67. La bohème

As early as six months before the first performance of Manon Lescaut (64), Giulio Ricordi, in a letter to Puccini dated 22 July 1892, mentioned La Bohème as a possible next opera project.Note: *Original in I-Lmp, a copy in MRC 1892/93 - 2/273. According to that letter, the idea to make Henri Murger’s magazine articlesNote: *Murger’s “scènes” were originally published between 1845 and 1849. about the lives of fictitious Parisian artists’ the subject of an opera originated even earlier.
In 1851, Murger collected those articles and published them as a sort of novel. Prior to that (in 1849), a dramatization written in collaboration with the playwrite, Théodore Barrière, was presented in Paris under the title La vie de Bohème. From this material, Giulio Ricordi commissioned an opera libretto, which he offered to Puccini along with other suggestions for operas after Manon Lescaut.
For some time this proposal was in serious competition with another subject Puccini was considering: an opera based on Giovanni Verga’s story La Lupa. The choice remained undecided until the summer of 1894, when Puccini travelled to Sicily to discuss the project with Verga. Ultimately, Puccini was determined to compose La Bohème (see Epistolario 1982 No. 36 of 13 July 1894), the libretto of which by that time had already progressed far beyond a scenario.
Puccini’s librettists Luigi Illica and Giuseppe Giacosa had already begun working steadily on the libretto at the end of March 1893, when it became known that Ruggero Leoncavallo was also about to set the Murger subject to music. The famous controversy between the two composers and their publishing houses of Ricordi and Sonzogno concerning who was the first to stake their claim to the subject cannot be reliably determined on the basis of the surviving documents. Much, however, suggests that Leoncavallo was interested in the Murger source material considerably earlier than Puccini. It is also possible that Leoncavallo himself might have written the first draft for the Ricordi libretto. That conjecture is also strongly supported by the surprising similarity in the selection of scenes for both operas,Note: *That composing La Bohème was Leoncavallo’s idea first, can also be derived from Marotti and Pagni 1926, pp. 47-48, a 
source close to Puccini; see also Maehder 1986, pp. 124-25.  independent of each other, yet finished virtually in tandem. In any event, Illica seems to have been involved with Puccini on this project no earlier than March 1893.Note: *See Ricordi’s correspondence to Illica from 1892-93 in I-PCc. Giacosa was also apparently in place as co-author of the libretto before the public controversy between Puccini and Leoncavallo began (see the Gazzetta musicale of 19 March 1893, p. 203, in which a report emanating from Lucca on 12 March, stated that “the libretto will be written in collaboration by two artists, whose names are already favorites and acclaimed for their art of writing for the stage”Note: *“Il libretto verrà scritto in collaborazione da due artisti, i cui nomi sono già cari ed acclamati nell’arte drammatica." — information which doubtless originated directly from Puccini).Note: *See also Giacosa’s first known comments regarding this matter in Carteggi 1958 No. 82 of 22 March 1893.
The development of the libretto of La Bohème turned out to be hardly less difficult than the libretto of Manon Lescaut. Again there was the matter of a complete act (set in the courtyard of 8 Rue Labruyère [“Il cortile della casa di via Labruyère 8”]),Note: *The original manuscript of the text of that act (not very correctly published in La Scala, December 1958, pp. 37-49) is preserved in the Museo Luigi Illica, Castell'Arquato (Piacenza); it is more accurately transcribed in Groos and Parker 1986, pp. 152-81.  which was ultimately eliminated. That decision was probably made in February 1894. It was to have been Act 3, preceded by what is now the third quadro (“La barriera d’Enfer”) as Act 2. What is now the first and second quadri were (together) to have formed Act 1. The end, of course, was what remains the fourth quadro, with Mimì’s death as Act 4. The elimination of the “Cortile” act lead to a break in dramaturgical logic, because Rodolfo’s motive for his separation from Mimì is no longer apparent. At times the project was on the brink of failure (see Carteggi 1958 Nos. 91-93).Note: *The somewhat confusing origins of the libretto have been examined in depth and are discussed in particular in Groos and Parker 1986, Maehder 1986, and Maehder 1987. A short description of the unpublished drafts of the libretto is given by Bice Serafini (in Critica 1976, pp. 119-23). They are preserved in the Archivio Giacosa in Colleretto Giacosa (Province of Turin), and have been examined more closely by Burton 1995 and Gillio 1998. Furthermore, a publisher’s proof of a libretto of the fourth quadro with autograph annotations (“un libretto dell’opera ‘La Bohème’, atto IV°, — bozze di stampa — con appunti autografi”), is included in No. 71 of the official Inventario degli oggetti e dei cimeli pucciniani conservati nella Villa Puccini di Torre del Lago of May 1952. However, it is among the documents missing from the Inventario Archivio Puccini in Torre del Lago of 1980.According to Puccini, he began the composition around 10 July 1894, immediately after his return from Sicily (see above). However, composition sketches were already written in a draft of the libretto (67.A.1), which certainly originated earlier. In typical fashion, the libretto was nowhere near being definitively finished, and revisions dragged on right until the completion of the full score.
Puccini completed the composition of the first quadro by the end of August 1894 (see Carteggi 1958 No. 114). The further course of his work that year is unknown. The composition of the second quadro probably essentially belongs to that period, when he also decided to use the Piccolo valzer (see 66, written for another purpose) as the source for the music of Musetta’s famous waltz, “Quando me'n vo.”
On 21 January 1895, Puccini began the instrumentation of the first quadro, which took him almost half a year. It was completed on 8 June 1895. In the meantime, he had obviously finished the composition of the second quadro, as proven by the note at the end of the first quadro in the autograph score (see 67.B.1). That means the second quadro was orchestrated with remarkable speed and ready by 19 July. But apparently it was changed once again, for Puccini scribbled “23 July” as the date of completion on a wall of the studio in his vacation villa of Castellaccio.Note: *Reproduced in Marchetti 1973, p. 224. Shortly before that, he began composing the third quadro (see Carteggi 1958  No. 125), which was finished at the beginning of September (see La Scala, No. 77, 1956, p. 28). In the almost incredibly short period of eight days (10–18 September 1895), this quadro was fully orchestrated. The appearance of the notation of 67.B.1 attests to that. The first two quadri for the piano-vocal score obviously had already been engraved (see Carteggi 1958 No. 128 of 9 August 1895). According to MRL, that work indeed began on 28 June. The engraving of the full score began on 9 October 1895. Simultaneously, some portions of the opera that were already finished were being revised (see Epistolario 1982 No. 47, which probably dates from the first half of September), and numerous changes in the libretto were made. This particularly pertains to the fourth quadro, the composition of which Puccini began on 20 September.Note: *See 67.A.IV.1.a, Marchetti 1973, No. 210, and Carteggi 1958 Nos. 131 and 132. Numerous letters (for the most part undated) demonstrate the changes made in the very last phase of composition.Note: *See Carteggi 1958 Nos. 133-146, Pintorno 1974, No. 21, Marek 1951, pp. 154-55, Epistolario 1982 No. 48. Unfortunately, those letters cannot be arranged in a reliable sequence.  The composition was finished in the wee hours of a night in November 1895 (“in una tarda notte del novembre 1895”).Note: *Reported by Ferruccio Pagni as a reliable eyewitness, in Marotti and Pagni 1926, pp. 70-72. 67.B.1 has 10 December 1895 as the date of completion, therefore the last quadro was also orchestrated only within a few days.
The first piano-vocal score (67.E.1) was doubtless published in the second half of January 1896. This is supported by the registration dates of the I-Rsc and US-Wc copies and the deposit at the Prefettura di Milano on 25 January 1896. In addition, Puccini had offered the dedication of the opera to Marchese Carlo Ginori Lisci (see Critica 1976, p. 191) on 18 December 1895;Note: *On Ginori see notes to 68. the first edition already contained that dedication: “Al Marchese / Carlo Ginori-Lisci / Giacomo Puccini”. However, the changes Puccini made during the rehearsals for the première of the opera in Turin were excluded from this edition (see Pintorno 1974, No. 22 of 11 January 1896), signifying that the first edition had already been printed. Those revisions were included in 67.E.2, published shortly thereafter.
After the performance in Naples on 14 March 1896, Puccini decided to make some changes in the middle and in the finale of the second quadro (see Carteggi 1958 Nos. 159 and 161), as well as a very small variant in the second finale he wanted to try out in the performance at Palermo (24 April 1896), which remained unpublished for quite some time. Not only was there no new edition of the piano-vocal score published to reflect those modifications, but the first publication of the piano-vocal scores in the English and German languages (67.E.2a and 67.E.2b) adopted the version of 67.E.2, only minimally changed in comparison with the first edition. Alterations began appearing in the score only with the publication of the second English edition (67.E.3), dating from the autumn of 1897, which manifests several changes, but not to the extent that Puccini had already considered by the end of March 1896. They were fully realized only in the French edition (67.E.4), published in April 1898, which also contains additional changes in the third and fourth quadro. This definitive version was published together with one further addition in the third quadro as the “new edition with additions by the author” (“Nuova edizione con aggiunte dell’Autore”, 67.E.5) in October 1898. Apart from certain details, all of the following editions agree with this one and those variants are just publisher’s errors or arbitrary returns to earlier versions (compare 67.E.5l). 67.E.6/67.E.6a only approximate a new critical edition.