Chapter 1
A Short History of Conducting
It didn’t begin in a particularly mysterious way. Any group of musicians—whether they were monks in a ninth-century monastery or instrumentalists hired by an eighteenth-century Hungarian prince—always needed a leader: someone to run a rehearsal; someone everyone looked at to start a piece and perhaps to indicate its ebb and flow. In most cases, it was the composer.
The job of the conductor as we know it today came into its own in the middle and late nineteenth century, and not without a certain amount of controversy from those who would have preferred a less overt organizer. Robert Schumann called conducting “a mania” and “a necessary evil” in 1836, and Giuseppe Verdi was horrified to learn that conductors were taking a bow, since he saw them only as necessary functionaries in the performance of his operas. In 1872, he wrote to his publisher, Giulio Ricordi, “But now it is the style to applaud conductors too, and I deplore it.”
You can sympathize with Verdi’s point of view. Most of his operas, including Rigoletto and La traviata, were performed before the era of conductors and before opera houses were redesigned to have orchestra pits. Led by the principal violinist from a position among the orchestra players in front of the stage, the “leader” sometimes had to interrupt his playing to make signals to the singers and musicians using his bow. Toward the end of Verdi’s long career, which extended into the 1890s, a conductor was for the first time taking charge of rehearsals and performances of his Aida, Otello, and Falstaff at La Scala, while Verdi sat in his seat—or ran down the aisle to protest something he did not like, which apparently was fairly often.
That the role of a conductor became essential is inextricably linked to the development of Western music itself and the proliferation of concert halls and opera houses in cities. It was both a practical and a musical solution. The practical one is the development of notation—the indications of how a musical work should be played, written down so that the performer can replicate them without having heard the piece beforehand, and without the presence of the composer himself.
It started out as a kind of security device, once the idea of performing a symphony without the composer’s physical presence became the new normal, an impossible task to execute without his intentions clearly represented on the pages of the musical score. Remember that the vast majority of music, ever since the very beginnings of humanity’s desire to sing and dance, was passed on and invented by rote and imitation. For example, we have no precise idea of what the music played in Cleopatra’s court sounded like—though we know what instruments were used from the drawings and paintings left behind—since the remnants of its notational system are not understood. Something of the sound probably exists in the current folk music of Egypt, passed down for thousands of years. We can only imagine it, as Verdi did in his score to Aida in 1871, as Alex North did for his score to the film Cleopatra in 1963, and as Samuel Barber did in his 1966 opera, Antony and Cleopatra.
This is not true of the music of Bach and Mozart, whose music exists and whose notation is understood. Though stylistic details will always be a matter of interpretation, we do know the entire scores of Bach’s St. Matthew Passion and Mozart’s Don Giovanni.1 Without notation we would have only vague echoes of Bach and Mozart in our current music; we could never have a performance of the repertory we hold so precious today.
The notation system that developed over the centuries and is currently used worldwide began with the Roman Catholic Church and its desire to standardize the chants sung throughout its domain. Much of that system was expanded in the last 150 years. It is comprised of three elements:
(1) The notes themselves appear to float over five parallel lines called the staff, the invention of a monk called Guido of Arezzo about a thousand years ago. The musical notes are read from left to right, like words in a book. The higher the note on the staff, the higher the sound it represents. When there are multiple staves, as in an orchestral score, the notes that are precisely above or below other notes are meant to sound simultaneously. Thus a conductor can see how the flute part fits with the oboe part as well as the double basses. The flutist, the oboist, and the bass section, meanwhile, have no way of knowing what other people are playing, because their printed music does not include anyone else’s part. The conductor alone knows the composer’s complete intent, because he alone possesses a full score.
At first, conductors did not think it necessary to learn to read these multiple, simultaneous lines of music. Indeed, many orchestral works were not published with full scores, just the individual parts and/or a “piano/conductor” score that combined all the orchestra parts into two staves, as if the work had been composed as a solo piano piece. In some circumstances, the conductor was a violinist and led the orchestra by following only the violin part. In other circumstances, the leader was a keyboard player who led by nodding his head and playing a piano reduction of the symphony while the orchestra played their individual parts. This was true even in Beethoven’s day, even though there is no part for a piano in his orchestration. Both of these traditions make sense. In earlier times, instrumentalists took their lead from the harpsichord player, who had the most complicated and fully harmonized part—and who played, unlike other instrumentalists, from the first moment of a piece until the end with no pauses. At other times, as in the opera house, it was the principal violinist who would lead with his bow, taking on the complicated task of coordinating the singers with the orchestra. In the early twentieth century, it was a violinist who led the hotel and theater orchestras for the silent-movie era, as Eugene Ormandy did at New York’s Capitol Theater and my grandfather Baldassare Mauceri did at the Waldorf Astoria.
As music got more complicated, and with the desire to imagine new and unusual instrumental combinations, anyone attempting to conduct without a full score could easily have no idea if the players (and their individual parts) were correct. Monsieur Albert Yves Bernard, a teacher of mine at the Berkshire Music Center at Tanglewood, enjoyed telling the story of the conductor who tried to rehearse Paul Dukas’s Sorcerer’s Apprentice when Bernard was a young violist in France before World War II. Most people know this tone poem from Walt Disney’s 1940 animated feature Fantasia, though it was composed in 1897.
When the French orchestra got to the place where the music’s riotous first climax comes to a complete halt (where Mickey chops an enchanted broom to bits and it slowly comes back to life), the orchestra was reduced to short grunts punctuated by silence. There is a bar of music with only a contrabassoon and a bass pizzicato. The rest of the orchestra—some ninety players—were sitting quietly and counting. What followed was a bar in which no one played anything at all. A second or so later, the third bassoon, the timpani, and the cellos and basses played a single note in unison.
The conductor thought there must be something wrong. Either the orchestra was not counting correctly or the parts were engraved after lunch (when copyists, having enjoyed too much wine, were more apt to get things wrong or leave things out). And so he tried the section again with the same result. It turned out he was conducting from a first-violin part, not a full score. The first violins do not play in this segment of the piece, leaving him with an indication of eighty-four bars of rest—i.e., with no music in them—and no way to know who should be playing. His frustration turned to fury; he could not believe this was the composer’s intent. He shouted something unprintable, stormed out of the rehearsal, and refused to perform the work.
(2) In addition to using the five-line staff to tell a musician precisely what note is to be played and for how long, the bar line was added approximately five hundred years ago: This is an up-and-down line drawn through the staff that demarcates a complete number of beats of time before the next strong beat. Compared to music of other cultures—Indian and African, for example—Western music uses a very simple rhythmic system. The bar line will let a musician know when the required number of beats—its “meter”—has been filled within a measure. A work that has three pulses—a strong beat followed by two weak beats—has three pulses from beginning to end. Other works might have four or two. Music in five or seven was practically nonexistent, although it is a common meter in non-Western cultures. Up until the twentieth century, entire movements of symphonies were in one meter from beginning to end. In fact, most Western music was composed in one recurring meter, and when it comes to popular music, it still is.
Imagine, if you will, an entire musical language in which the pulse or meter remains unchanging from start to finish and you will understand how very simple this aspect of Western music is. (If you think of dance music, or a dance band of any kind, you can see why there is no need for a conductor. The phrase “One-two-three-four!” is all that is generally needed to set the music in motion until the end of the song or dance.) If a work is “in four,” there is a measure line every time four beats have been filled. The first beat after the bar line is the strongest, and here is where a conductor would give a strong beat, usually indicated by a downward gesture of the right hand. To this day, it is known as the downbeat. The last note of the measure was, conversely, the upbeat.2
The bar line acts as an interim demarcation for the players. With the left-to-right reading of the notes on the staff and the up-and-down line to indicate a repeating meter, Western music had a kind of musical graph paper from bar one to the double bar line that indicates the work has come to an end.
Without a conductor, players simply nodded their heads toward each other as a way of communication, usually at each bar line—something one can see at chamber-music concerts to this day. But that method of communication would become impossible as composers expanded the number of players into what we call the orchestra. (Mind you, visual contact and physical gestures among members of an orchestra section—violins, horns, etc.—is still essential, as well as between section leaders: for example, the principal cellist with the principal second violinist.) It also became impossible to perform orchestral music without a conductor because musical performance was finding a new way to expand the expressive possibilities of what looked pretty simplistic on the page. It made the repetitive meter far more complex and therefore more interesting.
(3) While this very simple—and practical—notational system efficiently gave the musician essential information, it did not say how the notes should be played. It did not tell him how fast the piece should be played (i.e., the tempo). It did not indicate whether it was loud or soft, or if the notes should be smoothly linked to one another or sounded separately. As the years went by, new music had multiple melodies sounding simultaneously, and sharply different densities in orchestration (who and how many people were playing at the same time). In addition, there was a continuing desire for a musical utterance not to exist in a single, repeated number of beats, so that a measure with three beats in it might be followed by one with four or two, and so on.
More important, composers and the public seemed to desire far more flexibility in the performance of single-meter music (the vast majority of music), a blurring of the bar lines that would give infinite nuance to the notated single and inflexible tempo. In other words, it was not just the musical content that was becoming more complex, it was the performance of music, both old and new, with regard to massaging the pulse and creating new levels of expressivity.
The Italians used the word rubato, which means “stolen,” for this performance practice. It means that the performers steal a bit of extra time in playing the rigid musical notation on the page in front of them. This might involve taking more time in an upbeat, or accelerating a sequence of notes so as to make it more exciting. This new expressivity—borrowing time from one place and paying it back in another—created demands from an instrumental ensemble that were impossible to organize without someone in charge of it. Without someone to determine how much and when, freedom within a tempo would become chaos, with a violist wanting to slow down in one place while a bassoonist wants to forge ahead. At every moment there would have to be a meeting of the orchestra committee to decide how to manifest expressivity and how to be—here’s a nicely ambiguous word—“musical.”
Understandably, composers themselves wanted to indicate whether the music should accelerate or slow down, even though, as I said, the meter—the number of beats in a measure—remained constant. New kinds of accents were invented. Dynamics, which had once been reduced to the simple concept of “soft” (piano, abbreviated as p) and “loud” (forte, abbreviated as f) now went from “very, very, very quiet” (pppp) to “very, very, very loud” (ffff).
The metronome came into use so that the precise speed of the music could be stated unambiguously by the composer himself.3 And to help clarify overall tempos, there were words—andante (“walking”), for example, and largo (“spacious”)—that were sometimes used without the metronome marking. When that was not enough, adverbs and phrases were added to better express the feeling or impression the composer had in mind. A score might start with an allegro brillante or an andante mesto (a sad walk) or, in the case of Wagner’s Tristan und Isolde, a langsam und schmachtend (“slow and yearning”).
Of course you have already figured out the problem here: What does “slow and yearning” mean in objective terms? Does one conduct the entire prelude to Tristan in one “slow and yearning” speed, as Bernstein does in his complete recording of the work? Does the “yearning” part of the instruction override the “slow” part as it reaches its climax, as everyone else does?
Wagner’s exact contemporary (they were both born in 1813) Giuseppe Verdi was perhaps the most precise of any composer of his time—before the explosion of indications in the music of Richard Strauss, Giacomo Puccini, and Gustav Mahler. Unlike Wagner, who disdained the metronome, Verdi gave the conductor both a metronome mark (the objective speed of the music) and how it should “feel.” Thus, in his Rigoletto a pulse of 66 beats per minute (i.e., a little bit faster than one beat per second) is marked andante sostenuto (a controlled and sustained walking tempo) in the prelude to act 1; andante mosso (a moving walking speed) for the duet between Rigoletto and the murderer for hire, Sparafucile, that begins act 1, scene 2; adagio (slow) for the opening of act 3; andante for the famous quartet; and largo for Gilda’s final solo, just before she dies.
Verdi clearly is telling us not only the feeling he wants but also how to achieve it. That music called andante mosso, andante, andante sostenuto, adagio, and largo is expected to be achieved with the same tempo is indeed astonishing, and points to a multiplicity of meaning in describing the same objective tempo, a singularity of the “building blocks of time” that Verdi used to construct his operas. Even with this specificity, no two performances of Rigoletto will or can adhere to all these instructions, even if the conductor attempts to observe them. Very few do, it should be said—something I shall return to later.
In general, notation developments in music tended to describe the performance practices—the interpretation of the printed page—of a previous generation. Thus, if you want to know how Wagner was being performed in his day, look to the notational vocabulary used in the printed scores of his successors Strauss and Mahler. But even as it got more and more specific, the score still had to be interpreted and adapted to very varied conditions—artistic, acoustic, and technical—no matter how faithful the performers were to the engraved musical text before them.
And therein lies the conundrum: notation, no matter how precise, still has to be interpreted. Every performance of Beethoven’s Symphony no. 5 contains the same notes, no matter who is conducting it and what orchestra is playing it. Every performer is attempting to do what is written in front of him, and yet no two performances are the same—nor can they be.
In music written up to the middle of the 1700s, each player could easily figure out how his part worked in the totality of an instrumental work. The tempo was fixed at the beginning, and rarely did anyone have to wait long before his instrument joined in. Most of the time, everyone was continuously playing and so there was no need to count rests; one just started at the beginning and went on until the end. The ensemble was small. Music was performed in a tavern, a palace, or a large room at a university. In Leipzig, the orchestra rehearsed and performed on the third floor of a large store that sold clothing and textiles. (To this day, that orchestra is called the Gewandhaus Orchestra—Clothing Store Orchestra—even though it performs in a concert hall built in 1981.) By the end of the nineteenth century, proper concert halls had been built in major cities in Europe, Russia, the United States, and Latin America, accommodating larger audiences as well as larger instrumental ensembles. Composers dreamt big, and it was not unusual for an orchestral work to demand more than a hundred players, as well as a chorus and vocal soloists. On September 12, 1910, Gustav Mahler unveiled his Symphony no. 8 in Munich, which became known as the “Symphony of a Thousand.” A photograph of its American premiere in March 1916 (Leopold Stokowski and the Philadelphia Orchestra) gives a good impression of the necessity of having a conductor to lead the new music being imagined and created at that time.
A publicity photo commemorating the American premiere of Mahler’s Symphony no. 8 in Philadelphia
Even in the first quarter of the nineteenth century, it was barely possible to perform new music, like a Beethoven symphony, without a conductor standing in front of the players. In one well-known case, the importance of a conductor was demonstrated when the orchestra was instructed to ignore him, since he was sure to throw them off. This was the touching first performance of Beethoven’s Ninth Symphony on May 7, 1824. Although completely deaf, Beethoven insisted on positioning himself before the orchestra in an attempt to conduct his new symphony—and so did the orchestra’s music director, Michael Umlauf. The orchestra watched Umlauf, while the audience’s attention was fixed on the composer turning pages and making gestures that were markedly at odds with the dynamics and accents they were hearing. Clearly Beethoven was frequently in another place in the score. And when the symphony ended, with its raucous and cosmic coda of joy, Beethoven was still conducting, imagining a slightly slower performance, totally oblivious to the standing ovation he was receiving. Only when the contralto soloist, Caroline Unger, put her hands on his shoulders and turned him around did Beethoven realize the triumph he had achieved.
In 1836, Robert Schumann wrote that “a good orchestra…needs to be conducted only at the start and at changes in tempo. For the rest, the conductor can quietly stand at the podium, following the score and waiting until his direction is again required.” That may have been true for the straightforward way he wanted music to be performed, but it soon became literally impossible to interpret large works without an all-knowing maestro, because composers were writing music in which the orchestral forces not only were expanding but were also being used selectively.
In 1850, a cymbal player in the German city of Weimar opened his brand-new part to the act 1 prelude to Wagner’s Lohengrin, which lasts about ten minutes, to find that he had only four notes to play—the first one coming after fifty-three bars of music in 4/4 time that is marked with a single word for its tempo: langsam, “slow.” Indeed, as he counted to four fifty times, he raised his crash cymbals to prepare for his first entrance. More than six minutes had transpired since the violins first began playing at the top of the piece. What if he had miscounted? His left hand moved upward as his right descended, cymbals strapped to his hands. How exactly was he meant to enter correctly without some indications, both in his part and from a conductor—the only person who knew what everyone was doing from the first chord? It is not easy to hide a cymbal crash marked fortissimo, and so the player had better be correct. At the world premiere in 1850, the conductor was Franz Liszt (Wagner was in Switzerland, having been exiled from Germany), and we can only assume that he looked at the cymbal player just before the climactic moment and that all went well.
This story opens up the larger issue of how opera made the modern maestro essential. For centuries, it had been assumed that the Greek plays were presented with actors speaking their lines. Once it was determined that the actors sang their poetry, a group of Florentine men, the Camerata, decided around 1600 to recreate the musical theater of the founders of Western culture, the Greeks, by telling great stories through sung poetry, movement, scenery, and costumes.4 Since the men involved in this pursuit were cultured Italians, they chose a Latin word, rather than a Greek one, to describe this new-old art form by taking the plural of the word opus, or “work,” and called it “the works,” or opera.
In order to perform an opera, more than one leader was needed to prepare, coordinate, and synchronize its many elements. Like high mass and festival rituals, opera was a very complicated business to compose, rehearse, and perform. Once it was time to perform the musical drama, a conductor led the instrumental ensemble from the keyboard, gesticulating to the singers and setting the tempos for the dances while watching their choreography.
Two centuries later, it was an art form that had moved from the court to the public theaters and attracted both the aristocracy and a mass audience of passionate citizens. Early on, it was noticed that the orchestra, with its double basses sticking up from the floor, was blocking the sightlines and distracting that part of the audience seated on the floor level (as opposed to the wealthy patrons in the boxes above). Even more distracting was the conductor. He usually stood on the floor facing the singers and dancers, but close to the apron of the stage with his back to much of the orchestra, who also sat facing the stage. He was both a prompter and a leader. In many cases, he was also a follower of the expressive singing, in which great interpretive artists added trills, cadenzas, and passing notes that were not written in their parts but were expected by the composer in the ever-developing styles of the times. This is an important situation, because the orchestra could see and hear the singers and therefore treat opera like chamber music, reacting to a singer’s breathing and body language while making use of the conductor’s beat when necessary for changes in tempo.
There were no balance problems between the singers and the orchestra, because the orchestra’s sound went toward the singers and away from the audience. We experimented with this practice at the Pittsburgh Opera with a production of Rossini’s La Cenerentola in 2003. The orchestra pit was raised so that the orchestra could see the stage and the players faced the singers, with me, the conductor, close to the edge of the stage. At first there was panic from the singers because the orchestra was so loud. We turned off all the monitors, the speakers that currently amplify the orchestra from the wings; they were no longer needed. Yes, we got complaints from a few of my colleagues about the basses sticking up and intruding on the illusion of the stage. But I have never found it easier to perform an opera from this period, and ultimately the experiment was a success with the singers and the delighted orchestra musicians, who were totally engaged in the production and its remarkable performers.
Opera was frequently composed for specific singers, and during preparatory rehearsals, notes might be added to or subtracted from their vocal parts to show the singers to their best advantage. All of this was normal and negotiated with the composer. Sometimes, when other singers essayed these roles, other arias might be substituted so that the new singer’s talents were better displayed.
As the nineteenth century progressed, the era of the bel canto diva who used every aria to show off as a singer—and thus stop being the character—faded away. In the 1860s, the most famous soprano of the age, Adelina Patti, sang Rosina’s “Una voce poco fa” from The Barber of Seville for Rossini, who had composed it in 1816. Patti performed the aria with added ornamentation—the roulades and cadenzas we have now come to expect in performances of this famous aria. Rossini, a notoriously acerbic wit, said to her afterward, “Wonderful piece. Who wrote it?”
In 1972, when Leonard Bernstein was rehearsing Carmen at the Met, there was a major contretemps with Marilyn Horne, who was singing the title role for the first time. Horne possessed one of the great mezzo-soprano voices of the century, and in the last verse of Carmen’s Habanera, she shifted into her powerful lower vocal range—her chest register—to sing the line, “Prends garde à toi,” slowing the music to a dead stop. Bernstein said, “Jackie [her familiar nickname], please don’t do that. All you are doing is calling attention to yourself here instead of being Carmen.” Horne did not agree. In her way of being Carmen, that oft-sung line needed a final and memorable declamation: “If you should love me, be very, very afraid.” She also pointed out that her fans expected to hear this unique part of her voice, made famous in her spectacular performances of earlier music, mostly by Rossini and Bellini.
At the dress rehearsal, packed with opera donors and fans, I sat behind Bernstein in the front row of the orchestra seats at the Met in order to take notes. When Horne got to that place in the score, she opted to sing it louder and slower than any human before or since. Bernstein, always a great accompanist, waited as she got to the seemingly endless “à,” instinctively knowing what she was doing, raised his right arm in a huge extended upbeat—and then on “toi” he gave a gigantic downbeat (the music is marked piano), which made the Met orchestra play fortissimo. Over the din he shouted a two-word obscenity unheard by anyone but me. I did not write it down on my legal pad. It was, needless to say, an unforgettable moment for this twenty-seven-year-old apprentice. (Later on we will explore the question of who was right.)
The very idea of a fundamental, correct, and unchanging text for an opera was long in coming, and Verdi was perhaps the first to insist on it. His compositional career, which included more than twenty operas, ran from 1839 to 1893. In other words, when it came to producing and performing operas, he saw it all—from a time when singers ruled the stage to one when a composer and his surrogate, the conductor, became the final word on how an opera was to be performed (except when a diva was not to be dissuaded, even by a Leonard Bernstein). The techniques required to conduct opera, as I said earlier, are by far the most complex and demanding. It is not surprising, then, that almost all of the great conductors of the late nineteenth century and the first half of the twentieth emerged from working in opera houses.
The nineteenth-century solution to the annoyance of having an orchestra with lighted music stands and the tops of the basses and bassoons intruding on a clear view of the stage required a technical solution that all but ensured the necessity of having a conductor: the orchestra pit. Lowering the floor in front of the stage became a partially successful solution that found its ultimate fulfillment in 1876 with the opening of Wagner’s dream theater, his Festspielhaus—festival house—at Bayreuth: the covered orchestra pit, partially excavated under the stage.
Wagner had written about the idea in the 1850s, well before his dream solution was fulfilled. Verdi supported it, believing it would also help blend the sonorities of the orchestra. “This is not my idea but Wagner’s,” he wrote. “It is excellent. It seems impossible that today we tolerate the sight of shabby tails and white ties, for example [the conductor], mixed with Egyptian, Assyrian, and Druidic costumes, etc., etc. [the costumed singers onstage], and, even more, the sight of the entire orchestra, which is part of the fictitious world, almost in the middle of the floor, among the whistling or applauding crowd. Add to all this the indecency of seeing the tops of harps, the necks of the double basses and the baton of the conductor all up in the air…”5
The Bayreuth Festspielhaus, whose design was supervised by Wagner himself, seated the audience in semicircles, as in an amphitheater. The auditorium’s seats rose gently upward from the first to the last row, with no side boxes for the aristocracy and every seat with a direct sight line to the stage. The orchestra pit was not only lower than the first row of seats, but it continued downward under the stage, so that the players were seated on three different and receding levels. The edge of the pit—a curving hood made of wood and leather—was designed so that no one in the audience could see the players. The sound that came up from below the stage bounced off a leather-and-wooden wall and was diffused, making it the opposite of “stereophonic,” or binaural, sound. The orchestra could be heard easily by the singers on the stage, and its volume in the auditorium was slightly reduced because of its circuitous journey from its source in the pit to the hood behind the conductor, from which it bounced upward toward the stage and then finally back from the stage to the audience. Since the major building material was wood, the entire auditorium acted like a gigantic string instrument that could subtly vibrate in resonance with the music being emitted by the hundred-piece orchestra. No one in the pit, however, could see (or clearly hear) the singers onstage. The only person who could manage the process of performing the music was a single man, seated at a large desk, who could be viewed by all the performers but was invisible to the audience: the conductor.
The English and Americans called the invention, first experienced between August 13 and 17, 1876, for the world premiere of Wagner’s complete Ring des Nibelungen, “the pit.” The French called it la fosse d’orchestre, “the orchestra ditch.” The Germans named it der Orchestergraben, “the orchestra grave.” The Italians, who had, after all, invented opera in the first place, came up with the most poetic name for it: il golfo mistico, or “the mystic gulf.” Whatever it was called, it was here to stay, and so was the maestro.