Giacomo Puccini
Giacomo Puccini; drawing by David Levine


The history of Italian opera throughout most of the nineteenth and twentieth centuries was directly tied to the fortunes of the publishing house founded in 1808 by Giovanni Ricordi. Casa Ricordi of Milan issued the work of composers of Italian opera from Rossini through Puccini to Luigi Nono and continued its efforts until the firm was acquired in 1994 by Bertelsmann (BMG). Bertelsmann’s pursuit of immediate profit took precedence over the cultural pride and long-term investment that had characterized the old Ricordi company; the restructured BMG Ricordi largely stopped publishing contemporary music. True, commercial interests had also been important for the earlier generations of the family. Yet Giulio Ricordi was able to tolerate the fiascos and partial successes of the young Giacomo Puccini during the 1880s because he believed the composer would ultimately develop a voice that would bring honor to Italian culture and lire to the company’s purse.

Ricordi’s investment was richly rewarded by the royalties from Puccini performances. By the early 1970s, however, their copyrights on the three most frequently performed Puccini operas—La Bohème (1896), Tosca (1900), and Madama Butterfly (1904)—had expired, and with them a significant source of income. Taking account of recent efforts to produce critical editions of nineteenth-century Italian operas, especially the works of Rossini and Verdi, Ricordi decided to undertake a similar program for Puccini, and they asked me in the mid-1970s to edit Madama Butterfly. When I tried to give myself a crash course in scholarly work on the composer, there was practically nowhere to go. Two pioneering books about Puccini had recently been published in English,1 and there were several editions of Puccini letters, but no one had apparently thought about musical sources, let alone documented the complex interaction between Puccini and his librettists. Critical studies rarely went beyond the level of program notes.

I tried to compare the printed score of Butterfly with Puccini’s autograph manuscript, preserved in the Ricordi archives in Milan, but I soon recognized how futile this was. Madama Butterfly, as I learned to formulate the problem a decade later, after the appearance of Jerome McGann’s treatise on textual criticism,2 must be seen as the product of a variety of collaborators. Unlike Italian operas from Rossini through Verdi’s Aida (1871), where the manuscript in the composer’s own writing remains our best source, Puccini’s works were subject to a more complex process of development, in which the composer was actively engaged but which he did not fully control. With Puccini’s blessing, editors, librettists, conductors, and other collaborators helped to bring the work into print. When the readings of the autograph manuscript differ significantly from the printed edition, therefore, which should we believe is the composer’s definitive work? A musical score cannot be treated as an example of postmodern ambiguity: rehearsal time costs money.

The first act of Butterfly, for example, begins with a bustling fugue. In Puccini’s autograph manuscript and in the original vocal score, the first two notes are always slurred. When the orchestral score was published, however, the articulation was changed in the strings. Someone—we don’t know who—decided that slurring the first two notes on a single upbow and reserving a downbow for the third note, which falls on the downbeat, would sound like a weak attack, and so the slurs were replaced with a pair of staccatos, and the two notes were marked with a downbow and an upbow. While that may be a reasonable choice, no parallel changes were made in the doubling winds. Thus, the printed score gives the opening notes as staccato in the strings and slurred in the winds. Did Puccini or anyone else realize this? Should the contradictory readings be allowed to stand?

When Butterfly’s servant, Suzuki, prays to the Japanese gods at the beginning of the second act, her melody in D minor rocks gently between A and G. The preceding measure resembles a battleground in Puccini’s manuscript. After staring at it for hours, I realized that the composer called on one of the horns to descend from A to G, anticipating the melody of Suzuki’s prayer, but no edition of Butterfly has ever included that G. Did Ricordi make a mistake, which went unnoticed, or did Puccini remove this genial detail? There was no way to know. Nor in the critical climate surrounding Puccini in the 1970s did anyone think such questions important.

The situation has changed dramatically today. Although some critics continue to resist the allure of Puccini’s operas, the arrogance and intolerance that characterized much “informed” opinion before the 1970s has largely evaporated.3 Not only do La Bohème, Tosca, Butterfly, and Turandot (unfinished at Puccini’s death in 1924) flourish on the world’s stages, but opera houses are more likely than ever to program revivals of Manon Lescaut (1893), La fanciulla del West (1910), La rondine (1917), and Il trittico (1918), and even to produce Edgar (1889) and Le Villi (1884). Between Rent and Baz Luhrmann’s recent La Bohème, Puccini’s Bohemians have been up and down Broadway, as have various derivatives of Butterfly.


Instead of accepting critical theories that define operatic drama from a Wagnerian perspective, new scholarly work looks more closely at Puccini’s operas themselves. What emerges is a view of Puccini as having a keen dramatic sensibility, not at all a composer of shameless sentimentality. Puccini’s music and acutely theatrical timing revivify the late-nineteenth-century melodrama of Victorien Sardou and David Belasco, much as Rossini reanimates the theater of Voltaire, and Verdi that of Victor Hugo. Puccini’s melodic language is sumptuous but hardly ever coarse; in each opera he carefully develops a discrete number of musical ideas. He never presents an idea at random, not even the reprise of “E lucevan le stelle“—“When the stars were brightly shining”—at the end of Tosca; no audience fails to understand its emotional resonance. The composer’s carefully wrought orchestral texture often resembles that of chamber music, especially in the more experimental works of the 1910s, among them La fanciulla del West and Il trittico. His basically tonal harmonic language is inflected throughout in ways that show knowledge of his contemporaries in France (Massenet and Debussy) and Germany (Wagner and Strauss), but it remains deeply personal. The recent appearance of three books in English offers an opportu-nity to reflect on the man and what he accomplished.


Giacomo Puccini was the last in a family of Tuscan musicians stretching back to the early eighteenth century. They settled in Lucca, where Giacomo was born in 1858. His early life parallels Verdi’s: an excellent young musician is destined for local ecclesiastical service; well-to-do supporters make it possible for him to receive training in Milan; some youthful operas have a mixed reception; then a big success (Nabucco, Manon Lescaut) opens the way to the future. Throughout his career, Puccini, like Verdi, remained intensely loyal to his geographical and cultural roots. From the early 1890s his private life centered around his property in Torre del Lago, and he never abandoned its simple pleasures: hunting, cards, quiet camaraderie. Verdi the farmer discussing hydraulic projects, Puccini the hunter, his rifle and game birds proudly displayed; both composers cultivated these images, while genuinely enjoying provincial life.

Mary Jane Phillips-Matz has writ-ten a biography of Puccini with much anecdotal detail, some colorful, some tedious. She has traced Puccini’s movements, sought out primary sources where possible, and over the past forty years interviewed many of those who knew him directly, friends, relatives, and fellow artists, most of whom have emphasized the composer’s humanity and modesty. She is not always sufficiently selective: scouring auction catalogs for letters can be worthwhile, but Puccini’s effort to help a young corporal get a job in “some regimental office” adds little. Yet one savors the anecdote told by Arturo Buzzi-Peccia, a friend from Puccini’s student days, claiming that he and Puccini cheated at cards “by humming a melody that was in fact a code that corresponded to the values of the cards they held.”

When it comes to documenting Puccini’s intimate life, Phillips-Matz is unrivaled.4 There is much to tell. In the mid-1880s he fell in love with one of his piano students, Elvira Bonturi, the wife of a grocer and traveling salesman, Narciso Gemignani, with whom she already had two children. By 1886, pregnant with Puccini’s son Tonio, Elvira left her husband and children to join the composer. They married after her husband’s death and remained together throughout Puccini’s life, although their relationship was never easy. Burdened with Elvira’s consuming jealousy, the composer nonetheless had a series of relationships with other women. The Turinese “Corinna,” with whom Puccini was involved from 1900 to 1903, remains shadowy. It seems peculiar that no one has succeeded in identifying her, particularly since their affair ended with her threatening legal action against him for refusing to marry her.

In October 1904, during a trip to London, he met Sybil Seligman, wife of a London banker, with whom he had (according to Sybil’s sister) an intense but brief sexual relationship that was followed by a lifelong friendship. Selections from Puccini’s letters to Sybil, many about artistic matters, were published by her son in English translation in 1938, but the originals—some owned by an Australian collector—remain unavailable. There was no more than casual affection between Puccini and his serving girl, Doria Manfredi, but Elvira’s vociferous claim that the girl was a whore drove Doria to suicide (an autopsy established that she was a virgin).

While Phillips-Matz presents little new information about Elvira, Corinna, Sybil, or Doria, she is the first to quote from Puccini’s correspondence with two other women, Margit Vészi, a Hungarian writer, whom he encountered around 1912, and the singer Rose Ader, whom he met in 1921, writing to her in over one hundred letters Phillips-Matz describes as “all very passionate.” She also quotes liberally—in English translation—from the letters of Puccini to Toscanini, now in the New York Public Library.5 These letters (which describe, for example, how he “filled out” the final scene of Fanciulla with choruses) draw us closer to Puccini the composer, who alone justifies a biographer’s efforts. More thoughtful attention to documents of this kind and to Puccini’s other artistic relationships would have improved Phillips-Matz’s book.


Julian Budden and Michele Girardi, without neglecting the composer’s biography, concentrate their attention principally on his operas.6 What, they both ask, was a composer to do after Aida? The evolving musical language that had defined Italian opera from the 1810s through the early 1870s had played itself out. Verdi himself filled the operatic gap between Aida in 1871 and Otello in 1887 (followed in 1893 by Falstaff) with revisions of earlier works (Simon Boccanegra in 1881 and the Italian Don Carlo in 1884). Of the generation of composers who matured during the 1860s and 1870s, the most prominent were Amilcare Ponchielli and the Verdi librettist Arrigo Boito, known primarily as a composer for his often-revised Mefistofele (1868, 1875, 1881). Boito’s efforts to invent a new style alert to influences from north of the Alps soon languished, along with the entire musical faction that, under the banner of scapigliatura (“bohemianism”), hoped to rejuvenate Italian musical culture. Ponchielli’s art also looked north, especially in his La gioconda of 1876 (usually performed in its definitive 1880 version). The post-Wagnerian atmosphere that gradually pervaded Italy after the 1871 performances of Lohengrin in Bologna, however, offered little space for the techniques of French grand opera on display in Ponchielli: spectacular staging, divertissements (especially the “Dance of the Hours”), a strong choral presence, and a relatively fluid approach to musical form.

Composers who matured during the 1880s, the giovane scuola italiana (“young Italian school”), took a different path. Alfredo Catalani, Umberto Giordano, Ruggero Leoncavallo, Pietro Mascagni, and Giacomo Puccini each developed a musical language that drew on disparate influences but avoided being dependent on models from Germany (particularly Wagner) or France (particularly Massenet). Some of their works were labeled verismo, stressing a crude realism that became synonymous with the presentation of sensational plots, but the term proved elusive.

Some characteristics of the scuola can be identified. They used a freer, more conversational idiom, replacing the high diction of opera librettos with the language of the contemporary realist theater, taking Sardou and Belasco as their models. Plots often drew on lower-class or even tawdry characters, Bohemians, artists, actors or singers, and peasants (as in the stories of Giovanni Verga). A sense of “local color” became central, and composers adopted musical ideas that would suggest exotic geographical settings, often using preexistent melodies (Chinese, Japanese, American, church music, folk tunes). The reign of fixed musical forms for arias, duets, and ensembles that had dominated Italian opera was over, and composers sought new ways of providing continuity, often by exploiting a succession of orchestral motifs, as Ponchielli had done through the theme associated with the evil Barnaba in his La gioconda. These were far removed from Wagnerian leitmotifs: they rarely underwent symphonic development. Although the composers of this school rejected more florid bel canto elements, they developed a new kind of declamatory lyricism to help differentiate the Italian school from northern models.

It is no wonder that the first two great successes of the giovane scuola were relatively short operas about love and death: Mascagni’s one-act Cavalleria rusticana of 1890 and Leoncavallo’s I pagliacci of 1892 (written in two brief acts, usually performed without a break). Within these compressed forms each composer demonstrated strong lyrical gifts, as well as the ability to contrast music heard by the singers on stage as part of the story (songs and sacred hymns in Mascagni, a commedia dell’arte accompaniment in Leoncavallo) with intense dialogue. Neither composer, though, was able to sustain his art effectively in a more broadly conceived opera, and neither reproduced his youthful achievement.

Of the giovane scuola, Puccini was the most successful at composing full-length operas, but that success was hard won, as can be seen in the composer’s correspondence with librettists, publishers (first Giulio Ricordi, then his son Tito), performers, and the nonmusical friends with whom he discussed his art. These letters express uncertainties and self-doubts, alarming to collaborators, which often led to fallow periods. Even after an opera was well established in the repertory and its text fixed, Puccini followed his works all over Italy, Europe, the British Isles, and even North and South America. He did so ostensibly to stage them, but more often to gain respite from his external and internal demons.

Whereas Verdi from early in his career was self-assured in his dealings with theaters, publishers, and librettists, Puccini was often tentative. Verdi criticized the work of his collaborators mercilessly, but his demands were usually clear and precise: this subject is for me, he would say; this one is not; since I’ve already written the melody, use the following meter; provide an extra strophe for the soprano here and two verses for the chorus there. Puccini, on the other hand, vacillated constantly. For months he encouraged a librettist to pursue a subject, then abruptly changed his mind. Luigi Illica, who (together with Giuseppe Giacosa, a better-known poet and playwright) provided texts for Puccini’s best-loved operas, complained to Ricordi early in January 1893:

Puccini has confided to a friend of his that he could well do without my libretti…and, besides, nobody understands him because he longs for something…something …something…that…

You will see that this “something,” expressed like that, is very difficult to interpret. It means that amid this darkness I should have to grope around to find out just what is this “something” on which Puccini has set his heart, only to hear him answer “I don’t like it.”

For several years during the early 1890s Puccini pursued a story by Verga, La lupa (The She-Wolf), and began setting a libretto fashioned by Verga and Federico de Roberto. Then in June 1894 the composer went to Sicily to discuss the project with Verga, but came back discouraged:

Since my consultations with Verga, instead of being stirred to enthusiasm over La lupa, I’ve been assailed by a thousand doubts…. My reasons are the abundance of dialogue, carried to extremes in the libretto and the unpleasantness of the characters, without a single luminous figure that stands out…. My only regret is for the time lost, but I will make up for it by throwing myself heart and soul into La bohème.

Puccini perceived theatrical problems with keen intelligence, but the contorted ways by which he arrived at such decisions made life impossible for those around him. With Illica he spent ten years considering and three years working on an opera concerning the imprisonment, trial, and execution of Marie Antoinette; he then abandoned it.7

Even after a project was well advanced, significant dramaturgical changes could follow. Puccini insisted that an entire act—a scene in the cortile, or courtyard—be removed from the libretto of La Bohème,8 a cut Illica described as an “enormous wound.” The libretto of Turandot mutated from two acts to three while he worked on it.9 Not all the vast amount of material pertaining to the genesis of Puccini’s librettos is easily available: neither the Ricordi archives in Milan nor the Illica archive in Piacenza nor the Giacosa family archives in Colleretto Giacosa have been fully tapped. Recently discovered letters from Puccini to Carlo Zangarini (the author, with Guelfo Civinini, of the libretto for Fanciulla) remain to be published. While Julian Budden and Michele Girardi draw on these sources, one could hardly expect their panoramic surveys to do justice to the history of any particular opera. They nonetheless provide sufficient material to give readers an introduction to the circumstances in which each work was conceived and the issues of interpretation that have arisen concerning them.

However hard Puccini was on his collaborators, he was even more exigent with himself. The more one knows about the history of the operas, the more unfair is the charge of “laziness” or “superficiality” that is often brought against their composer. He would spend years bringing an opera from its earliest musical sketches to a producible version, but the première was only the initial step in a process of revision, contraction, and expansion, as the opera took shape during performance. Between 1904 and 1906 Puccini revised Madama Butterfly four times, shortening the early scenes with Butterfly’s relatives, presenting Lieutenant Benjamin Franklin Pinkerton as less odious than in the earliest version, and reducing the presence of Pinkerton’s American wife in the final scene. These revisions focus attention on the plight of the heroine and increase the opera’s emotional impact.

Although Puccini changed his original two acts to three, largely because of the unusual length of what had been the second act, he did so with great reluctance. Modern performances often return to Puccini’s original version, so that the audience waits through the night with Butterfly and Suzuki for the arrival of Pinkerton’s ship. Nor did Puccini hesitate to enlist the participation of trusted colleagues. The critic Gabriele Dotto has referred to La fanciulla del West as “Opera, Four Hands,” in view of the extensive interventions of Toscanini, who conducted the world première in New York. He also had a significant part in preparing the final revision of Manon Lescaut.10

While various versions of Puccini’s operas have often been discussed, most remain unavailable for study and performance. There is also little published work about the methods by which he composed. Some photogenic sketch pages are regularly reproduced, but the hundreds of pages of work on La fanciulla del West in the Koch collection at the Beinecke Library of Yale University still await study. Other sketches in Italian collections are currently inaccessible.

Those at Ricordi for the final scene of Turandot, left unfinished at the composer’s death, on the other hand, have attracted much attention, both from scholars (such as Teodoro Celli and Jürgen Maehder) and from composers (most recently Luciano Berio). All expressed doubts about the quality of the opera’s finale prepared by Franco Alfano on the basis of these sketches and still regularly performed, often in an abbreviated form. Yet work on the Turandot finale has been carried on without sufficient reference to material that might help clarify Puccini’s intentions. Examination of Puccini’s work on Fanciulla, for example, would have allowed scholars to compare his sketches with the completed score, thereby developing some understanding of the process that led from one to the other.


Julian Budden takes a standard approach to Puccini. Except for two introductory biographical chapters about the composer’s youth and a brief conclusion, each chapter of his book centers on a specific opera. Within these chapters Budden traces the composer’s life at that moment and the genesis of the opera, and he then analyzes its dramatic shape and music. The sections on biography and on the origins and different versions of Puccini’s operas demonstrate the mastery that has characterized Budden’s writing since his three-volume The Operas of Verdi.11 He is responsible for much of the new Puccini scholarship and his synthesis of this material is often elegant.

He is less successful in his treatment of the operas themselves. Providing a continuous commentary on the plot and the music worked reasonably well with Verdi, because the composer generally constructed his operas using separate musical numbers or scenes. Puccini’s works, on the other hand, depend on an intricate web of melodic fragments, which do not lend themselves to being separately named. As a result, a blow-by-blow analysis, laced with adjectives, results in the following typical paragraph, which describes the moment in the second act of Tosca when the heroine, unable to endure the torture of her beloved Cavaradossi, reveals to Scarpia the hiding place of Angelotti:

A cry from Cavaradossi, and Tosca reaches the end of her tether: “Nel pozzo…nel giardino” [“In the well…in the garden”], forestalled by the orchestra with Ex[ample] 7.10 fortissimo. Cavaradossi is carried in unconscious. Ex. 7.12 at its most funereal, with violas and bassoons on the melody, is dovetailed into the painful Ex. 7.13, over which Ex. 7.9a pours its balm as Tosca tries to revive her lover with kisses.

However accurate Budden’s observations may be, this kind of prose is tedious, even indecipherable. Budden is well aware of the musical issues in Puccini’s operas, such as the interplay of motives, his use of tonality, and his borrowing from his own work or from that of other composers (including Wagner). But when Budden begins to address these issues, whether as they concern an individual opera or as Puccini’s practice throughout his entire career, the tyrannical organization of his book does not permit him to develop his argument.

Girardi’s approach is bolder. He does not summarize the plot of each opera or give an extended discussion of its music, although he can deal clearly with both when he finds it appropriate to do so. Alongside his historical and biographical discussion, however, Girardi concentrates his attention on what he considers to be the most significant aspects of each opera. Thus, for Madama Butterfly, he discusses “Realism” and “Exoticism” in earlier nineteenth-century opera and analyzes Puccini’s use of actual Japanese and American melodies, as well as his invention of a Far Eastern style.

Girardi documents Puccini’s use in Butterfly of a complex web of themes that differentiates the imagined world of the heroine, in which she believes against all evidence that Pinkerton will return to her and her child, from the reality of her situation—her poverty, the curse of the Bonze, the efforts of Sharpless to make her face the truth. For Girardi, Puccini’s themes represent “conceptual fields”: each theme refers to different elements in the drama that group themselves into broad categories of human experience—among them death, love, illusion, destiny. This technique is different from Wagnerian leitmotif, as Girardi explains in his discussion of Manon Lescaut:

But Puccini’s engagement with Wagner…is particularly clear in the rigor and consistency with which he made use of leitmotivic technique, combining it with an Italian conception of musical drama whose mainstay was melody. The thematic material employed in the opera is set out in a clear system of relationships, one that ties the characters to the situations they experience and their relative states of mind. The music, liberated from simple narrative necessities, serves to create sophisticated symbolic associations.

Invoking this principle for Butterfly, Girardi offers many musical examples, but they are always part of a cogent argument, and he thereby avoids the trap into which Budden falls.

More controversial is Girardi’s approach to the way Puccini puts these themes together to create musical structures supporting the drama. Some of his diagrams analyzing La Bohème suggest an indigestible alphabet soup, with recurring themes labeled, for example, ABABABCADEE. Girardi imposes on Puccini’s music various patterns that are familiar from Rossini through middle Verdi (using terms such as “tempo d’attacco,” “cantabile,” “tempo di mezzo,” and “cabaletta”) but that are quite irrelevant to Puccini’s art. There is no reason to suppose that Puccini was either consciously or unconsciously turning to these earlier structural models for his operas, quite unlike Alban Berg’s explicit invocation of models from the Baroque and Classical periods in his Wozzeck.

Nor does Girardi always find his bearings in dealing with the equally contested question of tonal analysis. To what extent did Puccini employ different keys in his operas either as part of a structural plan or to suggest dramatic associations?12 While Girardi frequently invokes tonality, he is not always convincing. Particularly suspect is his interpretation of the ending of the first and third acts of Butterfly, including the “chords of the theme that accompanies the heroine’s dying breath, the theme that recalls her fatal destiny….”13 In general, Puccini’s harmonic practice deserves further study.

Between the 1970s and today, studies in Puccini’s life and art have flourished. Each of the books under review gives some idea of current research, but much else remains to be done.14 Most of all, the entire corpus of Puccini’s operas, in their various versions, needs to be made available in more accurate and complete editions. His sketches and manuscripts need to be investigated, together with the history of the librettos of his operas. Any such project will require developing new editorial standards, appropriate to Puccini’s repertory. BMG Ricordi should give such a project the highest priority. If the scholarship of the past thirty years has taught us anything, it is that the operas of Puccini deserve closer and more thoughtful attention than they have had so far.

This Issue

March 27, 2003