Recomposing Two Musical Settings of Goethe’s “Ein Gleiches” in Romantic Style

Stephen Rodgers

I’d like to begin with a brief audio clip.

What you just heard was the opening vocal melody of Robert Schumann’s famous song "Im wunderschönen Monat Mai," from his song cycle Dichterliebe—not, however, in its actual form, but instead in recomposed form. This is how Schumann’s vocal melody really ends:

Several weeks ago, I performed this recomposition as part of a presentation in a class on music and text in German culture. I wanted to highlight what makes the first song of Schumann’s cycle so expressive—the way he pulls the vocal melody ever upward, and pulls the phrase away from the home key, just as the poet speaks of his love "breaking forth"—and it seemed to me that the easiest way to do that was to play what Schuman might have done, and then compare that with what he actually did.

This is nothing new. Music analysts often use recompositions like this, above all in teaching but also sometimes in print.¹ However, they seldom reflect on the strengths and implications of the practice of recomposition. It’s easy enough to find individual examples of recomposed musical passages in studies that deal with other topics, but it’s much harder to find studies that focus on recomposition and use it as a primary analytical tool. There are some exceptions, of course. In the 1950s, the British musicologist Hans Keller created a system of "functional analysis," which aimed to present analytical observations without the aid of words; in essence, it involved isolating and juxtaposing various musical fragments from one work or many works and sometimes recomposing those fragments to highlight musical relationships.² The music theorist Kofi Agawu has argued that the analysis of music should be viewed as a hands-on, creative activity akin to the composition of music.³ More recently, another music theorist, Matthew BaileyShea, published an article that used recomposition to graft together several different settings of one of Goethe’s Mignon poems.⁴ And in May of this year, William O’Hara completed a dissertation at Harvard University entitled "The Art of Recomposition: Creativity, Aesthetics, and Music Theory," which explores the use of recomposition by music theorists from the eighteenth century to the present day.⁵

Still, recomposition deserves to be studied more closely and used more widely because it raises crucial questions about the goals and methods of music analysis in general. For example, what can recompositions do that conventional music analyses can’t? How do what we might call analytical recompositions (that is, those used as an aid to analysis) differ from more creative recompositions—like the collage-like pieces of Luciano Berio, which juxtapose quotations from other works with Berio’s original music; like Hans Zender’s Winterreise: A Composed Interpretation (1993), which reorchestrates and radically transforms the music of Schubert’s cycle;⁶ like the original "remix" that Chi Wang discusses in her essay; or even like the poetic "revisions" of a Goethe theme that Dorothee Ostmeier analyzes in her article? In these cases, the goal may be less to make an analytical point than to make a new, stand-alone work of art (although as Chi Wang points out, and as Zender’s subtitle "A Composed Interpretation" suggests, it’s possible to do both things at once). Granted, I had to use a small amount of compositional skill to produce the recomposition in Audio Example 1, but I would never describe it as a composition, not in the sense that Chi’s piece is a composition. I would, however, describe it as an analysis—and not just a supplement to a written analysis but a bona fide analysis in its own right, an elucidation of the special formal and expressive features of the music, conveyed through music.

Below I explore the strengths of analytical recomposition by fusing together two nineteenth-century musical settings of Goethe’s poem "Ein Gleiches," blending features of both songs in a characteristically Romantic style. The two settings are by Fanny Hensel (1835) and Robert Schumann (1850). I choose them for two reasons. My first reason is simple (and, admittedly, a little promotional): the songs deserve to be better known. If you know one setting of this poem you likely know Schubert’s version from around 1823. Schumann’s setting has received passing mention in the scholarly literature,⁷ and occasionally been programmed on recitals, but it is hardly revered in the way that Schubert’s is. (The Schumann scholar Eric Sams can only bring himself to say that it is "not negligible."⁸) Fanny Hensel’s setting is barely known at all; nor was it known during her lifetime, except by her inner circle, since the song wasn’t published until 1993.⁹

My second, more important reason for choosing these songs is that they are strikingly similar—similar enough, in fact, to give the impression that Schumann was explicitly modeling his song upon Hensel’s. Except that he couldn’t have been: Schumann almost certainly didn’t know Hensel’s song. Hensel and Schumann did meet one another, when Fanny traveled to Leipzig in 1843 to see her brother, and Clara Schumann apparently performed her husband’s compositions for Fanny. But we can’t know for sure if Fanny shared any of her music with Clara or Robert. In any case, the encounter didn’t make a strong impression on either composer: in a diary entry Fanny said that Schumann’s pieces "did not seem very beautiful to me" (mir nicht sehr schön vorkamen);¹⁰ Schumann also wrote little of the encounter, saying only that the depth of Fanny’s mind spoke through her eyes.¹¹ The connection between these songs, then, is likely not a matter of influence.

Yet it’s also not a matter of coincidence. I don’t mean that Schumann was clairvoyant, rather that the similarities between these songs are a function of the similarities between the composers’ styles. One of my main aims in fusing together these two songs (in addition to exploring the benefits of recomposition as an analytical strategy) is to highlight those stylistic similarities, which thus far haven’t been explored. I have often thought that Hensel’s songs sound much more like Schumann’s than Schubert’s, despite that Hensel seems to have known Schubert’s songs better. Much of this has to do with the composers’ similarly fluid sense of tonality—their tendency to undermine the status of the tonic (i.e., the home chord in any given key), not necessarily with the distant modulations that we see in Schubert, but with a kind of harmonic evasiveness, where keys are suggested but not fully articulated and cadences are subtly undermined.

This evasiveness is on full display in the songs in question. Hensel and Schumann make use of a compositional strategy that appears again and again in their songs: they write endings that call into question the very category of endings; their songs reach tonics that don’t feel like moments of closure. Many scholars have commented on this aspect of Schumann’s songs. Two give just a couple examples: David Ferris writes of Schumann’s characteristically "open endings," where (as he puts it) "there is closure at the end of the song, but its conclusiveness is compromised in some way,"¹² and Susan Youens describes how in his song "Warnung," op. 119, no. 2, Schumann seems to have set himself the task of "mak[ing] the tonic chord un-final."¹³ However, Hensel’s means of creating "inconclusive closure" and "making the tonic chord un-final" have received much less attention, even though they are just as inventive as Schumann’s, if not more so.¹⁴ A comparison of two songs of course isn’t enough to show the variety of Hensel’s and Schumann’s "open endings." But it can reveal how these two composers—who were acquainted but hardly intimate, aware of one another but not influenced by one another—relied on a favorite maneuver to tackle the same compositional problem: namely, how to find a musical analogue for a poem about the rest that you sense in the world around you but that you can only experience fully in the future—how to write a song that ends but also doesn’t.

So, let’s get to some music. I’ll begin with Hensel’s song, using a bit of recomposition to show how she departs from harmonic expectations and thereby calls into question the stability of the home key. I’ll then do the same with Schumann’s song. After that, I’ll discuss my blended recomposition. Finally, I’ll close with some comments about the broader implications of this kind of analysis.

Example 1 shows how Hensel’s song might have begun, if she had done something much more conventional. (Each musical example is accompanied by an audio example, so that you can see the score and hear the music at the same time. Unless otherwise indicated, the performances were done by me.) I’ve left her melody the same and only changed the harmony. In my recomposed version the home key is clearly articulated, with a conventional I–V⁷–I progression. We start from a place of rest (a root-position tonic), depart from it, and then return to it. If my musical phrase were a linguistic sentence, it would start with a clear subject (the I chord), lead to a clear verb (the V chord that stretches over measures 2–3), and end with a clear object (the final I chord).

Hensel’s version, however, is much less clear, and less restful. This is what she actually wrote:

Hensel starts with a less stable form of the tonic (an E-major chord in first inversion, i.e., with the third of the chord in the bass) and, more crucially, ends not with another tonic chord but with a dominant chord of another key (a V of C-sharp minor). If her musical phrase were a linguistic sentence, it would start in the middle, as if already underway, and it would end without a sense of closure. And indeed that is what Goethe’s sentence does. The sentence that stretches across lines 1 and 2 of the poem—"Über allen Gipfeln / Ist Ruh,"—begins with a preposition ("über") and, although it expresses a complete thought, it ends with a comma; furthermore, the subject ("Ruh") falls at the end, not at the beginning. (The lines could well have been "Ruh ist über allen Gipfeln.") Hensel’s opening phrase is thus the perfect analogue for Goethe’s opening thought: her musical syntax matches his verbal syntax.

Aside from these syntactic parallels, what makes Hensel’s opening phrase so fitting is that it sounds at once restful (with its simple accompaniment, gentle rhythms, and normative four-bar structure) and restless (with its slightly unstable opening and unexpected turn toward a minor key). Moreover, it places the most dissonant chord on what is arguably the key word of the entire poem: "Ruh" (calm, rest)—the very thing that the addressee, or the "you" of the poem, seeks and senses in the world, but cannot experience fully.

The ending of the song is similarly evasive. Example 3 shows the last two phrases of the song as Hensel might have written them, again with the melody unaltered and only the harmony changed.

Notice the strong cadences in E major in the first measure of the second system (a perfect authentic cadence, abbreviated PAC, in which a root-position V⁷ leads to a root-position I with scale-degree 1 in the melody) and two measures before the end (an imperfect authentic cadence, or IAC, where scale-degree 3 appears above the root-position I). If I gave Hensel’s melody to my music theory students and asked them to harmonize it, this is likely what they would come up with—and they’d get an A.

In Example 4 you’ll see what Hensel actually wrote. The two moments of closure in the previous example are weakened. The perfect authentic cadence (PAC) in E major at the beginning of the second system has now become an evaded cadence (EC) in C-sharp minor, a moment where we expect a PAC but don’t get one. And the final cadence is no longer really a cadence at all: rather than scale degrees 5–1 in the bass we hear 2–1, and rather than a dominant chord we hear a kind of passing chord, a chord that lacks a leading tone altogether, which creates a sense of closure but not the full closure we would get from a true cadence. (I’ve circled the pitch that "should" have been a leading tone, D#, but is instead the tonic pitch, E.) To put it simply, at the end of the song the music is sort of at rest but not quite—or not yet, if we imagine that true rest lies outside the bounds of the song, beyond the final barline. (Example 5 provides the complete score to Hensel’s song, if you would like to see it in its entirety; an excellent recording of the song—by baritone Tobias Berndt and pianist Alexander Fleischer—can be found on YouTube: Example 5 (YouTube).)

Schumann’s setting of Goethe’s poem is hardly identical to Hensel’s. (The declamation, for example, is much more irregular—which, as Harald Krebs has shown, is typical of Schumann’s late songs.¹⁵) Still, it’s hard to escape the fact that the two composers interpret the poem similarly, setting it to music that is balanced on a knife edge between tranquility and unease. This is most noticeable in the way they set the end of Goethe’s poem. Like Hensel, Schumann ends without a real ending, poignantly underlining the addressee’s inability to find true rest. Here’s the ending of his song as it might have sounded had it been written in a more conventional mode. Listen for the two strong cadences on "auch" (IAC, then PAC).

This is far too boring for Schumann. In the actual version he undercuts both of these cadences. The first moment of undercutting isn’t too surprising. We hear a deceptive cadence (DC—that is, a move to a vi chord rather than a I chord) in place of an authentic cadence, but with a move to an A-major chord rather than an A-minor chord. The second moment, however, is startling. The bassline slides down to E, which supports a weak first-inversion tonic rather than the anticipated root-position tonic (resulting in an evaded cadence), and the melody, rather than descending stepwise to C and giving in to musical gravity, rises a seventh to a higher C.

How strange, and how wonderful, is this moment when the voice leaps upward and the bass slides downward! It suggests resignation, with the bass sinking down to a I⁶, not the tonic we were expecting but perhaps the best we can do under the circumstances. But it also suggests frustration, with the melody leaping dramatically upward, as though the addressee were looking up enviously toward the mountain peaks and the treetops—and the "Ruh" that envelops them. It’s true that Schumann’s song produces an authentic cadence in its final bar, but the voice doesn’t participate, and what sticks with me, what lingers after the song is over, is not so much this "successful" cadence as the more dramatically "unsuccessful" cadence that happened four bars earlier. The close of Schumann’s song, like the close of Hensel’s, fills me with a paradoxical feeling of restful unrest, or unrestful rest, suggesting that just as spiritual and emotional rest is not fully possible in the confines of the addressee’s life, so is musical rest not fully possible in the confines of the song. (For the entire score to Schumann’s song, see Example 8; for a recording by tenor Peter Schreier and pianist Norman Shetler, go to Example 8 (YouTube).)

Because of these similarities—and also the similar opening and closing melodic shapes: scale degrees 5–4–3 at the beginning and 5–4–3–2–1 at the end—blending the two songs together ended up being a relatively straightforward task. What I present here is of course only one blend among innumerable possible options, but it’s the one that I think best showcases (through music) the close kinship between the songs. I will keep my analytical commentary to a minimum, with the idea that the relationships may be easier to hear than to hear described. I will, however, provide a brief road map to my recomposition.

In the introduction and phrase 1, I begin with Schumann’s accompanimental texture and also his chords. I leave his chords basically unaltered, until the last bar of the first system, where instead of leading to the tonic, as Schumann had done, I borrow a harmonic device from Hensel and lead to the submediant (A minor, or vi). Above this I place Hensel’s melody, modifying it only slightly so as to fit Schumann’s chords (it helps, of course, that the first vocal phrase of each song outlines a descent from scale degree 5 to 4 to 3). I also introduce the characteristic triplet rhythm of Hensel’s melody.

In phrase two (measures 8–14), I use Schumann’s texture and chords—until two measures before the end of this system ("Hauch"), when I basically copy Hensel’s texture and chords exactly, setting the bars in minor rather than major; fortuitously, Schumann’s song modulates to the tonic of Hensel’s song (i.e., it goes from C major to E minor, and Hensel’s song is in E major), so the temptation to basically cut and paste Hensel’s music at this point was too strong to resist. The melody in phrase two is a combination of Hensel’s and Schumann’s melodies: it starts with the stepwise descent (A–G–F–E; "Wipfeln spürest du") and leap up (to C, on "kaum") that we saw in Schumann’s song, and ends more or less identically to Hensel’s melody, save the change to minor mode.

I take a few liberties in phrase three, but the end result is that this passage evokes the texture and chords of Schumann’s middle section and places a modified version of Hensel’s middle-section melody above it, fitting it to the harmony.

Finally, phrase four does something similar to what I did at the beginning of my recomposition: it uses Schumann’s chords and Hensel’s melody—again, the similar 5–4–3–2–1 melodic shapes made this fairly easy. At "du auch" in the fourth and fifth measures of this system, however, I use Schumann’s surprising upward leap and Hensel’s brand of evaded cadence (V/vi—vi⁶). In the last three bars I borrow Hensel’s melody and chords and dress them in Schumann’s accompanimental texture.

As you listen to the recording of my recomposition, you may choose to follow my annotated score, tracking how I fuse together different aspects of each song. You may also choose to just listen without a score, experiencing a piece that (if I’ve done my job right) sounds like a 50/50 blend of both songs.

What does an "analysis" like this teach us? In closing, let me suggest four ways that our music analyses can be enriched by the use of recomposition.

1. Recomposition allows for a kind of immediacy, since it requires concretizing something that might seem rather abstract. Instead of only describing the relationships between the pieces, I am forced to demonstrate them, via a combination of composition and performance. As a result, analytical recompositions may be particularly useful for conveying musical ideas to a lay audience, who can hear the connections in real-time performance but may get lost in the notes or in overly complex analytical terminology.
2. Recall that I said this recomposition was not terribly challenging to create, because the songs were written around the same time, using a similar aesthetic, and because the links between them are so strong. This is no small point. Any attempt to fuse together two pieces or passages—two settings of the same poem, two sonata-form transitions, two structural cadences, two fugue subjects—will necessarily reveal precisely how similar the two pieces of music are. Where the task is challenging—as, for example, with Matthew BaileyShea’s virtuosic fusing of several different songs from several different eras—it will vividly reveal the extent of the gulf between the pieces, and it will require us to specify, more than we might do in a more conventional comparative analysis, what musical elements are incompatible and why. Where the task is easy, maybe even easier than anticipated, it can reveal stylistic connections that we may have sensed but been unable to articulate.
3. Recomposition seems to me an especially powerful tool for the analysis of art song. Songs, after all, are often described as "interpretations" of poems—one might even say "analyses" of them, expressed via the medium of music. Attempting to recompose parts of a song—whether making them more normative or more unorthodox—can help us to specify exactly why a given musical "interpretation" of a poem works the way it does; it gives us something specific to compare the song with (especially in the absence of another setting of the same text by a different composer or the same composer).
4. Finally, using recomposition as an aid to analysis can restore to analysis an element of play. Because recomposition involves composing and playing, rather than just looking and listening, it can help us and our students to learn a given musical style not by only reading about it—like learning German by merely memorizing vocab lists and common phrases—but also by using it, becoming fluent in a musical idiolect by "speaking" it, with our voices, with our hands, and with our entire bodies.

Notes