VII

The early years of opera, from Rinuccini through Metastasio, could reasonably be considered the Age of the Poet. For many years the written words were regarded as more permanent, and were treated more like art, than the scores. Next came the Age of the Singer (these "ages," of course, overlap), described and satirized in Benedetto Marcello's "Il teatro alla moda" and chronicled by many other observers. Lasting through much of the eighteenth and early nineteenth centuries, this was the age of Baroque and bel canto opera, when the services of uniquely gifted singers (including castrati) were regarded as indispensable.

Around 1860, by sheer force of will and popularity, Verdi and Wagner were able to impose, or superimpose, what one might call the Age of the Composer. At long last, the person most responsible for what mattered in an opera was able to give orders regarding the libretto, the singers, the staging, and the conductors he wanted. Composers regularly conducted their own works as well, at least for important premières and gala events, which extended that authority even further.

As opera composers with this kind of power and popularity began to die out (the last two were Puccini and Strauss), dominion shifted again, and the Age of the Conductor was born. For a few decades in the first half of this century, important productions of opera—almost all of old operas now, because the repertory had begun to cease growing—were identified with, and in a major way created by, their conductors: Krauss's Salome , Furtwängler's Tristan , Toscanini's Traviata , Busch's Così .

Now a fifth age in the history of opera appears to be upon us: the Age of the Producer. During the past forty years, the nature and variety of opera productions around the world have altered to an astonishing degree. Traditional productions more or less like those our parents and grandparents saw still account for most of the thousands of performances given each year. But audiences in several major centers, some festivals, and a few smaller cities are now regularly confronted with assertively, even defiantly novel reinterpretations and reconceptions of standard repertory works. All of these are the creations of a type of opera/theatre professional virtually unknown less than a century ago.

Most older operagoers grew tip on visual and musical incarnations of popular operas little changed since the days of the later Wagner and Verdi, the days of Shaw and Ernest Newman—little changed, that is, since the days when electric lighting and three-dimensional sets first seriously altered staging techniques. If the libretto declared that an opera was set in seventeenth-century Spain, efforts were made to dress the performers and design the sets according to someone's idea, at least, of the costumes and buildings of seventeenth-century Spain. Depending on the resources available, village plazas, mountain crags, Egyptian temples, or Paris cafés identified in the text would be recreated either sketchily, with painted back-

OPERA LOVERS . Anonymous illustration from Scrici (John H. Swaby),  Physiology of the Opera  (1852).
Courtesy of the Institute for Studies in American Music, Brooklyn College.

drops and a few movable thrones, tables, altars, or rocks, or in painstakingly naturalistic detail. In the larger houses, where extravagance for its own sake was sometimes the rule, garrets and boudoirs grew to be sixty or eighty feet wide—a convention audiences rapidly learned to accept.

A few librettists and composers (notably Wagner and Verdi) left explicit instructions on how they wanted their stages to be set and their actors costumed, moved, and motivated. If a libretto called for, or the music indicated, a sword fight, a dragon, a ghost, a little table, or a chaste kiss, that's what audiences saw. Almost from the start, a few operas (The Magic Flute , Wagner's Ring, Pélleas et Mélisande ) seemed to invite stylizanon, even abstraction, in design, and occasionally they got it.

The beginnings of the current revolution in the worldwide staging of opera, in the years after World War II, arc usually traced to the work of Wieland and Wolfgang

Wagner (the composer's grandsons) at Bayreuth, to the work of Walter Felsenstein in East Berlin, and to parallel innovations on the nonmusical stage by people like Peter Hall in England. For many operagoers this "revolution" will be something they have read about rather than experienced. A few companies in the United States have dipped a toe into the New Wave, but the "majors" have experimented only diffidently or rarely: one could cite producer Goeran Järvefelt's Bluebeard's Castle/Erwartung double bill at the Met, Peter Sellars's Tannhäuser in Chicago, Nicholas Joel's Parsifal , Nikolaus Lehnhoff's Salome , and a number of productions by Jean-Pierre Ponnelle in San Francisco. Opera productions in Spain and Latin America are still for the most part solidly nineteenth century. There are innovators at work in Britain and France, but the most radical of them have tended to work outside the largest houses. La Scala in Milan and the Comunale in Florence have taken a few eccentric stabs at postmodern production, but most of Italy remains determinedly traditionalist. The home of avant-garde production ideas remains essentially where it began, in the German-speaking countries—although ideas travel far and fast in today's opera world, partly because German producers travel also.

I describe a number of the new-style, postmodern productions in the essays ahead, so I won't rehearse them here. Catalogues of revisionist outrages are included in books such as A. M. Nagler's Misdirection: Opera Production in the Twentieth Century (originally titled Malaise in der Oper ) of 1981 and Henry Pleasants's Opera in Crisis of 1989. Nagler (who seems to be dissatisfied with any alteration from opera as it was produced sixty years ago) briefly labels and denounces more than 200 "misdirections" of 10 selected operas since 1950 (and sometimes earlier), most of them in Germany. Pleasants instances 29 British productions from 1977 to 1988—17 of them from the more "innovative" English National Opera—and (by way of secondhand accounts) 18 from other countries, including the 1976 and 1988 Bayreuth Rings and Peter Sellars's versions of the Mozart-Da Ponte operas, on which I comment in the pages ahead. Although Nagler casts his net much more widely, many of the same "producer-kings" come under attack from both men: David Alden, Ruth Berghaus, Peter Brook, Patrice Chéreau, Walter Felsenstein, David Freeman, Götz Friedrich, Joachim Herz, Gocran Järvefelt, Harry Kupfer, Yuri Lyubimov, Jonathan Miller, Jean-Pierre Ponnelle, David Pountncy, Luca Ronconi, Ken Russell, Peter Sellars, Andrei Serban, Peter Stein, and Graham. Vick. Pleasants writes:

The root of all this presumptuous, arrogant, and often licentious mischief is the crisis of opera in this century, i.e., the stagnation of the repertoire and the scarcity of great singers. . . .^[15]

Unable to update the music. . . . [opera managers] have looked to a new breed of producer, mostly from theater and film, to sustain the illusion of vitality and continuity by updating and altering the staging in cynical violation of tradition and oblivious of, or indifferent to, the consequent stylistic anachronisms, aberrations, abominations, and—not to mince words about it—vandalism.

Although all of the productions denounced in these books (as well as new postwar directions in opera production generally) have had their ardent defenders, many, perhaps most, local and international music critics, along with many apoplectic writers of letters to editors, have echoed the outrage of Nagler and Pleasants. Blame is sometimes spread to conductors and singers as co-conspirators; but more often they are regarded with sympathy as hapless puppets, imprisoned by tyrant-producers in their unspeakable new productions.

The new stagings most often criticized are of several sorts:

1. Those in which an opera originally set by its authors in one time and place is moved by the producer to another time and place (e.g., shifting Rigoletto to New York or Wagner's Ring to the nineteenth century), even if this forces the libretto into occasional anachronisms or nonsense.

2. Those in which specifically sited operas are moved "out of the world" altogether into a science-fiction or apocalyptic future, or a surrealistic fantasyland of the producer's and designer's own creating;

3. Those in which a producer flatly ignores or defies the original authors' own directions, the apparent and traditionally accepted meaning of the libretto, and the traditionally accepted "meaning" of the music;

4. Those in which an opera composed at an earlier time is used to argue or express contemporary, often controversial ideas, to make modern social or political "statements" the original work was obviously not intended to make. (This, in fact, often serves as a producer's justification for alterations that fall trader categories 1, 2, and 3.)

The French critic Guy Verriest traces the birth of the independent opera producer and set designer to postwar Germany, and in particular to Walter Felsenstein's influential concept of Musiktheater . Felsenstein (in Verriest's formulation), infusing Wagner's Gesamtkunstwerk with Brecht's antibourgeois "epic theatre," granted absolute priority to the overall dramatic effect of an opera by means of a minutely conceived and (usually) very original staging. What bothers this critic is that Felsenstein and his successors typically drew their "dramatic truth" exclusively from the libretto—even from its literary or historical source—and betrayed the original by creating a new work in conformity with their own taste and temperament.

They [the producers] often ignore the music—which they generally don't understand; and nothing is more destructive to the comprehension of an opera. In each opera, we can judge whether the musical and verbal messages reinforce or contradict one another. In every case, we must take the music as our guide and try to understand the composer 's conception of the text. It is this conception, and not the naked text, that defines the ultimate significance of the work.^[16]

Although original scores (or some modern scholar's reconstruction of them) tend to be more respected than in centuries past, producers, conductors, and singers are still criticized as well for cuts, interpolations, and transpositions. In general, however—except for Baroque opera—the music and words in these novel productions tend to be presented as written, or as rewritten, or as carefully reconstructed, or as "traditionally" edited and cut. (Many popular operas exist in several versions, each defensible oil diffirent terms.) What is being ignored or defied in postmodern productions, its critics argue, is not so much the score as the tradition of onstage performance practice—which is to say, the traditions of the bigger opera houses between 1880 and 1940—and the traditional "readings" or interpretations of text and score.

The range of works attacked by Nagler, Pleasants, Verriest, and like-minded operatic conservatives includes everything from productions that deviate only slightly from topographical realism, or the legacy of a librettist's minutest stage directions, to the most radical deconstructions of the original text. Because of this range, I find it impossible to declare the new breed of producers "right" or "wrong"; one would have to argue the case production by production. In any event, I would feel ill at ease debating the merits of any production I hadn't seen and heard myself. In my own experiences of the lyric (as of the spoken) theatre, I have found some extremely novel readings and interpretations of classics to work very well, to enlarge considerably my sense of the original. Others have struck me as reductionist and perverse.

Whatever one may think about such deformations of traditional operas—Bellini's Norma in a tank, Monteverdi's Ulysses in Vietnam—in one important respect the conservative critics and polemicists are right. Since 1950, the director or producer of an opera has become a far more significant figure generally than he was in the years before World War II.^[17] Despite the occasional and isolated emergence of a dominating, "heroic" opera producer, like Alfred Roller or Max Reinhardt in the early years of this century, the very idea of the nonmusical producer of

an opera becoming a creative force as important as, or even more important than, the singers or conductor of an opera is a phenomenon very much of our time.

A few conductors, like Herbert von Karajan, have served as their own producers. Some producers, like Franco Zeffirelli and Jean-Pierre Ponnelle, have taken on the role of designer as well. In extreme cases—which are themselves exemplary of the Age of the Producer—the producer assumes precedence over composer and librettist. Peter Brook's popular Paris entertainment of 1981 called The Tragedy of Carmen was a rewriting and condensation created by Brook out of Bizet's opera. In the case of John Adams's Nixon in China and The Death of Klinghoffer , it was the producer, Peter Sellars, who first conceived the idea of the opera, then assembled a sympathetic team of composer, librettist, and designer.

Two centuries ago, such a czarlike role would have been assumed not by an independent stage producer or by the composer but by the librettist. "It is therefore the poet's duty," wrote Francesco Algarotti in 1755, "as chief engineer of the undertaking, to give directions to the dancers, the machinists, the painters; nay even to those who are entrusted with the care of the wardrobe and dressing the performers. The poet is to carry in his mind a comprehensive view of the whole of the drama; because those parts which are not the productions of his pen ought to flow from the dictates of his actuating judgment, which is to give being and movement to the whole." Daniel Heartz has demonstrated that distinguished eighteenth-century poet-librettists like Metastasio, Goldoni, and Da Ponte all gave specific instructions for stage positions and movement, oversaw rehearsals, and "instruct[ed] the actors in the truth of the action and of the expression."

Critics unhappy about the new Age of the Producer will insist that these non-musical interpreters, these mere painters of scenery and shifters-about of singers, have assumed supremacy over composers and librettists in hundreds of cases less notoriously imperial than those of Peter Sellars and Peter Brook. Like old-line Catholic theologians, such critics refer to what they regard as the joint and equal authority of Scripture and Tradition: Scripture being what composers and librettists have in fact written down (which includes not only explicit or implicit stage directions in the libretto but also letters, transcripts of rehearsal notes, and production books); Tradition being the way in which their operas have "always" been done. ("Always," as I say, tends to mean the way operas were done at the Teatro alla Scala, the Royal Opera at Covent Garden, the Vienna Staatsoper, the Paris Opera, the Bayreuth Festspielhaus, and the Metropolitan Opera, between the introduction of electric lights and three-dimensional sets and the beginning of World War II.)

In earlier days, the poet told singers where to stand and how to move. The composer prepared and usually conducted (from the keyboard) at least the première performances. The theatre's resident musical director assigned roles and supervised the singing. Later on, chorus-masters came in to deal with crowds and

stage movement in general, although (according to Wolfgang Hildesheimer), "there was no theory of stage direction in our sense of the term; action, gesture, mime were hardly synchronized; everyone did what he could; improvisation was substituted for rehearsal. . . . Actors were allowed to follow their interpretive inspiration of the moment."

Old singers' memoirs, from the days when most singers maintained a single home base all their lives, suggest that young members in a company learned expression from voice teachers and stage deportment from senior colleagues—out of which grew a tradition and a minimal need for stage directors as such. In many cities (particularly Paris and Milan) the design of opera sets developed into an elaborate and independent craft of its own, basically unrelated to concepts of acting or musical interpretation.

Today, the music critic typically addresses first the (new) production of an opera—which is presumed, for better or worse, to aspire to some sort of conceptual unity: Peter Hall's Così , not Bernard Haitink's; Jean-Pierre Ponnelle's Così , not John Pritchard's; Peter Sellars's Così , not Craig Smith's. What does the production look like, what do the actors do (and wear), what appears to be their character and motivation? How does this differ from other productions of the same opera? What, on the whole, does it (the production, not the opera) seem to "mean" or "say"? Does it roll along with or fight against, enlarge or reduce our previous conception of the work? All of this has become a fascinating, sometimes a maddening game for music critics (and for audiences) to play. It is one that they rarely had the opportunity or obligation to play in the past.

If the visible/theatrical production is in fact "traditional" (or tired, or very dull), or if the opera features a famous singer—especially one undertaking a new role—a critic may start at once with the voices. If Karl Böhm or Carlos Kleiber is making a rare and impressive appearance on the podium, the conductor's interpretation may still be addressed first. But the new 1976–1988 Ring productions at Bayreuth—conducted originally by Pierre Boulez, Georg Solti, and Daniel Barenboim respectively, each making use of some of the best-known Wagnerian singers in the world—are inevitably referred to today as Patrice Chéreau's Ring , Peter Hall's Ring , and Harry Kupfer's Ring .

Rarely, unless it is new to the audience, is an opera itself evaluated. For most critics today, the standard repertory is simply a "given." Joseph Kerman protests what he regards as this overturned set of priorities. Discourse in operatic criticism, he complains, is seldom about meaning but rather about peripheral topics, like "modern production methods."

In an ideal world, the order would perhaps be reversed. A critic would deal first with the value of the opera, then with the singing and conducting (focusing, of course, on dramatic values), and only finally, if at all, with the physical production. But when I once ventured to include my own opinion of Strauss's Die Frau ohne

Schatten —not one of my all-time favorite operas—in a London Times review, my editor asked me to keep my private judgments of repertory standards to myself. "Please confine your remarks to the new production," he instructed me. "We presume our readers have already formed their own opinions about the opera and aren't interested in knowing yours."

George Bernard Shaw was never backward in coming forward with his opinions of the operas he reviewed, however "established" in the repertory they may have seemed. Like Kerman, he obviously hoped he might help to disestablish one or two. But though he almost never referred to "producers" as such—except insofar as conductors like Thomas Beecham also took responsibility for productions—he did, as a conscientious man of the theatre himself, care a great deal about proper interpretation, realization, mise-en-scène. Before he mentioned singers at all (however famous) in his reviews, he tended to deal with the egregious weaknesses and errors that prevented most operas produced in London in the 1880s and 1890s from being fully realized on stage: the wretched acting (or nonacting), the total misreading of a composer's deeper meanings, the casual and foolish cuts and alterations made in the scores. Kerman, too, when working as a journalist-critic—though always ready to condemn any opera he dislikes—inevitably attends first to dramatic values, both in individual singers' performances and in productions as a whole.

I have said that "the right production" is the second essential ingredient for a successful operatic experience, that opera-as-written demands a proper production if it is to be completed and brought to life. What, by my lights, makes for a proper production? Given the repertory—and that is a very large given—what is it that I look for in a production?

I find myself applying many of the same standards to a production that Kerman applies to written scores. Where he looks for optimal dramatic coherence, I look for optimal theatrical coherence. This usually involves some sort of intellectual concept, ideally one drawn from the text, a concept that is tangible (or felt as evolving) throughout the opera, as we are experiencing it. This concept cannot be something arcane and imported, undiscernible except by way of a director's explanatory notes. In fact, many of the best directorial visions (in all forms of theatre) are impossible to express in words. Like a conductor's "interpretation" of a score—for scores do not play themselves—this involves a "reading," the discovery of an emotional and intellectual structure that permeates, vivifies, clarifies, and unifies everything we see and hear happening on stage. Ideally, the conductor's and producer's interpretations—of words and music, in both cases—would be worked out together. One would then sense in the opera house the excitement of a close and sympathetic collaboration.

Opera is, or at least can be, "total theatre" in the Artaudian sense: theatre to the

max. Anything less than a total recreation, therefore, a controlled and integrated production in which every clement contributes to a single coherent vision, seems to me inadequate. An indispensable subset of total theatre is dramatic conviction; it is far more indispensable, I believe, than "perfect" singing or "flawless" instrumental playing. I expect opera singers to be able to act, with their voices first, but ultimately with their whole bodies, no less than I do speaking actors.

With all respect for their extraordinary vocal abilities, I therefore find myself out of sympathy with singers like Joan Sutherland when she declares (Luciano Pavarotti has said essentially the same thing), "I think it is tile sound of staging that people want when they come to the opera. If they want a good dramatic performance, they should go to a straight play." There are those of us who think we have a right to expect both—even at an opera. If singers can't or don't want to learn to act, perhaps they should stick to the concert stage.

That much said, I prefer to back off from more specific prescriptions about the ingredients of "the right production." Although there are important clues for production written into the words and music of every opera, these clues are not orders, nor do they yield a single "right" production. The detailed instructions left by some composers (Wagner's for Parsifal and the Ring , Verdi's production books for seven of his operas, Berg's for Lulu , Bartók's for Bluebeard's Castle ) are certainly important for a producer to know; but they are not the last words, despite the "orders" of the original publishers.^[18] Productions today that try to follow the originals precisely, such as the 1982 Parma Forza , are likely to be of primarily academic and historicist interest, like exact imitations of old buildings.

Within the score and its libretto is encoded a key for production—or rather several keys, since music is such an equivocal means of communication.^[19] Studied carefully by a producer who either knows music well, or undertakes the task together with someone who does, the established text can provide not only the essential unifying concept; but also innumerable ideas for individual characterizations, movements, and gestures, as well as for settings, costumes, and lighting. The spectator should feel that the visual enactment blooms organically out of the text, even when (as in many contemporary productions of Shakespeare's The

Merchant of Venice and The Taming of the Shrew ) the production is actively commenting on or criticizing the text.

One can legitimately search for and make use of external clues for a production concept, beyond the more or less obvious ideas conveyed by the text, when dealing with any opera written before our time. The "modern" operas of Alban Berg, for example, no longer strike me as contemporary, any more than T. S. Eliot's The Wasteland , or a classic 1920s film. They seem to me to call out for a production concept that takes into account the distance we have traveled since they were written.

If this is true for operas of the 1920s, it is even more true for older works—which is to say, virtually all of the repertory standards. Unless one is aspiring toward a historicist revival, a study in the taste and ideas of another time (or nothing more creative than a harmless "concert in costume"), it seems appropriate, even desirable, to interpret and illuminate works of the past in the light of a number of external facts:

· what we know surely of the composer's own life and times (I find something both logical and revealing in performances of eighteenth-century opera seria—even when the operas are supposedly "set" in ancient Babylon or Greece—in eighteenth-century European court dress);

· what we know of the composer's own historical and literary sources (there are good clues for producing Bizet's Carmen , I believe, to be found in Prosper Merimée's short story);

· the traditional (1880–1940) image or performance style of the work, even if one elects to discard or work against it;

· the historical substance of the years that have passed between an opera's composition and today;

· the technical means at one's disposal, as well as the musical and vocal resources. Although I realize that this is not always possible at the international houses, where leading artists fly in and out, I greatly respect producers who build their conceptions around (while working with) the available performers, as composers themselves often did.

The very fact of the "archaism," the cultural distance from us of most repertory operas—and yet of their undeniably persisting musical/expressive power—invites the producer to acknowledge that archaism, to incarnate, not some image from 1787 or 1876 preserved, as it were, under a glass bell jar, but an onstage world that takes that distance (and the intervening history) into account. As Jonathan Miller has said:

The author or composer is not necessarily the best authority on his own meanings. This is not to say that he is slipshod, or does not understand his meaning, but simply that no writer, no author, no composer has total access to all his own meanings. . . . The producer can provide this insight. He is the bystander, the intelligent critic who is in a position, sometimes, to identify meanings that were not directly accessible to the composer, and to extract new meanings with the passing of time. . . . One constantly has to make allowances for modern sensibilities, for unforeseeable changes in the perceptions of the modern audience. . . .

Composers and writers working before about 1850 had little reason to suspect that the future would be any different from the past, or that their work would pass into a cultural environment very different from their own. History was relatively stable then. Now it is not so, and therefore the future of any particular work cannot be foreseen.

There are a number of ways in which a producer can discover "new meanings" in an opera, meanings of which the original authors may have been unaware. One can play melodrama so tightly that it hurts; one can make stock suffering heroines really suffer. One can openly dramatize what is problematic in a text—for example, the rigid and sometimes frustrating class distinctions in Figaro and Don Giovanni , the limited roles and images permitted to women in earlier times, Superman Siegfried's disgust for lower orders of beings, the Freudian undercurrents of Elektra and Salome . A producer who attempts to do such things—by dream or nightmare contexts, by modern dress and stage settings, by the dramatization of actualized neuroses and psychoses, by vivid symbolism and external reference, by taking a plunge into new emotional depths—will be inviting attack from critics like those cited previously. But all of these approaches strike me as legitimate and defensible, as long as the written work itself is not belied or betrayed in some essential way.^[20]

This, I believe, is what happens (as I argue in chapter 6, on Peter Sellars's versions

of the Mozart-Da Ponte operas) when a producer imposes from without a directorial conceit alien to the score. The producer's goal in such cases may be to shock—to defy the expectations of audiences he regards as half-asleep; or to editorialize—to force the opera to serve as a vehicle for some urgent personal message of his own. Some of the new productions that have been most vigorously denounced appear as well to defy, to play against the apparent message or meaning of the text and score.

This troubles me primarily, I suppose, because of the sublime arrogance involved. The words and music of a long-established work are turned into a plastic, pliable vehicle for a producer-king's late-twentieth-century ideas; the ideas of a producer-king who appears to regard himself as a creator, a maker of symbolic statements equal or superior to those of the original creators.

A case can be made for such apparently arrogant and perverse directorial "statements." I know, because I have heard versions of this case made time and again. One may decide (1) that composers of operas are not necessarily more important artists than the producers of their works; (2) that because the composer and librettist are dead—no more than names attached to a text, their intentions indecipherable or irrelevant—their words and score are now ours to play with as we will; (3) that modern producers of other people's texts need no longer play the self-effacing, secondary role of mere réalisateurs; (4) that in any case there are a thousand ways to "read" whatever "directions" may be encoded into the music; or (5) that because all values today are disputable, there can be no canons or "right ways."

All of these ideas may be found in contemporary literary critical theory as well. But a new, "deconstructionist" reading of a classic literary text differs considerably in its audience, its cultural effect, and its reality status from a similar reading of an classic opera, made by a person in a position to produce it. Having to live with a producer's strange new interpretation of Wagner's Ring at Bayreuth for five or more years (and the many other productions it is likely to affect) is quite a different thing from being able to read (or ignore) a critic's strange new interpretation of a Baudelaire poem. The case is even more serious with willfully outré productions of little-known works, because this particular interpretation is likely to be the only one the audience has ever seen or will ever see.

The producer of a standard repertory opera who based his interpretation on premises like those just cited—for example, Peter Sellars, in his productions of The Magic Flute or Tannhäuser —could insist that the only legitimate way to object to such a production would be to do battle, not with the producer's "misconception" or "perversion" of what we regard as Mozart's or Wagner's ideas, but with his own apparent statement. All productions, even the most traditional, the most dutiful, are interpretations of a sort, and hence implicitly critiques. A postmodern producer opposed to what he sees as the values of a composer or his time may take this idea one step further and mount an anti -Puccini Madame Butterfly , let us say (I

have seen such productions), in which he attacks the composer and his values in the course of producing his work. (Of course, musical analysts, critics, and producers can debate endlessly over the nature of the original "meaning or message," in addition to the propriety or impropriety of defying it. There are almost as many musicologists' interpretations of the Mozart-Da Ponte operas as there are producers'.)

The chief weakness of such revisionist efforts is that the spectator cannot help but, in Kerman's words, "trust the music" first, because music carries so much more emotional impact than anything else: more impact than the words; more than buffoonish costuming, psychotic staging, cynical acting "against the text"; more than even the most vivid, the most hostile production. The producer, as I suggested, may even be presuming in the audience a shared dis trust, even disgust, for the score he has chosen to direct (Andrew Porter once suspected this of what he regarded as Jean-Pierre Ponnelle's trashing of Massenet's Manon ); in which case, we are all implicated in the cynical, destructive enterprise.

Many operagoers are far more easily satisfied than I am with predictable, routine productions—Carmen with the spit curl and the rose between her teeth, flashing her knees flamenco-style beneath ruffled fuchsia petticoats as she clacks her castanets and dances atop an inn table près des remparts d' an apparently authentic Seville . Beyond that, they ask only for good music, well played and (if possible) spectacularly sung.

My preferences for fresh, alert, conceptually unified, and dramatically credible total theatre are based as much on autobiography as on any reasoned aesthetic theory. A literary and architectural education in the 1950s, several years spent as a critic and teacher of drama, a heady exposure at a very susceptible stage to the work of such people as Wieland Wagner, Walter Felsenstein, Peter Hall, Giorgio Strehler, and Maria Callas all disposed me to think of opera as theatre first; as drama, only better. In a frequently quoted remark, Michael Billington of the Guardian has claimed that "the lyric theatre has absorbed much of the dramatic theatre's energy and talent. . . . To put it crudely, opera now possesses [he was writing in 1988] the controversial dynamism theatre had 15 or 20 years ago." David Pountney (one of the postmodern directors conservative critics love to hate) says, "On good nights in the opera house, the acting you see is as good [as], if not better than, the acting that you see at the National Theatre."

I can't quite buy that. I remain an ardent follower and a fan of non lyric drama at its best, as well as opera. I know that one continues to see better acting—taken all in all—at the National Theatre than at Covent Garden, and at the better American regional repertory theatres than at the Met, and that it is foolish to claim otherwise. But I early learned of the far greater expressive potential of sung over spoken drama, however rarely that potential is achieved. Much of my indoctrination came at the hands of Kurt Herbert Adler, the Viennese-born general director (1953–

1982) of the San Francisco Opera, an impresario considerably more adventurous in his choice of operas, singers, and producers than either Rudolf Bing in New York or Carol Fox in Chicago. If I had been educated to opera in New York or Chicago, my responsiveness to "total theatre" values might conceivably have been less, and the likelihood of my thinking of opera primarily in terms of fabulous voices all the greater. (Contemporaries I met from those two cities who cared about opera were inevitably violently partisan fans of this or that particular singer.) On the other hand, if I had grown up learning about opera in any one of several German cities, I might not care about fabulous voices at all.

An ideal, total production of a major opera, by my standards, is very nearly unattainable, although I have seen many that came close. Exceptional resources and rehearsal time, rarely available outside of a few prime summer festival situations (Glyndebourne, Salzburg, Bayreuth), are probably indispensable. A first-rate conductor sensitive to voices is required. Many of the world's best conductors are unwilling to invest the time and energy required for a top-quality opera production (outside their home ports) when they can earn the same fees and acclaim for less demanding orchestral concerts and recordings.

This ideal conductor should have the time and will to work out a line-by-line interpretation together with a congenial and musically sensitive director. Ideally, the director would be his own designer, or would be working with one in sympathy with his ideas. The two of them would want, of course, the best orchestra, soloists, chorus, and dancers available, which implies a resident ensemble accustomed to working together, all of its members pliant and theatrically skilled, able and willing to undergo a great many rehearsals.

One is rarely likely to get all that, which is why a "perfect opera" is so hard to pull off. Most of the time I'm more than willing to compromise—as composers, conductors, producers, and singers have always had to do. If nothing essential is seriously harmed, I will settle for an imperfect realization of a great opera—a Don Giovanni, let us say (the text and score offer so much that even a partial achievement can be a feast); but I hope for something closer to a fully realized, 100 percent incarnation of a Thaïs or an Arabella . The pleasure of a well-produced minor work can be considerable, since a near-complete realization of almost any work of theatre yields (to someone possessed of a "theatrical sensibility") a sense of fulfillment; an ultimate, if a lesser kind of satisfaction. (My presumption is that 100 percent realizations of theatrical works, like perfect grades of "20" on French examination papers, never occur. The nearest I've come to what felt like a perfect realization of a great work of the spoken theatre was a production of Chekhov's The Three Sisters given by the new National Theatre at Chichester in 1963; of an opera, a Der Rosenkavalier —which featured Elizabeth Schwarzkopf's American debut—at San Francisco in 1955, reprised with great style in subsequent seasons.)

AUDIENCE AT THE GLYNDEBOURNE OPERA FESTIVAL THEATRE , 1969. Drawing 
by Sir Osbert Lancaster.Courtesy of Lady Ann Lancaster and the Glyndebourne Opera Festival.

Lesser achievements can still teach us many things, give us many pleasures. Good music well played call offer, at the very least, the joys of the concert hall; fine voices well deployed call yield a deep and intimate satisfaction, even if the vehicle is weak, the production inadequate. In such cases, however, the frustration of what has not been achieved may well undercut the satisfaction of what has.