Comparative Literature
Nicholas Swett

A Theory of Musical Translation

  • Faculty Advisor

    Reginald Gibbons

  • Faculty Advisor

    Vasili Byros

  • Faculty Advisor

    Hans Jorgen-Jensen

Published On

May 2017

Originally Published

NURJ 2017-18
Honors Thesis

In his book On the Beautiful in Music, nineteenth-century music theorist and critic Edouard Hanslick claims that music “is a language that we understand and speak, but that is impossible for us to translate” (Szendy 49). He proposes that it is impossible to express the purely musical meaning that the language of music expresses in any verbal language. We could contradict this, pointing out instances where music contains quite specific designative meaning that has been “translated” into verbal language for us by the composer in the form of a narrative program accompanying the music. However, we can certainly grant Hanslick the belief, still held by many music theorists today, that music is simply not specific enough in its designations to be fairly rendered in verbal language. Even if we accept Hanslick’s assertion that the abstraction of music renders it untranslatable for us, his statement leaves one question unanswered; that is, whether music can be translated into other, equally linguistically abstract music.

Philosopher and musicologist Peter Szendy offers an answer to this question in his essay “Writing Our Listenings: Arrangement, Translation, and Criticism.” He argues that musical transcription, the setting of music to a new instrumentation, is a method of intra-musical translation. In this article, Szendy makes a strong case for why we should explore and define musical translation. He points out that in reading a translation, we observe the translator in a critical process – we are “reading their reading” (Szendy 50). He sees musical translation in the same light – that “arrangers have so many things to tell us…about what listening means” (Szendy 50). He defends transcribers admirably, but to do so relies exclusively on Walter Benjamin’s essay on translation, The Task of the Translator. Specifically, Szendy proposes that an original piece and its transcription together bring a listener closer to experiencing an ideal musical form, the “Work,” which is the unreachable amalgamation of all performances and realizations of the original that exist. Jacques Derrida and Robert Eaglestone, who maintain that either this ideal form doesn’t exist, or it is something that we can never experience and so it is not worth pursuing, present an alternate school of translation theory that throws Szendy’s perspective into question. In short, Szendy fails to give us an altogether convincing theory of musical translation, and so we must look elsewhere to develop a more sustainable one.

To do so, we can build upon the work of music theorist Joseph Straus in his article “Recompositions by Schoenberg, Strainsky, and Webern.” He does not discuss translation outright, but suggests that transcription alone is not a radical transformation of musical language. He instead points to examples of re-composition, a practice where a composer not only transcribes a work for a new instrumentation, but also changes the very organizing principles behind the music. In his books Emotion and Meaning in Music and Style and Music: Theory, History, and Ideology, music theorist Leonard Meyer defines periods in the history of music that can be distinguished from one another by looking at such organizing principles, what he calls “Style Systems.” Meyer’s style systems function much like the continuum within one language implied by poet and translator Ezra Pound in his essay “Guido’s Relations.” I propose defining musical translation as a re-composition from one point to another on this a continuum of style systems, rather than in terms of simply a change of instrumentation. This definition encourages us to look at twentieth-century adaptations of pieces from the musical past such as Stravinsky’s Pulcinella as translations. The result of such an investigation reveals that musical translations serve multiple functions: they can, like literary translation, bring a work and composer from a distant period, which might have faded into obscurity, to an audience. But they can also serve illuminate, define, and even create the musical language of the translator.

 Peter Szendy is one of the only musicologists to tackle the question of intra-musical translation directly. In “Writing Our Listenings: Arrangement, Translation, and Criticism,” Szendy makes a strong case for why we should explore and define musical translation. He points out that in reading a translation, we observe the translator in a critical process – we are “reading their reading” (Szendy 50). He sees musical translation in the same light – that “arrangers have so many things to tell us…about what listening means” (Szendy 50).

He uses Franz Liszt’s transcriptions of Ludwig van Beethoven’s symphonies as a case study in musical translation. In transcribing Beethoven’s iconic symphonies for the keyboard, Liszt proclaimed “I will be satisfied if I have accomplished the task of an intelligent engraver, the conscientious translator, who grasps the spirit of a work along with the letter” (Szendy 48). When he transcribed Symphony Fantastique by Hector Berlioz, he used similar language: “I scrupulously tried, as if it were a matter of translating a sacred text, to carry over to the piano, not only the musical framework of the symphony, but also the effects and the details” (Szendy 48). Szendy finds evidence for the analogy to translation in Liszt’s score of his Beethoven transcriptions. He points out that Liszt includes instructions as to the orchestral instruments that lines in the piano score represent. Of course, it is not possible to recreate the timbre of an oboe on the piano, and so Liszt must have been pointing to a sense of absence in his piece—indeed, to a sense of lack or loss with relation to the original text that is inevitable in any translation. Szendy argues that similarly, when Liszt provides two options full of different notes from the original that the performer must choose between, the impossibility of transferring all of the material that was playable by the orchestra to the new “language” of the piano becomes starkly clear.

Szendy moves a step further in his arguments when he begins to discuss Robert Schumann. Schumann said that he wanted his Six Etudes After Paganini Caprices “to give the impression of an autonomous piano composition, which would make one forget its violin origin, without the work losing any of its poetic Idea” (Szendy 64). This reflection of Schumann’s seems quite reminiscent of the task of the literary translator, who hopes to move the poetic effect of a piece from its native language to another. Here, Szendy relies quite heavily on Walter Benjamin’s theory of literary translation. Benjamin, in “The Task of the Translator,” suggests that languages depend on one another in their aim of expressing the pure language that lies behind all linguistic content. He says “kinship of languages rests in the intention underlying the language as a whole—an intention…which is realized only by the totality of their intentions supplementing each other: pure language…In translation the original rises into a higher and purer linguistic air” (Benjamin 77). Szendy argues that changes in instrumentation have the same ambition and effect; that in re-writing an original piece, transcribers attempt to move that piece, in the way that Benjamin proposes, closer to a higher and purer form: the timeless “Work” that is the ultimate, unattainable goal of all art.

Szendy does not discuss the opposing camp in translation theory, which does not subscribe to Benjamin’s notion of pure language and ideal forms. In his article “Levinas, Translation, and Ethics,” Robert Eaglestone lays out two possible approaches to looking at the relationship between languages, one of which is that of Benjamin. The theorist whom he focuses on, Emmanuel Levinas, is also in Benjamin’s camp. They both maintain that there is a pure, universal language that all languages we speak aspire to, and which translations help us to access (Eaglestone 135). In the opposite corner sit translation theorist Jacques Derrida, who takes for granted the “irreducible multiplicity of tongues,” (emphasis mine) and cultural anthropologist Clifford Geertz, who asserts that “what man is may be so entangled with where he is, who he is and what he believes that it is inseparable from them” (Derrida, 218);(Eaglestone, 134). These theorists argue that we simply cannot strip away the trimmings of language and culture to arrive at “pure language.” They suggest that all linguistic, literary, or musical acts are irreducible; if we translate them, it does not therefore bring them closer to this illusion of universal truth. Rather, in their model, each translation would become its own cultural entity—perhaps an analogy to the original that can serve to positively inform relationships between cultures, but not something abstractly and spiritually linked to its source.

Szendy claims that transcriptions help the music they recreate to move closer to achieving an abstract, timeless form. However, we could just as easily see these pieces not as evidence of the limitlessness of an abstract masterwork of music, but as examples of the particular limitations of individual instruments. Szendy allows Liszt’s statement about the “unlimited development” of the “harmonic power” of the piano to stand: “it can produce, with few exceptions, all the features…and leaves to the orchestra no other superiorities (although these are, it is true, immense) than those of the diversity of timbres and the effects of the massed forces” (Szendy, 59). But this is of course untrue; Liszt’s transcriptions make painfully clear the inability of the piano to sustain notes and to tune certain sections expressively, two features not included in “timbre and effects of massed forces.”

Equally problematic in Szendy’s essay is the audibility of the “translation” he is referring to. Most of the elements that Szendy cites as evidence for “translation” are things that are written into the score. For example, Liszt’s transcription of the iconic opening of Beethoven’s fifth symphony includes the indication “instruments a cordes et Clarinettes” or “string instruments and clarinets” (Liszt, m. 1). This sort of instruction for the performer, according to Szendy, reveals the distance between this new text and the original. However, these markings, and therefore this crucial distance, would not be audible to the audience. Reading the score to Liszt’s transcriptions is central to experiencing them as translations. And therefore, without extensive score study, it is difficult for the audience to “listen to [Liszt] listen” (Szendy, 60). We do hear him hearing the piece on the piano rather than for an orchestra, but all of the crucial philosophy and criticism that Szendy wants us hear in Liszt’s work is perceivable only to the pianist and the scholar. According to Szendy, the importance of musical translation, in the end, lies in Karl Wilhelm Friedrich Schlegel’s conception that “a judgment about art that is not itself a work of art has no validity in the realm of art” (Szendy 65). He uses Schlegel’s idea of critical judgment in art to emphasize his point that hearing transcription is a way to engage with music criticism and a piece of music simultaneously. But if we cannot hear the critical import of Liszt’s piece—which we can fully experience only through extra-musical score study—then we have not encountered a work that can be fully experienced as a translation. In short, Szendy’s approach to musical translation is both limited in its applications and held back by its narrow frame of reference in literary translation theory.

In his essay on re-compositions of the early twentieth century, music theorist Joseph Straus gives a different perspective on eighteenth- and nineteenth-century transcriptions. He brings up J.S. Bach’s transcriptions of Antonio Vivaldi’s music, as well as Liszt’s transcriptions of Beethoven. In order to differentiate the twentieth-century re-compositions he focuses on, including Stravinsky’s Pulcinella, from transcriptions, Straus claims that “the historical distance between the re-composer and his model has become considerably longer…spanning deep stylistic gulfs. Bach and Vivaldi and, to a lesser but still significant extent, Liszt and Beethoven had a common musical language” (Straus, 302). While there is certainly a difference in sound between Beethoven’s orchestra and Liszt’s piano, the stylistic differences are perhaps not great enough to call it translation. Instrumentation alone cannot always define a musical language, since for much of the history of Western Music, the arsenal of instruments has been relatively constant. In order to argue for a change in musical language, we would need to define how wide a “stylistic gulf” is required to call a musical transformation a translation.

In his book Emotion in Meaning in Music, music theorist Leonard Meyer concerns himself with the nature of the abstract meaning that is produced by music. He proposes that affect in music is generated by the affirmation or negation of expectations that are conditioned as “habit responses” to musical stimuli. He contends that we can identify a particular piece with a particular set of “habit responses” by considering stylistic probability. Pieces of music are made up of musical patterns on various levels of scale – from individual notes, to clusters of notes, to phrases, to entire sections and movements. Meyer argues we should determine the frequency with which certain patterns occur together throughout the entire corpus of music that was written, and therefore heard, at a given period of time. If the relationship between these elements is indeed probable within a time period, then that relationship is a sign that the piece belongs to that system of habit responses, what Meyer terms a “style-system.”

Elements of style are constantly peaking and receding as they gain and lose habitual relevance to the musical idiom of the time. We can find an almost infinite number of peaks and recessions when looking at the frequency of both small units of music and larger musical structures. Since this process is continuous, we can draw the lines between different style-systems wherever we choose. After large amounts of time—which create the “stylistic gulf[s]” along this continuum that Straus referred to—the difference between style-systems can become quite pronounced, even dramatic. This series of style-systems is quite similar to the history of a literary tradition within one language that has also been transformed over time.

In his essay, “Guido’s Relations,” Ezra Pound discusses his aesthetic priorities in translation. He claims that in order to write well, he must get rid of “the crust of dead English, the sediment present in my own available vocabulary” (Pound 85). He qualifies this statement by saying that one cannot “learn English, one can only learn a series of Englishes” (Pound 85). Later in the essay, Pound translates a poem by thirteenth-century Italian poet Guido Orlando into two different Englishes: his own, and pre-Elizabethan English. He shows that the differences between the “languages” within one language can be quite significant, and argues that this difference is particularly dramatic in English. He proposes that “Guido’s thirteenth-century language is to twentieth-century Italian sense much less archaic than any fourteenth-, fifteenth-, or sixteenth-century English is for us” (Pound, 92). While this comparison is not something that we should be so easily convinced of, Pound does illustrate well that one language can be viewed as a rapidly changing continuum that in fact includes different versions of itself. Pound’s essay provides a foundation on which we can build an understanding of contemporary English-language style-system translation.

We recognize the temporal distance between different English “style-systems” when we translate the tenth- and eleventh-century Old English of Beowolf or the fourteenth-century Middle English of Chaucer’s Canterbury Tales. The editors of “No Fear Shakespeare” see the same kind of stylistic gulf when they choose to translate the Bard’s sixteenth-century Elizabethan English, which is technically considered the same language as the English we speak today. Translation is in fact quite common along Pound’s continuum.

In the same way that we find examples of translation along the continuum of English, we can look for examples of musical translation from one style system to another. One candidate is Stravinsky’s Pulcinella, a collection of re-compositions of music by Giovanni Battista Pergolesi.

In writing Pulcinella, Stravinsky worked with a number of pieces by early eighteenth-century Italian composer Giovanni Battista Pergolesi, including several sonatas for two violins and harpsichord and various arias for voice and keyboard accompaniment. He arranged these pieces for a full orchestra, in order to accompany a ballet, and in most movements he maintained a clear measure-to-measure correspondence with the original. However, in many cases, he inserted a number of new notes and harmonies.

In his essay on recomposition, Straus discusses several elements of Stravinsky’s Pulcinella. He mentions Stravinsky’s aggressive characterization of his relation to his source text—the “metaphors of force, penetration, and possession” that the composer employs in his autobiography when he talks about reworking the pieces by Pergolesi (Straus 313). Stravinsky asked himself, “Should my line of action with regard to Pergolesi be dominated by my love or by my respect for his music? Is it love or respect that urges us to possess a woman? …In order to create there must be a dynamic force, and what force is more potent than love?” (Straus 313). Stravinsky’s own language with respect to his source material is violent to an uncomfortable extent. It reeks of misogyny and sexual violence. However, in using such language, he reminds us of the relationship between a translator and a source text, which is often one of struggle and conflict.

As George Steiner states in his essay “The Hermeneutic Motion,” after the first of four stages of translation, where we recognize value in a work we hope to translate, we begin to violate it: “After trust comes aggression…the translator invades, extracts, and brings home” (Steiner, 194). The struggle that Stravinsky’s words express relates his re-composition process in music to the work of literary translation, albeit through a quite masculinist and appropriative lens.

In Steiner’s four stages of translation, after violence comes reconstruction, and finally, a gesture of reconciliation and even reparation on the part of the translator. He argues that in this final stage, the translator can and should give something back to the original work or the culture from which the work came. All of the composers that Straus discusses in his essay claim in one way or another that their re-compositions highlight the elements of the original piece that make it musically valuable. Through their acts of “translation,” they ostensibly aim to honor the original and make it more accessible to their audience. However, Straus concludes that the violence wrought on the original piece is greater than this recuperation – that these “translations” highlight not homage to but rather conflict with past composers. They fight to overcome styles and composers of the past rather than celebrating or proliferating their works (Straus, 328).

Straus describes in detail the musical effect of Stravinsky’s changes in harmony. In the words of Straus, the subtle changes in the Serenate movement of Pulcinella “have the effect of super-imposing a new dimension of musical organization upon the old one” (Straus 316). Stravinsky adds an insistent repeating bass figure in the accompaniment part of the aria he bases this movement on. This repeated bass creates a new harmonic structure – in Straus’ words, a “characteristic Stravinskyan touch” (Straus 317). In short, Stravinsky has moved the piece out of Pergolesi’s style system and into his own – for better or for worse, he has translated Pergolesi.

Other movements of the Stravinsky illustrate this transformation of style even more clearly. Stravinsky arranged Pulcinella into Suite Italienne, a selection of movements from the ballet played by Cello and Piano. In doing so, he in fact makes the instrumentation closer to Pergolesi’s originals, while the harmonies, and so the language, are still securely in Stravinsky’s style. The penultimate movement of Suite Italienne, the “Minuet,” presents the full source melody twice. The first time through, Stravinsky’s additions to the harmonies are minimal. The main modification in this first presentation of the melody is the addition of a few drone pitches that keep the music from driving forward through harmonic progressions as the traditional tonal motion of the original would. The second time through the melody, the alterations are stronger. Stravinsky adds accented and plucked chords in the cello part that do not adhere to the eighteenth-century principles of smooth accompanimental voice leading. These insertions are full of surprising harmonies. Additionally, the second presentation modulates to an unexpected key and concludes there, completely altering the harmonic structure of the original aria. In short, while the first iteration of the melody sounds like a quirky arrangement, the second time through sounds like a true translation into Stravinsky’s language. This movement in particular allows the listener to hear in succession both the original, or something quite close, and a translation, the same way we might see in a book of literary translations the original and translation en face.

All of the changes above show Stravinsky’s Pulcinella to be a good example of a musical composition that we can view as a translation from one of Meyer’s style-systems to a later one. A more comprehensive survey of twentieth-century pieces that are deliberate reconstructions or “re-hearings” of earlier works will likely yield more and various examples of works that could fall under the umbrella of musical translation. Stravinsky’s piece, other re-compositions like Arnold Schoenberg’s Concerto for String Quartet and Orchestra, and likely many more works, could qualify as such translations.

There is a great irony to the assertions of twentieth-century composers that their re-compositions make the original more accessible. Contemporary performers in the classical tradition tend to program music written in the eighteenth and nineteenthcenturies more frequently than music of the twentieth century. It follows that audiences are typically more fluent in the musical languages of these earlier periods than in those of the twentieth century. I would say that the opposite of what composers like Stravinsky claim about their re-compositions is true. Hearing older works translated into the musical language of the twentieth century does not necessarily teach us about the musical works of Bach or Pergolesi, but it does help us to understand the “translator’s” language better. In Pulcinella, Stravinsky translates a text from a familiar language into a foreign one and in doing so he helps to teach his audience this new language. By relating the less familiar musical languages of the twentieth century to the more familiar musical languages of the past, musical translations could provide a missing link in relating complex music to audiences.


Nicholas Swett


Benjamin, Walter. “The Task of the Translator.” Translated by Harry Zohn. From Theories of Translation, edited by Rainer Schulte and John Biguenet. Chicago: University of Chicago Press, 1992, 218-227.

Derrida, Jacques. “From Des Tours de Babel.” Translated by Joseph F. Graham. From Theories of Translation, edited by Rainer Schulte and John Biguenet. Chicago: University of Chicago Press, 1992, 218-227.

Eaglestone, Robert. “Levinas, Translation, and Ethics.” From Nation, Language, and the Ethics of Translation. Edited By Sandra Bermann and Michael Wood.

Meyer, Leonard. Emotion and Meaning in Music. Chicago: The University of Chicago Press, 1956.

Pound, Ezra. “Guido’s Relations.” From Theories of Translation, edited by Rainer Schulte and John Biguenet. Chicago: University of Chicago Press, 1992, 83-92.

Steiner, George. “The Hermeneutic Motion.” From The Translation Studies Reader. edited by Lawrence Venuti. Routlege, 2004, 193-198.

Straus, Joseph N. “Recompositions by Schoenberg, Stravinsky, and Webern.” The Musical Quarterly, Vol, 72, No. 3 (1986), pp. 301-328.

Szendy, Peter. “Listen: A History of Our Ears.” Trans. Charlotte Mandell. US: Fordham University Press, 2008.

Liszt, Franz. “Beethoven Symphony No. 5 for Piano Solo.” Leipzig: Breitkopf & Hartel, 1865.

Stravinsky, Igor. “Suite Italienne: Five Movements from Pulcinella arranged for Cello and Piano.” London: Boosey and Hawkes, 1934.