There is a persistent myth in the operatic world that singers sing difficult roles too early in their careers, and a related myth that this is the principle cause of the vocal burnout among early- and mid-career opera singers that has been plaguing the business of opera for decades now. (To see how real, prevalent and long-standing the problem is, take a look at Will Crutchfield’s still pertinent “Vocal Burnout at the Opera” from the New York Times back in September of 1986.) At first blush, it may seem plausible that an early assumption of “heavy” operatic roles could cause vocal damage, and that too many singers have attempted the dramatic roles of Verdi, Wagner, Puccini and Strauss, leading to technical problems and shortened careers.
The real issue, however, is not as straightforward as it seems, and I have come to believe that the primary cause of distressed voices does not lie in a specific repertoire sung prematurely, but rather is a function of the inadequacy of modern vocal training. In fact, it seems increasingly obvious to me that the reason operatic voices don’t last is not that singers take on big roles too early, but that singers arrive at the age when they should be earning a living from singing without the technical and musical foundation that will see them through the rigors of singing opera, period. For, historically speaking, it is simply untrue that singers from the “Golden Age” waited longer and bided their time singing small roles and “easy” music before tackling demanding operatic roles. Yes, there are cases of singers who sang lighter parts as their voices developed. Kirsten Flagstad is often cited as an example of someone who didn’t sing her big roles (in her case, Wagner) until she was in her 30s, but in fact, Flagstad didn’t bide her time singing only operetta and Handel; she essayed Desdemona, Amelia in Ballo and that notorious voice-killer Minnie in La Fanciulla del West while still in her formative years.
Flagstad, however, seems a model of caution compared to the example of Rosa Ponselle, whose operatic debut took place when she was twenty-one, in the company of Enrico Caruso, at the Metropolitan Opera’s 1918 revival of Verdi’s La Forza del Destino. Imagine what would be said of a twenty-one year old making his or her debut under such circumstances today! It would be considered madness, and yet the most remarkable thing about Ponselle’s debut was how prepared she was for it, and how sustained was the success of a career that began so early.
Ponselle’s case may be something of an exception, but looking at the debut statistics of great singers from the past is both surprising and salutary. Callas debuted as Santuzza in Cavalleria Rusticana at the age of sixteen, sang her first Tosca at nineteen, and was offered a Met contract to sing Madama Butterfly and Fidelio just after her twenty-second birthday. (She turned it down.) Renata Tebaldi had performed in Mefistofele, L’amico Fritz, La bohème, Andrea Chenier and Die Meistersinger, all by her mid-twenties. The great 19th-century bel canto sopranos Giuditta Pasta, Maria Malibran and Pauline Viardot all made debuts while still in their teens.
Male singers, while not quite as precocious, also performed major roles at an earlier age than most of their modern counterparts. Enrico Caruso was twenty-two when he debuted in his native Napoli, and only twenty-seven when he sang Rodolfo in La bohème at La Scala under Toscanini’s baton. Caruso’s fellow Napolitano Fernando De Lucia debuted at twenty-five (as Gounod’s Faust) and sang extensively in Italy, Spain, South America and England before he was thirty. Baritones, too, started early. Antonio Scotti debuted as Amonasro in Aida at twenty-three, Giuseppe de Luca as Valentin in Faust at twenty, and Friedrich Schorr at twenty-four, not in some lyric role, but as Wotan in Die Walküre, a part he would own for decades.
Notice that all these singers began with major parts from the beginning, or very near the beginning, unlike today’s singers who often gain experience in Young Artist Programs, performing Parpignol and Annina while waiting to graduate to Rodolfo and Violetta. Astrid Varnay is another good example. She spent her late teens and early twenties learning languages and dramatic soprano roles, so that when the Met called on her to replace Lotte Lehmann as Sieglinde in its Walküre broadcast of December 6, 1941, she was not only ready, but could top the feat by replacing Helen Traubel as Brünnhilde in the same opera six days later! Not bad for a twenty-three year old who only began preparing for a career as a singer at eighteen!
Of course, almost all of these artists were born into a culture steeped in music and opera and had experienced it from an early age. Certainly, as daughters of Manuel Garcia, both Malibran and Viardot had been born into the art and business of opera. Both of Varnay’s parents were singers and her father the founder of the opera company where Flagstad sang early in her career. Moreover (and very much in contrast to today’s schooling of singers), operatic training in previous generations was daily and intensive. You didn’t see your teacher once a week, you saw her every day for several hours!
But the argument that we would do well today to return to intensive, daily training of singers, rather than the comparatively inadequate system employed at universities and conservatories today, while worth exploring, is only tangentially my point in this post. For I believe that by failing to turn out singers who can essay their properly chosen operatic roles in their twenties and early thirties, our musical culture is skewing the singing profession towards middle-age, and that singers and audiences alike are missing out on what should be one of opera’s greatest glories: the sound of a technically solid, youthful voice in full bloom.
In the normal course of things, and with proper training, singing should be a young person’s game. One often hears from voice teachers that the operatic voice doesn’t mature until one’s thirties or (they sometimes say) forties. I submit that this is not only untrue for the vast majority of singers, but that it is usually an excuse foisted on students by intellectually dishonest pedagogues who wish to explain away their trainees’ unreadiness to sing professionally.
What follows is an anecdotal survey, highly unscientific, meant to bolster my theory that with proper training, a normal operatic voice can take on leading parts in the singer’s twenties, and that these years, not the late thirties or early forties, represent the normal beginning of a singer’s prime. Moreover, I will demonstrate that the mid-to-late forties of even a good singer’s career often compare unfavorably with his or her late twenties and early thirties, in terms of vocal beauty, tonal steadiness, and stamina.
Let’s begin with the aforementioned estimable Wagnerian, Astrid Varnay. That remarkable debut at the age of twenty-three was, luckily, a Met Saturday broadcast, and we can hear it today. Varnay’s Sieglinde is partnered here by a fifty-one year old Lauritz Melchior, still in fine voice as Siegmund, though scarcely as disciplined and fresh sounding as he had been only six years before in Bruno Walter’s commercial recording of Act One of Die Walküre. (Better still are Melchior’s earlier commercial recordings of Siegmund’s music, which date from his thirties.) Varnay, as one would expect, is at her vocally freshest here at the very beginning of her career, but the voice sounds nothing like what the modern ear would associate with a singer in her early twenties. There is a depth and maturity in the sound that remind one of her idol and friend, Flagstad, despite no real tonal resemblance. She frequently attacks notes just under the pitch and then scoops up to the proper frequency, but she is hardly the first or last Wagnerian soprano to evince that tendency. She sounds like a mature, experienced dramatic soprano.
Fourteen years later, in August of 1955, we can hear Varnay again as Sieglinde, in a Bayreuth performance under Joseph Keilberth. She is now thirty-six, and sounds like the experienced performer she is. There is a generally more settled and “lived-in” quality to her singing here, even though Sieglinde was rarely her role in this opera. (She sang Brünnhilde far more frequently.) Vocally, however, the performances are remarkably similar and consistent. She still scoops, perhaps a bit less frequently now. However, just a few years later, by her early forties, Varnay’s voice began to betray the unsteadiness and intermittent tonal harshness that eventually led her give up hochdramatisch parts for lower-tessitura mezzo roles.
I can hear my voice teacher colleagues’ cries of foul! “Doesn’t this just prove,” they will ask, “that Varnay’s early assumption of Wagner roles hurt her voice?” My answer is that a run of twenty years singing Wagner’s heaviest parts at the highest international level is no career to sneeze at, especially as it was followed by at least another two decades as a beloved character mezzo. In fact, the length of Varnay’s prime as a singer is not so different from Flagstad’s or, later, Birgit Nilsson’s run as the world’s premiere Isolde and Brünnhilde. It’s just that Varnay had this prime in her twenties and thirties, where Flagstad sang her Wagner heroines in her forties and early fifties. Nilsson was actually born in the same year as Varnay, but did not begin an international career until the early 1950s, when Varnay had been singing at the Met for a decade already. Nilsson sang many of her great roles into the 1970s and 80s, but although she remained an audience favorite in New York, many of her later performances were plagued with intonation problems and a voice that had lost much of its beauty, though she never lost her clarion high notes. In short, each of these sopranos had her own glory years, but none of them was really viable in the roles that made them famous for much longer than two decades.
Let’s take the example of another supposed “late bloomer,” Joan Sutherland. The legend seems to be that she sang for a long time before she hit her peak years and then retained her great vocal form for decades. But recordings tell a rather different story. First of all, Sutherland made her debut at the not uncommon age of 25, and sang her first major role at Covent Garden (Amelia in Ballo) the following year. It’s true that she didn’t hit the “big time” until nearly a decade later, when she sang the title role in Zeffirelli’s production of Lucia di Lammermoor at Covent Garden under Tullio Serafin. But the years in between were not idle ones. She sang, mostly at Covent Garden, Desdemona, Gilda, Pamina, the Countess in Figaro, Donna Anna, Rodelinda, Eva, Lady Rich in Britten’s Gloriana, and Jenifer in Tippett’s Midsummer Marriage. These are not roles for a “developing” voice, and she was certainly already in her vocal prime when she sang them. What she wasn’t was famous. That all changed with the Lucia production, and in 1961, the “new” star recorded that opera for Decca. She was thirty-five.
And this is where the story gets interesting and where I part company with the usual legends about singers’ prime years. For although she sings astonishingly in that recording, she had in fact already recorded both the Act I scena of Lucia and the Mad Scene two years previously, in 1959, for her first recital album for Decca. Both the recital and the complete opera are marvelous recordings, but if you listen very carefully, you can hear on the earlier recording not only ever-so-slightly fresher tone quality, but better diction and a better integrated chest register. The loss is negligible on the complete recording, and both represent post-WWII singing at its finest. However, ten years later, Decca and Sutherland decided to record Lucia again, with the diva this time surrounded by a group of younger colleagues including Luciano Pavarotti and Sherrill Milnes, and conducted by her husband, Richard Bonynge. The critics were nearly unanimous in praising the new recording, at the expense of the 1961 version. In High Fidelity, Dale Harris proclaimed that “her first vocal splendor has yielded to a better-equalized scale, greater firmness of tone, and a sharper sense of attack.” I find none of that to be true. He further stated that “the years have diminished only minimally the freshness of her sound,” a defensive statement that falls far short of reality to my ears. The truth, plainly audible to anyone who cares to do an A/B comparison, is that in the 1961 recording, Sutherland sounds like a young woman, plausibly the age of the character, whereas in 1971 the immediate impression is of a matronly figure. In the love duet, the healthy “ping” and liquid sound of Pavarotti’s youthful voice (he was thirty-five) contrast uncomfortably with Sutherland’s rather lumpy phrasing and opaque tonal quality. Nor are her notes above the staff of the same pure and ringing quality that they were in 1959 or 1961; there is a sense of effort and “push” in launching them, though she gets there. As a young opera fan, believing the critics, I bought the 1971 recording, and I enjoyed it. The voice there is still beautiful and impressive, the coloratura still a model of clarity and legato. But to pretend, as the critics did at the time, that the ’71 version documented an improvement in Sutherland’s work is an example of the kind of wishful hearing that is often used when judging singers. “Well,” this line of thinking goes, “she’s at the height of her fame, she sells out the Met, everyone loves and respects her. Surely she’s at the pinnacle of her vocal prowess as well!” Well, no.
A cynical friend of mine says that record critics have always engaged in these dishonest shenanigans, simply to sell copies of the “new, improved” recording. Certainly that is sometimes the case: a prolific reviewer in the Gramophone used to embarrass himself monthly by basically proclaiming that each new recording of X was the greatest recording since the last recording of X. And, to be sure, there are occasional instances of a major singer actually continuing to improve musically and technically throughout a long stretch of his or her career: Placido Domingo sang generally better in the 1980s than he had in the 70s, and Karita Mattila went from being a middling Mozart lyric to a compelling singing actress who really sang in the 1990s. The exception proves the rule, though, and many of our greatest singers were at their best in their early years.
Both Dietrich Fischer-Dieskau and Gérard Souzay were masters of the art song who did most of their best work in their first decades of fame. I once made a list of my “desert island” Fischer-Dieskau recordings, and the latest of them, the 1966 Winterreise with Jörg Demus, was recorded when the singer was forty-one years old. All the rest of my choices came from Fischer-Dieskau’s twenties and thirties. It was a pleasure to see and hear him until the end of his long career, but his best recorded work dates from two plus decades before his retirement. The same is even more emphatically true of Souzay, who began with one of the most beautiful vocal timbres ever heard, but whose later recordings, meaning those dating from his forties and fifties, are frankly difficult to listen to.
Jon Vickers was another artist who was rightly hailed throughout a long career, but whose actual vocalism was undoubtedly strongest in the first ten or fifteen years that he sang. This did not prevent critics from clamoring for a second Vickers recording of Otello, which duly appeared in 1974 and was quickly forgotten. The 1961 RCA recording is the one people listen to and enjoy, even though Vickers had yet to sing the role of Otello onstage at the time he recorded it. I could go on and on citing many other beloved operatic artists: Callas, Tebaldi, Leontyne Price, Christa Ludwig, Mirella Freni, Fiorenza Cossotto, Nicolai Gedda, Sherrill Milnes, Nicolai Ghiaurov. All of them wonderful, all of them technically finer and more lovely of timbre in their early years.
One final example, with the aid of YouTube. I’m posting two versions of the love duet from Act II of Un Ballo in Maschera, both featuring Luciano Pavarotti. The first comes from his 1970 commercial recording for Decca, with Renata Tebaldi as Amelia. Tebaldi was in the twilight of her career at this time (a career that began in the 1940s) and is in far from her best vocal condition. The second recording, also for Decca is from thirteen years later and features Margaret Price. So these recordings represent Pavarotti in his thirty-fourth and forty-seventh year, respectively. They are both fine examples of his singing in a role that he performed often throughout his career. But note the remarkable freshness of timbre and liquidity of sound in the first recording, in comparison with the comparative stiffness and pressured quality of the second. For vocal beauty, finesse of phrasing and honest, unexaggerated vocal acting, I’ll take the early version.
The truth of the matter is that the wear and tear of an operatic career is hard on voices, and that whether a singer is young or less so, he or she must have a solid basis in a reliable technique to survive its rigors. Moreover, even fine and technically sound singers may begin to show audible signs of vocal erosion as they age. If singing is not, strictly speaking, a young person’s game, it is still a game in which many of our finest performers have thrived from their earliest professional days, provided they were technically and musically prepared.