This page is also available in French
See also Texts and Documents; Berlioz and his music: self-borrowings
There is a striking contrast between the history of the composition of the Symphonie fantastique and that of Berlioz’s second symphony Harold en Italie. In the case of the Symphonie fantastique, Berlioz had from the outset a clear conception of the character of the work: it was going to be a large-scale symphonic work of a novel kind, and it was inspired by his passion for the Irish actress Harriet Smithson. In the case of his second symphony, it came into being almost by accident: Berlioz had not planned initially to write a new symphony, but the work was begun and then evolved as a result of a commission by the violin virtuoso Paganini for a composition that would give a prominent part to a solo viola which he would play himself. The evidence of Berlioz’s writings (his Memoirs and his correspondence) shows that at the outset he had no preconceived idea as to how he would fulfill Paganini’s commission, and the project developed in unexpected ways until it reached its final form in the symphony Harold en Italie.
Berlioz noted that during his time in Italy, despite all the leisure he had, he found it difficult to compose music, and the works he produced were few in number and on a limited scale (Memoirs, chapter 39). During his stay he did conceive some large-scale projects for the future. The first started to take shape in Berlioz’s mind in April 1831 while in Florence, and he revealed it to his friend Humbert Ferrand once back in Rome in July: it was for an apocalyptic oratorio entitled Le Dernier jour du monde (The Last Day of the World), a project which looks forward to the future Requiem of 1837. Another project emerged at the very end of Berlioz’s stay in Italy, in May 1832 on his way back to France: while crossing the plains of Lombardy he passed the site of Napoleon’s battle against the Austrian army near the bridge of Lodi (10 May 1796), which suggested to him the idea of a military symphony on the return of the French army from Italy; this looks forward to the Symphonie funèbre et triomphale of 1840.
Neither of these two projects had any specifically Italian association. On the other hand, Berlioz’s Italian stay did have a deep and lasting impact on his subsequent musical output. The impressions he derived from his wanderings in the country were to permeate many of his later compositions, down to his last major work the opera Béatrice et Bénédict of 1862. But this only started to happen with some delay, as though Berlioz needed time and a period of reflection to be able to distance himself from his Italian experiences and absorb them.
To summarise the account of the Memoirs (chapter 45), the great violin virtuoso Paganini, who had recently become a dedicated convert to Berlioz’s music, commissioned from him in early 1834 a work for viola and orchestra, which was designed to show off to advantage a Stradivarius viola in his collection. Berlioz sketched out a work which gave due prominence both to the orchestra and to the solo viola, but Paganini seemed disappointed with the result, as he was expecting a more developed part for his own instrument. He soon left for Italy, and Berlioz then worked out his own plan; the ultimate result was the symphony which, in a deliberate reference to Byron’s Childe Harold, was eventually given the title Harold en Italie (the full text of Berlioz’s account is given on a separate page). This account compresses the perspective and omits some interesting detail, and it needs to be supplemented by the evidence of the composer’s correspondence, as well as that of the contemporary Parisian press.
The first mention of Paganini’s commission is found in a letter of Berlioz to his friend the critic and writer Joseph d’Ortigue, dated 24 January 1834 (CG no. 378):
You know that I am writing a work for chorus, orchestra and principal viola for Paganini. He came himself to ask me for it a few days ago. Could you get this announced in four lines in the album of the Revue de Paris? Le Rénovateur has announced it and I went today to ask the same favour from M. de Briant at la Quatidienne; he was not there.
Presumably d’Ortigue had seen the announcement in Le Rénovateur, which had clearly been prompted by Berlioz. An identically worded announcement appeared on 26 January in the Gazette musicale de Paris (cited by Julien Tiersot in Le Ménestrel, 17 July 1904, p. 227):
Paganini, whose health is improving by the day, has just asked M. Berlioz for a new composition of the same kind as the Symphonie fantastique, which the celebrated virtuoso intends to perform on his return from England.
This work will be called: The last moments of Mary Stuart, a dramatic fantasia for orchestra, chorus and principal viola. Paganini will perform the viola part for the first time in public.
As can be seen, there are two details that are not mentioned in the account of the Memoirs; first, in Berlioz’s original plan there was to be a chorus in addition to the orchestra and the solo viola; and second, the work was to be concerned with the death of Mary Stewart. Neither had been mentioned by Paganini in his commission, and this must have been Berlioz’s own idea. The subject of Mary Stewart is a surprising inclusion; it had not apparently figured in Berlioz’s projects previously. Nothing further is known of the project nor of any music that may have been written for it, and Berlioz never returned to the subject.
The next relevant letter, to his friend and confidant Humbert Ferrand, comes nearly two months later, and shows how Berlioz’s new plan was evolving. The work is now called a symphony with principal viola, though Berlioz does not yet make clear that it was specifically related to his Italian experiences (CG no. 384, 9 March 1834):
I am completing the symphony with principal viola which Paganini asked me to write. I was intending for it to have only two parts, but a third occurred to me, then a fourth; but I hope to stop there. I still need a solid month’s work.
Berlioz was kept busy with writing articles and reviews for newspapers. and the composition of the work took longer than expected, as the next letter to Ferrand shows (CG no. 398, from Montmartre; 15 or 16 May 1834). From its reference to the title of the second movement the letter reveals indirectly the Italian character of the work:
I have completed the first three parts of my new symphony with solo viola; I will now set out to complete the fourth. I believe it will be good and in particular strikingly picturesque in character. I have in mind to dedicate it to one of my friends you know well, M. Humbert Ferrand, if he will allow me. It has a March of pilgrims singing the evening prayer, which I hope will be famous in December. I do not know when this huge work will be published; at any rate, please see to it that M. Ferrand grants his authorisation.
The work had not yet received its title, as emerges from another letter to d’Ortigue, dated 31 May 1834 (CG no. 399, from Montmartre):
At home I never stop writing, whether for these rascals of the newspapers, or to finish my symphony which will be born and baptised before long.
The work was eventually completed by the end of June, as another letter to Ferrand shows (CG no. 408, 31 August; cited in the page on Montmartre). In this letter the symphony is given for the first time its name Harold.
With both the Symphonie fantastique and its sequel the Melologue, Berlioz freely admitted that for the new works he had used music composed earlier. This was of course perfectly legitimate. Writing about his early opera Les Francs-Juges Berlioz comments (Memoirs, chapter 11): ‘I have used there and there the best ideas of the opera by developing them in later compositions; the rest will probably get the same treatment, or it will be burnt’. Elsewhere in the Memoirs he makes a brief and dismissive mention of his overture Rob Roy and states that he burnt it immediately after hearing it performed at the Conservatoire (chapter 39). A copy of the overture, sent by Berlioz to the Conservatoire, has survived, and this has revealed what otherwise would be unknown: a substantial part of the first movement of the Harold symphony re-uses material from that overture, a work which did not have any Italian connotations.
This applies to the second subject of the movement (bar 167 onwards; compare Rob Roy bar 170 onwards), and especially to what can be called the Harold theme, presented first in the minor by the wind (bar 13 onwards), then in the major by the solo viola (bar 38 onwards; compare Rob Roy bar 275 onwards). In the overture the theme is presented by the cor anglais; in the symphony it is now assigned to the solo viola. The theme is, for Berlioz, unusually regular in shape – two symmetrical four bar phrases, each of which consists of two bars of a falling interval followed by two bars of a rising then falling arpeggio. It is also purely diatonic and free from chromaticism. It will be seen that a large part of the thematic material of the symphony is derived from elements of that theme, the falling intervals and the rising arpeggio, either singly or in combination. This applies to the first and second subjects of the first movement, the serenade of the third movement, and the beginning of the main theme of the last. Another part of the thematic material derives from the opening bars of the symphony, the chromatic passage in semiquavers in the lower strings (from bar 1), and the descending chromatic phrase in the bassoon (bar 3). For example, the main viola theme of the allegro (bar 130 and following) combines first the rising arpeggio figure and then the two chromatic elements of the opening of the symphony.
As with the Symphonie fantastique Berlioz deliberately delayed the publication of Harold en Italie till after he had been able to conduct himself a number of performances at home and abroad (see e.g. CG no. 741 in January 1841). In the event the symphony only appeared in 1848, even later than the Symphonie fantastique. It was dedicated to Humbert Ferrand, as Berlioz had decided from an early date (CG no. 398). Hence when writing to Ferrand about the symphony, Berlioz would refer to it as ‘our symphony’ (see for example CG no. 3244). The autograph score of the symphony was given by Berlioz to his friend Auguste Morel, who then bequeathed it to his favourite pupil Alexis Rostand, who in turn bequeathed it at his death in 1919 to the Conservatoire.
Unlike the Symphonie fantastique, Harold en Italie does not have (and does not need) a programme. Instead Berlioz merely gave titles to its four movements to indicate their character:
1. Harold in the mountains. Scenes of melancholy, happiness and joy.
2. March of the pilgrims singing the evening prayer.
3. Serenade of a mountaineer from the Abruzzi to his mistress.
4. Orgy of brigands. Memories of earlier scenes.
Of Berlioz’s four symphonies the first two are the most directly comparable. They are both purely instrumental without any use of voices, and both are based on the classical four-movement form of the symphony (the fourth and fifth movements of the Symphonie fantastique may be thought of as constituting a double finale). The very opening of Harold en Italie – the theme in the cellos and basses, with its alternating semitone interval (G, A flat, G) – carries a probable echo of the start of the Symphonie fantastique (bar 3). Berlioz’s own account of the origins of Harold en Italie makes a direct comparison between the two symphonies (Memoirs, chapter 45):
My intention was to write a series of orchestral scenes, in which the solo viola would be involved as a more or less active participant while retaining its own character. By placing it among the poetic memories formed from my wanderings in the Abruzzi, I wanted to make the viola a kind of melancholy dreamer in the manner of Byron’s Childe-Harold. Hence the title of the symphony: Harold en Italie. As in the Symphonie fantastique, a principal theme (the viola’s opening melody), is reproduced throughout the work. The difference is that whereas in the Symphonie fantastique the idée fixe keeps obtruding like an impassioned obsession on scenes that are alien to it and deflects their course, Harold’s melody is superimposed on the other orchestral voices, and contrasts with them in tempo and character without interrupting their development.
The comparison might be taken further. Harold en Italie is one of Berlioz’s most relaxed and poetic works, very different in atmosphere from the Symphonie fantastique of 1830. It conveys a new feeling of nostalgia and looks back to happy memories. But Harold’s melancholy is free from anguish and kept within limits: unlike Faust in La Damnation, Harold does not contemplate suicide. The brooding, chromatic music of the opening bars in G minor soon gives way to diatonic music in G major with the entry of the viola accompanied by the harp. Harold’s theme is singularly placid and almost free from emotion. As noted above, the theme is, for Berlioz, unusually regular in shape, made up of two symmetrical four-bar phrases. It is also totally free from chromaticism, unlike the idée fixe of the Fantastique. After its first statement in the minor at the start of the symphony (bar 13 and following), it soon moves to the major (bar 38 and following), and subsequently never returns to a minor key.
The prevailing mood of the symphony’s first three movements is joyful. The second movement regularly proved to be popular with audiences, as Berlioz had anticipated (CG no. 398), and after the first performance he remarked that ‘it can now aspire to be the (religious and gentle) counterpart to the Marche au supplice’ (CG no. 416). The third movement, entitled ‘Serenade of a mountaineer of the Abruzzi to his mistress’ contains a direct echo of Berlioz’s Italian experiences: it imitates the music-making of the ‘pifferari’ which Berlioz had admired in Rome and in the mountains (cf. also the first of the three pieces for Alexander’s melodium).
After the first three movements Berlioz evidently felt that the symphony needed a finale that would provide a contrast with them: the Orgy of brigands. During his trip to Italy, Berlioz had fantasised about the lifestyle of brigands, whom he idealised as men who had liberated themselves from the shackles of society. He gave expression to these ideas in one of the monologues of The Return to Life and the Brigands’ Song which follows, though in his travels in Italy he had not actually come across real brigands. In Harold en Italie the concluding Orgy of Brigands is the counterpart to the Witches’ Sabbath of the Symphonie fantastique. But it has none of the nightmarish quality of that movement. In relation to a performance he gave at Loewenberg in April 1863 Berlioz says he conducted this movement ‘in his own style, with fury; I was gnashing my teeth’. Yet the grinding dissonances and dislocated rhythms of this energetic movement are counterbalanced by passages of great delicacy (bars 231-268, 394-440). Berlioz’s brigands are human beings, not supernatural monsters. There are also nostalgic reminiscences from the earlier movements at the start (bars 12-98) and near the end of the movement (bars 464-500), a device adapted from the finale of Beethoven’s 9th Symphony. For all their iconoclastic ferocity the brigands seem to be enjoying themselves, and the end of the movement sounds almost triumphant. But the finale of Harold is musically perhaps less compelling than that of the Fantastique. The latter develops a relentless progression from one section to the next, whereas in the finale of Harold the main section of the movement (bars 118-247), is simply repeated almost note for note (bars 280-410). The movement concludes with an extended coda, with a last reminiscence of the pilgrim’s march and the final return of the brigands’ music (bar 411 to the end).
The use of a solo instrument to represent a person was an innovation on the part of Berlioz, in response to Paganini’s commission of a work that would give a prominent part to the viola. It may also have been suggested by the frequent use of a solo instrument in Weber’s operas (Der Freischütz, Euryanthe) to accompany a characteristic aria by one of the protagonists. The idea was later followed by Rimsky-Korsakov (Scheherazade) and Richard Strauss (Don Quixote, Ein Heldenleben). Berlioz had a special fondness for the viola and made frequent use of its special tonal qualities in many of his works – he had already used a solo viola to accompany Marguerite’s ballad of the King of Thule in his Huit Scènes de Faust of 1828-9, in the same key of G major (H 33, later incorporated in the Damnation of Faust, but transposed down to F major). Harold en Italie is of course not a viola concerto – hence the initial dismay of Paganini when Berlioz showed him the work (cf. CG no. 408). The viola part is free from any element of technical display for its own sake, which Berlioz always avoided in his instrumental writing, even in such a work as his Rêverie et caprice for violin and orchestra. The work is thus hardly a fusion of concerto and symphony, but rather one of chamber and symphonic music. This can be seen in a number of very lightly scored passages throughout the work, notably in the slow introduction to the first movement (bar 38 and following) and near the end of the finale (bar 473 and following).
Harold en Italie received its first performance in Paris on 23 November 1834; it was performed twice again that same year, and there were another three performances of the complete work the following year 1835. The first three performances were conducteed by Narcisse Girard, who had stepped in to take the place of Habeneck who was no longer willing to conduct for Berlioz. The choice proved to be unfortunate, as Berlioz relates in his Memoirs (chapter 45): Girard was less competent than Habeneck, and made repeated mistakes in his conducting of the work; this convinced Berlioz that he could not entrust his music to any conductor. The composition of Harold en Italie thus had the unexpected result that it forced Berlioz to take up conducting himself, which he did with conspicuous success: within a few years he had become probably the finest conductor of his age.
After 1835 the symphony received a number of performances in Paris, sometimes of the complete work (1836, 1838, 1840, 1842, 1844), sometimes of individual movements only (1839, 1845, 1850; the second movement was a frequent favourite). After 1850 the symphony disappeared altogether from concert programmes in Paris, but it had made its mark. The performance on 16 December 1838 was a momentous occasion: it was then that Paganini heard for the first time, under Berlioz’s direction, the symphony that he had instigated. He responded with a public accolade for the composer, hailed him in a famous letter as the successor of Beethoven, and presented him with a cheque for 20,000 francs which enabled Berlioz to compose the following year Roméo et Juliette (Memoirs, chapter 49; CG no. 602).
The popularity of Harold en Italie in Paris may not have lived up to Berlioz’s expectations, but in his travels abroad the work formed one of the staple items in his repertoire. In his first trip to Germany of 1842-43 it figured in concerts in a whole series of cities: Brussels, Stuttgart, Hechingen, Mannheim, Weimar, Brunswick, Hamburg, Berlin (twice), Hanover, and Darmstadt. In 1845 he performed it in France outside Paris in Marseille and Lyon. In his second tour of Germany and central Europe of 1845-46 he conducted it in Vienna, Pest, Breslau, Prague, and Brunswick. In 1853 he conducted it again in Frankfurt, Brunswick, Bremen, and Leipzig, and later in Baden-Baden (1861), Loewenberg (1863) and Cologne (1867). He had occasion to perform it several times in his visits to London (1848, 1853 and 1855). It figured in both his trips to Russia, in 1847 (St Petersburg, Riga), and again in 1868 (Moscow, St Petersburg). The performance in St Petersburg on 8 February 1868 was in fact the last work of the last concert he ever conducted.
The symphony also travelled abroad in Berlioz’s lifetime, even without his advocacy. One example was a number of performances of the work in New York in 1863 then again in 1866. For the first American performance of Harold en Italie on 9 May 1863 the conductor Theodore Thomas prepared a very careful concert programme, an image of which is reproduced on this site together with a transcription of the text. When news of these performances reached Berlioz he was delighted (CG nos. 2840, 2856, 3076, 3244).
A few technical points:
First movement: Berlioz’s metronome mark for the slow introduction is quaver = 76. As pointed out by Hugh Macdonald (Berlioz Studies, ed. Peter Bloom [Cambridge University Press 1992], p. 23-4) this speed works for the opening pages but seems slow with the entry of the solo viola in bar 38. In this version the tempo has been very gradually increased from bar 13 to reach quaver = 80 at bar 38.
Second movement: the arpeggios in the solo viola part in the middle section of the movement (bars 169-247) have been written out in full, to produce the required effect. No attempt has been made to reproduce the special sul ponticello sound of the viola in this passage.
Third movement: this movement is written throughout in 6/8 time, but with two different tempi, the first of which (Allegro assai, dotted crotchet = 138) is double of the second (Allegretto, dotted crotchet = 69). At the end of the movement Berlioz superimposes the two. The tempi given by Berlioz are somewhat faster than what is often heard in performance. To enable the listener to judge, the movement is presented here in two versions, the first with Berlioz’s original tempi, the second with somewhat slower tempi (dotted crotchet = 126 and dotted crotchet = 63).
Fourth movement: (1) To obtain the correct note values on playback it has been necessary to notate several passages in full and not in abbreviated form (triplets or sextuplets in bars 38-40 [violas], 107-9 [violins 1 & 2], 175-6 [strings], 278-9 [strings], 338-9 [strings], 449-52 [1st violins]). (2) Berlioz gives no tempo indications or metronome marks for the concluding part of the movement after bar 449. The present version follows current performing practice in slowing down for the return of the pilgrims’ march and then returning to the original tempo of the orgy (bars 464-505). A slight quickening of the tempo is also applied from bar 514 onwards up to minim = 112, without which the music risks dragging.
in Italy: 1st movement (duration 15'14")
— Score in large format
(file created on 28.05.2000; revised 5.12.2001)
— Score in pdf format
in Italy: 2nd movement (duration 7'2")
— Score in large format
(file created on 27.07.2000; revised 23.12.2001)
— Score in pdf format
Harold in Italy: 3rd
movement (duration 5'17" and 5'46"):
(a) Version with Berlioz’s original tempi
— Score in large format
— Score in pdf format
(b) Version with slower tempi
— Score in large format
(files created on 12.11.2000)
in Italy: 4th movement (duration 11'54")
— Score in large format
(file created on 1.1.2001)
— Score in pdf format
© Michel Austin for all scores and text on this page.
This page revised and enlarged on 1 December 2021.
Back to Berlioz Music Scores