Paternalism in China Daily ’s coverage of Chinese Muslims (2001–2015)

The corpus for analysis

The data for this study come from a corpus of 747 news articles about Chinese Muslims in CD, in which all the news was published from 11 September 2001 to the end of 2015. The 9/11 attack was taken as a starting point since the coverage about Muslims increased dramatically after the attack (e.g. Baker et al., 2013). The end of 2015 was the cut point of data collection since conflicts in Xinjiang dramatically decreased after 2016 (BBC, 2016). From this socio-political period between 2001 and 2015, our project reveals how the government and over 20 million Chinese Muslims are portrayed. The texts were first collated to compile two news archives by keyword searching of ‘Chinese Muslim(s)’, ‘Uighur(s)’, ‘Uyghur(s)’, ‘Ethnic Hui’ and so forth. ⁴ All the word items are usually used to describe Chinese Muslims. After the extraction, the news pieces irrelevant to the research were manually deleted.

A CD corpus of 485,327 words was compiled. In order to have a comprehensive picture of how paternalism is represented in the news discourse, news texts of all genres were included in the data. More than 95% of the chosen texts were news reports, which can appear in the columns of Business, Culture and Government in CD. News coverage of other genres (e.g. editorial, opinions) account for 5% of the data, and it is made explicit by the newspaper per se. About 67.17% out of 747 pieces of news in the CD originally come from CD. CD also represents views from other official newspapers in China, such as Xinhua and Beijing Weekly.

Data analysis

The data analysis was divided into three stages. The data were initially investigated manually before the use of corpus analysis. Corpus analysis was conducted to signify topics pertaining to governmental paternalism, and this enquiry addresses the question of what topics in the CD elicit governmental paternalism. Results from corpus analysis were further examined by qualitative analysis, construing its linguistic realisation in Chinese media discourse about Chinese Muslims.

Initially, 30 news texts in the CD were randomly selected and read by the analysts, which ensures an overview of pronounced topics in the data suggestive of governmental paternalism in the media discourse. For instance, in a news article titled Chinese Muslims head for Mecca, 13,800 Chinese Muslims were helped by the Chinese government to participate in Mecca pilgrimage in 2012. The subsidence from the Chinese government to the religious life of its fellow Muslim citizens exhibits paternalistic characteristics.

Topic modelling is the primary corpus analytical method in the analysis, which aims at identifying the topics in the CD corpus exhibiting governmental paternalism (i.e. paternalistic discourse), such as subsidence from the government towards Muslims’ pilgrimage (e.g. ‘state-organised visits to Mecca for the annual hajj pilgrimage’). Ideologies are inherent in thematic structures in discourse, that is, topics in discourse (Van Dijk, 2009). Topic modelling can discover topics in discourse in an inductive manner, which allows for the retrieval of hidden topics in corpus otherwise overlooked (Törnberg and Törnberg, 2016b) and express them with topic indicators (see Appendix 1). This study used LDA (i.e. Latent Dirichlet Allocation) to run topic modelling, which is a commonly used algorithm within this method. We used GUI for Mallet’s implementation of LDA, an online topic modelling tool. ⁵ Two parameters were important in the implementation. First, in line with Jockers (2013) and the length of most articles in the CD, each document in the analysis lies within the word limit of 500–1000 words. Extremely long and short texts were, respectively, cut and combined. For instance, a 1100-word text was divided into two documents of 1000 words and 100 words. The 100 words were then combined into another document with significantly less than 500 words. Second, in terms of LDA, the number of topics needed to be set before any analysis was applied using either statistics or researchers’ judgement. We followed Törnberg and Törnberg (2016a) who suggest experimenting with different topic numbers to find the most interpretable results. It was found that setting the number at 22 would be optimal for generating the most meaningful interpretation as to what topics were marked in the corpus. Setting a different number of topics, we could find multiple meanings in one topic. The automatically generated topics were unnamed and thus were labelled by the researchers. We examined the top 20 high-frequency indicators in each topic to decide whether the topic could reliably be indicative of governmental paternalism. The examination of the number of indicators lightly varying at either side of the figure of 20 did not significantly impact the results.

Having identified topics and their significance in the CD relevant to governmental paternalism, we then analysed the ideological construction by argumentative strategies. Ten discourse sections that include topic indicators were selected randomly for each topic related to the ideology, and 90 discourse sections were examined qualitatively. Such investigations were conducted to examine topoi in paternalistic discourse. They would link the argument together to address the ideological qualities of the content in the CD.

Topoi are prevalent in media or political discourses for persuasion, as have been evidenced in Wodak (2011) and Wright and Brookes (2019). The use of topoi can present positive self-presentation, negative-other presentation, revealing differences/similarities, warning of dangers and discrediting opponents (Wodak et al., 2009). For the identification of topoi, we followed the taxonomy developed in Wodak (2001, 2009) that provides a helpful interpretive framework, for example, the topos of advantage and the topos of responsibility. Notably, we found another topos – the topos of happiness recurrent in our data but no appearance in Wodak’s taxonomy. The topos, as is explored in Allen (2014), expresses a positive mood (e.g. satisfaction and pleasure) to accomplish persuasion. This scheme is overlapping with the topos of advantage since both argumentative strategies stress positivity to ingroups for a convincing conclusion (e.g. the implementation of an initiative). However, the topos of happiness focuses on positive emotions of people, while benefits to individuals (e.g. the betterment of social security) are stressed with the topos of advantage. Therefore, we explored the argumentative schemes that formulate governmental paternalism in the corpus.

Subjectivity in the analysis was reduced but not excluded. Researchers have independently read through the news articles to identify paternalistic discourse in the CD, specified the number of topics and topic names and identified the types of topoi in paternalistic discourse. Homogeneous results were re-examined until a consensus was achieved. Topoi are presented in the discourse in flexible linguistic patterns, so the naming and identification process cannot avoid subjectivity. To address this issue, typical topoi we observed were elaborated with examples.

Paternalism in the CD

Governmental paternalism towards Chinese Muslims is elicited in 10 topics that are grouped as four topic categories; a topic category comprises topics with similar meaning (Figure 1). Statistically, the contribution value of these topic categories is 0.45 out of 1, which indicates the importance of these topics in the CD corpus at the discourse level and a high degree of discursive paternalism in CD’s coverage of Chinese Muslims. In addition to these topic categories, there are 5 topic categories and 13 topics in the CD, such as community affairs that illustrate life stories of Chinese Muslims.

OPEN IN VIEWER

Figure 1.

Topic categories in the CD.

Chinese governmental paternalism is primarily revealed through the topic category of social transformation that emphasises Beijing’s subsidisation to improve Chinese Muslim society. The newspaper highlights the improvement in living standard (e.g. ‘home’, ‘work’), economy (‘economy’, ‘investment’), education (‘school’, ‘teach’) and healthcare system (‘public’, ‘health’) with the overall contribution value of these topics in corpus being 0.21 (⩾0.05 ⁶ ). Contribution value of a topic indicates its importance in the corpus at the discourse level, and thus 0.21 elucidates CD’s focus on social transformation in the coverage of Chinese Muslims, and on Chinese Muslims as recipients of the government’s help to improve their lives. An article of 2013 cites a top Chinese official: ‘Xinjiang should be established as a western port for China’s foreign trade’ (24 September). Economic support and achievements have the highest contribution value of 0.1 among the four topics in the CD, and such a focus in corpus agrees with Chinese national policies concerning Chinese Muslims – the Chinese government pays attention to the economic development in Xinjiang with the aim of promoting the modernisation of the region (Anand, 2018; Mackerras, 2014). The list of topic indicators is provided in Appendix 1.

The second topic category is Islam as a religion in the corpus. Beijing helps ethnic Muslims with religious freedom, such as the organisation of chartered flights or adjusting the market price of lamb during Ramadan. The topics of Muslim rituals and figures (e.g. ‘pilgrimage’ and ‘halal’) are included in the category, with the total contribution value of 0.12 (⩾0.05) (Figure 1). There are many examples in the text, for instance, a journalist wrote on 17 September 2014 that ‘the total number of Chinese pilgrims who go to Mecca will exceed 14,000 on a government-organised trip this year’ (also see Example 2). The topic salience in the corpus reflects Beijing’s respect for Islam and assistance to Chinese Muslims for the practice of Islam, thus reflecting governmental paternalism in such coverage. Also, as suggested from the topic category, CD helps Beijing to promote the idea that religious freedom should be implemented through government assistance (Kung, 2006). These topic categories reveal that, to a large extent, Chinese Muslims are projected as ingroup members to the Chinese government – Chinese citizens needing and enjoying governmental assistance. This is important because such identification is typical of the discursive traits pertaining to paternalistic ideology (cf. Le Grand and New, 2015).

The topic categories of terrorism/violence and reaction to terrorism/violence are inherent to Chinese governmental paternalism, whereby journalists set the boundary between most Chinese Muslims as an ingroup versus a few Chinese Muslims as an outgroup to the government. The ideology is reflected by addressing the vulnerability of citizens (Le Grand and New, 2015). More specifically, securitisation has been highlighted by the Chinese government in its paternalistic policies in Xinjiang, which emphasised the relative powerlessness of Chinese Muslim citizens, underlined the security challenges of Xinjiang, as well as justifying Beijing’s subsequent policies towards Chinese Muslims (Anand, 2018). As was revealed in topic modelling analysis, the topic category of violence/violence has the contribution value of 0.07 (⩾0.05) in the corpus. Correspondents cover terrorist acts inclusive of terms such as ‘terrorist’, ‘killed’ and ‘attack’, as well as reporting on consequences of these acts to the majority of Chinese Muslim citizens (e.g. ‘security’ and ‘injuries’). For instance, a scholar on Xinjiang issues denounces the slaughter of Chinese citizens in a news piece (China Daily, 2013). The foci on violent acts and their effects on most Chinese Muslims are portrayed as justifying Beijing’s protective initiative and reveal Beijing’s concerns about Chinese Muslim citizens, strongly indicating paternalistic approaches and the inclusion of the group within the national community. However, the reference to ‘terrorist’ indicates that parts of the community are politicised as an outgroup and are not considered a part of Chinese Muslims. Moreover, the category of reaction to attacks/violence (e.g. ‘police’, ‘crack’ and ‘down’) with a contribution value of 0.05 also indicates a high degree of paternalism in the news. As was shown from an official’s voice from Beijing ‘measures should be taken to prevent more terror attacks and guarantee people’s safety’ (also see Example 4), the journalist distinguishes between ‘people’ (i.e. Chinese Muslims) as an ingroup to the government and ‘terror’ – the juxtaposition of terrorism with religious identity – as an outgroup. There is both a legitimisation of Chinese Muslims as a conceived collective whole and isolation of its detracting elements – some Chinese Muslims participating in violence. The identification of an ingroup to Beijing and the distinction made for the ideological detractors indicate paternalism and provide, through discursive constructions, the need and obligation for the government to protect ‘good’ citizens.

Paternalism and topoi in the CD

Topoi in the CD were investigated for the linguistic construction of governmental paternalism in the media discourse towards Chinese Muslims. We probed into how argumentative strategies are used to construct and portray paternalism towards these citizens in the four topic categories previously highlighted. We also graded the relevance of these topoi to the formulation of ‘paternalism’ in the discourse. The contribution value of a topic category that entails the frequent occurrence of a specific topos indicates to what extent the topos helps with the expression of paternalism in the corpus (Figure 2).

OPEN IN VIEWER

Figure 2.

Grading the use of topoi for governmental paternalism in the CD.

The topos of advantage is an essential strategy in the CD corpus that portrays the majority of Chinese Muslims as recipients of Beijing’s help. The topos is typified by the following condition: if an advantageous outcome is generated, actions will be resumed. So, writers in discourse typically underpin the benefits to a social community to imply the necessity for the implementation of a practice (Wodak, 2001, 2009). In the CD, the topos of advantage is identified in most of the topics previously discussed, and the contribution value of these topics is 0.38 out of 1.04. Thus, the topos is used significantly to construct paternalism in the corpus (see Figure 2). With the topos, journalists highlight the government’s supportive intervention in the development of the Chinese Muslim community as well as the benefits given to its citizens (cf. Le Grand and New, 2015). Also, it reflects the prioritisation of needs towards the minorities that reflects the government’s ‘putting people first’ principle (Wang and Liu, 2018). The topos thus clearly contributes to building the Chinese government’s father-like status and paternalistic framing of responsibilities towards these citizens. The providing of benefits to the Chinese Muslims also indicates a need to intervene in the private lives of the ingroup and thus assumes the mantel of social guardianship. Therefore, with both the security aspect and the provision of care, we can say that paternalism is ideologically embedded in the CD’s representation. To illustrate, Example 1 shows Beijing’s supportive policies towards Chinese Muslims in the economic sphere.

Example 1

URUMQI – China’s far western Xinjiang Uygur autonomous region is becoming better connected with the international community after the implementation of the Silk Road Economic ⁷ Belt. (11 July 2015)

The topos of advantage is manifested by the narrative of Beijing’s launching of the Silk Road initiative and its goodwill to Chinese Muslims. As suggested by the topic indicators ‘economic’ and ‘the implementation of Silk Road Economic Belt’, Beijing has undertaken economic supporting initiatives for the social transformation in Xinjiang with the largest Chinese Muslim population. Notably, the supporting action is nominalised as ‘the implementation of Silk Road Economic Belt’. Fairclough (1989) claims that the use of nominalisation serves to leave the attribution of responsibility unclear. Through this expression, the government’s initiative in the supportive economic policy is backgrounded, and its interference in the regional economy is downplayed. Beijing is thus portrayed as a considerate father-like figure, assisting the regional economy with little interference in the autonomy of its citizens. The supportive economic measures modernise the local economy and aid local Chinese Muslim populations to broaden their capacity for interaction with the rest of the world since Xinjiang becomes ‘better connected with the international community’. Xinjiang has a large population of Muslim citizens, and it is known as a geographically remote region in Western China with insufficient communication with the outside world. So, it is imaginable that increasing socialisation meets the need for a large population of Muslim residents. Based on the analysis, the topos indicates the Chinese government’s paternalistic care to Chinese Muslims with ‘the Silk Road Economic Belt’, or the Belt Road Initiative. Therefore, such economic inclusion of Chinese Muslims reveals their identity as an ingroup to the Chinese government, which also indicates Chinese governmental paternalism in the sentence.

Governmental paternalism in China is shown by filial piety – citizens show respect and obedience to the central power (Wang and Liu, 2018). In the CD corpus, this feature is revealed by the use of the topos of happiness – satisfaction and appreciations from Chinese Muslims to the involvement of the Chinese government in their lives. The use of the topos indicates an alignment of Chinese Muslims with the government’s intervention policies so we can say that the CD’s exposition of filial piety legitimises governmental paternalism by showing citizen appreciations. The individual-level anecdotes also promote the use of pathos to create an emotional response from the reader. As suggested by Figure 2, the topo is salient in the topics of Beijing helping Chinese Muslims to improve their livelihoods, education, healthcare and religious freedom with the contribution value of 0.33 out of 1.04 in the CD. Given that these topics are inherent with the ideology, the contribution value indicates the significance of the topos to the construction of governmental paternalism in the data.

Seen from the following paternalistic discourse, journalists use voices from Chinese Muslims to demonstrate the appreciation for social development and the government’s help. Through the topic indicators of ‘apartment’ in Example 2, Chinese Muslims are delighted because the Chinese government subsidies families move into new apartments.

Example 2

The family of six had been living in a rented apartment that had barely enough space, Xirip said. She is happy to see local governments are now helping the family with two low-rent apartments. (5 February 2014)

Chinese Muslims all explicitly convey sentiments consistent with the topos of happiness in these examples – ‘she is happy’. The clear expression of happiness distinguishes most Muslim citizens from the small number who are engaged in violence (cf. Example 4).

In the above example, satisfaction is juxtaposed with the benefits of national policies towards Chinese Muslims. Contentment among Xirip and help from Beijing are obvious in Example 2. ‘Two low-rent apartments’ that are supported by the ‘local government’ and give satisfaction to the Muslim ‘family of six’. In a reported speech, the writer claims that ‘she (Xirip) is happy to see’ the betterment of the living standard in her family (i.e. the topos of advantage) that is achieved by governmental subsidisation to family’s living conditions and reveals her submissiveness to the government. The writer also makes it explicit that Beijing’s engagement addresses the needs of her family since they ‘had been living in a rented apartment that had barely enough space’. In this sense, filial piety from Xirip is related to the government’s prioritisation of people’s necessities. So, the appreciation of the ingroup justifies the government’s approaches by showing that the view of guardianship is positive, and intervention is mutually supported. The combined use of the topos of happiness and advantage here exemplifies filial piety (e.g. respect and submissiveness) from the Muslim citizens towards the government. Filial piety pertains to Chinese national policy of ‘putting people first’ – the government supports Chinese Muslims through addressing their needs and solving their difficulties in daily life. The Chinese government is thus portrayed as paternal to a largely obedient Chinese Muslim citizenry.

The topos of responsibility is frequently identified in CD’s coverage of Chinese Muslims to reveal the government’s responsibilities and obligations to assist and protect its citizens and identify citizens as an ingroup to the government. The topos of responsibility views that because a group or person is responsible for the way things are, that group or person should act to put things right (Wodak, 2001). In the corpus, this argumentative strategy is identified in the topics of social transformation and Beijing’s responses to terrorist attacks with the contribution value of 0.26 (out of 1.04) (see Figure 2). The paternalistic tradition from the Chinese government maintains the principle of parent officials (i.e. Fumu guan) – fulfilling the duty to take care of its citizens (Cheng et al., 2017; Ling, 1994). CD combines a dual approach of mutually inclusive support for transformative social policies and action on terror directed at the outgroup. This two-fold discursive strategy is the dominant approach of CD in legitimising paternalism.

Modal verbs are typical in the presentation of the topos of responsibility (see Wright and Brookes, 2019). The view is correspondent with Fairclough (2000) who believes that the use of modal verbs indicates ‘authoritarian language’ and suggests ‘moral toughness’ (p. 108). In paternalistic discourse here, we specifically investigated the modal verb ‘should’ in the conveyance of Beijing’s responsibility in assisting its Muslim citizens because of the higher frequency of ‘should’ (frequency = 709) than other modal verbs in the CD (e.g. ‘must’, frequency = 354; ‘need’, frequency = 480). The modal verb has both deontic and epistemic meaning, which respectively refers to obligations that ground demand in the subjectivity of the speaker and possibilities that intermediate the certainty and increase the objectivity of information or knowledge. In the corpus, the deontic ‘should’ is mostly related to the topos of responsibility. It conveys Beijing’s subjectivity and determination to get involved in Muslims’ lives in a supportive way, reinforces an ingroup identity of most Chinese Muslims and provides the opportunity structure for integration and collusion. For instance, in Example 3, the Chinese government is described as embracing the duty of curbing Chinese Muslim terrorism; a small number of Chinese Muslims participating in the Kashgar violence and the group is to be derided and no longer included in the Chinese Muslim ingroup.

Example 3

‘Strong and effective measures should be taken to prevent more terror attacks and guarantee people’s safety, their assets and regional stability’, he said.

‘(We should) resolutely punish terrorists according to the law, crackdown on terror activities, resolutely crack down on extreme religious forces and effectively contain illegal religious activities’. (2 August 2011)

In the voice of former Vice-Premier Zhang Gaoli, ‘crackdown’ makes the extract an anti-terror narrative. In the extract of anti-terrorism, he demonises the Muslims who have participated in the violence: ‘terrorists’, ‘extreme religious forces’ and conduct ‘illegal religious activities’. Specifically, the repeated labels of ‘extreme’ and ‘religious’ remind people of fundamentalism and global terrorism. The pejorative features portray the minority of Muslims as a peril to the majority of Chinese Muslims, that is, ‘people’, which subsidise the government in excluding the few outgroup members from majority ingroup members.

In the narrative of Beijing’s anti-terror activities, the topos of responsibility is first of all realised by two ‘shoulds’ whereby curbing terrorism is seen as an obligation for the Chinese government after violent attacks, and the objectified need is indicative of intervention. Despite the first ‘should’ appearing in the passive voice, the responsibility of taking ‘strong and effective measures’ against terrorism is directed from the Chinese government through Zhang’s voice. These measures are claimed to ‘guarantee people’s safety, their assets and regional stability’, thus beneficial to Chinese Muslims citizens in Xinjiang. Once again, the benefits presented in the discourse provide justification for the government’s interventionist policies. The next ‘should’ elaborates on the prevention of terrorism with strong measures, which reinforces Beijing’s father-like status. With the inclusive ‘we’ and ‘should’, the official treats it as the responsibility to protect its ingroup members from terrorism.

A list of three is used to strengthen Beijing’s duty to prevent terrorism after a series of attacks, as well as to level-up Muslims’ identity as an ingroup to the government. By definition, speakers in political speeches usually use the parataxis of at least three actions with similar patterns to reinforce the rhetorical force of delivery and increase persuasive qualities (Atkinson, 1989). The journalist in such an example juxtaposes ‘punish(ing) terrorists’, ‘crack(ing) down on terror activists’, ‘crack(ing) down on’ extremists and ‘contain(ing) illegal religious activities’. These four actions are listed to impress international readers and show that the Chinese government will bring lawless and evil terrorists to justice for safety and stability in Chinese Muslim areas.

The topos of reality is used in the CD to reveal governmental paternalism. According to Wodak (2011), the scheme is typically the documentation of terrible social situations in discourse (e.g. poverty) and serves to justify the need for specific actions or decisions. In the CD corpus, the topos is used frequently to frame the topic category of terrorism/violence with the contribution value of 0.07 (out of 1.04) (Figure 2), and topic indicators include ‘attacks’, ‘killed’ and ‘terrorists’. Regarding the use of the topos of reality, journalists emphasise actions in violence and consequences, and the scenarios position most Chinese Muslims as being in a worrying and dangerous situation. The plight of Chinese Muslims in the broadsheet justifies the government’s actions to punish terrorists, who constitute a minority within the Chinese Muslim population. Simply put, the topos of reality reveals the vulnerability of a large population of Chinese Muslim citizens bred by the few evil and lawless Chinese Muslims and the needing of governmental interventions. The topos thus construes governmental anti-terror campaigns that Beijing prioritises the necessity of its ingroup members and bring them security as well as constructs paternalism (cf. Le Grand and New, 2015; Saich, 2012; Wang and Liu, 2018). Example 4 about violence and terrorism elaborates on the use of the topos of reality to portray the victimised Chinese Muslim majority and the paternalistic Chinese government; topic indicators include ‘riot’, ‘terrorist’, ‘extremist’ and ‘destroyed’.

Example 4

Xinjiang Communist Party of China (CPC) chief Wang Lequan said Monday the riot in Urumqi revealed the violent and terrorist nature of the separatist World Uyghur Congress leader Rebiya Kadeer.

‘The riot has destroyed the spiritual support with which the terrorist, separatist and extremist forces cheated the people to participate in the so-called “Jihad”’, Wang said. (7 July 2009)

In the example, harm caused by terrorists to the majority of Chinese Muslim citizens is marked in the voice of ‘Xinjiang Communist Party of China (CPC) chief’. They are first expressed in related to ‘riot’ and ‘Jihad’ and explicitly labelled as ‘terrorist, separatist and extremist’, and the labels explicate their Muslim identity and malicious nature. These terrorists, separatist and extremist Chinese Muslims are harmful to the Muslim majority in China because they ‘destroyed the spiritual support’ and ‘cheated the people to participate in the so-called “Jihad”’. In other words, these Muslims are pictured as betraying Islam and misleading the majority (i.e. ‘the people’) to fundamentalism. The official’s voice and the statement of harm caused by this minority to the Chinese Muslim majority build a boundary between the majority as an ingroup and the minority outgroup. CD justifies the necessity for the government to protect the ingroup by revealing its members’ mental victimisation to the evil group. So, the topos of reality here predicts the government’s activities of anti-terrorism previously discussed, through which the ideology is discernible.