Ovid, in Theory

Here, I examine the ways in which Ovid’s poetry participates in the translation of Greek literature to Roman performance through a lens that brings gender and genre in dialogue with one another. To do so, I draw on the theories and vocabulary in scholarly literature in order to better understand Ovid’s process of gender and generic construction and its effect. For example, Judith Butler maintains that social identities are performative and theorizes the “assumption” of gender, sex, and kinship roles as an identification.¹² The roles which one may perform are defined and restricted by those recognized in one’s community. This identification is performative in as much as it is constituted through behavior and attire (e.g., by a Roman matrona) which act as a “citation” of the symbolic position with which the individual identifies.¹³ “Acts, gestures, and desire produce the effect of an internal core or substance, but produce this on the surface of the body, through the play of signifying absences that suggest, but never reveal, the organizing principle of identity as a cause. Such acts, gestures, enactments, generally construed, are performative in the sense that the essence or identity that they otherwise purport to express are fabrications manufactured and sustained through corporeal signs and other discursive means” (Butler 2008, 185). Applying Butler’s formulation, a Roman becomes a matrona because she persuasively performs this role. She dresses (by wearing the stola) and behaves (by advocating for one’s children) in a way which is recognizable to Roman society as a matrona. Just as the adaptations of Greek drama did for republican culture, so Ovid’s Phaedra and Medea offered new roles to his audience in the Augustan Age and provided a guide for how to cite these roles. Throughout his work the heroines continue to evolve as identities multiply performed by various characters in his corpus. However, Ovid’s verse demonstrates that Phaedra and Medea always ultimately represent monstrous identities, whose lust and rage must be punished and eliminated in order to protect the normative subjects with whom they interact.

In addition to Butler’s theory of the performance of gender, I draw on Julia Kristeva’s understanding of the abject to elucidate why Ovid introduces these novel identities only to repudiate them. Kristeva theorizes the abject as that part of the self or of a society which must be rejected and is then refused acknowledgment as part of the self or society. Instead, the abject is established as an other who embodies these distasteful characteristics in order to create a fictional individual or society—an ideal self or state which is free from characteristics now associated solely with the abject.¹⁴ Judith Butler expands Kristeva’s definition in her formulation of the abject: “The abject designates here precisely those ‘unlivable’ and ‘uninhabitable’ zones of social life,” excluded from subjecthood but necessary to delimit “the domain of the subject.” “It [the abject] will constitute that site of dreaded identification against which—and by virtue of which—the domain of the subject will circumscribe its own claim to autonomy and to life” (Butler 1993, 3).¹⁵ By embodying what is unacceptable for normative society, the female abject subject reminds a good Roman woman how not to act. At the same time, abject behaviors and desires originate in and are, in fact, undesirable characteristics of any normative Roman matrona. This places pressure upon a Roman woman to repudiate and associate these failures with an abject subject and introduces the potential for her failed performance of the normative identity matrona by admitting or expressing any of these abject characteristics. As we will see in this study, Ovid’s Phaedras and Medeas make efforts to find a place for themselves in society, but Ovid’s verse repeatedly constructs them as abject subjects, required to define what is normative in terms of gender and genre.

I have expanded my definition of Ovid’s Phaedras and Medeas to include those heroines who “cite” them, following closely Curley’s application of Conte’s concept of “code modeling.”¹⁶ He has demonstrated that characters in Ovid, who embody distinct tragic dilemmas or “code models,” act as “local models” or tragic paradigms. When characters who do not have their own tragic provenance, such as Byblis (Met. 9), take up such dilemmas, or tragic “codes” in Ovid’s corpus, “[their] story has been encoded as a tragedy” (Curley 2013, 16). When I use the term “tragic” in the following chapters, I refer specifically to the tragic genre. So as not to create confusion, I do not use “tragic” at any point in its vernacular sense, that is, mournful, pathetic, or disastrous.¹⁷ While it may seem strange to call Byblis, for instance, “tragic” when scholars do not have any evidence that her myth was represented in a tragedy, I employ the term to describe her and the other female desiring subjects who cite Phaedra or Medea. These heroines make up the corpus of Phaedra-like and Medea-like figures under consideration in this study.

The code-modeling which Curley identifies in these episodes is thus a generic paradigm, not a series of specific allusions (although we will see that allusions are in some cases very pointedly made to Euripides’s plays).¹⁸ Like Curley, I do not pursue line-by-line comparisons of Ovid’s use of a Euripidean source text or aim to establish Euripides as the primary source for the Ovidian representations I consider in the following chapters. Rather, taking up Ovid’s own cues, which he gives us through direct allusions to Euripides, general thematic correspondences, and lexical and rhetorical strategies which evoke the tragic stage, I consider how the echoes of the Euripidean Phaedra and Medea guide and inform the interpretation of these mythological figures in Ovid. More specifically, I consider how the two heroines, as established by their tragic representations, disrupt Ovid’s verse and are mastered by the poet.

Looking specifically at the tragic voices citing Phaedra and Medea in Ovid’s elegy and epic, we may see that the tragic heroine’s voice is a threat—sometimes playful, sometimes domineering, sometimes desperate. Each heroine comes to embody the abject self and song. Her excessive passions are always present as the narrator’s (and audience’s) own secret but disavowed emotions. Her tragic tradition is always already drawn from the same mythological matrix as the more reasonable Roman genres of epic and elegy. In this way she can be located spatially on the fringes of both community and canon—included as a member of the group but excluded from the inner circle. Her threat and marginal location are marked by her feminine gender and by her tragic genre.¹⁹

This monograph joins the growing work of scholars who have applied the theoretical concepts of abjection and monstrosity to Greek and Latin literature with great success. Their studies demonstrate how Greek and Roman identity was constructed and shored up through various processes of disidentification. Most recently, Spentzou (2018) interprets Lucan’s Caesar as an abject subject, destabilizing the categories of subject and object. Felton (2012), Del Lucchese (2019), and Mitchell (2021) see the monster as a marker of the boundaries defining what is human and normative. By contrast, Valladares (2021) traces an aesthetic of tenderness in Roman art and literature beginning in the first century BCE, which paradoxically embraced monstrous mythological figures. Valladares’s focus on the late republic and early empire joins those of Hardie (2009), Lowe (2015), and Pietropaolo (2021). Hardie’s (2009) collection, Paradox and the Marvellous in Augustan Literature and Culture, includes several contributions which broach this subject, especially those from Barchiesi, Platt, and Rosati. Lowe (2015) and Pietropaulo (2021) explore the monstrous and grotesque in Augustan poetry specifically. Bloomsbury’s Cultural History Series will soon include Felton’s forthcoming A Cultural History of Monsters in Antiquity. Felton is also the editor of the forthcoming Oxford Handbook of Monsters in Classical Myth.²⁰ This important work has illuminated Greek and Roman systems of categorization, marginalization, and oppression such as nationality, gender, ability, and sexuality. My reading of Ovid’s abject tragic heroines contributes to these studies by investigating how his poetic play both reinforces dangerous gender stereotypes and simultaneously reveals their fiction.

Phaedra and Medea in Greece

Euripides’s extant Hippolytus (Hipp. II), performed at the Athenian City Dionysia in 428 BCE, is his second and only surviving treatment of the myth.²¹ The action of the play takes place in Troezen while Theseus is away consulting an oracle. Phaedra, wife of Theseus, falls in love with her stepson, Hippolytus. The play begins with Aphrodite announcing that Phaedra’s desire is part of the goddess’s revenge upon Hippolytus. Phaedra has chosen to die rather than act on her desire, for the sake of her family’s reputation and her sons’ future, but her nurse approaches Hippolytus without Phaedra’s consent. Fearing the consequences if Hippolytus tells his father, Phaedra accuses her stepson of rape. This accusation is delivered in the form of a letter and read by Theseus after her suicide. Theseus, influenced by Phaedra’s false charges against Hippolytus, exiles and curses his son, who is killed by Poseidon’s bull, sent from the sea in response to Theseus’s curse.²²

In Euripides’s Medea, performed at the Athenian City Dionysia in 431 BCE, the eponymous heroine turns to murderous revenge when the hero, Jason, finds a Greek wife in Corinth.²³ We meet her in agony over her abandonment and watch her devise and implement a plot which results in the death, not of Jason, but of Jason’s children with Medea, his bride, and future father-in-law. Medea’s rage is provoked by Jason’s oath-breaking. She considers their marriage a sort of alliance between equals which he has violated. Her revenge appears to be a distorted version of the epic hero’s punishment of his enemies.²⁴ Unlike Phaedra, the Euripidean Medea has no gods motivating her excessive passion. The play ends with Medea flying from the stage on the mēchanē as if a god herself. Her myth continues with a career as a murderous evil stepmother to Theseus in Athens.²⁵

While Euripides tells one version of Medea’s and Phaedra’s myth, other poets in other genres foreground different themes and moments. Medea’s mythological tradition links her Euripidean revenge with Phaedra’s forbidden lust, for Phaedra’s story rehearses Medea’s at its beginning. In Apollonius’s famous epic, Medea is a maiden in love, who violates laws of kinship—betraying her father and killing her brother for love of a foreign (and therefore inappropriate) beloved.²⁶ Her erotic choice represents a violation of gender norms. It is the role of the male guardian, not a young girl, to choose a suitable match. Medea flees with Jason back to Greece, employing her magic to rejuvenate either Jason, his father, or the nurses of Dionysus, and her guile and reputation to persuade Pelias’s daughters to kill their father on Jason’s behalf.²⁷ Medea’s myth is expansive, encompassing a full life—maiden, wife, aged mother—and incorporating many tragic roles such as destructive lover, vengeful wife, and evil stepmother.²⁸ This Phaedra-like episode in Medea’s myth was fresh in the minds of the Roman literati as Ovid began his career, for Varro Atacinus had recently translated Apollonius’s epic Argonautica into Latin—Argonautae—sometime after 47 BCE.²⁹

Phaedra and Medea in Rome

As noted above, Phaedra and Medea would likely be familiar to a Roman audience from the stage. Moreover, elite Romans would know their myths and their literary tradition from school. Early Roman poets, beginning with Livius Andronicus, self-consciously chose and translated Greek models that were central to the Greek education of Magna Graecia.³⁰ Each act of translation was a judgment of the literary value of the text. Translations also consciously cited their Greek models and the later Roman models which mediated each new adaptation.³¹ As an example, Feeney (2016, 156) adduces Livius’s Odusseia, whose title and first line signals its model, Homer’s Odyssey, and its innovation, in Livius’s Latin Saturnians.³² This self-referential characteristic is especially strong in later translations. Euripides’s Medea, as I will discuss further below, was reworked by each subsequent generation of Latin poets, and we may see in successive translations expressions of an awareness of the past iterations as well as a desire to innovate.³³

We do not have any extant fragments of a Roman Phaedra or Hippolytus before Seneca’s Phaedra, a generation after Ovid’s death;³⁴ however, such an absence of the myth in Roman tragedy does not preclude access to and knowledge of the Greek plays. Suetonius (Vita Ter. 5) reports that Terence traveled to Greece to read Menander’s plays.³⁵ Ovid, at least, knew one of Euripides’s tragedies entitled Hippolytus, for this play is the first listed in a catalogue of tragedies which offer materiam … amoris (Tristia 2.382): “Is there anything in the Hippolytus except the blind passion of a stepmother?” (numquid in Hippolyto, nisi caecae flamma novercae, Tristia 2.383). Greek plays were being performed in southern Italy before Livius Andronicus’s Roman production; a Phaedra or Hippolytus may have been among them.³⁶ Feeney (2016) emphasizes the significance of Roman exposure to Greek theater in the third century as the Roman army expanded into Magna Graecia and the Greek mainland. In addition to southern Italians bringing Greek dramas to Rome and translating them into Latin, Roman soldiers spent a considerable amount of time in Sicily during the wars against Carthage and presumably attended Greek dramatic productions there.³⁷

While Phaedra is absent from our roster of extant Latin plays, the Roman tragedians produced several plays about Medea and Medea-like figures.³⁸ Medea herself appears in the tragedies of Roman playwrights Ennius,³⁹ Pacuvius,⁴⁰ and Accius.⁴¹ While Ennius’s Medea Exul and Accius’s Medea are believed to be adaptations of Euripides’s play, Ennius’s Medea may have been set in Athens and Pacuvius’s Medus in Colchis and Media.⁴² As Boyle (2006; 2012) and Cowan (2010) have argued, the Euripidean Medea offered the Roman audience the opportunity to assume or reject identification with a Greek character and Greek culture. Romans might have seen themselves in the Greek hero Jason facing a barbarian other, or in Medea for whom Greece is a foreign land and Greeks are enemies.⁴³

In the imperial period, tragedy seems to have been superseded to some degree by the pantomime.⁴⁴ The freedmen of Augustus and Maecenas, Pylades and Bathyllus, are credited with introducing the tragic and comic pantomime respectively to Rome in 22 BCE.⁴⁵ This was a more stylized performance with a single actor, which, nevertheless, incorporated arias and material from tragedy. Lucian advises the dancer of pantomime “not to be ignorant of anything that is told by Homer and Hesiod and the best poets, and above all by tragedy” (οὐδὲν τῶν ὑπὸ τοῦ Ὁμήρου καὶ Ἡσιόδου καὶ τῶν ἀρίστων ποιητῶν καὶ μάλιστα τῆς τραγῳδίας λεγομένων ἀγνοήσει, de Saltatione, 61).⁴⁶ He provides a list of popular pantomime subjects (37–61) which every good dancer should know, among which he includes the Athenian myths “concerning Theseus and Aegeus” (ὅσα περὶ … Θησέως καὶ Αἰγέως, 40), “the reception of Medea and her flight back to Persia” (τὴν Μηδείας ὑποδοχὴν καὶ αὖθις ἐς Πέρσας φυγὴν, 40), “the daughters of Pandion” (θυγατέρας … τὰς Πανδίονος, 40), and “Hippolytus’ suffering” (τὸ Ἱππολύτου πάθος, 40); the Corinthian myths, including “Glauce and Creon” (τὴν Γλαύκην καὶ τὸν Κρέοντα, 42); in Crete, “Phaidra” (τὴν Φαίδραν, 49); and in Thrace, “Jason” (τὸν Ἰάσονα, 52).⁴⁷

Generic Performances

With the exception of Ovid’s own tragedy, Medea, which is now lost but for two lines,⁴⁸ Ovid’s adaptations of Euripides’s heroines were twofold—translated from Greek into Latin and reconceptualized from tragedy into elegy or epic.⁴⁹ As Curley (2013) and others have noted, Ovid signals both strategies of adaptation through allusive play. In this process, Ovid’s construction of Phaedra and Medea as gendered abject subjects is also defined by the generic contexts in which Ovid locates them, inextricably linking their embodiment of characteristics undesirable to both the Roman woman and the genres of epic and elegy to their tragic provenance. They wreak havoc in Ovid’s Roman elegy and epic because they do not belong either at Rome or in the genres to which the Augustan poet has relocated them.

This play with tragic disruption of other genres by Phaedra and Medea is characteristic of Ovid’s thematization of genre throughout his poetry. There has been a great deal of scholarly attention to the dialectic of epic and elegy in Ovid.⁵⁰ This is no surprise as elegy is the genre most represented in his corpus, and elegy programmatically defines itself as epic’s antithesis.⁵¹ His elegiac poems include the erotic Amores; the erotodidactic Ars Amatoria, Remedia Amoris, and Medicamina; the epistolary Heroides; the calendar Fasti; and the exilic Ibis, Tristia, and Epistulae ex Ponto. Hinds (1987), reconsidering Heinze (1919), argues that Ovid’s poetry intentionally crosses the boundaries differentiating genres and signals these crossings while betraying anxieties over the resulting generic contamination.⁵² Hence, Ovid’s epic poem, Metamorphoses, has been called “a masterpiece of generic transformation” (Farrell 2009, 376).

Both undermining and observing generic expectation, however, rely on literary codes and conventions to which attention is drawn, especially in the process of their transgression. That generic codes are not the product of modern criticism is attested by, for example, Horace’s comments about genre in Ars Poetica (73–85), where the poet identifies themes and meters appropriate to each genre.⁵³ Horace contends that hexameter is suitable for kings and wars, elegiac couplets for lament and humble elegies, rage for iambics which are used by comedy and tragedy, lyric for gods, their children, athletic victories, and lovesickness. Ovid gives a similar description of the conventions of literary genre in Remedia Amoris (372–86). Responding to criticisms of his Ars Amatoria, the poet-praeceptor cites generic expectations as a defense for the content of his verse. He enjoins his pupil to “suit each [work] to its own meter” (ad numeros exige quidque suos, 372): war for hexameter; anger for tragedy; everyday life for comedy; insults for iambics; and love for elegy.

The passage, of course, invites suspicion. Ovid anticipates his audience’s laughter over such strict rules articulated in the context of a didactic poem written in elegiac couplets. Nonetheless, these programmatic rules and expectations accord with Horace’s earlier list. Ovid’s generic play relies on a familiarity with such traditional definitions of literary genre. A generic code is described by Conte (1986, 81) as a model in relation to which poets can define their own poetry. Reference to the “norm” of a generic discourse “delimits the common space within which new poetry can both emulate tradition and speak with a fresh voice.” Conte (1986, 93) sees genres as “definable precisely in their mutual relationship” and as “a horizon of historically formalized (and historically variable) literary expectations.” Codes would be familiar to an ancient audience, suggesting that the audience would recognize generic elements out of place, that is, in the wrong generic context. Therefore, Ovid’s crossing of genres in his elegiac and epic poems—incorporating tragic themes and language—could be identified by an ancient reader or auditor of his work.

While Roman erotic elegy programmatically defines itself against epic, Ovid also juxtaposes tragedy to elegy and other genres. In Amores 3.1, for example, the personified Elegy and Tragedy vie to be Ovid’s genre of choice.⁵⁴ The poet, on a stroll in the woods, is confronted by Elegy and Tragedy, who wear costumes emblematic of their genre:

uenit odoratos Elegia nexa capillos,

et, puto, pes illi longior alter erat.

forma decens, uestis tenuissima, uultus amantis,

et pedibus uitium causa decoris erat.

uenit et ingenti uiolenta Tragoedia passu:

fronte comae torua, palla iacebat humi;

laeua manus sceptrum late regale mouebat, (3.1.7–13)

Elegy came with hair bound and fragrant, and, I think, one foot was longer than the other. Her shape was becoming, her dress most thin, the face of a lover, and her feet’s imperfection was the reason for her beauty. Violent Tragedy also came with a giant stride: her hair was lying on her grim brow, her mantle on the ground; her left hand was brandishing widely a royal scepter,

Despite the dialectic of elegy and tragedy constructed so explicitly by the poet himself in his earliest literary endeavors, the function of tragedy in Ovid’s extant work has only recently begun to receive due consideration.⁵⁵ Most attention has been paid to the use of tragedy in the Metamorphoses⁵⁶ and Heroides,⁵⁷ in which scholars have identified and traced specific Greek and Roman tragic allusions and intertexts. More recently, tragedy in his exilic work has received attention.⁵⁸

Some recent scholarship considers how Ovid “stages” episodes in his Metamorphoses or the epistolary monologues of Heroides through consideration of dramatic elements: for example, creating a sense of spectacle with a setting which resembles a stage or an introduction which recalls a stage entrance or tragic prologue, including (in the case of the Metamorphoses) direct speech in imitation of dramatic dialogue, or employing the language of the theater in the narrative.⁵⁹ Thus far there has been only one monograph treatment of this topic, Curley’s Tragedy in Ovid (2013). My research expands on his important study of Ovid’s reception of Greek tragedy, which has demonstrated its pervasive influence throughout the poet’s work.

Ovid on the Intersection of Genre and Gender

As I argue in subsequent chapters, when Phaedra and Medea enter Ovid’s verse, they challenge the literary sophistication of his reader. Knowledge of both Greek and Roman tragedies, particularly their most famous songs, constituted “cultural capital,” and was taught in schools.⁶⁰ For this reason, individual characters could be invoked in a variety of contexts. Cicero, for instance, cites Ennius’s Medea in thirteen passages of his work for demonstrations of rhetorical techniques, emotional experiences, and philosophical arguments.⁶¹ In these cases, Cicero uses mythology as a symbolic system, adducing mythological characters as if they were words in a language to illustrate his arguments. Cicero relies on his audience’s familiarity with Medea’s myth in order to make his points clear. Charles Segal refers to this shared language of myth as a “megatext,” one which shows a “conscious awareness of sign systems” indicating “an advanced, if not explicit, semiotic consciousness” (1986, 51).⁶² Within this system, characters may act as paradigms for one another. Segal adduces the myths of young men transitioning (or not) to adulthood: Theseus, Perseus, Telemachus, Orestes, Phaethon, Hippolytus, and Actaeon (1986, 56). As symbols in a system, these myths and characters are also polysemic. For example, while at the beginning of their myths Telemachus and Orestes are both young men displaying great loyalty to their fathers, later in their stories Telemachus’s loyalty to his mother stands in stark contrast to the violent antipathy of Orestes to his. The megatext of myth is rendered even more nuanced when the speaker and addressee have carefully studied the literary provenance of each representation of mythological characters, as Cicero and his elite Roman audience had.

In my analysis of Ovid’s representations of the two tragic heroines, Phaedra and Medea, Segal’s formulation of this megatext shines a light on Ovid’s discursive strategies and those he attributes to his heroines. Considering the two heroines and their paradigms in his corpus, I investigate further what their generic origin connotes; how they manipulate their literary tradition, employing generic codes in order to redefine or reify their meaning in the megatext of myth; and how their gender affects the success of their attempts to control their connotation, specifically in their new, nondramatic poetic location. When Phaedra and Medea speak in Ovid’s work, their tragic lineage looms large, constituting a challenge to overcome or the inevitable outcome of the verses they compose.

My definition of Phaedra-like and Medea-like figures follows Curley’s formulation and includes those mythological heroines whose stories in Ovid’s corpus engage most closely with the plot and themes of Euripides’s plays. In the Ovidian Phaedra’s epistle (Her. 4), our heroine-writer makes reference to a number of traceable moments in Euripides’s extant play.⁶³ But Larmour has argued that the stories of Scylla, Byblis, and Myrrha—found in Metamorphoses 8, 9, and 10 respectively—incorporate those elements of Euripides’s Hipp. II which the Hippolytus-Virbius episode (15.479–621) does not.⁶⁴ In book 9, for example, “Caunus plays the part of Hippolytus”⁶⁵ by reacting to Byblis’s incestuous desire. Although in Ovid’s tale, the nurse is replaced by a letter, Caunus’s response (Met. 9.574–79) resembles that of Euripides’s extant Hippolytus (Hipp. II, 602–67) in its sudden and violent fury. Caunus’s flight also recalls Hippolytus’s exile in Hipp. II.⁶⁶ Other elements of the plot, such as Phaedra’s initial desire to die and the sophistic nurse who intervenes and mediates, show up in Myrrha’s tale.⁶⁷

As is the case for Phaedra, we may find traceable moments from Euripides’s Medea throughout Ovid’s work.⁶⁸ Potentially, the closest reworking of Euripides’s tragedy may have been found in Ovid’s own lost tragedy, Medea (cited above).⁶⁹ In the Heroides, both Hypsipyle’s (Her. 6) and Medea’s (12) letters anticipate the events of the Attic tragedy in what Barchiesi (1993) calls a “future reflexive.”⁷⁰ Ovid’s epic Medea follows Apollonius’s epic more closely, for her Phaedra-like tradition is foregrounded, while her Euripidean revenge is relegated to only a few lines (Met. 7.394–97).⁷¹ Nonetheless we find Euripides’s play restaged elsewhere in the Metamorphoses. Newlands (1997) interprets Medea’s tale as part of a larger “marriage group” found in books 6–8 (Procne, Medea, Scylla, Procris, Orithyia). Procne in particular completes her story by rehearsing Medea’s Euripidean infanticide at the end of book 6, preparing Ovid’s audience for his introduction of Medea at the beginning of book 7.⁷² Ovid’s epic Procne, moreover, is renewing the mutual influence of Euripides’s Medea and Sophocles’s Tereus in the fifth-century theater of Dionysus. Pointing to the chorus’s omission of Procne as a precedent for Medea’s crime at Eur. Med. 1282–84,⁷³ March (2003) has argued that Tereus could have been written after Sophocles saw the moving revenge staged by Euripides in 431 BCE. Alternatively, Sommerstein, Fitzpatrick, and Talboy (2006, 158–59) suggest that Tereus preceded Medea, and that Euripides pointedly suppresses the knowledge of Procne, the absence of which his audience would note. By incorporating famous elements of Euripides’s Medea in Procne’s tale, Ovid may be signaling his awareness of his tragic predecessors’ dialogue, further connecting both heroines in his Metamorphoses to fifth-century Attic tragedy.

The intersection of performance and performativity is crucial to Ovid’s construction of the Euripidean heroines. The literary representation of woman or girl, torn between her erotic desire and normative gender and kinship roles, acts as a citation of a very specific identity, that of the Phaedra of Attic tragedy. The representation of betrayed lover searching for a means of revenge cites the Medea of Euripides’s tragedy. Ovid’s engagement with his literary predecessors in the genre of tragedy highlights the dramatic performance of a role such as that of incestuous stepmother or murderous wife. His engagement with tragic paradigms also draws attention to the performative nature of gender and kinship roles, identities which must be repeated, for each citation is made under different circumstances (each individual production of the play) and by different subjects (each individual actor playing Phaedra or Medea).

Tragic Passions

The following chapters consider the effect of such citations of Phaedra and Medea in Ovid’s pre-exilic works. In chapter 1, I begin my study with Ovid’s erotodidactic poem, Ars Amatoria, where the myths of Phaedra, Medea, and similar heroines are distilled into examples that function like words in the megatext of myth. Their semiotic value is cited by the poet-praeceptor in instructive catalogues meant to provide lessons for his pupils. Chapter 2 explores the effect of Phaedra-like female narrating voices on the same megatext of myth which the poet-praeceptor speaks (Phaedra, Her. 4, and Byblis, Met. 9). Each heroine struggles to redefine her meaning as sign by shifting her citation into a new generic context. In chapter 3, I demonstrate how Medea-like female speakers, by contrast, embrace their traditional exemplary function and exploit the destruction it cites (Hypsipyle, Her. 6, and Procne, Met. 6).

Ovid’s representations of Phaedra- and Medea-like figures are creative products, meditations on gender, genre, and power in Augustan Rome. Nonetheless, his gendered discourses repeatedly reproduce, in their deep structure, gender hierarchies which justify the subordination of female bodies in the Roman world. Ovid’s play with the Augustan discourses of desire and femininity in these passages paradoxically reinvests cultural norms with authority through their very repetition. Before his exile, however, Ovid’s own, albeit hesitant, alignment with these marginal feminine voices attributes to poetry the power to influence Roman norms of gender and genre as he represents his heroines trying to manipulate the symbolic economy of myth and redefine their semantic value despite the weight of their literary tradition. Phaedra’s own Attic tragic heritage provides a precedent for the poet’s ability to redefine the paradigmatic value of a mythological character. Barrett notes that Euripides’s two Hippolytus plays are unusual among Attic tragic playwrights.⁷⁴ “He produced so radical and so successful a recasting of his original treatment that with it … he won one of the only four first prizes that he achieved throughout his career” (Barrett 1964, 13).⁷⁵

How radically distinct each of Phaedra’s tragic representations was is not clear from the extant evidence, but we may nevertheless surmise that each iteration of Phaedra by Euripides and Sophocles, like Butlerian citations, gradually modified her symbolic value. So too, we can interpret Ovid’s further iterations of Phaedra and Medea under different names in his own work as attempts to innovate in this tradition by redefining their symbolic value. Ovid’s temporary alignment with female writing subjects does not, however, represent the male poet’s identification with an abject feminine position. The poetic failure, punishment, and repudiation of female writers are constructed as the result of their “nature” as Woman, and are generically marked as “tragic.” His masculine poetic voice, by contrast, ultimately aligns itself with a controlled, rational, and authoritative ars, whose gendered and generic position enjoys a free reign to oscillate and play with poetic possibilities.