prefatory material | Chapter One


Any group which participates in the telling of a story (and that group must definitionally include both audience and speaker) creates an ad hoc community; while that specific community may not outlive the story itself, story-telling can also be a way of celebrating, defining, and strengthening community. Drama is a form of story-telling which is inherently communal; the actors form one community, the audience another, and the fact of performance brings these two communities together into a transitory but powerful group experience. Thus, drama is particularly well-suited to fictions which can represent a community to itself, and this characteristic is one of which early modern English drama made full use. The close connection between drama and community is visible even at the beginning of the secularization of drama in England; the Corpus Christi pageant-wagons were a way for guilds to perform their social identities, thus a way for their small communities to reaffirm both their value to the larger social construct of the towns and their strength as communities to themselves.

Drama can be more than affirmation of community, however. It can also provide a safe arena for discussion of a community's flaws and concerns. Stories about community do cultural work, whether they talk about divisions within communities or their borders: the interface between a community and the world around it, geographically, culturally, temporally--or, in microcosm, the interface between a community and an individual within it. Stories about the boundaries of community are also frequently stories about the destruction or failure of community, stories in which the pressures of the world outside exert a greater force than the community can withstand, or the single individual either destroys or is destroyed by the rules his or her community has created in order to assure its survival. It is no wonder that stories such as these are often stories which seek to evoke horror in their audience.

The revenge tragedies of early modern England are one such set of stories. Early modern revenge tragedy is both an art and an artifact of its time. It tells us certain things about its culture and the work its culture needs its stories to do. But at the same time, as an art-form, it follows its own inclinations. One cannot reason backwards from the communities depicted on stage to the cultural reality of the community which first watched them--except in certain careful, limited ways. There is a constant dialogic tension in understanding these plays. On the one hand is their own artifice, and the latitude that they enjoy as works of entertainment. Early modern drama uses its own metatheatricality to defuse the overt threat of its story or themes; one example would be the discussion of the London theatrical scene between Hamlet, Rosencrantz, and Guildenstern in the second act of Hamlet. By calling attention to their own fictionality, these plays make themselves look harmless. But on the other hand, they undeniably do cultural work of their own. Their cultural work is done through and by horror. They exaggerate in the direction of destruction and despair, considering the boundary of community in yet another way: what happens when the community has been broken? Revenge tragedy enacts and embodies its culture's fears about what happens when an individual and his or her community are set against each other, also about the repercussions of a failure between a community and its own past.

The most cogent example of this metonymic process of the embodiment of cultural fears on the theatrical stage is the figure of the ghost. Ghosts are very much about the boundaries of community, about where and how those boundaries are drawn. In a play, ghosts are characters who are both inside and outside the community; when they were alive, they were members of the community, but in death they are no longer. They are the epitome of the dangerous and uncontrollable world that exists outside a community's self-drawn boundaries. As victims, moreover, they represent a community's failure to protect itself. The social construct of the community is besieged on all sides, and ghosts signal ruptures in its defenses.

They trope that boundary by violating another: the essential epistemological boundary between the quick and the dead. Also, on the Renaissance stage, ghosts are a marker of another kind of community, that of the literary tradition by which the English playwrights traced the descent of their enterprise back to Lucius Annaeus Seneca. Thus, both intra-, inter-, and extra-literarily, ghosts and the actors who play them embody the weight of the past influencing the present.

The place where ghost meets genre is the idea of category violation. Both the concept of the ghost and the generic demands and effectiveness of horror are predicated on category violation. In the chapter "The Abominations of Leviticus" in Purity and Danger, Mary Douglas chases the concepts of holiness and pollution through the prohibitions of Leviticus and Deuteronomy and concludes that among the requirements for holiness is what she calls "completeness": "Holiness requires that individuals shall conform to the class to which they belong. And holiness requires that different classes of things shall not be confused" (Douglas 54). Categories must remain inviolate; things must be either fish or fowl. Hybrids are bad, unclean--abominable. Thus, the site of category violation is an indication of what a culture fears, an indication of where it is vulnerable to the power of story-telling.

Douglas's interests are anthropological, and her book explores the creation and use of order in religion. She explores the ways in which human societies categorize things, specifically the holy and the profane, in order to create social, religious and cosmological structures, resulting not in fear but in "positive effort[s] to organise the environment" (Douglas 2). I am considering the flip-side of her argument, taking the idea of categories as structures which impose order on the environment and using it to look at the horror and existential chaos which ensue when those categories break down.

The anthropological and epistemological concept of category violation is closely tied to the psychoanalytic concept of the abject. I follow Julia Kristeva's definition of the term, which allows me to use the abject as a mediating point between category violation and the workings of repression. The abject is that which will not stay repressed, and it is also that which emerges at the site of, and out of, category violation. Kristeva points out that boundaries which cannot be breached lead, in psychological terms, to psychosis, as "the constituting barrier between subject and object has ... become an unsurmountable wall" (Kristeva 47). Thus "category violation," as a concept, is itself paradoxical, that which we both fear and need; things which violate categories are abjected, but so is the idea of category violation itself.

As I use the concept in my argument, category violation can work in one of two ways. Either two things which the human mind constructs as separate, such as "good" and "evil," fail to stay distinct from each other, or one thing which the human mind constructs as a whole, such as the human body, fails to maintain its integrity.

Category violation can be either a source of pleasure or a source of intense discomfort; the extent to which it is one or the other depends on the degree to which it is contained, either by the fiction around it, or by the social and ideological pressures that shape our reactions. Thus drag queens, who violate the boundaries of socially-constructed gender, are cheered and applauded by some and feared by others. Our reaction to the phenomenon of category violation is largely governed by the degree to which the violation in question affects our own personal cosmology.

Category violations in horror frighten and disturb us because, through the specific vehicle of a ghost or mutant or Frankenstein-esque hybrid, they destabilize much broader matters of cosmology and epistemology. Faced with too many ruptures, humans react like chameleons placed on plaid, unable to process the world in any meaningful way. D. J. Palmer points out that in Titus Andronicus ritual is the response to the unspeakable and unbearable "because ritual is the ultimate means by which man seeks to order and control his precarious and unstable world" (Palmer 322). But the terrible thing about rituals in Titus Andronicus, as in other revenge tragedies, is that they do not help. Human beings crave order; revenge tragedy examines what happens--to society, to the individual--when that order is taken away. That examination is enabled and modulated by the relationship and the differences between ritual and theater.

My theorizing of the relationship between ritual and theater comes from Turner's in From Ritual to Theatre: The Human Seriousness of Play. Turner postulates that the most basic mode humans have for dealing with conflict is what he calls "the social drama" (Turner 11), which he categorizes as having four stages: breach, crisis, redress, and either reconciliation or "consensual recognition of irremediable breach" (Turner 92). He argues that this formal pattern is the basis for the structuring of rituals as the means by which a community reconciles itself to changes both internal and external. These rituals involve the use of liminal space, a space of ambiguity, a "sort of social limbo" (Turner 24) in which the rules and characteristics of ordinary social interaction do not apply. Liminality is a space for a community to consider its own history. Although this examination takes place through inversion, negation, and other techniques of disordering the orderly, liminal experiences are "ultimately eufunctional" (Turner 54). Liminality works under the "release valve" theory of subversion and containment. Its chaotic and seemingly anarchic nature is in fact a reinforcement of the rules and structures of the customary social space.

The liminoid, on the other hand, while it can be broadly described as a secular version of the liminal, is in fact much more complicated and valuable than that. The liminoid, argues Turner, is the province of artists; moreover, while the liminal is a central and fully integrated part of the society's functioning, the liminoid is not, and its expressions develop "along the margins, in the interfaces and interstices of central and servicing institutions--they are plural, fragmentary, and experimental in character" (Turner 54). Even leaving aside the questions raised by the plays themselves, Turner's words describe the situation of the Elizabethan playhouses to perfection. The liminal is respectable; the liminoid is risky.

Unlike the liminal, which depends on the structures of ritual, the liminoid is free-form. Although liminoid experiences can be collective, they are mostly produced by individuals. The liminoid does not fit into a subversion-containment model; it frequently goes hand in hand with social critique and even revolution. It is, most importantly, an arena for play. Kristeva, in identifying modern literature as a place for the jouissance of the abject, describes it as "a substitute for the role formerly played by the sacred, at the limits of social and subjective identity" (Kristeva 26). Her concept maps onto Turner's without impediment or flaw. The abject belongs to the liminoid and vice versa.

Theater, which comes from the liminal space of ancient Greek religion, is liminoid in nature. It is a space of play, in all senses of the word, and thus is more often than not a space of transgression. Social drama, by Turner's definition, always begins with the violation of some societal standard. Therefore, the social drama, and the ritual, liminal, and liminoid experiences which stem from it, is instigated by transgression. The same definition applies to horror, as a narrative genre; horror begins with transgression, whether the transgression of a boundary, such as that between the living and the dead, or the transgression of a taboo: incest, cannibalism, infanticide. So that horror, in its way, is another modality for dealing with violations of the categories with which human beings structure our world. I would argue here that social dramas, in Turner's sense, tend to be responses to specific transgressions--i.e., the action of a person within the community--while horror tends to be a reaction to the idea of transgression. And ritual is both, the place where these two things meet and come together. The Greek tragedians wrote for the festival of Dionysus as part of a ritual celebration in honor of Dionysus's transgression of the boundary between life and death.

Theatrical ritual requires both performers and audience, and the idea of "the audience" requires some unpacking. No two persons experience any given work, whether textual or performed, in the same way. This is especially true in horror, because not all horrors affect all audience members equally. As a random example, the fear of spiders can be an intense phobia for those who suffer it, but it is hardly universal. Works of horror involving spiders, such as the movie Arachnophobia (1990); the Buffy the Vampire Slayer episode "Nightmares" (1997); Stephen King's novels IT (1985) and Needful Things (1991), or his collaboration with Peter Straub, The Talisman (1984); the horrifying arachnids in J. R. R. Tolkien's works: the spiders in Mirkwood (The Hobbit 1937), Shelob (The Two Towers 1954), and Ungoliant the Unlight (The Silmarillion 1977)--these works will have a greater effect on an arachnophobe than on an entomologist. "Nightmares" even makes use of this:

WILLOW. How do spiders not ruffle you?
XANDER. I'm sorry, I'm unruffled by spiders. Now if a bunch of Nazis crawled across my face ...
(BtVS 1.10)

Fear is a deeply subjective emotion. We cannot judge a work of horror solely on how the audience--any audience--responds to it. Tzvetan Todorov, in The Fantastic, disallows emotional effect on the reader as a marker of genre, arguing that, were one to accept that criterion, genre would depend "on the sang-froid of its reader" (Todorov 35). I would argue, however, that it is not the actuality of the reaction that marks a genre convention, but our recognition that the reaction is expected. Genre conventions are part of a dance between poet and audience. Works of horror have expectations of their audience coded into them, regardless of whether a given audience follows those cues or not.

Horror, like other kinds of fiction, is dependant on what Samuel Taylor Coleridge called "the willing suspension of disbelief," and this process is one with which many narrative theorists engage. Peter J. Rabinowitz, in Before Reading: Narrative Conventions and the Politics of Interpretation, theorizes the process as part of a negotiation of three different audiences, actual, authorial, and narrative. The actual audience is the flesh and blood human being who reads the novel or watches the play; the authorial audience is the hypothetical audience for whom the author writes. The narrative audience is the audience generated within the text, the audience for whom the fiction is reality; Rabinowitz argues that the narrative audience is intrinsic to the work of fiction and is part of author and reader's awareness of fiction as fiction. As an audience, we negotiate between these three audience positions. We both believe and disbelieve; we simultaneously know that the events we witness are fiction and that they are true. Fiction is a precarious and delicately balanced mechanism.

And the balance of this mechanism becomes even more precarious when we examine it, not only as readers, but as critics. Revenge tragedy offers a particularly cogent example of this difficulty in the question of morality, which has been exercising critics since Fredson Bowers's Elizabethan Revenge Tragedy. Bowers assumes that morality is one criterion of great art, and since he will not argue that revenge tragedy is not art, he argues instead that it is moral. His definition of "moral" is narrow, and forces him into an argument that has impeded the critics who follow him. He cannot or will not admit the darkness of revenge tragedy morality: that the protagonists are damned before they even begin. Therefore, Bowers argues that blood-revenge was understood by Elizabethan playwrights and audiences as a sacred duty.

His evidence comes from Seneca's Agamemnon; he asserts that Aegisthus is duty-bound to revenge Thyestes by murdering Agamemnon. This argument fails to take into account the out-and-out evil of Thyestes's union with his daughter--evil which the ghost of Thyestes himself explicitly admits (Thyestes 28-36). In Seneca's conception there is nothing sacred about the burden of revenge placed on Aegisthus by Thyestes. Moreover, this reading, making Aegisthus the patron saint of early modern blood revengers, ignores two qualities of the Agamemnon which Bowers himself remarks on: "Aegisthus is of lesser stature [than Atreus]; it is inevitable that the unnatural revenge of Clytemnestra should overshadow that of her paramour. Atreus, Aegisthus, and Medea--Seneca's three great revengers--are all villains" (Bowers 264). Thus, first of all, Aegisthus is not really the point of the Agamemnon, despite Thyestes's prologue (a typically Senecan bait-and-switch), and secondly, he is incontestably a villain.1 Consequently, as Bowers will argue about the protagonists of the "villain plays" of the early seventeenth century, Aegisthus's revenge has--and can have--nothing of the sacred about it. The idea of revenge as a "sacred duty" and therefore morally condonable is a chimaera, undercut as soon as posited. But has nevertheless proved to have remarkable staying power as a point of dissension.

Eleanor Prosser, in Hamlet and Revenge, takes up arms against Bowers's thesis and sets out to prove that revenge was no more acceptable to Elizabethan audiences than it is to twentieth-century critics, arguing vehemently against the idea that revenge could be a sacred duty. She finds that Elizabethan moralists condemned revenge not only as "illegal, blasphemous, immoral, irrational, unnatural, and unhealthy--not to mention unsafe" but also "thoroughly un-English" (Prosser 10). Her work with the cultural context demonstrates that the authorial audience and the actual audience are not as widely separated as Bowers imagines.

Michael Cameron Andrews takes issue with Prosser, insisting on a distinction between the judgments we make as an audience during a play and the judgments we make, as critics, afterwards. He argues that "when we talk of withdrawing moral approval from what we were seduced into accepting during the play, we are probably saying that we don't like the dramatist's ideas, not that we were insufficiently alert to the nuances of his play" (Andrews 12). Our judgment when we consider the play as an object of study should not contaminate or override our judgment when we watch the play.

But even Andrews cannot resist the temptation of trying to make the quandary resolve. In contrast to Bowers, Prosser, Andrews, and other critics who try to make revenge tragedy either moral or immoral, I would argue that one reason for the horror generated by Elizabethan and Jacobean tragedies (as for Senecan tragedy before them) is that morality has ceased to be relevant within the worlds of the plays. Elsinore, Alicant, the "radically degraded, material world of carnival" which Arthur Lindley argues is more the setting of The Revengers Tragedy than any "notional Italy" (Lindley 45): these are worlds in which morality exists--the characters are aware of it--but has no force. It does not matter whether the actions taken by the revenger are moral, immoral, or amoral; the tragedy emerges because the actions are unavoidable and because by taking them, the revenger destroys himself. Revenge tragedy does not ask us to sit in moral judgment; it asks us to empathize with people trapped in untenable situations.

Revenge tragedy is always and inherently morally ambiguous. Some critics, such as Joost Daalder, Janet Adelman, Marjorie Garber, and Peter Sacks, feel that Hamlet's killing Claudius is unproblematically the "right" thing to do. There is a difference between "right" and "inevitable"; revenge tragedy is an inexorable two-handed engine, and the characters reach a point where they have no choices about their actions. But as with the classic paradigm of Greek tragedy, the fact that the catastrophe is inevitable does not make it any less a catastrophe.

Both sides of the debate about the morality of revenge tragedy also tend to elide the fact that revenge tragedies are fictions, entertainments. Just as today, people go to see movies or read books2 about people and actions which they would never admire or condone in real life, so too in the Renaissance. Uplifting moral messages are rarely entertaining--as Hobbes observes in a late Calvin and Hobbes cartoon, "Virtue needs some cheaper thrills" (Watterson 96). We cannot, therefore, assume that the popularity of revenge tragedy and the undeniable charisma of its heroes mean that those heroes were intended or regarded as moral exemplars.

In fact, what makes those characters attractive is more often their wrong-doing than their virtue. Rosslyn argues that the villain in a revenge tragedy is in fact acting for the audience, allowing them to enjoy wickedness, chaos, and transgression without having to be responsible for the consequences. (Rosslyn 3). The pleasure an audience gets out of revenge tragedy comes precisely from watching characters who are not "heroic," who follow their Ids rather than their Superegos. Rosslyn's argument, however, does not take into account the fact that the contract she postulates takes place in a very narrowly defined, liminoid space, and that the reason there are no "consequences," insofar as one can argue there are not, is because the villain is a fiction; on some level, he and the audience both know that he is only make-believe. We cannot extrapolate from this contract to cultural mores at large, except in ways as limited and carefully defined as the space of the theater itself.

But at the same time revenge tragedy does not support or condone its characters' behavior. Stephen Booth makes what I feel to be a flawed comparison between the revenger and the hero of a Western, arguing that:

The possibility that an audience's Christian belief that vengeance belongs only to God will color its understanding of revenge in The Spanish Tragedy is as unlikely as a modern film audience's consideration of a villain's civil rights when somebody shouts, "Head him off at the pass."
("Value" 152-53)

In drawing this comparison, Booth does not fully take into account the fundamental difference between the hero of a Western and the "hero" of a revenge tragedy. The hero of a conventional western, like the detective in a conventional mystery, is an unambiguous force for good. He "takes the law into his own hands" specifically because there is no other way to achieve justice; moreover, he remains heroic throughout, and the Western ends with a celebration of his achievements. Not so the revenger. Revengers may start out morally pure, but the nature of their undertaking inevitably and unanimously corrupts them. Unlike the Western hero, the revenger cannot stay clean of the corruption he fights, and the transhistorical validity of this point is demonstrated by the fact that revengers come to sticky and horrible ends. They are not rewarded for their acts of vengeance, nor even allowed to escape. Revenge tragedy playwrights and audiences might sympathize with the revengers and even delight in their stratagems, but they did not view them as having the pre-lapsarian innocence that Booth suggests.

As Gordon Braden points out, revenge is culturally ambiguous, "both sanctioned and interdicted ... the avenger is much more fully and consciously a member of the society whose restraints he violates than is the villain hero" (Braden 113). The ambiguity is the key here, and the fact that it is an ambiguity we are not meant to be able to resolve. A revenge tragedy can take place because the playwright has meticulously jiggered his protagonist into a state where the question of morality vs. immorality can no longer be resolved without misrepresenting the play. Critics seeking to resolve the dilemma can only do so by doing violence to the very phenomenon they are trying to study.

Happily, there are critical tools at our disposal which permit discussion of ambiguity--formal as well as moral--without forcing a resolution. Raymond J. Pentzell, in "The Changeling: Notes on Mannerism in Dramatic Form," proposes the artistic school of Mannerism as a model for thinking about early modern plays: "in [these] works we recognize an extreme tension between totally manipulative artifice (the terminus ad quem of maniera) and the 'shock of recognition'" (Pentzell 24). He argues later that "Our two 'strategies,' conscious virtuosity at the possible expense of meaning and engagement, and forced engagement at the possible expense of formal coherence, become two ends of the same rubber band" (Pentzell 23-24). This language could be applied equally well in the analysis of Seneca I will make in Chapter One, and therefore from two radically different starting points, Pentzell and I come to the same conclusion regarding early modern revenge tragedy, that it is an intensely self-conscious artform, always aware of its own performance.3 These are self-aware plays which deliberately court the grotesque and horrible. Critics must refuse simplicity and transparency as determinedly as the plays do themselves.

This diagnosis of Mannerism also explains why analysis of Elizabethan and Jacobean drama benefits from the tools provided by the Brechtian model of theater. Raymond Williams, in Drama in Performance, points out the inadequacy of the proscenium stage and its associated mindset for Brecht's non-naturalistic concept of theater (Williams 132), and naturalism is as misleading a lens for Renaissance revenge tragedy as it is for Brechtian drama. Though the dramatists were not using distancing effects with the highly theoretical self-awareness of Brecht, they nevertheless used them, and used them consciously.4 The frame of naturalism, of Aristotelian mimesis, is a set of blinkers as far as understanding early modern drama is concerned.

Naturalism is not synonymous with effectiveness in fiction. The audience does not need strict verisimilitude in order to identify with characters or to be moved by their stories. In fact, the deeper workings of the human consciousness do not always give credence to naturalism, as anyone may notice by considering their own dreams, or especially their nightmares. Horror can be readily generated by things that audiences know perfectly well are not "real." The logic of horror is much more closely akin to the logic of dreams than it is to Newtonian or even Einsteinian physics. The rational mind is not the one in charge.

Freud, for all his failings, observed very keenly the mechanics by which the repressed (or abjected) reasserted itself to the conscious self, and the reasons that that reassertion become frightening, unsetttling, unheimlich. He defines the unheimlich as "that class of the frightening which leads back to what is known of old and long familiar" (Freud 220), and moreover says of the word itself: "heimlich is a word the meaning of which develops in the direction of ambivalence, until it finally coincides with its opposite, unheimlich. Unheimlich is in some way or other a sub-species of heimlich" (Freud 226). Here, too, we find category violation to be intrinsic to the ideas under discussion; the boundary transgressed is that between the familiar and the unfamiliar, the safe and the threatening.

Freud's analysis of the unheimlich remarks on the uncanny properties of doubles and doubling--a thematic obsession, as we will see, of revenge tragedy--and the "compulsion to repeat"5 (Freud 238). He identifies the unheimlich fundamentally as "something repressed which recurs" (Freud 241, emphasis in original). Ghosts fit this pattern exactly. Although Freud claims that Shakespeare's ghosts are not uncanny (Freud 250), his grounds for doing so depend on a particularly idiosyncratic definition of what can and cannot cause feelings of horror in a reader of fiction.6 Without wishing to engage with Freud's particular categorizations, I will here try to define what I mean by "horror."

Horror is a category of fear, generatable by both real and fictional events. My interest is limited to the moments at which something patently and demonstrably "unreal" can generate, even momentarily, that sensation of fear and shock and revulsion. So "horror," in the context of my argument, is the emotion generated by watching or reading, in a work which we know to be fiction, something which we do not want to see or to know--and yet at the same time an emotion we have picked up the book, gone to the play, watched the film, expressly in order to experience. Nol Carroll, one of the few critics to attempt a rigorous theoretical analysis of the genre of horror, categorizes "art-horror" as "a genre that crystallized, speaking very roughly, around the time of the publication of Frankenstein" (Carroll 13). He defines his term as follows:

Assuming that "I-as-audience-member" am in an analogous emotional state to that which fictional characters beset by monsters are described to be in, then: I am occurently art-horrified by some monster X, say Dracula, if and only if 1) I am in some state of abnormal, physically felt agitation (shuddering, tingling, screaming, etc.) which 2) has been caused by a) the thought: that Dracula is a possible being; and by the evaluative thoughts: that b) said Dracula has the property of being physically (and perhaps morally and socially) threatening in the ways portrayed in the fiction and that c) said Dracula has the property of being impure, where 3) such thoughts are usually accompanied by the desire to avoid the touch of things like Dracula.
(Carroll 27)

Carroll's definition reads like a legal contract, and it is not one with which I entirely agree, although I think much of what it says is very helpful, especially in identifying the reasons that a particular text can generate horror. Obviously, I do not consider horror to be a genre which sprang full-grown from the turn of the nineteenth century. But also Carroll's carefully scrupulous if/then definition, by its very nature, misses a key characteristic of horror.

Horror, as I will argue over and over again in this dissertation, is no respecter of boundaries. One cause of horror is boundary collapse (as Carroll's sub-clause about impurity recognizes), and the emotion partakes of the nature of its causes. Stephen Booth remarks about Aristotle's Poetics that one reason for its enduring influence is that it makes tragedy "intellectually manageable" (KL,M 89). Carroll is attempting to do the same thing. But the entire point of horror is that it is not intellectually manageable, and thus a careful and stringent definition, by dissecting the concept so neatly, destroys the very quality that makes horror so effective.

Horror, as I define it, is predicated on category violation, on the transgression of intellectual and epistemological boundaries; it is inherently transgressive and thus cannot be neatly and completely articulated, or fully understood on a purely rational level.7 What horrifies us is what violates the categories of our world-view, and thus works of horror frequently examine the breaking of taboos. This quality is evident from the beginning, in Oedipus Rex (incest), Medea (infanticide), Thyestes (cannibalism). Good horror finds the nerve that lies under a culture's taboos and jabs it, hard. One reason that Stephen King's Carrie (1974) is horrifying, for example, is that it flouts the American social taboo against women's anger. Carrie's apocalyptic rage at the end of the novel is merely the overt acting-out of the seething anger and cruelty that make up the female culture of her high school. Horror, in general, like satire, works to reorganize our understanding of the world. It forces us to contemplate ordinary things from an extraordinary and frightening perspective. It transgresses the boundaries of our everyday epistemology.

Works of horror thus violate social assumption; they explore the destruction of community. This destruction is always multivalenced and overdetermined, so that the horror of social and political chaos in Macbeth, for example, is layered with the breakdown of psychological categories (Lady Macbeth's descent into madness), sexual categories (the witches, and Lady Macbeth's speech renouncing her sex), perceptual categories (both Macbeth and his wife have increasing difficulties with the category of sleep as the play progresses, and when Macbeth sees things that other characters do not, sometimes he is hallucinating, and sometimes the audience shares his hallucination), moral categories (Macbeth does evil, and yet we come to fear for him), and epistemological categories (the supernatural erupting into the natural world). Any of these category violations can be foregrounded in a discussion of the workings of horror in the play, precisely because they are all there and all interconnected. Horror makes boundaries permeable; it mixes things we feel ought to be kept distinct. It refuses decorum in every way it can.

One way of refusing decorum is to refuse moderation. Senecan horror is a genre of excess, excess in all possible arenas. If Death of a Salesman were a Senecan tragedy, Willy Loman would not merely die in an accident; he would poison his wife and dismember his sons and then, like Othello, stab himself. This is an intentionally silly example, but it demonstrates, through its exaggeration, what happens to tragedy when it falls into a Senecan world. Horror comes from a relentless, unstoppable proliferation of results, each worse and more appalling than the one before, an epistemological chain-reaction, tending ever more rapidly toward critical mass and explosion, annihilation, and death. There is no sense in Senecan horror of when "enough's enough." "Enough" is never enough.

This excess can all too easily trip horror into comedy. This permeability of the boundary between those two categories is why bad horror movies are so thick on the ground; for every Ridley Scott or John Carpenter, there are a dozen Ed Woods--or even worse, those whose visions do not even have the genius of their demented awfulness. At any appearance of any given monster, the audience is as likely to laugh as to scream. The Renaissance recognized this problem, as the mock ghost in Beaumont and Fletcher's The Knight of the Burning Pestle shows, and Seneca's plays constantly problematize and destabilize the gravity of their own undertaking, daring us to laugh at the same time they urge us to scream. Senecan horror is a constant tightrope walk between the appalling and the ludicrous, and the rope plays such as Thyestes walk, and dare themselves to fall off of, is that of the audience's belief. Senecan horror is characterized by its mixture of the tragic, the blackly comic, the camp, and the grotesque. It does not limit itself to one of these modes, but swerves vertiginously between them, indulging itself in an excess of moods as much as an excess of story or gore.

The indulgence in excess, the refusal of closure, has a metatextual dimension as well, janus-faced, looking at both comedy and horror. Monty Python's Flying Circus (1969-1974) excelled at the metatheatrical refusal of closure for comedic purposes. Sketches would end abruptly, unresolved, and frequently with a nonsequitur; or segue into "something completely different," or keep recurring throughout a particular episode. But even when enjoyable, lack of closure is also disturbing. The point of the Foucauldian subversion-containment model, after all, is the containment. True horror resists containment, and that makes it uncomfortable. The Poetics tells us that there are rules, and that good tragedy conforms to those rules. But any tragedy which generates true horror does not.

And the refusal to abide by certain rules of narrative makes King Lear, although only ambiguously supernatural, a great work of horror. Booth says:

When Lear enters howling in the last moments of the play, Shakespeare has already presented an action that is serious, of undoubted magnitude, and complete; he thereupon continues that action beyond the limits of the one category that no audience can expect to see challenged: Shakespeare presents the culminating events of his story after his play is over.
(KL,M 11)

An inability to stay within boundaries is thematically characteristic of horror as a genre, and it is also structurally characteristic of the genre of horror--and even more so of revenge tragedy; Cartwright suggests that even literal endings become contingent, arguing that endings may be less closural than an Aristotelian definition might desire; he describes endings merely as "resting places, they enable us to go home" (Cartwright 41). Part of the horror of revenge tragedy is in fact generated by our uneasy sense that it will not close, that the catastrophic events of the story will not be shut down. We will not be given, in Cartwright's terms, a resting place. "The Sorcerer's Apprentice" sequence from Disney's Fantasia (1940) is ultimately comic, but the nightmarish middle section demonstrates the kind of snowballing panic that plays like Titus Andronicus and Hamlet use. The closure of "The Sorcerer's Apprentice" in fact depends on a deus ex machina, the paternal figure of the Sorcerer himself. Revenge tragedies, which eschew the deus ex machina, can be closed only by death. And even then, as the figure of the ghost suggests, that closure is only partial, imperfect, reversible.

My specific interest, of course, is in those texts whose cultural work deals with the supernatural, the spectral. There are two different kinds of hauntings, benevolent and malicious. Both models focus on community: the benevolent arguing for the continuation of community, the hostile emphasizing isolation. In Gillian Bennett's study of the supernatural experiences of older women, she observes a clear distinction between what her respondents call "visitations" and what they classify as "hauntings. Benevolent visitations act as reassurance that community and love continue to exist beyond the grave; hostile hauntings recast that promise as a threat, turning the "visit" into an invasion (Bennet 24). Lise Saugères talks about contests between ghosts and the living for the space within houses; language of invasion is again prominent, whether it is the living person who feels threatened by the ghost, or the ghost responding hostilely to the advent of the living. The grave is a boundary that can be transgressed either to build community or to destroy it.

What is a ghost? As Marjorie Garber's book Shakespeare's Ghost Writers suggests, this is not as simple (or as simple-minded) a question as it appears. Garber describes the ghost as "a memory trace ... the sign of something missing, something omitted, something undone. It is itself at once a question, and the sign of putting things in question" (Garber 129). Ghosts are avatars, a priori, of the past. They represent the pressure of the past on the present, giving the past a voice. As a metonymy for the past, they also give it agency, the ability to act. In narrative, they enforce nonlinearity--the eruption of the past into the present perforce means that the present cannot move smoothly into the future. They also represent a story that is past but not finished; they incorporeally embody lack of closure. The ghost demands that the past's story be finished before the present's story can close.

Thus it is not surprising that literature returns again and again to the relationship between the living and the dead. What holds do the dead have on the living? What do the living owe the dead? Each culture has its own stories about how to negotiate this pain, beginning with Gilgamesh and his friend Enkidu.

In particular, ghosts are very popular figures in Renaissance revenge tragedy; there are over eighty plays listed in An Index of Characters in Early Modern English Drama Printed Plays, 1500-1660 as having ghosts in them. But there is no "typical" revenge tragedy ghost, for each ghost's appearance is shaped by the needs and aesthetic of its particular play. What revenge tragedy ghosts share is their structural role, the functions that they fulfill and the reasons that the plays need, over and over, to have them appear.

Theatrical ghosts emerge in part from the epistemology of theater. There is a very real sense, on the Renaissance stage, in which silence and death are equivalencies. They are both the realm of the abject; Kristeva describes the abject as death and writing (i.e., speech) as a resurrection (Kristeva 26). This is true in The Spanish Tragedy, where Hieronimo is forced to resort to silence when death is denied him; it shows up again in Othello, where of the maleficent doubles, Iago and Othello, one kills himself and the other chooses silence:

IAGO. Demand me nothing. What you know, you know.
    From this time forth I never will speak word.
LODOVICO. What, not to pray?
GRAZIANO. Torments will ope thy lips.
OTHELLO. Well, thou dost best.
(Oth. 5.2.309-13)

Although it is not clear to whom Othello's comment is addressed, by the end of the scene, he, too, will have chosen "from this time forth ... never [to] speak word." Speech, agency, and life are all intertwined in theatrical representation. Silence equates most readily to powerlessness and death.

Therefore, if silence equals metaphorical death, does death equal silence? Here, the answer is, Not necessarily. There is a convention available by which the dead can speak, by which the silenced can be given tongue. That, of course, is the convention of the ghost, and it is a convention that revenge tragedy utilizes to the fullest. Death and abjection are symbolically the same, as Kristeva remarks, and speech is the opposite of both.

Ghosts are the abject in the full sense of the word: that which is banished but cannot be exorcized, that which is denied only to return and speak. The space they inhabit is the space of the abject, the space of the liminoid, but also, as Alan L. Ackerman, Jr., points out, "the crucial space in-between [fiction and reality], the space of performance ... Ghosts, by definition, transgress boundaries, of present-past, material-spiritual, and real-imaginary" (Ackerman 136). Drama is peculiarly well-suited to the trope of the ghost, and vice versa.

Ghosts in revenge tragedy, as in other kinds of fiction and in the wider cultural understanding, are pieces of the past, unfinished business. The primary function of ghosts in revenge tragedy is as motivation for revenge. It is for this reason that they are ghosts: they are dead, murdered, and yet still possessed of will and the ability to communicate, whether verbally or through their actions. They are the mechanism by which hidden things may be revealed. Freud observes that one of his sources classifies as unheimlich "everything ... that ought to have remain secret and hidden but has come to light" (Freud 225). Ghosts in revenge tragedy, already unheimlich by virtue of being familiar/unfamiliar at the same time, furthermore enact this definition. Ghosts are puzzles; the percipient must take the first step toward communication before the ghost can make its needs known--as in Hamlet, where the Ghost can do nothing but terrify the guards until they understand that they must bring Hamlet to him. And finally, the ghost is a spectacle which induces horror: witness again the guards in Hamlet or the reaction of De Flores to the ghost of Piracquo in The Changeling. Ghosts are defined by the use the narrative makes of them and by the reactions of those characters who can see them. As such, they have a strange dual existence. They are both a category violation and a category unto themselves. Even before the categories which define order in a revenge tragedy begin to erode, the ghost is a special and peculiar problem.

For the characters in a play, ghosts are terrifying category violations, dead persons who nonetheless still affect the world of the living. Ghosts are unnatural, wrong, a sign that the order of the universe has been disturbed. They appear, in revenge tragedies, because something is out of place or unfinished. Therefore, to the characters who see them, they are both a sign that the world is askew and in and of themselves a violation of the natural order. They are frightening because they exist. They deny the stability of the material world.

For the audience of a play, on the other hand, the ghost is a category, a stock character who behaves in the ways I outlined above, an entity whose existence has rules of its own. When we go to see a play with a ghost, we have expectations about the way that ghost will behave. Standing outside the fiction, we see that the ghost does have a place in the structuring of the story, in the microcosm the author has created. We understand the Ghost as a category, a form on which variations can be performed, the same way in which we understand the Evil Tyrant or the Ingenue. Therefore, we may share the characters' fear, but our world does not totter; if the ghost remains within the rules the audience understands, it cannot horrify them. Many revenge tragedies thus generate horror by violating the category of ghost as it is understood by the audience.

An example of how this can work is offered by Thomas Heywood in "An Apology for Actors" (1612):

At Lin in Norfolke, the then Earle of Sussex players acting the old History of Fryar Francis, & presenting a woman, who insatiately doting on a young gentleman, had (the more securely to enioy his affection), mischieuously and seceretly murdered her husband, whose ghost haunted her, and at diuers times in her most solitary and priuate contemplations, in most horrid and fearefull shapes, appeared, and stood before her. As this was acted, a townes-woman (till then of good estimation and report) finding her conscience (at this presenment) extremely troubled, suddenly skritched and cryd out Oh my husband, my husband! I see the ghost of my husband fiercely threatning and menacing me. At which shrill and vnexpected out-cry, the people about her, moou'd to a strange amazement, inquired the reason of her clamour, where presently vn-urged, she told them, that seven yeares ago, she, to be possest of such a Gentleman (meaning him) had poysoned her husband, whose fearefull image personated it selfe in the shape of that ghost: whereupon the murdresse was apprehended, before the Iustices further examined, & by her voluntary confession after condemned. That this is true, as well by the report of the Actors as the records of the Towne, there are many eye-witnesses of this accident yet liuing, vocally to confirme it.
(Heywood Gv-G2)

Heywood frames this as part of his argument about the morality of theater, but considering the fact that in other parts of his treatise (such as the section on boy-actors) he is insistent that the audience does not confuse fiction with reality (Heywood C3v), his exemplar here is a two-edged sword. For my purposes, the interest in this anecdote is in the confusion between audience and fiction, the murderess's conflating the "ghost" on stage with the imagined ghost of her victim. The ghost ceased to be a character for her, and became a reality.

Heywood's example is an extreme one; in general, ghosts do not affect the audience so directly. But still, what makes them figures of horror is the way in which they reach out of the fiction to strike a nerve. This violation of boundaries is also what enables and signifies the cultural work they do.

Ghosts in Seneca's plays, as with other manifestations of the paranormal, serve to express the existential dread, uncertainty, and absurdity of Julio-Claudian Rome. Seneca's stance on the supernatural is always and deliberately self-contradictory; he uses the otherworldly as a metaphor to record the absurdity (in the existential sense) of the real world in which he lives.

In early modern revenge tragedy, ghosts represent the individual's troubled relationship to the past. The Reformation further inflects our understanding of ghosts in Renaissance England. Protestant doctrine abolishes Purgatory, "claiming that it lacked scriptural foundation,"and belief in Purgatory was considered "vain, perilous, and injurious" (Cressy 386). Protestant authorities insist that ghosts are demons come from Hell to tempt the living into eternal damnation, not the spirits of the dead. As Eamon Duffy remarks in his discussion of the 1552 prayer-book: "the dead were no longer with us. They could neither be spoken to nor even about, in any way that affected their well-being. The dead had gone beyond the reach of human contact, even of human prayer" (Duffy 475). There was no longer any continuum between the living and the dead; orthodox Anglican theology declared that boundary impervious. But nevertheless, the boundary continued to be as fluid and collapsible as it had always been. People continued to pray for the dead for several decades after the practice had been outlawed by the orthodoxy (Cressy 398). And if the Protestants made slow headway against traditional religious practices, it is no wonder that they made equally slow headway against traditional beliefs; People continued to see ghosts, and they continued to believe they were the spirits of the dead, as witnessed by the pamphlets of haunting from the mid-seventeenth century, a hundred years and more after the Reformation came to England.

There are nine of these pamphlets extant, published between 1643 and 1683.9 These are not the learned treatises of Lavater or Taillepied or Glanvil; rather, the pamphlets are the accounts, or purported accounts, of encounters between ordinary people and the spirits of the dead. These nine pamphlets provide us with nine snapshots of English Renaissance and Restoration culture, offering glimpses of how the people of that time thought about and interacted with the dead. These pamphlets emphasize societal obligations, the orderly transmission of property, the meting out of justice. Revenge tragedies will show us a different relationship with the dead, this one based on the individual, focusing on the consequences of transgressing the social framework suggested by the pamphlets.

The pamphlets demonstrate communitas: social bonds and obligations which continue to hold even after death; by honoring those obligations, both living and dead strengthen the community. Bennett's study from the 1980s discovered the same concerns and same values: "They therefore provide for the women the strongest possible evidence, not only of the survival of a personal identity after death, but also of the continuance of the important structuring relationships of family and kinship" (Bennett 64). In these pamphlets, as in Bennett's study, the important boundary is that between the community and the outside world, not that between the living and the dead.

The community-centered model of haunting is in stark contrast to the effect of ghosts in plays, which is to destroy community, to isolate the protagonist from those around him or her. The pamphlets as much as the plays show boundary transgressions, but ultimately in the pamphlets, the crucial categorical distinction is not between living and dead, but between the community and the wilderness. Many of the pamphlets are about the defusing of horror; ghost after ghost promises that the living have nothing to fear from them. There are exceptions, but those exceptions reflect a failure of community rather than its dismemberment. In a nutshell, the pamphlets are about the workings of community; the plays are about the relationship--and the dissolution of the relationship--between the individual and the community.

Theatrical ghosts reexamine the traditional valences of the community between the living and the dead; more and more in the theater, the dead are seen as dominating and oppressing the living. The obligation the living hold toward the dead becomes unhealthy and imprisoning, not a cementing of community but an entrapment in concerns that should have been laid to rest in the graveyard.

Also, early modern revenge tragedies enact the failure or unsustainability of community; the ghosts represent the failed community. These tragedies may or may not end with the establishment of a new community; if they do, it will not be one that includes ghosts. The cultural schism of the Reformation is reflected in the early modern plays discussed in this dissertation. Each play performs the unraveling of a community, its inability to maintain itself against the transgressions of villains and revengers. The plays ask the question of whether or not it is possible to reconstruct a community shattered by transgression; in most cases, their answers are not optimistic. And in all cases, transgression and redress are troped and worked through via the ghosts generated by the plays.

The argument of this dissertation is multifold. Starting from the understanding that horror is generated from category violations, it argues that early modern revenge tragedy, following in the Senecan tradition of theatrical horror, explores the disasters inherent in the breakdown of community and the violation of social categories. This breakdown is represented on the stage frequently by the figure of the ghost, itself a figure transgressing the boundaries of categories such as "dead" and "living." I argue that horror performs cultural work, allowing us to explore category violations within the safe containment of fiction--although, because it is horror, that containment is never stable nor complete. I also argue that revenge tragedy is a highly self-conscious, highly metatheatrical artform, that the effects it produces and seeks to produce are not nave or accidental.

I will begin to trace the literary tradition of horror on stage with the tragedies of Lucius Annaeus Seneca. Chapter Two will examine the seminal text of early modern revenge tragedy, The Spanish Tragedy. Chapter Three will discuss two of Shakespeare's early plays, Richard III and Titus Andronicus. Chapter Four is an in-depth examination of Hamlet, and Chapter Five will discuss Jacobean revenge tragedy, focusing on The Atheist's Tragedy, The Second Maiden's Tragedy, The Changeling, and The Revengers Tragedy. The conclusion will discuss the generic shifts in revenge tragedy from Kyd to Tourneur and the ways in which the critical model I have constructed can be applied to nineteenth and twentieth century horror in fiction and film. The dissertation will finish with an examination of modern Senecan horror in Peter Shaffer's play, Equus.


1. Although it is not part of my argument, I would like to note the confused double-standard which can admit that Clytemnestra overshadows Aegisthus, yet still label Aegisthus, rather than Clytemnestra, one of "Seneca's three great revengers."

2. E.g., movies: Psycho (Alfred Hitchcock, 1960), Pulp Fiction (Quentin Tarantino, 1994), Once Upon a Time in Mexico (Robert Rodriguez, 2003); books: The Talented Mr. Ripley (Patricia Highsmith, 1955), L. A. Confidential (James Ellroy, 1990), The Cipher (Kathe Koja, 1991).

3. Karen S. Coddon sees the same thing when she remarks: "It seems no cultural accident that the popularity of the trompe l'oeil in early modern Europe roughly coincides with the radical anti-mimesis of Jacobean tragedy" (Coddon 74).

4. For an discussion of distancing effects in Jacobean tragedy, though not couched in Brechtian terms, see Gruber 197.

5. Marjorie Garber identifies this compulsion at work in Hamlet (Garber 128-29).

6. E.g., he asserts that Hans Christian Andersen's stories do not partake of the unheimlich, whereas I, as a reader, would assert that they most emphatically do.

7. In this way, my conception of horror greatly resembles the Romantic conception of the Sublime. Horror, however, remains linked to the visceral, the bodily, which the Sublime does not.

8. All citations from Shakespeare are to The Norton Shakespeare.

9. Horrid and Strange News from Ireland (1643), A Strange and Wonderfull Discovery (1662), A True Relation of a Horrid Ghost of a Woman (1673), News from Puddle-Dock (1674), The Deemon of Marleborough (1675), The Rest-less Ghost (1675), Great News from Middle-Row (1679), Strange and Wonderful News from Linconshire (1679), The Demon of Spraiton (1683).

prefatory material (including copyright notice) | Chapter One

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