The music of Medtner has made a lasting and profound impression on me, first as a listener, then as an interpreter, and now as an aspiring composer studying music theory. I chose to analyze and present a work of Medtner's partly because I feel his music deserves to be better known — particularly by musically educated audiences — and partly because I felt impelled to dig deeper into his work. From my attentive listening, I was quite sure nearly any Medtner piece would make a fascinating study, and the question became where to start for purposes of a half-hour classroom presentation. I selected the Skazka, Op. 26 #3 for its beauty, density of material, and remarkable harmonization, all in the span of a tractable three pages.

Despite its brevity, the piece manages to be an approximation of a sonata-allegro movement. There is a Theme 1 (measures 1-16), a Theme 2 (measures 17-25), and between the two there is an ambiguity of key signature; there is a development section (measures 30-54) making some use of both themes; and there is a recapitulation of Theme 1 (measures 55-70) with some reference to Theme 2. The form ultimately deviates from classical sonata-allegro due to space constraints (for instance, Theme 2 is not unambiguously recapitulated), but the piece is clearly composed with that form in mind.

Fascinatingly, placing this ternary piece under the microscope reveals that its opening is itself in ternary form. Measures 1-4 spell out the motivic idea of Theme 1. Measures 4-8 develop the idea through a two-step sequence moving chromatically downward by half-step. Measures 9-12 repeat the motive in melodic inversion, ending on an ambiguous augmented fifth. Measures 13-16 recapitulate the motive, harmonized differently for the first two measures (VI - Neapolitan vs. i - iv), then identically for the final two. On a large scale, when a sonata returns to a theme, the listener recognizes it from previous occurrences. On a small scale, using the same technique, Medtner makes Theme 1 seem familiar after hearing it just once.

Theme 2 proceeds immediately from Theme 1. The left hand figuration used to transition from measure 16 to 17 is identical to that connecting measure 4 to 5. The shape of this figuration is “bouncy” and is used throughout the middle of Theme 1, which is to say, wherever harmony is in flux. As measures 17-25 use this LH figuration exclusively, it is no surprise that Theme 2 bounces between F minor and A♭ major, avoiding tonicizing either over the other. Theme 2 opens melodically an octave higher than Theme 1, with the notes F and E♭ held long. Theme 1 only reaches these notes in this register in one place: measures 4-7, part of its “development” section. Theme 2's melody, then, can be seen as a development on this portion of Theme 1. Both themes are melismatic (within phrases, there is no melodic leap greater than a fourth) and marked cantabile or cantando. In other words, Theme 2 follows directly from Theme 1, not only temporally but also logically. The Pedale marking held across measures 16 and 17 quite literally underscores the point!

The I - vi - V/V progression in A♭ at the end of Theme 2 (measures 23-25) leads into a brief transitional section (measures 26-29) outlining E♭ as a V, first in the RH, then the LH in contrapuntal inversion. Starting at measure 30, however, chord functions become less clear and stepwise linear motion becomes more significant for analysis. In this context, the rising stepwise motion of the transitional section should be read not only as a resolution of the chord progression it completes, but also as a premonition — in the Mixolydian mode, lacking the stability of a leading tone — of the linear melodic motion which follows. Just as Theme 2 derives from Theme 1, this transitional material is not completely new. Its first few notes in the RH melody are redolent, both rhythmically and melodically, of the first few notes of Theme 1; the LH also starts with Theme 1's characteristic eighth-note-rest delay followed by a widely spaced rising arpeggio.

At measure 30, the development begins in earnest. In measures 30-33 we have a figure centered around E♭that begins very closely spaced and expands simultaneously downward (LH) and upward (RH). The figure repeats in measures 34-37, this time handled solely by the RH in order that the LH can “sing” from Theme 2. The next part of Theme 2 begins at 38 on the F in the RH, switching pianistically to the LH in 39. Yet something is amiss. An exact diatonic copy of Theme 2 would have B♭ on beat 3 of measure 37 and would begin the stepwise ascent at beat 1 of 38 on a D♭. The expected B♭ doesn't arrive until beat 2 of measure 38. Meanwhile, the stepwise ascent starts on time, but it's raised by a third (starting on F) and takes us to the key of b♭ — though this tonicization is short-lived and lacks the feeling of a full modulation. In these few bars, by tearing the two phrases of Theme 2 asunder, Medtner lays the groundwork for the disorienting harmonic shifts to come.

Treating the b♭ interlude as a sort of iv, in measures 43-46 we reiterate 34-37 over F, a step higher. The RH opens with an eighth-note rest, which follows logically from the RH rhythm in previous bars, but has otherwise heretofore been identified solely with the LH. Things are getting topsy-turvy now. Just as in beat 3 of measure 37, something unexpected happens on beat 3 of measure 46. This time, however, it's even more unexpected. The A♮ in 37 is diatonic in the context of tonicizing b♭, but this B♮ is purely chromatic. Over a major-minor seventh chord rooted at B♮, a modified theme 1 returns in measures 47-48. The function of this chord is very difficult to characterize. Theme 1, this time almost intact (but a half-step upward), is then repeated in the RH over a nearly identical LH, giving a more typical major-minor seventh sound, though if it were a dominant seventh we would expect a resolution to E and there is no context to suggest that's where we're headed. Echoing the outward stepwise motion earlier in the development, the first RH note of measure 47 moves up a half-step to 49 while the first LH note of measure 48 moves down a half-step to 50. Measures 51-52 reiterate 47-48 note-for-note — one of the few places in the entire piece where nothing has been changed — followed by 53-54, wherein the RH rises stepwise, as does the last note of the LH from measure 52 to 54. In measures 53-54 we finally hear what sounds like it could be IV leading to vii/V, which — if true — would be enough information to determine that 49-50 is also a IV, and that perhaps 47-48 can be thought of as a Neapolitan relative to the b♭ tonality which we're leaving behind.

Is it true? Having built up such massive tension (not only harmonically but also by means of crescendo, accelerando, and widening spacing between the hands), Medtner makes us wait to find out. Only after the fermata do we hear the telltale signature of a V64 chord supporting the recapitulation of an unadulterated Theme 1; finally, a few bars later at measure 58, we arrive at an unambiguous resolution to i and the feeling that the remainder of Theme 1's original sixteen bars is on the way. Yet once again, something is amiss. Theme 1 is not in its original key, it's a half step higher! Nonetheless, everything returns to normal by measure 67. How is it done?

Comparing these bars with their earlier counterparts is instructive. The RH lowers C♯ to C♮ on beat 2 of measure 61, and B♮ to B♭ on beat 3 of measure 63. In the former case, this chromatic alteration occurs over an unmodified LH; in the latter, the LH has been changed from a simple triad (as in measure 9) to a major-minor seventh. In both cases, these alterations merely modify the expected resolutions of these “dominant” chords from major to minor, and this is particularly unobtrusive because the alterations occur in passing and the resolutions never occur. Furthermore, the major-minor seventh in measure 63 functions as a pivot: it's an augmented VI in f♯ until the RH plays B♭, when it becomes a V/ii in f. The RH A♭ in measure 65 (over a LH chord similarly modified from its counterpart in bar 11, this time a genuine dominant seventh) signals that normalcy has almost been achieved. By measure 66 we are solidly back in f and only one note differs from bar 12: the LH A♭ has become a B♭, making the chord function a clear V. While this is an inevitable result of the same linear stepwise motion found throughout the piece — in particular, of the stepwise descent of the bass in the original Theme 1 — it also has the desirable effect of suggesting that the piece isn't over, because it hasn't quite finished evolving.

True enough, after measures 67-70 reiterate 13-16 precisely, there is a small coda. Over a steady tonic pedal, the RH plays Theme 1 in diminution while the LH (above the right) seems inspired by both Theme 2 and the Dies Iræ. The hands descend together by step before resolving into F. One last quote from Theme 1, the “bouncing” figuration in the LH from measures 4-5 (this time moving to a lower register), ends the piece.

From having listened to a great deal of Medtner, I had intuitively concluded that “everything is related.” Studying this relatively simple piece not only supports my conclusion but also raises a new hypothesis about his music: “everything has to be exactly as it is.” It's an unprovable assertion, of course. But there's a whole lot of music to be studied, and I'll certainly enjoy investigating.