When I first offered my services as a translator to film and TV producers who did dubbing and postsynchronization here, in Montréal, I was told they did not need translators, but “adapters.” I remember arguing that translation was adaptation, that you did not just copy words from a dictionary automatically, that every sentence, every word, every comma, was the object of a decision, of a transposition into another context, not just linguistically, but also socially and culturally. But these people had a set idea of what a translator did, and they couldn’t be moved. It was only when I told them I also translated poetry and songs that I got their attention. That, they knew, required adaptation.

At first, I was given documentaries to “adapt” into French. Well, there wasn’t much adaptation to be done here. It was mostly straightforward translation. Except you had to take into account certain elements you would not necessarily pay attention to in a written translation. For instance, time and tempo. In the case of a narration, the translation has to follow the image and refer to its various elements as they appear on the screen. So, even if an inversion would be more appropriate in the target language, it may not be indicated if it does not suit the picture. The illusion to be maintained always is that the audience is watching an original production.
When a voice-over technique is used, the length of your text has to correspond to the length of the speaker’s text, and even be shorter. That’s when you see the speaker on the screen and a translation is supplied so that you hear the person speaking a foreign language in the background and an actor’s voice is “voiced over” that, drowning it and taking its place, though the audience always hears the foreign language in the back. This provides the illusion that an interpreter has stepped in and is simultaneously translating what that person is saying. The studio actor/interpreter does not play a role with emotions as if it were acted out. He or she is merely an interpreter who repeats what the person is saying. Here the difference between a translator and an adapter is that the adapter must make this speech—which is often improvised—sound as if it were a well thought out discourse. No hesitations, no ungrammatical sentences, no interrupted utterances, no mistakes—unless they contribute to the scenario. The studio actor must have a flawless text, even if the character on the screen is hesitant or speaks English as a second language, not mastered very well.
With documentaries, the constraints that require adaptation are thus mostly timing and grammatical soundness. Most documentaries have both narration and voice-over, but in all cases the main objective is to give the audience the illusion that they are watching an original production.
After a while, I was entrusted with documentaries where there was also some synchronous dubbing, then with feature films and TV series where everything was dubbed. This is where the screen actor says the lines in one language, but the audience hears them in another. The illusion to be maintained is that the studio actor’s voice belongs to the screen actor. This is achieved through various techniques. The one we generally use in studios here is the rhythmo band.

No, this is not a group of Latin music performers. It’s actually a 35 mm film strip on which the text and voice noises are hand written in black ink by a calligrapher, so that the studio actors can substitute their voices for those of the screen actors. My task is to provide the calligrapher with a text to copy. To do this, I am given a “mother band,” a dull matte white 35 mm film strip on which someone has written down all the text and voice noises as performed by the screen actors in synchronicity with the image. I write my translation/adaptation with a lead pencil above or below these markings. Everything is noted on the mother band, not only the text, but also the inspirations, expirations, hesitations, smacking of the lips (as in a kiss), everything. Because the studio actors will have to reproduce all this.
Sometimes the original soundtrack is not synchronous with the original image. Indeed, as you may know, most films and TV productions nowadays are shot without sound and the soundtrack is added afterwards in a studio. This is called postsynchronization. And with the increasing number of international casts in contemporary productions, often the voice you hear in the so-called original version is not the voice of the actor you see on the screen. Even as early as in La Strada, Anthony Quinn did not speak Italian and had to be dubbed. But I’m getting carried away and losing sight of my topic.
The writing out of the original text and voice noises onto the mother band is called “detection.” The “detector” uses an editing machine to unwind the white strip of 35 mm film synchronously with the film he’s “detecting” and writes the text with a lead pencil on this white strip at the exact place where the voice is heard in the film. He stretches stressed vowels and shortens the unaccented ones, writes “probly” for “probably” in English, or “habm” for “haben” in German, if that is how these words are actually pronounced. And so on and so forth. I say “he” because the detector I’m working with on the series I’m presently adapting from German is a man, but there are also women detectors (and they are as competent and as well paid as their male counterparts).
The detector also uses a number of conventional signs to indicate if the mouth is closed or open at the onset and at the end of each utterance. For instance, the word “stop” may be pronounced with a momentary closing of the mouth on the voiceless bilabial stop “p,” then an opening of the mouth after, when the air is released. But if “stop” is the last word of the utterance, the mouth may also remain closed after the “p” is pronounced. Try it. See? That means that the translator/adapter cannot add a sound after the bilabial. For instance, I couldn’t put “départ” or “débat,” since the audience would see the mouth closed and still hear a sound after the bilabial consonant. I’ll get back to this later.
The detector also underlines those bilabial consonants (b, m, p) wherever they occur, and puts a small circle underneath the semilabials (f, v, w, English retroflex “r”). I’m expected to put bilabial consonants over the bilabial consonants in the original text, so that the viewer can be tricked into believing that the screen actor is really pronouncing what the studio actor is saying. I can put a “b” over an “m” or a “p,” I can even put a semilabial over a bilabial. Or I can put a “b” over an “f” or a “w,” and so forth. And I can skip some of these when there is a long succession of such consonants. For instance, “probablement” has four bilabials while “probably” only has three. But that is part of the illusion of cinema.
We’re talking about illusion after all. I mean, we know that when actors are gunned down in an action movie, they don’t really bleed to death. And when they say “I love you,” they don’t always mean it in real life. In the same way, this whole undertaking is meant to maintain an illusion: synchronism.

This is where adaptation comes into play. Synchronism is coincidence in a point of time. There are three kinds of synchronism: phonetic synchronism, semantic synchronism, and dramatic synchronism. Phonetic synchrony is achieved when the lip movements of the screen actor match perfectly the sounds produced by the studio actor, not only words, but also breathing, grunts, screams, etc. Actors do that in the studio, even if they are invisible. They make gestures, and get into their roles. I’m almost surprised when I see they don’t wear costumes.

When I started translating/adapting TV series and telefilms for dubbing, I became obsessed with phonetic synchronism. I watched people’s mouths as they spoke, not only in films and on TV but at the grocery store and on the bus, everywhere. I was literally reading everybody’s lips. I looked at myself in the mirror and made faces as I pronounced words or sentences. I dug out my old books on phonology and brushed up on definitions of articulators and articulation points.
Word order is tricky when translating for dubbing from English into French. “Independence Day” is “Jour de l’Indépendance.” Sometimes, you can cheat: if the mouth of the man on the screen is open before he starts speaking, you can sometimes squeeze in a couple of syllables before the first word, provided they do not contain bilabial consonants which would require him to close his mouth. As long as his mouth is open, the audience does not know if he’s saying anything. They trust what they hear. Illusion. So, you could probably have the studio actor say “jour de” without anybody noticing.
As a result, my first adaptations were perfectly synchronous from a phonetic point of view, but the sentences were a bit twisted on the grammatical and lexical levels. I mean, no formal mistakes, but plenty of awkward constructions. I had to fit the word to the lips at all cost. I could do it. I experienced victory every time I managed to turn a phrase that would match the articulatory mechanisms of the source language, never mind if it sounded a bit weird in the target language. I figured the audience wouldn’t notice. They’d be so caught up in the action and so impressed by the perfect phonetic synchronism of this version that they wouldn’t notice it was strange for a cop to speak of the “décédé” (deceased) for the “dead,” because it fit the lip movement better than “mort.” I ended up writing what Jim Palmer, a friend and collaborator of mine, calls “dubbage.”
It wasn’t all that bad. I’m being hard on myself here, and I’m looking at my early work with the benefit of 15 years of experience. What I mean is that this phonetic constraint can be overwhelming and can lead to being blinded to other constraints which should in fact have priority.

Semantic synchronism

Obviously a translator’s main objective is that the translated text have the same meaning as the underlying text. So semantic synchronism is a priority, even more so than phonetic synchronism. But not always. For instance, there are cases where a number can be replaced with another number that better fits the lip movement if no damage is done to the overall meaning of the scene. Take the number two, “zwei” in German; in French, “trois” would be more synchronous than “deux,” but I have to ask myself, is it possible to replace it in this scene? Is the number crucial, or could it be any number? This is where judgment and adaptation come in.
Let’s look at an example taken from a German television police series I’m working on now. In one of the episodes, there was a reference to Frau Dusward, the owner of the apartment, “die Besitzerin der Wohnung,” but because of word order in French, I could not call her “la propriétaire.” There was no place to put the bilabial “p”s. So, I simply said she was the tenant (la locataire) of that particular apartment. It did not make any difference to the plot anyway.
In most cases, though, phonetic synchronism must be sacrificed in favour of semantic synchronism. When translating/adapting an educational video on mathematics or physics, it is essential that the vocabulary be scientifically accurate. The only phonetic constraint that must necessarily be observed is that the voice must not be heard once the speaker has finished speaking, nor must he be seen blabbing away while no sound is uttered. This can be achieved by condensing or filling the text as appropriate. There are also a certain number of culturally fixed phrases. “To be or not to be” cannot be translated any other way than “Être ou ne pas être.” So the adapter has to use his judgment and his sense of compromise in any situation.

Yet another constraint that takes precedence over phonetic concordance is dramatic synchronism. It’s important that the characters speak with a certain amount of realism. If they shake their heads (at least in most European languages), they should be saying “No,” if they nod assent, the sentence should be affirmative, even if there’s no phonetically congruent expression available. Language level, use of idiomatic expressions, realism are all factors to be taken into account. The character has to “sound” real. Then, if his lips don’t follow exactly the flow of the sounds, it won’t appear so strange to the audience. The audience must never be surprised by the text, unless that is the intent in the original audiovisual document.
Accents are difficult to deal with, because an equivalent cannot always be found. Generally to distinguish a character with a British accent from an American in a given scene dubbed into French, adapters rely on diction and choice of words. The British character will tend to speak like an aristocrat and thus set himself off from the other characters. But what if all the characters are soldiers and they’re all working class? The adapter must somehow give additional information on the origin of the character in the text itself, sometimes in a comment from another character. Usually, though, the characterization is achieved on the screen by the actor’s dress or by his way of delivering his lines or his physical attitude. I’m thinking of Steve McQueen with his chewing gum, his baseball and his mitt in The Great Escape/La grande évasion.
I remember translating/adapting into French an animated British feature film entitled Truckers/Les Voyageurs, where urban gnomes were contrasted with country gnomes. The country gnomes spoke in a characteristic Yorkshire dialect, which set them apart in their speech. I therefore gave their French voices a Canadian Acadian dialect, which is conveniently similar to the Berrichon accent in France. In French, the country gnomes used archaic syntactic forms, such as “j’avons, j’avions, j’aurions” for “j’ai, j’avais, j’aurais” and “il avont, il aviont, il auriont” for “il a, il avait, il aurait.” Of course, I had to fight a little with an editor at Radio Canada, our national broadcaster, to have them accept this apparent deviation from the grammatical norm, but I finally won, thanks to the support of Isabelle Laffont of Les éditions Robert Laffont, who were sponsoring this production.

Here are a few examples of adaptation taken from the German police series I’m currently adapting into French. In episode 19 of the Soko 5113 series, a man was unjustly convicted for a murder he hadn’t committed and spent seven years in prison before his innocence was recognized. Seven years, “sieben Jahre,” occurs repeatedly throughout the episode in different contexts. Each time the problem was to cover the “b” in French, because “sept ans” has no bilabial consonant (in spite of the spelling). Each time, I had to rely on a different scheme.

Er schuldet mir sieben Jahre.
Il me doit sept ans maintenant.

Ich habe sieben Jahre... über nichts anderes nachgedacht.
Tu t’imagines, pendant sept ans, j’ai pensé à rien d’autre tout le temps.

In sieben Jahren hat der es aber vom Verkaüfer weit gebracht
Notre vendeur en a fait du chemin depuis sept ans, tu trouves pas?

In France, at SACEM (Société des auteurs, compositeurs et éditeurs de musique), those who translate/adapt films and TV broadcasts for dubbing are called “dialoguistes,” screen writers. In fact, that’s exactly what we do. We write dialogues for the screen, except the lines have already been spoken by the screen actors and we have to find a text that fits their lip movements and the length of the utterance, as well as their gestures, the situation, the character, and the setting, not to mention what they are actually saying.
So I end up watching