The Routledge Handbook of Translation and Pragmatics

Critical pragmatic insights into (mis)translation in the news

Jan Chovanec

Introduction

The practice of translation in news journalism and in the production of media texts is a specific process that differs from the role translation plays in many other domains. Translation of fiction, as well as non-fiction, has traditionally been centred around the notion of equivalence between the source and target texts with a view to how specific forms, meanings and effects of the former can be appropriately rendered in the latter. News-related translation, however, is characterised by various kinds of textual transformations since the journalist-translator usually uses the original text to construct a new text. Various forms of translational non-equivalence are hence the norm. This has led some researchers to propose various other terms in order to point out the specificity of this process. In an early contribution to the field, Stetting (1989), for instance, proposed the broader term “transediting” as an alternative concept to refer to the modifications that news texts frequently undergo as a result of being translated and, thus, recontextualised into other linguistic environments. According to Stetting (1989: 377), this concept has three dimensions: cleaning-up transediting (i.e., adaptation to the standard of efficiency in expression), situational transediting (i.e., adaptation to the intended function of the translated text in the new social context), and cultural transediting (i.e., adaptation to the needs and conventions of the target culture).

Since she was among the first scholars to turn attention to this issue, Stetting’s ideas were subject to much discussion among translation scholars, as well as criticism. While acknowledging the role of the term to raise attention to a previously neglected issue, Schäffner (2012) rightly points out that an additional terminological extension is not actually necessary in order to capture the specifics of news translation since such an approach would be needlessly reductionist. It could imply – and perpetuate – the dated view that translation is concerned with the transfer of meaning and word-for-word equivalence, while modern translation studies theories have moved beyond such a view. She acknowledges that “[a]s any translation, news translation, or media translation more generally, is a textual and a sociocultural process which involves transformations” (Schäffner, 2012: 881).

The terminological difficulty of how to best refer to the text-production practices in news and other media contexts is a common bone of contention in translation studies. According to Valdeón (2014), these practices form a cline ranging from translation to adaptation, involving various degrees of transformation of the original text. He argues that linguistic and cultural transformations can serve the purpose of framing news items for the target audience, suggesting that in specific instances, they can even become appropriations of the originals. Through appropriation, “the foreign can become more palatable by preserving its origin” (2014: 58) – and conveying such an effect may actually be more important than merely adapting the text for the local audience.

Journalistic translations are typically produced by journalists themselves as part of their work. Indeed, as Bielsa and Bassnett (2009: 65) observe, “they [journalists] do not see translation as a separate process from the edition of texts”. The work of journalistic translators can be co-opted under the notion of language “mediators” (Pym, 2004: 55), i.e., individuals who produce novel texts (summaries, reports, explanations, consultations) on the basis of pre-existing texts in other languages (or, even, in different varieties of the same language). In this sense, such linguistic mediators operate as knowledge or information brokers working across languages. News translation has then been justly described as “a genre that locates itself between localisation and cultural mediation” (Orengo, 2005: 175) or “a transformative act” carried out by “intercultural mediators” (Hatim & Mason, 1990; Katan, 2004).

It is worth noting that the concept of mediation, which can be defined as the practice of linguistic adaptation across both inter- and intra-language communicative contexts (see also Valdeón, 2005, 2008), has recently been extended into other disciplines as well. This includes fields such as language education (Dendrinos, 2006; Chovancová, 2016), where mediation is treated as a distinct communicative skill. In sociolinguistics, a similar practice could be subsumed under the concept of “translanguaging”, particularly in the context of multilingualism (García & Wei, 2014). Nevertheless, whatever the specific conceptual framework, the various approaches indicate that the “trans” dimension (i.e., trans-lating/-posing/-editing/-forming) involved in the spanning of meanings across different linguistic codes, cultures and communities goes beyond the notion of equivalence. In news media contexts, journalistic translation operates as a specific text-processing and text-producing practice that tackles not only the localisation of the source text but also its transformation and metamorphosis into a new textual product.

In this context, two main dimensions of translation in relation to news need to be distinguished. First, there is “news translation” – simply the process of translating and republishing items in foreign language mutations. Here we have a source text and a target text that can be aligned and subject to analysis using the standard tools of translation studies. Typically, this process can also involve some degree of localisation since translated texts should, as suggested by Orengo (2005: 170), have “the feel and look of a nationally-manufactured piece of news”. The second dimension concerns “journalistic translation”. This is the process when, on the basis of translation work, a new text comes into existence. Clearly, more creativity is involved: journalists draw on foreign language source texts (news reports, quotes, speeches, etc.) and incorporate them in their own news texts. As Bielsa and Bassnett (2009: 10) note about modern news translation, “the dominant strategy is absolute domestication, as material is shaped to be consumed by the target audience, so has to be tailored to suit their needs and expectations”. Once again, this general approach is markedly different from, for example, literary translation, where issues of formal and stylistic equivalence play a much more important role.

Current literature does not make this terminological distinction, with authors preferring to use the concept of “news translation” generically in reference to any aspect of translation work in the production of news in another foreign language (Bielsa & Bassnett, 2009). Nevertheless, the distinction appears to be justified with reference to how journalists engage with the source media: ethnographic studies of journalistic news-text production processes indicate that their interaction moves along the continuum from “informational transmission at one end to practices of negotiation and entextualization at the other” (Van Hout & Jacobs, 2008: 78). Similarly, Valdeón (2014: 60) suggests that translation should be distinguished from other practices (such as adaptation and appropriation) that are involved as strategies of news translations.

Clearly, news translation is poised between two worlds with different professional rules and ideals. The requirements of news production mean that a news item need not be entirely original, and can be expected to contain wording and structures found in other texts (Bell, 1991; Cotter, 2010). Needless to say, this also applies to non-translated news items based on agency copy or some other preformulated external content that finds its way into the final news product (cf. Jacobs, 1999). Such textual recycling not only saves time but also routinises news production (Van Hout & Jacobs, 2008: 75). At the same time, this situation also means that some of the professional expectations stemming from ideals underlying translation work may have to be compromised: the journalistic news translator need not feel obliged to render the source text with the same professional considerations (i.e., engaging in translation proper rather than “transediting”) as professional translators would. Within the ethnography of the news production process (cf. Catenaccio et al., 2011), translation work has an ancillary function, though it is – paradoxically – at the core of many items about international hard news and human interest stories.

As a result, more radical textual transformations tend to be the rule in translation-related news, posing a challenge to classic translation studies approaches that address formal and stylistic equivalence between source and target texts. For Bielsa and Bassnett, news translation is primarily concerned with the transmission of information, where:

translation is one element in a complex set of processes whereby information is transposed from one language into another and then edited, rewritten, shaped and repackaged in a new context, to such a degree that any clear distinction between source and target ceases to be meaningful.

(Bielsa & Bassnett, 2009: 11)

Although news translation is almost omnipresent in today’s globalised world, it is certainly nothing new. The practice has a documented history of several centuries. Ever since their beginnings, early newspapers depended on content partly supplied from external sources and foreign correspondents. Brownlees (2011), for instance, documents how seventeenth-century English corantos were based on external news content that was translated for the domestic audiences mainly from Dutch and French sources. Nevertheless, there appear several significant differences between news translation in early newspapers and in modern news media. First, there was little textual transformation in the early days of print newspapers: news items were translated in a very literal manner. Second, the news was not contextualised or put into any perspective: it was simply presented without any elaboration or explanation.¹ Finally, there was the almost total absence of quotes – “events were reported but people were not heard”, as Brownlees puts it (2011: 51).

That situation contrasts with the present-day media, where (a) translation typically entails adaptations that are so extensive that it sometimes may be difficult to talk about actual “translation”; (b) translated items and textual segments are editorially reconstituted into new texts; and (c) there is an abundance of quotes, i.e., heteroglossic textual segments. Interestingly enough, out of the translated content, it is direct speech quotes of news actors that appear to be the most resistant to textual transformation: quoted passages, representing the seeming authentic external voices of news actors, are typically translated in a relatively faithful manner. In spite of that, however, it appears that when the issue of journalistic mistranslation is raised by various stakeholders (typically elite news actors, including politicians and government officials whose statements will have a significant international impact) in subsequent reactions to mediated news items, such accusations are directed either against translated direct speech quotes on the grounds of “misrepresenting” the actual words or “taking them out of context” (see section 3).

The role of modern news translation is, then, characterised by a functional duality. As Orengo (2005: 170) says, translation is not only fundamental for news transmission but also marginal as regards the process of news making. More specifically,

[w]ithin such a process, translated texts are dismembered, used as raw material and not viewed as target texts, since the journalist’s real goal is the production of a news story (i.e., a totally new text) and not the presentation of a target text in its own right.

(Orengo, 2005: 170)

It may appear somewhat paradoxical that translation is simultaneously so important, as far as the processing of news source texts is concerned, and so negligible, as far as the construction of the final textual end-product is concerned, be it a localised news item or an entirely new news report. However, this is the result of the fact that “the translated text is not a goal but a means either to the construction of another text, or to the distribution of another product” (Orengo, 2005: 173).

1 News translation between semantic transposition and textual transformation

Research into news translation is a relatively new field: it has emerged as a focus of attention by linguists and translation scholars only in the past couple of decades. Surveys of recent research on journalistic translation can be found in Schäffner (2012), Károly (2012, 2017), Holland (2013) and Valdeón (2015). Some of these scholars also pay attention to research that has been published nationally (particularly in Spanish and Hungarian), reflecting considerable attention to practices at the local level, attention that has received insufficient international exposure. In terms of their theoretical approach, it is probably of little surprise that most of the relevant work has been grounded in translation studies, but some other linguistic disciplines have provided useful insights as well. The present section reviews some of the studies that are particularly relevant for the purpose of this chapter.

Many scholars have dealt with the distinct forms and strategies of adaptation of translated texts that give rise to various kinds of shifts and non-equivalences between source and target news texts. Van Leeuwen (2006), for instance, studied 100 translations of news items from Vietnamese into English with the aim of identifying the changes that the source texts undergo in the process of translation. Reflecting on how the local becomes globalised, he has noted how “the local Vietnamese newspaper style [becomes] transposed into global ‘journalese’ and local cultural and ideological references transformed into globally understandable and acceptable versions” (van Leeuwen, 2006: 218). Van Leeuwen documents how editors modify the translated text, in order to have the English version conform to Anglo-American norms. This involves three dimensions: (a) micro-level linguistic features (such as reporting clauses, nominalisations, articles, tenses, etc.), (b) higher level phenomena (such as the modification of a flowery style and idiom towards a more succinct form of expression), but also (c) the repositioning of the reader, as attested by the need to perform cultural and ideological adaptation for readers unfamiliar with the local context. Contrary to expectations, van Leeuwen argues, such target-oriented adaptations of translations do not lead to a complete “loss of the voice” of the Vietnamese journalist. While many journalists/translators demonstrably like to see their style refined into “good ‘global’ English”, even if it “entails some slippage of voice and thereby perhaps also of meaning”, foreign sub-editors actually “favour (re)creating a local ‘accent’” (ibid.: 225).

Stylistic and ideological adaptation thus seems to stand between homogenisation, arising from the forces of globalisation, and the preservation of some features indexing the locality of the text’s origin. Moreover, Western cultural values can sometimes clash with the source text norms. A particularly interesting case mentioned by van Leeuwen concerns attributions, since foreign sub-editors, who like using lots of direct speech in target news texts, tend to reformulate translated speech in order to make it sound more idiomatic in English. However, this becomes problematic with government officials since it “will distort the exact meanings the leaders are trying to convey” (ibid.: 229). Hence, preference is given to replacing direct speech quotes with reported speech, even though it causes stylistic problems and can also entail, for instance, the introduction of reporting verbs that convey evaluation, i.e., the translating journalist’s subjective stance. From a pragmatic perspective, we see that the conscious avoidance of direct speech is actually used by the sub-editors as a hedge in anticipation of the possible semantic non-equivalence. Significantly, then, the text-producing strategy of the target news reflects the anticipated effect on the target audience and is motivated by the power and social status of the news actors concerned.

The restructuring of translated news texts according to the textual conventions of Anglo-American print media draws attention to some of the transformative processes at play along the movement from local to global and vice versa. As Orengo (2005: 173) observes,

[g]lobalisation involves opposing movements, since [. . .] one must paradoxically argue that globalising means making a local product global and then local again, over and over, so that more locales can use it. Localisation is the process of making a product that was designed to be marketed on a global scale usable locally.

A global news report, as argued by Orengo in his study of translations into the Italian locale, becomes subject to not only interlinguistic localisation but also intralinguistic adaptation “to suit readers’ political leanings within the same linguistic locale” (ibid.: 168). The news localisation process thus has a decidedly ideological dimension.

In a similar vein, Valdeón (2014: 60) suggests that “the original emphasis on language transfer and edition is of lesser importance than the political, economic and social implications of processes like adaptation and appropriation”. In fact, translation serves the function of framing, realised through such strategies as selection and omission of information, quotes and text segments. Moreover, since not all international wires that are received by a media office can be placed in the newspaper or online, translation serves a gatekeeping function (see also Vuorinen, 1995; Song 2017). In this manner, a media outlet can use translation of selective news items to support its own ideological positions and those of its target audiences, which constitutes a distinct appropriation of the source text for one’s own purposes. By contrast, a translated news item may also be used locally to undermine a foreign news source (Kang, 2010). All this indicates that the notion of perlocutionary equivalence (Hickey, 1998), i.e., the idea that translated texts should aim for an equivalent pragmatic effect in the source and the target audience (cf. the scopos theory in translation studies, Nord, 1997), cannot be readily applied to news translation. Issues related to power and ideology play a significant role in this process, despite the fact that they mostly remain hidden and buried in the institutional background of news production processes.

The ideological nature of journalistic translation has been observed by other scholars. Adopting a critical discursive perspective, Schäffner (2012: 879) suggests that the selection and transformation of information not only helps the readers to understand but also conveys ideologies. The intervention of translators in news texts can be documented on the level of lexical choices or syntactic structures, e.g. the passive voice used to avoid agency (Schäffner, 2002, 2010). Schäffner (2002: 34) argues more generally that “it can be said that any translation is ideological since the choice of a source text and the use to which the subsequent text is put is determined by the interests, aims, and objectives of social agents”. This perspective is, in fact, in harmony with the basic precepts of critical discourse analysis (CDA), which argues that any use of language is always given from a particular position, and hence is never entirely neutral. Any linguistic representation entails a point of view of its encoder (Fairclough, 1992; Wodak, 2001; Hart, 2010).

Most of the classic CDA studies have worked with news media data to reveal the ideological basis and underlying bias of language and representation. Though CDA was occasionally criticised in the past for its activist approach and social commitment, raising accusations of bias and non-objectivity (Widdowson, 1995), it provides a useful set of analytical tools that enable us to see the link between the levels of textual analysis and social practice, pursuing a contextualised interpretation of texts. Arguably, this kind of approach is helpful wherever critical reflection of one’s own semiotic practices is needed; translators can benefit from it by developing their own critical awareness of the effects that their textual work may have.

In news translation, the critical dimension has a dual level: the translator should be aware not only of the ideological positionings of the original news text and the constructive nature of representations contained therein (and perhaps render them in the target text), but also of the ideological implications of his or her own textual transformation into the target text. These two dimensions may be in harmony, as in the case of the media gatekeeping certain content favourable to their own view of the world (see above). Likewise, however, they may be in conflict. In that case, translators may position themselves – whether consciously or unconsciously – with respect to the stance or ideology communicated in the source text. Since – generally speaking – the professional ethos of journalistic translation is geared towards textual transformation rather than mere translation (i.e., the production of new texts rather than the remediation of prior texts from other sources), such a clash of ideologies makes the target news text internally dynamic, polyphonic and even dialogical (Bakhtin, 1981). After all, journalism itself is an “interpretive practice” and news production is “a process of entextualization involving multiple actors who struggle over authority, ownership and control” (Van Hout & Jacobs, 2008: 60).

2 Shifts in news translation

Let us consider some case studies to see how such diverging ideologies can be realised in news texts. It is noteworthy that sometimes, even a translation solution that is classifiable as an error may be congruent with a prevalent underlying ideology of the journalist or the target audience. Reporting on a case of mistranslation found in British news on speeches from the German Parliament, Schäffner (2002) documents how the choice of words creates a different impression from the original. For instance, the German phrase “ein fester Kern”, used in a German parliamentary document on the proposal for the formation of a core group of closely integrated EU member states, was rendered as “hard core” in the English translation in the Guardian. Schäffner argues that the original conceptual metaphor has a positive connotation in German, “suggesting a firm commitment to European integration”, while the English version “is frequently associated with people and things that are tough, immoral and incorrigible” (2002: 48). Consequently, the German proposals were perceived in a negative way as “an attempt of the core countries (and in particular Germany) trying to impose their ideas on all member states”. However, this may be more than a mere mistranslation because “the selection of information [. . .] fits into a traditional way of reporting about Germany and seems to reveal deep-seated perceptions and stereotypes about the Germans” (Schäffner, 2002: 55).

In another study, Calzada Pérez (2007) documents shifts in transitivity in translations of speeches made before the EU parliament in English and Spanish. Several systematic divergences between source/target texts are identified: shifts in agents’ animacy (i.e., their human or non-human nature), causation, voice and depersonalisation. Some shifts are obligatory, arising from the structural differences between languages. Other shifts, however, stem from translators’ individual choices. While some translation shifts are isolated, others have been found to be cumulative, resulting in more significant textual shifts over larger stretches of text. Apparently, “ST transitivity processes are not necessarily scattered at random but may cluster together, forming trends that respond to semiotic/ideological influences and reinforce isolated effects of particular processes” (2007: 147). Shifts of transitivity give rise to unwarranted effects, which can be unconscious or unintended, but may have significant “pragma-semiotic (ideological)” consequences (ibid.: 3).

The cumulative effect of deviations found in the target text has been commented on by other scholars as well, e.g. Pan (2014: 258), who combines Fairclough’s strand of CDA and appraisal theory to document how translations of news about Lhasa riots offer the Chinese target audience an ambiguous frame of the news event. Based on data from Reference News, a Chinese newspaper exclusively publishing professional, in-house translations of foreign reports, the analysis shows that news outlets resorted to a form of mediation that amounted to “filtering rather than translating news items” (ibid.: 260), due to the potentially sensitive nature of the events in Tibet for the domestic audience. The changes involved such strategies as labelling deviations, removal of ethnicity labels and downscaling (e.g. the omission of the adjective angry in collocations describing the attackers as mob). These deviations have served to alter the Western narrative of the events found in the source texts, changing it into a different narrative that is more congruent with domestic expectations, and defocusing from the political and ethnic nature of the riots. A survey carried out among the producers of Chinese media texts by Pan has indicated that their work does involve, as far as negative and sensitive news is concerned, taking into consideration the attitude of the government as well as the possible response from the Chinese audience. More specifically, it appears that for in-house translators, “filtering [is] a necessary means for guaranteeing the target reader’s proper reactions” (ibid.: 260).

From a combined cognitive and critical pragmatic point of view, such a practice consists of an intentional manipulation of symbolic discursive spaces. In the proximisation theory proposed by Cap (2008, 2013), for instance, similar strategic deployment of lexico-grammatical devices is understood as a means for the (de)legitimisation of various political or public policies. The symbolic space constructed through any discursive action along the temporal, spatial and axiological axes (cf. Chilton, 2004) can be skilfully manipulated in order to create a desired effect on the audience (e.g. make a threat feel more imminent, and thus “proximal”). Arguably, the textual transformations involved in news translation can be used to similar ends, in harmony with the ideological orientations of the news media outlet. With respect to the above-mentioned Tibet riots case, thus, the labelling deviations and other transformations in the target text function to remove the potentially sensitive nature of the news event or the incongruent (Western) ideology found in the source text. Such an entextualisation constitutes a forced cognitive construal that mitigates the assumed undesirable effects on the audience: it serves to increase the symbolic distance between the news event (as reported in the translated news report) and the target audience. In this sense, it is an act of dissociation or pragmatic “distanciation” (cf. also Wieczorek, 2013).

Nevertheless, the proximisation/distanciation effects need not result from intentional textual manipulation only. They may simply arise out of mistranslation. Bielsa and Bassnett (2009) note that news agencies in particular need to be vigilant in this respect because any errors can have damaging consequences due to the speed with which they spread globally. They give the example of a mistranslation into Spanish of a suggestion by Donald Rumsfeld in 2004 that there might be a risk of a terrorist attack in the US. Apparently, “Rumsfeld alluded to the attacks in Spain to refer to the possibility of the US being attacked, which was translated as a warning of possible new attacks in Spain” (Bielsa & Bassnett, 2009: 150). A similar case is reported by Holland (2013: 333), who mentions that CNN mistranslated the assertion, made in Farsi in 2006 by President Ahmadinejad of Iran, that the country had “a right to use nuclear technology” as “a right to use nuclear weapons”. In both cases, the mistranslation has the undesirable effect of reducing the symbolic distance through the proximisation of the threat to the target audiences, and thus having a potentially detrimental impact.

The misplaced effect of what are obviously translation errors should be distinguished from semantic indeterminacy in the source text. While vagueness and ambiguity may prove challenging for translators and both source/target text audiences, they may ultimately give rise to some unwanted perlocutionary effects as well. In 2016, for instance, a tweet by President Donald Trump became subject to some debate as to its intended meaning (Figure 7.1). The concern was directly linked to the fact that the pronouncement was bound to become global hot news, and translated into many local media contexts.

While Twitter communication by politicians seems to be ideally suited for succinct, impactful pronouncements on various issues, catering to the media’s interest in brief soundbites, it becomes frequently problematic on account of its lack of context. In Trump’s tweets, the decontextualisation combines with the president’s highly idiosyncratic communicative style (cf. Williams & Prince, 2017). The utterance in Figure 7.1 appears problematic not only on account of its brevity and bluntness, which run counter to the cautious and diplomatic language typically used by public figures informing about national policies (the previous US nuclear weapons policy consisted of a 64-page report, as noted by Fisher, 2016), but also due to its vagueness. Any of the phrases in the tweet can have multiple meanings and thus invite a whole range of possible interpretations.²

While being ideal for inclusion in news reports because they constitute “pre-formulated” textual segments (Jacobs, 1999), tweets and parts of public speeches by politicians have also become staple features of translated news, though they may create quite different effects in the target culture. In an opinion piece written for the Guardian, Doshi and McCurry (2017) report on the translation strategies adopted by many international translators and interpreters faced with the task of rendering faithfully the content and style of President Donald Trump’s speeches. In India, for instance, news and broadcast media have been consciously avoiding this problem by reducing Trump’s speeches to mere soundbites or paraphrasing them extensively. Also, they observe that “In English, Trump may not sound very intelligent, but when you translate him with context in Hindi, it makes him sound much better than he is” (Doshi & McCurry, 2017: n.p.).

On the other hand, Trump’s frequent colloquialisms – indicative of what has been referred to as a “restricted code” in sociolinguistics – may pose a problem for translation in cultures where the social identity of public figures is associated with the use of an “elaborated code” (Bernstein, 1971). The latter is associated with standardness and sophistication, being indexical of the speaker’s high social status, elite education and prestige. In the case of Trump’s quotes, for instance, phrases such as “nut job”, and the notorious “grab the women by the pussy” have left English-to-Japanese translators facing problems with how to reconcile the tension between being faithful to the original and not offending the normative sociolinguistic expectations of the target audience, eventually opting for neutral and non-offensive language (Doshi & McCurry, 2017). A different problem is posed by Trump’s “occasional absence of logic” and mangled sentence structure, which is particularly testing for interpreters for different reasons. Chikako Tsurate, an interpreter and professor of translation studies, concludes that:

I tell my students that with simultaneous interpretation, the trick is to anticipate the speaker’s intentions and tell a story, to be slightly ahead of the game. But when the logic is not clear or a sentence is just left hanging in the air, then we have a problem. We try to grasp the context and get at the core message, but in Trump’s case, it’s so incoherent. You’re interpreting, and then suddenly the sentence stops making sense, and we risk ending up sounding stupid.

(Doshi & McCurry, 2017: n.p.)

3 Some issues in the pragmatics of news translation

In recent years, the approach of translation studies has moved significantly towards a dynamic conception of the speech event as it is understood in post-Gricean pragmatics – particularly interactional but also cross-cultural. Baker (2006a), for instance, emphasises dynamically changing contexts, the fluidity of interactions as well as the joint construction of meanings that underlie the entire translation process. She points to the “active negotiation among participants with shifting agendas and unequal levels of control over the interaction” (p. 335).

In the present section this dynamic approach underlies the discussion of several issues related to translation work involved in the production of news texts. First, it deals with some common instances of localisation of the news for the target audience (section 3.1). Second, it identifies the presence of translation metadiscourse in news translation as an under-researched but potentially salient area of research (section 3.2). Next, it discusses how (mis)translation itself can become subject to news coverage (section 3.3), identifying fake translation as a distinct form of fake news and discourse manipulation (section 3.4). The chapter then describes how some features of participatory media, notably reader comments, complement the official voice of the paper (section 3.5). Finally, it briefly mentions the media trend towards multimodality in news stories, pointing out that more attention is needed for understanding the way the visual material is transposed across languages alongside text (section 3.6). Most of those issues are illustrated with translation-based data from the Czech language media. The discussion reflects the shift of translation research towards more interdisciplinarity, where increasing attention is being paid to the connections between translation, power, ideology and narrative construction (cf. Baker, 2006b; Schäffner, 2010; Sidiropoulou, 2013).

3.1 Localisation of news content

The localisation of news content is a very pertinent area for undertaking pragmatic and discourse analysis because underlying all such modifications and shifts is the translator’s conscious attention to the target recipients. The translator seeks to provide an optimal amount of information, which may mean resorting to various explicitation translation strategies (Pym, 2005; Kamenická, 2007), privileging certain aspects of news items (Valdeón, 2008), or adjusting the structure of news stories by making shifts in topical development and rhetorical structure of news (Károly, 2012).

Some typical instances of localisation are illustrated in Figure 7.2, which shows a screenshot from the Czech-language news site blisty.cz. The news item reproduces the source text – the original tweet by Nick Robinson in English – preceded by the target text, i.e., the Czech translation:

The target text is a relatively faithful translation of the English original, but it contains several deviations that attest to the translating journalist’s transformation of the text. The first difference is to be found in the attribution line. Since the average Czech reader can hardly be expected to know who Nick Robinson is, the following descriptive note was added in the Czech version in order to provide the missing cultural context: “Nick Robinson, moderator of the morning commentary programme on BBC radio”. This is an instance of pragmatic (cultural) explicitation in the target text (Kamenická, 2007: 48) that is evidently motivated by Gricean cooperation. At the same time, the translation has removed the reference to Robinson’s Twitter account (which is partly indexical of his institutional identity, cf. “@bbcnickrobinson”) and to the radio programme itself (“@BBCr4today”). The latter is replaced in favour of the phrase “on BBC” in the body of the tweet and becomes re-expressed in the attribution line. The final intervention by the translator consists in the specification of the temporal deixis of the tweet: where the original uses the adverbial of time “today”, which is fully sufficient since the tweet constitutes the news event itself, the translation elaborates by specifying “dnes ráno (v úterý)” (“today morning [on Tuesday]”). Once again, the translator opts for a greater explicitness in the translation because the target text remediates the tweet beyond its original context; the translator has thus deemed that a more specific temporal anchorage is needed for the benefit of the Czech readers, even though the translated news item appeared on the same day within hours after the English original.

The above-example also illustrates another phenomenon characteristic of modern online news, namely their multilingual character. It is increasingly the case that source texts (typically tweets or social media posts) are embedded – as accompanying visuals – within a foreign-language media text. This practice is affecting the structure of the news item, leading to the fragmentation of the traditional news texts (see also Facchinetti, 2012). The juxtaposition of the source text (typically in English) and its translated version (in a local language) within the target news text invites readers to engage their multilingual skills and pragmatic competence. Thus, they not only gain a more direct access to the utterance but may also evaluate the result of the media’s translation process.

3.2 Glossing translation through metadiscourse

In cases where a source text is not very semantically transparent, the media sometimes engage in some form of translation-related metadiscourse in their attempts to approximate the meaning. While such a strategy could be subsumed within the broad concept of “localisation” or “explicitation” strategies (see above), it may – interestingly – appear in other situations where the original formulation may not be lacking semantic determinacy. In other words, the metalinguistic gloss is essentially redundant because the meaning is sufficiently clear. For instance, in January 2017, Donald Trump made a widely-publicised comment in an interview for the British newspaper The Times, describing the Nato organisation as “obsolete”.³ In one of its reports on this issue, the Czech online news site idnes.cz included the following metalinguistic comment on the translation of the expression:

(1) Kritická slova nového amerického prezidenta Donalda Trumpa na adresu evropských spojenců o tom, že je NATO „zastaralé“ či v tom horším překladatelském významu „překonané“, zapůsobila mezi evropskými lídry jako třaskavina. (https://zpravy.idnes.cz/ruska-media-o-trumpove-kritice-nato-d6w-/zpr_nato.aspx?c=A170122_200537_zpr_nato_inc, 22 January 2017)

[BACK TRANSLATION] The critical words of the new American president Donald Trump addressed to European allies stating that NATO is “obsolete”, or in the worse translated meaning “overcome”, have had an explosive effect among European leaders.⁴

Since the English adjective “obsolete” has a relatively straightforward equivalent in the Czech expression “zastaralý” (lit. “dated”), this was the expression used by most of the media to report on Trump’s statement in translation. However, idnes.cz took the editorial liberty of elaborating on the meaning, suggesting a less common, though possible translation equivalent “překonaný” (lit. “overcome”). This has an undeniably stronger effect and serves to increase the negative news value of the story; as also attested by the translation-related metadiscursive comment made by the paper itself (“in the worse translated meaning”). The unnecessary elaboration on the additional equivalent meaning of the expression in the target language seems to indicate the media’s conscious attempt to intensify the pragmatic effect of the utterance on the audience, essentially shifting the meaning from “Nato being old-fashioned” to “Nato being a thing of the past”. Clearly, the study of translation-related metadiscourse in the media promises to bring some novel insights into the media’s metapragmatic awareness.

3.3 (Mis)translation in the news

Another issue concerns situations when translation itself becomes the subject matter of news coverage. More often than not, this obtains in negative contexts, e.g. in alleged cases of mistranslation. Many politicians and other public figures frequently make claims of “translation errors” or “their words being taken out of context” when their utterances, made for foreign media channels or in foreign-language contexts, result in some dispreferred reactions from the public and the media. Such claims give the speakers a limited possibility to control the meaning and to distance themselves from the unintended perlocutionary effects of their utterances – they are a kind of post-factum hedging that mitigates the negative impact of their words.

A more insidious situation arises when mistranslation is suspected of being used by the media in order to pursue a particular agenda. Special problems arise particularly where the source text is either not available or where a little known or non-European language (such as Arabic) is involved, making it very difficult to assess the degree of equivalence and to trace any shifts in translation (see also Holland, 2013: 344).

Sometimes, mistranslation – or allegations thereof – can be involved in the generation of heated public debate on unrelated issues and contribute to the formation of some aspects of shared cultural knowledge of a given community. For instance, in February 2016, the nationwide Czech channel TV Prima ran a news story on a 17-member Christian family from Iraq that had just been resettled, thanks to help from an endowment fund Generace 21, into the city of Jihlava from a refugee camp in Lebanon. The news report included a section of an interview with the head of the family, with the voiceover rendering his statements translated from Arabic into Czech as follows:

(2) Chtěli bychom tady zůstat a později i pracovat, ale nejdůležitější je pro nás bydlení. Než abychom bydleli v přemalovaném kravíně, tak to se raději vrátíme do Iráku.

We would like to stay here and later also work, but what’s most important for us is accommodation. Rather than staying in a redecorated cowshed, we’d prefer to go back to Iraq.

It is perhaps little surprising that the news report caused significant criticism from the public as well as other media, culminating in a complaint filed with the Council for Broadcast Media. The Generace 21 endowment fund accused TV Prima of manipulation and tendentious reporting, arguing that the problematic words had never been uttered, while the TV reporter countered by alleging pressure from Generace 21 not to air the report. The endowment fund produced its alternative translations from Arabic, as did the TV station, each availing themselves of officially certified, court-appointed interpreters to prove their case. It was eventually pointed out, at the micro-level of linguistic form, that the Iraqi was using a conditional construction, referring to a hypothetical scenario rather than the actual situation, though the unedited interview also showed that the topic of inadequate housing was brought up by the interviewee himself several times during the interview without a direct prompt from the reporter.

Nevertheless, what came to be known as “the redecorated cowshed case” (“kauza přemalované kravíny”) quickly extended from a case of contested correctness of translation to an argument against immigration. With respect to the media, it sparked intensive debates about whether or not, through the skilful editing and juxtaposition of words and images, the TV channel manipulated the report in order to create a negative impression of the Iraqi family, implying that they were ungrateful and undeserving economic migrants rather than genuine refugees. The interpretation was also alleged to fit the unstated editorial line of the channel (as demonstrated by a leaked recording of an internal meeting laying out the channel’s preferred policy towards immigration). Ultimately, thus, the case helped to delegitimise the issue of immigration, as well as the alleged political correctness of the national (as well as Western) mainstream media, in the eyes of the Czech public. When the family moved to Germany to seek asylum only two months later, it was seen as the ultimate confirmation of their undeserving status of refugees.⁵

3.4 Fake translation

In translation, errors may and do occur. However, errors differ from intentional acts of mistranslation: the latter concern such deliberate floutings of the Gricean quality maxim that are intended to mislead the target audience, while falsely giving the impression of being genuine communicative acts. Such wilful miscommunication can also involve intentional fake translation.

In December 2013, the memorial service for the late South African president Nelson Mandela was marred by a bizarre incident involving the official sign interpreter of the event. For hours, the interpreter stood alongside world leaders, making nonsense gestures that were subsequently analysed by sign language experts without any success (see, e.g. Laing, 2013). Despite earlier complaints about his inability to sign (submitted by the National Federation of South Africa to the ANC), the fake interpreter was allowed to appear on stage at a high-level event. While it was evident to the deaf community throughout the duration of the ceremony that the man was an imposter, the vast majority of the worldwide audience was unaware of this fact until subsequent media accounts emerged days later uncovering the botched act of interpreting. Though the fake interpreter eventually tried to defend his poor performance by suggesting that he had suffered a schizophrenic episode in the act, the incident pointed to the broader problems with professional incompetence of official interpreters/translators in South Africa (Pienaar & Cornelius, 2016).

The case shows that as long as the audience is not in command of either the source or the target code, be it a different language or some other signifying system, they have to rely on the translator’s/interpreter’s professional integrity in rendering the translated text in a faithful matter. While some translation shifts will inevitably occur, wilful mistranslation breaches the fundamental trust that the audience places in the language mediator. In pragmatic terms, intentional mistranslation flouts the cooperative principle, resulting in a conscious manipulation of the intended effect on the audience. In media contexts, such intentional mistranslation can form a specific type of fake news.

The Nelson Mandela example, moreover, also highlights the “emblematic” role of some translation/interpreting acts, where the semantic function is backgrounded in favour of other considerations. The fake interpreter was, in fact, engaged in meaningless glossolalia, producing “a text reduced to pure textuality, that is to pure pragmatic value: not a series of expressions that convey representation, but an act, an intervention in a situation” (Lecercle, 1999: 191). The fake sign interpreter even became the subject of humorous treatment in the serious media, e.g. the British Telegraph came up with humorous “translations” of some of his hand gestures (cf. Read, 2013).

Yet another issue concerns cases of remediation that changes the genre status of the news texts. Occasionally, translations of either spoof news (cf. Ermida, 2012) taken over from satirical websites or of non-serious news items published in regular newspapers are included among serious news in the target culture. While the original humorous or non-serious intent may not be readily perceived in some cases, such practice constitutes a hoax and may be used to further a newspaper’s agenda, or to promote an alternative narrative (Baker, 2006b; e.g. in matters relating to immigration).

3.5 Participatory journalism

The advent of the online world has resulted in a profound transformation of the role of the media. Presently, we all experience a situation of “transworld simultaneity and instantaneity” (Bielsa & Bassnett, 2009: 20). In the past, the media had a monopoly on information, enjoying a privileged access to foreign news sources, news wires, etc. that served as source texts for their textual transformations into target news texts. While some of that privileged role still remains, media audiences nowadays have instant access to many of the same sources as the media. That is particularly the case where such English-language global media as the Guardian, CNN, BBC, etc. publish news items on their websites that serve as a basis for either direct translation or reuse in media on the national level.

Thanks to instant access to source texts, readers frequently juxtapose the local translated text with the original material. In this sense, they serve as watchdogs over the local media’s news text-producing practices. Reader discussion forums in national newspapers become sites where not only meanings are contested, (news) ideologies revealed and news sources corrected but also the correctness of the media’s translations are commented on.

Sometimes, readers have a partisan approach in correcting the media, uncovering their alleged bias. Readers often complement the news stories by providing direct links to the original source articles in foreign language media, enabling others to obtain unmediated access to the source texts and see beyond the mediated representation in the national (local) press.

There are situations when the media either cannot or will not provide complete information. In the UK, for instance, there have been many recent cases when courts have issued so-called super-injunctions, forbidding the press from revealing the identity of celebrities (cf. Chovanec, 2018). Aside from legal restrictions, media in other countries may sometimes prefer not to disclose the identity of news actors in translated news items for some other reasons. For example, where the provision of full names could reveal the ethnic origin of crime victims or offenders, some media will not use their full names, supposedly not to activate the audiences’ ethnic and racial stereotypes and to prevent racist hate speech in online comments. The post-media discourse space of reader comments is then frequently used for sharing links to the source news items and for disclosing the full identity (and hence often the ethnicity) of news actors. Users in discussion forums then not only correct mistranslations but also bridge gaps in reporting and enrich the mediated news content by alerting others to modifications and omissions of information, often speculating on the ideological significance of such practices.

Viewed from a CDA-perspective, this kind of audience participation exposes, on the one hand, the prevailing public ideologies and power and, on the other, opens up the world of the news media to critique. As pointed out by Baker (2014: 22), it is not only the mainstream industries (publishing, news, etc.) that seek to influence global publics in the globalisation era, but also “amorphous groups of fans and activists who wish to pose a challenge to the dominant world order [and who] also use translation and interpreting to undermine existing structures of power”.

3.6 Translating multimodal news

While it is easy to concentrate fully on the linguistic aspect of translated news texts, we should not forget that news and other media texts are multimodal units that have a textual as well as a visual component. In modern online media, other elements are present as well, e.g. links to video material, social media and other hyperlinked content (see Figure 7.1).

Traditional news has relied on nuclear images, often sources from international news agencies, around which the news story is constructed with its structural (headline, lead, image caption) and other textual elements (Bednarek & Caple, 2012; Chovanec, 2014). Modern media, particularly the popular press, rely on extensive visual presentation of the news story: the visual component often prevails and the story is increasingly told through a series of sequenced images and their captions or in accompanying photo galleries (Chovanec, 2019).

What happens to the visual component of news in translation? While language is understood to be traded cross-linguistically in a matter-of-fact fashion, it is a different matter with images since they are proprietary and subject to copyright protection, often of third parties. Where the source news text uses documentary photographs taken on location, the local news text that renders the news story on the basis of a partial translation from a foreign-language media typically will not have such images at its disposal. The images available to local media may be obtained through news agencies, but where such material is not available, the newspaper may complement the translated news item with photos taken from image banks. In exceptional cases, local media can even recycle photographs from different news stories that had been published previously and, thus, selectively appropriate them for their own purposes (cf. Baker, 2006b: 114). In my data on Roma immigration news in the British and Czech press, for example, there is even an instance in which an original documentary photograph was reused by the same Czech newspaper to accompany an entirely unrelated news item on a similar topic at a later point.

The way photographs are handled in such translated items is somewhat counterintuitive: visual content that is specific in the source text is replaced with generic content in the target text. Replacing the specific with the generic may appear as something of a paradox: as we have seen, on the textual level, the process of news localisation typically results in translation explicitation. On the visual level, however, we encounter the contrary process: generic images from photo banks change the status of the visual content from “documentary” to merely “illustrative”. Precise documentary meaning is sacrificed in favour of symbolic representation, where the visual element ends up having a decorative rather than a communicative function.

While it is not surprising that translation studies have tended to concentrate on the linguistic component of media messages, modern pragmatics-oriented analyses, which deal with the complex effects of the transposition of media texts into different linguistic codes and cultural contexts, face a different challenge. These analyses need to start paying more systematic attention to how such non-verbal modes participate in the construction (as well as shifts) of meanings in and between the source and the target texts, among other things.

Concluding remarks

This chapter shows that news translation, as a part of the ethnography of the news production process, is accompanied by frequent strategies of textual transformation. Research into news translation needs to take into account a combined product- and process-approach to analysis. Discourse analytical approaches, which have traditionally focused on a close textual analysis, can be aptly enriched by more ethnographic approaches that can take into account the processes of news production. As a matter of rule, texts become adapted for the target culture, and the process is governed by different professional imperatives than those usually followed by translators.

A broad pragmatic approach to news translation calls for a systematic study of the relationship between the producers, the texts, and the recipients, taking into account the intended and actual effects that the translated media texts have on the audience. That is in harmony with the current post-Gricean pragmatics, which is concerned with how speakers interact through language in view of such broader social phenomena as identity, power, gender, etc. and where the cross-cultural and inter-cultural dimensions play an increasingly central role. Nevertheless, given the complexity of the modern media discourses at a time when translingual flows of texts, signs and meanings across the globe are both institutionally mediated and personally accessible in real time, news translation research also needs, first, to accentuate the critical dimension that texts play in their trans-local contexts, and second, to become sensitive to the conflicting ideologies. The ideological dimension is located at multiple points: in the construction of original news texts, in their transformation into localised and translated news products, in their reception in new local contexts, as well as in their reflection and deconstruction by multiple stakeholders. Such a combined critical pragmatic research agenda is particularly acute at a time when intentional and unintentional shifts of meaning, misrepresentation and mistranslation are increasingly coming into the foreground in connection with media bias and manipulation.⁶

Notes

1 This phenomenon may be related to the fact-collecting approach that characterised early newspapers. In the past, newspapers served for the aggregation of external content (cf. Ungerer, 2002; Chovanec, 2017), often with minimal editorial intervention. This differs markedly from modern newspapers where translated news is not only contextualised and transformed within the news text itself but also frequently presented in a cluster of related articles that provide additional information and angles on the news event, leading to the fragmentation of the traditional news text (Lewis, 2003; Chovanec, 2014).

2 Writing for the New York Times, Fisher (2016) gives such possible interpretations as the following: “greatly strengthen” = “modernise existing nuclear forces”, or “expand qualitative nuclear capability”, or “deploy existing weapons systems closer to adversaries”; “expand” = “move some warheads from reserve stockpiles to active deployment”, or “build and deploy new warheads”, or “build new warheads, but immediately stockpile them”, etc.

3 The full text is available at www.thetimes.co.uk/article/full-transcript-of-interview-with-donald-trump-5d39sr09d.

4 Translation into English by the author. As the literal translation indicates, the metacomment is somewhat confusing since it refers to a “translated/translation meaning”. What is meant, rather, is a “translation solution/equivalent”.

5 Some time after the incident, Generace 21 cancelled the whole project for the resettlement of Iraqi Christians in the Czech Republic. Owing to their alleged dissatisfaction, the Iraqi family, which had already obtained asylum in the Czech Republic, left for Germany, where it faced the possibility of a lawful deportation to a Czech detention facility (and eventually, a forced return to Iraq). However, after availing themselves of the ecclesiastical protection of the Evangelical Church, their application for asylum was finally accepted for processing by the German authorities. (Source: https://zpravy.aktualne.cz/domaci/iracane-kteri-utekli-z-ceska-mohou-zustat-v-nemecku-pozadaji/r~4f91c160d99011e699ee0025900fea04/.)

6 This publication was supported with a research grant No. GA16-05484S of the Czech Grant Agency.