The problem of translating Chinese policy-related expressions: a case study of wenming (‘civilised’)

3.1 Wenming-related research

Various sinologists have explored wenming from socio-political perspectives, without, however, considering translational issues surrounding this expression (see e.g., Boutonnet 2011; Nguyen 2012). Another relevant area of research includes sociological explorations of wenming as a concept denoting civil behaviour (e.g., De Seta 2018). Such research on civility has also not considered translational issues.

In research published in Chinese, one can witness a surge of interest in wenming, including studies on translational issues relating to this expression. For example, Deng et al. (2014) compared wenming as a noun with the English word civilisation, by arguing that these two expressions are radically different. He (2019) argued that although wenming as a polysyllabic word existed in Classical Chinese texts before this expression was borrowed from the West through Japanese, the Classical Chinese wenming expression should be strictly distinguished from its modern Chinese counterpart. Liang (2020) argued that both proper nominal and common nominal translations of wenming and its singular and plural uses should be distinguished from one another in translation.

Along with such examples of critical empirical research, several more essentialist studies should also be mentioned here. Sun (2006) made some sweeping overgeneralisations about wenming by relying on dictionary definitions of the nominal wenming and its English counterpart civilisation. Liu (2010) conducted essentially ideologically-driven research on wenming, by assuming an East–West dichotomy. He (2007) approached wenming primarily as a concept describing etiquette and using it in a moralising manner.

In summary, a significant amount of research has been dedicated to wenming and its importance in the Chinese linguaculture. However, no multimethod, bottom-up and corpus-based approach which underpins the present study has been attempted.

3.2 Translation: theoretical background

Since we are studying the problems associated with the translation of linguaculturally embedded expressions, it is worth revisiting the classic writings of Jakobson (1959) and Nida (1964; see also Nida and Taber 1969). As Jakobson (1959: 234) famously argued,

All cognitive experience and its classification is conveyable in any existing language. Whenever there is deficiency, terminology may be qualified and amplified by loanwords, neologisms or semantic shifts, and finally by circumlocutions.

Notwithstanding this ‘law of universal translatability’, we should nevertheless not forget that certain real translatability limits do exist (see House 1973, 1997, 2018; see also Schäffner 2004, 2010 and Pym 2006 regarding the translation of political texts). Jakobson recognises this problem by explicitly referring to ‘all cognitive experience’, i.e., the possibility that translation will be severely restricted if we take connotations into account (see also Chafe’s 2000 notion of ‘shadow meanings’). Connotations defy explicit definitions: they even vary in one individual’s mind as a person’s moods and experiences change. Also, connotations cannot be delimited from denotative meanings. So, connotative meanings are far too elusive to be captured in translation.

Translatability is also limited whenever the form of a linguistic unit takes on a special importance. Connotations and the special importance of the form of a linguistic unit relate to the translational issue that we will study in this paper by considering the case of wenming. We must qualify Jakobson’s dictum of universal translatability, as Nida and Taber (1969: 4) have done:

Anything that can be said in one language can be said in another, unless the form is an essential element of the message.

This is why translation-related research like our study benefits from departing from an essentially contrastive pragmatic approach, the latter relying on the dictionary meanings of the expressions being studied (see also House 2015).

Another classic work discussing the limits of translatability is Catford (1965: 93–94):

Source language texts and items are more or less translatable, rather than absolutely translatable and untranslatable. For translational equivalence to occur, both source language and target language text must be relatable to the functionally relevant features of the situation. A decision, in any particular case, as to what is functionally relevant must … remain to some extent a matter of opinion.

This means that when judging the appropriacy of any translation there will always be a remnant of subjectivity. House (1977, 1997, 2015 tried to minimise this subjectivity by providing a systematic framework of categories to assess the quality of a certain translation. Figure 1 displays the essence of this framework. Our repertoire of translational terms is rooted in this model.

Figure 1:

House’s (2015, 2018 system of comparative text analysis and evaluation.

As Figure 1 shows, the model regards translation as an act of recontextualisation in a new situational context. The ‘context of situation’ (Malinowski 1935) is opened up through the register categories of ‘Field’, Tenor’ and ‘Mode’, as in the Hallidayan tradition (see e.g., Halliday and Hasan 1985). While we do not use all the fine details of this model, it provides an important component of our own framework (see Section 4) because it allows us to examine why the translations of certain uses of wenming into English turn out to be problematic. The corpus part of House’s model in particular – indicated by the lower right-hand box and arrow in Figure 1 – precludes speculation about idiosyncratic translations, a point which is also important in our framework.

In the process of recontextualisation, ‘overt’ and ‘covert’ translations need to be emphasised in particular (see also House 2018). Briefly, an overt translation is a translation in which the original text is left intact as far as possible – for example when wenming is translated as ‘civilised’ in any context – whereas a covert translation takes into account the needs of a new foreign audience, with the translator making changes to the original text by applying a ‘cultural filter’. We pay particular attention to whether wenming collocations are overt or covert in scope.

4.1 Data

Our methodology (see Section 4.2) consists of three phases, and we rely on multiple corpora in our research. The corpora used during the first phase of our study include the Balanced Chinese Corpus (BCC; http://corpus.zhonghuayuwen.org/CnCindex.aspx) and the British National Corpus (BNC). We agree with Sharoff et al. (2013) that no corpus is perfect nor representative. In our case, this imperfection includes differences in the size of the two corpora, as well as certain differences in the genres featured in them. For example, BCC includes a larger percentage of literary texts than BNC. The results of our Chinese and English corpus searches include 200 valid examples per linguaculture.

The corpora used in the second phase of our research involves Chinese–English bilingual news outlets featuring the expression wenming in Chinese, and its English translations. We only included reports on domestic Chinese events to ensure that the bilingual news featured ‘native’ uses of wenming and their English translations. We drew our corpus from the following two news outlets:

1. China Daily: https://language.chinadaily.com.cn/

2. Global Times: https://www.huanqiu.com/; https://www.globaltimes.cn

We chose these sources rather than larger corpora of Chinese newspapers as the latter would not permit the study of translations. Table 1 summarises the frequency of occurrence of the expression wenming in our corpus of source media texts.

In the third phase of our research, we rely on data drawn from a translational task. Here, we asked the participants to translate occurrences of wenming in utterances drawn from the BCC. This task was followed by post-hoc interviews with the participants. The overall length of these interviews was approximately 100 min. We removed all personal information from the dataset, and followed all the standard ethical procedures, including asking for the consent of our participants.

4.2 Methodology

We focus on the adjectival uses of wenming and compare them with those of the English adjective civilised. In Chinese, an adjective can either modify a noun directly, implying a strong relationship between the modifying adjective and the noun, or with the aid of the particle de 的, indicating a weaker adjective–noun relationship. We only examine wenming–noun phrases because we are assuming that the context of policymaking implies a strong relationship between wenming and the noun it modifies. Indeed, as our corpus-based analysis will illustrate, a typical policy-related expression is wenming-shenghuo 文明生活 (lit. ‘civilised life’, i.e., ‘progressive lifestyle’), while wenming-de-shenghuo 文明的生活 (‘progressive-DE-lifestyle’) is not used in policy communication.

Our approach is tri-partite:

1. We first conduct an individual corpus search of the adjectival uses of wenming and civilised, followed by a contrastive pragmatic analysis of these uses and collocations. The corpora that we use in this phase allow us to consider the broader context of the uses and collocations of civilised and wenming. We sampled a representative set of 200 examples in each language. We chose sets of 200 examples because our previous experience showed that it is important to select an optimal data size like this in contrastive pragmatic analysis in order to make sure that the results are both representative and manageable (see House and Kádár 2021). This investigation allows us to empirically work out contrastive criteria of analysis. For example, our present study shows that the uses of civilised and wenming can be differentiated according to the following criteria:

   1. whether they relate to the behaviour of an individual or a group;

   2. whether they describe a state already achieved or a state to be achieved.

   We use such criteria as focal points in annotating the uses and collocations of civilised and wenming in our corpora. In the current study, our teams involved both Chinese and Western members, and as part of our bottom-up procedure our Chinese team worked exclusively on the Chinese corpus and our Western team on the English corpus, without initially discussing the outcomes of the annotation. Our aim was not to create hard and fast annotation categories: rather, we wanted to arrive at an empirically grounded set of categories that would be useful and manageable in the ensuing contrastive analysis of the uses and collocations of wenming and civilised. The two teams then convened to harmonise the annotation. For example, the Chinese team annotated certain uses of wenming as ‘polite’, whereas the Western team labelled the comparable annotation category as ‘well-mannered’. We ultimately decided on the annotation category ‘well-mannered’, simply because ‘politeness’ has a specific academic meaning. We also harmonised the number of such categories, by keeping their quantity below the threshold of five in each linguaculture, and by merging less frequent categories into more frequent ones. The ultimate goal of this contrastive research is to identify problem sources for translation.

2. During the second phase, we study how wenming is rendered in the English version of Chinese news outlets. We first categorise wenming-translations in a corpus of bilingual media texts according to the wenming annotation categories obtained during the first phase. We then examine whether wenming really turns out to be problematic to translate into English, by relying on House’s (2015) model of translation (Figure 1). In other words, we consider whether wenming is translated overtly or covertly in our corpus, and we analyse overt solutions with no cultural filtering through the lens of register concepts, in particular Tenor. While judging what is ‘problematic’ from a translational point of view can never be completely ‘objective’ (see House 1997), the fact that our team consisted of two native speakers of Chinese and two ‘Westerners’ helped us gain among the team members in identifying cases which may trigger translation problems. This procedure also helped us diagnosing the overt versus covert nature of a translation.

3. The third phase of our research consists of a translation task and follow-up interviews, conducted in China with a panel of 10 expert translators. The aim of this follow-up approach is to triangulate our findings (cf. House and Kádár 2021): i.e., we test whether the translational tendencies (cf. overt versus covert translation) in phase two can also be observed in the translational solutions provided by the translators involved. This is an intriguing question because previous experience has shown that Chinese public media sources follow preset conventions when it comes to communicating and translating policy-related expressions (see Kádár and Zhang 2019). Thus, here we aim to find out whether translational tendencies can be interconnected with the Mode involved (see Figure 1). The task we provide for our participants consist of short examples drawn from our Chinese corpus (cf. phase one), featuring wenming uses according to the categories identified in phase one. Following the test, we interviewed our participants regarding their translational choices.

Figure 2 provides a brief summary of our tripartite approach:

Figure 2:

Our methodological procedure.

5.1 Phase one: corpus-based examination of civilised and wenming

Here we first examine the use of civilised and wenming in our English and Chinese corpora, and then conduct a contrastive analysis.

5.1.3 Contrastive analysis

Table 6 summarises the differences and similarities between uses of civilised and wenming:

Table 6:

Similarities and differences between the uses of civilised and we nming.

Table 6 is divided into two different parts: Part 1 considers uses that exist in both linguacultures, while Part 2 considers uses existing in only one of the two linguacultures.

Wenming and civilised have two comparable uses (Part 1 of Table 6), with the first one being the most frequently employed in both linguacultures. While in this use both wenming and civilised refer to the state of groups, it is only wenming which refers to a state to be achieved, a fact which stems from its policy-related nature. The second comparable use of wenming is the least frequently employed one in the Chinese corpus (see Table 4). This use is listed second in Table 6 because, to a certain degree, it is comparable to that of civilised. While in this case wenming still refers to groups, whereas civilised refers to individuals, these uses are essentially comparable because both describe a state that has already been achieved. The examination of such uses of wenming has revealed that this category is the only non-policy-related one in Chinese, and this is why it does not relate to a future state of affairs. In summary, Part 1 of Table 6 consists of two comparable uses of wenming and civilised, with the first one being the most frequently employed in both linguacultures. However, these uses differ significantly and, as such, might lead to translational difficulties.

The uses detailed in Part 2 of Table 6 might imply different translational difficulties than those in Part 1 because they have no linguacultural equivalent. At this point, we hypothesise that this lack of linguacultural equivalence might trigger a preference for covert translation and cultural filtering in comparison with the first use in Part 1.

In the following discussion, we will only consider the first use in Part 1 and those uses of wenming in Part 2 for which there are no equivalent uses of civilised, i.e., the first two lines in Part 2. These uses of wenming are all group-related and indicate states to be achieved, i.e., they are typically policy-related expressions.

5.2 Phase two: the English translations of wenming

Table 7 provides a summary of the wenming translations in our Chinese–English corpus of media texts:

Table 7:

Translations of wenming.

Table 7 has been aligned with Table 6: we divide the translational categories into two parts, as outlined above. We highlight those uses which, according to the research outcomes of phase one, may be particularly problematic for translators. As Table 7 shows, by far the most frequent translational solution in our corpus of Chinese–English translations includes instances of overt translation and the consequent lack of cultural filtering: in 158 cases, representing 83.6% of the translations, wenming is translated according to its ‘civilised’ dictionary meaning. This tendency can be explained by referring to the category of Tenor in House’s (2015, 2018 translation model (Figure 1): since wenming tends to be used in communicating governmental policies and is reported as such in Chinese news, both the source texts and translated texts are characterised by a social relationship between an authoritative power (the government) and its citizens (cf. Kádár and Zhang 2019).

In the following, we first examine translations of the ‘Cultured, developed and well-mannered’ use of wenming, representing Part 1 in Table 7. We then examine the other two uses of wenming in Part 2.

5.3 Phase three: translation task and follow-up interviews

During the third phase of our research, we asked 10 professional Chinese translators to translate eight utterances including wenming drawn from the BCC. We chose eight examples because phase one had revealed that wenming affords four different uses, i.e., we provided two examples for each use. Our goal was to test whether translational tendencies (in particular, overt versus covert translation) can also be observed in our participants’ translations. We were particularly interested in finding out whether the fact that the task was unofficial – allowing ‘private’ solutions – influenced translational preferences. We chose BCC rather than texts drawn from media because we wanted to see whether general uses of wenming in seemingly ‘simple’ utterances triggered any difficulties for the translators involved. Table 8 shows our partcipants’ translational choices:

As Table 8 shows, while the translators occasionally applied a cultural filter, e.g., by translating wenming in its ‘Cultured, developed and well-mannered’ use as “model” and “harmonious”, they still mostly chose the overt translation “civilised”. This outcome surprised our team because our partcipants were expert translators. One can explain this preference by referring to the category of Tenor: despite the private nature of the task, the usually formal character of texts featuring wenming and the close connection between wenming and policy communication turned out to be dominating our translators’ choices. Due to space limitations, here we provide only one example for overt and another one for covert translations of the use ‘Cultured, developed and well-mannered’:

The following are excerpts from our interviews with the translators of the above examples:

Translator 2 (overt translation of example 27)

“这个十大文明路口” … 我就觉得既然人家都这么使用了呢(.)，我就也这么使了。

Regarding this “top ten civilised intersections” … since others use this form (.) all the time, I opted for it as well.

Translator 10 (covert translation of example 28)

我觉得外国人会不理解文明指的是什么。

I feel that foreigners may not understand the meaning of wenming in this context.

As the second excerpt shows, the translators who applied covert solutions and cultural filtering usually referred to the fact that translating wenming as ‘civilised’ in domestic Chinese policies may sound alien to foreigners. For us, this was a particularly important outcome: while in the case of our previous corpus of bilingual texts we did not want to speculate about the reasons why a small number of translators chose covert translation, the translators involved in the translation task revealed awareness of the difficulty foreign readers might have in understanding the overt translation of wenming was responsible for their choices.

Some translators revealed their awareness of the complex meanings of wenming, arguing that even they themselves did not always fully understand all of such meaning:

Translator 9

其实我也不太明白 “星级文明户” 到底指的是什么。

In fact, I myself do not understand exactly what “star-level wenming household” describes.