D. Kern Holoman
[text of an article to appear in The Berlioz Companion, ed. Peter
Bloom, forthcoming from Cambridge UP.]
Berlioz left posterity an admirable performance legacy. The scores and parts published under his supervision and, for the most part, to his satisfaction are sources that typically offer unambiguous direction as to his intent. They often reflect years of perfecting the manuscript materials in conjunction with live concerts under his own baton. His personal involvement with multiple performances of the symphonic works, unusual for its time (far greater, for instance, than Beethoven's), was met with meticulous compositional response; and with his orchestration and conducting treatises he left useful guides to the performing forces at his disposal and his notions as to their most effective deployment.(1) His sensitivity to the practical issues of live music-making, if not always to the cost of music and musicians, makes his work feel somehow welcoming to those who undertake it. With the exception of perhaps a half dozen passages of legendary difficulty, the music lies well beneath the fingers and is rewarding to discover and recreate--that is, to perform.
Berlioz the conductor left across Europe a generation of professional musicians schooled in how his music was supposed to go--though, by the same token, too few conductors committed to his cause. By the end of his life, most of the completed works had been well performed. A good proportion of these had been heard often and were familiar to serious listeners both in Paris and elsewhere; a few--the Fantastique, Pilgrims March from Harold in Italy, Roman Carnival overture, Hungarian March from Faust, and portions of L'Enfance du Christ--were even popular: hummed in the streets, known to hundreds. After his death thinking musicians continued to promote the Berlioz legacy, at least so far as they could acquire the performance materials, and those who knew his life's story did it both from enthusiasm for these "lovely pages" and out of a sense of atonement for the difficulties the master had encountered in being understood.
Jules Pasdeloup (181987) began to popularize some of the orchestral excerpts in his mass-market Popular Concerts from 1861 and can be credited with the universal popularity of the so-called "Three Pieces from The Damnation of Faust:" the Hungarian March, Ballet of Sylphs, and Minuet of the Will-o'-the-Wisps. Édouard Colonne (18381910), whose orchestra concerts began in 1873, premiered a complete, well-rehearsed Damnation de Faust in February 1877, repeated it for six consecutive weeks, and before his death had conducted more than 160 performances. The best orchestra in France was the Société des Concerts du Conservatoire, to which, despite its wary relationship with Berlioz as an active composer and potential first conductor, he donated all his performance material in 1863. By 1918 they had used his collection to master, in the systematic fashion that was their habit, most of the central repertoire.(2) Their version of Roméo et Juliette, undertaken in 1873 with the Scène d'amour, was completed in January 1879 under E.-M.-E. Deldevez (181797) and became a staple of their repertoire: they were the only orchestra in the world, it was said, capable of playing the Queen Mab scherzo accurately and at sufficient speed.
The better French conductors (Colonne, Danbé, Taffanel, Gaubert, and later Dervaux and Prêtre; the notable exception is Messager) were for the most part familiar with Les Troyens, Berlioz's single best work, and there were important complete productions in Germany from 1890.(3) In England, plans for a Covent Garden production were delayed by World War II, but in 1947 Thomas Beecham led a radio broadcast from the Maida Vale studios of the BBC, to which London enthusiasts flocked. In short succession came Sir Jack Westrup's 1950 reading of the work with the Oxford University Opera Club, the Westminster recording of Les Troyens à Carthage with the Société des Concerts under Hermann Scherchen (and a post-recording concert performance at the Palais de Chaillot on 10 May 1952), Rafael Kubelik's 1957 Covent Garden production of the complete opera, and a two-year project of the Chelsea Opera Group (La Prise de Troie, 1963; Les Troyens à Carthage, 1964). These undertakings did much to shape modern enthusiasm for Les Troyens in specific and for performing Berlioz in general: among those who participated in the Chelsea Opera Group at the time were David Cairns, Colin Davis, Roger Norrington, and John Eliot Gardiner, all of whom have played critical roles in later stages of the Berlioz "revival." By the Berlioz centennial in 1969 a coherent and visionary approach to the work was in place, resulting in Hugh Macdonald's published edition for the New Berlioz Edition, the lavish Covent Garden production of 1969, and the release of the first complete recording as, essentially, the flagship of the Colin Davis Berlioz Cycle for Philips Records. Musicians and music lovers alike thereupon discovered the majesty of Les Troyens: as a summary of the composer's art, as the last of the great lyric tragedies, as a worthy companion of Tristan and Otello. If live performances remain exceptional (not, in itself, a bad thing), the main reason is the ongoing shortage of tenors trained for and capable of mastering the part of Aeneas--a difficult and taxing role, notably in the fifth act.
Today, as a generation of scholars and performers looks back on careers of promoting these masterpieces, they have reason to take pride in having established the order and amount of Berlioz's artistic accomplishment. A vibrant critique of his life and works has replaced the old, unseemly jousting of partisans and detractors. The three operas (Benvenuto Cellini,Les Troyens, and Béatrice et Bénédict), left in 1869 without anything approaching standard texts, can now be brought to life more or less routinely. Lost works and artifacts of the composer's study, notably the Messe solennelle recovered in 1992, have resurfaced and taken their place in the Berlioz lore. Now it seems foolish to argue that Berlioz is any longer "misunderstood and misperformed."
But performing Berlioz still has its particular challenges. For one thing the Berlioz repertoire demands rigorous forethought as to venue and personnel--and, as I have written elsewhere, not a little carpentry. The Te Deum cannot be done effectively without a pipe organ to the rear, behind the audience, an arrangement common only in French cathedrals and basilicas; La Damnation de Faust demands a choral force large enough that the men can split into two distinct groups for one of its central moments; Roméo et Juliette requires its choruses of Capulets and Montagues (and an intermission for them to take their place onstage), a third chorus for the recitatives, and a contralto and tenor soloist who sing briefly at the beginning and then disappear. Berlioz's interest in musical instruments led him to employ novelties of manufacture and curiosities of antiquity that failed to achieve permanence in a typical orchestra's inventory. And still only a half-dozen titles are to be found in an the orchestral library: the three pieces from Faust, the Royal Hunt and Storm from Les Troyens, the Roman Carnival and Corsaire overtures, the Fantastique, and L'Enfance du Christ.
So even if performers and listeners can take satisfaction in the frequent live performances of the central repertoire and the wide dissemination of its recordings, it would be wrong to consider the repertoire fully discovered nor the aesthetic issues of performing Berlioz fully engaged. One would not talk of a rich performance tradition for either the Te Deum (pipe organ and basilica required) or Benvenuto Cellini (a single recording; adequate materials just becoming available). And there is much fine music to be found in repertoire hardly performed at all, perhaps especially in the songs Berlioz developed for his own public concerts; here I think particularly of Zaïde and Sara la baigneuse, where the only explanation as to why they are so frequently overlooked must be the difficulty of fitting single short works for voice and orchestra onto modern concert programs. Such ceremonial patriotic works as Le Cinq Mai, the Hymne à la France, L'Impériale, and even the Chant des chemins de fer are worth at least an occasional hearing.
One might summarize the challenges of performing Berlioz as those of finding the hardware and personnel on the one hand, and the software on the other.
Generally a modern symphony orchestra (90-some musicians: a dozen players in each strings section, quadrupled winds) with affiliated large choral society (150200 singers) is ample for the Berlioz repertoire. The Société des Concerts, Berlioz's paradigm, numbered about 80 players and 80 professional singers including opera virtuosi. Most of the necessary hardware--piccolo, English horn, E and bass clarinet, light percussion--is in keeping with 19th-century norms. Among the exceptional requirements are:
WOODWINDS. Four bassoons are customary for the French orchestral repertoire, owing to the smallish envelope of the instruments of the era; Berlioz occasionally writes four-voice chords for bassoon. The contrabassoon part in the Francs-Juges overture may be omitted. There are instances of two and three simultaneous piccolos. While Berlioz arranged a demonstration piece for Sax instruments including saxophone, the Chant sacré version III (Hol. 44C, 1844), the source is lost; saxophones are not required in Berlioz.
BRASS. Here lie some of the critical issues of Berlioz performing practice, affecting all the sections. Though the piston- and rotary-valved chromatic horns were claiming a place during the epoch of the Fantastique, Berlioz himself preferred to write for natural horns, where the key of the instrument was established once per movement with a crook of appropriate length. Owing to the chromaticism of Berlioz's harmonic rhetoric, he thus often needs horns pitched in multiple keys: in Roméo et Juliette, for instance, the Love Scene calls simultaneously for horns I in E, II in F, III in high A, and IV in D. While such parts are not intrinsically difficult for professional musicians either to read or to play, they do make the horn staves a particular mental gymnastic for the conductor. Additionally there is some loss of color when out-of-series pitches that would have been achieved by stopping the bell are played in conventional fashion; in any event, it is important to execute the notated sons bouchés with the hand and not a mute.
The Berlioz trumpet section usually consisted of two trumpets and two piston cornets, the latter--valved bugles--having a sweeter, somewhat more delicate sound than modern trumpets. It makes sense to try pairs of modern trumpets and cornets in these cases; additionally there is a lovely added solo part for piston cornet in the waltz from the Fantastique, originally composed for the great J.-J.-B. Arban (182589).
Many informed listeners think that the greatest loss from the tone-color spectrum of 19th-century orchestras is caused by the modern practice of using three large, triggered double trombones in place of the altotenorbass trio favored from Mozart to mid-century; the better compromise is to use two tenor trombones and a bass. From the Fantastique (1830) through Faust (1846), Berlioz generally calls for ophicléide (a keyed bugle, of airy tone quality and dubious pitch) or ophicléide and the old French Revolutionary serpent still in use during that era for ecclesiastical chant. (The Requiem calls for grande ophicléide monstre, amusing to imagine but impractical to duplicate.) Starting with the Marche funèbre pour la dernière scène d'Hamlet (1843) he calls for ophicléide or tuba, or ophicléide and tuba. (In the manuscript parts for L'Impériale, he replaces "ophicléide" with "bass saxhorn" and "tuba" with "tuba (contrabass saxhorn)" (see next paragraph). Tubas need to be made to suffice for all these parts, but the baritone model is sometimes, as in the case of the Dies irae in the Fantastique, a better solution than the powerful all-purpose doublebass tuba.
Berlioz asks for a double quartet of saxhorns (soprano, contralto, tenor, and contrbass) in the Marche troyenne and a quartet of tenor saxhorns in the Royal Hunt and Storm from Les Troyen. Visually and for tone color, the latter--the most frequently performed excerpt from Les Troyens--might well be played on baritone horns; one recommended solution for the former is to use E cornet or trumpet, flugelhorn, French horn, and baritone tuba. The parts for saxhorn suraigu in the Te Deum and Trojan March are best played on a piccolo trumpet. A number of period-instrument brass bands in the United States own sets of saxhorns (and a matched set is pictured on the 7.7¢ stamp released by the United States Postal Service on 20 November 1976).
PERCUSSION. Each timpanist--Berlioz often calls for multiple players--requires a minimum of three pairs of mallets: hard, medium, and soft, to answer the composer's call for baguettes de bois, de bois recouvertes en peau, and d'éponge. The implication of Berlioz's sometimes puzzling use of these terms is that the leather-covered medium stick in the norm.(4) Recent developments in coating plastic drumheads duplicate the sound of 19th-century skin heads quite satisfactorily.(5) The ordinary orchestral snare drum has little use in Berlioz: for movement IV of the Fantastique, the Marche funèbre pour la dernière scène d'Hamlet, and the Marche pour la présentation des drapeaux from the Te Deum, players should use snare-less field or tenor drums.
The bells in the last movement of the Fantastique present the most notable of all the challenges in performing Berlioz, since the symphony is the second most frequently performed (after the Roman Carnival overture) of all his works. The customary acceptable solution is to use large suspended metal plates, available from percussion suppliers by rent; the customary unacceptable solution is to use standard tubular orchestral chimes. A better course of action still is to borrow the largest moveable G and C from a local carillon. (A half-dozen major orchestras in the United States have their own bronze bells cast for the Fantastique, but those of the San Francisco Symphony, for example, sound an octave higher than Berlioz probably intended.)
The tuned antique cymbals in Roméo et Juliette and Les Troyens, seen by Berlioz at the museum in Pompeii, are generally replaced by crotales struck with a plastic mallet. (The other "antique" instruments in Les Troyens--double flute, sistrum, and tarbuka--are onstage visual props, matched respectively by oboes, triangles, and the one-headed Provençal tambourin.(6)) An anvil (petite enclume), struck with "a small sculptor's hammer," is required for one of the smiths' choruses ("Bienheureux les matelots") in Benvenuto Cellini. The Hamlet march reaches climax in a "volley" [of rifle fire]; theatre companies are usually equipped to provide an appropriate effect.(7) The jingling johnny, or pavillon chinois, needed for the Symphonie funèbre is usually to be found at a local Shrine band--in France, at the Foreign Legion band.
HARPS, KEYBOARDS. The 19th-century Érard pedal harp was considerably smaller and quieter than the standard modern Lyon- and-Healey, but the effect is roughly the same. Berlioz calls for pair of harps from the second movement of the Fantastique forward; both the solo exposition of Harold en Italie and the mezzo-soprano strophes in Roméo et Juliette feature important harp solo work. In the Fête chez Capulet from Roméo et Juliette and the scene of the Trojan women in Les Troyens ("Complices de sa gloire," finale of act II), some three to six pairs of harps are envisaged, though a single pair will suffice. (Wagner probably got the idea for the similar effect at the close of Das Rheingold from having heard and seen Roméo et Juliette at one of the 1839 Paris premieres.)
In L'Enfance du Christ Berlioz asks for the harmonium organ he knew as the orgue mélodium d'Alexandre, the inventor having been an acquaintance. Purists will need to look around for a foot-pumped model; a more practical solution is the electronic synthesizer or Baroque positive.
The guitar is needed for Méphistophélès's scene in the Huit Scènes de Faust and Somarone's in Béatrice et Bénédict (and was first envisaged for the strophes in Roméo et Juliette); two are needed for Benvenuto Cellini.
STRINGS. The period from the Fantastique to Les Troyens saw the complete redesign of the violin family for increased power, developments in which the Parisian violin maker J.-B. Vuillaume (17981875; and the bow-maker F.-X. Tourte, 17471835) took the lead. New instruments were built to the bolder specifications and older instruments refitted. Amount and frequency of vibrato certainly increased as the decades elapsed. Berlioz would have been happy enough with this apparent progress, and it makes little sense to ask most players to adopt a substantially different approach to bowing, articulation, and vibrato than they use for the Beethoven-to-Mahler repertoire.
SINGERS. As indicated previously, a large chorus of some 150 to 200--that is, twice the size of the standing chorus of the Société des Concerts--will usually suffice. The problem is the balance of voice parts, since Berlioz typically envisages a little over one-third women, with equal numbers of tenors and basses. For the choruses in Roméo et Juliette he suggests 70 Capulets and 70 Montagues (30-20-20); the same sort of force would work for Sara la baigneuse, with half of all men and women assigned to chorus I, the remaining women to chorus II, and the remaining men to chorus III. Faust calls for similar or larger numbers, but with two significant movements for men alone and added children's chorus for the final Apothéose; the Te Deum specifies a double chorus (40-30-30 x 2) and a massed children's chorus (600). Works thereafter are for S-A-T-B, the cast list for Les Troyens calls for "a good hundred singers" ("une centaine de choristes surnuméraires").
The published Berlioz performance material follows three avenues: (1) scores and parts descended from the original publications contracted with and overseen by the composer; (2) those descended from the Berlioz Werke published by Charles Malherbe and Felix Weingartner, now generally called the Old Berlioz Edition (Leipzig: Breitkopf & Härtel, 190007); (3) those based on the New Berlioz Edition, with Hugh Macdonald as General Editor and published by Bärenreiter Verlag of Kassel, Germany (1967 ).
While at least a few copies of the first publications are preserved in research libraries, very little of the performance material available today for hire or purchase descends photographically from the Berlioz originals. (Exceptions are some of the vocal scores and a few sets offered for sale by Kalmus, including those for The Marseillaise. But the vocal scores tend toward disorder and aberrant readings, and the parts, lacking rehearsal letters and measure numbers, are viable only for the shortest works.) By contrast descendants of the Breitkopf & Härtel edition are widely available at attractive price: Kalmus and Luck's Library parts, and Dover, Kalmus, and Broude scores. These have become essentially the standard texts, and with a little work by conductor and librarian--addition of rehearsal indications and the blotting out of extraneous text--remain highly serviceable.
Scholarly considerations aside, the practical triumph of the New Berlioz Edition is in how it provides musicians with scores and parts for works that theretofore were not to be had at all. The project began with two such works, the Symphonie funèbre et triomphale (1967) and Les Troyens (196970), and has recently unveiled two others: the newly discovered Messe solennelle (1994) and, more than 150 years after the first performances, a workable edition of score and parts for Benvenuto Cellini (199496). (All four editions were the work of the general editor, Hugh Macdonald.) Bärenreiter's preferred business practice is to rent the performance materials, withholding them from direct sale. End-users dislike this form of capitalism, since it is thus impossible for orchestras or conductors to own sets with their own markings; and the corollary attempt to control performance rights, practiced aggressively for the first performances of the Messe solennelle, seems particularly objectionable. One obvious result is fewer, rather than more, performances using these important new materials; another is to favor entertainment cartels and the major opera houses and orchestras over regional, local, and educational institutions. Nevertheless we must be patient as private enterprise recuperates its investment: the cost of preparing these materials is something on the order of $100 a page, hence (it is said) in excess of $50,000 for the parts to Cellini alone. As of this writing, about half the published volumes of the NBE have accompanying parts (see Appendix).
Given that on the whole there is comparatively little in the New Berlioz Edition that changes the overall sound of the familiar works, it is important to emphasize that the NBE achieves a considerably more accurate representation of the composer's notions than did Malherbe and Weingartner, and represents a significant corrective to misleading characteristics of the materials in common use. For one thing, the old edition attempts, unsuccessfully, to present trilingual readings (German at the head, then French and English) of titles and lyrics. In score and parts for the Fantastique, for instance, we find at the head of the first movement:
Rêveries Passions Visions and Passions
--not all that egregious an English translation, to be sure, but misleading
as to the reveries and missing the pointed suggestion that the reveries
and the passions are two different parts of the movement (the slow introduction
and the allegro, respectively). The English rhymed texts are barely viable
for singing and useless as translations. For Zaïde:
|«Ma ville, ma belle ville,
C'est Grenade au frais jardin,
C'est le palais d'Alladin,
Qui vaut Courdoue et Séville.»
[Qui vaut Courdoue,
Qui vaut Courdoue et Séville.]
|Granada my native city
'Tis the home of all that's fair
Bright as a gem past compare.
Though some may other towns more splendid
There's naught fairer than Granada!
(Roger de Beauvoir)
(Trans. Percy Pinkerton)
Then, too, Malherbe and Weingartner adopted principles of layout that actually run counter to Berlioz's musical thought. The chief of these was to place the horn staves above the bassoon, woodwind-quintet fashion, suggesting a philosophy of orchestral choirs that has more to do with post-Wagnerian ideals than anything Berlioz ever thought or espoused. Further, in my view, Berlioz's manner of notating the trombone parts (often with the bass trombone on one line and the tenor and alto on a second line above) suggests his understanding of voice- and chord-function, such that the Breitkopf re-notating of the two lower parts in the bass clef and the upper part in a C-clef can confound the reader as to what is really meant to happen. Jacques Barzun, at the close of his two-volume biography of Berlioz, presents a useful list of dozens of "Errors in the 'Complete' Edition of the Scores"; quite a number of these imply an audible difference between Berlioz's conception and that of Malherbe and Weingartner.(8)
The New Berlioz Edition presents no-nonsense, French-only scores, with each staff identified on every page, measure numbers in the upper left corner of each system, and on the average at least one rehearsal letter for every two-page opening. The guiding principles of editorial policy were formulated with performers in mind: notes and critical apparatus are short, important, and of practical use. (It thus makes sense for conductors to use the full, clothbound NBE volumes as opposed to the paperback "Urtext" scores furnished with the rentals but lacking the critical matter.) The overall look of the NBE has mutated according to rapid changes in the technology of music typesetting and in the global labor force. The crisp, clean character of the early volumes, prepared in-house in Kassel with presstype (rub-off transfers), was replaced by a bolder and less attractive typeface with scores produced in Asia. Bigger, blacker notes led to wider layout and more frequent page turns (and more pages, thus higher prices for longer volumes); compare, for instance, pp. 16263 of the Fantastique (vol. 16, 1972: 17 measures) with pp. 26667 of Roméo et Juliette (vol. 18, 1990, 11 measures). Benvenuto Cellini (vol. 1, 199496) returns to a somewhat tidier look, thanks in large measure to the use of computerized typesettting and page design.
Otherwise there is little to quibble over in the accuracy and usefulness of the New Berlioz Edition scores but for the fact that the silver ink on the spines wears quickly away. One grows accustomed to the minor idiosyncracies--beams, not flags, for the vocal syllables; slurs, not beams-and-slurs, for melismas; dots dropped for dotted-quarter flagged triplets and dotted-half flagged sextuplets (see the illustrations below)--and accepts the inevitable discrepancies in width of wedge accents and crescendo/decrescendo markings. The clarity of the NBE often surpasses that of the precedent publications, and the extra space can be seen to have advantages that perhaps outweigh the frenetic page-turns demanded by the NBE Queen Mab scherzo.
Compare, for instance, the first page of the published viola part for Roméo et Juliette in the editions of Brandus et Cie (1847), Breitkopf & Härtel (1901), and Bärenreiter (1990).(9) The musical text is substantially the same in all three, the chief difference being the divisi called for in bar 41 of the Brandus part, reflected in neither of the subsequent editions. (The bar in question is no different, technically, than what precedes and follows. Whether or not to divide for these effects is best left to the players; professional musicians will generally choose the double stops.) The Brandus part is from a first drawing of the set used by Berlioz for performances in 1846 and corrected in his hand and that of his copyist. (The pencilled indication "No 2" in the upper right corner is autograph, as are the rubrics on the tissue wrapper, not pictured.) The part carries the rubber stamp of the Société des Concerts at the top, where it arrived in 1863 and probably served for the performances of 1877 and thereafter; at the bottom is the stamp of the Bibilothèque Nationale (where the Société des Concerts archive arrived beginning in 1974) and at the lower left the shelfmark there, Rés. Vma 215, in the hand of the BN conservator Jean-Michel Nectoux. 102 measures are given on 11 staves of the first page, compared with 68 for the OBE and 61 for the NBE parts with 10 staves each.
The NBE part gives a more accurate representation of Berlioz's intent for the wedge decrescendo as carrying through the sixteenth note of the main figure in the fugue subject, where OBE's reading looks more like an accent; by contrast bars 38 and 39 are more successfully represented by OBE. Note, too, the dot-saving scheme NBE uses for the sextuplet figures beginning in bar 24. Both OBE and NBE give a "courtesy" A in m. 44 lacking from the Brandus part. Brandus has no rehearsal or measure numbers; OBE gives rehearsal numbers 1 to 3; NBE gives both measure numbers at the start of each line and rehearsal letters A to F. The cued snippet, mm. 8791, of the trombone recitative (the "intervention du Prince") as given by Brandus at the foot of the page also appears on p. 2 of the OBE part (not pictured); but Brandus lacks the new tempo-character indication, and OBE gives it in Italian: "Fieramente, un poco ritenuto, col carattere di Recitativo misurato." NBE (p. 2, not pictured) gives the correct French original, "Fièrement, un peu retenu et avec le caractère du récitatif," but, curiously, only mm. 9091 of the trombone cue.
Whether or not these kinds of differences matter to the everyday performing viola player is open to speculation. The more spacious layout of the NBE offers certain psychological advantages but yields 30 page turns as opposed to 28 in OBE; musicians always seem reassured by the traditional Breitkopf & Härtel look. In this particular case the primary advantage of the NBE parts would appear to be the line-by-line measure numbers and frequent rehearsal letters. In fact what might make the most positive difference to the overall shape and concept of Roméo et Juliette from using the NBE parts is simpler still: the proper numbering of the seven movements, as opposed to OBE's contorted and thoroughly wrongheaded attempt to squeeze the structure into four Roman-numeralled symphonic movements.
In sum one must hire the NBE materials for the operas and such less-familiar and available works as the Messe solennelle: there is no other viable choice, and the enterprise deserves supporting. For the symphonic canon it makes good sense to rent the NBE materials for works not already in the library and to base other decisions on considerations of budget and rehearsal strategy. Note that the only parts presently for direct sale by Bärenreiter are those for the Fantastique, co-published with Breitkopf & Härtel of Wiesbaden. Since these amount to the old B&H parts with measure numbers added (but no rehearsal numbers or letters), one would make the decision based on the cost of the materials and labor to prepare them for rehearsal.(10)
In the case of the Requiem I strongly recommend full NBE materials, with vocal scores purchased by the chorus. All too commonly, modern productions of this work bring together mismatched parts owned by the many groups who need assembling for the final rehearsals and performance: the orchestra plays from descendants of the Breitkopf & Härtel parts, the chorus sings from the cheap vocal scores (Schirmer and Kalmus), and if the conductor reads the NBE score, three rehearsal-letter schemes are afoot. The vocal scores, descended from a Brandus publication of 1882 based on the 1838 original, embrace a passage in the a cappella Quaerens me deleted from the 1853 second edition published by Ricordi, as well as dozens of details of voice-part disposition and declamation subsequently improved by Berlioz himself. Further, the timpani parts, as published by Breitkopf & Härtel, have been redistributed and bear little resemblance to Berlioz's own scheme: using them embraces unnecessary compromises to both the visual and (to my ear) audible impact of the timpani choir, particularly if drums are re-tuned, as it were, en route. The timpani parts can be arranged for four players each controlling four instruments, but who, having assembled the rest of the performing force, would want to bypass this central effect? (Additionally, there is the problem of the disposition of the cornets in the Sanctus and the trombone players after the Lacrymosa for the final movements, points on which Berlioz is somewhat ambiguous and the parts more ambiguous still.)
Enhanced accessibility to Berlioz materials is not uniquely the work of the NBE. In the recent past, for instance, facsimiles of two piano-vocal mélodies (Élégie en prose, La Captive) were included in Garland's series of Romantic French Song, 18301870, ed. David Tunley (vol. I, pp. 4960; 1994), and Peter Bloom published an edition of the piano-vocal Nuits d'été based in part on newly recovered manuscript material (Paris: Éditions Musicales du Marais, 1992). For the French bicentennial ceremonies in 1989, D. Kern Holoman and a team of graduate students at the University of California, Davis, published the newly discovered Chant du 9 Thermidor in score and parts, along with editions of the Marche pour la présentation des drapeaux and the Berlioz arrangement of the Marseillaise. A fascimile reprinting of the 1863 Collection de 32 Mélodies in conjunction with the Berlioz bicentenary in 2003 would go a long distance toward making these songs available to vocalists far and wide.
The "performance practice movement" reached Berlioz with Roger Norrington's performances and recording of the Symphonie fantastique in 198889, a natural outgrowth of his famous Beethoven recordings just preceding. Norrington's London "Berlioz Experience" of 46 March 1988 featured, in addition to the Fantastique (and lectures, roundtables, and recitals), a vibrant Roméo et Juliette--broadcast but not released on disc--and a Francs-Juges overture later released in a collection of Early Romantic Overtures. In 1991 John Eliot Gardiner's Orchestre Révolutionnaire et Romantique audio- and video-recorded a Fantastique in the Salle des Concerts du Conservatoire, where it was first performed. Gardiner's period orchestra and Monteverdi Choir released the first-ever Messe solennelle in 1993, likewise including laser-disc; 1998 saw publication of a remarkable pick-and-choose recording of the various versions of Roméo et Juliette. Both Norrington and Gardiner were already admired for Berlioz recordings with conventional forces.(11)
Efforts to approximate the conditions of a live performance during the composer's lifetime--the use of period instruments or replicas thereof, corresponding phrase and bowing strategies, and the composer's preferred layout of performing force and specified tempos--were in the late 1980s, and remain, very much in vogue. The European Community could boast a cohort of properly equipped young professional musicians available to travel the relatively short distances involved to constitute more-or-less authentic performances in the era's main venues. The big recording companies, anxious to acquire novel audio and video "content," arranged for the mixed-media packaging.
All five "performance practice" recordings are worthy additions to the Berlioz discography, the Gardiner Fantastique perhaps especially so by virtue of a video component that preserves something of the look and feel, if not in fact the acoustic, of a beloved hall now consigned to other uses. Even though the sounds of Beethoven-era strings and old-fashioned woodwind and brass have become familiar from recordings of music of other composers, one cannot help being intrigued by the woodiness of the woodwind, the stern brass voices of the Berlioz era, the short bow strokes and limited vibrato of the strings. These serve to remind us of a certain loss of personality that resulted from the late-19th- and 20th-century quest for a united, rich-and-round sonic ideal--the near equivalence in tone quality, for instance, of the tenor register in horn, bassoon, saxophone, and even cello--and for the "long line." If the net result seems "stringy" or "harsh" by comparison with traditional recordings, that is in many respects the point.
Norrington's particular focus has been on scrupulous attention to a composer's metronome markings. It does not take much to see the limitations of manuscript metronome marks: composers are notorious for attaching them without actually consulting a metronome; the old-fashioned clockwork mechanism can be unpredictable outside mid-range; and speed as imagined by a composer in his workshop is often markedly different from speed as sensed by working musicians in live concert venues. Berlioz himself favored cautious use of the metronome: music played by imitating the "mathematical regularity of the metronome," he wrote, "would be of glacial rigidity." But in his precedent-setting cycle of the Beethoven symphonies, Norrington demonstrated incontestably (except, perhaps, in the case of the famously slow scherzo of the Ninth) that a fundamentally musical use of the metronome markings, combined with other levels of musical rethinking, could afford intellectually provocative readings of old, familiar works. His readings are surprisingly novel, challenging to the intellect, and--often as not--obviously "right." Norrington's "nice and slow" March to the Scaffold from the Fantastique--"because that's the right speed"--has fostered a vogue for such accounts and perhaps a certain rivalry among conductors to see who can go slowest.
But is slow "right," and is Norrington's account of the last movement, at 10½ minutes, "nice and slow" or simply ponderous? For that matter, do the "performance practice" recordings really follow the composer's scheme? Berlioz's indications are as follows:
Movt. IV: Marche au supplice
Allegretto non troppo [halfnote] = 72
Movt. V: Songe d'une Nuit du Sabbat
Larghetto [quarternote] = 63
Allegro [dottedquarter] = 112
("mockery of the Beloved" I, m. 21)
Allegro assai [wholenote] = 76
("roar of approval," m. 29)
Allegro [dottedquarter] = 104
("mockery" II, m. 40: presumably slower
owing to technical difficulty)
sans presser ---
(Dies Irae, m. 127)
animez un peu (from m. 223) ---
Un peu retenu [dottedquarter] = 104
(Ronde du Sabbat, m. 241)
[HB: "The tempo, which should have
picked up a little, returns here to
that of m. 40, [dottedquarter] = 104."]
animez (m. 492) ---
Both movements consistently struggle to break free of the composer's metronome indication: owing, in the March, to the change of character between the lugubrious, sinister first theme and the swashbuckling second, as well as the mounting frenzy toward the end; in the Ronde du Sabbat, to the general sense that witches must dance faster than [dottedquarter] = 104. (A Sousa march in [6/8]--Marines, not witches--travels at [dotted quarter] = 120.)
The received French performance tradition takes both movements considerably faster that Berlioz's metronome marks. Charles Munch, who left the most persuasive Berlioz recordings of the first half of the century (his cycle with the Boston Symphony and the Harvard/ Radcliffe/New England Conservatory choruses has recently appeared as an 8-CD set),(12) begins the March at 76 and gathers consistently in speed to reach 94 at the end, having skipped the repeat. He takes the Mockery of the Beloved at a relatively strict 104 and the Round Dance just short of 132--i.e., a great deal faster than Berlioz says. The young Spanish conductor Ataulfo Argenta, who left the later of two recordings by the Société des Concerts,(13) follows much the same tempos as Munch. Colin Davis and the Concertgebouw take the March at 80, the Round Dance at 130.
Norrington, by contrast, begins the March well under 72, reaching only 74 by the end; the Round Dance goes at between 116 and 120 throughout. Gardiner's March to the Scaffold at begins at 80 and settles at 84, with a strict 104 for the "mockery of the Beloved," and a Round Dance at "traditional" speeds, i.e., starting at about 120 and reaching 132 toward the end.
During a London roundtable discussion of conducting Berlioz held in October 1995, Norrington, Macdonald, David Cairns, and Holoman discussed the matter of finding workable tempos for these two movements.(14) David Cairns found Norrington's reading of the Marche to the Scaffold powerful, because it was "more brutal and obscene." (Norrington himself remarked that "I haven't changed since I discovered this lovely metronome marking--how it kind of ponderously goes on.") But Cairns thought even the 120 of the last movement, as recorded by Norrington, "too held back."
The conversation turned naturally to other unaccountably slow markings in Berlioz: the [eighthnote] = 76 at the start of Harold en Italie, which, according to Macdonald, "has to speed up" at the entry of the viola or else is, according to Norrington, "distressingly slow"; the [dottedquarter] = 63 for La Mort d'Ophélie, which instead might well go "swingingly along" (RN: "She was floating down the river. Very nice and then she sank."); the [quarternote] = 50 for the "Te ergo quæsumus" in the Te Deum, a thoroughly impractical speed for the tenor soloist. Both the Love Scene from Roméo et Juliette and the Shepherd's Farewell from L'Enfance du Christ often seem lethargic when begun at the marked tempo, but this is more commonly the result of slipping from dotted-quarter into eighth-note pulse than a function of the calibration itself. The only reasonable explanation of a performance time left in Berlioz's hand for the Queen Mab scherzo, 10 minutes for a movement usually played in 7, is that it reflects an earlier, longer musical text.(15)
Norrington summarized an approach to these questions that amounts to giving first priority to Berlioz's metronome marks, but only so long as they can be made to work:
People get the impression that one sees a metronome mark or hears of a timing and then uses it for some religious reason. I only do it when they appeal. And sometimes I just don't do it. Harold is one of those cases: I don't see how I can do the metronome marking. You have to really be convinced. What's nice is that one so often is. The Beethoven symphonies are a case in point. Except for the Ninth, every single marking in the first eight symphonies is a revelation. It suddenly sounds right.(16)
"It's a revelation," I remarked, "if you're Roger Norrington and if you make such beautiful music from the revelation."
The London roundtable discussed matters of layout of the Berlioz orchestra, favoring the composer's preferred antiphonal placement of the two violin sections. For the rest, his attitudes were conditioned largely by the cramped quarters of the Salle des Concerts, where the violins filled the forestage and the chorus continued outward over the covered pit; a desk or two of bass strings was wedged between the violin sections, with the remainder of the orchestra consigned to steeply rising platforms reaching back to the walls of a removable shell.(17) (The one legible illustration of the 19th-century Société des Concerts at work clearly shows the double basses on the top platform, far removed from the violins.(18)) Few of these accommodations to an unusual room seem to merit duplicating for modern performance in more spacious quarters. Even the chorus-in-front strategy for Roméo et Juliette (and by extension for the other choral works up to Faust), though extraordinary in both visual and acoustic effect, is probably too costly in terms of inherent dangers for conventional modern performance. A large, wrap-around chorus accomplishes the same effect and can be controlled by the chief conductor.
The discussants--all of them conductors--closed by advocating a return to another lost tradition of 19th-century performance: applause between the movements. Norrington attributed the demise of applauding between movements, and of on-the-spot encores, to Sir Henry Wood's discouragement beginning in the 1930s. But "it's wonderful," said Cairns, "when people applaud when they shouldn't."
"It seems to me," said Norrington, "that a concert should be a good
deal more fun." Later than evening, in the Royal Festival Hall, I led a
salvo of applause after the Pilgrim's March from Harold en Italie--over
a chorus of shushing from the London regulars. Norrington turned and acknowledged
our corner of the parterre with a satisfied nod. But he did not
grant an encore.
Performance Practice Recordings
Symphonie fantastique (Norrington, 1989). London Classical Players, Roger Norrington, conductor. EMI CD CDC 7 49541 2. 1989
Les Francs-Juges ov. (Norrington, 1990), London Classical Players, Roger Norrington, conductor. On disc Early Romantic Overtures, with works of Weber, Mendelssohn, Schubert, and Wagner. Recorded 1988. EMI CD CDC 7 49889 2. 1990.
Symphonie fantastique (Gardiner, 1991). Orchestre Révolutionnaire et Romantique, John Eliot Gardiner, conducting. Recorded and filmed at the Salle des Concerts of the old Conservatoire (now the Conservatoire National Supérieur d'Art Dramatique), September 1991. Philips video 440 070 254-1 (laser disc), ... 254-3 (VHS cassette). 1991. Philips CD 434 402-2. 1993.
Messe solennelle (Gardiner, 1993). Donna Brown, soprano; Jean-Luc Viala, tenor; Gilles Cachemaille, bass-baritone; Monteverdi Choir, Orchestre Révolutionnaire et Romantique, John Eliot Gardiner, conducting. Recorded live at Westminster Cathedral, London, October 1993. Philips video 440 070 272-1. Philips CD 442 137-2. 1993.
Roméo et Juliette (Gardiner, 1998). Catherine Robbins, contralto; Jean-Paul Fouctrecourt, tenor; Gilles Cachemaille, bass-baritone; Monteverdi Choir, Orchestre Révolutionnaire et Romantique, John Eliot Gardiner, conducting. Philips 454 454-2 (2 discs). With alternative and variant readings. 1998.
Symphonie fantastique Performance Times
RCA 1954; rpt. CD 1996
London 1955; rpt. CD 1998
New Edition of the Complete Works
General Editor: Hugh Macdonald
Bärenreiter (Kassel, Basel, London, New York, 1967 )
Issued by the Berlioz Centenary Committee London in association with the Calouste Gulbenkian Foundation, Lisbon. Except for Symphonie fantastique and Te Deum, copublished with Breitkopf & Härtel of Wiesbaden, parts are identified with the Bärenreiter edition number (BA 5441ff.) plus 72; vocal scores with edition number plus 90. Scores with rental sets come with paper bindings and without commentary and notes. An "Index to the New Berlioz Edition" appears on pp. viiiix of NBE vol. 1a: Benvenuto Cellini. For latest information, search the on-line catalogue at www.barenreiter.com.
1 Benvenuto Cellini, ed. Hugh Macdonald (4 vols.: 1ad). 199496 [vol. 1d forthcoming]
BA 5441. Parts. Overture separately available.
2 Les Troyens, ed. Hugh Macdonald (3 vols.: 2ac). 196970.
BA 5442. Parts. "Chasse royale et orage" ("Royal Hunt and Storm") separately available. Eulenberg pocket score EE 6639 (1973); "Chasse royale," EE 1371 (1978).
3 Béatrice et Bénédict, ed. Hugh Macdonald. 1980.
BA 5443. Parts. Vocal score ed. D. Müller, W. Konold, J. E. Durek (BA 5443a, 1985). Overture separately available.
4 Incomplete Operas, ed. TBA. Forthcoming.
[BA 5444.] Les Francs-Juges, La Nonne sanglante.
5 Huit Scènes de Faust, ed. Julian Rushton. 1970.
BA 5445. Parts forthcoming, c. mid-2000.
6 Prix de Rome works, ed. David Gilbert. 1998.
BA 5446. Parts (by cantata title). Fugue (1826), La Mort d'Orphée, Herminie, Fugue à trois sujets (1829), Cléopatre,Sardanapale.
7 Lélio, ou Le Retour à la vie, ed. Peter Bloom. 1992.
BA 5447. Parts.
8 La Damnation de Faust, ed. Julian Rushton (2 vols.: 8ab). 1979, 1986.
BA 5448. Parts. Vocal score ed. E. Wernhard (BA 5448a, 1993).
9 Grande Messe des morts (Requiem), ed. Jürgen Kindermann. 1978.
BA 5449. Parts. Vocal score ed. M. Töpul, D. Woodfull-Harris (BA 5449a, 1992). Study Score TP 332 (1992).
10 Te Deum, ed. Denis McCaldin. 1973.
BA  and 5782. Parts. Vocal score ed. Otto Taubmann (BA 5782a; copublished as Breitkopf & Härtel 8061, 1978).
11 L'Enfance du Christ, ed. David Lloyd Jones. 1998.
BA 5451. Parts forthcoming, c. 1999.
12 Choral Works with Orchestra. 2 vols.: 12a, ed. Julian Rushton, 1991; 12b, ed. David Charlton, 1993.
BA 5452/I, vol. 12a: Resurrexit, Scène héroïque,Chant sacré, Hélène, Quartetto et coro dei maggi,Sara la baigneuse,Le Cinq Mai.
BA 5452/II, vol. 12b: Tristia (Méditation religieuse, La Mort d'Ophélie, Marche funèbre pour la dernière scène d'Hamlet), Vox populi (La Menace des Francs, Hymne à la France), Chant des Chemins de fer, L'Impériale. Parts: Chant des Chemins de fer, Tristia (others forthcoming).
13 Songs for Solo Voice and Orchestra, ed. Ian Kemp. 1975.
BA  and 5782, 5784. La Belle Voyageuse, La Captive, Le Jeune Pâtre breton,Nuits d'été, Le Chasseur danois, Zaïde, Aubade. Nuits d'été parts. Nuits d'été vocal score with transpositions for mezzo-soprano, ed. D. Woodfull-Harris (BA 5784, 1995).
14 Choral Works with Keyboard, ed. Ian Rumbold, 1996.
BA 5454. See NBE index, vol. 1a.
15 Songs for One, Two, or Three Voices with Keyboard, ed. TBA. Forthcoming c. 2001.
BA 5455. See NBE index, vol. 1a.
16 Symphonie fantastique, ed. Nicholas Temperley. 1972,
BA  and 5781. Parts (BA 5781 [65 winds; 74,75,79, 82, 85 individual strings], copublished as Breitkopf & Härtel 4929). Study score TP 331 (1972).
17 Harold en Italie, ed. Paul Banks, forthcoming 2000.
18 Roméo et Juliette, ed. D. Kern Holoman.. 1990.
BA 5458. Parts. Vocal score ed. E. Wernhard (BA 5458a, 1995). Study Score TP 334 (1996).
19 Grande Symphonie funèbre et triomphale, ed. Hugh Macdonald. 1967.
BA 5459. Parts. Eulenberg pocket score EE 6642 (no. 599) (1975).
20 Overtures, ed. Diana Bickley. 1999.
BA 5460. Waverley, Le Roi lear, Rob-Roy, Le Carnaval romain, Le Corsaire. (See above for Cellini and Béatrice ovs.) Parts forthcoming.
21 Other Orchestral and Instrumental Works, ed. Hugh Macdonald. Forthcoming 2003.
BA 5461. Includes Rêverie et Caprice; see NBE index, vol. 1a.
22 Arrangements. 2 vols.: 22a: Arrangements of Works by Gluck, ed. Joel-Marie Fauquet.; 22b: Arrangements of Works by Other Composers, ed. Ian Rumbold. Forthcoming.
BA 5462/I, 5462/II. See NBE index, vol. 1a.
23 Messe solennelle. ed. Hugh Macdonald. 1994.
BA 5463. Parts. Vocal score ed. E. Wernard (BA 5463a, 1994). Study score TP 333 (1993). "O Salutaris" for chorus, organ, BA 6394.
24 Grand Traité d'Instrumentation et d'orchestration modernes, ed. Hugh Macdonald and Peter Bloom. Forthcoming.
25 Catalogue of the Works of Hector Berlioz, by D. Kern Holoman. 1987.
26 Portraits, ed. Gunther Braum. Forthcoming 2003.
1. See Grand Traité d'instrumentation et d'orchestration modernes (Paris, 1843; conducting treatise added for second edition, 1855), Hol. A2 (Holoman, Catalogue of the Works of Hector Berlioz, New Berlioz Edition vol. 25, 1987). I have treated some of these same issues in "The Present State of Berlioz Research," Acta musicologica 67 (1975), 3167; and in Berlioz (Cambridge, Mass., 1989), notably pp. 34861. For clarifying various points in 1999 I am grateful to David Cairns, Hugh Macdonald, Michael Steinberg, and the editorial staff of Bärenreiter Verlag, Kassel.
2. To summarize: the sylph's scene from La Damnation de Faust (from 1849), the duo-nocturne from Béatrice et Bénédict (from 1863), La Fuite en Égypte (part II of L'Enfance du Christ; from 1864), Le Carnaval romain (from 1873), the Francs-Juges ov. (from 1874), La Mort d'Ophélie (from 1875), Roméo et Juliette (excerpts from 1877; complete from 1879), Le Corsaire (from 1880), Benvenuto Cellini ov. (from 1895), Le Roi Leari (from 1899), Marche funèbre pour la dernière scène d'Hamlet (from 1899), Sara la baigneuse (from 1903), Harold en Italie (Pilgrim's March from 1874, complete from 1907, a vehicle for the celebrated violist Maurice Vieux), Symphonie fantastique (from 1917).
3. To summarize: Karslruhe (1890, Mottl), Munich (1895, Levi; 1907, 1908, Mottl), Cologne (1898, Klessel), Leipzig (1900, Gorter). A list of performances of Les Troyens through 1987, compiled by Louise Goldberg, appears in the Cambridge Opera Handbook Hector Berlioz: Les Troyens, ed. Ian Kemp (Cambridge, 1988), pp. 21627.
7. In a 1974 performance with the Orchestre de Paris at the Kennedy Center, Washington, DC, Daniel Barenboim pulled a revolver from his belt and fired it. This silliness provoked laughter from the audience and large and particular guffaws from those who made a connection with the conductor Artur Rodzinksi, said to have packed a pistol at orchestra rehearsals.
10. My figures for the Fastastique, at 1999 exchange rates, are as follows, for a set including 6 of each strings parts: NBE: $462, Kalmus: $607.50. Details: NBE: score, 110 DM; wind parts 230 DM, strings parts 36 x 12 DM = 432 DM; total 772 DM = $412, + $50 labor (add rehearsal letters); does not include clothbound vol. 13. Kalmus: score, $70; set $250; extra strings, $187.50, labor, $100 (add m. nos. And rehearsal letters). Exclusive of shipping and handling.
11. Gardiner with two different important recordings of the Berlioz Mélodies (Oiseau-Lyre, 1968, Musifrance, 1990), L'Enfance du Christ (Erato / RCA, 1988), La Damnation de Faust (Philips, 1989), and Gluck's Orphée et Eurydice in Berlioz's arrangement (EMI, 1989); Norrington with Songs for Chorus (Argo, 1969).
12. Christian Wasselin notes "I don't think that Charles Munch was a great Berliozian conductor. ... [He] was very different from one evening to another. When you hear the different recordings of the SF by Charles Munch, you have three or four different conceptions with different timings and different tempos." See the question-and-answer sessions following his paper on "The Culture of Paris at the Time of Berlioz," Berlioz Society Bulletin 155 (Spring/Summer 1996), p. 26.