When music meets language: a guide to music-linked translation

When music meets language

Only in recent years have scholars begun to investigate what Harai Golomb (2005: 121–162) calls “music-linked translation.” One explanation for this delay lies in the methodological puzzles the field presents. It blurs familiar boundaries between translation, adaptation, and rewriting, it shakes received ideas about authorship and the source text, and it unmistakably demands a interdisciplinary lens (Susam-Sarajeva 2008: 188–189). Two collected volumes — Gorlée 2005 and Susam-Sarajeva 2008 — have recently joined a handful of earlier articles dedicated to music and translation. These works show that translating musical texts extends well beyond opera and assumes many guises: literal renderings of CD booklets or printed libretti, sung translations, rewritten song lyrics, surtitles, subtitles, and dubbing. This overview concentrates chiefly on opera translation in its various forms — the area that has garnered the most scholarly attention — as well as on the translation of songs and musicals.

Translating opera for the stage

An opera creates meaning through a collaboration of music, performance, and verbal text. Yet the commanding presence of music has traditionally framed opera as essentially a musical genre rather than a dramatic art. This bias is visible in sung translation: the music is normally regarded as untouchable and serves as the axis around which translation decisions revolve. The verbal text is subordinated to the score, forcing the translator to adhere to the notes and tempo of the original. Strategies are therefore constrained by phonetic elements such as vowel length, sound quality, rhythm, prosody, and word or sentence stress. Herman and Apter (1991) offer a thorough analysis of these constraints.

Rhetorical devices like rhyme, alliteration, repetition, and onomatopoeia also frequently shape the choice of verbal text, though most experts on opera translation advocate for flexibility. Each language presents its own difficulties and achieves “singability” — a central concern — through different means. Characterization is conveyed by the singer’s register, so it is vital for performers to receive a text they feel comfortable singing and that forms a unified whole with the musical pattern.

The opera translator needs not only a strong command of the languages most often used for libretti (primarily Italian, German, French, English, Russian, and Czech) but also some familiarity with music, vocal technique, and rhythm. At the same time, the translator must recognize the dramatic nature of the script and respond to each composer’s individual style. Analyzing opera as a multimedia text, Kaindl (1995) insists that a work is not simply the sum of several codes and proposes a holistic approach — one that attends not only to the tight link between words and music but also to aspects of stage performance. Therefore, the translator should collaborate closely with the director, the conductor, and those in charge of scenic elements so that translation decisions can be tailored to a particular production.

Opera translation is often surrounded by a debate about its feasibility. Those who oppose sung translation argue that because music is the pivotal element, there is no need to understand the text, or else that the text is an inseparable part of the music and cannot be altered (see Herman and Apter 1991). Supporters counter that while sung translation inevitably forfeits the original alignment of words and music that the composer intended, it also brings considerable gains: the performance is enhanced, allowing opera to reclaim its genuine nature as musical drama. Music history indeed supplies countless examples of operas performed in translation for theater audiences across many target languages, and there are also recorded translations of famous arias sung by important voices.

Language policy in opera has paralleled the genre’s development. Italian, the language in which opera was born, served as the main operatic language until the nineteenth century, when German and English ended the near constant presence of Italian and French in European opera houses (Desblache 2007). Up to the Second World War, however, it was not unusual to hear an opera sung half in Italian and half in the local language (Desblache 2007: 159–160). Financial reasons drove this practice and still influence decisions today: singers now perform in productions worldwide, making it more economical to have one common language of performance for each opera — usually the original libretto language. Some countries nevertheless alternate productions in the original language with performances in translated versions.

Reasons for translating or not translating opera are actually tied more closely to sociological, historical, and economic factors than to linguistic or technical ones: the inherent incongruity of characters communicating through song; the perception of opera as an essentially musical genre; the claim by frequent attendees that they know the works by heart; the conception of opera as a “difficult” genre meant only for a social and cultural elite. All these factors have probably contributed to modern audiences’ frequent resistance to sung translation.

Translating opera libretti

Before the twentieth century, audiences wanted to understand the libretto and could follow it in translated printed versions. Indeed, the text and librettists held a prominent place in opera until the nineteenth century. But things changed dramatically in the middle of that century, when a black-out was introduced in the auditorium — preventing audiences from reading libretti during performances — the music took precedent over the words, composers took control of the libretto, and the long-standing practice of adapting texts from foreign operas became unacceptable (Desblache 2007).

Today, opera libretti in foreign versions are commonly included in CD recordings, which routinely offer inserts in two or three languages — typically English, French, German, or Italian. Bilingual reader-oriented publications also exist, and some vocal scores print a translation below or above the original text. These target texts are usually literal translations, intended for opera lovers’ reading or singers’ study. Unfortunately, audiences and critics have often judged libretti based on these semantic renderings, which were never meant for performance, or else on readings of the original libretto apart from its music. Yet the hybrid nature of opera means that the text is not an autonomous entity but just one element in a complex semiotic whole. The quality of these literal libretto translations can be fairly assessed only when considered alongside this fact and their very specific skopos (see Functionalist approaches).

Opera surtitling

The non-autonomous nature of the libretto also becomes apparent in a new translation method that arrived in the 1980s to complement sung translations and prose versions of libretti. Registered as SURTITLES™ by a team of Canadian professionals, this technological innovation — aimed at making opera understandable and widely accessible — has radically transformed the experience of opera audiences.

Target texts produced through surtitling are projected on a screen, usually above the proscenium arch, while the source text is being sung on stage. Although surtitling does not present the same verbal difficulties as sung translation, it is still constrained by technical, format, and timing factors. The written text on the screen must match the performance on stage and be subject to the timing demands of the music. The functional nature of surtitles — whose essential goal is to help audiences understand the libretto without overly interfering with the overall reception of the opera — also strongly shapes translators’ textual choices. As a result, the principle of economy is paramount, and surtitles typically involve a significant reduction of the source text, ranging from one-third to one-half. This reduction is always dependent on the music, which is the most decisive factor: it dictates how the verbal text is divided into separate captions, their synchronization, and how long each is displayed. Each title should appear during its corresponding passage of singing while allowing a comfortable reading speed.

The surtitling process and the audience’s reception of the titles are also influenced by the live, transient, and unrepeatable nature of an opera performance. At the micro-textual level, this requires simplifying vocabulary, syntax, and punctuation; condensing verbal repetitions; removing secondary details; and structuring each title as a consistent, meaningful unit. The fact that the audience simultaneously has access to the sung source text and the projected target text further restricts the translator’s choices and increases the difficulty. The translator must frequently decide which content to keep or which roles to prioritize, judgment that requires constant attention to the libretto, the music, the production, and the audience. Surtitlers must therefore work in collaboration with other agents involved in the production. As in most forms of Audiovisual Translation, teamwork is vital both for creating the titles and for projecting them during the performance.

Despite the controversy that surtitles initially sparked in some countries — such as Great Britain and the USA — most audiences have welcomed them enthusiastically. Surtitling has markedly increased audience sizes, making opera more accessible and changing what opera-goers expect from a performance (they now expect to understand the story, the characters’ emotions, and their verbal exchanges). Surtitling has also widened opera house repertoires by introducing new languages, works, and composers. Lastly, it has brought about a new way of watching opera — more active and fragmented — that nevertheless enhances the experience. (See Desblache 2007; Mateo 2007; Virkkunen 2004.)

Other accessibility techniques have recently been introduced in some opera houses, such as audiodescription and audio introductions. For blind audiences, these methods provide, respectively, a full narration or an introduction to the visual dimensions of a performance (see Media accessibility). Both have already received attention from scholars of Audiovisual Translation (see Díaz Cintas et al. 2007).

Foreign musicals in a target context

Translating stage musicals involves similar problems to those of opera translation for performance, though musicals tend to alternate between sung parts and spoken dialogue more frequently. Both musicals and opera share the semiotic complexity, the ephemeral nature of reception, and the many agents involved in a production that typify drama translation (Mateo 2008: 321). Today, however, different translation policies prevail for each type of musical drama: most opera houses — at least in the Western world — use surtitles for foreign operas, while sung translation is the norm for stage musicals in most target systems (Mateo 2008: 320). The divergent perceptions of each genre in the music and theater worlds — one considered canonized and elitist, the other non-canonized and popular — have likely shaped a different attitude in audiences toward the language of performance. Whereas most opera-goers are quite content with the condensed translation offered in surtitles, translated Anglo-American musicals would probably not have reached such wide audiences in European countries if they had been exported “only” with surtitles. This differing attitude may also be explained by the stronger association of opera with convention and artifice, whereas musicals are perceived as more realistic — for instance, in matching singers to roles.

Economic and cultural factors also come into play, not just in the chosen translation method but also in the translation process and in selecting source texts — as a handful of articles on musical translation have shown. Financial considerations have driven the unsystematic and partial transfer methods used to import American film musicals into some European countries, like Italy. By contrast, the successful introduction of Anglo-American stage musicals into other countries, like Spain, has been based on commercial strategies and something as straightforward as filling a perceived “cultural gap” in domestic artistic production (see Di Giovanni 2008; Mateo 2008).

These macro-level approaches to musical translation are complemented by John Franzon’s micro-level and semiotic analysis (2005). Taking a functional view of lyric translation and accounting for the multimedia nature of stage musicals, Franzon argues that even when the purposes of the source and target texts are similar — here, a musical performance — one can only expect similarity at the contextual and functional levels, not at the textual-semiotic one.

Translating songs

The translation of songs tends to be commissioned less frequently than that of other musical genres (Franzon 2008: 373). It typically occurs for film subtitling, occasionally for theater, or when lyrics are quoted in publications and CD booklets. The technical challenges of song translation are quite similar to those of sung translation for opera or musicals, but researchers generally agree that the translator’s priorities may differ considerably. Moreover, songs are often translated by professionals other than translators: singers, songwriters, opera specialists, playwrights, and, increasingly, fans of the source song on the internet (Franzon 2008: 373–374).

A wide range of (sub)genres fall under song translation: folk songs, German lieder, choral works by classical composers, religious hymns, French chanson, pop songs, children’s songs, and poems set to music. Music history offers examples of all of them in translation.

Franzon proposes a functional view of “singability,” stating that “a song might be recognized as a translation if it is a second version of a source song that allows some essential values of the source’s music and/or its lyrics and/or its sung performance to be reproduced in a target language” (2008: 376). The translator thus has several theoretical options (2008: 376–389): leaving the song untranslated, adapting the translation to the original music, or translating the lyrics without accounting for the music, among others. The principal factor that determines which option to select is the delivery mode (printed format, subtitling, sung performance for recordings or the stage, etc.).

From a functional standpoint as well, Peter Low proposes the “Pentathlon Principle” for studying song translation, which he sees as a balance of five criteria: singability, sense, naturalness, rhythm, and rhyme (2005: 185–212). Like Franzon, Low insists that the approach must be flexible and pragmatic across all these features, with particular attention to the overall effect of the translated song, which should create the illusion that the original music was actually conceived for the target-language lyrics. Focusing on popular songs, Kaindl argues that song translation must be understood as a product in which words, music, and image are interdependent, which means its study should be interdisciplinary (2005: 234–262).

Looking forward

Music-linked translation offers a broad terrain for investigation: the specificities of diverse musical genres and various translation methods, micro-textual analyses of source and target musical texts, descriptive studies of texts and contexts, the role of cultural and sociological factors, reception studies, intertextuality, creativity, and questions of identity in music translation, among others (see Susam-Sarajeva 2008: 191–195). Whatever the focus, solid research in this area will require a multidisciplinary approach — drawing on musicology, theater studies, semiotics, sociology, literary history, and translation studies — in order to fully account for the translational norms and processes that span linguistic, discursive, aesthetic, socio-cultural, historical, ideological, economic, and technical dimensions.

References

“My Fair Lady.” In Gorlée (ed.), 263–298.

Franzon, John. 2008. “Choices in song translation: Singability in print, subtitles and sung performances.” The Translator 14 (2): 373–399.

Golomb, Harai. 2005. “Music-linked translation (MLT) and Mozart’s operas: Theoretical, textual and practical approaches.” In Dinda L. Gorlée (ed.), 121–161.

Gorlée, Dinda L. (ed.). 2005. Song and Significance: Virtues and Vices of Vocal Translation. Amsterdam/New York: Rodopi.

Herman, Mark & Apter, Ronnie. 1991. “Opera translation.” In Translation: Theory and Practice. Tension and Interdependence [ATA Scholarly Monograph Series 5], Mildred L. Larson (ed.), 100–119. Amsterdam/Philadelphia: John Benjamins.

Kaindl, Klaus. 1995. Die Oper als Textgestalt: Perspektiven einer interdisziplinären Übersetzungswissenschaft. Tübingen: Staufenburg Verlag Brigitte Narr.

Kaindl, Klaus. 2005. “The plurisemiotics of pop song translation: Words, music, voice and image.” In Gorlée (ed.), 235–262.

Low, Peter. 2005. “The Pentathlon approach to translating songs.” In Gorlée (ed.), 185–212.

Marschall, Gottfried (ed.). 2004. La traduction des livrets: Aspects théoriques, historiques et pragmatiques. Paris: Presses de l’Université Paris-Sorbonne.

Mateo, Marta. 2007. “Surtitling today: New uses, attitudes and developments.” Linguistica Antverpiensia 6: 135–154.

Mateo, Marta. 2008. “Anglo-American musicals in Spanish theatres.” The Translator 14 (2): 319–342.

Susam-Sarajeva, Sebnem. (ed.). 2008. Translation and Music. Special Issue of The Translator 14 (2).

Virkkunen, Riita. 2004. “The source text of opera surtitles.” Meta 49 (1): 89–97.