Translations for dubbing as dynamic texts Strategies in film synchronisation *

This article aims to describe how written translations of audiovisual products change along the dubbing process until they reach the audience, focusing mainly on the synchronisation stage. The corpus is made up of the first reel of the Catalan and Spanish translations of three different films released in 2006 (Casino Royale, Good night, and good luck, and The Da Vinci Code). After a brief summary of the dubbing process and a short overview of the approaches to synchronisation found in the literature, the specific strategies found in a bottom-up analysis are highlighted, offering both a wide analysis of the general changes throughout the process and a closer analysis of the specific changes that occur in the synchronisation stage. A comparison of the strategies found in Catalan and Spanish is also included in order to assess whether different languages result in different strategies.


Introduction
Dubbing, next to subtitling, has been one of the modalities that has captured the interest of most audiovisual translation researchers for many years.However, there is still need for empirical research to help us understand more profoundly this transfer mode.This article deals with dubbing and aims to carry out systematic corpus-based research on how translations for dubbing can be viewed as dynamic texts which change along a complex process.The corpus is made up of the first reel of three fiction films released in 2006, namely Casino Royale, The Da Vinci Code and Good night and good luck.The translations into Catalan and into Spanish have been analysed and compared to both the synchronized script, the linguistically revised script (if available) and to the final broadcast version in order to identify the changes the text undergoes.Special attention has been paid to the synchronisation stage.
After a brief summary of the dubbing process ( §1), this article presents an overview of the approaches to synchronization found in the literature ( §2) and then highlights the strategies found in the analysis of the corpus ( §3), offering both a closer perspective on the changes made in the synchronisation stage and a wider perspective which takes into account other key stages such as the language revision or the actual recording.A comparison of the strategies in Catalan and in Spanish closes the article and is used to assess whether different languages result in different strategies being employed ( §4).

Dubbing: the process
Dubbing is a complex process in which various agents take part.From translators to dubbing actors a series of professionals work towards a common final goal: offering an audiovisual product in the target language that can be accepted by the audience as a credible illusion.According to Ledesma and López (2002), in Catalonia the following steps can be identified in the process: (i) the original version is sent to the dubbing studio: it includes a master (digital betacam) with the international soundtrack and dialogues.A script might be included, although it is not always available; (ii) a sound engineer makes copies to work with; (iii) the artistic director views the product to be dubbed; (iv) the production manager chooses the translator, the dubbing dialogue writer and the linguist; (v) the audiovisual product is translated and adapted; (vi) the translated product is synchronised; (vii) sometimes there is a language revision: this is normally the case for Catalan television (TVC) but it is not a common occurrence when translating into Spanish; (viii) the spotting is done by the director's assistant while the dubbing director chooses the actors; The result at this stage is a written script in the target language, with indications both for actors and sound engineers, ready for actual recording in the studio.The process continues as follows: (ix) Dubbing, with the dubbing director, actors, a sound engineer and if necessary, a linguist.
The result after the recording is an oral version in the target language. (x) The linguist (in the case of Catalan) revises the oral version and asks for retakes, if necessary; (xi) the sound engineer takes care of the mixing; (xii) the artistic director views the final result and asks for retakes, if necessary; and (xiii) the client (or a supervisor appointed by the client) views the final result and asks for more retakes, if necessary.
After all this long and expensive process, the result is a dubbed version in the target language in which many agents have taken part.Even though all participants in this chain are essential in order to produce a final successful result, my interest lies in the role of translators, dialogue writers and linguists, which will be described in the next paragraphs.
Translators deliver a written version of the original film taking into account it is to be interpreted orally by actors and to be received audiovisually by the audience.The translator captures all the nuances of the original audiovisual product and transfers them into the target language, delivering a version which mimics the features of the original in terms of language variation and adequacy.
The process of synchronisation, which can be done by another professional or by the same translator, consists in adapting the text to achieve different types of synchrony, including a series of symbols to facilitate the actual dubbing, as will be explained in section 2. Some authors add that these professionals also adapt the script so that it sounds natural but from my point of view this task should lie with all agents: that is, the translator must produce credible dialogues and changes made by the dialogue writer must also fulfil this requirement.
Finally, certain clients such as Catalan television (TVC) require a specialised professional to check both the written script and the final recorded version: this is the task of a linguist, who verifies the linguistic adequacy of the text and monitors the pronunciation.

Synchronisation in dubbing
Synchronisation can be defined as "one of the features of translation for dubbing, which consists of matching the target language translation and the articulatory and body movements of the screen actors and actresses, as well as matching the utterances and pauses in the translation and those of the source text" (Chaume 2004).This definition envisages visual synchrony as an umbrella which encompasses lip synch, isochrony and kinesic synchrony, a view recurrently found in the literature (Agost & Chaume (1996: 208), Agost 1999: 65-67, Chaume 2003, 2004, Espasa 2004).
Other approaches to synchrony that probably inspired this classification are those of Fodor (1972) and Whitman-Linsen (1992), two milestones in the description of synchronisation in dubbing.Fodor (1972), in a book whose initial version was published in 1962 proposes the differentiation between phonetic synchronyconcerning lip movement-, character synchrony -regarding the voice of the dubbing actors-and content synchrony -which points to the coherence between the translation and the situation on screen.Whitman-Linsen (1992) A wider overview of the literature in the field can be read in Chaume (2004), who states that four approaches towards synchronisation are to be found: a professional perspective (Martín 1994, Ávila 1997, Gilabert, Ledesma and Trifol 2001), a functionalist perspective (Fodor 1976, Mayoral et al. 1988, Kahane 1990-1991and Zabalbeascoa 1993), a polysystemic approach (Goris 1993, Karamitroglou 2000) and a cinematographic approach (Chaves 2000, Bartrina 2001, Chaume 2003and 2004, Bravo 2003).According to Chaume, the first group proposes a functional viewpoint where dialogue writers are responsible for synchronisation, the second group coincides with the previous group although adopting a more academic approach, the third group focus on naturalization as a translational norm in dubbing and synchrony as the most prominent feature of this naturalization, and the fourth group try to establish links between film language and translation and analyse synchronisation as "one of a set of elements in the broad network of signs that make up the message, the film and the narration" (Chaume 2004: 42).
As for national differences in this practice, Pommier (1988) offers a thorough introduction to the technical procedures used in France, where the bande rhythmo is used, and Chaume (2007) presents a comprehensive overview of the different practices used in Europe that comes to show the absence of common standards.

Strategies used in synchronisation
Although as shown in the previous section some authors have described synchronisation -also called adaptation or adjustment-and a few have even listed some of the strategies used in order to synchronise the original text (Chaume 2005, Mayoral 2003), to the best of our knowledge no corpus analysis have been provided to demonstrate the changes translations undergo before they reach the audience.The aim of this article is to fill this void, by providing a categorisation which focuses on synchronisation but also includes further changes such as those that take place during the language revision or even later in the process, in the actual recording.Instead of starting from an already established categorisation, a bottom-up approach is followed, proposing a classification resulting from the data analysis.
As already stated, the corpus consists of the first reel of three fiction films for each film, and a common analysis taking into account the language (Spanish/Catalan) is also offered.
Figure 1 summarises the changes observed in the first reel of the Da Vinci Code both in Catalan and in Spanish.In the first column one can see whether the change belongs to the synchronisation stage (i.e.changes found in the synchronised scripts that are compared with the written translated script), to the language revision (i.e.changes found in the linguistically revised script compared with the synchronised script) or to the actual recording (i.e.changes included in the broadcast version but not in the written script, probably due to improvisations or last-minute changes proposed by either the dubbing actor, the dubbing director or the supervisor).In the second column different strategies have been identified following a bottom-up approach.
• Both in the synchronisation stage and in the recording: (2) Original: I was supposed to have drinks with him earlier this evening.(3) Original: Professor Langdon, do not react to this message.(6) Original: Ah, the Grand Galerie.
Repetition, with 11.63% in Catalan and the significantly lower percentage of 3.85% in Spanish, could be considered a variation of the last strategy since it is used to make sentences longer, as in the following example, in Catalan: often used to reach lip-synch, whilst modifications are hardly present and in some cases, such as (9), belong to linguistic issues ("vespre", literally 'evening', is more adequate than "nit", literally 'night', due to the time sequence of events).On the contrary, modifications in Spanish occupy third place with a remarkable 34.61% (10).(9) Original: Tonight this will be our quest.As far as language revision is concerned, only the Catalan synchronised translation is altered because language editors do not usually take part in the Spanish dubbing process.In Catalan, most changes belong to issues related to both style and register, and also include grammar and vocabulary problems.Some minor changes affect the content or amplify/shorten the translation.
Finally, with regard to the final changes that find their way into the broadcast version, it is remarkable to notice that a high number of changes are found in the Spanish version, probably due to the fact that no language revision is foreseen in the process, whilst in Catalan these changes are kept to a minimum.

Fig. 2. Casino Royale
In this case, most changes occur in the synchronisation stage, although the percentage is higher in Spanish probably due to the fact that the text does not undergo a language revision.As for the strategies in synchronising the translation, reduction ranks first in both languages, with the most used ones being the deletion of unnecessary verbs, the deletion of discourse markers the deletion of expressive units and the use of synonyms (see 13 in Catalan and 14 in Spanish).
(13) Original: I wonder if bomb makers are insured for things like that.
Translation: No sé si els fabricants de bombes tenen una assegurança que els ho cobreix [I don't know if bomb makers have got insurance that covers that].Concerning language revision, very few changes are made in the Catalan version, with those that are, mostly concerning vocabulary and style.In (18) the translator used the noun "banquers" instead of the adjective "bancari", which is the most appropiate in this particular context.(20) Original: I believe in a reasonable return of rate.

Synchronised version:
Translation: Creo en una tasa de retorno razonable [I believe in a reasonable return of rate].

Synchronised version:
Creo en una tasa de retorno razonable óptima [I believe in an optimal return of rate].
Finally, figure 3 presents the results for the film Good night, and good luck.probably due to the fact that translations are done from English into Romance languages such as Catalan and Spanish which use longer constructions to deliver the synthetic English speech.However, it is also striking to notice that in all cases amplification, i.e.
adding information, is the second most frequent strategy used in order to adapt the translation.There is no doubt that wider research which encompasses more translators and dialogue writers would shed more light on this issue and would go beyond personal practices.

Dubbing into Catalan and into Spanish: a comparison
In this section we aim to analyse whether different languages result in the use of different strategies; hence, instead of analysing the data taking each film into account, results are grouped by language and summarised in figure 4.  Taking a wider perspective, a striking difference is found between Catalan and Spanish: whilst in the former most changes are carried out in either the synchronisation stage or the language revision stage, in the latter the maximum number of modifications occur during synchronisation, followed by changes made during the recording.This comes to show that language revision can be envisaged as an extremely useful filter which allows the recording to run smoothly and faster.
Looking closer at the data in each category, it can be seen that in the first stage, i.e. synchronisation, the two main strategies are the same -reduction and amplification-and their percentages are quite similar, probably hinting at the importance of isochrony.On the contrary, minor strategies such as repetition, modification or changed orders present variable percentages.
As for the specific strategies used in each category, it is not the aim of this article to present a detailed list of them but figures 5 and 6 can offer an overview of the main ones found in the analysis.As seen in figure 5, when making translations shorter, deletion and the usage of synonyms are found in similar percentages whilst changed structures or units are employed in lower percentages.As for amplifying strategies, these are generally achieved by adding different types of units (whole structures, emphatic units, interjections, discourse markers, verbs, pronouns, etc.) although longer synonyms are also used and verb tenses are sometimes changed, as shown in the following percentages.

Strategy Occurrences %
Addition of structures such as "I think", "I'm sure", etc.
1 1.69%  Finally, in the final stages of the process, changes are kept to a minimum in Catalan whereas in Spanish reductions, modifications and amplifications are abundant.
This could serve to demonstrate the usefulness of having language editors in a previous stage as a means of avoiding unnecessary changes at a later stage, although it should not be forgotten that external supervisors usually exercise stronger control over Spanish versions, usually to be shown in cinema theatres, rather than Catalan versions, more often released on DVD.

Conclusions
In summary, this article has presented a small-scale study into the changes translations undergo before they reach the final audience, by focussing mainly on the synchronisation stage but also on other significant phases such as language revision and recording.Although only the first reel of three fiction films has been used, some interesting findings on recurrent patterns have arisen, hopefully opening new research avenues in the field of dubbing.Further detailed and systematised studies would not only contribute to a thorough description of the discipline but could also prove extremely helpful in teaching a craft which is slowly finding its way onto university courses.
In our opinion, these future studies could tackle three issues: (i) a description of the specific strategies used in each category; (ii) the usage of a wider corpora, incorporating other languages, and (iii) an analysis of the relationship between the synchronisation strategy and each shot (close-up, long shot, actor on/off-screen, etc.), stating which type of synchrony is achieved (lip-synch, isochrony, kinetic synchrony).
released in 2006 (Casino Royale, The Da Vinci Code and Good night and good luck), both in their Catalan dubbed version and Spanish dubbed version.I have compared the translated script with the synchronised script, the linguistically revised script (if available) and with the final broadcast version.The results are then presented separately

:
Aquesta tarda havia quedat amb ell per prendre una copa [This afternoon I was meeting him for a drink].Synchronised version: Aquesta tarda havia quedat amb ell [This afternoon I was meeting him].

( 5 )
Original: Is there a restroom I could use?Translation: Podria anar al lavabo?[Could I go to the restroom?]Synchronised version: Escolti, podria anar al lavabo?[Listen, could I go to the restroom?].

( 7 )
Original: Why would someone do this to him?Translation: Per què li han fet això?[Why has someone done this to him?]Synchronised version: Per què, per què li han fet això?[Why, why has someone done this to him?]Repetition is the third strategy used in Catalan in this film next to a change of order (8),

( 8 )
Original: I'll take another look when I come back.Translation: Ja m'ho tornaré a mirar quan torni [I'll take another look when I come back].Synchronised version: Quan torni ja m'ho tornaré a mirar [When I come back, I'll take another look].
No sé si els fabricants de bombes tenen assegurança [I don't know if bomb makers have got insurance].(14) Original: If M was so sure I was bent, she'd have sent a Double-O.Translation: Si M estuviera tan segura de mi corruptibilidad, habría enviado a un doble cero [If M was so sure of my corruptability, she would have sent a Double-O].Synchronised version: Si M supiera que soy corrupto, habría enviado a un doble cero [If M was sure I'm corrupt, she would have sent a Double-O] Amplification is the second most frequent change made to the translation in both languages ( 15 and 16), whilst modification and altered orders occur with some frequency in Catalan (17) and infrequently or not at all in Spanish.(15) Original: Looks like our man.Burn scars on his face.Translation: Sembla el nostre home.La cara cremada [Looks like our man.Burnt face.]Synchronised version: Sembla el nostre home.Té la cara cremada [Looks like our man.(He) has a burnt face.](16) Original: He's on the move.He's on the move and he's heading straight for me.Translation: Se mueve.Se mueve y viene directo hacia mí [He's moving.He's moving and heading straight for me.]Synchronised version: Se ha levantado.Se ha levantado y viene directo hacia mí [He's stood up.He's stood up and heading straight for me.](17) Original: And I can access it anywhere in the world?Translation: Podré accedir-hi des de qualsevol punt del món? [I can access it anywhere in the world?]Synchronised version: Hi podré accedir des de qualsevol punt del món? [I can access it anywhere in the world?]

( 18 )
Original: I suppose our friend Mr. White will have told you that I have provided reliable banking services for many other freedom fighters over the years.Translation: Suposo que el nostre amic, el senyor White, ja li ha comentat que fa anys que presto serveis banquers fiables a molts defensors de la llibertat [I suppose our friend Mr. White will have told you that I have provided reliable banking services for many other freedom fighters over the years].Synchronised version: Suposo que el nostre amic, el senyor White, ja li ha comentat que fa anys que presto serveis bancaris fiables a molts defensors de la llibertat [I suppose our friend Mr. White will have told you that I have provided reliable banking services for many other freedom fighters over the years].Finally, regarding the changes made either by actors, external supervisors or dubbing directors in the final stages of the process, the aim is to amplify the text in the Catalan version (19) and either reduce it or modify it (20) in the Spanish dubbed version.(19) Original: I suppose that's something.Translation: Ja és alguna cosa [(That) is already something].Synchronised version: Això ja és alguna cosa [That's already something].

( 23 )
Original: And if there are any historians about fifty or a hundred years from now.Translation: I, si d'aquí a cinquanta o cent anys queden historiadors... [If in fifty or a hundred years there are historians...].Synchronised version: I, si d'aquí a cinquanta o cent anys encara queden historiadors... [If in fifty or a hundred years there are still historians...].

Figure 4 .
Figure 4. Results in Catalan and Spanish

Fig 6 .
Fig 6.Strategies used in amplification (synchronisation stage) changes are due to either erroneous translations which alter the meaning of the original or wrong data.The third, fourth and fifth column contain the information concerning the Catalan version: the number of occurrences for each strategy, the percentage of times this strategy is used within each group (synchronisation/language revision/recording) and finally the percentage of changes found in each of these stages.The same data is offered for the Spanish version on the right.
the number of syllables of the original is reduced; o repetition: some words are repeated; o amplification: the text is made longer; o modification: the length of the utterance is kept the same but the sentence is altered; o changed order: the same words are kept but in a different order, and o deletion: some part of the utterance is deleted but the rest remains the same.• In the revision stage: o amplification, as defined previously; o reduction, as defined previously; o grammar, referring to changes made due to grammatical reasons; o vocabulary, meaning lexical units that have been changed; o style/register: changes made in order to follow the style-sheet of the television network or because the register is not adequate, and o content: when (1) Original: Your Embassy called Division.Translation: La seva ambaixada ha trucat a la divisió [Your Embassy called Division].Synchronised version: La seva ambaixada ens ha trucat [Your Embassy called us].
We don't pick and choose which rules to follow.
(10) Original: Interpret for me, please, this symbol.First thing that comes to mind, anybody.Translation: Synchronised version: ¿Me está diciendo que el último acto de Saunière en la tierra ha sido dibujarse el símbolo de una diosa en el pecho?[You're telling me that, uh, Saunière's last act on earth was to draw a goddess symbol on his chest?] Modifications found atthis stage in the Spanish version mainly include amplifications (11) and reductions(12)which are achieved by means of longer or shorter forms, repetitions and deletions, addition or deletion of emphatic words, etc.(11) Original: You're telling me that, uh, Saunière's last act on earth was to draw a goddess symbol on his chest?Translation: ¿Me está diciendo que lo último que ha hecho Saunière ha sido dibujarse el símbolo de una diosa en el pecho?[You're telling me that, uh, the last thing Saunière did was to draw a goddess symbol on his chest?]Thepreviousdata corresponds to the first reel of Da Vinci Code but two other films were also analysed.The data for the first reel of Casino Royale is included in Figure2and is summarised next.
Again, most changes occur in the synchronisation stage, although a remarkable percentage of changes are made in the language revision of the Catalan version.As for the alterations in the actual recording, they are kept to a minimum in both versions, with few reductions, modifications and amplifications.It is my desire, if not my duty... Es mi voluntad y mi deber... [It is my desire and my duty].