Volume 13, no. 1:
Jean-Baptiste Lully: Armide: tragédie en musique. Edited by Lois Rosow. Edition of the livret: Jean-Noël Laurenti. Œuvres complètes, Série III, Opéras, vol. 14. Hildesheim: Georg Olms Verlag, 2003. [xxxi, 333 pp. ISBN 3-487-12524-2. €315.]
Reviewed by Graham Sadler*
2. The Edition
1.1 A hundred years from now, seventeenth-century music specialists (if they still exist) will surely wonder how their twentieth-century counterparts could have failed to produce a complete Lully edition that was anything like complete. Love him or loathe him, Jean-Baptiste Lully has long been regarded as one of the towering figures of the Baroque period, the creator of a style whose influence dominated French music until the eve of the Revolution and spread well beyond the borders of his adopted county. Yet the proportion of Lullys output published in reliable editions is still remarkably small. In this regard, indeed, many of his acknowledged masterpieces were better served in his own day than they have been in ours.
1.2 The history of Lully publication in modern times, at least in terms of collected works, got off to a fairly dismal start in the 1870s with the appearance of eleven of his operas in the series Chefs-dŒuvre classiques de lopéra français.1 Given their inelegant presentation and lack of inner orchestral parts, these reduced scores can have done little to enhance Lullys reputation, still less encourage performance or scholarly curiosity. More promising was the critical edition begun by Henry Prunières, the Œuvres complètes produced in the 1930s.2 But that monumental effort was left sadly unfinished at Prunièress death in 1942, comprising only three of the operas, three volumes of comédies-ballets and two each of ballets and motets.3 In the late 1970s, there were plans for a collected edition to be published by Broude Brothers. Three decades later, this brave venture has still produced only a single, if admirable, volume of motets, published by the Broude Trust in 1996.4
1.3 With a new century has come a new Œuvres complètes, this one published by Georg Olms Verlag with two of the key figures in current Lully research as general editors: Herbert Schneider, whose thematic catalogue has done so much to facilitate and stimulate that research,5 and Jérôme de La Gorce, author of numerous studies on Lully and his period, among them the most comprehensive and up-to-date biography.6 With several volumes already issued and a first-rate team of editors, the Olms Œuvres complètes gives every sign of measuring up to the challenge.
2.1 It is particularly good that the first opera to appear in the new venture is Armide (1686), Lullys last completed tragédie en musique. This work, widely regarded as the composers operatic masterpiece, has never previously been published in an adequate modern edition. Apart from the unsatisfactory reduced score in the Chefs-dŒuvre series noted above, we have had to make do with Eitners full score of 1885, Frank Martins version of 1924, and an undated vocal score prepared by Busser.7 None of these is as useful as the facsimiles of the first edition (Paris: Christophe Ballard, 1686) available from Broude and on-line from the University of North Texas.8
2.2 As editor of the present volume no one could be better qualified than Lois Rosow. She first nailed her colors to the Armide mast with her doctoral dissertation,9 a penetrating source-critical study of the operas 80-year performance history which, among other things, prepared much of the ground for the present edition. She has since refined and extended her research in a series of publications, many with direct relevance to this project. It is gratifying that such painstaking and perceptive scholarship should reach its logical culmination in this exemplary critical edition.
2.3 Armides very success both complicates and simplifies the editorial process. Among the vast quantity of surviving materials (none, alas, autograph), three sources emerge as especially important: (a) the Ballard full score cited above, issued soon after the 1686 premiere at the Paris Opéra (the Académie royale de musique); (b) the first edition of the libretto, on sale at the time of the premiere; and (c) a huge collection of the Opéras performing parts, variously dating from different stages of the works history up to its final revival there in 1766 and incorporating to a greater or lesser extent revisions made during that period.10 As it happens, Lully had no hand in the great majority of these revisions: he died little more than a year after the Armide premiere. Interesting though such posthumous adaptations may be, they have limited relevance to the present edition. (One could, however, envisage a separate editorial project, suited to on-line publication, with hyperlinks among different chronological stages of the works performance history.) The 1686 Ballard edition—a collaboration between composer, publisher, and librettist (Philippe Quinault)—necessarily becomes the principal musical source. As the editor has shown, however, two of the dessus de violon parts in the collection of Opéra parts can be dated to the time of the premiere. These, together with the first edition of the libretto, have been judiciously used to supplement the principal source or clarify ambiguous readings.
2.4 One aspect of the Ballard edition already revealed in Rosows dissertation is the number of stop-press corrections preserved in surviving exemplars. Further evidence of last-minute changes takes the form of in-house manuscript corrections, these occurring in sufficient numbers for the editor to be able to differentiate between at least seven of Ballards scribes (their work carefully distinguished from annotations by subsequent owners of the exemplars in question). She has thus been able to posit a hypothetical ideal copy, one that takes into account all relevant corrections. This in turn serves her editorial goal—to reconstruct the state of Armide during the initial production at the moment when the composer stopped adjusting details and the work assumed a stable form (p. xxiv). The full extent to which this represents the authorial intention which Rosow aims to recover is impossible to judge. It is, however, as near to that intention as anyone could now get, given the current source situation. Everything we know about Lullys authoritarian personality and working methods suggests that, even if he was not directly responsible for this or that aspect of the Ballard score, he was happy to approve the result. His paraphe, scribbled on the majority of surviving exemplars, was in effect his imprimatur.
3.1 It goes without saying that this score of Armide is designed to be studied, analyzed and interpreted but, above all, performed. The generous extent to which the performers needs are taken into account is thus particularly pleasing. It is not merely that a vocal score and orchestral parts will eventually be available. The editor provides abundant help with the elusive matter of choral and orchestral scoring. In this vitally important area, the Ballard edition may often strike the non-specialist as frustratingly imprecise. Over the past half-century or so, however, scholars have learned much about how Lullys Opéra chorus, orchestra, continuo section and dance troupe functioned, how his clef codes reveal much about instrumentation, how woodwind doubling was usually organized, how the scoring of trio interludes in choruses was adapted, and much else. Indeed, Rosow herself has been in the vanguard of this research. Her introduction to the present volume provides an authoritative five-page distillation of the current state of our understanding of these matters (pp. xxvi–xxx), supported by copious reference to further material. In the score, meanwhile, the choral and instrumental labelling at the start of each movement makes a neat typographical distinction between scoring that is explicit in the Ballard score or can be confidently deduced from it, and scoring judiciously derived from the set of parts. Further clarification of specific passages in the score is usefully provided by on-page footnotes.
3.2 Another performance-related matter addressed by Rosow is that of continuity. She stresses the seamless nature of Lullian opera, where adjacent passages [flow] directly from one to another without break (p. xxx). Where appropriate, her score conserves Ballards simultaneous use of single and double barlines, a feature which performers should find helpful and which should discourage the excessive sectionalization that mars some performances of French Baroque opera and has given it a reputation for being short-winded.
3.3 As to the music notation, this has been largely modernized (p. xxv), but original meter signatures (essential to the performer) have been retained, as have irregular rhythms associated with the flexible dot. Details of the editorial method emerge during the course of the editors long and informative introduction. Still, a useful feature for future volumes would be an itemized list distinguishing all original notational elements retained in the edition from those that have been modernized. True, the General Preface (p. vi) includes brief remarks to this effect, but nowhere in the edition is there a statement on, for example, the treatment of accidentals. The user is left to deduce that these have been modernized on the staves but not in the continuo figuring.
4.1 This latter policy (retention in the figuring of flats and sharps that would be modernized as naturals) seems questionable. As a continuo player, moreover, I find it unhelpful. Admittedly, the player can adjust to this system with little difficulty, and I am aware of counter-arguments to do with transposition (though how relevant these are in the present context is even more questionable). But in giving out mixed messages—on-staff naturals versus flats or sharps in the figuring—the edition appears illogical. The system can also lead to misreadings, as on p. 40 (m. 124): had the single flat in the figuring been modernized as a natural, the impossibility of this reading would have been immediately obvious. From the context—a diminished triad above a bass F-sharp—the correct reading must be ♭, or better still, ♮.
4.2 This example raises the question of the extent to which the continuo figuring requires editorial attention. Rosow wisely avoids adding anything to the figuring except where this can be deduced or corrected from parallel passages, in which case the additions are placed in square brackets. There would, however, be an argument for completing the figuring in cases where the interval in question is raised or lowered by an on-staff accidental. A case in point occurs on p. 159 (m. 20), where the figuring is given as 6 but where an accidental in the haute-contre shows that it should be 6[♯].
4.3 One pleasing aspect of the figuring policy is that, in chords involving tritones, the edition retains the accidentals in 4♯ and 5♭ even where the tritone is already, as Rosow neatly puts it, a by-product of the key signature (p. 320). Like many French Baroque composers but certainly not all,11 Lully maintained a distinction between the perfect fourth and fifth (never figured with an accidental, whatever the signature) and the tritone (always indicated by 4♯ or 5♭, whatever the signature). This useful distinction could, indeed, have been exploited in the editorial process, as on p. 100 (m. 62), where the principal source has an erroneous 5♭. The Critical Notes reveal that this figuring has been removed, yet it could equally have been emended to a 5 without accidental. Like some of his contemporaries, Lully sometimes use this figuring in a cautionary sense, to warn the continuo team of the need for a perfect triad rather than the first inversion which the falling third in the bass might lead them (and, indeed, the unwary modern player) to expect.12
5.1 Among its many invaluable features, the prefatory material includes an edition by Jean-Noël Laurenti of Quinaults libretto. Users with limited experience of reading seventeenth-century French will appreciate the modernization of the spelling and most of the capitalization. The original punctuation and poetic structure are, however, retained.
5.2 In the musical score, by contrast, the textual underlay reproduces the original orthography except in cases of inconsistency. Such underlay could well prove problematic to many performers. It will be welcomed, of course, by those who wish to reconstruct period pronunciation (a minority, I suspect), for whom the original spelling provides clues. And native French singers who prefer modern pronunciation will have no more difficulty with Quinault-Lully orthography than anglophones would with, say, the Dryden-Purcell equivalent. But Lully performance is not limited to French singers or the period-pronunciation lobby. Individually, the modern pronunciation of spellings like connoistre (i.e., connaître), destournez (détourné), or preveu (prévu) is easy enough to establish; but given that each page may include several comparable examples, one wonders at the cumulative rehearsal time occupied by such matters, especially with non-specialist choral singers. By retaining the original orthography, the new Lully edition allows the needs of a few (for whom a facsimile is, after all, only a few mouse-clicks away) to take precedence over the convenience of the many.
6.1 It is no surprise that Rosow, with her unrivalled experience of the Armide sources, supports her editorial work with an exemplary critical apparatus. The source discussion is a pleasure to read, even if the point-size is a bit too small for comfort, and the dozen or so pages of critical notes and associated materials are remarkably easy to use. Details of emendations and other editorial decisions are admirably pithy yet never ambiguous, while the task of locating the relevant passage of music is facilitated by the scores use of running headlines which include act/scene numbers.
7.1 In short, Rosows sensitivity and her care over detail will ensure that this Armide volume remains at the elbow of all subsequent Lully editors, setting the standard for the remaining opera volumes of the new Œuvres complètes. More important still, her edition proves a trustworthy starting point for staged and concert performances of this outstanding work. To see the fruits of her research is gratifying enough: how much more so to hear them.
* Graham Sadler (firstname.lastname@example.org) is Professor of Musicology at the University of Hull. He is a member of the editorial committee of the Rameau Opera omnia (Bärenreiter), for which he is currently preparing a critical edition of Zaïs (1748).
1 This series, edited by Théodore de Lajarte (Paris: Michaelis, 1878–83; reprint, New York: Broude Brothers, 1971), includes Lullys Cadmus et Hermione, Alceste, Thésée, Atys, Isis, Psyché, Bellérophon, Proserpine, Persée, Phaëton, and Armide.
2 Henry Prunières, ed., Jean-Baptise Lully: Œuvres complètes (Paris: Éditions de la Revue musicale, 1930–39; reprint, New York: Broude Brothers, 1966–74).
3 A third volume of motets, edited by Prunières and revised by Michel Sanvoisin, was published by Broude Brothers in 1972.
4 Jean-Baptiste Lully: The Collected Works, series IV, vol. 5 (New York: The Broude Trust, 1996); it contains Jean-Baptiste Lully, Quare fremuerunt gentes (LWV 67), Notus in Judaea Deus (LWV 7/177), Exaudiat te Dominus (LWV 77/15), edited respectively by Lionel Sawkins and Carl B. Schmidt; John Hajdu Heyer; and Anne Baker.
5 Herbert Schneider, Chronologisch-thematisches Verzeichnis sämtlicher Werke von Jean-Baptiste Lully (LWV) (Tutzing: Schneider, 1981).
6 Jérôme de La Gorce, Jean-Baptiste Lully (Paris: Fayard, 2002).
7 Robert Eitner, ed., Jean Baptiste de Lullys Armide und Alessandro Scarlattis La Rosaura (Leipzig: Breitkopf & Härtel, 1885; reprint, New York: Broude Brothers, 1966); Frank Martin, ed., J.-B. de Lully: Armide (Geneva: Édition Henn, 1924); Henri Busser, ed., J.B. Lully: Armide (Paris: Lemoine, [ca.1957]).
8 Lully, The Tragédies lyriques in Facsimile, vol. 13 (New York: Broude International Editions, 2000); and http://www.library.unt.edu/music/assets/lully/Armide.pdf (accessed 1 November 2007).
9 Lois Rosow, Lullys Armide at the Paris Opéra: a Performance History: 1686–1766, (PhD diss., Brandeis University, 1981).
10 Unlike a number of other French operas of the Lully-Rameau period, no production score of Armide survives. The term production score has been coined for scores used by the batteur de mesure or others involved in the performance of a given opera, these indicating the various cuts, additions and other revisions made during successive rehearsals, performances, and revivals.
11 See, for instance, Monsieur de Saint-Lambert, Nouveau Traité de laccompagnement du clavecin, de lorgue, et des autres instruments, rev. ed. (Amsterdam: Estienne Roger, [ca.1710]), 19, on the ambiguity in French continuo figuring of chords involving a tritone. Although Saint-Lambert discusses only the augmented fourth, his remarks apply equally to the diminished fifth. For examples of the sort of figuring that Saint-Lambert would have considered defective, see Robert Zappulla, Figured Bass Accompaniment in France (Tournhout: Brepols, 2000), 121, 206.
12 A recent discussion of such cautionary figuring is included in Graham Sadler, Idiosyncrasies in Charpentiers Continuo Figuring: Their Significance for Editors and Performers, in Les Manuscrits autographes de Marc-Antoine Charpentier, ed. Catherine Cessac (Liège: Mardaga, 2006), 157–76.
Copyright © 2008 by the
Society for Seventeenth-Century Music. All rights reserved.