If someone had told me a decade ago that I was going to produce the first (and, so far, only) published reconstruction of Chopin's posthumous Mazurka in F Minor that includes every uncanceled measure of the composer's sketch, I would have laughed. But apparently when a performance opportunity inspires a research obsession, unexpected things can happen.
I'm no piano virtuoso, but for some years I've performed chamber music and solo programs on a church recital series. As a student I learned Fauré's B-Minor Nocturne; Scriabin's Preludes, Op. 74; and Scott Joplin's Magnetic Rag—each of which is its respective composer's last completed composition for solo piano (and in the case of the latter two, the composer's last completed composition). Eventually I had the idea of performing an entire recital of such works, adding Schubert's Sonata in B-flat, D. 960; Schumann's Ghost Variations; Brahms's Klavierstücke, Op. 119; and the piece that is generally considered Chopin's last: the Mazurka in F Minor, Op. 68, No. 4.
Of these last completed piano works, the mazurka was the least finished. Chopin customarily composed at the piano, then notated a sketch, and lastly wrote a fair copy. But for the work published posthumously as Op. 68, No. 4, he only got as far as the sketch, a confusing one-page document that was discovered after his death in 1849.
Virtually all editions of the complete Chopin mazurkas include the same sixty-two-bar A–B–A version published in 1855 by former Chopin copyist Julian Fontana, based on a fair copy by cellist Auguste Franchomme, a close friend of the composer and dedicatee of his Cello Sonata. (The Henle edition exactly follows Franchomme's copy, which is more accurate in several respects than Fontana's published version.) But I was aware of a longer version from pianist Arthur Rubinstein's 1967 RCA recording of the complete Chopin mazurkas—a 1965 reconstruction by Polish pianist-musicologist Jan Ekier that adds a C section in F major. I purchased a copy of this A–B–A–C–A version and assumed I'd play it on my recital.
Looking up the work in Maurice Hinson's Guide to the Pianist's Repertoire, however, I learned of a more complete 1975 reconstruction by British pianist Ronald Smith as published by Hansen House. It turned out to be out of print, but I tracked down a copy via interlibrary loan. The C section in Smith's A–B–A–C–A version is thirtytwo bars, versus sixteen in Ekier's; in his foreword, Smith specifically criticizes Ekier's version, saying that it "takes a disastrous short-cut in the F major episode."
Wondering whether there had been any other published reconstructions, I of course did a Google search, which led me to "Chopin's Last Style," an article by the eminent Chopin authority Jeffrey Kallberg that appeared in the Summer 1985 issue of the Journal of the American Musicological Society and was republished with slight emendations as a chapter in his 1996 book Chopin at the Boundaries: Sex, History, and Musical Genre. Kallberg contends that the piece's form is actually A–B–A–C–A–B–A, which struck me as the solution to one riddle in the sketch— the two different versions of the left hand of the B section's last measure (the second version is printed as an ossia passage in the Ekier and Smith editions). It seems likely that one version of that measure is intended for the first B section, the other for the second B.
Kallberg also provides a detailed summary of the various attempts to reconstruct the piece, which in turn led me to two other versions: an edition by Polish composer-pianist Milosz Magin, published in 1983 by Editions Concertino of Paris, and British pianist John Vallier's 1986 edition (now out of print) of a reconstruction by British musicologist Arthur Hedley, who as the first modern scholar to view the sketch in 1951 had discovered the long-missing C section. The thirty-two-bar C section in Magin's A–B–A–C–A version is structurally identical to Smith's. The form of the Hedley-Vallier version is (as Kallberg proposed) A–B– A–C–A–B–A, with the aforementioned slight variation at the end of the second B; but its final A section is oddly truncated, and the thirty-two-bar C section includes material that is clearly crossed out in the sketch.
In "Chopin's Last Style," Kallberg states that "numerous attempts have been made to decipher the document, none entirely successful." Playing through the four published versions, I couldn't help drawing the same conclusion; I could not see performing any of the four verbatim. So I thought I'd put together a composite version, just for my own personal use, incorporating the most convincing readings from the various editions.
But how does one make those decisions? My own sense of musical style, and of Chopin's style in particular, provided a general basis. Ultimately, though, I had to study the sketch itself, a facsimile of which was printed in several of the published editions.
One of the first spots I examined was a segment that Kallberg cites as having been omitted in previous reconstructions (Ekier, Magin, and Smith all fail to include it): the bass-clef passage that Chopin labeled 3ci (an abbreviation for trzeci, Polish for "third"). Kallberg's article interprets this segment as the left hand for the transition to the second B section, and the Hedley- Vallier edition adopts that interpretation as well. I took one look at this passage in the sketch and—to my great surprise—drew a completely different conclusion.
The passage consists of a quarter note and four eighth notes, connected by an ascending slur (the only slur in the entire sketch) that continues up and to the right of the last note; the beam over the eighth notes also continues up and to the right of the last note. To me, this looked like shorthand for an ascending tonic arpeggio that continues upwards (despite the fact that the last notated pitch is a D-flat, which sounds wrong even if the passage is used as a transition to the second A; Kallberg now conjectures that Chopin intended this note to be a C). And the function of the passage seemed clear to me as well: a little extra flourish to distinguish the final iteration of A as the end of the piece. Indeed, the A section ends three different ways—first leading to B, later leading to C, and finally ending the piece—and Chopin specifically labels this ending the "third." (I find it highly unlikely that the passage represents the second A-to-B transition. My examination of all the Chopin sectional dances—the mazurkas, polonaises, and waltzes—reveals that in A–B– A–C–A–B–A rondos where the first and second B sections are in the same key, the second transition from A to B is always identical to the first.)
But the 3ci passage isn't the only sketch segment that is typically omitted. At the end of the C section, Chopin wrote one measure above the staff in (presumably) treble clef and one measure plus a quarter note below the staff in bass clef. Ekier, Smith, and Magin include the former but omit the latter; Hedley-Vallier includes the latter but omits the former. No previous reconstruction had included every measure of the sketch that Chopin had not crossed out. So I made up my mind that my version would be the first—and this required figuring how to incorporate both these segments at the end of the C section. Thinking more as a composer than as a musicologist, I came up with a solution that seemed to make musical sense.
Barely had I drawn these conclusions than I began to question how I could have solved riddles that had stumped a whole series of eminent Chopin specialists. So first I sent my reconstruction to an esteemed scholar who has dealt extensively with another composer's manuscripts. She was quite encouraging, as were the other editors and teachers I then asked to review my work. Eventually I got up the nerve to send my version to Jeffrey Kallberg himself, who responded, "You've solved a number of puzzling passages that have eluded other researchers, including myself (I'm thinking especially of your very sensible rendition of measure 95 [the end of the C section])." He also generously provided a great deal of critical feedback that led to numerous improvements in both my reconstruction and the accompanying editorial notes.
By this time, I realized that my version might well be of interest to other pianists and scholars, so I began submitting it for publication, and fortunately Richard Walters of Hal Leonard decided to issue it as part of that company's Schirmer Performance Editions series. It appeared in print in 2012—the year after I gave its premiere performance on my last-completed-pianoworks recital.
All music schools try to link performance and scholarship. But sometimes, by pure serendipity, one leads seamlessly to the other—and then back again.