An examination of ideology in translation via modality: Wild Swans and Mao’s Last Dancer

1. Introduction

It is often said that the Anglophone book markets are resistant to translated works, as reflected by the well-known Three Percent Problem: only around 3% of books published in the United Kingdom and USA are translations (Donahaye 2012, 6; Ban 2015, 160). Such resistance may have encouraged attempts by some migrant writers to write directly in English. Since the late 1980s, there has been a number of successful English works by Chinese migrant writers (Cheng 2010; Ha 2000; Qiu 2003; Li 2003/2009; Chang 1991/2003). Amongst them, Wild Swans, by Chinese-British writer Jung Chang (1991/2003), is at the summit of success, having sold 13 million copies worldwide and been translated into 37 languages (Chang 2016). In Australia, Li’s (2003/2009) book has become a best-selling and award-winning book: with a young reader’s edition recommended in the Australian school curriculum, it is a nationwide household name.

Despite the phenomenal success and their cross-cultural nature, these two books can hardly be considered as tokens of Sino–Western friendship. Rather, they may be indicative of ongoing conflicts of ideologies. Both books have been banned in Mainland China due to their explicit criticism of the Chinese government. This ban extends to the translations: although both books have been translated “back” to Chinese (Chang 1992; Li 2009), they have only been permitted for publication outside of Mainland China.

To date, the Chinese translations have received little scholarly attention, even though these books are highly valued (Halliday 1988, ix) and politically controversial. A previous study (Li 2017) of Wild Swans found that Mao has been represented dramatically less and as less agentive in the Chinese translation. This paper further investigates the shifts of political ideology in the Chinese translation of Wild Swans, with a new focus on the interpersonal metafunction within systemic functional linguistics (SFL) (Halliday and Mattiessen 2014). As judgements are often considered to be expressed through choices of modality within the interpersonal metafunction, an examination of modality has the potential to reveal ideological shifts in the translation.

The overall objective of this paper, therefore, is to explore the influence of political ideology in translation through investigating potential regularities in the translation shifts of modality. It proposes a quantitative–qualitative approach to reveal the ensemble of semantic consequences of these linguistic choices on the broad front (quantitative results) and the localised ideological shifts on the narrow front (quantitative/qualitative results).

To achieve this objective, three research questions have been raised:

1. What are the typological differences that may affect translation studies? Are there even commensurable systems?

2. Can patterned shifts be identified in the comparable domains or systems of meaning in each translation? How do the two translations compare?

3. Do the regularities in such shifts have ideological consequences? In particular, are the regularities non-random, in the sense that may seem to be directed, whether unconsciously or deliberately?

This study therefore sets out to contribute to a better understanding of the link between modality and ideology in translation. A subsidiary goal is a more delicate comparison of modality across two major world languages – English and Chinese – from the perspective of SFL. Although modality has been extensively studied in both English (cf., Lyons 1977; Butler 1988; Matthiessen 1995; Biber et al. 1999; Krug 2000; Palmer 2001; Huddleston and Pullum 2002; Leech 2003) and Chinese (cf., Chao 1968; Zhu 1996; Lu 1942/1952/1956/1982/2002; Ma 2005), descriptive linguistics cannot be mechanically applied to translation studies without the bridge of contrastive linguistics. However, existing typological descriptions often lack the delicacy to handle specific translation issues, because language is “inexhaustible” in its system potential (Halliday 1994, xiii).

2. Modality and its relations to mood in English and in Chinese: an SFL perspective

3. Data and methodology

3.2. Methodology

After the chosen chapters were scanned from hard copies and converted into Word documents, the texts were divided into clause complexes and clauses, before being imported into a computational tool named SysFan (Wu 2000, 111–194; Yu and Canzhong 2016, 7).² As shown in Figure 1, SysFan allows the analysis of modality in combination with speech function: the systems of mood, polarity, modality type, value, manifestation, and orientation can be manually selected for each free major clause. Once all clauses were analysed, SysFan would generate quantitative results. Clauses containing modality were exported to Excel for a closer qualitative analysis. In addition, SysFan was used to visualise the distribution of modality.

Figure 1.

System network for analysing mood and modality in SysFan.

4. Results and discussions

4.2. Ideological shifts through the translation of modality on the narrow front

Quantitative and qualitative analyses have been carried out within each of the four modality types to further investigate the localised ideological shifts in translation. Considerable ideological shifts have been found in the translation of Wild Swans, through choices of probability, usuality, and obligation. Therefore, inclination will not be discussed in this paper due to lack of space. In addition, as detailed analysis of the translation of Mao’s Last Dancer has found no clear patterns of overall ideological shifts, only the translation of obligation in Mao’s Last Dancer will be discussed, as examples of how shifts of two opposite trends may result in a dynamic balance and an overall non-shift of ideology.

4.2.1. Probability

As previously mentioned, the instances of probability have decreased in both translations, which is foregrounded against the overall increase of modality in both translations. This decrease has resulted in higher certainty of propositions in Chinese, because polarity should be considered as more certain than high probability.

Probability in Wild Swans has decreased dramatically by 38%: from 29 in ST (15 median/11 low/3 high) to 18 in TT (7 median/8 low/3 high). Out of the 29, only 11 have been translated into Chinese modality of the same type and value, and another 11 have been translated into polarity. Propositions, therefore, become more certain in Chinese.

Because the overall trend of probability is a decrease from ST to TT, any notable additions of probability in Chinese should be considered as a foregrounded pattern. For instance, Figure 2 shows a cluster of clauses in the Chinese translation, where two cases of added probability have been identified in adjacent sentences. In the same cluster, one also finds two cases of the foregrounded shifts of orientation, from “objective” to “subjective,” a case of elevated value from low “could” to high “kěndìng huì,” and an omission of the word “probably.” This is a crux where shifts of various subsystems of probability have converged, and it has been visualised in SysFan in Figure 2.

Figure 2.

Visualisation of the distribution of modality in SysFan: Zhou.

(Left: English ST; right: Chinese TT; Lighter shade: lower modality; darker shades: higher modality)

This cluster is related rhetorically to Premier Zhou Enlai, who is depicted as a mediating character: he is positively evaluated for protecting people, including Chang’s father, but negatively evaluated for never having openly challenged Mao. Examples 1–4 show a clear ideological shift, that Zhou is evaluated less negatively in the Chinese translation through choices of probability.

Example 1

ST: But perhaps he had acted … in order to prevent an even more horrendous disaster.

Modality: perhaps: objective/implicit/low.

TT: 但是我想，他也可能是在盡量防止更大、更可怕的災難

Back Translation (BT): But I think, he also possibly be avoiding bigger, more horrendous disaster.

Modality: wǒ xiǎng: subjective/explicit/median; kěnéng (in bound clause): objective/implicit/low.

The primary probability in TT is “wǒ xiǎng” (I think), an interpersonal metaphor. The TT steps away from being an objective historian; instead, it presents this evaluation as a personal view of Chang. This makes her sympathy for Zhou explicit.

Example 2

ST: Like a civil war, [[which an open challenge to Mao could have brought on]]

Modality: could (in downranked clause): subjective/implicit/low.

TT: 比如公開對毛澤東挑戰後||肯定會產生的大規模內戰。

BT: For instance, after openly challenging Mao, certainly will cause large-scale civil war.

Modality: kěndìng huì (in downranked clause): objective/implicit/high.

Although these two examples are not in free clauses and therefore do not form part of the quantitative results, they clearly show an ideological shift. The author shows understanding of Zhou’s dilemma: there would be the possibility of a Chinese civil war should Zhou have openly challenged Mao. As this possibility of war is presented as much higher in the TT, it presents Zhou as a more sensible leader. These shifts suggest more sympathy for Zhou in Chinese.

Example 3

ST: But probably also saved the country from total collapse.

Modality: probably: objective/implicit/median.

TT: 但也使國家避免了完全崩潰。

BT: But also enable the country [to] avoid total collapse.

Without modality, Zhou is given definite credit in Chinese for having saved China from a total collapse.

The examples above show a closer alignment of the character “I” in the book with Zhou Enlai and a more favourable evaluation of his contributions in the Chinese translation. The more positive evaluation of Zhou in Chinese echoes the general perception of him as a well-respected figure in Mainland China, despite the fact that the translation is published in Taiwan.

In short, shifts in probability have led to a more favourable evaluation of Zhou in the Chinese translation of Wild Swans. In comparison, the translation of modality in Mao’s Last Dancer is considerably more faithful. Both translations have elevated certainty of propositions due to the decrease of probability.

4.2.2. Usuality

Quantitatively, expressions of usuality in the TT (14: 6 median/8 high) appear to be rather similar to those in the ST (13: 6 median/7 high). However, a closer examination finds several subtle shifts. Out of the 13, 8 have been translated into the same modality type and value. The rest comprise five decreasing shifts and five increasing shifts.

Decreasing shifts	Increasing shifts
Translating high usuality “always” into polarity Translating “always” into median usuality “píngshí” (usually) Translating high usuality “never” into polarity “gēnběn bù” (not at all) Translating “always” into a verb “ài” (love + doing sth)	Translating polarity into high usuality “zǒng” (always) Translating the adjective “numerous” into median usuality “shíbùshí” (occasionally) Translating median inclination “would” to high “zǒng” Translating median “frequently” (in bound) to high “lǎo” (always)

There are two shifts in usuality which clearly reveal a changed evaluation of represented participants:

Example 4. Mother’s fight for a good memorial speech for father’s funeral.

ST: This was the traditional time for a Chinese family || to engage in a bit of emotional blackmail.

Modality: “the traditional time for x to do”: objective/explicit/median usuality.

TT:死亡對中國家庭來說，可能是唯一的機會，可爭取到[[非此得不到的]]東西。

BT: Death for Chinese families, can be only opportunity, can strive [to] achieve things [that are otherwise unobtainable] [kěnéng and kě].

Modality: “kěnéng” and “kě”: both subjective/implicit/low probability.

The mother’s intention to threaten the authorities for an acceptable speech is presented in the ST as “usual,” normal, and expected, with the choice of median, explicit, and objective usuality, hinting that this chance to fight is not uncommon following someone’s death in China. With such normality, the mother can be expected to succeed. However, in the TT, it is expressed with low, implicit probability “kěnéng” and “kě,” which implies more obstacles and a lower chance of success. In turn, having won unexpectedly, the mother must have had to put in more effort. This contributes to a more agentive and resolute image of the mother in Chinese.

Example 5. Madam Mao (Mao Zedong’s wife, who is 21 years junior)

ST: Since she was frequently seen together with … || people said she had taken them as “male concubines.”

Modality: “Frequently”: objective/implicit/median [in bound clause].

TT: 從宣傳媒體裡，大家 老 看見||她和…

BT: From propaganda media, people always saw [that] she [was] with….) [lǎo]

Madame Mao was thus blamed for causing China’s sexual repression, on the one hand, and openly taking “male concubines” on the other hand. The Chinese expression “lǎo” expresses high usuality, and it clearly exceeds the median value expressed through the English “frequently,” which could have been rendered as “jīngcháng,” “chángcháng,” or “cháng.” A higher frequency of Madame Mao being seen with handsome young men is highly charged with a negative evaluation of her promiscuity and hypocrisy.

In fact, this shift in usuality is supplemented with a shift beyond modality: an added direct speech of Madame Mao in Chinese, which expresses her approval for women to take male concubines. This higher usuality in the TT and the added direct speech result in a more negative evaluation of Madame Mao in Chinese. The harsher criticism echoes the general perception of her as a notorious leader in Mainland China, even though the translation is published in Taiwan.

In summary, a number of subtle shifts have occurred in the translation of usuality in Wild Swans, although the overall statistics are similar. Some cases in usuality have led to a more negative evaluation of Madame Mao and a more resolute image of the mother. In comparison, no expressions of usuality are used in the ST or TT of Mao’s Last Dancer.

4.2.3. Obligation

Amongst the four modality types, obligation shows the greatest shifts in both translations, with significantly more obligation in both TTs, in particular high obligation.

4.2.3.1. Obligation in wild swans

Obligation in Wild Swans stands out as the most dramatic shift, with an increase from 31 in ST to 51 in TT. High obligation has more than doubled from 22 to 45 instances. Out of the 31 English expressions of obligation, 21 have been translated into a Chinese obligation of the same value. Another six instances of English obligation have been translated into Chinese obligation with higher value. Other factors that contribute to this increase include

• translating one English clause with obligation into two Chinese clauses with shared obligation;

• translating English bound clauses with obligation into Chinese free clauses with obligation; and

• translating English clauses without modality into Chinese clauses with obligation.

The increases in the instances, percentage, and value of obligation clearly have ideological consequences: they reconstruct an inner context that imposes more and harsher restrictions, which may serve to either express stronger resentment and frustration towards the lack of freedom and justice, or implicitly portray certain characters as more resolute, resourceful, and brave for battling against harsher obstacles. Some significant ideological shifts via choices of obligation include as follows:

1. Restrictions on women in Chinese society

Example 6

ST: We women students were supposed to keep our hands under the table, and [were supposed to] sit motionless.

Modality: Supposed to x 2: objective/implicit/median obligation.

TT: 我們女學生得把手放在桌子下面，[得] 坐得絲紋不動 [dei]

BT: We women students have to have hands placed under the table, [have to] sit motionless.

Modality: dei: objective/implicit/high obligation.

Dei expresses a stronger value of obligation than “supposed to”: it implies stronger necessity for an adherence to such rules. As this obligation was directly imposed on the author herself, who was one of the Chinese women, the shift from median-to-high obligation can be taken to express stronger personal resentment towards restrictions on women in the Chinese society. Several similar cases can be found in the same cluster of clauses, as shown in Figure 3.

Figure 3.

Visualisation of the distribution of modality in SysFan: Restrictions on Women (Lighter shade: lower modality; darker shades: higher modality).

Resources to express either high or median obligation are abundantly available in both English and Chinese, as shown in Tables 1 and 2. However, whereas the entire cluster of English modality items is of median value, all in the corresponding cluster in TT are of high value. The author ‘I’ clearly expresses stronger interpersonal judgements in Chinese on an issue that concerns herself: women’s role in the society.

• (2) Restrictions on “me” imposed by the teachers

The TT also expresses stronger resentment towards the teachers who forbid “me” from boarding a foreign ship for a dinner banquet, as shown in the explicitation of obligation in the two examples below:

Example 7

ST: (a) my teachers said that || no one was allowed on board a foreign ship.

(b) I was told to say || I was busy that evening.

Modality: No modality in free clause.

TT: 但是，老師們不准我去 [bù zhǔn]

他們要我去撒個謊，[yào]

BT: But, the teachers did not allow me (to) go.

They want(ed) me go tell a lie.

Modality: Both bù zhǔn and yào express explicit/objective/high obligation, with the teachers as the explicit source.

In the ST, there is no obligation in the free clause: the high obligation “neg + allowed” is in a bound clause and is used in passive voice, which conceals the source of obligation: the rule may not have been set by the teacher. In addition, “no one” blurs the target of the rule. In contrast, the source of restriction is explicated in Chinese as the teachers, and such restrictions are imposed directly on “me.” A stronger resentment towards the teachers is thus detected in the TT.

1. Restrictions on mother

Example 8

ST: She wanted to concentrate on….

Modality: Subjective/implicit/median inclination.

TT: 她得集中精力…[dei] Increasing

BT: She had to concentrate energy….

Modality: dei: objective/implicit/high obligation.

This examples echoes Example 4, in that both Chinese examples imply a context that is harsher and more restrictive for the mother, which then suggests a more resolute and resourceful image of her for having won a tougher battle.

In short, the instances and percentage of obligation, especially high obligation, are considerably higher in Chinese. This has resulted in the reconstruction of the external restrictions as harsher and more challenging in the Chinese version, and a stronger personal resentment towards the societal expectations.

4.2.3.2. Obligation in Mao’s Last Dancer

Contrary to the case of Wild Swans, the shifts in obligation in Mao’s Last Dancer show no consistencies of ideological shifts in the evaluation of the Chinese government officials. One reason is the dynamic balance brought by two translation trends that tend to counteract the ideological consequences, as discussed in the following:

Trend one: from imperative or polarity to obligation

It comes as little surprise that shifts between imperative mood and indicative mood with obligation are found in translation, as both realise the speech function of proposal/command. However, being the incongruent realisation, proposal realised through modality tends to be more polite. Many commands come from the direct speech of the Chinese consulate staff; and two cases have been found, in which the Chinese opt for a more polite form, instead of faithfully rendering the imperatives; as shown in Examples 9 and 10:

Examples 9 and 10

ST	TT	Shift in obligation
You follow Chinese laws (Consul Zhang)	必須遵守中國法律 (you) must obey Chinese law	Imperative (marked) > modality (politer)
Have faith in the party (vice-consuls general)	要相信黨 (you) must believe (in the) party	Imperative > modality (politer)

There is an additional example of a shift from polarity to obligation. Similarly, even high obligation presents more potential for negotiation than polarity, as in Example 11:

Example 11

ST	TT	Shift in obligation
Everybody listens to the party (vice-consuls general)	每個人都得聽黨的 Everybody all has to listen (to the) party	Polarity > modality (politer)

Based on Examples 9–11, one may assume that the Chinese consulate staff appear softer in Chinese, by allowing more negotiation. However, three other shifts contradict such image.

Trend two: from obligation to imperative or polarity

Examples 12–14

ST	TT	Shift in obligation
You simply have to listen to [[what we say]] (vice-consuls general)	你老實聽我們的話吧 You obediently listen to us + “ba: imperative mood tag”	Modality > imperative (unmarked theme) The interpersonal tag “ba” somewhat softens the harsh imperative “you obediently listen….”
It’s not for you to ask me questions	現在不是||你問我問題 Now neg. + be you ask me question	Modality > polarity (more strident)
Could’ve killed that bastard!	真該殺了這混蛋！ Really should kill + aspect: complete this bastard	Low inclination [ability] > median obligation (higher value/more strident)

Examples 12–14 present the Chinese consulate officers as being more strident and blunt, allowing less negotiation, which counterbalances the opposite effect brought by Examples 9–11. Consequently, no significant shift in the image of the Chinese consulate officers can be concluded.

In short, whereas both translations have shown more obligation, especially high obligation, clear ideological shifts have been identified in the translation of Wild Swans, which help construe an external environment with harsher restrictions on the Chang household. In Mao’s Last Dancer, however, no clear patterns of overall ideological shifts have been identified.

5. Concluding discussion of the influence of ideology on the choices of modality in translation

An examination of the translations of modality in Wild Swans and Mao’s Last Dancer has found more similar shifts than differences. In terms of similarities, both Chinese translations use more instances of modality than do their respective parallel STs; and this is particularly true for high modality and obligation. In addition, both translations use less instances of probability. These suggest a preference in the Chinese language for more strident expressions of modality, which present prepositions as more certain, and emphasise more the speech function of modulation and circumscription.

In terms of differences, however, translation shifts in Wild Swans are much more dramatic than those in Mao’s Last Dancer. A considerably lower percentage of English modality items in Wild Swans have been translated into expressions of Chinese modality of the same type and value. In addition, greater shifts have been found within each modality type in the translation of Wild Swans.

Furthermore, shifts in modality in Wild Swans frequently reflect shifts in the ideological stance. Through meaningful choices of usuality, the author/translator present harsher criticism of Madame Mao, who is an “accomplice” of the Gang of Four. This confirms previous findings (Li 2017) that the Gang of Four is represented as being more agentive, and hence more culpable, in Chinese. Considering the Chinese translation is published in Taiwan, the harsh criticism of Madame Mao and favourable evaluation of Premier Zhou Enlai set up an interesting question as to why the adjusted evaluation of Zhou and the Gang of Four in Chinese appears closer to the general opinions of them in Mainland China.

The English ST is primarily interested in revealing the absolute agency and culpability of Mao Zedong. To highlight his agency, the thinking capacities and agency of other “accomplices,” such as the Gang of Four, are somewhat blurred. The Chinese translation, however, shows less interest in singling out Mao. Instead, more people are represented as having thinking capacities and agency, and thus as being held accountable for the catastrophic Cultural Revolution: for example, the Gang of Four, other central and provincial party leaders, and the collective destructive power of the youth – the Red Guards and general students. The Chinese version can be considered a somewhat more balanced account than the English book, as the ST appears to be an oversimplification of Chinese history with a disproportionate contrast between the agency of Mao and the passivity of all others.

Wild Swans has been criticised for its interpretation of Chinese history (Lin 1992, 121–128; Gao 2002, 419–434). Such dramatic shifts found in both the ideational and interpersonal metafunctions in the Chinese translation may well be an attempt by the translator/editor, and even the ST author, to adjust a certain over-interpretation of history and bias in the STs, as the Chinese translation is intended to address readers who potentially possess higher levels of knowledge about the inner context of the book.

Another previous finding is that the Chinese translation is not only less interpretive of history but also a more personalised account. This has been supported in this study, in that stronger resentment has been expressed in Chinese when it comes to restrictions that would personally apply to the author.

In contrast, the translations of modality in Mao’s Last Dancer not only are more faithful but also show no notable ideological shifts. Although the increase in the instances and value of modality may have resulted in the Chinese translation being slightly more dramatic and strident than the English book, this may be largely influenced by stylistic preferences and does not affect the evaluation of represented participants. The lack of ideological shifts may be because the ideological stance of the book was not considered controversial or problematic for the Taiwanese readers: the chosen chapter is a recollection of closed-door personal experience, which is hard to verify or challenge.

Another difference between the two translations is tenor. Wild Swans was translated by the author’s brother, Zhang Pu. As a represented character who shares many of his sister’s experiences, he has more power than ordinary translators. In addition, as Chang was well educated in China and had completed a university degree by the time she left China at the age of 26, her Chinese proficiency is believed to be high enough to potentially be actively contributing to the translation decision-making; in this case, Chang may potentially have allowed or even proposed translation shifts that an ordinary translator would not dare to consciously introduce. In comparison, although the translator of Mao’s Last Dancer was based in the same city with Li, had interviewed Li, and consulted him for advice, he was not personally related to Li. In this sense, he is like an ordinary translator. In addition, as Li was only a young graduate of a dance academy when he left China, his Chinese literary proficiency may not be high enough to be thoroughly involved with the translation; in this case, the translator may opt for safety by staying close to the ST. These factors in tenor provide potential explanations for the relatively faithful translation of modality in Mao’s Last Dancer.

6. Concluding remarks

This paper has carried out a systematic comparison of modality in the Chinese translations of two politically controversial English works. The significance of this paper is fivefold. First, this study makes an original contribution to categorising Chinese modality in the subsystems of value and orientation; and the lists of modality items in English and Chinese will hopefully be useful for future translation studies and contrastive linguistics. Second, it contributes to a better understanding of the link between modality and ideology in translation: shifts in the value, orientation, and manifestation in probability, usuality, and obligation can significantly alter the interpersonal evaluation of and the author’s alignment with represented participants. Third, it has highlighted the importance of a balance between a quantitative and qualitative approach to the study of modality: quantitative alone may disregard potential dynamic balance underneath the surface; and qualitative alone may appear unsystematic. Lastly, by exemplifying the complex translation shifts of modality in various subsystems, it will hopefully raise awareness of different dimensions of equivalence in modality in translation teaching and practice.

A limitation of this study is that the quantitative approach needs to be manually adopted and cannot be applied automatically to a large corpus within a concordance tool. Some major challenges need to be addressed before the proposed lists of modality items can be applied to larger corpora: for instance, the potential of a modal auxiliary to express multiple types, interpersonal metaphor and the plethora of variants of Chinese modality.