IN: News & Features

Chopin: A Progressive in Tonality


Though standard harmony textbooks don’t mention it, you can find “progressive tonality” on Wikipedia; I think it’s a misnomer, but the term is often heard in loose contexts, and there’s a basic definition that seems useful enough. “Progressive tonality” was apparently first described by Dika Newlin in her book, “Bruckner, Mahler, Schoenberg,” which arose out of her Ph.D. dissertation in the 1940s. (I knew Dika Newlin slightly from AMS meetings. She was distinctly an eccentric personality, if that’s the right description; but she was a good musician, a loyal follower and chronicler of Schoenberg, and her book, though dated, is still usefully readable today.)

Chopin’s F Major Ballade, with substantial sections oscillating between F major and A minor, and ultimately ending in A minor, is cited as an example; tonally speaking, it would seem to dislodge the idea of closure, i.e., that a piece of music should end in the same key in which it began. Yet most of us are open-minded enough not to regard Chopin’s different ending as a defect, one that we can actually hear. 

Chopin’s Scherzo No. 2, Op. 31, offers “progressive tonality” of a different kind. The first phrase of the piece is all B-flat minor. The second phrase, mm. 9-22, is a paradigm of the problem that isn’t a problem. The upward four-note motto, recurring many times throughout the work, is clearly in B-flat minor. But bars 13-17 are D-flat major, the first appearance of that relative-major key in the work; its ff A flat in bar 13 descends from the corresponding Bflat in bar 5 of the previous phrase (not shown). The cadence on D flat in bar 7 is immediately contradicted by the dominant of B-flat minor (bars 18-22). B-flat minor and D flat major are constantly referenced back-and-forth throughout the entire Scherzo, with D-flat major eventually predominating and finally concluding triumphantly. This Scherzo has always been published as “in B-flat minor”, but Heinrich Schenker’s analysis in Der freie Satz specifically identifies it “in D-flat major.”

That there thus seems to be a dispute about the principal tonality tells you something about what it means for a piece of music to be “in” a particular key. Short pieces, such as any of Schubert’s sixteen-bar waltzes, are usually “in” a single key. A slightly longer piece, like the Minuet from Eine kleine Nachtmusik, can be “in” one key, be followed by a Trio section in another key, and in the Da capo return to the first key (in this instance, G major – D major – G major). Any greater length, and change of key surely becomes an aesthetic necessity, and there’s no limit to what can happen to the key then. (Take a look at Beethoven’s Two Preludes, op. 39, for piano or organ, “through all the major keys.” At one point you find the key signature changing in every bar, all the way around the circle of fifths, twice. The “or organ” takes on a significance, if you consider these pieces to be used in church, to get from one key to another. But both these curious pieces are considered “in C major”, as a whole. Then think of a much larger work, such as Meistersinger or Falstaff, each opera beginning and ending in C major, with umpteen other keys in between beginning and end, and ask what it means to say that one or the other might be “in C major.”)

A particular key is established for a certain time by the way the music defines it. This normally requires a cadence from dominant to tonic — establishment, and confirmation, made stronger by the presence of the tonic note in the upper part of the tonic triad, by rhythmic strength, by repetition, and by other factors, and of which can be dislodged, diverted, or dissimulated — but that’s what music is supposed to do.

In Chopin’s Scherzo the tonal strength is bifocal — B-flat minor at one moment and D-flat major the next, constantly shifting not just back and forth but modulating to a multitude of other keys as well. In the opening section (mm. 1-132) and its slightly varied repetition (133-264), the initiating Bflat minor motto is repeated eight times, but D-flat major prevails overall. The Trio section that follows (Sostenuto, mm. 265-365) likewise has a repetition (366-467); this is in A major, modulating to E major (leggiero), but with substantial emphasis on C-sharp minor in between, especially gravitating melodically to the C sharp pitch itself. Some of that gravitation is even stressed in the notation, to make sure that the pianist doesn’t miss the point. (285)

Following the repeated Trio section, Chopin decides to include a Development of the Trio themes, a wild ride through successive keys, including chromatic sequences, eventually arriving back at B-flat minor (sempre con fuoco, m. 544). This goes on all the way to m. 583, with plenty of dominant of B-flat minor — a counterbalance, in timing, to the D-flat major of the opening sections. The da capo begins at mm. 584-8, sotto voce, corresponding to mm. 1-5, and so on to m. 716, where the Coda begins with a sudden, painful interruption in A major — the key that began the Trio section (we might have forgotten about it from long ago):

We do glance at the A in the very last bars, which give a parting shot at B-flat minor in the form of its dominant.
If there is a progressive lesson in this work, which is still a favorite for its pianistic brilliance, it is that the tonal cosmos is all-encompassing. B-flat minor and D-flat major are closely related; A major and E major are distantly related to B flat and D flat, but close to each other, and C-sharp minor forms the connector; the synthesis is tonal grandeur, and Chopin leaves no doubt. 








Mark DeVoto, musicologist and composer, is an expert in Alban Berg, also Ravel and Debussy. A graduate of Harvard College (1961) and Princeton (PhD, 1967), he has published extensively on these composers and many music subjects, most notably, harmony.


8 Comments [leave a civil comment (others will be removed) and please disclose relevant affiliations]

  1. Quite fascinating. I am reminded of Charles Ives’s remark that a piece doesn’t need to end in the same key as the one in which it began any more than you are obliged to die in the same house (or, nowadays, hospital) as the one in which you were born.

    But if Schenker is right and op. 31 is “really” in D-flat, then is Mahler’s Fifth Symphony “really” in D major? And we can come back to one of my favorite questions, is a classic ragtime a piece that begins in the tonic and ends in the subdominant, or one that begins in the dominant and ends in the tonic?

    Comment by Vance Koven — November 27, 2020 at 6:20 pm

  2. The song “Pirate Jenny” from “Die Dreigroschen Oper” by Bertolt Brecht and Kurt Weill starts in C minor and ends in F#, so at the end of every verse there is very jarring truck driver gear change back to the original key.

    Of course there are many pop songs that use a more conventional truck driver gear change to go up a half or whole step to add excitement, and many church organists do it for the last verse of the hymn.

    Comment by Mark Lutton — November 27, 2020 at 7:58 pm

  3. The following is surely known to the harmonically savvy here, but I had an interesting exchange some years ago concerning Bach’s ‘Keyboard Workout’, which spanned two decades, four volumes, and all styles, aimed at ‘connoisseurs seeking spiritual renewal’. The earlier Volume I comprises galant partitas, II notionally French- and Italian-influenced works, IV the Goldbergs. In vol III the E-flat Prelude and Fugue surround a German Mass, 25 exemplary chorales and other pieces, the entire volume organized along the lines of both the Lutheran liturgy and a Bach organ concert (the prelude and fugue enclose chorales).

    The first three chorales are Kyrie hymns traditionally arranged, the tune moving from soprano through tenor to bass, with full organ in the third. S.669 and S.670 are old-style, S.671 much grander, tune in the pedal.

    What vexed me was that the three chorales are all in three flats, like S.552, but do not sound in any particular key, certainly not E-flat, nor, as even some references have it, in C minor.

    So I asked a good acquaintance, organist and professor Stephen Roberts (then at Western Connecticut State), who gave me a useful history lesson about ‘being in a key’:

    ‘The chorales on which these three pieces are based are slightly simplified versions of the troped plainsong kyries from Mass II in the Liber usualis. The three kyries are not in major or minor “keys” at all, since they are modal. These chorales are in plagal church modes, and the organ settings are also in plagal modes transposed so they contain three flats, this part of the Clavierübung being full of Trinitarian symbols. … In keeping with the antique character of the plagal modes, the final chords therefore are based on the “final” of the plagal mode of the chorale on which each of the three kyries is based. … The Gregorian chant on which this is based is the troped “Kyrie summum bonum: Kyrie fons bonitatis” of Mass II for Feasts of the First Class. The “key” of the first kyrie is therefore a transposed G-Phrygian plagal mode. The second is treated more freely in 18-century counterpoint and hence has more of a feeling of major/minor. The third kyrie uses a great deal of inversion in the imitation, with the cantus firmus in the pedal; since it completes the cycle, it returns to the tonality of the first kyrie.’

    Comment by David R. Moran — November 28, 2020 at 11:01 pm

  4. There’s a lot to respond to here. I’ll speak to David Moran’s observations first. Bach’s modal chorales are often based on chant melodies, sixteenth century or earlier, which are constructed in a hexachordal system of modes; there’s a lot of well-known theory on that subject but I’m not historian of theory and it’s mostly way too complicated for me. There’s a book on the subject, Bach’s Modal Chorales, by my Canadian colleague Lori Burns. (Some of the modal notation survives in the so-called “Dorian” key signatures, e.g. Bach’s G minor Sonata for unaccompanied violin, with a one-flat instead of a two-flat key signature.) But I have no problem comprehending a chorale like “Ach wie flüchtig, ach wie nichtig” (no. 48 in the 371 Chorales)as a tonal, not modal, chorale that readily alternates between relative major and minor, with ample supporting dominants (A minor and C major). The accidentals supplied for sixth and seventh degrees in A minor (harmonic and melodic minor as needed) are there because our notational system of key signatures isn’t flexible enough to accommodate them. There are many such instances. Some of this is also discussed in my “Russian Submediant in the Nineteenth Century” article which you can find on my website if you have the patience for some dry analysis.

    It’s interesting that Chopin’s B-flat minor ( = D-flat major) Scherzo has the same key signature as the slow movement of the Balakirev Symphony I discussed a few weeks ago. But D-flat major was a favorite key of Balakirev, and it was also favored by Tchaikovsky in some of his most memorable passages, e.g. the Love Music in Romeo and Juliet, as well as the famous Piano Concerto no. 1, which begins with a four-note repeated “motto,” like Chopin’s, in B-flat minor, immediately followed by the crashing D-flat major entry of the piano. Everybody calls this concerto “…in B-flat minor.” (In case anyone remembers that far back, this beloved concerto received its world premiere in Boston, in October 1875. I don’t recall that the composer was present.)

    Vance Koven’s point about Mahler’s Fifth Symphony is well taken; but in that case we’re dealing with a multi-movement work in which one expects different keys in different movements. (Mahler takes such differences to extremes in the Ninth Symphony, with four movements in D major, C major, A minor, and D-flat major.) When I referred to Falstaff and Meistersinger, I intended to imply that the C major tonalities for beginning and end are really symbolic rather than developed internally. The same would be said of Wagner’s Ring, with the first appearance of the gods in Scene 2 of Rheingold in D-flat major, and their final “twilight” appearance 17 hours later at the end of Götterdämmerung, likewise in D-flat major. There must be something about those five flats. (Mark Lutton’s comment, with the very appropriate “truck driver gear change” reference, applies to the “Pianists” movement in Saint-Saëns’s Carnival of the Animals, beginning with the scale exercise in C major, and then truck-shifting up to D-flat major, then D major, etc.)

    Comment by Mark — November 29, 2020 at 12:17 pm

  5. >> it’s mostly way too complicated for me


    Comment by David R. Moran — November 30, 2020 at 1:46 pm

  6. Mark,
    Great reading your many essays. Strange about the term “progressive tonality”. I’ve heard it most often used with Sibelius, Nielsen and Vaughan Williams. These three composers, some of my very favorites, need a better term to describe their process where a key center is a final goal that the composer strives for, that they spend the entire composition either in close proximity or in distant realms before finding the final tonality. “Progressive” doesn’t describe it.

    Regarding the Ring, Wagner primarily describes Valhalla in the key of D flat, Siegfried’s Horn Call is in F major etc. So, as pointed out to me many years ago by the great Wagner scholar Dr. Robert Bailey, Wagner not only used motifs to described action, he also used their respective key centers.
    There’s also a lot of mention of the Rhinemaidens at the end of the Ring, and that’s why there’s a lot of E flat major (and the Rhinemaidens theme) as well as D flat (with Valhalla burning) when Grane hits the pyre. So yes, you have the D flat major of the key centers, but also the motif (tune) that Wagner uses when he is eliciting Valhalla.

    Regarding the world premiere of the Tchaikovsky Concerto: Hans von Bulow agreed to perform it on his American tour, and yes, the first performance was at Boston’s Music Hall, now the Orpheum Theatre. Bulow did have a great success with the piece, in spite of the trombones not coming in the critical point the first movement, causing the soloist to hurl German invective at them for everyone to hear at the concert. The soloist conveyed to the composer the success of the Concerto, it was the first time a telegram had been successfully sent from Boston to Moscow.

    Comment by Brian Bell — November 30, 2020 at 3:03 pm

  7. Mark DeVoto: “In case anyone remembers that far back, this beloved concerto [T1] received its world premiere in Boston, in October 1875. I don’t recall that the composer was present.”

    Ha – saw what you did there, Mark! (And great discussion here…seriously Bb-Ab-Db! Thank you.)

    Comment by nimitta — December 2, 2020 at 10:32 am

  8. Regarding Classic Ragtime and the Tonic-Subdominant vs. Dominant-Tonic Modulation Concept, a performance fact I stumbled on a few years ago regarding minuets etc. with trios. By the 1960’s it was firm practice to do them AABB-CCDD-AB. There IS evidence, as in a few such pieces (OK Haydn symphonies!) where it was AA’BB-CCDD-etc. (#104) which would imply a repeat of all sections; I think #97 has AA’BB’-CC’DD’ which would require AA’BB’ for the repeat. So Original practice was Do Everything! Now towards the end of the 19th century performance practice had gotten to the point of eliminating the return repeats entirely. Most of Sousa’s marches follow this form and many of them end with trios in a different key; Sousa (1854-1932) was writing to late 19th-century practice. Willem Mengelberg (d. 1950 I think) the Dutch conductor was virtually the last conductor to do this practice with classical symphonies. Think Mozart’s #40’s minuet with its vigorous G-minor drive; then that light G-Major trio–then nothing. Seconds later Mengelberg would start the G-minor finale up.
    In the last decades performance practice seems to be going back to doing all repeats.
    So that is what I think those rags were doing, merely reflecting the current performance practice of no return repeat after the trio–just like in the Sousa marches–rather than a new modulation practice. No one expected a return so composers didn’t give one.

    Comment by Nathan Redshield — December 8, 2020 at 7:37 pm

RSS feed for comments on this post.

Sorry, this comment forum is now closed.