Chapter One | Chapter Three

Revenge in London: Kyd's Spanish Tragedy

Over the course of the sixteenth century, the cultural work done by English drama began to shift. Mystery plays and morality plays did not die out, but they were joined, and by the end of the century outnumbered, by plays on secular themes performed, not as part of a civic ritual, but purely for entertainment. Plays could tell different kinds of stories, and as drama became more emancipated from the liturgical calendar, the space for new plays became much more generous. Thus, playwrights began to look for new stories to tell and new ways to tell them; it was only natural that they would look first and most enduringly to Seneca, the only Roman tragedist whose works have survived, and the classical tragedist most readily available to sixteenth-century English authors. His dramatic tricks and, more importantly, his tragic sensibility had a profound effect on the plays of the English Renaissance and most especially on the revenge tragedies that are the era's most famous and controversial theatrical genre-child.

Before examining the manifestation of Seneca in Thomas Kyd's Spanish Tragedy, I wish to make the boundaries of my argument clear. Renaissance revenge tragedy is heavily influenced by Seneca, but it is decidedly a genre to itself. It applies Seneca to its own generic and cultural needs, so that while the line of descent is very clear, so are the ways in which it is self-invented. One thing which sets Renaissance tragedy apart from classical tragedy is the later period's love of structural complexity. Seneca's plays, like those of his Greek models, follow a very simple narrative arc: one action, one plot, one catastrophe. The English playwrights of the early modern period have a passion for polyphony; the stories they tell are far more complicated and involve far more characters. This intricacy illuminates another way in which the narrative drive toward destruction proliferates as the genre moves from Rome to London. In Seneca, as in the Greek tragedists, we get one downfall, one isolated destruction. In the English Renaissance, plays often interweave two or more tragic arcs as, for example, The Revengers Tragedy does. Even in plays in which the plots do not proliferate, the number of characters is larger and more of them end up dead. Renaissance tragedy no longer often goes to the extravagant extremes of hubris and apocalypse found in Seneca--although there are certainly individual exceptions--but it casts its nets more widely.

Revenge tragedies also employ more complicated storytelling devices. They adore plays-within-plays and other metatheatrical conceits; Kay Stockholder makes the connection between "the play-within-the-play, which tended to erode the boundaries between the stage and the audience, and the ghost, which merges the past into the present, and the personal into the public" (Stockholder 95). Category violation is again the crucial mechanism of the drama. Aside from providing spectacle, plays-within-plays are places where the line between fiction and reality gets smudged and redrawn, places where the audience's certainties can be ambushed and upset. We are, after all, watching ourselves in microcosm. Both interior plays and ghosts operate to erode distinctions between categories whose stability underlie and reinforce the stability of the society from which the plays emerge.

This same problem, the erosion of ideally-distinct categories, is strikingly visible in the sine qua non of revenge tragedy: the revenger. Revenge tragedy is obsessed by the problem of the heroism of its protagonist. The Spanish Tragedy, like the revenge tragedies which follow it, deploys a deliberate ambiguity around the figure of its revenger protagonist, a figure who is sometimes a hero, sometimes a villain, and most often both at once.

Another important characteristic--and one not drawn from Seneca--is the use of spectacle. In Renaissance theater, the horrifying becomes as strongly based in the visual as in the verbal. When the messenger in Thyestes describes the grove at the center of Atreus's palace, the horror is generated by what he says; when Marcus in Titus Andronicus describes his raped and mutilated niece, the horror is generated by the juxtaposition of his words with the physical, visible presence of Lavinia on the stage. Another example would be the ceremonial procession of ghosts at the end of Richard III, where the play's obsession with ritual carries over into the spirit world.

Leaving aside Octavia, which I have argued is in important ways not Senecan, ghosts in Seneca are part of the dramatic structure. Tantalus and Thyestes do not haunt the characters of Thyestes and Agamemnon; they appear to the audience to explain to us how the forces of the past act on the living characters whose tragedy we are about to watch unfold. The ghost of Achilles, which behaves in a much more ghostlike fashion in Troades (i.e., he is actually haunting the Greeks), is a metonym for the causeless malevolence of the universe which Seneca sees to be part of human existence. And he does not appear on stage.

Ghosts in the Renaissance, both fictional and nonfictional, come out of a different understanding of the afterlife, a different understanding of the relationship between the living and the dead. As with ghosts in Seneca, ghosts in Renaissance plays are embodiments of the past (literal embodiments, since they are represented by the physical, material bodies of living actors). However, while Senecan ghosts have a structural dramatic function, because they do not interact with the living characters, they do not raise the questions of community and identity that surround early modern ghosts. Early modern dramatic ghosts are puzzles for the living characters, as are the ghosts in the pamphlets I discussed in the Introduction, but the riddles they set are not community-centered. These dramatic ghosts are never explicitly damned souls,1 but they are associated strongly with the revenger's alienation and eventual downfall, and with the destruction he wreaks along the way.

The vogue for revenge tragedy in England begins with Kyd's The Spanish Tragedy. Kyd is very clearly and deliberately working with the conventions of Senecan tragedy; he quotes Seneca and, as Prosser points out, uses Hieronimo's misreadings of the passages to signal Hieronimo's irrecuperable commitment to revenge (Prosser 50).2 The Spanish Tragedy is not, however, merely a pastiche. Kyd is practicing aemulatio in the best Renaissance sense of the word, not imitation but outdoing. He uses Seneca's formal tricks such as stichomythia and the Chorus--but he adulterates them freely with other classical influences (Hill 144-45). Kyd invokes the classical authors as he pleases, and the uses to which he puts his sources are distinctively his own. Kyd is not writing from a mythic/epic exemplar; his characters are not legendary heroes and demigods. Horror is evoked not by the monumental nature of his characters but by the intensity of the tragedy Kyd brings forth on the stage.

The ghost in The Spanish Tragedy exemplifies Kyd's adaptation of classical models to his own ends. Kyd's Andrea is a ghost very much like Seneca's Tantalus and Thyestes. Like them, he appears at the beginning of the play; like them, he functions metatheatrically as a prologue. Kyd outdoes his model by making Andrea the Chorus as well. Like Tantalus, Andrea is accompanied by a figure from the underworld, although the unabashedly allegorical Revenge is working with Andrea rather than against him.

In other ways, however, Andrea fails to function as a ghost, the principal one being that he, in himself, is not a source of horror. He is not visible or tangible to any of the characters in the play. They cannot see or sense him, and therefore one important context for horror--their reactions--is absent. Looking ahead to Hamlet for a moment, the ghost in that play is introduced to the audience through the fearful reactions of Marcellus, Francisco, and Barnardo; Andrea introduces himself to us directly. Andrea also has a degree of interiority lacking in other ghosts in the period; he reacts to what he sees and expresses feelings and thoughts. Rather than a ghost per se, he might better be described as a character who just happens to be dead.

Andrea's status as a dead person affects his function in the play significantly, but he does not function as a ghost. Hill contends, close-reading the first ten lines of the play, that the Elizabethans would not even have necessarily found him sympathetic (Hill 148). I do not accept Hill's argument whole-heartedly, but his argument emphasizes the human-ness of Andrea, the way in which he can be read purely as a character, rather than as a spectral phantasm. Recognizing the ways in which Andrea is heimlich enables us to see that the unheimlich in the play does not come from the simple existence of a ghost. Andrea is part of the play's production of horror, but he is not its sole cause.

The nature of horror in the play is troped by the metatheatrical nature of its Chorus, the conglomeration in them of this archaic and awkward theatrical convention and their intense relationship with the play's action. Andrea and Revenge are not a Chorus in the standard use of the word, a group of people interested but only peripherally involved in the action of the tragedy; their position is more like that of the monarch at a masque: a privileged audience for whose pleasure this particular story is being told:

ANDREA. ... fair Proserpine began to smile,
    And begg'd that only she might give my doom.
    Pluto was pleas'd, and seal'd it with a kiss.
    Forthwith, Revenge, she rounded thee in th'ear,
    And bade thee lead me through the gates of horn,
    Where dreams have passage in the silent night.
    No sooner had she spoke but we were here,
    I wot not how, in twinkling of an eye.
REVENGE. Then know, Andrea, that thou art arriv'd
    Where thou shalt see the author of thy death,
    Don Balthazar, the Prince of Portingale,
    Depriv'd of life by Bel-Imperia.
    Here we sit down to see the mystery,
    And serve for Chorus in this tragedy.
(ST 1.1.78-91)

Andrea is both deeply invested, as a character, and detached, as a ghost, from the action of the play. Peter Sacks, who remarks on the framing of the play as Andrea's dream--"Forthwith, Revenge, she rounded thee in th'ear, / And bade thee lead me through the gates of horn, / Where dreams have passage in the silent night" (ST Pro. 81-83)--observes that Revenge and Andrea are "little more than versions of the playwright and his audience" (Sacks 585). Andrea is ghost, chorus, audience, and dreamer.

Andrea's overdetermined multiplicity of roles makes him both a part of the action and an impotent witness, as his horrified and angry reaction to the events of Act III shows:

Awake, Revenge! if love--as love hath had--
Have yet the power or prevalence in hell!
Hieronimo with Lorenzo is join'd in league,
And intercepts our passage to revenge.
Awake, Revenge, or we are woebegone!
(ST 3.15.12-16)

Here Andrea is neither a part of the play proper--his distress comes from what he, like the theater audience, has witnessed on stage--nor a part of the audience, since he demands that the course of events he anticipates be averted. He exists in the doorway, neither entirely actor nor entirely audience, metatextually calling the audience's attention to the fictionality of what they are watching. And metatheatricality in The Spanish Tragedy is not merely a device; it is the mechanism by which the play moves from pity to horror.

Kyd begins this movement by allowing the audience to pity Hieronimo. The scenes of his madness (and Isabella's suicide) are scenes of pathos, where we watch an old man grieving for his son. The alternate title for the play, Hieronimo is Mad Again, suggests that these scenes were one cause of the play's amazing and enduring popularity. But Kyd is not writing a sentimental melodrama; unlike Nahum Tate, he is not content to let his audience stay comfortably distanced from the suffering of the characters.

Kyd uses the audience's natural desire to admire and sympathize with the protagonists to put them in the wrong. Prosser points out that this maneuver was popular among Renaissance playwrights, citing Bussy D'Ambois and Tamburlaine (Prosser 35). Kyd, like Marlowe and Chapman, lures the audience into sympathizing with stances they cannot condone. Hieronimo and Bel-imperia are admirable and sympathetic; Lorenzo and Balthazar are hateful, selfish and immoral. We want Hieronimo's revenge to be successful, because we like the murderer and hate his victims. The moment when Hieronimo murders Castile forces an abrupt and violent reappraisal by the audience of its own sympathies and moral stance. We discover that we have been rooting for a crazed murderer, that we are culpable in Hieronimo's crimes. The backwash from the play's central moment of metatheatrical collapse stains the audience's hands blood-red.

This metatheatrical collapse is achieved through the use of revenge tragedy's favorite metatheatrical technique: the play-within-a-play. Hieronimo's play is itself a revenge tragedy: Perseda kills Soliman in revenge for the death of her husband Erastus, who was killed by one of Soliman's bashaws so that Soliman might woo Perseda (ST 4.1.106-24). Even in its flat, harmless rendition, this plot has strong parallels with the main plot of The Spanish Tragedy, although Hieronimo distorts that mapping, so that all of the characters who are villains in the main play conveniently become victims in the interior play. McMillin points out that Hieronimo himself takes the place of the villainous Lorenzo (McMillin 46). As with Hamlet, Hieronimo must become the thing he hates in order to succeed in his revenge.

The play uses its own spectacle to disguise the truth of its purpose; real deaths are camouflaged as theatrical deaths. The methodology of the end of the play is foreshadowed in the fate of Pedringano, who is similarly ensnared and killed by a metatheatrical trap. Pedringano believes his execution is merely for show, whereas the truth is that he is both audience and victim of a nested fiction, one which by its very fictionality converts Pedringano's own fiction into nonfiction. This interplay between stories, this shifting back and forth between the real and the fictional--all within the fiction of the play--emphasizes, as Hieronimo's play will, the frightening fluidity of the boundary between reality and make-believe. As Shapiro puts it, Pedringano's death reveals a disturbing truth about the theatrical representation of death: "neither the actor to be executed nor the spectators who witness the execution can be entirely sure that the violence is not real" (Shapiro 103). We are taught this principle by Pedringano's death; we see it put into action in Hieronimo's play, in which neither actors nor audience can know if death is real or fictional, as the fictional roles violate the category of fiction to become truth.

The on-stage audience to Hieronimo's play (the King, the Viceroy, the Duke), unaware that the play is anything more than a play, are very pleased, but they have the natural reaction of any audience:

KING. Well said!--Old Marshal, that was bravely done!
HIERONIMO. But Bel-imperia plays Perseda well!
VICEROY. Were this in earnest, Bel-imperia,
    You would be better to my son than so.
KING. But now what follows for Hieronimo?
(ST 4.4.68-72)

The King's confusion in the last line is telling. He means, What happens to the murderous bashaw?, but he has conflated Hieronimo with his role in a way that exactly though unknowingly reflects the way Hieronimo has conflated all of the actors with their roles. This conflation is also echoed, with gruesome double-entendre, in Hieronimo's line (4.4.69 above), where Bel-imperia "plays Perseda well" because nobody can play a dead person better than a corpse.

The King, the Viceroy, and the Duke watching Hieronimo's play represent the audience to itself (as Hill puts it: "The King's fatuous misreading of the dumb show serves as a warning to Kyd's audience" [Hill 160]), and so when this conflation is revealed, the reality of the seeming fiction echoes back at the theater audience. Castile in particular, having had no part in Lorenzo and Balthazar's schemes, represents the victimization of the audience. Like the audience in the theater, he has done nothing more blameworthy than coming to watch a play. Hieronimo's murder of Castile is thus an attack on the audience in the theater, suggesting that playwatching, in and of itself, is neither innocent nor safe.

Castile's death also emblematizes the impossibility of stopping the play's downward spiral into darkness. A trope of revenge tragedy is the way in which death proliferates with oxymoronic fertility, and the end of The Spanish Tragedy uses that trope to its full extent. Hieronimo's engine of revenge has become an engine of destruction. Horatio's death has led to Isabella's death and Pedringano's death; now Hieronimo's revenge for Horatio's death has slain Balthazar, Lorenzo, and Bel-imperia. In Bel-imperia's death, crucially, the collapse between fiction and reality has escaped Hieronimo's control; Bel-imperia carried through the story despite Hieronimo's wish that she should not. Hieronimo tries to enact closure by killing himself, but is stopped by the interference of his audience. Their action, rather than closing the chain, leads to the death of Castile and the death of Hieronimo, which only then leads us, finally, to the end of the play. It is peculiar and disturbing that this level of violence is required to maneuver the story into a position from which it can reach closure, and it is from this bewildering but inescapable violence that The Spanish Tragedy blossoms forth from pathos into horror.

This horror continues through the last lines of the play, as Kyd questions and deconstructs the audience's sympathies, this time forcing a reappraisal of Andrea, who has been the audience's surrogate throughout. In the play's final scene, Revenge asks Andrea if he is satisfied. Having outlined the pleasures awaiting the "good" characters in the Elysian fields, Andrea asks Revenge what will become of the villains. Revenge says they will be dragged to Hell, whereupon Andrea says:

Then, sweet Revenge, do this at my request:
Let me be judge, and doom them to unrest.
Let loose poor Tityus from the vulture's gripe,
And let Don Cyprian supply his room;
Place Don Lorenzo on Ixion's wheel,
And let the lovers' endless pains surcease
(Juno forgets old wrath and grants him ease);
Hang Balthazar about Chimaera's neck,
And let him there bewail his bloody love,
Repining at our joys that are above;
Let Serberine go roll the fatal stone,
And take from Sisyphus his endless moan;
False Pedringano, for his treachery,
Let him be dragg'd through boiling Acheron,
And there live, dying still in endless flames,
Blaspheming gods and all their holy names.
(ST 4.5.29-44)

The audience may very well be taken aback by Andrea's bloodthirstiness. We have been sympathetic to Andrea throughout, and now we discover that the classical trappings of Kyd's underworld are not mere trappings, but significant pointers for reading Andrea's character. There is no Christian rhetoric of forgiveness here. Moreover, the self-aggrandizement with which he asserts the crimes of his enemies--especially, again, the blameless Don Cyprian, Duke of Castile--are worse than those of the archetypal classical villains may perhaps make us nervous.

Kyd takes the proto-genre of revenge tragedy from Seneca and mutates it to match the idiom of sixteenth-century English theater. At the end of the play Kyd remaps what he has made back onto his exemplar and shows us, very clearly, that they no longer fit. We are reminded again that Andrea is not English, not familiar, not perhaps someone with whom we should be sympathizing. Kyd ends the play by alienating the audience from the two heroes, and via this gesture, he alienates the audience from itself.

This final alienation is one source of horror in the play, quiet psychological horror rather than overt gory horror, but horror nevertheless. The Spanish Tragedy does not produce horror as Seneca did, by confronting its audience with over-the-top monstrosity; instead it works to undermine audience certainties about the difference between reality and fiction, right and wrong.

Significantly, too, the play's ending is not an ending that produces closure. Just as Hieronimo's domino chain of violence seems unstoppable, so too the vengeful energies of the play extend beyond its ending. The King of Spain and the Viceroy of Portugal exit mourning the destruction of both their heirs in much the same way that Edgar and Albany have to trail dismally off the stage at the end of King Lear, but even that lame and limping closure is not the end of the play. We go on to the coda between Revenge and Andrea in which Andrea expresses his delight in the unending torments awaiting his enemies, and the last lines of the play turn the entire enterprise inside out. Revenge says: "For here though death hath end their [the villains'] misery. / I'll there begin their endless tragedy" (ST 4.5.47-48). The ending is not the ending at all; it is the beginning of a tragedy without end. Kyd makes no effort to return the society represented on his stage to the state of order and normalcy it occupied, however tenuously, at the beginning of the play. The destruction and death begun by Hieronimo's revenge have entered the endless feedback loop of the underworld, and they will not stop. This first English revenge tragedy is as relentlessly nihilistic as any play that will follow it; Kyd both creates and masters his genre.


1. Stephen Greenblatt discusses the ambiguities of the Ghost's status in Hamlet in Hamlet in Purgatory (Greenblatt throughout).

2. Other critics, most notably Daalder, read this scene very differently, but in this instance I agree with Prosser: the use of Seneca is signaling Hieronimo's progress along the path toward bloody and horrific tragedy.

Chapter One | Chapter Three

© Sarah Monette 2004     Feel free to link to this academic trial-by-ordeal, but please do not reproduce it without permission.