Show Less
Open access

Multilingualism in Film

Edited By Ralf Junkerjürgen and Gala Rebane

Multilingualism is a phenomenon that has become increasingly visible in popular cinema and thus is currently a very novel object of academic inquiry. The present volume is a cutting-edge collection of cross- and transdisciplinary takes on this phenomenon and its different aspects. Its topics range from translation theory to political and aesthetic quandaries of audiovisual translation and subtitling, to narratological function of multilingualism in fiction, to language ideologies and language poetics onscreen. Its authorship is a worldwide body of perspectives, whose contributions span a distinctive collection of international, national and regional film traditions.

Show Summary Details
Open access

“That’s just what we need, a fourth language”. Multilingual humour in film and television translation1

Patrick Zabalbeascoa

“That’s just what we need, a fourth language”. Multilingual humour in film and television translation1

Abstract: With the benefit of insight into multilingualism in fiction within the field of audiovisual translation, and careful analysis of just a few examples, I show that there is a compelling argument to be made for updating traditional approaches to translation. In all actuality, the core model of translation may be flawed or incomplete, as multimodal, audiovisual, multilingual and humorous factors and variables must be incorporated into translation models from the beginning, rather than as an afterthought. Otherwise, we will not be in the best position to make a positive scholarly contribution to the worries and needs of the profession today, nor for the foreseeable future.

Initial considerations

Translation is typically seen as an operation performed on language, usually written, generally well-written (e.g., prestigious prose), in order to transmit the meaning expressed in one language (L1) into the words of another language (L2), thereby involving two languages (L1=>L2 interlingual translation) to reflect the meaning of a single message. In western cultures, this tenet regarding translation is deeply rooted in the tradition of Bible translation (the requirement of a faithful, non-interventionist 100% ideal rendering of a “superior” source text), even when discussing translation issues of other texts, sacred or otherwise (e.g., Nida [1964] 2003). Thus, there is a pervasive influence of the principles of Bible translation in almost all approaches to translation in Europe and its former colonies. In short, these principles involve a view of the importance of being faithful to the words of the source text (as “the Word of God”). The trouble with models of translation based on this kind of thinking is that so much is left ←15 | 16→out, especially as societies and their types of text production and communication evolve. This is precisely what makes audiovisual humour and language variation such interesting, challenging topics (I prefer the term challenge to problem, as I see translating more like searching for solutions to different puzzles than something troublesome or somehow “philosophically” unattainable) simply because these three aspects of humour, audiovisual communication, and multilingualism are not, to put it mildly, typical traits of the Bible. First, then, we need to accept that translators are faced with such issues on a daily basis. Second, we need to come up with an account of translation that can include these challenges and their related issues.

Another “given” in many traditional explanations of how translation works and how translators should work is that there is no room for creativity (and, hence, authorship) or subjectivity while translating. Translators should be objective and faithful, which seems to mean that they should work mechanically. This objective, mechanical quality of ideal translation is the stuff of dreams of computerised, automatic machine translation tools. Two points on this. One is that we might be tempted to say that such dreams are ultimately impossible given the inherent ambiguity of natural human (pragmatic) communication. The other is almost the opposite: we already have such popular internet translation tools as Google Translate that do quite a decent job in many cases. Curiously, these seemingly opposite views about machine translation – one, its “theoretical impossibility”, the other, the reality of its real-life success and efficiency – actually, somehow, lead to the same logical conclusion. If human translators wish to stay in business they will probably find more work precisely where more creativity is required, as the space in translation that humans can claim for themselves, admitting that machines can already translate “Biblically” very efficiently, and more and more so, is the space kept for texts that call for creative, imaginative solutions for their translated versions, and when more creativity is required, the “theoretical impossibility” of machine translation is that much more apparent.

So, as we break free from the narrow view imposed by looking at all translations as if they are governed by the same laws as Bible translation (the Nida-esque principle of: what is good for the Bible and its translations must also be good enough for any other type of translation) we will not ←16 | 17→only find that alternative models can be proposed but even that it is ultimately for the good of human translation as a profession, and for its preservation. For Steiner (1975: 234) the infallibility of machine translation is questionable insomuch that translation is a hermeneutical task, “not a science, but an exact art.”

Table 1. Extremes of ideal writing (ST) and ideal translating (TT) vs. a cline of real-life in-betweens and variation

Ideal TT


0% original, 0% manipulation, 100% equivalent. Only input: ST. Mechanical, objective application of translation “techniques”.

Ideal ST


100% original. Only input: blank page and author’s mind. Creative, authorial writing process.

Real-life variability


A range of degrees of originality, creativity and intertextuality (i.e., input from other texts and other sources of inspiration) for producing both STs and translations.

Objective (nonmanipulative) 100% rendering that requires no interpreting on the translator’s part as a model for Bible translation and general translation theory is paradoxical and unrealistic because it overlooks the wealth of divergent interpretations of the Bible. This model seems to demand that the translator should do the job of providing “the (single) interpretation” of the text (in the words of a second language for those who cannot read the first) for all to be able to grasp the full meaning of the Scripture. This amounts to an expectation that the translator could settle all disputes over how to interpret the Holy Scripture, thus surpassing all Biblical scholars and authorities.

In his “For an Abusive Subtitling”, Mark Nornes (1999), somewhat like George Steiner (1975: 236–296) before him, likes to use metaphors of violence to make a forceful argument. Although “abusive” is probably used so as to sound deliberately misleading, given that the alternative Nornes offers is “corrupt”. Abusive is borrowed from Philip E. Lewis ([1985] 2000) to express that translators must cope with dissimilarities between languages, “proposing strong, forceful translation that values experimentation, tampers with usage, seeks to match the polyvalences or plurivocities or expressive stresses of the original by producing its own”. Corrupt subtitlers respond to the abusive challenge by “conspiring to hide ←17 | 18→their repeated acts of violence through codified rules and a tradition of suppress… feigning completeness in their own violent world”.

Aspects of multilingualism and language variation in fiction

Back in the 1980s, when one of us two started looking into literary bilingualism, scholars working on the topic often felt obliged to justify their interest in such an unconventional domain of study. Bilingual writers and multilingual texts were still very much frowned upon, being freak-like exceptions to the unwritten rule of monolingualism in the literary realm, notwithstanding the (by now well-documented) fact that every century and every genre has seen its share of language-related experiments. (Delabastita & Grutman 2005: 11)

The first and most important aspect of multilingualism in fiction is that it is part of the fiction, in the sense that it is a rhetorical device. And we know that it is a rhetorical device because its alternatives are also rhetorical devices. Thus, authors have several means of reflecting the presence of different languages either in the prose narrative, or in the dialogues, through reported speech or direct speech (e.g., Sternberg’s 1981 widely accepted distinction between homogenizing and vehicular matching). An even earlier strategic decision to be made by any author is whether or not they actually want to reflect language diversity, depending on their choice of topic, its treatment, the point of view and the other ingredients of prose fiction or screenwriting. What follows from this is the realization that language variation, in composition (e.g., novels, film scripts) can take on different forms and perform different functions. Now that there is more research available on this topic, and presumably much more to come, we are in a position to say that there are more variables and possibilities than might be apparent at first glance. For example, typically and traditionally, little prior thought had gone into the issue, so when it came to Bible-style translation norms and criteria, the unwritten rule was literally not to do anything with “other languages” that may appear in a text; i.e., transcribe such passages or pieces of dialogue as if they were a “strange body” within the text, similarly to what some would do with proper nouns and numerals, as if one could translate “around” these textual elements (words denoting numbers or names, or foreign words) and just deal with the “normal” parts of speech.

←18 | 19→

The phenomenon we are interested in looking at is the use of languages other than the text’s main language. I think the focus should lie on language variation in films (Ellender 2015) rather than multilingualism as defining a type of film. Types of films, like polyglot films, multilingual films (Heiss 2004; de Higes 2014; Beseghi 2017), present the theoretical need to define conditions for type membership, such as Díaz-Cintas (2016): “there must be at least one character speaking more than one language”.

Typologies can sometimes be problematic, so rather than offer multilingual films as a new type of film (like Díaz-Cintas 2016), or having to define the borderline between what would constitute an instance of an additional language regardless of any specific film or novel, it might be more productive (theoretically and practically) – and less contentious – to speak of instances of language variation (Corrius & Zabalbeascoa 2011) as a means of referring to a phenomenon characterised by the co-presence, mixing or code-switching of different languages, dialects, sociolects, creoles, made-up languages, diglossia, jargons, slang, and even special cases of speech disorders or temporary speech impediments (conditioned speech, see Zabalbeascoa & Sokoli 2018: 2, 13, 19), such as a film character being gagged, emotionally upset, anxious, or under the influence of medication, drugs or alcohol (Parra, forthcoming). Specifically, what I mean by language variation as a challenge for translation is the presence of relevant features within one or more parts of a text that make certain utterances stand out from the main variety, usually referred to as the language of the text. We might even dare to go so far as to regard all texts as being potentially multilingual – whereby ST=L1+L3 and TT=L2+L3. L1 and L2 representing the main language, and, following the Trafilm project (Zabalbeascoa & Sokoli 2018: 5), L3 is a notation for the purpose of translation analysis to indicate the presence – in the source text and the target text – of any language or any form of language variation other than the main language of each text.

Aspects of humour translation

Just as the case of language variation and intratextual multilingualism, humour is another feature of writing and fiction that is not at all prominent in the Bible, and even less so in Bible translation recommendations. If ←19 | 20→anything, humour of the unwanted variety is strongly ill-advised. Nida warns us not to use any words in our translation that might give rise to unwanted puns or other such humorous interpretations (e.g., for Jesus riding into Jerusalem on Palm Sunday, use the word donkey, not ass). For many scholars, there is no place for humour in the Bible.2 What does that tell us about theories and norms for translation based on the Bible model? Even if we cannot say for sure whether or not there is any humour in the Bible, no one can deny the pervasiveness of humour in all kinds of texts, particularly, but certainly not exclusively, comedy, much of which has to be (and is) translated. There is a gaping hole in translation “prescriptions” that do not really cater to humour other than by coincidence. The fact of the matter is that humour is a feature of human communication that often requires considerable creativity to be translated. The all-too-easy cop out is to claim that humour is by and large impossible to translate, and that successful foreign-language versions do not really constitute translation (according to some narrow conception inspired by Bible translation practices) but adaptation that is not really translation proper.

Furthermore, it also seems to be the case that the most challenging form of humour for the purpose of translation is humour based on wordplay. Overlapping with the issue of wordplay we might add that something that is always tricky to render in translation is the presence of metalinguistic features. This is not surprising given that translation (largely interlinguistic) involves changing the wording from one language to another. Therefore, that part of a text or message which can easily be expressed in several different ways (in one or more languages) because the propositional value or the intention of the message (the function) is much more important (and univocal) than a specific choice of words (the form) is usually easier to translate (especially by Google or Bible scholars, or literal-minded people) mechanically and “objectively” than the metalinguistic references or intense exploitation of language-specific grammatical features (e.g., phonemes, graphemes, polysemy, multiword expressions) ←20 | 21→and their various forms of interaction. A text talking about or stylistically exploiting aspects of a particular language is not such a straightforward matter for translation as a weather report or an explanation of the solar system. So, if humour is not easy to translate (except for happy coincidences) and language-based texts or passages are also not easy, it might be fairly safe to say that humour based on wordplay, aspects of a particular language and linguistic commentary must presumably be very challenging. There are a number of different factors that can make a translation challenging or difficult (e.g., lack of means or motivation) but what I mean here is challenging in the sense of textual and crosslinguistic complications calling upon a translator to be more resourceful, including creativity as well as among other skills. Creativity is a taboo in some areas of translation, but it should not be that way for all areas. On the contrary, it is becoming increasingly urgent for translation studies to either ditch the prescription of translation “techniques” (with a name that implies that they are mechanical and objective) or at least complement them with creative alternatives for more challenging cases.

Metalinguistic humour and humour based on wordplay must include humour derived from intentional comical use of language variation and combinations of different languages (L3, as defined above).

Aspects of audiovisual translation (AVT)

We have pointed out that language variation and humour are the poor relatives of reflection on translation criteria and possible strategies for interlingual transference, due to the historical weight of sacred text translation traditions. So too is AV communication, because sacred texts tend to be written, and the Bible predates the invention of cinematography by many centuries. Of course, the bias towards written texts is not only thanks to the Bible but also promoted by linguistic science, that starts off with studying phonemic, graphemic and lexical units and combinations used for the construction of (monolingual and serious) messages by grammatical rules, while proposing that there are only two modes of linguistic expression: oral and written, leaving out for a much later date the fact that real-life natural (and artistic) human communication relies heavily on body language, paralinguistic items, context (setting, occasion, etc.) ←21 | 22→and social norms of interpersonal interaction and textual composition and publication, not to mention multimodality. Just as translating “other languages” (L3) within a text constituted a non sequitur among translation scholars, and humour was largely ignored, so too any attempt to update mainstream translation theory based on the reality of audiovisual translation has been delayed until quite recently, especially if we consider that film was invented more than a century ago, and translation and multilingualism have been associated to film production and distribution almost since its birth.

The title Chaume (2018) chooses for his review of the state of the art is already revealing in itself: “Is audiovisual translation putting the concept of translation up against the ropes?”. Chaume largely picks up from where Zabalbeascoa (2012) left off in considering how research into audiovisual translation reverberates around the whole area of translation studies and its theorization. Chaume focuses in particular on the need to find the right place in the field for new terms and concepts, such as: localisation, (trans)media adaptation, transadaptation and transcreation. I think this needs to be done within the field of translation studies, and it should not become an excuse to strengthen a caricature or a prejudiced view of an ideal extreme definition of translation (Table 1) that is incapable of recognising anything other than linguistic equivalents of written words as actual translations.

AV text composition is a complex process, undertaken by the film director, its intellectual author, (although the situation is often the result of important contributions from others, like the scriptwriter), and this process involves creating a semiotic construct that combines verbal and non-verbal elements which are conveyed through audio and visual sign systems (words, music, sound effects, photography…). AV translators are rarely allowed to change non-verbal semiotic elements, but at the same time their work is expected to fit in seamlessly with the rest of the fabric of the audiovisual composition. This can become a daunting endeavour in the case of humour translation, given that humour can be created and communicated through any combination of all sorts of semiotic means.

There are various modes of AVT, the most frequent being dubbing and subtitling. Each mode presents limitations and possibilities of its own, with a wide range of solutions (real or hypothetical) available to the translator for rendering humour faithfully or creatively. Some of these include a ←22 | 23→combination of dubbing and subtitling within the same version; this practice is normally considered a strategy within translation for dubbing but never for subtitling. Subtitling is not only used within dubbing but also may be used for humorous purposes by the film director, i.e., as part of the AV source text. Examples of this kind of practice can be found in Monty Python and the Holy Grail, among many other titles listed in Wikipedia (subtitles as a source of humour).

In any discussion on the options available for translation, and AVT in particular, three main factors must be taken into account. First, the technology of the moment; certain things can only be done if the necessary technology with which to do them is available, and this is as true for translation as for anything else. Actually, all forms of translation are dependent upon technology, with the exception of liaison (as represented, albeit farcically, in example 1) and whispered interpreting without note-taking. Writing necessarily involves some form of technology (pen, paper, keyboard), as well as voice recording and transmission, too. Nowadays, computer technology is pervasive in producing, publishing and receiving translations. We need to explore and be aware of how existing and available technology might affect or promote certain types of (creative) translational solutions. However, technology might be expensive, so, the second factor to consider is the financial feasibility of any translation project or particular translation strategy. A traditional notion of financial aid is called patronage (Lefevere 1992); more recent terms would include the need to make a profit. We cannot propose translation solutions that are too expensive for our budget, even if the technology has developed sufficiently. Connected to the financial aspect is also the labour market; some problems pointed out about dubbing, for example, have more to do with a shortage of different voice actors and actresses than any theoretical shortcoming.

The third essential factor to make a translation project work becomes relevant once we have overcome technological and financial problems, and it has to do with social acceptance, including the aesthetics and ethics of the proposed solution. In the case of AVT any solutions proposing “radical” changes (manipulation) of the non-verbal elements may be faced with firm opposition by stakeholders and/or the general public, who may defend a position that audiovisual texts should not be “tampered with” to preserve their integrity. In short, in searching for AVT solutions a translation ←23 | 24→project first has to find the necessary technology, a feasible budget, and short- or medium-term social acceptability.

Once we have sorted out these three essential contextual factors, it is then time to concentrate on the textual make up both of the source text and its translated version(s). Audiovisual composition, as we have pointed out, entails developing a complex web of relationships among all the AV elements, words and otherwise, audio and visual. While admitting that there may be others, these relationships include: complementarity, redundancy, contradiction, incoherence, separability, aesthetic value (Zabalbeascoa 2008: 33–35). Using a different terminology, Yus (2019) applies a similar approach regarding relationships between words and nonverbal items to the study of memes and multimodality.

[…] in search of different categories that these memes might fit into depending on the relationship existing between picture and text and its impact on the quality of the eventual interpretation. An underlying assumption in the chapter, broadly within a cyberpragmatic framework […], is that different text-picture combinations will have an impact on eventual relevance by yielding different balances of cognitive effects and mental effort, the latter sometimes compensated for by an offset of additional cognitive effects in the shape of implications. (Yus 2019: 105).

Critically, and unfortunately, Yus uses the term “text” with the same meaning as “words”.

Presentation of examples

Example 1: Frasier, season 2, episode 21. Scene Seven, “Get out your dictionaries.”

In the living room of Niles’s house, Gunnar – a tall, athletic young German – is getting his fencing gear ready for Maris. Niles Crane comes down the stairs, followed by his brother Frasier, and his Guatemalan maid, Marta.

Niles:   There you are! [Gunnar looks up] Yes, I’m talking to you, strudel boy! No one seduces my wife and gets away with it! You probably thought because of my refined bearing and swimmer’s build that I wouldn’t put up a fight for the woman I love. But you’re dead wrong, because real men have a thing called “honor!” [Gunnar stands up; he’s a full head taller than Niles] Yow! You wouldn’t know about that, would you? You wouldn’t know how decent people behave. You wouldn’t know the meaning of the word “rectitude!”

Frasier:   Niles, he wouldn’t know the meaning of the word “dog,” “cat,” or “pencil!” He doesn’t speak English, remember?!

←24 | 25→

Gunnar:   [to Marta] Wieso ist er so böse?

Marta:   Ich weiß es nicht.

Niles:   Marta! You speak German?

Marta:   ¿Qué?

Frasier:   ¿Uh, habla alemán?

Marta:   ¡Sí! Yo trabajé para una familia alemana que llegó a Guatemala después de la Guerra.

Frasier:   Apparently, she worked for a German family that turned up in Guatemala… [deep voice] just after the war.

Niles:   Well, well. Good, good! She can translate for me! Tell her to tell him–

Frasier:   Niles, Niles, just wait! Look at him! God, if he knew you were calling him “strudel boy,” he’d be wiping his feet on your face!

Niles:   Hang that, Frasier! If there’re going to be scuffs, they’ll be scuffs of honor. [to Gunnar] How dare you steal my wife! [to Frasier] Translate!

Frasier:   Oh, all right. [to Marta] Senor Crane quiere que preguntas a Gunnar, uh, “¡Cómo se atrevez a robar mis zapatos!” (sic.)

Marta:   [to Gunnar] Was fallt dir ein, meine Schuhe zu stehlen?

Gunnar [draws his sword on Niles]: Schweinehund! [Frasier and Marta jump back in alarm.]

Niles:   All right, fine, you want to challenge me? [throws off his jacket and grabs the other sword] En garde!

Frasier:   Oh yes, Niles, that’s just what we need, a fourth language!

Example 1 shows metalinguistic traits (talking about language) as an important ingredient of the scene, which links up a series of interdependent jokes. The TV series’ main ST language, L1, is standard American English, and one of its manifestations of language variety is L3-British English, in the speech of the in-house physiotherapist character, from Manchester, England, who appears in most of the episodes. In this particular scene, L1 is combined with other languages (L3) in the construction of the jokes as well as the development of the plot and further reinforcement of the portrayal of the characters. L3, then, includes German (L3Ge), Spanish (L3Sp) and French (L3Fr), each one an interesting case in its own right. For a successful sitcom such as this one, foreign languages may be potentially funny simply because they are foreign, almost regardless of which language it is and what is being said in L3. From this point of view, we might say that this kind of technique is better used sparingly because the effect will probably soon wear off, so it is used in small doses ←25 | 26→or very skilfully, yet we might also say that the fact that there are so many in example 1 has at least the potential of making it funnier, due to the contrasting sounds that interact and seem funny when thrown together in such a manner. The next step towards a slightly more refined analysis would be to ask how each foreign language can be funny in a different way or for different reasons.

For the purpose of translation analysis, a key variable is the relationship between L1 and L3 in the ST. German, Spanish and French have different relationships with English (and its speakers, originally American in this case), morphological and sociohistorical, and these can be exploited for humour. An important factor is comprehensibility, and a scriptwriter can “plant” L3 words at certain points according to their being readily understandable, both phonologically and semantically. Some American viewers may even know the whole sentence “Ich weiß nicht”, and most will at least be familiar with the Schwein part of Schweinehund. Other instances of the L3Ge and L3Sp utterances will not be understood, and others may be deduced from situation, setting and body language (like shrugging) combined with paralinguistic features (like raising one’s voice and sounding angry). This leads us to another consideration. Who uses each L3, and how does their behaviour and appearance confirm any stereotype or preconception the audience might have about a language and its native speakers? The German speaker is tall and blonde; the native Spanish speaker is short and dark.

Languages are not necessarily spoken by their native speakers, and another source of humour can come from languages being “butchered” by non-natives, and, even worse, by parodies of the language so far removed from grammatical correctness that they must be regarded as pseudolanguages. An instance of this might be Frasier’s “¡Cómo se atrevez a robar mis zapatos!” (i.e., “How dare you steal my shoes!” with a mis-spelling of atreve for atrevez possibly to indicate poor command of the language, just as senor for señor) with the conspicuous “z” and the total disparity of actual and intended meaning. In the film A Fish Called Wanda an uncouth character has a prominent trait of speaking pseudo-Italian because his girlfriend likes to hear foreign languages while having sex… and he does not really know any. In example 1, Marta is a native speaker of Spanish, a fairly proficient speaker of German, but a poor speaker of ←26 | 27→English, if we are to believe that she is not just pretending. However, the real joke is on the two psychiatrist brothers, Frasier and Niles, who are caricatures of pedantry, and one aspect of this is their belief that they can speak foreign languages much better than they really can. The humour related to these aspects is in keeping with a well-defined theme/joke that runs through the whole series: Frasier and Niles frequently find that their displays of pedantry often get them into trouble and they tend to be embarrassed by people who are less assuming but more knowledgeable. In this case, Marta is the only one who can act as a competent interpreter and she leaves the impression (not only in this scene) that she does not do a better job with her English oral skills because it is to her benefit to feign ignorance.

What Frasier calls the “fourth language” in example 1 – the French expression en garde – provides additional proof of the complexity of an issue (L3 sprinkled at certain points of an otherwise monolingual text) that may seem more straightforward at first glance. This short expression shows how easy it is to assign a language to a phrase too quickly. The words may come from French, they are French, but they are also the only way to say en garde in English, and from that angle the expression is a legitimate part of the English language. So, is it L1 or L3 (and how should it be rendered in translation)? The answer to this intriguing question is, like so many things in translation practice and analysis, that it depends; it depends on the context of the words and the text it appears in. This means that translators and translation scholars should not be called upon to do a linguist or a lexicographer’s job of establishing which words and expressions belong to which language or of drawing the lines that divide one dialect from another, or upgrading a dialect to a language, and so on. For translators and translation researchers, what matters is the interpretation of the text as a whole and the parts of the text as they help construct the overall meaning of the text, as a vital part of the interpretation stage of the translating process.

Sometimes a foreign word or expression will be given L1 status as a loan word, but not always. In example 1, en garde is analysed as L3Fr because there is textual evidence: Frasier literally says “not a fourth language” thereby highlighting its L3 status, even though some English speakers will simply think of en garde as the only (L1) way you can say ←27 | 28→that in English. However, we cannot rule out the possibility that part of the joke relies on this paradox (irony) that en garde can be either a foreign language (L3) or an L1-English expression (borrowed from French). Even in this case, and for the purpose of translating, not en garde, but “not a fourth language” the translator will have to present the viewers of the TT (dubbed or subtitled) with something that can plausibly be interpreted as four different languages or otherwise incorporate changes from the ST proposal of combining utterances in four different languages into a smaller number and compensate accordingly for the “not a fourth language” utterance. This particular challenge might be easier if en garde were the usual expression in L2, but this is not the case for Spanish at least, which uses “en guardia”. So, the TT for Spain cannot use the strategy of the ST script, that is, exploiting the ambiguity of whether an expression borrowed from a foreign language can be ascribed to both the main language and the L3.

Valdeón (2005) demonstrates in a case study of Frasier how a particular combination of main language+L3 can be – and is – reversed in the target text (especially when L3ST happens to coincide with L2) to maintain the effect of language variation or multilingualism; his analysis shows that if the priority is to embarrass the Crane brothers for their pedantry by speaking foreign languages, and this is often done in the English-speaking world (for historical reasons) by resorting to French words or sentences, in France the same effect can be achieved to portray pedantry via English dubbing over the utterances that were originally L3 French. So, the ST version combining L1Eng+L3Fr is rendered in the TT as L2Fr+L3Eng, thereby avoiding a conflation of languages, given that the whole point of the joke is to combine and display two languages.

One key factor that would affect this instance of “four languages” is the scenario where the ST is translated for Spain and Germany, and even France, and/or other countries that use one of the languages of example 1, Spanish, German or French for their dubbing or subtitling. This poses the following problem: if, for the case of Spanish dubbing, L1-English is translated into L2-Spanish, in all instances, and the various L3 instances are all kept the same, then we only have three languages because L2-Spanish will replace both L1-English and L3Sp, and likewise for ST L3Ge for TT-German, and ST L3Fr for TT-French. The dubbing practice in Germany is more systematic than in Spain when dealing with L3 and every effort is ←28 | 29→made to keep L3 unaltered in the TT, whereas Spanish dubbing practices show room for some creative solutions, although the general tendency is the same. Example 1 is one of the rare examples where the German dubbed version indeed changed ST-L3Ge to a different language in the TT despite keeping ST-L3Ge as German in translation in almost all other cases. The German dubbed version for example 1 has created a fencing instructor who is Danish instead of German, enabling “four languages” to be kept: L1-English rendered as L2-German, L3Sp and L3Fr remain the same, and ST-L3Ge is substituted for TT-L3-Danish. In Spanish dubbing, L3 status can be kept by the same means, if the maid’s native language were changed to Portuguese, for example, making her Brazilian in the TT. Another option would be to make her variety of Spanish stand out as quite different from standard Iberian Spanish, in an attempt to make ST-L3Sp be perceived in the TT as L3-Guatemalan Spanish dialect while L1-English is rendered as L2-Iberian Spanish. This would be forcing the idea that there are “four languages” unless she mostly used dialect words that are not known by most Spanish citizens. It is striking that in Spain Spanish utterances reflecting varieties of Spanish from other parts of the world are often redubbed with Iberian Spanish accents, resulting in a mismatch of Iberian Spanish accent spoken by characters supposedly coming from Argentina, Mexico, and so on. This kind of mismatch is never funny nor is it meant to be (even when dealing with comedy). On the contrary, Latin American varieties of Spanish are avoided across the board for dubbing (Menéndez-Otero 2013) in Spain because of the fear among distributors that they will sound ridiculous, or off-putting, as has historically been recorded when such attempts were made.

Example 2. The Hudsucker Proxy (script corresponding to minutes 66’47” to 67’11”).

The hula-hoop has suddenly been discovered by the American public and become hugely popular overnight in the USA. This scene is part of a newsreel report about the new craze.

NEWSREEL 72:   A scientist with a Van Dyke beard, wearing a laboratory smock, is facing the camera. Behind him we see other scientists studying a hoop that has been hooked up to a gyroscopic-looking device that analyzes its various movements and properties.

NEWSREEL ANNOUNCER:   What scientific principle explains the mind-bending ←29 | 30→motion of this whipping wheel of wonder? A title supered over the Scientist’s chest identifies him as Professor Erwin Schweide.

SCIENTIST:   Ze dinkus is kvite zimple, really. It operates on ze same principle zat keeps ze earth spinning ‘round ze sun, and zat keeps you from flying off ze earth into ze coldest reaches of outer space vere you vood die like a miserable shvine! Yes, ze principle is ze same, except for ze piece of grrrit zey put in to make ze whole experience more pleasant (the spelling here is copied from the script).

(Spanish dubbed version 1:06:47 - 1:07:10. SCIENTIST: En realidad es un chisme muy sencillo. Se basa en los mismos principios que mantienen a la tierra girando alrededor del sol y que les impiden a Uds. salir volando de la Tierra a los fríos confines del espacio donde morirían de forma miserable. Sí es el mismo principio… excepto por el pedazo de tierra que le han metido dentro para que la experiencia resulte más agradable.)

Although Hollywood, among others, tends to use L3Ge mostly as part of the Nazi stereotype (see, for example, Bleichenbacher 2008: 47), there are occasions when the stereotype is a different one. In example 1, we saw a German fencing instructor, who was not overtly playing on the Nazi stereotype. Christie Davies (2006), in his response to Dirk Delabastita’s (2005) otherwise brilliant study of Shakespeare’s exploitation of multilingualism in drama (so relevant, too, to the Henry V film adaptation), offers an interesting word of caution about rushing into racist or colonialist explanations for all comic use of foreign languages. Example 2 shows the stereotype of the (mad) scientist. While there does seem to be a stereotype of Germany as producing much good science and engineering, it is less clear whether the German scientist sounds a bit sinister because he is a scientist or, maybe, because (as in the Frasier reference to post-war Nazi refugees) he might be a Nazi, after all; or a former refugee escaping from Nazi Germany. The funniness of the scene relies more on the sinister overtones of the scientist’s utterances than his foreignness, even though there is room for a possible sinister stereotype of (German or all foreign?) scientists. The joke relies on the incongruity of explaining something that is pleasant and fun in terms (and tones) that are quite the opposite. In the Spanish dubbed version, unfortunately, there is no trace of foreignness, nor the sinister tone and tempo having applied standardised (dubbing) speed and intonation, as well as normative pronunciation. All that is left, then, are a few words “morirían de forma miserable” [you would die despicably] that sound more out-of-place with the rest of the ←30 | 31→discourse than sinister. In example 2, L3 is not German proper, but non-native English, sprinkled with the odd German word. The Spanish version not only leaves out the few German words but also misses the opportunity to create a caricature of (off-putting) scientific discourse with such key words as grit. What this means is that it was not even totally necessary to reflect so clearly the German origin of the scientist, especially if Spain does not share the German-scientist stereotype, for a successful or efficient dubbed version of the humour intended in the ST. The translation strategy would focus on reflecting (somehow, with or without L3) the incongruity of a stereotypical scientific discourse that is completely stripped of empathy, yet used to explain something that is meant to be a toy and fun for all. The Spanish translation for the dubbed version shows how this was not the case because it does not attempt to achieve an unpleasant discourse through L3 (the accent mostly, plus Schwein), nor through the other ST rhetorical devices with carefully chosen words like “grit”, and the weird AV-ST pauses in midsentence that sound more threatening than instructive.


With the benefit of a growing number of studies on the issue of multilingualism in audiovisual fiction within the field of translation studies, and careful analysis of just a few examples, I hope to have shown that there is a compelling argument for revisiting and updating traditional thinking and models applied to translation. I believe that the case presented here does not only affect the small area of translating language combinations that appear in audiovisual comedy, but any general claims to be made about translation as well. It is no longer an attempt to convince scholars and professionals that we need a more refined theory and set of guidelines, but a realisation that the cornerstone, the core model of translation, may be flawed or incomplete. In other words, multimodal, audiovisual, multilingual and humorous factors and variables must be incorporated into translation models from the very beginning, not as an afterthought or a concession to a late trend. Otherwise we will not be in the best possible position to make a positive scholarly contribution to the worries and needs of the profession today, nor for the foreseeable future.

←31 | 32→


A Fish Called Wanda (1988). Dir. Charles Crichton. Metro-Goldwyn-Mayer.

Frasier “An Affair to Forget” (1994). Dir. Philip Charles MacKenzie. Grub Street Productions.

Henry V (1989). Dir. Kenneth Branagh. BBC Films.

Monty Python and the Holy Grail (1975). Dir. Terry Gillian and Terry Jones. Python (Monty) Pictures.

The Hudsucker Proxy (1994). Dir. Joel Coen. Silver Pictures.


Beseghi, Micól (2017): Multilingual Films in Translation. A Sociolinguistic and Intercultural Study of Diasporic Films. Oxford: Peter Lang.

Bleichenbacher, Lukas (2008): Multilingualism in the Movies. Hollywood Characters and Their Language Choices.Tübingen: Narr Francke Attempto Verlag.

Chaume, Frederic (2018): Is audiovisual translation putting the concept of translation up against the ropes? In: Jostrans: The Journal of Specialised Translation (30), 84–104.

Corrius, Montse & Zabalbeascoa, Patrick (2011): Language Variation in Source Texts and their Translations: The case of L3 in Film Translation. In: Target 23 (1), 113–130.

Davis, Christie (2006): Response. Comic Welsh English in Shakespeare: A reply to Dirk Delabastita. In: Humor: International Journal of Humor Research 19 (2), 189–200.

De Higes, Irene (2014): Estudio descriptivo y comparativo de la traducción de filmes plurilingües: el caso del cine británico de migración y diáspora. Castellón: Universitat Jaume I. PhD Dissertation.

Delabastita, Dirk (2005): Cross language comedy in Shakespeare. In: Humor: International Journal of Humor Research 12 (1), 1–32.

Delabastita, Dirk & Grutman, Rainer (2005): Fictional representations of multilingualism and translation. In: Linguistica Antverpiensia 4, 11–34.

Díaz-Cintas, Jorge (2016): Multilingüismo, traducción audiovisual y estereotipos: el caso de Vicky Cristina Barcelona. In: Prosopopeya 2014-2015 (9), 135–161.

←32 | 33→

Ellender, Claire (2015): Dealing with Difference in Audiovisual Translation: Subtitling Linguistic Variation in Films. Oxford: Peter Lang.

Heiss, Christine (2004): Dubbing multilingual films: a new challenge. In: Meta 49 (1), 208–220.

Lefevere, Andre (1992): Translating, Rewriting, and the Manipulation of Literary Fame. London & New York: Routledge.

Lewis, Philip E. ([1985] 2000): The Measure of Translation Effects. In: The Translation Studies Reader, ed. Lawrence Venuti, London: Routledge.

Menéndez-Otero, (2013): Linguistic pluralism and dubbing in Spain. Mise au Point 5. European cinema and languages.

Nida, Eugene ([1964] 2003): Toward a Science of Translating: With Special Reference to Principles and Procedures Involved in Bible Translating (2nd Edition). Leiden, Holland: Brill.

Nornes, Abé Mark (1999): For an Abusive Subtitling. In: Film Quarterly 52 (3). 17–34.

Parra López, Guillermo (forthcoming): “A map of drug-induced language variation in cinema and its translation.”, InTRAlinea: Online Translation Journal, Special Issue: The Translation of Dialects in Multimedia IV.

Steiner, George (1975): After Babel. Oxford: Oxford University Press.

Sternberg, Meir (1981): Polylingualism as Reality and Translation as Mimesis. In: Poetics Today 2 (4), 221–239.

The Trafilm Project.

Valdeón, Roberto (2005): Asymmetric representations of languages in contact: uses and translations of French and Spanish in Frasier. In: Linguistica Antverpiensia 4, 279–294.

Wikipedia, Subtitles as a source of humour. Online under: https://en.

Yus, Francisco (2019): “Multimodality in memes. A cyberpragmatic approach.” In: Analyzing Digital Discourse: New Insights and Future Directions, eds. Patricia Bou-Franch & Pilar Garcés-Conejos. Blitvich: Palgrave, 105–131.

Zabalbeascoa, Patrick (2008): The Nature of the Audiovisual Text and its Parameters. In: The Didactics of Audiovisual Translation, ed. Jorge ←33 | 34→Díaz-Cintas. Amsterdam: John Benjamins Translation Library 77, 21–37.

Zabalbeascoa, Patrick (2012): Teorías de la traducción audiovisual. Un viaje de ida y vuelta para progresar. In: Reflexiones sobre la traducción audiovisual. Tres espectros, tres momentos, ed. Juan Martínez. Valencia. Publicaciones de la Universitat de València, 187–199.

Zabalbeascoa, Patrick & Sokoli, Stavroula (2018): The Trafilm Guide: how the trafilm database works and how to feed in more metadata into the system, with basic trafilm concepts Universitat Pompeu Fabra. CC-Attribution 3.0 Spain. Online under:

←34 | 35→

1 This chapter was written as part of the Trafilm Project, (FFI2014-55952-P) funded by the Spanish Ministry of Economy and Competitiveness. This publication reflects the views only of its author.

2 There are also scholarly attempts to claim the presence of humour in certain passages, but they seem more the result of first accepting humour as a positive value for any text and then setting out to find it in the Bible, which surely must encompass every literary virtue.