David Charlton

© 2007 David Charlton

I: Tasso’s source II: Boismartin’s text III: The Epilogue IV: Berlioz in 1828

V: The idée fixe theme VI: ‘Beethoven has turned his brain’

The aim of these informal thoughts is to consider Berlioz’s second Prix de Rome cantata, especially with reference to its source, Tasso, and then in the context of 1828. Herminie was written at a complex moment in the composer’s development, and is arguably more interesting than its traditionally slender critical attention suggests. Nevertheless, there has been a newer move towards more detailed focus on his apprentice years; and the presence in this cantata of the earliest known form of the idée fixe from Symphonie fantastique will surely always guarantee it a special place in the Berlioz canon. The crop of valuable Berlioz literature since 2000 includes David Cairns’s second edition of The Making of an Artist and Julian Rushton’s The Music of Berlioz, together with Peter Bloom’s 1998 biography, Oliver Vogel’s full-scale study of the earlier works (in German) and Rémy Stricker’s 650-page monograph (in French).1

I: Tasso’s source. It is of course true that Berlioz had no control over the subject-matter of the cantatas. But it is also true that the composer required a degree of self-identification with any musical project he undertook and composed. This worked in the cantatas in at least two ways: Berlioz needed to imagine points of identity with the poetic sources provided and, when the cantatas were finished, he would transfer out of them whatever music he wished to transform and re-employ; they would arrive at different yet related contexts which would allow for better self-identification but, of necessity, retained core features with the emotional and expressive worlds decided on by the Institut des Beaux-Arts, at least in portions of each of the four cantatas undertaken.2 All recent commentators, when writing in any detail, have made similar points.

It is therefore necessary to understand something of who Herminie was, and how Berlioz might have construed her. His first mention of Torquato Tasso in the surviving letters comes in an Italian letter from 1831, and is merely a passing mention.3 In the feverish world of 1820s literature, one may legitimately wonder how far a young Romantic had read or discussed Gerusalemme liberata. This epic had been completed around 1575, and of course had furnished the plots of celebrated operas, some known to Berlioz: probably not Campra and Danchet’s masterful Tancrède of 1702, but without doubt the Armide of Quinault set first by Lully and then Gluck. ‘Sur Armide et Gluck’ is the title of one of Berlioz’s earliest publications, proclaiming in December 1825, ‘Je sais tout Gluck par cœur’, etc.4  In 1831 he visited Tasso’s tomb with Mendelssohn. But of his acquaintance with the actual epic we know not much.

In history, crusader states were founded in 1099 at both Antioch (by Bohemond I) and Jerusalem, where, as Tasso recounts, Godfrey of Bouillon established a kingdom that eventually lasted until 1291. Bohemond’s nephew was Tancred, who ruled Antioch from 1111 to 1112 after a period as regent.5 Erminia (Herminie) is introduced by Tasso in Canto 3; then she appears in the episode glossed by Berlioz’s Prix de Rome cantata, in Canto 6. Finally we meet her in Canto 19. So she is present throughout, but always on the margin of the action. A Muslim, like her best friend Clorinda, Erminia is nevertheless a stranger to the environment of the poem, having been born princess of Antioch, the daughter of king Cassano. In the flashback sequence in Canto 6 Tasso reveals that even as Antioch was destroyed by Tancred and his knights, the hero chivalrously protected Erminia’s person and her wealth.

But he received her as some sacred flower,

Nor harmed her shrinking leaves; midst outrage keen

Pure and inviolate was her virgin-bower. […]

The generous knight in every act and word

Honoured her, served her, soothed her deep distress,

Gave her her freedom, to her charge restored

Her gems, her gold…

But this freedom for her comes at the price of emotional turmoil: ‘never did Love bind / With his most charming chain a more devoted mind. / Thus, though in person free, her spirit ever / Remained his willing thrall…’.6

Later, having taken her ailing mother to Jerusalem, where they seek asylum, Erminia meets the warlike Clorinda, a Persian by origin. Her central role in the epic begins in Canto 2, when she rescues the unfairly condemned Christian woman Sophronia. Already she is a warrior leader.

All feminine attractions, aims and parts,

She from her childhood cared not to assume […]

Whilst yet a girl, she with her little hand

Lashed and reined in the rapid steed she raced,

Tossed the huge javelin, wrestled on the sand,

And by gymnastic toils her sinews braced.7

In Canto 3, the contrast between the two females is fully established. The Frankish knights are marching on Jerusalem; Erminia, alone since her mother’s death, ‘lovely as the dawn’, by ‘her winning charms’ maintains respect at the court, and remains a bystander as Clorinda leads an early attack on the knights. Aladine, ruler of Jerusalem, answers Erminia’s question at this point, as both observe events from a high tower: ‘Who then is this, that in fierce grace outstrips / All other knights?’ Erminia can only hide her emotions as Tancred’s identity is revealed. Later in the epic it is Clorinda who will die first, at the hero’s unwitting hand, in Canto 12: she receives baptism from him before losing consciousness. Erminia’s final role will be to cure the half-dead Tancred by means of her knowledge of herbs (Canto 19), following his final and successful duel with the Muslim strong-man, Argantes.

Indeed it is Argantes who, in Canto 6, has duelled at length with Tancred: evenly matched, they are separated as night starts to fall, and resolve to fight again after several days’ respite. Erminia, hearing of the intended renewal of the conflict, becomes agonized in the imagining of Tancred’s future wounds, but confident in her ability to cure them. Her inner conflict is expressed by the arguments first of Honour, who berates her for wanting to actually visit Tancred and offer him succour; then of Love, who encourages Erminia to be true to her gentle nature, follow her instincts, and seek him out to heal him. All this remains a secret to Clorinda, in spite of the women’s long conversations. One day Clorinda is for some reason elsewhere, and this is when Erminia, seeing her friend’s armour hanging up, is tempted to wear it, try to leave the city, and gain access to the Christian camp. Her female servant has at first to help her to walk, and inevitably ‘The hard cold steel oppresses and offends / Her delicate smooth neck and golden hair.’8 She sends her squire ahead to meet them with horses; darkness is again falling, and they reach the city gate after winding through the back streets. But again, she does not meet the hero: although the gates are opened, and the squire sent on to apprise Tancred of the visitor, Erminia and maid are forced to take flight from Polypherne, an Italian who attacks them because his own father has recently been killed by Clorinda, whom he takes Erminia to be. Erminia is eventually looked after by a Syrian shepherd, and spends time on the banks of the river Jordan.

II: Boismartin’s text. All observers, Berlioz included, seem to have commented on the unchanging structure of Institut prize poems: ‘the inevitable dawn, then the first recitative followed by the first aria…’,9 to a total of three of each. Berlioz’s published memoirs contain little about the cantata Herminie apart from the way he amplified the final aria text – P. A. Vieillard de Boismartin’s ‘Air de mouvement’ – into an ABA’ structure by expansion, making as the centrepiece Erminia’s prayer to the Christian deity. More deserves to be noted, even within a cantata universally voted to be conservative by the composer’s lights. The text, as recorded by the Institut, appears in NBE vol. 6, Appendix I. The first aria text looks back to Erminia’s love even as it was born at the moment of her rescue in Antioch; its best line refers to her long-kept secret, ‘Mon cœur brûle! … et ma bouche est réduite au silence!…’. The following recitative, using tortured neoclassical diction, refers to Tancred’s recent duel with Argantes, and Erminia’s fears for his present physical state; the following lyric movement was supposed to be a ‘Cavatine’, its single quatrain oddly marked ‘agitato’. Here Erminia rhetorically begs Tancred to avoid further risks: ‘Arrête, cher Tancrède, arrête’. Without apparent later censure by the judges, Berlioz consequently decided to borrow the two first lines from the following (i. e. the third) recitative to create a contrasting middle section for his aria, again effectively expanding the field of expression and character. These two lines were originally couched as a single alexandrine followed by an octosyllabic four-stress line; but Berlioz omitted the bracketed words below, a possible reason for which is considered later.

J’exhale en vain [vers lui] ma plainte fugitive,

Je l’implore; il ne m’entend pas.

Peter Bloom’s translation reads, ‘In vain I give voice to my evanescent plaint, I entreat him, he hears me not.’10 In the third recitative Erminia, having envied Clorinda’s strength, notices her armour, and imagines her own plan of action. Boismartin’s lines are strong and dramatic this time, within the limits of ancien régime habits of diction. It suffers not at all from having had its first lines removed, save in the now-incomplete rhyme scheme. But Boismartin’s text for the third and final aria does not actually show Erminia doing anything: the first quatrain apostrophizes Clorinda’s armour, Erminia girding her own courage, and the last octet constitutes her prayer to the god of the Christians, swearing religious conversion should Tancred somehow be permitted by Heaven to love her. This idea was perfectly false to Tasso, who, as we said, used Honour and Love as conflicting inner voices. Neither would the original Erminia convert: in Canto 19 her heart-rending speeches no longer express the possibility of love, but of her own proximity to death, at least before she heals Tancred’s wounds; it seems that she remains chaste to the end. Obviously Boismartin hoped to create an acceptable and plausible resolution, given the absence of any death at this point in the epic, and yet the static nature of the imagined final scene is at wholesale variance with Tasso: he made it to be Erminia’s greatest moment of resolve, and indeed her long-delayed conversion of passion into action. Berlioz, from familiarity with one source or another, must have known the actual course of events in the poem. Perhaps the Institut procedure allowed for an informal contextualisation of the chosen source of the set text, just to make sure candidates were fully able to respond to it. Or perhaps everyone then was better-read than we give them credit for now. Consequently, Berlioz once again expanded the given imaginative field, placed his Prayer in the middle of a final ABA’ aria, and had Erminia’s closing high cadences indubitably lead towards her desperate actions. For the same reason, he once again changed the text, borrowing and slightly modifying another two lines from the preceding (third) recitative, and turning Boismartin’s semantic ‘what if’ into his own ‘I shall’:

[Original recitative]: Si j’osais m’en couvrir!… si, trompant tous les yeux,

Sous cette armure aux périls consacrée,

Je fuyais d’Aladin le palais odieux,

Et du camp des Chrétiens, allais tenter l’entrée!

[Final aria, inserted within text of reprise:]

Oui! Sous cette armure aux périls consacrée,

Du camp des chrétiens je vais tenter l’entrée.

And his orchestral epilogue, mentioned below, completes the representation of the heroine’s active exit.

III: The Epilogue. Although we have seen that the orchestral epilogue – functionally, a coda – corresponds to an authentic response to Tasso’s epic, the music itself is to me curiously ineffective. This is in spite of the composer’s use of what he would later call ‘rhythmic modulation’, in order to recall the ‘Motif de la Prière’ (from its G major Largo at bar 119) at ‘Allegro impetuoso vivace’ tempo, making it sound as if the Prayer is intoned or imagined at its original pace, though now in the main tonic of E major. As Macdonald notes, this is a ‘device he had so successfully worked in the overture to Les Francs-juges’,11 recently promoted publicly in Berlioz’s concert of 26 May 1828, conducted by his friend Nathan Bloc. Moreover, in 1825–26 he had composed the same technique into Scène héroïque, Final, where the tempo becomes twice as fast at bar 60 and, later, the main theme ‘Le monde entier’ returns in triumphal layered phrases, notated at half the original speed. The premiere of this work was also on 26 May 1828. For these reasons alone the epilogue to Herminie must have come easily as a device; the figurative power of the accompanying string motif (derived from the vigorous dotted rhythm of the third aria) is to me simply too baldly uncharacteristic: the original figure is ‘liquidated’ (in Schoenberg’s term) rather than retained. Contrapuntal drama in this manner relies on our immediate identification of the contrasting meanings, yet Herminie’s string figures seem too generalised. Maybe, also, there was lurking a logical contradiction known to Berlioz, inasmuch as Erminia, in the epic poem, does not immediately ride off on horseback. Having donned the armour, with her maid’s help, she finds that she needs the latter’s support merely in walking, due to the added weight. It is only when the women manage to rendezvous with their squire that they are able to mount horses and proceed to the city gate.

Oh, how fatiguing every moment grew

The unequal weight! how slow her faltering pace!

Faint to her handmaid for support she drew,

And by her help moved onward a short space.12

So there may have been a visual and conceptual confusion here that Berlioz could not overcome in time. In any case, in the coda, why are Erminia’s thoughts still on the prayer, instead of more focused on physical realities, including the object of her adventure?

However, everyone in recent years is convinced to the contrary: for David Cairns, for example, it is ‘pure Berlioz’, and the ‘long diminuendo suggests Erminia galloping away on her desperate errand of mercy’.13 Julian Rushton is neither for nor against it: Erminia goes ‘forth to battle’ (an undoubted slip) ‘over a throbbing continuation of the allegro…’; ‘This passage foreshadows many effects of simultaneity in later works.’14 Hugh Macdonald finds it so ‘dramatically imaginative’ that he considers Berlioz ‘should also have mentioned the close’, i.e. in the Memoirs.15 Rémy Stricker goes the furthest in justifying our admiration, by analysing what seem at first curious features of the Prière, as heard against the tremendous previous sequences of rhythmic activity in arias 2 and 3: he especially observes the gradual introduction of timbral detail which comes to effect a poetic duality between ‘stasis and iridescence’.16 Thus in the epilogue-coda, when the prayer theme is counterpointed with the disappearing sound of the horse’s hooves, ‘held down over a pedal note’, the result is ‘a stroke of genius’.17

I persist in thinking that the composer may have been overly affected by the hearing of his Tableau, fading into nothing, which closed the first prize cantata, La Mort d’Orphée, and was a great success in the rehearsal for the 26 May concert (the actual public hearing had to be cancelled). There is also a passage in the Berlioz letters from this time which ought not to be overlooked, and suggests why he will avoid, seemingly at any cost, normal final cadences of resolution: reporting on the May 1828 concert to his father, he wrote,

In general I avoid like the plague those commonplaces which all composers (except Weber and Beethoven) put in at the end of their movements; it’s a sort of confidence trick that announces: ‘Get ready to applaud, we’re near the end’; nothing is more pitiable to me than those banal, conventional phrases which make all music sound the same.18

Thus a dogmatic insistence on finding some alternative form of ending almost guaranteed that Berlioz needed to find some more visionary solution. ‘He closed all his cantatas with an instrumental representation of the end of the drama.’19 But the coda to Herminie was far too ordinary ever to be re-used elsewhere later.

IV: Berlioz in 1828. Commentators labour under the fact that rather few of the composer’s letters survive from this period; but then it may not have been a period when many more private letters were written. In 1827 Berlioz had undergone the Shakespearean coup de foudre at the performances of Kemble’s company at the Odéon; he had conceived a passion for Harriet; and he first heard the Eroica and Beethoven’s Fifth when the Société des concerts du Conservatoire’s first season opened on 9 March 1828. Thus, he tells us,

Ravaged day and night by my Shakespearean love, which the discovery of Beethoven’s music, so far from alleviating, seemed only to make more painfully intense…perpetually in a dream, silent to the point of dumbness, unsociable, unkempt in appearance… I… put my name down for the examination at the Institute.20

This is a gross concealment of his phenomenal mental activity, and of the tremendous efforts which made his concert of 26 May such a success, even politically speaking: it was attended by many names in the musical world, and attracted a reasonable press coverage.21 Even the letters to Ferrand that allude to his despair also allude to his activity. That of 28 June, for example, opens with a confession of nine months of anguish ‘through which only music has allowed me to survive’; and ‘of many things on the go at the moment, nothing positive’, followed by running back from Villeneuve-Saint-Georges, and ‘de la joie, morbleu!’.22 The phrase ‘que la musique seule me fait supporter’ suggests at least some measure of compositional activity in reaction to emotion, although Herminie, as noted above, remains his first known composition following the hearing of the first Beethoven programmes. At the same time, the laconic and objectified accounts surrounding the 1828 Prix de Rome (in the Mémoires) are mirrored by the letter of 15 July in that year to Ferrand, during the composition period, when matters of ongoing life are foremost in his mind, and ‘for a fortnight I have been shut up at the Institute; this awful competition is my final recourse because it is a source of money and one cannot do anything without the filthy stuff’.23 A job had to be done. Yet it was a supremely professional one: the musical achievement of Herminie is surely far wider that of Orphée, and to me more satisfying in the wealth of polished responses to the text itself: many pages do attain a feeling of complete and controlled Berliozian maturity.

Although we perforce rely on guesswork in our interpretation of the composer’s capacities in 1828, we can perhaps nuance the composer’s account, and also the estimate of David Cairns, who writes, ‘He might be in no state to compose, but he could organize’, namely the concert of his own music on 26 May.24 Harriet Smithson was near the apogee of her fame, while Berlioz was depressed by ‘the gulf that separated them’. When it came to responding to Tasso’s Erminia, as Cairns speculates, the situation was analogous to his own, with the genders reversed: he the sidelined aspirant and she the heroine of the hour. This is surely reasonable, and encourages our consideration of the famous theme in the light of its original context. At this point a new theory by Hugh Macdonald comes into play. In this article, which coincidentally uses a very similar approach to my own and deals with the same kind of materials, Macdonald reviews the historical evidence for an early cycle of compositions in 1827–28 that Berlioz may have sketched and developed, and entitles this Quatre scènes de Roméo et Juliette.25 The argument, in brief, is that Berlioz seems likely to been have engaged in compositional work in 1827–28 that he chose not to divulge in Memoirs, and that must have been related to Shakespeare’s Romeo and Juliet. We have various independent evidence for at least two such settings, best regarded as Scènes in the same sense as the later Huit scènes de Faust: music drawing on episodes in the play, just as Berlioz later designed ‘Faust’ music around the lyrics within Goethe’s text.26 Macdonald’s conjectured four ‘scenes’ are: Mercutio’s Queen Mab speech; the Capulets’ ball; the balcony scene or similar declaration of passion, in which the future idée fixe theme must have played a part; and Juliet’s meditation ‘How if, when I am laid into the tomb’.

Any love music that represented, say, the balcony scene would also be his own ardent declaration to the object of his passion, Harriet Smithson, and it would be natural to represent this passion musically with a theme of an impetuous, striving character [e.g. the idée fixe theme][ … ] The first fifty-three bars of Herminie could easily be borrowed from a Roméo et Juliette love scene without damaging its integrity even if the work itself had still not been rejected, since Prix de Rome cantatas did not qualify in his mind as the composer’s authentic work.27

In the following section an attempt will be made through literary analysis to discriminate more precisely where Berlioz felt ‘authentic work’ to have taken place in these cantatas.

V: The idée fixe / Tancrède theme. The cantata Herminie uses a theme which for ease of reference could be labelled after its definitive employment in Symphonie fantastique, but which I will hereafter also label as the ‘Tancrède’ theme, in order to bestow as much integrity on the cantata as a conception as possible. The question has often been raised as to whether the ‘Tancrède’ theme was first conceived in August 1828 and, whether or not if so, what might be revealed about its original nature. Julian Rushton observed that ‘It is not known whether the best ideas in these cantatas occurred to Berlioz under examination conditions, or whether he took them with him as a legitimate aid to speedy composition.’28 He shrewdly notes that among the manifold possibilities concerning musical themes of which we are ignorant, Berlioz had been working on now-lost or uncompleted projects like Richard en Palestine (after Scott), which in this case was a crusader subject, like Herminie. Elsewhere, he thinks that ‘we might guess that its [the Tancrède theme’s] original context was the Florian opera [Estelle et Némorin, mid-1823] rather than the oratorio [Le Passage de la Mer Rouge, 1823–24], or Richard en Palestine’.29 If we navigate historically in oceans of ignorance, we should never underestimate the possibilities, especially following the discovery in our own time of astonishingly mature music in parts of the Messe solennelle.

David Cairns, rather similarly, feels that

More likely…when composing … under competition conditions he tried out ideas already sketched that lent themselves to the particular requirements of the moment. Though it is not impossible that the idée fixe first came to him ‘en loge’ at the Institute, my feeling is that it already existed, associated with Harriet Smithson and his unrequited passion for her … meanwhile he used it to express the lover’s exalted yearning for a beloved who is both unresponsive and physically distant.30

Since it is part of our task to account for the creative status, as well as the treatment, of the idée fixe / Tancred theme, I will here put forward a form of analysis that relies on close reading of the Mémoires, and approach the question by surveying the degree of apparent self-identification in the cantata scores. Breathtakingly original music was invented for the first and third prize cantatas, inspired by authorial identification with, first, the god of Music himself, with masculine sentiments after the loss of a beloved, and,second, with the heroic sublime of the closing Tableau, when Orpheus’s violent death is succeeded by fragments of his earlier aria played by an imagined shepherd, and set against wispy arpeggios from Orpheus’s half-broken lyre. Then, in Cléopâtre, the force of identification with the suicidal heroine is virtually Shakespearean. (One wonders whether Berlioz saw that ‘The barge she sat in’, Enobarbus’ speech, finds an exact equivalent when Boismartin’s text has Cleopatra looking back in Aria 1 at past glories: ‘Où sur le sein des mers, Comparable à Vénus, / D’Antoine et de César réfléchissant la gloire, / J’apparus triomphante aux rives du Cydnus!’)

It is probably significant that in the Mémoires Berlioz chose to mention only one or two specific elements from each cantata, and that these elements apparently relate to music which was composed ‘en loge’, as well as which was good enough for re-use. These are surely portions which can be claimed as having engendered self-identification. If this musical autobiographer conceals much, it seems to me, that which he reveals is likely to contain the truth, and the narrative flow appears to direct attention in these chapters to what was written at the time, during the competitions. In Orphée ‘I think the final movement had its points’ – this was the Tableau just discussed, which Macdonald has called ‘a page of the highest poetry’.31 Berlioz’s detailed footnote in the cantata score, explaining the scenic landscape and meaning of the music, makes it clear that this music was imagined in situ. Furthermore, in Cléopâtre the autobiographical account leaves no doubt that the movement entitled ‘Méditation’ (‘Grands Pharaons’) was inspired and written en loge; and he even reports re-using this music in Lélio.32 In Herminie it is just the Prayer in Aria 3 which the composer allows himself to mention, writing, ‘and without doubt if there was anything good in my score, it was that andante’; it was in fact re-used both in Neuf mélodies and much later.33 He reinforces this immediacy by telling the story of the janitor Pingard’s appreciation of it.34 Perhaps another point supports this pattern: the Prayer in Herminie is one section that, historically, the composer might have wished not to mention on any grounds of particular merit, since he had already written (though not published) at least one other Prière, also in G major, also to the Christian deity, also conceived for high voices and also following music in a faster tempo. This prayer occurs in the Scène héroïque as the third movement, ‘Astre terrible et saint, / Guide les pas du brave!’

Notable for its absence from this small roster of cantata references is the presence of the Tancrède / idéé fixe theme; even if we are not surprised that the relatively obscure ‘Chant sacré’ is not mentioned at this point, one might have thought that the Symphonie fantastique ought to have deserved to be. Neither a sense of immediacy nor a sense of inspiration attaches in the Mémoires or other document to any section of Herminie in which the above theme occurs.

The feeling of some commentators that the Tancrède / idée fixe theme could have predated Herminie is surely strengthened by the facts of its musical construction and way that the voice is attached to it in the second aria of the cantata: see also the example on pp. 85–86 of Rushton’s The Music of Berlioz. The voice sings what it can, but it is the orchestra that guarantees its identity. If Berlioz had composed the theme for ‘J’exhale en vain’ en loge he would, arguably, have found one that worked better for the voice, and that incorporated the missing words ‘vers lui’ (see above, section I), especially since the couplet had to be borrowed from the text of the following recitative. These two words are not entirely without meaning and force, though are expendable, and were necessary to the metre. One further thought may be relevant: the instrumental nature of the theme itself. In its lengthy, winding phrases, if not its ‘impulsive neurotic bulges’,35 the theme closely resembles certain well-established achievements in Berlioz’s orchestral works: the second subject of the Francs-juges overture (32 bars) and the cello melody in the slow introduction to the Waverley overture (about 50 bars), music which in both cases had been publicly presented on 26 May 1828. Whether or not this Francs-juges theme dated back to youthful chamber music, the idée fixe theme could well have been in stock already.

Oliver Vogel’s approach to this long-breathed melody assumes that it was specially composed, on the lines of the second subject in Les Francs-juges, although the reasons are bound up with his implied belief that Erminia was actually a warrior herself, which is a misconstruction.

The lengthy work on Les Francs-juges left behind it some traces on the theme [i.e . the idée fixe / Tancred]. As an expression of the heroine’s inclination to fight, the opening (in its triadic outline and its rhythm) preserves something of the warlike gesture of the hero’s theme in the overture, itself emerging residually in the music of the chief of soldiery at the start of the first act [at ‘Guerrier, ton noble cœur’, bar 40].36

Vogel’s music example, on page 357 of his study, usefully juxtaposes the general similarity of these musical outlines; we can easily accept that, even if never a warrior, Erminia at least must think of herself as capable of taking on Clorinda’s appearance. Nevertheless, as we shall argue below, the way the theme is employed shows that this music cannot actually represent herself, but rather Tancrède.

It is thus essential to approach the theme’s identity by thinking closely about its functions in the cantata. In Orphée, the central melody of the work had maintained a reasonably straightforward musical and textual integrity: Berlioz used a shortened form of Orpheus’s aria melody (starting at bar 100) both for his orchestral introduction and the epilogue (the ‘Tableau’). The words to which it is sung by the soloist form its essential message of consolation and sublimation, ‘Ranime mes accents, Seconde mon délire, Viens, ô ma céleste lyre! Retentis dans ces bois.’37 The Orphic lyre is the guiding thread of the cantata, even though the theme’s tonality varies on each rehearing, from D in the Introduction, to E and A in the aria, and A flat in the Tableau. The melody’s meaning is made clear and does not vary, even as spread across the whole extent of the work. So the use of a recurring theme (not to mention several other features) was of itself nothing new in Herminie.38 However, the poietic uses of the Tancred / idée fixe theme bear no resemblance to events in Orphée. Their varied functions are dramatically diverse, and rather encourage the speculation that the composer was using his time-limited energies to weave in a pre-composed theme in more sophisticated fashion. In any case, the instrumental introduction to Herminie sets neither the general scene in old Jerusalem (whereas that to Orphée clearly painted the antique landscape) nor the time of day (approaching dusk), even if Vogel’s analysis does allow us to hear the echo of crusading strife. The theme (G major) is however fixed in counterpoint against agitation-motifs in lower strings; the latter must relate to human movement or heart-beat39 while the Tancred / idée fixe theme expresses inner restlessness. Two other observations: the theme becomes extended in length only when it is heard a second time (bar 20). It then grows in slightly unwieldy fashion, unworried about its harmonic direction, retreating quickly from A major at bar 37. Was Beethoven to blame? With this procedure one might well juxtapose the feature of the winding and growing main theme in the third movement of Beethoven’s Fifth Symphony, together with the presence in this music of some harmonic ambiguity. Precisely these grouped characteristics had inspired E. T. A Hoffmann’s fascinated commentary on the movement in 1810, and his music example.40 My feeling is encouraged by the independent observation of Vogel: although not mentioning the Fifth Symphony, he points to the same kind of coincidence:

Following the exposition of the antecedent, the developed consequent brings a contrasting instrumental effect. During the second hearing (bars 20 to 55) Berlioz in addition displays a motivic-thematic working of his theme, which dies away in separated statements of the main motive. An immediate taking of measure of the German symphonies presented in the performances of the Société des Concerts is perceptible here.41

Secondly: a small but important clue occurs in bar 7, namely the winds’ octave interpolation of Berlioz’s ‘fingerprint’: V–flat VI–V, which is also a basic motif in the substructure of the Symphonie fantastique.42 This fingerprint, as it derives from a later portion of the Tancred / idée fixe theme which is otherwise absent from the introduction, suggests in retrospect that the full expressive nature of the theme was present in the composer’s experience, but was simply not yet divulged. As in other respects, this temporary limitation of the expressive treatment is consonant with the theme’s possible pre-existence, and its quite complicated function in the cantata.

Recitative 1: Rushton’s sensible commentary reasons that ‘The common factor of these passages [Tancred / idée fixe occurrences] is not an image of Tancred, but an image of Herminia’s feeling for him’ and he duly compares the Symphonie fantastique wherein the ‘idée fixe does not represent the beloved, but the protagonist’s perception of her’.43 Nevertheless, for some reason he implies that the first appearance of the Tancred / idée fixe theme is at bar 72. It actually comes at bar 57, and is therefore fixed conceptually as follows: ‘Quel trouble te poursuit, malheureuse [a bass variant of the theme’s first bar now enters in unison], malheureuse Herminie?’ [the variant is then repeated, in developed variation] ‘Tancrède est l’ennemi’. So the theme ‘pursuing’ Erminia is soon identified with Tancred’s image. At bars 72 to 82 it intensifies into a chromaticised metaphor of obsession (Rushton’s apt word: and this is the meaning of the French phrase idée fixe), even ‘enslavement’ through love (bar 85). So closely is the Fantastique adumbrated here that it is easy to forget how new, for the Berlioz known to us in 1828, this type of motivic manipulation was. The bass variants themselves could have derived operatically from works like Cherubini’s Médée, but neither Cherubini nor Méhul could have invented the chromatic variants.

Aria 2: The central section, and the Tancred / idée fixe theme, are in F major. The theme’s sequential rising phrases are now heard for the first time, forming the ‘consequent’ of the opening phrases. Erminia’s music, texted, represents her ‘evanescent plaint’; so the familiar music now performs a new function, even more complicated (narratively) than in the Fantastique. There, the audience interiorises the music and its virtual ‘narrator’, who invisibly ‘voices’ the story of the programme. In the cantata, which must be imagined to be acted in costume,44 the singer actualises music which in the opening sections was, by operatic convention, ‘unheard’ to her as music, but heard by us. (Otherwise, we might have to assume that Erminia was musically trained, etc.) This must be why, when she sings ‘J’exhale en vain’ etc., she first listens to the violins, then imitates them; it is as though she first hears an inner voice – the voice of nature, as the eighteenth century called it – then reproduces the sounds. By the same token, the inability of the voice line to cover all the violin notes seems psychologically right; and, in any case, the sheer power of Tancred’s image is by now more than she can easily withstand. It would in fact have been unconvincing had she enunciated her complaint to an exact form of the theme, because it remains her (our) imagined image of Tancred, which cannot be the same as a set piece. (In Lélio the composer could have had the same theme sung, somehow; but the actor-speaker is deliberately disallowed from singing.)

Finally: It is worth noting that the end of the idée fixe of 1830 is not present. Had it already been composed in or before 1828, it would just have been omitted as less than appropriate. Holoman reminds us that composer-candidates, like painter-candidates, sometimes practised for the Prix de Rome.45

Recitative 3: in this classic scena of vacillation, Berlioz brought back four bars of the Tancred / idée fixe music on solo flute (bar 205), in D, ‘en mourant’, pianissimo violins giving the barest support. Just before this, Erminia has considered the likelihood of her own death (dressed as Clorinda, she may be attacked: which indeed in Tasso she is). But care for him overrides care for herself. So she imagines Tancred’s imminent death from injuries received in the first duel with Argantes and, since she knows that she has the power to cure him (as will be seen later in the epic), this fantasy spurs her into action. The music therefore functions as a transformed reminiscence; an actively supposed memory, still perhaps ‘pursuing’ her, but in a newly varied guise; and because she is really on the point of making a decision, we might also claim that she calls up the image, as much as it calls on her.

VI: ‘Beethoven has turned his brain’: these words were supposedly uttered by Berlioz’s judges.46 Perhaps they are fictional, or else were chosen by a member of the Institut simply as a form of denigration. Yet they are mediated through Berlioz’s pen, and therefore register that he wanted it thought possible (by posterity) that he might already have been affected creatively by the auditory revolution of hearing the Third and the Fifth Symphonies. Although critics often mention the influence of Beethoven, the Fifth is less often evoked in this way than the others. Cairns mentions several detailed features in Berlioz relating to that work, including the idea that it is ‘conceived in response to a single governing thought’, though he cautions that its influence ‘will be more general than specific … not the style or process of composition so much as the principle.’47 Beethoven’s varied recollections and uses of his opening theme, and his recall of the third movement in the finale, however, may easily be paralleled with his admirer’s choice of a recurring and diversified idée fixe theme. Stricker, although without having time to analyse each different appearance of Berlioz’s cantata theme, can claim simply on internal evidence that ‘[It is] therefore already an idée fixe, with all the structural ambitions which that leitmotiv alla Berlioz entails.’48 The ingredients for the Fantastique were in some sense in place by 1828, although Berlioz had not yet had time to study the earlier work intensively. He had already shown in Herminie how his Tancred / idée fixe theme could assume a variety of conceptual and musical functions, stretching across three sections of his cantata. Its psychological ‘profile’ was ready, and its malleability proven, in both musical and other ways: Tancred’s music ‘pursues’, ‘tortures’, prompts Erminia into utterance and is called up in fantasy. Blend the symphonic unity of Beethoven’s Fifth with the techniques of Herminie and you have a recognizable matrix.

David Charlton


1. See relevant footnotes below. My thanks are offered to Hugh Macdonald and Julian Rushton for helpful readings of this paper.

2. They were: La Mort d’Orphée (1827), Herminie (1828), La Mort de Cléopâtre (1829); and Sardanapale (1830). These scores are edited by David Gilbert in NBE vol. 6: Berlioz, Prix de Rome Works (Kassel, etc., 1998). These all gave rise to various instances of later self-borrowing. Peter Bloom goes so far as to say Berlioz regarded them as bank ‘deposits’ from which he could later make regular ‘withdrawals’. Since the final cantata is partly lost, we cannot know all the uses that may have been made of it.

3. Correspondance générale [hereafter CG], ed. Pierre Citron et al., 8 vols (Paris, 1972–2002), I, p. 463.

4. ‘Sur Armide et Gluck’, Le Corsaire, 19 Dec. 1825, ed. in Critique musicale, ed. Anne Bongrain et al. (Paris, 1996–, 5 vols issued to date), I, pp. 9–11.

5. John E. Morby, The Wordsworth Handbook of Kings and Queens [originally issued as Dynasties of the World](Ware, 1994), pp. 171–72.

6. J. H. Wiffen (transl. and ed.), The Jerusalem Delivered of Torquato Tasso, 5th ed. (London, 1854), pp. 132–33.

7. Jerusalem delivered, p. 34.

8. Jerusalem delivered, p. 141.

9. Berlioz, Mémoires, transl. David Cairns (London, 1969), p. 115.

10. Peter Bloom, The Life of Berlioz (Cambridge, 1998), pp. 46–47.

11. Hugh Macdonald, Berlioz (The Master Musicians)(London, 1982), p. 83.

12. Jerusalem delivered, verse 93, p. 141.

13. David Cairns, Berlioz, vol. I: The Making of an Artist, 1803–32 (Berkeley, 2000), p. 287.

14. Julian Rushton, The Music of Berlioz (Oxford, 2001), p. 24.

15. Macdonald, Berlioz, p. 83.

16. ‘La fameuse Prière […] est, en revanche, d’un étrange hiératisme: la fermeture sur soi-même de chaque membre de phrase fait penser à un lento caractéristique de la même fixité, chez deux hommes que Berlioz n’aimait guère, Haendel et Haydn. […] mais, de phrase en phrase, le continuel renouvellement des timbres orchestraux finit par opérer une fois de plus de manière inimitable, en soumettant l’oreille à deux forces contraires: immobilisme et irisation’: Rémy Stricker, Berlioz dramaturge (n. p., 2003), p. 227.

17. ‘Lorsque ce thème vient se superposer finalement à la fuite au galop vers le silence, son inscription sur le rythme de chevauchée, impérativement bloqué sur pédale, est un trait de génie’: ibid., p. 227.

18. ‘J’évite en général comme la peste ces lieux communs que tous les compositeurs (excepté Weber et Beethoven) mettent à la fin de leurs morceaux; c’est une espèce de charlatanisme qui veut dire: ‘Préparez-vous à applaudir, ça va être fini;’ et rien à mes yeux n’est plus pitoyable que ces phrases banales et de convention qui font que toutes les musiques se ressemblent’: CG, vol. I, p. 192 (29 May 1828). Stricker also makes use of this passage, p. 225. Rushton, Music of Berlioz, p. 19, quotes the unexpected ending of Aria 2 of Herminie, taking it as an illustration of ‘choking with emotion’ after the high B flat. I would also see it as a particular solution to Berlioz’s ‘problem’ of finding endings, and here determined by word-painting, since the words ‘trancherait mes jours’ mean literally ‘would sever my connection with life’. A later explanation (Rushton, op. cit., p. 92), namely that Berlioz is here ending ‘with a fragment of recitative’, seems perhaps overdetermined.

19. Rushton, Music of Berlioz, p. 17.

20. Memoirs, p. 109. The Fifth was given on 5 April; other works performed in this season included the Violin Concerto, parts of the Missa solemnis and the Egmont overture.

21. Cairns, The Making of an Artist, pp. 279–81.

22. CG, I, pp. 197–201.

23. ‘[C]et abominable concours est pour moi de la dernière nécessité, puisqu’il donne de l’argent et qu’on ne peut rien faire sans ce vil métal’: CG, I, p. 203 (15 July 1829).

24. And ‘something not attempted in France before’: Cairns, The Making of an Artist, p. 260, from which the following citation is also made. Stricker explicitly endorses this: Berlioz dramaturge, p. 228: ‘Si Harriet est encore plus vraie que nature en Juliette ou en Ophélie, pourquoi pas déjà en Herminie?’

25. Hugh Macdonald, ‘Berlioz’s Lost Roméo et Juliette’, in Peter Bloom (ed.), Berlioz: Scenes from the Life and Work (Rochester, 2007), forthcoming. I am most grateful to Hugh for allowing me to see an advance copy of this important text.

26. This is the reason why the character of Faust never appears in Huit scènes: Goethe gave no lyrics to Faust to sing within the play. This simple point is sometimes missed.

27. Macdonald, ‘Berlioz’s Lost Roméo et Juliette’.

28. Rushton, The Music of Berlioz, p. 24.

29. Rushton, Music of Berlioz, p. 84.

30. Cairns, The Making of an Artist, p. 288.

31. Hugh Macdonald, Berlioz (Master Musicians) (London, 1982), p. 82.

32. Ibid., p. 122.

33. Berlioz, Mémoires, Chap. 23, transl. Cairns, p. 116. It became a ‘Chant sacré’ for chorus, first with piano (two versions: NBE vol. 14) and then with orchestra (NBE vol 12a) in the 1840s.

34. ‘of course I know nothing about music, but I’ll be damned if that "God of the Christians" of yours didn’t fair churn me up inside’: Mémoires, Chap. 23, transl. Cairns, p. 116.

35. Macdonald, Berlioz, p. 83.

36. ‘Im Thema hat die fortwährende Arbeit an den Francs-Juges Spuren hinterlassen: Als Ausdruck der Kampfeslust der Heldin bewahrt der Themenkopf (in seinen Dreiklangsbrechungen und in seinem Rhythmus) etwas von dem kriegerischen Gestus des Heldenthemas der Ouvertüre, der übrigens auch in der Ansprache des Soldatenführers zu Beginn des ersten Aktes der Oper auftaucht’: Oliver Vogel, Der romantische Weg im Frühwerk von Hector Berlioz (Beihefte zum Archiv für Musikwissenschaft, 53) (Stuttgart, 2003), p. 357.

37. The whole aria forms an analogy to the words of the composer cited earlier (‘la musique seule me fait supporter’): the incipit reads, ‘O seul bien qui me reste! ô ma céleste lyre!’ In the concert of 26 May, Berlioz’s Orphic identity already spoke covertly of his personal (emotional) situation.

38. With which Stricker would be in agreement: this theme is ‘[une] autre préfiguration de l’"idée fixe" de la Symphonie fantastique’: Berlioz dramaturge, p. 222.

39. Not, for me, the ‘figure of death’ identified by Frits Noske and suggested here by Kern Holoman in Berlioz (London, 1989), p. 79, by analogy with Spontini; Vogel (Der romantische Weg, pp. 350–51) too considers it as a death-motive; the interesting citation he makes, from the ‘Prière’ bars 146–47, is not decisive evidence: for one thing, the supposedly deathly rhythm is absent from Aria 2, where Erminia specifically sings about Tancrède running the danger of death.

40. Schriften zur Musik, ed. Friedrich Schnapp (Munich, 1977), p. 47; E. T. A. Hoffmann’s Musical Writings transl. Martyn Clarke, ed. D. Charlton (Cambridge, 1989), p. 248. Rushton elsewhere muses that the idée fixe of 1830 was ‘hardly what an audience accustomed to Beethoven would expect’, being ‘a fully cadenced melody’. But Beethoven’s longer symphonic melodies include the Prometheus theme in Op. 55

41. ‘Nach der Exposition des Vordersatzes bringt der entwickelte Nachsatz einen kontrastierenden Instrumentaleffekt. Im zweiten Anlauf (T. 20–55) führt Berlioz noch motivisch-thematische Arbeit am Thema vor, das in Abspaltungen des Hauptmotivs verklingt. Der aktuelle Massstab der in den Aufführungen der Société des Concerts vorgestellten deutschen Symphonik wird hier fühlbar’: Vogel, Der romantische Weg, p. 359.

42. As Edward T. Cone discussed in ‘Schumann Amplified: An Analysis’, in his edition of the Fantastic Symphony (Norton Critical Score)(London, 1971), pp. 249–77.

43. Rushton, Music of Berlioz, p. 84.

44. Berlioz’s articles on the Prix de Rome state this more than once: e.g. ‘toutes les classes réunies de l’Académie des Beaux-Arts […] entendent de nouveau, toujours au piano et sans action, les cantates dramatiques composées pour un acteur et un orchestre […]’: Revue et gazette musicale de Paris, 19 June 1836, ed. in Critique musicale vol. II ed. Yves Gérard et al. (Paris, 1998), p. 480.

45. Holoman, Berlioz, p. 52. Vogel’s study, for all its sensitive attention to the word-setting and meaning of this section of Aria 2 (Der romantische Weg, pp. 358–59) semantically deals with the 1828 theme as though it were simply identical to the 1830 idée fixe, in spite of the 1828 absence of the last nine bars of the symphony’s theme.

46. The Mémoires, Chap. 23, recounts the surreptitious report made of the judges’ meeting by the aged porter, Pingard. One musician reportedly said ‘All he cares about is that wild fellow Beethoven’: ‘Go and ask Cherubini […] Beethoven has turned his brain’: transl. Cairns, p. 117.

47. Cairns, The Making of an Artist, p. 266.

48. ‘Déjà idéé fixe donc, avec tout ce que comporte de visées structurantes ce leitmotiv alla Berlioz’: Berlioz dramaturge, p. 228.

* We are most grateful to Professor David Charlton for this invaluable contribution to our site.
David Charlton is Professor of Music History at Royal Holloway, University of London. He has been editorial board member of the New Berlioz Edition  since 1982. He edited vol.12b of the NBE, Hector Berlioz, Choral Works with Orchestra (II), which was published in 1993. He was one of the organisers of the "Interpreting Berlioz" Conference held in London on 15-17 November 2002 as part of Berlioz bicentenary celebrations. A selection of papers presented to the conference appears in The Musical Voyager: Berlioz in Europe, edited by David Charlton and Katharine Ellis (Peter Lang Press, 2007).

Related page on this site:

Berlioz Mémoires – Chapter 22

Berlioz in Italy – Rome

Berlioz Libretti – Herminie

Berlioz Discography – Herminie

The Hector Berlioz Website was created by Monir Tayeb and Michel Austin on 18 July 1997; this page created on 1 March 2007.

© 2007 David Charlton. All rights of reproduction reserved.

Back to Original Contributions page
Back to Home Page

Retour à la page Contributions Originales
Retour à la Page d’accueil