(above) From left: Tenor Brett Sprague as Pedrillo, soprano Sarah Hayashi as Blondchen, tenor Stephen Steffens as Belmonte, and soprano Halley Gilbert as Konstanze, in a scene from Act II of the Bronx Opera’s recent revival production of Mark Herman and Ronnie Apter’s English translation of Mozart’s The Abduction from the Seraglio. (Photo used by permission of Hannah Spierman, social media director of the Bronx Opera.)
What if your translation had to be singable to a pre-existing melody that couldn’t be changed? What if it had to be understood on the fly while being performed? These are the sort of “what if” questions that my guests for this column, Ronnie Apter and Mark Herman, have been facing for years.
Ronnie Apter, Professor Emerita of English at Central Michigan University, received her PhD from Fordham University in 1980. Among many other accomplishments, she is the author of two books, Digging for the Treasure: Translation After Pound (1984) and A Bilingual Edition of the Love Songs of Bernart de Ventadorn in Occitan and English: Sugar and Salt (1999).
Mark Herman is a literary translator, technical translator, playwright, musician, and actor. For over two decades he has written the “Humor and Translation” column in The ATA Chronicle. He has a BS from Columbia University and an MS from the University of California, Berkeley, both in chemical engineering.
In 1979, the two were commissioned by the Bronx Opera to translate Mozart’s The Abduction from the Seraglio into English for performances in New York. The production and their translation were reviewed favorably in The New York Times. Since then, Mark and Ronnie have translated 23 additional operas, operettas, choral works, and songs, many of which have been performed in the U.S., U.K., and Canada. They have also collaborated on many translations of poetry and children’s books and co-authored Translating for Singing: the Theory, Art and Craft of Translating Lyrics (Bloomsbury, 2016).
Thank you both for joining us today. To begin, please tell us a little about the circumstances of that first commission in 1979.
We had done a singable translation of the lyrics to Carl Orff’s 1936 choral cantata Carmina Burana (ironically, for copyright reasons, the translation can’t be sung publicly) and sent it to various opera companies that performed at least occasionally in English translation. We didn’t receive responses from most of them, but we did hear from the Bronx Opera, a small company in New York now celebrating its 50th anniversary. They were looking for a new English version of Mozart’s The Abduction from the Seraglio and decided to take a chance on us, knowing that if we didn’t deliver, or they didn’t like our translation, they could still use an existing one. However, they were hoping that we would come up with a translation they preferred to the existing versions.
How did you handle what I assume was a relatively new challenge for both of you? Had you already translated much poetry at that stage? Any songs?
As mentioned, we had translated Orff’s Carmina Burana, a substantial work. We found that matching words to sung phrases came naturally to us. Also, we had both written metered, rhymed poetry, some of it published. Mark had written a couple of librettos, and some music, for musicals and Ronnie had written several plays. Both of us had also sung in high-quality amateur choruses for years. As for translations, in addition to Carmina Burana, Ronnie had translated several poems and Mark was translating technical documents for the library at Exxon Engineering, where he worked as an environmental engineer. And so we dove into the Abduction translation.
You’ve said that opera translation requires both musical and linguistic training. Tell us about your musical training at that point, and since then.
As children, both of us had instrument lessons—Ronnie on piano and Mark on saxophone and clarinet. As adults, we both had professional vocal training and, as mentioned, both of us were singing in choruses. Also, Mark had taken courses in music theory. We test our translations for singability by actually singing them before we submit them anywhere.
You’ve translated works originally written in Latin, French, German, Italian, Russian, and Czech. Are you fluent in all of them?
Unfortunately, we’re not truly fluent in any of them. We proceed with the help of dictionaries and, when necessary, native speakers of the source languages. In school, Ronnie studied Latin and French, and Mark Spanish, German, and Russian. In graduate school, Ronnie studied Old Occitan (also called Old Provençal, which is closely related to contemporary Catalan), and some ancient Greek. Between French and Latin, Ronnie was able to pick up Italian. When the Bronx Opera asked us to translate a Czech opera, Ronnie said, half jokingly, that we would do it if the company paid for Czech lessons. They said yes, and Ronnie studied with a native Czech speaker for about six months.
What did you learn from that first assignment in 1979 that encouraged you to keep translating operas?
We learned that we loved struggling to make meaningful, singable phrases, the thrill of working toward an actual performance, getting input from the directors, conductors, and singers, and hearing our words come alive in the voices of the performers. We also loved bringing our children to performances. After all, they had heard us test singing the lyrics for months.
Who usually commissions your translations?
Our work is commissioned by publishers, opera companies, and school music programs. We also do some translations on our own. Our publishers include Ricordi in Milan, Italy (the original publisher of operas by composers such as Verdi and Puccini), as well as Musica Russica in San Diego. For Ricordi, our work is subject to scrutiny by editors, who, initially, were conveniently located at the University of Chicago. We send them a translation, they make suggestions, and, usually, after three or four go-rounds the translation is considered publishable. When we work for an opera company or school, we submit each act as we finish it to the stage director (or conductor, if there is no stage director, such as for a choral work), who makes suggestions. Once rehearsals start, we also receive suggestions from the performers and conductor, most often still via the stage director.
When several people have a say in the performance of your translated lyrics, how are differences of opinion resolved? Who makes the final call?
The final say is with the editor of a published work. For an opera company, it’s usually the stage director. However, we require that no changes be made without consulting us first. Often, we just make changes in line with the criticisms. In the remaining cases, we argue; sometimes we win, sometimes we lose, and sometimes we compromise.
Do you attend rehearsals of an opera you’ve translated, or attend opening night? Was there one performance that was especially memorable?
If possible, we attend both rehearsals and performances. However, distant locations often make it impossible for us to attend either. Most memorable for us, because they were our first, were the Bronx Opera performances of The Abduction from the Seraglio. We attended the opening night on Friday and the second performance on Saturday. The audiences were enthusiastic. After the Saturday night performance, we went with friends to Greenwich Village, ate a late supper, and then bought an early morning edition of the Sunday New York Times. In it was a favorable review of both the Friday performance and our translation: “Highly singable.” We were soaring pretty high ourselves.
What can you tell us about how critics react to foreign opera sung in English?
Reactions vary widely. Some critics hate the very idea of opera performed in translation. Others blame the translators for everything they don’t like about the performance, or, if they like the performance, don’t mention that it was sung in translation, much less mention the translators. The New York Times, however, has mentioned us every time it has published a review of an opera sung in our translation.
Opera lyrics challenge the translator in all the ways that poetry does, but with the added burden that the words must also fit into a pre-existing—and largely unforgiving—musical structure. The poetry translator must prioritize meter, rhyming scheme, syllable count, and rhythm while keeping the meaning and allusions in mind. What must the opera translator do in addition?
In addition to functioning as poetry, words in an opera serve theatrical and musical functions. Therefore, in addition to the elements of poetry you mention, we must consider diction level and characterization and how the words fit the music.
You’ve said that “During the Victorian and Edwardian eras, the unwritten rules for the translation of all poetic forms into English called for an exact imitation of the rhyme scheme (and of the original meter), whether or not required by the music and however convoluted the result.” What about the fashions in vogue today?
A revolution in the writing, and translating, of poetry occurred early in the 20th century, discussed in detail in Ronnie’s book Digging for the Treasure. As a result, contemporary English poetry, both original and translated, is mainly free verse. The idea is that a translation can, and frequently should, emphasize important aspects of the original other than form. However, the preponderance of free verse means that few contemporary poets (unlike lyricists) acquire the metrical and rhyming skills needed for translating opera. But the fashion pendulum is swinging: forms like hip-hop are reviving interest in meter and rhyme, driven by the fact that human beings delight in patterns.
You’ve also said, “While rhymed originals should generally be translated into rhymed English, the spirit rather than the letter of the relationship between words and music should dictate the stanza form.” Can you give us a brief example of a translation “in the spirit”?
Consider a quatrain from Giuseppe Verdi’s Il trovatore: “Era già figlio prima d’amarti, / non può frenarmi il tuo martir! / Madre infelice, corro a salvarti, / o teco almeno corro a morir!” The rhyme scheme is abab, where the a rhymes are feminine (not accented on the last syllable) and the b rhymes are masculine (accented on the last syllable). The literal translation of the first two lines, sung by Manrico to his beloved Leonora as he is about to leave her, is “I was her son before I loved you. Your suffering will not hold me back.” The literal translation of the last two lines, sung by Manrico to his (absent) mother Azucena, is “Unhappy mother, I run to save you, or at least run to die with you!” Our translation, which omits the a rhymes, is “Though you may suffer, I cannot stay here; / she is my mother, you are my bride. / Unhappy mother, I swear to save you / or to die with you there at your side.” The missing a rhymes are compensated for by a weak rhyme (“suffer-mother”), repetition (“mother-mother,” “you-you”), and assonance (“may,” “stay,” “save,” and “die,” and “side”).
In your book Translating for Singing you claim that English is a “rhyme-poor” language. Please explain.
English has more individual words than most other languages, but lacks large classes of words with identical endings. A plethora of identical endings make rhyming in some languages easy. In the example above, the Italian endings “arti” (lines 1 and 3) and “ir” (lines 2 and 4) rhyme. Another problem in English is that “love,” an extremely important word in opera, has very few English rhymes, in contrast to “amor,” a word with many rhymes in Italian and Spanish, or “amour,” a word with many rhymes in French.
What are the abilities required to be a competent translator of singable lyrics? What prepares or trains the mind to develop these qualities?
As with all literary writing, improvements come with practice and work, but there must be some inherent talent. Translators should have at least some knowledge of the source language; relying entirely on dictionaries or previous translations is usually insufficient. Then, as for all literary translation, the translators must be excellent literary critics in order to discern tone and diction level, as well as understand how the lyrics mean in addition to what they mean. For example, should the words be considered straightforward or sarcastic? Were the words contemporary, archaic, or up-to-the-minute at the time of the work’s creation? If a character in an opera uses clichés, is it because it was customary for some librettos to be generally clichéd, as in early 19th-century Italy, or is it because that particular character, and only that character, actually speaks (sings) that way, as in late 19th-century Italy?
Next, the translators must have some knowledge of music and the requirements of vocal production. Translators should also be skilled lyricists in the target language. Translators also need to be prepared to work long and hard (translations put together in a weekend by changing a few words in someone else’s translation are usually terrible). Last, opera translators, often ill paid and frequently ignored or blamed for everything wrong with a performance, must consider their work a labor of love.
As quoted in your book Translating for Singing, Oscar Hammerstein said, “The sad fact is that when you hear opera in English it is in pretty bad English. The [translators] … are not poets, nor dramatists, nor showmen. A good adaptation of an opera requires a librettist who is all of these.” How do you two work together to perform these different roles?
To begin, the person who knows the source language best does an extremely literal translation, including all multiple meanings and odd turns of phrase, and explains any literary or historical allusions. Then, one of us does drafts of singable versions, after which we work together on the final version, every word of which is checked for singability. Ronnie usually sings the high parts (soprano and tenor) and Mark sings the lower parts (alto and bass). Mark usually starts at the beginning of a section and works toward the end, while Ronnie sometimes decides on what she wants the final phrase to be and works toward that. Mark is more interested in literality and Ronnie in overall feeling—a good set of checks and balances. We fight a lot. The entire process usually takes at least four months for a full-length opera. After our version is complete, a piano-vocal score (almost always dual-language) is made so that the translation is fully usable. Once a translation is done and performed, we advertise it, together with any favorable reviews, to other companies who might use our work on a royalty basis. We also self-publish dual-language piano-vocal scores for all of our translations not already published by Ricordi or Musica Russica, and sell them to music libraries and other interested parties.
Most operas currently performed were composed in the 19th century or even earlier. Haven’t most of them already been translated by now?
Yes, and some of them have been translated many times. But, as pointed out by Oscar Hammerstein, many of the existing translations are terrible. Also, translations can become dated even if the originals do not. Finally, a new translation may be needed for a particular stage production. For all these reasons, there is plenty of work remaining for contemporary translators.
Sung words are harder to understand than spoken words. Can enough of the words in a singable translation be understood to make all the effort that goes into making and performing a singable translation worthwhile?
Yes. In a translation that pays close attention to singability, as many words will be understood as are understood by native speakers listening to the original opera. Unlike operas, musicals are rarely performed in languages other than those of their actual audiences, and few complain about comprehension. Great composers such as Mozart, before they set the repetitions of a line on high-flying and incomprehensible coloratura, almost always set the line on notes in the performer’s middle range, which allows the words to be easily understood. Alhough the audience cannot re-read a missed sung word as in a book, usually one or more of the repetitions are themselves comprehensible, allowing the audience more than one chance to hear and understand the words. Finally, although we disagree with the practice in most cases, there is the possibility of projecting English captions (surtitles) even for works sung in English. Audiences can pay attention to the stage and glance at the captions only when necessary, as opposed to projected English captions for works sung in another language, which require the constant attention of the audience.
Opera translators probably pay a lot more attention to the prosody of the original than non-literary translators pay to the nuances of, for example, the technical language of a software manual. But in either case the patterns in the source and target languages have to be considered. What have you learned about the rhythmic arrangement of words in your field that might be useful to other translators?
It’s true that operatic music is strongly influenced by the prosody of the source language. However, it’s also true that music, with its regular accent patterns and fixed number of beats per measure, at least partially distorts the prosodies of both the source and target languages, moving both closer to the meter of the music. One reason for the awfulness of many Victorian and Edwardian opera translations is that the translators, believing they had to reproduce the prosody (and rhyme scheme) of the libretto exactly, set themselves a much more difficult task than if they had allowed the music, in addition to the words, to influence the prosody of their translations.
One area that is as important in the translation of literary prose as it is in the translation of opera lyrics is the “foreignization/domestication” dichotomy. Please give us your views and a couple of examples.
Much over-simplified, “foreignization” means bringing the audience to the original, and “domestication” means bringing the original to the audience. The merits of each, whether designated by these exact terms or not, have been debated for centuries. We’re in favor of both, used thoughtfully, and keeping in mind that an audience must understand the words as they are sung. A weird example of “domestication” is our translation of The Sorcerer’s Daughter from nonsense Italian into nonsense English. Sometimes we foreignize an entire translation by writing our version in non-modern English when we believe it’s the only way to convey the true flavor of the original. For example, Charles Gounod’s 19th-century French operetta The Doctor in Spite of Himself, based on Molière’s 17th-century play, involves 17th-century customs that could strike the audience as very odd if spoken of matter-of-factly in modern English. Because of this, we translated it into 17th-century English. Here’s a sample regarding an elaborate social ritual involving hats:
Lucas: (to Sganarelle)
Straightways ha’ we come from the village, sir, for to summon your skill.
Sganarelle: I am your very humble servant.
Valère: But, sir, shade your head from the sun’s febrile ray.
Lucas: I’ truth the sun be fiercely beating—on with your bonnet, pray.
Sganarelle: (Sganarelle puts on his hat.) Hi-hey! Pox pester this pother of greeting!
The archaic language informs the audience that the ritual they are observing is similarly archaic and therefore acceptable on its own terms. They will usually not be mentally caught up asking “What was that about?” as they might be had the words been in contemporary English.
In addition to translating an opera’s dialogue and lyrics, you sometimes also contribute lines and dialogue that were not in the original. Tell us about that process and give us a few examples.
Sometimes added words are necessary to explain something that the original librettist omitted—to add a footnote directly into the performed libretto because there is no chance for the audience to read one. For example, in François Boieldieu’s French operetta My Aunt Aurore, a satire on the early 19th-century craze for gothic novels, the heroine Julie tells her aunt that she has followed her instructions and keeps a dovecote in honor of the one kept by the title character in Samuel Richardson’s novel Clarissa. The original early 19th-century audience would have known that Clarissa kept a dairy, so we added a footnote:
Julie: Good heavens, aunt, you yourself desired me to keep a dovecote in honor of the one Clarissa had.
Marton: Clarissa kept a dairy.
Aurore: Yes, but cows are so very large. I explained all that…
Jokes are often not translatable locally; that is, at the same places in the translation that they are in the original. For example, the French comic operetta Madame Angot’s Daughter contains the following lines about Madame Angot: “Captured in Malabar, she is believed to be a widow, alas! They want to burn her alive!” We added a joke to compensate for those lost in our translation by rendering the lines as, “The natives tried to burn her / in Hindu Malabar. / They took her for a widow, / but she refused to char.” This joke refers to the fact that “char,” in addition to its usual meaning, is a British verb meaning “clean,” and cleaning women, also called “chars,” were often widows.
Presumably, like the original librettist, you fall in love with some of your words. Please identify one of your favorite lines and tell us why it pleases you.
One of our favorites is a couplet from Mozart’s The Magic Flute, with a German libretto by Emanuel Schikaneder. The meaning is “A man is of strong spirit, he thinks about what he may say.” The tone is slightly humorous. We kept getting stuck on the rhymes, working with “weak-speak,” “man-can,” and others. Finally, we hit upon “A man is made of sturdy stuff: / he knows when he has said enough.” Our version also adds a joke: the singers repeat the second line four times, suggesting that they should listen to what they are singing and shut up.
Despite the concern, some might say obsession, of many operas with sex and violence, albeit occurring almost exclusively offstage, most audiences would still give most operas a “G” rating. Is this because audiences don’t understand the foreign words? Have you yourselves ever run into censorship?
Some “G” ratings do arise in part because audiences do not understand the foreign words. Also, some translations are censored. Mostly, however, the sanitized ratings come from the fact that opera is considered “high art,” much as Shakespeare’s plays today usually get the same free pass despite bawdiness and violence (though there was a period when Shakespeare’s works were heavily “Bowdlerized”).
We have encountered a few taboos over the years. For example, in Mozart’s The Magic Flute, the character Sarastro sings lines meaning, “A man must govern your heart, / for without him / every woman is likely to step outside her [rightful] sphere of activity.” We originally wrote, “Your heart needs guidance from a husband. / Without a man, / a woman will exceed the sphere of woman’s work.” We were required to change this to “Your heart needs guidance from a husband, / or, like your mother, / you will never learn to wed desire and duty.”
What are the most significant lessons you’ve learned by doing this work? Do you have any tips for translators who might be interested in working in this area?
We’ve learned that opera translation, as a branch of show business, is essentially a collaborative endeavor, that the translation is not finished until the performance is over, and that new productions may require additional tweaks. Translators interested in working in this area, regardless of inherent talent, should probably start by participating in sung performances of English lyrics. These can be amateur affairs, such as community choruses or community theater productions of musicals. Then the skills mentioned above should be acquired. After that, the hard work begins.
- Apter, Ronnie, and Mark Herman. Translating for Singing (Bloomsbury Publishing, 2016),
In addition to its own treatment of the subject, the book has an extensive bibliography.
- Low, Peter. Translating Song (Routledge, 2017), http://bit.ly/Translating-Song.
- Excerpts from performances in Apter and Herman’s English translations are available on YouTube, including:
The Bartered Bride
This features Mařenka and Jeník’s Act III duet from Bedřich Smetana’s Czech opera The Bartered Bride, sung by soprano Adelaide Boedecker and tenor Matthew Grills. The Bartered Bride was produced by the Eastman School of Music in Rochester, New York.
This features the aria with chorus for Don Carlos from Act II of Giuseppe Verdi’s Ernani, with baritone Russell Malcolm as Don Carlos, the King of Spain. The excerpt from Ernani is audio only, although stills from the production are shown during the aria.
Ernani was produced by the Tayside Opera in Scotland.
Tony Beckwith was born in Buenos Aires, Argentina, spent his formative years in Montevideo, Uruguay, then set off to see the world. He moved to Texas in 1980 and currently lives in Austin, Texas, where he works as a writer, translator, poet, and cartoonist. Contact: firstname.lastname@example.org.