Are Shakespeare's plays always metatheatrical?
The ambiguity of the term "metatheatre" derives in part from its text of origin, Lionel Abel's 1963 book of the same name. By his own admission, Abel's use of the term was "loose and sometimes erratic" (v). If we use the term in its broadest sense—to describe any theater that in some way draws attention to its own artifice—it becomes evident that early modern drama is always "metatheatrical" to some extent: these plays are designed never entirely to lose sight of the material realities of their performance, or of the physical co-presence of their audiences. If this is the case, how useful is the term "metatheatre"? Indeed, are Shakespeare's plays always metatheatrical? This article unpicks some of the conflicting notions of metatheatre suggested in Abel's book, and suggests a modified conceptual model based on the work of Arthur Koestler. Arguing against the tendency to see early modern theatrical self-consciousness as a form of proto-Brechtian alienation, it uses Koestler's concept of bisociation to think about the delight produced by "universes of discourse colliding, frames getting entangled, or contexts getting confused" (40). It considers several examples from performance, especially moments from productions at the reconstructed Shakespeare's Globe, to argue that metatheatre functions as a kind of imaginative game. This game may be prompted by cues in the written text, but it is one that can be played only in performance. While Harry Newman's essay for this special issue argues that metatheatricality was available to early modern readers "on the paper stage of printed playbooks" (104), my essay posits a decidedly more theatrical definition of the term, contending that the agency of the actors plays a central role in determining the metatheatricality of particular moments on stage.
Some definitions of "metatheatre" are more specific. One is the play-within-the-play, or more broadly, the play that stages some kind of sustained exploration of the nature of dramatic art. Abel opens his book [End Page 19] with a chapter on Hamlet, and when he turns to Shakespeare's wider body of work, his first observation is that "Shakespeare experimented throughout his whole career with the play-within-a-play, sometimes introducing play-within-a-play sequences in his tragedies, almost always introducing such sequences in his comedies" (140). By this definition, plays like Hamlet, Love's Labour's Lost, A Midsummer Night's Dream, and Henry IV Part 1 might be considered metaplays alongside more recent examples such as Chekhov's The Seagull, whose first act revolves around the characters' preparations for, and aborted staging of, an experimental play on a makeshift lakeside stage at a Russian country estate. Indeed, like some of its Shakespearean counterparts, The Seagull's play-within-a-play is framed by discussions of art, the imagination, and the responsibilities of the dramatist.
But though Abel recognizes that some of the plays he is discussing can "be classified as instances of the play-within-a-play", many of them do not employ the device. "Yet the plays I am pointing at do have a common character," he continues:
All of them are theater pieces about life seen as already theatricalized. By this I mean that the persons appearing on the stage in these plays are there not simply because they were caught by the playwright in dramatic postures as a camera might catch them, but because they themselves knew they were dramatic long before the playwright took note of them.(134–5)
There are two important aspects to this description, the closest Abel comes in the book to offering a definition of his term. First is the notion of life as "already theatricalized": Abel opens the book with an analysis of Hamlet not because of The Murder of Gonzago, but because he reads the play as a text populated with characters who behave either like dramatists (Hamlet, Claudius, Polonius, and the Ghost) or like actors (Gertrude, Ophelia, and Laertes). In this sense, we might see a play such as Ibsen's A Doll's House as a metaplay, featuring as it does a protagonist who stage-manages her family life, constructing and attempting to maintain a precarious identity as her husband's "little skylark" before recognizing, as the play reaches its climax, that her "home has never been anything but a playroom" (98). "I've been your doll-wife," Nora tells her husband Torvald, "just as I used to be papa's doll-child" (98). In a symbolic moment towards the end, she leaves the stage to remove her "fancy dress" (96).
But I have cited these parallels from the naturalistic theater tradition in order to draw attention to what I take to be the defining feature of Abel's concept of metatheatre, and it is one that Chekhov and Ibsen's characters [End Page 20] typically do not share. Abel's metatheatrical stage figures "are aware of their own theatricality" (135). When the actor playing Peter Quince emerges onto a stage in front of a tiring house, pretends that he is in a wood, observes "here's a marvellous convenient place for our rehearsal," and then suggests "this green plot shall be our stage, this hawthorn-brake our tiring-house" (3.1.2–4; all citations from Shakespeare are from Taylor and Wells's Oxford edition), he is doing something very different from Chekhov's portrayal of frustrated lakeside thespians. Similarly, Nora and Torvald's domestic role-playing is not of the same order as Edmund's, when he notes of his hapless half-brother Edgar, "on's cue out he comes, like the catastrophe of the old comedy; mine is villainous melancholy" (1.2.129–31). Chekhov and Ibsen's stage figures are unaware of their real-life audiences, or of the material realities of the theatrical effects that sustain the illusion that they are real people. Shakespeare's, on the other hand, are not bounded by a self-contained naturalistic world. They know they are in a play: they could not talk to us in asides and soliloquies if they did not.
This degree of self-consciousness in Shakespeare's work (and indeed in the early modern theater more broadly) is often discussed in terms of disjunction and alienation, and sometimes even seen as a form of proto-Brechtian alienation. Thus, for James L. Calderwood, Hamlet's many "instances of theatricalization … serve as Brechtian alienation devices to shatter our illusion of Danish reality and cut the cord of our imaginative life there" (167). Similarly, for Graham Holderness, the Chorus speeches of Henry V "are there to foreground the artificiality of the dramatic event, placing a barrier between action and audience" (History 137), while the Sly frame in The Taming of the Shrew might have worked "in the self-reflexive, metadramatic and ironic manner of Brecht's epic theatre" (Performance 25). While I do not entirely disagree with Holderness or Calderwood on these examples, I wonder whether we might usefully question our eagerness to reach for the word "Brechtian" and its related concepts of alienation and critical distance whenever we see self-conscious theatricality in early modern drama.1 Writing about "metadrama" in general, Richard Hornby describes the "experience for the audience" as "one of unease," leading at times to "the most exquisite of aesthetic insights, which theorists have spoken of as 'estrangement' or 'alienation'" (32). I do not doubt that Hornby's argument holds true for a great deal of post-naturalistic metadrama, but I am resistant to the notion that metatheatrical moments in early modern drama necessarily produce audience "unease" or "alienation." I want to argue that moments [End Page 21] like the Quince and Edmund lines are more likely to produce delight than distancing, and I would like to suggest a conceptual model that allows us to consider this delight in greater detail.
I suggest that Shakespearean metatheatricality functions along similar lines to a pun. Let us consider a simple example of the form: "A good pun is its own reword." Like all puns, this one brings two separate ideas into humorous collision: in this case, the cliché that a good thing can be "its own reward," and the fact that a pun can be expected to "re-word" something. Appreciation of a pun of this sort relies upon the reader's ability to recognize the co-existence of these two separate ideas at once in the same utterance. (The fact that this is a pun on the subject of punning lends it a further layer of charm: perhaps we should call it a meta-pun.) Let us take another example from Shakespeare's own "corrupter of words," Twelfth Night's Feste (3.1.35). When Feste claims to "live by the church" (3.1.3), spectators may initially be as puzzled as Viola: we know that he makes his living as a Fool and not as a "churchman" (3.1.4). But when he reveals "my house doth stand by the church" (3.1.6–7), a second meaning of "live by" springs into view, producing a flash of insight, and, hopefully, a laugh. Our perceptual lens has been suddenly and playfully shifted.
One of the neatest explanations of the delight provoked by verbal humor appears in the 1964 book The Act of Creation by Arthur Koestler. Koestler's wide-ranging study of creative thought begins with a discussion of humor. Describing jokes as "universes of discourse colliding, frames getting entangled, or contexts getting confused" (40), he uses his discussion of humor to outline his concept of bisociation. Koestler explains this as
the perceiving of a situation or idea, L, in two self-consistent but habitually incompatible frames of reference, M1 and M2 […]. The event L, in which the two intersect, is made to vibrate simultaneously on two different wavelengths, as it were. While this unusual situation lasts, L is not merely linked to one associative context, but bisociated with two.(35)
"M" here stands for "matrix," the word that Koestler uses to denote "any ability, habit, or skill, any pattern of ordered behaviour governed by a 'code' of fixed rules" (38). For Koestler, everyday life usually requires us to think on only one plane, or matrix, at a time, whereas the creative act "always operates on more than one plane" (35). Bisociation, he argues, is a means of "escaping our more or less automatized routines of thinking and behaving"; it is "the spontaneous flash of insight which shows a familiar situation or event in a new light, and elicits a new response to it" (45). A phenomenological "flash" of this sort is an intrinsic part of imaginative [End Page 22] delight: I remember, for example, my two-year-old niece barely able to contain her excitement at discovering a misshapen strawberry with two points, introducing it to anyone who would listen as a "strawberry rabbit." At once both a strawberry and a rabbit, the object was resonating, in my niece's eyes, on two perceptual planes at once.
When Koestler addresses theatrical performance much later in The Act of Creation, he understands it as working rather differently from the "flash" produced by a joke. Theater, he argues, functions as a much more continuous state of bisociation in which the knowledge that the events of the play are fictional is held in mind as spectators simultaneously respond to those events with participatory emotions. "The aesthetic experience," he says,
depends on that delicate balance arising from the presence of both matrices in the mind; on perceiving the hero as Laurence Olivier and Prince Hamlet of Denmark at one and the same time; on the lightning oscillations of attention from one to the other, like sparks between charged electrodes. It is this precarious suspension of awareness between the two planes which facilitates the continuous flux of emotion from the Now and Here to the remoter worlds of the Then and There, and the cathartic effects resulting from it.(306)
The way in which Koestler maps theatrical performance onto his theoretical model is thus clear. On one plane (let us call it M1) is the "Now and Here" of the real-life actor (Laurence Olivier), the theater and the audience; on the other (M2) is the "Then and There" of the character (Prince Hamlet of Denmark) and the fictional setting. Before and after the performance, these planes are entirely separate, but for the duration of the performance, they are bisociated, so that the stage figure is perceptible as both Hamlet and Olivier at once.
In their book The Way We Think, the cognitive scientists Gilles Fauconnier and Mark Turner take Koestler's concept of bisociation as one of the starting points for their own theory of "conceptual blending." The human brain, they argue, is capable of "running multiple conceptions simultaneously, some of them conflicting with each other, and it seems that the brain is very well designed to run such multiple and potentially conflicting conceptions" (232). In the theater, then, conceptual blending allows the spectator to perceive Hamlet and Olivier as separate entities, existing on different perceptual planes, but also to construct a third identity, the "blended" Hamlet/Olivier, who occupies both planes at once. Fauconnier and Turner discuss theater only briefly in their book, but they give a clear [End Page 23] sense of their understanding of this dynamic. "Dramatic performances," they argue, "are deliberate blends of a living person with an identity":
They give us a living person in one input and a different living person, an actor, in another. The person on stage is a blend of these two. The character portrayed may of course be entirely fictional, but there is still a space, a fictional one, in which that person is alive.(266)
In watching drama, they continue, "we are simultaneously aware of the actor moving and talking on stage in front of an audience, and of the corresponding character moving and talking within the represented story world" (266). Bruce McConachie has extrapolated a fuller theory of theater spectatorship from Fauconnier and Turner's ideas, and he articulates this in his own book Engaging Audiences. McConachie uses the notion of conceptual blending to discuss the blended "actor/character". Like Koestler, McConachie uses the metaphor of "oscillation" to describe the mental process that produces this blended stage figure, arguing that "theatre audiences oscillate between counterfactual blends and perceptions of their actual, material circumstances" (50). However, there is an important difference between their uses of the metaphor: whereas for Koestler the oscillation is between the "Now and Here" and the "Then and There," these planes remaining cognitively separate from one another, for McConachie the oscillation is between "blending" and "unblending." As McConachie explains it, theater spectators "oscillate millisecond by millisecond among blends and singular identities, not between skepticism and faith" (44).
These thinkers present bisociative theories of theatrical spectatorship in general; their theories describe the normal functioning of drama, at least as they understand it. For a consideration of the ways in which bisociation in metatheatre might work differently from theatrical bisociation more broadly, we might turn to the Polish theorist Slawomir Swiontek.2 Swiontek uses an idea much like Koestler's to describe metatheatre. For Swiontek, drama operates simultaneously on two planes: the "stage-stage" axis in which characters speak to one another within a fictional world, and the "stage-house" axis in which the performance communicates to its real-world audience. As he puts it, a theatrical act is
a presentation, by a real executor (actor) in the presence of a real addressee (spectator), of the fictional acts of communication between the characters, in order to provoke a real act of communication across the actor-spectator (stage-house) axis.(129)
[End Page 24] "Metatheatricality," he argues, "appears when the two addresses (two axes of communication) and two destinations of the utterances that constitute the dialogue are revealed or thematized" (136); in Jenn Stephenson's words, "any time the audience becomes aware of this higher-order pragmatic function of dialogue" (117). Stephenson's translation of Swiontek uses numerous synonyms for "reveal" to describe what metatheatre does to the stage-house axis: to "make apparent," to "expose," to "unveil." At the metatheatrical moment, argues Swiontek, "[t]he theatrical situation, hidden until now in the text, begins to be articulated. all this makes apparent the meta-enunciative character of dramatic dialogue" (136). Once again, this is a fundamentally Brechtian model of metatheatre in which spectators find themselves pulled out of a participatory mode of engagement with the fiction and distanced from it. This is even clearer in Stephenson's own elaboration on Swiontek, in which she discusses "the detached awareness that comes from metaphorically stepping back from engagement with the fictional properties of the art work into a cooler, more detached stance that perceives the process of its construction as art" (121). "For the audience," she argues, "the revelation of the secondary meta-enunciative axis is what brings us 'back to ourselves,' back to an awareness that the play is just a play, an event in the actual world" (118).
The trouble with the theoretical model outlined above is that it assumes that realism and a self-forgetful absorption into its fiction are the baseline modes of drama and spectatorship, and that any acknowledgement of the presence of the audience or other deviation from reality effects will disrupt this absorption. Swiontek and Stephenson both presuppose that drama typically persuades its audiences to lose sight of the "meta-enunciative" stage-house axis, to forget that "the play is just a play." For Swiontek, "all revelation of the theatrical situation destroys the scenic illusion" (131). The idea of "scenic illusion" is itself telling. Indeed, the very notion of a "stage-house axis" is rooted in a post-nineteenth-century concept of theater: in her summary of Swiontek's work, Stephenson describes the "dual function" of theatrical dialogue as "two communicative vectors forming perpendicular axes; one traversing the stage (stage-stage) and the other arcing past the proscenium (stage-house)" (117). When she discusses "the geography of the theatre" and of "[j]oining the stage to the house," her assumption of a picture-frame stage is clear (118). Elsewhere, she refers to "the audience-eavesdropper" as opposed to "the character-participant" (117).
While these assumptions may well hold for naturalistic drama, I am not convinced they work so well for early modern plays. What if we start [End Page 25] with the opposite assumption, namely that audiences are treated by stage figures as present and in on the game of imaginative co-creation from the start? What if the character is always already communicating to an audience, and this communication was never hidden in the first place? What if, in other words, the audience is not an "eavesdropper," but a "participant"? Swiontek's theory assumes that mimetic dialogue is the normal mode in drama and that soliloquy is a sort of breaking of that convention. But in early English drama, soliloquy is more often than not the mode of speech that frames plays: in medieval mystery plays and morality interludes, or in Tudor plays like Jacke Jugeler or Roister Doister (both first performed in the 1550s), spectators are addressed at the beginnings and ends, and indeed throughout, by fictional characters who openly acknowledge the presence of their real-world audience. Shakespeare's own plays make frequent use of this mode, characters like Richard III, Falstaff and even Hamlet turning their attentions throughout the performance to the mass of spectators gathered around them. It is hard to distinguish the "stage-house axis" from the "stage-stage axis" when a stage figure who is aware of the audience's presence stands amongst them on a thrust stage in shared light; under such circumstances, acknowledgement of the play's "meta-enunciative" function hardly seems much of a revelation, let alone any kind of disruption.
I would like to return to Koestler, then, to suggest a different model for thinking about the kind of bisociation enacted by metatheatre. Fiction and reality are of course largely separate, and separable, in the theater. Spectators of early modern drama knew just as well as those of naturalism that when a character dies on stage, they need not worry that the actor has died for real: Bottom's insistence on a prologue to assure the audience "that Pyramus is not killed indeed" (3.1.17–18) is funny precisely because he does not understand this. There are, then, two planes at play in drama just as Koestler suggests, the "Now and Here" (M1) and the "Then and There" (M2). Spectators perceive these planes as separate, and "oscillate" between them (or perhaps, as McConachie suggests, between "blending" and "unblending" them) over the duration of the play. But metatheatrical moments are not, I suggest, the moments at which the audience notice M—they almost certainly never stopped noticing it in the first place—but those moments in which M1 and M2 become entangled, a line or theatrical moment resonating on both levels at once. In 2015, I directed a touring production of Macbeth in which five actors performed the play as a kind of low-tech, semi-parodic film noir: they used handheld battery-powered floodlights to suddenly shift the scene, shining a [End Page 26] pair of parallel lights towards the audience to suggest car headlamps, for example, or holding one light above the head of another actor and pointing it downwards to evoke a lamppost.3 Audiences tended to laugh when they recognized the film noir tropes that were being sketched in, and I assume their laughter was like that provoked by a pun: these stage images were registering at once both as the material work of the actors and as the imaginary world they were creating, the worlds of "Now and Here" and "Then and There" coming into creative collision. I am under no illusion that spectators were reflecting on the play from a critical distance here: they were expressing pleasure. But this pleasure need not be purely comic. My production evoked the stormy weather of Act 2, for example, by having an actor scrunch a plastic carrier bag against the surface of an onstage microphone, creating a rumbling, pattering sound a little like that of wind and rain. Spectators rarely laughed at this, but sometimes commented afterwards that they had found the effect enjoyable. We could easily have played a recorded sound effect instead, but audiences seemed to derive pleasure from having their attention drawn to the means by which the effect was produced at the same moment that they registered its fictional import.
Shakespeare's plays are full of moments that invite this sort of bisociation. Stephenson gives an example from Hamlet in which the title character, hearing the Ghost cry out from under the stage, refers to him as "this fellow in the cellarage" (1.5.152); she notes that the word "cellarage" works both "as a theatrical metaphor to indicate under the ground" and as a literal reference to the area beneath the stage, so that the line "resonates with metatheatrical duality" (125–6). At the Globe, when Hamlet looked up at "this majestical roof fretted with golden fire" (2.2.302–3), the line was at once both a reference to the decorated heavens of the stage canopy, and to an imagined sky. Something similar may have happened later in the play when Polonius recalls having played Julius Caesar at university, and having been killed by Brutus (3.2.94–101); Andrew Gurr argues that Polonius must have been played by the same actor who had played Caesar at the Globe the previous year ("probably John Heminges"), while Hamlet, with whom Polonius is speaking at this moment, would have been played by that production's Brutus, Richard Burbage (106). For Gurr, in-jokes of this sort reflect Shakespeare's expectation that his audiences "would be well aware of their environs, and that the fictions were to be seen as open mimicry whose pretence at deceit was obvious" (106).
Bert O. States applies Koestler's theory to Shakespearean performance in his seminal work on theater phenomenology, Great Reckonings in Little Rooms. "In productions of The Two Gentlemen of Verona," argues States, [End Page 27]
Launce's dog Crab usually steals the show by simply being itself. Anything the dog does—ignoring Launce, yawning, wagging its tail, forgetting its "lines"—becomes hilarious or cute because it is doglike. The effect here is comic because it is based on a bisociation, in Arthur Koestler's term. We have an intersection of two independent and self-contained phenomenal chains.(1985: 33)
The image of the dog onstage operates at once on both matrices: the fictional Crab and the real dog playing him. So far, so much like any theater performance. But the metatheatrical element here is that spectators know that the dog cannot be aware that it is "acting." Any behavior it exhibits is inescapably its own behavior, and the comedian playing Launce must incorporate that element of the "Now and Here" into the "Then and There" of the fiction, in which (in Launce's eyes, at least) it becomes a sign of Crab's hard-heartedness. The matrices become muddled. "The 'flash' at the intersection," says States, is "equivalent to the punch line of a joke" (1985: 33).
Tim Carroll gives an example of a comparable (though less overtly comic) incorporation of the "Now and Here" of an animal's real-life behavior during a performance of his Macbeth at Shakespeare's Globe in 2001:
During one performance, just after Macbeth had learned that his wife had died, a pigeon landed on the stage just in front of him. It was, of course, completely incongruous. Some people in the audience giggled nervously: surely this was going to spoil the moment. But did it? No. Jasper [Britton], being the remarkable actor he is, immediately saw his opportunity. He looked at the pigeon as though its landing merely summed up the undignified absurdity of life. Then, when the pigeon began to walk along the front of the stage, it was as though this made a strange thought occur to Macbeth. He said "Life's but a walking shadow, a poor player / That struts and frets his hour upon the stage" (5.5.23–4) and then he waited till the pigeon flew off before saying "And then is heard no more" (5.5.25).(39)
Carroll interpreted the audience laughter that accompanied this interaction as "a moment of beautiful revelation, vouchsafed by the beauty of chance" (40). Another collision, then, that despite its deeper philosophical implications—or perhaps because of them—produced the laughter of delight that often accompanies the metatheatrical "flash." This moment was clearly not planned for, nor anticipated in the text in the same way that the behavior of the dog playing Crab generally will have been; in this instance, the actor simply saw a one-off opportunity to bisociate the "Now [End Page 28] and Here" with the "Then and There," and took it, brilliantly. Macbeth's speech is, of course, a deeply metatheatrical one, but it hardly invites the specific bisociation for which Britton found opportunity on that particular day. As this suggests, metatheatrical moments can be invited by the text, but are not delimited by it.
Nor is metatheatre confined to references in single lines or to individual moments in performance. Often in Shakespeare's plays, bisociative collisions are sustained over whole scenes. Koestler describes jokes as "universes of discourse colliding, frames getting entangled, or contexts getting confused" (40), and suggests that it is possible to "dissect any specimen of humour […] by discovering the type of logic, the rules of the game, which govern each matrix" (64). I want to argue that early modern metatheatre can be analyzed in much the same way. The concept of metatheatre as a game of repeated bisociation is hinted at in Carroll's discussion of Malvolio's gulling scene (2.5) in his 2002 production of Twelfth Night at the Globe:
The whole situation is so unbelievable that I had a disagreement with one of my colleagues at the Globe, who said, "You have the box-tree too close to Malvolio." He said, "Surely they should be in the musicians' gallery, where it is easier to believe that Malvolio would not hear them." I said, "Wherever we put them, it is impossible to believe that Malvolio would not hear them. It's a game."(38)
Gulling and overhearing scenes are metatheatrical, I suggest, not primarily because their characters occupy spectator-like positions, but because the scenes are violently bisociative: as every theatergoer knows, they are funnier the more audacious the game becomes.4 The game is being played on two planes at once, and both play and performance orchestrate repeated and humorous collisions between the two. One rule governs the behavior of the characters on the "Then and There" matrix: the characters in Twelfth Night are playing the game of trying to remain hidden. A contradictory rule prompts the decisions of the actors on the "Now and Here": the actors, generally speaking, are playing the game of putting their characters' chances of remaining hidden into comic jeopardy. Both are on view at once in the same sequence of stage behavior, and humor results from the spectators' simultaneous awareness of both games. Thus, in Carroll's production, the eavesdropping characters broke out into cooing bird impersonations when Malvolio almost heard one of Sir Toby's noisy interjections; they picked up the whole box-tree in which they were hiding (something, of course, that one can only do with a prop box-tree) [End Page 29] and moved it closer to Malvolio in order to better overhear him; having recklessly leant out of it, they ducked back in when Malvolio turned around. Each near miss, each audience laugh, signaled a bisociation. In a successful performance, spectators relish watching skilled performers doing the acting equivalent of spinning plates, repeatedly juggling the conflicting demands of "Now and Here" and "Then and There." We enjoy the scene for its audacity in openly stretching our credulity, extending the play of the "Now and Here" while pushing the "Then and There" further and further towards a breaking point at which the fiction will collapse into nonsense. It is no coincidence that Shakespeare draws attention to this in another bisociative line later in the play, when Fabian observes, "If this were played upon a stage, now, I could condemn it as an improbable fiction" (3.4.125–6).
This is an example of what the theater practitioner John Wright describes as "declaring the game" (45–7). In his book on physical comedy, Wright uses "game" as a metaphor for the rules governing an actor's stage behavior:
Games have the same function as an objective. The plot gives you the dramatic context but the game gives us the life of the interaction. Games give you a simple structure in which to play an objective—and you must play to win.(88)
A Stanislavskian actor, then, who decides on her character's objective and performs actions in order to achieve that objective, is playing a game that is hidden from the audience, creating an impression for the audience of her character as a realistic person with hidden desires and compulsions. But a non-naturalistic actor has the option of "declaring the game" to the audience. Under such circumstances, says Wright, there will be "no illusion, and all your actions will be valued for what they are rather than for what they imply. When you declare the game, you play it so as to have an effect on the audience" (46).
Throughout his book, Wright makes use of terminology very similar to Koestler's "Now and Here" and "Then and There," drawing repeatedly on Hamlet as an example. Audiences enter "a curious split-mindedness" whilst watching a play, he suggests, in that they "flip from the 'here and now' to the 'there and then,' and back again, with consummate ease" (78). Wright recalls the final scene of a production of Hamlet that he directed, in which the actor playing Hamlet exploited this doubleness by mocking the ways in which the other actors had played their death throes, before performing his own character's death. Wright describes the audience reaction: [End Page 30]
We laughed because it was such a relief, after the carnage we'd just seen and we also laughed because, on another level, the final scene of Hamlet is dangerously over the top for a young theatre audience in a small provincial arts centre. But in declaring the game, Hamlet kept us in the "here and now." They were all just actors playing dead—it might be horrible but really no more than a few gurgling noises and dramatic twitching.(47)
But Wright goes on to explain that this foregrounding of the "here and now" did not, in fact, create emotional distance. "Having laughed with Hamlet," explains Wright, "now his death was a loss to us"; his declaration of the game "pushed our capacity for empathy to its limits and put us all on a theatrical knife-edge" (47).5 Wright concludes his chapter on "Theatre as Game" with an extended discussion of Mark Rylance's 1999 performance as Cleopatra at the Globe, a performance in which "[t]he game was clear right from the start" but which, like Wright's example from Hamlet, "touched tragedy" as the play neared its climax (95). As Rylance/Cleopatra shared his/her grief at Antony's death with the audience, says Wright, the performance "flipped from the 'there and then' to the 'here and now' and, for a few moments at least, this was for real" (96).
Wright's examples hint at the liminality of the metatheatrical stage figure. The reader will notice that this article has drawn on various hybrid constructions to describe the stage figure who speaks to the audience—"actor/character," "Rylance/Cleopatra," and in the case of the latter, "his/her." This is because the stage figure who "declares the game," crossing from the there-and-then to the here-and-now and back again, is tangling and confusing the two matrices to such an extent that he or she sometimes becomes hard to locate in one or the other. McConachie analyses the scope for repeated "blending" and "unblending" in the role of Viola in Twelfth Night, arguing that Shakespeare prompted his original audiences "to unblend the boy actor/Viola/Cesario mix by reminding them both of the dramatic masquerade and the theatrical casting convention at strategic moments in the play" (139). As I have already argued, I am not sure that audiences would have needed "reminding" of the male actor—they would not have forgotten him—but I agree with McConachie that this role is structured in order to repeatedly shift the nature of the relationship between the there-and-then of the cross-dressed female character and the here-and-now of the male actor. When a boy actor dressed in masculine clothing says the line "As I am man" (2.2.36), the audience whom he is addressing may well for a moment hear a stage figure speaking from the here-and-now of the player. When the same speaker qualifies this by continuing, "My state is desperate for my master's love" (2.2.37), the line [End Page 31] becomes both an acknowledgment of the character's disguise on a fictional level and, perhaps, of forbidden homosexual desire on a metafictional one. The next line, "As I am woman" (2.2.38), refigures the speaker as female and fictional.
McConachie describes this speech as "intentionally playful and dizzying" (140). His metaphor recalls the description by New York Times critic Ben Brantley of Viola's next scene in Carroll's 2002 production: "Mr. Carroll stretches the scene to the breaking point, so that different levels of perception swim in and out of focus. It is sexy, uncomfortable and highly disorienting" ("Boys Will Be Girls").6 In Carroll's all-male production, Act 2 Scene 4 contained a constantly-shifting relationship between the matrices of "Now and Here" and "Then and There." As Orsino and Viola/Cesario sat on a bench together to listen to Feste's song, Orsino suddenly clasped his page's hand; the two turned inwards slightly, making and then breaking eye contact. At the end of the song, Orsino demanded they be left alone, and as their subsequent conversation became more intimate and impassioned, Viola/Cesario seemed on the verge of revealing something:
Ay, but I know—ORSINO
What dost thou know?(2.4.103–4)
James C. Bulman has perceptively unpicked the simultaneous meanings that were in play during this moment of theatrical danger:
If Cesario is taken for a woman in disguise, her interrupted line ("Ay, but I know—") becomes almost an accidental revelation of her female identity, and the hypothetical condition ("were I a woman" ) a poignant confession of her scarcely hidden passion for the man whom she is addressing. If Cesario is taken for a man, however, then the condition and the confession—"it might be, perhaps, were I a woman / I should [love] your lordship" (108–109)—reveal something more unnerving for both of them: a love that dare not speak its name.(236)
As Bulman notes, with a male actor (Michael Brown) as Viola, "[b]oth interpretive options were in play" in this production (236). But this was not an interpretive choice, with either option available for spectators depending on their inclination—both meanings presented themselves at once, simultaneously and irreconcilably. As Viola/Cesario went on to reveal that "I am all the daughters of my father's house, / And all the brothers too" (2.4.120–1), she/he began to weep, at which Orsino took his page into a comforting embrace. The clinch lasted a little too long, and the two seemed on the verge of kissing, before Viola/Cesario [End Page 32] broke away. Different gender, sexual, fictional and metafictional identities were repeatedly blending, colliding and overlapping in this scene in many sorts of creative and bisociative ways. For Catherine Silverstone, "the gap created by the 'not-quite' kiss" allowed the production to admit "the possibility of a range of desires and sexual identifications between actors, characters and spectators which do not invite easy categorization as homo- or heteroerotic" (41).
I would like to conclude by returning to the question which forms the title of this article: are Shakespeare's plays always metatheatrical? In one sense, the answer would seem to be "yes": as we have seen, spectators are regularly treated as being in on the game by being directly addressed by the plays' characters. But in the model I suggest, metatheatricality cannot be a constant state: it is always the result of a shift in the ways in which the two planes relate to each other. If a play addresses its audience in the same mode from start to finish, as in some medieval drama, for example, or modern pantomime, we do not have metatheatre but a stable convention, a particular and constant mode of bisociation. Metatheatre, I argue, emerges when an element of the "Then and There" starts to interact with the "Now and Here" in a way that it previously had not, playfully rearranging the rules by which those two planes had been co-existing. If we think of metatheatre as a game of bisociative collisions between the "Then and There" of the story and the "Now and Here" of performance, it becomes clear that Shakespeare's plays are not always metatheatrical, and that staging them according to the principles of realism can in fact efface their metatheatrical potential altogether. Metatheatre is a game that is, in many cases, invited by the text, but one that can be played only in performance.
1. Tiffany Stern makes a similar argument, from a rather different perspective, in "This Wide and Universal Theatre."
2. Swiontek's work is not well-known in Anglophone scholarship and has yet to be fully translated into English. I am therefore indebted to the translation and summary of Swiontek's work on metatheatre provided by Jenn Stephenson in a 2006 issue of Journal of Dramatic Theory and Criticism.
3. I directed this production for The Pantaloons theater company; it opened at the Blackfriars Arts Centre, Boston, and toured to 31 venues in England and Wales over October and November 2015.
4. Tom Bishop has identified a number of Shakespearean scenes that work along similar game-like lines, among them Richard and Buckingham's theatrical one-upmanship in Act 3 of Richard III, the gulling scenes of Much Ado About Nothing, and the "clifftop" scene in King Lear. [End Page 33]
5. Bridget Escolme describes a comparable dynamic in Mark Rylance's performance as Hamlet at the Globe in 2000, in which the character's gradual withdrawal from direct audience address constituted a "bereavement for the spectator" (73). Indeed, it is worth noting that numerous thinkers cited in this essay—Abel, Koestler, McConachie, Stephenson, Swiontek, Wright—cite Hamlet in order to illustrate theater's doubleness. This may be simply because it is a famous play with which a general reader is likely to be familiar, or because of the play's own numerous reflections on dramatic art, but I suspect it may also be a tacit acknowledgement of the ways in which the play's dramatic effects often hinge upon the protagonist's fluctuating awareness of the presence of the audience. It may be that there is something especially bisociative about this play, with its protagonist who seems painfully aware that he is not the hero of revenge tragedy that he—and, in some performances, he seems to think, we—would like himself to be (see Escolme, 62–73).
6. Brantley's notion of a scene being pushed "to the breaking point" echoes my own description, above, of the metatheatrical games of Malvolio's gulling scene. To me, this "breaking point" suggests not the shattering of theatrical illusion (what Stephenson calls "the traditional boundary model" of metatheatre, 118), but the shattering of the internal consistency of the fictional plane—the point at which the planes of fiction and reality become so entangled that the stage action ceases to be legible as the representation of a fictional world at all. Metatheatre tends, playfully, to approach this point, even to defer it, but the examples discussed in this article do not, I think, quite go beyond it; performances that do so might be considered with reference to Hans-Thies Lehmann's concept of "postdramatic theatre," but this is not the space for that discussion. On the relationship between Shakespearean metatheatre and the postdramatic, see Robert Shaughnessy's article in this special issue.