merit. See also Leigh 1997, 607.

22. Ross 1975, 78. See further Conte 1986, 57–69; Hinds 1998, 1–5; Miller 1993; and Curley 2013, 154, on the Alexandrian footnote and other techniques for signaling intertexts and allusions in Roman poetry. See chapter 2 n. 25 for an example of an Alexandrian footnote in Phaedra’s epistle at Her. 4.5. See OLD, fero, 38, “to get (from a source), derive.”

23. Also identified by Michalopoulos (2004, 97) and Vaiopoulos (2013, 128). See Miller 1993, 157–58 n. 10 for an example of narratur as an allusive cue at Met. 7.826–27.

24. Conte 1986, 57–63; Hinds 1998, 14–15; and Miller 1993, 153. See also Conte 1986, 57–63; Solodow 1988, 227–28; Miller 1993; and Hinds 1998, 3–16, for further discussion and examples.

25. Cf. Ars am. 2.551: “I remember, her own husband had given kisses” (oscula uir dederat, memini, suus).

26. See Armstrong 2006, 32–37, for Jason’s role as cunning storyteller at Ap. Rhod. 3.997–1004; 1083–101.

27. Noted by Knox 1995, 176. “He having fixed his eyes on the ground” (ὁ δ᾽ ἐπὶ χθονὸς ὄμματ᾽ ἐρείσας, Ap. Rhod. 1.784).

28. Ag. 861–65: “First of all, it is a terrible evil for a woman without her man to sit alone in the house, hearing many injurious reports: and for one man to arrive, then another to bring an additional evil worse than the last, shouting out misfortune for the house” (τὸ μὲν γυναῖκα πρῶτον ἄρσενος δίχα / ἧσθαι δόμοις ἔρημον ἔκπαγλον κακόν, / πολλὰς κλύουσαν κληδόνας παλιγκότους˙ / καὶ τὸν μὲν ἥκειν, τὸν δ᾽ ἐπεσφέρειν κακοῦ / κάκιον ἄλλο, πῆμα λάσκοντας δόμοις). Likewise, Sophocles’s Deianira receives news about Heracles from a messenger (Trach. 180–290).

29. Jacobson (1974, 101) comments on the “dramatic” nature of the stranger.

30. The messenger in Trachiniae, for example, is Deianira’s son Hyllus. He describes the gruesome death of Heracles (749–812).

31. Lines 31–38 are considered by some to be spurious. See Jacobson 1974, 99–100, for an explanation of this opinion and a persuasive argument for their authenticity.

32. Noted by Knox (1995, 186). Cf., Sen. Med., 123: incerta vecors mente non sana feror; Leo 1878, 166–67; Knox 1986, 211. See Curley 2013, 37–49, on Ovid’s Medea, the chronology and further bibliography.

33. Knox (1995, 191–2) accepts Koch’s emendation se iubet. Other readings include avet, favet, iuvet.

34. See the introduction “Generic Performances” and chapter 2 n. 134 for direct speech as structurally dramatic, with further bibliography.

35. Following, as it does, a direct quote here, the close reader of Ovid may suspect a quotation or close reworking of lines from another poem, as happens at Met. 14.812–16, where Mars quotes the Ennian Jupiter of Ann. Skutsch fr. 54 (cited above). Conte 1986, 57–63; Miller 1993, 153; and Hinds 1998, 3–16. Bloch (2000, 205) suggests that this instance of memini cites Medea’s epistle at 12.85–91.

36. Dangel (2002, 351) notes further correspondences with Ap. Rhod. 4.338–39, 345–50, 360–65, 376–84, and lists variants with bibliography.

37. Eur. Hipp. 1028–31: “I wish to die without glory, without a name, without a city, without a home, wandering the land as an exile, and I wish no sea or land welcome me” (ἦ τἄρ᾽ ὀλοίμην ἀκλεὴϲ ἀνώνυμοϲ / [ἄπολιϲ ἄοικοϲ, φυγὰϲ ἀλητεύων χθόνα,] / καὶ μήτε πόντοϲ μήτε γῆ δέξαιτό μου). Knox 1995, 201. See further Hinds 2011, 29–33, on this “exile pattern” of speech in Her. 6 and in Greek and Roman tragedy with further bibliography.

38. Hecuba after discovering Polydorus’s murder: “without a city, all alone, the most unfortunate of mortals” (ἄπολιϲ ἔρημοϲ, ἀθλιωτάτη βροτῶν, Eur. Hec. 811). See also similar language which Deianira uses to describe the captive women of Oechalia in Soph. Trach. 299–300: “I, seeing them unlucky in a foreign land, wandering without a home and without a father” (ταύταϲ ὁρώϲῃ δυϲπότμουϲ ἐπὶ ξένηϲ / χώραϲ ἀοίκουϲ ἀπάτοράϲ τ᾽ ἀλωμέναϲ). Deianira, an accidental Medea, embodies the unstable boundaries between female erotic passion and rage, for the outcome of each is indistinguishable. She herself warns sensible women against rage: “but as I said, it is not noble for a sensible woman to be angry” (ἀλλ᾽ οὐ γάρ, ὥϲπερ εἶπον, ὀργαίνειν καλὸν / γυναῖκα νοῦν ἔχουϲαν, 552–53). On the “desperation speech” in Greek and Roman literature, see, e.g., Fowler 1987; Dué 2006, 50–53, 120; Curley 2013, 147–51.

39. I am applying Conte’s (1986, 29–31) formulation of code models and exemplary models in my reading. Conte 1986, 31 and passim, understands two levels of literary imitation. The “exemplary model” is the work “precisely imitated” by an auther. The “code model” is the work that is representative of the genre and imitated generally by an author. Ovid’s Hypsipyle intentionally writes a letter which is most productively interpreted in terms of the Greek tragic code. Euripides’s Medea, moreover, serves as the preeminent exemplary model. See the introduction “Ovid, in Theory” for the application of Conte’s formulation of generic code models to Ovid by Curley (2013).

40. Lindheim (2003, 114–35) argues that Hypsipyle identifies herself in opposition to her rival, but also likens herself to her rival in an attempt to approximate the type of woman to whom Jason is attracted.

41. See chapter 1 for a discussion of this process in Ovid’s erotodidactic poetry.

42. Michalopoulos (2004, 110–11) points out Hypsipyle’s participation in Medea and Jason’s crimes at Ap. Rhod. 3.1203–6 and 4.424–34 in the form of her gift of a cloak to Jason, as well as Ovid’s suppression of these details.

43. Cf. Eur. Med. 1078–80: “and I know what sort of evil I am likely to do, but my passion is stronger than my deliberations, which is the origin of the greatest evil for humans” (καὶ μανθάνω μὲν οἷα δρᾶν μέλλω κακά, / θυμὸς δὲ κρείσσων τῶν ἐμῶν βουλευμάτων, / ὅσπερ μεγίστων αἴτιος κακῶν βροτοῖς). See Foley 2001, 253, for an alternative interpretation of this line as a combination of reason and emotion, not a triumph of passion over reason. In his later epic, the Ovidian Medea exclaims: “I see the better path and I approve it, but I follow the worse” (uideo meliora proboque, / deteriora sequor, Met. 7.20–21).

44. Ovid’s poet-praeceptor calls Medea barbara at Ars am. 2.382 in a short list of examples of cheated women: “The barbarian of Phasis avenged by means of her own children” (barbara per natos Phasias ulta suos). Fulkerson (2005, 45) notes that Medea later refers to herself as barbara (12.70, 105). See further Jacobson 1974, 101; Bloch 2000, 202; and Lindheim 2003, 119.

45. Lindheim 2003, 121–22; Michalopoulos 2004; Fulkerson 2005, 50; and Vaiopoulos 2013, 140. See Newlands 1997, 189–91, for Medea’s witchiness as dehumanizing in Met. 7.

46. Cf., Jason’s words to Medea (“wishing to save you and to beget princes as brothers to my children, a safeguard for the house,” σῷσαι θέλων / σέ, καὶ τέκνοισι τοῖς ἐμοῖς ὁμοσπόρους / φῦσαι τυράννους παῖδας, ἔρυμα δώμασιν, 595–97) and Medea’s own words to the chorus (“but I am all alone without a city and I am being injured by my husband, carried off as a prize from a barbarian land,” ἐγὼ δ᾽ ἔρημος ἄπολις οὖσ᾽ ὑβρίζομαι / πρὸς ἀνδρός, ἐκ γῆς βαρβάρου λελῃσμένη, 255–56). See also 536–38, 1330, where Jason claims to have done her a service by taking her from a barbarian land. Michalopoulos 2004, 98.

47. See Knox 1995, 192.

48. Ovid will later identify his abject land of exile, Tomis, as gelidus axis: Trist. 5.2.64; Pont. 2.10.48, 4.14.62, 4.15.36; Knox 1995, 193. See Rosenmeyer 1997 for the exiled Ovid as a Heroidean abandoned maiden.

49. Noted by Jacobson (1974, 101).

50. See also Lindheim 2003, 120.

51. See Verducci 1985, 63. See also Lindheim 2003, 120, on Hypsipyle as chaste wife.

52. See chapter 1, n. 42, for vir as “husband” with further bibliography.

53. See also Michalopoulos 2004, 113, on Hypsipyle’s status.

54. See Knox 1995, 187 ad 80, for Argolica and Argolidas meaning “of mainland Greece” generally.

55. Nepos, Atticus 12.1. See chapter 1, n. 8 for further discussion of the importance of Roman affinitates. We may see another famous example of this in the imperial family when Tiberius was forced to divorce his wife, Vipsania, and marry Augustus’s daughter Julia: “He was compelled to take Julia, Augustus’s daughter, as a wife immediately, not without great anguish in his heart” (Iuliam Augusti filiam confestim coactus est ducere non sine magno angore animi, Sue. Tib. 7).

56. Vaiopoulos (2013, 133–34, 146 n. 50) draws attention to connotations of p(a)elex (6.81), which can mean prostitute (from the Greek παλλακίς) and can connote a lower social status (Gellius 4.3.3), further strengthening Hypsipyle’s discursive construction of Medea as a bad wife—both because of her inferior social status and her sexuality. Michalopoulos (2004, 98) points out the irony of the common root in the names Alcimede and Medea—μῆδος. See further McKeown 1987–1998, vol. 2, 380, and Knox 1995, 154 ad 5.60, 187 ad 6.81.

57. Cf. Ap. Rhod. 3.639–40, where, in Medea’s monologue, she commands herself and Jason to look to their own homes and families: “Let him court an Achaean maiden far away among his own people, and may my virginity and my parents’ home be my concern” (μνάσθω ἑὸν κατὰ δῆμον Ἀχαιίδα τηλόθι κούρην, / ἄμμι δὲ παρθενίη τε μέλοι καὶ δῶμα τοκήων).

58. Jacobson (1974) reads Hypsipyle’s letter as preoccupied with gender and gender roles. He argues, however, that Medea and the Lemnian women represent one version of femininity (misandric and violent) which she has failed to perform, and that Hypsipyle wants to be Medea in order to be victorious over men.

59. Pace Fulkerson 2005, 54, who reads these strategies as both a testament to the poetic authority of the heroines and a means for “the reader to trap them in their canonical narratives.” Vaiopoulos (2013, 139–40) also notes Hypsipyle’s purposeful distortions and omissions of her story in her self-presentation. Helen makes this connection in Her. 16.95–98, where she cites Hypsipyle and Ariadne as abandoned maidens; Fulkerson 2005, 63. Medea’s resemblance to Ariadne in Met. 7, noted by Curley (2013, 127), further recommends this reading.

60. Although Theseus’s famous epithet in Catull. 64, perfidus, is used three times (132, 133, 174) to describe him, it only appears once in Hypsipyle’s letter as the abstract noun perfidia (perfidiae pretio qua nece dignus eras, 146), but the deceit is attributed to Jason. Hypsipyle also points out that, in one respect alone, the boys do not resemble their father, for “they do not know how to lie” (fallere non norunt, 124).

61. Tereus, whose story is discussed later in this chapter, uses false tears on two occasions in order to persuade his addressees of his feigned sincerity: to Pandion at Met. 6.471; and to Procne at 565–66. See further Westerhold 2019.

62. Lindheim 2003, 122–23. Fulkerson (2005, 53, n. 42) and Vaiopoulos (2013, 141) note the similarity between Hypsipyle’s curses on Medea and Catullan Ariadne’s curses on Theseus (64.188–201), and the irony they create. In both cases the heroines appear to be the magical cause of an outcome demanded by mythological tradition. Fullkerson (2005, 53) further notes genetics may be responsible for the effectiveness of both heroine’s curses, as Ariadne is the grandmother of Hypsipyle (Her. 6.115–16, discussed below). See Michalopoulos 2004, 111–16, for an analysis of the form of Hypsipyle’s curses in addition to intertexts with other sources.

63. Fulkerson 2005, 122–42. See chapter 2 for a discussion of Ariadne in Her. 4.

64. Fulkerson (2005, 2 and passim) interprets the fictional writers of the Heroides as “composing their texts together and with reference to the poetic issues of that community” (2).

65. See chapter 2 for this discussion.

66. Vaiopoulos 2013, 141. Knox (1995, 194–95 ad 115–16) does not accept line 115.

67. See, e.g., Catull. 64.256–64 or Met. 4.1–30 for a description of women’s ecstatic worship of Bacchus. Apollonius’s Lemnian women are likened to maenads when they attack the Argo: “resembling flesh-eating bacchantes,” Θυιάσιν ὠμοβόροις ἴκελαι, Ap. Rhod. 1.636. See further Panoussi 2019, 150–51, on the influence of Ap. Rhod. 1.627–39 and Dionysus as helping divinity in Her. 6. See also chapter 1, n. 31 for the charges attending the Senate’s actions against the Bacchanalia, which included sexual licentiousness.

68. See Gregory 1999, xxxv, on furies and revenge in Attic tragedy. For furies as tragic figures in Latin literature, see Gildenhard and Zissos 2007, 1.10–22 and n. 4.

69. Eur. Med. 1–8; Enn. Med. Joc. fr. 103; Catull. 64.171–72.

70. Leigh 1997, 606 and passim. Leigh remarks a second allusion to Dido, another abandoned, tragic figure in Augustan poetry, whose curses act as a passive revenge against her departing lover, at Verg. Aen. 4.657–58: litora … tetigissent … carinae. For Dido as a “tragic” voice in Vergil’s epic, see, e.g., Conte 1986, 158–84; Keith 2002, 263; Panoussi 2009; Curley 2013, 52–57; and Gildenhard and Zissos 2013. See Armstrong 2006, 52–61, for correspondences between Vergil’s Dido and Catullus 64.

71. Curley (2013, 14–16, 92–94, 146–47) and Larmour (1990) have identified this pattern, which Curley calls a modesty-desire topos. See further chapters 1 and 2, as I apply his formulation to my reading of these women as interchangeable abject desiring subjects.

72. Cf., e.g., Ars am. 3.33–40, where Ariadne appears in a list of women deceived—Medea, Ariadne, Phyllis, Dido—or Rem. am. 743–45, where Phaedra and Ariadne are cited as examples of wealth facilitating poor romantic decisions.

73. Vaiopoulos (2013, 140) argues that Hypsipyle’s adoption of the curse constitutes an imitation of Medea’s magic. See also Fulkerson 2005, 50–55, and Lindheim 2003, 124, on Hypsipyle’s magical curse.

74. Jacobson 1974, 104.

75. Vaiopoulos 2013, 134–35. He further notes a possible intertext with Ap. Rhod. 3.776 where Medea is called an Ἐρινύς. See also Michalopoulos 2004, 100 and nn. 38, 39.

76. See n. 55 above for a discussion with further bibliography of the importance of marriage in cementing political relationships in ancient Greece and Rome.

77. See, e.g., Eur. Med. 20–23, 492–98, 1391–92; on Medea’s engagement with the masculine language of oath-swearing, see, e.g., Williamson 1990. See Newlands 1997, 184, on Jason’s oaths in Met. 7 to Medea.

78. Williamson 1990, 18–19 and passim.

79. See Bloch 2000, 202, for intertextual allusions to Her. 12.

80. See n. 55 above.

81. In Her. 7, Dido makes a similar offer to Aeneas—Carthage as dowry (149–50). I owe this insight to an anonymous reviewer for Cornell University Press. Cf. also the first of the anonymous Elegies for Maecenas (Miller 2009, 92–93): hic modo miles erat, ne posset femina Romam / dotalem strupri turpis habere sui (53–54). In this poem, Actian Apollo prevents Cleopatra from taking Rome as her dowry.

82. Knox 1995, 175 ad 17–18; Bloch 2000, 199; and Michalopoulos 2004, 100.

83. Cf. also Her. 7.133–38; Bloch 2000, 199, n. 18.

84. Apollonius offers hints of the erotic relationship between Hypsipyle and Jason. At 1.850–52 the narrator says that Cypris, acting on behalf of Hephaestos, inspired “sweet desire” (γλυκὺν ἵμερον); at 1.872–73, Heracles urges the heroes to get back to their own affairs while Jason stays in “Hypsipyle’s bed” (ἐνὶ λέκτροις / Ὑψιπύλης); at 3.1206, Jason puts on a robe given as a token of their “frequent or loud-voiced beddings” (ἀδινῆς μνημήιον εὐνῆς).

85. Bloch 2000, 200–201.

86. Cf. Eur. Med. 475–90; Ap. Rhod. 4.355–67; Bloch 2000, 201 and n. 23; see Bloch 2000, 201, n. 24, for further bibliography on Jason as a disappointing hero in Apollonius.

87. Bloch 2000, 201. Cf. Eur. Med. 488–89; Am. 2.18.23–24, where Ovid lists the epistles, either Her. 6 or 12 or both are identified as “what the ingrate, Jason, might read” (quod male gratus Iason … lega[n]t); Her. 17.193; Bloch 2000, 197 and n. 4.

88. Fulkerson (2005, 45–46) notes that, in the Heroides, both Hypsipyle and Medea refer to their paelices as hostes (6.82; 12.182), a unique occurrence in the collection.

89. Jacobson 1974, 104, and Verducci 1985, 62. Lindheim (2003, 123–24) argues that Hypsipyle’s self-presentation is a wish, pointing to the imperative and subjunctive moods in the final lines of the poem. Indeed, her potential Medea-ness is contradicted by her own mythological tradition, where she refuses revenge upon her father.

90. Commenting on 6.139 (“I condemn the Lemnian women’s crime, I do not admire it, Jason,” Lemniadum facinus culpo, non miror, Iason), Fulkerson (2005, 52) suggests that Hypsipyle’s failure to kill Thoas was due to naivety and that, having learned what men are really like, she has assumed the vengeful jealousy inherent in all Lemnian women. Vaiopoulos (2013, 139) argues that Hypsipyle’s participation in the Lemnian androcide already makes her a Medea.

91. Cf., e.g., Ars am. 3.672, where the praeceptor figures his warnings against infidelity to his female pupils, whom he calls Lemniasi, as “swords” for his own death. In this passage, the Lemnian women are paradigms of betrayed man-killers. He describes his own suicidal behavior in language used by Ovid for women in love (quo feror insanus, 3.667), suggesting that it is a mad, abject passion which would drive a man to be honest with a woman. See Gibson 2003, 352 ad 667–68, who compares this phrase to Dido at Verg. Aen. 4.595, Byblis at Met. 9.508–9, and Myrrha at Met. 10.320.

92. Clauss (1997, 161–64) compares Medea’s conflict between shame and love in book 3 of Ap. Rhod. with Phaedra’s in Eur. Hippolytus II.

93. In Medea’s letter, if we accept its authenticity, we may see a dramatization of this metamorphosis from desiring woman to avenging woman, and from Hellenistic epic to tragedy. See Verducci 1985, 66–81; Newlands 1997, 179 n. 3; and Fulkerson 2005, 40–66. See, e.g., Jacobson 1974; Hinds 1993; and Knox 2002, 120–21, on the debate over Her. 12. Verducci (1985, 17) notes that Her. 12 is the only treatment where a mature Medea is represented as a sympathetic woman in love. Newlands (1997, 182–83 and n. 8) notes that Medea’s choice to pursue deteriora despite her recognition of meliora (at Met. 7.20–21) transforms the Euripidean victory of anger over reason (Eur. Med. 1078–79) into the victory of love over reason. Foley (2001, 253) has argued that θυμός should be read as “seat of rationality” and not anger, giving the Euripidean lines another meaning entirely.

94. Conte 1986, 159–84.

95. Feldherr (2010, 224–32) traces this “textualization” of Procne, Philomela, and Tereus in the episode and in Fasti 2 in connection with the theme of reception. Feldherr (2010, 232) argues that reducing the characters to their literary exemplarity facilitates the objectifying “gaze” of Ovid’s audience. See the introduction “Ovid on the Intersection” for Segal’s (1986) theory of Greco-Roman myth as a megatext.

96. See further Ant. Lib.; Apollod. 3.14.8; Hyg. 45; Paus. 1.5.4; and Thuc. 2.29.3. See March 2003, 143–51, and Panoussi 2019, 247 n. 2, for a good overview of the mythological tradition.

97. Cf., e.g., Aesch. Supp. 57–66; Aesch. Ag. 1140–49; Eur. Heracl. 1021–23; Eur. Hel. 1107–12; Eur. Phoen. 1514–18; Soph. El. 107, 148–49, 1077; Soph. Aj. 629–30; Soph. Trach. 962–63. See further, e.g., Loraux 1990, 84–100; March 2003; and Monella 2005 for Greek and Roman treatments of the tale.

98. Sutton 1984, 132; March 2003.

99. Sutton 1984, 129–30.

100. On Sophocles’s Tereus, see Sutton 1984, 127–32; Curley 1997, 2003; March 2003; Monella 2005, 79–125; Boyle 2006, 133–37; and Coo 2013. On Livius Andronicus’s and Accius’s Tereus, see Boyle 2006, 133–37; Dangel 2002, 346–47. See Anderson 1972, 206–37; Sutton 1984, 130–32; Ciappi 1998; Curley 2003, 2013, 28–29; and Monella 2005, 173–220, for a comparison of the Greek with the Roman play.

101. Boyle 2006, 134, 158–59. Gildenhard and Zissos (2007 n. 4) and Feldherr (2010, 215) cite Cic. Att. 16.2.3 and 16.5.1.

102. See further, Sutton 1984, 129, and March 2003.

103. See Pavlock 1991 for Ovid’s characterization of Tereus as a stereotypical tyrannus/rex.

104. See March 2003, 149, on the possible absence of Bacchus in Sophocles’s play, but Bacchic elements in other variants.

105. Feldherr (2004) and James (2004) both argue that Marsyas’s tale expresses the challenges of literary innovation and generic limitations in the Augustan era.

106. “He strikes the ground stained with his criminal blood as he dies” (tundit humum moriens scelerato sanguine tinctam, 5.293); “to this day the stains of murder have not left their chest, and their feathers are marked by blood” (neque adhuc de pectore caedis/ excessere notae, signataque sanguine pluma est, 6.669–70).

107. For the tragic nature of Ovid’s version of this tale, see, e.g., Curley 1997, 2003, 189–97 and passim; Gildenhard and Zissos 1999, 2007; Schiesaro 2003, 83–85; Feldherr 2010, 199–239; and Hinds 2011, 5–6 and passim.

108. Curley 2003, 192.

109. The two anaphoras (hac aue, 433, 434; quaque, 436, 437) which begin the lines, interrupted by a single line of variation (435), draw our attention back to the first anaphora announcing the tragic theme and sound almost like a tragic choral ode in its repetition of an idea which offers an interpretative guide for the audience—“Eumenides … Eumendies … with this bird … this bird … that day … that day.”

110. See n. 34 above.

111. Curley (2003, 192–93) notes her speech is “tantamount to a dramatic rhesis.”

112. I count, inclusively, lines 424–674.

113. See Curley 2013, 96–107, and Westerhold 2014, 3–4, for a similar prologue and dramatic setting in the Hecuba episode, Met. 13.429–38. See further, Segal 1969; Rosati 1983, 129–52; Hinds 1987, 33–42, and 2002, 136–40; and Curley 2013, 95–133, for Ovid’s variations on locus est as a signal of a locus amoenus, and the dramatic characteristics of these scenes in his poetry. See Coleman 1977, 132 ad 4.6, with further bibliography, on the myth of Saturn’s flight and exile in Italy, called tempora Titan at Ov. Met. 6.438 and Saturnia regna at Verg. Ecl. 4.6.

114. See Anderson 1972, 174, 211 ad 6.165, 451–54; Curley 2003, 192; 2013, 99; and Feldherr 2010, 211, on ecce as a signal of a dramatic entrance in Vergil, Ovid, and in this passage in particular.

115. Ovid uses gravior and maior to describe tragedy at, e.g., Am. 3.1.23–24. See chapter 2 for a discussion of generic terms in this poem. See further, Hinds 1993, 39–40, and Trinacty 2007, 67. Descriptions of late Republican and early Imperial performances recommend this reading. Cicero, for example, in a letter to Marcus Marius (Fam. 7.1), is critical of the ostentatious tragic costumes, sets, and props presented for the opening of Pompey’s Theater in 55 BCE. He mentions the six hundred mules included in Accius’s Clytaemestra and the three thousand mixing bowls featured in a Trojan Horse (7.1.2). Shackleton-Bailey (1977, 326) notes that both Livius Andronicus and Naevius were known to have written a Trojan Horse.

116. Hardie 2002, 268–69, and Curley 2003, 192. Citing March (2003, 157–61), Curley (2003, 191 n. 49) further notes Procne’s self-diguise may reverse the disguise of the Sophoclean Philomela, whom March posits was dressed as a house servant and kept in the palace.

117. See further Rosenmeyer 2001, 61–97, for the embedded letters in these tragedies.

118. See Panoussi 2019, 140–46, for Bacchic ritual in this passage. Curley (2003, 178–89 and passim) argues that Bacchus played an important role in Sophocles’s Tereus, perhaps as the deus ex machina resolving the tragedy (TrGF 4.589). He cites Fitzpatrick (2001, 99), who argues for Apollo. Were it Apollo, the adoption of Bacchus by Ovid would resemble other reversals which Curley (2003, 190–91 and n. 49) remarks in the Ovidian adaptation of Sophocles’s tragedy. He notes but rejects the conjecture that Ovid is adopting an Accian innovation (Dangel fr. 445, cited above; Curley 2003, 181).

119. March (2003, 149), citing Apollod. 3.14.7, notes that Pandion’s reign witnessed the introduction of the worship of Dionysus. Itys’s murder may have been a sacrifice to the god in an earlier variant. She cites further Burkert (1983, 179–85). Such a variant, if known to Ovid, would create a stronger Bacchic resonance to Procne’s filicide.

120. Fitzpatrick 2001, 98; Curley 2003, 193–94; and Newlands 2018, 160. See Newlands 1997, 206; Curley 2003, 193–95; and Natoli 2017, 74, on Philomela’s carmen miserabile as a text. See March 2003, 160, for the importance of literacy as a mark of civilization and difference in Sophocles’s play. See Natoli 2017, 74, for the intertext with Verg. Geo. 5.111–15.

121. Curley 2003, 194; Feldherr 2010, 209–10. See further Schiesaro 2003, 74 and n. 13. Newlands (1997, 206 and n. 47), in her comparison of Orithyia and Medea in Ovid’s Met., argues that the speech of Medea, represented by her magical carmina, is both the source of female power and a threat to male power. By analogy, therefore, we may interpret Tereus as silencing Philomela for fear of her speech, but her carmen reveals his crime and subsequently destroys the family.

122. Feldherr 2010, 211–12.

123. Gradivus is also an archaic epic epithet. Ovid uses it again for Mars at Romulus’s apotheosis (Met. 14.820) and at the narrator’s prayer to Mars (and the other gods) to delay the death of Augustus (15.863). Anderson 1972, 208–9 ad 426–28.

124. For these terms as generic signals, see n. 115 above.

125. Also noted by Newlands (1995, 164).

126. Cf., e.g., Cephisus’s rape of Liriope: uim tulit, 3.344; Sol’s rape of Leucothoe: uim passa, 4.233; Neptune’s rape of Caenis: uim passa, 12.197; or Vertumnus’s anticipated but unnecessary rape of Pomona: uimque parat: sed ui non est opus, 14.770. Per OLD, uis is used in the singular as “violence” (1) and “a force used to obtain sexual gratification” (2); in the plural as “hostile strength” (21). See Kennedy 2012, 194–95 and passim, on the play of literal and metaphorical meanings of vis and other words in Ovid’s elegy. On Ovid’s rapes, see Curran 1978 and Richlin 1991.

127. See n. 115 above for the tragic connotations of Philomela’s Athenian attire.

128. Feldherr (2010, 209) identifies this generic shift earlier, at Procne’s first direct speech (440–44). Curley (2003, 167 n. 6) suggests that Ovid’s emphasis on Tereus’s gaze adapts the Sophoclean pun of ἐπόπτην (watcher) and ἔποπα (hoopoe) and Tereus and τηρέω (to watch over).

129. Currie 1981, 2729; Dobrov 1993; Barsby 1996; James 2012; and Curley 2013, 46 and n. 130, 71.

130. Characters like Plautus’s Lysidamus in Casina are foolish because they are not the right age to be lovers.

131. An oracle tells Thyestes that the child of an incestuous union with his daughter, Pelopia, will take revenge for his brother’s revenge—the cannibalistic meal. This child is Aegisthus, the coagent of revenge in other tragedies. Feldherr (2010, 213) notes that the comparison of the spectacle to the metaphorical food for desire (omnia pro stimulis facibusque ciboque furoris / accipit, Met. 6.480–81) anticipates Tereus’s actual meal. In book 10 of Ovid’s Met., we see another mythological father committing incest when Cinyras is tricked into taking his daughter, Myrrha, into his bed.

132. See chapter 1 for the theatrical connotation of this phrase in Ars am.

133. See further Kennedy 2012, 195–96.

134. By contrast, Feldherr (2010, 212) interprets Pandion’s credulity as a result of his nationality, for an Athenian would be accustomed to a man playing the part of a woman (sub illa).

135. Pavlock 1991, 507, and Hardie 2002, 266. This may be an allusion to the mythological traditions related by Apollod. Bibl. 3.14.8 and Hyg. 45, in which Tereus claims Procne has died and takes Philomela as a second wife. For this tradition, see further Pavlock 1991, 507, and Hardie 2002, 266. Feldherr (2010, 217–25) has demonstrated that Ovid’s epic version, when read alongside Ovid’s tale of Lucretia in Fasti, explores the importance of family and marriage to Roman myth and ritual. See also Newlands 1997, 192–95, and Panoussi 2019, 141–42.

136. Anderson 1972, 211–12 ad 451–54. Apollo gives voice to his own fantasy of generic transformation of Daphne from epic nymph to elegaic puella: “He looks at her unstyled hair hanging down her neck, and [asks] ‘What if her hair were arranged?’ ” (spectat inornatos collo pendere capillos, / et “quid si comantur?,” Met. 1.497–98). See further Hardie 2002, 260–62, and Newlands 2018, 155.

137. See chapter 2 for a discussion of this passage and its theatricality.

138. OLD, fingo, 1–4 for literal meanings of “forming,” “shaping,” 5–10 for metaphorical meanings of “devising,” “inventing,” “contriving.”

139. In the last line of the simile (518), the captured hare (capto), Philomela, is surrounded grammatically by the lack of escape (nulla fuga) and its captor (raptor). Capto, spectat surrounding the caesura emphasizes both his gaze and its power. What is seen is possessed (capto … sua praemia, 518).

140. Ovid may be asking us to look even further back in the epic canon to Hesiod’s Catalogue of Women and its more proximal mediators, the Homeric Hymns and Callimachus’s Hymns. See, e.g., Hardie 2005 and Ziogas 2010 for the influence of epic catalogues on Ovid’s poetry.

141. See the introduction “Ovid on the Intersection” and chapter 1 on the paradigmatic status of these figures and further bibliography.

142. Curley (2003, 177–78) argues that her cottage imprisonment is an Ovidian innovation, for a tragedy could neither accommodate a second setting nor a time lapse of a year, nor would it be necessary. March (2003, 158–59), contra Calder (1974, 89), posits that Sophocles’s play followed the variant recorded by Ant. Lib. Met. 11 and introduced Philomela into the palace as a slave, with the short hair, costume, and mask of a slave.

143. Also noted by Feldherr (2010, 214).

144. Also noted by Pavlock (1991, 39). Newlands (2018, 157–58), noting the elegiac connotation of domina (660) to refer to the mutilated Philomela and an intertext between this violent scene and one found in Verg. Aen. 10.395–96, argues that Tereus’s genre blending diminishes his epic status and possibly foreshadows Philomela’s future revenge.

145. See the introduction “Generic Performances” for the metapoetic resonance of a forest in Am. 3.1.1 (stat uetus et multos incaedua silua per annos), where, as I argue elsewhere, Ovid’s amator strolls through an ancient forest in order to find a new literary project. Ovid’s verse is in dialogue with two famous examples of metapoetic silvae, which Tereus’s scene also resembles. Cf. incedunt arbusta per alta, Enn. Ann. 175 Sk.; itur in antiquam siluam, stabula alta ferarum, Verg. Aen. 6.179. Both forests allude to Homer’s forest at Il. 23.114–20; Westerhold 2013. Like Tereus’s woodland hut, each Latin forest has depth—alta. Verg. Aen. provides an even closer intertext, for his ancient forest acts as a deep abode for wild animals. In Vergil, the forest as stabula alta offers timber for the pyre of Misenus. Tereus’s stabula alta may also be seen to memorialize the metaphorical death of Philomela the maiden, the lost sister of Procne. For the figurative use of silva, see further Bright 1980, 20–49; Hinds 1998, 11–14; Keith 1999, 41–62; 2008a, 125; Petrain 2000, 409–21; Wray 2007, 127–43; Newlands 2011, 6–7; Walters 2013; and Westerhold 2013.

146. See chapter 2 for Julia Kristeva’s theory of poetic semiosis and my application of her “thetic stage” to the Ovidian letters of Phaedra (Her. 4) and Byblis (Met. 9).

147. Tereus’s performance before an Athenian audience further emasculates him for the Romans. Pandion remarks that Tereus shares the desire of the sisters: et uoluere ambae (uoluisti tu quoque, Tereu) (497). In his discussion of this passage, Feldherr (2008, 38–39; 2010, 204–7), citing Webb (2005), adduces Juv. 3.93–97, which describes a Greek actor becoming the woman he plays. Ovid’s tale, like Juvenal’s satire, expresses the threat of mimesis for both actor and spectator, which destabalizes the barrier between self and other, including especially gender barriers. Moreover, Ovid tells us in Ars am. 1.276 that women dissimulate better: uir male dissimulat, tectius illa cupit (see chapter 1 on this passage). See further Kennedy 2012, 196–97, on the danger of mimesis as it is explored at the end of Ars am. 1.

148. Well noted by Panoussi (2019, 140–46).

149. See Feldherr 2010 for a sustained consideration of focalization in the Metamorphoses and its effect on audience identification.

150. See further Gildenhard and Zissos 1999, 165–67, on the use of accusative participles and the gaze in this passage; and Curley 2013, 117, on looking at Hercules’s pain in Met. 9.

151. Pavlock 1991, 39. March (2003, 154–55) suggests that Sophocles’s Tereus represented Procne as the strong sister (i.e., Antigone or Elektra), citing Sophocles’s wont to contrast a pair of sisters, strong and weak. She notes, moreover, Procne’s privileging of sibling over husband and child (155).

152. Eur. IA 1375–76, 1555; Eur. Hec. 347, 548.

153. Cf. “he has put a sharp knife in his hand after drawing it from its sheath … the priest taking the knife” (ἔθηκεν ὀξὺ χειρὶ φάcγανον cπάcαc / κολεῶν ἔcωθεν, … ἱερεὺc δὲ φάcγανον λαβὼν, Eur. IA 1566–67, 1578); “Then taking a gilded knife by its handle he drew it from its sheath” (εἶτ᾽ ἀμφίχρυcον φάcγανον κώπηc λαβὼν / ἐξεῖλκε κολεοῦ, Eur. Hec. 543–44).

154. Iphigenia aspires to prevent the rape of all Greek women by her death, IA 1378–84; cf. Eur. Hec. 367, 548–52.

155. Natoli (2017, 33–79) argues that speechlessness in the Met., including Philomela’s, attends characters who are removed from human communities.

156. See Newlands 2018, 152–64, who also comments on Tereus’s misidentification with divine erotic violence. Pavlock (1991, 39–40) compares the two passages and further notes the simile at 6.516–17, which aligns Tereus with Jupiter through his bird.

157. Jupiter’s dark forest is the lair of wild beasts (latebras intrare ferarum, Met. 1.593), perhaps offering another intertext to the forest of Verg. Aen., which was the deep dwelling of wild beasts (stabula alta ferarum, Aen. 6.179). Noted by Newlands (2018, 154).

158. Natoli (2017, 54–65) includes Io and Philomela (65–79) among the silenced characters that use writing to overcome their obstacles to communication. He further connects this with Ovid’s own use of poetry to reclaim a voice from relegation.

159. See Dodds 1960, xvii; Schiesaro 2003, 85; Curley 2013, 143–45; and Panoussi 2019, 117–67, for the the moral confusion associated with Dionysian worship and tragedy. Schiesaro (2003, 76 n. 18) further connects Procne’s confusion of good and bad to the god of tragedy through an intertext with Hor. Carm. 1.18.7–11, where excessive worship of Bacchus leads to an intoxicated confusion. The adjective Sithonia is used in both loci (Carm. 1.18.9; Met. 6.588).

160. Curley 2003, 185–86; Fantham 2004–2005, 123; and Westerhold 2014, 310 n. 69.

161. Feldherr 2010, 209–10. See also nn. 95 and 106 above.

162. Feldherr 2008; 2010, 199–239.

163. See, e.g., Curley 2003, 2013, 134–41, 229–30; Feldherr 2008, 37, 2010, 203 and n. 6 for intertexts with Attic Greek tragedy and further bibliography. Gildenhard and Zissos (2007, 4.36–37) interpret Procne as an author who innovates on Philomela’s carmen miserabile (6. 582) and provides a model for future authors (4.37).

164. Curley (2003, 187) argues that Procne’s maenadic disguise is transferred by Ovid from the Pentheus passage in Met. 3. Curley (1997, 320) further notes that Itys’s repeated cry “mother” (6.640) resembles that of Pentheus who cries to his own mother at Bacchae 1118 and 1120. Itys’s cry, Curley (1997, 320–22) argues further, may also replace Procne’s own traditional lament of “Itys, Itys.” See further Feldherr 2008, 44–45, 2010, 230–31, and March 2003, 141.

165. TrGF 5.1.391–97. Sophocles may have written three plays about the brothers, one of which told the story of Atreus’s revenge: TrGF 4.247–69. See Jocelyn 1967, 418–19, and March 2003, 150 and n. 33, for a list of Greek playwrights known to have written a play by this name with further bibliography. Schiesaro (2003, 70–138) traces Seneca’s reception of Ovid’s Tereus through allusions in his Thyestes, further confirming the recognition of Ovid’s own allusions.

166. See Boyle 2006, 79, for the political importance of Epirus as the possible setting of Ennius’s Thyestes. See Boyle 2006, 127–28, for the Accian genealogical connection of the Atreidae with the Romans via Evander and the association of Atreus with Tiberius Gracchus when Accius’s play, Atreus, was produced. See Bilinski 1958, 44–45, on the first production of Accius’s Tereus as an attack on Marius and his party.

167. For Accius’s Atreus, see Jocelyn 1967, 414–15, and Boyle 2006, 127–33. For Varius Rufus’s Thyestes, see Boyle 2006, 161–62, and Hollis 2007, 256–58, 275–78, frr. 153–56. We know of seven Roman treatments of this myth, including one by a Gracchus (Hollis 2007, 276, 335–37, fr. 200). Ennius wrote a Thyestes (Joc. frr. 149–160), which, Jocelyn (1967, 412–19) argues, covered the events predating the feast, but he also provides an overview of arguments for the inclusion of the feast in the plot. See further Boyle 2006, 78–83.

168. The association of Tereus and Tarquinius in the Roman mind is demonstrated by the substitution made of Accius’s Tereus for his Brutus at the Ludi Apollinares by Gaius Antonius, the brother of Marc Antony. Boyle (2006, 158) notes that, although Brutus, Caesar’s assassin, intended the praetexta about his ancestor, the substitution “affected little the political semiotics of the occasion.” See Cic. Phil. 10.8 for a contemporary account of the play. See further Boyle 2006, 158–59, and Feldherr 2010, 215–16.

169. See Boyle 2006, 12–13; Feldherr 2010, 217–25; and n. 135 above.

170. Hinds 1993, 39–40. Williams (2012), citing this line and other signals in Met. 7, argues that “moreness” is the fundamental characteristic of the Ovidian Medea. See also chapter 2 for further discussion of Hinds’s interpretation of Her. 12 and the metapoetic resonance of this line with reference to Her. 4.19. Schiesaro (2003, 81 and n. 25) identifies an intertext with Procne’s statement and Seneca’s Atreus at Thy. 269–70 and further cites Prop. 2.34.66.

171. These terms are borrowed from Conte (1986). See n. 39 above. For the intratexts with Her. 12 and further intertext with Seneca, see Curley 2013, 226–27.

172. See also Larmour 1990, 131–34; Curley 1997, 320; March 2003, 155–56; and Gildenhard and Zissos 2007, 3.29. Newlands (1997, 193–94) sees less conflict in Procne’s decision by comparison to Euripides’s Medea.

173. In both Euripides and Ovid, the mothers are momentarily overcome with a maternal impulse, which nearly breaks their resolve. This maternal impulse, in each case, is precipitated by the bodies of their children. Medea says goodbye to her plans after her children gaze and smile at her (1041–42; 1048). For Procne, Itys’s embraces, coos, and kisses bring tears to her eyes (6.625–26; 628). Both women just as quickly restrain their emotions and affirm their original intentions (Med. 1049–51; Met. 6.629–30).

174. Sommerstein, Fitzpatrick, and Talboy (2006, 158–59) also note the influence of one play on the other but argue that Tereus need not precede Medea. See also the introduction “Ovid on the Intersection.”

175. Procne’s assumption of Tereus’s role has been well noted by previous scholars. See Joplin 1984, 45; Larmour 1990, 133–34; Pavlock 1991, 40–46; Segal 1994, 267, 269, and passim; Newlands 1997, 194–95; Gildenhard and Zissos 1999, 167, 2007, 3.30–36 and passim; Schiesaro 2003, 82–83; and Feldherr 2010, 202.

176. In book 6, ater is used only three times. The first instance is to describe the Niobids’ mourning attire (uestibus atris, 288) as Apollo and Artemis kill each one. The next is found in this passage, just ten lines earlier, to describe the dark earth into which Philomela’s dying tongue murmurs (terrae … atrae, 558). Has her tongue called up the fury that Procne will become?

177. Hardie 2002, 268–69; Curley 2003, 192. Citing March (2003, 157–61), Curley (2003, 191 n. 49) further notes Procne’s self-diguise may reverse the disguise of the Sophoclean Philomela.

178. See, e.g., Boyle 2006, 21; Manuwald 2011, 69.

179. See Shaw’s 1975 seminal article on the “the female intruder” in Attic tragedy; see, e.g., Foley 1982 and Easterling 1987 for responses to Shaw.

180. See further chapter 1 for a detailed discussion of maenadic and Dionysiac imagery in Ovid and Latin literature in general, and Panoussi 2009 for tragedy in Vergil’s Aeneid. It has also been noted that Ovid’s maenadic Procne looks back to two “tragic” characters in Vergil’s epic, Dido and Amata. Hardie 2002, 269 n. 20, and Curley 2013, 27. Vergil’s Amata, driven, like Procne, by a fury (Aen. 7.351), costumes herself as a bacchant in order to sneak her daughter, Lavinia, away from the palace and prevent her marriage to Aeneas (7.385–88). See chapter 1, n. 39, for further discussion of this passage. Dido is likened to Pentheus and Orestes chased by the furies (Aen. 4.469–73). Dido is further likened to a maenad when she learns of Aeneas’s impending departure (4.301–3). This allusion is especially strong: in both cases the women experience dolor (Met. 6.595; Aen. 4.296); the rites are introduced by a temporal adverb (quo, 587; ubi, 4.302); they are described as trieterica in the same sedes (trieterica Bacchi, Met. 6.587; trieterica Baccho, Aen. 4.302); and the rites are located temporally at night (Met. 6.588–90; Aen. 4.303), and geographically by the name of a mountain (Rhodope, Met. 6.589; Cithaeron, Aen. 4.303), which resounds with the celebrations (sonat Rhodope tinnitibus aeris acuti, Met. 6.589; nocturnusque uocat clamore Cithaeron, Aen. 4.303). Procne is called regina for the first and only time in this tale (Met. 6.590), a frequent description of Dido, used in the passage under consideration (Aen. 4.296), and crucial to her role in Vergil’s epic. Dido, however, charges Aeneas with pretending (“dissimulare etiam sperasti”, 4.305), while the narrator tells us that Procne herself pretends (Bacche, tuas simulat, 6.596).

181. Anderson (1972, 174–75 ad 6.165), in a comment about ecce as a “dramatic expletive” for introducing important scenes in epic, also notes Vergil’s use of “crowd scenes” to draw attention to an important figure. He cites Vergil’s use of the phrase magna comitante caterua for Laocoon at Aen. 1.497 and Dido at Aen. 2.40. While Anderson does not refer to this at 6.594, Ovid’s use of the crowd for Procne may offer another intertext with Dido: cf., Aen. 2.40. A more proximal intratext may be found between Procne’s “chorus” of companions with Hecuba’s: captiuarum agmina matrum, Met. 13.560. As noted above, Procne and Hecuba also share an exemplary status as poenaeque in imagine tota est, 6.586; 13.546.

182. See n. 68 above for the association of furies with tragedy.

183. Also noted by Natoli (2017, 77) and Panoussi (2019, 143–44 and 248 n. 16).

184. Noted by Gildenhard and Zissos (1999, 167).

185. In this respect, Tereus embodies the threat of drama to the actor who becomes what he plays. Tereus acted the part of an Athenian princess, Procne, whose family is the victim of his violence. He becomes Pandion, an Athenian victim of his violence. Tereus also embodies the threat of drama to the spectator who identifies with an effective actor and becomes what he watches. Tereus, through identifying with Pandion, eventually becomes him.

186. See Westerhold 2019 for the role of tears in this passage.

187. Panoussi (2019, 144) argues that maenadism creates a bond between the sisters.

188. Joplin (1984, 45–46 n. 36) cites Achilles Tatius, Leukippe and Keitophon, where “passionate women” like Procne are said to prioritize the pleasure of revenge over the harm done to themselves and their loved ones.

189. See Adams 1982, 197–98, on gaudia with sexual connotations. He cites Ovid Am. 3.7.63; Catull. 61.110; Tib. 2.1.12.

190. Aesch. Ag. 1389–92: “and spouting a swift gush of blood, he hit me with a black drop of gory dew, rejoicing no less than the sown corn rejoices in the water granted by Zeus in the bursting of the bud,” (κἀκφυσιῶν ὀξεῖαν αἵματος σφαγὴν / βάλλει μ᾽ ἐρεμνῇ ψακάδι φοινίας δρόσου, / χαίρουσαν οὐδὲν ἧσσον ἢ διοσδότῳ / γάνει σπορητὸς κάλυκος ἐν λοχεύμασιν).

191. Boreas’s appearance at the end of Procne’s tale may be a vestige of the Sophoclean and/or Accian play. Tragedies commonly ended with a deus ex machina. March (2003, 161 and n. 54) suggests that Sophocles began his play with Ares and ended it with Athena. We may see Tereus, son of Mars, as a trace of this god’s Sophoclean prologue, with Boreas, a real god, apparently resolving the contradiction of Procne’s Terean behavior.

192. Newlands (1997, 203–7) identifies this tale as the last of the “marriage group,” in which Ovid investigates gender inequities in love and power. She notes both that Orithyia, unlike Procne, Medea, Scylla, and Procris, offers no resistance to Boreas, and that Boreas “spells out here what the other tales imply: passion and force rule in human affairs as in divine ones” by eventually ignoring rules governing marriage and taking what he wants (205). See further Feldherr 2010, 233–35, who connects this myth with Roman foundation myths and Romulus in particular.

1. Pace Feldherr (2010), who traces the multiple, sometimes contradictory models of spectatorship which Ovid’s Metamorphoses presents to his audience.

2. See the introduction for a more detailed discussion of generic codes as defined by Conte (1986) and applied to Ovid by Curley (1999, 2013).

3. See the introduction for a detailed discussion of this theoretical term.

4. Cf. Butler 2006, 190. See the introduction for Butler’s explanation of the social pressures on individuals to perform normative gender roles as a society or state defines them.

5. See the introduction and chapter 1 for further discussion of this term.

6. See chapter 2 for Kristeva’s formulation of semiosis, the semiotic chora, and poetry’s power to modify language.

7. Enterline (2000) has argued that such an alignment in his Met. is a symptom of Ovid’s anxiety over poetic authority. Identifying Echo (among other female figures in the Met.) as a surrogate for Ovid, Enterline notes that “although the poet pictures his own survival on his reader’s lips, his own earlier story of the same circumstance [i.e., Echo] stresses two problems the final lines occlude: even the most faithful, literal revoicing alters the original” (57). See also Natoli (2017), who examines representations of writing as a means for recovering a voice.

8. Williams (2012) has made a similar argument with regard to Medea in the Met. He interprets her as a poet competing with Ovid for authorial control over the poem. Her failure proves Ovid’s mastery.

9. Cf. the myth of Achilles at Skyros, whose masculinity is too strong to be threatened by a girl’s costume. See Cyrino (1998) for a discussion of and further references to the Roman fascination with gender reversal and transvestism, with a focus on this myth in particular. On this myth in Rome, see also Heslin (2005). See Gold (1998) on transvestism in Roman comedy.

10. See Most (1992) on the figure of dismemberment in Neronian poetry, including Seneca’s tragedies, as an exploration of personal identity.

11. For Ovid’s relegation, see, e.g., Thibault 1964, 20–32; Hollis 1977, xiii–xvii; Williams 1994, 3–8; Gaertner 2005, 14 and n. 43, 24; Claassen 2008, 2–3; and Fulkerson 2012, 340 and n. 2.

12. See also Davisson 1993 and Johnson 1997, 403, on shifting exemplarity in the exile poems.

13. See chapter 2 for a discussion of Byblis’s tale in Met. 9. Tissol (2014, 75–76 ad 5–12) sees a resemblance between Byblis’s hesitation to identify herself by name at the opening of her letter to Caunus (Met. 9.530–34) and Ovid’s own ostensible hesitation to name himself at Pont. 1.2.5–12 (vereor ne nomine lecto / durus et aversa cetera mente legas, 7–8); 1.7.1–6; and 3.5.1–4.

14. As Claassen (2008, 4) notes, Pliny the Elder (HN 32.152) is the first instance of a vague reference to Ovid’s relegation to Tomis, with Statius (Silv. 1.2.254–55) being the second, and more explicit, reference.

15. Ovid’s exile was questioned by, e.g., Brown (1985), who includes in his argument the resemblance of the epistles from exile to Ovid’s own fictional epistolary poems by mythological heroines and heroes. See Williams (1994, 3–8); Claassen (1999, 19, 34 and 265 n. 125; 2008, 229–30); and Tissol (2014, 13–18) for a discussion of this theory.

16. It has been well noted that Ovid’s verse thematizes the contrast between his exilic location and Rome. See Evans 1983, 50–73, esp. 69–73; Videau-Delibes 1991, 151–60; Williams 1994, 11–12; and Larosa 2013, 35–37.

17. See, e.g., Williams’s (1994, 26–34) reading of Pont. 1.8, where Ovid contrasts the fertile landscape of Italy to his own barren Tomis, which lacks the garden in which Ovid was accustomed to write (37). Ovid’s complaint locates poetic production and agricultural production in the same space.

18. Cf. “an exile safer and a little more quiet” (tutius exilium pauloque quietus, Tr. 2.577–78); “a changed place” (mutati loci, Tr. 3.5.54).

19. See especially chapter 1.

20. See chapter 1 for this construction in Ovid’s erotodidactic poems.

21. These are the only tragic heroines treated with any depth in the exile poems. Evans 1983, 62; Oliensis 1997b, 187–88; and Ingleheart 2010. Natoli (2017, 108–22) argues that Philomela provides Ovid with a model for exile overcome by poetry.

22. See Jacobson 1974, 101; Bloch 2000, 202; Lindheim 2003, 119; and Fulkerson 2005, 45.

23. Oliensis (1997b, 190 and 193 n. 31), citing Ahl (1985), connects this etymology with Caesar, “the cutter,” cutting Ovid off from his home and from his work: “his body will never be reunited with his literary corpus” (190).

24. Ap. Rhod. 4.452–76 narrates the story of an adult Ab(p)syrtus deceived and murdered by Jason and Medea. We may also see in Ovid’s intertext with his own Met. (cf. Medea’s exclamation, “vicimus” inquit: 3.9.23, to Tereus’s: “uicimus!” exclamat, Met. 6.513) a hint that Medea is also inspired by the Thracian tyrant.

25. FGrH 3 fr. 32; Cic. Leg. Man. 22; Apollod. 1.9.24. Jason in Eur. Med. 1334 charges that Medea killed her brother “by the hearth” (παρέστιον). See Bremmer 1997, 85, for Pherecydes’s version, and passim for all extant versions of his murder.

26. Oliensis (1997b, 188) interprets the metapoetic message as Ovid-Medea’s “disowning of certain of his own ‘family members’ ” and Ovid-Aeetes mourning the loss. She further reads the collocation of legere and tristis in 32 as a reference to his Tristia—Medean poems meant to delay Augustus. Oliensis (1997b, 189–90) further notes a reference to Marc Antony’s famous treatment of Cicero’s hands and head, which he ordered nailed to the rostra. Sen. Suas. 6.17–21; Plut. Cic. 48–49. Oliensis (1997b, 193 n. 28) cites Schubert (1990, 97).

27. See further Filippi 2015, 206–8, for intertexts with Enn. Alexander, Ov. Her. 6, and Ibis.

28. Ovid’s self-associations with epic heroes can be found at Tr. 1.1.99–100; 1.2.4–12; 1.5.19–24, 54–84; 1.9.27–33; 3.5.37–42; 3.8.1–10; 3.11.62; 4.1.15–18, 31–32; 4.3.29–30, 63–78; 4.4.61–88; 5.1.53–64; 5.2.13–16; 5.4.11–12; 5.5.3–4, 41–58; 5.6.7–12, 25–28; 5.14.35–40; and Pont. 1.3.5–6, 61–82; 2.2.25–26; 2.3.41–48; 2.4.22–23; 2.6.25–30; 2.7.60; 3.6.17–20; 4.10.9–28, 78. Harrison (2002, 90–91), remarking Ovid’s generic alignment with epic from Tomis, compares Tr. 1 with Hom. Od. and Verg. Aen. See McGowan 2009, 177–201, for Ovid’s self-association with Ulysses in the exile poems. See, e.g., Videau-Delibes 1991, 51–105, Claassen 2008, 174–76, and Tissol 2014, 6–10, for a discussion of Ovid’s use of mythological exempla and his generic interaction with epic in the exile poems. See further Kenney 2001, 262–67, and Tissol 2014, 7 and 103–4, ad Tr. 1.4.

29. Claassen (2008, 174–75) has identified Jason as the most topical myth for Ovid’s exile. See further Videau-Delibes 1991, 54–56.

30. Elsner and Sharrock 1991, 159–61 and passim. Elsner’s study considers the Pygmalion episode from Met. as a metaphor for creating and viewing realistic art: “The ivory statue … generates him as a viewer-lover, just as the Metamorphoses generates us as its readers” (155). “Ovid suddenly brings the reader up against the boundaries of his own desire as generated by the text. What can happen in a story can’t happen in life. The erotic myth of Pygmalion’s statue turned to flesh is as much an assertion of absence as it was of fulfillment” (165).

31. This relationship is expressed explicitly through their representation as writing subjects (see especially chapter 2), and implicitly through their self-construction as an elegiac puella who is the materia of Roman erotic elegy. On this topic, see Wyke 1987, 2002; Gold 1993; Keith 1994; and chapter 2.

32. The realism of Am. 3.12 translates this into Corinna’s promiscuity, for which Ovid, her writer, is at fault: “she sells her body with my talent” (ingenio prostitit illa meo, 8). Horace expresses a similar (mock) fear for the future of his book once it has been published at Epist. 1.20, e.g., dirtied by the hands of commoners (manibus … volgi, 11), eaten by moths (12), or used as a school text (17–18). He addresses the book as if it were a separate being altogether from its author (1–9): “You seem, book, to be looking at Vertumnus and Janus” (Vertumnum Ianumque, liber, spectare videris, 1).

33. Pygmalion, notes Elsner (1991, 159 and passim), is both the creator and viewer. He can, therefore, completely control the interpretation of his statue as artistic creation. Ovid is not so fortunate. “If Pygmalion the artist is a figure for Ovid the writer, then the myth as a whole undermines the integrity of Ovid’s writing since he is not its only reader, he cannot control its meanings as Pygmalion does those of the statue” (Elsner and Sharrock 1991, 167 n. 28).