Reaching a Foreign Audience: Cultural Transfers in Audiovisual Translation

This paper examines some of the issues involved in the intercultural transfer of films. It focuses on the translation of culture-specific references and questions in particular the validity of the notions of foreignisation and domestication, brought to the fore of Translation Studies by Venuti (1995), as a conceptual framework traditionally used to discuss the strategies applied when translating cultural specifics. Drawing on the findings of a pilot study consisting of three French films dubbed and subtitled into English, this paper suggests a theoretical challenge by proposing a more pragmatic approach to the study of cultural transfer in audiovisual translation (AVT). More particularly, it will examine whether it is possible to observe any form of consistency in the strategies used for the translation of culturally-bound references and what this implies for the dialogic relationship between Self and Other, and the representation of alterity.


Introduction
One of the greatest challenges for a movie after its domestic release is reaching an international audience and being successful abroad.In this process of internationalisation, linguistic difference is one of the major obstacles, and translation for the cinema (mainly in the form of subtitling and dubbing)1 has thus taken on major economic and social importance.However, language and culture are deeply intertwined, and translators obviously do not translate individual words deprived of context, but whole texts which are culturally embedded and based on a community of references predictably shared by most members of the source culture -thus creating 'moments of resistance' for translation.Since it brings cultures into contact with one another, translation for the cinema in particular, and the audiovisual world in general, raises considerable cross-cultural issues.Disregarding them may lead to a translated programme which is unintelligible for the target viewers.The issues involved in the cultural transfer of films are manifold, ranging from the very choice of movies to be distributed abroad to the marketing strategies employed and the techniques used to translate culture-specific material.The translation of cultural specifics in particular constitutes one of the most challenging areas of intercultural transfer, to the extent that cultural references are traditionally regarded in the literature as being 'untranslatable' (Catford, 1965;Cornu, 1983;Arson, 1988), therefore touching on the very limits of translation.Particularly interesting is the issue of the impact that translation strategies may have on audiences' perceptions of the source culture.When starting to investigate these issues, I was naturally drawn to Venuti's notions of 'foreignisation ' and 'domestication',2 and his claim (1998a: 67) that '[t]ranslation wields enormous power in constructing representations of foreign cultures'.This led me to conjecture that translation for the cinema, because of its tremendous social impact and visibility as a mode of intercultural exchange, may in fact affect cultural representations to a greater extent than other types of translation -both in the way a national cinema is perceived abroad and, more importantly perhaps, in how cultures perceive each other and themselves.The foreignisation/domestication model has been acclaimed as a powerful tool to conceptualise the interface between the source culture (SC) -seen as the 'Self' -and the target culture (TC) -seen as the 'Other' -but has also sparked wide debate in the field.According to Venuti (1998b: 240), foreignisation and domestication as overall translation strategies take place at two levels: the macro-level -with the selection of foreign texts to be translated -and the micro-level, i.e. the actual methods used to translate them.For Venuti (1992; 1995a; 1995b; 1998b), domestication is a natural tendency of translation and consists in translating in a fluent, idiomatic, and transparent way which tends to erase the foreignness of the source text and to conform to the needs and values of the domestic/target culture.In his own words: [A] fluent strategy performs a labor of acculturation which domesticates the foreign text, making it intelligible and even familiar to the target-language reader, providing him or her with the narcissistic experience of recognizing his or her own culture in a cultural other, enacting an imperialism that extends the dominion of transparency with other ideological discourses over a different culture.(Venuti, 1992: 5) Foreignisation, on the other hand, takes the target reader towards the source text with a defamiliarising effect, and consists in "preserving linguistic and cultural differences by deviating from prevailing domestic values" (Venuti, 1998b: 240).Following Schleiermacher (1813) and Berman (1985), Venuti claims that the foreignising method is "highly desirable [as a way] to restrain the ethnocentric violence of translation" (1995b: 20) and "to mak[e] the translated text a site where a cultural other is not erased but manifested" (1998b: 242).Different terms have been used by Venuti and his followers, such as 'naturalisation' or 'assimilation' (used in place of 'domestication'), and 'exoticism' or 'exoticisation' (as near synonyms of 'foreignisation').In fact, as Kwiecinski (2001: 13) notes, these terms "tend to be used rather loosely and to refer to different phenomena potentially leading to terminological gaps and inconsistencies".More particularly, Kwiecinski (2001: 15) distinguishes between foreignisation/domestication and exoticism/assimilation, and stresses that the foreign and the exotic have often become merged.Interestingly, Zlateva (2004: 2) also notes that, in reality, domestication and foreignisation "refer to two different entities" and cannot be compared as symmetrically as they are: "what is domesticated is the form and the content of the source text [�c and] what is foreignised and exoticised [�c] is the form and content of the translated text" (ibid.).3Although originally meant by Venuti as a political act aimed at drawing attention to the translators' invisibility, the foreignisation/domestication model has been reappropriated by many scholars working in the area of intercultural transfer.As will be demonstrated in the paper, the strategies used for the translation of culture-specific material in particular are typically classified according to this model.4It is not my intention to question here the quality of Venuti's arguments as such, i.e. his advocacy of foreignising translation projects, but rather the notions of foreignisation and domestication as a conceptual framework traditionally used to discuss cultural transfer in translation.In this paper, I propose to report the findings of a pilot study consisting of three French films dubbed and subtitled into English, and designed to test the validity of the foreignisation/domestication model in empirical situations.More particularly, I will examine whether it is possible to observe any form of consistency in the strategies used for the translation of culturally-bound references and what the answer to this question implies for the dialogic relationship between Self and Other, and the representation of alterity.My intention, therefore, is not to question the foreignisation/domestication framework and some of Venuti's arguments in the same way that Robinson (1998), Pym (1996) and others have done -i.e. by arguing that foreignising strategies may in fact have potentially negative effects5 -but to present a theoretical challenge by exploring whether this widely used model can adequately inform the cultural issues involved in my corpus and, by possible extension, in the specific context of audiovisual translation (AVT).It is beyond the scope of this paper to examine the applicability of the foreignisation/domestication model to large-scale issues such as the selection of films to be distributed abroad or the role of marketing and paratexts in familiarising audiences with potentially unfamiliar content, although these are obviously important to better understanding of how a film reaches a foreign audience and affects cultural representations.In this paper, I investigate the suitability of the notions of foreignisation and domestication to lowerleveltranslation issues -in this particular case, the strategies used by dubbers and subtitlers for dealing in English with French culture-specific references.I will first outline the way that translation procedures for culturally specific references are traditionally regarded in the literature.

Definition
The theoretical and practical issues raised by the notion of cultural specificity are particularly complex.Franco Aixelá (1996: 56-57), for instance, rightly points out that: The first problem we face in the study of the cultural aspects of translation is how to devise a suitable tool for our analysis, a notion of 'culture-specific item' (CSI) that will enable us to define the strictly cultural component as opposed to, say, the linguistic or pragmatic ones.The main difficulty with the definition lies, of course, in the fact that in a language everything is culturally produced, beginning with language itself.For the purposes of this paper, I am adapting Olk's definition (2001) of culture-specific material, given in the context of textualtranslation,6 to the particular context of film.I understand culture-specific material to encompass the verbal and non-verbal (visual and auditory) signs whichconstitute a problem for cross-cultural transfer because they refer to objects or concepts that are specific to the original sociocultural context of the film -i.e.that, at the time of distribution, do not exist, or deviate significantly in their connotational value from similar objects and concepts in the target culture(s) considered.

Classifications of procedures
With the notable exception of Tomaszkiewicz ( 2001) and Nedergaard-Larsen (1993), who have examined in a systematic way the strategies used for the translation of culture-specific references in the context of film, most of the research work on culture-specific material has been carried out in the context of textual translation (Ivir, 1987;Florin, 1993;Newmark, 1995;Franco Aixelá, 1996;Mailhac, 1996;Kwiecinski, 2001;Olk, 2001).Although not the focus of this paper, it is of particular interest to explore to what extent the classifications of translation procedures suggested by these authors are applicable to the context of AVT.My main argument here, however, is that most classifications of translation strategies follow a common general progression from the most exoticising to the most domesticating, even if this progression is sometimes only implicit yet obvious from the order in which translation procedures are presented as is the case in Ivir (1987) andNewmark (1995).Although not all possible translation strategies can be presented in Figure 1 since scholars themselves do not agree on the number of procedures available to translators, or on how to label them, the typical progression can be visualised as follows: According to this model, translation procedures are distributed along a scale with two poles, usually termed foreignisation and domestication, but also referred to as exoticism/assimilation, source/target, foreign/familiar, Self/Other, etc.Each translation procedure is situated on this spectrum according to the extent to which it accommodates the target reader/viewer's own cultural background.This basic model is described, for instance, by Franco Aixelá (quoted in Kwiecinski, 2001: 151), as "a continuum of various degrees of intercultural manipulation, defined as a scale of conservation vs. substitution strategies".My point is that, despite the variations and terminological overlaps in the way these terms are understood and used, the model charted out to present the range of translation strategies available is essentially the same: the two extremes of the spectrum represent tendencies or general strategies in relation to which each translation procedure (explanation, calque, cultural substitution, omission, etc.) is situated according to its degree of cultural mediation.The model is therefore clearly based on a polarisation with each translation procedure tending towards one pole or the other, thus presenting Self and Other as mutually exclusive.

Cultural impact of translation strategies
The argument put forward by most scholars working in this area is that translators have at their disposal a whole range of strategies to translate cultural specifics, and that the type of strategy used will impact on TC -perceptions of the SC -preserving 'local colour', perpetuating (positive or negative) stereotypes, undermining or highlighting cultural specificities, possibly even creating cross-cultural misunderstanding -and therefore on the way a text or film is perceived in the foreign market.Venuti (1995a: 23) clearly argues, for instance, that translation is inevitably domesticating since it is usually made to conform to the needs and values of the domestic culture and, therefore, famously advocates foreignising strategies because they retain the foreignness of the original and encourage readers/viewers of translations to become more open to cultural differences.Similarly, Olk (2001: 54), in the particular case of culture-bound references, argues that translation is a discursive practice and that translation strategies can "influence the way the source culture is perceived in the target culture" (ibid.: 56), a standpoint also shared by authors such as Jacquemond (1992), Herrero (2000), andWitte (1994).Yet the assumption behind these claims -an assumption rarely made explicit by any of the aforementioned authors -is that the use of translation strategies is always consistent, therefore creating a norm with the potential impacts alluded to above.For instance, Franco Aixelá (1996: 60) states that: we will obtain a frame which will allow us to discover quickly the general tendency of a translation as regards the double tension discussed at the beginning of this paper (being a representation of a source text and being a valid text in itself).This enticing argument served as a basis for my research project and initially led me to identify the various specific strategies used in subtitling and dubbing when translating culture-specific material, as well as their frequency and impact on cultural representations of the SC.I will now present the results of a preliminary study designed to verify whether this assumption and, in particular, the foreignisation/domestication model and its impact on cultural representations, are supported empirically.

Problems arising from the empirical study
It is important to stress that my empirical study was only carried out on a small corpus of three films, both subtitled and dubbed into English: Le dîner de cons (1998), Astérix et Obélix contre César (1999), and Le pacte des loups (2001).However, since each film has been translated by different subtitlers and dubbers, it is expected that the problems outlined below would be similar with a larger corpus.When trying to apply the typical classifications of cultural reference procedures described above, three types of empirical problems arose.

Labelling of strategy
It was first of all difficult, in some cases, to precisely identify the translation procedure used.In Le dîner de cons, for example, when Pierre Brochant hurts his back, François Pignon ('le con', who accidentally caused him to fall) exclaims: 'Là, à part un voyage à Lourdes...' [Now, except a trip to Lourdes�c], which has obviously a most ironic effect.The subtitles read: 'You'll need a miracle!'.7 Whether this is a form of explanation (i.e.people going to Lourdes are actually seeking a miracle) or neutralisation (the cultural reference to Lourdes is omitted; only the idea that Brochant is now in a hopeless situation is preserved) is difficult to determine.Besides, translation procedures are more often than not combined together, also making it difficult empirically to identify with precision the strategies used.Interestingly, this problem of labelling may in fact account for the plethora of labels and classifications found in the literature, since most researchers seem to propose their own classifications of procedures after highlighting inconsistencies, overlaps in labels, or the lack of clarity in previous typologies, only to fall back again into the same problems when charting out their own classification.8

Foreignising or domesticating procedure?
Notwithstanding this problem of labelling, I still tried to apply the classifications described above to my corpus.The second problem arising from the empirical study was that, once the procedures had been more or less identified, it was difficult at times to situate them on the foreignisation à domestication spectrum.For instance, it is not clear whether neutralisation and omission are forms of domestication or foreignisation.They tend to be culturally neutral and, therefore, to erase somewhat the specificity of the SC.The same holds true of explanation, which does try to accommodate the needs of the target viewers by "reducing the unknown to the known and the unshared to the shared" (Ivir, 1987: 38), but still retains some of the foreignness of the original, since it is often combined with another procedure.In Le pacte des loups, for instance, 'le pays du Gévaudan' is dubbed as 'the Gévaudan region',9 which is a combination of literal translation and explanation.Situating the explanation strategy on the foreignisation à domestication scale is therefore difficult.It follows that most procedures are not assimilating or exoticising in and of themselves but that these potential characteristics -if relevant at all -can only be determined in context (see 4 below).Classifications such as the ones described in 2.2 above appear, therefore, to be of limited use, especially in the context of screen translation.

Lack of consistency in the use of strategies
Finally, despite the problems emphasised above (i.e.decontextualisation and difficulty of labelling the strategies), a quantitative study of the procedures used in the three films, both in the subtitled and the dubbed version, was carried out in order to further test the usual operational models based on the notions of foreignisation and domestication.
Notably, it appears that a relatively large number of literal translations and transferences is used in all three films.However, proper names (of places or people) amounted to more than half of the occurrences:10 because names are mono-referential by nature, they usually can only be borrowed into another language, i.e. transferred without change or with only minor spelling amendments.The main conclusion of this study, however, is that thewhole repertoire of translation procedures is used in the three films, making it therefore impossible to demonstrate any form of consistency as far as foreignising or domesticating norms are concerned.This finding seems to be contrary to what Venuti and other scholars have claimed happens in the context of literature.Consequently, and as research in the area has already demonstrated (see note 5), it is particularly difficult to draw any conclusion about the type of impact that the translation strategies used in the subtitled and dubbed versions may have on representations of the SC.

Problem with taxonomies
In this empirical study, the foreignisation/domestication model does not appear to convincingly reflect the pragmatic realities of AVT.The findings presented above seriously question the suitability of rigid models and classifications based on a belief in the consistency of the procedures used for the translation of culture-specific material, and on the underlying assumptions made about translators' or distributors' agendas (see 2.3).In reality, it seems that audiovisual translators may select translation procedures on an ad hoc basis.The problem with the classifications described earlier actually arises from the very drive to classify.Taxonomies, by nature, decontextualise.In this particular case, they present cultural references and their translation in relative autonomy from their context and their function in the text.Kwiecinski (2001: 10), while still providing his own classification of translation procedures, points indeed to its limitations: "the foreignness/domesticity of a transeme is highly cotext-and context-sensitive and thus cannot be adequately captured solely by means of a formal taxonomy of procedures".My belief is that contextual factors play a much more important role than is typically believed, especially perhaps in screen translation since the nature of the medium is characterised by particular technical constraints, and implies strong visual and contextual embeddedness.Ivir (1987: 37) is among the few to note that: in planning his translation strategy, the translator does not make a one-time decision on how he will treat unmatched elements of culture; rather, even if he has established an overall order of preferences, he usually makes a new decision for each such element and for its each use [sic] in an act of communication.

Importance of context
It is, therefore, necessary to adopt a more pragmatic approach to research in AVT and to highlight the crucial importance of context in the selection of translation strategies.Context in AVT must be understood very broadly and should include: • Linguistic co-text.
• Function and relevance of the culture-specific reference in the larger context of the film/diegesis, taking into account Ivir's notions (1987: 42) of culture-in-focus vs. culture-as-background.
• Technical constraints at that particular moment in the film (e.g.close-up shot causing difficulty in the case of dubbing, long and fast-paced lines of dialogue for subtitling, etc.).• Genre of the film.
• Target audience of that particular film.
• Distribution context (e.g.paratexts accompanying the release of the film, which may influence the larger cultural environment in which the movie is received, etc.).• General cultural context (to what extent are cultural references shared between SC and TC?What relationships exist between SC and TC?), and so on.
As Niemeier (1991: 151) rightly points out, "at any moment [the translator] has to consider the context of the whole film in order to reach a successful transcultural translation".
Determining with precision what context in AVT encompasses and how it may affect the choice of translation strategies is yet to receive greater academic consideration.

Findings from interviews with subtitlers
This theoretical challenge grounded in empirical analyses also seems to be confirmed by interviews carried out in February 2004 with subtitlers at the Special Broadcasting Service (SBS), an Australian multicultural channel which broadcasts movies and programmes in approximately sixty different languages and runs its own subtitling unit.Subtitlers claimed to systematically select strategies on a case-by-case basis, and not to have any form of ideological, aesthetic or didactic agenda.When asked whether they were following particular guidelines for the translation of culturally specific references, one of the subtitlers replied: "No.We don't have any particular guideline, so [�c] we do it on a case-by-case basis".Another subtitler stressed: 'Ça dépend vraiment du film et de la scène.Il faut établir des priorités' [It really depends on the film and the scene.It is necessary to set priorities].This seems to suggest that audiovisual translators try to select the best strategy for eachindividualtranslation problem and do not necessarily have a pre-established general strategy, at least not for the translation of cultural specifics.In view of the immediacy of film and the technical constraints of audiovisual language transfer, this may imply giving priority to communicative translation over cultural considerations.That is, translators may aim for a translation which is immediately accessible by the target viewers, therefore omitting or neutralising, if necessary, cultural references which would otherwise be difficult for the target audience to understand.

Conclusion
As stated earlier in this paper, my initial hypotheses were based on the widely used foreignisation/domestication model proposed by Venuti and reappropriated by other scholars working on the translation of cultural specifics.After undertaking a pilot empirical study to test those assumptions, it soon appeared that the model and, more particularly, the taxonomies of translation strategies based on that model cannot, in fact, adequately capture the pragmatic complexity of cultural transfers in film, as confirmed by reports from practitioners.
Ultimately, the concept of norms itself may need to be questioned.
This conclusion may admittedly only apply to screen translation because of the many constraints of that particular mode of translation and the various contextual factors involved in the choice of strategies, which may well take precedence over possible ideological or didactic agendas.Although these issues need to be confirmed by a more extensive study of the entire corpus, my objective here is to open the debate and suggest a different -more pragmatic -approach to research in this aspect of translation studies.This does not imply, of course, that the notions of foreignisation/exoticism and domestication/assimilation are to be totally rejected, or that translation is not a discursive practice.These stimulating ideas have indeed made a worthy contribution to the field by drawing attention to translators' ethical responsibilities and the potential implications of translation decisions.However, instead of regarding Self and Other, Familiar and Foreign, as fixed notions, it is more fruitful to see them as mutually defining since, according to the existentialists, the Other is always more or less constructed from the Self and defined in relation to the Self.As Cryle (2000: 40) puts it, "L'autre, loin d'être un simple référent, est toujours l'autre de quelqu'un" [The other, far from being a mere referent, is always someone's other].Seen in this light, foreignisation and domestication occur in realityat the sametime, and those seemingly fixed identities are therefore negotiated identities, as aptly reflected by Pym's description (1997: 14) of the translator as a mediator of interculturalité: "l'espace du traduire -le travail du traducteur -se situe dans les intersections qui se tissent entre les cultures et non dans le sein d'une culture unique" [The space of translation -the translator's work -is located in the intersections established between cultures, not within a single culture].Instead of being subjected to either a movement from the viewer to the film (foreignising strategy) or from the film to the viewer (naturalising strategy), the translator is allowed to stand on the frontier, in a middle-ground position.
Viewers themselves are also in an intercultural position.As observed by Turner in Film as Social Practice (1993: 79), "[t]here is a high degree of cross-cultural coding where audiences agree to accept an imported system of meaning for the purposes of enjoying the film".Translators and distributors are, therefore, not the only ones responsible for the cultural transfer of films; viewers as well play a role.Indeed, audiences' intercultural skills and readiness to accept the Foreign might too often be underestimated, and the fact of sitting in a dark room to watch a foreign movie may in fact already bear testimony to viewers' openness to accept this negotiation.