5.1.1Answer: “It’s my strategy. But not a written rule”

It is perhaps not surprising that no single overarching written norm was overtly claimed by the interviewees. For example, IN7F49, a freelancer and translation manager, when asked whether she received any explicit instructions from her government clients, indicated that specific translation guidance had rarely been given by the foreign-affairs office that her company worked for:

Rather than explicit written guidance, the translation traditions seem to be a rule of thumb. When asked if they would use an official or ‘authentic’ translation if available, most translators agreed that they would indeed adopt an existing official version, but they also admitted that there was no mandatory rule about that:

5.1.2Answer: “As a staff translator, I prefer a faithful translation”

Although the translators had only rarely received specific guidance on how to translate China’s political discourse, the idea of being ‘faithful’ was mentioned frequently, seemingly serving as a mental image of what a ‘good translation’ is in practice:

One conclusion to be drawn here is that the norm of ‘faithfulness’ seems to concern general professional ethics: as a translator, your role is restricted. At first glance, the insistence is a strong start-oriented defense: the original is something that is untouchable. Answers from other interviewees suggest that attempts to preserve a visible echo of the original also concern identity. The adherence to the default norm is largely motivated by the translator’s ideological stance:

This change of ‘stance’ is a sign of a hierarchy of norms: when translating Chinese political discourse, the identity of the institutional translator has priority over other ethical considerations, and this unsurprisingly results in restrictions on translatorial freedom. As institutional translators, the interviewees are subject to constraints imposed by internal rules and power relations. For them, it would be risky to intervene and give a new sense to the original.

5.1.3Answer: “It depends on what kind of faithfulness you want to pursue”

Most of the interviewees noted the value of ‘faithfulness’, but there was little agreement on how this fuzzy concept should be defined. Perhaps the concept itself is messy, mixed up with beliefs about formal correspondence, emphasis on functional equivalence, and insistence on the author’s intention – all these interpretations are competing for space and prestige. Some translators, but not all, were aware of the co-existence of contradictory definitions of the ‘faithfulness’ principle:

Here communicative effects challenge the prevailing norm of faithfulness in a formal sense. The translator points out that the idea of calling a representation in toto true is itself questionable: the possibility of linguistic and cultural untranslatability should be considered. FG1M41 also suggests that, at the very least, norms should be sensitive to reception. In this view, a literal translation like M1−INT is inaccessible to foreign readers:

In addition, IN7F49 recognized that faithfulness could be applied in a strict or loose sense, depending on the purpose of the start text – a view that recalls functional equivalence. For relatively low-sensitive topics, omission is fairly acceptable:

What can we conclude from these examples? First of all, most Chinese translators gloss ‘faithfulness’ as the dominant norm, as a part of ideological fidelity or translator ethics. Second, the nature of ‘faithfulness’ is far from clear, making the idea of ‘being faithful’ a matter of degree. This is not very surprising, owing to the pragmatic considerations associated with linguistic or cultural differences. In fact, rather than arbitrarily classifying faithfulness as falling into one camp or another (e.g., formal correspondence vs functional equivalence), we might also think of how the ambiguity of the norm allows translators to make interventions. When the Chinese translators justify solutions such as omission in opposition to literalism, the intrinsic ambiguity in the norm can account for functional appropriateness, expanding the concept to include the translator’s fidelity to communicative effects. In this sense, norms themselves are not something absolutely standard but are open to being interpreted by human agents.

5.1.4Answer: “The translation should sound ‘formal’”

Making a translation sound ‘formal’ is observed as another norm, evident in the translators’ stylistic preferences for neutralizing colloquial expressions:

This argument sounds like Toury’s (2012)Toury, Gideon 2012 Descriptive Translation Studies – and Beyond (Benjamins Translation Library 100). Revised edition. Amsterdam: John Benjamins. “law of standardization.” In Toury’s definition, standardization exerts a stronger influence when the text is translated into a target culture where the translation (or translation in general) is of relatively minor importance (see Pym 2008aPym, Anthony 2008a “On Toury’s Laws of How Translators Translate.” In Pym, Shlesinger, and Simeoni 2008, 311–328.). But do the Chinese translators really think the status of their translations is peripheral in the target culture? I suggest that the standardizing is possibly due to the “rhetorical salience threshold” postulated by Chesterman (2007b)Chesterman, Andrew 2007b “The Unbearable Lightness of English Words.” In Text, Processes, and Corpora: Research Inspired by Sonja Tirkkonen-Condit, edited by Riitta Jääskeläinen, Tiina Puurtinen, and Hilkka Stotesbury, 231–241. Joensuu: Joensuun yliopistopaino.: rhetorical adjustments are needed in order to “tone down” or “tone up” the target text to avoid undesired rhetorical effects perceived by the translator.

In the same vein, target-language proficiency is given weight, in the sense of the translation having to be “well-written English” (IN7F49). The translation company manager, reflecting on the feedback received from her government clients, argues that translators must pay careful attention to stylistic register:

IN7F49 suggested that the requirement for language proficiency was a sign of the government’s target-side awareness; in other words, their awareness of the responses of the international readers:

The translators’ understanding of ‘being formal’ offers some interesting asides. One interviewee suggested that she would want Chinese political leaders’ translated speeches to sound more like Barack Obama than Donald Trump. The translator hence tended to omit traces of exoticism and to avoid ‘vulgar’ expressions, for fear that the target-language readers might not be able to cope or may not like the style.

On the one hand, the Chinese translators tend to ‘internationalize’ their outputs. The attempt to standardize the text is in line with the features of institutional translation identified by Schäffner, Tcaciuc, and Tesseur (2014Schäffner, Christina, Luciana Sabina Tcaciuc, and Wine Tesseur 2014 “Translation Practices in Political Institutions: A Comparison of National, Supranational, and Non-governmental Organizations.” Perspectives 22 (4): 493–510. , 494) as “collective, anonymous and standardized.” However, the demand for a formal (stylistically more polished) target text may not truly be an expectancy norm in Chesterman’s sense. The peculiar emphasis on stylistic preferences suggests that a given translation is presumed to have the same effect on all target-language readers: it is assumed that everybody agrees that the more formal the English the political leaders speak, the more respectable they will appear. But how valid is this assumption for all target cultures? Will American, British and Australian readers trust the Chinese premier less if he sounds ‘colloquial’? Since there is little empirical evidence for such a claim, attempts to level out the register might be rather futile. It seems that the Chinese translators are primarily governed by start-culture norms, while the expectations of the target-language readers hardly enter the picture.

5.1.5Answer: “Different foreign-affairs offices do things differently”

It seems that start-culture norms, and consequently the degrees of flexibility that the translators are granted vary significantly among different institutions. IN5F32 was working as a senior translator in the same local government’s foreign-affairs office that IN7F49 and her translation company were working with. During the interview, IN7F49 reiterated that the local foreign-affairs office paid serious attention to the readability and proficiency of their translations. In a later interview with IN5F32, I had the opportunity to ask whether a target-oriented attitude could lead to new norms for translating China’s political discourse:

IN5F32 explained that she was part of a group of translators at the foreign-affairs office who had a similar educational background and agreed “on the nature of translating.” This team was responsible for doing the majority of the in-house translation work and for proofreading large translation projects such as the annual government reports outsourced to translation companies:

However, IN5F32 suggested that, to her knowledge, different foreign-affairs offices had different norm priorities. Apart from the profiles of the translators, interpersonal factors at the management level also affect norms and sometimes might exert a decisive influence. In the case of the local office, the emphasis on communicative effects was attributed to the heads of the local government:

Rather than any actual demand made by the target-language readership, the translators’ motivation for coping with stylistic weaknesses is somehow determined by internal criticism on the start side. Recall Chesterman’s (2006aChesterman, Andrew 2006a “A Note on Norms and Evidence.” In Translation and Interpreting – Training and Research, edited by Jorma Tommola and Yves Gambier, 13–19. Turku: University of Turku., 15) argument: “Norms are not personal.” However, the hierarchical relation of a norm authority implies that a higher-ranking individual’s personal preference can heavily influence a general translation practice. Here, pragmatism replaces dogmatic defenses of fidelity to the original as a new norm, thanks to the influence of an individual in management. Personal opinions feed into norms, reflecting them and affecting them. In this sense, norms can indeed be “personal” to a degree, and we see how norms are negotiated between agents: the translators can educate their clients (or vice versa); one might even suggest that more enlightened clients contribute to better translations and new norms (see Chesterman 1999Chesterman, Andrew 1999 “The Empirical Status of Prescriptivism.” Folia Translatologica 6: 9–19.).

This may also partly explain why the language quality of Chinese institutional translations varies between agencies: although the expectations of the foreign reader probably remain the same, the organizing codes of the institutional settings are different, and thus norms vary.

5.2.1Answer: “It depends on where you want to use the translation”

It would be naïve to suggest that institutional translators have no awareness of reception at all. A target-oriented approach starts to appear when the Chinese translators take a utilitarian view of translation. One frequent question I was asked during these interviews was, “Where do you want to use this translation?” This implies that the translators have a mental image of possible prospective readers, stimulating a forward-looking orientation with respect to possible reception conditions.

This functionalist awareness might motivate the translators to adopt different solutions rather than to stay close to the start text. Hermans (1998)Hermans, Theo 1998 “Some Concluding Comments on the Debates and the Responses.” Current Issues in Language and Society 5 (1–2): 135–142. suggests that the translation system distinguishes between a valid or a non-valid representation of the start text. However, what regulates the Chinese translators here is not whether a faithful representation of the start text is possible, but whether it is necessary according to the translators’ perceptions of the target side. IN7F49 said that she would change the register of the original to suit the target conventions. For instance, she would translate the English phrase best wishes as 顺颂商祺 shun song shang qi, a classical Chinese idiom meaning ‘best wishes to someone’s business’ – “Our Chinese people are more formal,” the translator added. This kind of “minor change” (IN7F49) in the register was justified in order to create similar contextual effects on the part of the target-side addressees, minimizing any potential misunderstanding. In another example, IN7F49 argued that when encountering cultural differences, translators should take readers’ ‘feelings’ into consideration:

It seems that the professionals indeed have an awareness of different readerships and their very different expectations about what a translation should be like. However, the issue of directionality may affect the translators’ predictions of communicative effects: both examples here were translated into the translator’s L1. There is little evidence that the translators would have the same confidence to outright adapt the original when the direction changes, even when they are aware of the importance of reception.

Now, if it is true that most translators more or less attach importance to the target side, who are the receivers that they are thinking about?

5.2.2Answer: “There are two sorts [of readers]”

In dealing with political jargon, FG2F35 roughly categorized the intended readers into two groups: the general target public and the readers who “already have some basic knowledge” of Chinese political discourse:

This is a simplified, idealized classification of two possible target-culture readerships: for “a general public,” literalism is less acceptable than if the translation is designed for experts (e.g., China watchers). Hence, explicitation seems to be considered as an efficient, cost-effective solution to offset the negative impact of literal translation: more explicit information may break contextual constraints in a general sense.

Let’s put the experts aside for a moment and ask what is considered a “general public”?

5.2.3Answer: “It doesn’t matter whether the reader is male or female”

When asked whether the translators had any socio-demographic profiles of the general readership in mind, most translators were confused and hesitated to answer:

The gender question was designed to elicit whether the translators were aware that the target readership is not homogeneous and could have different expectations. Not very surprisingly, it seems that the internal diversity of the readership is not a major concern to the Chinese translators. Nevertheless, an interesting point is that, for the translators, the readers should bear the responsibility to have contextual knowledge to understand the translation. This can be linked to Schleiermacher’s two mutually exclusive translation methods. Schleiermacher argued that translators may either domesticate the foreign text or retain its foreignness by “leave[ing] the writer in peace, as much as possible, and mov[ing] the reader toward him” (translated in Lefevere 1992aLefevere, André ed. 1992a Translation, History, Culture: A Sourcebook. London: Routledge., 49). Schleiermacher clearly favors the second method, as he regards foreignization as a tool to educate and build up German culture. Make the reader work! In the eyes of the Chinese translators, if the receiver is considered as a kind of “stakeholder” (Pöchhacker 2016Pöchhacker, Franz 2016 Introducing Interpreting Studies. Second edition. London: Routledge. , 178) in the communicative event, some basic motivation to comprehend the start culture is presupposed.

For the Chinese translators, the real or at least more accessible readers are not an abstract, imaginary general public but a visible group whom the translators can confidently recognize and effectively address. When asked, “Who are the important receivers for Chinese political discourse?” expert readers come to the surface:

In the end, a general readership may be no more than an illusion, since there has been no real attempt to identify their needs. The translators seem to be neither interested nor motivated to walk across the minefield: the general readers’ expectations of various kinds are deemed to be too complicated to be met by one translation, making the communicative effect difficult, if not impossible, to gauge.

5.2.4Answer: “It depends on who I am going to work for”

Another way of classifying the readership is into internal and external modalities. This leads to the distinction between start-language and target-language readers, which challenges the idea that an assumed ‘target reader’ is always from ‘the target culture’. That is, the addressees of the translated Chinese political discourse are not only external (e.g., target-language-speaking readers) but also internal (e.g., the Chinese government clients, professionals, and reviewers). Of these two, the internal readership sometimes overrides the translators’ attempts to conform to target-culture norms. For example, FG4F26 argues that the foreign reader who is directly addressed is not the most important receiver, let alone the only one – how her employer will receive and judge her translations is what ultimately counts, rather than any external readership per se:

For the translators, since their Chinese employer would usually give more weight to formal equivalence than to pragmatic considerations, any special effort beyond that internal expectation is usually deemed unnecessary, since little appreciable gain will result.

5.2.5Answer: “‘Sounding local’ is not always positive”

What happens if even the translators realize that their external readers might not fully understand the translation because of foreign elements? A striking response is that “it doesn’t matter.” Some translators argue that “sounding local,” in the sense of naturalizing the start language to target conventions, should not be a universal norm:

One explanation for the refusal to naturalize to target conventions lurks at the roots of institutional translation: losing start-culture identity is a serious risk. For example, one translator points out that there is no reason to naturalize the Chinese leader’s words to make him sound like his Western counterparts:

It would be unjustified to see the emphasis on constructing a national identity or retaining an original ideology merely as assuming cultural superiority. For the Chinese translators, since each culture has its personality and uniqueness, keeping those foreign references in place may help to attract interest, which also resonates with the agenda of the state’s international communication policy – to export Chinese culture to a wider readership. Seen in this light, the Chinese translators ostensibly have good reason to claim that target-culture traditions are secondary and that the start culture’s own voices are primary; in fact, this priority leads to the optimistic and ambitious view that translations can have effects on a wider target culture via “individual cognitive, emotional or aesthetic reactions” (Chesterman 2002Chesterman, Andrew 2002 “Semiotic Modalities in Translation Causality.” Across Languages and Cultures 3 (2): 145–158. , 151).

More than a few scholars might applaud this position for giving more power to translation and the profession (see Chesterman and Baker 2008Chesterman, Andrew, and Mona Baker 2008 “Ethics of Renarration: An Interview with Mona Baker.” Cultus 1 (1): 10–33.; Dam and Zethsen 2009Dam, Helle V., and Karen Korning Zethsen eds. 2009 Translation Studies: Focus on the Translator. Thematic section of Hermes: Journal of Language and Communication Studies 42: 7–166.). Nevertheless, if there is no empirical evidence to show that the exposure to others as others is well-received by the target-side readers, this argument would probably be mere wishful thinking.

5.2.6Answer: “Nobody would read it”

One of my assumptions was corroborated by the interviewees: not every institutional translation has or needs to have a real reader. When evaluating Task 3, FG3F41 argued that M3−INT seems to be a typical, official translation published on the Chinese government website. As an answer to the question, “Why are there so many literal translations around?”, the lack of real readership provides a causal explanation. The translators further suggest that the high number of translations published on the government website underscores that “nobody will read it” (FG3F41). Rather than address the need for any real target reader or real use, sometimes the institutional translations are translated for the sake of being translated, functioning as symbolic rather than pragmatic practices:

Apart from the possible lack of a real readership, the nature of institutional translation makes it hard to give the target culture priority over authoritative identity:

This makes some sense. For institutional translators, it is not always the target-language reader’s views that prevail or even matter; the authorities cannot be wrong. As Koskinen (2000Koskinen, Kaisa 2000 “Institutional Illusions: Translating in the EU Commission.” The Translator 6 (1): 49–65. , 51) ironically remarks on the European Commission translation services, “[s]ometimes the primary function of the translation of a particular official document is simply to be there, to exist.” The institutional component seems to be particularly marked in this “existential equivalence” (ibid.); it raises the question of whether the apparently poor-quality literalist translations can really be separated, not just from conceptual debate about the virtues of foreignization, but also from the conspicuous absence of any real target-culture demand. If the external, target-language reader does not matter, whether the translation should adhere to target conventions is a moot point.

5.2.7Answer: “Interpreting is different”

Different translation modes seem to affect the ways the prevailing norm is validated. Before evaluating the tasks, many participants asked whether the texts were written translations or transcripts of (conference) interpreting. These interviewees self-reported that they work in both modes. Interestingly enough, the professionals clearly distinguished between what ‘interpreting’ pursues and what a ‘written translation’ is for:

In their view, the norms that regulate interpreting are different from those governing written translations. In this (over)simplified dichotomy, a communicative function is important for interpreting, but written translation should be rendered as closely to the original as possible.

There are at least three reasons why different norm priorities are given to written translation and interpreting. First, the reliance on linguistic faithfulness could be loosened in interpreting settings, given that the time constraints on working memory make a full representation less possible and it allows limited capacity for many explanations (see Gile 2009Gile, Daniel 2009 Basic Concepts and Models for Interpreter and Translator Training (Benjamins Translation Library 8). Revised edition. Amsterdam: John Benjamins. , especially 42).

Second, in most cases, the conference interpreter’s audience is more visible (remote interpreting is an exception). Compared to their translator counterparts, interpreters tend to have a clearer picture of who their ‘here and now’ receivers could be and what they would expect:

That is, the audience attending an interpreted event is more likely to be a uniform group, in the sense of being motivated by similar purposes. The identifiable profile of the audience resonates with what Pöchhacker (1995Pöchhacker, Franz 1995 “Simultaneous Interpreting: A Functionalist Perspective.” Hermes: Journal of Linguistics 14: 31–53., 49) calls the “cultural common ground” that is shared by the participants in (conference) interpreting settings. In terms of the pre-familiarity with the subject matter of the communicative event, the interpreting audience is by nature more homogeneous; there are fewer cultural barriers actually separating them. This might make it easier for interpreters to adapt the message to the receivers’ comprehension levels and expectations.

Third, some clues may be found in the pedagogical ideology of interpreting. More than a few interviewees used the term 脱去语言外壳 tuo qu yu yan wai ke ‘deverbalization’ to justify their preference for freer renderings in interpreting transcripts. For them, a literal, somewhat unnatural written translation is taken as normal, while an interpreting performance is expected to be fluent and must make sense. The emphasis on decoding rather than transcoding harks back to Seleskovitch, who advocated that conveying intended meaning (or vouloir dire) rather than linguistic features should be the main component of the interpreter’s work (cited in Pöchhacker 2016Pöchhacker, Franz 2016 Introducing Interpreting Studies. Second edition. London: Routledge. , 59). For trained interpreters, communicative success is determined by target-side response. Thus, in order to produce acceptable target-language utterances, some deviation from linguistic equivalence is inevitable, and “filtering” to enhance the communicative impact of the text is necessary (Gile 1992Gile, Daniel 1992 “Basic Theoretical Components in Interpreter and Translator Training.” In Teaching Translation and Interpreting: Training Talent and Experience. Papers from the First Language International Conference, Elsinore, Denmark, 1991, edited by Cay Dollerup and Anne Loddegaard, 185–194. Amsterdam: John Benjamins. , 189). In this sense, interpreters tend to give priority to the needs of their audience rather than to the strict form of the original. Here we see that trainers (and textbooks, perhaps) are indeed powerful norm authorities, as they not only issue norms but also provide professionals with a metalanguage to justify their solutions in their real-life work.

However, a prescriptive idealization of target-oriented norms is not rare in translator training, especially within Skopos theory. So why do interpreters still appear to be relatively free from the constraints of the start text? Here the causality of norms works from the other direction: the professionals consciously or unconsciously choose the norms in terms of how well they are compatible with their interests, in other words, legitimizing the translators’ solutions and reducing the actual risks involved.

5.3.1Answer: “Trust me, there are many worse start texts than this”

When do translators willingly intervene in the start text? One immediate answer is, “when the original is bad.” The Chinese translators do not shirk from criticizing the quality of the Chinese start text. ST₃ describes Chinese premier Li insisting on wearing two football scarves, one for each of the opposing Australian football teams. The Chinese text indicates that the premier “said with humor” when explaining his purpose that, even though the weather was hot, he wanted to show his support for both sides by still wearing two scarves. When asked which translation made them think that the premier had a better sense of humor, none of the participants even thought that the start-text joke was funny. Some translators pointed out that the start text was poorly written. The political joke was not successful even in the original, leaving translators little scope to render it literally:

Aside from politicians’ occasionally unfunny jokes, as IN7F49 pointed out, Chinese political discourse is conventionally imbued with linguistic and semantic ambiguity, making its meaning sometimes hard to grasp. IN7F49 said that she was familiar with this dilemma:

The axiomatic superiority of the start text is here shaken by an awareness of its own deficient quality, leaving room for the translator to justify the ‘necessity’ of their norm-breaking.

5.3.2Answer: “The premier just wanted to amuse his audience”

A second parameter is the intended function of the start text. When reflecting on how to translate premier Li’s greeting to a full-stadium Australian audience, the translators believe that some freedom of choice should be allowed. The underlying idea is that, since image building is the main impetus for this cultural diplomacy, the communicative effect should have more weight:

For the translators, breaking with start-text faithfulness sometimes has a problem-solving function if there is no symmetry between linguistic forms. If an adaptation is deemed compatible with the communicative intention of the start text, the attempt to intervene is generally seen as acceptable.

Nevertheless, it appears to be unreasonable to speak about great freedom in highly sensitive texts. The translators tend to justify intervention in terms of cultural and linguistic relativity, basically when they are aware that a given stylistic feature (e.g., tautology) may have different effects on the readers in another culture. However, once a given formal feature is linked to the start-side ideological agenda, intervention suddenly becomes a ‘no-go’ zone. When asked whether the translation of a Chinese leader’s speech was allowed to be adapted to local cultural conventions, IN7F49 said that if the speech was about Communist ideology, the answer should be ‘no’:

For IN7F49, risks have a hierarchy of sorts: for a lower-risk culture-specific topic, communicative effects have priority over literalism; for translations of ideology-bound content, the virtues of linguistic loyalty are strongly highlighted, shunning any attempts to intervene.

This observation brings us back to the dichotomy of foreignization and domestication. Target-language readers are presumed to be fascinated at seeing a new culture, but any attempt to highlight a foreign ideology could scare them away. In this sense, neither foreignization nor domestication is the prevailing norm; they are both legitimate risk-management strategies, and their status largely depends on how many rewards might be granted by the most important receiver perceived (e.g., the institutional employers in this case). For the translators, if there are no substantial rewards, then there is little reason to take risks.

5.3.3Answer: “I was asked to translate the title as ‘Governor’ rather than ‘Secretary of the Provincial Party Committee’”

Another impetus for breaking norms could lie in national image-building as one purpose of the translation. This too is functional. When literalism is equated with negative effects on the national image, the translators are more confident about legitimizing intervention:

As mentioned, the translators make a conscious effort to faithfully and truly retain the original ideology in high-risk texts, like a good mirror. On the other hand, in some cases, the norm authority seems to grant translators some freedom to play down ideological differences for the sake of a better national image. When evaluating M3−INT, many translators disliked the literal ‘Chinglish’ translation because it made the Chinese premier sound aloof:

FG3F42 suggested that the local government she worked with somehow did not want to remind foreign investors that they were doing business with a country with a strikingly different ideology. FG3F42’s government client, fearing that the interpreting audience might react negatively to ideological alterity, gave her a clear (but oral) instruction to “weaken the Communist Party’s color” by not interpreting the leaders’ Communist Party titles when introducing them to foreign investors:

This is somewhat paradoxical, though, as it seems that the authority makes a distinction between pragmatic expediency and a lasting written record: they are aware that ideological alterity may have a negative impact on the target-culture side, so when engaging with the ‘on site’ investors, they eliminate or minimize the ideological element for immediate use, whereas the Communist values are left unchanged in written translations, where future checks are hard to control.

5.3.4Answer: “I definitely won’t rewrite the text”

We might safely conclude that intervention takes place when translating Chinese political discourse, albeit in a varied and limited sense. But do the translators appreciate being granted creative license? Not much, it seems. To talk about what affects the translators’ motivation to break norms is to consider where the limits of translator intervention lie. I suggest that the translator’s primary concern is the translation-specific risk of losing credibility (see Pym 2008bPym, Anthony 2008b “On Omission in Simultaneous Interpreting: Risk Analysis of a Hidden Effort.” In Efforts and Models in Interpreting and Translation Research: A Tribute to Daniel Gile (Benjamins Translation Library 80), edited by Gyde Hansen, Andrew Chesterman, and Heidrun Gerzymisch-Arbogast, 83–105. Amsterdam: John Benjamins., 2015Pym, Anthony 2015 “Translating as Risk Management.” Journal of Pragmatics 85: 67–80. ). This can be interpreted in two senses: professional and personal.

The first kind of risk, professional credibility, concerns all legitimate matters pertaining to a prescriptive statement of what a ‘translation’ is. Some translators refuse high-level intervention since they think the result would no longer be a ‘translation’ but a ‘rewriting’:

This high intervention lays itself open to the translator’s criticism, as it seems to extend the boundaries of ‘what-is-a-translation’ too far. So what could go wrong if a text is translated too freely, or too creatively? Along this path, we find the traditional focus of loss and betrayal: translated texts are secondary and the originals are primary. The opposition to intervention in the rewriting sense also seems to mean that the translators want to trim the edges of a translation, frame it and give it neat and clear borders, so that nothing except original messages remains within its margins.

Contemporary scholars in Translation Studies have proposed various ways to define translation in terms of different epistemological stances (see Chesterman and Arrojo 2000Chesterman, Andrew, and Rosemary Arrojo 2000 “Shared Ground in Translation Studies.” Target 12 (1): 151–160. ). The notion of equivalence is pervasive, whereas Lefevere argues that translation is indeed a kind of rewriting, highlighting translation’s similarity with anthologizing, paraphrasing and summarizing (see Lefevere 1992bLefevere, André 1992b Translation, Rewriting and the Manipulation of Literary Fame. London: Routledge.). The Chinese case seems to suggest that a more conservative view is prevalent among practitioners. The notion of what Newmark (1988Newmark, Peter 1988 A Textbook of Translation. Hemel Hempstead: Prentice Hall., 204) called the “authoritative text” that “[translators] have no right to improve” is still fairly influential.

Apart from fidelity to the start text, what else could be sacrificed when translator intervention goes too far? For FG1F34, the answer is “our collective voice”:

The behavior of institutional translator is constrained and regulated. This resonates with the definition of institutional translation offered by Schäffner, Tcaciuc, and Tesseur (2014Schäffner, Christina, Luciana Sabina Tcaciuc, and Wine Tesseur 2014 “Translation Practices in Political Institutions: A Comparison of National, Supranational, and Non-governmental Organizations.” Perspectives 22 (4): 493–510. , 494) as a “collective [and] anonymous” activity. In the case of institutional settings, translators’ individual styles are both practically unnecessary and politically risky. Any high-degree intervention is perceived to be unrewarding, rendering further efforts to make translations more creative simply not worth the time.

5.3.5Answer: “Literalism is a safer choice for staff translators”

The second kind of risk concerns the translator’s personal credibility. An example is omission. As might be imagined, since anyone on the start-culture side could compare the original with the translation and note the absence, the translators have to make an extra effort to justify the necessity of their intervention:

Given that the potential risks of losing the start-side trust are more palpable, low intervention is usually perceived by the translators as the ‘safest’ choice. This recalls Toury’s (2012Toury, Gideon 2012 Descriptive Translation Studies – and Beyond (Benjamins Translation Library 100). Revised edition. Amsterdam: John Benjamins. , 285) formulation of “environmental feedback”: the feedback the translators receive is not only from the target recipients but also from those on the start culture. This kind of feedback is “normative in its very essence” (Toury 1998Toury, Gideon 1998 “A Handful of Paragraphs on ‘Translation’ and ‘Norms’.” Current Issues in Language and Society 5 (1–2): 10–32. , 26). The elaborate result of such a personalized social and cultural experience is the translator’s habitus (see Simeoni 1998Simeoni, Daniel 1998 “The Pivotal Status of the Translator’s Habitus.” Target 10 (1): 1–39. ). The norms that are embodied explicitly or implicitly prescribe what is supposed to be done and what is ‘improper’ behavior that should be avoided. In other words, the decisive factor for breaking or adhering to certain norms is the translator’s habitual risk management. The translator adheres to a prevailing norm because that is what others typically do, and hence it is a safe choice; the motivation underlying a translator’s choice of norm-breaking derives a real risk (or a reward) perceived in a given situation.