Hyperion Records

The only mystery which surrounds Liszt’s transcription of Harold in Italy is why, having produced it so promptly, he took so long to publish it. As with the Symphonie fantastique, it is also clear that Liszt’s version was made before Berlioz effected some slight alterations in the score prior to publication—as so often with Berlioz, there was a considerable delay between composition and performance and the appearance of the printed work, and Liszt’s version preserves the original text. Thus the viola is heard in several notes just before the final chorale in the Pilgrims’ March which no longer feature in the score, and Liszt has allowed the viola to participate in the Allegro assai sections of the Serenade with double-stopped chords in imitation of a piper’s drone. However, Berlioz’s brilliant second thought of foreshortening the repeated chords by half a bar at a time in the peroration of the first movement (four occasions) has been adopted in this performance in accordance with his published score.

Although Liszt’s partition could scarcely be called chamber music, it is undeniable that, without Berlioz’s orchestration, the viola part is heard to much greater advantage than is usually the case, and from time to time a real chamber music texture emerges, leaving one to regret that Liszt expressed himself relatively little in the medium, and then usually in transcription. Liszt seems to have thought of the piece as a piano transcription of the same stamp as that of the Symphonie fantastique, and the demands upon the pianist are similarly acute. (Liszt did make a solo piano version of the second movement of Harold, which is recorded in this series on Volume 5.) At any rate, Paganini, who declined to play Berlioz’s original because he felt that the viola had too little to do, might have felt less oppressed by Liszt’s version, at least in the first three movements. And, as with the Symphonie fantastique, Liszt is fully in tune with Berlioz’s ideas of programme (they were both completely enamoured of the works of Byron) and colour. Berlioz’s original is too well known to require further description, but it will be observed that Liszt was not tempted to alter the degree of participation of the viola in the final movement, even though, at the entry of the menacing trombones and Liszt’s extravagant tremolo which covers over half the keyboard in its attempt to recreate the violin parts of the original, the pianist could well have used some assistance.

from notes by Leslie Howard © 1993