The Male Actor of Greek Tragedy: Evidence of Misogyny or Gender-Bending?

By Nancy Sorkin Rabinowitz
Hamilton College
Clinton, New York

The recent flurry of interest in cross-dressing--think of Marjorie Garber's Vested Interests, Julia Epstein and Kristina Straub's Body Guards, as well as 'Farewell My Concubine,' 'The Ballad of Little Jo,' and 'The Crying Game'--clearly amounts to something of a cultural obsession; my participation in that obsession as well as my study of Greek tragedy has led to this paper. Most theory of Greek tragedy assumes the maleness of the actor and then ignores it; I will first question the convention and then focus on it. By examining the ancient Greek convention that assigned male actors to female roles, I hope to unsettle our too-easy belief in received opinions.

But perhaps we exaggerate the significance of the convention: was imitating a woman any different for an actor of Greek tragedy than imitating an epic hero? It seems valid to assume that the transformation of male to female was significant to the ancient Greeks because they took gender differences as a framing dichotomy through which to interpret the world. And while the categories slave/free and non-Greek/Greek provided similar axes, gender is especially prominent as an overt issue in the plays' plots. Thus, although it is doubtless contemporary interest that leads me to ask these questions, they are not inappropriate ones to pose.

The psychologically realistic conventions of modern theater may also lead us to exaggerate what was required of the actor: did the actors have to look like real women? An answer must be based on the goals of tragedy, to what extent did it work as a realistic imitation and to what extent did it work as a conventionalized form. I claim that the audience of Greek tragedy responded both on what Peter J. Rabinowitz (1987) has called the narrative level, that is, taking what it saw presented as real and happening for the first time, and on what he has called the authorial level, that is, aware that they were watching a re- presentation constructed by an author. Both were important in the ancient response to tragedy, and both can be related to the maleness of the actor.

Evidence from Plato, Aristotle and Aristophanes suggests that the common belief was that the poets attempted to create lifelike imitations, thus invoking Rabinowitz's narrative audience. But tragedy was also a highly conventionalized form, with masks, chorus, elevated language, musical accompaniment and connection to ritual, all of which would draw attention to the play as a work of art and to the authorial level of audience response. Our contemporary western immersion in a realist aesthetic, where emotional response is not only crucial but also based on individualistic psychology, predisposes us to see a conflict between the imitative and the conventional ways of viewing tragedy. The two levels readily coexist, however, if with Aristotle (Poetics 15; 1454a) we take the characters as types not individuals. Thus, a slave, a barbarian, a woman were kinds of beings with appropriate behavior (conventions) to be imitated, signs to be adopted.

From this same modern aesthetic, the conventional aspect of tragedy might also seem to conflict with emotional engagement; it may be difficult for a modern western reader to accept the fact that in its own day Greek tragedy had a powerful emotional impact --whether because of or despite its non-illusionistic form. But Aristotle's emphasis on 'pity and fear' as the appropriate end for tragedy as well as Plato's fears and other ancient testimony (Gorgias, Helen 9, Diels-Kranz 82B11) indicates that the poets just as clearly strove for strong emotional reactions to their creations (Stanford 47; Bain 1-7; Taplin 1978: 169-70; 1986: 164, 170; Rehm 30).

I would argue that in Greek tragedy the cross-dressed actor was crucial, a sign of the conventional nature of the drama. Male playing female is inevitably distanced from the role and makes it clear that the role of the woman is an idealization, not a realization. In addition the male playing female indicates a possible relevance of the form to Dionysos, a god associated with masks and characterized by softness, woman's curls and dress (Bacchae) and worshipped with transvestite ceremonies at the Oschophoria (Seaford; Segal 1982: 10-20, 158- 68, 214).

These matters are not without political ramifications. It is possible for a pessimistic critic looking at tragedy to see misogyny; the male actors speak for a male playwright and a male city; tragedy is revealed to be an ideological apparatus of masculine political power predicated on the silence and invisibility of women. Sue-Ellen Case (318) states this very clearly: '`Woman' was played by male actors in drag, while actual women were banned from the stage. . . . The classical plays and theatrical conventions can now be regarded as allies in the project of suppressing actual women and replacing them with the masks of patriarchal production' (cf. Ferris 30).

A second position would take the maleness of the actor as an instance of gender-bending, and therefore see it as constituting a challenge to the fixity of gender and sex, a moment of subversion. Thus, Garber finds 'one of the most important aspects of cross-dressing . . . the way in which it offers a challenge to easy notions of binarity, putting into question the categories of `female' and `male,' whether they are considered essential or constructed, biological or cultural' (10).

Withal there is an implication that something must be either misogynist or gender bending, that the two are posed as opposites. Each of these perspectives requires modification. On the one hand, the misogynistic reading is problematic: the representation of women in tragedy has everything to do with women because it inscribes certain behavioral norms. On the other hand, Garber's position seems to lump together all instances of cross-dressing, and to assume that they are always progressive. But cross-dressing does not necessarily challenge the binary of male/female.

According to the misogynist interpretation, the form is and was conservative and kept women in their place. Plato, the first of the anti-theatrical critics, on the other hand, thought that all imitative poetry was threatening to his ideal state and ought to be outlawed for encouraging behavior in large part belonging to women and therefore inappropriate to a ruler or any male citizen: imitation leads to the reality. From Plato's perspective, theater seems to be a wild zone, a site where inappropriate actions might be represented and volatile emotions aroused. If, as Jack Winkler (1990: 50) argues, gender was on a continuum for the Greeks, and there was a risk that the normative male could by being softened end up female, the male actor of female roles might be one site of crossover or leakage.

In this paper I will explore the possibility that cross- dressing in Greek tragedy played a double role, both reinscribing patriarchy and providing a site of resistance to it. My hypothesis is that the narrative audience saw a woman on the stage; simultaneously, the authorial audience was aware that it was watching a play, saw a man playing a 'woman.' I want to imagine another, more integrated level, on which the spectators heard a middle voice, saw a third gender, so to speak, that was strictly speaking neither male or female, a figure removed from real life although referring to it, as the actor was removed from the character. The masked actor was not simply the man he was in real life, nor simply the woman he portrayed, but something new. Since the audience was not one but multiple (both narrative and authorial, that is, believing in the reality of what was enacted on stage but also aware of the conventions), tragedy could both have served the state and been subversive. I am claiming that the cross-dressed figure was emblematic of the significance of tragic imitation. If the cross-dressed actor was a sign of the non- illusionistic nature of tragedy as well as of the relationship of the form to Dionysos, then his experience must have been significant even though nobody else in the culture shared it precisely. Thus, the construction of gender and its divorce from biology was at the heart of the effect of tragedy, itself crucial to the construction of Athens.

Let me turn now to two plays of Euripides that explicitly raise questions about the desirability of women's access to language. Hippolytos on one level supports a misogynistic reading. For the narrative audience, Phaedra is at first a virtuous woman, silently starving herself rather than give in to her illicit desire for her stepson. In the end, she gives up her noble aims and instead leaves a suicide note accusing Hippolytos of rape, and to a large extent she loses her good name. Although Artemis does exculpate her, her fate is to be known for having loved Hippolytos. Hippolytos on the other hand is faithful to the vow of silence made unwittingly to Phaedra's nurse, and, although he suffers excruciating pain, he is rewarded with a cult of marriage in his name and praise from his father. To the extent that the narrative audience believed that what was represented on stage was real, it saw norms of noble silence and chastity established and perceived that women don't meet them, while the male hero does. This level of response thus reinscribed what ancient Greek culture defined as femininity: a dangerous duplicity.

The narrative audience does not, however, have the only experience that counts. Let's first look at the actor. On the one hand, to the extent that this was an imitative action, and the actor of Phaedra performed 'woman,' he experienced what it was to become female. The actor playing Phaedra acts out her pain in his body. The male actor of female roles then temporarily goes through the path the culture ascribes to women; in this case, he must viscerally experience Phaedra's shame as he represents it. On the other hand, by maintaining the conventionality of the genre, the tragic actor would have been distanced from any character, doubly so, I am hypothesizing, from his female character. That distance could have enabled him to make a critique, for his own remove from events might predispose him to recognize that it was not simply nature, as in the 'ill-favored nature of woman' (tai dustropwi gunaikwn harmoniai, 161- 62), but convention that was to blame for Phaedra's actions--just as convention drove his re-enactment of her plight.

Second, in this play, the conventional aspect of tragedy is underlined by divine prologue and passages explicitly problematizing speech; thus the role of the authorial audience was also heightened. What did it mean to the authorial audience when Hippolytos called women a counterfeit coin (616)? Do they remember the parade of the actors and think 'yes,' Phaedra is a counterfeit, only an actor, and a male one at that (Zeitlin 1985; Bergren; Bassi 1993; in progress).

In one sense, the authorial response simply bolsters the misogynist narrative reading; Hippolytos' wish that women not speak publicly or even have servants to speak for them (645-58) is granted by the convention that eliminated women's actual voices and bodies from the production. As the character Phaedra is displaced by the staged reunion between father and son, so 'Phaedra' is played by the male actor, and what seem to be her wishes (to stifle her desire, to die) are clearly cultural requirements reinforced by poet and script.

It is precisely here that I see a possibility for resistance, however: the audience which integrates these experiences, having become aware of the conventions behind the author and actor, can recognize the misogyny it accepted on the narrative level and can understand the power such conventions can exercise. If the ancient audience did not draw these conclusions, a modern audience can be moved to do so; aware of the male disguised as female, the audience can gain an understanding of the workings of our own institutions and the ways in which our desires are engendered. Phaedra's gender--her values and conception of appropriate feminine behavior--does not depend on or correlate with 'her' sex. Thus, on the authorial level, ancient or modern, cross-dressing in tragedy can make visible the construction of gender and its separability from sex.

On the narrative level, the Medea has similar effects to the Hippolytos. While the narrative audience is led to have sympathy for Medea at first, the later reaction of the chorus suggests that it would ultimately recoil at child murder. Sympathy shifts from Medea to Jason as she moves from victim to victimizer, prey to predator; this shift supports a misogynist interpretation of this play as well by making the audience turn against the wronged woman when she turns out to be strong.

The audience simultaneously responds strongly on the authorial level because of the self-referentiality of the play. For instance, when the chorus reflects on the power of poetry and the wrong done to women by male poets, they must have affected how the ancient audience responded to the events on stage. Did the authorial audience recognize that 'she' is a 'he' when Medea describes a woman's lot? Did that description then seem like a cultural prescription, since it actually came from a man, and would it have made the audience more or less sympathetic to the character? What about the scene in which Medea--having convinced Kreon to give her a reprieve of one day--abandons her obsequious posture and tells the chorus she would not have fawned on him if she did not have other intentions (368-70)? Would the ancient audience have imagined the face of a man when Medea speaks thus 'in her own voice'?

If this speech along with the other authorial elements made them aware that they had only been watching a 'woman' throughout, then her extensive and dramatic conflict over killing the children might appear a manifestation not only of a split in her character, the manly heroic (she voices the Homeric hero's code when she refuses to be made a laughingstock and addresses her warrior's soul [Burnett, Foley, Knox 1977]) and feminine maternal, but also between constructions of two genders, underlined by the split between the male actor and the female mask he wore.

How does 'the play' settle or unsettle the ancient Greek ideology of male supremacy? The audience responding on both the narrative and authorial levels might have had an uncanny sense of the man in the woman and could have taken the two together. Such a realization would not, of course, have had only one effect. It might have been reassuring to men in the audience, by underlining the fact that this was no ordinary woman asserting her power but a male actor playing a woman. Like a fetish, the male actor replaced an actual woman who would have been perceived as a threat if she acted as Medea did; the staged Medea is a phallic woman who is actually a man in costume, because 'she' has a penis, and thus 'her' power is 'his' (Garber 119-21; Ackroyd 13, 21; Tyler). On the other hand, such a male Medea might have been frightening or unsettling to the normative male audience either because she seemed to contain a mixture of traits normally taken to be separate, or because she suggested that there was a repressed man inside women in daily life; then it would seem to reveal that historical women are masked and only pretend to be weak to get what they want.

The women in the ancient audience taking Medea as real on the narrative level might have been frightened or found her despicable, because of what her behavior and situation suggested about actual female lives--for although Medea has the last laugh, she gets it only by inflicting almost as much pain on herself as on Jason. Thus, they might have feared for themselves if they thought that the only way to get revenge was to kill their own progeny, or metaphorically, to have to give up something very dear to them. On the authorial level, they would have taken her as a male actor in costume, which might have alerted them to the constructed nature of what was presented as the natural feminine role.

In a paper so full of questions, it seems appropriate to end with more questions, and particularly this difficult one: just how radical is even the most radical possibility for tragedy? The probability that women were played by men generates many possible experiences and meanings. My underlying assumption has been that the Greek audience must have responded both to the imitation and to the knowledge that it was watching an imitation; as a result the gap between biology (male actor) and culture (female character) might have been opened up on the stage. The practice was not necessarily subversive, though, for the cross-dressed man could have reinforced gender norms if the signifier was not pried away from the signified, that is if the audience failed to make the inferences I have hypothesized. Having men play women does not seem to me likely to have been liberating for actual women in antiquity, although it might have helped free some men from the need to be masculine. If the man in the female robes was experiencing something analogous to initiation, when he came back to the norm, he has been expanded as well as purified by the experience; for women, the experience of masculinity was suppressed because it was dangerous to the culture (Delcourt 5-6, 12; Loraux 1990). There is an underlying asymmetry: actors put on the feminine, and male characters enact femininity, but when female characters act like men (Cltyemnestra and even Medea), they suffer for it.

But as Oliver Taplin (1978: 171) says vigorously, 'We are now the audience of Greek tragedy.' Those of us teaching and staging these dramas today must make decisions as to what we want to highlight. And this opens up many opportunities for feminist theorists. We could choose to stress the narrative level, and reveal the misogyny behind it, so as to show its powerful place in our culture, or choose to recuperate the strength of these female characters, in a way that would be useful for modern women. Or the modern teacher, producer or actor could play up the authorial level, projecting a voice for the actor that would emphasize the gap between sex and gender, thereby making the audience aware of the possibilities for new configurations of sex and gender. Even if Brechtian distance was not maintained in antiquity, we can insert it. In the present, we can perhaps get the most critical version of the plays by taking Phaedra, Medea and Hekabe as un-natural (poetic) entities and deploying them to denaturalize our own mythologies through that explicitly art-full status.

For the contemporary audience, mindful of the possibilities and varieties of sex and gender, using male characters to play the female parts could definitely make for sensitivity to the separability of sex and gender. In each play, the audience, responding on both narrative and authorial levels, will take away different impressions of the consequences of that split--asking who makes Phaedra want what she wants and us want what we seem to want, asking who decrees that Medea must punish herself to get back at Jason. Focusing on the role of the actor may sharpen the reader's awareness of structures of domination even though by itself it won't eliminate that domination (cf. Butler 1993: 231). How far this awareness takes us depends on other factors, such as who the actual audience is and what they do with the realization; neither the director nor the teacher nor the critic can control the effect entirely, but we can try. On another level, by offering such readings of old canonical texts we make an intervention in the academy itself, on the 'authorial' level if you will. No, these texts are not universal and unchanging as the humanists still claim, but vital and open to contestation. And we members of the opposition intend to stake out our claim as well.


Ackroyd, Peter. 1979. Dressing Up: Transvestism and Drag: The History of an Obsession. Simon and Schuster.

Bain, David. 1977. Actors and Audience: A Study of Asides and Related Conventions in Greek Drama. Oxford: Oxford University Press.

Bassi, Karen. 1993. 'Helen and the Discourse of Denial in Stesichorus' Palinode.' Arethusa 26: 51-75.

Bassi, Karen. 'Gendered Histrionics: Essays on Greek Theatrical Culture.' Forthcoming from Michigan.

Bergren, Ann. 1983. 'Language and the Female in Early Greek Thought.' Arethusa 16: 69-95.

Burnett, Anne. 1973. 'Medea and the Tragedy of Revenge.' Classical Philology 68: 1-24.

Butler, Judith. 1993. Bodies that Matter: On the Discursive Limits of 'Sex.' New York and London: Routledge.

Case, Sue-Ellen. 1985. 'Classic Drag: The Greek Creation of Female Parts.' Theater Journal 37: 317-27.

Delcourt, Marie. 1958. Hermaphrodite: Mythes et rites de la bisexualite dans l'antiquite classique. Paris: Presses universitaires de France.

Epstein, Julia and Kristina Straub, eds. 1991. Body Guards: The Cultural Politics of Gender Ambiguity. New York: Routledge.

Ferris, Lesley. 1989. Acting Women: Images of Women in Theatre. New York: New York University Press.

Foley, Helene. 1989. 'Medea's Divided Self,' Classical Antiquity 8: 61-85.

Garber, Marjorie. 1992. Vested Interests: Cross Dressing and Cultural Anxiety. New York: Routledge.

Knox, Bernard. 1977. 'The Medea of Euripides.' Yale Classical Studies 25: 193-225.

Loraux, Nicole. 1990. 'Herakles: The Super-male and the Feminine.' In David M. Halperin, John J. Winkler, Froma I. Zeitlin, eds., Before Sexuality: The Construction of Erotic Experience in the Ancient Greek World, 21- 52. Princeton: Princeton University Press, 1990.

Rabinowitz, Peter J. 1987. Before Reading: Narrative Conventions and the Politics of Interpretation. Ithaca, N.Y.: Cornell University Press.

Rehm, Rush. 1992. Greek Tragic Theatre. London and New York: Routledge.

Seaford, Richard. 1981. 'Dionysiac Drama and the Dionysiac Mysteries.' Classical Quarterly 31: 252-75.

Segal, Charles. 1982. Dionysiac Poetics and Euripides' 'Bacchae'. Princeton: Princeton University Press.

Stanford, W. B. 1983. Greek Tragedy and the Emotions: An Introductory Study. London: Routledge & Kegan Paul

Taplin, Oliver. 1978. Greek Tragedy in Action. Berkeley: University of California Press.

Taplin, Oliver. 1986. 'Fifth-Century Tragedy and Comedy: A Synkrisis.' Journal of Hellenic Studies 106: 163-74.

Tyler, Carol-Ann. 1991. 'Boys Will Be Girls: The Politics of Gay Drag.' In Diana Fuss, ed. Inside/Out: Lesbian Theories, Gay Theories, 32-70. New York: Routledge, 1991.

Winkler, John J. 1990. The Constraints of Desire: The Anthropology of Sex and Gender in Ancient Greece. New York: Routledge.

Winkler, John J. and Froma I. Zeitlin, eds. 1990. Nothing to Do with Dionysos: Athenian Drama in Its Social Context. Princeton: Princeton University Press.

Zeitlin, Froma. 1985. 'Playing the Other: Theater, Theatricality, and the Feminine in Greek Drama.' Representations 11: 63-94. Reprinted in Winkler and Zeitlin 1990.

Nancy Sorkin Rabinowitz
Hamilton College