JAVS Spring 2013

“Berlioz and his Harold Symphony,” but also with his transcription of the work for viola and piano in 1837. Liszt’s viola part follows Berlioz’s original almost exactly, except in the third movement, “Serenade,” where the viola takes the clarinet parts as a drone for the opening thirty-one bars (ex. 1) and in the return of the same material in measures 136–66. It is implied in a letter from Berlioz that Liszt want ed to add more music for the violist to play. In their correspondence, however, Berlioz insisted that Liszt keep the viola part just as isolated from the piano as it is in the orchestral version, writing, “Don’t you think the part you give to the viola, being more extensive than the part in the score, changes the physiognomy of the work?” 22 Here again is further evidence for the physical separation between the viola and the orchestra, as well as the violist’s style of performance emulating the isolated character Harold, the observer, and not a chamber music part ner. As with many of Liszt’s orchestral transcriptions for piano, Harold in Italy is a handful for the pianist, and with so little for the violist to do, the work is slightly awkward in performance. If musicians choose to perform the viola and piano version, the Liszt transcription nonetheless seems more appropriate on a recital program than the piano reduction by Hugh Macdonald in the New Berlioz Edition . The purpose of this reduction, according to Macdonald, is to assist violists prepar ing for a performance with orchestra. 23 Also, since Harold in Italy is neither a concerto nor a sonata and does not aim to feature the viola part specifically as would be the case in those genres, Macdonald’s piano reduction does not serve its purpose in public performance. The Liszt transcription, on the other hand, exemplifies the contrast between the virtuosity of the orchestration and the simplicity and isolation of the solo viola. This contrast offers the audience a

more interesting and colorful hearing as well as a more faithful narrative of Berlioz’s literary intent associated with the work.

The Paganini Version

The newest issue to address regarding the viola solo of Harold in Italy is the twenty-two-measure episode of Paganini-style passagework for the viola from the original version, which was probably written for the violinist. 24 After Paganini abandoned the project, however, Berlioz revised this passage and omitted the virtuosic variation on the Harold theme, reducing it to a simpler statement in octaves. In his Memoirs , Berlioz explained his new approach to the work after Paganini’s refusal of the solo part: “Realizing that my scheme would never suit him, I set to work to carry it out with a different emphasis and without trou bling myself any more about how to show off the viola in a brilliant light.” 25 Since 2001, Macdonald’s piano reduction in the New Berlioz Edition has made this version available. This score and part format includes the “alternate version” of measures 73 to 94 of the first movement as an ossia passage in the viola part (ex. 2). 26 By printing this ver sion above the definitive viola part, the editor makes it a more accessible alternative for violists to play. Without knowing much about the historic back ground of the work, violists may choose this optional passage instead of the version Berlioz included in the final draft. The choice of performing the Paganini ver sion for these measures is problematic since it mixes a draft version with the definitive Berlioz text. In August 2011, a recording was released by violist David Aaron Carpenter with the Helsinki Philharmonic conducted by Vladimir Ashkenazy, which incorporates this alternative passage. While the Macdonald edition acknowledges the passage as

Example 2. Hector Berlioz, Harold in Italy (“Paganini version” passage), movt. I, mm. 92–94 (viola part).

V OLuME 29 NuMBER 1 29