previous sub-section
Chapter One— Introduction: The Difference Is They Sing
next sub-section


Independent of productions and spectators, is it possible to classify works of the lyric stage absolutely on a scale from best to worst?

In 1956, the London-born, Princeton-trained, Berkeley-based critic and musicologist Joseph Kerman (then thirty-two) published a short, original, extremely readable, and notoriously provocative book entitled Opera as Drama , in which he proposed to establish standards for distinguishing good operas from bad.

He began (as I have begun) by referring to the antiopera. critics. As far as he was concerned, the only respectable way for serious defenders of opera to counter their attacks was to become more selective, to separate the few grains of wheat in the repertory from the mountains of chaff. Writing in 1956—but retreating very little in a second edition prepared thirty years later[10] —Kerman has nothing good to say for the "hardening repertory," "the slag-heaps that have kept opera thriving."

Flabby relativism is certainly the danger, as anyone knows who buys an opera season ticket. Under the tacit assumption that everything is all right in its own terms, extremes of beauty and triviality are regularly placed together. In our opera houses, art and Kitsch alternate night after night, with the same performers and the same audience, to the same applause, and with the same critical sanction. Confusion about the worth of opera is bound to exist when no distinction is drawn publicly between works like Orfeo and The Magic Flute on the one hand, and like Salome and Turandot on the other.

If a case is to be made for opera as a viable form, he insists, it can be only in terms of the best operas, operas that work "dramatically," in the same way that good plays do. The radical difference is that in good operas (on the evidence of this book, Kerman appears to think there are about twenty; Hans Keller, writing in a wholly different context, puts the number at "about 25"), the basic elements of drama—characterization, action, psychology, and so on—are defined and conveyed by the music , not by the words. These few operas, he sometimes appears to be saying, are the only ones that really ought to be produced. He takes music critics to task for wasting their time judging new productions of standard repertory works, when they should be trying to reduce that repertory by judging operas—old operas as well as new ones—according to more rigorous standards.


Kerman begins his own sifting of the repertory with a quotation from the musicologist Edward Cone: "In any opera, we may find that the musical and the verbal messages seem to reinforce or to contradict each other; but whether the one or the other, we must always rely on the music as our grade toward an understanding of the composer's conception of the text. It is this conception, not the bare text itself, that is authoritative in defining the ultimate meaning of the work."

Whenever there is an apparent conflict between the drama implicit in the libretto and the drama implicit in the score, Kerman insists, it is the composer who wins.

We trust first what is most emphatic in the music. ( Orfeo )

In opera we trust whatever is musically forceful. ( The Marriage of Figaro )

In opera, we trust what is most convincing in the music. ( Così fan tutte )

In opera we trust what is done most firmly by the music. ( Don Giovanni )

As always, the dramatist is the composer, and the Rake's progress is articulated by a progress in the music.

Most of Opera as Drama is devoted to close, selective investigations of about a dozen operas Kerman frankly admires. ("The significant operatic canon is not large. Monteverdi, Purcell, Gluck, Mozart, Verdi, Wagner, Debussy, Berg, Stravinsky, and a few others have left a body of musical drama which is rich and various, but not large.") In these chapters, he describes and analyzes differing ways in which drama can be created and conveyed by the score. His explications of the "musical dramaturgy" of certain individual scenes—the Act II trio of Don Giovanni , the Act II finale of The Marriage of Figaro , Otello's murder of Desdemona, Tristan's long Act III "Delirium"—are about as good as purely verbal musical-dramatic analysis can get. In them, he demonstrates how music works simultaneously to create, compel, and contain action, to define and express subtle and evolving changes in and among characters. Especially impressive is his relation of the simple "progress forward" of music—the alterations of structure, rhythm, and tonality in time—to the parallel forward progress of the drama, the evolution of feeling in the characters.

The standard Kerman applies to opera, in his attempt to distinguish better from worse, is closely related to literary criticism of the post-World War II years (his early essays on music, which evolved into this book, were published alongside the work of leading literary New Critics in The Hudson Review ), and more specifically to theoretical works on drama by critics like Francis Fergusson, Eric Bentley, T. S. Eliot, and Una Ellis-Fermor, all of whom he cites. If he is to demonstrate the viability or ultimate worth of an opera, he must prove that it contains rich


characterizations, a coherent progress or plot, and a unified world, like those of the best spoken drama. "The spoken theatre serves as a court of appeal," he writes, "or at least of analogy, when the dramatic efficacy of opera or any other non-verbal medium is likely to be tested." Unfortunately, as Herbert Lindenberger once pointed out, "as long as opera is seen from the point of view of spoken drama, it is always likely to seem wanting. . . . As soon as opera is compared to drama, the former comes to look deficient in 'intellectual' content."

Beyond that, Kerman depends considerably on the ideals of formal coherence, integration, and unity, the kind of "organic economy" proposed and promoted by leading British and American literary critics of the 1940s and 1950s (whom today he labels "classics"). By the standards of many literary critics of the time, one could ask for nothing more than maximum complexity under maximum control; and it was presumed that one could demonstrate this kind of perfection by the close analysis of a text.

But when Kerman applies similar standards to opera, the works of Baroque and bel canto composers art inevitably seen as wanting. "The fault, almost alvvays," he writes, "was in the abysmal lack of integration of lyricism into a sensible dramatic plan." Purcell, by contrast—one of Kerman's two seventeenth-century exceptions—understood "how to organize arias into a total, coherent dramatic form." The action in the operas of Mozart—Kerman's model of ultimate success—is "infinitely more complex . . . included within a single musical continuity, and unified by it."[11]

This particular set of ideals leads to what appears to be an aesthetic of historical progress—at least up to Wagner and Debussy—whereby operas became better as they moved away from separable "numbers," away from the expressively limited alternation of arias and recitatives; and toward greater musical/dramatic continuity: which, of course, is what the Wagnerians have been telling us all along. "Conflict, passage, excitement, and flux could [with the development of the Classical sonata formal be handled within a single musical continuity," Kerman writes; and at the same time be "made to cohere, to present unified impressions," permitting a much richer presentation of the human psyche. While acknowledging that in certain circumstances traditional arias may still "work" dramatically, he appears to give his most wholehearted approval to operas that manage to do without arias, or to depend more on ensembles.

But it is neither his theory of what makes for good opera nor his short list of works that succeed that has made Opera as Drama the most often-quoted book on


the subject of the last thirty years. It is his near-absolute dismissal of the works of Giacomo Puccini and Richard Strauss, which still account (as we have seen) for a huge part of the standard repertory—eleven of the 100 leading works, including three of the Top 10. He is arguing, in effect, that critics and impresarios should stop granting these two composers serious attention.

I do not propose to analyze the musical texture of Tosca [he writes, after savagely disposing of its ultimate scene]; It is consistently, throughout, of café-music banality. If Joyce Kilmer or Alfred Noyes had taken it into his head to do a grand poetic drama on Tosca, that would have been something analogous in the medium of language. . . .

But if Turandot is more suave than Tosca musically, dramatically it is a good deal more depraved. . . . [The score] is consistently, throughout, of café-music banality. . . . Rarely has myth been so emptily employed as in this absurd extravaganza. Drama is entirely out of the question. . . . Puccini clings to his limited ideas and repeats them protectively. . . . There is almost a sense of despair in the meaninglessness of Turandot .

Strauss, Kerman concedes, was a more advanced and innovative composer than Puccini, masterly in power and technique. But as a musical dramatist of human situations, Kerman finds Strauss cynical, shallow, false, and sentimental. The conclusion of Salome he regards as "the most banal sound in the whole opera," the final coming together of the young lovers in Der Rosenkavalier "the poorest thing in the opera."

Was it for this minimal level of consciousness that we have had to stiffer the Marschallin's self-pity and to sacrifice Ochs? for this, the silver rose and the white suit, the Three Noble Orphans and four finicky hours of leitmotivs, modulations, and program-musical wit? . . .

No one who understood The Marriage of Figaro could ever have taken Der Rosenkavalier seriously. . . . Salome and Der Rosenkavalier  . . . are insincere in every gesture, meretricious and doubly meretricious on account of their show of outer formal integrity.

"From the start," he concludes, "Puccini and Strauss revealed a coarseness of sensitivity and a deep cynicism towards true dramatic values. . . . The operas of Strauss and Puccini are false through and through; the trouble runs much deeper than mere faults in conception or technique. . . . The response, the quality of the action, is insensitive or simply sham all the way. . . . In the deepest sense the operas of Strauss and Puccini are undramatic, for their imaginative realm is a realm of emotional cant. They are unable to match any action, however promising, with anything but the empty form of drama."


Kerman also dismisses all opera between Monteverdi and Gluck (except Purcell's short Dido and Aeneas ) as of "unparalleled dramatic fatuity." He makes no mention of Handel beyond a line of very faint praise in the epilogue of the second edition. He is generous to La traviata and Rigoletto; rejects Bellini (but for the end of Norma ), Rossini, and Donizetti; tosses out Aida; and appears (in Opera as Drama , although not in a later essay in The Hudson Review ) unfavorably disposed to Wagner's Ring .[12]

If I am to begin working toward a definition of "the right opera," Kerman's standards and strictures are as good a place as any to begin. I am obliged, in any case, to live with the existing repertory. Like other critics and opera lovers, I must therefore come to some sort of terms with his responses to Puccini and Strauss.

Although creditable and well argued, Kerman's aesthetic of opera seems too restrictive for me, too limited, too purist. It is only partially able to account for my (and I presume other people's) actual experience in the opera house. He has (in Opera as Drama ) almost nothing to say of the specific power of the singing voice alone, in ensemble or chorus, or with an orchestra—which I would have thought was the fundamental distinction of an operatic experience. Outside of his reviews, he makes no reference to production or performance, to the impact of live actors on stage, to the legitimate power of melodrama in opera, to the legitimate place of


spectacle. In fact, he often simply labels such concerns as "vulgar" and "coarse," compared to musical-dramatic integrity.

Musical effects in opera intended primarily to be exhibitionistic, shocking, or bathetic, to move an audience to tears or gasps or bravos, are no more (and no less) admirable than the high-wire turns of a circus acrobat, the scream scenes in a movie thriller, a tear-jerking soap. But I tend to think that certain extradramatic, even vulgar, aspects of opera production—melodrama, vocalise, the stage presence of real actors, spectacle and setting, a conductor's or a director's particular reading or concept—are at least as essential to potential success as whatever drama may be crafted into the score.

Kerman's focus on overall coherence and dramatic integrity leads him to ignore or belittle our ability, perhaps even our innate tendency, to separate out and enjoy individual scenes, voices, and moments in opera. The sum of such pleasures may or may not be less intellectually worthy than the pleasure we take in an overarching musical-dramatic unity. But by the very creative conditions of opera, such unity is likely to be rare. "Opera by its very nature is a gigantic series of compromises," writes Winton Dean in Handel and the Opera Seria . To this, Herbert Lindenberger adds, "The whole institutional setting within which opera has traditionally flourished requires so many compromises that only a few operas demonstrate the intensity and evenness of craftsmanship that nontheatrical music can more easily attain."

For some critics, historians, and theorists of opera, the list of fully realized musical-dramatic unities appears to come down to one: The Marriage of Figaro . Rather than apply such reductive standards, I prefer to make a case for a broader base of appreciation: for trying to enjoy opera other than generically—in bits and pieces, if necessary, rather than as a coherent musical-dramatic whole. This is what Kerman might call "hedonistic," rather than "aesthetic," appreciation—a distinction he made in an early analysis of Virgil Thomson's critical standards.

In drawing his ideals of music drama largely from mid-twentieth-century theorists of spoken drama canonically certified as "literature," Kerman slights the fundamental importance of melodrama to opera. He does have favorable things to say about scenes in Rigoletto, La traviata , and Norma that might strike one at first glance as melodramatic. But a commitment to high-dramatic standards may lead a critic to prefer works drawn from more "respectable" sources, or works that make use of more subtle and less hysterical plots. Despite its contemporary position of disfavor as a "low" and pop-obvious genre, melodrama—as I argue in chapter 8, on Victor Hugo's opera plots (borrowing my case in part from Eric Bentley)—is closely related to human needs, is sometimes ideally adapted to operatic rendering, and, as the word melodramme originally implied, can be unusually viable when sung.

There is also a place for, and considerable pleasure to be taken in, the cultivation of a historical or cross-cultural imagination at the opera house, by which we make


an effort to place ourselves into the situation of audiences not our contemporaries, and try to enjoy the conventions of other times and places. ("With a strong exercise of historical imagination," Kerman writes of Donizetti's Anna Bolena , "one can perhaps see why the piece made its mark in 1830." I recommend such exercise.) In the attempt to apply mid-twentieth-century standards of unity, continuity, controlled richness, and overall coherence to pre-twentieth-century works—admirable as such standards may appear to us today—we run the risk of condemning unheard, or hearing unsympathetically, a great deal of opera created at a time when such standards counted for less.

Beyond a single reference to "the emotional power of the human voice" (and a passing jibe at "the vulgar taste for vocal virtuosity"), there is virtually no mention in Opera as Drama of the affective force of singing, which is surely—even more than the orchestral score—what most clearly distinguishes opera from any other form of drama, both as an art form and as all experience. At times, in fact—sounding rather like Verdi or Wagner on a bad day—Kerman makes it seem as if live singers and physical productions are a hindrance, an obstacle to the composer of operas, when in fact they are the means by which his works are brought to lift.[13]

Kerman admits, in the preface to the second edition, that there is a "total absence of any discussion of performance values in Opera as Drama ." In this, he is simply behaving like most musicologists, who tend to see printed notes on a page (and, in the case of opera, the accompanying printed words) as composing something finished and complete, ready to be analyzed and judged and, perhaps, somewhere along the way, enjoyed. In trying to snatch Alban Berg's Wozzeck back from more microscopically analytic musicologists, Kerman reminds them that "the ultimate judge is the ear, not the eye, and that the work is destined for the opera house, not the analyst's study." But there is little reference to that destination elsewhere in the book.

It is here that Kerman's search for a code of values, an aesthetic for opera, most widely diverges from mine. By nature or training I am disposed to regard operatic scores, as I regard the texts of plays, primarily as scripts for production, incomplete until performed. Kerman (quite justifiably) regards both Così fan tutte and Don Giovanni as imperfect works, because the librettist provided texts that were either cynical and empty-hearted (Così ) or clumsy and improbable (Giovanni ), to a composer who was unable to take emotions other than seriously.

But what Kerman sees as flaws I see as opportunities; as, in fact, two of the greatest challenges the producer of an opera can face. (I use the term "producer" as


a kind of shorthand, rather as one speaks of a film's "director." I am fully aware that a whole team of creative individuals may be responsible for an opera production—often beginning with a conductor, rather than with what the French call a metteur en scène .)

Can Mozart's and Da Ponte's two magnificent, problematic works be produced in such a way as to realize simultaneously, even to reconcile, their warring worlds of discourse; in such a way as to make these operas "work," somehow, despite their troublesome décalage? Kerman complains that he has yet to see an explanation (or, I presume, a production) of Così that makes sense of both the action and the music—as George Bernard Shaw lamented that he would probably never in his lifetime see an adequate production of Don Giovanni . I put my faith in the ingenuity of producers, and wait in hope.

I regard works such as Carmen, Boris Godunov , and any so-called opera by Handel in a similar, tentative way: can they be made to work, made to matter, made to hurt? Can a producer fight the sentimental, lockstep stage directions Strauss and Hofmannsthal inscribed into Der Rosenkalvalier? Can he or she make Puccini's Butterfly really suffer? Jean-Pierre Ponnelle's jaunty, knife-edged, totally rethought production of I pagliacci (1976) resurrected for me an old war-horse I had given up for dead. Other producers, conductors, and singers have performed similar miracle cures on similar works.

Even so, I do have standards of my own by which I distinguish better operas from worse, apart from their productions. Other things in the theatre being equal (which they never are), I too prefer coherence, a felt sense of continuity and unity, the overall arch of a single musical-dramatic conception, to a sequence of ill-matched musical scenes (like that of Boito's Mefistofele ) that seems to defy one's dramatic expectations. (There are, of course, disunified works, like those of Alban Berg, that make good emotional sense.) I also agree that this drama should be achieved through an artful fusion of sung music, orchestral commentary, and onstage action, rather than simply through the libretto itself.

A good opera, I believe, is one written by a composer who is fresh, ingenious, and inventive; who can come up with interesting musical ideas that illuminate, illustrate, and help to tell his story.

Ergo, there should be a story. I find I am emotionally and intellectually compelled by stories that deal with characters whose situations represent or reflect those of human beings, rather than ideas or abstractions. I respond with greatest fervor and commitment to a dramatic action. I can be made to care about (in the way one "cares about" actions in a theatre), to works with a high potential for characterization (which need not mean "realism" or "naturalism"); to works in which I feel myself potentially implicated in the persons and plights of the characters on stage. They can be animals—as in  Janácek[*] 's The Cunning Little Vixen —or


the nonhuman creatures of Wagner's Ring , as long as they allow me some means of emotional entry into the characters they represent. In any good production of the Ring , I find myself caring quite a lot about Alberich and Fasolt. There is also something quite wonderful about music dramas that offer a rich and still unified combination of tragedy and comedy—if only because "lift is like that," as we say, and the challenge to the composer is greater. This may partly explain the enduring appeal of the Mozart-Da Ponte operas.

I realize that singing one's feelings aloud, along with or over an orchestra, to a theatre full of one or two or three thousand people invites a degree of exaggeration in both characterization and plot. That's all right with me. This in part explains why "melodrama" often seems more appropriate than realism, than so-called verismo. The onstage worlds of Operaland virtually demand people considerably larger and certainly more passionate than arc we, their poor wizened counterparts in real lift, who don't have to sing our emotions or pretend to sustain them for hours on end.

Dedicated producers, musicians, and singers, I have discovered, all of them working very hard, can maximize the faint, relatively uncomplicated human potential of opera seria. But most of the pleasure I take in it is musical, rather than dramatic, so I tend to regard it as a lesser form of opera.

I seem unable to respond with the same kind of fictive "belief" or intellectual approval that others can muster to operas like Pélleas at Mélisande, Parsifal, Dic Frau ohne Schatten , or Michael Tippett's symbolist works (the best I can do, dramatically, is to regard them as some kind of dream); to the ldeological elements of The Magic Flute or Wagner's Ring , or even to the "inspiring message" of Fidelio . However much I may enjoy their music (or the more "human" parts of their progress), I have a hard time warming to onstage abstractions, or crediting musical works for the supposed depth or righteousness of the ideas they express.

I prize operas that provide good occasions for singing (solo, ensemble, and choral), but preferably singing that makes dramatic sense. I grow impatient at mere circus-turn inserts, even when sting by large and famous canaries with priceless golden throats. I dislike irrelevant choruses and interpolated dance numbers, unless the producer has found a way to fit them into the plot. I have learned to live with, even enjoy, all manner of operatic conventions. But I most admire composers, like Mozart and Verdi, who call turn the conventions they are saddled with to genuine dramatic point.

The best opera composers are those who love, as I do, the human voice staging. They learn how to write for it, how to take advantage of it, and how to use it to captivate and compel, to win our attention and assent by exploiting either its sheer beauty or its potential for resonant dramatic expression.

The best operas are those in which the music—vocal and orchestral—is remarkable and captivating all by itself, but is also heart seizing and transcendentally


CLARAMAE TURNER  (center) AND CHORUS MEMBERS  backstage during a rehearsal of
Wagner's Die Meistersinger , San Francisco Opera, 1959.
Courtesy of the San Francisco Opera.

affecting when experienced as the vehicle for a human drama about which one can care. The orchestra should be intelligently and dramatically used; it should contribute to, and sometimes create, the action. It can (as many writers before me have pointed out) describe inner thoughts, comment on events, foreshadow and recall, and create fear or suspense far more potently than words alone can do. It can motivate characters, define precise social relationships, bring new worlds into being, and (as in Mozart) even undercut the apparent meaning of the text. I am happiest when an opera orchestra does these things in interesting and original ways, with something less than movie-music obviousness.

Potential for good theatre is nice to have as well. It is no accident that the three greatest opera composers—Mozart, Verdi, and Wagner—were all theatrical geniuses, regardless of whether (as Shaw pointed out) the people who undertake to


stage their works always realize that. I find many operas, in and out of the standard repertory, to be fundamentally uninteresting both dramatically and musically, to the point where I doubt that any new production, however ingenious, could shake me out of my lethargic inability to respond. I once attended such works, when reviewing a whole season of which they were a part, as a professional critical chore. I no longer do. "I have difficulty with stupid works where bombastic music exceeds dramatic necessity," producer/impresario Michael Hampe has said. (The example he cited was Andrea Chenier .) I couldn't agree more.

There are other popular, and in few cases critically acclaimed, works that 1 actively dislike, usually because of a combination of what strikes me as either insipid (Louise ) or aggressively "ugly" (Elektra ) music, together with an overwrought, artificial human story that simply cannot compel my attention, however pertinent or meaning-laden it once might have seemed; or because they embody (inescapably, in the music) attitudes and values that repel me. (See chapter 9, "Sex and Religion in French Opera.") But most of my nays are still tempered and tentative, decisions not altogether closed. I try to leave room in almost every case for the possibility that a great conductor working with good singing actors and an insightful director may yet open my cars and my mind.

A few operas—operas my list is probably no longer than Kerman's or Keller's—do seem to me beyond critical question. That is to say, I know they are great, and need no production to prove it. They may, of course (and often will), be abysmally produced and performed. But I can always try to imagine the angels staging them, the composer directing them, through all of the squawking, scraping, and stumbling on stage: then go home and read the score, or listen to a good recording, and imagine the perfect production staged in the Opera House Under My Hat.

The greater part of the standard repertory—and a good many operas not in it—I like potentially . I tend, far more than harsher critics, to accept the repertory as it has evolved, and to worry less about the perfect opera, or the twenty tolerable operas, or some Olympian standard of acceptability. The bent of mind that seeks to form "canons" seems to me, in the 1990s, archaic and uncomfortably authoritarian.[14] My job, both as critic and as simple hedonist, is not to narrow the possible inlets of


joy, but to remain as open as possible to what a good production team can make of a tolerable text and score. I am at least as fascinated by the challenge of making a problematic opera "work" as I am by the imputed dramatic perfection or imperfection of its score.

George Bernard Shaw once wrote, "Nobody has ever greeted a performance of Tristan und Isolde by such a remark as, 'We shall never be able to go back to L'elisir d'amore after this'; or declared that [Donizetti's] Lucrezia was impossible after Brünnhilde." Actually, I think people may very well have said something like that; I know I have thought things like that, after an especially moving evening of Wagner. But Shaw's intelligent point is that work of the first class does not displace work of the second class, any more (the analogies are his) than Ibsen displaces melodrama or Shakespeare displaces the circus. We need and enjoy good Shakespeare; we need and enjoy a good circus. We cannot breathe permanently the air of Olympus. After one good seventeen-hour Ring cycle, I find I don't require another for quite a while. After the Rhine has finally overflowed and Valhalla has gone up in flames, I find myself turning to diversions of a considerably lighter sort (which, of course, is simply a form of self-definition. There are good people who can happily sit—or even stand—through three Ring cycles in a row).

There is nothing arbitrary about the size of the standard operatic repertory. It takes about 150 operas, plus new additions and discoveries, to keep the existing houses and companies running, the audiences contented, the singers and orchestras paid. Although I have about had my fill of Toscas and Bohèmes (but only after a great many enjoyable evenings), there are new opera lovers starting out every season who should have their chance to see, and perhaps one day reject, these works. Revisionist productions in recent years of Butterfly and Turandot , cross-grained productions that stressed the cruelty against the lyricism, have affected me strongly. I never want to experience another Faust , which is one of the few top repertory standards I really would like to see dumped. But I will keep on going to Rigoletto and Carmen until I see and hear a production as powerful as the version in my head. Well-performed Offenbach (which is hard to find) I love, the five satiric operettas more than Hoffmann . Experience in the theatre has taught me that any of a dozen Verdi operas, the more durable bel canto tragedies, Monteverdi and Gluck, much of  Janácek[*] and Britten, Der Rosenkavalier, Ariadne , and Arabella , and almost any well-wrought opera buffa can be turned into silver (if not gold) by the right team of stagers, musicians, and producers.

But that is my point: almost any tolerable script and score, I believe, can be made to work, can be rendered into a memorable theatrical experience, given the right production. And within "tolerable" I include Rossini and Vivaldi serias, Massenet and Meyerbeer, Henze and Shostakovitch, Penderecki and Poulenc: a list, in the end, so long that to speak of canons is to make no sense at all.


THE WAGNERITES . Drawing by Aubrey Beardsley from  The Yellow  Book (1894).
Courtesy of the Bancroft Library, University of California, Berkeley.


previous sub-section
Chapter One— Introduction: The Difference Is They Sing
next sub-section