The interplay between discourse and ideology has become a controversial issue not only in humanities (e.g. Eagleton, 1991, Chapter 7, pp. 193- 220) and social sciences (e.g. Purvis & Hunt, 1993, also see Section 3.2 and Section 3.3 above for an overview) but also in critical linguistics. There is even a body of literature on this topic in critical linguistics (e.g. van Dijk, 1995, 1998a; Kress & Hodge, 1993). In critical linguistics the
72 debate centers on whether and how the ideology of discourse producers, agents and groups of the ideological adherents is essentially embedded in texts. A number of studies have been conducted and these discussed below.
In his study on the ideology of liberalism and discourse studies, Beaugrande (1999) attempted to reconcile the issue of the ways ideology is embedded in discourse. He began his exploration by arguing that even sciences cannot be free from a particular ideology of the institution to which the scientists are attached (see Althusser (1971) for the dichotomy between science and ideology and Kuhn (1991) for the ‗vagueness‘ of scientific knowledge in relation to institutions). The same holds true for linguistics whereby some linguists have tried to idealize the study of language as non-ideological. Such a paradigm of language studies according to Beuagrande (1999) is based on the static dichotomies ‗langue‘ versus ‗parole‘ or ‗competence‘ versus ‗performance‘, which draws the line between ideal language and real language.
Beaugrande (1999) argued that the ideology of linguisticism, the study of language as a pure science, was authorized by the theoretical linguists who separate the language from the speech community. He posits that sciences, such as modern linguistics also impinge on an ideology; it is not the ideology of content but the ideology of methods, which he calls ―scientism‖ (Beaugrande, 1999, p. 260). He does not accept the attitude of mainstreaming linguistics which attempts to be free from an ideology by idealizing language. Theoretical linguistics as an academic discipline studies human language scientifically, but in isolation, which also needs to be conceived as linguisticism.
To challenge the paradigm of formal linguistics, which was pioneered by de Saussure (1961) and Chomsky (1965), Beaugrande turns to critical linguistics which gives rise to a new interdisciplinary science of text and discourse (Beaugrande, 1997). Consequently,
73 he relies on corpus linguistics, because according to him the order of language is dynamic and transitory. Language is always created and negotiated in discourse. Corpus linguistics enables linguists to explain how language operates in society. This is so because corpus linguistics is considerably involved in data-driven or bottom-up theories. Working with corpus data leads the linguists to real language of the speech community. To provide evidence for his position, Beaugrande (1999), relying corpus analysis, unraveled how the ideology of ‗liberalism‘ is manifested in three corpora: the UK, the U.S. and South Africa sources. He makes use of systemic functional linguistics, particularly the system of grammatical colligations and lexical collocations. The former refers to morphological and syntactical conditions of linguistic elements to be combined and the latter is to do with co-occurrence of individual lexical items (Matthiessen, 1998; Quirk, Greenbaum, Leech & Svartvik, 1985). In other words, the grammar and the lexical support each other.
Beaugrande‘s (1999) analysis of authentic discourse showed that there was a different tendency of the use of the key term ‗liberalism‘ in the three corpora. The UK data revealed that the term ‗liberal‘ was closely related to the context of economy, policies and free market. The contextual meaning of ‗liberal‘ has deviated from the meaning enacted by the conservative ideology, which often refers to a political type rather than the economic and social policies. The U.S. corpus indicated that ‗liberal‘ or ‗liberalism‘ was rarely used in connection to economic meaning. This could be inferred from the data, because in the U.S. politic there is no respected ‗Liberal Party.‘ Consequently, the term has been used in a blurred sense, discrediting the reinforcement of the rights of minorities, intellectuals, and environmentalists. In the South African data, the study revealed that the terms ‗liberal‘ and ‗liberalism‘ occurred more frequently compared to the UK and U.S. data. The terms with a specific meaning were used to promote the
74 rights of Black Africans by the White Africans. The terms ‗liberalize‘ and ‗liberalization‘, which carry the economic meaning, were not frequently used. Different from the UK data, the South African data indicated that ‗liberal‘ did not associate with the ‗conservative‘, but was associated with the ‗communist.‘ These findings suggest that there exists the process of what Hasan (2003) refers to as Re-semanticisation in the English language. The meaning of the word has shifted.
The study by Beaugrande (1999) also revealed that the adjective ‗liberal‘ was far more variable and contested in its collocations than the noun ‗liberalism‘. The analysis also indicated that the subjects of the verb ‗liberalize‘ were almost always the government, institutions, or countries and objects of the verbs were nearly always abstractions, such as prices, trade, market, tariffs and financial system. The word ‗economy‘ constitutes as the head word.
What Beaugrande (1999) implies is that ideologies of discourse participants might reside not only in texts (Kress and Hodge, 1993) but also in other social practices (Fairclough, 1992, 2003)—economies, politic, cultural and social— which is partly constructed in texts. This is so, because ‗meanings are produced through the interpretations of texts in relation to other practices, such as economy and politic and texts are open to different interpretation, which can differ in their ideological import‘ (Fairclough, 1992: 89; emphasis added). The data from the three corpora also suggests that uses of language served to establish, sustain and change social relations are ideological and words are the most straightforward way of producing ideological effect. What Beaugrande‘s (1997, 1999) studies lack is the role of pragmatic and presupposed meaning that is difficult to be generated through corpus analysis. It was only based upon the study of lexical items that ignore the wider social and political contexts. As a result,
75 the analysis is less critical. It does not attempt to make the implicit explicit nor does he take a genuine social political stance.
Other studies that relate discourse to ideology of dominance and in particular racism are van Dijk (1998a, 1998b) in his exploration of a number of editorials in newspapers, particularly the New York Times, the Washington Post and a political treatise, The End of Racism by Denise D‘Sauza (1996). He based his study on the major triad relationship between cognition, society, and discourse (see Section 3.3 above for this triad paradigm). The study suggests that discourse meaning constructed during production and comprehension is reliable to embody opinions that derive from underlying ideologies. Discourse meaning is accounted for only in terms of abstract concepts, lexical structures or mental models—representations as based on social knowledge. In these studies it was revealed that lexicalization, such as the use ‗fundamentalist‘, ‗terrorism‘ and ‗communist‘ are the most dominant and straightforward way of expressing ideology. The more complex discursive strategies are structures of propositions, implication and overall meaning or topic of discourse (van Dijk, 1998a, 1998b). Racist ideology, calling a certain group of people ‗terrorist‘ or ‗fundamentalist‘ and another group ‗civilized‘ or ‗democratic‘ is conceived as one impact of globalization on modern society (Fairclough, 2006; Steger, 2005). That is, globalization affects how people construct their ideology and how they represent and position themselves in discourse in the age of globalization, as revealed by studies of the role of language in new capitalism (e.g. Chiapello & Fairclough, 2002; Fairclough, 2006, 2009). Different from the current study, van Dijk‘s (1995, 1998b) studies on ideology in newspaper Op-Eds and the political treatise did not focus on globalization. Instead, he looked at the ideology of racism in modern society.
76 In a collaborative study, Chiapello & Fairclough (2002) explore the way to approach the issue of language use in the new capitalism in a transdisciplinary paradigm, i.e., sociology of new capitalism (Chiapello) and Critical Discourse Analysis (Fairclough). They specifically examine the ideology of new management as represented by Robert M. Kanter, a celebrated management expert at the Harvard Business School.
They analyzed the extracts of Chapter 9 ‗Leadership for Change‘ from the book Evolve! by R.M. Kanter (2001). The collaborative analysis suggests that transdisciplinary research not only brings together different theories and analytical frameworks but also provide a dialogue between two disciplines and framework. This dialogue provides a basis for development of each discipline. CDA, for example, can obtain a contribution from new sociology of capitalism the theoretical concept that relates to the way of meaning making (Fairclough & Thomas, 2004). That is, the substantive change in new capitalism constitutes the change in ideology and order of discourse. The same holds true for the new sociology of the capitalism. Analyzing texts from a CDA perspective enlightens the development of the new sociology of capitalism.
Another relevant study to the present work is Flowerdew (2002). He examined the discourse of globalization from the view of a specific Eastern country, Hong Kong. Using CDA for textual analysis, he analyzed a speech that was addressed by Tung Chee Hwa, head of Hong Kong Administrative Government, in London in October 2000. The analysis reveals that Hong Kong as a Semi-Autonomous Special Administrative Region needs to tackle the challenges of globalization on its own rights and takes advantages of it. The examination also indicates that in Hong Kong, globalization is discursively constructed and the government has to face the phenomenon and its consequences. The study also makes clear that there are a number of discursive strategies that Tung Chee
77 Hwa deployed in the discursive construction of globalization, such as modality, nominalization and system of pronominalization.
Also central to the framework of the current work is Fiss and Hirsch (2005). In their study on the discourse of globalization in the United States from the sociology of media perspective in relation to the concept of framing and sense-making, Fiss and Hirsch (2005) explore how globalization discourse emerged in response to U.S. involvement with the international economy and how the frame concept contributes to the meaning- making of globalization in respect to groups‘ interests. The theoretical foundation was built upon two concepts of frame, i.e. positive and negative and sense-making. Framing is the process by which actors influence the interpretations of reality among various audiences. In other words, frames are schemata of interpretation that positively, neutrally or negatively organizes experiences and guide action. Meanwhile, sense- making is defined as response to action in regard to positioning of the actors themselves on what they are for or against and what they need to do (Hirsch & Fiss, 2000).
They based the study on an analysis of newspaper articles on economic discourse published from 1985 to 1998 in the Washington Post, the New York Times and the Wall Street Journal. They also surveyed the press releases on the same issue available from the PR Newswire and extracted from all these sources of data through the Dow Jones Interactive database available in the similar period. The data set was segmented using the key word ‗globalization‘ and its derived forms, such as ‗globalized‘ and ‗globalizing.‘ In the end, the newspaper segment contained 1,805 documents and press releases segment contained 1,735 press releases.
Using both the quantitative and qualitative measurements, Fiss and Hirsch (2005) found that there was a substantial increase of awareness of globalization in the United States.
78 That it, the discourse of globalization emerged in the early 1980s, attained a small increase in 1989-90, and then receded until 1994. However, the newspaper articles and press releases on the discourse of globalization increased sharply by 1995. They also revealed that there was a combination between material process and public sense- making of the globalization discourse they examined.
From the perspective of the scope of data and the time range, Fiss and Hirsch‘s (2005) study was fairly comprehensive. Unfortunately, although they used both quantitative and qualitative analysis, the study seems to have relied more on the statistical graphs than on the uses of text extracts to justify their findings. They only examined the data in terms of the computer-generated texts, which is vulnerable to the distortions of pragmatic meanings.
Two recent studies of discourse and ideological manifestation that also rely on CDA are Holborow (2007) and Vaara, Sorsa & Pälli (2010). While Holborow looked at language and the ideology of neoliberalism in the documents, brochures of the asylum seeker process and the interaction between the ‗customers‘ and the official in the call centers in Ireland, Vaara et al. (2010) examined the discursive aspects of strategy texts deployed by the municipality on a plan of the City of Lahti in Finland. Holborow (2007) firstly argues that language and ideology is interconnected but later concludes that the ideology of liberalism cannot be sufficiently described as a discourse. She found that there was a gap between the ‗speak as if you are smiling‘ of the training manual and the actual attitudes of the operators in the call centers. In other words, there exists a customer metaphor of a top-down process; at the top process, asylum seekers are regarded as customers, but in reality at the lower level they are not referred to as customers.
79 Vaara et al. (2010) analyzed the strategy texts and their power effects. They referred to interviews, emails, diaries and presentation slides for the sources of data. Using CDA as a research methodology particularly from Fairclough‘s version of CDA, they found in their inductive analysis that the strategic plan of the city was enacted through five main discursive features: self-authorization, special terminology, discursive innovation, forced consensus, and deonticity.
Self-authorization acts as a basis for textual agency that the city organization has to rely on giving directions in the city planning. Special terminology refers to specified lexicon that is used to plan, present and to interpret the final product of the strategy genre. The terms such as scenarios, vision, SWOT, and change factors were used in the interaction. These terms were used to sustain the power relation in which the authors construct the self-identity as experts and that the readers deserve an explanation of those lexicons. This is what in Boudieu‘s (1991) terminology is referred to as field—the expertise of the discourse producing subject.
The third aspect that Vaara et al. (2010) reveal as a crucial feature of the strategy plan is discursive innovation. What it is meant by discursive innovation is the statements that clarify the idea that the city‘s strategic decision is part of individual responsibility. The fourth feature of the strategy text is forced consensus. This discursive feature provides space for the involved individuals, the City Council, elected leaders, and political parties to agree on the disagreement. And the most apparent forced consensus was directed to the individual responsibility.
The last feature of the strategic planning genre that Varaa et al. (2010) found was deonticity. This discursive feature of the strategy planning emerged from the text when the declarative statements were used for imperative aims, especially when the text is to
80 lead future actions. The declarative statements become the imperative ones. That is, linguistically the authors use mood or the grammatical metaphor for giving directions (Halliday, 2004). Vaara et al. (2010) conclude their study by emphasizing that CDA can be applied in analyzing strategic genres and be used by any research inquiry that is dependent on linguistic aspects and that needs a critical analysis. Central to this important point and its closest relevance to the present study are the findings that suggest that strategic texts constitute power relations among discourse participants and that although language and ideology are different, they are not independent.
The other studies which are much earlier such as Fowler (1991); Kress and Hodge (1993); van Dijk (1993, 1995) and Fairclough (1992, 2010) have also attempted to investigate the interrelationship between ideology and discourse. Each of these studies referred to CDA as a research paradigm. Fowler mainly focused on newspaper features; Fairclough was interested in a number of discourse types, such as newspaper features and pregnancy and baby books; van Dijk mostly relied on newspaper editorials and newspaper Op-Eds while Kress and Hodge looked at a variety of texts including interviews and news headlines. Despite a difference of sources of data, they affirmed in their exploration that ideology and discourse could not be entirely separated. However, what ideology is shaped and how it is actually fixed in discourse still remain not only as a contentious matter but also as a quest in scientific social research. In addition, although there is a great body of literature on globalization— either from the dominance or counter-discourses—those studies often lack evidences from concrete textual analysis (e.g. Pieterse, 1995; Koh, 2005). Consequently, the present study is an attempt to explore the discourse-ideology dialectical relation which is mainly based on both the textual and social analysis. The expectation of it relies on the enlightenment it contributes to the development of critical social research from the interdisciplinary
81 paradigm. The discussion above has taken a position that ideologies are discursive constructions. Therefore, ‗the question of ideology is part of the question of how discourse relates to other moments of social practices‘ (Chouliaraki & Fairclough, 1999, p. 26), be they economy, politic, and culture. This constitutes the main issue raised in the current transdisciplinary work.