Silvia Lelli

Željko Lučić as Macbeth in Peter Stein’s production of Verdi’s Macbeth, Salzburg, Austria, 2011

Verdi adored Shakespeare. Besides the three operas he took from him—Macbeth, Otello, and Falstaff—he considered (though briefly) doing a Tempest or Hamlet or Romeo and Juliet. He considered for a very long time, and came near to creating, an opera from his favorite play, King Lear.1 He did not take lightly the duty of being true to Shakespeare. When he read the score of Ambroise Thomas’s Hamlet, he said of the librettists, “Poor Shakespeare! How they have mistreated him!” He did not mean to mistreat the great dramatist himself.

Hundreds of operas were derived from Shakespeare’s plays—even more than from the works of Schiller, Goethe, or Walter Scott. Phyllis Hartnoll and her collaborators in Shakespeare in Music counted over 180 Shakespeare operas, but admitted they were missing some.2 The editors of The Oxford Companion to Shakespeare claim the number is closer to three hundred.3 Most of the Shakespeare operas in nineteenth-century Italy, France, and Germany were taken from the plays indirectly, from parallel sources, or from poor translations. Rossini’s Otello (1816), for instance, was based on a French adaptation of Shakespeare’s own Italian source, Cinthio’s Hecatommitthi. Bellini’s I Capuleti ed i Montecchi (1830) came from a play by Luigi Scevola.4 Verdi was the first Italian composer who worked hard to get back to Shakespeare’s authentic text.

Verdi could not read English—though his wife, who helped him, could—but he carefully compared the latest and best recent translations (some made by his friends or acquaintances).5 He had not been to England when he composed Macbeth, but he had acquired, from friends like Andrea Maffei, solid information on the way Macbeth was staged in the country of its origin. For Macbeth, he cut the play down to opera size himself, creating a prose synopsis for his librettist, Francesco Maria Piave, to versify. He was dissatisfied with Piave’s work, correcting it, adding suggestions, above all trimming it. He wanted no wasted words. He insisted to Piave, Poche parole! Poche parole! Poche parole! (Few words! Few words! Few words!) Finally, in his exasperation with Piave, he had his scholar friend Andrea Maffei, an expert translator, correct portions of the libretto.

Verdi worked himself so deeply into Shakespeare’s mind that in revising Macbeth for a Paris premiere, he took the gem of this performance—Lady Macbeth’s aria “La luce langue”—directly from Shakespeare, in collaboration with his wife: “Verdi himself actually wrote the text for this aria—not only the detailed prose version he first sent to [his librettist] Piave on December 15, but the verses themselves, to which the librettist made only a few minor changes.”

Most of the many operas made from Shakespeare’s plays are failures. Loose adaptations have been more successful—overtures, fantasias (like Tchaikovsky’s Romeo and Juliet), incidental music (like Mendelssohn’s for A Midsummer Night’s Dream), variations (like Berlioz’s Béatrice and Bénédict). The rare success of a complete Shakespearean opera—like Benjamin Britten’s A Midsummer Night’s Dream—is a one-off for its composer. Verdi is the only one who created three solid masterpieces from Shakespeare plays. They not only succeeded at the time of their premieres but have grown in reputation over the years, standing out even from his own extraordinary line of great works. The last two—Otello and Falstaff—are arguably the greatest things he ever wrote. He composed more operas from Schiller’s plays (four) than from Shakespeare, and some of those are very impressive—especially Don Carlos. But none towers up above his Shakespeare operas.

Verdi, across time and language barriers, obviously felt a great affinity with the dramatic ideals of his Elizabethan predecessor—and with good reason. The two men worked in theater conditions with many similarities. Both were supplying performances on a heavy schedule, to audiences with a voracious appetite for what they wrote. In a career of little over twenty years, Shakespeare turned out thirty-eight plays (along with some collaborations). Verdi had a longer career of fifty-four years—but with a sixteen-year inactive period between Aida and Otello—in which he created twenty-seven operas (along with important revisions).

These men were producing two major theatrical works a year during their most intense times, and were engaged in other poetic or musical compositions, as well as managerial and directorial work along the way. Shakespeare was composing narrative poems and sonnets. Verdi was composing religious and ceremonial and chamber music. Some of their contemporaries were even more prolific, especially Rossini in Verdi’s youth.6 The Elizabethan theaters were continually buzzing with new work, from dozens of aspiring playwrights, making the playhouses “pestiferous with plays” (as Bernard Shaw put it).7 The pace of professional life was unlike anything we see today, when a single play is kept in performance for long runs:


In the month of January, 1596…the Admirals’ Men played on every day except Sundays and presented fourteen plays. Six were given only one performance in the month, and no play was presented more than four times. The shortest interval between the repetition of any single play was three days, and the next shortest five. Although all except one were old plays, this record represents an achievement that would almost certainly be beyond the capacities of actors in the modern theatre.8

We can only be stunned at the memory powers of the actors on such a schedule. The opera houses of Verdi’s time were just as bustling with new works and crowded seasons.

Shakespeare and Verdi were creative volcanoes. But mainly they were men of the theater, engaged in the companies they worked with, active at each stage of the production of the plays and operas that filled their lives—Shakespeare as an actor in his own and other men’s plays, Verdi as a vocal coach and director of his works. Theirs was a hands-on life of the stage, not a remote life of the study.


Those who doubt that Shakespeare wrote Shakespeare are working, usually, from a false and modern premise. They are thinking of the modern playwright, a full-time literary fellow who writes a drama and then tries to find people who will put it on—an agent to shop it around, a producer to put up the money, a theater as its venue, a director, actors, designers of sets and costumes, musicians and dancers if the play calls for them, and so on. Sometimes a successful playwright sets up an arrangement with a particular company (Eugene O’Neill and the Province- town Players) or director (Tennessee Williams and Elia Kazan), but the process still begins with the writer creating his script, before elements are fitted around it, depending on things like which directors or actors are available for and desirous of doing the play. Producers complain that it is almost impossible to assemble the ideal cast for all the roles as the author envisioned them in his isolated act of creation. The modern writer owns the play by copyright and can publish it on his or her own, whether produced or not. None of these things was true of dramatic production in Shakespeare’s time.

Then, the process began with the actors. They chose the playwright, not vice versa. They owned the play, and could publish it or withhold it from publication. Each troupe had limited resources—often, nine to twelve adult actors (all male), and far fewer boy actors (sometimes as few as two). A Swiss traveler in 1599 saw “about fifteen” players handle the forty-five speaking parts in Julius Caesar.9 An aspiring playwright had to bring his idea to these actors (or their representatives) with a plot accommodated to the number and talents of the particular troupe. The parts he was describing had to be so arranged as to allow for multiple doublings. A man playing two roles could not meet himself on stage, or even come back in as someone else too soon to allow for costume and other changes (a beard, wig, spectacles, padding, and so on). “For some thirty-five years from 1547–8 plays advertise, usually on the title-page, the number of actors required and how the parts may be doubled, trebled, and even septupled.”10 In a 1576 morality play, The Tide Tarrieth No Man, the Vice character is told to prolong his duel “while Wantonness maketh her ready” in the tiring-house to come back out as Greediness.11 The plot had to be tailored for the company from the very outset.

If the actors liked the concept of a play, they would normally recommend it to a theatrical entrepreneur (Philip Henslowe was the most famous of the half-dozen or so working at a time) for an advance to the playwright while he finished the work. This advance was a loan, which the actors would pay back later, preferably from the proceeds of the play when performed. When the author finished writing his work, he read it to the company, which either accepted or rejected it at this point. If accepted, the script had to be presented to the Master of the Revels for state censorship, with payment for his reading it. He would often demand certain changes—or in some cases turn it down entirely. Only then, if cleared, could the play be put on. If, despite such screening, the play seemed seditious or libelous in the actual presentation, the actors were responsible along with the author and could be fined, suspended, jailed, even mutilated (by branding or ear cropping or nose cropping), or their theater could be closed.12


Thus, in the modern theater, performers are fitted to the play, but in Shakespeare’s time, the play was fitted to the performers. If the playwright had an ongoing relationship with the troupe—like Shakespeare’s with the Lord Chamberlain’s Men (later the King’s Men)—he could create his text for the known strengths of particular actors, as Shakespeare did for the talents of the great Richard Burbage. Shakespeare wrote comic scenes in different ways for the famous clown Will Kemp and for the intellectual jester Robert Armin. He even took advantage of animal performers available to the cast. When the troupe had a trained dog, he wrote the part of Crab into Two Gentlemen of Verona. When it had a young polar bear at hand, he wrote a scene-stopper for The Winter’s Tale: “Exit, pursued by a bear.”13 When he had two sets of players who looked alike, he wrote The Comedy of Errors. In modern productions, with an established text, producers can shop around in a large pool of unattached actors to find two couples who are plausibly similar, but Shakespeare began with the four men already in his company and wrote the play to use them.

The trickiest job was to write for that rare commodity, the boy actors who played women. These were hard to come by and train in the brief time before their voices broke. That is why women’s parts make up only thirteen percent of the lines in the plays. The playwright had to know what stage of development each apprentice had reached. There were usually just two or three boys in the public plays (though more were available from choristers when a play was given at court or in a great family mansion). The boys’ memories were such that Shakespeare wrote shorter parts for them than for adult actors—an average of three hundred or so lines to the adults’ 650 or so lines per play. But when he had a spectacular boy like John Rice, he was able to write as big a role for him as that of Cleopatra (693 lines). Nothing could be more absurd than the idea of the Earl of Oxford writing a long woman’s part without knowing whether the troupe had a boy capable of performing it.14 Only Shakespeare, who knew and wrote for and acted with and coached John Rice, knew what he could do and how to pace him from play to play.

An acting company could not just pick up any boy off the street. The boy had to have a good voice and memory and diction—and preferably an ability to sing, dance, and perform on a musical instrument (like Lucius in Julius Caesar). After the Swiss traveler saw Julius Caesar, two of the adult actors and the play’s only two boys came out and danced the after-show jig. Where to get such talented boys, and how to train them? Of course, there were all-boy theatrical troupes in the chapels and schools, very popular and with their own professional writers—but some parents resented even those boys’ playing in secular dramas.15 The public theaters had much more trouble finding players for their female parts. Sometimes they could persuade a boy from the chapel or school to join them, or find a middle-class family with surplus boys willing to apprentice one—or even buy an apprentice from a troupe about to be dissolved. It was clearly hard finding and keeping boys of the requisite skills. (How many had to be dropped because the promise of talent proved illusory?)

Each boy had to be adopted as an apprentice, to live with an adult actor’s family. The boys were given board, food, and training, but no wages. The adult master had to swear to the good morals of his charge, to fend off Puritan attacks on the immorality of the theater. Shakespeare never had an apprentice of his own, since he did not have his family with him in London. When any boy’s apprenticeship ended, he could take advantage of his training by becoming a paid adult member of the troupe, perhaps a sharer in its property and profits. After marriage, he could acquire his own apprentice.

There are many signs of Shakespeare’s crafting roles for particular boys. In three plays of the late 1590s, A Midsummer Night’s Dream, Much Ado About Nothing, and As You Like It, he had one boy who was short and dark and another who was tall and fair. The contrast was so striking that Shakespeare made his lines play on it.16 He began with particular boys’ talents, and then wrote his scenes around them. He must have had a boy from Wales when he wrote I Henry IV, in which a woman speaks and sings Welsh. One of the experienced boys, in As You Like It, was good enough for Shakespeare to create his second- longest woman’s role for him—Rosalind (686 lines).

Very young apprentices were cast as little boys rather than as women, and they were given small parts to memorize—Macduff’s son in Macbeth (twenty-one lines), Lucius in Julius Caesar (thirty-four lines), Prince Edward in Richard III (fifty-one lines), and so on. As they matured in age and training, Shakespeare could give them larger roles as boys, like that of Prince Arthur in King John (120 lines) or the clever Moth in Love’s Labor’s Lost (116 lines). Only as they advanced farther into their teens could he trust them with important women’s roles, and with the doubling made necessary by the small number of boys for female roles. A boy recruited at age eleven or twelve had perhaps five years of training and performance before him.

Such experienced boys, a rare resource, had to be used with great economy. Lady Macbeth, as important as she is, has only one brief appearance (twenty lines) in the last two acts of the play. That is because John Rice was needed to double Lady Macduff (forty-five lines) in Act 3. It is poignant that Lady Macbeth, who was not in on the murder of Macduff’s wife, somehow learned of it before the sleepwalking scene when she says, “The Thane of Fife [Macduff] had a wife—where is she now?” (Act 5, Scene 1).

Cordelia in King Lear is absent from the play for an even longer time than Lady Macbeth. After the play’s first scene (forty-one lines), Cordelia is gone from the rest of Act 1 and all of Acts 2 and 3. She shows up in the last two acts to speak only forty-eight lines. This seems a very uneconomical use of a trained boy, until we notice that Cordelia exits the first scene well before the Fool shows up, and the Fool disappears before Cordelia returns. The Fool is an innocent “natural,” sexless and accidentally wise, unlike the “allowed fools” played by Robert Armin (Feste, Touchstone). Armin, it seems, played the “mad” Edgar, while a boy doubled Cordelia and the Fool.17

Shakespeare was not a full-time writer without other responsibilities, like O’Neill or Williams. But what might look like a distraction for such authors—acting in his own and other people’s plays, coaching fellow players, helping manage the ownership of the troupe’s resources (including its two theaters, the Globe and Blackfriars)—was a strength for Shakespeare, since it made him a day-by-day observer of what the troupe could accomplish, actor by actor. The company was, after all, mounting plays with bewildering rapidity, studying, memorizing, and rehearsing in the morning and evening while performing in the afternoon. Without that experience, Shakespeare could not have written as he did. Lord Bacon or the Earl of Oxford, writing in their homes, could not have known such things. As Ivor Brown says, “Shakespeare was as much on and around a stage as in his study.”18


The working methods of a composer of operas in the nineteenth century had much in common with those of an Elizabethan playwright—enough to make them ideal subjects for a study in comparative dramaturgies. The playwright had to tailor his drama to the resources of a particular acting company. The composer had to fit his music to the voices available to him in a particular opera house. Rossini, in his long association with the Naples theater, the San Carlo, had to give most male parts to tenors, since that was where the troupe was strongest. In his version of Otello, the Moor is a tenor, Iago is a tenor, Roderigo is a tenor, the doge is a tenor, and the gondolier is a tenor. Only Desdemona’s father, not a very important role, is a bass.19 Among the women singers, Rossini had to compose with the skills of Isabella Colbran in mind, since she was the house star (and she would be his mistress, later his wife).

When Verdi did his version of Macbeth, he had only one weak tenor at hand, and he gave him the minor role of Macduff, with only one aria late in the opera. He did not let him sing his own climax-song (cabaletta) alone, but had the chorus join in with him. He had chosen Macbeth as a subject because he knew he could get by with only a minor contribution from the tenor, so long as he had a great baritone—which he demanded from the theater—to play Macbeth. When he received at first his commission for an opera at the Teatro della Pergola in Florence, he was initially undecided over setting Schiller’s Die Räuber or Macbeth. For the Schiller he would need a good tenor, and the manager said the house could not get the one Verdi had in mind (Gaetano Fraschini). So it had to be Macbeth. In modern performance, it is hard to get a star tenor to sit out the night waiting for his one aria. Only in a recording studio will Pavarotti or Domingo come in for the chance to sing “Ah la paterno mano.” When the opera was put on in Paris, the impresario defied Verdi’s express ban and expanded the tenor role, giving him an extra cabaletta and letting him repeat Lady Macbeth’s drinking song.

Even with a proven performer like Felice Varesi singing the role of Macbeth, Verdi was tailoring and adjusting the part as he composed—he gave Varesi three differently scored versions of his final scene, asking him which he thought would fit his voice better. When he sent her first music to his Lady Macbeth (Marianna Barbieri-Nini), he wrote, “If there should be some passage that lies badly [for her voice], let me know before I do the orchestration [for the passage that needs change].” He did the same with Varesi, writing him: “I’m convinced that the tessitura [range] suits you well, but there could be some notes or passages that are uncomfortable for you, so write to me before I orchestrate it.” He asked Barbieri about the state of her trill before writing trills into her role—after her assurances, he gave her many trills in her drinking song. Just as it would make no sense for Shakespeare to write a Cleopatra if he did not know John Rice’s ability, it would make no sense to write trills for a part unless Verdi knew he had a singer capable of them.

This was typical of the tailoring to specific voices that was demanded of the opera composer. He, like Shakespeare, had to be a man of the stage. He was expected to coach the singers and direct the first performances of his work. “According to Pacini,” Julian Budden writes in The Operas of Verdi, “it was the custom at the San Carlo theatre, Naples, for the composer to turn the pages for the leading cello and double bass players on opening nights.” The composer had to change his score to fit new voices if there were substitutions caused by illness or some other accident. In subsequent performances, he was expected to take out or put in arias for the different houses, transposing keys, changing orchestration. He was not a man of the study but of the theater.

This “hands-on” approach to composition had long been the case. As Budden writes:

In the eighteenth century…the singer, not the composer, was the starting-point. When Mozart was a youth no one would dream of composing an aria until he had first heard the artist who was to perform it; and this might be no more than a fortnight before the premiere. Thus, for instance, Leopold Mozart to his wife during the composition of Mitridate Rè di Ponte in Milan in 1770—“Wolfgang has composed only one aria for the primo uomo, since he has not yet arrived and Wolfgang doesn’t want to do the same work twice over.” More than sixty years later, when Bellini was writing I Puritani for the Théâtre des Italiens in Paris, the situation was no different. “The whole of the first act is now finished, except for the trio, because I want first to try it out (provarlo) on [the tenor] Rubini.” Provare is the word used for trying on a suit. Bellini’s contemporary, Giovanni Pacini, one of the most prolific operatic practitioners of his day, wrote in his memoirs that he always tried to serve his singers as a good tailor serves his clients, “concealing the natural defects of the figure and emphasizing its good points.”

Early in the nineteenth century, temperamental sopranos (or those sleeping with the theater manager) could just ignore the score if they did not like it and put in a favorite aria from another work—what were called “trunk arias” (arie de baule) since the singer carried around scores to be inserted at her whim. As Budden says, “It was only in England that Handel was able to hold a prima donna out of the window until she complied with his wishes.” Handel was escaping what Budden wryly calls “the tyranny of good voices.”

Some composers, Budden writes, kept singers from using other men’s work by yielding to their wish for a special aria from the maestro himself: “Verdi wrote a new tenor cabaletta for I Due Foscari and a new cavatina for Giovanna d’Arco to oblige the singers Mario and Sofia Loewe respectively.” In fact, he regularly composed substitute arias for singers in his first seven operas—abandoning the practice only as his prestige grew.20 Verdi was able to gain greater creative control over his work as his career developed. Later on, he could threaten a manager that he would withdraw his work from performance, even after the last rehearsal, if he did not get the performers, the rehearsal time, the direction and costumes and lighting he wanted.

There are other similarities between the conditions of Shakespeare and of Verdi, besides their writing for particular performers. Both had to put up with a demanding and prior censorship from the political authorities. Verdi’s problems were greater than Shakespeare’s. The British companies had to deal only with the state machinery of Queen Elizabeth or King James. But since Italy was divided into various political jurisdictions—papal, Austrian, Spanish, French—and each jurisdiction had competing authorities overseeing morals, the approval process was a marathon of trials. What was approved in one place could be rejected when the work was taken to another city’s opera house. This was not just a matter of minor changes. Often the whole libretto had to be rewritten in order to be politically acceptable. The time of the story, the locale, the characters had all to be changed. In Rome alone, these kinds of changes were required. As David Kimbell put it in his Verdi in the Age of Italian Romanticism:

Verdi’s Giovanna d’Arco became Orietta di Lesbo, Rossini’s Guillaume Tell became Rodolfo di Stirling. Throughout Italy, after the failure of the wars of independence in 1848–9, La Battaglia di Legnano became L’Assedio d’Arlem; Le Roi s’amuse was eventually rendered respectable by being transformed into Rigoletto.21

When Verdi planned an opera on the adultery of Gustavus III of Sweden, he was first compelled to change the story to avoid the assassination of a Swedish monarch. For performance in Naples, the opera became the tale of a Pomeranian duke, Una Vendetta in Domino (A Masked Revenge). But any European assassination was considered unacceptable, and the opera finally appeared in Rome as Un Ballo in Maschera (A Masked Ball), the story of a colonial governor of Boston—though, as Philip Gossett says, “To place Ballo in Boston is like setting La traviata in Munich.”22

Another resemblance between the theater of Shakespeare and that of nineteenth-century opera is the crossing of gender lines. The Puritans had kept women off the stage in England, and the papacy kept women from choirs in Rome. Boys had to play the roles in England, castrated men in Italy. Though the use of castrati was disappearing in the nineteenth century, it was suggested that the brilliant boy soprano Gioachino Rossini should become one.23 The castrato tradition shifted into the “pants roles” of male parts sung by women—like Rossini’s Tancredi, Donizetti’s Smeton, or Meyerbeer’s Urbain. Verdi has few women sing as boys, but there is Oscar of Un Ballo in Maschera and the boy apparitions of Macbeth. He also planned to make the Fool in his version of King Lear a woman.

All in all, the combined illusion and professionalism, the artifice and heightened reality, the searing poetry and melodic ambition, make the theater of Shakespeare and that of Verdi similar in many ways, obvious and hidden.