Duke Bluebeard’s Castle, the lone opera by Béla Bartók, was begun in 1911 and premiered in 1918. The piece is barely longer than an hour, and usually appears on a double bill with another short opera. It has only three characters: a Bard who recites a brief prologue and then disappears, Duke Bluebeard, and the Duke’s new wife, Judith. Bluebeard brings Judith to his castle. It is dark, and Judith wants to let the light in. The dark castle has seven doors, and Judith forces Bluebeard to give her the keys to each one. The first door opens on a torture chamber, the second on an armory, and the third on a glittering treasury, wondrous and strange. There Judith notices Bluebeard’s finest crown is stained with blood. Quickly she opens the fourth door. It reveals a garden in bloom with lilies, carnations, and roses. But when Judith pulls up one of the roses, she finds its roots tipped with blood, for the plants have been drinking blood from the ground. Judith is undeterred by these phantasms, and demands that Bluebeard open the fifth door. Behind the fifth door lies his Kingdom in all its majesty. Light pours into the fortress, though a bloodstained cloud lurks in the distance.

Don’t open any more doors, Bluebeard asks. Open the next door, Judith enjoins, and she gets her way. It gives onto a lake of tears. Open the final door, says Judith. Don’t, don’t, says Bluebeard. But at the end of the opera the final door opens. Behind it are Bluebeard’s former wives, though they have not been murdered, as rumor had had it. They are still alive, trapped in his castle, frozen in a great mass of jewels. Judith is astonished by their beauty. Bluebeard explains that he met the first at dawn, the second at noon, the last at twilight. Bluebeard tells Judith that she is the most beautiful of them all, the night to their day. Before she knows it, she is being crowned with the crown of the night, crushed by a mantle of diamonds. She becomes the last wife, the end of the diurnal cycle, locked away in the room like the others. There is darkness everywhere, as before.


Duke Bluebeard’s Castle, the stage play by Béla Balázs on which the opera is based, was completed in 1910, and dedicated to Bartók and the composer Kodály, perhaps in the hopes that one of them would make an opera from it. It is a symbolist play, heavily influenced by Maeterlinck, in which a woman tries to open the doors to a man’s soul. “Bluebeard’s castle is not a realistic stone castle,” Balázs wrote. “The castle is his soul. It is lonely, dark, and secretive: the castle of locked doors.” He intended the major stage effects to be streams of light issuing from the seven doors, each a different color: blood-red, yellow-green, gold, blue-green, white, darkening white, moon-like silver. Sometimes he wrote stage directions like, “The door opens silently and reveals a blood-red rectangle, like a wound, in the wall.” Sometimes Judith remarks that the castle is crying, sweating, sighing. Bluebeard tells Judith that the castle’s foundations tremble and shiver with pleasure.


Balázs had listed four characters at the beginning of his play: the Bard, Bluebeard, Judith, and the Castle. Bartók, when he set the play to music, used almost the entire text word-for-word, but he cut the Castle from the list of characters.

What’s the point of the Castle? Balázs thought it was alive. Bartók did not, though he does at one point in the score require it to sigh. Perhaps he thought that, in an opera, the aliveness of the setting needed no acknowledgement.


Still from Georges Méliès's 1901 silent film   Barbe-bleue

Still from Georges Méliès's 1901 silent film Barbe-bleue

Bartók’s music is intended to suffuse—like perfume, or light. This last comparison is made explicit, since both Balázs and Bartók were emphatic about using great long streams of light, pouring forth from each door, as the only major stage effect. And each stream of light has its own distinct timbre and key. F♯ is Bartók’s code for darkness, and it begins and ends the piece. C Major—a tritone away from F♯, which is to say, its opposite in musical pitch space—is his key marker for pure, white light, and it radiates from the fifth door, in the middle of the opera when Bluebeard’s kingdom is revealed. (In the 1988 BBC film version by Leslie Megahey, a miniature landscape rises up out of the floor like the setting for a huge model train-set.) Each door is its own tone poem, evocative and programmatic. The clatter of the xylophone announces the torture-chamber. A solo violin, ravishing but also corrupting, courses through the golden treasury, a snake gliding over a shimmering, tremulous background of subdued glitter and sparkle. The garden, meanwhile, is depicted with harps and tremolo strings, and a handsome, stately horn. The libretto also suggests another image for suffusion: Bartók’s music must cover the whole theater in tears and blood.


Perhaps this is why Bartók took for granted that his opera’s setting was alive, and didn’t need to be mentioned in the dramatis personae—all the more so since the setting and main character were nearly synonymous. Consider an example closer to home: the way music functions in film. I am thinking especially of those scenes when a character is shown walking down the street or preparing for battle, accompanied by an overwhelming soundtrack of non-diegetic music. What this signals to us is that, at that moment, the character’s subjectivity has completely saturated everything surrounding him. His living presence has filled out the limits of the world. This happens usually in moments of exaltation or transformation: when the character is in love, in control, or, during a montage, gradually mastering the skills that will prepare him for the final showdown. It is the epic mode in movies, the moment when character and reality are in alignment, and it was lifted directly from the practice of opera and ballet. Duke Bluebeard’s Castle uses this trope from start to finish, projecting the protagonist’s aliveness onto the very walls.

(Clothing retailers have borrowed this, too. When you walk into an Abercrombie store, the all-pervading bass is meant to make you feel fully co-extensive with the store and, ideally, everything in it. It’s no accident that the all-pervading music—sometimes called the soundtrack to your life—is accompanied by great waves of perfume, also intended to suffuse the air with a single continuous presence.)


"Lake of tears" scene from Michael Powell's 1963 television adaptation of   Bartók’s   Duke Bluebeard’s Castle

"Lake of tears" scene from Michael Powell's 1963 television adaptation of Bartók’s Duke Bluebeard’s Castle

Bartók’s music in Duke Bluebeard’s Castle is directly, almost naively atmospheric. He wrote at a period in history when that kind of music, and that kind of musical description, was going out of fashion. In the early 20th century there came a fundamental re-working of the relationship of words to music. You could not say anymore, This music reminds me of a deer running through a forest, and that part there is a rushing stream, and this motif is a hunter coming to kill the deer. In an equal and opposite way, a composer found it harder to write music, like Berlioz once had done, which had a directly programmatic intention: this motif is the deer running through a forest, and this the rushing stream. But Bartók wrote an opera in which exactly this is the fundamental structural principle: one passage is bristling spears, another passage is blooming flowers, yet another (beloved by the composer György Kurtág) is the lake of tears.


Some of this is due to the very nature of opera. It is very hard to escape the musical description of reality in a composition with a story and a setting. If the opera takes place somewhere exotic, it can be difficult to resist adding some local color to the score. If fate is at hand, there must be fateful music. There must be tender music for the lovers’ duet. The genre, as a simple matter of form, requires music and events to walk hand-in-hand. (You can maybe get away from this, but you’d have to do a lot of fancy footwork.)


Bartók, though, had a second reason to write music that was tied to the world: as a response to his pioneering research into folk music. In one of his essays on folk music he writes that “a folk melody is like a living creature: it changes minute by minute, moment by moment.” By this he meant that variations in how a performer renders a folk melody should not be understood merely as mistakes; they were a necessary part of a tradition in flux. The music is alive because it responds flexibly to the incredible complexities of its culture. For this reason, Bartók writes,

the ideal folklorist possesses an erudition that is virtually encyclopedic. Knowledge of linguistics and phonetics is necessary in order to perceive and record the most subtle dialectical pronunciation; he must be a choreographer to describe accurately the interconnections of music and dance; only a general knowledge of folklore permits him to determine in minute detail the relationships between music and customs; without sociological preparation he would be incapable of checking the influence on folk music that is exerted as a result of changes which now and then disturb the collective life of the village.

Folk music is part of the organic whole of a native culture, a culture which involves dance, customs, landscape, settlement, mores, violations of mores, myth. It is elastic and flexible and highly responsive, but situated within a totality. The ensemble of folk culture, you might say, resembles an opera. Or, anyway, opera bears some resemblance to the totality of a folk culture, at least as Bartók describes it.


For too long we’ve considered opera as only music and text. For many hundreds of years a somewhat facetious debate was carried on as to which was more important. Now it seems music has won this debate, and musicians often voice their disapproval of a given operatic production by saying that the direction must serve the music.


Anonymous 19th-century newspaper adaptation of the Bluebeard tale, Metz, France

Anonymous 19th-century newspaper adaptation of the Bluebeard tale, Metz, France

But the terms of the debate are wrong. Opera is not only about music and text. If it were only music and text, then it would be indistinguishable from a song cycle, or a cantata. It is also about some third element which synthesizes music and text, though it may actually, in some sense, precede them. This third thing is mostly closely identifiable with the production, the staging, the design, the costumes. But that is an inadequate description. For it is actually much closer, at least in the ideal, to Bartók’s description of the folk music context. It is the sense that music and language only join together if they exist as part of the same three-dimensional world, if they inhabit a space already marked out and amplified by the structures of culture.


The critic Jean Starobinski has suggested that opera is an art form tied to a longing for a Golden Age, in which all aspects of life were fully integrated. The earliest practitioners of opera projected this image of a purely harmonious world onto Greek tragedy, which they understood to be not so much a synthesis of the arts as a kind of medium that didn’t need any form of differentiation into music or poetry. With their elaborate pageants based on the story of Orpheus, the disciples of the Florentine Camerata thought they could resurrect ancient tragedy. Wagner’s Gesamtkunstwerk also aspired to be not merely a fusion of the arts, but Art without any distinction. Otherwise he would have called it—and German is a language that might allow this—musikalisches Sprachbildwerk.


Another important source for opera’s founders, less often invoked than the pseudo-Grecian pageants of the Renaissance, was the Catholic Mass. In the Mass words and music are paired together by the mysterious third element, a production of sorts, involving costumes and gestures and a narrative arc. But what is being established, or reaffirmed, is the physical space of culture, a replication on the human scale of the three-dimensional nature of the physical world. The staging of the Mass—not only the movements on the altar, but also the positioning of the worshippers and the choir and the decorations on the walls of the church: all of this delimits a space. The text gives the space its meaning. And the music—that full sonic blanket that fills up everything like a perfume—provides a thickening agent, for no corner of the space can escape its permeation (accompanied here, of course, by that other permeator, incense). The earliest operas are all about stories of resurrection and transfiguration: Orpheus, Daphne, Armida. It may not be a coincidence that, although always expressed in the idiom of classical antiquity, they are very close to the main narrative of the Christian Mass, the resurrection and transfiguration of Jesus.


Bálazs and Bartók were not so heavily invested in this third thing, which might be called artistic or cultural space-time (though Bálazs did originally call Duke Bluebeard’s Castle a “Mystery Play”). They intended it to be presented very sparsely, as a drama between two figures engaged in a fatal and inescapable dance. The only exception was, again, the lighting, which had come to dominate experimental theater in the period. When each door opened, the stage was to be suffused with light. At the beginning and end, there was to be complete darkness.


Still from  Blaubart , Pina Bausch's 1977 adaptation of  Bartók's opera

Still from Blaubart, Pina Bausch's 1977 adaptation of Bartók's opera

Most productions of Duke Bluebeard’s Castle have stayed true to its creators’ intentions. I once saw a student production in a church basement in which nothing happened until the very end, when a cupboard swung open and three masks, symbolizing the three wives, bounced up and down on strings.

But the recent staging at the Metropolitan Opera, directed by Mariusz Treliński, wasn’t much more than that, either. A bare stage, cold and dark, a few video projections of white tiles and a pair of eyes, a row of kitchen knives, a shower-stall, a bathtub full of jewels. But mostly just Judith and Bluebeard, moodily interacting as if they were characters in a taut, naturalistic stage play.

But Duke Bluebeard’s Castle, short as it is, is not a taut, psychological stage play. Nor is it a poem, nor a symphony. If it were a poem, the splendor of the language would be enough to bring the piece to fruition. We would be overwhelmed by the words and their conjuring of diamonds and lilies and reservoirs of tears. If it were a symphony, the music itself would be enough, suggesting an abstract field of affective registers. If it were a stage play, the psychological conflict between the two actors on stage might be sufficient. But in an opera, these things are not sufficient. The slowness and strangeness of opera require something very different.


Opera, like many great theatrical traditions, sits uncomfortably with a theory of art based on imitation. The modern Western theatrical canon, exemplified by playwrights like Shakespeare and Ibsen, is an outlier in its commitment to what we perceive as a certain kind of realism. Most theatrical styles the world over are committed to extreme stylization. Opera, with its singing and stomping around and extended hand-gestures and often perversely slow pacing, fits into this broad category, alongside the Noh theater of Japan and Kathakali theater of Southern India, among others. There is an important paradox here. Theater is potentially the most realistic of all art forms. Its primary medium is the human body, moving through space and time—exactly the same thing people do every day in the real world. So to compensate, theater tends towards extreme distance from the real world. The body acquires new modes of movement estranged from walking and running; the voice acquires a declamatory style far removed from its normal range.


But there is a danger in that complete estrangement of body and voice. The end-result risks becoming utterly unrecognizable, the texture of reality completely alien, like something from science fiction. Many non-Western theatrical genres provide a solution, and that solution corresponds to Bartók’s description of folk culture. They situate artistic production within a thick and highly stable web of conventions, which provide a surface upon which the drama plays out. They are a star-map, a users’ manual, a glossary. More than that, they are a filler, a thickener of the world: they crowd the stage with their invisible presence, borne unto the air by music and incense. For this reason, many ancient theatrical traditions have little use for set-design, and present a very sparse and schematic stage: the imaginary world of the play (what it looks, feels, tastes like) is already common knowledge, shared by actors and audience. This is what it means to cry culturally (in Noh: drop your hand in front of your face); this is what it means to bawl culturally (in Noh: drop both hands at once).


Opera is an art form as hopelessly stylized as Noh or Kathakali. In their splendidly deadpan history of opera, Roger Parker and Carolyn Abbate put it like this: “Opera is in a basic sense not realistic—operatic characters go about their business singing rather than speaking.” Elsewhere, they elaborate only slightly: “That’s it. That’s opera. Just a lot of people in costumes falling in love and dying.”

One of illustrator Kay Nielsen's  Bluebeard    drawings, 1913

One of illustrator Kay Nielsen's Bluebeard drawings, 1913

But opera, though structured like Bartók’s definition of folk culture, also participates fully in Western high art’s commitment to novelty, revolution, and innovation. And herein lies a problem, a second paradox. How do you make legible an art form that is at once stylized, and yet is always revising the code that translates the stylized into the everyday? Imagine a religion which every week changed most of the gestures and much of the language of its ceremony, but still expected the audience to follow it, to understand it. (Not a religion that did away with ceremony, like Quakerism, but one which produced every week a whole new set of robes, gestures, songs, chants, and flavors of incense, all as complex and as inscrutable as the ones from the week before.) We would probably not call that a religion. We would say, precisely because of its bizarre mixture of ritual (which hinges on repetition) and innovation (which relies on singularity) that it was, by our recognizably Western and modern definition, art.

One of the main reasons why Wagner is the central figure in the history of opera is because he understood this problem. He realized that a new kind of opera meant creating an entirely new apparatus surrounding it, a thick web of conventions, traditions, signs, gestures, and designs. He saw that opera resembled a whole culture in miniature, and that opera always brought that whole culture, its own little planet, onto the stage. Wagner codified that culture, tried to put a stop to any further, unchecked evolution. He wanted to regulate every aspect of it, not just music and text, but this third element, the distribution of space-time, the dimensions of the world. When Wagner laid the groundwork for a new kind of music drama, he was trying to produce a Unified Field Theory of opera. The idea was not just to write new operas, but to lay down the laws of physics for the universe of New Opera: to provide an explanation that linked everything within the opera to everything else. Wagner hoped that this complete universe of opera would be as long-lasting and unchanging as Noh, and that its center and circumference would be Bayreuth, the theater built exactly to his specifications.


An American opera-lover named Kate Douglas Wiggin understood this very well, and wrote about it in a short satire from 1914 called Bluebeard: A Musical Fantasy. The book pretends to be the transcript of a lecture delivered before performances of a newly discovered Wagner opera, based on the life of Bluebeard. (Bluebeard was, in fact, a popular figure among the fin-de-siècle decadents: he was the subject of Paul Dukas’ opera Ariane et Barbe-bleue, a goofy operetta by Offenbach, a Satanist novel by Joris-Karl Huysmans, and a number of tongue-in-cheek short stories, one of which was later made into a movie starring Gary Cooper.)

Wiggin explains that a housemaid working at a villa where Wagner had once stayed noticed a halo shining over a certain bed, the “said halo absolutely refusing to remove itself when attacked with a feather-duster.” The halo, it turned out, was pointing to the manuscript of a glorious unknown opera in a hatbox tucked under the mattress. Now the work is to be performed before the New York public, but like all of Wagner’s operas, it requires initiation. “One could no longer leave one’s brains with one’s hat in the coat-room when the ‘Nibelungen Ring’ appeared!” she writes.

Learned critics, pitifully comprehending the fathomless ignorance of the people, began to give lectures on the ‘Ring’ to large audiences, mostly of ladies, through whom in course of time a certain amount of information percolated and reached the husbands—the somewhat circuitous, but only possible method by which aesthetic knowledge can be conveyed to the American male.

And so, too, with the new Blaubart opera, which must be explained in full beforehand. Wiggin takes her imaginary audience on a helpful tour through the leitmotivs: the Immer-wieder-heirathen Motiv (the always-getting-remarried motif), the Verdammungs Motiv (the damnation motif: just one note, 14 lines below the bass clef), the Ausgespielt Motiv (the all-played-out motif: three bars of silence).

Wiggin mocks two things about Wagnerian opera: its imperiousness, its attempt to establish a monopoly over meaning; and more interestingly, its odd naiveté, the leitmotiv system’s assumption of a one-to-one correspondence between music and meaning, or between meaning and representation in general. She is mocking the fact that this preeminently modern art form—at least for its day—harkens back to a much older, even primitive, idea of what art should be.


Opera’s difficulty as an art form is in part due to this naïve relation to meaning. Because of its profound alienation from the experience of daily life, opera needs a stable code to explain itself at the primary level: it needs a consistent account of its own reality, which makes all the strange things in it cohere. And because of the continuous transformation and energy of the genre (even now that the energy is largely devoted to the reinterpretation of earlier works), that complete account needs to be rearticulated every time an opera is staged. Bluebeard’s Castle cannot be taken for granted as a psychological symbol: it must appear in a thousand details.

This doesn’t mean at all that opera-design should be more realistic or feature convincing special effects—in fact, quite the opposite. Such measures inevitably fail in their attempt to annex opera to the standards of the plausible by which we live in our everyday, not-sung world. Instead, opera-design should be full of detail, busy, thick, in the sense that Clifford Geertz used the word to describe the density of context in cultural phenomena. An opera production must create a vocabulary of movement, of interaction, of scales of gesturing and reacting, all complex enough that they no longer seem random or unmotivated. Now they become part of a system, governed by their own laws of physics. They function like a complete, foreign society revealing itself to an attentive, though maybe bewildered, anthropologist.

Many contemporary productions in the United States make the mistake of assuming that coherence is a purely stylistic matter. They assume that if one single aesthetic reigns, or if a place and time (fascist Spain, or roaring twenties New York) is recreated meticulously enough, then that will suffice. But it doesn’t. Rather, the entire vocabulary of movement, action, and experience needs to unfold on a commensurate scale. If one gesture is outsized, then every gesture must be outsized. The finest productions in experimental Regie-theater achieve this. They usher you into a reality whose strangeness is so complete that opera’s inherent strangeness no longer seems like strangeness. Instead, it makes perfect sense, reading as a natural consequence of the universe in which it exists.


Armory scene from Powell's adaptation of     Duke Bluebeard’s Castle

Armory scene from Powell's adaptation of Duke Bluebeard’s Castle

The British filmmaker Michael Powell, infamously blacklisted for his 1960 thriller Peeping Tom, was a master of such altered worlds. In addition to a stunning Tales of Hoffmann, he produced an exceptionally strange Duke Bluebeard’s Castle for West German Television in 1963. Despite its low-budget production values, the conjuring of another world is splendid. This Judith and Bluebeard live on an alien planet from a B movie backlot. The scenes move seamlessly and irreverently between painted drops and full sets. Their planet is richly detailed but self-contained. It resembles an extra-terrestrial coral reef, though one made entirely by an intergalactic origami master. When Judith unveils the armory, she is suddenly surrounded by swords floating cheaply through the air, pointing at her like so many angry muppets. Then, coyly, they point away, and again, suddenly but not very threateningly, they point back. The literalness of this depiction should not be mocked, because it’s literal only in a world where “literal” means something completely different than ours, where the literal is far stranger and more distant than any figuration.


A literal world that is stranger and more distant than any figuration—opera shares this with fairy tales. That is not the same as the supernatural; neither fairy tales nor operas require that. Charles Perrault’s famous version of the Bluebeard story, the first to be written down in the West, features almost no magic at all, except for the bloody key which cannot be washed off, indicating that Bluebeard’s wife has visited the forbidden chamber where the corpses of his earlier wives are hung. But at every turn the tale situates itself firmly in a world where the rules of motion and language and behavior are profoundly different from our own. It does so first with its ritual opening: il était une fois—once upon a time. Then it maintains a continuous and unbroken thread of extremity and exaggeration: Bluebeard, we learn, was impossibly rich; but we also learn he was so ugly that there was not a woman in the world who would marry him. (Never mind that three already have, and another one is about to.) No one would visit him because of his terrible blue beard, but his possessions were richer and more beautiful than anything else imaginable, which drew him admirers from far and wide.

Bluebeard warns his new wife that there is nothing she cannot expect from his anger if she opens the forbidden chamber. Nevertheless, her curiosity is absolutely unhindered, and she opens it as soon as she gets a chance. (She can’t even wait that long, running away during a party “without even considering that it was impolite to abandon her house-guests.”) Later, when she is about to be killed and her sister Anne goes up to the tower to scan the horizon for rescuers, the two women start speaking to each other in little poems, apparently ignoring the fact that Bluebeard is waiting below with his cutlass.

The brilliance of fairy tales, their atavistic pleasure, lies in this unembarrassed confidence in the other world, with its rules of deportment, its extremities, its ritual poems, its own laws of chemistry (for instance when it comes to the washing away of blood). Fairy tales fully accept these parameters, neither disguising nor justifying them. Everyone in the opera world today frets about how to make the art form more relevant. But they should take their lessons not from novels or television, but from folktales. Don’t make it new; don’t make it spectacular; don’t make it traditional, either. Make it strange, because strange is opera’s middle name.


But the world of folktales, strange as it is, has a consistency across time and space. We are initiated at an early age into this world, and its parameters remain quite distinctly defined in our minds. The world of opera, in contrast, has been a far more mutable and contested thing, its borders far more insecurely felt. For many people going to the opera feels like listening to someone else’s dream. It is the dullest thing imaginable, and it could only become exciting if they could somehow enter it, to inhabit the dream as their own.

One of a set of engravings by Winslow Homer for a Bluebeard novelty  tableau vivant , 1868

One of a set of engravings by Winslow Homer for a Bluebeard novelty tableau vivant, 1868

This was as true in Perrault’s day as in ours. His contemporary La Bruyère writes, “I don’t know how the opera, with its perfect music and regal expense, could succeed in boring me.” But he’s being disingenuous, since in fact he goes on to say that he knows exactly how. The problem is the one-sidedness of certain productions. They are barely anything more than concerts; marvelous stage effects and storytelling have been dismissed as being for children. Against this idea he protests: opera only gets its energy from the complete immersion of all the senses, from the creation of a sense of wonder. We cannot simply think about the lives and habits of others in an opera; we must feel completely swallowed by their experience. La Bruyère was intrigued by a festival held by the Prince of Condé where everything was conceived according to a plan: the design encompassed poetry, music, stage-machines, and even the theater itself, which was specially built for the occasion. The event was spoiled only by the incessant bragging of those who played a part in the collaboration; their egoism, we infer, ruined the apparent unity of the project.

Perrault himself loved the all-embracing nature of modern opera, and defended it in a fictional dialogue called “Critique de l’opéra.” In this short work a detractor named Aristippe argues that opera is not worth seeing because it is so much worse than Greek tragedy. A supporter and Perrault stand-in named Cléon counters that opera is good on its own merits, and even compares favorably to Greek tragedy. The detractor Aristippe objects that opera relies too much on the artificial, whereas Greek theater was perfectly natural. Cléon replies that the opera is not like a comedy or tragedy: in a comedy, realism is paramount; everything must be plausible. In tragedy, only select moments of the marvelous—divine intervention, prophecy—are possible. But in an opera, he explains, the plausible is the least interesting, and least successful, part of the performance. The marvelous and unbelievable must be given pride of place. Machines, wonders, exaggerated conventions, outrageous gestures—these, he suggests, are not opera’s flaws. They are opera’s defining features.


Part of the reason why opera may be in such precipitous decline as an art form is because we have forgotten this fact. Opera cannot attempt to be true to life, to show us a mirror of ourselves. Neither should it have special effects which try to make something implausible seem realistic. It must glory in its artificiality, and its artificiality must be fully self-explanatory; it must have its own system of justification. For its own survival opera must reject all the platitudes about art teaching us to recognize who we are, to make us reflect on human nature. No one should go to a Verdi opera to learn how to love. No one should go to a Wagner opera to learn how to rule (people tried that in the 1930s, and it turned out disastrously). Opera can only teach us to be who we are not, to demand a complete transformation, in which the whole of experience undergoes a great estrangement. Only once we’ve stepped into the circle of transformation—a kind of spiritual transvestitism—can we learn something. Opera says: you must believe this is the way the world is, even though it obviously isn’t like that at all. And this is why it mirrors a culture, the total sum of a society’s reinvention of the world. And this is why it is at once the most complete and most impossible of art forms.


What if Perrault and La Bruyère collaborated on a production of Duke Bluebeard’s Castle? What if they convinced the Prince of Condé to mount it during a six-day festival in honor of the Dauphin at Chantilly? What if they brought Bartók all the way from Budapest in a carriage and six just for the occasion? (“Why can’t I just take a train?” Bartók would ask, reasonably.)

Garden scene from Powell's adaptation of     Duke Bluebeard’s Castle

Garden scene from Powell's adaptation of Duke Bluebeard’s Castle

They would have to build their own theater in the gardens, of course, and, of course, it would have to be in the shape of a castle. On the stage might stand an immense machine, with wheels and cranks and ropes and pulleys. How like a Piranesi drawing it is! It has its own decorations: festoons of Spanish moss made from Chinese silk, and decrepit statuary from Egypt and Assyria. Huge chains criss-cross the stage, and in the center, a fountain of quicksilver plays quietly.

Judith and Bluebeard are dressed in outrageous baroque splendor: Bluebeard in an immense wig, Judith in a long train. Yet their castle is filled with all kinds of bewildering domestic objects. In between the complex gears and pumps of the machine are chess sets, ironing boards, a doll’s house waiting for an unborn child. Between the opening of each door, Judith and Bluebeard return to domestic tasks: paying bills, playing cards, sewing, serving lunch. There is no escape into the mundane because the mundane is wrapped in the fabric of the extraordinary.

And the doors? In the center of the great machine lies a bank of keyholes. Each time Judith gets another key she plugs it into a hole, and the machine whirrs to life, twisting and cranking and sighing. It brings down a door on a chain from the flyhouse above, which Judith pulls open. A lone and sprightly skeleton comes dancing out of the torture chamber. An army of swords drops on strings to adorn the armory. A parade of handsome men covered in jewels come filing out of the third door to signal the treasury. The same handsome men return bearing huge flowering branches to symbolize the garden. Wait and see what splendors are prepared for the kingdom, and the lake of tears! But certainly not a real kingdom or a real lake. Perhaps Perrault and La Bruyère take inspiration from the last shot of Andrei Tarkovsky’s film Nostalghia, in which a ruined church suddenly appears to contain within it a vast landscape, infinitely bigger than what might actually fit inside a collapsing nave.

The final key turned in the final lock brings down not a door but three beauteous women suspended on chains. Or are they giant puppets? Or are they the handsome men in drag? It is impossible to tell, but quite possible to tell that the confusion is part of the effect.

Matthew Spellberg studies the literature and anthropology of dreaming at Princeton. He also writes on opera and architecture. His work has most recently appeared in the Yale ReviewLos Angeles Review of Books, and Southwest Review.