A critical discourse analysis on how four English-language outlets reported the 2018 Taiwanese local election MA journalism program dissertation Oscar Cheng Kai Wu, S3613178 RUG journalism ‘18-’20 (International track) Supervisor/First reader: Dr. Rik Smit Second reader: Dr. Ansgard Heinrich Hand-in date: 2019 October 21
Table of contents
Table of contents ... 2 Introduction ... 4 I. Theoretical framework and literature review ... 8 1.1. Critical Discourse Analysis (CDA), ideology and news ... 8 1.1.1 Discourse studies and the analytical target ... 8 1.1.2 Critical Discourse Analysis and ideologies ... 9 1.1.3 News as ideological discourse ... 10 1.2. Carvalho’s framework ... 12 1.3. CDA and Taiwanese studies ... 14 II. Research design: Carvalho’s framework ... 16 2.1. Sample/Corpus ... 16 2.2. The analytical framework ... 17 2.2.1 Textual analysis ... 17 2.2.2 Contextual analysis ... 20 2.3. The procedure of data analysis ... 21 III. Political and Socio-cultural Context of Taiwanese elections and referenda .... 23 3.1. The 2018 election ... 24 3.1.1 The road to 2018 ... 24 3.1.2 The 2018 election ... 27 3.2. Issues at stake in the 2018 election and referenda ... 30 3.2.1 The China cleavage ... 30 3.2.2 The Olympic referendum ... 33 3.2.3 Same-sex marriage and gender equality ... 35 3.2.4 Nuclear and energy politics ... 37 3.2.5 Third forces parties ... 40 3.2.6 Ko Wen-Je, the mayor of Taipei ... 42 3.3. Corporate background and ideological standpoints of media outlets .... 44 3.3.1 The New Bloom ... 44 3.3.2 The New York Times ... 44 3.3.3 The South China Morning Post ... 45 3.3.4 The Hong Kong Free Press ... 46 IV. Results and discussion: textual and contextual analysis ... 47 4.1. Layouts and structural organizations ... 47 4.1.1 Visual elements ... 474.1.2 The lead ... 53 4.2. Objects ... 57 4.3. Actors ... 58 4.4. Language and discursive strategies ... 60 4.4.1 Those with “one actor” ... 61 4.4.2 Those with “four actors” ... 64 4.4.3 Those with “five actors” ... 66 4.5. Ideological standpoint(s) ... 68 Conclusion ... 72 References ... 75 1. Corpus ... 75 2. Academic and non-academic sources ... 78 Appendix: Coding sheets ... 88
Introduction
After the 2018 Taiwanese local election was held, reporters from the South China Morning Post (Chung and Huang, 2018) contacted Yu Keli, a former director of the Taiwan research center at the Chinese Academy of Social Sciences. Keli was asked to provide context to the election results: The Democratic Progress Party (DPP) was defeated by its pro-Chinese competitor, the Chinese National Party (Kuomintang, KMT). In the report, Keli said that the results “would have a major impact on future cross-strait relations, namely the Sino-Taiwanese relations” and that: The Democratic Progress Party (DPP) is at a crossroads, we will see whether there will be a more rational voice in the party to reflect its policy on relations with mainland China, but we shouldn’t build our hopes up too much. (Chung and Huang, 2018) It seemed that Chinese influence could not be separated from the outcome of the election. To rebuke this logic, another academic, Lin (2018), wrote an article to the Hong Kong Free Press after spending a few months contemplating the link between the electoral outcome and the factors that led to the DPP’s failure:
Quite a number of mainstream media outlets saw China as the biggest winner in this year’s Taiwanese elections as the nationalist Kuomintang (KMT) party won a majority […] The analysis has made a mistake by assuming that voter support for the KMT means support for China […] In fact, some analysts, like sociologist Thung-Hong Lin, see the DPP’s loss at the polls as the result of unpopular domestic policies, like annuity reform, more than any positive public sentiment for the cross-strait relationship. (Lin, 2018)
As the two opinions above demonstrate, the result of the 2018 Taiwanese elections has sent a shockwave through both scholarly and journalistic communities. Both communities did not expect the defeat of the DPP were so deep and profound, and thus were baffled and thus eager to analyze the unexpected landslide defeat of the current governing party, the Pro-Taiwanese DPP and the triumph of the Pro-China KMT. Some link the landslide defeat to Chinese meddling which (allegedly) has waged a disinformation war across the Formosan Strait (Hioe 2018; Horton 2018), whereas others attributed it to the frozen cross-strait
relations and the unpopular reform conducted by the DPP government (Chung and Huang, 2018; Hioe, 2019).
Journalists can wield their pens as weapons. While they adhere to journalistic practices and norms concerned with objectivity and facticity, they also incorporate their political ideology, their understanding of their own and Taiwanese national identity, and information from local sources to their news coverage (Sullivan and Lee, 2018: 273-278). This perplexing news event constitutes a great challenge to those local and foreign correspondents. It tests the limit of their journalistic professional attitude and practices to place their understanding on the map of issues. This thesis will address and analyze the challenge of relaying complex and extraordinary political events and discussions to the general public and those local and foreign journalists through an investigation into local and foreign news coverage (one week prior and after the elections), as the “reality” of the election was constructed by the media being the major contributor to the publics’ perception of the political event. Hence, the research questions are “(1) How do the international and regional outlets, International New York Times (NYT), the South China Morning Post (SCMP), the Hong Kong Free Press (HKFP) and local Taiwanese news magazine, The New Bloom (NB), represent the Taiwanese 2018 municipal election, spanning from the pre-election debate to the post-voting analysis? (2) What do these journalistic representations tell us about the aforementioned news outlets’ views on the Taiwanese national identity? (3) Is there any difference between the outlet’s corporate ideological leaning and the ideological standpoint embedded and shown in their news articles? In order to answer these two related questions, this author will focus on four English outlets: (1) The New Bloom which represents voices of the Taiwanese left-leaning youngsters; (2) the center-left (Pew, 2016) International New York Times which had dispatched a foreign correspondent to Taiwan a few years ago in order to extract local viewpoints and knowledge; (3) the centrist Hong Kong Free Press (Baiocchi, 2015); (4) and the Pro-China South China Morning Post. The reason for selecting these four outlets is that this thesis compares three international outlets (one Pro-China, one centrist in terms of covering Taiwanese affairs, and one left-wing), and a local left-wing online magazine dedicated to young voices. Will there be an overlapping view of the Taiwanese election because of similar left-leaning ideology shown in The New Bloom and the International New York Times? In terms of the future of Sino-Taiwanese relationships, what are the difference of viewpoints in the coverage between those four? To what extent do those ideologically different outlets clash with each other in covering Taiwanese
politics? What are the collision points? Answering these questions is important in order to understand the current Taiwanese political and ideological landscape as well as how those political ideologies embedded in numerous issues shape the past, the current, and the future of Taiwanese politics.
Before diving into those questions regarding local knowledge and details about Taiwanese domestic and foreign politics, it is necessary to shed light on the overall context and history of Taiwan. Therefore, after the first three chapters (the introduction, the theoretical framework, and the research design), the fourth chapter will provide a brief historical overview of the Chinese-Taiwanese political relationship as well as other important contextual issues, ranging from marriage equality to energy politics. A basic understanding of this historical context is required in order to analyze and situate the coverage of the 2018 elections. After setting up the stage, the fifth chapter will present textual and contextual analysis and discuss the results of the analysis.
In terms of research design, this thesis conducts a Critical Discourse Analysis (CDA). This method is suitable for its ability to dissect those news articles and reveal hidden and implicit ideologies, social conditions, and power relations (Carvalho, 2008). In order to systematically perform such an analysis, Carvalho’s CDA framework (2008: 167) will be employed. In the framework, Carvalho discusses two key aspects of discourse analysis: textual analysis and contextual analysis. Under the category of textual analysis, Carvalho introduced six dimensions that could be utilized as parameters of analyzing news articles. This thesis starts by discussing two fundamental dimensions: (1) objects, and (2) layout and structural organization
First, the dimension of “objects” concerns those topics news articles discuss. After the selection of objects, which defines the topic, one proceeds to pick an angle. Second, as discussed by Carvalho (2008: 167), the dimension of layout and structural organization is mainly made up by elements such as the media outlet, page number, size, and visual elements. Headlines, the lead and the first few paragraphs could also be included in this dimension.
The remaining four dimensions are particularly important for this thesis because they illustrate the link between the usage of language and the construction of meaning in those news articles: (3) actors, namely the identity of those actors and how those actors were represented; (4) language, grammar, and rhetoric (e.g. metaphors): how language is utilized in depicting the journalist’s ideology, and what are the intended effects the correspondents want to achieve; (5) discursive strategies: This strategy, as defined by Carvalho, is made up of five elements: selection, composition, positioning, legitimization and politicization
(Carvalho, 2008: 169). The first two elements: “selection” and “composition” were put together into one dimension because of their similar nature: arrangement of facts; and also bundles the late three into another dimension that could be used to analyze the relationship between the usage of language, symbols, and hidden meanings; and the sixth element, (6) the ideological standpoints: which tell us what ideological leaning those journalists stand for in writing their articles. Ultimately, the goal of this thesis is to show how newspapers that are operating in a highly contentious political climate discursively construct reality. This thesis explores an unchartered territory of Taiwanese politics: the 2018 Taiwanese local election. This election is not merely an election of governors and city councilors, but it also contains more than a dozen of referenda which are constituted by various issues. On those specific issues, each outlet has their own political and ideological standing. Therefore, through employing the CDA approach, this thesis will be able to excavate and thus reveal hidden ideological leaning inside those news articles. But before laying down the foundation of contextual knowledge of Taiwanese politics, it is vital to discuss the theoretical framework that serves as another foundation of the CDA methodology.
I. Theoretical framework and literature review
This chapter addresses fundamentals: the theoretical framework around critical discourse analysis (CDA) and a literature review. Since CDA can be both considered a theoretical and a methodological approach, this chapter reviews the literature on this approach. CDA has a long history in a number of disciplines and has been conducted in a range of ways. It is therefore necessary to provide clarity about how this thesis is situated in this tradition. 1.1. Critical Discourse Analysis (CDA), ideology and news 1.1.1 Discourse studies and the analytical target Before diving into the field of critical discourse analysis and its relations to ideology and news, it is important to address two fundamental question: “what is the difference between discourse analysis and critical discourse analysis?”; “what is the analytical and empirical target of discourse studies/analysis?”
First, While the analysis of discourse is the study of the structure of a message in a communication or an examination regarding various functions (pragmatics), CDA is an effort or a parsing process to give an explanation of a text (social reality) and being reviewed by a person or a group of dominant tendency who has a specific purpose. CDA aims at explaining how social inequalities are reflected or created in language itself, as social semiotic, but in the use of language as social action (Habibie, 2016: 13; Widdowson, 1995: 167-168). Second, although discourse analysis originally developed from predominantly linguistic approaches to discourse analysis, discourse analysis, and studies, it is not merely textual studies like linguistics (Kelsey 2018:246). The analytical unit of discourse studies is not limited to formal grammar and abstract sentences. They include the natural use of languages by real language users in real social situations. That is, structures and strategies that are embedded in social interaction and communication are the main focus of discourse analysis. Besides those verbal dimension of the text, discourse studies also utilize a multimodal research approach that could also be applied to non-verbal elements such as images, sounds, and videos (van Dijk, 2008: 192).
Therefore, given the fact that there are various analytical units to study, it is less likely that one academic discipline could encompass and facilitate all discourse studies. Unlike other studies with unitary discipline or theory as backbone, discourse studies are comprised of levels of theories with a variety of disciplines. On the one hand, these levels may be described by micro-level analyses, and by macro-level analyses, on the other (van Dijk, 2008: 192).
To be specific, the micro-level of analysis relates to discourses as language use and its presupposed cognitive aspects of production and comprehension, which involves various kinds of mental strategies, knowledge, mental models and other representations in memory. On the other hand, discourses that are studied in relation to various kinds of “situation,” such as interactional, social, communicative, political, historical and cultural frameworks, are interpreted by the participants as relevant macro “contexts.” (van Dijk, 2008: 192)
One basic principle of these analyses is that of “sequentiality”, which means each analytical unit at each level (word, sentence, meaning, speech act, turn, etc.) of discourse is produced, interpreted and analyzed as being conditioned by previously interpreted units. That is, if the process of production, interpretation, and analysis is prolonged and repeated, it will become a reproduction of certain discourse that reinforces and enhances the meanings the text/discourse tries to convey. This will result in, in general, the reproduction and enhancement of a certain social knowledge that comes from a certain social group. In short, discourse will transform into a form of domination that might help to facilitate social inequality (van Dijk, 2008: 192). Through the help of employing the analytical dimension of sequentiality, researchers are able to identify those discourse dominations, which thus tackles a part of social inequality derived from it. 1.1.2 Critical Discourse Analysis and ideologies Similar to discourse studies/analysis, Critical Discourse Analysis (CDA) also views discourse as “social practice” (Fairclough & Wodak, 1997: 258), and takes a special interest in the dialectical relation between language and society, which is normally hidden and implicit. A dialectical relationship refers to a two-way relationship in which discourse is shaped by social contexts and it also shapes them (Fairclough and Wodak, 1997). However, the relationship between language and society can be described, interpreted, and explained through studying and analyzing written and spoken texts. CDA could reveal hidden and implicit ideologies, social conditions, and power relations (Yang, 2013: 84-85; 2018:137). The key CDA theorists include Fairclough (1992, 1993, 1995), van Dijk (1993, 2001,2015), Gee (2005), van Leeuwen (1993), and Wodak (2001). Although they hold somewhat distinct views on CDA (Rogers, 2004), it is generally agreed that CDA should be viewed as an approach to investigate the relationship between the use of language and social context (Yang, 2013: 84-85). However, CDA is actually made up of different principles and methods. The divergence on this perspective caused many CDA theorists to develop their own CDA definition and facilitation
(Van Dijk, 1993; Meyer, 2001). Some of them share similar points of view, while others are rather different from each other.
For example, Fairclough (1995:54) claimed CDA often involves a search for aspects or dimensions of reality that are obscured by an apparently natural and transparent use of language. Unlike the linguistic approaches of Fairclough and Wodak (1997), van Dijk (1995) develops methods which are more focused on cultural and social contexts. The methods are based on four categories: action, context, power, and ideology. Discourse analysis is, from van Dijk’s (1995: 17) point of view, ideology analysis, because “ideologies are typically, though not exclusively, expressed and reproduced in discourse and communication, including non-verbal messages.” Then, he further claims that ideologies “indirectly influence the personal cognition of group members” (van Dijk 1995: 19) and they are manifested through discourse. The manifestation is particularly presented by Us versus Them dimensions, in which people of one group are likely to use positive terms to describe themselves while they present others in negative ones (van Dijk, 2008:193-194). Besides ideology, CDA also emphasizes the issue of “power” as its analytical goal. The researcher then tends to be alerted to power relations being exercised through discourse and aims to overcome the normal opacity of social practices (Carvalho, 2008: 162). In a society, the relationships of “dominance, discrimination, power, and control as manifested in language” can be described, interpreted, and explained (Wodak, 2001: 2). In other words, every discourse is produced and interpreted historically in a society where powerful groups dominate discourse structures. Thus, three major themes shared by all CDA models are “the concept of power, the concept of history, and the concept of ideology” (Wodak, 2001: 3).
1.1.3 News as ideological discourse
Journalists work within organizational structures and adhere to editorial values and pressures that influence the work they produce. Furthermore, journalists are people who are part of the shared cultural practices that readers and audiences are engaged within society (Kelsey 2018: 246). The social actor that is able to forge the aforementioned three concepts—power, history, and ideology into one pot is the media. As Fairclough (1995) and Van Dijk (1988) consider the media as key ideological brokers, media are able to reproduce and maintain discourses of dominant social order.
Van Dijk understood that media studies were already concerned with the economic, social, and cultural contexts of news media, but he also stressed the
importance of conducting a structural analysis of news stories as a “qualitative alternative to traditional methods of content analysis” (1988, vii). He claims that news discourse is “explicitly linked to social practices and ideologies of news-making and, indirectly, to the institutional and macro-sociological contexts of the news media” (1988, vii).
Hence, news discourse may be considered as an ideological and significant influence within society (Costelloe, 2014). For instance, it has been argued that news coverage shaped the direction of a protest movement as a general consequence (Lin, 2014). Furthermore, Bednarek and Caple (2012: 6) describe news discourse and media influencing governments, major institutions and the reader audience; in short, the discourses and linguistic choices made by the media outlets may have an influential impact on society. The journalistic response to certain issues is of vital consequence as to how the public comprehend and participate in socio-political events (Brindle, 2015: 5-6). However, the influential impact the media possessed is neither entirely neutral nor objective. The media construct representations of events, which may be biased according to the political affiliation of a particular media organization or the stance of the journalists writing the articles (Baker et al., 2013). Particular language usage results in specific interpretational patterns for the comprehension of social issues, relations and identities (Habermas, 1999)
Numerous CDA approaches that contribute to the studies of news, discourse, and ideology can help to identify those biases through interpretational patterns. For example, news structure analysis shows us where and how ideologies preferably manifest themselves in news reports. Van Dijk’s (2008) new socio-cognitive approach explains how underlying ideologies control more specific group attitudes and how personal mental models of journalists about news events control activities of news making, such as newsgathering, interviews, news writing, editing, and final makeup. That is to say, these news-making activities are ultimately controlled by the specific, ongoing context model of the journalists about the relevant aspects of the social and political situation. Such context models of news-making include current setting (location, deadlines, etc.), news participants (reporters, editors, news actors, sources, etc.) and their roles, as well as current aims, and the social knowledge and ideologies of the participants (van Dijk, 2008:195).
Another approach to this matter is Fairclough’s three-layer model of CDA (1995, 2003) which covers textual analysis, discursive practices, and social practices. This framework transcends “the division between work inspired by social theory which tends not to analyze texts, and works which focuses upon the
language of texts but tends not to engage with social theoretical issues” (Fairclough, 2003: 2–3).
The first layer of Fairclough’s model is concerned with textual analysis. This is more than a linguistic or grammatical analysis of texts. It concerns the ideological role of language as a social product. Language is used by powerful groups to reinforce dominant ideologies and therefore needs to be studied as a site of ideological struggle (Simpson, 1993). Therefore, the second layer of analysis is discursive practice. Attention to discursive practice means that processes of textual production and consumption are scrutinized. In terms of news discourse and journalism, textual features are all seen to occur through decision-making processes in their cultural and professional contexts. Journalists and news organizations are wrapped up in cultural practices of time and place as well as their employers having their own editorial interests and socioeconomic agendas to fulfill (Kelsey 2018: 250). Third, the level of social practice expands beyond media texts and examines some the wider social contexts of which the discourse is a product of and responding to. This is often the level where CDA takes an explicit normative or political position in response to a problem, dominant ideology, or exploitative social relations (Richardson, 2007).
To conclude, journalism is typically a discursive and ideological reconstruction of reality. Rarely do journalists witness events or get to know reality in a way that does not involve the mediation of others. The media representation of social issues seems to be very much a function of social actors to organize their claims and to project attention to ‘‘happenings’’ and problems (Carvalho, 2008: 164). This also means that whatever other professional and social ideologies (including norms, news values, etc.) may be at work in news production, the constraints of the now relevant context, as defined by the participants, are the crucial filter that makes news more or less appropriate in the current social and political situation (van Dijk, 2008:195).
1.2. Carvalho’s framework
Despite the analytical and methodological advancement CDA has achieved in the past decades, two groups of critics emerge. For the first group, some have claimed that it is too ideologically committed (Tyrwhitt-Drake, 1999), while others have suggested that its methodological diversity should be overcome (Toolan, 1997). However, most of these criticisms do not diminish the theoretical and analytical value of CDA. Methodological pluralism, for instance, can be seen as a strength rather than a weakness, and the ideological commitment, as being
discussed previously, is an explicit agenda and an indispensable element of CDA and does not equal analytical distortion (Carvalho, 2008: 162).
For the second group of criticism, Carvalho argued that CDA has not been able to account for the full cycle of news discourse, which involved the integrated analysis content and processes of production, reception, and circulation. Nonetheless, this has not been answered by another scholar’s work. In order to answer the criticism she came up with, she developed three critical aspects: the time plane in discourse analysis, the discursive strategies of social actors and the effects of mediated discourse.
First, the time plane in discourse analysis refers to the historical context involved in analyzing discourse as previous forms of CDA do not account for the time sequences of texts nor do they clearly explain the implications of previous discursive positions on subsequent ones. Likewise, the same phenomenon also exists in most studies of media discourse, which are like snapshots examining some news items in detail but covering a short time span (Carvalho, 2008: 163-164).
While this short-spanned approach may be relevant for some events, most public issues have a significantly long ‘‘life’’. Understanding the evolution of matters such as authoritarianism, nationalism or climate change, and the ways they are interdependent in relation to the media is one of the most important contributions to be made by researchers (Carvalho, 2008: 164). However, there are obstacles to doing this. Examining the various stages of the public life/media coverage of such issues, from the conditions of emergence in the public arena to their constitution into political problems, formulation of answers, adoption of measures, implementation, and evaluation, requires analyzing media(ted) discourses for a relatively long period of time. This means a large amount of work and man-hours.
Analyzing the second aspect: the discursive strategies of social actors is also time and resources consuming. The reason that it involved in a great deal of work is that a variety of social actors serve as sources of information for media professionals, in a direct or indirect way. Dissecting the web of discursive strategies entangled by social actors is a huge task. However, if this task is fully fulfilled, researchers will be able to create a crystal-clear analytical framework in which presents the news values in operation, preferences, and options of those media professionals largely depend on depicting social problems (Carvalho, 2008:164-165).
Critical Discourse Analysis should also account for those two levels of discursive intervention over a certain ‘‘object’’, namely the sources’ or social
actors’ intervention, and the journalists’ intervention. This breaking-down method would help to understand and make explicit the alignment of news media with a given side and increase awareness of the plurality of views, as well as of the biases both in the media and in social actors’ discourses (Carvalho, 2008:164-165). Third, while discourse analysts have concentrated attention on the text a constitutive view of discourse has to encompass the analysis of discourse’s concrete means of extra- or supra-textual effect, which means they are realized beyond or independently of a given text. Media discourse is an especially important arena for social and political action (Carvalho, 2008:165).
To address the issue of extra- and supra-textual modes of operation of discourse, Carvalho proposed the category of discursive effects. Discursive effects are processes that are linked to texts but occur outside the text or ‘‘above’’ it. They cannot be ‘‘found’’ in one single text. Moreover, discursive effects are not the direct consequence of one actor’s discursive interventions but are often dependent on a variety of (discursive) causes and circumstances and show the constraining force of discourse (Carvalho, 2008:164-165). In conclusion, three necessary elements can be identified for critical scrutiny: the historical frame, the social actor and their discursive strategies, and discursive effects. These elements have been integrated into Carvalho’s newly developed “framework for analysis of media discourse” : the “discursive effects” element was integrated into the contextual analysis, and the element of “historical actor” and the “social actor and its discursive strategies” were assimilated into the textual analysis. This thesis will briefly explain this framework for the analysis of media discourse in Chapter 3, then apply the framework to all the news coverage about news coverage of the 2018 local election in Taiwan in Chapter 5. 1.3. CDA and Taiwanese studies CDA has been applied to the Taiwanese media landscape. Kuo (2007a; 2007b), for example, used a corpus-based approach to analyze discursive change in media discourse. Other studies of Taiwanese newspaper discourse include Wang (2009), who adopted a critical discourse analysis (CDA) approach to study gender stereotypes in Taiwanese tabloids, and Kuo and Nakamura (2005), who compared how two ideologically different papers in Taiwan gave a different account of the same event: an interview with the Taiwanese First Lady. The authors show that such specific linguistic differences of news report may be explained in terms of the reunification vs. independence ideologies of the two newspapers.
In the second decade of the 21st century, studies about Taiwanese media discourse resonate with the huge societal and political changes in recent years.
For example, Calello and Chen (2013) utilized Allan Bell’s (1998: 59) news discourse structure model to examine three newspaper articles’ coverage of Costa Rica shifting its diplomatic recognition from Taipei to Beijing. China, likewise, appeared again as the main topic in the next critical discourse study. Yang (2013) showed how suitable van Dijk’s socio-cognitive model is in applying to describe, interpret, and explain the relationship between language and power in the national debate on the issue of a major economic agreement between Taiwan and China—the Economic Cooperation Framework Agreement (ECFA).
Unlike ECFA’s smooth ratification passage in the Taiwanese legislature, the vetting procedure of the CSSTA (Cross-strait Service Trade Agreement) was obstructed by several angry protestants, and later prompted students and citizens occupied the parliament in protest for leaving Taiwanese industry vulnerable to Chinese political pressure. English newspapers in Taiwan, however, according to their own political affiliation and situational context exposed to their media professionals, construct different discursive strategies despite working on the same story. Brindle’s study (2015) employed a corpus-based approach to discourse analysis to examine two leading local English newspaper in Taiwan, in corpora of articles published in the six-month period after the protest began. In response to the student protest—widely known as Sunflower movement— then-president Ma of Taiwan gave a speech and released statements to the whole nation. Those texts become samples that were subject to further critical discourse studies. Yang (2018) complied and examined President Ma’s press releases and official statements. It shows the power of language could be presented differently by different sources of identities; and the strategies of language could also be applied variously for different purposes (Yang, 2018: 141).
II. Research design: Carvalho’s framework
This section aims at introducing Carvalho’s analytical framework. In the second section I described Carvalho’s framework as a crystallization of previous debate on CDA methodology. That is why this thesis intends to employ this framework as its own tool to dissect and analyze the corpus, which will be introduced later in section 2.1.
2.1. Sample/Corpus
As mentioned previously in Chapter 1, the first research question of this thesis is
How do the international and regional outlets: The International New York Times, the South China Morning Post, the Hong Kong Free Press and local Taiwanese news magazine, The New Bloom, represent the Taiwanese 2018 municipal election, spanning from the pre-election debate to the post-voting analysis?”
This question already includes several elements that pinpoint the size and content of the sample. The first element is the label of the sample, which can break down into four categories of different newspapers and internet outlets: the International New York Times, the South China Morning Press, the Hong Kong Free Press and The New Bloom, a Taiwanese local internet outlet. These outlets, despite adhering to the similar journalistic principle such as objectivity and facticity, they nonetheless represent various ideological leanings due to the outlet owners’ and the journalists’ political affiliation. For example, outlets who had been bought by Chinese capitals would present pro-China viewpoints; on the other hand, those who run without Chinese capitals would present more pro-Taiwan opinions. This will result in they will hold divergent views on presenting the same political or social event— either explicitly show or implicitly embed in their production, sourcing and editorial agenda.
This leads us to the next element: the political event this thesis aims to analyze. The news reports are naturally limited in the timeframe of the Taiwanese local election and the simultaneous referendum, in which numbers of news coverage peaked at numerous debates, the voting and the announcement of the (un)official electoral/referendum result. Those events are roughly situated in the two weeks—one week before and after the Election Day: 24th, November. There are some outliers as well, such as two HKFP articles which are about the
referendum result, are written in the midst of January 2019. However, most of them are produced in between two weeks range. Thirdly, the outlets, the political event, and the electoral timeframe constitute three layers of filters that help this thesis sorting out samples. The whole sample size is comprised of 35 articles, with 21 articles from The New Bloom, 3 articles from the New York Times, 6 from the Hong Kong Free Press and 5 from the South China Morning Post (a Google Drive link to all articles is provided in the appendix). 2.2. The analytical framework
Carvalho’s framework is the crystallization of previous CDA critics and debates as this thesis had discussed it in Chapter 1. That is why this thesis intends to apply this methodological framework to a timely issue— the 2018 Taiwanese local election and referendum. But before heading toward the actual analysis and result, it is important to introduce Carvalho’s framework first. Carvalho’s framework (see below graph 1) has two main backbones: textual analysis and contextual analysis. The former one consists of six elements— layout and structural organization; objects; actors; language, grammar, language and rhetoric; discursive strategies; and ideological standpoints. On the other hand, the latter has two approaches to conduct the analysis— comparative-synchronic and historical-diachronic. This thesis will dissect the concept, explain the meanings and apply those frameworks to analyze aforementioned corpus. Figure 1 (Carvalho, 2008: 167) 2.2.1 Textual analysis (1) Layouts and structural organization In this category, there are “Surface’’ elements of the newspaper and of the text itself, which contain the section in which the article was published, the page number, the size of the article, how the visual elements are arranged and what are
those photographs, graphics or others about. According to Carvalho (2008: 167), those elements say something about the valuation and categorization of the issue by a given news outlet, with implications for the audience’s perception (2) Objects The notion of “objects’’ is close to topics or themes. Furthermore, rather than just ‘‘refers to’’ the realities at stake, the term ‘‘object’’ has the advantage of enhancing the idea that discourse constitutes. However, objects of discourse are not always obvious, but clearly identifying them is an important step towards deconstructing and understanding the role of discourses (Carvalho, 2008: 167-168). (3) Actors The core questions related to the category “actors” are: who does the article mention? How are those actors represented? The term ‘‘actors’’ in this analysis means both social agents (someone who has the capacity of doing something) and characters in a (staged) story (which is ultimately what news reports are). One principle is clear: this thesis is focused more on the individuals or institutions that are either quoted or referred to in the text (Carvalho, 2008: 168). (4) Language, grammar, and rhetoric This category has three linguistic elements: language, grammar, and rhetoric. For the first element, the vocabulary used for representing a certain reality, for example, verbs, adjectives, adverbs, and the writing style such as formal/informal, technical, conversational are important dimensions of the constitution of meanings. Linguistic analyses of journalistic texts pay close attention to issues of pragmatics, semantics, and syntax in discourse in a much more detailed manner (Fowler, 1991). However, such analysis is not the focus of this framework/thesis and the analysis of language will be limited to the aspects listed above (concepts, vocabulary/lexical choices, and style) (Carvalho, 2008: 168-169).
Secondly, the study of a text’s grammar can reveal many of its underlying (ideological) presuppositions. For instance, nominalizations and active/passive sentences are among the most important syntactic features in news discourse(Carvalho, 2008: 168-169). For the third element: rhetoric, it contains metaphors, other rhetorical figures and persuasive devices employed in the text (van Dijk, 1988b). An emotionally charged discourse, with an appeal to readers’ emotions, for instance, is often found in the press and can have an important rhetorical role. In the analysis of language and rhetoric as this thesis has argued in
theoretical chapter, there are formulations advanced by social actors, and by journalists as well (Carvalho, 2008: 164,168-169). Therefore, this thesis will follow this notion that linked (social) actors with linguistic features.
(5) Discursive strategies
Discursive strategies are forms of discursive manipulation of reality by social actors, including journalists, in order to achieve a certain effect or goal. Here manipulation does not have the sense of an illegitimate alteration of a certain reality (van Dijk, 2006). Rather, Carvalho (2008: 169-170) used the term to mean a discursive intervention. This intervention and the procured goal can be more or less conscious. Specifically, the goal is to map discursive strategies in relation to social actors: which actors use which discursive strategies? How are different actors involved in the discursive strategies of others (how are they constructed by others)?
The main intervention the journalists makes is enacted in the selection of an angle of the (complex) reality. This is an important part of the act of ‘‘framing’’ a certain reality. Framing is to organize discourse according to a certain point of view or perspective. In the production of texts, framing involves selection and composition (Entman, 1993) which together become one of our analytical tool this thesis will utilize in the upcoming chapters. To be specific, the selection is an exercise of inclusion and exclusion of facts, opinions and value judgments. The composition is the arrangement of these elements in order to produce a certain meaning.
The second analytical tool includes legitimization and politicization. Legitimation consists of justifying and sanctioning a certain action or power, on the basis of normative or other reasons (van Leeuwen and Wodak, 1999). Politicization is the attribution of a political nature or status to a certain reality, such as climate change. Some of these strategies have a reverse, such as de-legitimation and de-politicization (Carvalho, 2008: 170).
The discursive strategy is the undercurrent flowing beneath all the other aforementioned elements. The discursive strategies of news professionals are implicated in the layout and structure of the text, in the construction of objects and actors in discourse, and in the language, grammar, and rhetoric (Carvalho, 2008: 171). (6)Ideological standpoints It is important to make ideologies manifest because they involve fundamental motivations and justifications for keeping or changing a certain status quo.
However, one should expect the ideological standpoints of an author not to be always explicit in the news texts. Appearing natural is at the core of the ideologically shaped work of representing reality in the news media. Therefore, the analyst has to learn to identify ideological standpoints from relatively subtle mechanisms and devices. Looking at alternative constructions of the same reality, such as applying comparative-synchronic analysis to different media reports can be a helpful strategy (Carvalho, 2008: 170-171).
Ideology is embedded in the selection and representation of objects and actors, and in the language and discursive strategies employed in a text. That is to say, the analysis of both discursive strategies and ideological standpoints in journalistic texts is not independent of the analysis of the other elements enunciated previously. In turn, ideological positions have to be inferred from all the other elements, as well as from discursive strategies (Carvalho, 2008: 170-171). 2.2.2 Contextual analysis There are two analytical approaches in the category of contextual analysis. First, a comparative-synchronic analysis means looking at various representations of an issue at the time of the writing of one specific news text (the unit of analysis). More specifically, this approach entails comparing one text with other representations of the issue: texts published on the same day (or another time unit) by different authors, both in the same news outlet and in others (Carvalho, 2008: 171-172).
Second, the historical-diachronic analysis involves examining the course of social matters and their wider political, social and economic context. It also involves examining the temporal evolution of media(ted) discourses and to produce a history of media constructions of a given social issue. This involves looking at the sequence of discursive constructions of an issue and assessing its significance: how did representations of real impact on subsequent ones? How were they reproduced or contested? What were the political, social and/or cultural implications of dominant discourses? What alternative arguments and proposals were excluded from public agendas and why? (Carvalho, 2008: 171-172) Given the feasibility of data and research, it is important to note that this thesis will only apply the first approach: the comparative-synchronic analysis to the corpus. There are several reasons for this decision. First, the comparative-synchronic approach is a confrontation of alternative depictions of reality that mainly aims to enhance the critical reading of news discourse and help identify the specific discursive traits of a given news outlet. The same event depicted by
multiple authors in different reports, using various linguistic and discursive strategy that contains numerous and conflicted ideological standpoints, will be examined thoroughly through this approach.
Second, the historical-diachronic approach requires a great deal of data in order for the researchers to rebuild a timeline of discursive strategies. The fact that this approach is beyond the scope of this thesis is because examining all the other local and general elections in Taiwan is unaffordable. To be specific, this thesis will not fulfill all two levels of historical-diachronic analysis. Rather than fully investigated the discursive strategy employed in all the coverage of Taiwanese election held until so far, it can only apply the first level, which is aimed at examining the related social and political context of one social event. These issues will be introduced and discussed later in the chapter of context.
2.3. The procedure of data analysis
In previous sections, we had already stated clear the analytical aim, the corpus and the methodological framework this thesis will apply to analyze the corpus. This section, however, will explain how to piece together all the necessary elements into one standardized analytical procedure in order to make the whole critical discourse analysis smoother and more efficient. First, by using the date filter, this thesis gathers all 35 news reports of which almost all were written within two weeks—one week before and after the election. Although the rest of them were not written within the two-week range, their topic is relevant to the election and the referenda. Second, it is necessary to either copy-and-paste or directly download it from the internet and make them into word (.doc) files. This includes all the visual elements: graphics and photographs, which play a huge role in our critical discourse analysis, along with texts, the title, and the lead. This thesis also takes record of all the authors’ name and the release date of the news text for which will benefit upcoming CDA that includes rebuilding a timeline and identifying the link between a particular author and the ideological standpoint s/he possesses. The third step is applying Carvalho’s framework to the texts. Based on the instruction from Carvalho, this thesis design and use the coding form showed below. Starting from the basic categories such as the tile, analysis of the lead and all the visual elements will be in the former part of the form. The central part of the form is the backbone of critical discourse analysis, which includes linguistic and discursive elements that embed in the news reports. The last part: the ideological standpoint will be treated as a short judgmental conclusion of the news report as ideology—no matter explicitly shown or implicitly hidden— is not
independent of other elements of CDA and will only be dug out after all the elements had been carefully examined. (Number) Title of the text Layout and structural organization Object Actors Language, grammar, and rhetoric Discursive strategies 1. Selection and composition 2. (De) legitimization and politicization Ideological standpoints Figure 2: the coding form
The fourth step is to conduct contextual analysis: comparative-synchronic approach and the first level of the historical-diachronic approach. Beginning from the ideological category, we examine and classify all the standpoints across each particular issue. For example, some reports are about the local election, but others emphasize gender equality, energy, and the Olympic referendum. This thesis records down these issues one by one, and later crosstabs the result with each author and outlets. That is to say, this will generate the result similar to an ideological spectrum that shows each author’s ideological standpoints on each specific issues. This thesis will then compare the general context of those particular issues, and the ideology backed up by the outlets’ ownership background to the result of CDA.
Before commencing with the CDA coding forms (see appendix 1) and the result coming out of them, it is necessary to discuss another layer of context and integrate it into the thesis. Namely, the context of issues ranging from energy, gender equality, the Olympic name-changing controversy, and the general local election, Discussing these issues will help understand the news reports better and make a more in-depth CDA possible.
III. Political and Socio-cultural Context of Taiwanese
elections and referenda
There is no academic research done on the latest Taiwanese election because so far it is only less than one year after the local election. Therefore, a description of context has to rely on journalistic news articles, analysis, and blog posts written by Taiwanese affairs observers. Some of those observers work in academia, but that does not equalize their own blog posts as academic projects. There is, however, a merit of choosing those articles as sources. First, those posts are free from malicious partisan competition currently plaguing Taiwan. Despite not being entirely ideology-free, the scholar Nathan Batto, who runs a blog used to inform this context chapter, knows how to avoid partisan biases. The evidence can be shown in his posts about the monthly renewed presidential poll in which the most biased data set were treated as inconsequential. This principle also applied to other qualitative posts: Batto (2018) did not conclude the article without obtaining the first-hand data and peer-reviewed information, such as the one from the legislature or information from the Academia Sinica, where Batto currently works. Along with other journalistic works such as The Economist, which also upholds the notion of treating “facts” with caution and sourcing professionals including Lo Chih-Cheng (The Economist, 2018), both a DPP lawmaker and a political science professor, they provide precious contextual insight that is helpful to this thesis.This chapter makes up a significant part of this thesis because the socio-cultural context—namely the detailed information about Taiwanese politics, elections, referenda, and history—is important in understanding the topics covered in the analytical framework, which are called “objects” in Carvalho’s CDA framework. Moreover, rather than just ‘‘referring to’’ the realities at stake, the term ‘‘object’’ has the advantage of enhancing the idea of the discourses (Carvalho, 2008: 167) Therefore, in order to gain more background knowledge about what these topics entail in the article, the author of this thesis classifies and selects six main topics based on the results obtained from the corpus on which CDA framework is applied. Those six categories are the China cleavage, the Olympic referendum, same-sex marriage, nuclear and energy politics, third forces parties and Ko Wen-Je, the mayor of Taipei city. Those topics will be discussed in detail in this chapter in section 3.2.
3.1. The 2018 election
3.1.1 The road to 2018Since the democratization in the 1990s, Taiwanese politics has been dominated by several political parties, ranging from the former authoritarian party, the Kuomintang, which had ruled the island for decade after 1945; to the one who were born as the descendent of Dangwai1 movement, namely the DPP; to those who did not satisfy with the mainstream party lines and thus splintered from the two main parties: the New Party (NP), the People First Party (PFP) and the Taiwan Solidarity Union (TSU); and to those who were assembled after the Sunflower movement: the New Power Party (NPP) and the Social Democratic Party (SDP) (Fell, 2016: 87-99; Hioe, 2017). The history of political parties, which includes their formation, peak and eventually fading away or growing strong, reflects the political history of Taiwan (see below figure 2). It is important to pave some historical context before diving into the 2018 election, which is the main theatre of this thesis. Figure 3: the timeline of Taiwanese politics and relations with China (AFP, 2015) After the death of the longtime despot Chiang Ching-Kuo in 1987, the son of another dictator Chiang Kai-Shek, the Chiang dynasty which saw Taiwan being ruled by iron-fist, patron-client relationship, and political and economic monopoly, had ended. The successor of the presidency was Lee Teng-Hui, who was also the first President of the Republic of China (ROC) born in Taiwan instead of China
(PRC). In his term, Taiwan experienced slow but steady democratic reform which was the result of taking advice from the opposition DPP lawmakers and some liberal-leaning young student activists. Several elections, including the first direct presidential one in 1996, and the first re-election of the parliament after 1949, was held in accordance with the reformists’ expectation. Representing the KMT reformist wing, Lee eventually won the 1996 presidential election in a landslide victory despite facing blatant threats from China due to their unwillingness to recognize Taiwan electing their own leader (Jacobs, 2016: 51-59). Taiwan, in the spring of 2000, would experience another critical juncture to test if the democracy is resilient enough to accommodate DPP’s victory, President Lee could not seek another re-election for the constitution only promulgates tow terms for the presidency, which is why the election must be held at the end of his term. The election proved to be rather smooth and peaceful, but the result was beyond many people’s expectation: Chen Sui-Bien, from the Democratic Progress Party, had won by taking 39% of the votes. His KMT and independent-turned-PFP rivals conceded defeat (Jacobs,2016: 59-63). The reform-minded president Chen faced several challenges right after the beginning of his first term. He needed to tame the restless military and bureaucracy, which were once two backbones of the old KMT regime. He had to mitigate the gap between those who strived for Taiwanese independence, an old anti-KMT and anti-China movement that was born in the early 20th century (Fleischauer, 2016: 68-82), and those who advocated political pragmatism toward both ROC, PRC and the US. However, despite its narrow victory in the 2004 election which sees his term extended to 2008, his presidency proved to be scandalous. His family and government were mired in corruption allegations. Those incidents led to the DPP’s grave defeat in the 2007 local election and 2008 presidential one (Jacobs, 2016: 63-64).
President Ma Ing-Jeou, who represented the KMT, had assumed office in spring, 2008. His government proved to be inefficient and incompetent. It is still debatable whether the disasters Typhoon Morakot had brought could be attributed to Ma’s administration’s incapability of preparing for a natural catastrophe. But one thing is for sure, one month after the typhoon menace, the premier resigned and Ma’s reputation plunged. These facts, however, did not lead to his defeat in the 2012 election. He won by a margin of 800,000 votes to the second candidate: Tsai Ing-wen (Jacobs, 2016: 64-65).
Ma’s second term was trapped in waves of student protests. The most prominent was the Sunflower movement, which was sparked by the controversial reviewing of a trade deal with China. This resulted in students storming the
parliament and occupying it for several weeks. Including the anti-nuclear energy protest, which was also being crushed by the riot police, those incidents induced wide dissatisfaction that culminated into KMT’s defeat in the 2014 first-ever nine-in-one local election. After the election, KMT found themselves governed only six out of the 22 cities and prefectures (Rich and Sullivan, 2016: 131).
In 2016, Ma could no longer seek re-election bid. Therefore, KMT had to nominate another political heavyweight to run against the same DPP candidate: Tsai Ing-wen. This time, due to several public safety incidents and the student protests that shattered the Ma administration to the core, the KMT president candidate was defeated by Tsai in a landslide. Tsai won an overall 56% of the presidential vote (see below graph 3) (Jacobs, 2016: 66). Other new parties, including the NPP and the SDP that born from waves of protests, also participated in the election. The former one took five seats in the parliamentary election and thus became the third-biggest party. Figure 3: the 2016 presidential and legislative election (The Economist, 2016)
Tsai administration was eager to establish its reputation as the most reform-minded government since the beginning of democratization. Pension of the military, public service people and teachers were slashed in order to make fiscal sense (Shih, 2018). Cases of past human rights abuses conducted during the Chiang administration were put under serious transitional justice review, which includes the removal of statues of Chiang Kai-Shek from all the schools and local government branches. Other issues such as same-sex marriage and denuclearization were also high on the parliamentary and administrative agenda. Those reforms, however, sparked controversies and later emerge as a bellwether issue in 2018 local election and referendum. 3.1.2 The 2018 election In 2014, the DPP had won 13 of the 22 cities and counties, accounting for 62% of Taiwan’s population. The KMT had only won 6 cities and counties, covering 25% of the population. This year, those numbers were reversed. The DPP could only hold 6 cities and counties, encompassing 27% of the population, while the KMT won 15, with 62% of the population. There was a similar reversal in votes. In 2014, DPP candidates won about 840,000 more votes than KMT candidates; this year, KMT candidates outpolled DPP candidates by over 1.2 million votes (Batto, 2018). The DPP saw its gravest failure in some of the largest cities and traditional strongholds. The DPP’s most stunning loss came in Kaohsiung City, which it had not lost in 20 years. Yet the DPP not only lost Kaohsiung, but the race also wasn’t particularly close. The KMT candidate won by 9.1%. In the other major deep green city, Tainan City, 29.6% of voters cast a protest vote for one of the minor candidates, holding the winning DPP candidate to a mere 38.0%. If the DPP did poorly in its strongholds, it also suffered heavy losses in the traditional bellwether districts. The partisan balances in Taichung and New Taipei Cities are roughly similar to the overall national pattern, so if you can win Taichung and New Taipei, you will probably win all of Taiwan. The DPP was blown out in both races, losing by 14.2% in Taichung and 14.3% in New Taipei (Batto, 2018; The Economist, 2018). Why did the DPP suffer such as stinging loss against the KMT? What might be the reasons behind its landslide defeat which caused a total reversal of seats and votes compared to the result of 2014? Nathan Batto (2018) does not want to take side in explaining the reasons:
While it is tempting to explain away each outcome as an independent confluence of the local factors and issues that dominated the campaign discourses, the overall trend is too broad and too deep to attribute to idiosyncratic factors such as unequal regional economic development or poor air quality. However, Batto’s words contradicted The Economist’s view on which issue was to blame in explaining the DPP ‘s defeat. First, The Economist (2018) stated this race had nothing to do with China, a prominent cleavage issue that had plagued Taiwan ever since the beginning of election and democratization. A large majority still support the DPP’s position of maintaining the status quo and edging towards independence over the KMT’s desire for unification. But they still voted for the KMT in this election, because this election centered on disgruntlement with President Tsai and her reforms, not an endorsement of the KMTs policy (Evans 2018). More details can be found in Evans’ (2018) article about which reason should be accounted for the iconic loss of the seat of Kaohsiung mayor. He dissected his analysis into three layers: the local one, the national one and the international one. All three of them were intertwined with DPP politics and policies conducted by the Tsai administration. The local factor, for example, concerned the past two DPP mayor of Kaohsiung, despite pledged to finish their second mayoral term, broke their promises and left to become other high-ranking bureaucrats in Taipei. Furthermore, the latest election showed the DPP‘s complacency for the party nominated a longtime Southern policy expert Chen Chi-mai to run for the mayor. They thought the experience he has will be the merit that could compensate his lack of charisma, compared to his KMT rival Han Kou-yu. But this calculation backfired (Evans, 2018). For the second layer—the national factors—the author attributes the loss to the hasty reform conducted by the Tsai administration. Among some of the most prominent reforms, the public pension reform irritated the people to the core, As Evans (2018) has explained in several sentences why it struck such a deep chord in ordinary Kaohsiung people live: “Around four percent of the overall Taiwanese population works in the public sector, but in Kaohsiung, that graph is higher. As Kaohsiung’s population skews older than that of Taiwan’s northern metropolises, there is a higher proportion of retired public sector workers as well. It is this demographic that the pension reforms have hit hardest.”
Furthermore, what makes the reform so fatal and toxic to the DPP election result is another factor. The government failed to communicate with people about
the political and fiscal necessity of such a reform. It did not alarm the people about the financial difficulty these pension fund faced. The way they dealt with the pension reform is also fatal to the election: They removed a large chunk of many people’s income in one fell swoop, rather than phasing in any changes gradually, leaving a significant proportion of the population poorer and angry about it (Evans 2018). Big external factors, such as disinformation spread online, was another one contributed to the DPP ‘s failure. The disinformation campaign capitalized on the public resentment on Tsai unpopular pension cut along with other reforms as well. Then Evans (2018) wrote the nature of the disinformation and its relation to the social media in Taiwan: In these online worlds, they exist in a very insular environment where like-minded people share stories, often in closed-channel networks, which support the preconceptions that people already hold. But while the dissemination of a huge amount of fake news and disinformation on social media and other online platforms has been proven, the influence this had on people’s voting intentions is less clear-cut. It is inconceivable that it has not played a role in swaying the views of at least some voters (Evans, 2018).
However, it is not just China which has been trying to subvert Taiwanese democracy. There is also evidence of the involvement and finance of U.S. Christian organizations supporting the anti-equal marriage referendum campaigns. Again, it is hard to prove that this had a direct outcome on results. Unlike in the United States, in Taiwan, there is no clear correlation between money spent and votes earned. But it is still likely to have had some impact. Further in-depth investigations will be needed to clarify if any laws have been broken and punish those responsible (Evans, 2018).
As Ngerng (2018) had argued, Taiwan's battle was fought on a very simple concern – wages. However, this concern has two mutual contradictory logic embedded. On the one hand, Tsai recognizes she needs to increase wages to ensure the youths are given hope, but on the other, she feels she has to attract Taiwan's companies, most of which are in China, back to Taiwan to ensure they have jobs. But she also has to ensure costs are low so that those companies will return rather than chase low wages elsewhere, which consequently demands the continued depression of wages here. With bearing this loggerhead logic in mind, it is extremely difficult for Tsai to come up with a policy that could please all kinds of constituents at once. This also reflected the DPP ‘s and Tsai administration’s dilemma; Despite they are progressive and reformist from head to toe, they, from time to time, have to compromise due to external and internal pressure from all sides. Any failure to mediate those political interests will only result in angering
political groups across the spectrum. This is the main reason why Tsai administration is so unpopular (Ngerng, 2018).
3.2. Issues at stake in the 2018 election and referenda
After applying the CDA framework to the corpus which consisted of 35 articles, six clusters of topics emerged from the classification and selection. Therefore, it is vital to discuss and introduce those topics in this section which are related to history and contextual knowledge. It is important to note that the issues discussed below are interconnected. This thesis will begin with introducing the most prominent issues across all electoral cycles after the democratization. Then it will move to the related issue of Olympic name controversy and referendum. Second, the issue of same-sex marriage, nuclear energy politics and third forces politics including Ko, the mayor of Taipei, all have their own specific cleavage nature, which means they have the potential to divide constituents in ways different than the traditional China electoral cleavage. The former two issues represent the traditional China cleavage line itself, while the second half are the actors enacting and combining both the new and the old cleavage. 3.2.1 The China cleavage Batto (2018) argues Taiwan has one and only one dominant political cleavage: Taiwan’s relationship with China, which is broadly conceived as more important than all of those other issues put together. This China Cleavage fundamentally shapes every aspect of Taiwan’s politics, most notably the party system. Moreover, it is impossible to understand aNYThing else unless you understand how it is embedded within the China Cleavage.
The China cleavage is constituted by four related but somewhat different dimensions. The first is ethnic background – the conflict between native Taiwanese, including those who were already here during the Japanese colonial era and mainlanders that came to Taiwan after WWII and the Chinese civil war. This dimension was most important during the authoritarian era but has faded in importance over the past few decades. The second is national identity, whether a person self-identifies as Chinese or Taiwanese. The most important measurement in this issue is the percentage of respondents identifying as Taiwanese only in the NCCU Election Study Center tracking polls. The third dimension involves Taiwan’s future status. Should Taiwan unify with China, become a formally independent state, or something else. This dimension seems to have faded in importance in recent years. The fourth dimension is how to manage day-to-day relations with China. On what basis should Taiwan have economic
