A Challenge in Audiovisual Translation
By Rocío Baños-Piñero & Frederic Chaume (London Metropolitan University & University 'Jaume)
Abstract & Keywords
Creating fictional dialogues that sound natural and believable is one of the main challenges of both screenwriting and audiovisual translation. The challenge does not lie so much in trying to imitate spontaneous conversations, but in selecting specific features of this mode of discourse that are widely accepted and recognised as such by the audience. The main purpose of this article is to analyse and describe the linguistic code in an audiovisual corpus, focusing on what is specific to audiovisual texts and, therefore, to audiovisual translation. Although this code is common to all texts that need to be translated, it stands out further in audiovisual texts since they are “written to be spoken as if not written” (Gregory and Carroll, 1978: 42). We are therefore dealing with texts whose orality may seem spontaneous and natural, but which is actually planned or, as Chaume (2004a: 168) terms it, ‘prefabricated’. Since this is a characteristic that is common to most audiovisual fictional texts regardless of their origins, our aim here is to describe the main features of the linguistic code in native and foreign productions (dubbed from English into Spanish) and to highlight the trends when writing and translating these texts, in order to compare them at a later stage.
Keywords: audiovisual translation, dubbing, oral discourse, fictional dialogue, screenwriting, conversation, prefabricated orality
©inTRAlinea & Rocío Baños-Piñero & Frederic Chaume (2009).
inTRAlinea Special Issue: The Translation of Dialects in Multimedia
Edited by: M. Giorgio Marrano, G. Nadiani & C. Rundle
This article can be freely reproduced under Creative Commons License.
Permanent URL: http://www.intralinea.org/specials/article/1714
Creating fictional dialogues that sound natural and believable is one of the main challenges of both screenwriting and audiovisual translation, especially in the case of dubbing. The challenge does not lie so much in trying to imitate spontaneous conversations, but in selecting specific features of this mode of discourse that are widely accepted and recognised as such by the audience. The implementation of those features should not hinder the understanding of the dialogue, but must succeed in recreating a spontaneous-sounding conversation. As a result, many authors (Chaume, 2001 and 2004a; Mason, 1989; Whitman, 1992, among others) have concluded that fictional dialogue consists of a combination of linguistic features used in both spoken and written texts, and that both translators and scriptwriters should aim to achieve a balance between speech and writing.
Bearing in mind the critical role that fictional dialogues play in audiovisual texts and their translations, it is worth delving into the study of the resources available to recreate spontaneous dialogue on the screen. Considering the multiple semiotic codes which operate simultaneously in audiovisual texts, the main purpose of this article is to analyse and describe the linguistic code in an audiovisual corpus, the focus being on what is specific to audiovisual texts and, therefore, to audiovisual translation. Although this code is common to all texts requiring translation, it stands out more in audiovisual texts since they are “written to be spoken as if not written” (Gregory and Carroll, 1978: 42). We are therefore dealing with texts which are characterised by a strange kind of oral discourse, an orality which may seem spontaneous and natural, but which is actually planned or, as Chaume (2004a: 168) terms it, ‘prefabricated’. Since this is a characteristic that is common to most audiovisual fictional texts regardless of their origins, our aim here is to describe the main features of the linguistic code in native and foreign productions (dubbed from English into Spanish) and to highlight the trends when writing and translating these texts, so that they can be compared at a later stage. We will then follow a descriptive empirical methodology by analysing, selecting and comparing some oral and written features of fictional dialogues from both a domestic sitcom in Spanish (Siete Vidas) and a translation for the dubbing of the TV series Friends. This comparison between two products in Spanish – one original, the other a translation – will reveal the extent to which translations try to imitate, on the one hand, domestic audiovisual genres and norms and, on the other, the spoken target language. Moreover, a comparison of these texts will allow us to determine if, as suggested by other studies (Baños, 2006; Romero Fresco, forthcoming), domestic audiovisual texts bear more resemblance to spontaneous conversation than dubbed texts.
2. Prefabricated Orality in Original and Dubbed Fictional Dialogues
Prefabricated orality is common to most original and dubbed audiovisual programmes based on a script to be interpreted as if it had not been written, especially fictional texts. Scriptwriters have numerous linguistic resources at their disposal with which to elaborate believable dialogues that, despite having been carefully planned in advance, can be identified by viewers as true-to-life conversation. Depicting realism through dialogues seems to be one of the keys to creating a successful audiovisual programme, according not only to professionals in the field (Comparato, 1993; Toledano and Verde, 2007), but also to scholars (Chaume, 2004a; Whitman, 1992) and viewers.
However, several factors seem to put a limit on scriptwriters’ creativity and actors’ improvisational skills. Firstly, scripts must comply with the constraints of audiovisual media, as well as with the conventions of the specific genre the programme belongs to. Secondly, the production of audiovisual discourse is governed by norms which have been consolidated since the creation of cinema and television – such as the use of formulaic language (Pavesi, 2008: 93) – and which might differ depending on the media where the text is to be broadcast (television, cinema, Internet). The selection of particular linguistic features aimed at mirroring spoken speech will ultimately depend on what is considered as acceptable in the system the audiovisual text belongs to, and on the varied factors which operate on that system (see Karamitroglou, 2000: 71-81).
The specific mode of discourse of audiovisual programmes also has direct consequences on its translation: translators must be aware that the original script they receive from the translation agency or dubbing studio, which needs to be translated taking into account the multiple signifying codes that operate simultaneously in audiovisual texts, has been written to be spoken in such a way as to appear spontaneous. Audiovisual translators must therefore be skilful at imitating spontaneous-sounding conversation in the target language. In a sense, the translator of audiovisual texts could be considered as a second scriptwriter, whose task is to transfer the exchanges on screen in such a way that they sound believable in the target language, and could thus be identified as true-to-life dialogues and easily understood by the target audience.
Mirroring spontaneous conversation in dubbing might not seem to be too complicated as, unlike subtitling, spoken linguistic features of fictional dialogues are also transferred using the spoken mode in the target language. However, in addition to the above-mentioned restrictions to scriptwriters’ creativity and actors’ improvisational skills, and the constraints of dubbing as a type of audiovisual translation (lip synchrony, kinesic synchrony and isochrony, see Chaume, 2004b: 43-45), one must take into consideration the tacit norms set by dubbing studios in dubbing countries that ‘enforce’ the linguistic and stylistic standardisation of dubbed fictional dialogues (Ávila, 1997: 25).
Concerns about the corruption of the Spanish language through dubbing and subtitling have been manifested through the publication of style guides and books aimed at regulating the correct use of Spanish in the media. Taking these concerns into account, the absence of a publication providing guidance for translators specialised in dubbing and subtitling in Spanish is paradoxical, to say the least. However, a work of this kind has been published by Televisió de Catalunya (1997) to control the quality of audiovisual programmes dubbed into Catalan: Criteris lingüístics sobre traducció i doblatge (Linguistic Criteria for Translation and Dubbing; see its updated online version esadir.cat The authors of this style guide argue that both translators and dialogue writers should imitate spontaneous spoken speech using appropriate spoken linguistic features in Catalan, but without failing to observe the grammar rules and conventions of this language (Televisió de Catalunya, 1997: 11). The differences between the discourse of domestic and dubbed audiovisual texts are also pointed out in these linguistic criteria, which suggest that some linguistic features of spoken Catalan should be exclusive to native programmes and thus not be used in audiovisual texts dubbed into Catalan.
This statement brings us back to the key concept of this article: prefabricated orality in both original and dubbed sitcoms. It also invites us to reflect on what makes domestic programmes stand apart from dubbed productions from a linguistic point of view and on what kind of norms govern the production of their discourses.
Based on Baños (2006) and Chaume (2004a), a polysystemic and descriptive approach has been adopted in this study, according to which the native and dubbed texts under study are seen as part of the Spanish audiovisual polysystem. We thus consider that audiovisual texts should not be analysed as an isolated phenomenon, but as elements belonging to a complex, dynamic system that in turn is made up of numerous subsystems. All kinds of audiovisual formats and texts, regardless of their genre and origins (domestic or foreign), could be accommodated within this audiovisual polysystem. For the purposes of this article, a clear distinction could be made between the subsystem comprised of native texts and that comprised of (translated) foreign texts, that is, between domestic and foreign productions. Therefore, the texts of our corpus would belong to different subsystems (Siete Vidas belongs to the domestic production or native texts subsystem, whereas Friends would belong to the foreign production or translated [dubbed] texts subsystem), but the two would coexist within the same polysystem (i.e. the Spanish audiovisual polysystem).
Following Toury’s terminology (1995: 59), our study focuses on the analysis of ‘textual-linguistic norms’. The aim is therefore to describe the linguistic features used in the elaboration and the translation of spontaneous-sounding fictional dialogues in audiovisual programmes – domestic and dubbed sitcoms (translated from English into Spanish). The ultimate purpose of this description is to identify trends or norms in the production and translation of prefabricated orality in both discourses. This involves accepting that the production of dialogues mirroring spontaneous spoken speech in original and dubbed texts is governed by specific textual-linguistic norms which, despite being active in the same target polysystem, might not coincide (Baker, 1996: 177; Toury, 1995: 59).
In order to carry out the descriptive analysis of textual-linguistic features of prefabricated orality in Spanish, a monolingual and comparable audiovisual corpus (consisting of four episodes of the North American sitcom Friends and two episodes of the Spanish sitcom Siete Vidas) was compiled taking into account several criteria regarding their suitability and their similarities (see below). An analytical framework was also developed from the findings and works of several scholars (Biber et al., 1999; Briz, 1996; Briz and Grupo Val.Es.Co, 2000; Chaume, 2001 and 2004a; Vigara, 1980 and 1992), with the aim of describing the main features of prefabricated orality in both native and foreign productions at the microtextual level.
3.1. Audiovisual Corpus
Our corpus consists of dubbed and original domestic fictional television series, both broadcast in Spain, belonging to the subgenre of situation comedies. Thus, it could be clearly divided into two sub-corpora: domestic texts and dubbed texts. The native sub-corpus comprises two episodes of the Spanish TV series Siete Vidas, as aired on Spanish television, and their respective scripts, which have been compared with the final audiovisual product. The dubbed sub-corpus consists of four episodes of the North American TV series Friends, and their respective scripts. The duration of each episode is the main reason for the difference between the number of episodes included in each sub-corpus, as episodes of Friends normally last 25 minutes, whereas the duration of the Spanish sitcom is nearly double (approximately 50 minutes). Both components were thus considered as comparable as far as their duration was concerned, i.e. we are then comparing 100 minutes of each TV series.
Several well-thought-out criteria were taken into consideration when selecting the corpus: a) the text type and subgenre; b) the two specific series; and c) the six specific episodes. The parameters taken into account when selecting fiction television series belonging to the subgenre of situation comedies are related to the suitability and potential of these products as candidates for the study of Spanish prefabricated orality. Fictional dialogue is a clear example of prefabricated discourse: dialogues have been carefully planned but are intended to emulate a natural and spontaneous-sounding conversation. In addition, sitcoms aim to recreate typical domestic situations by using dialogues which resemble everyday speech, and they constitute an excellent example of genre influences between systems (Gómez Capuz, 2001).
Friends and Siete Vidas were selected because of their similarities as well as their position in the Spanish audiovisual polysystem. A North American series was chosen due to the overwhelming presence of these products on Spanish television (Eurofiction, 2004: 13). Regarding the position of the chosen texts in the system, these products seem to reflect the trends governing the Spanish audiovisual fictional polysystem: Friends is considered to be a semi-cult series in Spain, followed by a small number of very loyal fans, whereas Siete Vidas has not only achieved very high ratings, but is also perceived as a domestic sitcom of excellent quality, supported by authentic and meticulously written scripts.
With the aim of compiling an up-to-date, homogeneous and valid comparable corpus, the resemblances between the series chosen were one of the main criteria that guided the selection process. Consequently, it was determined that both series would have a similar profile and bear significant resemblances regarding genre conventions, themes, broadcasting characteristics, and viewership. Finally, the six episodes selected bear some resemblances with regards to the field of discourse, and the year when they were aired in Spain; the episodes selected were broadcast in 2003, and the main plot is built around the pregnancy of Rachel (in Friends) and Carlota (in Siete Vidas).
3.2. An analytical model for the study of prefabricated orality in sitcoms
An analytical framework was developed from the findings and works of several scholars (Biber et al., 1999; Briz, 1996; Briz and Grupo Val.Es.Co, 2000; Chaume, 2001 and 2004a; Vigara, 1980 and 1992). These authors have mainly analysed spontaneous oral discourse. Chaume (2004a), however, has selected those features of spontaneous oral discourse which also appear in dubbese, as well as features typically belonging to written discourse which are also characteristic of dubbese. The model, then, comprises both features belonging to spontaneous oral discourse and features that are typical of dubbese. These are the main characteristics of the model:
4. Findings – prefabricated orality in domestic and dubbed sitcoms in Spanish
4.1.Phonetic and prosodic level
Results point out that, at the phonetic and prosodic level, the domestic audiovisual production bears more resemblance to spontaneous conversation than the dubbed foreign production. Nevertheless, the analysis also showed some resemblances between the prefabricated discourses under study.
The resemblances registered between domestic and foreign prefabricated orality are as follows:
- Clear and correct pronunciation: cacophonies, metathesis and prosodic ambiguity are thus avoided in both corpora;
- Marked and emphatic pronunciation, to facilitate the viewers’ understanding of on-screen conversation in both corpora;
- Marked intonation units in both corpora are used as a means of organising prefabricated speech.
- Phonetic articulation is overall relaxed in the domestic sitcom. Typical phonetic phenomena of spontaneous spoken discourse such as consonant and vowel reduction and elision, as well as assimilation, are frequent in Siete Vidas but completely absent in Friends.
4.2. Morphological level
Regarding the use of carriers of orality at the morphological level, although prefabricated, the Spanish sitcom discourse is closer to spontaneous conversation than that of Friends. The correct, standardised discourse of the dubbed sitcom contrasts with the occasional use of non-standard and seemingly spontaneous morphological features in the domestic production, such as grammatical disagreements. These occasional features, supposedly included by actors while filming, are not at all frequent in Siete Vidas, but help to achieve convincing dialogues.
4.3. Syntactic level
Trends identified in the analysis of the corpora at the syntactic level again suggest that the domestic series is less planned than the dubbed series. However, resemblances between domestic and foreign production are greater than on the above-mentioned levels, and the main divergences seem to lie in the frequency and variety of false spontaneity markers used in both TV series. As expected, these features are overall more diverse and recurrent in the domestic production.
- Predominance of short, simple sentences
- Frequent use of stereotypical structures of conversation
- Sentences are frequently linked without conjunctions; predominance of juxtaposed structures as opposed to subordinated structures, which are more common in writing
- Frequent use of discourse markers
- Predominance of canonical word order, but emphasis through fronting and inversion
- High level of repetitions and additions
- High level of ellipsis and deixis, which are characteristic of spoken discourse
- The elision of conjunctions, considered to be a feature of real spoken discourse, is avoided in Friends but not so in Siete Vidas
- Syntactic dysfluencies (pauses, hesitations, repeats, digressions, expressive paraphrases, fillers, etc.) are more common in the domestic audiovisual production
- The use of the following features is overall more marked and varied in the corpus of native texts: discourse markers, conversational routines, pragmatic word order, colloquial particles used as fillers, deictic units typical in spontaneous conversation, etc.
4.4. Lexical and semantic level
Similarities outweigh differences at the lexical and semantic level, since both native and dubbed audiovisual productions use very similar features to imitate real spoken discourse and to achieve spontaneous-sounding dialogues. Prefabrication allows various colloquial characteristics of spoken texts to be included at the lexical level. However, these features seem to be more standard in the dubbed sitcom, this being the main difference between the two discourses at this level. This trend is reversed in the case of some word-creation procedures, which are more frequent in the foreign TV series.
- Use of generalised vocabulary and colloquial lexis
- Avoidance of unnecessary specialised terminology
- Use of features aimed at increasing expressivity and lexical creativity: metaphors, phraseology, comparisons, word play, intertextuality, clichés and stereotypical structures of colloquial conversation, etc.
- Avoidance of dialectalisms and anachronisms
- Frequent use of argot (especially youth lingo)
- Loans and specialised terminology are more widely used in the domestic sitcom, in comparison to the dubbed sitcom
- Swear words and non-standard vocabulary are more common in the domestic sitcom, in comparison to the dubbed sitcom
- Non-standard features are used more frequently in Siete Vidas than in Friends
- However, some word-creation procedures (the use of prefixes and shortening processes typical of spoken language) are more common in the dubbed subcorpus.
5. Conclusions and suggestions for further research
Findings reveal that the oral features of the Spanish sitcom are ‘pretended’ and prefabricated, and that domestic fictional dialogues are not as natural and spontaneous-sounding as viewers would expect when watching the TV series. The conventions and restrictions of audiovisual media, and the meticulous planning that the scriptwriting of this series is subject to, clearly tip the balance in favour of prefabricated discourse. However, both scriptwriters and actors use several features, either planned or triggered by improvisation, which are targeted at making viewers believe that the conversation they are witnessing is a spontaneous and colloquial exchange between friends. These features aimed at mimicking spontaneous speech include, among others, the use of a relaxed articulation which mirrors authentic spoken Spanish (phonetic level); the introduction of dysfluencies by actors while interpreting the written script (morphological level); the varied mechanisms that disrupt the typical fluency of written discourse (hesitations, fillers, false starts, etc.), as well as the implementation of word order patterns that are characteristic of spontaneous conversation (syntactic level) or the non-standard, non-specific and colloquial language used by most characters (lexical-semantic level). Results have also pointed out that the mirroring of spontaneous spoken speech tends to be higher at some levels of language: similarities with spontaneous spoken speech are outstanding at the lexical-semantic level, marked at the syntactic level, notable at the phonetic and prosodic level, and very limited at the morphological level.
As far as the dubbed discourse of the TV series Friends is concerned, findings similarly reveal that its orality is prefabricated, planned and falsely spontaneous. The aim of mirroring conversational speech is manifested through the use of linguistic features which normally occur in spoken texts, such as the high occurrence of repetition, the frequent use of short, simple syntactic structures at the syntactic level, and the introduction of suffixes, slang, prefabricated fillers and simple vocabulary in fictional dialogues (lexical-semantic level). As suggested by Chaume (2001; 2004a), these ‘carriers of orality’ (Pavesi, 2008: 90) are more prevalent at the lexical-semantic level, which appears to be ‘overloaded’ if compared to other language levels, where features of spoken language are limited (phonetic-prosodic level) or practically non-existent (morphological level). Thus, the foundations of prefabricated dubbed fictional dialogue rest upon the partial and selective imitation of the syntax and lexis of spoken language.
In addition to the conventions and restrictions of audiovisual media, and the consequences of the scriptwriting and translation of Friends’undergoing meticulous planning, other factors lessen the feeling of spontaneity and credibility of dubbed dialogues in Spanish. These factors include the difficulties involved in improvising in the dubbing studio, the space and time constraints restricting the dubbing process, and the tacit target norms governing the dubbing of audiovisual texts which recommend the use of standardised language (Ávila, 1997: 25).
Although findings drawn from the descriptive analysis prove that both the native and the dubbed sitcom feature a prefabricated spoken speech, findings drawn from the comparative analysis reveal that domestic audiovisual texts bear more resemblance to spontaneous conversation than dubbed texts. Dissimilarities often lie in the frequency and variety of the orality markers used in dialogues, which tend to be higher in domestic texts, with only some exceptions (such as the expressive use of creative neologisms, especially the use of prefixes and shortening processes that are typical of spoken language). However, the leeway that scriptwriters and Spanish actors are allowed to mirror spontaneous conversation tends to be greater when compared to that of translators and dubbing actors.
Differences and resemblances between the two discourses studied, as well as the concessions granted to written and spoken discourse, could be represented on a continuum delimited by two poles referring to one of the characteristics of register: the mode of discourse, and more specifically, the degree of spontaneity, as shown in the following figure:
Drawing on the results of the current study, further research will be aimed at the implementation of statistical tools (i.e. corpus management tools) in order to quantify observed recurrences, increase the precision of results and conclusions drawn from the contrastive analysis, and provide a more accurate description of trends identified with regard to the production of prefabricated discourse.
Corpus processing software should also be used to analyse specific features of native and dubbed fictional dialogue and verify the conclusions obtained in the present study, in line with research carried out by scholars in the field of audiovisual translation (see Matamala, 2004; Pavesi, 2008; Romero Fresco, 2008). Taking this research one step further, source texts could be added to constitute an additional sub‑corpus, to be compared against the sub‑corpus of target texts. The purpose of comparing source and target texts would be to identify whether those carriers of orality featured in the target text are also present in the source texts, or whether they are a result of the implementation of domesticating strategies and norms.
As far as the methodological framework is concerned, we believe that the proposed analytical model for the study of prefabricated orality in original and dubbed sitcoms could be adapted and used to study other genres and even other types of audiovisual translation, such as voice-over and subtitling, even though this type of translation involves converting spoken features, originally expressed in a written script, back into writing.
Ávila, A. (1997). El doblaje. Madrid: Cátedra.
Baker, M. (1996). “Corpus-based translation studies: the challenges that lie ahead”, in H. Somers (ed.), Terminology, LSP and Translation: Studies in language engineering in honour of Juan C. Sager. Amsterdam/Philadelphia: John Benjamins, 175-186.
Baños, R. (2006). Estudio descriptivo-contrastivo del español oral en una serie de TV de producción propia y en una serie de TV de producción ajena. El caso de Siete Vidas y Friends. Unpublished Master’s Thesis, Universidad de Granada.
Biber, D., Johansson, S., Leech, G., Conrad, S. y Finegan, E. (1999). Longman Grammar of Spoken and Written English. London: Longman.
Briz, A. (1996). El español coloquial: situación y uso. Madrid: Arco Libros.
Briz, A. y Grupo Val.Es.Co. (2000). ¿Cómo se comenta un texto coloquial?. Barcelona: Ariel.
Chaume, F. (2001). “La pretendida oralidad de los textos audiovisuales”, in R. Agost and F. Chaume (eds.), La traducción en los medios audiovisuales. Castelló: Publicacions de la Universitat Jaume I.
Chaume, F. (2004a). Cine y traducción. Madrid: Cátedra.
Chaume, F. (2004b). “Synchronization in dubbing: A translational approach”, in P. Orero (ed.), Topics in Audiovisual Translation. Amsterdam/Philadelphia: John Benjamins, 35-52.
Comparato, D. (1993). De la creación al guión. Madrid: Instituto Oficial de Radio y Televisión Española.
Eurofiction (2004). Informe anual sobre la actividad televisiva en España, [Online]. Available from
Gómez Capuz (2001) “La interferencia pragmática del inglés sobre el español en doblajes, telecomedias y lenguaje coloquial: una aportación al estudio del cambio lingüístico en curso”. Tonos Digital, Revista Electrónica de Estudios Filológicos, 2, [Online]. Available from
Gregory, M. y Carroll, S. (1978). Language and Situation. Language Varieties and Their Social Contexts. London: Routledge.
Karamitroglou, F. (2000). Towards a Methodology for the Investigation of Norms in Audiovisual Translation. Amsterdam: Rodopi.
Mason, I. (1989). “Speaker meaning and reader meaning: preserving coherence in screen translating”, in Kölmel, Rainer & Jerry Payne (eds.), Babel: the Cultural and Linguistic Barriers Between Nations. Aberdeen: Aberdeen University Press, 13-24.
Matamala, A. (2004). Les interjeccions en un corpus audiovisual. Descripció i representació lexicogràfica. PhD Thesis, Universitat Pompeu Fabra.
Pavesi, M. (2008). “Spoken language in film dubbing: target language norms, interference and translational routines”, in D. Chiaro, C. Heiss & C. Bucaria (eds.) Between Text and Image. Updating Research in Screen Translation. Amsterdam/Philadelphia: John Benjamins, 79-99.
Romero Fresco, P. (2008). “Naturalness in the Spanish Dubbing Language: A case of not-so-close Friends”, Meta, 59:1, pp. 49-72.
Televisió de Catalunya (1997). Criteris Lingüístics sobre Traducció i Doblatge. Barcelona: Edicions 62.
Toledano, G. and Verde, N. (2007). Cómo crear una serie de televisión. Madrid: T&B Editores.
Toury, G. (1995). Descriptive Translation Studies and Beyond. Amsterdam/Philadelphia: John Benjamins.
Vigara, A.M. (1980). Aspectos del español hablado. Madrid: SGEL.
Vigara, A.M. (1992). Morfosintaxis del español coloquial. Esbozo estilístico. Madrid: Gredos.
Whitman, C. (1992). Through the Dubbing Glass. Frankfurt: Peter Lang.
This study was funded by the Spanish Ministerio de Ciencia e Innovación (MICINN), by means of a research project called Estudio empírico y descriptivo de las normas profesionales de la traducción audiovisual en España (HUM2007-65518).