The first attempt to publish Jean-Philippe Rameau's complete works goes back to 1766, two years after the composer's death. The brainchild of his admirer Jacques-Joseph-Marie Decroix, this project, while never completed, allowed a partial reconstruction of Rameau's oeuvre through manuscript copies. The second attempt, under the direction of Camille Saint-Saëns and Charles Malherbe for the music publisher Durand, was first slowed by Malherbe's death before being interrupted by the First World War. Out of thirty projected volumes, only seventeen were published. The third attempt, known as Jean-Philippe Rameau, Opera omnia, was initiated by the present writer; its first volumes were published in 1996 thanks to the support of Musica Gallica (French Ministry of Culture and Foundation Francis et Mica Salabert) and several private patrons. Since 2003 the edition has been published by the Société Jean-Philippe Rameau and distributed worldwide by Bärenreiter. Sylvie Bouissou, as Editor-in-Chief, leads the following scholarly team: Nathalie Berton-Blivet (main assistant), Thomas Soury (publication assistant), an editorial committee (Cécile Davy-Rigaux, Denis Herlin, Davitt Moroney, Yvon Repérant and Graham Sadler), François Saint-Yves (in charge of the piano-vocal scores) and Pascal Denécheau (archival and documentary research). The edition is divided into six series:
Series I: Instrumental Music (three volumes)
Series II: Religious Vocal Music (one volume)
Series III: Secular Vocal Music (one volume)
Series IV: Dramatic Music (thirty-one volumes)
Series V: Fragments, Incomplete Works and Doubtful Works (one volume)
Series VI: Supplements (notably the Catalogue thématique des œuvres musicales (thematic catalogue of musical works) and the Dictionnaire des copistes (dictionary of music copyists)).
The publication of each vocal work is accompanied by performance materials, a piano-vocal score and a collection of symphonies (descriptive instrumental pieces and dances) derived from the score. Moreover, beginning in 2015 the Société Jean-Philippe Rameau and the Centre de Musique Baroque de Versailles (CMBV) have been jointly publishing an eight-volume anthology of Rameau's operatic vocal airs. At the time of writing, the Opera omnia numbers fourteen full scores, fourteen piano-vocal scores, four collections of symphonies and six volumes of the operatic-air anthology. Seven new titles, completed and already used for staged performances, will soon be published: Les Indes galantes, Naïs and Le Temple de la Gloire, and two versions each of Dardanus (1744) and Castor et Pollux (1754). Each score, in full-cloth binding, includes:
-
1 a historical introduction in French and English detailing the work's genesis, performance history and reception
-
2 a critical edition of the vocal text
-
3 a critical edition of the musical score
-
4 a critical apparatus including a description of sources, a report of significant textual variants and the editorial method
-
5 appendices containing sketches, drafts and fragments
-
6 facsimiles of sources.
The critical edition avoids expressing an opinion about the musical quality of a specific version of a work or justifying a textual change such as a cut or an addition. In order to determine which source(s) will be used for the edition, the editor must reconstruct the history of the libretto, score and performance materials by referring to autograph manuscripts, copyists’ manuscripts (including those intended for publication), and fragmentary prints and manuscripts. When they exist, performance materials allow one a priori to reconstruct a single version of the work authorized by Rameau, if not several of them. Contemporary copies supplement and clarify indications provided by performance materials, or furnish them when they are missing. As a result of this research, the editor is able to determine which available version or versions to edit. One should not be led by the quest for a certain type of ‘authenticity’ to offer a sort of commercial ‘best of’ by reconstructing a version of a work that Rameau never intended. Like many other opera composers of his time, Rameau altered his works during rehearsals, performances and later revivals. Performance materials may therefore reflect discarded or newly added sequences that were desired by him at a specific moment in the work's performance history. The editorial policy of the Opera omnia is to make the various authorized states of a work available to modern musicians in a chronological sequence. For instance, the critical edition of Naïs is based on the April 1749 version of the work. Nine supplements provide sequences that were discarded during rehearsals and added or modified for the 1764 revival.
In the case of works surviving in two distinct versions authorized by Rameau, such as Castor et Pollux, Dardanus and Les Surprises de l'amour, the Opera omnia offers these versions in two separate volumes (see the 1739 and 1755 versions of Dardanus in volumes IV.5 and IV.8). However, when multiple versions have much in common, the score is based on a single version and invites musicians to select modified or added sequences from among the supplementary materials gathered at the end of the volume. For instance, the March 1736 version of Les Indes galantes, edited in volume IV.2, 7, follows the prologue with the sequence Les Incas du Pérou, Le Turc généreux, Les Fleurs and Les Sauvages. The same volume includes nineteen supplements containing material sung by the character l'Amour at the work's 1735 première, the first version of Les Fleurs and the modified or newly added sequences for the 1751 and 1761 revivals. Similarly, the edition of Le Temple de la Gloire (volume IV.12) is based upon the 1746 Paris version, with supplements including the 1745 Versailles versions of Bélus and Bacchus, as well as excerpts from Trajan (shortened in 1746).
The editor's critical insights are crucial when sources are missing or contradictory, or when a choice must be made between two equally valid versions of a work. For example, modulations in the celebrated Trio of the Fates (trio des Parques), ‘Quelle soudaine horreur’, from Act 2 of Hippolyte et Aricie, were found so extraordinarily difficult during rehearsals for the opera's 1733 première that the performers refused to sing it. Writing with open bitterness in his Génération harmonique (1737, page 155), Rameau claimed to have included this trio – a musical ‘experiment’ dear to him – in the printed score of the opera ‘so the curious could judge’. Since we are curious to follow the composer's invitation, it seems obvious that the trio should be included in the edition of Hippolyte et Aricie. Similarly, because the first version of the earthquake in the Incas du Pérou from Les Indes galantes posed excessive technical challenges for the performers at the time of the première, Rameau replaced it ‘with common music’. By twisting the editorial rules a bit, we have retained this unpublished version of the earthquake in the edition, and placed the ‘common’ one among the supplements.
Though the editorial policy of the Opera omnia is assuredly scholarly, it nevertheless listens to the voices of knowledge and history that illuminate what is truly at stake in any critical edition: scholarship in the service of art and the musician.
