Just as there is inconsistency between the title page and music head titles of the sonnets—in one place ‘di’ and the other ‘del’ Petrarca—the work commonly called the ‘Dante Sonata’ is described both as ‘une lecture de’ on the title page and ‘une lecture du’ at the head of the music and in the amended title on the manuscript. The original title of the piece was Paralipomènes à la Divina Commedia—Fantaisie symphonique pour piano, and the first version (which is in two parts) is probably what Liszt first played in 1839. A first layer of revision in the principal manuscript may well belong to the second projected title Prolégomènes (still in two parts), and Liszt seems to have performed a version of this work under the title Fantasia quasi sonata (Prolégomènes zu Dantes Göttlicher Comödie). A further, much more extensive layer of revision carries the final title and one-movement structure, but a good many final corrections and alterations were made at the proof stage to produce the present work.

The principal manuscript with its revisions in Liszt’s hand is actually in a copyist’s hand and contains several errors which went uncorrected by Liszt through the various stages of revision. Problem bars include: 65 (first left-hand group may be incorrect—the second left-hand chord should perhaps have a B flat instead of an A); 102 (second harmony should surely have E sharp—MS has a (redundant) natural sign); 255 (first left-hand chord should probably have an F sharp instead of an E); 262–3 (almost certainly B flats and hence E flat major—otherwise the augmented triad is the only such chord in the work in all versions, and it is a chord to which Liszt normally grants particular importance in a musical structure—furthermore, this theme is always extended elsewhere by common triads); and 297 (the right hand should certainly have G sharps on the third crotchet, as in the earlier version—the lack of them in the rewriting, which otherwise preserves exactly the same progression, is clearly a slip of the pen). (All but the first of these problem passages are rendered according to these observations in the present performance.)

Many commentators have essayed a description of the particular reading of Dante which Liszt has chosen to represent, although he himself gave no specific clues. (The case of the Dante Symphony is quite another matter: each movement represents Liszt’s reaction to Inferno and Purgatorio, with a hint of Paradiso in the concluding Magnificat, and the musical text is laid out upon occasion to fit various quotations of Dante’s work.) Clearly, the diabolus in musica tritone—heard at the outset, and at all the important structural junctions—suggests Inferno, and suggestions have been made concerning the Francesca da Rimini episode. But calling the reprise of what amounts to the second subject (10 bars of ethereal tremolo at bar 290) a representation of Paradiso as some commentators have done is surely wide of the mark, and the piece as a whole is much less celestial or purgatorial than it is relentlessly infernal. Formally, the structure is a much tighter sonata-form than the epithet Fantasia might suggest, and the musical form outweighs any attempt there might have been to convey a poetic narrative rather than just a general reaction to Dante’s work. (Howard)

Pf: Goran Filipec

свернуть