Interview with Peter Low1

Lauro Meller and Daniel Padilha Pacheco da Costa*

Peter Low is a Senior Fellow at the University of Canterbury in Christchurch

(New Zealand) and his fields of research are Translation Theory and French
Poetry and Song. Both as a translator and as an academic researcher, he is
one of the leading specialists in song translation. Faced with the challenge of
producing singable translations, and based on his own background as a
lecturer of French at the University of Canterbury and as a trained pianist,
he devised an original method which he calls the “Pentathlon Principle”.
In his book Translating Song: lyrics and texts2 (2017), Peter Low tends
10.17771/ PUCRio.TradRev.50547

towards the notion that song lyrics should usually be translated in a way that
they could be sung over the original melody – what he calls “singable
translation”. Although this sounds almost impossible, thousands of songs
are translated – often well translated – all over the world. Many of them are
in fact so well worded that listeners do not even suspect these songs were
originally sung in a different language. Here is the interview we’ve done
with Peter Low.3

Lauro Meller (LM): Dr Low, please tell us a little bit about your academic
background and how you got involved with Translation Studies.

1 E-mail: [email protected].
* Universidade Federal do Rio de Grande do Norte (UFRN) e Universidade Federal de Uberlândia (UFU).
2 See LOW, 2017.
3 This interview is the result of three different phases. It started in October of 2019, when Lauro Meller

travelled to Christchurch-NZ in order to spend two weeks working with Peter Low. After Low’s
acceptance of being interviewed for a special issue of the Brazilian journal Tradução em Revista on the topic
of Translation and Music, Daniel P. P. da Costa sent Peter Low some complementary questions, which
were answered by email over July and August of 2020. Finally, Lauro Meller and Daniel P. P. da Costa
interviewed Peter Low together by videoconference on the 5th August of 2020.

Submetido em 01/06/2020
Aceito em 15/08/2020
 MELLER and COSTA Interview with Peter Low

Peter Low (PL): After studying in New Zealand and France, I got a
university job teaching French language and literature. One writer whose
works I taught was Prévert, and to me his popular poems and texts seemed
to demand to be translated into natural English, so I worked on them. Later
I saw a need to teach students to approach translation as a specific skill (a
craft, sometimes an art). So I developed a course on French-English
translation. That was in the 1990s. I also read a lot of translation theory,
which later made it possible to teach translation to students working in other
languages. There are general principles, and above all there are general
questions the translator ought to think about.

LM: Besides being a reference in Translation Studies, notably singable

translations, you also worked as a translator. Could you tell us a little
about the songs or genres you translated?

PL: I had a job as an academic; so translating was an extra thing, and I was
10.17771/ PUCRio.TradRev.50547

seldom paid for it. I often translated non-literary prose texts, even a book
about depleted uranium. I never did interpreting or subtitling, almost
never. But surtitling, yes, I provided surtitles for several French and Italian
operas. Poetry didn’t scare me (perhaps it should have!). I collaborated
recently on translations of the poet Évariste de Parny4. I often did what I call
“recital translations”: non-singable versions of French poems and song-lyrics
intended to be printed in concert programmes or CD inserts where the song
is sung in the original. But that is a different genre – certainly a different
skopos – from “singable translations” which are intended for performance in
the target language. I guess it was my various articles about song-translating
that led to an invitation to write a book on the subject.

Daniel P. P. da Costa (DC): You said you translated poetry before you
started translating songs. Many problems discussed in literary translation
aren’t excluded from song translation. What are the differences between
literary and song translations?

4 See PARNY, 2018.

Tradução em Revista 29, 2020.2 299

 MELLER and COSTA Interview with Peter Low

PL: One translates a poem for the printed page, and sometimes for recitation.
But a song-lyric is one component in the hybrid art-form called song. That
skopos dictates the greater importance of sounds and rhythms and the
absolute need for singability.

DC: You have surtitled six French and Italian operas. You have also written
about surtitling or subtitling opera on screens both in articles5 and in the
chapter called “Translations to read” of your book6. What are the main
difficulties of this specific task?

PL: The common issues of meaning and naturalness, plus a concern for good
communication in the particular skopos situation – but fewer difficulties than
with singable translations.

DC: Since the beginning of the Romantic movement until now, translated
10.17771/ PUCRio.TradRev.50547

opera has been strongly undervalued in Brazil, in spite of some

remarkable advocates, as the modernist writer and musicologist Mário de
Andrade7. But Italian operas have been often sung in translation not only
in Brazil, but also in Germany, France and English-speaking countries.
“Nowadays, however”, as you state in your book, “most translations of
opera are done not for singing but for reading on surtitle or subtitle
screens” (LOW, 2017, p. 7). Why aren’t there more performances of
translated opera? Is this the reason why your theory doesn’t deal much
with singable translations of operas or musicals? Would you draw the
main differences between singable translations of operas and musicals
and singable translations of songs?

PL: The Italians invented opera, and their language, with few vowels and
few closed syllables, is wonderful for singing. The decision to use
translations lies mostly with the directors, who can usually understand the

5 See LOW, 2002.

6 See LOW, 2017, p. 40-62.
7 See KAISER, 1999.

Tradução em Revista 29, 2020.2 300

 MELLER and COSTA Interview with Peter Low

Italian words. But there is a good case for presenting the story in a language
that the audience can follow. To me, the best discussion of translating opera
is in the book Translating for Singing, by Apter and Herman8. But my book
makes a good point: “In operas and musicals every song needs to cohere
with the extended stage-work that it belongs to, a need not covered by my
five criteria” (LOW, 2017, p. 110). The TT should play its intended role in a
complex storyline, and in the fictional character’s emotional world. These
considerations are likely to lead to more compromises, to greater flexibility
with literal meaning.

DC: I’d like to ask you about the criteria behind the choice of ST (source
text) in song translation. Many interpreters (for example, Ella Fitzgerald)
improvise in many ways, including with the lyrics of the songs, modifying
them or adding stanzas. There are often significant changes between one
performance and another in the several recordings of American jazz
standards. We observe the same phenomenon in Brazilian music. The
10.17771/ PUCRio.TradRev.50547

composer Noel Rosa is an example of it: depending on the recording, the

lyrics can vary substantially. How do you choose the original lyrics used
in the translation? What criteria do you use to determine the ST of the
songs you translate?

PL: I usually choose a published text, presumably authorised by the

songwriter. But when there are various recorded versions, translators may
follow their own preferences.

LM: In one passage of the book you ponder whether songs should be
performed in the language of the public or not. To give you a Brazilian
example, singer-songwriter Caetano Veloso has launched many songs –
even full albums – singing in languages other than Portuguese: English,
Spanish, Italian, French. Who do you think would be the presumed public
of such recordings?

8 See APTER and HERMAN, 2016.

Tradução em Revista 29, 2020.2 301

 MELLER and COSTA Interview with Peter Low

PL: My view is that if a song is performed in the language of the audience,

the hearers and singers have a different experience. If I hear a song sung in
Russian, then I might enjoy it because of Tchaikovsky’s romantic melody –
but there is a whole verbal dimension of the song that I am missing, it might
as well be performed on a violin. My book does not contend that songs ought
to be always performed in translation, but it refuses the opposite – which is
the pretension that every song should be sung in its original language
whether the audience understands it or not – which is the approach that
sometimes seems to happen. I do know that some top opera composers
wanted the language of the audience. Puccini said he wanted it in French for
Paris. Wagner wanted his operas in French for Paris. If you’re hearing the
voices in a foreign language, you are missing part of the meaning of the
work, the verbal dimension.

DC: At the end of the second chapter of your book, you say: “It is likely
that some songs – a small minority – are impossible to translate at all well,
10.17771/ PUCRio.TradRev.50547

for reasons that are linguistic, cultural or both” (LOW, 2017, p. 38). Have
you ever given up translating a song because of its “untranslatability”?
From your experience, do you consider some songs should not be
translated, but left untranslated?

PL: I once gave up on a song of Brassens. I even mention a song of Perret as

being untranslatable. Some texts are too embedded in the source language
and culture. Some are too exquisite to touch. And then there are verbally
weak texts which don’t merit attention.

LM: When you mention “songs that don’t merit attention”, would that be
the opposite of “songs with staying power”, as Malcolm McNeill [a jazz
singer in Christchurch] puts it?

PL: Malcolm would record songs of Hoagy Carmichael because he thought

that these songs had staying power in time, i.e., they could be good songs in
different times, in different cultures, whereas some pop songs you hear are
very much throwaway – they’re only gonna be popular for a while, and then

Tradução em Revista 29, 2020.2 302

 MELLER and COSTA Interview with Peter Low

they disappear. I think “staying power” implies a quality and a merit that
will give durability to this song. So, you wouldn’t have a political song from
this last election, because that wouldn’t have staying power. You know, as
well as I, that it is a hard job to do a good translation of a song, a singable
translation. Therefore, you don’t want to spend your time on something that
will be easily forgotten.

LM: Aren’t we here talking about the canon? I know this may sound
impressionistic, but at the end of the day there are good songs and bad
songs, aren’t there?

PL: Of course there are! Subjectively I can say that “Summertime” is a good
song because I like it; I can also say objectively that hundreds of people have
thought that “Summertime” is a good song. It has been recorded dozens of
times, it sticks in people’s memories, and when you walk down the street
and happen to be singing [hums the melody to “Summertime”] people will
10.17771/ PUCRio.TradRev.50547

recognise it. There’s objective evidence. You don’t have to have to analyse
the strange harmonies that Gershwin uses, you don’t have to know the show
Porgy and Bess from which it comes, the song has proven some sort of quality
and merit – actually, whether you like it or not.

LM: But aren’t there “bad songs” – and by that I mean without depth or
originality – which nevertheless sell millions of copies and stick in
people’s minds over time? Couldn’t these also be considered “songs with
staying power”?
PL: For me “staying power” means something other than the fact that it got
well-known. There’s a Christmas song called “The little drummer boy”
which I think is rubbish. It has been recorded and you could hear it in
shopping malls and places like that. I have the prerogative of rejecting it from
my canon of good songs, but I can’t deny that it is well-known. And
conversely, there are things from early European music which I think are
wonderful – things like “Amarilli, mia Bella”,9 which hardly anybody

9 “Amarilli, mia Bella” (1601), music by Giulio Caccini and lyrics by Giovanni Battista Guarini.

Tradução em Revista 29, 2020.2 303

 MELLER and COSTA Interview with Peter Low

knows, but which is an exquisite Italian song of the late 16th century. That’s
how things go. Popularity seems to bring popularity, but then staying power
– I use a different phrase from my friend Malcolm: a good song is “a song
that feeds me”. A wonderful song of Schubert, or Fauré, or – I don’t know –
somebody from Italy. If I feel fed by that, in my heart and mind or both, then
that’s a good song.

DC: You’ve mentioned the opera Porgy and Bess,10 in which the variations
in the dialect of the characters play an important role in denoting racial
and class divisions. In the second chapter of your book, when discussing
the “Problems of non-standard language (dialect, sociolect, slang,
colloquialisms)” (LOW, 2017, p. 28), you quote a well-known example of
non-standard English in that opera: “I loves you”. As a general rule, you
apply to songs Newmark’s advice of “‘processing’ only a small proportion
of the SL dialect words” (NEWMARK, 1988, p. 95). Even following this
“moderation principle” in the translation of non-standard language in
10.17771/ PUCRio.TradRev.50547

songs, are there cases in which this problem should be left aside in favor
of other things to attend to?

PL: Non-standard words are not uncommon in songs, and a TT can often
handle that. But often they’re unimportant and a translator can ignore them.
But here (as often elsewhere) we need to ask the question: “Is this a
significant feature of this particular ST, one which needs to be replicated for
a true translation?” When Bob Dylan sings not “changing” but “a-changin’”,
I wouldn’t bother to replicate that in another language, since to do so would
detract from the aim of translating the song’s core message.

LM: The Pentathlon Principle defines five criteria for translating songs.
How did you arrive at the pentathlon metaphor? What was the inspiration
for it?

10 Porgy and Bess (1935), music by George Gershwin and lyrics by Ira Gershwin and DuBose Heyward.

Tradução em Revista 29, 2020.2 304

 MELLER and COSTA Interview with Peter Low

PL: Many classical songs (such as German Lieder) are printed with so-called
“singable English versions”. But most of these are mediocre or worse, and
they are seldom actually used by singers. Some people even object to
“singable translations” on principle. I felt that I could identify big strategic
problems in the translators’ approach – they failed to properly consider all
the relevant criteria, which are usually five in number. In developing my
own approach, chiefly to help me do better myself, I considered words like
“juggling” and “trade-offs” before stumbling onto the “pentathlon”
metaphor, which has the virtue of suggesting an “overall score for five
events.”11

LM: “Singability” is, in your own words, “relative ease of vocalisation”.

And you add that it should be “[...] the first criterion, because any target-
text that scores poorly in this criterion is a failure, and would not be sung,
or even pronounced, whatever its other virtues may be” (LOW, 2017, p. 82).
By “first criterion” do you mean “priority number one”, if compared to the
10.17771/ PUCRio.TradRev.50547

other four criteria of the Pentathlon?

PL: Not number one in importance, but sine qua non, it’s a criterion that has
to be met at an adequate level, otherwise you’re wasting your time. If the
thing is not going to be performed because of the troubles in vocalising,
because a translation into Polish contains the word such and such that can’t
be sung, even though it’s spoken in Polish – Slavic languages seem to be
worse than others – you should revise the translation so that it can be sung.

LM: Still according to you, the best people to assess if the singability is
good are not the translators, but the singers themselves.

PL: Singers, or singing teachers, or choir masters – people who are really into
that. Once a famous singer approached the translator and said: “I want a

11By using this sporting metaphor, Low implies that the same way no athlete will excel in all five events,
no translation will be 100% perfect. Then, some degree of commitment has to be made according to the
difficulties and characteristics of each song: some of them will lean more heavily onto the rhythmic aspect,
some on the rhymes, etc.

Tradução em Revista 29, 2020.2 305

 MELLER and COSTA Interview with Peter Low

different vowel for that high note – can you give me a different vowel?”. That
seemed to be a totally reasonable choice [for the translator] and there may be
reasons why the translator couldn’t oblige, but she knew what she could sing
well. Some languages have wonderful vowels all the way through: Italian is
pretty good, Maori is pretty good, English has difficult vowels for high notes,
words that shouldn’t be sung on a high note, words like “the”, which should
never have a stress, because it is always a weak word, maybe the most
common [word] in the language. English has a lot of diphthongs – a word
like “today”, with this /ei/ like that – that’s not a bad vowel – but they can
create problems in singing. Other languages like Italian have a very pure
vowel and so you can sing it long – besides, that syllable will not usually
have a consonant at the end, and that can create problems, perhaps more in
English than in Portuguese.

LM: Should then song translators start off with the criterion of singability
in mind? I believe some song translators might give extra importance to
10.17771/ PUCRio.TradRev.50547

rhythm in the early stages of the translation (i.e., keeping the same number
of syllables and the same stresses in both the ST and the TT), in order to
make sure the “new” text can be sung over the original melody.

PL: I would say singability should be the first criterion for evaluating a song
translation. That’s not the first thing I work with. These five criteria I
developed as I was actually doing a job of making singable translations. It
was only after that that I considered they could be used for evaluating and
comparing. In fact I wrote an article12 in which I compared three translators
of Schubert’s songs – I tried to tally up how well they had managed with my
pentathlon – and of course some were better than others and some were
weaker. The reason why I put it first in the evaluation is that if the singability
is poor, then you’re wasting your time. That’s all.

12 See LOW, 2008.

Tradução em Revista 29, 2020.2 306

 MELLER and COSTA Interview with Peter Low

DC: Many song translations and operas sung in translations don’t strictly
respect the rhythmic patterns of the ST and, nevertheless, they are
singable. How strict should the song and opera translator be about it?

PL: To me, rhythm of any translated line can be measured easily (though
crudely): if the stresses fall well in the TL and the syllables match the SL, then
I would say “rhythm 100%”. But I often find that a little flexibility in syllable-
count enables me to achieve a better overall result.

LM: You mention that rhymes are “[...] the easiest criterion to assess, and
usually the least important” (LOW, 2017, p. 80). Is rhyming really so
secondary?

PL: I accept that sometimes it is very important, but you can’t say “this
songwriter always places a great importance on rhymes,” because any good
songwriter will vary, and some might even make songs without rhymes or
10.17771/ PUCRio.TradRev.50547

may be reluctant. One of my favourite songwriters, Tom Lehrer, does

brilliant things with rhymes. He is an American Jew from New York, he
wrote only about twenty songs, but all classic satirical songs. I wanted to
include one in my book, but I couldn’t get access to him or his agents, it was
impossible, and then somebody said: “that is a piece of black humour, some
of your audience won’t like it.” Here’s the start: “I hold your hand in mine,
dear / I press it to my lips / I take a healthy bite / From your dainty
fingertips”. [...].13 Of course that’s humorous, but how would you translate
it? He also had a wonderful song where he had about twelve rhymes all
ended in “-ality” – and then a song about the Periodic Table, which is a
wonderful replacement text. One can’t simply translate it without rhyme,
because that’s part of the whole point – You don’t have a standard formula

13 The exact lyrics to the song “I hold your hand in mine” are as follows: “I hold your hand in mine, dear,/
I press it to my lips / I take a healthy bite / From your dainty fingertips / My joy would be complete, dear
/ If you were only here / But still I keep your hand / As a precious souvenir / The night you died I cut it
off / I really don't know why / For now each time I kiss it / I get bloodstains on my tie / I'm sorry now I
killed you / For our love was something fine / And till they come to get me / I shall hold your hand in
mine”. Retrieved from: https://genius.com/Tom-lehrer-i-hold-your-hand-in-mine-lyrics. Access on Oct
06, 2020.

Tradução em Revista 29, 2020.2 307

 MELLER and COSTA Interview with Peter Low

to translate a song, you have to look closely at the song in question. In any
case, my point is that sometimes the rhyme gives a gusto, an energy to a song
and if you miss that, you’ll lose the clinching effect of a rhyme at the end of
a stanza, i.e., the final line of a group of four or six, where the instrumental
interlude comes in – a clinch, as if you were shutting the door and you hear
a click – and the rhyme gives that in verse. So that’s a structural value, it tells
you where the end of a section of the song is. And also probably at the end
of a sentence, it’s a grammatical thing. But it’s clear to me that there are lots
of vocal music that don’t have rhymes, and that varies in different cultures:
Maori music doesn’t have rhymes, Latin songs, church songs didn’t usually
have rhymes.

LM: In contrast with the examples you have just mentioned, I would say
rhymes are still relatively important in Brazilian Popular Music. Can we
then conclude that the importance of rhymes in songs depends on the
language?
10.17771/ PUCRio.TradRev.50547

PL: That’s interesting – there is a cultural expectation, in Brazilian

Portuguese, and it actually is in English there was as well, i.e., popular songs
have to rhyme. It’s no longer expected in English that poetry will ever rhyme,
but in the past people would say “that’s not a poem, it doesn’t rhyme”, I
mean 200 years ago, and poets were given that status probably because they
knew how to do rhyme. In French rhyming is particularly important because
it provides some of the rhythm of a poem or song. It is not so important in
English, and you can speak for Portuguese. I ask questions like: “Does my
target text need to rhyme at all? Is rhyme essential?” And there are various
reasons why it might be. Rhymes will make the song more memorable, more
euphonic, perhaps more punchy. I think a lot of witty songs are quite like
Tom Lehrer – he’s partly driven by rhymes, and if you don’t rhyme at least
half as much your version is going to lose that punch. Now, this means of
course that you are deciding to use rhyme at the expense of some meaning –
and that’s what this chapter here14 is talking about. I have a defence here on

14See chapter 6 – “Singable translations (B) – rhythm and rhyme” in Translating Song (LOW, 2017, p. 95-
113).

Tradução em Revista 29, 2020.2 308

 MELLER and COSTA Interview with Peter Low

page 104 of a translation of a Shakespeare song by Schubert, because the

word “Kindheit” doesn’t mean “kindness”, it means “childhood” – but it
rhymes perfectly and it fits well enough in my opinion, even though it looks
like a mistake. Above that, I’ve got this important claim: I will stick to a
margin of flexibility – the rhymes won’t have to be as numerous or as perfect
as in the source. The original rhymes may not be observed. If there are, say,
eighteen French words rhyming with “amour”, that’s a sort of virtuoso
thing, I can’t do this is English, you might be able to do it in some languages,
but I will try to get a lot of the rhyming, and if I end up with half the lines
unrhymed, maybe that doesn’t matter provided the clinching rhymes at the
end can give us that particular satisfaction.

LM: Talking about the differences between translations, adaptations and

replacement texts,15 you defend the view that the latter should not be the
object of Translation Studies.
10.17771/ PUCRio.TradRev.50547

PL: Yes. As regards meaning, and although I claim to be an adept of the

fidelity paradigm, I agree that some licences can be taken in favour of some
other aspect that is more important – but not to the point where translations
become adaptations (not to mention replacement texts – which, although a
legitimate cultural phenomenon, do not lie within the scope of Translation
Studies). Once I had an argument with some academics in Britain about that.
I said, “No translating is taking place here, therefore they’re not a proper
object for Translation Studies”. I am not denying the cultural importance of
that – and in fact a good song like “Lili Marlene”, which has been replaced
by many texts. In fact, I’m pretty sure the New Zealand soldiers who
marched during the War were singing a replacement text on the tune of “Lili
Marlene” – it was not a translation, it was a vulgar anti-German song.

LM: Some authors distinguish translations from adaptations in terms of

percentages – texts with a high percentage of sense fidelity between source
language and target language would classify as translations, whereas

15 See chapter 7 – “The place of adaptations”, in Translating Song (LOW, 2017, p. 114-127).

Tradução em Revista 29, 2020.2 309

 MELLER and COSTA Interview with Peter Low

those that take too many liberties would be deemed as adaptations. What’s
your take on that?

PL: I don’t take that approach. My approach is this: “If all significant details
of meaning have been transferred, it is a translation” (LOW, 2017, p. 116) –
and the word significant, of course this is a matter of opinion. But in order to
say that this is not a translation, I would have to say, “this reference to her
mother’s death has been omitted, and that’s a significant point, therefore this
is an arrangement, an adaptation”. In this definition of an adaptation, I’ve
said “some significant details have not been transferred”, with the extra
proviso, “which easily could have been”. I’ll take an example from famous
French song by Jacques Brel, “Ne me quitte pas”. It is significant to me that
that is an imperative – “don’t leave me”. And when the American translator
translated it as “If you go away”, there was a significant part of the utterance
that has been lost. It became a conditional – it assumed that the woman was
thinking of going away, but hadn’t gone away yet. In order though to prove
10.17771/ PUCRio.TradRev.50547

that this is a mere adaptation, I have to say that you could have easily started
with “Don’t walk out on me”, “Don’t leave me”, “Don’t abandon me”. That
could have easily been done, and so the other was an unforced deviation,
addition or modification. It was a good, successful song in English, but the
litmus test was to compare the actual wording. The adaptation has wilfully
modified the source text, and the translation has not wilfully modified it. It
might have reluctantly or accidentally modified it, of course.
I’m not talking about things like “oh, the word order was changed”
– of course the word order was changed, this is a different language! I’m not
saying things like, “Oh look, the source text has three words beginning with
a ‘z’, and the target text has only one” – well, hang on, are these three words
beginning with a ‘z’ actually significant as part of the source? And they
probably aren’t. I’ve got an example on the next page16, which is “Frère
Jacques”. There is a replacement text at the end of the page that goes [he
sings to the melody of “Frère Jacques”] “In the forest / In the forest / There
are trees / There are trees / Johnny saw a squirrel / Johnny saw a squirrel /

16 See LOW, 2017, p. 117.

Tradução em Revista 29, 2020.2 310

 MELLER and COSTA Interview with Peter Low

And some bees / And some bees”. That’s not a translation – it’s a replacement
text! I do take liberties with insignificant details in my versions, and I will
still call them translations. And if you want to prove I have made a bad
translation, you’ll have to show that it could have easily been improved.
That’s pragmatic. But we are not working with perfection here.
But one aspect I don’t really cover is that a solo song is often created
as a vehicle for a performer to add value to sell it to the audience. Of course,
some bad songs get sold so well by a good performer that they become
famous. But if you have a translation that your singer doesn’t feel able to sell,
then you’ll have to do better, I think. I know with some translations you sit
down with a short story and translate it all on the computer by yourself – but
songs are a different matter, there are different people involved, the
stakeholders, including of course the performers and the listeners.
I’ve got a lovely example of an old German song about the woman
who gets flea bites. If you sing it in German to an audience who doesn’t know
German, they are not going to get it, are they? [Sings a fragment of the song
10.17771/ PUCRio.TradRev.50547

in German] – it sounds like German, that’s all! [laughs]. [Quoting from his
book:] “The notion of ‘unforced deviation’ excludes normal changes in
word-order, or the other standard procedures outlined in textbooks and
used regularly by good translators, such as those termed ‘transposition’ and
‘modulation’” (LOW, 2017, p. 116). There is a nice case in English: you don’t
translate “Village vert” as “Village Green”, you translate it as “Green
Village” – because a “Village Green”, in England, is a strip of grass that is
used for the marketplace or for playing cricket on Saturdays – that’s the
“village green”. And of course the word order tells you – in English – the
second word is the noun, and the first one is the adjective.

DC: We are reaching the end of the interview. Could you give us your
input on the evolution of song translation since you started working and
researching it?

PL: I think with novels and short stories that translation theory has improved
the quality of what is published. With song-translation I don’t see that yet –

Tradução em Revista 29, 2020.2 311

 MELLER and COSTA Interview with Peter Low

indeed I see a confusion of adaptations, good translations, and replacement

texts which are not translations at all.

DC: What advice do you have for someone attempting a singable

translation?

PL: Recognise that some features of the song will be lost in translation. So
examine carefully what the songwriter was doing in this particular song, and
identify which of its features are the most important to retain. This analysis
will help you to make the necessary trade-offs in the optimal way. And read
p. 109 of my book!

DC: Everything considered, would you like to compose a “coda” for the
interview?

PL: My hope is that we the translators of songs can increasingly see our role
10.17771/ PUCRio.TradRev.50547

as making people aware of the artistic output of other languages and cultures
– and not (as has often happened) as appropriating, adapting and exploiting
their creativity for our own ends.

References

APTER, Ronnie; HERMAN, Mark. Translating for Singing: The Theory, Art
and Craft of Translating Lyrics. London: Bloomsbury Academic, 2016.
KAISER, Andrea. Óperas no Brasil. Versões em português. 1999. 118 f.
Dissertação (Mestrado em Musicologia) – Programa de Pós-Graduação em
Artes. Escola de Comunicações e Artes. Universidade de São Paulo, São
Paulo, SP. 1999.
LOW, Peter. Surtitles for Opera: A Specialised Task. Babel, Amsterdam, vol.
48, n. 2, p. 97-110, 2002.
LOW, Peter. Translating songs that rhyme. Perspectives: Studies in
Translatology, London, vol. 16, n. 1-2, p. 1-20, 2008.
LOW, Peter. Translating Song: lyrics and texts. New York/London:
Routledge, 2017.

Tradução em Revista 29, 2020.2 312

 MELLER and COSTA Interview with Peter Low

NEWMARK, Peter. A Textbook of Translation. New York, Prentice Hall,

1988.
PARNY, Évariste. Selected Poetry and Prose of Évariste Parny: In English
Translation, with French Text. Edited by Françoise Lionnet. Translated by
Peter Low and Blake Smith. New York: The Modern Language Association
of America, 2018.
10.17771/ PUCRio.TradRev.50547

Tradução em Revista 29, 2020.2 313