A New Callas

Maria Callas converted me to opera. I am sure I am not unique in this, except in the particulars. In my early college years I immersed myself in recordings of the nineteenth-century symphonic repertory—Beethoven, Schubert, Brahms, Bruckner, the Russians—but for a long time I refused to listen to opera, would listen to an overture and then rush to change the record before the singing started. Then one day my roommate put the 1953 Tosca with Callas on the turntable and dropped the needle onto “Vissi d’arte.” I had no idea what she was singing, but near the conclusion of that imploring aria, as she comes to the end of the arching wordless phrase that soars from an A down slightly to a G, there is an audible intake of breath. She gasps—or is it a sob?

Maybe she misjudged her breathing, but I can no more assess it in such terms now than I could then. It sounded—it still sounds—to me like pure emotion. I didn’t need to know, at that moment, that Tosca was asking God why, despite all her devotion to art, to the Virgin, she was being forced to submit to the advances of the Roman police chief in order to save her lover from execution. That gasp—not a note, not even something musical—opened up an operatic world to me, a world of feeling pushed to an extreme, that in large measure became defined for me by Callas.

Yet if I hadn’t been hooked by a gasp, I would eventually have been hooked by her way with something else—a word, a phrase, an aria, an entire scene. For what made Callas a great singer was her fierce commitment to the voice as a means of dramatic expression. This must be stressed: her fame as an artist came not from gossip about her temper or her cancellations or her relationship with Aristotle Onassis. She was not great because of her “acting,” some mesmerizing physical gestures that we must trust those who actually saw her to verify. If that was primarily what she brought to her performances, she would be a dim legend by now. (The few extant videos of her that we have—mostly in recital—suggest rather how inward her identification with her characters was, how restrained her gestures onstage must have been.)

It is the other way around. Callas was a great actress because she was a great musician, because she understood that for a singer the drama is above all in the voice: in phrasing, in shaping the vocal line, in executing a perfect descending scale, in giving every word—even, or especially, in recitatives—its proper emphasis. Nothing is passed over, nothing ignored or slighted. (Listen, for example, to the monastery scene from Verdi’s La Forza del Destino: you need only to hear her to understand Leonora’s desperation, that her very chance for salvation is at stake.) Callas’s voice is instantly recognizable, intensely focused, it is theater; it is why, nearly forty years after her death, nearly fifty after her last stage performance, she remains her studio’s best-selling classical singer, as her recordings constantly find new admirers and provoke fresh arguments over their merits.

Before her first performances of Leonora in Verdi’s Il Trovatore in 1950, Callas’s mentor, the conductor Tullio Serafin, refused to help her learn the role for another conductor. So she “plunged into the score…absorbing every note and expression mark,” writes John Ardoin in The Callas Legacy (1977), with the result that “the wealth of detail her voice unearths here is breathtaking. It is as if an old painting, familiar but dim, has been cleaned to its original tints.”

The same could be said of Maria Callas Remastered, the new edition of all of her studio recordings—twenty-six operas and thirteen recital discs, the earliest from 1949, the last from 1969. Performances long familiar from LPs or the two earlier, less careful CD transfers now leap from the speakers with vivid clarity of sound and virtually no background noise even in the mono recordings. Engineers at Warner Classics (which acquired EMI, the company that produced most of them, in 2013) went back to the original master tapes and production notes from the recording sessions to create what finally must be, at least until the next advance in the technology of sound reproduction, the definitive digital version. (The book included in the complete set gives a detailed account of the process.)

Maria Callas Remastered, by bringing us as close as we can come to what she must have sounded like, at least in the studio, is an opportunity to rediscover what made her so electrifying. In her prime her voice had a range of nearly three octaves and a very particular timbre; it was moreover able to negotiate elaborate vocal lines with precision, power, and a variety of tonal shading. She simply sounded like no one else. One Italian critic described her as “a star wandering into a planetary system not its own.” But her voice had more than technical facility. It was also uncommonly expressive: it was what Leo Lerman called “the most haunting voice I have ever heard…filled with lost joys, permeated with present despairs.”

It was, to be sure, a voice that was not conventionally beautiful in the way that, for example, her contemporary Renata Tebaldi’s was. (Tebaldi, who mostly sang late Verdi and Puccini roles, could not match Callas’s skill in coloratura or more importantly her theatrical sense, though admirers of sheer vocal beauty often held her up as a counterexample to what they considered Callas’s shortcomings.) From the beginning Callas’s voice could be uneven; as Ardoin writes, it “divided into three distinct registers: a bottled, covered low voice; a reedy middle; and a top that was brilliant at times to the point of stridency and that would, without warning, threaten to wobble out of control.” Over the two decades represented in this collection she underwent a marked vocal decline; by the 1960s the wobble had become persistent and her tone could turn raw.

You can hear the changes in her voice in the two studio versions of Norma from 1954 and 1960. It was the role she sang more than any other—nearly ninety times between 1948 and 1965. Its demands are not only vocal, from coloratura to martial declamation, but emotional, from mystical reverie to fury to interior monologue to supplication. Over almost twenty years, Callas developed for Norma her most penetrating dramatic insights, meticulously differentiating the Druid priestess, the mother, the warrior, the anxious, then raging, then reconciled mistress. By 1960 she no longer sings with the same security as in 1954, though there are compensations for the sense of vocal strain: her interpretation is more affecting, the phrases more poignantly inflected. I do not always prefer the later recording, but I would not want to be without it.

It is sobering, listening to these performances separated by only half a dozen years, to recall how short a career Callas had. Though she sang her first operas onstage while a conservatory student in Athens just before and during World War II, her first professional engagement outside Greece was at the Verona Arena in 1947, when she was twenty-three. In her early years she was cast primarily in Wagnerian and verismo roles like Isolde, La Gioconda, and Turandot that required a voice powerful enough to soar over a large orchestra. Then in 1949, in Venice, where she was singing Brünnhilde in Die Walküre, she substituted on short notice for an ailing soprano as Elvira in Bellini’s I Puritani, a role that requires a completely different kind of voice, one with vocal agility in coloratura passages and the ability to sustain long lyrical lines suffused with pathos.

Other sopranos have sung both Brünnhilde and Norma without attempting the fragile Elvira. Yet it was more than a virtuosoic stunt: it was a revelation. Callas, by fusing the weight and color of her dramatic soprano voice with the nineteenth-century bel canto technique she had absorbed as a student, rediscovered the emotional depths in roles that had long been performed, if they were performed at all, by light-voiced sopranos as pretexts for mindless vocal display. When she sang Lucia di Lammermoor, Violetta in La Traviata, Anna Bolena, or Imogene in Il Pirata, they once again became fully dimensional tragic figures.

It was the beginning of the revitalization of an entire repertory: in the following decades operas by Bellini, Donizetti, and Rossini that had fallen out of favor over the previous century were pulled from dusty archives and became vehicles for Leyla Gencer, Joan Sutherland, Montserrat Caballé, and Beverly Sills, among others. Many of them had superb voices; hardly any could approach Callas’s gift for seizing the dramatic possibilities in such works.

For a few magical years in the 1950s she seemed unconstrained by conventional vocal categories, able to adapt her voice equally to the wistful tones of Anina in Bellini’s La Sonnambula and the menacing ones of Verdi’s Lady Macbeth. But her conquests of such a wide range of roles came at a cost. Her vocal flaws became harder and harder to control. She gave her last performance of a complete opera in 1965, but by then her voice had been unreliable for more than five years.¹ Her greatest accomplishments were packed into barely more than a decade.

Callas was fortunate that her emergence coincided with the introduction of the long-playing record in 1948, which meant that recording companies were keen to build their catalogs in the new format. She began recording for the Italian label Cetra in 1949, with an album of arias, followed by complete recordings of La Traviata and La Gioconda (all of which are included in the newly remastered set). That first Cetra album, with Isolde’s Liebestod and the “Casta Diva” from Norma as well as Elvira’s mad scene from I Puritani, already displays her extraordinary range. By 1953 she had been lured to the much bigger British multinational EMI, for which she recorded until 1969, and which enabled her to reach a far larger audience across Europe and the US.

Maria Callas Remastered does not, of course, give us Callas whole: there are also many recordings of live performances from opera houses and concert stages, beginning in 1949, that preserve important roles that she never recorded in full in the studio, such as Abigaille in Nabucco and Elena in I Vespri Siciliani; in many cases where both exist, the live recordings capture her in better voice or more dramatically engaged. Her best Norma, for example, is neither of the studio versions but the live performance from La Scala’s opening night in 1955—widely available on CD—that caught her in commanding form.

Vagaries of sound quality in these live recordings aside, her 1955 EMI Aida doesn’t have the circus thrills of Mexico City in 1951, when she ended the triumphal scene with an unwritten E-flat—though the third-act scene between Aida and her father (Tito Gobbi) is far more intense in the studio. The 1953 studio Traviata came too soon, before she had fully plumbed the role, but she is heartbreaking as Violetta at La Scala in 1955 and even more so at Covent Garden three years later. Her raging Medea in Dallas in 1958, just after she was fired from the Metropolitan Opera, makes her exhausted studio version from the year before nearly superfluous. But she dazzlingly surpasses her 1951 performance in Florence of the fifth-act Bolero from I Vespri Siciliani on the 1954 album Lyric and Coloratura Arias, her voice seeming to echo itself in its cascading roulades of sound.

I found myself, however, almost morbidly listening over and over to what has widely been considered the least successful of her recordings, the second, stereo Lucia of 1959—one of the very few I had never heard until I got the new set. So it comes as some surprise to read, in the book accompanying Maria Callas Remastered, her producer Walter Legge’s comments about it: “Unless I am very much mistaken it will be a sensational success. Callas has never sung so well….”

One imagines that Legge must have been reassuring the executives at EMI, for the performance lacks the fearless brilliance of her 1953 studio recording (in mono) of the role. It also lacks the plaintive despair of the live performance from Berlin in 1955, in which the notes veering in and out of control in the mad scene seem to purposefully reflect Lucia’s own unhinged mind. Where once the fragility of the voice seemed calculated to convey the fragility of the character, it is now too easy to hear the singer’s fragility. Some high notes waver uncomfortably, there are unaccustomed compromises in some of the more florid passages, and the supporting cast sounds more provincial than in her previous Lucias.

So one wonders why this was a priority for EMI at the time. Yes, the sound is better, but the singing isn’t. Did the advent of stereo recordings in 1958 really mean that EMI needed to replace the splendid earlier mono Lucia rather than have Callas set down before the microphones any number of her other important roles that she hadn’t recorded? Why a new Lucia and not Macbeth or Anna Bolena, in both of which she triumphed at La Scala but of which she recorded only excerpts?

Such inscrutable commercial decisions haunt her entire recording career. Why record Manon Lescaut and La Boheme—roles she never sang onstage—instead of I Vespri Siciliani or Il Pirata, which she did? Presumably EMI’s assessment of the market dictated giving priority to better-known works even if she had no association with them. Perhaps no one expected her career to be so brief, even as she was performing less and less frequently and less and less consistently well.

With the exception of a few arias recorded in 1969 and 1972, Callas stopped recording as well as performing in 1965. For the rest of the decade she and EMI discussed many projects that never came to fruition, primarily and frequently a new Traviata; the booklet reproduces letters from EMI executives laying out plans for performances and recordings, though one also hears their exasperation at the priority she was giving her social life—attending premieres with Georges Pompidou or Elizabeth Taylor—over her musical commitments. But it is equally likely that she had taken the measure of what she could accomplish and the disadvantage of competing with herself; her voice was too well documented on recordings that her audiences could be counted on to have heard wherever she sang.

She eventually agreed, unwisely, to a worldwide recital tour in 1973–1974 with the tenor Giuseppe di Stefano, with whom she had often sung in the 1950s. By then not only was her voice in ruins, but she seemed to many who heard her to have lost as well her technique and expressivity, which had long endured even as her voice betrayed her. Far more memorable were the master classes she gave at Juilliard in 1971–1972, at a time when she was trying to realign her voice and studying scores anew. There she worked with young singers of all vocal types, before an audience that included famous singers, conductors, instrumentalists, directors, critics, and fans. It was the hot ticket of the season.²

In her classes she dutifully instructs her students to observe meticulously the details of the score and to understand the meaning of the words they are singing, advises them on their breathing and even on their clothes. She is frank and friendly and sometimes funny. The most thrilling moments, though, come when she is so caught up in the music that she takes over and sings along with them to demonstrate her points. Even hoarse her voice flashes to life. Has any baritone sung Rigoletto’s “Cortigiani, vil razza dannata” with as much ferocious urgency as she does while teaching it? Her student can’t sing it “like an animal,” as she tells him—and shows him—he must. Yet her musical intelligence may have been too intuitive, too singular, too inimitable to pass on.

One must resist living entirely in a golden operatic past accessible only through recordings. Opera requires the excitement and uncertainty of live performance and an openness to new singers who can bring their own unexpected insights to their characters. But we learn to listen to them by listening to their great predecessors. Callas wasn’t perfect, to be sure. (And perfect can be boring, as some of her successors have demonstrated.) But even when she falls short of her best, she gives an intimation of what an ideal performance might sound like. Few more perfect singers have managed to do that.