64. Manon Lescaut

Very early in Puccini’s career (in March 1885, just at the beginning of the composition of Edgar), his attention was drawn to Manon Lescaut as a possible subject for an opera by his first librettist, Ferdinando Fontana (see Quaderni 1992 II, No. 6). But only after the first performance of Edgar did Puccini and his publisher begin preparing their next opera. Diverse subjects were taken into consideration, one of them already being Sardou’s drama La Tosca. On 28 April 1889, one week after the first performance of Edgar, Ricordi’s Gazzetta musicale informed their readers that Puccini had received a commission for a new opera. In the issue of 9 June there was even the talk of two operas — certainly meaning Manon Lescaut and Tosca.
On 15 July 1889, Ricordi entered into a contract with the authors Marco Praga (1862–1929) and Domenico Oliva (1860–1917) for a libretto for Puccini, adapted from the novel Histoire du Chevalier Des Grieux et de Manon Lescaut (1731) by the Abbé Antoine-François Prévost (see Morini 1983, p. 90). Neither Praga nor Oliva had ever written an opera libretto, nor did they ever attempt one again. Perhaps this explains the problems that arose during the opera’s gestation. Another problem was that a recent opera on that subject already existed: Jules Massenet’s Manon, crowned by success all over the world since its first performance in 1884. (It only appeared in Italy much later, after the first performance of Puccini’s opera.) The Italian rights to Massenet’s Manon had been purchased in May 1888 by Ricordi’s rival, Sonzogno. The fact that one year later Puccini, Praga, and Oliva were assigned the same subject by Ricordi, was perhaps also a manifestation of Ricordi’s desire to compete with an Italian treatment of Prévost’s popular subject.
According to Praga, the scenario of the libretto Puccini desired consisted of: “[Act 1:] the meeting of Des Grieux and Manon. [Act 2:] Then the miserable dwelling of the two lovers with Lescaut’s interest in protecting them, his dastardly game, his boastful immorality, his skeptical advice. [Act 3:] Then Manon in the luxury procured for her by Geronte, the intervention of the cavalier lover, the attempted theft and escape, the discovery, the arrest. [Act 4:] Finally the desert and the death” (quoted by Adami in Epistolario 1982, p. 72).Note: *“[Act 1:] incontro di Des Grieux e Manon. [Act 2:] Poi la misera casa dei due amanti con l’interessata protezione di Lescaut, il suo perfido gioco, la sua fanfaronesca immoralità, i suoi scettici consigli. [Act 3:] Poi, Manon nel lusso che le procura Geronte, l’intervento del cavaliere innamorato, il tentativo di furto e di fuga, la sorpresa, l’arresto. [Act 4:] Infine il deserto e la morte.”   The rather early libretto drafts in I-MrArchivio Ricordi, Milan  (old shelf mark PP 650-1-1-d), bearing several of Puccini’s autograph annotations, make Adami’s account somewhat dubious, for they already contained most of the Le Hâvre scene (set in a different location, namely a “piazza presso l’ospedale [a public square near the hospital]”).Note: *A libretto draft of Act 1 is also mentioned in No. 71 of the official Inventario degli oggetti e dei cimeli pucciniani conservati nella Villa Puccini di Torre del Lago of May 1952. It also appears (on p. 51) in the Inventario Archivio Puccini in Torre del Lago of 1980. However, it is no longer listed in the Verbale di Inventario of 1992.
Puccini began the composition in the summer of 1889, for by September he was already sketching the first meeting of Manon and Des Grieux (confirmed by the date in the continuous composition draft, see 64.A.1/1). In October 1889, three acts of the libretto were delivered to the publisher by the librettists (see MRC 1889/90 - 7/226 and 8/79). According to a note in the full score, Puccini began the orchestration in March 1890, which means that at least the composition draft of Act 1 had already been completed by that time.
However, Puccini did not at all regard the existing libretto as ready, for also that March Oliva provided a new version of Act 2 (see Carteggi No. 39), and Praga spoke of a new “ultimo atto [last act]” (certainly what is now Act 4), in a poem of 3 June 1890 sent to the composer (see Quaderni 1985, p. 212). Puccini was not in agreement and complained about the many revisions (see Epistolario 1982Giacomo Puccini: Epistolario, edited by Giuseppe Adami. 2nd ed. Milan: Mondadori, 1982  No. 28, probably written on 15 June 1890). That is why a new collaborator was recruited at that time: Ruggero Leoncavallo, then still completely unknown, who actually wanted to have his opera project in several parts entitled Crepusculum published by Ricordi. According to Leoncavallo’s own accounts, not always very reliable, he was enlisted by Puccini and Ricordi to draft a new scenario of Manon Lescaut. According to the documentary evidence, this seems to have happened in the summer of 1890.Note: *See Puccini’s letters quoted in Marek, pp. 115-116, and Epistolario 1982 No. 29, all certainly originating from the summer and autumn of 1890. 
By September, Puccini had orchestrated almost all of the first half of Act 1, as shown by the date in 64.B.1; that date also indicates, however, that he obviously had not worked on it for some time, as he explicitly recorded the continuation of his work in the autograph score. The libretto then still contained a second part of the third act, i.e. the scene of the roll call and embarkation in Le Hâvre (now Act 3); it was described by Oliva as being finished on 19 October 1890 (see Quaderni 1985, p. 213). According to his own information (quoted in Epistolario 1982, p. 73), at that time Praga no longer participated in the work, but a notice in the Gazzetta musicale of 9 November 1890 (p. 712) still referred to Oliva and Praga as the librettists, and forecast the completion of the opera for the autumn of 1891. Oliva continued in the enterprise with considerable doubts about Puccini’s intentions, while Giulio Ricordi and Leoncavallo worked on corrections of the plot (see Morini 1983, pp. 94-95 = MRC 1890/91 - 7/92), and Puccini complained about Oliva’s hesitant way of working (see Pintorno 1974, No. 2). Nevertheless, Act 1 was now definitively finished, so Puccini was able to complete the orchestration of the act in January 1891.
On 21 May 1891, Ricordi recorded the state of work: libretto ready; Act 1 including its instrumentation finished as well; the first half of Act 3 and Act 4 composed; and the second half of Act 3 already drafted, with only Act 2 missing (see MRC 1890/91 - 19/350). So all could be finished soon. The lack of continuity in this muddle confirms Puccini’s difficulties with the composition. On 30 July 1891, there was again some talk of a new libretto of Act 2 to be sent by Oliva (see MRC 1891/92 - 2/463), indicating a renewed revision of the part of the opera not yet composed. 64.A.II.9.a proves that Puccini had already done extensive work on the old version of the second act, including the instrumentation.
Probably by the end of 1891, an additional author for the libretto (the fourth one) was recruited: Luigi Illica (1857–1919). In contrast to Praga, Oliva, and Leoncavallo, Illica was a rather experienced librettist who had worked for Smareglia, Catalani, and Franchetti. With his collaboration, the original second act (rewritten several times) was totally jettisoned, resulting in a dramaturgically abrupt break between Act 1 (finished long ago) and the new Act 2, already set in Manon’s world of luxury (formerly the first part of Act 3).
At the beginning of 1892, this new dramaturgical structure was firmly established.Note: *See Carteggi 1958 Nos. 59 and 60, as well as Puccini’s unpublished letters I-PCc, Fondo Illica Nos. 9, 11, 25, and 126, certainly dating from this time.  Illica worked very thoroughly, while Puccini harmonized the considerable changes in his composition necessitated by the revisions in the text (see Carteggi 1958 Nos. 68–72, and Pintorno 1974, No. 3). At first, Illica wanted to abandon Act 4 and make the opera end filled with hope, with the lovers united on the deportation ship. However, this idea was soon dismissed.
Puccini finished the instrumentation of the “I^a Parte Atto 2°” through the end of the intermezzo on 5 June 1892 (see 64.B.1). But he also later revised it, for which Leoncavallo was called in again at the end of the work on the libretto to provide some additional lines (see Pintorno 1974, No. 4, of 2 August 1892).
The official information disseminated by Ricordi that Puccini had finished the opera was excessively optimistic (see the Gazzetta musicale of 3 July 1892, p. 429). In particular, the Le Hâvre section created many difficulties. For Giulio Ricordi, the ending of the scene seemed too long and implausible.Note: *See MRCCopialettere Ricordi, Milan  1892/93 - 2/273, 3/9, and 3/87 (published in part in Marek 1951, pp. 118-120) from the end of July through the beginning of August 1892.   So much so, that the publisher himself contributed four lines for it — though it cannot be ascertained if they were used. (According to Adami 1942, pp. 91-92, Praga maintained they consisted of what is now the last three lines of the act.) This part was orchestrated at the beginning of October (see 64.B.1).
In the meantime, there had been a legal agreement with Oliva concerning the authorship of the libretto, apparently mediated by Giuseppe Giacosa (see Carteggi 1958 Nos. 73 and 75). It is not correct, as is often asserted, that Giacosa himself was a collaborator on the libretto of Manon Lescaut. During the summer of 1892, Oliva, together with Illica, prepared the publication of the libretto, which differs in part from the text Puccini ultimately set to music.Note: *See MRCCopialettere Ricordi, Milan  1892/93 - 3/9, and Marek 1951, p. 121; see also Morini 1993 with new documents.  Until recent times, the names of the librettists were not even mentioned in any printed edition, probably due to some complicated copyright matters between Puccini’s collaborators and the publisher. 
At the same time, the piano-vocal score was already engraved and printed in single acts. Act 1 was ready on 17 September (with considerable changes compared to the autograph 64.B.1, which Puccini made in the meantime); Act 2 was in the course of being engraved (see the letter of 17 September 1892 in Marek 1951, p. 121). The entire full score may have been finished in October.Note: *Carteggi 1958 No. 76, suggesting that the completion occurred in mid-November, certainly is dated incorrectly. It most probably dates from 11 September; see also the letter of 19 September quoted in Marek 1951, p. 122, which certainly refers to it.   The earliest known printed copies of the piano-vocal score have blind stamps of December 1892.
The origin of the “Preludio Atto II” (64.E.0), edited in 1983 by Pietro Spada, is somewhat uncertain. Prior to that publication, and even in the most recent studies, this prelude was not mentioned in any of the Puccini biographies or catalogues of his works; nor does it figure in any of the printed full or piano-vocal scores of the opera, and certainly not in the autograph full score (64.B.1). In any case, according to Puccini’s autograph cancelled heading of the manuscript (64.A.II.1.a), it was written earlier than the dramaturgical reorganization of the act, that is, either at the end of 1891 or the beginning of 1892, perhaps shortly after Ricordi’s exhortation of 31 May 1891 (see above). It was obviously eliminated when Puccini began to orchestrate the completely revised act which is now Act 2. Then Puccini apparently gave the manuscript to an unknown recipient. 
The history of the variants after the printing of the first piano-vocal score is extremely complicated. Despite the commendable efforts by HopkinsonNote: *The information in Hopkinson 1968 (pp. 8-13 and 71) is much too cursory and somewhat confused; he simply did not understand some of the complicated connections between the various editions (absolutely excusable in view of the state of Puccini research in those days).   and Scherr, up to now not all of it has been described precisely. There is also a possibility that some of the information for the individual editions listed in the present catalogue might be incomplete, just as it is also possible that other unreported revisions exist and may someday be recovered.
The rehearsals for the first performance, with Puccini present from the very beginning, were based on the first piano-vocal score (64.E.1). Already during the rehearsals, however, changes were made. They were documented in the second piano-vocal score (64.E.2), published a few weeks after the first performance and probably representing the version actually performed. Puccini continued revising and correcting the score soon thereafter and in particular provided a completely new finale for Act 1. Ricordi was generous enough to produce a third printed version of the piano-vocal score (64.E.3) the same year. That version was performed for the first time on 21 December 1893 in Novara.Note: *See the Gazzetta musicale of 31 December 1893, p. 866, which reported that the new finale of Act 1 had been played there. Perhaps this minor opera house was chosen because Carlo Carignani, the arranger of the piano-vocal score and one of Puccini’s closest musical collaborators, conducted there.   Puccini was only able to evaluate this new version on the occasion of the production of the opera in Naples, on 21 January 1894, where he supervised the rehearsals.
Immediately afterwards, Puccini and Illica obviously considered the insertion of a completely new second act designed to eliminate the gap in the dramaturgy (originally deliberately desired by Puccini) and to better establish Manon and Des Grieux as ardent lovers,Note: *Carteggi 1958 No. 52, describing this intention, obviously is dated incorrectly as “Primavera 1891“. Its contents and the entire context suggest that the letter was written on 1 February 1894.   but Puccini and Illica never got around to doing so.
The printed editions of the full and piano-vocal scores issued since 1894 manifest a large number of variants of considerable proportions, certainly not limited to minutiae. In an effort to eliminate the chaos, near the end of 1920 Puccini began a revision, which he apparently hoped would lead to something akin to a definitive edition (see Carteggi 1958 No. 781). In particular, he emphasized the necessity to preserve Manon’s aria in the last act, which had been cut completely in the last printed piano-vocal score of 1909 (64.E.5) and in the first publicly sold full score of 1915 (64.E.6). That revision was newly reworked in the summer of 1922, with the concrete goal of a definitive performance at La Scala.Note: *See Puccini 1981, No. 105, and Carteggi 1958 No. 813 — the latter, according to the original in I-Mr to be dated 14 August 1922 (and not 1921, as erroneously reported by the editor). For that purpose, Puccini even requested some new lyrics. On 1 November 1922, he asked Giuseppe Adami, who was then collaborating on the libretto of Turandot, for “altre sentite parole [other heartfelt words]” at the beginning of Manon’s aria in the last act.Note: *See Epistolario 1982 No. 34, certainly dated this way, and not 1923.   By acquiescing to this request, Adami made the total of librettists for Manon Lescaut an even half dozen. He, in turn, supplied Manon’s lines “in landa desolata! ... Intorno a me s’oscura il ciel ... Ahimè, son sola!” in the current version. Since 64.E.7 does not yet contain those lines, while the later edition 64.E.8 does, Adami apparently provided them only during the course of 1923.
The La Scala performance, conducted by Toscanini, who had been involved in the revision of the score as early as 1910, took place on 26 December 1922.Note: *See also Barblan 1972, p. 229.  It was followed shortly thereafter by the publication of a new edition of the piano-vocal score (64.E.7), and soon after that, however, by Puccini’s latest revision of Act 4 (see 64.B.3), which Puccini doggedly asked Ricordi to include in the piano-vocal and full scores.Note: *See Carteggi 1958 Nos. 868 and 869, as well as Puccini’s unpublished letters to Clausetti of 23 June and 14 July 1923 in I-Mr.  The editions 64.E.8, 64.E.8a, and 64.E.8b, finally based on this revision, were published only shortly before or even after Puccini’s death. They contain some obvious shortcomings and careless mistakes, perhaps even overlooked by Puccini (for at that time he was intensively occupied by Turandot and probably not particularly interested in his work of old), if indeed he ever saw them. There were probably some performances of this so-called definitive version in Puccini’s lifetime, i.e. the three performances he attended in the autumn of 1923.
The piano-vocal scores of Manon Lescaut, including the most currently available edition, were mainly printed from the plates of 1892/1893. Although it is actually not possible to clearly distinguish different versions of the opera, the history of its publication leaves one with an impression of an extremely muddled situation. Real authenticity can only be ascribed to the version of the first performance, to the reworking of the first finale (64.E.3), and to the late revision of 1920/1923 (the final state of which is represented by the editions 64.E.8, 64.E.8a, and 64.E.8b). All of the interim versions serve to provide vignettes of the various stages of development insofar as Puccini’s definitive intentions are concerned. Since the latest editions also differ from each other in numerous details (far more than in any of Puccini’s operas), a critical edition of Manon Lescaut taking into account all of the variants is imperative.