I wrote earlier in these pages about Schumann’s Fourth Symphony, which before it was revised had a successful but heavily criticized premiere. The original version has now been recorded and published, and can be appraised alongside the revised version, which has been performed exclusively for 182 years. I now move on to extend my remarks after Jeremy Denk’s fine recital on December 9th.
Critical opinion has never weighed in very much about the various revisions to Schumann’s major compositions for piano, and this is as regrettable as it is incomprehensible. I think especially of the Davidsbündlertänze,, op. 6; Symphonic Etudes in the form of Variations, op. 13; Kreisleriana, op. 16; and most recently the three-movement-long Phantasie, op. 17. All of these have been available in many publications for a hundred years or more. On my own shelves are the Kalmus reprint of the Breitkopf edition in seven volumes of Schumann’s “Complete Works for solo piano…edited according to manuscripts and from her personal recollections by Clara Schumann,” a “second edition,” which has become standard for more than a century and includes the texts of the first editions of Schumann’s lifetime in ossias and smaller type. The Symphonic Etudes on my mother’s piano were a Schirmer edition by Max Vogrich; five extra variations, excluded by Schumann from the initial edition but published separately after his death, were added in this printing as an Appendix.
Many performers today play the second edition (12 etudes, some of which aren’t variations) and add the five posthumously published variations interspersed, making 17 in all, including a rondo Finale. Differences between first and second editions are mostly minor details in the variations, and in the second-edition Finale, the three A sections are all essentially identical except for the Coda; but in the first edition, the A` section is entirely different and thoroughly interesting, and this version of the Finale I have yet to hear in public performance. Why is the less-interesting version the one always performed, and, equally to the point, why does it exist at all? I have long heard it said that Schumann made the changes in the second edition at Clara’s behest. If so, her reasons would surely not have been simplifications of difficult passages; plenty of places in the Etudes steer close to the limits of playability but remain in both editions.Differences in the two editions of Kreisleriana, op. 16, have to do with a few cadential measures, repetition of formal sections, and, in the second piece, a reordering of certain phrase groups. These are note in small notes and footnotes in Clara’s edition. Four of the eight pieces in the set have no differences at all between editions.
Clara’s second edition of the Davidsbündlertänze, op. 6, is a fully re-engraved score of the entire work; but the only immediately obvious differences from the first edition are changes in tempi and omission of the quaint rubrics of Florestan and Eusebius, who individually initial each of the 18 pieces. No. 3 in the first edition is marked Etwas hahnbüchen, which my oldest dictionary renders as “somewhat clumsy” (Hutcheson gives it as “rather cockeyed”); the second edition changes this to Mit Humor.
Charles Rosen first brought to public attention the changes that Schumann made in the final bars of the Phantasie, op. 17, just before first publication in 1837. Here is the original ending as Rosen gives it in The Romantic Generation, p. 110, and as he recorded it (at right). One can listen to a recent recording HERE.
Everyone who knows the standard version of the ending of the third movement will recognize that these Adagio measures recapitulate the ending of the first movement, with its melodic quotation from Beethoven’s An die ferne Geliebte, but with one subtle, telling difference: on the downbeat of mm. 148 and 151 Schumann has changed the strongly cadential tonic six-four (I64) to the precadential supertonic seventh (ii7). The standard ending omits these fifteen recollective measures entirely, substituting downward C major and F major — “three bars of perfunctory arpeggios,” Rosen calls them — to the final C major bell chords. “If an editor had made this change, we would call him a vandal.” But an editor didn’t make this change; Schumann the composer made it, whether at Clara’s or anyone else’s suggestion or not, we shall never know. Nor did Schumann make the change after publishing a “first version.”
As a member of the Romantic Generation myself I beg to disagree with the Fassung letzter Hand, in the same way that I prefer all the earlier versions, the first editions, of the Schumann piano works discussed above. (I have not yet made up my mind about the Fourth Symphony, other than to reinforce a basic impression that both versions, 1841 and 1851, have major defects.)
It is well worth subjecting the omitted Phantasie measures to a close examination. As Jeremy Denk pointed out in his brief remarks, the first movement (“Passionate and fantastic”) clearly states its most important thematic material only on its last page, after more than 11 stormy and very diverse developments and notable dissonances. That thematic material is literally identical to mm. 142-156 in the example above, with the exception of tonic six-fours (as in 143 and 145) at mm. 148 and 151. The supertonic sevenths (ii7) of the omitted version, each changing one note in the melody, add an instability that is more than merely charming: they are a deflection of the “distant beloved.” The tonic note in the upper part is no longer a stabilizing tonic root, but a psychological dissonance, and it dissolves to an F by a non-resolving downward leap, which requires absorption into the dominant seventh that follows it. The lack of assurance that this implies was apparently too much for Schumann. It shouldn’t be too much for his legacy. I will continue to urge all pianists to restore the original ending. .