Lost in Translation, starring David Lewin and Four Singers; or, What Happens at the End of Schubert’s “Morgengruss”?

Benjamin Binder (Duquesne University)

David Lewin’s legendary 1974 monograph-length analysis of Schubert’s “Morgengruss” (Morning Greeting)—the eighth song of Die schöne Müllerin, D. 795 (1823)—has at long last been published (posthumously, in 2015), and I’ve recently written an extensive review (forthcoming in Nineteenth-Century Music Review).  The analysis is magnificent:  engaging, insightful, provocative—everything we’d expect from Lewin writing on music and text.  But, in my review, I express some concerns about the relationship in his argument between analysis, performance, and interpretation, particularly when it comes to understanding the song’s conclusion.  Here I’d like to go a little farther with that critique.

“Morgengruss” opens with the miller lad greeting the object of his desire from beneath her window, at which point she immediately turns away; thereafter, he stands from afar and cajoles her to reappear.  The final lines of the song cap off his urgent plea, but the precise meaning of Müller’s words is impossible to pin down:

Die Lerche wirbelt in der Luft,
Und aus dem tiefen Herzen ruft
Die Liebe Leid und Sorgen.

For the first line, I would give, as most translations do, something like “the lark warbles in the air.”[1]  For the final two lines, the grammatical subject has to be “die Liebe,” since the verb “ruft” is singular.[2]  One could simplify the word order as follows:

Und die Liebe ruft Leid und Sorgen aus dem tiefen Herzen.

“And love calls pain and sorrow from the depths of the heart,” more or less.  But two unanswerable questions remain.  First, whose love and heart are we talking about—the miller lad’s, or the maiden’s?  Second, is this love calling forth pain and sorrow (that is, bringing them into existence), or is it calling those feelings away (that is, getting rid of them)?

As I discuss in my review, there are many plausible interpretations and translations here.  Lewin, however, doesn’t draw any attention to the ambiguity of the lines and presents his own translation as definitive:  “and, from the heart’s depths, call love’s pain and cares.”[3]  For Lewin, the whole point of the song is for the miller lad to finally recognize that the turbulent pain and cares that he bears within his loving but unrequited heart are what might have scared the maiden away at the beginning of the song.  To greatly oversimplify Lewin’s analytical argument:  on our first hearing of Schubert’s strophic setting, the moments in the melody that grab our attention most are the striking high F in m. 9 and the E in m. 16 to which it ultimately resolves, as the miller lad agrees (temporarily) to leave the maiden’s window.  But by the time we hear the final strophe, our attention to the F–E connection fades, and we focus more on the D that concludes the first phrase of the song in m. 10, not to be resolved definitively until the C in m. 17 (on “Sorgen”).  In Lewin’s reading, then, Schubert’s miller lad comes to realize by the end that his own “Leid und Sorgen” indeed pulled him toward the maiden’s window in the first place.

Morgengruss Score

Singers performing this song have to decide for themselves how to translate the final two lines and express their own dramatic conception of the song’s conclusion.  As Lewin himself reminded us many times, a performance is itself an analysis, an interpretation, and can articulate musical and dramatic insights and perceptions at least as well as the written word.  So I find it striking that Lewin never mentions any performances that might have influenced his own particular hearing of the song, especially its ending.  That leads me to wonder how Lewin’s interpretation might sound in performance, and moreover, what performances Lewin might have had in his ear when he drafted the Morgengruss essay in 1974.

It turns out that Dietrich Fischer-Dieskau and Gerald Moore released a recording of Die schöne Müllerin in 1972, and to my mind, that performance translates the end of the song in much the same way that Lewin does.  Fischer-Dieskau makes a clear contrast between the bright, tremulous energy of the lark in mm. 12–13 and the languorous depths of the heart in mm. 14–15.  By connecting mm. 14–15 to mm. 16–17 without a breath, and making a gradual crescendo of volume and expressive intensity through all four bars that climaxes on “Sorgen,” Fischer-Dieskau’s miller lad releases the anguish stored within his loving heart as it now cries out to the maiden—“and, from the heart’s depths, call love’s pain and cares,” as Lewin would have it.

But there are also other ways of translating the ending, and performers can show us the way.  Consider Olaf Bär’s 1986 recording with Geoffrey Parsons.  Bär’s lark also “warbles” with vitality, and he infuses a bit of plangent despair into his rendering of “tiefen Herzen.”  But by taking a breath before “die Liebe,” Bär highlights the importance of that word and its function as the subject (and perhaps even the goal) of the sentence.  There is no crescendo into “Sorgen,” and most crucially of all, Bär and Parsons treat the repeat of the phrase in mm. 18–19 as a gentle echo.  Here I cannot help but hear the echo as a tender, inward lullaby—“trust in love, for it will sing the heart’s pain and sorrow to sleep,” Bär’s miller lad seems to say, perhaps to himself, perhaps to the maiden.  By returning to a fuller sound for mm. 19–21, Bär ruefully acknowledges the existence of the heart’s “Leid und Sorgen,” but this doesn’t invalidate his basic message:  love will call pain and sorrow away from the heart.  This is rather the opposite of Lewin and Fischer-Dieskau’s translation.

The most striking thing about Michael Schade and Malcolm Martineau’s 2005 recording is the way they handle that fermata on “ruft,” just before “die Liebe.”  Like Bär, Schade takes a breath, but it is animated by a sudden, vigorous impulse that leads purposefully into “die Liebe” in m. 16.  The shift to flowing triplets in the piano also contrasts more sharply with Martineau’s hushed, hesitant duplet eighths in mm. 14–15, lending a new energy and an optimistic shine to “die Liebe Leid und Sorgen.”  Even more than Bär and Parsons, Schade and Martineau make “die Liebe” into the focal point of the song’s conclusion, rather than “Sorgen.”

Jonas Kaufmann brings an entirely different sort of vocal instrument to “Morgengruss” in this performance with Helmut Deutsch from 2009.  Kaufmann doesn’t use the full intensity of his Heldentenor here, but there’s still something of the suave operatic leading man in the way he handles “die Liebe Leid und Sorgen”:  a kind of heroic, ringing quality throughout the entire phrase that makes me feel as though this miller lad is singing directly to the Müllerin here, promising to take away all her pain and sorrow if she would only give in to the promptings of love and fall into his arms.  Kaufmann doesn’t take a breath before “die Liebe,” but he blooms on that word after restraining himself in the previous phrase, and the way he lingers on the “L” in “Leid” creates a certain resistance which he can then valiantly overcome by the time he gets to the end of the phrase at “Sorgen.”  Kaufmann’s final “Leid und Sorgen” in mm. 19–21, like Bär’s mm. 18–19, are quieter and more inward, and I can’t quite decide if his miller lad is briefly retreating back into his own pain or imagining and depicting for the maiden the way that both of them will feel their pain dissolve once their love is consummated.  Like all of these performers, then, Kaufmann shows (to me at least) that more than one reading is possible, and multiple, even contradictory emotional and dramatic meanings can emerge from the same performance.

To finish up, let’s return to Fischer-Dieskau, still with Gerald Moore but this time back in 1951 when he was about 26 years old.  Fischer-Dieskau seems to be the only one of these performers who sides strongly with Lewin’s translation.  Did he always read the words that way?  From this performance, I’d say yes.  Certainly Fischer-Dieskau’s younger voice conjures up the dewy, tender Jüngling here, rather than the more adult and assertive vocal persona of the 1972 recording.  But as in the later performance, the focus is on “Sorgen.”  Fischer-Dieskau again takes no break before “die Liebe,” and while there is no crescendo into “Sorgen” in mm. 16–17, there is a hint of delay before that word, and the phrase moves steadily towards it, with no lingering or special emphasis on “Liebe.”  Instead, the young Fischer-Dieskau saves the crescendo for the repeat of the phrase in mm. 18–19, releasing onto “Sorgen” after a great deal of anxious vibrato.  His treatment of the final “Leid und Sorgen” in mm. 19–21 is the quietest and most inward of all these performances.  I see the miller lad just standing there, immobilized by his pain and loneliness, with no Müllerin anywhere to be seen—for Fischer-Dieskau, this moment isn’t really about her.

These are just my reactions to these performances, of course.  No performance can be translated definitively, just as no text can, musical or poetic, especially the texts of Müller and Schubert’s “Morgengruss” that I’ve been considering.  My little point is that performers and performances ought to be in the conversation from the get-go when we’re translating or explicating song lyrics, especially ones that are particularly ambiguous.



[1] A more contemporary meaning of “wirbeln” is “to circle,” but given the simile between the lark’s warbling—a morning call like the poet’s “Morgengruss”—and the heart’s call, both acoustic events, I’m inclined to stick with “warbles.”  Contemporary support is found in J.C. Adelung’s 1811 Wörterbuch.

[2] If “die Lerche” were the subject, we would expect the word “sie” in the second line and commas in the third:  “Die Liebe, Leid, und Sorgen.”

[3] Problematically, Lewin’s version of the text gives “der Liebe” instead of “die Liebe,” a reading with very little, if any, philological support.  In fact, “der Liebe” is ungrammatical; for “Leid und Sorgen” to be the grammatical subject of the final two lines would require a plural verb:  “rufen,” not “ruft.”  However, Lewin’s translation is still plausible even if the words are “die Liebe” instead of “der Liebe.”


PAIG Meeting at 2017 SMT Conference

We are looking forward to seeing you soon in Arlington! The PAIG meeting will take place on Friday, November 3, 12:15–1:45 in Studio A. Our program will comprise three short papers (chosen through a blind review process) followed by a business meeting and general discussion. Thank you to everyone who submitted proposals and/or assisted with the review process!

  • 12:15  Welcome
  • 12:20 “Using Embodiment Schema to Help Student Performers Relate to Their Theory Work” (Bonnie McAlvin, CUNY Graduate Center)
  • 12:40 “Three Case Studies In Search of Holistic Performance Research” (Jonathan Dunsby, Eastman School of Music)
  • 1:00 “Paradox of Interpretation and the Resolved(?) Dualism” (Wing Lau, University of Arkansas)
  • 1:20 Business Meeting and General Discussion
    • We hope to keep the Business Meeting very brief so as to use the remaining time for some General Discussion. Members are invited to share a bit about their current work, recent publications, or other interests so members of the PAIG community can get to know one another.

Please scroll down to the bottom of this post for abstracts for all three papers. During the general discussion,


This brief talk illustrates how embodiment schema (Lakoff and Johnson 1980) can be used to engage performance students in their music theory work, by bridging students’ analytical work to the development of their narrative imagination and mastery of grouping and expression. An important function of analysis is that of revealing potential groupings. In many cases, groupings map readily onto embodiment image schema such as UP/DOWN, CONTAINER, and BARRIER. We are able to empathize with motions UP, DOWN, IN and OUT, so groupings which map onto these schema are in turn readily anthropomorphized. An approach to a passage might ask whether a middleground reading moves UP or DOWN, and what the anthropomorphized pitch-group hopes to achieve in moving UP/DOWN? Contextualizing the slope is part of this work: is the anthropomorphized pitch-group navigating a slippery cave, battling a current, or wafting a hillside? As students master analytical tools, mappings can become sophisticated. What awaits the pitch group at the end of the PATH: a tonic chord or a betrayal? Which of the steps UP/DOWN are in the diatonic CONTAINER, and which are chromatic? Of the scale degree steps, which are part of the governing harmonic CONTAINER? How does each step relate to the various metric CONTAINERS and how does this enliven the narrative? The talk suggests some group activities that use embodiment schema mapping to prime students toward higher engagement in the intricate analytical work of harmonic, voice-leading, Schenkerian, and set class analysis.



I offer three short case studies:

1) a critical aspect of microtiming studies, specifically Llorens’ recent article researching onset asynchrony in Brahms performance, with a minimum perceptibility threshold of 100 ms;

2) Gould’s ‘prelude’ to Webern’s Op. 27i, which to my knowledge has not been studied theoretically before, despite the wealth of recent performance research on the Piano Variations; and

3) another look at Schenker’s graphic analysis of Chopin’s Fourth Prelude and  Schachter’s durational reduction, in light of actual ‘readings’ old and new (for example here, Pollini’s and Trifonov’s).

I briefly air my position that what typically appears in theoretically-informed performance studies is a disassembled practice, the investigation of only those elements of a performance that can be measured, assessed, or manipulated, always in danger of being called ivory-tower rather than real-life. Equally, the concentration on surface analysis, on the results of the application of what followers of Lewin call the ‘technology’ of music theory, may be considered to have evaded a concrete engagement between interpretive practice and musical meaning in its deepest, inclusive senses.

In the first two case studies we see Kerman’s ‘positivism,’ through, first, excusable, totalizing assertion of perceptual pertinence, and, secondly, the understandable exclusion of obscure but intriguing creative evidence. The second two cases show how an engagement with precompositional materials, or structurally remote features of the ‘inner’ form, can be part of interpreters’ volitional agency and affect all parameters of the sounding score.


The paradox of interpretation, the belief that performance serves the work and the composer but the performer’s subjective contribution is inevitable, has a long history in performance discourse (Rosen 2002 and Cook 2013). The attempt to reconcile such paradox, the notion “to play as if from the soul of the composer” represents an early Romantic subjectivity (Hunter 2005), a resolved dualism that still influences the classical music culture today.

In this paper, I suggest that the attempt to reconcile such paradox, the resolved dualism, is paradoxical in and of itself. I illustrate discrepancies between the recordings of Artur Schnabel, a pianist who strives to reach the “same free spiritual height” as the composer, and his interpretative rationales documented by his students in Wolff 1979. In addition to the dissonance between his rationales and the interpretive choice documented in his recordings, I demonstrate the dissonance between his rationales and his perceived intention of the composer. I ask, can one be truly immersed in the soul of another? Can we not argue that the composer’s supposed intention is actually Schnabel’s own subjectivity masked behind authenticity, consciously or not? As this paper shows, the performers’ subjectivity and their perceived intention of the composer could be hard to distinguish, and that the resolved dualism could be more apparent than real. Recognizing this paradox sharpens our sensitivity towards performers with the authentic reputation and help us better understand the epistemology of the faithful performance.