69. Tosca

Victorien Sardou’s drama La Tosca, first performed on 24 November 1887 in Paris, subsequently toured all over Europe, with Sarah Bernhardt in the title role. As early as February and March 1889, Puccini and his librettist at the time, Ferdinando Fontana, attended performances of the play in Milan and Turin, shortly before the world première of Edgar (62).Note: *See Quaderni 1992 II, No. 135, and Burton 1995, pp. 418-419. Fontana was probably the first to suggest Sardou’s play as a possible subject for an opera for Puccini (see Carteggi 1958 No. 31 of 7 May 1889). Shortly afterwards, Ricordi made inquiries in Paris at Puccini’s request concerning the legal situation for the operatic rights to the Sardou play.Note: *See Emanuele Muzio’s letter to Ricordi of 29 May 1889, published in Abbiati 1959, Vol. 4, p. 406. The information in Ricordi’s periodical, the Gazzetta musicale (9 June 1889, p. 363), that Puccini was going to compose two new operas, apart from Manon Lescaut (64), also referred to Tosca.
Two and a half years later, Ricordi explicitly confirmed that his publishing company had bought the rights to Tosca, so Puccini could transform it into an opera with a libretto by Luigi Illica.Note: *See the Gazzetta musicale of 13 December 1891, p. 799. However, Puccini postponed work on Tosca for many years. In the meantime, Ricordi assigned the composition to Alberto Franchetti.Note: *See Carteggi 1958 No. 98 of 5 January 1894 and Maehder s.d., Nos. 48-52, of 8 August through 19 November 1894. Finally Franchetti abandoned the projectNote: *See Maehder s.d., Nos. 55/56, of 26/27 March 1895. and it was definitively placed with Puccini.Note: *See Marek 1951, p. 151, of 26 July 1895, and Carteggi 1958 No. 127 of 9 August 1895. For a detailed account of the relevant events, see Burton 1995, pp. 418-431.  On 9 October 1895, Puccini attended Sardou’s La Tosca for the third time, at Sarah Bernhardt’s new guest appearances in Florence.Note: *See Carteggi 1958 No. 141 and Marek 1951, p. 155.
By the end of 1895, Giacosa was already engaged as co-author of the libretto.Note: *See Carteggi 1958 No. 149 of 14 December 1895. As usual, the intensive and lengthy work of all of the people involved in creating the text (Illica, Giacosa, Puccini, and Giulio Ricordi) can be rather easily reconstructed on the basis of an abundance of extant source material.Note: *Apart from the drafts described in 69.A.1 and 69.A.2, there are other documents in I-Mr (V.I.), and particularly in the Giacosa archives at Colleretto Giacosa (Province of Turin; see Critica 1976, pp. 123-126, Gillio 2000, and Biagi Ravenni 2001). A complete publication of these drafts, sure to provide instructive insight into Puccini’s working methods, is in preparation for publication by the Centro studi “Giacomo Puccini” in Lucca.According to a contemporary newspaper article in Il Giornale di Sicilia, Puccini is said to have played “alcuni brani del primo atto [some excerpts of the first act]” as early as 1 May 1896, at a banquet in Palermo.Note: *Quoted in Ceresa 1982, after No. 143 This means that Puccini worked sporadically on Tosca immediately after the first performance of La Bohème (67), which is also confirmed by a notice in the Gazzetta musicale of 21 May 1896 (p. 352). On 22 August 1896, he officially informed Illica: “Incominciata ‘Tosca’ [Tosca has begun]” (Carteggi 1958 No. 168), but not in a way that resembled continuous work on the piece with a definitive text. Puccini sent the “prime note Toscane” (see 69.A.I.1.b) to Giulio Ricordi on 15 October 1896.
There are a few definite dates for the phases of the composition but only scattered hints of the actual origins of the components.Note: *E.g. Pintorno 1974, No. 35 of 23 November 1897, as well as Carteggi 1958 Nos. 177 and 178 of December 1897 and 17 January 1898. Puccini apparently began the continuous notation of the score in January 1898,Note: *See Schickling 1989, p. 141. although it is presumed that by this time a large number of sketches for individual passages throughout the opera already existed.Note: *This is confirmed by an article by Eugenio Checchi in Nuova Antologia (December 1897, pp. 470-481) The first page of the autograph full score (69.B.1), dated “gennajo 98”, comprising the first three bars of the opera, suggests that Puccini had already completely orchestrated this thematically significant starting point with Scarpia’s motive right from the very beginning of his work on the continuous composition.
At the end of July 1898, Puccini began writing the full score at his vacation home, Monsagrati. The autograph (69.B.1), in conjunction with Epistolario 1982 No. 60 (31 July 1898), suggests that Puccini began the instrumentation in the middle of the duet of Tosca and Cavaradossi (between figs. 37 and 40). From 18 August, he composed the Act 1 finale and resumed work on the beginning of the act later that month. Although Ricordi’s Gazzetta musicale reported on 6 October (p. 569) that Puccini had already delivered Act 1, it can be concluded from 69.A.I.75.a that the instrumentation of that act was finished no earlier than 4 November 1898.
While 23 February 1899 is given in 69.B.1 as the start of the orchestration of Act 2, Puccini obviously worked on it immediately after finishing the first act.Note: *See Cecchini 1980-1993 No. 1324, the second part of which certainly dates from the autumn of 1898. Perhaps this letter concerns the middle part of the act, which has an autograph numbering of the leaves in capital letters in 69.B.1.  The instrumentation of Act 2 was finished on 16 July 1899, the same day Ricordi (according to MRL) began the production of the printed piano-vocal score. On 19 July Ricordi acknowledged the receipt of Act 2.Note: *See MRC 1899/1900 - 2/123. 69.B.1 reveals that Puccini did not orchestrate this act from start to finish either. He omitted the brief trio after Cavaradossi’s “Vittoria! Vittoria!!” at first, then finished it later. In this second act, Puccini also used some music composed by his younger brother Michele, specifically the Gavotta played backstage and the melody of Cavaradossi’s “L’alba vindice appar”.Note: *See Schickling 1998b. Michele Puccini had died a few years earlier in exile in South America.
The instrumentation of Act 3 apparently followed immediately (at Puccini’s holiday resort in Boscolungo Abetone).Note: *See the unpublished letter of 31 July 1899 to Cosimo Pavone, quoted in part in the catalogue of Sotheby’s
auction of 16/17 May 1991, No. 384. 69.B.1 confirms that the major part of the full score of this act (from fig. 9 to approximately fig. 29) had already been delivered to the publisher, while the beginning and the ending of the act were still missing. On 19 September, Ricordi acknowledged the receipt of the “[Tosca-Cavaradossi] duetto”,Note: *See MRC 1899/1900 - 5/384. which was so contentiously discussed in two famous letters by Giulio Ricordi and Puccini (i.e. Carteggi 1958 Nos. 208 and 209) because portions of it were taken from the deleted fourth act of Edgar. The end of the opera (from fig. 30+5) was completed and orchestrated in the final days of September.Note: *See 69.B.1; Carteggi No. 205; and an unpublished communication to Illica known in 1992 to be offered for sale by the music antiquarian Ulrich Drüner, Stuttgart — all written between 27 and 29 September 1899.  Puccini orchestrated the missing opening of the third act last of all because he was searching for some folk-like verses in Roman dialect for the Shepherd Boy. He finally obtained them from a local writer, Luigi Zanazzo, but (probably for reasons of copyright) never entered them in the full score. Ricordi acknowledged the receipt of this “preludio” on 19 October 1899;Note: *See MRC 1899/1900 - 7/217. Puccini’s date of 17 October in 69.B.1 may indicate the end, not the beginning of the work — perhaps even both. It is certainly possible that Puccini could compose approximately 150 bars of a rather simple nature in little more than one day. However, what Puccini was devising for the prelude to the third act was already known to the publisher, for on 14 October, Giulio Ricordi complained about Puccini’s use of too many, too heavy, and too expensive bells.Note: *See MRC 1899/1900 - 7/95.
In mid-December 1899, Puccini went to Rome to supervise rehearsals for the first production of the opera. In all probability, all of the changes distinguishing 69.E.2 from 69.E.1 were made by that time, so the performance version was that of 69.E.2, published shortly thereafter. The first printed full score (69.E.2a) corresponds to 69.E.2 and therefore probably was also published within the first months of 1900. Thus, Puccini’s definitive version of Tosca was settled for the most part and also firmly established in nearly all of the editions in other languages (with the exception of a few variants) prior to Puccini’s death.
In the succession of the full scores of Puccini’s operas in octavo format produced for public sale, Tosca (69.E.2g) was the fourth, published in 1924. For this, Puccini made some minor changes in the instrumentation, dynamics, and agogics, and sent them to the publisher on 1 February 1924.Note: *See the unpublished letter to Carlo Clausetti in I-Mr; see also 69.C.2. Apart from a few small differences, this full score was the basis for the later editions of the music in that format and (together with 69.E.2) for the later piano-vocal scores. In any case, 69.E.2g must be regarded as the last edition personally supervised by the composer, at the same time overriding the changes conveyed in 69.E.3 (if indeed they were ever authorized by Puccini).
It can be assumed that the performance at the Teatro Costanzi in Rome, a few days after Puccini’s death, was the first to be based on that definitive version.Note: *See Musica d'oggi, December 1924, p. 393. There had already been a performance at Milan’s Teatro Carcano in September 1924, but that was merely a revival from the previous year and, therefore, probably used Ricordi’s old material.Note: *See Musica d'oggi, September 1924, p. 285, and ibid., August/September 1923, p. 271.