A Riddle in Chopin’s Preludes, Op. 28

Toward the end of Chopin’s Prelude in E Minor, the music comes to a halt in m. 23 on what spells out to be a dominant 4/2 chord. Yet the B♭ in the bass “resolves” upwards across the fermata to B♮ in m. 24 rather than down to A—a kind of syntactical non sequitur (ex. 1). Chopin’s spelling in m. 23 puzzled the editors of the 1949 “Paderewski” edition to the extent that they opted to normalize the notation by changing the B♭ into an A♯. The editors justify their decision by pointing out, in a sense correctly, that A♯ is “the fundamental, altered note of the subdominant in E minor (A–C–E) with the seventh G added.”¹Examples 2a and 2b juxtapose the two spellings and illustrate the different meanings they impart to the pause. Spelled as an A♯, the chord in m. 23 functions as a German augmented sixth and anticipates an upward resolution to $5^$ across the silence. Chopin’s spelling, on the other hand, projects a Neapolitan 6/3 harmony with F♮ over A into the silence. It is only in retrospect that the harmonic function attributed by the 1949 edition’s editors becomes clear.²

That the chromatic voice leading underlying the E-Minor Prelude resists straightforward harmonic analysis is evident from the wide range of approaches to its harmonic structure in the literature.³ One feature of this resistance, as Carl Schachter has observed, is that “chord-by-chord successions fit most uncomfortably” into the normative functional-harmonic framework. He points out that chords in the Prelude often project harmonic functions that connect not to what immediately follows but to non-adjacent harmonies.⁴ Moreover, the functional implications of these chords often derive from their enharmonic spellings. The visually induced non sequitur across mm. 23–24, therefore, stands as a culmination of this feature of the Prelude. The critical difference with m. 23 is that its break in the surface continuity affords a private, unmediated, real-time contemplation of the chord’s dual harmonic identity, particularly for those privy to the notation.

Acknowledging that the B♭ of m. 23 “could have gone down to A as the bass of a Neapolitan sixth (F-major) chord,” Schachter writes:

Admittedly, it is an uncommon exegetical gambit to contemplate what in fact is a non-sounding musical event, but I’d nonetheless like to venture an account of what may lie in Schachter’s “unknowable.” To begin, I suggest that the notation of m. 23 embodies a “negative action”—that is, a gesture defined by an underlying intention not to carry out a premeditated act. Given that the Prelude’s previous B♭s have consistently moved downward to As,⁷ we may speculate that Chopin would have heard and felt in his hands a “phantom” Neapolitan during the fermata in m. 23. Similarly for the performer privy to the notation, internal knowledge of Chopin’s spelling invites a private contemplation of the Neapolitan sonority along with its structural and expressive implications. Yet, crucially, the visual notation calls upon the agency of the performer to subvert the harmony at the last possible moment, to pull the plug—or more precisely, the hands—on the harmony’s materialization, rendering it inaudible for any listener who lacks access to the visual stimulus of the notation.

This “riddle” at once invokes several questions. What is the significance of the Neapolitan harmony and its implied emergence here? Why is its literal and public manifestation deemed undesirable, especially at this point of the form? Why invoke the harmony privately only to call upon the performer’s agency to subvert it? And what are the possible expressive meanings associated with this “negative” action?

One might begin to seek clues to these questions in a surviving sketch-leaf—dated 28 November [1838], Palma—whose two sides are reproduced as plate 1.⁸ On one side, the sketch contains the working out of the Prelude in E Minor and the Mazurka in E Minor, op. 41, no. 1 (accompanied by the inscription of the date); on the other, it contains an earlier draft of the A-Minor Prelude and a pair of unfinished fragments.⁹ The intimate proximity that these works occupied in Chopin’s mind can be gleaned from the close spatial proximity they occupy on this sketch-leaf. The document is a record of once palpable creativity in the form of shared sounds, touches, and structures. Although this sketch-leaf has been featured in several studies with different focuses, the two preludes on it have not been explored with the kinds of structural and affective connections that concern me.¹⁰ In what follows, I argue that a kind of meta “antecedent/consequent” dynamic underlies the A-Minor and E-Minor Preludes. The Prelude in A Minor, through a pair of disruptive elements, establishes a set of structural and expressive antecedents that are treated to a series of counterresponses from the E-Minor Prelude. At the culmination of this process is the riddle embodied in m. 23 of the E-Minor Prelude.

The right-hand melody of the A-Minor Prelude undergoes three phases: the establishment of a process, a critical disturbance, and a variation of the original process.¹¹ There is an unusually modular conception of the melody that makes ubiquitous use of the (025) trichord as its basic motivic unit.¹² While I do not imply set-theoretic sensibilities on Chopin’s part, T/I operations do capture the generative processes that underlie the initial phase of the right-hand melody and spotlight critical breaks from such processes later in the Prelude. During the first phase (mm. 3–12), four successive statements of the trichord are chained together by a common pitch-class, each alternating with its inverted form.¹³ The four trichords then group into two larger pairs of (0257) tetrachords—an important configuration for both preludes, as we will see—related by fifth-transposition. The momentum gained from these tight-knit processes heightens the dramatic impact when they subsequently break down.

During the second phase (mm. 14–16), the right hand enters in m. 14 on a weak beat, tentative and uncertain. For the first time in the Prelude the first pitch of a new trichord group establishes no common-tone linkage with the concluding pitch(-class) of the previous trichord. As a result, the A4 in mm. 14–15—the longest sounding pitch of the entire right-hand melody—creates uncertainty as to whether it will repeat the previous (025) trichord of mm. 10–12, raising the prospect of a shift in the melodic discourse. After nearly seven quarter-note beats, the A4 succumbs to its own sonic decay at the piano. With the left hand already derailed from an expected D-major cadence in m. 11, the right hand is unable to reclaim F♯ to reconstitute an (025) trichord. Instead, it gives way to the melody’s first semitone and interval-class 1, giving rise to an isolated (015) trichord. F♮ as a destabilizing agent displaces F♯ and encapsulates the dramatic turning point for the right hand as well as the global tonal trajectory of the Prelude. The third phase (mm. 17–21) then picks up from the F♮4 of m. 16 and reestablishes the linking of (025) trichords. The reprise of chained trichords now appears only in inverted form. The effect is an accelerated descent in pitch space through a pair of descending minor thirds between mm. 17–21 that pull inexorably downwards.

The left hand, too, begins by establishing a process that subsequently breaks down. A critical difference, however, is that the breakdown “is conclusive, in the sense that the old process is not reestablished.”¹⁴ The left hand first establishes a sequence: an initial E-minor horizon leads to a cadence in G major with the entire progression then transposed up a fifth to set up an expected D-major cadence in m. 11. Yet, for reasons that remain mysterious, the sequence is derailed at the expected resolution to D. The harmony instead homes in on the structural dominant of A minor, arriving on a discernible dominant $46$ harmony at m. 15 notwithstanding the continued surface dissonance.

Aspects of harmony, however, constitute only one of the left hand’s multiple dimensions—another is its somatic aspect. A correspondence might be sought here with the “Funeral March” movement of the Sonata in B♭ Minor, op. 35, whose left-hand ostinato chords—likewise spanning a tenth pivoted around a chromatic inner voice—induce a similar oscillation at the keyboard. Chopin’s Funeral March, as Lawrence Kramer has argued, encodes references to three cultural “logics” underlying each of the newly formed public institutions—the catacombs, the Paris Morgue, and the modern cemetery—that reestablished a symbolic place for death during the first years of the French First Republic.¹⁵ Moreover, from its inception through the twentieth century, the march has been coopted across social histories of death, especially public death, far beyond the social world of its origin. Yet the Funeral March (unlike its acknowledged model, the Funeral March of Beethoven’s Piano Sonata in A♭ Major, op. 26) equally encodes a private notion of death. One aspect that lends it its unusual transmutability, as Kramer observes, is the sonority of the ostinato in the left hand, which proceeds without change or interruption so that its “military topics lose much of their imitative value and become means of symbolization . . . as part of a process of privatization and internalization, rather than throughout (as in Beethoven) as part of a depiction of a public event.”¹⁶ I would further contend that this private, first-person perspective is inscribed in the somatic dimension of the ostinato, accessible in the literal sense only to the performer, and that it is this dimension that the left-hand of the A-Minor Prelude shares with the Funeral March. Chopin’s idiosyncratic beaming in the prelude suggests that the left-hand figure could be realized as a composite of two distinct physical and sounding elements: a finger legato in the inner part (beamed upward), framed by a non-legato “walking” ostinato spanning a tenth (beamed downward).¹⁷ The combined figure induces in the forearm of most pianists an oscillating motion from side to side. Chopin, who did not have particularly large hands,¹⁸ would have been attentive to the poetics of this uncommon choreography.

This allegory aligns with the frequently suggested identity of the inner four-note motive as deriving from the Dies Irae chant.¹⁹ In the upper-left portion of the sketch that contains the A-Minor Prelude, one witnesses Chopin’s conscious crafting of the notation of this left-hand figure. After rejecting an initial attempt, he arrives at a conception immediately to the right of the rejected version, which he carries through to the Stichvorlage (“engraver’s copy”), reproduced as plate 2.²⁰ Chopin’s experiments with notation show his concern not only for how the figure should sound in performance but equally for how its meanings can be expressed by the visual representation of its embodiment.²¹ Yet modern Urtext editions of the Prelude alter what I believe is an essential feature intrinsic to Chopin’s calligraphy: the confinement of the Dies Irae motive within the outer boundaries of the ostinato.²² By placing the upper beams beneath the upper notes of the ostinato, Chopin imparts a sense of suppressed latency and sonic murkiness to the Dies Irae motive that is key to understanding how it functions later in the Prelude. The 1992 National Edition of the Complete Works of Chopin (ex. 3), like other modern Urtext editions,²³ alters Chopin’s notation by raising the beaming of the inner voice above the top G of the ostinato. By bringing the Dies Irae motive out of its enclosure, the typesetting prematurely “brightens” the timbral and expressive meaning of the motive.²⁴ Moreover, the motive is “fuzzy” at its outset; it replicates the contour but not (yet) the precise intervals of the chant melody—it is one semitone too wide at the boundary.²⁵ Not until m. 15 does the Dies Irae motive crystallize on the “correct” scale degrees (as $3^$–$2^$–$3^$–$1^$ of A minor) over a discernible dominant $46$ harmony. Here the motive “bottoms out” as it were and progresses no further, marking the goal of the left-hand figure.

Building on Kramer’s insight that the Prelude in A Minor is a “many-sided study in dialectic” and that “on the largest scale, dialectical reversal appears in this prelude as a gradually unfolding antagonism between melody and harmony,”²⁶ I suggest that the Prelude combines its oft-discussed tonal structure with the action of the pianist’s hands on one another. The left hand, embodying powerful associations of death and decay, first catches the right hand off-guard on the downbeat of m. 11, when it steers away from a projected D-major cadence, while the right hand follows the left hand’s lead and succumbs to its tonal and expressive pull. In m. 15, when the Dies Irae motive reaches its goal on A♮ as its $1^$⁠, the right hand responds by breaking its chain of (025) trichords with an (015) trichord containing F♮. The transformation from (025) to (015) models the transformation of the Dies Irae motive from a (014) at the outset to the “correct” (013) in m. 15—both adjustments involve one semitone in their intervallic contents.

The agency of the left hand continues its domination over that of the right through to the final cadence. After the dramatic events of mm. 11–16, the right hand returns to its former process of chained (025)s. It now does so without alternating with its inverted pairs and hence traverses a linear descending-thirds sequence. The result is a descent in contour that suggests a sense of resignation after the displacement of F♯ by F♮ in m. 16. What I and others call the “chorale” music that follows (mm. 21–22) contrasts sharply with the rest of the Prelude.²⁷ It represents an oasis of harmony, counterpoint, and rhythm at their purest—a momentary relief from the Prelude’s chromatic density. This fleeting passage seems intent on reversing the two elements—F♮ and A♮ (A minor)—that have altered the course of the Prelude. It seems to suggest, however tentatively, an E-major tonicization with F♯ reinstated as a diatonic pitch class. While we might bask in the chorale music’s momentary purity, it ultimately proves unsuccessful at resisting the cumulative forces of an insurgent A-minor horizon via F♮ in combination with the Dies Irae motive’s crystallization.

The final cadence in A minor (mm. 22–23) abruptly revokes any hopes of E becoming tonic. Yet once again, tonal structure is only partially responsible for the ending’s sense of “tragic desolation,” as Robert Hatten calls it.²⁸ Contributing equally to this sense is the way the Dies Irae motive manifests itself somatically in the pianist’s right hand at the structural cadence. For as long as the left-hand figure lasted, the walking ostinato had kept the motive tightly in check within the middle fingers, rendering it (and its associations with death and decay) latent rather than explicit. In mm. 22–23, however, the Dies Irae motive can be heard to break out of its physical confinement in the left hand to clinch the upper voice of the structural cadence. Now set free, it takes possession of the right hand, the Prelude’s speaking voice. This final gesture forms both the culmination and the completion of the left hand’s dominance over the right throughout the course of the Prelude. The motive here is truncated from a four-note incipit into a three-note figure (B–C–A) that both fits with the right hand’s motivic profile and highlights the closing function of a motion from $2^$ to $1^$⁠.²⁹ As a final manifestation of its appropriation of the right hand, the Dies Irae motive here “speaks through” the dotted-quarter rhythmic figure that had throughout defined the right hand’s musical voice.

Finally, the visual design of Chopin’s notation in these final measures additionally bears out the opposing trajectories between the hands. Plate 3 reproduces the Stichvorlage alongside the “Paderewski” edition’s misguided layout (ex. 4). Chopin’s notation suggests a contextual “threshold” on the pitch B3. Given that each of the final three events contains B3 as a constituent pitch, B3’s shifting association with either the upper or lower staff imparts meaningful expressive distinctions to these gestures. In m. 21.1, the B3 as the final pitch of the last audible (025) trichord remains above the threshold. The B3s of the chorale music that follows, however, sink into the lower staff entirely. Lastly, when the Dies Irae motive clinches the structural cadence, the B3 rises back above the threshold along with C4. The notation paints a particularly vivid depiction of the Dies Irae’s dominance over the right hand’s discourse: the two B3s at m. 21.1 mark the right hand’s final stand before disappearing into the chorale music, while the B3 at m. 22.2 surges from the murky depths to clinch the structural cadence.

Schenker’s well-known interpretation of the A-Minor Prelude as an exemplar of large-scale auxiliary cadences suggests (as the analytic concept implies) a progression from relative tonal instability toward stability, a homecoming from a dissonant point of departure.³⁰ Yet any sense of comfortable resolution seems far removed from the common perception that the Prelude’s ending is contingent and volatile.³¹ A fuller account of the Prelude would need to reconcile the tension between the outward tonal stability of its ending with its equally palpable sense of “otherness.”³² Chopin’s far-reaching strategy here is not to resolve destabilizing surface elements at more background levels of structure, but rather to let those very elements become constitutive members of the structural background. By effectively “stabilizing” the pair of initially disruptive elements—F♮ in conjunction with (015) and A♮ in conjunction with the Dies Irae motive—Chopin makes closure maximally unstable, inviting us to recognize that these elements have taken over the deepest foundations of the work.

If the Prelude in A Minor enacts the forces of death overtaking life, the Prelude in E Minor is an expression of profound grief over that loss. With its chromatic descending bass line and a melody marked by descending semitones, the E-Minor Prelude has all of the principal markers of the musical lament established since at least the seventeenth century. The close proximity of both preludes on the sketch-leaf—we can safely assume that both sides were written in Majorca—further encourages us to imagining that one was born out of the sounding backdrop of the other even if the two preludes are neither adjacent as published nor intended to be performed back-to-back.³³ If the A-Minor Prelude concludes with the terrifying triumph of death channeled through the destabilization brought forth by F♮ and A♮, the E-Minor Prelude attempts a counternarrative that sets out to reverse the course laid down by those structural and expressive markers. “Failed” elements from the earlier prelude are reinscribed into deep levels of the later, while “successful” counterparts are reappropriated in ways that strip away their original expressive potency.

Like many of the op. 28 Preludes, the E-Minor Prelude is in the form of a large period. As both antecedent and consequent approach their cadential goals, the music moves away from its characteristic chromaticism toward diatonicism. The semitones that initially structure the melody grow wider and more consonant with their supporting harmonies, while the left hand homes in on structural pre-dominant and dominant harmonies. These trajectories from instability to (provisional) stability are key to how the E-Minor Prelude reappropriates elements from the A-Minor Prelude. Example 5 recalls the three principal events that altered and subsequently cemented the course of the latter: the onset of a (015) trichord displacing F♯ with F♮, the crystallization of the Dies Irae motive in the scale degrees of A minor, and the Dies Irae’s final emergence at the structural cadence. The example shows how the E-Minor Prelude reawakens all three events along with their destabilizing impulses and subjects them to transformations that neutralize their original implications.

Perhaps the most immediate connection between the two Preludes runs between the ending of the A Minor and the opening of the E Minor. Both Schachter and Eigeldinger observe that the octave upbeat of the E-Minor Prelude, which appears cramped on the autograph sketch compared with the Stichvorlage, must have been realized at the midpoint of the compositional process.³⁴ If the E-Minor Prelude is played and heard with the A Minor in mind, this crucial detail can have the effect of reawakening the Dies Irae motive from the end of the latter. If we imagine someone working through the published score, it is not far-fetched to suppose that the affinity between its first and second minor-key numbers would be discernible. The ending of the intervening G-Major Prelude might even facilitate the recognition, since the top note in its final chord is the B4 to which the octave leap in the E Minor brings one.

Example 5 shows that the upbeat figure helps transfer the three-note motive of the A Minor and its underlying rhythmic shape up an octave to the principal register of the E Minor. The Dies Irae motive thus can be heard to lurk in rhythmic augmentation within the initial melody of the E-Minor Prelude. Yet the identity of the motive begins to dissolve almost as soon as it is reinvoked. The melody, crucially, forges no direct foreground connection from C5 to A4—a connection that is intrinsic to the identity of the original chant melody. Instead, the E-minor melody reiterates the first two notes of the motive (B4–C5) three times, as if searching in vain for the A4, only to have the left hand slip away from its initial E-minor horizon, prompting a new response from the melody. The resistance against materialization of the Dies Irae motive might be said to foster the pervasive neighbor-note figure B–C–B that saturates multiple levels of the Prelude. The figure avoids a direct connection from C to A, which would otherwise have brought back the pitches and contour of the Dies Irae motive. Such a connection, at this early stage, would have short-circuited the E-Minor Prelude’s reappropriation processes.³⁵

Moving on, ex. 5 shows that all members of the previously displaced (025) trichord return in the E-Minor Prelude as stable motivic units in the latter parts of both phrases. This shared feature draws attention to the centrality of F♮ and F♯ in shaping the narrative of not only the A-Minor Prelude, but also that of the two preludes combined. The uppermost arrows of ex. 5 show the transformation of the (015) trichord back into the (025) trichord it had earlier displaced while the F♯ of the (025) trichord now (re)gains structural footing as diatonic $2^$⁠. In contrast with the left hand of m. 11 of the A-Minor Prelude, the (025) configurations here are given consonant support by structural pre-dominant and dominant harmonies (mm. 9–10 and mm. 18–19).

Finally, ex. 5 illustrates the E-Minor Prelude’s reappropriation of the left hand in mm. 15–19 of the A Minor, where the Dies Irae had first crystallized on the tonic scale degrees. The principal pitches—A, C, E—can be heard to return in mm. 9–11 of the E-Minor Prelude as iv⁶ in the left hand, functioning as the structural pre-dominant of the antecedent. In the consequent, the harmony recurs in m. 18 with F♯ now welded to the triad to make for ii$56$⁠. Solely within the context of the E-Minor Prelude, the added coloration and change of bass of the latter harmony might be taken as the consequent’s registral and dynamic intensification over the antecedent. Yet these details can take on greater significance when heard against the A-Minor Prelude. They highlight the leveraging roles of F♮ and F♯ and how shifts between the two can tip the balance of entire tonal and narrative trajectories. Example 5 shows that, in mm. 15–19 of the A-Minor Prelude, the A-minor triad in the first instance was strongly bonded with F♮. The bottom arrows show that, in the E minor, this sonority is progressively reappropriated by the new context into stable diatonic entities. The antecedent first decouples the A-minor triad from its association with F♮ (m. 9.2–4), and the consequent then brings in F♯ to reinforce it (m. 18.2). Throughout these processes, elements and events that were originally destabilizing are rendered stable. The cumulative effect is a kind of neutralization of their earlier expressive implications.

Example 6 illustrates the converse, whereby elements that had “failed” to attain structural status in the A-Minor Prelude are now assimilated into deep levels of structure. Schachter’s analysis of the E-Minor Prelude calls attention to an unusual feature of the music’s melodic unfolding. By his account, the fundamental line, which starts on $5^$⁠, bypasses $3^$ to make for a distinctly “gapped” Urlinie in both antecedent and consequent.³⁶ The $4^$⁠, which occurs over the subdominant in m. 9, bypasses $3^$ to arrive on $2^$ over V.³⁷Example 6 juxtaposes the E-Minor Prelude’s fundamental structure next to elements from the A-Minor Prelude. The four tones of the “gapped” Urlinie replicate, in pitch, the second pair of (025) trichords from mm. 8–12 of the A-Minor Prelude that were earlier denied their expected D-major cadence but are now inscribed into the background. The sympathetic resonances shared between identical notes on the piano impart a consistency of range and timbre to the right-hand melodies of both preludes, which in turn reinforce their relationship. Lastly, the example shows that left-hand elements from the A-Minor Prelude that had invoked an “E-centric” horizon are likewise absorbed into the new fundamental structure.

Although the antecedent and consequent phrases are comparable in length (twelve and thirteen measures, respectively), the two phrases of the Prelude in E Minor project markedly different internal proportions. The consequent achieves its greater sense of dramatic and emotional breadth from the ways it compresses the structural motions of the antecedent. The structural cadence therefore arrives “prematurely” in m. 21, and a third musical space is carved out in mm. 21–25 without impairing the fundamental two-part design of the period. In light of the foregoing discussion, what transpires in these final measures is nothing short of extraordinary.

In m. 21 the structural dominant deflects onto VI and is articulated by a crescendo hairpin placed across mm. 20–21. Significantly, the hairpin—which in the nineteenth century often had an agogic implication in addition to a dynamic one³⁸—originates from the end of m. 20, underscoring the sense of a subversive agency surging forth from within the dominant in the left hand. As is the case with many deceptive cadences, the upper voice/right hand remains on scale degree $1^$ at the “resolution,” lending the impression that the right hand’s agency remains “unaware” of the “deception.”³⁹ We learn more about the profile of this new agency when the left hand ushers back the downward chromatic voice leading reminiscent of the opening, offsetting the upward gestures at the beginning of the stretto, mm. 16–17, and the ensuing diatonicism of mm. 17–20. The return of B♭ in m. 21, now as a seventh over C, induces the same downward pull characteristic of the B♭4 of m. 4. And, as Schachter notes, the return of B♭ in mm. 21 and 23 completes a three-part narrative arch marked by the enharmonic spelling of the pitch-class B♭/A♯.⁴⁰ Yet the B♭ of m. 21 marks neither a structural nor an emotional return. As the chord tone of a harmonic span and not merely a passing phenomenon, B♭ threatens to usher in F♮ and thus to undo the Prelude’s recuperative processes thus far.⁴¹ Indeed, one senses F♯ losing its footing in mm. 21–22. In mm. 19–20, F♯ had planted itself firmly on downbeats as a consonant member of the structural dominant; in mm. 21–22, however, F♯ is pushed onto beat four as a dissonant upper neighbor to E with no harmonic support.

Although Schachter is well aware that the different spellings of the pitch class B♭/A♯ hold great expressive purpose for the Prelude, he ultimately chooses to “normalize” the B♭ of m. 23 to an A♯ in his final analysis and to interpret it as the end point of an “augmented-sixth” span stretching back to m. 21.⁴²Example 7, however, develops an alternative analysis that envisions the structural and expressive implications of Chopin’s B♭ spelling.⁴³ The analysis considers what would happen if B♭ were to take hold and a Neapolitan 6/3 to materialize in the space of the pause. In this hypothetical scenario, the B♭ of m. 23 would no longer function as a chromatic lower neighbor to the dominant but as a passing tone that serves to connect C3 (from m. 21) to an implied A♮2 at the pause. This would constitute the only instance in the Prelude whereby the pervasive B–C–B neighbor-note figure might not potentially “seal.” Example 7 shows that in such a scenario the four-note Dies Irae motive would have been allowed to materialize deep in the middleground between mm. 16–23, bringing back at pitch its first crystallization in mm. 15ff. of the A-Minor Prelude. Similarly, a resolution to F♮ in the melody at m. 23 would have brought back the disruptive (015) trichord of the A-Minor Prelude, likewise at pitch, and inscribed it into the middleground of the upper voice. To be sure, the hypothesized voice leading is necessarily awkward, since the example aims to capture graphically a road decisively not taken. Among other things, the status of E4 in m. 24 as a passing tone grinds against its gestural presentation. But this very fact further adds to the notion that a manifestation of the Neapolitan harmony at this late point would not only have reinstated the principal pair of destabilizing elements from the A-Minor Prelude—F♮ and A♮ along with their associated motives—but also unraveled the restorative processes of the E-Minor Prelude.

The foregoing analysis describes a three-part structural process underlying the E-Minor Prelude that cuts across the work’s two-part outer form. The Prelude progresses from reappropriating destabilizing structures from the Prelude in A Minor, to opening the doors for their resurgence, to suppressing that resurgence. When this progression is experienced in the context of the Prelude’s lament topic, there emerges a possible correspondence with the general contours of human grief: the need to overcome past memories of trauma, the proclivity for those memories to resurface in present consciousness, and the will required to mitigate their force. And if we further hear the Prelude addressing us “in the first person singular of a stylized singing voice” and infer “an emotional character from the tone of that voice,”⁴⁴ then the performer may fuse with that character and adopt its tone in much the same way that an actor embodies the words and actions of a fictional character. One might go as far as to say that it is by way of the performer’s embodiment that the sonic materialization of the Neapolitan is suppressed and shielded from public view.

But what possible context and specific expressive content, if any, might Chopin have adduced? The final moments of the E-Minor Prelude mirror a distinction Chopin is known to have maintained between music for his private use and music for public consumption. There was a critical separation in his mind, as Kallberg notes, between improvisation and sketch, spontaneous outpouring and compositional labor.⁴⁵ It is generally understood that Chopin formulated initial ideas privately at the keyboard. He would then move to an intermediate stage where he subjected these initial ideas to constructive process of conscious composition, during which he refined and crafted his musical materials into a form suitable for public presentation. The final stage thus involved preparing manuscripts, Stichvorlagen, suitable for wider dissemination.⁴⁶ Chopin’s practice of avoiding public use of private musical output may therefore shed further light on the implications of the silenced chord in the E-Minor Prelude.

In view of these compositional habits, we could well imagine that the unmistakable first-person “tone” of the finished E-Minor Prelude first came to life during Chopin’s improvisations at the piano. In the published work, one might identify remnants of these original improvisations in, for instance, the dramatic intensification in m. 12 of the opening B3–B4 upbeat⁴⁷ and the build-up to the stretto in mm. 13–20. I find Schachter’s description of these latter measures—“the sudden storm that builds up in the next few bars [mm. 16–18] is all the more gripping in its contrast to the phrase’s beginning, like a violent outburst of tears from a person not given to demonstrations of feeling”⁴⁸—an apt description for the kind of psychological impulse that may have inspired these passages during their inception at the keyboard.

Similar kinds of emotional outburst could likewise have been the initial inspiration behind the interrupted cadence at m. 21. One might conjecture a scenario in which a state of emotional vulnerability opened the way for disquieting elements explored in a prior improvisation to claw their way back into present consciousness. The interrupted cadence at m. 21, with its added B♭3 as seventh accompanied by a crescendo hairpin, ushers us back onto the horizon of A♮ and F♮, the pair of disruptive elements from the A-Minor Prelude. The final measures of the E-Minor Prelude therefore invite speculation as to whether any personal experiences might have triggered their recollection.

I am well aware that any relationship between musical structure and a composer’s personal experiences can (in default of positive testimony) be no more than hypothetical. And while my intention here has not been to lay the groundwork for a hermeneutic interpretation through the “authority” of structural analysis, I am nonetheless encouraged to explore the intriguing three-way intersection between the musical-structural evidence on the one hand, Chopin’s known compositional proclivities on the other, and his circumstances around the date of the autograph sketch at the center.

Table 1 presents a timeline and extracts from letters sent by Chopin and George Sand in the immediate days surrounding the date of the autograph sketch.⁴⁹ (See Table 1.) The selection foregrounds the deterioration of his health during this period. In the letter addressed to Julian Fontana on 15 November 1838, Chopin reports on his generally positive disposition. On that date he had been in Majorca for roughly one week and had recently decamped to Son Vent, a summer house in a small enclave northwest of the port of Palma. However, his sunny disposition did not last long. From what we can surmise, it was shortly after this letter that his health dramatically deteriorated while he was living in Son Vent. In both his and George Sand’s letters on 3 December, the adverse state of Chopin’s health was first revealed to members of their inner circle on the continent. In his 3 December letter, again to Fontana, Chopin reports that he had been “sick like a dog for the last two weeks” and gives reports (though with a sense of sarcasm) of the doctors’ uniformly grim prognoses. He claims to have fallen ill the day after he last wrote to Camille Pleyel, which we can presume to be 22 November, given that we know he last wrote to Pleyel on 21 November. The date inscribed on the autograph places it precisely five days prior to the report of 3 December and six days after the day Chopin presumably fell ill. It is thus almost certain that formative ideas for the E-Minor Prelude and the E-Minor Mazurka, op. 41, no. 1 (the two works on the side of the sketch that contains the date inscription), were worked out during this period of physical struggle and accompanying psychological duress.

Later, Sand related that the visibility of Chopin’s illness—especially his coughing of blood—created anxiety among the Majorcan locals. The fear of contagion determined much of Chopin’s and Sand’s dealings with the provincial populace, upon whom the two Parisians (with Sand’s two young children) depended for their everyday needs. On 15 February, Sand wrote bitterly to Carlotta Marliani from Barcelona after having left Majorca on the same day:

Notwithstanding cultural and other factors that might have contributed to the Parisian Sand’s bitterness toward the Majorcan locals, it is reasonable to think that her sense of besiegement on the island was real and owing in great part to the Majorcans’ perception of Chopin as a danger to public health. His visible coughing, which posed a significant obstacle in their everyday dealings on the island, would have been still more pronounced during the earlier days in Son Vent before they relocated to the more remote Valldemossa monastery. It is reasonable to suspect that Chopin would have felt a persistent and practical need to conceal signs of his perceived consumption in public.

We might ask what kind of human predicament could inform the peculiar structural trajectories of F♮ and A♮ across the two preludes examined here. Apprehending them as markers for symptoms of consumption and death, especially in conjunction with the Dies Irae, we might well ponder whether Chopin’s social impulse to conceal the signs of his consumption during his time in Majorca could have in part shaped his compositional impulse to shield the resurgence of F♮ and A♮, so fully present in the A-Minor Prelude, from public view in its E-minor companion.

The relationships between the A-Minor and E-Minor Preludes, I have shown, derive in large part from how F♮ and F♯ exert different kinds of harmonic leverage on their respective tonics. This by itself is not extraordinary, given the many available points for contact between any two fifth-related keys. What is beyond the ordinary are the surface and deeper-level designs that provide a rich set of contexts for F♮ and F♯ to shape their relationship in unpredictable ways. The autograph sketch sheds further light on this aspect of the two preludes.

The placement of the Mazurka, op. 41, no. 1, immediately above the E-Minor Prelude suggests that Chopin worked on the Mazurka first, before starting the Prelude on the space left at the bottom of the leaf. Example 8 reproduces the opening phrase of the finished Mazurka. The two initial two-measure units comprise two PAC cadential progressions with equal weighting and emphasis, the first in A minor and the second in E minor. They harbor an inherent ambiguity: which of the two keys takes precedence? Taking the latter to be tonic, Schachter observes that the melodic structure of mm. 1–4 duplicates the gapped Urlinie of the E-Minor Prelude in miniature and that the two works “contain two quite different realisations of a single compositional idea.”⁵¹ He suggests that “[the] inspiration [of the Prelude] was generated by work on the Mazurka, a supposition confirmed by the initially absent upbeat at the beginning, which makes the Prelude resemble the Mazurka even more.”⁵² What is still more astonishing is that the Mazurka’s “structural cadence” not only lacks a dominant but also leans heavily on F♮. Although such chromatic inflections are not uncommon in Chopin’s mazurkas, the substitution of an expected dominant—and attendant F♯—with an unharmonized F♮ of the Phrygian undermines E as tonic and allows A minor to cloud the final sonority. The onset of F♮ keeps alive the earlier ambiguities through to the very last chord—and beyond.⁵³

Accepting the likelihood that the Mazurka was drafted before the E-Minor Prelude (but some time after the A-Minor Prelude), we can also imagine that it served in part as a testing ground for working out the pitch-structural mechanics between the two preludes before implementing them in the E-Minor Prelude. Taking Schachter’s point a step further, we may speculate that Chopin used the occasion to hone his control over musical relationships that balance fifth-related keys by way of their shared ♭$6^$/♭$2^$ complex. The three works form a continuum with respect to this shifting tonal balance. In each case, F♮ and/or F♯ serve as a fulcrum that activates a different balance point on the tonal axis between A and E. In the A-Minor Prelude, the displacement of F♯ by F♮ nudges the piece toward an A-minor closure. In the E-Minor Prelude, the suppression of F♮ at the pause ultimately allows F♯ and the dominant to assert a successful closure in E minor (although we may now wish to question its ultimate success). The Mazurka, lastly, could be said to stand midway on this continuum: F♯ and F♮ bear almost equally on its tonal fabric, producing a “seesaw” effect corroborated by the modularity of its outward phrase rhythm. Of the three works, the E-Minor Prelude may be said to embody this interplay with the greatest subtlety.