It was not until much later in the twentieth century that anyone compared Busoni's version to the manuscript. In 1991, the Scottish pianist and writer
Kenneth Hamilton was the first to publish the results of such a comparison. In
The Opera - Fantasias and Transcriptions of Franz Liszt - A Critical Study, he writes: As the chief weakness of Liszt's Fantasia on Figaro and Don Juan is the introduction of the minuet from Don Juan into the middle of a piece which up to that point seemed to be a fine fantasia on motifs from Figaro alone, Busoni's solution was to cut the Gordian knot entirely and remove the minuet section, despite the fact that he could not excise the third minuet theme from the coda. By this means the piece gained greater structural coherence, although some fine music was inevitably lost. In addition to suggesting dynamic markings and marks of articulation, Busoni filled out some bars of the introduction (bars 28-37) by adding chromatic scales and tremolo. There is nothing in this which conflicts stylistically with Liszt's music, but the greater elaboration does tend to reduce the improvisatory effect important to this section. Soon after (bars 48 ff.), Busoni altered Liszt's figuration for what seems like no good reason other than his personal fancy [...] The task of providing a suitable transition between the variations of "Non più andrai" and the coda was carried out most dexterously by Busoni, as was his completion of the ending. Indeed, the unobtrusiveness of Busoni's hand to someone unfamiliar with the manuscript version is probably the main reason that most scholars took Busoni's title page at face value. There is no reason why both Busoni's and Liszt's original version of the fantasia should not find favour with pianists and audiences, the more satisfying structure of the one being offset by the fascinating minuet transcription of the other... Later, after learning of Hamilton's findings, the Australian-born
pianist, and notable scholar of Liszt's music,
Leslie Howard, attempted to reconstruct the work as he believed Liszt intended, recording, and in 1997 publishing, the newly-recast work. His intention was to "publish... the whole of Liszt's
Fantasy with an authenticity of text, supplying and clearly indicating the few bars... which are necessary to render the work performable." As reconstructed by Howard, the piece includes the music based on the dance scene from the Act I finale of
Don Giovanni. The dramatic opening is based on the music from
Figaro: a free paraphrase of "Non più andrai" followed by an arrangement of "Voi che sapete" in
A-flat major instead of Mozart's
B-flat major. This is the only appearance of Cherubino's music. Figaro's aria returns, initially in its original
C major, but is quickly varied both harmonically and pianistically. However, instead of proceeding to Mozart's
coda, the aria transitions to the dances from
Don Giovanni. In the opera, the dances are a
minuet in Time signature|, a country dance in , and a quick
waltz in . Liszt keeps the minuet in
F major, and combines it with the country dance in the same key. He does not add the waltz as Mozart does, but treats it separately, eventually combining it with the country dance and, "excellently", and the final 15 bars to complete the piece. He has also added expression marks, including "dynamics, marks of articulation, numbers indicating numerical groupings, m.d. and m.s. marks, pedallings, tempo and other textual directions. The evident deficiencies of the manuscript (time signatures, accidentals, stems, beams, rest signs, fermatas) have been tacitly supplied throughout." He uses square brackets to differentiate his additions or suggestions from the original text. The completion was published by Editio Musica Budapest (EMB) in 1997. ==Performances==